<%BANNER%>

Urban Economies and Globalization

Permanent Link: http://ufdc.ufl.edu/UFE0044464/00001

Material Information

Title: Urban Economies and Globalization Exploring Externally-Induced Urban Expansion and Impact on Peri-Urban Land in Accra, Ghana
Physical Description: 1 online resource (381 p.)
Language: english
Creator: Adu-Brempong, Kofi
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: alonso -- economic -- economies -- expansion -- fdi -- financial -- flows -- geography -- iuseem -- land -- remittances -- spatial -- theory -- urban
Geography -- Dissertations, Academic -- UF
Genre: Geography thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Mainstream urban expansion theorizing and modeling neglect influences of major economic forces associated with globalization, indicating a less than comprehensive understanding, given the current era of intensifying globalization. This dissertation explores the theoretical underpinnings and methodological approach to understanding and quantifying globally induced urban spatial expansion and implications for peri-urban lands in Accra, Ghana. The major outcomes are three-fold: (1) FDI induced urban spatial expansion theory is proposed as an extension to the Thunen-Alonso theoretical framework; (2) The Integrated Urban Spatial Expansion Estimation Method (IUSEEM) is proposed as an appropriate method for quantifying FDI induced urban spatial expansion; (3) The study demonstrates that integration of geographical concepts and themes and urban economic theory help place FDI induced urban spatial expansion in the broader context of human-induced environmental change. FDI and remittances were ‘inserted’ in current extensions of Alonso’s theory.  Propositions developed were tested using log transformed multiple regression (OLS), for a sample of 107 cities. FDI, remittances, city population and GDP per capita explained 82% of the variation in built up area. More importantly, FDI correlated positively with urban built area, statistically significant at the 95% confidence level; a 10% increase in FDI corresponded to almost 1% (0.087%) increase in the urban built up area. This co-efficient is non-trivial since FDI can multiply dramatically within a short time, affecting urban expansion through multiple pathways. Remittance was found to be negatively correlated with built up area, contrary to literature and theoretical analyses in this dissertation. Results from the IUSEEM formulated to quantify FDI induced expansion indicate that the method could be useful. This dissertation provides the theoretical foundations and methodological framework for analyzing contemporary urban land expansion as induced by FDI, facilitates a theoretically grounded geographical analysis of FDI induced urban land expansion; with implications for urban planning,  policy and conservation.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Kofi Adu-Brempong.
Thesis: Thesis (Ph.D.)--University of Florida, 2012.
Local: Adviser: Fik, Timothy J.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044464:00001

Permanent Link: http://ufdc.ufl.edu/UFE0044464/00001

Material Information

Title: Urban Economies and Globalization Exploring Externally-Induced Urban Expansion and Impact on Peri-Urban Land in Accra, Ghana
Physical Description: 1 online resource (381 p.)
Language: english
Creator: Adu-Brempong, Kofi
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: alonso -- economic -- economies -- expansion -- fdi -- financial -- flows -- geography -- iuseem -- land -- remittances -- spatial -- theory -- urban
Geography -- Dissertations, Academic -- UF
Genre: Geography thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Mainstream urban expansion theorizing and modeling neglect influences of major economic forces associated with globalization, indicating a less than comprehensive understanding, given the current era of intensifying globalization. This dissertation explores the theoretical underpinnings and methodological approach to understanding and quantifying globally induced urban spatial expansion and implications for peri-urban lands in Accra, Ghana. The major outcomes are three-fold: (1) FDI induced urban spatial expansion theory is proposed as an extension to the Thunen-Alonso theoretical framework; (2) The Integrated Urban Spatial Expansion Estimation Method (IUSEEM) is proposed as an appropriate method for quantifying FDI induced urban spatial expansion; (3) The study demonstrates that integration of geographical concepts and themes and urban economic theory help place FDI induced urban spatial expansion in the broader context of human-induced environmental change. FDI and remittances were ‘inserted’ in current extensions of Alonso’s theory.  Propositions developed were tested using log transformed multiple regression (OLS), for a sample of 107 cities. FDI, remittances, city population and GDP per capita explained 82% of the variation in built up area. More importantly, FDI correlated positively with urban built area, statistically significant at the 95% confidence level; a 10% increase in FDI corresponded to almost 1% (0.087%) increase in the urban built up area. This co-efficient is non-trivial since FDI can multiply dramatically within a short time, affecting urban expansion through multiple pathways. Remittance was found to be negatively correlated with built up area, contrary to literature and theoretical analyses in this dissertation. Results from the IUSEEM formulated to quantify FDI induced expansion indicate that the method could be useful. This dissertation provides the theoretical foundations and methodological framework for analyzing contemporary urban land expansion as induced by FDI, facilitates a theoretically grounded geographical analysis of FDI induced urban land expansion; with implications for urban planning,  policy and conservation.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Kofi Adu-Brempong.
Thesis: Thesis (Ph.D.)--University of Florida, 2012.
Local: Adviser: Fik, Timothy J.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044464:00001


This item has the following downloads:


Full Text

PAGE 1

1 URBAN ECONOMIES AND GLOBALIZATION: EXPLORING EXTERNALLY INDUCED URBAN EXPANSION AND IMPACT ON PERI URBAN LAND IN ACCRA, GHANA By KOFI ADU BREMPONG A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Kofi Adu Brempong

PAGE 3

3 To the memory of our beloved late broth er, Dr. Kwadwo Konadu Agyemang

PAGE 4

4 ACKNOWLEDGMENTS Glory is to God! F or through God all things are possible. Indeed, I have lost count through human agency without which this work would not have materialized. In short, in completing this dissertat ion research I have stood on the shoulders of many kind and caring people. To these people, therefore, I owe tons of gratitude and great indebtedness. I sincerely thank Dr. Timothy J. Fik, my Advisor and Chair, for accommodating ieces as I started to explore the subject matter of this dissertation. Even though things were murky in the beginning, Dr. Fik continued to have confidence in my intellectual abilities; specifically pointing out books that will help me sharpen my thought s in the area of research I was leaning towards as well as painstakingly reviewing my documents but allowing me to think freely. In the economic front I was one lucky beneficiary of the Edward Fik foundation which came in a very difficult time of financial stress. Thus, all things considered, Dr. Fik has been a blessing to me. I thank no prior appointment and they would attend to me. Their show of interest in the subject mat ter of this dissertation has been very encouraging and their suggestions were very helpful. My thanks go to Dr. Zwick, my committee member and Minor Representative, whose class I took prior to requesting him to be on my dissertation committee, helped me r elate my urban and economic geographical interests to urban planning. I should like to mention specifically, his questions for my written comprehensive examination bordering on modeling, which helped lea d me into further exploration into the subject matter I thank Dr. Peter Waylen, former Chair of the Department of Geography for his

PAGE 5

5 exemplary leadership, creating an enabling academic environment for budding geographers like me to flourish. Going beyond his academic and administrative call, Dr. Waylen cont even from a different university for which I would never forget. Once again, thank you, Sir. To all the friendly people and well wishers in the Geography Department, I say thank y ou and God bless. I should like to make particular mention of the financial which made it possible for me to go back to Ghana to say a fin al goodbye to my beloved father. I wi ll forever be grateful to you al l. Many thanks to Desiree, Julia and Rhonda for their support services that sustain de partmental functions and efforts. I thank Dr. Shlomo Angel and colleagues at the Lincoln Institute of Land Policy for making freely availa ble online a rich and extensive database on urban expa nsion which contributed immensely to my dissertation. Thanks to my wife, Cynthia Adu Brempong and son, Kwadwo Adu Brempong, for enduring ye ars of my absence Sorry for the emptin ess and loneliness I cau sed you Thank God, you were finally able to join me in G ainesville, Florida, in 2011. While still rejoicing in this family reunion, o n 19 th November 2011 my family was blessed yet again with a beautiful daughter! I could not find any befitting name other Brempong, we warmly welcome you. Last but not the least I thank my family, especially, Dr. Owusu Dommey, Dr. Kwame Obeng, Mr. Kofi Appiah Kubi Brenya and Mr. Kwadwo Ampofo Brenya as well as all the people who stood up for me and prayed for me in the wake of the fatality that I suffered on March 2, 2010. I cannot express my gratitude in words so all I can wish for

PAGE 6

6 TABLE OF CONTENTS Page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ .......... 10 LIST OF FIGURES ................................ ................................ ................................ ........ 11 LIST OF ABBREVIATIONS ................................ ................................ ........................... 12 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 GENERAL INTRODUCTION ................................ ................................ .................. 15 Globalization, Spatial Fin ancial Flows and Urban Spatial Expansion ..................... 15 Spatial Financial Flows in a Globalizing era. Africa in Context ......................... 18 ................................ ................................ ... 27 Statement of the Problem and Objectives ................................ ............................... 30 Research Questions and Hypotheses ................................ ................................ ..... 36 Questions ................................ ................................ ................................ ......... 37 Hypotheses ................................ ................................ ................................ ...... 39 The General Conceptual Framework and Theoretical Structure ............................. 39 General Conceptual Framework ................................ ................................ ....... 39 Theoretical Structure ................................ ................................ ........................ 42 Study Area, Data a nd Methods: An Overview ................................ ......................... 45 Study Area ................................ ................................ ................................ ........ 45 Ghana ................................ ................................ ................................ ........ 45 Accra ................................ ................................ ................................ .......... 51 Data and Methods ................................ ................................ ............................ 55 Significance of this Dissertation ................................ ................................ .............. 56 Overview of Chapte rs and Organization of the Dissertation ................................ ... 60 2 FDI AND URBAN SPATIAL EXPANSION: TOWARDS A SYSTEMATIC CONCEPTUAL FRAMEWORK ................................ ................................ ............... 67 Backgro und ................................ ................................ ................................ ............. 67 Globalization, Africa and FDI Flows in Context ................................ ....................... 72 A Review of the FDI Theoretical Literature ................................ ............................. 79 Perfect Competition and FDI ................................ ................................ ............ 79 Market Imperfections and FDI ................................ ................................ .......... 81 The Product Life Cycle and Beyo nd ................................ ................................ 83 FDI and Urban Spatial Expansion: A Conceptual Framework ................................ 92 FDI Impact on Land Conversion via Income and Population ............................ 93 FDI Impact on Land Conversion: Direct Land Demand ................................ .... 94

PAGE 7

7 Summary ................................ ................................ ................................ ................ 97 3 CONCEPT UALIZING THE LINK BETWEEN REMITTANCES AND URBAN SPATIAL EXPANSION ................................ ................................ ......................... 100 Background ................................ ................................ ................................ ........... 100 Globalization and Remittances: An Overview ................................ ....................... 103 Remittances Theoretical literature ................................ ................................ ........ 105 Mainstream Theories ................................ ................................ ...................... 105 Th e Asset Accumulation Motive of Remitters ................................ ................. 109 Conceptualizing Remittances Urban Spatial Expansion Relationship .................. 110 Summary ................................ ................................ ................................ .............. 113 4 ....... 117 Background ................................ ................................ ................................ ........... 117 Economic Theory and Urban Space: The Thunen Alonso Theoretical Framework ................................ ................................ ................................ ......... 123 ................................ ................... 124 Al ................................ ................................ .... 125 ................................ ..... 131 On Income ................................ ................................ ................................ ...... 134 On Population ................................ ................................ ................................ 134 Related Theoretical Propositions ................................ ... 137 Towards Globa lly Induced Urban Spatial Expansion Analysis: An Empirical Example ................................ ................................ ................................ ............. 141 Data ................................ ................................ ................................ ................ 141 Method ................................ ................................ ................................ ........... 142 Urban built up area expansion regression models. ................................ .. 145 Regression Estimation Results ................................ ................................ ....... 147 Discussion ................................ ................................ ................................ ............ 152 Urban Spatial Expansion Analysis and Globalization. Overlooked Factors? .. 152 Do FDI and Remittances Drive Urban Spatial Expansion? ............................. 158 ................................ .............. 163 Issues with Remittances and FDI as Urban Land Expansion Drivers. ............ 164 Globalization Induced Urban Spatial Expansion? ................................ .......... 168 Summary ................................ ................................ ................................ .............. 172 5 FDI INDUCED P ERI URBAN LAND EXPANSION IN ACCRA, GHANA: THEORETICAL AND QUANTIFICATION CONSIDERATIONS ............................ 179 Background ................................ ................................ ................................ ........... 179 FDI Induced Urban Spati al Expansion in the Context of the Literature ................. 184 Human Environment relationship Induced Environmental Change .............. 186 Globalization and Environment ................................ ................................ ....... 188 Urban Environmental Change ................................ ................................ ........ 196

PAGE 8

8 Globalization Economy Environment: Geographical Perspectives and Urban Economic Theory ................................ ................................ .............. 200 The Thunen Alonso Theoretical Framework ................................ .................. 205 Premises for the initial hypotheses ................................ .......................... 207 Quantifying FDI Induced Urban Spatial Expansion ................................ ........ 208 Modeling ................................ ................................ ................................ .. 209 Overview of Current urban spatial ex pansion modeling ........................... 215 Exponential growth function (Model) and the Allometric law: The population growth model. ................................ ................................ ...... 220 The Allometric l aw (or Allometric growth principle) ................................ .. 221 The IUSEEM: General Considerations ................................ ........................... 227 Developing the IUSEEM: The main steps ................................ ...................... 231 IUSEEM Vs. the Population Growth Model: Similarities and Differences. ...... 235 Methodology ................................ ................................ ................................ ......... 238 Overview ................................ ................................ ................................ ........ 238 Research Design ................................ ................................ ............................ 240 Data ................................ ................................ ................................ ................ 242 Method ................................ ................................ ................................ ........... 242 Results ................................ ................................ ................................ .................. 252 Discussion ................................ ................................ ................................ ............ 268 Accra: Highly Urbanized, Globalizing and Spatially Expanding ...................... 268 Is Population a Major Driver of Rapid Spatial Expansion in Accra? ............... 270 Is Income a Major Driver of Rapid S patial Expansion in Accra? ..................... 273 How likely is FDI a Major Driver of Rapid Spatial Expansion in Accra? .......... 275 To what extent is FDI a M ajor Driver of Urban Spatial Expansion in Accra: Interpreting the IUSEEM Results ................................ ................................ 280 Average growth rate. ................................ ................................ ................ 280 Varying FDI concentr ation levels ................................ ............................. 283 The IUSEEM Results Compared to Independent Data ................................ .. 285 FDI and the Thunen Alonso Theoretical Framework ................................ ...... 290 The Peri Urban Land Market in Accra under Globalization ............................ 2 97 FDI and Human Induced Environmental Change ................................ ........... 305 Summary ................................ ................................ ................................ .............. 309 6 GENERAL CONCLUSIONS ................................ ................................ ................. 330 Summary ................................ ................................ ................................ .............. 330 Contributions of this Dissertation ................................ ................................ .......... 333 Integrated Geographical Theory Construction ................................ ................ 333 Urban Spatial Expansion Model ing and Estimation Methodology .................. 335 Thinking Across Geographical Scales and Disciplines ................................ ... 337 Implications of this Dissertation Resea rch ................................ ............................ 338 Urban Planning and Policy ................................ ................................ ............. 338 Urban Environmental Conservation ................................ ................................ 342 Suggestions for Further Research ................................ ................................ ........ 346 FDI, Remittances and Cost of Low Income Housing in Accra ........................ 346 IUSEEM, GIS and Empirica l Research ................................ .......................... 348

PAGE 9

9 Expanding the Scope of this Research Modeling ................................ ........... 349 Scale of Data Capture ................................ ................................ .................... 350 Remittances and Urban Spatial Expansion ................................ .................... 350 Some Limitations of this Research ................................ ................................ ....... 351 Data ................................ ................................ ................................ ................ 351 Scale of Data Capture ................................ ................................ .................... 352 Scope ................................ ................................ ................................ ............. 352 APPENDIX A DATA USED FOR REGRESSION ANALYSIS ................................ ..................... 354 B SUPPLEMENTARY MAPS ................................ ................................ ................... 361 C GENERAL NOTES ON DATA SUITABILITY ................................ ........................ 364 LIST OF REFERENCES ................................ ................................ ............................. 367 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 381

PAGE 10

10 LIST OF TABLES Table page 1 2 Populat ion growth in the ten largest cities in Ghana1970 2000 .......................... 66 4 1 Regression results with population of individual cities and GDP per capita ...... 177 4 2 Regr ession results with population GDP per capita FDI and remittances ........ 177 4 3 Stepwise regression results and R 2 change ................................ ..................... 177 4 4 Regression results with built up area regressed on FDI and remittances ......... 177 4 5 Regression results with built up area regressed on remittances ....................... 178 4 6 Regression results with built up area regressed on FDI ................................ ... 178 5 1 Spatial expansion o f Accra and annual growth rate 1985 2002 ....................... 324 5 2 Portions of regression estimation results in Chapter 4 ................................ ..... 324 5 3 IUSEEM estimated buil t up area of Accra FDI coefficient alone ....................... 324 5 4 USEEM sensitivity to varying FDI concentrations with F DI co efficient only ..... 325 5 5 IUSEEM estimated built up area of Accra FDI GDP and population ................ 325 5 6 IUSEEM results sensitivity to varying FDI concentrati ons (all coefficients) ...... 326 5 7 Five year averages for the regression explanato ry variables ........................... 326 5 8 Geographical distribution of FDI projec ts among the regions of Ghana ........... 327 5 9 Comparing IUSEE M results with Mo ller Jensen et al. FDI coefficient alone ..... 327 5 10 Comparing IUSE EM results to Moll er Jensen et al. all coefficients .................. 328 5 11 Confi dence interval for IUSEEM with 80% FDI concentration 2001 2011 ........ 328 5 12 IUSEEM results compared to independent data ................................ ............... 329 A 1 Extrac ted Data for the Sample of 120 cities and urban areas. ......................... 354 A 2 Socio economic dataset used for regression estimation. ................................ 357

PAGE 11

11 LIST OF FIGURES Figure page 1 1 Transmi ssion of SFF into urban spaces and consequences on ecosystems. .... 62 1 2 Study Area Location of Ghana ................................ ................................ ......... 63 1 3 Study Area Location of Accr a. ................................ ................................ ......... 64 1 4 Study Area The Greater Accr a Metropolitan Area. ................................ ............. 65 2 1 Pathways linking FDI to urban spatial expansion ................................ ............... 99 3 1 Outline of mainstream remittance determination theories ................................ 115 3 2 Pathways linking remittances to urban spatial expansion ................................ 116 4 1 ................................ ........................ 175 4 2 Alonso's theory and model bid rent curves and urban spatial structure ........... 176 5 1 Spatio temporal growth of the city as a spatial /geographical system .............. 315 5 2 T he Integrated Urban Spatial Expansion Estimation Method (IUSEEM). ......... 316 5 3 Spatial expansion of Accra1985 2002. ................................ ........................... 317 5 4 A s ection of Accra and a diagrammatic scheme of recent internal structure 318 5 5 Sprawled residential developments in the peri urban areas of Accra .............. 319 5 6 Trends in FDI inflows to Ghana in US$ 19 93 2007. ................................ ..... 319 5 7 Growth in FDI proj ects in Ghana 2003 2007. ................................ .................. 320 5 8 Sectorial distribution of employment to be generated by FDI i nflow to Ghana 321 5 9 Geographical distribution of FDI projec ts among the regions of Ghana ........... 322 5 1 0 FDI induced peri u rban land conversion city. ........................ 323 B 1 Location M ap of the sample of 120 cities. ................................ ....................... 361 B 2 Map of Accra showing main Central Busine ss Districts ................................ .. 362 B 3 Map of Accra and part of GAMA showing points of interest ............................ 363

PAGE 12

12 LIST OF ABBREVIATIO NS FDI Foreign direct investment FSA Firm specific asset GAMA Greater Accra metropolitan a rea GIPC Ghana investment promotion center GIS Geographic information systems GSS Ghana statistical service IUCN International union for conservation of nature IUSEEM Integrated urban spatial expansion estimation method MNC Multi national corporation MNE Multi national enterprises OLS Ordinary least squares SAP Structural adjustment program SFF Spatial financial flow TNC Trans national corporation UN United nation s UN CTAD United nations conference on trade and development UNDP United nations development program WCED World commission on environment and development

PAGE 13

13 Abstract of Dissertation Presented to the Graduate School of the University of Florida i n Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy URBAN ECONOMIES AND GLOBALIZATION: EXPLORING EXTERNALLY INDUCED URBAN EXPANSION AND IMPACT ON PERI URBAN LAND IN ACCRA, GHANA By Kofi Adu Brempong August 2012 Chair: Timoth y J. Fik Major: Geography Mainstream urban expansion theo rizing and modeling neglect influences of major economic forces associated with globalization, indicating a less tha n comprehensive understanding given the current era of intensifying globalization This dissertation explo res the theoretical underpinnings and methodological approach to understanding and quantifying globally induced urban spatial expansion and implications for peri urban lands in Accra, Ghana T he major outcomes are three fold: (1) F DI induced urban spatial expansion theory is proposed as an extension to the Thun en Alonso theoretical framework ; ( 2 ) The I ntegrated Urban Spatial Expansion Estimation Method (IUSEEM hereafter ) is proposed as an appropriate method for quantifying FDI indu ced urban spatial expansion ; (3) The study demonstrates that i ntegration of geographical concepts and themes and urban economic theory help place FDI induced urban spatial expansion in the broader context of human induced enviro nmental change FD I and rem ittances were in s theory. P ropositions developed were tested using log transfor med multiple regression for a sam ple of 107 cities. FDI remittances, city population and GDP per capita explained

PAGE 14

14 82% of the variat io n in built up area. More i mportantly, FDI correlated positively with urban built area, stat istically significant at the 95 percent confidence level; a 10 percent increase in FD I corresponded to almost 1 percent (0.087 ) increase in the urban built up a rea. This co efficient is non trivial since FDI can multiply dramat ically within a short time, affecting urban expansion through multiple pa thways. R emittance was found to be negatively correlated with built up area contrary to literature and theoretical analy ses in this dissertation Results from t he IUSEEM formulated to quantify FDI induced expansion indicate that the method could be useful This dissertation provides the theoretical foundations and methodological framework for analyzing contemporary urban l a nd expansion as induced by FDI; facilitates a theoretically grounded geographical analysis of FDI induced urban land expansion; with implications for urban planning, policy and conservat ion.

PAGE 15

15 CHAPTER 1 GENERAL INTRODUCTION Globalization, Spatial Financia l Flows and Urban Spatial E xpansion This dissertation proposes the theory of FDI induced urban spatial expansion as well as the Integrated Urban Spatial Expa nsion Estimation Method to quantify the impact of FDI on built up area expansion over time. The car dinal reason for this effort is that mainstream urban expansion theorizing and modeling neglect influences of major economic forces associated with globalization, indicating a less tha n comprehensive understanding of how globalization related economic forc es such as foreign direct investment ( FDI ) and remittances influence ur ban land use values and have spatial ramifications. In effect, therefore, t he central underlying argument of this dissertation research is that rapidly increasing levels of FDI and rem ittances flows, when significantly concentrated in an urban location, drive urban land values up, increase the opportunity cost of holding lands in agricultural and natural uses in peri urban areas, thereby boosting rapid land conversion to urban uses whic h could lead to loss of urban ecological resources and contribute to environmental degrad ation. Consequently, it has been suggested that these global economic forces be incorporated in the classical Alonso theory to help account for current realities of e conomic globalization. In the view of many researchers, including geographers, increasing globalization is rapidly shaping our world (Dicken, 2003). As Peter Dicken s rightly puts it We live in a world of increasing complexity, interconnectedness and vol atility; a world in which the lives and livelihoods of every one of us are bound up with processes operating at a larger geographical scale (Dicken s 2003: 13). surface have, and continue to impact localities and regions, is emerging as one of the

PAGE 16

16 key issues of concern among geographical scientists as indicated in NR report, for example. In this regard the overall aim of this dissertation resonates with the N s strategic direction question number 7 and ideas transformi Essentially, therefore, this dissertation has sought to link contemporary urban spatial expansion in a primate city, as a dynamic geographical system, to external economic forces FDI and remittances as globalization related external funds flow seeking to ground the process in standard urban economic theory along the lines espoused by William Alonso (1964) By so doing the study would have linked global economic processes to local spatial changes, with the potential to increasing our understanding of human induced environmental change, albeit in urban areas in developing countries. In very broad terms, therefore, this disse rtation research has att empted to answer the question: to what extent are FDI and remit tances ( described as spatia l financial flows in this work SFF, hereafter), potent drivers of contemporary urban spatial expan sion and urban environmental degradation ? It is widely recognized that globalization has defied complete and precise definition even though the word has become a common usage among researchers, policy makers, development agencies and ordinary people alike (Lim, 2010 and Carmody 2010). Lim notes Globalization has become one of the most frequently talked about subjects in academia, the mass media and the publi c policy arena in recent years. And yet w hat it means is not clear (Lim, 201 0: 13). However, lack of concise definition has not prevented globalization discourses from attracting increasing attention from researchers. Some authors have alluded to three main components of globalization cultural, political and economic (Short & Yeong Hyun, 1999) others posit an integrated view, arguing that

PAGE 17

17 Although a highly contested term, globalization is often defined as the increased interconnectedness between places in terms of trade, investment, and information flows in particular (Carmody, 2010: 2). In this regard, it is obvious that the lives of many people in distant places are being affected by globalization as Dicken s (2003) has argued. This notwithstanding, globalization has not been accepted wholeheartedly by all, in that there have been p roponents on one hand who tout the benefits of global ization. Consequently, w hereas supporters have suggested that foreign investment (associated with globalization) can bring capital and technology, develop s kills and linkages and increase employment and incomes; there are opponents on the other hand who l ament the negative consequences of globalization, espec ially on the plight of the poor or low income households in developing countries, suggesting that globalization has led to dependent, or restricted, development (Lim 2010). Even though important, t hi s contentious debate is on going and is not the focus of this work per se. What is less contentious, however, is the realization that increasing levels of international trade, foreign investments as well as the dramatic growt h in multinational corporation s and firms have come to be closely associated with the globalization phenomenon. In this respect, this st udy shares the assertion that The debate on globalization usually su rrounds economic globalization. In this line of thinking, it refers to an integr ation of the world economy, removing trade barriers and allowing freedom of interaction. (Lim, 2010: 14). by production and 225). From a geographical perspective, therefore, increasing spatial interconnectedness and rising

PAGE 18

18 levels of spatial financial flows have characterized the global economy, making spatial financial flows cardinal aspects of globalization (Dicken s 2003). Two main forms of spatial financial flows which form the focus of this study are (1) Foreign direct investment m from one country buys a controlling investment in a firm in another country or where a firm sets up a branch or s, 2003: 51); and (2) Remittances also simply defined as om migration destination to the 5). In effect, therefore, this dissertation research asserts that these two forms of external economic impulses could ha ve significant spatial ramifications in distant localities and regio ns other than their origins, giving impetus for systematic geographical analysis of this emerging globalization induced urban spatial expansion process. Spatial Financial Flows in a Globalizing era. Africa in C ontext Increasing foreign investment can be us ed as one measure of growing economic globalization, as noted above. Whereas world output has approximately doubled since 1970, FDI has quadrupled during the same period (Dicken s 2003; 1997). And according to the World Investment Report (2007), FDI incr eased from US$400 billion in 2000 to a record US1, 833 billion in 2007 (UNCTA D, 2007; Bank of Ghana, 2009). The current global economic downturn notwithstanding, the global economy as a whole is predicted to experience a rebound soon. In this regard, th e United Nations note that (UNCTAD, 2010: xxiii). UNCTAD further noted that Global inflows are expected to pick up to over $1.2 trillion in 2010, r ise further to $1.3 1.5 trillion in 2011, and head towards $1.6 2 trillion in 2012 UNCTAD, 2010: xvii).

PAGE 19

19 But FDI prospects are fraught with risks and uncertainties, including the fragility o f the global economic recovery. (UNCTAD, 2010: xvii). Consequently, the rebound of the global economy is not totally doubtful and it seems reasonable to assume that associated FDI flow levels will be sustained, i f not enhanced albeit with varying concentration levels for different countries Traditionally, developed or industrialized countries are known to receive the largest proportion of global FDI flows, with the bulk occurring between North America, North Wes t Europe and Japan (Hoogvelt, 1997). As a result, in 1970 1980, large parts of Asia, Africa and Latin America had little or no FDI inflows (Hanink, 1994). The situation back then has been aptly captured by authors such as Ankie Hoogvelt (1997) who desc flows and trade almost invariably were concentrated in the rich / industrialized countries and some parts of Asia. However, the situation is rapidly changing. For example, accordi ng to the World Investment Report, FDI inflows and outflows of developed countries plunged in 2008, with inflows declining by 29%, to $962 billion, and outf lows by 17%, to $1,507 billion (UNCTAD, 2010 ; 2009 ). The UNCTAD continues that developed countries d id not return to FDI growth in 2010. show that FDI flows to this group of economies fell some 7% to $527 billion, despite the ro bust recovery in some countries such as the USA. What is even more revealing is the gradual increase in receipts of FDI in developing and transition economies and its resilience even in the face of global economic crisis (UNCTAD, 2010). Thus, note d UNCTAD, despite worldwide economic downturn, the global FDI inflows to developing

PAGE 20

20 and transition economies rose to 43% in 2008, after six years of continued growth (UNCTAD, 2009). This compares favorably with the general situation for developed countries as a group in terms of resiliency. In this regard UNCTAD asserts that these countries saw a decline in F DI of about 44% (UNCTAD 2010). Moreover, according to UNCTAD global FDI inflows kept rising: for the first time ever, developing and transition economies are now absorbing half 2010: xviii xix). Hence, it appears that contemporary FDI flows are becoming more geographically Whereas some may argue that the increasing volumes of FDI flows to the Global South point to increasing global economic integration, others such as Sutchcliffe (2001), however, caution that the bulk of this expansion may be accounted for by the privatization of state enterprises, rather than the setting up of new factories as the definition of FDI implies It is to be noted that this debate is not the focus for this dissertation research. The relev ant question for this work is: are FDI flows to developing and transition economi es merely ephemeral or enduring? The United Nations seem to suppor t the latter view, stating that There are some major changes in global FDI patterns that preceded the global crisis and that will most likely gain momentum in the short and medium term. Firstly, the relative weight of developing and transition economies as both destinations and sources of global FDI is expected to keep increasing. These economies, which absorbed almo st half of FDI inflows in 2009, are leading the FDI recovery (UNCTAD, 2010: xvii). H owever, these flows exhibit remarkable geographic al inequality with some regions and countries attract ing more FDI than others. For example it has been estimated that in 1997 about 71% of the foreign direct investment in developing

PAGE 21

21 countries, described g enerally as the Global South was received by only 9 countries, with China topping the list with receipts of about 30% of the FDI inflows (UNCTAD 2010). Globalization in general has been highly uneven and so has associated FDI flows (Dicken s, 2003 ). So w ith the inequality characterizing global FDI flows, how has Africa fa red? In other words, has Africa been bypassed by globalization and FDI flows ? Africa has had a long history of interactions and interconnections with places far and beyond, ostensibly p redating the current era of glob al economic integration (Cooper, 2001) from slave trade, colonization to economic exploitat ion as argued by Bond (2006). In contemporary development discourses, two main views on Africa prevail (Bond, 2006). The first vi ew asserts that Africa has been bypassed by economic glo balization; with the second alluding to the fact that Africa has suffered from glo balization (Bond, 2006; Carmody, 2010:1). It must be noted that until recently independent countries in Africa were su spicious of their former colonial masters, fearing that the Global North would seek to recolonize them under the disguise of economic reforms. Thus, after gaining independence, following the end of the Second World War, most countries in Sub Saharan Africa suspected that foreign investments in the form of foreign ownership of production in their economies were neocolonial strategies that have been instituted in order to recolonize the region (Killick, 2010; Carmody, 2010: 41 42 ; Bro adman, 2007; Voltoirenet, 2006; ). This prompted many leaders of the newly independent countries to nationalize companies with foreign ownership; in some cases privately owned companies by citizens were not spared. For example, Tony Killick (2010) has noted that the first independ ent government of Ghana followed this path to some degree. To a very large extent, this hostile attitude towards foreign business

PAGE 22

22 interests in their domestic economies and the general aversion for private entrepreneurship effectively restricted FDI inflow s (Killick 2010 ; Carmody, 2010 ). Currently, however, the general consensus emerging among developed countries and developing ones alike is that FDI is desirable, if not essential for both economic growth and poverty reduction ( Carmody, 2010: 41 42 ; Br oad man, 2007; Voltoirenet, 2007 ). participation in the global econ Largel y as a consequence of these developments, among others, some advances have been forged aimed at through natural resources exploitation and export of primary product from the continent and increasing demand for mobile phones, with some modest achievements in attracting transnational corporations (Carmody, 2010: 34; Ndulu et al., 2007). For exam ple, according to Padraig Carmody I enty fastest growing economies and there are now thirteen middle income countries in sub Saharan Africa (Carmody, 2010: 34, citing Ndulu et al., 2007). Sector wis e, the United Nations note that Contraction of investment in the services sector in Africa was less pronounced than in other sectors. Sustained by expanded activity, the telecommunications industry became the largest recipient of FDI inflows. Recovering commodity prices and continued interest from emerging Asian economies are expected to feed a slow upturn in FDI flows to Africa in 2010. (UNCTAD, 2010: xix). Consequently, m receipts have been relatively small there are indications that the continent is increasingly beco ming FDI destination, believed to be currently attracting about three

PAGE 23

23 times the levels i n the 1990s FDI (Broadman, 2007; Voltoirenet, 2007; Carmody, 2010: 40 Africa has evolve d significantly, particularly, in the past decade, as a result of the rise of 3). Current globalization in Africa is being led, predominantly, by Asia, particularly, India and China and predicated largely on natural resource exploitation (Carmody, 2010). The author has argued that China is playing a cardinal role in the globalization of Africa but is dismayed that mainstream research works have largely neglected to assert this fact. As are result he has suggested that The rise of China in Africa, through the intensification of economic and ideational flows and migration, should be seen as an evolution of globalization on the continen t rather than separate from it. (Carmody, 2010: 3). To put this in pers pective, Padraig Carmody notes Chinese trade with Africa grew an astonishing 45.1 percent in 2008 alone to US$107 billion (Center for Chinese Studies, 2009). China is now the largest trading partner, after the United States ; and its si ng le la rgest source of imports. (Carmody, 2010: 3). 4 5), Chinese domination notwithstanding. Another im portant development is that it is generally held that globalization on the African continent has moved beyond resource exploitation to encompass trade, investment and migration (Carmody, 2010: 4 5). More importantly, it has been noted d, the Wo rld Bank claimed that 2006a, cited in

PAGE 24

24 Carmody 2010: 34]). To the extent that Africa continues to globalize, FDI inflows is likely to co ntinue. Like FDI, remittances, particularly f rom developed to developing countries have been on the increase following increased international migration (International Organization for Migration IOM, 2011 ) and economic reforms which allow for free flow of financial resources across countries. The World Bank has estimated that migrant worker remittances were US$182 billion in 2004, representing an increase of 5.7% and 34.5% over the figures recorded for 2003 and 2001, respectively ( Maimbo & Ratha, 2005 ; World Bank 2 004). Even given the current global economic downturn, remittance flows have shown a better resilience compared to private capital flows such as FDI and portfolio investments. Thus, Mohapatra and colleagues have asserted that The decline in remittances during the global financial crisis was modest compared to a 40 percent decline in foreign direct investment (FDI) between 2008 and 2009 and an 80 percent decline in private debt and portfolio equit y flows from their peak in 2007 (Mohapatra et al., 2011: 1). countries are estimated to have fully recovered to the pre crisis level of $ 325 billion in 2011: 1) and the authors further assert t hat : R emittance flows to developing countrie s are expected to increase by 6 .2 percent in 2011 and 8.1 percent in 2012, to reach $346 billion in 2011 and $374 billion in 201 2 respectively. (Mohapatra et al. 2011: 1). These d evelopments point to the ever increasing importance of remittances to developing countries. However, remittance funds, like other forms of global funds flow, are unevenly distributed. Thus, even though remittances have increased significantly throughout developing countries as whole, regional variations do exist. Singh and colleagues note that

PAGE 25

25 A number of African countries are among the largest recipients of remittances relative to their GDP, and for some of them remittances represent a m ajor source of foreign exchange (Singh et al. 2009: 3). For exa mple, it has been noted that Sub Saharan Africa received estimated US$19 billion in remittances in 20 07 accounted for only about two and half percent of GDP for the sub re gion, which represented about five percent of total remittance s to developing countries in general, far behind the amounts received in the Middle East and South Asia as regions (Singh et al. 2009: 3). Additionally, remittances in 2007 in terms of GDP, six were in Africa (Cape Verde, Comoros, 2009: 4). Moreover, f or some countries in Africa remittances constitute a substantial sour ce of foreign exchange earnings as has been noted by Singh and colleagues: As a source of foreig n exchange, in Benin, Cape Verde, Comoros, Eritrea, Gambia, Lesotho, and Uganda, remittances in 2006 represented more than 25 percent of ea (Singh et al., 2009: 4). Increasingly, it seems, remittances flow to developing countr ies is assuming an important dimension as a vital source of external finance for many of these countries (Mohapatra et al. 2011). Beyond development financing, such remittance funds continue to be critical in terms of providing economic sustenance for nu merous families as well as savings towards the acquisition of real property such as houses and lands in migrant home countries (Addison 2004). To the extent that globalization reinforces international migration and given the ageing economically active po pulation in rich countries; and so far as economic liberalization ensures almost seamless spatial flow of economies to less developed ones) is likely to continue into the future (Chami et al. 2008).

PAGE 26

26 Ghana is a beneficiary of A conservative estimate of oversea remit tances by the Bank of Ghana put at US$ 1billion in 2003 (Addison 2004), representing ab out 13 percent GDP ( Mazzucato et al., 2008: 104 ; Addison, 2004 ). Remittances to Ghana amounted to over $4.5 billion in 2005, making it the largest f oreign exchange earner (Sophism, 2006). Of particular importance is that remittances as a percentage of GDP r ose from 3.2 percent in 1990 to 13.3 percent in 2003. Also, as a percentage of export, remittances rose from 22.5 percent to 40 percent within the sam e period (Sophism, 2006). Some researchers have asserted that local aggregate incomes formed only 64% of household spending in 1996 in Ghana, with the rest being derived from external sources such as remittances and external jobs (Grant 2007), indicating that a significant portion of local consumption is related to foreign derived incomes. Similarly, FDI inflows to Ghana have increased significantly in volume ( G hana Investment Promotion Center G IPC 2009 ; Grant 2001 ). Concentrating on current figures, in 2008 Ghana received a total of GH¢ 3.17 billion in FDI funds (about US$ 3.5 billion), representing more than 160% increase over the total FDI received in 2007 (GIPC, 2009). Additionally, the GIPC reports that Ghana received GH¢ 527.63 million (US$ 351.75 million) in FDI funds, which represent 92.95% of the total estimated value and a rise of about 118 % over what was received in the same quarter of 2010 US$161.34 million (GIPC, 2011 ). Unfortunately for Ghana, even though increased inflows of remittances and FDI should be a welcoming development for the better, the restricted geographical concentration of the bulk of these funds in the Greater Accra region and

PAGE 27

27 the Accra metropolis in particular, may not bode well for the environment (Moeller Jensen et al. 2005 ; Grant & Yankson, 2003 ). Primate C it y The contention that Africa is being bypassed by globaliza tion, to a large extent, does not seem to be tenable, given the discussion above. If African economies are connected to the global economy, are cities on the continent equally connected? Generally, urban research have mostly foc such as London, New York, Tokyo, among others, in regard to globalization ( Grant 2007; 2009; Sassen, 2006; Merwe 2004). Merwe has particularly lamented the lacuna of information about African cities as world cit ies researchers have paid l ittle or no attention to these cities in general (Merwe, 2004). This situation is changing, however. Current research efforts have aimed at situating the African urban system s in the global urban system since these cities can no longer be ignored considering their important economy. Thus, researchers with such orientation have argued that, increasingly African urban spaces are being shaped by glo bal forces (Grant 2009; 2007; Robinson 2002). Focusing on the external economic forces, for many countries FDI related activities ha ve been concentrated in large, usually, primate cities, which more often than not, are also the designated capital citie s of these countries (G rant, 2009). I t has been documented that the majority of FDI related projects established in Ghana have consistently been over concentrated in the capital city, Acc ra (GIPC, 2011; Grant, 2009; Yeboah, 2003). For example, in the fir st qua rter of 2011(January March), th e GIPC reports that six out of the ten regions in Ghana benefited directly from FDI related projects. Of these the Greater Accra Region received the majority share, amounting to

PAGE 28

28 about 78% (GIPC, 2011). It is import ant to note that almost all of such FDI related projects in the Greater Accra R egion would normally be located in and around the city of Accra (Grant 2009; 2007; Grant & Yankson, 2003; Yeboah, 2003). In a similar vein, Richard Grant has asserted that the remittance receipts have been invested in some form of real estate property in the city of Accra (Grant, 2009) In effect, t he over bearance of the city of Accra as the main economic hub for the Ghanaian econom y has helped positioned it in a considerable advantage in terms of receipts of these external funds, compared to other cities and towns in Ghana. Historical factors may be cited for this situation. In particular, Accra has enjoyed increasing attention sin ce its inception as the colonial seat of government in the 1870s, benefiting from construction of relatively superior quality of infrastructure. Additionally, the boom in the cocoa industry in Ghana even in the pre independence era has been noted to have had positive impacts on the city which served as the main port for exp orting the commodity (Grant & Yankson 2003). Indications are economic ion trends lead to resources allocation consequences and these in turn result in Bae, 2010: 2). This dissertation research, along with Profe ssor Richard Grant and many others, share the view that this is akin to what has happen ed in Accra since the implementation of the Structural Adjustment Program (SAP) in Ghana in the 1980s, especially the rapid loss of peri urban natural areas which has become greatly intensified in the era of economic globalization (Grant, 2009; Grant & Yan kson, 2003). This necessitates the consideration of global economic forces (FDI and remittances) in mainstream urban

PAGE 29

29 land use theorizing and urban expansion modeling in primate cities in developing countries such as Accra in Ghana, to better understand dr ivers of urban land use change in the era of globalization. However, to a large extent this has not yet materialized. Three main issues come to the fore, following a review of the existing literature: (a) inadequate representation of forces of econo mic gl obalization in mainstream or the classical urban land use theory (b) lack of systematic articulation of spatial financial flows FDI and remittances as integral components of economic globalization in urban land use theorizing; and (c) externally induced p eri urban little attention in the global urban environmental ch ange literature, especially, concerning sub Saharan African cities. In very general terms, these issues c onstitute the problem which is the focus of this study. The timeliness and imperative of this research are underscored by the NRC (2010) report regarding the need to understand the human induced aspects of the changing earth by stating that Although previ ous research has documented shifts in climate, soil erosion, habitat loss, and water degradation, the human role in these changes is often inadequately understood, hindering abilities to predict the magnitu de and timing of future change. (NRC 2010: 3). M oreover, regarding geographical input to globalization studies, Coe and Yeung (2001) have called on geographers, thus: It is imperative for future researchers to continue to evaluate the uneven geographical outcomes of globalization processes in order to a rrive at bette r informed development policies. (p. 14). The exploration of the contemporary physical expansion of Accra and associated loss of peri urban natural areas as globally induced spatial change at the local scale being

PAGE 30

30 advanced in this dissertati on will, hopefully, add some insight s in this important emerging area of geographical research. Statement of the Problem and Objectives The rapid spatial expansion of Accra has been breathtaking by all accounts ( Angel et al., 2011; 2005; Gr ant, 2009; Mol ler Jensen et al. 2005; Grant & Yankson, 2003 ). For example, Shlomo Angel and colleagues note that Between 1985 and 2000, the population of Accra, the capital of Ghana, increased from 1.8 to 2.7 million, a 50 percent increase. Its urban land cover increa sed from 13,000 to 33,000 hectares, a 153 percent increase: Urban land cover in Accra grew more than twice as fast as its population ( Angel et al., 2011: 5). This is not unique to Accra, however. There have been instances in which population dynamics in a gi ven geographic area appear to have little to do with the spatial expansion of the geographical region. For example, Burchell and colleagues have noted that From 1982 to 1997, the U.S population grew by 17 percent while urbanized land grew by 47 percen t. Between 1970 and 1990, the Chicago metropolitan area grew by 46 percent while its population grew by 4 percent. The Cleveland urban area expanded by 33 percent in that same period while its regional popu lation declined by 11 percent. (Burchell et al., 2 005: 38). Obviously, at least to an extent, expa nsion of urban built up area may not have much to per se. It is therefore a welcoming development when researchers are able to isolate important drivers according to so me criteria such explanatory variables (population and GDP or GNP per capita growth). For example, Seto et al. (2011) have concluded that population increase was responsible f or much of urban land expansion in developing countries, whereas in developed countries GDP per

PAGE 31

31 capita growth rate was the most important variable. However, the apparent high significance of population as urban land use change driver in developing countri es may be masking some other not so obvious underlying economic factors. For example, inflows of FDI into an urban economy may be a very significant source of employment for people, affording them the needed purchasing power in order to be able to express r housing and land. In such a situation the advent of globalization and attendant FDI inflows may be contributing to increasing pressure on demand for housing and land. Indeed, even though Seto et al. (2011) did not include FDI in their analysis they rightly pointed it out that FDI could be responsible for some propo rtion of urban land expansion. From the foregoing, therefore, it is obvious that urban land expansion research efforts that will incorporate FDI will help increase our u nderstanding. Similarly, remittances inflow to Ghana has been on the increase and anecdotal these funds (Tipple et al., 1998) Consequently, increasing remittances inflow into Needless to say, therefore, t he combined forces of FDI and remittances could constitute a formidable force of urban spatial expansion, especially if they are over concentra ted in a single primate city or urban area such as Accra But, a re these forces adequately represented in the standard economic theory of urban land use/land value? Objective 1 : Theorizing urban l and u se land v alue under g lobalization: Most urban expansi on models lack a coherent theoretical structure to explain the underlying economic drivers of urban land use / land cover change. Such models are usually

PAGE 32

32 simulated and include most Cellular Automata (CA) and Agent Based Models (ABM); Attempts to link urban expansion modeling to a coherent urban land use /land value theory have invoked the postulates of the neoclassical economic theory of urban spatial structure (Angel et al. 2011; 2005; McGrath, 2005; Brueckner & Fansler 1983). These approaches have most often emphasized local characteristics such as population growth, transportation costs and income (GDP per capita). However, global linkage of economies have resulted in the exposure of the urban economies and their biophysical environment to vagaries of g lobal economic forces and these forces have the tendencies of influencing local land use decisions in favor of conversion of agricult ural lands to urban uses. Given the fact that traditional factors emphasized by the neoclassical economic land use theory lag behind urban land conversion, it is important to explore beyond the local characteristics. However, for the most part the neoclassical economic theory of urban land use/land value still focuses on local factors pertaining to the city ; and cities of all sorts are grouped together and attempts are made to account for the variations in their land cover by invoking traditional factors such as population growth, Gross National Product (GNP) per capita and transportation costs (Angel et al. 2010: 9). Even t hough these factors have been important for a long a time, their relative significance Consequently, urban land use / land value theory rely ing on these factors as the most important predictors to atte ties such as Accra tells only a part of the story thereby failing to provide a fuller understanding of the dynamics of the contemporary land expansion, especially as pertaining to primate cities in dev eloping countries. To better understand contemporary urban expansion,

PAGE 33

33 Ghana, it is imperative that the standard economic theory of urban land use be refined to incorporate the influences of external economic factors flowing from economic globalization. In view of this the first and overarching objective of this study is: to postulated by Alo nso (1964) to reflect realities of the current era of economic globalization whereby global financial resources flow almost seamlessly into urban spaces impacting urban physical growth In order to properly pursue this objective, however, it is important that FDI and remittances are systematically conceptualized as drivers of urban land conversion, thereby helping to identify the pathways through which FDI and remittances would impact contemporary urban spatial expansion. Objective 2 : Linking external e c onomic forces to urban land c onversion : That economic globalization has brought in its wake almost seamless flow of financial resources across locations is no longer a bone of contention in globalization discourses. In particular, as elaborated on above, F DI and remittances continue to flow into urban spaces in developing countries. expansion, therefore, it is necessary that the global economic impulses are traced into the urban spaces by unraveling the path ways through which FDI and remittances impact urban land conversion. In this regard, a number of excellent research efforts have been devoted to describing the linkages between the physical expansion of Accra as influenced by forces of globalization. (Gra nt, 2009; 2007; Yeboah, 2003; 2001 ). Ind eed, Yeboah (2003) has argued

PAGE 34

34 there is a growing body of works that have s ought to analyze the globalization spatial expansion nexus in Accra, via the interplay of global and local forces. Undoubtedly, these studies have provided rich detailed information and analytical insights about how globalization and economic liberalizati on policies have combined to help shape the contemporary expansion of Accra (Grant, 2009 and Yeboah, 2003). However, these attributions have not systematically conceptualized the linkages between FDI and y spatial expansion. Laying out systematic conceptual frameworks that identify the specific pathways through which FDI and remittances impact urban land conversion will enhance understanding as well as facilitate their formal incorporation into the standa rd classical economic t heory of urban land use (Alonso, 1964), which could be developed further to help account for externally induced urban land expansion. In view of this, the second of objective of this dissertation is: to develop two conceptual framew orks specifically aimed at unraveling the pathways through which FDI and remittances impact land conversion in cities where they are over concentrated focusing on Accra It appears that the overconcentration of FDI and remittances in an urban setting tr ansmits their impulses into the landscape, which in the absence of effective land use controls and urban planning could result in loss of peri urban natural areas. This necessitates quantification in order to understand the extent to which this cons titute s a problem. Widely used spatially explicit urban expansion modeling and estimation approaches such as Cellular Automata (CA) and Agent Based Models (ABM) largely fail to incorporate global economic influences, hence may not be appropriate for quantifying

PAGE 35

35 externally induced urban spatial expansion. An appropriate and suitable model is called for. Objective 3 : Q uantifying e xternally induced built up area expansion over time : Land use / land cover changes, especially land conversion, form an important aspect of environmentally significant consumption since these have direct and indirect impacts on the biophysical environment (NRC 1999). Consequently, it has been argued that economic globalization, urbanization and environmental change are three fundamental, intertwined worldwide processes impacting almost all livelihoods on planet earth 2005; Dicken s, 2003; UN 2005 ). Even though the process of urbanization and attendant loss of urban vegetation cannot be said to be occurri ng only in recent times, the advent of economic globalization appears to be intensifying the process. The global linkages of urban economies and the exposure of urban environments and ecosystems to external forces of economic globalization make it imperati ve that we better understand contemporary peri urban environmental degradation, especially in the developing world, in the proper contexts of overriding external influences following economic globalization. More importantly, perhaps, the external economic dimensions of contemporary peri urban environmental change should be properly contextualized; rate o f land expansion with respect to global economic forces should be identified and quanti to a very large extent, urban gro wth models have focused on local factors using probability models and remotely sensed land cover change data to predict urban land use /land cover change, to the neglect of underlying global forces as potential drivers of urban land cover change. In effec t, therefore, specific linkage of globalization forces and loss

PAGE 36

36 of urban vegetation cover through land conversion has been considered less adequately in quantitative terms. As a result our understanding is inadequate. Also, environmental managers, in t heir attempt to curb loss of peri urban ecological resources may be targeting the wrong variables. This situation needs to be addressed given that globalization and urbanization are what we are going to live with into the future if current projections hol d ( Angel et al., 2011 ; 2005 ; UN, 2005; 2003 ). The third objective in this study, therefore, is: to lay the foundations for a simple analytical model of urban spati al expansion (the IUSEEM) which, when properly calibrated, could be used to compute the rate s of loss of peri urban natural areas in Accra as induced by e conomic forces of globalization; emphasizing the increasing importance of FDI as a potent driver of contemporary urban land ex pansion in primate cities such as Accra The main que stions formulat ed in pursuant of the foregoing objectives are given below Research Questions and H ypotheses The following questions are designed, firstly, to identify the pathways through which remittances and FDI affect the urban expansion process. Secondly, to inves tigate the specific statistical relationships between remittances and FDI on the one hand and urban built up area on the other hand. Thirdly, to interrogate how the classical Alonso theoretical framework could be refined to accommodate remittances and FDI as emerging potential urban land conversion driv ers. Finally, to inquire into how an appropriate method can be developed to quantify incremental changes in the urban extent over time as induced by external (global) economic influences in the form of FDI i nflows The questions are followed by related specific hypotheses.

PAGE 37

37 Questions RQ1: What are the specific pathways through which FDI impact urban spatial expansion? Due to relatively less attention paid to researching the FDI urban spatial expansion relation ship, the specific pathways through which FDI impacts the urban built up area have not been identified for systematic conceptualization in order to aid theory building. Answering this question will provide more insight on the specifics of the FDI urban sp atial expansion relationship which will help facilitate urban land use theory building in the face of increasing economic globalization. [Answer attempted in Chapter 2] RQ2 : What are the specific pathways through which remittances impact urban spatial expa nsion? Following the relative lack of consideration, the specific pathways through which remittances impact urban spatial expansion are not properly understood. To properly understand the remittances urban spatial expansion relationship it is imperative that the specific pathways of transmission of remittances impulses in the urban landscape are properly identified and systematically conceptualized, hence the relevance of this question. [Answer attempted in Chapter 3] RQ 3 a : What is the nature of the corre lation, if any, between FDI and remittances on the one hand, and urban built up areas, on the other hand? FDI and remittances constitute the spatial financial flows as conceived in this dissertation. lands in destination countries have b een examined mostly from the perspective of extractive industries such as mining, plantation agriculture and the likes, with less attention paid to FDI impact on urban lands. However, increasing global FDI flows, with the tendency of overconcentration in u rban agglomeration enclaves in developing countries, would warrant that more attention be paid to this FDI urban built up area

PAGE 38

38 relationship in view of economic globalization. Similarly, even though researchers are increasingly paying attention to remittan ces and their impact in origin countries of migrants, the potential impact of remittances on urban land demand has received little attention. In other words, this question seeks to understand the specific nature of the FDI urban spatial expansion relations hip as well as that of remittances and urban spatial expansion. [Answers attempted in Chapter 4] RQ 3 b : In what ways can the classical Alonso urban land use theory be refined to accommodate remittances and FDI as drivers of urban land expansion? This questi on phenomena of remittances and FDI in order to provide a rigorous and theoretically grounded explanation for contemporary urban spatial expansion in light of economic globaliz ation. The rationale here is that the classical Alonso urban land use land value theory is considered by urban economists as the most useful theory of urban land use and land value. However, the advent of globalization seems to have raised new questions as to how realistic the theory reflects current realities of economic far distant places. This underscores the need for the re evaluation of the Alonso theoretical framework land value change drivers like FDI and remittances. [Answer attempted in Chapter 4] RQ 4 a : What is the nature of the relationship between FDI flow volume and rate of expansion of the urban e xtent over time? RQ 4 b : How best can FDI induced increas e in the urban extent be quantified? As noted above, models and methods directly concerned with urban spatial expansion fail

PAGE 39

39 to sources, hence the need for this question. Hypotheses H1 : FDI and remittances are expected to be positively correlated with urban built up area; increasing inflows of both would correspond to expanding urban extent, and statistically significant at the 95 % confidence level. [Statistical tests in Chapter 4]. H 2: The Thunen Alonso theoretical framework can usefully be extended to reflect the influences of FDI and remittances on urban land use values by framing these external economic variables in the context of urban land demand and urban land values c and Chapter 5 ]. H 3: FDI induced urban spatial expansion exhibits non linear and exponential growth characteristics ; increasing FDI volumes correspond to rapid increase in the rate of expansion of the urban extent [Tested in Chapter 5 using the IUSEEM ]. H 4 : The I USEEM is most appropriate for quantifying FDI induced urban spatial expansion. [Tested in Chapter 5, using the IUSEEM]. It is important to not e that since the 4 main chapters of the dissertation have been designed as independent but related essays, the objectives, questions and hypotheses above may be broken into sub units in the appropriate essay if necessary. T he General C onc eptual Framework a nd Theoretical S tructure General Conceptual Framework economies, integrating them together, thereby making these economies interdependent on each other (Dicken s, 2003), as w ell as exposing the biophy sical environment (Leichenko &

PAGE 40

40 can often be traced to global forces s u ch as FDI (Grant, 2009); for example, whereby foreign capital is used to exploit loca l natural resources such as oil (Carmody, 2010). To some extent, therefore, this study shares in the assertion that One of the ways in which the world has changed profoundly in the last two centuries is that sustainability of the biosphere increasingly is threatened by human activities. Globalization contributes to the threat of environmental destabilization by tying distant places closer together, by diffusing new technologies faster to distant lands, and by encouraging changes that may not fit harmonious ly into local cultures or may produce unanticipated effects (Johnson et al., 2010: 46). It must be noted, however, that this author is not opposed to globalization per se. It is the tendency of advocates to attempt to decouple the phenomenon from nature that is troubling. More importantly, therefore, this author belie ves, with numerous others, that H umans cannot be separated from nature for analytical purposes as easily as one might think. Humankind lives in and is an integral part of the natural world; our species is both nurtured by nature and an influential driving force in altering the natural environment (Johnson et al., 2010: 46). This turn of thinking necessitates the reconceptualization of the nature of the human environment relationship ( Cons tanza et al., 1997; Da ly, 1991, for example). In resonance with these views the general conceptual approach to this study is that of a hierarchical systems perspective in which humans are considered integral, but small part of nature ( Costanza et al., 19 97; Rees, 1992; Daly, 1991; Georgescu Roegen 1971 ). Essentially, w ith the intensification of economic globalization the urban biophysical system or environment is exposed to the full vagaries of global economic forces. Thus, local land use/ land cover c hange becomes directly or indirectly link ed to global economic dynamics. This is because one important pathway through which global economic forces manifest in the urban landscape is through real estate development and other land consuming activities such as infrastructure and r oad

PAGE 41

41 construction (Leichenko & en, 2008; 2005; Keivani & Mattingly, 2007; Parnell et al., 2007; Seto, 2005 ). For example, Leichenko and Solecki (2005) have argued that globalization of consumption (housing in this case) has led to the exportation of the American style of single family housing units with large open spaces to the dev eloping countries of the world. The authors concluded that changing global consumption culture is shaping the urban spaces of developing countries el sewhere with the attendant accelerated peripheral land conversion as growing incomes enabled the adoption of American styled housi ng Since urban land conversion usually forms an important part of environmentally significant human consumption activities, the process affects the biophysical en vironment (NRC, 1999). Figure 1 1 shows the linkage between economies (global and urban) to their natural support systems the physical environment (global finite and regional ecosystems). With regard to the broade r global finite ecosystem in relation to the global economy Daly (1991), Rees ( 199 9; 1997; 1992), Sanderson et al. (2002) and Rees and Wackernagel (1994), among others, have elaborated the linkages and consequences of unconstrained economic growth o n globa l ecosystems. Sanderson et al. (2002), in particular, have asserted that the burden economic growth puts on ecosystems is resulting in global disappearance of wilderness through increasing human footprint globally. Land conversion impact is borne directl y by the regional physical environment, unlike other foo he inverse relationship between economic growth and land conversion becomes obvious. It must be noted, however, that these research efforts have largely taken the broadest view based on global scale dynamics. However the same principles could be applied on the

PAGE 42

42 regional and local scales as well for the integration of the urban economy and environment into the larger global economy. Figure 1 1 depicts t he four w orld systems impacted by globalization and macroeconomic reforms. Globalization pushes global economic growth, resulting in capital accumulation which subsequently flows into the urban economy, much of which would be t ranslated into built structures How ever, global and regional ecosystems are fixed. Consequently economic globalization drives urban spatial expansion at the expense of the regional ecosys tem in which a city is located. pty 1994: 24 25). By this conceptualization, this dissertation research advances the argument that spatial financial flows on a global scale, in the form of FDI and remittances can and do affect urbanization conte mporary urban land expansion in urban spaces far and near, especially when such funds are disproportionately concentrated in a single, primate city such as Accra in Ghana Theoretical Structure The systemic conceptualization and theorizing of land use in the context of globalization will enable the explicit incorporation of global economic dynamics into urban land use theoretical framing as well as modeling of contemporary urban spatial expansion; and simultaneously give solid foundation for a better unde rstanding of the role of global economic forces in the urban environmental change process, especially, in rapidly expanding cities in developing countries. useful starting point (Cadwallader, 1985: 24 50). Current formulations (Angel et al. 2011 ; 2005, for example) have sought to reflect external economic dimensions but

PAGE 43

43 inadequately, since they failed to consider the combined impact of FDI and remittances inflows, even though these forms of global funds flow have assumed important dimensions under globalization. For cities of sub Saharan Africa that exhibit some form unds are geographically constrained, often over concentrated in these large cities, thereby fueling rapid expansion through increased demand for housing, infrastructure and related activiti es that consume lands(Grant & Yankson, 2003). At the minimum, it ca n be argued that in addition to recognized local drivers of urban expansion, it is important to consider factors such as the urban hierarchical structure and the relative position of the particular city under study [lumping cities like Cincinnati (USA) and bulk of FDI and / or remittances. Again, the extent of the integration of particular economies into t he broader global economy should have some influence on the city is more likely to be compared with prices in other similar sized cities (or cities of similar status) with prices in cities in the same country (Sassen, 2006). For example, international demand for real estate would compare prices in Accra to those in Lagos rather than comparing pr contract motives as well as self interest motives in which remitters invest in real assets and businesses ( Cha mi et al., 2008; Addison 2004 ). Remittances increase as senders

PAGE 44

44 maximize their utility with respect to self interest as they seek to acquire real assets in their origin countries. The combined effect of these funds inflow manifest in the urban landscape through real estate boom and other construction activities, usually in the p rimate / large cities (Yeboah, 2003; Grant &Yankson, 2003; Tipple et al., 1998; Diko & Tipple 1992); making it imperative that these external variables be integrated into st andar d urban land use theories. Hence, this study seeks to theorize the globalization urban spatial particular reference to Angel et al. (2011; 2005), McGrath ( 2005), Fans ler (1987) and Brueckner & Fansler (1983). After contextualizing FDI and remittances in the standard economic theory of urban land use, the study further explores how the induced global financial flows are implicated in the rapid loss of peri urban natural a reas. The global and urban economies Sanderson et al., 2002; Rees 1999; 1997; 1992; Costanza et al., 1997; Rees & Wackernagel 1994; Daly, 1991 ) Urban expansion, largely driven by economic p rocesses, proceeds in tandem with global economic growth ( Wackernagel & Rees, 1995; Daly, 1991; Georgescu Roegen, 1971); putting stress on ecosystems ( Wackernagel et al., 2006; Burchell et al., 2005; IHDP, 2005; UN, 2003; Goldblum & Wong 2000 ; NRC, 1999 ). Consequently, urban land transformation constitutes an important environmentally significant consumption process, partly driven by flows of external funds into cities. Thus it is important that urban growth models properly incorporate these external dri vers. Harvey (1969) notes that the Allometric law could be used to mathematically link a growth

PAGE 45

45 process over time to spatial change with the potential to enhancing geographic theory aw states that the rate of relative growth of an organ is a constant fraction of the rate of relative growth of the total organism (Savageu, 1979; Nordbeck, 1965 ). Coffey (1981) adds that Allometry refers to the study of size and its consequences, and rela tes the differences in proportions of one component of a system to changes in either the absolute magnitude of the system or a second component of the system. (p. 185). negatively impacting regional ecosystems. To help better understand the underlying global economic drivers of contemporary urban expansion this study attempts to link the ic based exponential function which is integrated with regre ssion analysis to develop the IUSEEM Thus, an important focu s of this dissertation research an attempt to link standard economic theory to urban spatial expansion modeling which is mostly lackin g in many current spatially explicit urban model expansion models but would be most suitable for studying contemporary urban physical expansion in the chosen study area Study Area, Data and Methods: An Overview Study Area Ghana Ghana is a West African co untry with a population of about 23 million people according to year 2008 estimate (CIA 2009) ; and occupying a land area of about 238,533 sq. km The physical size of Ghana is estimated to be equivalent to the size of United Kingdom or a little less tha n the state of Oregon in the United States The

PAGE 46

46 northern boundary is shared with Burkina Faso, with the south of the country bordered by the Gulf of Guinea (in the Atlanti boundary coincide, approximately, with latitudes 4 30' N and 11 N, respectively. The eastern boundary coincides with longitude 1 12' E, with the western boundary coinciding with longitude 3 15' W Engli sh is the official language of Ghana, a colonial legacy inherited from Great Britain but country is home to more than 50 ethnic and linguistic groupings. Incidentally, the Greenwich Meridian (Longitude 0) which passes through London also goes through Te ma, a prominent industrial hub in Ghana, which is gradually assuming a twin city status with Accra. Speaking countries, following colonization by France. rally of the warm, tropical type with drier conditions along the southeastern coastal areas. The southwestern coast is comparatively hot and humid with the northern part of the country being generally hot and dry. Mean annual rainf all ranges between 1,10 0 mm (approx. 43 in) in the north to about 2,150 mm (approx. 83 in) to the south The southwestern part of Ghana is wettest with annual mean rainfall of about 2,100 mm. The Accra Plains, in the Greater Accra Region, within which the city of Accra is locat ed has mean annual rainfall ranging between 750mm 1100mm, considered among the drier areas in Ghana, besides the northern part of the country. The mean annual temperatur e for Ghana ranges between 26 29 degrees Celsius. As would be expected t he vegeta tion types in general, follow the rainfall and temperature patterns. Thus, wooded grasslands are found mostly to the north and south east drier parts of the country, with forests occupying roughly the south central and the southwestern parts of the countr y.

PAGE 47

47 economy is generally recognized by its development partners such as the IMF and these agen cies. As of 2011 according to the CIA (2012) GDP wa s as follows: agriculture (29.9 percent ), industr y (18.6 percent) and services (51.4 percent ). Even thou gh the services sector surpasses the agricultural sector in terms of c still heavily dependent on agriculture which employs about 56% of the estimated labor force of 10.56 million people, with the industry and th e services sectors employing 15 percent a nd 29 percent respectively (CIA 2012 ). Major agricultural products exported from Ghana include cocoa, timber, oil palm, coffee, cotton, shea nuts. Gold has been a major source of foreign exchange for Ghana for a long time. It is important to note that commercial quantities of crude oil has been found in Ghana and drilling is currently (2012) underway which may change the composition of Ghana exports and mainstay of the economy, which currently dependents largely on primary products exports with cocoa US$31.08 billion (at official exchange rat e), with a real growth rate of 7 .7% annually, according t o the CIA (2011 ) ; and estimated to grow at the rate of 13.5% for 2011, ranked 3 rd fastest growing economy in the world (CIA, 2012). The estimated GDP per capita in 2010 was US$2,500 (Purcha sing Power Parity), ranked 180 th in the wor ld, one of the lowest (CIA, 2011 ), even though speculations fastes t growing economy abound The low ranking of the GDP per capita may be partly due to inflation, estimated at 10.7% (at consumer prices) in 2010. For instance, Nigeria, a neighboring

PAGE 48

48 country, had the same estimated GDP per capita of U$2, 500 in 2010 but was ranked 175 th on the same sca le (CIA 2011 ). GPD per capita is estimated to be US$3100, moving up to the 170 th position in terms of world ranking (CIA, 2012) which represents a modest improvement in economic performance over the 2010 pe rformance level. For international trade, a vailable data US$ 10.95 bil lion in 2010. E xports amounted to US$ 5.84 billion in 2009, increasing to US$ 7.892 billion in 2010 (C IA, 2012). Clearly, Ghana stil l imports more than it exports, The annual budget is even more revealing. The estimated revenues for 2010 were US$ 5.381 billion while expenditures totaled US$ 7.757 billi on, creating a deficit of 7.6% of GDP. The population below poverty line (usually defined as people living on less than US$1 per day in the Ghanaian context) is estimated at 28.5% (CIA, 2012). Positive highlights of the Ghanaian economy, compared to its neighbo rs, include unemployment rate estimated 11% which ge ria (5.8%) (Baabereyir, 2009, citing CIA, 2008). Attempts are being made to transform the signing of the Millennium Challenge Corporation Compact in 2006. Furthermore, estimated oil reserves discovered in Ghana is about prospects. roug h colonization, nationalism; and modernity thanks, largely to the implementation of the Structural Adjustment Program SAP (Killick, 2010; Aryeetey Attoh et al. 2010; Grant, 2009 ). Consequently, t he linking of the

PAGE 49

49 Ghanaian economy to the global economic order has been intensified especially after the implementation of the SAP in the 1980s and the 1990s ( Killick, 2010; Grant, 2009; Grant & Yankson, 2003; Yeboah, 2003 ) facilitating rising volumes of FDI inflows into Ghana with high concentration in Accra Even though urbanization level in Ghana is generally low, re cent indications are that the situation is rapidly changing, with increasing numbers of people now living in According t o the Ghana Statistical Service (GSS) data an population increased from 23 in 1960 to 43.9 in 20 00 (GSS, 2002). Further, the number of urban settlements increased from 98 to 364 dur ing the same period (Table 1 1) Even though population numbers have been rising in some urban centers in Ghana, most urban settlements are For example, in 2000, Accra, the largest urban center in Ghana had a population of a litt le more than 2 million (Table 1 2), with Kumasi, the second largest recording just a little over 1 million peo ple (GSS, 2002, cited in Baabereyir, 2009). The statistics given above show that even though low, the level of urbanization in Ghana is on the increase. Seve ral factors have been put forward to explain the increasing urbanization levels, with the key ones being the rural urban migration and natural population increase (GSS, 2002; Songsore, 2003, cited in Ba abereyir, 2009). Related to the rural urban migration phenomenon is the fact that the post independence industrialization drive pursued by Kwame Nkrumah president, favored large cities, particularly Accra, enhancing economic opportunities in tract mor e rural urban migration, and at the same time neglected rural areas helped push people into large cities. Urban

PAGE 50

50 ization level in Ghana has also been impacted by reclassification of settlements usually after census, whereby hitherto rural settleme to urban status. One important observation about the urbanization discourse among urban researchers in Ghana is the fact that more often than not, urbanization has been described in terms of the demographics, with little attention paid to the spatio temporal dynamics until recently. Consequently, explanations of urban expansion have usually attributed the phenomenon to demographic factors of rural urban migration and natural population increase (for example, So ngsore, 2003, cited in Ba abereyir, 2009). Even spra 2000) as is currently happening in Accra. Indeed, one may wonder: what proportion of rural urban migrants has the requisite purchasing power to afford these quality homes that are eating up the peri urban lands? As a consequence of the relative lack of the systematic analysis of the sp atio temporal aspects of urbanization in general in Ghana, the externally induced urban environmental degradation on Accra in the advent of economic globalization may escape detection and may not be mentioned as one of the urban challen ges facing Ghana. One major urbanization challenge in Ghana often cited in the literature is rapid population increase, (grow th rates estimated at about 4.1percent for urban areas and 2.6 percent nationally). It is argued that the rapid growth in popula tion puts pressure on infrastructure and other amenities in urban centers. More specifically, traffic congestion, inadequate and poor housing, overcrowded hospitals and schools, rising unemployment

PAGE 51

51 levels, uns anitary conditions and the likes ( Baa bereyir, 2009 ). What is important for the purposes of the arguments advanced in this dissertation is that urban environmental degradation resulting from urban land conversion would almost certainly not make the list of environmental problems in Ghana Essentially therefore, with the exception of a few researchers like ( Angel et al., 2011; 2005 ; Moller Jensen et al., 2005; Grant & Yankson, 2003 ), most urban expansion analyses have continued to attribute the rapid physical expansion of Accra, for example to demogra phic factors, including rural urban migration. This can no longer be accepted without question, in view of the realities of economic globalization whereby urban land demand may be pushed by forces originating afar. For example, demand for real est ate in Accra may be boosted by employment in FDI related jobs. Moreover, remittances, which is now reputed to be a major foreign exchange earner for G hana (CIA, 2012), has been estimat ed to constitute about 20% of real estate demand in Accra (Tipple et al., 1998) In short, t hese developments warrant a closer examination of the demographic based explanation of urban expansion, especiall y in Accra and similar cities in Ghana and Sub Sah aran Africa generally Accra occupied predominantly by the Ga people, to an urban settlement, is traced back to the a rrival of the European colonialists the Dutch, Dan es and the English in the 15 th century (Adu Boahen, 1975, cited in Baabereyir, 2009). The advent of the Europeans saw the establishment of three main trading posts, namely, Christiansburg Castle (Danes), Usher Town (Dutch) and Ja mes Town (English). After the acquisition of all

PAGE 52

52 these trading posts by the English, t of activities from Cape Coast to Accra, which was later to become the capital of the colonized Gold Coast from where the country was admi nistered. as the capital of Gold Coast was to remain even after the attainment of independence from colonial rule in 1957. Thus, the relatively well developed infrastructure, including roads, government offices and accommodation, schools, health centers, among othe rs, in the city received a major Today, t he city of Accra contains the bulk of and financial institutions as well as government ministries, pa rliament and foreign missions, among other things. The concentration of infrastructural development in Accra did not the most diversified urban economy in Ghana, attracting people from all walks of life including rural urban migrants. Consequently, the city has continuously been me nested in the city of Accra (Songsore, 2003, cited in Baabereyir, 2009). According to the Ghana Stati stical service (2002), about 25 percent in Accra in 2000, with the city contributing about 2 0 percent to t he GDP of Ghan a and offering employment to 10 percent ; cited in Baabereyir, 2009). From humble beginnings, therefore, Accra has been positioned as the main economi c hub of Ghana and also the gateway to the global community through air linkage (for example, airport built in the 1960s is located in Accra). Therefore by the time of the

PAGE 53

53 G hana was well established but the SAP is widely held to have helped Accra maintain the status quo Currently the urban area of Accra is usually defined to include the Accra Metropolitan Area, the Tema Municipal area and the Ga District (Grant & Yankson, 2003), with a population of 2.7 million in 2000, estimated to be 3.2 million in 2005 and projected to increase to 4 million in 2010 (GSS, 2002). In general, Ghana has had to deal with its fair share of environmental challeng es ranging from severe and rec urrent drought s in the northern savannah areas. Th e north is further plagued by intensive cultivation and animal grazing as well as rampant and widespread seasonal bushfires which exacerbated the already precarious environmental conditions. These issues have combined to create serious environmental degradation in the northern part of Ghana. The southern, mostly forested areas are grappling with different set of environmental problems, including unregulated timber exploitation, pollution from mining, as well as soil infertility resulting largely from traditional farming practices (Baabereyir, 2009) Whereas these environmental issues are known and well documented in Ghana, the emerging urban environmental degradation involving the loss of peri urban natu ral areas which is the focus of this dissertation research, has not been systematically explored. Alm ost invariably, therefore, the phenomenal urban growth such as has been experienced by Accra would be faced with environmental problems as well as health issues relating to poor environmental and unsanitary conditions, which according to Songsore (2008) also hold true, to some extent, for urban areas in Ghana in general

PAGE 54

54 Not surprisingly, Professor Songsore cites some of the most pressing environmental pr oblems as I nadequate access to potable water, sanitation facilities, and solid waste disposal as well as smoky kitchens from the use of biomass fuels and insect infestation among others ( Songsore, 2008: 8). He further asserts that These hazards directl y threaten the health of citizens arising from the life and death immediacy of malaria, respiratory illness, diarrhea, and other infectious and parasitic diseases (Songsore, 2008 : 8 ). therefore deserve the attention of urban researchers, spatial ramifications associated with the rapid expansion of Accra such as sprawl ed urban development, in the absence of effective land use controls and efficient physical planning, is emerging as a formidable environmental challenge with the potential to undermining urban en vironm ental sustainability (Grant & Yankson, 2003 ). It is the view of many globalization (Aryeetey Attoh, 2010, Grant, 2009; Moller Jensen et al., 2005; Grant & Yankson, 2003). In particular, the potential undermining of ecosystems services resulting from re moval of vegetation cover co uld also be considered dium to long term through loss of ecological resources and consequently environmental degradation. drivers of population and income growth may be blamed, it seems that this is being driven, at least in part, b y external economic forces of globalization su ch as FDI and remittances whose advent has been greatly facilitated by the SAP. In this regard the following three statements are worth quoting to buttress the main argument of this dissertation research :

PAGE 55

55 Ac cra has the most diversified economy of any area in the country. The the introduction of structural adjustment policies (SAPs) in 1983, which has resulted in a sprawl along all frontie rs (Grant & Yankson, 2 003, citing AMA, 1999; Amuzu & Leitmann, 1994: 5). One of the most visible impacts of the liberalization program is the dramatic rise in the number of new foreign companies that have established operations in Accra, which is currentl y headquarters for 655 foreign companies (Grant 2001: 1005, cited in Grant & Yankson, 2003: 69). T he liberalization of the financial sector has enabled foreign currency transactions, and much of this money originates from Ghanaians working abroad and ca n be used for invest ment in Accra housing (Grant & Yankson, 2003: 69). The statements above, at least to some extent, seek to link the rapid physical expansion of Accra in recent times to forces of economic globalization which trigger spatial financial flo ws (FDI and remittances) of which Accra is generally believed to be the major beneficiary in the Ghanaian context Consequently, the need to explore the apparently emerging phenomenon of externally induced spatial expansion in Accra and possible negative repercussions on peri urban ecological resources cannot be overemphasized, giving further impetus to the imperative for this dissertation research. Data and Methods The four essays comprising the main focus of this dissertation are designed as inde As a result, the data and methods relating to particular essays are fully described and detailed where appropriate. For brevity therefore it is suggested that readers refer to these data and description of methods in Chapter 3 and Chapter 4. Consequently, only very brief overviews are given here. Statistical dataset covering annual FDI flow amounts and annual remittances flow amounts have been ob tained from United Nations ( 2011; 2012 ) online databases. Also,

PAGE 56

56 processed remotely sensed urban land cover change data for Accra based on published 2011 ; 2010 ) and Moller Jensen et al. (2005). The methods employed in this research are quantitative in nature. The theoretical propositions derived from the refined Alonso theoretical framework are tested using regression modeling and analysis. Specifically, FDI and remittances are included in log transformed cross sectional multiple regression models ( Chapter 4). The log transformation made the data suitable for the application of the Ordinary Least Squares (OLS) regression estimation method. Another desirable characteristic of the log transformation is that the co efficients can be interpreted as elasticity (percentages), thus eliminating the problem of unit of measurement. T he study takes advantage of this to sum up the estimated co effici ents in the (IUSEEM) formally developed in Chapter 5. IUSEEM represe nt s an attempt or approach to contemporary urba n expansion modeling, potentially suitable for a rapidly expanding primate city in a developing country context ; as induced by economic forces, with particular reference to FDI. The model integrates regression into an allometric based exponential function, which is simple, analytically tractable, dynamic, non linear and spatially explicit in nature. (Full description of this estimation method or model the IUSEEM, is presented in Chapter 5 of the dissertation) Significance of this D issertation Firstly, this research is designed to contribute to theory building in contemporary urban and economic geographical research in the context of economic globalization by examining the significance of the external economic dimensions in the spatial expansion process. Thus, a classical economic theory of urban land use to incorporate FDI and remittances as

PAGE 57

57 and emerging external economic drivers of conte mporary urban spatial expansion, especially in primate cities in developing country setting. This has the potential to making immense contributions to theory construction in geographical research in general and urban geography in particular, from an in ter disciplinary perspective; utilizing concepts, ideas and theorie s from geography and economics. Secondly, a methodological contribution is also envisaged as the study adopts a new approach to develop an Allometric based urban growth model which can be used to monitor urban environmental degradation as induced by economic forces, mostly originating from global sources such as FDI. This model could be potentially replicated in primate cities elsewhere with similar characteristics as Ac cra. Some of the main leveled against traditional urban analytical models by advocates of more modern methods (Batty 2005) include being static and linear However, the modern spatially explicit, dynamic, non linear simulation models such as Cellular Automata (CA) gen erally lack the suppo rt of standard economic theory; can be computationally intensi ve and methodologically complex, sometimes placing them beyond the means of many researchers, especially those in developing countries. Desirable characteristics of the IUSE EM formulated in this study include simplicity, dynamism, non linearity and spatially explicit. In this regard the model will be potentially useful to many modelers, especially those interested in the contemporary urban spatial expansion in primate cities in developing countries as induced by economic forces (FDI in particular). Thirdly, Geographical research, especially the tradition that focuses on understanding human environment interactions could be usefully employed to help understand the human dimensi ons of global environmental change.

PAGE 58

58 The human dimensions of global environmental change are assuming significant proportions as world population grows and human consumption expands as well as continued acceleration of urban ization ( IHDGEC, 2005 ; NRC, 1999 ). Thus, this approach to urban geographical research, in addition to attempting to make urban land use theory more realistic in view of growing global interconnectedness and increasing spatial interactions, has the potential to contribute meaningfully to efforts aimed at mitigating current urban environmental problems, especially in large cities of developing countries. The study seeks to highlight the magnitude of peri urban land consumption that could be attributed to the external forces of economic glo balization, particularly, FDI. Assessing the globally induced aspect of the vegetation cover depletion will contribute to our understanding of the human dimensions of the current environmental change in the peri urban areas which are leading to environmen tal degradation. Further, this study seeks to integrate concepts and ideas across the di sciplines of urban geography, economic geography and urban economics as well as drawing on concepts from biological science. Such an approach is likely to be effective since no single discipline can adequately explain the complex issues of globalization and urbanization and their combined impact on the urban biophysical environment. In addition to broadening the scopes of the respective disciplines through integration o f concepts and ideas, a better understanding would be gained in the area of linking temporal process to spatial form, thereby facilitating thinking across geographical scales. Primate cities in developing countries such as Accra in Ghana are grappling wit h the environmental consequences of globalization and urbanization (Gr ant, 2009; Moller Jensen et al., 2005; Grant & Yankson, 2003 ). For exa mple, Moller Jensen et al.

PAGE 59

59 (2005) have asserted that effective environmental policy and urban planning are urgently needed in Accra. It is not too difficult to see, therefore, that the continued intensification of global linkages of cities and economies makes it imperative that urban policy makers and managers be cognizant of the impact of external forces on their objec t of interest the city and its environment if the challenges posed by these forces are to be adequately taken care of and in order to stem the tide. The broader perspective of this research, therefore, is the anticipation that urban planners and urban e nvironmental managers and conservationists will find insights from this research usefu l in their decision making, giving due consideration to the influences of external economic forces of globalization. At the minimum this will help ensure that the extern al dimension is recognized and anticipated. Moreover this dissertation research is designed to produce 4 publishable peer reviewed journal articles. These planned journal publications, when realized, will be of great benefit to the advancement of knowledg e in general and geographical research in particular in addition to enhancing my prospects as a budding geographer in academia teaching and research. Finally, just as any worthwhile piece of scholarly work is invaluable to the academic community, it is the fervent hope of this writer that this dissertation research would contribute to energizing urban and economic geographical theorizing with the potential to opening up new avenues for intellectual pursuits in geographical research aimed at increasing th e understanding of contemporary economic globalization and spatial financial flows as well as their ecological consequences in urban areas of developing countries.

PAGE 60

60 Overview of Ch apters and Organization of the D issertation This dissertation is composed of 6 chapters, 4 of which are designed as independe nt but interrelated publishable articles (Chapters 2, 3, 4, and 5), anticipated to be published in peer reviewed journals Chapter 1 provides an overview of the dissertation, and a general statement of the r esearch ques tions, objectives and hypotheses ; as related to the topic of assessing the impact of Foreign Direct Investment and remittances, as driven by globalization and economic liberalization, on urban peripheral change and the expansion of the urban fr inge (study region: Accra, Ghana). Chapter 2 offers a conceptual and theoretical analysis on the specific pathways through which FDI impact urban spatial expansion; basically, expansion of the employment base as associated with FDI related jobs, the multip lier effect, induced employment for the population and migrants who gain entry to FDI related jobs in stimulated sectors, and demand by expatriate workers and FDI firms. These pathways increase the demand for housing and land, especially for those workers in the skilled portions of the labor force (middle and upper income segments; those who have the greatest purchasing power) typically a segment of the economy who prefer locations in the peripheral areas of the city, which leads to urban sprawl. Chapter 3 examines the relationships between remittances and urban expansion, as enhanced by globalization and economic liberalization policy; with an overview of the determinants and consequences of remittances as they relate to such variables as economic growth rates and urban land consumption land value/use to examine the theoretical links between intensification of globalization and liberalization policies and its impact on urban spatial stru cture using a mathematical framework; focusing on the disproportional impact FDI and remittances have on primate

PAGE 61

61 cities such as Accra. Empirical evidence is presented to support the model for a selected time period (via a Regression analysis). Chapter 5 m akes the case that FDI can be directly implicated in the physical expansion of Accra and the associated loss of peri urban land and related ecological resources. The argu ment is that FDI driven change is responsible fo r rapid urban change based on a re view and synthesis of the literature and the formulation of an Integrated Urban Spatial Expan sion Estimation Method to estimate the impact of FDI induced expansion. A series of results and data are presented t o analyze the expansion of Accra over the peri od span ning from the mid 1990s to 2001 2011 Chapter 6 provides a summary and overview of the dissertation, its implications and limitations, an d directions for future research Note: The reader is reminded that ideas, arguments and statistical figures may be repeated across the dissertation chapters. The reason is that the four main chapters (Chapter s 2, 3, 4 and 5) have been designed to be independent, publishable journal articles. This necessitated the inclusion of all relevant facts, arguments and stati stical figures in the relevant chapters to render each one coherent and understandable on its own, resulting in the repetitions across the chapters since they all address the one topic of the dissertation I t is hoped that readers interested in specific a sub topic of this dissertation can read the relevant chapter and gain useful insights without necessarily having to read the whole document.

PAGE 62

62 Figure 1 1. Transmission of spatial financial flows into urban spaces and co nsequences on ecosystems

PAGE 63

63 Figure 1 2. Study Area Location of Ghana from the Environmental Systems Research Institute (ESRI).

PAGE 64

64 Figure 1 3. Study Area Location of Accra from the Environmental Systems Research Institute (ESRI).

PAGE 65

65 Figure 1 4. Study Area The Greater Accra Metropolitan Area (GAMA) Source: Adapted from Mo ller Jensen et al. (2005)

PAGE 66

66 Table 1 1 Growth i n Gh ana's urban population: 1960 2000 Y ear Total population Percentage urban No. of urban settlements 1960 1970 1984 2000 6744234 23.0 98 8555211 28.9 135 12296081 31.3 203 18912079 43.9 364 S ource: Ghana Statistical Service (2002). 2000 Pop ulation and Housing Census Table 1 2. Population growth in t he ten largest cities in Ghana 1970 2000 City Total Population 1970 19 84 2000 Accra Metropolis 634809 1005834 2045125 Accra C entral 624091 969195 1658937 Kumasi 346336 496628 1170270 Seko ndi Takoradi 143977 188203 289593 Tamale 83653 135952 202317 Ashaiman 22549 50918 150312 Tema 60767 100052 141479 Obuasi 31005 60617 115564 Koforidua 46235 58731 87315 Cape Coas t 56 601 6 5763 82291 Wa 13740 36067 66654 Source : Ghana Statistical Service (2002). 2000 population and housing report. S pecial report on urban localities

PAGE 67

67 CHAPTER 2 FDI AND URBAN SPATIA L EXPANSION : TOWARDS A SYSTEMATIC CONCEPTUAL FRAMEWORK Background C hapter 2 is guided by the question: What are the specific pathways through which FDI impact urban spatial expansion ? Inv estigating this is important because t he link between FDI and contemporary urb an spatial expansion is still less understood; largely lacking a well established theoretical framework which take account of current realities of economic globalization. Develo ping a formal theory to help address this situation would benefit from a well constructed conceptual framework which will help clarify the linkages for formal articulation into a theory. In this regard, Eg an (2002), has asserted that most of the literatur e point to the fact that conceptual writing is a necess ary component of theory writing; hence the necessity for this chapter which seeks to unravel the linkages between FDI and urban spatial expansion through the construction of a coherent and systematic c onceptual framework. Economic liberalization has coupled with economic globalization to reinforce global capital flows across geographic space without hindrances as barriers are removed to enhan ce economic integration (Dickens, 2003). There are various ty pes of global financial flows those meant for investments and loans for which returns are expected but there are also those for which returns are not expected including remittances and aid. In this chapter the focus is on the capital category. Two main forms of capital flows are distinguished in the literature investment and lending (Hanink, 1994). Investment comes in two main forms portfolio and direct investment. According to Hanink (1994) it entails control of the

PAGE 68

68 Thus, an important component of the global capital flow FDI in which foreigners bring their financial resources into a country with the aim of establ ishing business operations, going into partnerships with local entrepreneurs or some combinations of both. Such operations are sometimes generally termed as M ulti National E nterprises (MNEs) or Trans National Corporations (TNCs) by some researchers (Hanink 1994). Increasing FDI flows have become a key indicator of economic globalization. Whereas world output has approximately doubled since 1970, FDI has quadrupled during the same period (Dicken, 1997). And according to the World Investment Report, FDI i ncreased from US$400 billion in 2000 to a record US1, 833 billion in 2007 (UNCTAD 2007; Bank of Ghana, 2009). Undoubtedly, currently, the global economy has taken devastating hi t, he current FDI recovery is taking place in th e wake of a drastic decline in FDI flows worldwide in 2009 (UNCTAD 2010: xvii). The organization 37 per (UNCTAD 2010: xvii). This has led to wha t has been lobal inflows are expected to pick up to over $1.2 trillion in 2010, rise further to $1.3 1.5 trillion in 2011, and head towards $1.6 2 trillion in 2012. Thus, in very general terms, the current global economic downturn notwithstanding, the global economy as a whole is predicted to experience a rebound soon (UNCTAD, 2010: xxiii). This expected global economic recovery will almost invariably be associated with increasing levels of FDI f lows, even t has been noted espite its impact on FDI flows, the global crisis has not halte d the growing

PAGE 69

69 that multinationals will continue to look for viable economic opportunities in different geographic locations. In this regard it is heartening (at least to those interested in international business) to know that: The reduction in sales and in the value added of foreign affiliates of transnational corporations (TNCs) in 2008 and 2009 was more limited than the co ntraction of the world economy (UNCTAD 2010: xviii). More specifically, the UNCTAD (2010) has asserted that foreign affiliates of multinationals are contributing significantly to the GDP of countries around the world, ached a n historic high of 11 per cent. foreign employment increased slightly in 2009, to 80 (p. xviii). For most developing and transitional economies in particular the increase in multinational operations in their economies is a blessing as they bring much needed capital in the form of FDI, which unlike loans, they do not have to worry about repayments. According to the UNCTAD (2010): addition, they acco unted for 28 per cent of the 82,000 TNCs worldwide in 2008, two percentage points higher than in 2006. This compares to a share of less than 10 per cent in 1992, and reflects their growing import ance as home countries as well. (UNCTAD, 2010: xviii). Ther efore, it would be rational to expect that economic globalization and associated FDI flows across geographical space, particularly the flow into developing economies will continue to increase. These almost seamless flows of capital across geographical spa ce have spatial ramifications in the urban landscapes in FDI destination countries (Harvey 1985) which are often overlooked in the theorization and analysis. Harvey (1985) has considered this link of global capital and urban development, from the perspec tive of the capitalist mode of production, asserting that urban expansion has

PAGE 70

70 been the result of excess accumulation of capital, which is then invested in real assets in the urban economy, to put it simply. Even though the purpose for this chapter is to l ink global capital to urban land conversion, this author does not see it necessary to use the approach by Harvey (political economy Marxism) because cities have been in existence for as long as human civilization, long before there was a formal articulat ion of Capitalist mode of production (capitalism). For instance, before colonization of Ghana from the 1800s and before the introduction of Western system of governance, the Ashanti people of Ghana had a functioning city Kumasi, which had become well esta blished and had nothing to do with Western capitalism as far as one tell. Yes, money does have the potential to cause urban growth, but this does not necessarily mean that those who build structures in the city are capitalists. Consequently, this chapte r focuses on unraveling the pathways of transmission of FDI impulses into the urban land expansion process, without interpreting it as being the result of capitalism or otherwise. The FDI determination theoretical literature is well developed with a long history, even predating the current era of economic globalization. This aspect of the FDI theoretical literature can be loosely described as focusing ma inly on the question of: what motivate entrepreneurs to set up operations in foreign countries as oppo sed to producing in their domestic eco nomies and exporting elsewhere? As would be expected this literature offers virtually nothing in the way of explanation of the potential impact of FDI on urban lands in destination economies or countries. Thus, theo retical frameworks bothering on FDI impact on destination economies should provide the platform for understanding FDI impacts such as economic growth, poverty alleviation,

PAGE 71

71 household welfare, as well as impacts on specific localities. Even though commendab le research have been done in this direction, in general terms, little has been done by way of helping lay the theoretical foundations for the comprehensive understanding of the specific pathways through which external economic impulses such as FDI inf low influence the urban extent. Among other things the geographical concentration of oth er areas, which could potentially put undue pressure on demand for housing and land which would in turn necessitate the need to construct new infrastructure such as roads. It is imperative that a better understanding is gained in this emerging area of ex ternally induced urban change by developing a formal theoretical framework which will enable urban researchers (within geography and beyond) to account for realities of economic globalization as it affects distant geographical spaces, albeit urban spaces i n primate cities in developing countries. The cardinal objective of this chapter, therefore, is to build a conceptual framework for linking FDI and urban spatial expansion under globalization, which would provide the platform for the formal theoretical art iculation of FDI as an urban land conversion driver in an era of intensifying globalization. Specifically, the chapter seeks to identify the pathways through which FDI would influence contemporary urban land conversion, especially where FDI is over concen trated in large / primate cities in developing countries and lays, in a systematic manner, the conceptual framework for understanding the globalization urban expansion relationship. According to Rudestam and Newton (2007):

PAGE 72

72 A conceptual framework, which is simply a less developed form of a theory, consists of statements that link abstract concepts (e.g. motiv ation, role) to empirical data. (p. 6). For Lynha m (2002 ) conceptualization serves as the first step in developing a theory. Following along these lines, therefore, unraveling the linkages between FDI inflows into urban spaces and urban spatial expansion will help pave the way for the explicit incorporation of FDI as a plausible explanatory variable in urban land use theorizing as well as urban expa nsion modeling. Therefore, the importance of this exercise cannot be overemphasized. In broader context of this dissertation, this chapter is one of the two chapters that are designed to provide the conceptual frameworks for developing a formal urban land use land value theory (Alonso, 1964). Globalization, Africa and FDI Flows in C ontext Traditionally, developed or industrialized countries are known to receive the largest chunk of global FDI flows, namely, North America, Nort h West Europe and Japan. Consequently, in 1970 1980, large parts of Asia, Africa and Latin America had little or n o FDI inflows (Hoogvelt, 1997 ; Hanink 19 94). Ankie Hoogvelt (1997), in particular, that capital flows and trade almost invariably were concentrated in the rich / industrialized co untries and some parts of Asia. However, the situation is rapidly changing. For a large measure, developed countries still account for siz eable share of global F eveloped countries did not return to FDI flows to this group of economies fell some 7 percent to US $ 527 billion, despit e the more than 40% over 2009 levels, representing an increase of US $ 56 billion, and has

PAGE 73

73 been described as the single biggest increase in FDI among the major economic region 2011: 4). For developing countries overall increases in FDI inflows have been modest but nonetheless significant. What is even more revealing is the gradual increase in receipts of FDI in developing and transition economies and its resilienc e even in the face of global economic crisis. Thus, notes the UNCTAD, despite worldwide economic downturn, the UN reports that the global FDI inflows to developing and transition economies rose to 43 percent in 2008, after six years of continued growth (UN CTAD, 2009). Th is was however to decline by 27 percent to US $ 548 billion in 2009 (UNCTAD, 2010: xviii xix). Even so (UNCTAD2010: xviii xix) claims that this group of countries fared better than developed countries, as a group, in terms of resiliency, whi ch suffered a decline of 44 percent More over, according to UNCTAD Put in perspective, for developing and transition e conomies, t heir share in global FDI inflows kept rising: for the first time ever, developing and transition economies are now absor bing h alf of global FDI inflows (UNCTAD, 2010: xviii xix). It appears, therefore, that contemporary FDI flows are becoming more geographically mostly develop ing and transitional economies. Are FDI inflows to developing and transition economies merely ephemeral or enduring? The United Nations seem s to suppor t the latter view, stating that There are some major changes in global FDI patterns that preceded the global cr isis and that will most likely gain momentum in the short and medium term Firstly, the relative weight of developing and transition economies as both destinations and sources of global FDI is expected to keep increasing. These economies, which absorbed al most half of FDI inflows in 2009, are lea ding the FDI recovery. (UNCTAD, 2010: xvii). How does Africa fare in terms of economic globalization and receipts of FDI?

PAGE 74

74 Africa has had a long history of interactions / interconnections with places far and beyond, ostensibly predating the current era of global economic integration (Cooper 2001) from slave trade, colonization to economic exploitation as argued by Bond (2006). In contemporary development discourses in relation to globalization, two main views on Africa prevail (Bond, 2006). The first view asserts that Africa has been bypassed by economic globalization; with the second one alluding to the fact that Africa has suffered from globalization (Bond, 2006; Carmody 2010:1). It must be noted that until re cently independent countries in Africa were suspicious of their former colonial disguise of economic reforms. Thus, after gaining independence, following the end of the Second World War, most countries in sub Saharan Africa suspected that foreign investments, foreign ownership of production were neocolonial strategies that have been instituted in order to recolonize the region (Killick 2010, Broadman, 2007; Voltoirenet, 2006, C armody, 2010:41 42). This prompted many leaders of the newly independent countries to nationalize companies with foreign ownership; in some cases privately owned companies by citizens were not spared. For example, Tony Killick (2010) has noted that the fir st independent government of Ghana followed this p ath to a large extent (Killick, 2010). To a very large extent this hostile attitude towards foreign business interests in their domestic economies and the general aversion for private entrepreneurship eff ectively restricted FDI inflows (Killick 2010). Currently, however, the general consensus emerging among developed countries and developing ones alike is that FDI is desirable, if not essential for both economic growth and poverty reduction (Broadman, 20 07; Voltoirenet, 2007, Carmody, 2010:41 42 and NEPAD 2001). Indeed,

PAGE 75

75 EPAD, 2001). Largely as a consequence of these developments, some global mainstream economy, albeit through natural resources exploitation and export of primary product from the continent and increasing demand for mobile phones, with some modest achievements in attracting transnational corporations (Carmody, 2010: twenty fastest growing eco nomies and there are now thirteen middle income countries in sub Saharan Africa (Ndulu et al., 2007 cited in Carmody 2010: 34). Sector wise, the United Nations note d that Contraction of investment in the services sector in Africa was less pronoun ced than in other sectors. Sustained by expanded activity, the telecommunications industry became the largest recipient of FDI inflows. Recovering commodity prices and continued interest from emerging Asian economies are expected to feed a slow upturn in F DI flows to Africa in 2010 (UNCTAD, 2010: xix). In terms of economic growth in general, Africa narrowly missed the projected 7 percent annual growth rate which has been predicted to ensure accelerated growth and enhance sust ained poverty reduction by 2015 (Carmody, 2010: 40 41; UNDP, 2005; NEPAD, 2001). Africa in general is not faring too badly. In the case of Sub Saharan Africa, it is estimated that Overall growth was 6.9 percent for sub Saharan Africa according to the IMF (2008), just shy of the 7 perce nt that some international plans argued was necessary fo r sustainable poverty reduction (NEPAD, 2001; cited in Carmody, 2010: 34). To some degree globalization in Africa so far has been mostly extractive focused on outward flows of wealth and resources such as oil and agricultural products like cocoa.

PAGE 76

76 evolved significantly, particularly, in the past decade, as a result of the rise of china and the global technological revo 3). Thus, c urrent globalization in Africa is being led, predominantly, by Asia, particularly, India and China and predicated largely on natural resources exploitati on (Carmody, 2010; Klar, 2008). In particular, Padraig Carmody has argued that China is playing a cardinal role in the globalization of Africa, even though mainstream research works have largely negle cted to assert this fact. As a result he has suggested that T he rise of China in Africa, through the intensification of economic and i deational flows and migration, should be seen as an evolution of globalization on the continent rather than sep arate from it (Carmody, 2010: 3). Chinese trade with Africa estimated to have grown by about 45.1 percent in 2008, worth about US$ 107 billion (Center for Chinese Studies, 2009). Indeed, China is widely believed to be the single most influential trading partner now; with Africa trailing only USA in terms of imports from China (Carmody, 2010: 3). In sum, current developments on the African contin en t in relation to globalization seem encouraging, leading Carmody (Carmody, 2010: 4 5) For example, w hol e of inf the global technological revolution has swept over the continent, through the wide scale adoption of mobile phones (Carmody, 2010: 4 5). Some researchers have suggested top spot of mobile phone producing countries and its consequences on price has contributed to making mobile phones accessible to more Africans (Carmody, 2010): Thus, the rise of China and the spread of the information technology revolution to Africa are not empirically dis tinct events, but interconnected

PAGE 77

77 axes of the wider process of globalization generated, mediated, and promoted through state, corpo ral and other social structures. (p. 3). transitory but is rather a structural break from the past (20 06a, 2 cited in Carmody 2010: 34]) Professor Carmody continues that Nonetheless, the economist Intelligence Unit (predicted that seven of the top ten fastest growing economies in 2009 would be in sub Saharan Africa (Malawi, Angola, Ethiopia, Congo Brazzaville, Djibouti Tanzania, and Gambia (Carmody, 2010: 6). This is a welcoming development for the continent as whole and individual countries such as Ghana. It can be argued therefore, that t he economic success of Africa is related to mainstream economic globalization, considering the flow of global capital into African economies, particularly in the form of FDI. FDI inflows into the continent have had something to do with the rapid transform ation of most economies that implemented liberalization policies in the 1980s and the 1990s. In this regard, many researchers are of the view that risen to 1.8 percent, attracting absolute amounts roughly three times what they were per year in the 1990s ( Carmody, 2010:40 41 ; Br oadman, 2007, Voltoirenet, 2007 ). For example, regarding Accra in Ghana, Grant (2009) has asserted that The establishment of new foreign companies accelerat ed to unprecedented levels in the liberalization period: over 90 percent of all companies currently active h ave been established since 1983 (p p 27 28). In terms of jobs for Ghanaians it is estimated that FDI projects has created over 7 3,000 jobs as of 2004 (Mmieh & Owusu Frimpong 2004: 591). Moreover, foreign firms have been noted to have accounted for about 75 percent of increase in export earnings (UNCTAD 2002:16). It seems, therefore, that even though the bulk of the FDI flows have taken place pr edominantly among developed countries such as among UK, USA,

PAGE 78

78 Japan and Germany ( Dicken s, 2003 ; Beenhakker 2001; Hanink, 1994 ) such funds have become an important source of development financing in developing countries in view of the associated job creatio n and other economic benefits to the destination countries in the developing world ( GIPC 2009 ; Glick 1998; Mallampally & Sa u vant, 1 994 ). Among developing countries, therefore, there is intense competition for such private funds flowing from external sou rces, chiefly from developed and newly industrialized countries ( Grant 2009; Grant & Yankson 2003; Grant 2001 ; Malla mpally & Sauvant, 1999; 1994 ) as this quote exemplifies: With liberal policy frameworks becoming commonplace and losing some of their trad itional power to attract FDI, governments are paying more attention to measures that actively facilitate it. Still, the economic determinants remain key. What is likely to be more critical in the future is the distinctive combination of locational advantag es and, especially, created assets that a country or regio n can offer potential investors (Mallampally & Sauvant 1999: 37) For example, following the doctrine of the economic liberalization brought on by the SAP, Ghana established a unit called the Ghan a Investment Promotion Center (GIPC) to help create conducive investment climate to at tract FDI into the country (Grant 2009; 2001; GIPC 2009). It is estimated that Ghana has benefitted from over US$ 2.8 billion of foreign investments, largely due to th e activities of the GIPC and administration of political processes in the country. It would be interesting to review the theoretical underpinnings of FDI flow What drives FDI? This will aid the interrogation about the linkage, if any, betwee n FDI flow determination theories and urban spat ial expansion. In other words: are there any theoretical signals to the effect that FDI could feed into housing and land dema nd in developing host economies ?

PAGE 79

79 A Review of the FDI Theoretical Literature FDI ca n be divided into three main components, namely, extraction, manufacturing and services (Moran, 2011). This FDI theoretical literature review has been categorized into two main groups FDI flow determination theories and theories on the impact of FDI on destination countries/economies. This latter category will provide the platform for the new conceptual framework which is the subject of this chapter identifying the link Perfect C ompetition and FDI Early theorizati on of FDI derived mainly from the Heckscher Ohlin background where perfect competition underlies the postulates. In these approaches, FDI has been modeled as capital flows from one location to another based on the concept of relative capital abundance The basic tenet of the reasoning was that there is spatial inequality in the distribution of capital among countries some countries h aving more capital than others. In essence, therefore, this result in di fferential r eturns on capital, implying higher returns on capital in capital scarce countries and lower returns on capital in abundant Consequently, capital would move from capital abundant countries (rich or developed countries) to capital scarce countries (poor or less developed countries), thereby exhibiting the notion of comp arative advantage of some sort. Mundell (1957) attempted a formalization of the perfect competition approach to FDI theorizing. He explained the effects of factor movements in a two sector, two countries and two factors (222) framework principally deriving from the Heckscher Ohlin model. Holding constant the possibility of factor equalization constant Mundell postulated that product and factor prices wo uld remain constant following capital inflows (Mundell, 1957; Lattore, 2008). A further deduction was that capital inflows and imports

PAGE 80

80 were substitutes in that increasing capital inflows would tend to reduce the imports of the recipient country, resulting in what has been characterized by the catch 2008: 3). However, these outcomes did not sit well with some researchers, particularly, the one way of FDI; and the notion that fact or prices are not responsive to capital inflows was particularly disconcerting (Caves, 1971; 2007; Brown et al., 2003 ; MacDougall 1960). The general argument here was that empirical evidence pointed to two way movements of capital. This observation, to incapable of explaining the empirical evidence, needing a modification or reformulation. MacDougall (1960) took up this challenge, postulating a one sector economy receiving capital inflows. However, unlike Mundell, MacDougall allowed for the responsiveness of labor productivity and local capital to capital inflows whereby rent on local capital gets reduced whereas labor productivity increased (MacDougall 1960; Lattore 2008), thus allowin g for two way flows of cap ital or FDI It must be noted that both theoretical contributions (Mundell and MacDougall) still alluded to the fact that capital would move from capital abundant (rich, developed) countries to capital scarce (poor developing) countries. However, the empi rical evidence once again pointed to the contrary as it was observed that the so called capital rich countries were simultaneously origina tors and recipients of capital. I ndeed, it has been noted that t he bulk of FDI flows originates in (and is directed t o) developed economies, which should be capital abund ant (Barba Navaretti & Venables, 2004 ; Markusen, 2002), alluding to the existence of mar ket imperfections for FDI flows.

PAGE 81

81 Market Imperfections and FDI According to Kindleberger (1969) everyone agrees on one point, in a world characterized by perfect compet ition, FDI would no longer exist. Thus, it was realized that the assumption of perfect competition was no longer tenable as it became increasingly obvious that individual firms may have monopolistic c ontrol over their innovations and inventions, at least in the sho rt to medium term (Kindleberger, 1969). countries, this could cause a break from the capital movement s framework (Kindleberger 1969 ). It is also important to note that early theoretical frameworks did not distinguish between portfol io and FDI per se Stephen Hymer (1976) demonstrated that the view of FDI as capital movement which followed higher rates of return was contrary to what empirical data presented. Drawing on his original doctoral dissertation, Hymer argued that due to market imperfections the type of assets possessed by multi national enterprise s cannot be transferred easily. He went on to i dentify two main types of market imperfections operations claiming that this violated the assumptions of perfect competition); and (2) resulting from costs of transaction s (Lattore, 2008 ; Denisia 2010). In a nutshell, H ymer argued that if a firm decides to set up a subs idiary abroad against all odds ; the firm must have some advantages relative to firms operating domestically in the foreign country. For example, the ind igenous firm has knowledge of indigenous market, consumer tastes, the legal and institutional framework of business and local business customs (Denisia 2010 ; Lattore 2008 ; Kindleberger, 1969 ), which on the surface of it should put the MNEs subsidiary pu rporting to operate in the domestic economy at a

PAGE 82

82 disadvantage. Hymer explained that control of foreign enterprises was important to appropriate fully the returns on advanta ges the firm possessed (Denisia, 2010). Thus, FDI becomes an alternative choice fo r minimizing the cost of uncertainty. Furthermore, in order to survive the imperfect market, direct ownership (affiliates) is the viable strategy to be adopted. Essentially, therefore, Hymer concentrates on firm specific advantages (FSA) which can be ex ploited in foreign countries by MNEs (Lattore, 2008). These advantages may include ownership of a brand name, innovations, unique marketing know how, patents, management skills and perhaps secure sources of finance as well as economies of scale derived fro m plant size, etc (Lattore, 2008). Therefore, industries where vertical integration takes place will be characterized by high seller concentration, which creates barriers to entry (Lattore 2008). It is also important to note that Stephen Hymer (1976; 1 960) was the first scholar to have distinguished FDI from other foreign capital movement such as portfolio investments ( Denisia, 2010; Moran, 2010 ; Lattore 2008). path breaking efforts to advance his line of ar gument. For example, Cave (1971) has features of the markets in mu ltinational corporation system. According to Cave, a firm can enter a foreign market through horizont al or vertical integration Horizontal integration describes a situation whereby MNEs produce the same kinds of products in foreign markets as being produced in the home country of the multinational corporation. Cave argued that product differentiation wo uld underlie horizontal FDI integration which is enabled by the possession of assets/advantages not available to local firms in the host country. On the other hand, vertical FDI integration is engendered by the MNEs

PAGE 83

83 attempts to avoid uncertainties over sup plies, and prices. The involvement of MNEs in this type of integration, therefore, assures that supplies are carried out by affiliates. Consequently, Cave has argued that the most important skill is the ability to differentiate a product. The Product Life Cycle and Beyond Raymond Vernon (1966) worked along sim ilar lines of thinking as Hymer but elected to describe the expansion of multinational enterprises (or FDI flows) as distinctive stages. His theory is widely known as the Product Life Cycle theory PLC (Vernon, 1966; Lattore, 2008) and entails three main stages. Vernon argued that most production processes went through these s tages or product life cycle. The first stage is characterized mainly as innovation phase whereby firms produce for the local mar ket (US in this case). The key importance of this stage is that it enables the firm to satisfy local demand with sufficient profits to allow for it to engage in research and development, thereby enhancing product quality. In this stage therefore, resear ch, development and the production processes are e fficiently coordinated (Lattore, 2008). The second stage focuses on market seeking outside the home country (mainly Western Europe, at the time of the formulatio n of this theory). In the stage, the product begins to be exported (to Western Europe). In a third stage, some competitors arise in Europe. If conditions are favorable the firm will establish foreign subsidiaries there to face the increased competition and it may also establish subsidiaries in le ss developed countries to have access to cheaper labor costs to enhance its competitiveness. Vernon (1979) himself recognized that the circumstances had changed rapidly since his theory was developed and that this had considerably weakened its predictive power. However, the product cycle theory provided a framework under which a number of

PAGE 84

84 authors dealt with crucial questions about FDI. Hirsch (1976), for example, worked on to serve the foreign market. His model takes into account the costs of managing production abroad as well as the asset specificity of the capital owned by MNEs in a simple but complete framework (Denisia, 2010; Lattore, 2008) The cardinal importance of the product life cycle theory was the novel attempt to give relevant explanation specifically relating to the expansion of US enterprises in foreign markets, especially after the Second World War. Consequently Lattore (2008) notes that the theory s FDI as a reaction to the threat of losing markets as a product matures, and as a search of cheaper factor cost Again, this theory is concerned mainly with FDI flows among capital rich countries or capital abundant countries not comprehensive enough to accoun t for other types of FDI flows, especially those flowing from developed to developing countries. In response to this deficiency, John Dunning (1977) attempted to synthesize the existing FDI theories to formulate a more c omprehensive theory (Denisia, 2010 ; Lattore, 2008 ) resulting in what has been described as the Eclectic Theory of foreign direct investment (Denisia, 2010; Lattore 2008 ; Dunning, 1977 ). His theory is based on the Ownership specific adva ntages, Locational advantages and Internalization Drawing on existing theories the eclectic theory attempts to incorporate the three main strategies through which firms try to get involved in the global economy or operate in foreign mark ets, namely, exports, FDI and contracts ( i.e. licensing, technical assistan ce, management and franchising) ( Denisia, 2010; Lattore 2008 ). Essentially, the Eclectic Theory posits that firms that would seek to engage in

PAGE 85

85 FDI must have some advantages whic h include ownership location and internalization sim ply abbreviated as OLI (Lattore, 2008: 57). Ownership advantages refer to intangible assets, which are, at least for a while exclusive possessions of the company and may be transferred within transna tional companies at low costs, leading either to higher incomes or reduced costs (Dunning 1977). Dunning (1973; 1980; 1988) theorized that there are three main types of specific advantages that can be recognized Monopoly advantages in the form of privi leged access to markets through ownership of natural limited resources, patents, trademarks; technology, knowledge broadly defined so as to contain all forms of innovation activities; economies of large size such as economies of learning, economies of scal e and scope, greater access to finan cial capital and so on (Denisia, 2010: 57). Locational advantages are mainly derived from endowments such as natural resources and cheap labor. Location advantages of different countries are the key factors to determin ing who will become host countries for the activities of the transnational corporations. The specific advantages of each country can be divided into three categories (De nisia, 2010) e conomic, political and social: (a) The economic benefits consist of qu antitative and qualitative factors of production, costs of transport, telec ommunicati ons, market size, among others; ( b) Political advantages: common and specific government p olicies that affect FDI flows; ( c) Social advantages: includes distance between the home and host countries, cultural diversity attitude towards strangers, among other things Supposing th e first two conditions (OL ) are met, it must be profitable for the company to use these advantages, in collaboration with at least some factors outside

PAGE 86

86 the co untry of origin (Dunning, 1988 ; 1980; 1973 ). This third characterist ic of the eclectic paradigm offers a framework for assessing different ways in which the company will exploit its powers from the sale of goods and services to various agree ments that might be signed between the companies As cross border market benefits is higher the more the firm will want to engage in foreign production rather than offering this right under license, franchise. Eclectic paradigm shows that OLI parameters are different from company to company and depend on context and reflect the economic, political, social chara cteristics of the host country. These advantages are assumed to increase the wealth creating capacity of a firm an d hence the value of its asset. FDI will occur only if the ownership specific advantages can be profitably internalized This means that a firm will get more advantages to use its outputs by itself rather than to sell or lease them to foreign firms. The basic tenets can summarized as follows: Enterprises which have the greatest opportunities for internalizing activities will be the most competitive in foreign markets. If only ownership specific advantages are possessed, a firm will co nsider FDI exporting, and licensing as equal. In case those ownership specific advantages can be internalized the firm will prefer FDI and exporting to licensing. Lastly, if the ownership specific advantages can be profitably internalized across national boundaries because of the location specific f actors of a foreign location, then the firm will definitely prefer foreign direct investment t o both exporting and licensing. Essentially, therefore, productions can be explained by these three condi tion s. The possession of ownership advantage is a necessary prerequisite of foreign involvement. According to the internalization advantages, the enterprises will exploit ownership advantage by way of

PAGE 87

87 export of FDI rather than by contractual resource ex changes, whereas the FDI preferred to exports will be chosen where location advantages favor a foreign rath er than a domestic production. Several variants of the OLI paradigm have been formulated. For example, the Multinational Enterprise Theory is develo ped from the ecle ctic theory by Markusen (1991). Markusen argued that the most basic question that any theory of the multinational must answer is why multinationals exist (Denisia 2010). Why are markets not served by exports from foreign firms or by pro duction by locally owned firms? Markusen invoked the concept of know ledge based, firm specific assets (FSAs), arguing that these are proprietary assets of the firm embodied in such things as the human capital of the employees, patents or the reputation of the firm. There are two reasons why these knowledge based assets are more likely to generate FDI. First, knowledge based assets can be transferred easily back and forth across space at low cost. Second, knowledge often has public goods characteristic in that it can be supplied to additional production facilities at very low cost. The knowledge based assets have important implications for the efficiency of the firm and in turn for market structure. FSAs are important in the theory of the multinational. I t implies that a major component of trade between parent firms and subsidiaries is management, e ngineering, and other services. This theory can apply to both export and import competing industries, and indeed non traded industries as well. The Multinatio nal Enterprise Theory explains what types of firms and industries are likely to be dominated by Multinationals. This theory has thr ee major motives which are as follows; ( 1) market ori ented FDI; ( 2) raw material based FDI ; ( 3) co st reducing FDI Market o r iented FDI is naturally oriented toward countries with large home markets, often in response to rapid market growth

PAGE 88

88 and/or t h e threat of tariff imposition. Much of this type of investment takes place in the advanced industrialized countries or in the lar ge rapidly growing less developed countries, and some in the newly industrializing countries (Denisia, 2010; Lattore, 2008) Raw material based investment has traditionally been the source of much satisfaction between host and home countries. Linked very closely with colonial patterns, the historical factors which have led to extraction technology reposing in advanced countries and the raw materials in colonized, less developed countries have increased tension. The expropriation and indigenization have le d to massive restructuring of operations and the search for new forms of institutional arrangements (Denisia, 2010; Lattore, 2008). Cost reduction is one motive through which FDI c an be generated (Denisia, 2010; Lattore, 2008). In any industry labor fac tor is a major element of cost and one which can be reduced by the act of relocation (Denisia, 2010; Lattore, 2008) The search for cheap labor has led to multinationals reorganizing their operations so that the labor intensive stages can b e relocated (Mar kusen, 1991; Vernon, 1966). Markusen concluded that multinationals arise in industries which have certain technological and informational characteristics, but that trade orientation of multinationals in a particular country is determined by more tradition al principles of comparative advantage and trade barriers. The foregoing review, which is an attempt to trace the beginning of theorization of FDI, shows that among the FDI de termination theories, the OLI, t he PLC and the Multinational Theory hold sway an d that there is no single agreement as to which is better. Together they constitute the FDI determination theoretical framework. However, unlike remittances determination theories that hint on asset accumulation in migrant

PAGE 89

89 home country, which relate to ur ban land demand, these theories do not offer insights regarding how FDI feed into urban land demand as a result of over concentration of FDI in geographical space in destination countries. What might be m ore relevant to the objective in this chapter would be the theoretical literature that examines the impact of conversion/expansion. Unfortunately, such theoretical framework does not exist in the FDI theoretical literature What is available have focused mainly on empirical research. However, the need for such a theory that links FDI to urban expansion cannot be overemphasized, given the realities of economic globaliza tion currently. Even though developed countries receiv e the bulk of FDI, it is generally acknowledged that FDI is crucial for developing countries in their bid to better the lot of their people. Thus, Mallampally and Sauvant (1999) have noted that While FDI represents investment in production facilities, its significance for developing countries is much greater. Not only can FDI add to investible resources and capital formation, but, perhaps more important, it is also a means of transferring production technology, skills, innovative capacity, and organizationa l and managerial practices between locations, as well as of accessing inter national marketing and networks ( Mallampally & Sauvant 1999: 35 36). Economic growth naturally entails the use of resources and in developing countries, mostly characterized by extractive and manufactur ing FDI, the use of natural resources is usually paramount in attracting the FDI in the first place In this regard, Borensztein and colleagues note : The literature on the FDI growth relationship is vast for both developed and de veloping countries. The basis for most of the empirical work focuses on neoclassical and endogenous growth models. It is often claimed that FDI is an important source of capital, that it complements domestic investment, creates new jobs opportunities and i s in most cases related to the enhancement of technology transfer, which of course boosts economic growth. While the positive FDI growth linkage is not unambiguously

PAGE 90

90 accepted, macroeconomic studies nevertheless support a positive role for FDI especially in particular environments. Existing literature identifies three main channels through which FDI can bring about economic growth. The first is through the release it affords from the binding constraint on domestic savings. In this case, foreign direct invest ment augments domestic savings in the process of capital accumulation. Second, FDI is the main conduit through which technology transfer takes place. The transfer of technology and technological spillovers lead to an increase in factor productivity and eff iciency in the utilization of resources, which leads to growth. Third, FDI leads to increases in exports as a result of increased capacity and competitiveness in domestic production. Empirical analysis of the positive relationship is often said to depend o trade regimes and the degree of openness (Bo rensztein et al., 1995; 1998, cited in Ajayi 2006: 2). Considerations for resource usage should i nclude the need for land resources for crops/agriculture as well as land for settlement expansion where g lobalization fuels urbanization, for example. Lambin and Meyf roidt (2011) have asserted that Addressing global land availability is made more complex by the processes of economic globalization. In particular, cropland expansion and forest conversion are accelerated by global scale spatial dynamics caused by the displacement, rebound, cascade, and remittances effect s (Lambin & Meyfroidt, 2011: 3466). Thus, economic globalization has the tendency to reduce natural land availability, if FDI should influence urban land demand (imp acting urban settlement sizes). It is obvious that urban land expansion has not been considered here. However, in their own co nceptu alization (Lambin & Meyfroidt, 2011: 3465 ) natural land availability is represented as: Land for nature = Total land area (Agricultural area + Settlements) 2 1 This would imply that change in settlements affect total land area and consequently change in land for nature. However, urban land expansion has not been linked to FDI resulting from intensifying economic glo balization in a manner such as advanced in this

PAGE 91

91 dissertation However, Seto et al. (2011) have alluded to the plausibility of FDI induced urban land expansion, but identified GDP per capita growth rate as the major contributor to urban land expansion in d eveloped countries, and population increase as most relevant driver of urban land expa nsion in developing countries. Essentially, what abling aspect that is brought on through FDI jobs creation and related activities. In effect there is the need to know exactly how population in crease fuels urban land demand. In cities where new settlements are characterized by tenements at the periphe ry one can safely assume that people do not need to be relatively rich to be able to acquire and develop the land for residential purposes. Hence, in this situation, even new rural urban migrants will contribute significantly to the physical expansion of the city t hrough demand for land. However, if the contemporary expansion of the city shows peri Accra, it is not entirely tenable to attrib ute the expansion to absolute population numbers per se, since in most cases ordinary local residents normally do not have the purchasi ng power to acquire such homes. Konadu Agyemang (2001) provides a fuller discussion of the housing issues in Accra. It is a gainst this backdrop that C hapter 2 asserts that FDI provides the necessary purchasing power through employment. Essential ly, C hapter 2 argues that population increase must be backed by purchasing housing and land, in order to affect the urban extent, other things being equal. Even though it is difficult to untangle the demand for land generated by FDI via employment for rigorous analysis, it is worth

PAGE 92

92 trying. Untangling the pathways linking FDI a nd urban land expansion could potentially lead to a more complete understanding. This is imperative since our understanding in this area of research is incomplete, given the recent assertion by Seto e t al. (2011) to the effect that A lthough demographic a nd economic factors capture a fair amount of urban land expansion in China and India, much of the observed expansion in other regions cannot be accounted for by the explanatory varia bles of the model (Seto et al., 2011: 1 ) r efering to their work on the m eta analysis of ur ban land expansion studies. The authors further note t hat additional factors that may interact with the fundamentals of population and economic growth in determin ing urban expansion. First, the role of international capital, be it foreign direct investment, overseas development assistance, or other types of financial instrument, is key in driving development and especially urban expansion in developing country cities and is excluded from the analysis (Seto et al. 20 11: 7). This is precisely why this dissertation in general and this C hapter 2, in particular is timely and potentially useful exercise, especially developing well thought out conceptual framework for subsequent theoretical development. FDI and U rban Spatial Expansion: A Conceptual F ramework into urban agglomerations could have important spatial ramifications, including the e xpansion of the urban extent. in hand with rapid rates of urbanization and expansions of the built 27). Even tho ugh no causal inference has been advanced it is not too difficult to see that opening up of the Ghanaian econom y to external impulses could have played some role in this process.

PAGE 93

93 The impact of foreign direct investment on urban land conversion can be c onceptualized at 3 levels which feed into land demand and subsequently land conversion, other things being equal: via unemployed people already living in the city who get employment from FDI related jobs; people who are drawn from other parts of the countr y, and even from outside the country, into the city and get employed in FDI related jobs, and land requirements of foreign investors for expatriate housing and also to establish operations and other related activities, constituting direct demand for land b y foreign investors. These are outlined below. FDI Impact on Land C onversion via Income and P opulation At any point in time it is conceivable to reason that there is a surplus labor in the us in the urban hierarchy of their country, given their attractiveness to potential migrants who perceive relatively made available, new income earners enter the u rban economy, which would have the impact on land demand. The implication here is that even though the general income levels in the country for individuals as measured by G DP per capita may remain city increases as a result of gaining employment in FDI related operations. Population migration from rural areas to urban centers is not a recen t phenomenon by any stretch. In particular, colonial legacy and governmental urban biases have helped create overconcentration of infrastructure and other amenities in urban centers of most (Aryeetey Attoh 2010); a nd contributing to the current situation whereby such cities receive disproportionately

PAGE 94

94 large portions of FDI funds. Once FDI starts flowing into a country and it is co untry both from other towns and cities as well as rural areas into the city receiving disproportionate share of FDI. It must be noted that this FDI induced type of migration is distinct from rural urban migration in the literature in the sense that these people are likely to be relatively highly skilled, may be changing jobs into a high wage paying FDI urban migration. It is crucial to emphasize that the latter group of migrant would find it difficult to exert any appreciable pressure on demand for land that would cause new land conversion, given the income status and lack of skills to command significant purchasing power. At best such migration will only fuel demand for housing at the lower end of the income stratum which flies in the face of the high cost residential buildings in the periphery which has ation or greenfield If demand is satisfied through the latter channel it will fuel demand for new land and consequently land conversion and land expansion. FDI Impact on Land Conversion: Direct Land D emand Even though classical economic theory recogniz factors of production, mainstream economics has not directly considered FDI impact on urban land expansion. In this study land is literally interpreted to mean the physical space on which some infrastructure could be constru cted to facilitate the operation of the firm. For example, Direct FDI demand for land entails the requirement to establish warehouses, offices, plants, housing for expatriates among other things Figure 2.1 illustrates the pathways through which FDI tran smits impulses into the biophysical

PAGE 95

95 environment. Essentially, incidence of FDI in the urban economy results in job creation and arrival of expatriates. Job creation offers employment for both unemployed city residents and potential migrants from other pla ces, sometimes including from neighboring countries. Thus, more people gain the purchasing power. On the other hand expatriates demand housing. Expatriates, together with the newly employed in FDI related jobs will increase the demand for housing. Part o f this demand may be met through gentrification and occupation of existing homes. However, if the bulk of this demand is made up by middle to upper income earners, their demand will consist of quality, spacious, usually new homes. This will fuel demand f or new lands, usually. Added to this demand is the demand for land by FDI firms for their operations such as warehousing. Part of this land demand may be met through infilling and use of existing structure s. But almost invariably, a significant part of t his land demand may be met through the conversion of new land / natural areas, given the lower prices of land in the urban periphery for the most part (Hanink, 1997). The bro ken lines and boxes in Figure 2 on use of existing homes, infilling and use of existing structures absorb some of the demand pressure on housing and land. not likely to be significant enough to acc ommodate substantial part of the land demand generated by FDI Thus, the argument is advanced to the effect that the bulk of this FDI induced demand will be translated into the conversion of new lands / natural areas, especially in the absence of effective land use management policy and ecologically conscious planning practices. Assumptions : It is assumed that this city exhibits

PAGE 96

96 infrastructure, population, economic a gglomerations, among other things. Consequently, it is not far fetched to suggest that the bulk of FDI will flow into this Another important assumption is that agricultural land market, usually assumed in standard urban economic model ing is v irtually non existent (Chapter 4 ), hence demand for urban land overrides all other consideration, constrained only by customary land tenure practices in the case of Accra and other urban centers in Sub Saharan Africa In a nutshell, land will be made available once there is effective demand for it. Implications of the conceptualization are that: (a) Even though city population may increase in absolute numbers, this may not necessarily result in the type of demand for land that will result in new lan d conversion. Invariably, however, population increase power contribution afforded by the inflow of FDI as an external economic stimulus. It is important that this is untangled in order to properly understand the role of FDI in ties of developing countries. The statistical evidence of positive association does not give the complete picture therefore our understanding is less complete. (b) In primate cities: Policy interventions that target urban population control / management i n lieu of urban expansion should be augmented with effective FDI management policy otherwise the triggers described above will likely undermine their efforts. A comprehensive FDI management is in order, to ensure a more geographically diversified FDI dist ribution into smaller urban centers. Of course such an approach will

PAGE 97

97 come at the expense of benefits of economic agglomeration, perhaps. This makes it even more imperative that a more concerted effort be geared towards a more complete understa nding in th is area of research. For example, at what level does agglomeration result in negative consequences? What levels of FDI can be accommodated by the overconcentration? What proportion of land demand can be attributed to FDI activities? employment in FDI related jobs? These are questions that will help illuminate further the complexity of economy environment relations in general and in the urban setting in particular. Summary It is widely held that conceptualization enhances the proces s of theory construction. C hapter 2 has attempted to provide a conceptual framework which can be used to aid the t heoretical framing of FDI induced urban spatial expansion, especially in primate cities in developing countries se ttings. The chapter identifies and describes the pathways thr ou gh which FDI flow s into an urban space may influence the expansion of the urban extent: i.e. via employment of previously unemployed city residents who gain employment in FDI related jobs; via employment of induced population / migrants who gain employme nt in FDI related jobs; via direct demand by expatriate workers and demand by FDI firms for operations. Together these pathways contribute to increasing demand for housing and land and if this demand is not satisfied by existing housing stock or construc tion of high rise high density structures on existing lands or the use of existing lands and structures, new natural areas must be converted to urban uses.

PAGE 98

98 Another dimension of the externally (FDI) induced demand for housing and land is that it is likely that middle and upper class housing will dominate the market because FDI related jobs usually employ skilled, educated workers who more than likely will not earn low wages. and can ex quality housing and land, which in the case of Accra, get constructed in the peri urban areas. Preliminary indications are that it does appear that FDI overconcentration in a large/primate city could feed into the expansion o f the urban extent, even though there is no FDI related theory that seeks to link this increasingly important spatial financial flow to contemporary urban spatial expansion. It is argued in this dissertation that FDI and remittances both have the potent ial to impact the urban land conversion process. Given the foregoing conceptual analysis which gives more insight into the FDI land conversion relationship, it can be argued that similar insights would be needed for a better understanding of the remitta nces urban land conversion relationship. This is the focus of Chapter 3

PAGE 99

99 Figure 2 1. Pathways linking FDI to urban spatial expansion construct Unemployed city residents employed Exp atriates Urban s patial/land expansion Densification: G entrification & existing homes Demand for housing FDI jobs New land /natural area conversion Induced population employed Demand for Land Densification: Infill & existing structures FDI Firm op erations Net Foreign Direct Investments (FDI) inflow

PAGE 100

100 CHAPTER 3 CONCEPTUALIZING T HE LINK BETWEEN REMI TTANCES AND URBAN SPATIAL EXPANSION Background C hapter 3 seeks to explore and conceptualize the relationship between remittances and the urban spatial expansion process. Specifically, the analysis is guided by the question: What are the specific pathways through which remittances impact urban spatial expansion? The main argument advanced is that r emittances inflow, if significantly concentrated in a single city, impact urban land expansion positively (more remittances would reflect in more land conversion) Thi s process is especially accentuated with the advent of economic globalization and its atte ndant improved communications; fuelling unprecedented levels of human migration which generate rounds of remittances flows (IOM, 2010). Addison (2004) defines remitt ances Money transfers by migrants from their domiciled countries to their countries of origin, especially from wealthy countries to developing ones, cont inue to increase in rate and volume every year (Ratha et al., 2009; Pablo et al., 2009; Maim bo & Ratha, 2005; Faini, 1994). For example, Ma i mbo & Ratha (2005) note that Officially recorded remittance flows to developing countries exceeded US$125 billion in 2004 making them the largest source of development finance a fter foreign direct investment. (p. 2). Beyond development financing such funds continue to be critical in terms of providing economic sustenance for numerous families ( Mallampally & Souvant 2008 ; Addison 2004 ). The rate and size of remittances flow are closely tied to economic globalization (Maimbo and Ratha 2004) and liberalization policies implemented in most developing countries such as Ghana which facilitate easy inflows of funds acr oss locations ( GIPC,

PAGE 101

101 2011 ; Grant & Yankson 2003 ). It is widely held that remittances are of immense and increasing importance to developing countries ( Maimbo & Ratha 2005 ; Addison 2004; Ratha, 2003 ). Dillip Ratha notes that in 2001 remittances flow t o developing countries totaled US $72.3 billion, which was second only to flow of FDI Ratha (2003) Moreover, remittance has been found to be a more stable flow than other forms of funds such as portfolio and even FDI (Ratha 2003). Thus, for many develop ing countries, remittances from citizens abroad have assumed important dimensions as inflows of private capital and official aid decrease steadily (Ratha 2003). In effect, therefore, remittance has been recognized as one of the reliable financial flows w ith increasing importance as a form of development financing for developing countries ( Acosta et al., 2009; Ratha et al. 2009; Maimbo & Ratha, 2005 ; Faini, 1994). Moreover Maimbo & Ratha (2005) note that b eyond development financing, such remittance fun ds continue to be critical in terms of providing economic sustenance for numerous families. An important emerging aspect of remittances flow is remitters quest for investments in small businesses and acquisition of real property such as house s and lands ( World Bank 2010 ; Amuedo Dorantes & Pozo, 2006; Addison, 2004). According to the World Bank (2010) data Ghana received about US$17.210 million in 1995 in remittance funds which increased to about US$ 99.185 in 2005, representing about 476 percent increas e within the ten year period. It seems that these positive attributes of remittances have contributed to origin countries. Such potential impacts on urban land dem and and consequent land conversion to urban uses, especially if the funds are concentrated in primate cities of developing countries is closely associated with migrants asset accumulation motive

PAGE 102

102 articulated in the literature (Amuedo Dorantes & Pozo, 2006 ), but less articulated directly. Thus, even though Amuedo Dorantes and Pozo (2006) have emphasized the increasing importance of the motive of migrants to acquire real assets in their origin countries, including housing construction in cities, a formal the ory of remittance induced urban land expansion is yet to materialize. However, it appears as migration is projected to increase going forward and remittances levels continue to rise in quantum and importance in developing countries, the emerging phenomeno n of remittance induced urban land expansion is likely to assume hig h importance. To the extent that globalization reinforces international migration and given the ageing economically active population in rich countries; and so far as economic liberalizati on ensures almost from developed economies to less developed ones is likely to continue into the future (Chami et al., 2008). In order to effectively manage urban land use in lieu of urban sustainability given that globalization has come to stay, it is imperative that understanding about globalization related urban spatial changes be improved especially remittances induced urban spatial expansion which appears to be ch aracterizing primate cities in developing countries. Developing a formal theory to explain the relationship between remittances and contemporary urban spatial expansion in primate cities of developing countries will aid our understanding of how external e conomic forces in the form of remittances contributes in shaping the contemporary primate city in a typical developing country setting. A systematic con ceptualization will benefit such a theoretical development process (Egan, 2002).

PAGE 103

103 The goal in C hapter 3 therefore, is to start this theoretical journey by providing a conceptual framework, systematically formulated and designed to unravel the pathways through which remittances impact urban spaces they happen to be concentrated. The importance of conceptual frameworks for theorizing has been underscored by frameworks are developed to account or describe abstract phenomena that occur under rt that: Generally speaking, a research dissertation is expected to contribute to the scholarly literature in a field and not merely solve an applied problem. Thus, identifying a conceptual framework for a research study typically involves immersing onese lf in the research and theor etical literature of the field. (Rudes tam & Newton, 2007: 6). In a nu tshell it is plausible to conceptualize, to the effect that remittances will generate rounds of economic activities through direct employment as well as dem and for construction and other materials. Combined, these activities will exert pressure on housing and land deman d resulting in land conversion, especially with the intensification of economic globalization. Globalization and Remittances: An O verview Econ omic globalization forces and economic liberalization policies have combined to reinforce unprecedented spatial flows of economic resources almost seamlessly across locations (Dicken s, 2003; Hanink 1994). One of such economi remittances This private, unrequited flow of funds has assumed important dimensions for developing countries, in particular, as other financial flows dec line in relative terms (Addison, 2004). Remittances and migration are two phenomena which have been immensely im pacted by increasing interconnectedness of economies and deepening

PAGE 104

104 spatial inequality in global economic prospects (Dicken s 2003 ; Abler et al., 1 971). Thus, Abler and colleagues have pointed out that E conomic change is never ubiquitous in kind or intensi ty. Migrations are the ways people smooth out the peaks of per capita opportunity which differs from place to place (Abler et al., 1971: 197 198). In effect, migration and perceived enhancement of economic well being of migrants at appear to be intertwined. The International O rganization for Migration asserts that Migration is considered one of the defining global issues of the early twenty first century, as more and more people are on the move today than at any other point in human history. (IOM 2010). Not surprisingly, therefore, the unprecedented international migration has brought in its their countries of origin (IOM, 2010). Thus, remittance flows are closely tied to interna tional migration, which has been enhanced by globalization as well as economic liberalization policies, especially in developing countries (MOI 2010; Maimbo & Ratha, 2005). Moreover, remittance has been found to be a more stable flow than other forms of funds such as portfolio and even FDI ( Addison, 2004 ; Ratha, 2003 ). As noted previously, b eyond development financing, such remittance funds continue to be critical in terms of providing economic sustenance for numerous families as well as savings towards the acquisition of real property such as houses and lands in migrant hom e countries (Amuedo Dorantes & Pozo, 2006; Addison 2004). Consequent to this has been the growing interest in this phenomenon among researchers (Addison 2004), largely following the seminal work by Lucas and Stark (1985). R esearch on remittances can be grouped broadly into (i) determinants of remittances and (ii) impacts /consequences of remittances in recipient economies. In

PAGE 105

105 the latter group efforts have emphasized remittance impa cts on variables such as economic growth (Chami et al. 2008), poverty reduction ( Mazzucato et al. 2008 ; Stark,1985); and real exchange rates appreciatio n in receiving economies ( Acosta et al. 2009 ; Faini, 1994 ), among others. Since these have been exte nsively discussed in the literature already but do not explicitly/directly treat the impact of remittances on ur ban land expansion, this work focuses mainly on building the conceptual framework based largely on the asset accumulat ion motive (Amuedo Dorante s & Pozo, 2006); derived from the remittances determination theoretical literature. Remittances Theoretical literature Mainstream Theories The mainstream remittances determination theoretical literature, for the most part, has concentrated on the motive s behind remittances (Lucas & Stark 1985). These theories generally recognize the following underlying motives of migrants altruistic, implicit family contracts loan repayments and insurance; an d self interest s (Addison, 2004; Stark, 1991; Stark & Luc as 1988). It has been noted, however, that majority of the remittances flow theories have bordered mainly on the altruistic and self interest motives (Chami et al., 2008). The theories built on altruistic motive posit that migrants feel a moral obligati on to help cater for their families they left in their home countries; as Addison 2004; Stark, 1991 ; Lucas & Stark 1988; 1985 ). Theories based on the altruistic motive dominated research efforts on remittan ces in earlier times (Addison 2004). For example, Lucas and Stark (1985) have stated that Certainly the most obvious motive for remitting is pure altruism the care of a migrant for those left behind. Indeed, this appears to be the single notion underlyi ng muc h of the remittance literature. (p. 902).

PAGE 106

106 This line of reasoning led Lucas and Stark (1985) to formulate a utility function of family members left at home. St ark and Lucas (1988) have argued further that altruism, apart from having merit on its own, aids self interest motivations of migrants through a re enforcement mechanism whereby recipients of remittances reciprocate by taking housing and land (Chami et al., 2008). Current theoretical frameworks give increasing consideration to the self interest motives (Chami et al., 2008). With regard to self interest motivated remittances flow the ories (Chami et al. 2008), the main assertion is that remitters send money purely for self interests such as investing in business or long term projects (Lu cas & Stark,1985; Addison, 2004). For example, Lucas and Stark (1985) have noted that remittances are In this regard families are used as agents who then are compensated by the migrants in monetary form as the recipients take care of S tark 1985). for high levels of remittances by ensuring that their shares of inheritance of family lands are increased. In effect, therefore, remittances serve as an indirect investment in real assets for the benefit of the migrant. This is corroborated by Amuedo Dorantes and Pozo (2006) in their study of the behavior of Mexican migrants in the USA, but articulation of the asset accumulation motive in this case was mor e direct as opposed to the indirect approaches discussed earlier in the literature (Hoddinott 1994; Lucas and Stark, 1985) prior to Amueodo Dorantes and Pozo (2006).

PAGE 107

107 From the foregoing, the general state of the theoretical literature on remittances, there fore, is that research efforts appear to fall under one of the categories above, with mutual exclusivity an either or situation This has been pointed out by Chami et al. But one point that the recent literature does not sufficiently ack nowledge is have argued that both motives could co exis t, stating that Indeed, a critical reading of the literature on the theory of remittance determination finds ampl e support for the idea that multiple motivations and intended uses for remittances can coexist (p. 23). This, in effect, would imply that a remitter could ask the recipients that part of the money be used by the family for their own needs and part be inv ested in some business or be used to purchase real assets for him/her (the migrant), in the origin country. It has generally been acknowledged that altruistic motive reinforces self interest (or exchange) motive and these are not nece ssarily exclusive (Ch ami et al., 2008; Stark & Lucas 1988). Thus, increasingly, current research efforts have been geared towards models that allow for the co existence of both altruistic and self interest motives. Following along these lines, Chami et al. (2008) specifies a UE ( cE, a, UH) 3 1 Where; cE = action the recipient takes, and UH is the Chami et al., 2008: 24). In short, according to this utility function, a remi behavior (applying remittance funds as would be directed by the migrant) and the utility mo tivation to remit which has profound implications for empirical tests of remittan ce

PAGE 108

108 24). What is relevant to this dissertation research is umption is recognized, implying that remittances are not necessarily only altruistic but increasingly becoming a mode of wealth accumulation in origin countries; be it in small businesses, housing construction or some combinations of these. These encouragi ng developments in the remittances determination theoretical literature provide the premise for the conc eptual framework presented in this part of the dissertation research Figure 3 1 depicts a diagrammatic outline of the current state of remittances dete rmination theoretical literature and where the conceptual framework advocated in this chapter fits into the overall theoretical framework. The striped box denotes that remittances impact, as treated in the mainstream remittances literature is not the main focus in C hapter 3 and in this dissertation in general. Thus, the conceptual framework presented here is based on the determination theoretical literature for the following reasons. Firstly, as noted above the remittance impact aspects have been discussed extensively elsewhere. Secondly, the research on remittances impact does not directly link remittances to land demand and conseque nt spatial expansion in cities. Finally, it should be noted that approaching the conceptualization of remittance induced urba n easy to understand the link between r emittances and demand for land in that acquisition of real assets such as housing construction directly impact land demand. Even if business establishment is the motive, it is not too difficult to see that this will employ demand for housing and land.

PAGE 109

109 The Asset Accumulation Motive o f Remitters S ome a uthors have argued that the remittance determination research has conditions in the home c 2004: 64). Thus, Blue (2004), has argued that Building on the behavioral theoretical models advanced in Lucas and Star k (1985), Blue (2004) incorporates structural characteristics thereby shifting the focus of analysis onto the mitigating economic factors in the origin country of migrants. This paves the way for the inclusion of diverse factors such as state policies, ge nder, e conomic crisis, family ties, among other things (Blue, 2004). Along these lines Chami et al. (2008) identified nominal exchange rate, income gap (proxied by GDP per capita differential between remittance source and destination countries) and investm ent opportunities (indicated by interest rate differential between remittance source and destination countries) as important determinants/drivers of remittances flow (Not the main focus of this work, though). More important for this dissertation research, however, are the other works that have emphasized the notion of insurance under the self inter est motive (Amuedo Dorantes & Pozo 2006; Lucas & Stark, 1985). Amuedo Dorantes and Pozo (2006) describe two forms of insurance, namely, family provided insuran ce and self insurance (Figure 3 1); arguing that remittances sent from USA to Mexico showed that remitters were partly motivated by asset accumulation as a f orm of self insurance (Figure 3 1) whereby remitted moneys are used to purchase physical assets suc h as houses and plots of land. These authors attribute the self insurance motive to risks in host countries of remitters whereby high

PAGE 110

110 economic risks in the host country motivated increased remittances to the home country to purchase physi cal assets (Amued o Dorantes & Pozo, 2006). Whereas this attribution holds true to a large extent, it is also conceivable to reason that migrants do not necessarily have to experience or anticipate risks in their host countries in order to start acquiring real assets in the ir origin countries. For example, migrants who intend to return to their home country would likely want to remit for wealth accumulation for their own welfare. Consequently, in conceptualizing the relationship between remittances and urban land conversion in migrant origin countries, this study does not assume that migrants send money to th eir origin countries due to risks in their host countries per se Conceptualizing Remittances Urban Spatial Expansion R elationship to fund business establishments or operations which could in turn fuel land demand. Additionally, a portion of remittances is used for building houses in origin countries of migrants (Amuedo Dorantes & Pozo, 2006). In this lat ter sense it is plausible to reason that remittances will generate rounds of economic activities through direct employment. For example, t hose previously un employed or changed jobs for higher wages in construction related jobs become able to effectively d emand housing and land Combined, these activities will exert pressure on land demand resulting in conversion of new lands if the increasing demand for housing and land are not met through some form of densification (infilling, use of existing homes and s tructures and / or gentrification in general). The specific aspects of the remittances urban spatial expansion linkages are described below. Figure 3 2 outlines the pathways through which remittances potentially impact the urban spatial / land expansion process. A ccording to the wealth or asset accumulation

PAGE 111

111 motive of remitters, part of remittances would be used to acquire real property in the origin country, includ ing housing (Amuedo Dorantes & Pozo, 2006). Conceptually, t his results in Land Demand 1 in Figure 3 2. Thus, Land Demand 1 is essentially for construction of homes which would in turn create jobs in the construction sector of the urban economy. It is also conceivable that part of the remittances would be invested i n some businesses (Chami et al., 2008), adding jobs to the urban economy. Previously unemployed city residents and induced migrants will find work related to remittances, other things being equal. Having gained employment from remittances related economic activities, these people become purchasing power enabled, and are able to ef fectively demand housing and land. Part of the housing demand could be met through densification by way of gentrification into high rise buildings, and use of existing homes. However, there is likely to be surplus demand which would fuel into Land Demand 2 Part of Land Demand 2 may be fulfilled by infilling but likely surplus demand would require construction of new homes on new lands, which in the absence of effective land use management, would often result in conversion of natural areas into urban uses usually in the peri urban areas. Part of Land Demand 2 is also generated by remittances related businesses that require new lands. Back to Land Demand 1 part of the demand will be satisfied by infilli ng but more often than not there will be surplus demand for land for housing construction that will require new lands, which usually occur on /natural areas in the peri urban areas of the primate cities in developing countries. It must be noted that there is a distinction between la nd demand for housing construction that fuel land demand through job creation which in turn enable people to It should be n ote d that some of

PAGE 112

112 these previously unemployed people m ay be living with friends and family members prior to becoming gainfully employed). Also, i t is assumed that migrants would want to town, as pertains to some degree, in Accra. One reason for this is the possibility of Another reason may be the intention to eventually return home and entertain the hopes of finding employment in the is the hub of economic activities in the country). Also, it could be a source of pride for one to own a home in the capital city of the country. Finally, it should be noted that the bro ken lines and boxes in Figure 3 xpansion. Or more specifically, they are processes that if in operation, would work to mitigate the rate of new land conversion or the rate of loss of natural areas in the peri urban area to urban uses. To sum up, the main pathways through which remitta nces would impact urban spatial expansion are: (1) via direct demand for land for housing construction in the large/primate city; (2) via jobs boom in the building and construction sector. A related activity to migrant direct demand for land for house cons truction is the likely boost to the building and construction which offers employment to previously unemployed city investment in small businesses. Part of remittances may be used to fund small scale businesses which offer employment to people who can now express effective demand for housing and land. The combination of the 3 pathways would potential ly fuel the demand for housing and land.

PAGE 113

113 Summary The advent of economic glob alization has brought in its wake increased financial flows across geographic space, including remittances. Remittances have been particularly touted as one stable source of fund flow, usually not affected severely by economic downturns. However, its pote ntial impact on urban biophysical environment through the fuelling of lan d demand is often masked. C hapter 3 has attempted to unravel the relationship between remittances and contemporary urban expansion, most probably in primate cities in developing count ries. The chapter has presented a conceptual framework which clarifies the pathways through which remittances could potentially impact the urban land expansion process, if such funds are unevenly distr ibuted across urban areas in a given country. The main demand for land for house construction; (ii) demand for housing by people employed in construction and related jobs relating to remittances ; and (iii) demand for housing by people employed in small businesses established wi th r emittance funds. Thus, it seem s that there are some linkages between remittance flows on the one hand, and urban land expansion process on the other, especially as pertains in many primate cities in developing countries. The increased demand for housing a nd land must be met through densification gentrification into tall/apartment styled buildings, or use of existing vacant lands and buildings. Otherwise, new lands must be converted from agricultural and natural areas, usually in the peri urban areas, give n the general absence of effective land use planning. So far based on the literature on Accra in Ghana, it appears that the urban planning process has failed to take advantage of gentrification, infilling and activities that may lead to densification. C onsequently, it can be argued that increased remittance inflows to Ghana and considerable concentration in Accra may be

PAGE 114

114 contributing to the rapid physical exp ansion of the city. As a result holding other things contemporary land expansion in these cities of the developing world. The m ain contribution of Chapter 3 therefore, is the provision of the a systematic conceptual framework which has sought to unravel the links betw een remittances and contemporary urban spatial expansion, alluding to the notion of remittances induced urban spatial expansion, in both direct and indirect ways. Going forward, Chapter 3 ( this chapter ) together with chapter 2 have laid the conceptual fo undations for theorizing externally induced urban expansion, premised on the classical urban land use theory, for mally espoused by Alonso (1964). Thus, h aving unraveled the pathways and systematically conceptualized the relationships between FDI and remitt ances on the one hand and urban land expansion on the other hand the stage is set for the formal articulation of these emerging contemporary urban land conversion drivers into a relevant economic theory. In chapter 4 therefore, both FDI and remittances will be nd use, land value (Alonso, 1964) ; drawing extensively on current extensions which have derived the urban extent as a function of socioeconomic factors operating in the city (Angel et al. 2011 ; 2005; McGrath, 2005; Brueckner & Fansler 1983)

PAGE 115

115 Figure 3 1 Outline of mainstream remittance determination theories construct. Remittances Altruism Self interest Inheritance Insurance Loan repayment Self: Asset accumulation Family provided Impact Investments: small businesses Real assets: Housing, land Determin ation

PAGE 116

116 Figure 3 2 Pat hways linking remittances to urban spatial expansion construct. Employment: Previously u nemployed city residents and Induced population Construction Jobs Urban s patial/land expansion Demand for housing Land Demand 1 : Housing construction New lands /natural areas conversion Land Demand 2 : Miscellaneous De nsification: Infill gentrification, use of existing homes & other structures Other jobs Remittances

PAGE 117

117 CHAPTER 4 GLOBALIZATION AND AL ON OVERLOOKED FACTORS? Background The main objective in C hapter 4 is to attempt to extend the Alonso (1964) theoretical framewor k to facilitate the geographical analysis of contemporary urban spatial expansion as driven by FDI and remittances flows into u rban spaces. The economic theory of urban spatial structure, formally formulated by Alonso (1964) has not been sufficiently refi ned to incorporate contemporary external economic forces. Thus, influences of the forces of economic globalization have not been systematically examined under this generally accepted standard economic theory of urban land use and land val ue It is imperat ive therefore, that this important economic theory of urban land use be refined to reflect current realities of economic globalization, to make it particularly suitable for explaining contemporary urban spatial expansion in developing country settings, w here almost invariably such external funds become attracted by existing urban agglomerations in large or primate cities The twin forces of economic globalization and liberalization in developing countries have helped made their urban economies more open to global economic influences, exposing the urban biophysical environment to the vagaries of global economic forces as well. Thus, it is not too difficult to see that urban spatial change could be linked to forces external to the urban area and even outsi de the country (Chapter 2; Chapter 3) Global urban built up areas are growing at a very rapid pace such that even in some cities, the growth in the built area have outstripped the growth i n population (Angel et al. 2011 ; 2005 ). For example, A ngel and c olleagues note that Accra, the capital of Ghana saw expansion in its built area from 13,000 hectares in 1985 to 33,000 hectares

PAGE 118

118 i n 2000, representing over 150 percent During the same period, According to the authors, the population of the city i ncreased from 1.8million to 2.7 mil lion representing about 50 percent increase (Angel et al. 2011) In contemporary urban spatial expansion analyses such a dramatic increase in the urban extent normally is attributed to population increase for cities in developi ng countries and to GDP /GNP per capita growth for developed countries (Seto et al., 2011). Thus, the traditional suspects have, almost invariably, been population increase and growth rate of GDP per capita ( Angel et al., 2011; 2005; Seto et al., 2011; McM illen & McDonald, 2010; McGrath, 2005 ; Brueckner & Fans ler, 1983; Muth, 1969; Alonso 1964 ). Even though consistent with the classical Alonso urban land use theory, the advent of economic globalization and attendant potent external economic forces, requir es that new questions are raised with regard to the extent to which these traditional are able to give adequate explanation for contemporary urban land expansion. This is of particular importance to land expansion in primate cities of developing countries, where more often than not many primate city residents may not have the requisite purchasing power to effectively demand housing and/or land. For instance, in Accra many researchers agree that respectable, quality housing being put in the pe ri urban areas is usually above the means of the ordinary Ghanaian living and earning incomes in the local economy ( Moller Jensen et al., 2005; Grant & Yankson, 2 003; Yeboah 2003 ; Konadu Agyemang, 2001 ). Indeed, overall economic growth in some of the de veloping countries has lagged rapid urbanization and urban spatial expansion. It seems, therefore, that the rapid physical expansion of primate cities like Accra, characterized by 2003), might have some connections w ith external

PAGE 119

119 economic stimuli associated with intensification of economic globalization, fostered and enhanced by economic liberalization policies pursued in the 1980s (Grant & Yankson 2003). Thus, it would be interesting to know whether the explanations based on these variables are adequate given the current era of economic globalization whereby location and land use decisions could be driven by underlying global economic forces beyond the borders of a particular city or country. For example, demand fo r urban land and housing could be driven by increase s in FDI and to some degree, remittances inflow into a country as migrants seek to own real assets in their home countries. Therefore, i t is not gainsaying the fact that primate cities in developing coun tries are being bombarded by external economic forces of globalization as increasing spatial interconnect edness and close integration of economies have generated almost seamless flow of economic resources such as FDI and remittances across countries. Al most invariably for developing countries, these funds are often concentrated in and amenities are concentrated (Grant & Yankson 2003). On average, it is estimated tha t between 75 85 percent of the FDI funds concentrate in the Greater Accra Region; with metropolitan Accra being the main center of eco nomic activities (GIPC, 2001 2011). The disproportionate concentration of such funds in such a geographically restricted manner help fuel demand for land and housing, directly and indirectly; as firms locate in the urban space and can also induce people to move into the city as jobs are made available through FDI and remittances (Chapter 2; Chapter 4) Similar findings have been reported for other cities in developing countries (Seto 2005). It appears, therefore, that some kind of externally induced expansion of large or a primate

PAGE 120

120 city in developing countries is emerging with the advent of economic globalization. Some att ributions have been mad e in this regard (Grant, 2009 ; 2001; Leichenko & 2008; 2000; Yeboah, 2003 ). For example, Richard Grant has noted that Accra has become an international city of some sort whose growth has been significantly impacted by globa l forces (Grant, 2009). To the dismay of Grant, however, it appears, most typical African cities, including Accra, are not conceptualized in the global context, Accra i s ver ty African city conceptualized primarily in local and regional terms in the earlier nationalist era. This more internationally oriented city represents an incomprehensible city to government policy makers and others who fai l to grasp the extent of the transformation, and who instead still situate Accra in national, nati onal regional, and local terms. (Grant, 2009: 3). In essence, Accra has not been marginalized by globalization as other researchers have argued regarding A fri can cities in general (Merwe, 2004). Thus, even though researchers like Saskia Sassen (2006), most researchers would agree that they have not escaped the forces of ec onom ic globalization ( Killick 2010 ; Grant 2009; Grant & Yankson 2003; Yeboah, 2003). As a result externally induced urban growth may be emerging as a new phenomenon in the primate city of Accra in Ghana and might well be underway elsewhere in cities in other developing countries exhibiting similar characteristics. However, as noted by Grant (2009), this has note d been appreciated generally. All these developments support the notion that the classical land use theory cannot adequately account for curr ent realities of economic globalization which impact urban land expansion. Questions have been raised about the suitability of the classical

PAGE 121

121 land use model before (Batty 2005) but usually these questions have bordered on the so called unrealistic assump tions of rational economic behavior and the static nature of the model (Aniya Masamu, 2010; Batty, 2005) The contention in Chapter 4 however, is in respect of the external economic influences which appear to be influencing contemporary urban land use an d land expansion, particularly, in primate cities of developing countries. In particular this study does not argue for the abandonment of the classical urban land use theory / model but seek to refine it to reflect current realities of economic globaliza turn impact the urban extent. As noted from above dissatisfaction with the classical economic theory of urba n spatial structure (Alonso, 1964), however, is not new. In particular, the static nature and monocentric city assumption s do not sit well with many analysts, with some advocating the use of complexity theory as the most representative of the city growth dynamics (Aniya Masamu, 201 0; Batty, 2005 for example ). Laudable as these approach es have been in regard to helping increase our understanding of urban expansion modeling, explicit representation of space and time, they do not seem to represent external economic drive rs of urban change adequately. Admittedly, there is a myriad of facto rs that affect urban growth, it is generally agreed that economic forces dominate (Pacione, 2009). Moreover, the complexity approaches when us ed to simulate urban expansion do not explicitly show how individual explanatory variables contr ibute to the expa nsion process. Furthermore, the transition rules seem to be based on some arbitrary decisions by the modeler, not necessarily based on any economic theoretical propositions as infor med by socioeconomic processes. The implications of all these

PAGE 122

122 would seem to suggest that completely abandoning the standard economic theory of urban spatial structure may not be the way to go with regard to enhancing our understanding of contemporary urban expansion dynam ics. It is in this regard that this dissertation in gene ral and C hapter 4 in particular, advances the argument that to some reformulated to incorporate the influences of external economic forces on urban spaces, especially in develo propositions can be tested by estimating reg ression equations (Angel et al., 2010; 2005 ). Coefficients of the estimated regressions can be input in a simple exponential model which can then b e used to estimate the built up area of the city over time. Combined, these should be able to reasonably account for contemporary urban spatial expansion, especially in cities that receive disproportionately large amounts of the external funds inflow as pe rtains in primate cities of developing countries Such an approach will be theoretically informed, dynamic, explicit representation of contributions of individual explanatory variables as well spatially explicit. This allometric based integrated urban sp atial expansion model is formally articulated in Chapter 5 of this dissertation. For now, Chapter 4 into the classical economic theory of urban spatial structure, essentially building on curr ent extensions of the model that derive the urban extent as a function of socioeconomic forces operating in the city (Angel et al., 2011 ; 2005; McGrath 2005 ; Brueckner & Fansler, 1983). In this regard the main questions motivat ing the analysis in Chapte r 4 are: (1) what is the nature of the correlation, if any, between FDI and

PAGE 123

123 remittances on the one hand, and urban built up areas on the other hand ? And (2) i n what ways can the classical Alonso urban land use theory be refined to accommodate remi ttance s and FDI as urban land use value change drivers? In seeking answers to the questions above, this study shares the view that mo st economic activities ( be it globally or locally generated) can hardly be dissociated from the land or the biophysical system ( Wackernagel, 2006; Sanderson et al., 2002; Constanza et al., 1997 ), even though Alonso might not have considered his work from this broader perspective. Economic Theory and Urban Space: The Thunen Alonso Theoretical Framework Urban spatial change is essen tially the physical manifestation of the aggregation of land use decisions made by individuals and authorities over time, as individuals and societies attach different values to lands depending on expected economic benefits. The classical explanation to sp atial patterns and change, from the perspective of urban geographical research, has relied mainly on two important land use / land value theories /models which have been widely recognized among researchers (Cadwallader 1996; 1985 ) the agricultural land u se theory, the classical one being Thunen's agricultu ral land use theory (Von Thunen 1926; 1826) and urban land use theory with Alonso's (1964) bid rent theory being the classical example. It is generally agreed that these two theories have informed much of the urban land use theorizing and modeling research efforts to the present and according to Cadwallader (1985; 1996) provide useful theoretical foundations for understanding urban land use and change dynamics. The key underlying assumption of rational economic behavior in which individuals are believed to maximize profits from their use of the land (von Thunen, 1926; 1826) or maximization of utility ( Alonso, 1964; Muth, 1969) underlies these theories.

PAGE 124

124 heory Johann H useful starting point for theorizing land use in general and urban land use in particular (Cadwallader 1985: 24 is that tr ansportation costs and accessibility exert overbearing influence on what type of use a piece of land will be put to in order to obtain the maximum economic or location rent ; hence putti ng the land to its best use. This theoretical framework resulted in c oncentric zone arrangement of agricultural land use around the market, centrally located in the agricultural plane Basically, The von Thunen model as it was first introduced and used in geography was largely descriptive. It was not until the quantitative revolution that researchers with the requisite mathematical and model building background were able to examine the full range of geographic implications. This descriptive model was, in 1954, operationalized in the form of a mathematical model by E.S. Dun n, Jr. (1954). Dunn concurred with von He formalized the von Thunen model in a functional relationship expressed as: R = E (p a) Efk, Where R = rent per unit of land; E = yield per unit of l and; p = market price per unit of commodity; a = production cost per unit of commodity; f = transport rate per unit of distance for each commodity; and k = distance from market. (Dunn, 19 54 cited in Mabogunje, 1997 ) Also, i t has been noted that del, however, summarizes a one product linear model in which a given lands use is capab le of producing a net return up to that point where its marginal revenue and marginal costs are equal He postulated receipts will be maximized at a site as close to the market as pos sible and will be minimized at that distance wh ere marginal returns equal zero (Mabogunje 1997: 52). The most important elements of this theorizing are the concepts of economic /location rent accessibility and maximization of profits from the use of the land (Cadwallader 1996; 1985 ), essentially implying that economic or location rent is a measure of the value of land in its best use.

PAGE 125

125 One essential outcome of Von Thunen's land use theory, in spatial terms, is that land values declined wi th d istance from a central location; in this case the isolated market located centrally in the agricultural landscape (Figure 4 1), thereby representing a first serious pioneering effort aimed at theori zing land use and land values from a sp atial perspect ive (Cadwallader, 1996; 1985 ). xtensions The foregoing outline of the von Thunen theory/model, to a large extent, lays the foundation for analyzing distance land value relationship in general. This motivated William Alonso (1964) to extend and formalize this theoretical framework in the urban context. Essentially the complexities of urban landscapes do not make the explanation of their associated land use dynamics easily amenable to the simplified Thunen the oretical framework, impl ying theory of urban spatial structure ( Watkins, 2011 ; Reid, 2002; Mabogunje, 1997 ). It is generally agreed that th e standard urban spatial model is the mono c entric city model (McMillen & McDonald 2011: 79; Alonso, 1964) Basically, Alonso (1964) theorized the existence of a center of economic activity in a city known as the Central Business District (CBD), equally accessible from all parts of the city. Three main urban uses are recognized commercial, industrial and residential. He postulated a bidding process in the urban land market whereby the use to w hich a piece of land is put would be determined by its highe st and best use the land use or activity deemed to fetch highest economic return or value will command the highest price (McMillen & McDonald, 2010; Cadwallader 1985; 1996 ; Alonso, 1964). Alonso further postulated that individuals seeking to use the la nd will behave rationally so as to maximize their

PAGE 126

126 overall utility or satisfaction and in this case to minimize transportation costs. Again, accessibility becomes a key determinant of land use arrangement in the Alonso theoretica l framework. Consequently, concentric zones arrangement of land use patterns can be envisaged (Figure 4 2). Land use in more accessible areas such as close to the central business district (CBD) receives the highest bid for its use. Consequently, comme rcial land use (especially retail and higher rents), f ollowed by industrial land use; with the outlying areas being occupied by residential land use with higher density fir st and lower density bordering the agricultural land/ natural areas. tes and rightly so, the consequence, as per Alonso (1964), is the creation of Bid Rent C urves (Figure 4 2). It is also important to note that Von Thunen's theo ry in a way alluded to the bid rent curves concept in which land users offered or bid according to the economic value they would expect from a particular land use given the price of the crop to be cultivated and distance from the market (Mabogunje, 1997; Cadwallader 1996; 1985 ). Figure 4 2 depicts the bid rent curves and their associated spatial manifestation concentric zones derived from eoretical postulates (Reid, 2002 ; Mabogunje, 1997 ). This simplified representation shows the close li agricultural land use theory/model the distance/accessibility land value/land use relationships ; thereby serving as a critical link between the economic theory of urban spatial structure and geographic al analysis of urban land use. In this regard relies strongly on the concepts of accessibility and transportation costs to analyze the

PAGE 127

127 cost minimization behavior of individuals as they seek optimal locations for different urban economic activities. More succinctly, according to Mabogunje (1997): The bid price function answers the question: As the individual considers residential locations at different locations in th e city, i.e. at increasing distances from the city center, what price of land would allow her to buy sufficient amount of land (and other goods) to enjoy as much satisfaction as a given (starting) price (and amount of land) at the city center (Mabogunje, 1 997: 50 52). Based on current extensions with particular reference to work of Shlomo Angel and colleagues (Angel et al., 2011 ; 2005) formulation is outlined as follows: House households denoted by (L) occu py land in concentric zones which surround the CBD. Households, who have similar preferences, have at least one member who travels a distance of ( x ) to the CBD to work in order to earn (y) income annually, paying (t) as annual cost per unit of tr avel to work with the annual commuting cost totaling t* x It is further postulated that households budget their income such that quantity of housing (q), total commuting cost (t* x ) and composite good (c), (assumed to be the same through out the city), exhaust the incomes. Further, h ousing price (p) would vary with distance from the CBD. both housing and composite good ( c ) can be represented by a utility function v(c, q) which is quasi concave (Angel et al., 2 011; 2005 ). The equilibrium condition of city residents requires that all households are settled and that a household at any location attains a common utility level (u), inside the city. If households are assumed to be rationale and will operate on the h ighest indifference curve, choosing a combination of affordable housing (given income constraint) and the composite good that will maximize their utility, the equilibrium can be represented as in (Angel et al., 2011; 2005 ):

PAGE 128

128 4 1 Partia l differentiation of Equation 4 1 : change in price and quantity with respect to distance from the CBD can be d erived as in the Inequalities 4 2 4.2 The Inequalities 4 2 posit that price of land ( or cost of h ousing ) declines as one moved outward from the CBD but quantity of housing increased as the price of land declines with distance from the CBD. On the housi ng supply side, a gain based on Angel et al. (2011 ; 2005 ) producers combine capital (N) and land (l), operating with the production function H(N, l) which is concave (i.e. housi ng production obeys the law of diminishing marginal returns of capital and land) and exhibits constant returns to scale. If producers compete in a perfect market, with free entry and exit, an equilibrium land rent function r( x ) and capital land ratio S( x ) floor area ratio, or building density can be derived. Th is equilibrium is dependent on d istance from the CBD Taking partial derivatives of land r ent and the capital land ratio with respect to distance from the CBD will yield the Inequalities 4 3 which imply that both land rent and capital land ratio (otherw ise denoted by building density) decrease as distance increased from the CBD (Angel et al., 2011 ; 2005). 4 3

PAGE 129

129 ( x ) from the CBD is denoted by D( x ) and it is assumed that a household has only one member, population density declines as one moved from the CBD, away from the city center since houses would become larger ( perhaps due to changing taste?) and building density also decrease as shown in Inequality 4 4 as per Angel et al. (2011 : 42 46). 4 4 for the use of peri u rban land. Once h ousing producers outbid farmers, the peri urban land will be converted from agricultural use to urban uses, most probably to residential use. (It must be noted that land on the peri urban areas may not be necessar ily in agricultural use per se; it could b e natural area or lying fallow) A major question is: w hat happens in the absence of viable agricult ural uses for peri urban lands? It is not too difficult to envision an unbridled city expansion, driven mainly by demand for land for u rban uses, such as Accra seems to be experiencing in recent times. Representing the distance to the outer boundary of the city (radius) by and denoting agricultural rent in the peri urban area by (r a ) ; and given that 4 5 it can be deduced that r(x) > r a ; undary and that r( x ) < r a outside the city boundary, yielding the equilibrium condition:

PAGE 130

130 r( ,y,t,u) = r a 4 6 Thus, a circle with r adius (urban population, denoted by (L), should be accomm odated (or housed or settled) at equilibrium. building in a r ing x distance away from the CBD, the equilibrium can be derived as (Angel et al., 2011) : 4 7 Thus, according to Shlomo Angel and colleagues: The classical theory thus provides an endogenous solution f or the extent of the area that a city occupies, given its population L, the income of the agricultural rent on the urban periph ery r a ( Angel et al., 2011: 40). Resolv ing the equilibrium Equation 4 7 results in several testable propositions as presented in Angel et al. (2011 ): 4 8

PAGE 131

131 To sum up, according to the authors, increasing levels of agricultural r ent, transport cost and share of buildable area will tend to decrease the radius of the city (reduce size of residents will lead to increasing radi us (expanding built up area). Consequently Angel and colleagues assert that More generally, it follows that the total area of the city A, will decrease if the and will increase if the city population (L) increases and the income y of that population increases. ( Angel et al., 2011: 41). T his is a profound conclusion which is one of the underpinning pillars of this dissertation in that we can now formally deriv e the urban extent from the economic theory of urban spatial structure ( Angel et al. 2011; 2005; McGrath, 2005 ; Brueckner & Fansler, 1983 ); and that one can analyze the dynamics of the urban extent as a function of economic factors operating in the city Essentially, it implies the analysis in this dissertation can focus on the shifting boundary (extending radius) of the city and examine the responsiveness of the urban extent (built up area of the city) to changes in underlying economic forces; specifical ly for this dissertation research, as induced by external economic forces of globalization in the form of FDI and remittances flows lobalization: Initial Propositions al structure which have been confirmed population and income as explanatory variables ( Angel et al. 2011; Seto et al. 2011 ; McMillen & McDonald 2010 ; McGrath, 2005; Brueckner & Fansler 1983 ). In a meta analysis of over 300 studies, Karen Seto and colleagues concluded that GDP per capita did influence urban land expansion in developed or advanced countries but for developing countries, population

PAGE 132

132 increase wa s found to be the main driver of urban land expansion in contemp orary times (Seto et al., 2011). In a nutshell, conventional urban spatial/land expansion studies continue to arrest cts) for urban land expansion. However, the advent of economic globalization appears to hav e made it imperative that a closer look large/primate cities of developing countries. Thus, in the era of increas ed interdependence of economies, unrestricted flow of funds and recognition of property rights by individuals, several issues need to be examined in respect of co ntemporary urban land expansion; especially with respect to primate cities in developing count ries. So how does the city accommodate external economic shocks in the form of FDI and remittances inflows, for example? This dissertation research introduces the concept of Remittances purchasing which is described in this work simply as the proportion of people resident in a city who were previously unemployed but find jobs resulting from FDI and remittances related economic activities as well as those induced to move into the city by the existe nce of F DI and remittances related jobs The plausibility of FDI Remittance purchasing power enabled population is articulated as follows: FDI has a tendency to flow into existing agglomerations, especially so with manufacturing and services FD I (Chapter 2 ). Implications of this tendency are: drawing of population into the economic enclave of jobs; previously unemployed city residents eventually may cause demand for land to increase (Chapter 2); direct demand for land

PAGE 133

133 for FDI firm operations and expatriate housing demand. Similarly, remittances inflow into a country may be unequally distributed from a geographical perspective, usually in loping country setting. If a substantial portion of such remittances is over concentrated in a single large/primate city, the real estate (building and construction) sector of the city receives a boost which would lead to increased demand for land (Chapter 3) Furthermore, jobs created in the building and construction sector as a result of the remi ttance induced demand for land employ people who then densification, will result in demand for new lands, as has been exp lained in Chapter 3 of this dissertation research Consequently, the combined effects of FDI and remittances may reflect in direct and indirect demand for land for urban uses, via population and / or income. As detai led in Chapters 2 and Chapter 3, the unequal geographical distribution of external funds inflow in favor of the primate city in a country would potentially impact population ( through induced migration ) and incomes of city residents. It does appear theref ore that inflows of external funds in the form of FDI and remittances could induce expansion in the built up area of the city. However, neither FDI nor remittances have been formally considered as important explanatory variables in the analyses of contem porary urban expansion, with the exception of Se to (2005), for the case of FDI Chapter 4 loss of generality and importance. Alonso theoretical framework will be more relevant and potentially applicable to cities in

PAGE 134

13 4 disproporti onately large amounts of external funds in the form of FDI and remittances. On I ncome Let y = income of city residents who are currently employed (before the advent of FDI and remittances inflow this is what is usually denoted by GDP per capita in most analyses). Previously unemployed city residents may find jobs in the advent of FDI and remittances. mand for housing and / or land. Let y F = previously u nemployed city residents and induced migrants who find work in FDI related jobs and are earning incomes. Let y R = previously unemployed city residents and induced population who find work in remittance related jobs and are earning incomes. (Both y F and y R include part of expatriate incomes spent in the city) Thus, total income of city dwellers receiving FDI and remittances, denoted by (y T ) can be expressed as: y T = y + y F + y R 4 9 On P opulation Let L = population of the city before FDI and remittances (as defined in Equation 4 7 ). FDI and remittances inflows will be associated with movement of population into the city who are in search of jobs (note that this is not the typical rural urban migrati on, who may not be skilled ) as well as expatriates (C hapter 2; Chapter 4) Let this L F and L R for FDI induced population and Remittances induced population, respectively. Thus, in the wake of increasing flows of

PAGE 135

135 FDI and remittances, the total population of the city denot ed by ( L T ) changes and can be expressed as: L T = L + L F + L R 4 10 Thus, total city population consists of residents ( L ) before FDI and remittances and induced population (migrants) L F and L R respectively; following FDI and remittances inflows into the c ountry and city. Holding prices constant for now and assuming that the bulk of demand for housing will be high end, quality housing (by virtue of the possibility that these people who find jobs related to FDI and remittances are relatively higher skilled workers than the average rural urban migrant); which are usually new houses found in the peri urban areas of the city (barring any appreciable densification and use of ex isting vacant houses and land). Logically, t he city expands to accommodate the new po and consequently increased demand for land, resulting in new land conversion (assuming little or no densification). It is assumed for the purposes of the arguments in this dis sertation, that buildable land is available if there is effective demand (land market in most of the developing countries in Africa is complex; overseen by chiefs, elders and families with some market reforms underway but these are unlike the situation in more advanced economies) Also, agriculture is not considered a viable economic activity at the periphery of the city since land in urban use commands high economic value (implying that land owners holding land in agricultural use in the pe ri urban areas uses, including housing development, have higher present values on land near

PAGE 136

136 1997: 128 129). In short it is assumed that urban developers who find it profitable to build houses will find land within their budget. Consequently, demand for housing in the wake of increased FDI and remittances inflows into the urban space economy will mos t likely be translated into demand for new lands in the peri urban areas of the city. Consequently, t he boundary of the city shifts outwards (radius lengthens). This implies that the equilibrium condition: [ r( ,y,t,u) = r a ]; m ay no l onger be ten able and necessary; and may change to the inequality below, at least in the short to medium term: r( ,y, y F ,y R ,t,u F,R ,u) > r a 4.11 Where; u F,R = the utility of newly employed city residents, induced m igrants as well as expatriates. Definitions for the rest of the variable s remain the same as previously Note that so far no fundamental change has bee n effected to the Alonso theory: only that FDI and remittanc es have been inserted into the original equation a s espoused in Angel et al. (2011 ). who were gainfully employed before inflow of external funds (FDI and remittances), it is now p ossible to include people who are employed in FDI and remittances related jobs in the advent of these external funds. These FDI remittances related purchasing power enabled city residents will add to the original demand for housing thereby increasing tot al demand for housing and land, as they seek to maximize their utility (u F,R ) Since urban rent is likely to increase, housing producers will find incentive enough to build more houses. The city may expand until a new zone of low agricultural rent is rea ched

PAGE 137

137 (note that expansion at the periphery will cause urban rent to reduce). Eventually, equilibrium will be re established, spatially, far into the surrounding natural areas. This equilibrium is demand driven and is subject to change as soon as there is of urban land demand, especially as induced by FDI and remittances inflows. Consequently, t he radius economic globalization drives equilibra of land rents in th e peri urban areas of the city. Density reduces as in the origin al classical model. As a result, the city population urban land demand equilibrium (Equa tion 4 7) can be modified as in Equation 4 12. 4 12 It is important to emphasize that the t he city receiving large amounts of FDI and remittances should accommodate th L T ) within the city limits: Fortunately or unfortunately land supply in the city is not constrained by availability of buildable land since it has been assumed that in a globalizing primate city located in a developing country land will be made available for deve lopment into urban uses as long as there is effective demand. Related Theoretical Propositions Resolving the equilibrium E quation 4.12 concentrating on impact of FDI and remittances will yield the Inequalit (propositions) (after Angel et al., 2011) : L T (= L + L F + L R ) y F ,y R ,u F R

PAGE 138

138 Proposition 1 : FDI lenthens the radius of the city / y F > 0 4 13 Proposition 2 : adius / y R > 0 4 14 Proposition 3 : FDI in adius / L F > 0 4 15 Proposition 4 : Remittances induce population; lengthens the / L R > 0 4 16 Assumptions : (a) Substantial portion of both FDI and remittances must be through infilling, gentrification, use of ex isting structures and homes must be minimal such that the bulk of FDI and remittances induced housing demand would by supplied through construction of new homes on new lands at the peri urban interface; (c) Agricultural land uses at the peri urban interfac e are not economically feasible (low agricultural rent), enough to compete with bid s of urban housing producers; (d) Lengthening radius of the city is equivalent to increasing increasing urban extent; (e) S upply of developable land is not constrained. In essence, it is demand driven in that when demand for housing is generated through FDI and remittances inflow, housing producers will find it economically viable to build more houses. (Note that the

PAGE 139

139 developable land can be bought cheaply at the peri urban i nterface because potential agricultural land use market is virtually non exi s tent). The situation of agricultural land market at the peri urban interface in many primate cities of developing countries, including Accra in Ghana is akin to what Dean Han ink (1997) has put it Any one holding land in agriculture (or any other use) that could earn higher returns if used in a different way is incurring opportunity costs. They are effectively losing money even if their farming operations are prof itable (p.129). Left to market forces then, urban land expansion continues unabated as housing demand is constantly fuelled by induced process es via FDI and remittances. This is the cities in developing countries. For example, in Accra, like all other lands in Ghana, lands are held by chiefs elders and family heads in trust and are not deemed to be sold. However, if any chief considers the sale of lands as contributing to the development of their areas, arra ngements are made for the transfer of the lands to private developers. The point here is that more often than not individuals are not key players and in many cases the livelihoods of the current users of the land have been compromised compete with urban land uses in the peripheral areas of cities such as Accra in Ghana. Consequently, FDI inflow can influence demand for land in a more direct manner. It would n ot be too far could drive urban land expansion, should the remittance funds be geographically restricted in distribution in favor of the primate city. Given the foregoing it is not too d ifficult to see the imprints of economic globalization in the urban built landscapes of

PAGE 140

140 funds (GIPC various years ). Such funds inflows, if not geographically diver sified, will increase the pressure on demand for housin g and land in the urban area. It is argued o have been relatively successful at attracting FDI and increasing levels of remittances but have failed to geograhically diversify the inflows into other cities/towns. From the discussion so far, it can be said that e not explicitly derive the urban extent, recent formulations have been able to derive the urban extent as a function o f the classical determinants population growth, income growth, distance from the CBD, urban land rent, agricultural land rent, and c s utility (Angel et al., 2011; 2005; McMillen & McDonald 2010; McGrath 2005 ) ; with population and income being consistently identified as most significant. But these extensions have still largely neglected the external economic dimensions, especially, e conomic drivers of globalization (including neoliberal economic reforms) and associated increasing international remittances which is now believed to be influencing contemporary urban spatial expansion in cities such as Accra. This underscores the importa nce of the main motivation in Chapter 4 of this dissertation research which attempts to refine the Alonso theoretical framework to reflect current reali ties of economic globalization; which will likely impact positively on our understanding of how globali zation related economic forces impact the urban extent. Consequently, a nalyses in Chapter 2 and Chapter 3 as well as discusision so far in Chapter 4 lay the foundations for systematically interrogating the plausibiltiy of

PAGE 141

141 externally induced contemporary l and expansion. This paves the way for preliminary empirical testing of t he propositions derived in Inqualities 4 13, 4 14, 4 15 and 4 16 which could potentially add to our understanding of the emerging phenomenon of externally/globally induced urban land expansion as economic globalization relentlessly ; and could have practical implications for urban plann ers. Consequently, the next section presents an empirical example, attempting to put the zation related urban land expansion propositions to test, as Alonso theoretical framework. Towards Globally Induced Urban Spatial Expansion Analysis: An Empirical Example Even though the classical economic theory of urban land use / land value has been criticized from many quarters (Aniya Masamu 2010; Batty, 2005), this study shares in the believe that the Thunen Alonso land use theoretical framework still offers n land expansion analysis ( Angel et al., 2011 ; McMillen & Mc Donald, 2010; McGrath, 2005; Brueckner, 1987 ; Brueckner & Fansler, 1983 ) which can be improved upon to reflect current realities of economic globalization as has been attempted Furthermore, the outcome of the regression estimates can be further implemented in an urban land expansion estimation model / method particul ar cities, formally articulated in Chapter 5. Data Data for the empiri cal analysis of this study have been obtained from several reliable and authoritative sources. Built up area data were sourced from the Lincoln Institute for Land Policy Atlas of Urban Expansion (2010 ). Within this database is

PAGE 142

142 fo und a sample of 120 citi es with built up areas pre selected by Shlomo Angel and his associates; located at ( http://www.lincolninst.edu/subcenters/atlas urban expansion/urban national data tables.aspx ) This study took advantage of this and used this sample for our analysis. (Full documentation on the sample of 120 cities and related data as well as the underlying ratio nale for the sample selection is pr esented in Appendix A ). For this dissertation research 13 cities were removed from the original list of 120 due to unavailability of FDI and remittances data for some countries in the sample. Also, some countries experienced net negative flow of remittances and / or FDI, re ndering the data relating to such countries unusable for log transformation implemented in the regression analysis. Consequently, a total of 107 cities were included in the regression estimation presented in this chapter. Data on an nual FDI inflows (net) in US$ gross domestic product (GDP) per capita and remittances inflows (net) were gathered from United Nations databases (UN, 2011), available at: http://data.un.org ( Detailed documentation for the dataset has been p rovided in Appendix A ) Method Regression Models of urban structure and expansion have mostly implemented local factors as explanato ry variables (Angel et al., 2011; 2005 ; McMillen & McDo nald, 2010; McGrath, 2005 among others ). These have generally in cluded population and income changes, GDP per capita growth rates, geographic and topographic factors, transport costs, distance from the center of the city and in recent times, air links (Angel et al., 2005) ; and FDI (Seto, 2005). However, consistently, the most important urban land use change drivers have been identified as population and income, which have

PAGE 143

143 been found to be highly statistically significant in the regression models estimated. Most recent studies buttressing this a ssertion are (Angel et al ., 2011; 2005 ; Seto et al., 2011). Regression analysis has been found to be amenable to urban land expansion modeling as exemplified in st udies such as Angel et al. (2011 ; 2005); McMillen and McDonald (2010) and McGrath (2005), Braimoh and Onishi (2007); for example. According to Rogerson (2005), regression analysis refers to a more complete process of studying causal relationship between a dependent variable and a set of independent, explanatory variables. Relating to land use analysis, Millington et al (2007) have stated that Regression analysis is used to improve explanation of the mechanisms and processes of change (by examining the statistical significance of the influence of the predictor variable upon the dependent variable) and / or (ii) predicti on of change itself (derived relationships may be used to project future land use/cover from the current values of the independent variable) (Millington et al., 2007: 564). The choice of regression analysis in this dissertation research, therefore, is wel l founded and grounded in the literature. In a specific study including Accra in the sample of cities, Shlomo Angel and models were able to explain 93 to 95 percent of the variations in urban land cover amo ng coun regression in land expansion analysis. It is important to note that t his dissertation research does not divide the study area into discrete units as has been the approach in m ost land cover change modeling (Millington et al., 2007). Cross sectional multiple regression models have been constructed using urban built up area in individual cities in the sample of 107 cities. Thus, in all models the dependent (respondent) variable i s city built up area in 2000 or a

PAGE 144

144 close year, city population in 2000, GDP per capita in 1999, net foreign direct investments inflow into countries in 1999, and Workers remittances i nflow into countries in 1999. These 2000 1999 1999 1999 respectively. It must be noted that the built up area of the city in 2000 (Built 2000 ) remains the response variable in all equations (even though some cities may have their built up areas derived in some other year close to 2000) All the variables are log transformed using ordinary logarithm (Log 10 ). Log transformation is generally undertaken to attempt to achieve objectives such as ensur ing linearity as well as normalit y of the dataset (Chatterjee & Hadi, 2006: 151 174 ). Moreover, it could help remove problems of het eroscedasticity (Chatterjee & Hadi, 2006: 168). GDP per ca pita, FDI and remittances inflows have been lagged for one year, hence the use of 1999 figures as op posed to 2000. The rationale underlying variables are not likely to produce change in the response variable instantaneously. For example, it seems rational to expect that increased levels of FDI inflow in a given year may not necessarily effect change in built up area in the same year. However, the city population variable has not been lag ged due to data unavailability to cover all members of the sample of countries f or 1999. Once all the model variables have been log transformed, the Ordinary Least Squares (OLS) is applied to esti mate the equations formulated. All the assumptions of the classical OLS method, which may be found in standard statistics texts appl y (Chatterjee & Hadi, 2006; Rogerson, 2005; Burt & Barber,

PAGE 145

145 1996 ; Agresti & Finley, 1994 for example ). The regression models estimation were implemented in the SPSS software (2010). Urban built up area expansion regression models The basic models (structural equations) are presented below to aid comprehension. log Built 2000 1 ( log CityPOP 2000 2 ( log GDP 1999 4 17 Where ; 1 2 are estimable para the error term. All the variables hav e been defined above. Equation 4 17 attempts to capture the extent to e and spatial structure (Alonso, 1964); specifically their impact on changes in the urban extent (Angel et al., 2011 ). It is hypothesized that city population and GDP per capita are positively correlated with built up area of the city implying that increases in population and per capita GDP will correspond to increase in built up area of the city, all things which are not the main focus of this analysis but brought in to help build the case for our ariables FDI and remittances Thus, since the main focus of this study is to explore the external economic influences on the urban extent, we implement a second model. log Built 2000 1 ( log CityPOP 2000 2 ( log GDP 1999 3 ( log FDI 1999 ) + 4 ( log REMIT 1999 4 18 Where ; 3 4 are estimable parameters and all variables r emain as described above. Equation 4 18

PAGE 146

146 population and income growth, the city which is linked to the global economy is now receiving global impulses in the form of FDI and Remittances. Since these variables have the tendency to affect land demand in the city th r ough housing demand and demand for other built infrastructure, it seems rational to expect that they would im pact the urban extent. Specifically, in Equation 4 18 it is hypothesized that part of the increase in the extent of the built up area of the city in the current year may have been as a result in part, of increasing levels of FDI and remittances flows in to the city in the previous year. It is expected that increasing levels of FDI and remittances in a previous year would correspond to an increase in the extent of the built up area o f the city in the current year; and that co efficients would be stat istic ally significant at the 95 percent confidence level. In addition to the foregoing it seems rational to try to isolate the two gl obalization related economic variables, namely, FDI and remittances, in order to examine their potential effect when population and income are excluded from the model (s). This is approached in two steps. Firstly, the combined effect of FDI and remittance on the built up ar ea of the city is modeled as in Equation 4 19 : log Built 2000 1 ( log FDI 1999 2 ( log REMIT 1999 ) 4 19 All variables remain as defined above. Equation 4 19 states that built up area in 2000 is positively correlated with FDI and remittances inflow in 1999; Secondly, the potential impact of FDI and remittances separate ly, are modeled in Equation 4 20 and Equation 4 21 ), respectively. log Built 2000 1 ( log REMIT 1999

PAGE 147

147 4 20 Equation 4 20 posits that remittances inflow and built up area are positively correlated, increasing remittances inflow corresponds to expanding built up area. log Built 2000 1 ( log FDI 1999 4 21 According to Equation 4 21, built up area and FDI are positively correlated, implying that increasing levels of FDI should induce corresponding increase in the built u p area Regression Estimation Results The results of the estimation of the urban expansion regression models formulated in Equation 4 17 through to Equation 4 21 are presented below For Equation 4 17, t he Adjusted R 2 is .793 impl yi ng that this model explains 79 percent of the variation in the urban buil t land cover of cities in 2000 (Table 4 1) Overall model fit is quite good with F value of abou t 201 and significant at the 95 percent confidence level (.000). Regarding individu al explanatory v ariables, the coefficient of .779 obtained for c ity population implies that a 1 percent increase in city popul ation corresponded to about .78 percent increase in urban land co ver (built up area); 10 percent increase in city popul ation corre sponded to about 7.8 percent increase in the built up area of the average city These results closely resonate with the results ob tained by Angel et al. (2011 ), which also confirms the theoretical and empirical population and income in explaining urban spatial structure ( Angel et al., 2011; McMillen & McDonald, 2010; McGrath, 2005; Brueckner & Fansler 1983; Alonso, 1964 ) and urban expansion I n particular, Angel et al. (2011 ) have extended the original theoret ical model and estimation to cover a global universe of over 3000 cities (Angel et al., 2011 ; 2005) which have included both

PAGE 148

148 developed countries and developing countries explanatory variables have been found to be robus t and highly significant. To an extent theoretical model of urban spatial structure (Alonso, 1964) to reflect current world realities of globalization but still l eaves out FDI and remittances as important economic forces driving urban expansion The work presented in this chapter of the dissertation research draws extensively on this timely effort. However, it is important to reiterate that the main focus i n this dissertation in general variables (population and income) per se, but on emerging variables closely associated with economic globalization as they imp act the u rban spatial expansion Generally, the Adjusted R 2 increa sed from .793 i n the previous results (T able 4 1) to .820 (Table 4 2). This imp lies that the model explains 82 percent of the variation in urban land cover compared to 79 percent explained by the previous model which included only city population in 2000 and country GDP per capita in 1999 as independent variables. In terms of the co efficients, population and GDP remained variables, results show that 10 percent increase in FDI inflow woul d result in increase of about 1percen t (0.087) in the built up area of the city. In contrast the results indicate a negative relationship between built up area and remittance inflows. Specifically, 10 percent increase in remittances inflow would lead to about 1.5 percent decrease in the bui lt up area of the city. This is a surprising result given the fact that anecdotal evidence fo r Accra indicates that about 20 percent of remittances inflow to Ghana found its way into the real estate market of Accra (Tipple et al., 1998) which would be expe cted to fuel demand for land directly and through income

PAGE 149

149 (job creation) and possible inducement of migration into the city as jobs are created. Theoreti cally, therefore remittance induced urban land expansion is plausible as explained above but evidence fr om the regression estimation points to the contrary. May be remittances inflow has a threshold point which m ust be attained before showing a positive impact on the built area. One thing that may be taken into consideration is that, remittances unlike FDI may not be influenced so much by agglomeration hence its geography is likely to show a m ore diversified distribution, compared to FDI which could be significantly drawn to clusters of economic activities which more often than not are 1983). Next to be examined is the change in the explanatory power and overall fit of the models sequentially as the two globalization related variables foreign direct investments and remittances are included in the same model, toget her with population and income. For this purpose, s tepwise regression modeling approach was implemented the results of which ar e presented in Table 4 3. From Table 4 3, the two traditional variables population and i ncome (GDP per capita) explain about 79 percent (.793) of the va riation in urban built up area. The addition of annual remittances inflow (net) as a predictor results in a slight increase of the explanatory power of the model to about 80 percent (.804). Model 4 incorporates foreign direct investments as a predictor variable, in addition to population, GDP per capital and annual remittances flow. Adjusted R 2 cha nges from .804 to .820, thus Model 4 explaining 82 percent of the variation in city buil t up a rea as opposed to 80 percent in Model 3, 79 percent in Model 2 and 66 percent for M odel 1. Together, the two variables related to external economic forces, have added about 3.4 percent to the explanatory power of the model over the

PAGE 150

150 model incorporating o nly population and income. Additionally, the standard error of the Adjusted R 2 decreased by 6.7 percent {(.25347 .23643)/.25347 100)}, thus improving accuracy o f the explanatory power of Model 4 relative to Model 2. However small this change is, it is i mportant to note that the overall fit of the model remained highly stat istically significant at the 95 percent confidence level. Also, it is important to note that globalization is an on going phenomenon. Even though it has faltered somewhat currently ( 2008 2009), the general consensus is that it is picking up again. This implies that FDI will continue to flow into urban spaces and regions, including Accra in Ghana. Similarly, international migration is expected to be on the ascendancy, especially migr ation from developing to developed countries. Thus, other things held constant, remittances flow from developed regions/countries to developing ones is likely to rise significantly (Chami et al., 2008). It is from these consideration s that the results for the globalization related potential urban land expansion drivers may not be trivial. Specifically, the theoretical underpinnings and policy implications call for further investigation Results from the foregoing indicate that the addition of foreign dir ect investments and remittances as explanatory v ariables in the regression analysis (Model 3 and Model 4) alter the outcome, however small the change may be. At this pansion can stand on their own. The rationale here is to determine the extent to which the built up area responds to FDI and remittances inflow, leavi ng out population and income. Co nsequently urban built area is regressed on FDI and remittances inflows.

PAGE 151

151 Table 4 4 depicts the results or regressing built up area on FDI and remittances. The R square is .130 impl ying that the model explains 13 percent of the variation in built up area which is highly significant at the 95 percent confidence level. The co efficient of FDI has im proved markedly in this model; .193 as compared to .087 in the 2 implying that 10% increase in FDI inflows wo uld lead to an increase of about 2 percent in built up area. However, the remittances coefficient is negatively signed and not stat istically significant at the 95 percent confidence level. Overall fit of the model is quite modest (F = 8.853) but stat isti cally significant at the 95 percent confidence level. It appears that FDI is more robust than remittances, with superior degree of explanatory power. But h ow woul d these variables fare, each on its own? Table 4 5 displays the results of regressing built up are a on remittances. The results show that remittances alone do not have any statistically significant impact on built up area in cities in 2000. Specifically, the model explained less than 1% of the variation in built up area, with significance of .23 8. The constant (intercept), even though not considered in the analysis is, however, positive (3.512) and highly significant at the 99% confidence level (.000). Furthermore, the sign of the coefficient which has been negative when remittances were includ ed with the other explanatory variables has turned positive in this model. The positive sign is consistent with the theoretical postulates above. This notwithstanding, both the overall fit of the model (F value) and t value are statistically insignificant At this point it appears that remittances inflows do not impact urban built area. If model specification is assumed to be correct then remittances may not be playing a significant role in the conversion of land to urban uses (Further investigation is warranted).

PAGE 152

152 Next, built up area is regressed on FDI in order to determine the extent to which FDI alone, as an explanatory variable, would potentially impact the urban spatial expansion process. This model explains 13 percent of the variation in built up area compared to less than 1% explaine d by remittances alone (Table 4 5), which was not statistically significant at the 95% confidence level, in any case. The co efficient shows that 10% increase in FDI corresponds to about 1.7% increase in urban built a rea. Also, the F value of about 16.77 is highly statistically significant (.000). Compared to remittances, therefore, FDI appears to be a significant urban land cover chan g e driver in contemporary times. population and income (GDP /GNP per cap ita) FDI inflow may be assuming some importance in the era of globalization. However, remittances inflow produced surprising results, showing an inverse relationship with built up area contrary to the theoretical postulate advanced in this dissertation res earch. Discussion U rban Spatial Expansion A nalysis and Globalization Overlooked F actors? As noted above, population and income show up as highly significant drivers of urban land expansion in numerous studies with the general observation to the effect that urban built up areas increase in tandem with growth in population and income (as measured in GDP or GNP per capita). Thus, for e xample, it has been documented that population growth of about 4.5 percent per annum is way beyond th e national average of about 2.3 percent (GSS 2002 ). However, it is highly doubtful how significant this is to the physical expansion of the city during the period covered by this

PAGE 153

153 study (1985 2000/2002). Indeed, whether the population growth is driven by natural increase or rural urban migration or both it is the opinion of this investigator that this would not have any significant contribution to the contemporary expansion of a order for the people to be able to express effective demand for housing and land. Thus, when urban expansion is viewed narrowly, focusing on the population size living in the urban area, then population increase (presumably without purchasing power) can explain the expansion. In this regard an important fact to be taken into consideration is that, with respect to urban spa tial expansion in a primate city, particularly in the case of Accra in Ghana, the urban periphery is usually dominated by relatively high income dwellers contrary to the norm in many developing countries where the urban periphery ents or tenements (Aryeetey Attoh, 2010; Yeboah, 2003). For example, Professor Ian Yeboah has asserted that the (Yeboah, 2003) implying that affordability is usually above th e means of the average Ghanaian worker (Konadu Agyemang, 2001). This, in effect, rules out any substantial demand by average income earning Ghanaians for those quality houses found mainly in the peri urban areas of the city. Similarly, the importance of the usual traditional explanatory variables including m aximum slope, agricultural rent and shallow gr oundwater are highly questionable given technological advancements that enable even the draining of wetlands for urban/real estate development Consequen tly, this study takes the view that even though these factors are still relevant, they do not contribute significantly to the contemporary urban expansion process in many fast

PAGE 154

154 globalizing primate cities; at least, not in Accra. The argument being advanced in this dissertation research is that the physical expansion of primate cities in developing countries such as Accra in recent times, usually characterized by overconcentration of external funds, may have more to do with external economic forces flowing f rom population and income growth (GDP or GNP per capita) Nonetheless, g lobal urban land expansion analysts have consistently identified population and income growth (as m easured by GDP /GNP per capita) as main drivers of urban land e xpansion (Seto et al., 2011; Angel et al., 2011). Angel et al. (2011 ), in their study of a global sample of over 3000 cities found that city populat ion and GN P per capita explained over 75% of the variation in urban built up areas across countries. Similarly, Seto et al. (2011) concluded that GDP per capita explained most of the variation in urban land expansion in developed countries, but among developing countries population was found to be t he most significant driver of contemporary urban land expansion. The authors note that: Our model shows that urban land expansion in the fastest growing regions China, India, and Africa is driven by different mixes of factors. Annual growth in GDP per capi ta is related to approximately half of the observed urban land expansion in China but moderate or no expansion in India and Africa. Instead, urban land expansion in India and Africa is related more too urban population growth. (p 6). Moreover, the regres sion estimation results in this study suppo rt these findings. From Table 4 1, city population and GDP per cap ita, combined, explain about 79 percent of the variation in urban built areas. More specifically, when city built up area was regressed on city p opulation, the variable alo ne was able to explain about 66 percent of

PAGE 155

155 variables as drivers of urban land expansion c annot be overemphasized However, as revealed by t he conceptual frameworks (Chapter 2; Chapter 3) the theoretical or extension Chapter 4 several questions can be raised regarding the extent to which change drivers are able to explain contemporary urban land expansion in the light of intensified global linkage of economies with attendant interdep endence of places far and near, triggering unrestricted spatial financial flows. For example, could the continued importance of population Are there hidden underlying drivers that work through population numbers but which are difficult to untangle for analysis? Yes, population numbers will cause increase of the urban extent should the new development be characterized by low qualit y housing in the peri urban area; but this may not be the case in many instances. I ndeed it has been shown that the urban extent has increased in the face of population declines that may be attributable to changing life styles or affluence ( Burchell et al ., 2005). Thus, it is obvious that even though population increase may be the proximate cause of increasing urban extent, the underlying driver might be something other than pop ulation. Again, in respect of the contemporary growth of the city of Accra, Y eboah (2003 in 1985 the urbanized areas of Accra (including Tema) constituted 216 square kilometers, increasing to 555 square kilometers i n 2002 and the growth has largely occurred on the fringe lands of the city ( Moller Jensen et al., 2005; Grant & Yankson, 2003 ). In particular with respect to urban land expansion in developing countries, the purchasing power of the majority of the increa se in population should be taken into

PAGE 156

156 accretion (or compounding) at the urban periphery character ized by highly priced homes ( Aryeetey Attoh, 2010; Yeboah 2003). There is the need, therefo re, to re examine this issue, especially under economic globalization for a more comprehensive understanding. To take a more closer look at population, the conceptual and theoretical analyses in this dissertation have attempted to decompose the population of an assumed, typical globally linked city, into : resident city population before the advent of increased inflows of FDI and remittances as well as induced population migration; attracted by FDI and remittances rge/prima above. It is important to re iterate that t hese FDI and remittances induced migrants are likely to be more skilled tha n the traditional average rural urban migrant and will therefore be able to express effective demand for relatively higher quality housing since they will earn all things being equal. It is arguable that remittance purchasing power enabled Attributions have been made to this effect in the case of Accra (Grant, 2009; Grant & Yankson, 2003) In regression analyses of urban land expansion, however, the variable of city population used does not distinguish between resident ci ty population before FDI and remittances started flowing into the city. Consequently, population as an explanatory variable performs impressively but the relevant question one may want to ask is: I s population increase with little or no purchasing power a ble to express effective demand for housing and land to result in rapid urban spatial expansion such as observed f or Accra during 1991 2000/2002 ?

PAGE 157

157 It is a cardinal argument of this dissertation research that the requisite purchasing power, to some extent, has been afforded by inflows of external funds in the form of FDI and remittances. Similarly, a closer look at GDP per capita as an explanatory factor of urban land count ry/national ) level capture the total income and theoretical analyses presented in this dissertation research indicate that total income of city residents in the wake of inflows of external funds could be decomposed into 3 ca tegories, namely; (a) those city residents who were gainfully employed before the advent of FDI and remittances; (b) those previously unemployed people w ho find FDI and remittance related jobs; and (c) those people who migrated into the city and get employ ed in FDI and remittances related jobs such as construction and real estate. Obviously, t contain some induced aspects that could arguably be linked to el ements of FDI and remittances Unfortunately, GDP/GNP per capita usually used to denote income levels in urban expansion modeling and remittance are likely to enjoy. To this far, therefore, it appears that conventional analyses of contemporary urban expansion may be overlooking important factors related to the economic impulses transmitted by globalization into urban spaces which potentially feed into the urban spatial expansion process su pporting the general conceptual framework for this dissertation research as outlined in Chapter 1 In view of the foregoin g considerations, it seems reasonable to ask whether FDI and remittances could have direct or indirect impacts on urban land

PAGE 158

158 expansio n, especially pertaining to large / primate cities in developing countries which are fast globalizing In an era of globalization such as we are witnessing currently, it becomes more imperative that a better understanding of the full impacts of global eco nomic forces as key drivers of urban spatial change is sought since these potentially increase the rate of land conversion from agricultural uses / natural areas to urban uses. Do FDI and Remittances Drive Urban Spatial E xpansion? Few quantitative analysis of contemporary urban land expansion have attempted to directly implement FDI as an explanatory variable Seto (2005) included FDI as an explanatory variable in a regressi on analysis of the external dimensions in urban land use change modeling in a study of the urban land transfo r mation of the Pearl River Delta in China and the Re d River Delta in Vietnam. The inclusion of FDI, according to the author was motivated by the increasing levels of unrestricted capital inflows into the study areas as a result of increasing economic globalization. The study specifically investments and deve 198). The FDI related and upon testing, the variable was found to be positively correlated with both conversion of agricultural lands to urban uses and conversion of natural areas to urban uses, with sta tistical significance at the 95 percent confidence level (Seto, 2005: 205). Based on the conclusion reached by Seto (2005) with respect to the external economic influences on the urban lan d conversion process, it is obvious that foreign funds invested in construction activities in the local economy induced urban spatial expansion in the

PAGE 159

159 areas studied. However, t he potential impact of FDI on u rban spatial expansion more than likely goes bey ond direct investments in capital projects and into less obvious or demand as well as increasing levels of housing demand by expatriate wor kers attracted by FDI inflows. In short area than identified in the study by Seto (2005). For example, in addition to such investments in construction activities, expatriates demand housing, and FDI firms would require la nd for the building of offices and operational units such as warehouses and the likes which would all increase demand for land induced tendencies which also feed into the urban spatial expansion process (Chapter 3) motives as well as self interest motives in which remitters invest in real assets and busines ses ( Chami et al., 2008; Addison, 2004 ; Chapter 3 o f this dissertation rese arch). Remittances increase as senders maximize their utility with respect to self interest mot ives. The implications are that housing, real estate and other land consuming activities would increase considerably. The key reason for this is that in addit ion to desire of migrants ( Ghanaians in particular) estate/housing sector offers a relatively safe investment environment. Some research findings have indicated that remittances from citizens of developing co untries such as Ghana living and working in developed countries such as UK, USA and Canada form a substantial part of real estate demand in the large cities of their home countries (Grant 2003; Yeboah, 2003; 2000; Tipple et al. 1998 ; Diko & Tipple, 1992 ) Similarly, real

PAGE 160

160 urban areas of ci ties elsewhere (Sassen, 2006). regarding d eveloping countries ( Grant, 2007 ; Yeboah 2003 ; Grant 2001; Tipple et al. ,1998; Diko & Tipple 1992 ). ( levels, demographic changes and physical constraints of topography are o f secondary importance and are not significant determinants of spatial expansion, especially for cities such as Accra which are fast globalizing ) The rationale for this reasoning is that, firstly, the forces driving the real estate markets in such global ly linked cities relate more real estate development. To sum up, the combined effects of FDI and remittances inflow manifest in the urban landscape through real estate boom and other construction activities, usually in the primate / large cities (Yeboa h, 2003; Grant & Yankson, 2003; Tipple et al. 1998 ; Diko & Tipple, 1992 ); making it impe rative that these external variables be integrated into the standard urban land use theory (Alonso, 1964) to facilitate empirical research. In pursuant of the above, t related potential urban land conversio n drivers have been i mplemented in regression Equation 4 18 above, together with the more familiar or traditional explanatory v ariables The regression estimation results indicate that the adjusted R 2 change d from .79 (explaining about 79 percent the variation) in the c ase for populati on and income combined (Table 4 1) ; to .820 (explaining about 82 percent of the variation) ( Table 4 2) It is important to note that the model is still stat istically significant at the 95 percent confidence level,

PAGE 161

161 implying that the model t hat contained F DI explained 82 percent of the variation in the built up area of cities representing about 3.4% increase in explanatory power over the explanatory power of th e regression model that has only the combined population and GDP /GNP per capita as explana tory variables. T he stepwise regression e stimation results shown in Table 4 3 give further indication regarding the gradual change in the explanatory power (Adj R 2 ) of the overall model as FDI and remit tances are included. It is clear that adding FDI and remittances does not change the significance of the model (from statistically significant to insignificant), but rather they improve the overall explanatory power of the model, however small the change may be. In terms of the co efficients, FDI had 0.087 ; positively correlated with city built up area and stat istically significant at the 95 percent con fidence level, implying that 10 percent increase in annual FDI inflows corresponded to about 1percent increase in urban built up area. Essentially therefore, i f annual FDI doubled in amount th is would correspond to about 10 percent in the built up area of the city according to regression estimation results It is important to note that FDI inflows can often increase dramatically, such as has been the case for Ghana in re cent years. Figures from the G IPC indicate that during 1999 2007 FDI increased by about 290 percent With this versatility of change in FDI inflows, impact could be small cross sectionally, but may increase rapidly over time hen ce warrant ing further investigation. To further examine the potential of FDI as an increasingly important explanatory driver of contemporary land expansion, a bivariate regressio n Equation 4 19 was constructed and the regression estimation result s are sh own in Table 4 6. The results indic ate that FDI alone explained 13 percent (Adj R 2 = .13) of the variation in urban built up area. Also,

PAGE 162

162 the co efficient of FDI, when implemented alone as explanatory variable remained positive, with a value of .169, imp lying that a 10 percent increase in annual FDI i nflow corresponded to about 1.7 percent increase in the built up area. This is an improvement ove r the co efficient of FDI of .087 obtained when FDI was implemented together with population, GDP per capita a nd annual remittances inflow. Overall, the results of FDI, both when implemented together with the other variables and when impl emented alone, indicate that the variable may be a significant driver of contemporary urban expansion, which confirms the con ceptual formulations in Chapter 2 as well as giving some preli minary empirical credence to two of the initial (new) theoretical propositions (Proposition 1 and Proposition 3 ) ; following from the refinement of the Alonso (1964) theory. It is also important to reiterate that the empirical results obtained from the regression estimation in this chapter implementing FDI as explanatory variable are in consonance with what Seto (2005) obtained, even though in her case FDI amounts were not modeled directly or explicitly. Equation 4 19 implements FDI and remittances as the only explanatory variables of urban built area, with the results shown in Table 4 4. The model explains 13 percent (Adjusted R 2 =.130) and highly sig nificant (.000) The overall fitness of the model (F= 8.85) is significant (.000) at the 95 percent confidence level. Co efficient for FDI urban built up earlier was relatively more significant in magnitude; a 10% increase in annual FDI inflows corresponded to 1.9 percent increase in the urban built up area in a subsequent year. To the contrary, however, remittances co efficient continued to have a negative sign ( 0.078 )

PAGE 163

163 revealed, in quantitative terms, in this study may not be surprising to some geographers like Professor Ian Yeboah ( 2003 ; 2000 ) who, even though agrees that external s spatial development in recent times, has stopped short of attributing this to the external economic forces of globalization as described in this dissertation research (i.e. FDI or remittances). To a certain degree, the regression estimation results demo nstrated in this study show that such an attitude may be justified but as shown in the discussion a closer examination of remittances and FDI as explanatory variables (drivers) of urban expansion, indicate that the latter variable could have a statisticall y significant positive impact on contemporary urban expansion. Consequently, it can be argued that even though the preliminary empirical regression results presented in this dissertation research cannot be considered conclusive, the potential influences of external economic forces on the urban spatial expansion process cannot be dismissed entirely. The New Theoretical Propositions in Context Based on the regression estimation results and discussion above, it is obvious t hat FDI and remittances differ i n their influence on the urban expansion process. In essence, FDI appears to be more relevant whereas remittances appear not relevant. Consequently, it can be argued that Proposition1 and Proposition 3 have been validated indirectly, to an extent. Pr oposition 1 posited a positive correlation between FDI and the radius of the urban area Since, radius was not implemented directly in the regression models as a response variable, the built up area or urban land cover becomes the proxy. In effect, theref ore, the positive correlation between FDI and the built up area resulting from the regression estimation will also hold true for the radius. In other words, the expansion of the built up area of the city will also entail an increasing

PAGE 164

164 radius. I n short P roposition 1 is confirmed by the preliminary regression estimation results. For Proposition 3, the argument is that FDI attracts people into the city by virtue of jobs created; employment affords these in migrants the needed purchasing power to be able to express effect ive demand for housing and land, which would then feed into In the context of the regression analysis, therefore, city population implemented as an explanatory variable could be in fluenced indirectly by FDI inflows into the urban economy. In effect, there fore, part of the estimated co efficient for population may be attributable to FDI. As explained earlier, FDI gives the requisite purchasing power and will therefore not be entir ely out of place to suggest that the magnitude of impact of population to urban spatial expansion may not be that high without the influences of FDI. To sum up, FDI inflows into the urban economy may fuel demand for land, increasing urban land values. But peri urban lands remain cheaper compared to lands closer to the city core. At the same time, peri urban land owners would be incurring increasing opportunity cost for holding lands in agricultural use. It would make sense to convert to urban uses, contri buting to the expansion of the urban extent into the periphery. This explanation tentatively suggests that FDI induced urban land expansion is grounded in the Thunen Alonso theoretical framework under globalization (detailed explanation is given in Chapte r 5). Issues with R emittances and FDI as U rban Land E xpansion D rivers. Can the poor performance of remittances as an explanatory variable in urban expansion be explained? A few reasons readily come to mind, the combination of which could help explain the preliminary empirical results presented above.

PAGE 165

165 urban land conversion may be due to the fact that the bulk of remittances may not be invested in the real estate sector of urban areas and may be remitted for purely altruistic reasons which would confirm a popular view in the literature (Lucas & Stark, 1985). Secondly, it may be due to the fact that the asset accumulation motive of remitters which under Unlike FDI which responds to urban agglomerations and will therefore have the tendency to concentrate in already economically vibrant locations within a country, remittances may not necessarily follow agglomerations. The implication of this, from a geographical perspective, is that remittances are more evenly distributed spatially wi thin a country, compar ed to FDI; implying that enough remittance funds may not be concentrated in any single or a few urban centers (s) to enable a significant influence to be exerted on the urban expansion process. In essence, it is conceivable that some remitters will build or establish business in towns or in their home towns which are not necessarily large urban centers and therefore could not have been captured by the sample of 107 cities for the regression estimation in this dissertation research. Finally, it is importan t to note that the remittances flow data implemented in the regression that only documented remittances could be captured in such records. It is widely acknowledged, ho wever, that a substantial portion of total global remittances flow go through unofficial channels such as through friends and family members trav elling back Addison, 2004) Even though these reasons may help

PAGE 166

166 sign ( ) of the remittances co efficient cannot be readily explained and is very perplexing, for which more empirical investigations will aid our understanding. urban land expansion. It has been identified concep tually and theoretically in Chapter 2 Chapter 4 that FDI flow into an economy would afford residents purchasing power to enable them express effective demand for housing and land which could potentially contribute to spatial expansion; via demand from previously unemployed residents, induced migrants (indirect c hannels) as well as expatriate demand (more direct channel) which have been analytically decomposed above. However, the available dataset currently, do not allow for this decomposition for closer examination. Consequently, the dataset used for the regress ion estimation in this work may not be capturing all of the three pathways conceptualized and theorized in this work. Secondly, induced incomes earned by city residents, which would be more relevant for the argument advanced in this dissertation research, are not captured; neither can the GDP per capita used be decomposed into the different income earners of a globally linked city to enable robust analysis. Lastly, city population has not been decomposed into the various components arising in the wake of i ncreasing inflows of FDI. A very important fact is the failure by urban expansion analysts and modelers to recognize the purchasing power enablement afforded by FDI, thereby continuing to focus on population numbers (in absolute terms) and their growth ra tes used as explanatory variables. Indeed, the foregoing conceptual and theoretical analysis suggests that

PAGE 167

167 which have been overlooked by global urban expansion analysts. More importantly, population increase superficially. But it seems obvious, based on our conceptual and theoretical analyses that FDI (and remittances, even though not found to b e relevant in this work) do afford purchasing power to city residents which in turn enable them to On a more general note, therefore, this study does not purport to have carr ied out a comprehensive empirical analysis of these issues, which are better left for future research. The figures used for remittances and FDI used in this study are national figures but the urban built up areas used as a response variable are measured at the city level. Hence, there is no direct correspondence between the explanatory variables on the one hand and the response variable on the other in terms of scale of data measurement. One consequence is that city level dynamics of the explanatory variab les might not have been captured. Also, even though the importance of geographical distribution have been emphasized in the conceptual frameworks (Chapter 2; Chapter 3) an the data did not reflect this. Hence, the preliminary empirical analysis presented in this dissertation might have missed the potential impact of unequal geographical distribution of the external funds inflow into the urban spaces of the sample of cities used for the regression estimation. In a nutshell, this chapter might have ended up raising more questions than answers or geograph er, however modest.

PAGE 168

168 Globalization Induced U rban S patial E xpansi on ? T he world is urbanizing at a faster rate and developing countries have been forecast to have the highest proportion of urban settlements in the near future. According to some projections about 60 percent urban areas by 2030 (UN 2005; 2003) and scholars have devoted considerable of the inc reasing ecological footprin ts associated with cities and other ecological problems created by rapid urbanization (Keys et al., 2007 ; Rees & Wackernagel 2006; 1994; IHDP, 2005; Rees, 1992 ). More importantly, u rbanization is increasingly being recognized as one of the formidable asp ects of environmental change, posing one of the greatest challenges of the twenty first century ( IHDP, 2005 ; Turner et al. 2004; Sanderson et al., 2002 ; Alberti, 2003 ; Rees 1997; 1996; 1992 ). For example, Rees (1997: 63) asserts that besides growth in p opulation numbers increasing concentration of people in urban settlements constitute the most significant event for a century. Not surprisingly, therefore, u rbanization and its attendant environmental changes have engaged the attention of researchers aim ing to offer deeper understanding of human dimensions of gl obal environmental change (IHDP, 2005). For example, the IHDP (2005) has recently focused on the environmental change consequences of n and Global and its impact on the urban environment as well as human responses to changing urban env iro nments (IHDP, 2005). Current rates of urbanization and attendant lo ss of environmental resources such as depletion of vegetation cover have raised concerns among researchers largely due to the negative consequences for sustainable urban

PAGE 169

169 development ( Keys et al. 2007 ; McKinney, 2006; Wackernag el, 2006; 1994; Rees & Wacke rnagel 1994; R ees, 1999; 1992 ) which could undermine human well being in general (UN, 2 005; Alberti, 2003). T of ecosystems implications for human well being concluded in no uncertain terms that human well b eing was inevitably linked to quality ecosyste ms and their services to humans; impacting for example food availability, energy and good quality air (UN 2005: 1 25), just to mention a few. The advent of economic globalization, boosted by economic libera lization policies have helped ensure that urban economies are linked to the larger global economy resulting in unrestricted flows of funds across geographic locations globally, some of which are invested in the built environment. Consequently, the contemp orary expansion of built up areas, particularly in large / primate cities in developing countries, may be influenced by inflows of external funds into the urban economies. Not only does the expansion of the urban built environment lead to increasing human load on distant ecosystems, the process actually could have more direct and dire consequences on the immediate biophysical environment. Research findings have indicated that urbanization (expanding urban extent) have negative impact on the biophysical en vironment through loss of vegetal cover and consequently habitat loss, fragmentation and general loss of biodiversity ( Alberti, 2003; McKinney 2002 ; Marzluf 2001; Vale & Vale 1976; Luniak 1995). For example, McKinney (2002) has asserted that the speci es of many taxa often declines along the urban rural gradient with the lowest richness to be found in the urban core (p. 889), implying that urban expansion has negative impact on biodiversity. In more general terms, this has negative implications for eco system functioning as well as ecological integrity thereby

PAGE 170

170 impacting long term environmental sustainability (UN, 2005 ) So, does economic globalization impact the contemporary urban expansion process? This question is particularly imperative if the urban extent change is formulated as a function of the changing economic activities in the urban area such as increasing inflows of FDI and attendant population inducements as elaborated on previously in this dissertation research. From the general conceptual framework laid out in Chapter 1, an urban economic system linked to the global economy entails the exposure of the urban biophysical environment to external economic forces of globalization (Dicken s 2003). Also, in many cases, disproportionately large am ounts of external funds such as FDI are nt, 2007; Grant & Yankson, 2003 ; Yeboah, 2003) Thus, it seems plausible to argue that in an era of intensifying globalization, economic forces gen erated far away from the urban spaces they are affecting, may be contributing to their spatial changes and re organization. Not surprisingly, therefore, concerns for a better understanding of the changing planet in relation to socioeconomic forces and gl obal flows are high on the research agenda of geographical scientists (NRC 2010). In this regard, Julie Winkler has suggested that one area of research that should pre cope with the rapid spatial re organizati 317). Spatial re organization of economy and society entail, almost invariably, land use directions question number 7 bears directl y on the pertinence of globaliz ation and associated flows ( Chapter 1). Thus, urban spaces are continuously being shap ed by external forces (Grant & Yankson, 2003; Dicken s 2003), in addition to the traditionally

PAGE 171

171 recognized regiona l and local forces (Cadwa llader, 1985). This is especially pervasive in developing countries. Partly due to colonization, large cities in many countries in developing countries, in Sub Saharan Africa have assumed the status of primate cities. This has been further re enforced by economic liberalization and globalization as multinational corporations over concentrate their activities/operations in such cities by virtue of availability of infrastructure and relatively skilled pool of labor. Added to t his, is the increased migratio n which has characterized the era of globalization with associated increase in the flow of remittances into migrant origin countries, substantial portions of cities in developing countries. The co mbined effect of these can be seen through the rapid land conversion from agricultural use to urban uses, especially at the urban fringes, for residential and industrial purposes. Thus, for example, the rapid spatial expansion of Accra, the capital city of Ghana, like many similar cities in West Africa and Sub Saharan Africa, has an important external component as drivers of ur ban land use change. Essentially, spatial changes observed at the local setting are more likely the combination of many forces, som e of which are external and beyond the control of urban land managers, among other things Even though the external aspect of urban change is generally recognized by researchers, research effort has largely concentrated on the proximal (or local) causal a gents. Thus, the external dimensions of urban spatial change are still less understood. This is especially worrisome given the fact that increasing spatial interconnectedness of economies has helped exposed the local (urban) environments to the vagaries of global economic forces, which would have implications on the urban spatial change process. In essence, therefore, most urban land / spatial expansion

PAGE 172

172 (population and income growth). However, the an alys is presented in Chapter 4 should help draw attention to global economic influences on the urban land expansion process. It is gratifying that the process can be embedded in a standard economic theory in order to facilitate further empirical investigations for a fuller understanding of spaces. Summary So to what extent are FDI and remittances important as drivers of contemporar y urban spatial expansion? Chapter 4 has sought to advance a cogent argument for a refinement of the standard urban land use theory/ model (Alonso, 1964) in the con text of economic globalization. The overriding aim has been an attempt to refine the Thunen urban spatial expa nsion f rom a geographical perspective. Thus, the theo retical postulates advanced in Chapter 4 are informed by the classical economic theory of urban spatial structure, focusing mainly on the urban extent expansion which has been derived from the theory ( Angel et al. 2 01 1 ; 2005; McGrath, 2005 ; Brueckner, 1 987; Brueckner & Fan sler 1983 ). Specifically, a nnual FDI and remittances inflow were al. 2011; 2005 ). P roposit ions relating FDI and remittances to the urban extent were put forward and tested, using log transformed OLS regressi on estimation approach. FDI was found to be positively correlated with urban built up area, confirming the conceptual framework and theore tical postulates in this dissertation; and also supporting findings by Seto (2005). T o the contrary, however, remittances, as a driver of urban expansion w as not found to be relevant since the estimate d coefficient was negatively signed as

PAGE 173

173 opposed to the hypothesized positive correlation. These findings imply that, empirically, Hypothesis 1 is partially supported; with respect to the FDI urban expansion correlation but the empirical evidence does not support the hypothesized positive correlation between r emittances and urban spatial expansion. On theoretical and conceptual grounds, however theoretical framework, for useful and informed analysis, based reasonable argument s ther eby giving prelimin ary support to H ypothesis 2. An important conclusion so far ( Chapter 2; Chapter 3; Chapter 4 ) is that the more subtle underlying factors most likely induced by FDI (and rem ittance) inflows which afford many the purchasing power to be able to express effective demand for housing and land which in turn could result in increasing urban extent. It is als o suggested that the use of GDP/GNP per capita as explanatory variable ma y not capture incomes that city residents might enjoy which have been made possible through FDI and remittances related jobs. I n effect, therefore, Chapter 5 has argued that more attention needs to be paid t o the growing importance of FDI in particular as an of contemporary urban spatial expansion, especially in places where such external funds are likely to be attracted into existing urban agglomerations. T he case for the inclusion of FDI in the standard economic theory and empirical est imations appears to be supported by the preliminary empirical regress ion estimation presented in Chapter 4 and is also resonates with previous research (Seto, 2005) C hapter 4 of the dissertation research contributes to theory building in contemporary urb an geographical research and urban economic geography in the

PAGE 174

174 context of economic globalization; examining the significance of the external economic dimensions in the spatial expansion process. The continued intensification of global linkages of cities and economies makes it imperative that urban policy makers and managers be cognizant of the impact of external forc es on their object of interest since m any primate cities in developing countries such as Accra in Ghana are grappling with the environmental con sequences of globalization and urbanization (Grant 2009, Yeboah 2000, 2003, Grant and Yankson 2003 and Moeller Jensen et al. 2006). The bro ader perspective of Chapter 4 therefore, is the anticipation that urban planners and urban environmental managers w ill find insights useful in their decision making, with due consideration given to external forces of glo balization in the form of FDI. At the minimum this will help ensure that the external dimension is recognized and anticipated. Additionally, Chapter 4 provides the necessary economic theoretical grounding f or the formulation of the Integrated Urban Spatial Expansion Estimation Method (IUSEEM) in Chapter 5, implementing the estimated co efficient of FDI in an allometric based exponential growth model to attempt to cumulate FDI induced built up area of Accra over time.

PAGE 175

175 Figure 4 1. based on Reid (2002). Distance from Market Transport cost/ unit output / unit distance Market Perishable good Less perishable good Non perishable good

PAGE 176

176 Figure 4 2. Alonso's theory an d model, bid rent curves and urban spatial structure

PAGE 177

177 Table 4 1. Regression results with population of individual cities and GDP per capita Table 4 2 Regression results with population, GDP per capita, FDI and remittances Table 4 3 Stepwise regression results and R 2 change Model R R 2 Adj R 2 S.E F Sig 1 .815 a .665 .662 .32387 206.21 3 .000 2 .893 b .797 .793 .25347 201.730 .000 3 .900 c .809 .804 .24654 144.439 .000 4 .909 d .826 .820 .23643 120.270 .000 1 Predictors: (Constant), CityPOP 2000 2 Predictors: (Constant), CityPOP 2000 GDP 1999 3 Predictors: (Constant), CityPOP 2000 GDP 1999 REMIT 1999 4 Predictors: (Constant), CityPOP 2000 GDP 1999 FDI 1999 REMIT 1999 Dependent Variable: Built 2000 (Bui lt up area of sample of cities ). Table 4 4. Regression results with built up area regressed on FDI and remittances Co eff Std. Co eff S.E t Sig VIF Constant 1.556 .292 5.402 .000 CityPOP 2000 .782 .804 .035 18.373 .000 1.002 GDP 1999 .290 .359 .043 8.209 .000 1.002 N = 107 F=209.343 (.000) Adj R 2 = .797 S.E of est = .252 D W= 2.025 Co eff Std. Co eff S.E t Sig VIF Constant .959 .417 2.301 .023 CityPOP 2000 .787 .810 .040 19.458 .000 1.019 GDP 1999 .215 .267 .046 4.645 .000 1.94 0 FDI 1999 .087 .162 .032 2.763 .007 2.034 REMIT 1999 .134 .139 .041 3.245 .002 1.080 N = 107 F = 121. 70 6 (.000) Adj R 2 = .820 S.E of est =.238 D W =2. 096 Co eff Std. Co eff S. E t Sig VIF Con sta nt 2.895 .796 3.638 .000 FDI 1999 .218 .405 090 4.345 .000 1.068 REMIT 1999 .063 .065 050 .697 .487 1.068 N=107 F =9.522 (. 000) Adj R 2 =.139 S.E of est. =.52034

PAGE 178

178 Table 4 5 Regression results with built up area regressed on remittances onstruct. Table 4 6. Regression results with built up area regressed on FDI Co eff Std. Co eff S.E t Sig VIF Constant 3.512 .672 4.918 .000 REMIT 1999 .036 .037 .094 .378 .706 1.000 N = 107 F = 1.406 (.238) Adj R 2 = .008? S.E of est. = .56289 Co eff Std. Co eff S.E t Sig VIF Constant 2.432 .394 6.860 .000 F DI 1999 .209 .388 .041 4.095 .000 1.000 N = 107 F = 18.650 (.000) Adj R 2 = .143 S.E of est. = .51906 D W = 1.887

PAGE 179

179 CHAPTER 5 FDI INDUCED PERI URBAN LAND EXPANSION IN ACCRA, GHANA: THEORETICAL AND QUANTIFICATION C ONSIDERATIONS Background In Chapter 5 the IUSEEM is formulated; FDI induced urban spatial expansion theory is more formally articulated and the phenomenon is placed in the broader context of human induced environment al c hange. It has been demonstrated in Chapter 4 that FDI and remittances can be embedded in the classical Alonso (1964) theory in order to help account for realities of economic dels employed, like many regression based urban expansion analysis, have been cross sectional, static and linear. Thus, m ultiple regression models such as employed by Angel and colleagues (2011; 2005) ; and in Chapter 4 can represent the explanatory variab les for cross sectional data, therefore not capable of modeling urban spatial expansion spatio temporally in addition to being linear. Even more un settling, perhaps, is that current urban spatial expansion quantification methods and models in general, do not allow for the incorporation of rapid changes in predictive variables such as FDI, in order to reflect thei r dynamic impacts on the urban land expansion process through time Furthermore the most popular spatially explicit urban expansion models suc h as cellular automata (CA) and Agent Based Models (ABM) focus mostly on local whose land use decisions aggregate to impact physical expansion of the urban area (Batty, 2005 ) Desirable attributes of such models are their ability to model the cit y as a complex system, effectively handling non linearity and being dynamic, which according to advocates most closely approximate real urban dynamics (Batty, 2005 ). However, n ot only are the underlying motivational forces masked, but these models, for t he most

PAGE 180

180 part, have not been particularly linked to economic theory even though economic considerations seem to weigh heavily in t he decision making process that lead to urban growth (Pacione, 2009 ). For example, in a rapidly expanding and globalizing city decisions made by peri urban agricultural land holders to convert to urban uses may have been driven by increasing demand for urban land re sulting from FDI inflows (Chapter 2; Chapter 4) but would normally not be represented in CA and ABM. In short th e emerging, globalization related FDI induced urban land expansion phenomenon has neither been systematically the orized nor quantified, leading to a less than complete understanding of the urban spatial expansion process, even though it is widely acknowle dged that globalization has led t o increased interconnectedness of u rban economies; opening up these economies and their lands to economic forces flowing from global sources (Dicken s 2003 ) Grounding FDI induced urban land expansion in standard economic theory and quan tifying the process will increase understanding of urban spatial expansion of primate, fast globalizing and rapidly expanding cities in the developing world, especially, where FDI flows have consistently concentrated. The broad goal in this chapter therefore, is to seek to understand the extent to which increasing FDI inflows to Ghana might have contributed to the rapid rate urban areas, as Ghana liberalized economi cally, with Accra fast globalizing and attracting unprecedented levels of FDI inflows (GIPC, various years) In other words, it is suggested in Chapter 5 that Accra has been experiencing FDI induced physical expansion since about 199l. For the purposes of the analysis in Chapter 5 and tentatively, FDI induced urban land expansion may be described as expansion of the

PAGE 181

181 urban extent primarily driven by FDI and related economic activities in the urban economy which impact demand for land related amenities such as housing and infrastructure that necessitate the conver sion of new lands to urban uses pital city ha s undergone dramatic and rapid physical expansion in contemporary time s with a marked acceleration of urban annual rate of growth begi nning 1991, according to a study by Moller Jensen and colleagues (Moller Jensen et al., 2005) T he annual rate of physical expansion of Accra far outstripped the rates of growth for population of Accra (Angel et al., 2011) and incomes in Ghana (Table 5. 7 ) during the same period. In short, expansion seems to defy explanation based on the two most important traditiona l outside the the analysis given the imperatives of economic globalization which unleash global capital flows that have the tendency to influence urban land use decisions in distant areas In some respects, t loc akin to the impacts of climate change in local pla ces, which cuts across multiple geographical scales. Such globally induced local changes are becoming increasingly common with the in t ensification of globa lization, advanced technologies and increased economic flows. Global e conomic flows have tendencies to cause spatial re organization of economies and societies (Dicken s 2003 ), which almost invariable contribute to land use changes ( including land convers ion to urban uses ) In this regard, some researchers ( Aryeetey Attoh, 2010; Grant, 2009; Grant & Yankson, 2003; Konadu Agyemang, 2001; for example) have rapid physical expansion,

PAGE 182

182 but such ana lysis have not been grounded in economic theory and have not sought to It is t he main argument Chapter 5 that the increasing levels of FDI inflows to Ghana during the era of intensifying globalizatio n and economic liberalization (im plementation of SAPs), which has had the tendency to geographically concentrate in Accra, may be contributing to the rapid rate of physical expansion of the city, especially beginning from 1991, which may be feeding into pe ri urban environmental / land degradation. Arguably, land degradation may not be the most pressing problem affecting urban residents in Ghana, in general, and residents of Accra in particular. For example, researchers and commentators have sought to hig hlight the worsening situation of urban poverty, malnutrition (Songsore, 2003 cited in Baabereyir, 2009), lack of proper waste disposal management, (Baabereyir, 2009), among other things. Much as these issues are important, meriting attention, the issue o urban land degradation in Accra in recent times (Moller Jensen, 2005; Grant & Yankson 2003 ) cannot be wished away either, given the rapid pace and the haphazard nature of contemporary physical expansion o f the city ( Aryeetey Attoh, 2010; Moller Jensen et al., 2 005; Grant & Yankson, 2003). So, c an FDI be implicated in the peri urban land conversion process in Accra? In Chapter 5, attempt will be made to present reasonable arguments as to whether or not FDI can be implicated in the rapid physical expansion of Accra into the peri urban areas and specifically develop a simple spatial expansion estimation method to quantify the FDI induced expansion 2011 In Chapter 4 of this dissertation, initial proposi tions relating FDI to urban built up

PAGE 183

183 sectional regression models and FDI was found to be positively correlated with urban built area and stat istically significant at the 95 percent confidence level. This, in very general terms, would suggest a link between FDI and urban land conversion such as found by Seto (2005) for the Pearl River Delta in China and the Red River Delta in Vietnam. However, su ch analyses offer limited understanding from a temporal perspective. Given the dyna mic nature of FDI flows, it is reasonable to suggest that FDI contribution to urban built up area expansion would vary according to the flu ctuations in the flow volume and time, other things being If urban land conversion is virtually irreversible and adversely affects the biophysical environment (Seto et al., 2011), it seems imperative that we are able to account for just how much of a given natural/agricultural area is transformed into built up areas through time, as driven by some identified underlying drivers such as FDI. Indicat ions are that the FD I linkage to urban land expansion is theoretically founded, which can be modeled analytically for estimation of FDI induced urban built up area o ver time. Furthermore, preliminary estimation results using the allometric based urban expansion model (IUSEEM) formulated and implemented in Chapter 5, indicate that the role of FDI as a potential land cover change dri physica l expansion cannot be ruled out completely, thereby suggesting that further empirical research in this area ma y be warranted. Thus, even though the an alysis presented in Chapter 5 cannot be considered empirically rigorous by any stretch it is hoped the w ork will help lay the conceptual framework and theoretical foundations for modeling and estimating the linkage between FDI and urban land conversion over time, which in the absence of proper land use and effective environmental manageme nt

PAGE 184

184 policy, could lea d to land and environmental degradation such as is believed to be happening in Accra currently. Moreover the analysis in this chapter has the potential of motivating further research into the construction of simple estimation models and methods more real istically reflecting the dynamic FDI urban land conversion relationship which are also theoretically grounded. Research outcomes in this area would likely have important research applications and policy implic ations beyond Accra and Ghana, p arti cularly, for urban planners policy makers and urban environmental conservationists unintended consequences for the urban biophysical environment This need would be especially imperative for urban plann ers and urban land managers tasked with especially in sub Saharan and West Africa FDI Induced Urban Spatial Expansion in the Context of the Li terature FDI flows have characterized glo balization, with the tendency of concentrating in urban locations largely due to benefits of agglomerations. FDI induced urban spatial expansion has received little attention in the mainstream urban expansion analys is litera ture, yet it is difficult to deny that economic globalization increases economic activiti es drive urbanization usually resulting in increasing urban extent Hence, the FDI urban expansion linkage should be obvio us. However, a very basic but impor tant question still needs to be answered : How does foreign generated money (FDI) impact urban spatial expansion at destination locations such as Accra? One aim of this dissertation is de monstrating the need for urban and economic geographers to seriously begin considering the emerging globalization related phenomenon of FDI induced urban spatial expansion, especially as relating to primate cities in d eveloping country settings.

PAGE 185

185 Consequently, t his literature review will seek to contextualize the emerging ph enomenon of FDI induced urban spatial expansion in the relevant literature and demonstrate the need to formally incorporate FDI in the Thunen Alonso theoretical framework I t will be shown that it is about time FDI is incorporated in standard urban econom ic theory, particularly analyzed from a geographical perspective especially bordering on the fundamental geographical themes /concepts of FDI is a m ovement of global capital across locations on the surface of the earth. Similarly, urban land expansion is an instance of human environment interactions (one sided though), whereby human made capital is used to convert agricultural and natural lands to hu man made structures; in turn, the existence of such structures induce further rounds of human economic activities resulting in more land conversion. Arguably, therefore, FDI induced urban spatial expansion should form a major focus of urban and economic ge ographical research as a cardinal contribution to human dimensions of global environmental change (IHDGEC) research efforts. The main themes of the literature review are t hree fold: (1) Human environment relationship context of FDI induced urban spatial ex pansion which provides an overview of the economy environment relationship which generate human induced environmental change (environmental degradat ion) to which, this dissertation argues FDI induced urban spatial expansion could be contributing; (2) th eoretical background to FDI induced urban spatial expansion outlines the relevant theories that have been concerned with human land relationship, considering land as a vital input to human economic endeavors, particularly for housing and other similar purp oses The focus has

PAGE 186

186 been limited to perspectives from economic geography, urban geography and urban economics. The overall aim here is to identify the theoretical gap in the literature necessitating the formal theorization of FDI induced urban spatial expa nsion; and ( 3 ) Quantifying FDI induced urban spatial expansion is aimed at identifying the need for an appropriate method/model that can be used to estimate urban built up area expansion in response to increasing FDI inflows into the urban economy. Being a relatively new area of theoretical exploration, quantifying FDI induced urban spatial expansion will aid understanding as well as enhance the process of theory development in this area. Consequently, a new model/method is suggested, drawing on regression analysis, exponential growth modeling (similar as espoused in population ecology) as well as the principles of allometric growth (empirical law in biological science ). This integrated urban spatial expansion estimation model (IUSEEM) proposed in this stu dy is capable of handling non linearity, is truly dynamic and is suitable for quantifying human induced environmental change, with specific reference to FDI urban land expansion over time Human Environment relationship Induced Environmental Change Human induced environmental change and degradation have been a major concern of geography since the days of George Pe rkins Marsh, through to Carl Sauer and to the present. The United Nations International Strategy for Disaster Reduction (UNISDR) defines e Going by this definition of environmental degradation the concern with ecological needs is obvious but, perhap s, a more comprehensive definition would be that Environmental degradation is the deterioration of the environment through depletion of resources such as air, water and soil; the destruction of ecosystems and the extinction of wildlife (Wikipedia, 2012)

PAGE 187

187 Environmental degradation is increasingly being recognized as a formidable global problem an d is noted to be one of the Ten Threats officially cautioned by the High Level Threat Panel of the United Nations T Millennium Ecosystem Assess ment concluded that as much as 15 of 24 identified ecosystem services had been degra ded within the last 100 years (UN, 2003) The report noted that key human activities such as increasing use of fo ssil fuels and natural resources were undermining critical ecosystems services which could have adverse consequences on food supplies, freshwater resources, wood and fiber, among other things. Thus, environmental degradation is staring humanity in the face. Important components of environmental degradation inclu de changes in the atmosphere, biosphere and specifically, the land and its resources. Deterioration in the conditions of the last category, usually described as land degradation has some resonance with FDI induced urban land expansion as espoused in this dissertation. According to the United N any process that results in (United Nations, 1994 cited in Adewuyi, 2011: 32). However, land degradation can be attributed to numerous causes, spanning natural occu r rences and human made activities. It is widely held that land degradation is an important issue plaguing mankind currently, having potentially negative implications for agricultural p roductivity, environmental sustainabil ity and food security Even though land degradation may result from numerous causes anthropog enic causes have been found to be mo re enduring and even accentuate naturally occurring events that d egrade the land (UN, 2 009) For example, it has been noted that structures built across watercourse s and

PAGE 188

188 wetlands for urban development h ave contributed significantly to wors en flooding in Accra (Grant & Yankson, 2003 ). From this perspective, the major causes of land degradat ion may include l and clearance, depletion of soil nutrient s o vergrazing, dumping of industrial wastes, mineral exploitation, oil spill, urban development in general and sprawl, in particular. That these human activities are driven mostly by economic moti vations is not too difficult to see. For example t he World Bank degradation can be a by product of eco (Croitoru and Sarraf, 2010: 1). What i s mor e contentious though, is the fact that economic globalization drives environmental change ; and the extent to which globalization can be implicated in environmental change and degradation. Globalization and Environment Jennifer Clapp and Peter D auvergne write : Globalization is a multidimensional process, broadly restructuring and dynamic, ongoing, and accelerating process that is increasing the links among actors, as wel l as the structures within which they operate, both within states and across borders. Trade, production, and finance are now more globally integrated than ever before, as are global organizations and social movements In simple terms globalization means that the events and actions in one part of the world are affecting people in distant lands much more quickly, and with greater frequency and intensity (Clapp & Dauvergne, 2005: 20). The It further suggests that we are movin g toward an effectively borderless world, especially for ideas and mon ey 20). Two important observations relevant to the argument in this chapter can be drawn from the foregoing: (i) consequences of decisions in one location ma y generate c hange, inclu ding land use/land cover changes in locations far away from the origin within a short time ; and (ii)

PAGE 189

189 almost seamless transmission of ideas and money across the surface of the earth. In a similar vein investment decisions made elsewhere tr igger the flow o f global money (mostly in the form of capital such as FDI ) across the surface of the earth Thus, globalization also embodies economic flows whose impact on origin and destination locations alike could be analyzed. However, this is a daunting task due to t he complexity of the phenomena involved, a source of worry to analysts as has been rightly pointed out by Clapp and Dauvergne(2005): Separating the drivers from the consequences of globalization is difficult, because the consequences are themselves consta 21) It is generally accepted that economic activities do ha ve environmental repercussions as has been noted by Lelia Croitoru and Maria Sarraf : Industrial production often discharges pollutants into clean rivers and air, p reventing the use of these resources for other purposes and harming the health of those exposed them. Unsustainable agricultural practices can reduce crop productivity and cause dam sedimentation. Overexploitation of groundwater increases pumping costs and if it leads to saltwater intrusion, may make aquifers unusable (Croitoru and Sarraf, 2010: 1) However, what is not too clear is the globalization environmental change linkages usually bordering on the causes and impacts In effect, w hereas environmen tal change driven by human economic activities may be less contentious, it is not easy for people to agree on a set of environmental problems that can be linked to causes due to globalization, resulting in people having their own preferred worldviews. Four of such worldview categories presumably most dominant, have been described by Clapp and Dauvergne (2005) which may provide further elucidation for the present argument Firstly, Social greens make social and environmental justice their main focus with th e

PAGE 190

190 (Clap p & Dauvergne 2005 : 14 ) According to proponents of this worldview l arge scale industrial ization such as brought on by economic globalization leads to exploita tion and degrades the environment. Thus, globaliza tion according to social greens, accelerates exploitation, inequalities, and ecologica l injusti ce; arguing that the acceleration of globalization enables overconsumption by d eveloped economies which put en ormous stress on the environment (Clapp & Dauvergne, 2005: 14 ). Secondly, Bio environmentalist s focus their attention on e cosystems, believing that human survival is capaci ty is exceeded by overpopulation, excessive economic growth and overconsumption. To the bio environmentalist, therefore: E conomic globalization drives unsustainable growth, trade, investment and increasing debt, which accelerate the depletion of natural r esources as wel l as problems of waste disposal. (Clapp & Dauvergne, 2005: 14) It is to be noted that the social greens and the bio environmentalists both emphasize human consumption as a contributor to global ecological crisis but differ on the specifics. Bio environmentalists believe strongly that growth should be limited, population controlled and consumption reduced but social greens do not specifically concern themselves with these believing that social justice will mitigate problems resul ting from globalization, including the ecological crisis. Again, w hereas ecological crisi s is only one of the concerns for the social greens, it is the main concern of the bi o environmentalists. Thirdly, Institutionalists do not acknowledge that there is already ecological crisis but believe there is the potential if humanity fails to strengthen and improve institutional capacity in order to address environmental problems and underdevelopment, among other things They believe that economic globalization c an be managed to promote stronger institutions that will ultimately improve human welfare

PAGE 191

191 and enable people to manage the global env ironment. In effect therefore, I nstitutionalists believe economic globalization would impact positively on the environment if properly managed through stronger institutions, but advocate a cautious approach to human development This view contrasts sharply with both the social greens and the bio in essence. Lastl y, Market Liberals focus their attention on the free market functioning of economies, ome inevitable problems but overall modern science, technology, ingenuity, and money are i mproving the global environment. Clapp & Dauve rgne, 2005: 14). M otivated by neoclassical economics, Market L iberals hold the view that economic growth and increasing levels of per capita force, because it promotes eco lapp & Dauvergne, 2005: 14 15). economic growth, globalization, trade, foreign investment, technology and the noti on of Dauvergne, 2005: 4 ). One cardinal assertion of market liberals is that economic growth and poverty alleviation can best be pursued with increasing globalization not less (Clapp & Dauvergne, 2005). Since market liberals favor increased globalization, per haps it is fair to say that their worldview as described above has dominated the economy environment relationship thinking as has been exemplifi ed by postulated an inverted U shaped relationship between economic growth and environme ntal quality in what has been described as the environmental Kuznets c urve ( Antle & Heidebrink, 1995 ; Gross man & Krueger, 1991 ). The basic tenet of the theory (or hypothesis) is that at the initial stages, economic

PAGE 192

192 growth would impact negatively on environ mental quality, but as economic growth and development proceed, countries would be better equipped to take good care of the environment leading to improved environmental quality. Essentially, therefore, the Environmental Kuznets Curve (EKC) theory attempt ed to associate economic growth and environmental quality degradation but, in the view o f some researchers (Clapp & Daugverne, 2005 ; Costan za et al., 1997, among others tulating that even though there would be deterioration in environmental quality society would be ab le to protect the environment in the end The flaws inherent in this argument have been document ed in the literature (Clapp & Daugverne, 2005; Costanza et al., 1997 ), and widespread environmental quality deteri oration has been observed vironmental problems. This conference on Environment and Development in 1987, giving birth to perhaps the most popular phrase in global development environment al discourse currently in the famo us It appears that the concept of sustainable development captures many of the concerns from the four camps enumerated above at least, literally. The World Commission on Environment and Development (WCED) defines sustaina without compr om Nat ions, 1987, cited in Clapp & Dauvergne, 2005: 4). Sustainable development is usually conceived o f as comprising of the three components of economic social and

PAGE 193

193 en vironment (Costanza et al., 1997 ) and its appeal cuts across traditi onal disciplinary and political boundaries, attesting to its perceived usefulness to the social economic environment disc ourses. For instance, Betten court and Kaur (2011) note that The concepts of sustainable development have experienced extraordinary success since their advent in the 1980s. They are now an integral part of the agenda of governments and corporations, and the ir goals have become central to the mission of research laboratories and universities worldw ide. (p. 1). Many researchers would agree that the ent in sustainable development has contributed immensely to the new field of Sustainability Sci ence that is currently gaining enormous popula rity among a wide community of researchers. Judging from its scope, Sustainability Science appears to be the most relevant discipline yet that tackles the social economic environmental con flicts head on. For example, Bettencourt and Kaur assert that T he main themes that define the field, the concept of integrated management of human, social, and ecological systems and of the engineering and policy studies that support and enable them, are t he true crosscutting subjects that unify the field. (Bet tencourt and Kaur, 2011: 4). Thus, Sustainability S cience embodies researchers from traditional disciplines both in the physical and social sciences, includi ng geographers, economists and urban planners, attempting to use scientific tools and methods to effectively address issues bordering on human environment relationship. Importantly t development is not shrouded in secrecy according to Bettencourt and Kaur (2011); asserting that The single most important feature of growth in the field is the steep rise in its growth rate in the late 1980s and early 1990s. This corresponds to the years that follow ed the publicat ion of the Brundtland report, a widely acknowledged formative document for the field published in 1987 and

PAGE 194

194 around the time of the important publication of Agenda 21 at th e Rio Earth summi t in 1992 (Betten court & Kaur, 2011: 5 ). This is not to suggest that the problems have been resol ved to the satisfaction of everyone, at least not to some urban planning scholars and practitioners For example, Scott Campbell, a well know n urban planner and professor of Planning has a take: In the coming ye ars planners face tough decisions about where they stand on protecting the green city, promoting the economically growing city, and advocating social justice. Conflicts among these goals are not superficial ones arising simply from personal preferences. No r are they merely conceptual, among the abstract notions of ecological, economic, and political logic, nor a temporary problem caused by the untimely confluence sustainable development aspires to alluring, holistic way of evading these conflicts, they cannot be shaken of f easily. (Campbell, 2003: 435 458). Coming from a seasoned urban planner, the concern would imply that cities have not been spared by the controversies surrounding the economy environment relationships and that there is yet to be an enduring solution Thus, w ith particula r reference to the economy environment linkage in the urban setting it is not clear Sustainability S cience has developed tools and methods that can be used t o trace specific economic drivers of globalization to some specific types of urban change in destination locations in quantitative terms. For example, how does foreign money invested in urban economies of countries oth er than the origin of the money affec t the urban biophysical environment by way of spatial expansion ? Understanding such a linkage is imperative because there are good indications that increasing globalization fuels high levels of economic activities Furthermore, h igh levels of economic a ctivities would require more environmental resources as input and also for waste disposal both of which have been known to increase en vironmental stress. Also, that the city is growth for nations has long been recognized High l evels of economic activities with

PAGE 195

195 increasing levels of urbanization, could impact nega tively on the environment potentially degrading the natural system. Consequently, in very general terms, considerable research efforts have been geared towards understa nding the impacts of economic growth on the physical environment, especially on land and its biological resources ( Turner et al., 2004; McKinney, 2002; Sanderson et al., 2002). Concerns about global land cover change has increased considerably, particularl y since the 1980s when remotely sensed data became available and other methods of analyzing land cover change became sophisticated (Lepers et al., 2005: 115; Turner, 2002). Land cover chang e can simply be described as the alteration of the surface of the earth by humans in order to enhance their well In so doing, therefore, natural ecosystems and bioregions are greatly altered and modifi ed. The process has contributed significantly to species habitat loss and consequently to a general decline in biological diversity biodiversity loss (IUCN, 2006). In particular, i t is generally acknowledged that global biodiversity is seriously under t hr eat (UN, 2003 ). Other ramifications of rapid land cover change glo bally that have been identified include modification of hydrological processes and climate change (Defries et al., 2004), for example Biodiversity can be conceived at three levels ec osystems, species and genetic biodiversity (Spray et al., 2003: 29 33) and land cover change, especially, those resulting in impervio us surfaces such as urbanization (urban land conversion) has the potential of affecting all three levels of biodiv ersity ( Defries et al., 2004) mostly driven by increasing economic activities These are just a few examples which go to buttress

PAGE 196

196 the assertion that economic activities are increasingly impacting negatively on the biophysical environments of the world. As noted a bove, urban areas are at the center of global economic growth and therefore the fact that natural areas of cities have been caught in the global environmental change process is no longer in dispute Urban Environmental Change V iewed from a purely econo mic perspective, urban spatial expansion can be considered as progress for people since such urban growth usually occur s when the national and urban economies are booming. However, urban size and its change over time have seve ral implications, including i ncreasing costs of commuting, rising levels of carbon dioxide emission loss of sensitive ecological lands, among other things, often resulting in poor quality environment or environmental degradation (Set o et al., 2011; Wackernagel, 2000 ; Costanza et al., 1997 ). Major urban environmental quality deterioration (often termed environmental degradation) may be grouped into : (1) pollution (2) congestion, and (3) degradation of natural support systems (World Bank, 20 04). The latter is generally held to encompas s pollution, loss of biodiversity, waste or toxic accumulation and disposal, and general depletion of natural resources through activities such as deforestation, depletion of fresh water resources, paper consumpti on, among other things (Jha & Murthy, 2003, cited in UNDP, 2006: 35). In order to measure the extent to which urbanization impact these important ecological constituents a number of studies have focused on identifying and quantifying the extent of human A particula rly relevant effort which resonates with this work is the human footprint or ecological footprint analysis (Wackernagel, 2006; Sanderson et al. 2002 ). Sanderson et al. (2002 ) were able to map the human footprint based on factors such as land transformati on rates, human access to resources, human

PAGE 197

197 population density and power infrastructure (pp. 892 895). According to the authors, more economically advanced countries such as USA and the Western European countries were found to score higher (high impact) th an developing or poor countries. The general conclusion was that human beings (Sanderson et al., 2002 ) This supports the fact that increasing urbanization levels, usually associated with economic growth could also lead to inc reasing depletion of natural / environmental resources. Similarly, Wackernagel et al. (2006) have argued that rapid addition to land conversion for city expansion in creasing ly, resources from distant locations are required to support the urban population. Therefore, even though the ds, high urbanization levels usually correlate with high levels of economic growth, implying that corresponding ecological footprint would be high in high economic growth, highly urbanized areas. For instance, it was found that the average Canadian needed 7.5 hectares of land to support his/her consumption. The average urban dweller in Mexico and India, on the other hand, required about 2.4 hectares and 0.7 hectares, respectively ( Wackernagel et al., 2006: 105). Thus, it would seem that increased economic growth and attendant increased urbanization exert larger ecologi cal impacts and would have the potential of lowering environmental quality and sustaina bility, if not properly managed Arguably, therefore, within a given country, urban dwellers would appropriate more ecological resources compared to their rural counter parts. Generally, therefore, as di scussed above, Sanderson et al. (2002) and Wackernagel (2006) have shown that

PAGE 198

198 resources could come from afar and encompass large areas far in excess of their physical sizes, thanks to international trade and globalization. Hence, it has been asserted that even though urbanize d areas constitute only about 2 percent o f the land surface of the earth, urban dwellers consume over 75% of the e generate over 70% of wastes which would require increasing capacities of assimilation and natural recycling by the environment ( Wackernagel, 2006 ; Turner et al., 2004; Costanza et al., 1997 ). A particularly disturbing aspect of urbani zation in relati on to urban natural areas te removal of the entire vegetation cover when pristine and agricultural lands are converted into urban uses. More formally, as more lands are transfo rmed into built up area, a proportionate amount of vegetation cover is lost in the biophysical system. This is plausible because unlike other components of ecological footprint such as the food footprint, built footprint Wacker nagel, 1996) ; implying that the full impact of the spatial expansion must be borne by the regional biophysical system of which the urban area is a subsystem (Chapter 1 ) the ecological consequences o f urban spatial expansion according to the argument advanced in this dissertation, which could result in env ironmental and land degradation in the absence of effective urban growth management policy. It would not be gainsaying to suggest that such a situation would have negative implications for environmental quality and undermine urban sustainability efforts and degradation is a long recognized environmental issue, which straddles both the physical and socia (Adewuyi, 2010: 29), placing it in the purview of sub jects like urban economics, economic geography and urban geography. Many researchers in these traditional

PAGE 199

199 disciplines who study land change processes and consequences are contributing to S ustainabi lity S cience under Land Change Science Integrating GIS, re mote s ensing and statistical modeling, scholars have studied land change dy namics in agro ecosystems, forest systems with some few attempts relating to peri urban l and conversion However, it seems that b y focusing on local agent behavior and masking eco nomic motives such approaches may be giving less than fuller explanation of contemporary urban land expansion, especially as pertains to peri u r ban land expansion in primate cities in developing country set ting. Again, e ven though such works may be groun ded in economic theory they fail to incorporate influences of global economic forces explicitly in the peri urban land conversion process Given the preponderance of FDI to concentrate in urban areas, human induced urban environmental change studies would benefit from contextualization in a broader Globalization Economy Environment framework, rather than the usual economy environmen t framework. This is because the latter would have the tendency of limiting the focus to the urban economy since it correspon ds to the urban environment. However, in an increasingly globalizing era with linked urban economies and liberalized national economies, economic forces of globalization may be more potent as drivers of urban land use decisions than traditionally recogniz ed giving rise to what has been described in this study as FDI induced urban spatial expansion which may contribute to urban environmental degr adation via land degradation. Nonetheless, FDI may be found in places other than urban areas, and contribute to environmental degradation in such areas. For example FDI in mining have been known to cause land degradation and pollution of water bodies in rural areas in Ghana However, w hat makes urba n

PAGE 200

200 environmental degradation more precarious in an era of intens ifying economic globalization is the fact that economic globalization has been characterized by increasing FDI flows which have tendencies to concentrate in existing economically viable locations that are usually urban areas. In effect, FDI could contribu te to urban environmental degradation as economic activities increase d in the wake of increased FDI inflows into the urban economy ; by accelerating the land conversion process but has not been contextualized in standard urban economic theory that would all ow for expected Consequently, it may be worthwhile to integrate concepts, ideas and theories from some relevant disciplines, which perhaps, may add to the efforts aimed at co nstructing an acceptable and useful theory of sustainability. We are reminded that: We are in an era of globalization. As a process that touches on many aspects of our economies, societies, and cultures, it is important to investigate how it interfaces w ith global environmental change. (Clapp & Dauvergne, 2005: 19). In this regard economic geography, urban geography and urb an economics appear to have some prospects worth considering Globalizat ion Economy Environment : G eographical P erspective s and Urban E conomic T heory Economic globalization is a major driver of economic activities across the surface of the earth ; therefore, economic globalization is partly to blame for global environmental change, falling into the realm of human induced environmenta l ch ange. However, the globalization economy environmental change relationship is a complex one with different worldviews bordering on the causes and ways to mitigate adverse consequences as has been note d above It seems obvious that geography as a discipl ine that straddles the socioeconomic and the biophysical processes is relatively

PAGE 201

201 better positioned, from a single subject perspective, to offer valuable insights. Thus, human impacts on ecosystems, including lands and its resources, can be best understood from a geographical perspective. Indeed, it is common knowledge that geography, from the beginning of its inception as an academic discipli ne has concerned itself with human environment interactions (Ruhl et al., 2007; Johnston et al., 2002 ; Johnston, 2000 ). However, i t is no secret th at geography lacks a single all encompassing definition but it is hardly the on ly discipline plagued with this; it is especially so among the social sciences Nonetheless, p erhaps definition of Human G eography will best suit the purpose here. Johnston has defined Human Geography as That part of the discipline of geography concerned with the spatial differentiation and organization of human activity and its interrelationships with the physical envi ronment. (John ston et al., 2002 : 353 ). Thus, even from a human geographic phenomena the need for geographical analysis to link the human activity being studied to the physica l / environmental basis is stated without ambig uity. One cannot help, therefore, but to agree with J.B. Ruhl and colleagues when they note that Although terms have changed and concepts have become more precise over the decades, geographers have been keenly interested in natural capital and ecosystem s ervices, and in human impacts on these, throughout modern time s. (Ruhl et al., 2007: 36). Essentially, environment relationship led to two important but relegated debates in geography wh ich have come to be known as Environmental D eterminism and Environmental P ossibilism Proponents of Environmental D eterminism (William Morris Davis, Ellen Chruchill Semple and Griffith Taylor, for example) argued that humans were passive and

PAGE 202

202 responded to environmental and climatic influences. Indeed, Ellen Semple is noted to 2001: 42), implying that human behavior is governed by p hysical environmental dictates. That this was an import ant endeavor in its era has been underscored by R.J Johnston, thus: Instead of merely presenting information in an organized manner, either topically or by area, geographers sought explanations for the patterns of heir major initial source for explanations was the physical environment, and a theoretical position was established around the belief that the nature of human activity was controlled by the parameters of the physica l world within which it was set. (Johnsto n, 2001: 42) However, Environmental Determinism was found to be overly simplistic as well as containing racist connotations which were in bad taste drawing adverse criticisms Consequently, reactions to Environmental D eterminism gave birth to the opposi te side of the same coin, Environmental Possibilism According to R.J Johnston (2001) Environmental Possibilism P resented the individual as an active rather than a passive agent Led by French geographers, followers of the Annales s chool historian Luci en Febvre, P ossibilists presented a model of people perceiving the range of alternative uses to which they could put an environment and selecting that which best fitted their cultural disp ositions (p. 43). The key notion and important point, is that the environment offered humans possibilities that with culture (tools and implements) humanity can achieve their economic agenda. scholarship in this regard. For example, J.B. Ruhl and colleagues note: Carl Sauer, in his 1925 The morphology of the Landscape conceptualized between nature and society. His work contrasted strongly with the environmental determinist s chool dominant at the time (Ruhl et al., 2007:37).

PAGE 203

203 Thus, before concerns that human activity was causing major environmental changes gained popularity beginning in the 1960s, geographers had been grappling with how to better conceptualize and analyze huma n environment relationships. Admittedly, both Environmental Determinism and Environmental Possibilism went to extremes, t hey cannot be ruled irrelevant. Even with regard to the so called infamous Environmental Determinism it is not too clear if we have t otally abandoned the notions expressed. It appears that even people outside mainstream geographical research are still holdi ng on. For example, Ruhl et al. (2007) concur that Environmental Determinism was resoundingly rejected in the 1940s due to its asso ciation with racism and imperialism (p.37) But the authors continue Nevertheless, Jared Diamond, in his Pulitzer Prize winning book, Guns, Germs and Steel (1999), has reported the issue of the powerful influence of geographical relationships and ecos ystem characteristics on the unfolding of human history in a manner that rejects racism while embracing the archeological record, modern genetic analysis, and history as a natural laboratory. Diamond theorizes, for example, that the existence of domesticab le wild plants such as wheat and barley and wild animals such as goats, sheep, and cattle launched the Fertile Crescent, the northern part of what is now termed Middle East, onto a trajectory of social development that includes urbanization, political hier archy, metallurgy, writing, and, unfortunately, pandemic diseases derived from livestock. The longitudinal diffusion of these social innovations east and west, and to similar latitudes in the Western and Southern Hemispheres, explains, according to Diamond why Europeans, and to a lesser extent East Asians, have been able to dominate indigenous peoples of the Americas, Australia, and southern Africa. (Ruhl et al., 2007: 37 38). That a book containing this assertion should win a respectable prize is food fo r thought. Perhaps, a middle ground between Environmental Determini sm and Environmental Possiblism would have put geographers ahead in the area of human environment relationship study but this never materialized. As a result, the assertion that : F or over a century geographers have not com e to agreement over the raison society

PAGE 204

204 relationships, and this leaves geography lacking a core theory that can be applied to the subject of ecosystem service s. (Ruhl et al., 2007: 38, citing Turner, 2002 ). may not be too far fetched In effect there is no core geographical theory explaining human environment interactions However, human economic activities continue to change the physical environment. Global capital continues to flow into distant locations (including urban areas) and e ffect environmental changes in these localities Even without a core theory, the discipline of geography has rich perspectives to offer in understanding globally induced urban e nvironmental change such as FDI induced urban land expansion (and degradation) especially from the perspective of economic and urban geography More specifically it has been noted: As a field of study, economic geography focuses on the flows in the econ omy: t The economic flows of interest in economic geography are not confined to goods and other materials, however, because flows of people, services, and information a lso are necessary in an economy. (Hanink, 1997: 2). Global capital moves from one location to the other in search of high returns on investments. Basically, f rom a geographical perspective, there is a spatial differentiation (a core focus of geography) in the distribution of returns on investments, necessitating movement (an important fundamental theme of geography) Similarly, the sub discipline of urban geography takes an integrative approach seeking to understand human environment interactions ( an important fundamental theme of geography) an u rban setting. Consequently from a geographical perspective FDI flow across the surface of the earth is an instance of geographical imperative of spatial differentiation at work whereby foreign investors seek out locations with potentials for high returns on capital Once such a place is located and capital flows in human economic activities requiring the use of some environmental resources, including land, get reinforced or get

PAGE 205

205 started giving rise to an instance of human environment interactions In the context of the concept of FDI induced urban spatial expansion inflows of FDI leads to more lands being converted to built up areas; once built these may attract more rounds of FDI inflows more land conversion follows, and so forth. However, more detail s can be added for a more comprehensive unde rstanding by integration of the geographical perspectives with urban economic theory, particularly relating to the Thunen Alonso theoretical framework (Chapter 4 of this dissertation) drawing on current extensio ns such as provided by Angel et al. (2011; 2005) The Thunen Alonso Theoretical Framewor k The basics of the Thunen Alonso Theoretical framework drawing heavily on current extensions capable of modeling urban extent expansion as a function of socioeconomi c variables, have been laid down in Chapter 4 of this dissertation. Distance, a key geographical imperative, is an important unifying concept affecting the geographical distribution of urban land value s .lands close to the CBD command higher prices compar ed to lands in the urban periphery generally This reflects an original attempt by economists to input geographical considerat ions in economic theory, after years of relative neglect (Alonso, 1964) I n formalizing the classical urban land value theory, therefore, William Alonso Distance was to play a pivotal role in the explanation of differential land values in the city. Even though a progr ess, urban land value theory constructed this way leaves out ma n y important considerations from the perspective of geography which when include d may give a more complete understanding especially for the emerging phenomenon of FDI induced urban land expans ion For instance, the concept of spatial differentiation may help

PAGE 206

206 explain why FDI is /can be concentrated in the city, which in turn may partly explain increasing demand for land, even when local population and incomes (traditionally most important urban land expansion drivers) may remain unchanged or even decline. In this regard geography offers the critical perspective needed to look beyond what is happening in the local urban economy and country. The importance of this cannot be overemphasized given e ver increasing economic globalization in the current era. On the national and local scales spatial differentiation may help explain why FDI concentrates in primate cities within the national urban systems unequal geographical distribution of potential h igh returns on investments such locations offer better prospects compared to other urban areas in the country. However, the details still requires economic theoretical analysis of agglomeration economics for example. For now, focusing on what FDI does on ce in the urban system: Economic activities rejuvenated, demand for land increases (Chapter 2; Chapter 4), and driving urban land values higher. Peri urban land values are lower compared to inner city lands but higher than agricultural land because return s of peri urban land put to urban uses would earn higher returns than in agricultural uses In short, holding lands in the urban periphery in a rapidly expanding and fast globalizing city may not be a wise land use decision since one would be incurring ri value and peri urban land expansion gets accelerated. Following from the fo regoing discussion prel iminary hypotheses may be developed by the integration of economic geography, urban geography and urban economics. U rban economic theory informs the geographic perspective just as the

PAGE 207

207 geographic perspective s and themes inform economic theor y akin to what t he geographer, I. Burton said many years ago and concurred by Harvey (1969): of an economic geographer is to refine and adapt available economi 1963: 159; cited in Harvey, 1969: 119). P remises for the initial hypotheses Distri butions of Potential Returns on investm ents are geographically uneven: locations with high (at least, perceived) returns attract high inflows of FDI Foreign investors move global capital to attractive locations resulting in the spatial differentiation of F DI distribution globally: FDI peaks and troughs. Urban areas usually characterized by agglomerations, skilled labor pool, infrastructure and the likes, serve as attractive foreign investment locations drawing increasing volumes of FDI FDI inflows into t he urban economy rejuvenate economic activities; create jobs, contribute to increasing demand for housing and land. Increasing demand for housing and land cause urban land values to rise: But peri urban land is cheaper than land in t he inner city. P eri urban land value is less than inner core land value but highe r than agricultural land value. For urban use purposes, therefore, peri urban land is cheaper, and profitable. To the peri urban land owner, the land will be put to its highest and best use by co nverting to urban uses ( sell to developers or self build ), at least to avoid increasing opportunity cost associated with holding the land in agriculture. Based on the outlined premises above, the fol lowing initial hypotheses are advanced: Hypothesis 1 a : Su stained FDI concentration in urban location accelerates rate of expansion of the urban extent Hypothesis 1b : V arying levels of FDI concentration will result in differential rates of urban land expansion

PAGE 208

208 Hypothesis 2 : Rate of FDI inflows to urban location and rate of expansion of the urban extent are related by exponential function; best explained by exponential model. Testing the hypotheses above would require an appropriate method capable of quantifying the rate of urban land expansion as a function of t he change in the volume of annual FDI inflows. Quantif ying FDI Induced Urban Spatial Expansion It is common knowledge that t he socioeconomic system and the biophysical system are interrelated in comple x ways (Costanza et al., 1997; Jansson et al 1994). Humans affect and are affected by the environment as has long been forged by geographers and relationships are more likely non linear and dynamic, especially when put in temporal perspective Tracing the environmental footprints within a given region of an economic activity, generated even within the same region, is a daunting task which is further compounded by the fact that observed environmental change in a s made far away before impacting local l and use decisions elsewhere FDI induced urban spatial expansi on is a typical case in point i nvestment decision s made by people all over the world result in the flow of global capital into desirable high returns locations, with u rban areas being particular ly attractive locations The following review will lay the foundations for developing an integrated urban spatial expansion method or model (IUSEEM) which is non linear dynamic and stochastic in functional form, in order to closely approximate the process interfacing the socioeconomic biophysical systems Moreover, and perhaps more importa nt, it would be desirable if such a method is effective but simple enough to be widely accessible to inte rested researchers as well as being firmly grounded in some stand ard or acceptable theory, especially in an urban economic theory The question

PAGE 209

209 then is: is such a model / method readily available? Indications are that such a method or model, specifically capable of quantifying FDI induced urban spatial expansion over ti me in a manner described a bove is not currently available, h ence the need for t he proposed model / method formulated in this work Modeling Models often are structured by some theoretical framework, and yet theories have been defined in a multitude of way s in the socia l sciences (Jaccard & Jacoby, 2010: 28). Jaccard and Jacoby (2010) cite Hollander (1967: 55) thus: Basically, a theory consists of one or more functional statements or propositions that treat the relationship of variables so as to account fo r a phenomenon o r set of phenomena. (Jaccard & Jacoby, 2010: 28). In spite of the numerous definitions it is generally believed that at the core of all theories is made up of concepts and relationships, leading Jaccard and Jacoby (2010) to simply defin is widely recognized and appreciated among researchers to the extent that some hav e argued t he development of theory is at the heart of all explanation, and most writers doubt if observation or description can be theory light of the foregoing, it is not too difficult to understand why a model or estimation method with solid theoretical foundation is likely to be viewed favorably within the scientific research community and would have the potential for practical applications. Regarded as a process, theory consists of formulating concepts and developing them i nto s ymbolic expressions (Jaccard & Jacoby, 2010: 28). It must be noted that in the social sciences the dividing line between a theory and a model is oft en thin and blurred (Jaccard & Jacoby, 2010; Harvey, 1969) A more simplified perspective on theories

PAGE 210

210 shared by thi s author is that: (Zetterberg,1965, cited in Harvey, 1969: 87). Just as there are numerous definitions of what a theory is there are multitude of theories that purport to explain urban gr owth and spatial structure. These include but not limited to, location theory, the rank size rule and 2010: 12 13). The elaboration of the theory types is not the focus of this chapter, however. Aniya Masamu (2010), for example, provides more insight s in this regard. But in seeking to model urban spatial expansion, as driven by an external economic f orce (FDI), the study 1969: 486). Consequently, this chapter attempts to link spatially explicit urban growth modeling to th e standard classical economic theory of urban land use, in the light of economic globalization to constr uct a simple urban spatial expansion (urban land conversion) estimation method. However, the complexity of the human environment interactions study req uires gaining some control by isolating only the potentially relevant variables for closer scrutiny. Consequently, model construction becomes inevitable, if not desirable. In other words, the complexity of the relationships between socioeconomic variables and biophysical systems (to which this writer argues FDI induced urban spatial expansion belongs) necessitates some simplifications in order to focus on the most important aspects, hence the importance of modeling in order to achieve the goal s in this cha pter. Ther e is no single definition for what model is. However, in very broad terms, according to the renowned geographer Ron Johnston and colleagues, a model can be

PAGE 211

211 20 02: 508). Most important from the geographical viewpoint, it has been argued that it can also include reasoning about the real world by means of translations in space (to give spatial models) or in time (to give historical models) Mea dows (1957 cited in Chorley & Haggett, 1970). Thus, spatio temporal relationships study which occupy exists as a patterned and bounded connexity which has been explored by the use of simpli fied patterns of symbols, ru Haggett, 1970: 22). For Chorley and Haggett (1970), therefore, a model is thus a simplified structuring of reality which presents supposedly significant features or relationships in a gen eralized f orm (p. 22). It is commonly accepted, therefore, that a model is an abstraction from real world phenomenon in an attempt to generalize in order to focus on the fundamental (presumably most important) aspects of some real world process. This requires some degree of selectivity, leading some researchers to characterize models as highly subjective approximations in that they do not include all associated observations or measurements, but as such they are valuable in obscuring incidental detail and in allowing fundamental aspects of reality to appear (Chorley & Haggett, 1972: 22). In effect therefore model building is a worthwhile exercise which can result in rounds of increasing understanding about real world phenomena since it is simplified and may guide f urther development ; and in this regard the author of this dissertation re search agrees with other researchers enough to suggest novel hypotheses and speculations in the pr (Black 1962: 232 233, cited in Harvey, 1969).

PAGE 212

212 Models may be causal or correlational and in time meaning that the change in the explanatory v ariable precedes the change in the response variable, ideally. Correlational models, on the other hand, normally assert some correspondence between the response variable on the one hand and the explanato ry variable(s) on the other w ithout making causal i nferences in an e xplicit manner Either implicitly or explicitly models contain propositions in order to test their validity and fitness based on observed real world behavior. Generally, two main propositions may be assumed by the model builder which wo uld result in being classified as deterministic or stochastic Firstly, based on observation of the process to be modeled a modeler may assume that the observed variation in the dependent variable can be explained by the selected set of explanatory variab les in that an exact relationship is postulated between the dependent variable on the one hand and the set of selected explana tory variables on the other In short the change in the dependent variable is determined by the explanatory variable (s) This is what has been described as a deterministic model. More formally, a deterministic model can be defined as mathematical representation in which every variable alters according to a mathematical http://www.biology online.org/dictionary/Deterministic_model ) A key aspect of such models is that they do not allow for randomness in the change process in that any randomness may be assumed a way or may not be deemed to have any important eff ect on the dependent variable. In practical terms these models work well when they are applied to physical processes such as those studied by the disciplines that deal with physical phenomena

PAGE 213

213 such as phys ics, chemistry, computer science and the likes. With regard to decision making such models would be useful if all or almost all required information (data) are available and can be quantified with some reasonable accuracy (Malczewski, 1999). Deterministi c (mathematical) models have been utilized extensively in optimization, goal programming as well as simulation approaches. For example, the usefulness of optimization modeling in decision making is well known and are believed to yield best results under gi ven constraints approach to identify the best solution (1999: 66, citing Wilson et a l., 1981 ) Secondly, the observed variation in a phenomenon (dependent variable) may not be com pletely explained by a set of selected explanatory variables. This may result from inadequate information, difficulty of measurement (quantifying) or the existence of general randomness in the process. Given this situation it would be reasonable for the mo deler to incorporate randomness in the explanatory portion of the model in order to account for the random fluctuations in the variation. Thus, the inherent nature of the process could be reasonably accounted for. Models built on such assu mption are known as stochastic models. model is a mathematical model which takes into consideration the presence of some randomness in one or more of its parameters or variables ( http://www.biology online.org/dictionary/Stochastic_model ) It must be noted that, unlike deterministic models, even though such stochastic models are characterized as mathematical they do not posit exact relationsh ip between dependent variable and independent variables. By design, therefore, stochastic models are well suited to deal with social and economic

PAGE 214

214 phenomena most of which show high degree of probability and randomness in their behavior. Generally statisti cal models are mostly stochastic in form and function Regression modeling, by far the most popular explanatory modeling technique in the social sciences, is based largely on the assumption of randomness in the behavior of the response variab le (Rogerson, 2005; Agresti & Finlay, 1999). According to Rogerson Of particular importance in stochastic modeling is the recognition of the existence of random behavior and its explicit incorporation in the models which is directly included in regression models. Thus in such models part of the observed variation in the response variable is deemed explainable and can be accounted for by the selected independent variable (s) whereas part of the variation cannot be explained by the available data in the model represen ted by an error term (to depict random ness). The integrated urban spatial expansion estimation method to be formalized below incorporates stochas ticity by implementing coefficients of estimated regression model as the exponent in an exponential growth model ; and at the same time is deterministic because it utilizes the exponential function Since geographers deal with a combination of subject matt er cutting across both the physical and the social sciences, modeling appears unavoidable. Richard Harvey has been particularly emphatic stating that With very weakly developed geographic theory and a highly complex multivariate subject matter, it is ine vitable that the model concept should play a pa rt in geographic explanation. (Harvey, 1969: 168). Harvey further advances the argument that given the peculiar situation of geography that deals with complex phenomenon, we cannot help it but rely extensive ly on models since theories are lacking, if our research efforts are aimed at explanation and

PAGE 215

215 prediction (Harvey, 1969). In the final analysis, Harvey seems to believe that the route to developing a geographical theory is none other than through modeling when he building in (p. 168). In this regard models become a means to an end; an end which continues to be vi able in the face of mounting societal problems bordering on the socioeconomic biophysical environment interface which in the opinion of this writer (agreeing with a broad spectrum of geographers) is firmly grounded in the fundamental geographic theme of human environment interactions. Even though complex, one is encouraged by the fact that applying scientific models afford us the necessary control over such complex phenomena (Ackoff, 1962: 108, cited in Harvey, 1969: 168). In this sense, therefore, the approach adopted in this study to model contemporary urban spatial expansion would be an attempt to exercise control over a complex phenomenon, isolate important response t o increasing flows of FDI through time. Overview of Current urban spati al expansion modeling Ideally, the cardinal goal of urban researchers would be geared to understand, in a comprehensive manner, all the interrelated, complex functions of the city but it is generally acknowledged that it will be next to impossible to deal with all the important variables and elements at any particular point in time. Thus, Aniy a Masamu has stated that Understanding the dynamic process of urban growth should be based on the linkage with decision makers. The interaction between the spatial, temporal and decision making process is much more complicated and decision making behavior is subjective and fuzzy (Aniya Masamu, 2010: 22).

PAGE 216

216 And so has it been with the urban growth mo deling process. Very generally, urban modeling prospered in the Post World War II era, largely as a result of the quest to find solutions to urban blight and deteriorating health con ditions in many cities (Hall & Tewdr Jones, 2010). Even though the moment um slowed in the 1980s, there has been current resurgence, with particular focus on dealing with the inadequacies of the traditional urban models that had been largely static, linear, and cross sectional, among other thing s. One important programmatic out come of this turn of events is the focus on urban models, with proponents like Professor Michael Batty (2005). Thus, the dawn of complexity theory approaches in urban mo deling (Batty, 2005; An iya Masamu 2010) were upon us. Usually used in conjunction with GIS, the defining characteristics of these computationally intensive models are believed to be dynamism, non linearity, disaggregated micro, bottom up, process oriented and the mergi ng of space and time (Sui, 1998, cited in Aniya Masamu, 2010: 22 23). In very basic terms, proponents argue that the city is a complex system (Batty, 2005), with the processes characterized by irreducibility, emergence, instability and unpr edictability, among others ( Aniya Masamu, 2010 ; Batty, 2005 ; Sui, 1998 ). The most successful, so far, among the grained scale of equal ly sized grids; and simulated according to some pre defined rules, generally known as Cellular Automata (CA) models. to change the state of a cell given some const raints physical, economic, and the likes also known as the Agent Based Modeling (ABM). Implemented in GIS, these

PAGE 217

217 computationally intensive modeling approaches have promised to better simulate urban growth, urban sprawl in particular, in a most accurate manner w ith some researchers believing they could better handle spatial dynamics than the traditional urban growth models (Batty, 2005 ) A particularly ambitious objective has been the claim to accurately represent time and space dimensions, which are integral to the urban system, for a comprehensive understanding of the city through time and space (Batty, 2005 ) Even though these are noble objectives, they are still yet to be realized as affirmed by Batty (2005), and w itnessed by Aniya Masamu : Batty (2004) note s that most of the new generation models are not inspired by the kind of practical imperatives that forced planners and modelers first to si mulate the city in the 1950s and 1960s but technical developments in GIS, urban remote sensing, and by what st ill remain largely abstract arguments concerning the need to explore cities as emergent complex systems from the bottom up (Aniya Masamu, 2010: 24). It is further noted that T he new generation of models provides a great potential for understanding urban complexity. However, they remain theoretical exploration and still have a long way to go to model urban systems effectively (Aniya Masamu, 2010: 24). of these so called modern urban growth models are not grounded in standard urban economic theory. These observations should be unsettling given that these models were meant to replace the traditional models. It appears that urban simulations have been fun so far but are still not the only way to go. Thus, it seems therefore, that there is a need to re models which are steeped in urban economic theory such as the classical economic theory of urban spatial structure postulated by Alonso (1964). Motivated by these developments the argument is advanced in this chapter to the effect that in modeling we

PAGE 218

218 are reminded that one does not necessarily have to understand all the individual behaviors of the actors in the system, albeit cells, in order to understand how the urban system works such as the approach taken by most researchers in the so called Land Change Science community most relevant approach to understanding land conversion f rom agricultural to urban uses. Methods include use of GIS, Remote S ensing and Statistical modeling. However, Land Change S cience approaches seem to be more concerned with the effects of decisions made by lando wners that result in the change in the cover of the land, focusing on individual parcels of land as they change over time, sometimes land cover change modeling, a lmost in variably, the force driving the decision is neglected Sometimes attempts are made to link survey results to the land change process but the focus will still be limited to processes at the local urban or national scale whereas the major driver may be flowi ng from sources originating outside the countr y in which the city is located. Additionally, this leads to increasing complexity and error prone not to mention the ambiguity explaining exactly how the simulation model works However, the situation can be simplified if one could link the underlying driver (such as FDI) to the land cover chang e in a systematic way. The argument is as follows: (a) L and owners are interested in maximizing their utility, and will put the land to activity that brings the highest return (of course there would be ex ceptions). (b) P articular socioeconomic conditions would motivate land owners to put their lands to some particular use s For instance, agricultural land owners in the peri urban areas of a fast globalizing city with exp anding real estate demand will more than likely profit from the

PAGE 219

219 land by putting it to urban uses. (c) If such a situation prevails in many instances, one can bypass the decision maker and link the land cover change to the specific socioeconomic conditions suppo sedly motivating the land owners to seek t o turn the land into urban uses. This is the key underlying assumption for the estimation method developed in this chapter current extensions (Angel et al., 20 11 ) comes in handy. Such theoretical postulates can be formalized using mathematical formulations (calculus) and validated using cross section regression models (Angel et al. 2011; 2005; Chapter 4 of this dissertation). With log transformation, the estim ated coefficients can be interpreted as elasticity or percentage change, denoting the percentage change in the built up area in response to percentage change (s) in relevant socioeconomic explanatory var iable(s) In other words the estimated coefficients of the log transformed regression models can be regard ed as This is a very important, desirable attribute that can be used to integrate regression models and exponential growth modeling as espoused in the IUSEEM aimed at overcoming the linear and static nature of the cross sectional regression models. The allometric growth principle (or Allometric Law) is drawn upon to help account for the system subsystem relationship between urban biophysical system and economic sub system which is pos tulated in this dissertation to the effect that the economic subsystem (urban economy dominated by built structures) grows at the expense of the biophysical system. Thus, FDI induced urban land expansion depletes natural vegetation proportionally since la nd conversion to urban uses removes the vegetation cover almost completely.

PAGE 220

220 Exponential g rowth f unction (Model ) and the Allometric l aw : The population growth m odel Generally the expo nential function is denoted by e x and used as a standard model approxim ating the relationship in which an independent variable induces a proportional change in a dependent variable. T he exponential growth function has been applied in subjects like p hysics, f inanc e and Biology among others More relevant to this study is th e application in the study of human population (ecology) whereby the number of births and deaths per person per year can be used to estimate the size of a given population at some given po int in time. In this regard, a very popular version of the exponen tial growth function is the Malthusian growth model also known as the s imple exponential growth model For example, the formula below is used to compute the population growth rate: 5 1 Where, ln = natural log, P = population, t = time (in years). To estimate the population at current time (t), therefore, the following formula is usually employed: P ( t ) = P 0 e rt 5 2 Where P 0 = Initial Populatio n, r = growth rate, t = time (in years). When r is positive decreasing population and when r is zero population r ema ins the same. Note that in Equation 5 1 from Equation 5 1).

PAGE 221

221 The exponential function can be linked to the Law of Allometric growth to attempt to capture the impact one system has on another system or the impact of the change in one part of a system on the whole system. This linkage is necessary for the conceptual approach taken in this dissertation (Chapter 1) in which the urban economic system is some formulations of the Allometric Law or Allometric growth principles. The Allometric law (or Allometric growth princip le) The Allometric Law is said to have been originally discovered by the biolog ist John Huxley (Savageu, 1979). I n its simplest form, it states that the rate of relative growth of an organ is a constant fraction of the rate of relative growth of the total organism (Savageu, 1979; Nordbeck, 1965). A common allometric equation may be represented as: 5 3 organism. This formul ation of the law of allometric growth speaks to the notion of the growth of a component of a system impacting the growth of the whole system ( or organism ) ( Coffey, 1981 ; Savageu, 1979; Nordbeck, 1965 ). Of particular importance to geographical research is that as far back as in 1969, David Harvey observed that: t his allometric law has recently been introduced into geography by Nordbeck (1965), Woldenberg and Berry (1967) and several others, to be characteristic of both human and physical geographical syste ms. (Harvey 1969 : 466). He further explains that

PAGE 222

222 If we take measurements on some parts of a system and compare them with measures taken on other parts of the system (or over the whole system), we frequently find a basic mathematical relationship holds go od, namely the ALLOMETRIC LAW, which has the form: Y = aX b (Harvey, 1969: 465). Essentially, therefore, this dissertation research share s in the belief of such luminaries in the geographic discipline regarding the potential usefulness of the allometric la w for analyzing geographical systems, he nce the attempt to adopt the postulates of the allometric law as the underlying principles for estimating quantity of land consumed by a city at any point in time, driven by economic activities. Even though the allo metric growth law is an empirical one the theory of allometric growth s uffice here to say that many researchers have attested to the fact that postulates of the allometric growth law would hold true for both biological / physical systems on the one hand a nd social and economic systems on the other hand (Harvey, 1969). Consequently, since its formalization the allometric law or principle of allometric growth has been refined and modified to study different systems, both physical and social economic, as the se systems have been identified as generally obeying the postulates of the law ( Warntz, 1975 cited in Thomas & Huggett, 1980; Nordbeck, 1965 ). In current formulations and applications, particularly, in urban geographical research which is of relevance to this study, is the relationship between the population density and distance to the core of a city ( Wa ntz, 1975, cited in Thomas & Huggett, 1980 ; Nordbeck, 1965 ). Essentially, it has been observed that population density of a city declines as the built up area increases. In other words, population density decreases with distance from the center of the city based on a negative exponential logarithmic function (Wantz, 19 75, cited in Thomas & Huggett, 1980). Furthermore, in broad terms, Coffey (1981) has argu ed that

PAGE 223

223 Allometry refers to the study of size and its consequences, and relates the differences in proportions of one component of a system to changes in either the absolute magnitude of the system or a second component of the system (p. 185) The forego ing outline of the allometric law (principle of allometric growth) shows that the allometric equation has been conceived and used as a special form of general exponential functions This conception opens the equation to numerous applications in both the n atural sciences and the social sciences and will be useful for spatio temporal modeling of urban spatial expansion, but has not been attempted yet Thus with reference to exponential functions in relation to their applications in social science, Jacard an exponential function to create functions that reflect growth or change with certain fact that any number raised to the power of 0 is equal to 1, the following equation can be used to describe exponential Jacoby, 2010: 195 197): Y = S 0 e (kx) 5. 4 Where Y = population size at a given point in time, x = the distribution in time since a predetermined start time, and S0, e (natural log) and k are constants. The value of S0 is fixed at a value equal to the population size at a predetermined start time. This formulation is particularly useful for the urban spatial expansion analytical model being proposed in this chapter of the dissertation research. The list of applications for this kind of formulation of exponential functions and variants has been noted to be long and cuts across disciplinary boundaries. For ex ample, Yuri Shestopaloff notes that

PAGE 224

224 In psychology and biology, the sensitivity of different organs to irritation is de scribed by exponential laws; Statistics, probability theory, economics, social sciences and many other areas of science and t echnology use these functions as the main components of adequate models. (Shestopaloff, 2010: 33). For geography, one only has to look a little back into history when David Harvey owth process For the purposes of C hapter 5 the formulation of the allometric law (equations) this instance, urban growth is conceptualized as the development of a single system (the built up area) over time. In other words, the focus is on the quantitative change (additions of b uilt structures, translated into land area consumed) over time or the outward shifting of the urban boundary over time. Consequently, the spatial expansion of a city (land conversion from agricultural usage to urban uses) could be considered as a growth pr financial parlance. A particularly importa nt aspect of the Allometric Law, of importanc e to the modeling exe rcise undertaken to develop the IUSEEM is the version of the allometric equation which utilizes the natural logarithm to form a logarit hmic exponential function as in Equation 5 2 This formulation has been found to be typically usef ul for the study a but also in economic ones (Coffey, 1981) implication is not necessarily limited to only the increasing aspect of phenom ena. Thus,

PAGE 225

225 similar to Coffey (1981) as in IUSEEM) and negative. Consequ ently, urban spatial expansion could be interpreted as decline in urban vegetation cover since the city is embedded in the regional biophysical system and also the fact that increasing built area (as pertains to new land conversion) amounts to decreasing regional vegetatio n cover Therefore, tracking and quantifying built up area expansion over time would reflect the exten t to which regional vegetation cover is being lost to urban development By extension if the expansion process could be linked to FDI, it can be concluded that the loss of the vegetation cover has been induced by FDI Admittedly, processes that impact u rban expansion are n umerous, encompassing physical geographic influences technological advancements social and economic processes which influence the evolution of the city, it is not too far fetched to agree with the urban geographer Michael Pacione t hat economic activities and processes hold sway in the city (Pacione, 2009). For instance, technological advancements have made it possible to overcome physical cons traints, to the extent that wet lands can be effectively drained for urban development (eff ectively removing a physical geographic barrier) whereas economic agglomerations continue to influence decision making affecting urban land use, drawing global capital to selected, preferred urban enclaves. It is not too difficult to see, therefore, that any model of urban growth that misses out on the linkage to urban economic theory may be te lling only half of the story, litera lly speaking. This provides one main rationale for the urban expansion model advocated in this study and in this chapter in par ticular. In addition to being computationally less intensive, it is simple and analytically tractable as well as theoretically grounded. Such a model should be useful and handy to urban expansion

PAGE 226

226 modelers in developing countries, especially, who are ofte n challenged by unavailability of resources, including new technologies needed for simulation Urban expansion researchers; and should also be informed by urban economic theory, hence the analytical model being proposed in this chapter. Many researchers have observed that the ease of analytical models and their relatively less demanding computational and data requirements make them easily accessible to researchers in dev eloping countries and are being utilized to study current urb an expansion (Angel et al., 2011 ; 2005; Seto, 2005). The attraction of analytical models stem from some advantages over other modeling approaches such as simulation including cellular automata a nd agent based modeling. Importan tly, even though the se complexity inspired spatially explicit urban models take a systems perspective of the urban expansion process, it leads to an incomplete conceptualization; theorization and understanding since global economic influences are largely ignored and do not usually link to economic theory As a result, the linkage among the urban biophysical environment, the urban economy and the broader global economy (economic globalization influences) are neither explici tly articulated nor represented, potentially leading to a less comprehensive understanding. In particular, such models are likely to underestimate the increasing influences of external economic forces on the urban expansion process, which are assuming unp receden ted relevance with intensifying economic globalization. Inspired by the allometric principles, the model proposed in this chapter, IUSEEM, allows for incorporation of all kinds of explanatory variables, provided they are borne out by economic theo r y, (following from the Thunen Alonso theoretical framework, for

PAGE 227

227 example); regression estimation results are found to generally fit the data and significant at the 95% confidence level. Such a method is currently lacking hence the relevance of the Integrat ed Urban Spatial Expansion Estimation Method (I USEEM) proposed in C hapter 5 The IUSEEM : General Considerations Basically, IUSEEM is a modified exponential growth function, utilizing the natural ribed above and premised on ideas underlying the Allometric Law IUSEEM attempts to reflect the transmission of country in which the expansion process being modeled is located) translates this into change in the built up area over time. Further, IUSEEM invokes the Allometric law (or principle), which can be formulated as a form of exponential function. The rationale for the lin k with the Allometric law is that urban land conv ersion is deemed as essentially a compounding process, which .rarely declines in absolute size. In other words, urban built up area expansion is more o Essentially, the IUSEEM i s informed by the allometric principle (or law ) whereby the urban built up area expansion is conceptualized as an allometric growth process, by urban extent expansion as opposed to the description of the land use patterns in the urban area. Urban size and its change over time have several implications, including the costs of commuting, emission of carbon dioxide, loss of sensitive ecological lands, among other things (Seto et al. 2011; Wackernagel 2000; Costanza e t al., 1997 ). The IUSEEM implements the estimated co efficient of FDI in the formulated allometric function as the exponent, to estimate the total b uilt up area in any given year as well as

PAGE 228

228 the cumulated built up area up to a given point in time. The conc eptualization model ing and methodological approach presented in this chapter could be very useful to urban geographers in the area of theory development to help explain contemporary urban spatio temporal modeling and analysis as well as suitable method f or estimating FDI induced urban spatial expansion. It is important to reiterate that the IUSEEM is suitable expansion process In other words, the focus of IUSEEM is on estim ating the total area of peri urban land likely consumed, resulting from increasing FDI inflows into urban areas It is not meant to observe the behavior of individual parcels or particular cells, per se. It is also not meant to desc ribe spatial patterns. These are useful intellectual pursuits, no doubt, but spatio temporal estimation methods for urban expansion currently in existence fail to link to standard economic theory and poorly incorporate M ore formally the IUSEEM methodological approach advanced in this chapter invokes the principles of allometric growth to integrate the results of log transformed regression model in a simple allometric (exponential) equation which can be used to estimate t he total built up area of a city at any point in time, given the total area at some organism whose behavior could be reasonably approximated by th e postulates of the Allometri c L (Coffey, 1981). The importance of the allometric growth law to the analysis and understanding of geographical systems chan ge over time has long been noted or alluded to in the geograph ical research literature ( Coffey, 19 81; Warntz, 1975 cited in

PAGE 229

229 Thomas & Huggett, 1980 ; Harvey, 1969; Nordbeck, 1965 ). Richard Harvey in particular, gone to the extent of appearing to encourage geographical researchers to make product ive use of the allomet ric law in asserting tha t The allometric law promises to become an integral part of geographic theory, since it can be used to account for such phenomena as population density gradients the rank size rule, and so on (Harvey, 1969: 466). He further notes that Nordbeck thus shows that the area of a city is related to its total population by the allometric law and this suggests that a process of growth over time can be connected mathematically to size in space (Harvey, 1969: 466). To build on this, this dis sertation research advances a further perspective to the effect approaches (statistical modeling such as regression analysis counterparts such as t he exponential growth models ( function ) which will be useful for studying human induced (socioeconomic driven) changes in geographical / environmental systems over time such as urban land conversion If the contributions of socioeconomic variables to chan ge (s) in a physical system can be estimated with reasonable accuracy deriving the rate of change in the physical system with respect to the socioeconomic variables it would be reasonable to expect that one can use this rationale for the spatial expansion estimation method outlined in this chapter of the dissertation. Essentially, the city is viewed as a Spatial Socio Economic Unit SSEU (following Frank et al., 2001) changing through time and space, and the change has co 1 ), and whose expansion may be circular or semi circular depending on presence or absence of physical constraints.

PAGE 230

230 the reasoning is that initial bu ilt up areas are rarely removed even in the face of decline of the urban economies or populations (Burchell et al., 2005) which is akin to interest rate compounding from the perspective of money and banking In effect, urban extent expansion is character istically additive In developing the IUSEEM, a ttention is focused on the ever shifting boundary of the city viewing urban spatial expansion as essentially a moving boundary and modeling urban expansion as a continu ous motion of the boundary as opposed to the snapshot approaches in many land use land cover change modeling in general and urban land cover change in particular. processes) of the phenomenon through time. Usually, such studies implemented by detecting the change in the phenomenon of interest (say land cover) through the use of a few images selected for two or more years. The change is then associated with the Consequently behavior of the variables between time periods ar e largely unknown; implying the all explanations i n geography need not employ genetic explanation (temporal) the explanation of growth, as a matter of necessity, will be better served if genetic explanation is empl oyed. Urban land conversion over time could be viewed a s a cardinal aspect of the spatio tem poral evolution of the city, as peri urban natural areas up areas. More specifically, urban growth could approximated by the postulates of the Allometric law. One rationale for focusing attention on the shifting city boundary is that it affects the size of the built area which in

PAGE 231

231 turn has implications for other variables of interest such as transportation systems, infrastructure, urban ecosy stems, urban hydrology and climate, among other things. The importance of the above cannot be overemphasized. From ecological perspective urban land conversion or spatial expansion greatly ecological health due to its encroachment on natural areas with the potential to undermining sustainable urban development efforts in the long run. Beyond the practical needs of sustainable urban development, it is often desirable in geographical analysis to better understand the evolution of some spatial unit such a s a city over time ( Coffey 1981; Harvey 1969 ). The relevance of the IUSEEM is placed firmly in these contexts. D eveloping the IUSEEM : The main steps Step1 : The first step in developing the IUSEEM is the formulation of a multiple regr ession with all variables log transformed using the common logarithm (log to base 10). As much as possible one must ensure the use of economic variables (since these are likely to possess multiplier properties which would resemble the behavior of exponent ial growth, non linearity, among other things. More importantly, these variables must be borne out by economic theory. For the purposes of implementation in the IUSEEM, the explanatory variables must result from the theoretical propositions of the classi cal economic theory of urban spati al structure (Angel et al., 2011 ; 2005 ; McGrath 2005; Brueckner, 1987; Brueckn er & Fansler 1983; Muth, 1969; Mills 1969; Alo nso, 1964) with particular reference to their current extensions deriving the urban extent as a function of re levant socioeconomic variables population, income, among others

PAGE 232

232 Step 2 : The second step entails the regression estimation and diagnostics. The regression model so formulated (log transformed) can be estimated using the standard Ordinary L east Square (OLS) regression method. The estimated model should be subjected to the usual regression diagnostics particularly focusing on overall fitness of the model, multicollinearity, among other things. The co efficients are also examined for statist ical significance which must be, at least, at the 95% confidence level for all estimated co efficients. Also, for the purposes of this chapter, only the co efficients which have positive values (contributes to urban built area expansion) are considered to be relevant. Step 3 : The third step involves the extraction of the estimated regression coefficients that have positive values. A caution is in order here. If one is interested in the totality of the impact of the socioeconomic drivers modeled in the re gression, then one should sum up all the estimated co Alternatively, one can focus on individual explanatory variables, one at a time. In all cases the value obtained must be translated into a decimal by dividin g by hundred (100), since the estimated co efficients of the log transformed regression are percentages. Another important issue to be noted is that even though the regression itted from the analysis. This is not to suggest that urban growth would stall in the absence of the socioeconomic variables. It is simply that the issue gets complicated beyond the scope of this dissertation research. Step 4 : The final step is the implemen tation of the regression estimation results (extracted co efficients) in the allometric based exponential function which will be used

PAGE 233

233 to estimate the built up area of the city at any point in time, given the current size of the build area. The I USEEM has t he natural log (e) as the base of the exponent whereby e = 2.718. It is widely believed that the natural log (with value of 2.718 is most suitable for growth processes, with wide applications across both the natura l and the social sciences Figure 5 2 ill ustrates More formally, the general mathematical formulation of the stages outlined above is summarized below: Given an estimated log transformed (OLS) cross sectional regression, with urban built up area s of a representative sample of cities: t 0 1t 1 X 1t 1 2 X 2t 1 +. n X nt 1 5 5 t = the estimated mean of the response vari able, built up area of the city; 1 n estimated co efficients of log transformed multiple regression model; X 1 n = explanatory variables (pr eferably related to economic activities, or socio economic variables, likely to have multiplier effects in an economy). The explanatory variables are lagged one step behind the response variable to reflect the notion that change in the explanatory variabl es may not necessarily reflect instantaneously in the re sponse variable. In other words, economic processes take some time to effect noticeable chan ge in some other system (s). The IUSEEM is formally stated as: B = B 0 e c (1+f) ; B 0 > 0, c > 0 5 6 B = the current urban built up area, e = the natural log, with the numerical value, 2.718, c = the estimated FDI co efficient from Chapter 4 of this dissertation. Note, however,

PAGE 234

234 efficients as well, i f the modeler is interested in more than one explanatory variable. For the purposes of IUSEEM, based on the arguments advanced in this chapter, however, the focus is on FDI. It is important to note that the IUSEEM (E5.4) above incorporates the tem poral as pect of the urban land conversion process as well as the spatial aspect, akin to what Harvey alluded to, which because of its centrality to the goal in this dissertation research, is quoted here again to could be used to mathematically link a growth process over time to spatial change with the potential to enhancing stages: (1) the formula tion and estimation of log transformed regression (OLS) from which the FDI co efficient is extracted. The use of estimated regression coefficients implies that the urban spatial expansion process entails stochastic (random) elements, requiring provision fo r an error term in the quantification proce ss, (2) annual adjustment of the co efficient. In this case the change in the volume of FDI inflows, relative to the volume at the beginning period (1999 in this case), is computed and used to adjust the FDI regre ssion co efficie nt estimated for the base year (1999). T his allows for the efficient in order to reflect the current flow levels of FDI. In IUSEEM, therefore, c (1+ f ) = adjusted FDI induced change in urban built up area. This is the p arameter t hat controls the extent of FDI impact on the urban spatial expansion process. Even though IUSEEM and the classical population growth function (Malthusian population growth model) are similar in many respects, there are many differences between t he two.

PAGE 235

235 IUSEEM V s. the Population Growth M odel : Similarities and Differences. The Similarities are : (a) use of natural l og to model growth processes in both models; (b) both have the exponential function with (c) both descr ibe a nomenon Thus, similar to human population, even though the growth rate may decrease, the absolute size of the phenomenon of interest is likely to continue to increase, albeit at a slower pace. For example, even though world populat ion growth rate may decline, still some numbers will be added in absolute terms. Similarly, for IUSEEM, even though FDI flow volume (rate of inflow per annum, hence relative change) may decline, net additions t o built up area may be positive; and (d) Both have fixed, initial value that is a given. The Differences are : what makes the model dynamic is the annual relative change in net FDI inflows, derived fr om the measurement of the FDI; (b) the rat e of change (estimated regression co efficient) updated on annual basis usin g actual FDI inflows.in IUSEEM; (c) estimates built up area, using the rate of change a s induced by a second variable. It has been noted in the formula (5. 1 ) above, that PGR = the or the population growth rate. In IUSEEM, the regression c o efficient is simila r to PGR the rate of change. However, in the c ase of the population growth model, the rate of change is not derived with respect to a seco nd, separate variable. Thus, different elements of population birth rate, death rate, net migration are used to compute the size of population at any given time. It is important to note that population growth model does not reference any explanatory variable outside population itself Consequently, and in essence, the formula explains change in quantum of population based on internal

PAGE 236

236 dynamics of population itself growth per unit of time. On the contrary, IUSEEM does explicitly refer to a second, explanatory variable in the form of FDI. Consequently, the explanation of expanding built up area is attributed to inflow of FDI from the economy as a subsystem, flowing int o the urban biophysical system ; (d) G rounded in standard urban economic theory, refined to reflect reali ties of economic globalization; (e ) IUSEEM exponent partly results from regression estimation, allowing for the stochastic nature of social and econ o mic variables such as FDI; (f ) The IUSEEM method embeds e lasticity such as income elasticity of demand or cross elasticity of demand quantity demanded of one commodity changes in response to change in the price of another (second) commodity. In the case response to percentage change in annual FDI inflows into the urban economy With respect commodity, change is induced by a factor exogenous to the commodity whose (price of another commodity) Similarly, in IUSEEM the built up area expansion is quantified based on change in FDI, which in this case has origi nated from sources outside the city (and outside the country where the city is located) whose physical expansi on process is being quantified; (g ) IUSEEM is also premised on the Allometric principle to exemplify th e fact that FDI originates from the global economic system, a subsystem of the global ecosystem, into the urban economy (which is also a subsystem of th e regional biophysical system). As FDI inflows influence the urban economy, via real estate, housing and other land consuming activities, the regi onal biophysical system is being changed accordingly (through vegetation cover depletion) In eff ect, theref ore, the expanding built up are a

PAGE 237

237 process, by virtue of the fact that conversion of land to urban uses almost invariable entails complete removal of the vegetation. In short, IUSEEM quantifies the FDI induced change in urban vegetation cover, via a proxy In other words, IUSEEM results can be interpreted as environmental impact through loss of vegetation cover, indirectly quantified by estimating the expanding urban extent. It is particularly important to emphasize found in exponential functions The reasoning behind this is that in the case of IUSEEM, accumulate in a given urban space and it takes time to manifest in increase of urban extent time is not t he key element attracting the FDI and as a result may not be appropriate to implement directly in the exponenti al function. Nonetheless, IUSEEM time into consideration by using annual FDI flow figures and computing the relative change in FDI, for the flows are reflected and explicitly modeled which is then used to adjust the coefficient estimated by r egression It is expected that thi s estimation method (IUSEEM) potentially will be capable of s urban spaces. For an individual city, the IUSEEM will be useful for des cribing and explaining the change in the spatial extent of the city as a dynamic geographical system; with the potential to contribute to a noble objective in geographical research. A

PAGE 238

238 preliminary operationalization of IUSEEM using data for Accra, the rapi dly expanding and fast globalizing capital city of Ghana, has been attempted below. Methodology Overview expansion? As has been discussed elsewhere in this dissertation, citing a uthoritative 2002, with a particularly steep growth rate beginning from 1991 (Table 5 1) The full physical extent of Accra is generally defined to encompass the Greater Accra Met ropolitan Ar ea (GAMA) as shown in (Figure 1 3 in Chapter 1). Estimations for the urban extent of Accra and change vary somewhat. However, IUSEEM will draw on two main authorita tive sources (Angel et al., 2011; 2010 ; 2005; Moller Jensen et al. 2005). In 1 985 the urbanized areas of Accra (the GAMA) constituted 216 sq. km, increasing to 276sq.km in 1991 and to 555 sq. km, in 2002 (Moeller Jensen et al., 2005). According to Moeller Jensen et al. (2005), this growth has largely occurred on the fringe lands of the city w hich can be seen from Figure 5 4 According to t raditional urban expansion analysis such as the stud ies by Angel et al. (2011; 2005) and Seto (2005) would normally attribute such rapid increase in urban extent to p opulation and income growth. However, there is a discrepancy between income growth rate ical expansion ( Table 5 7 ). Hence, it seems reasonable to question the extent to which the traditional urban land expansion drivers population and income growth contributed to the process and to explore the possibility of some other driver (s) being responsible, at least in part, for the rapid physical expansion of Accra.

PAGE 239

239 ex pansion rate coincides with the implementation of economic liberalization policies (SAPs) in Ghana, which facilitated inflow of foreign capital (FDI) into the country. A quick review of historical data suggest that the FDI flow rate (annual change in FDI flow volume ) and rapidity of change (Table 5 7) closely resemble the characteristic of rate of physical expansion of Accra during the period under study Furthermore, i t is documented that Accra absorbs about 80% of annual FDI inflows to Ghana (GIPC, vario us years) Consequently, i ndication s are that this may be resulting from unequal given the urban population of Accra compared to all urban areas in Ghana combined (Chapter 1). A recap : Based on analysis in Chapter 2 and Chapter 4 of this dissertation, FDI appears to be a potent driver of numerous economic activitie s in receiving urban locations; increasing demand for land via housing demand and demand for land for FDI related activities, which if not mitigated by urban growth management policies and tools could result in conversion of new lands to urban uses, almost inevitably at the peri urban areas where urban land is comparatively cheaper c ompared to inner locations. It is therefore reasonable might have been induced in part by FDI inflows to Ghana. This is what the IUSEEM will seek to quantify. The claim is that the Rapidity of physical expansion of Accra in contemporary times espe cially resonates with the rapidity of change in volume of FDI inflows to Ghana during the same period The evidence adduced so far can be summed up simply that : Accra is sprawling and is characterized by high concentration of FDI funds, compared to other u rban areas in Ghana. The question then is : how much of the sprawled development (rapidly increasing built up area) of Accra may be attributed to

PAGE 240

240 FDI inflows to Ghana and its disproportionately large concentrations in Accra? In order to ascertain the exten t of FDI involvement, therefore, it would be necessary to be able to estimate the built up area of the city as induced by FDI inflow over time. Such a quantification exercise, at the minimum, may help create awareness, drawing the attention of urban planne rs to FDI as a potential driver of urban land conversion as well as contribute to the conversation on the all important topic of human induced global environmental change, especially in the area of urban environmental change As argued above, urban expan sion modeling have taken various approaches but in this chapter a new model / method (the IUSEEM) is suggested which is deemed more appropriate for Accra and like cities which are fast globalizing and being characterized by sprawl but situated in countrie s generally characterized by generally sluggish population and economic growth Research Design A four phase appr built up area expansion to the growth rate in population of Accra, GDP p er capita growth rate of Ghana and annual change in the volume of FDI (FDI g rowth rate), is formulated ( Equation 5.7 ). Population and GDP/GNP per capita (income) have been consistently identified as the most relevant urban land expansion drivers (Angel et al., 2011; 2005; Seto, 2005; McGrath, 2005 ). FDI is included based on the argument of this dissertation about the increasing importance of FDI as a potential urban land expansion driver. However, for the case of Accra, this study has argued that the tradit ional explanatory variables population and income may not be as relevant as it appears in model estimation s Consequently, the theoretical model is redu ced to a more relevant one ( Equation 5 8 ) which specifically links built up area expansion and FDI, to be quantified

PAGE 241

241 by IUSEEM. Secondly, to zero in on the most relevant land expansion driver in Accra potential variables are considered and eliminated or kept based on how closely their growth rates have come close to the rate of physical expansion of Accr a. S imple calculations have be en employed to compare the population growth rate of Ghana and income growth (GDP per capita) period 1985 2002, to highlight the discrepancy mentioned above This is to butt ress the point made above to the effect that population and income might not have significantly 2002. This paves the way for sical expansion in recent times. Subsequently, annu al FDI inflows to Ghana will be compared to the physical growth rate of Accra during 1985 2002 (Table 5 7) Thus, based on the simple calculations, it is suggested that since population growth has not bee n as dramatic and GDP growth has been sluggish, it would be worthwhile to focus on FDI, given the discussion and arguments advanced so far in this dissertation. FDI induced urban spatial expansion seems to be in order at this point. Thirdly, therefore, the IUSEEM is implemented to quantify the incremental growth in the built up area of Accra from 2001 2011. The IUSEEM results are aimed at testing the initial hypotheses developed above (end of the literature review) specifically emphasizing the rate of grow th in the urban extent Finally, a comparative analysis is attempted, pitching the ducted by Moller Jensen et al. (2005). It is important to note that the data used in IUSEEM were obtained fro m Angel et al. (2011 ) and are therefore independent of t he data in Moller Jensen et

PAGE 242

242 al. (2005). Comparing IUSEEM results to data from an independent source is necessary in order to assess the extent of FDI impact on the land expansion process Data The IU SEEM is operationalized using processed remotely sensed land cover change data sourced from reliable, published d atabase ( Angel e t al., 2011 ). Furthermore, the city of Accra is captured in the sample of 120 cities used in the regression model s in Chapter 4 in this dissertation. Hence, the use of this data for preliminary operationalization is consistent. Note that all that is required for IUSEEM computation are (a) built up area of the city at the beginning period (b) the relevant estimated regression coe fficient and (c) the annual FDI flow figures. Method The estimation focuses on quantifying the cumulative build up area and the rate of physical expansion of Accra through time as induced by increasing levels of FDI inflows to Ghana, with an estimated 80 percent concentration in urban Accra Most urban land expansion analysts /researchers would agree that population and income growth are the two most importa nt urban land expansion drivers even though the relative importance of the variables differ dep ending on the economic development level of the countries being studied. (Seto et al., 2011). However, as has been argued in this dissertation, FDI appears to be an emerging potent driver of urban land conversion ysical expansion in recent times. Thus, the relationship between the expandi ng built up area of Accra on the one hand and population, income and FDI may be expressed as in Equation 5.7: B gr = f (P gr Y gr FDI gr ) 5 .7

PAGE 243

243 Where, B gr = Rate of growth in the built up area of Accra (annual average growth), P gr = Annual population growth rate of Accra, Y gr = Annual rate of growth in income (measured by GDP per capita), FDI gr = annual growth rate of FDI inflows to Ghana Essentially, the Equation 5.7 posits that r ate of growth of urban land in Accra is a function of population growth rate, income growth rate and FDI growth rate, respectively other things being equal. In other words, i t is postulated that increases in to Ghana will accelerate the growth in the built up area of Accra. However, it has been argued in this dissertation, and docume nted evidence from published research to support the argument, to the e ffect that in the case of Accra, population and incomes physical growth of Accra. Therefore, for the purposes of the arguments in this chapter attention will be foc used on the built area FDI relationship Consequently Equation 5 7 may be reduced to: B gr = f ( FDI gr ) 5.8 Equation 5.8 posits that the growth rate of the built up area is a function of the rate of in FDI (annual fluctuations in the v olume of flow). This is not to suggest, however, that population and income are not relevant u rban land conversion drivers in given that in real terms their rates o f growth did not experience increases during the period Accra expanded rapidly Hence, in the IUSEEM estimation procedure the estimated co efficient of population and income will be adjusted for incorporation in the

PAGE 244

244 IUSEEM. Consequently, the IUSEEM formul ated above and restated below is applied from this perspective 0 e B 0 > 0, c > 0 5 .9 = e stimated bui lt up area for a current period, 0 = e xisting urban land cover at beginning of estimation period (same as year of urban land cover used for regression modeling and estimation) e = natural log ( 2.718 for the IUSEEM computation purposes), justed estimated regression = relative change in annual FDI inflows to Ghana, with reference to the base year used for the regression estimation (1999 for the p resent discussion). The term (1+ Using this term ensures that the exponent of annual FDI flows). It is also important to note that is the parameter controlling the FDI induced urban bui lt up area expansion over time. Essentially, therefore, t he IUSEEM as formulated above posits that urban built up area expansion in Accra as induced by FDI inflows follow ed an exponential growth path. It states that since the urban e, estimate of the current built up area takes into account the built up area val ue in the start of the period, since already built up area agglomeration and infrastructural enhancements would serve to attract more economic flows, increased land demand and increasing expansion. This is analogous to a single

PAGE 245

245 organism; say an animal, growing in size as a function of food intake, for example. In terms of the IUSEEM, the urban economy of Accra fed by increasingly high FDI concentration level s, will manifest in the acceleration of the physical expansion of the city over time. This is the rationale for linking IUSEEM to the Allometric Law (or the allometric growth principles). (Of course, this is simplistic but it is the esse nce of the whole enterprise of modeling, more or less). The exponent or the parameter u sed in the IUSEEM, is composed of two parts: the estimated FDI coefficient from regression model and the calculated annual relative change in FDI inflows I used the results for FDI in the regression estimation presented in Chapter 4 as input to Equation 5.9 (also shown in Table 5 2) Again, I used the estimated FDI co efficient as part of (the expone nt in the IUSEEM) because the interpretation of the log transformed regression posits elasticity, implying rate of change i n built up area with respect to FDI (pe rcentage change in built up area induced by percentage change in annual FDI in flow s ). I reasoned that I could use this characteristic to formulate the exponential function, since the "exponent" in regular exponential functions (such as the population growt h model outlined above) usuall y denotes "the rate of change". Consequently, to reiterate, the IUSEEM works by 'translating' change in FDI inflows into change in urban built up area, taking into consideration the compounding nature of urban built up area e xpansion. The estimated co effi cients of regression Models in Chapter 4 are inser ted in Equation 5 10 Built = 0 .959 + 0.787 ( POP ) +0 .215 ( GDP ) +0.087 ( FDI ) 0.134 ( REMIT ) 5.10

PAGE 246

246 All variables remain as defined in Chapter 4. The constant and coefficient for remittances are negatively signed, contrary to the theoretical postulates advanced in Chapter 4, needing further investigation. However, this has not been under taken in this dissertation so these coefficients have been eliminated and not included in the I USEEM implementation. In IUSEEM, the estimated coefficients of the cross sectional, static, linear regression model have been used to create an urban spatial expansion model (IUSEEM) which incorporates temporality, dynamism and non linearity in a simple an d analytically tractable manner. Also important is the fact that IUSEEM is linked to postulates of standard urban economic theory since the regression model whose coefficie nts have been used as exponent is firmly grounded in the Thunen Alonso theoretical framework and extensions (Chapter 4). For a better understanding of the workings of the IUSEEM, it is important to make the following clarification with regard to quantifying change in systems (variables of interest). Three main approaches may be envisaged Firstly, absolute change in a variable can be measured or quantified by observing the variable from two successive stages and documenting the change that has occurred. For example, basic urban land cover change data derived using remote sensing methodolo gy falls into this category. In of change in some element of the variable of inte rest or in a second variable. Secondly, change in a variable can be measured with reference to rate of change in some example, the computation of the size of a populat ion using Equation 5 2 utilizes the

PAGE 247

247 population growth rate (based on birth rate, death rate and migration as in Equation 5 1 to estimate the size of the population (absolute) at a given time. Finally, the change in a main variable can be quantified with re ference to the rate of change in some other variable. For example, in regression analysis, prediction in a response variable is sought through the implementation of a single or set of inde pendent explanatory variables. A typical example in this work is im plemented in Chapter 4 to quantify the relationship between urban land cover and a set of explanatory variables cross sectionally In this case the change in the response variable with respect to change in the predictive / explanatory vari able is measur ed, which may be interpreted as quantifying change in a system of interest (response variable: urban built up area in this case) with respect to change in another variable (the explanatory variable) deriving the rate of change The IUSEEM builds on this latter category, attempting to quantify the change in urban built up area over time based on an established rate of change involving the built up area and a second variable ( FDI in this case ) obtained from regression estimation (Chapter 4). In short res ults of regression analysis is integrated into exponential growth function (and drawing on allometric growth principles) to estimate change in urban built up area over time. In some sense, therefore, the IUSEEM takes ordinary cross section regression esti mation results to the next level, adding the temporal dimension without having to deal with problems associated with time series regression analysis and also allowing for the appropriate modeling of rapid physical expansion in a primate city characterized b y nonlinear growth In the case of Accra, the IUSEEM seems the appropriate method based on the following premise. 2002 has been especially rapid,

PAGE 248

248 involving the conversion of large agricultural and natural a reas to urban uses, characterized by declining density (sprawl). For example, Angel et al (2011) note that 2000 but its physical size increased over 150%. Furthermore, two distinct phases of the physical expansion of Accra have been identified with regard to the change in annual growth rate of urban land conversion (Moller Jensen et al., 2005). For the period 1985 1991, it has been urban land cover increased from 216sq.km to 276, a veraging an increase of 10sq. per annu m (Moller Jensen et al., 2005). However, for 1991 2002, the physical size of Accra increased from 276sq. km to 555sq. km, with an average increase of 25sq. km per annum. Clearly, this is quite a rapid chan ge within a v ery short As noted elsewhere in this dissertation, s uch rapid urban expansion, theoretically, would be attributed to population and income growth in the country where the city is locat ed. Howeve r, in the case of Ghana, it is important to emphasize that none of the traditional urban land expansion drivers population and income changed so dramatically around the same period ( Table 5 7 ). Incidentally, however, during this same period FDI inflows into Ghana appear to be the most highly relevant economic variable that experienced similar dramatic growth ( Table 5 7) periodic averages) as the rate of physical expansion of Accra E ven though levels may fall in some years, increases in FDI f lows can be dramatic within a relatively short period of time. For example, the average annual inflow of FDI to Ghana for the period 1986 1990 amounted to US$8,760,000. The average annual FDI inflows for the periods 1992 1996 and 1997 2001 were US$121,400 ,000 and US$149,620,000, representing about 1054% and

PAGE 249

249 1608%, respectively; above the average annual FDI inflow figure. From the foregoing, therefore, the rapidity of change in FDI flow levels in Ghana around the same period Accra underwent an unprecedented spatial expansion is obvious. In recent times the situation has not been much different. Generally, since 2006 FDI inflows to Ghana have generally been on the asce ndance (UN, 2012; GIPC, various years). Thus, FDI inflows to Ghana during this period app ear to exhibit rapidity of growth rate similar to the physical growth rate of the capital city of Ghana, Accra, where the bulk of FDI has been concentrated. Are the two related in any significant way? How best can the relationship be modeled to incorporate the temporal aspect as well as reflect the fluctuations in FDI flows? What are the implications of FDI concentration levels for the rate of physical physical expansion be descr ibed as FDI induced? These are the main underlying motivations for the formulation of the IUSEEM which will attempt to provide answers to these questions, in addition to the potential importance for economic/urban geographical theory construction as well a s urban policy. It is important to note, from Chapter 4 that FDI has multiple pathways through which it can induce urban land conversion. Again FDI is likely to be associated with multiplier effects. Thus, with the dramatic change in urban growth coincid ing with equally dramatic increase in FDI inflow to Ghana, with Accra attracting disproportionate amounts, both can be said to have has been established in Chapter 4, with high statistical significance. These considerations, among others, informed the choice of the functional form for the IUSEEM.

PAGE 250

250 contemporary physical expansion bears some resem blance to exponential growth process, which has more to do with inflow of FDI than population and income growth. In other words, FDI is playing a significant role in urban land conversion in Accra but has not been adequately recognized as of now, hence th e need to conceptualize, theorize, and quantify FDI induced urban expansion under intensifying economic globalization. land use theory can be linked to an integrated allometric based exponential function and simple regression to build a non linear, stochastic, and dynamic estimation method The use of the exponential function for estimation of some quantities has been noted in the literature ( Jaccard & Jacob y, 2010 ) However, the approach advanced in this chapter is more in depth integrating regression analysis with exponential function and allometric principles ; and a deliberate attempt has been made to link the method to the Thunen Alonso theoretical framework Specifically, t he IUSEEM uses the co efficient of FDI estimated from a cros s section regression which together with the relative change in annual FDI inflows into Ghana are implemented as the exponent of the allometric based exponential function ( Equation 5.8) U allometric analogy is well suited to the built u p area expansion process since this is The use of IUSEEM, an allometric based, integrated regression and exponential growth built up area estimation method could be potentially useful for tracking FDI impacts on urban spat ial ex pansion in Accra, thereby detailing the loss of agricultural and rural lands to urban growth

PAGE 251

251 Table 5 2 shows relevant portions of the regression estimation results in Chapter 4. The results indicate the high significance of population as an important driv er of urban land conversion compared to income; and FDI in part icular. The results in Table 5 2 are used in the IUSEEM calculations for the built up area of Accra over time as will be explained below All procedures carried out in IUSEEM (apart from the re gression analysis) were implemented in the Mi crosoft Excel software (2010). The three main test questions are: (a) Do high volumes of FDI inflow into Ghana accelerate the growth rate of the urban extent of Accra? (b) Do varying degrees of FDI concentratio n result in varying growth rates of the urban extent? (c) Do IUSEEM results support the hypothesized exponential relationship between built up area and FDI? For question (a), if IUSEEM results show that accelerated rate of physical growth corresponded to rapid change in FDI flow volumes, it may indicate that overconcentration of FDI may be the key driver of the rapid rate of physical expansion, all things being equal In the case of question (b) if the IUSEEM results show that varying FDI concentration l evels produce differential growth rates in the built up area of Accra, it could be inferred that varying FDI concentration levels among urban locations in Ghana may help explain, in part, why Accra might have grown so rapidly compared to other urban center s in Ghana This would suggest that a more geographically diversified FDI distribution among urba n centers within Ghana might have helped reduce the rapid expansion Finally, for question (c), if IUSEEM results confirm the h ypothesized exponential growth relation ship IUSEEM may be deemed an appropriate me thod for modeling and estimating FDI induced urban spatial expansion in

PAGE 252

252 a primate city which could then be replicated to study other urban area s with similar characteristic s as Accra Results Table 5 3 displays the results of IUSEEM computation using only the estimated FDI co efficient of 0.087 resulting from the regression estimation in Chap ter 4, adjusted to 0.0696 (assuming 80% FDI concentration in urban Accra) (Table 5 2) Before the de tails of the IUSEEM results are analyzed, however, understanding of the operation of the IUSEEM may be further enhanced by explaining the underlying mechanisms. Thu s, specifically, how does t he IUSEEM work? T he extracted estimate d regress ion results in Table 5 2 are deemed to be the outcome o f regression as presented in Chapter 4 and in this chapter ( above ) According to the results, 1 percentage increase in FDI inflow to a given country (in 1999) induced a corre sponding 0.087 per centage increase in the urban land cover of that country in 2000 This inference holds for Ghana as well. In the case of Ghana, however, only the capital city, Accra, is contained in the sample of cities (Appendix A ) (Note that one country may have more t han one city in the sample used for the regression estimation). Hence, the inference refers to the urban land cover of Accra as opposed to the aggregate of all urban land cover in Ghana. Howe ver, since the FDI data were captured at the national level, it is logical to assume that the estimated FDI case). Even though not all of the FDI amounts received in Ghana are invested in Accra, reliable sources have indicated tha t the geographical distribution of FDI funds in Ghana is highly skewed in favor of the Greater Accra Region ( GIPC, 2011: various years ) in More specifically, i t has been estimated that

PAGE 253

253 on the average between 75 to 85 percent of annual FDI inflows to Ghana concentrate in economy (GIPC, 2001 2011). For the purposes of the IUSEEM computations, therefore, the estimated FDI coefficient is adjusted using an estimated FDI concentration level of 80% for Accra, im plying that about 80% of all annual FDI inflows to Ghana is concentrated in urban Accra and its environs. Thus, the adjusted estimated regression coefficient for FDI = 0.087 0.8 = 0.0696 (or 0.000696 as decimal). With this adjustm ent, it can be inferred that 1 percent age increase in the volume of FDI inflow to Ghana in a given year (1999 in this case) would indu ce 0. 0696% increase urban land cover in the subsequent year ( 2000 in this case) In other words, an increase of 10 percent in FDI inf low to Ghana corresponded to an increase of about 0.7% in t he built up area of Accra It is important to note that the urban land cover data for Accra used to estimate the regress ion in Chapter 4 were for 2000 and all FDI figures from 1999. Thus, these ye ars serve as the base years for the operationalization of the IUSEEM It is interpreted that FDI inflow to Ghana in 1999 would induce change in the built up area of Accra most probably in 2000. By the IUSEEM convent ion and as per underlying the regression models presented in Chapter 4, therefore, FDI flow in any given year can induce change in the urban land cover in the subsequent year (s) Consequently, it is not necessary to estimate the built up a rea for Accra in 2000, since FD I figures beginning from 2000 are used in IUSEEM Hence, the IUSEEM estimation period in this exercise begins in 2001, using the relative change in FDI inflows to Ghana in 2000. U sing the 1999 FDI flow to Ghana as the base year, relative changes in FDI fl ow ( f ) in subse quent years are calculated by the formula:

PAGE 254

254 = FDI y FDI 1999 / FDI 1999 5 11 FDI y = FDI flow volume in the current year ( in US$), and FDI 1999 =Base year (1999) FDI flow volume. For e xample the for 2000 is calculated as follows : = (165900000 243700000) / 243700000 = 0.319245 This means that the FDI inflow to G hana in 2000 decreased about 32 percent compared to FDI inflow in 1999 (the 1999 FDI inflow remains the common denominator for calculating throughout th e IUSEEM computation exercise ) This also means that the exponent in IUSEEM cannot remain constant (as in a regular exponential function) because FDI flow levels change every year. Consequently, for the purposes of the IUSEEM computations, the esti mated and adjusted FDI co efficient for Accra must the actual FDI inflow level in 2000 as follows: = (1+ ) = 0.000696 [1+ ( 0.319245)] = 0.000696* (0.680755) = 0.000474 It is obvious from the foregoing calculation that due t o change in FDI flow levels, the IUSEEM exponent gets reduced, from 0.000696 (1999 FDI flow level) to 0.000474 (2000 FDI flow level), a decrease of about 32%, thereby updating accordingly And this would be reflected in the urban land cover change in 2001. In this way actual FDI flow level is used to estimate the built up area in the subsequent year. Thus, relative change in FDI for 2000 is used to compute the built up area in 2001, reflecting a lag in order to buttress the point that FDI inflow w ill not be translated instantaneously into built up area expansion in the same year. So for example, in estimating the built up area in 2001, the change in FDI in 2000 relative to the 1999 FDI flow level is used to update the

PAGE 255

255 estimated regression coeffici ent as shown above to derive (the exponent). Similarly, to estimate the built up area in 201 1, the FDI flow level of 2010 is compared to the 1999 flow level to obtain the relative change in FDI flow ( ) which is then used to update the FDI coefficien t (1999) and used as the exponent to estimate the built up area for 201 1 (Table 5 3 ) a two step lag. For example, in 1999 t he adjusted coefficient was 0.000696 (changed to decimal), which decreased to 0.000474 in 2001 because the FDI volume in 1999 dec reased about 32 percent in 2000 (as indicated by the calculated relative change). Realistically, therefore, this fall in FDI flow volume should reflect in a decline in the rate of FDI induced urban spatial expansion in Accra. I n Table 5 3 relative change in 2000 is shown in italics in 2001, relative change in 2001 is shown in bold in 2002 in that order, with relative change in 2010 shown in 2011. Thus, in IUSEEM, the FDI inflows to Ghana for 2000 2010 are used to estimate the built up area of Accra (co eff icient adjusted by 80% FDI concentration in Accra) fo r the period 2001 2011 More specifically, t he built up area in Accra over time is estimated by feeding the initi al urban land cover data, 328sq.km in 2000 (Angel et al., 2011; 2010 ), into the IUSEEM (T able 5 3 column 6) which is then used to estimate the current level of FDI induced built up area (Table 5 3 column 7). This figure (also shown in column 6) becomes the input ( 0 ) for the estimation of FDI induced urban built up area in 2002 over time. Now, the temporal dimension comes in when is recalculated annually, reflecting the current amount of FDI (relative to the value of FDI in the starting period 1999 FDI level ) ; is used to

PAGE 256

256 Following from the above, the built up area of Accra in 2001 is estimated using the IUSEEM as follows: B 0 = 328 sq .km (approx.), e = 2.718, = 0.000696 = 0.319245 (from 5. 9 ), 0.319245)] = 0.000696* (0.680755) = 0.000474 Substituting in model (5.9 ) : = 328 e = 328 2.718 ( 0.000 474 ) B = 328 1.000473889 = 328.1554355 = 328.1 6 sq. km (approx.) The outcome of the calculation above is show n in column 7, row 1 in Table 5 3 In 2000 the built up area of Accra was ab out 328sq. km (Angel et al., 2011 ; 2010 ). Using this as the initial input, the IUSEEM result s show tha t the estimated built up area yields 328.24sq.km (Table 5 3 column 7 ). The rest of the entries in column 7 are computed using the procedure described above. In terms of FDI inflows to Ghana for the period 2000 2010, a total of US$ 7 627 630 000 in FDI funds were received averaging US$ 693 420 909.1 per annum. For the period 2001 2011 according to IUSEEM estim ation results, a total of abou t 7sq.km land converted to urban uses in Accra could have been induced by FDI inflow to Ghana (and a conce ntration of 80% of it in Accra); averaging 0.66 sq. km (or about 0.7sq.km) of FDI induced urban l and conversion annual ly The cumulative built up area at the end of 2011 as estimated by IUSEEM with the 80 percent FDI concentration amounted to 335.22 sq. km, from 328 sq. km in 2000 On the sur face of it this change is not quite impressive. However, the results highligh t several important

PAGE 257

257 characteri stics which seem to support the hypotheses formulated above which also appear to support the conceptual and theoretical arguments of this dissertation. Firstly, the manifestation of the fluctua tions in FDI inflows to Ghana i n the built up area expansion of Accra is obvious from Table 5 3 (column 8), which is calculated by subtracting column 6 (existing built up area at beginning of the year of estimation) from column 7 (the estimated built up area at the end of the year) In 2001 the estimated built up area was 0.16sq. km which reduced to 0.08sq. km in 2002, with a further reduction to 0.06sq. km in 2003. However, in 2004 the estimated built up area rose to 0.10 sq. km further increasing to 0.13 sq. km in 2005 and to 0.14sq. k m in 2006, with the increasing trend continuing up to 2011. Even with all positive changes, the values for the respective years differ, reflecting the dynamics (fluctuations) of FDI inflows to Ghana, from year to year. This is a very important outcome bec ause it appears the IUSEEM results then confirm the postulate that the model is capable of reflecting the fluctuations in annual FDI flows to Ghana ; increasing volumes of FDI inflow to Ghana This is made possible through the implementation of the relative c hange in FDI flow annually ( ) as part of the co which would not have reflected adequately the FDI fluctuations in the built up area expansion process. If unit of time had replaced (f) in IUSEEM, the estimated coefficient would have stead of actual FDI flow levels which are subject to fluxes (fluctuations) from year to year. Secondl y, t he exponential expansion process is al so indicated in Table 5 3 (column 8). For example, the estimated built up area for 2003 was 0.06sq.km (note: for

PAGE 258

258 one year only). In 2006, the estimated built up area was 0.14 sq. km, repres enting an increase of a bout 146 percent over the estimated built up area in 2003. This change occurring within a period of 3 years can be considered quite rapid Similarly, in 2007 the estimated area was 0.60 sq. km. In 2011, the built up area was estimated to be 2.41sq. km, rep resenting a dramatic increase of about 303% over the estimated figure for 2007. Again, this change occurred within 4 years and can be considered exponential growth if not explosive. (Note that exponential growth here refers to the annual change in urba n built up area, not the cumulative or the absolute size of the urban extent). It is important to emphasize that these results correspond to periods of dramatic increases in FDI flows to Ghana as d isplayed in column 2 of Table 5 3 It can be deduced that FDI induced rate of change in built up area follows an exponential growth rate path Thus, the IUSEEM results so far exhibit the postu lated non linearity of growth as well as the exponential nature of the FDI induced urban expansion process. IUSEEM result s su ggest, tentatively, that FDI inflow into Ghana exhibited an exponential growth in volume the refore the same exponential growth manifested in the rate of phys ical expansion of Accra which absorbs estimated 80% of all annual FDI inflows to Ghana So f ar the results from IUSEEM obtain ed from the assumed 80% FDI concentration in Accra even though not impressive, do appear to give some credence to the basic argument in this dissertation However, it would be even more insightful to examine whether or not the estimated IUSEEM results will be sensitive to varying FDI concentration levels. Specifically, how would different FDI concentration levels in Accra impact the urban built up area expansion over time? For example, what would have

PAGE 259

259 s physical expansion if the city had been receiving 10 percent of all annual FDI inflows to Ghana? Table 5 4 depicts the results of the IUSEEM estimation based on assumed varying concentrations of FDI in Accra using only the estimated FDI coefficient of 0 .087, adjusted to 0.0696 for Accra. It is shown that when FDI concentration level in Accra increased to 100%, (from the 80% estimated concentration based on the literature) IUSEEM estimates that the cumulative built up area in 2011 amounts to about 337sq. km, with the total FDI induced change in built up area amounting to about 9sq.km and averaging about 0.82 sq. km per annum for the period 2001 2011(Table 5 4 ). This means that increasing FDI concentration by 20% (from 80%) resulted in an increase of 2 sq. km in the built up area of Accra with the average at 0.82sq.km annually (compared to the average of 0.66 sq. km for 80% FDI concentration in Accra) ; about 24 percent increase in annual growt h rate has been induced by 20% increase in FDI flow volume. With FDI concentration of 10% in Accra, the corresponding cumulative built up area is estimated to be about 329 sq. km for the period 2001 2011, totaling a change of about 0.9sq.km, with an annual average of 0.08 sq. km. Clearly, IUSEEM shows sensitivity to va rying FDI concentration levels reflected in the annual rate of FDI induced spatial growth Tentatively, therefore it can be deduced that within a given city different FDI concentr ation l evels in different years would have differential impacts on the rat e of physical expansion of the city over time ; by implication, unequal geographical growth rates in the urban system

PAGE 260

260 From the foregoing, IUSEEM results do confirm t he initial hypotheses proposed. However, it is important to further interrogate whether or not the full impact and significance of FDI have been captured, using the estimated FDI coefficient alone as implemented so far. There are reasonable grounds to sugg est that the magnitudes of the estimated results might have been underestimated (Chapter 2; Chapter 4). Thus, the IUSEEM estimation results shown so far appear to underestimate the full impact of FDI on the urban spatial expansion process in Accra It may be recalled that m ultiple pathways through which FDI impact urban land conversion have been identified, conceptualized and theorized in chapters 2 and 4 of this dissertation. In particular, this study has argued that FDI inflows into a city enable people to gain or increase their purchasing power to enable them demand the kind of quality residential units such as driving the rapid rate of urban land conversion and expansion process in Accra. It is therefore suggested that the purchasing power behind popul ation numbers must be properly considered when analyzing drivers of contemporary urban spatial expansion in (Chapter 4) In eff ect, this study questions the overwhelming importance and statist ical significance of population as a driver of urban l and conversion as shown in Equation 5. 9 Similarly, the increasing concentrations of FDI in primate cities would have the tendency of increasing per capita incomes of the city resident s or, at the minim um would lead to more people earning relatively higher incomes It sounds reasonable; therefore, to assert that in the globalizing city receiving dispro portionate share of FDI inflows would lead to relatively higher income levels above per capita which may have more to do with increasing FDI inflows than the general economic growth in the country. In effect,

PAGE 261

261 for the case of Accra, contribution of GDP per capita to the spatial exp ansion process shown in Equation 5.9 might have been brough t on largely through FDI related economic activities. However, these cannot be captured by the current measurement of GDP per capita necessitating an adjustment To sum up, FDI inflows may trigger the transmission of economic impulses through the traditio nal land expansion drivers population and income resulting in an indirect FDI impact on demand for housing and land which feed into the urban expansion process, especially when the overall national economy has not experienced equally rapid rate of growth as the g rowth rate of the urban extent In view of these considerations and depending on the nature of the spatial expansion process and the regional context of the urban location, it seems plausible to reason that some FDI influences might have been tran smitted through population and incomes. Thus, for example, in the regression estimation results in Chapter 4 the coefficient for the population of the city is 0.787 implying that 1% increase in population induced about 0.8% increase in urban built up area, an almost one to one correspondence, with high statist ical significance (95% confidence level) The real situation in places like Accra is far from this whereby low income residents find themselves bulked together in high density areas such as Nima, Ch orkor and Newtown, to mention just a few. The same argument goes for the case of GDP per capita since the housing units constructed in peri urban Accra is widely acknowledged to be above the means of the ordinary Accra residents ( Aryeetey Attoh, 2010; Yebo ah, 2003; Konadu Agyemang, 2001) It seems fair to suggest that t he contribution of the low income populace to the rapidly expanding built up are a, into the peri urban areas must be very minimal, if not insignificant. For the purposes of analyzing the pe ri urban

PAGE 262

262 land expansion in Accra, therefore, it is important that both the population and GDP per capita resulting from the r egression estimation (Table 5 2 ) are adjusted to reflect th e of FDI impulses thereby attempting to captu re the indirect effects of FDI on the urban land expansion process Bas ed on the reasoning so far, FDI, in its high concentration, is deemed to be a major source of income directly or indirectly, for many residents in a rapidly globalizing city such as Ac cra. Consequently, the regression co efficients for population and GDP per capita must be adjusted to reflect the purchasing power push they may be deriving indirectly from FDI. Due to lack of an established approach to this problem as of now, tentative calculations can be made based on the estimated percentage of FDI received annually in the city under consideration. Thus, for Accra where it is estimated that about 80% of the FDI inflows to Ghana annually is received, this percentage can be applied as f ollows: (a) GDP per capita co efficient 80% ; and (b) city population co efficient *8 0% The results of this exercise may be described as the indirect FDI related purchasing power It should be noted that the use of the assumed 80% FDI concentration to a djust the population and GDP per capita coefficients is not to suggest that population and GDP per capita are not important land expansion drivers in Accra The argument is that estimated regression coefficients such as obtained in Chapter 4 of this disser tation would normally be asc ribed to population and income. s economic growth (income) and population growth do not match the rate of urban extent growth of Accra (Tabl e 5 7). Incidentally however, FDI flows into Ghana in contemporary times ap pear to be the only economic variable whose growth rate paralleled that of the physical expansion of Accra. For the purposes of the analysis in

PAGE 263

263 this chapter therefore it i s suggested that as much as 80% (following the assumed FDI concentration level in Accra as described above) of the induced built up area expansion ascribed to population and GDP per capita (as per the estimated regression coefficients in Chapter 4) can be attributed to indirect influences of FDI inflow to Ghana ( It is to be noted th at the relevant interpretation here is limited to the results of the regression model formulate d in this dissertation only. This attempt is being made to capture the possible masking of FDI influences on urban land expansion i n peri urban Accra only; and s hould not be taken out of context). Following from the foregoing discussion, the population and GDP per capita coefficients are adjusted by the estimated 80% FDI concentr ation level in Accra as follows: Population: 0.787 0.8 = 0.6296 (0.006296 as a deci mal); and GDP per capita: 0.215 0.8 = 0.1720 (or 0.001720 as a decimal). In short it is assumed that about 8 0% of the induced change in built up area of Accra in 2000 attributed to population and GDP per capita, in reality, is likely to have been brough t on by FDI related economic activities in the economy of Accra. (It would be recalled that i n Chapter 4 these co efficients were interpreted as percentages or elasticity and therefore being unit free. As a result, they can be summed up). Put in perspect ive, it is the argument in this chapter that the adjusted co efficients for population and GDP per capita must be added to the impulses transmitted through these explanatory variables. He nce, the total adjusted co efficients deemed to better reflect the full impact of FDI on built up area expansion in the ca se of Accra will be the sum of the adjusted co efficients for all three explanatory variables: 0.6296 (Population) + 0.0696 (FDI) + 0. 1720 (GDP per capita) = 0.8712 (or 0.008712 as

PAGE 264

264 o 0.008712 (as opposed to 0.000696 for FDI alone shown in Table 5 3 ). Table 5 5 displays the results of the IUSEEM estimation for the period 2001 2011, implementing the combined adjusted coefficients of population, FDI and GDP per capita. According to the results, the total change in built up area of Accra for the period as induced by FDI and related activities is 102.81 sq. km, a dramatic increas e compared to the total FDI induced c hange in built up area of 7.22 sq. km for the F DI co efficient alone (Table 5 3 ). The average built up area change per annum is about 9sq.km (about 0.7 sq. km for FDI alone) which is also a remarkable increase. Of part icular importance to the arguments of this dissertation are the values in column 8 in T able 5 5 ) which conform s to the non linearity and exponential growth postulates illustrated b y the IUSEEM results above, just as in the case where with only the adjusted coefficient of FDI The estimated built up area in 2001 is 1.95 sq. km, decreased to 1.05 sq. km in 2002 and further decreasing to 0.70 sq. km in 2003 These were the manifestation of t he fluctuations in the annual FDI flow volume Comparing the est i mated built up area in Table 5 5 and Table 5 3 reveal that adjusting for the coefficients of population and GDP per capita dramatically increases the magnitudes of the estimates. For instance, the total change in estimated built up area in the combined coeffici ents scenario is 102.81 sq. km; compared to 7.22 sq. km for the FDI coefficient alone, an increase of 1323 percent. In terms of the average (per annum) estimated values, the combined coefficients results came up to 9.35 sq. km per annum, compared to 0.66 sq km per annum when only th e adjusted FDI coefficient was implemented. This represented 1316 percent increa se. These results show that using the FDI

PAGE 265

265 coefficient alone in the IUSEEM computation may lead to gross underestimation of the potential impacts of F DI and related activities on the spatial expansion of Accra. Even though not conclusive, these results highlighting the dramati c change in the built up area expansion incorporating the adjusted coefficients of population and GDP per capita, cannot be dismi ssed given the conceptual and theoretical analysis relating FDI to urban land conversion as advanced in thi s dissertation (Chapters 2; Chapter 4 ). Thus, in addition to supporting the non linearity and exponential growth, the IUSEEM estimation results would suggest that population and GDP per capit a may be masking some of the economic impulses transmitted by FDI which are translated into rapid urban land conversion, especially in primate cities in developing countries such as Accra in Ghana. (It should be noted that the choice of 80% for the adjustment of population and GDP co efficients were based solely on the estimated 80 percent FDI concentration level for Accra. This may not be the accurate figure or the best adjustment pr ocedure. But for an explorato ry study, the outcomes seem to suggest that FDI may be a potential driver of spatial expansion in Accra ). Further insights can also be gained by computing the resultant estimated built up area of Accra for the period 2001 2011 at different FDI concentratio n levels (percentages) using the combined a djusted co efficients. Table 5 6 displays the IUSEEM results from 10% to 100%, at intervals of 10. It is shown that at 100% FDI concentration level in Accra, a total of 133sq.km of land would be converted to urba n uses during 2001 2011 (compared to 9.05 sq. km previously estimated for FDI alone). The annual average FDI induced land converted at the 100% concentration level is 12 .11 sq. km, a huge incr ease compared to the annual average of 0.82 sq. km per

PAGE 266

266 annum indu ced by the FDI co efficient alone (Table 5 4 ) On the other end of the scale, 10% FDI concentration in Ac cra would induce an average of 1.03 sq. km in built up area per annum, compared to average of 0.08 sq. km per annum estimated from the FDI coefficient alone. It is obvious that the higher the concentration levels of FDI, the more rapid the physical expansion process; highly accentuated when population and GDP coefficients were adjusted and added to the FDI coefficients This sugges ts that FDI coefficie nt alone that result from normal regression analysis is more likely to underestimate the full impact of FDI in the urban expansion process in Accra, given the high levels of FDI inflow to Ghana with equally higher concentration in Accra To sum up, IUSEEM results show close positive correlation between FDI concentration levels and rate of urban land / spatial expansion. Higher FDI concentration levels result in more rapid land conversion in Accra. When viewed from the perspective of rapidity of rate of phy sical expansion, therefore, FDI does appear to contribute to the acceleration of the urban physical growth process in Accra with the magnitude of impa cts varyin g with varying FDI concentration levels. Again, the magnitudes of estimated built up for Accra appears to be grossly underestimated if only the estimated regression coefficient of FDI is implemented in IUSEEM. Overall, with the regression adjusted R square of .820 IUSEEM can be said to have a reasonable fit for the data used so far. When viewed fr om this perspective, it may therefore not be surprising why Accra has been expanding at such a rapid rate in contemporary times, far outstripping the growth rates of the traditional urban expansion dr ivers population and income The results confirm the va rious conceptual and theoretical postulates underpinning the IUSEEM, which would seem to explain the rapid rate of built up area

PAGE 267

267 expan sion in Accra. It does appear that the results obtained from the operationalization of the IUSEEM resonate with character istics and nature of both the physical expansion of Accra and FDI inflows to Ghana during the economic liberalization era in Ghana with attendant globalization of Accra. It is important to note that the overall contribution of FDI alone to the built up are a expansion may not be very impressive, about 7 sq. km in 11 years, averaging about 0.7 sq. km ann ually during 2001 2011. However, the situation changes dramatically as the coefficients of population and GDP are adjusted and added to the coefficient for FD I seemingly buttressing the assertion that population and income when used as urban land expansi on drivers, might mask the contribution of FDI to the urban spatial expansion process at least in the case of Accra Based on the results from the built up ar ea expansion estimation using IUSEEM, therefore, it can be deduced that even though the contribution of FDI to built up area expansion in Accra may not be huge in absolute terms, it can multiply within a very short time, underscoring the need to begin to p ay attention to this external economic f orce in considering the rapidity of physical expansion in Accra and other similar cities Furthermore, seems to fit the preliminary data for Accra, with some resonance to historical precedence for the spatial expansion of Accra between 1985 2002 as documented b y Moller Jensen et al. (2005). Therefore, IUSEEM seems to be capable of modeling nonlinear and exponential urban spatial expansion process as induc ed by high levels of FDI inflows into a restricted urban space where population and income growth have been far outstripped by the rate of physical expansion of the urban area under study It appears that FDI inflow to Ghana and its characteristic tendenc y to concentrate in urban

PAGE 268

268 agglomerations in Accra and its environs has contributed in some measure to the during the era of economic liberalization and intensifying economic globa lization 1991 2011 I t is important to emphasize that FDI is more likely to engender a more rapid expansion in built up area compared to population and GDP per capita. One geographical implication of these results is that spatial inequality in the distrib ution of FDI could have significant unequal impacts on the rate of built up area expansion as induced by levels of FDI concentration. Consequently, it sounds reasonable to suggest that urban areas receiving disproportionate share of FDI inflows should exp ect to experience rapid rate of physical expansion, all things being equ al. Discussion Accr a: Highly Urbanized, Globalizing and Spatially Expanding Farvacque Victovic et al (2008) have observed, generally for Ghana that Although urbanization rates have been lower than in neighboring countries of West Africa, Ghana is following similar trends where by 2030, 65 percent of the population will be located in urban areas, averaging an annual growth rate of 3.1 percent from 2000 2010 and 2 .2 from 2020 203 0, re spectively (Farvacque Victovic et al., 2008). Such urban expansion assertions have generally tended to bother on demographic aspects, but i n this case the authors went further to state that The urbanization pattern reveals strong physical growth, which i s typified by moderate and patchy densification within the inner city core, involving the replacement of residential by commercial users, and uncontrolled and low density peripheral growth (Farvacque Victovic et al., 2008: 3). And more specifically for Ac cra: A gradual extension of the urban boundaries towards the rural space is being observed. For example, localities adjoining Accra such as Dome, Taifa, Gbawe, New Achimota, Anyaa, Santa Maria, Amanfrom, Nii Boye Town, Mallam, Kissieman, and Agboba, which were classified as rural in

PAGE 269

269 the 1984 census, have attained urban status in the 2000 census (Farvacque Victovic et al., 2008: 2). Accra rose from a cluster of small fishing villages during the colonial era to the major urban center and agglomeration hub in Ghana ( Chapter 1 ), currently accounting for of manufacturing firms in Ghana; the seat of government and largest concentration of urban population in Ghana and continues to attract in migration than any other u rban area in Ghana ( Chapter 1 ). To reiterate, i t is the rapid rate of land conversion from rural and agricultural areas to urban uses (Grant & Yankson, 2003) could pa rtly be the result of high volumes of FDI inflows into Ghana in recent times. To sustain this argument it would be necessary to establish that the rapid physical expansion of Accra has been characterized by sprawled devel opment. According to Moeller Jense n et al. (2005) has largely occurred on the fringe lands of the city (in the peri urban areas). Furthermore, Professor Ian Yeboah has described the contemporary physical expansion of Accra as being characterized by l 2003). By quality residential sprawl it is being asserted t hat the new residential developments in the urban periphery of Accra were being patronized larg ely by relatively rich people. Moreover, Professor Samuel Ary eetey Attoh lends credence to the assertion of sprawl residential development in Accra (Aryeetey Attoh, 2010) and t his is illustrated in Figure 5 5 and Figure 5 6. Figure 5 5 depicts a representation of Accra, showing some important areas and localities. In the north east corner of the map a emergence of urban sprawl in Accra, eating into peri urban natural and agricultural

PAGE 270

270 lands, judging from the distance from the c ore areas of Accra (close to the coast) such as Ja mestown. Even though Figu re 5 5 shows the emergence of sprawled development in Accra, it may give the impression that the process is restricted to the small area on the map shown. Thus, to show that urban sprawl is widespread in Accra, Professor Aryeetey Attoh again provides a more revealing picture (Figure 5 6 ). Essentially all directions, with the housing units being predomina ntly patronized by upper income people as labeled in F igure 5. 6 Indeed, this seems to be in line with the conceptual and theoretical analyses advanced in th is dissertation (Chapters 2; Chapter 4) where it has been argued that such quality housing are mo re likely to be owned by relatively rich people, who are purchasing power enabled in large measure by FDI as opposed to rural urban migrants and other less economically endowed people who flock into the city. It must be troubling; therefore, if a blanket a ttributio n is made to population growth as the main driver of urban land expansion (land conversion) in Accra, neglecting to examine the purchasing power aspect of the whole urban land demand dynamics. Is Population a Major Driver o f Rapid Spatial Expans ion i n Accra? Firstly, it is generally acknowledged that contemporary physical expansion of Accra has been driven mainly by the construction of new residential housing units mostly in the peri urban areas of the city, much of which are of high quality (Y eboah, 2003), indicating relative affluence, which by all indications cannot be attributed to the population numbers per se, without the backing of an effective purchasing power. rth con sidering. According to the study by Moller Jen sen et al. (2005), between1985 1991, the physical size of Accra grew from 216 sq. km to 276 sq. km, thus growing at an annual average

PAGE 271

271 rate of 10 sq. km. For the period 1991 2002, Accra grew from 276 sq. km to 555 sq. km, averaging 25 sq. km per annum (Moller Jensen et al ., 2005). The annual growth rate of Accra for 1991 2002 showed 150% increase over the annual growth rate recor ded for the period 1985 1991. Going by the conventional explanation of urban land expansion, population and income growth in Ghana (and Accra) around these periods should indicate similar dramatic increases in order to effect the dramatic and rapid rat e of physical growth of Accra. There appear, however, to be a wide gap between these two rates of growth as will be shown. Table 5 7 displays five year periodic averages for all the three explanatory variables used for the present analysis: FDI, population and GDP per capita For the period 1980 grew at the rate of 3.12 percent per annum, declined to an annual growth rate of 2.78 percent for the peri od 1986 1990. For the period 1992 1996, the population grew at an annual rate of 2.72 percent, declining to an annual growth rate of 2.38 percent for the period 1997 2001. Comparing the population growth rate to the physical growth rate of Accra, the annual physical growth of 10 sq. km (1985 1991) corresponded with the annual population growth rate of 2.78 pe rcent for the 1986 1990 period. The physical growth of Accr a increased dramatically to 25 sq. km for the period 1991 2002, which would correspond to annual population growth rate of 2.72 percent (1992 1996), a decline of 0.13 percent compared to growth rate for 1986 1990; and 2.38 percent (1997 2001), again, a dec line of 0.14 percent. Thus, for the period 1992 population declined at an annual rate of about 0.13 compared to the growth rate for the period 1986 1990. However, around the same period (1991 growth rate per annum was 25 sq. km, compared to the annual growth rate of 10 sq. km

PAGE 272

272 for 1985 1991. In short, sq. km. With regard to population grow th of urban Accra, speci fically, it has been noted that: 50 percent increase, while its urban land cover expanded from 13,000 to 33,000 hectares, a 153 percent increase. Urban lan d cover in Accra grew more than twice as fast as it s population. (An gel et al., 2011: 39). Clearly, built up area expansion rate outstripped population growth rate. I t appears that in the case of Accra, population numbers (growth rate) might not h ave contributed much to the rapid physical expansion of the city in contemporary times since the population growth rates both for Accra and Ghana have not matched the physical growt h of Accra within the same period Hence, increase or decrease in populatio n extent. Nonetheless, it is tempting, s uperficially to attribute such rapid physical growth to population growth because it makes sense : people need shelter; increase in population must induce increase in the number of shelter s increa sing demand for housing and land. Consequently, the statistical evidence would seem to support this reasoning. Thus, the regression model estimated in Chapter 4 and implemented in IUSEEM, i ndicate that 1% increase in population induced 0.787% increase; 10 percent increase in population corresponded to about 7.9% increase in the built up area This type of results is common in the literature (An gel et al., 2011, for example). However, it is important to look beyond these seemingly significant pieces of statistical evidence for this study area, since actual evidence based on historical data portray a different picture It is possible that the population co efficient i nfluences of some more important land expansion driver (s) ( This will be revisited

PAGE 273

273 later ) For now it would suffice to focus attention on t he other most popular traditio nal urban land expansion driver, because population does not appear to be a major driv er of land expansion in Accra Usually in rich countries, in such a situation does not match population growth affluence (growth i n income or wealth), which lead to changing lifestyles (housing tastes) are usually i nvoked to explain the rapid physi cal expansion ( Burchtel et al., 2005; Zwick & Carr, 2007). Basically, a s people become relatively rich they demand housing that requires more space in and around the building, necessitating the need for relatively cheaper lands usually located in the urban periphery As this type of demand increases, the ur ban extent expands spatially at a rapid rate ( Burchell et al., 2005) Thus, for income to explain the rapid physical expansion of Accra, (per capita) m ust have experienced growth commensurate with the rate of physical expansion as usually pertains in advanced countries period If population does not seem to be a major driver of co ntemporary urban land expansion in Accra, placing Is I ncome a Major Driver of Rapid Spatial Expansion in Accra ? Table 5 7 shows the GDP per capita for Ghana in 5 year periodic averages. For 1986 1990, the GDP per capita was US$589.17, increased to US$614.66 (1992 1996) and declined to US$557.16 (1997 2001). Compared to the GDP per capita for 1986 1990, GDP per capita increase d 0.25 percent (1992 grew at the rate of 25 sq. km per annum (1991 2002), from a previous growth rate per annum of 10 sq. km (1985 1991). Thus, domestically, Ghanaians in general did not see

PAGE 274

274 appreciable increase in i ncomes during the period that Accra rapidly expanded its physical extent. This notwithstanding, the kind of quality housing units that have been put up in the peri urban areas of Accra, which in many respects can be said to be the main contributor to the physical expansion of the city, would suggest that the owners and /or occupants are not low income earners (Aryeetey Attoh, 2010). It is also housing market and that more often individuals must save up on their own in order to own a house ( Grant & Yankson, 2003; Gough & Yankson, 2000; Konadu Agyemang, 2001 ). As has rightly been asserted b y Karen Seto and colleagues, income seems to be more relevant to the explanation of u rban land expansion in developed co untries (Seto et al., 2011). Angel e al. (2011) came to similar conclusion in their study using a sample of over 3000 cities in a regres sion analysis and found that increase in GNP per capita is associated with a 1.8 percent increase in urban l and 2011: 45). Similarly, the preliminary empirical analysis presented in Chapter 4 of this dissertation shows that 1% increase in per capita income corresponded to about .215% increase in the urb an built up area. Important as these ma y be misleading. It is important, therefore, that the source of e conomic sustenance the main source of purchasing power t o the people which e nable them to buy the lands and real property in the first place must be recognized. FDI appears to be a good candidate because even though GDP per capita for Ghana as a whole might not have increased in any might have gone u p in the wake of increasing FDI inflows (Chapter 2; Chapter 4 of this dissertation)

PAGE 275

275 How likely is FDI a Major Drive r o f Rapid Spatial Expansion in Accra ? That Accra is rapidly globalizing is now reasonably established, at least to some e xtent (Gr ant, 2009; Grant & Yankson, 2003 ). It is common knowledge among researchers studying the urban form and function of Accra that the city has been attracting increased international attention following globalization and economic liberalization in t he 1980s and 1990s to the present ( Aryeetey Attoh 2010 ; Grant 2009 ; Grant & Yankson, 2003 ; Yeboah 2003 ). In particular, the Structural Adjustment Program (SAP) the country embarked on in the 1980s has had profound impact on FDI inflows to Ghana and Acc ra (GIPC, various years), even though Accra benefited from external economic inflows even in the colonial era when it served as the main port for the export of cocoa, and enjoyed the attention of both colonial and post independence governments. FDI, as a form of spatial financial flow, has come to characterize the current era of globalization, with increasing importance, particularly to developing countries, with substantial concentrations in urban economies of recipient countries. Accra, largely by virtu compared to other cities in Ghana such as Kumasi and Sekondi Takoradi, has attracted the bulk of FDI (Grant, 2009; Grant & Yankson, 2003; Konadu Agyemang, 2000). Thus, if FDI inflows are used as a rough indicator of a liberalized economy, Ghana would be accounted as a success story among West African countries (Grant, 2009; 2007; 2001; GIPC, 2009), even though not the largest recipient. For example, according to the GIPC (2009), Ghana receive d a total of GH¢3.17 billion in FDI funds (about US$3 .5billion) in 2008. This amount is reported to be unprecedented and represented more than 160% increase over the total FDI the country received in 2007. Generally, FDI inflows to Ghana post SAP has bee n impressive to a large extent. Figure 5 8

PAGE 276

276 shows that in 1993 FDI inflows to Ghana amounted to about US$120 million, rising to about US$230 million in 1994 and falling to about US$110 million in 1995 Thus, i t must be point ed out that throughout 1993 2005 characterized by fluctuations. However, it can be seen that since 2006 FDI inflows into Ghana has been trending upwards. The upward trending of FDI is more explicitly shown in Fi gure 5. 9; between 1992 1996 the average annual FDI inflows to Ghana amounted to about US$130million, increasing to about US$150million in 1997 2001 and finally increasing to about US$230 million during 2002 2006. A very important and desirable attribute of FDI inflow is that it creates or helps create jobs, thereby contributing to the reduction of unemployment and enhancing economic growth prospects of the recipient economies. For Ghana, this assertion holds some truth, according to GIPC statistics. From Figure 5 7 a total of 79 FDI projects were cre ated in Jan June of 2003. For the corresponding period in 2007, 150 projects were created, representing an increase of about 89.9% over the 2003 figure, with increments observed for the intermediate years over the previous estimate. As expected, FDI relat ed projects have been providing jobs to increasing number of Ghanaians. Figure 5 8 shows that the percentage of Ghanaians employed in FDI related jobs in all sectors of the economy were high compared to employment offere d to foreigners In Figure 5 9 in terms of expected jobs to be created, Building and Construction came in 3rd behind manufacturing and services; surpassing agr iculture and tourism This means that t he increasing FDI flows have also been reflected in the building and construction sector of In 2001, according to GIPC data, the building and construction sector registered a total of 10 projects, representing 85%

PAGE 277

277 of all projects, with an estimated value of about US$5.16milion (5.3%). In 2010, 30 building and construction proje cts were registered, representing 7.98% of all projects, with an estimated value of US$123 million, representing 9.69% of value of FDI funds. It is not too difficult to see that the building and construction sector of the Ghanaian economy has been booming with increasing FDI levels. Even though the foregoing cann ot be considered an exhaustive analysis, it would be difficult not to recognize the importance of FDI in real estate development, with associated jobs increases in the building and construction sec tor. It has been noted that the increased inflow of FDI into Ghana is highly skewed in favor of Accra, the capital city of Ghana, implying that Accra has attracted a more than proportionate share of FDI flows into its urban space. For example, according t o GIPC data, out of the total of 95 FDI projects created in the 2nd quarter of 2007, 81 were created in the Greater Accra Region, representing about 85%, with urban Accra (GAMA), being the main destination for these projects. It follows that the unequal g eographical distribution of FDI projects is even more revealing as shown in Table 5 8 for example, depicting the regional distribution of the FDI projects in the second quarter of 2007. Again, the pattern of over concentration of FDI projects in the Great er Accra Region and the city of Accra has been sustained, if not enhan ced in recent times (Figure 5 9 ). According to the GIPC (2011), the Greater Accra Region hosted a total of 105 projects, representing about 82% of all projects created for the 2nd quart er of 2011. The estimated value of 105 projects in the Greater Accra Region (majority of which are concentrated in the GAMA) was about US$579 million. This is a rough indicator pointing to the fact that FDI inflows to Ghana continues to increase and tha t the Greater

PAGE 278

278 Accra Region ( and the city of Accra) have been disproportionately favored in terms of the geographical distribution of this all imp ortant external funds. Again, in Accra, foreign building and construction companies are active participants i n the local real estate develop ment enterprises (Grant, 2009). In general, therefore, in the advent of increasing FDI inflows into Ghana, FDI related projects and employment have all been on the rise. A particularly important aspect of the FDI employment is that employment for expatriates, very important for the argument of this dissertation and the goal in this chapter is that even though small compared to Ghanaians employed in FDI relate d jobs, is increasing (Grant & Yankson, 2003). This is an importan t development which could impact the composition of residential housing demand in Accra since such expatriates are usually better paid and can therefore express effective demand for the new houses usually constructed in the urban per iphery. This, in a way gives some credence to the conceptual and theoretical analyses (Chapters 2; Chapter 4), consistent with the theoretical postulates. The foregoing discussion i s buttressed by the historical data on FDI presented in Table 5 7 The table displays five yea r periodic averages for FDI (1980 2001). For the period 1986 1990, US$8760000 flowed into Ghana annually, increasing to US$121400000 for 1992 1996 and again increasing to US$149620000 for the period 1997 2001. In comparative terms, average of annual inflo w of FDI to Ghana for 1992 1996 represented an increase of about 1055 percent over the average annua l FDI for the period 1986 1990. Similarly, the annual FDI inflow for the period 1997 2001 represented an increase of about 1608 percent over the average an nual FDI flow for the period 1986 1990.

PAGE 279

279 of 10 sq. km per annum for the period 1985 1991 (Moller Jensen et al., 2005 ; Table 5 1), corresponding to the FDI inflow to Ghana for the period 1 986 1990. Incidentally, the dramatic increases in FDI inflow for the periods 1992 1996 and 1997 2001, roughly size for the period 1991 2002, 25 sq. km per annum, 150 per cent increase over the annual physical growth rate for the previous period (1985 1991), at the rate of 10 sq. km per annum. According to these historical data, it appears that rapid rate of FD I inflow to Ghana during 1992 2001 roughly corresponded to an equally rapid physical expansion of Accra between 1991 2002. Whereas these alone may not prove conclusively that note that FDI seems to be the only economic variab le whose behavior around this period came close to the character of the physical expansion of Accra back then (compared to population and income growth). At the minimum this is a good reason for further interrogation of the FDI urban spatial e xpansion re lationship in Accra, given the multiple pathways of FDI impact on urban land expansion and possible multiplier effect s (Chapter 2; Chapter 4 ) Two important observations can be deduced from the analysis, documented published evidence and historical data presented above : (i) both FDI inflows to Ghana and physical expansion rate of Accra can experience dramatic increases in size/volume ; and (ii) both can change in a very short period of time, resulting in a sort of coincidental rapid growth rates. Conseque ntly, it would not be surprising for both FDI inflows to Ghana and the physical expansion of Accra to exhibit non linear and exponential growth characteristics. And with the high concentration of FDI in Accra, compared to other urban areas in Ghana, it c an be argued that FDI may

PAGE 280

280 have more to do with the rapidity of physical expansion in Accra than population and income growth The surge in FDI inflows have even been accentuated in recent times. For example, for the period 2000 2010, the average annual FD I inflows to Ghana topped US$6million, even though with declines in some years (Table 5 7 ). Accordingly, physical expansion if estimated with an appropriate method Gi ven the appar ent potential importance of FDI in the urban economy of Accra and the multiple pathways that FDI may impact the urban expansion process, the FDI urban land expansion relationship was modeled and estimated. The Integrated Urban Spatial Expansi on Estimation Method (IUSEEM) was designed to help achieve the goal of quantifying the dynamic relationship between FDI concentration in Accra, with urban primacy status, and the rapid expansion of the urban extent over time. To what extent is FDI a Major Driver of Urban Spatial Expansion in Accra: Interpreting the IUSEEM R esults Average growth rate USEEM results indicate that rapid rate of increase in FDI would correspond to rapid rate of physical growth of the urban area (Table 5 3 through to Table 5 6 ) This also resonates with the characteristic of FDI which responds to economic opportunities in the form of returns on capital, be it savings on labor or environmental quality rules and regulations. This ensures that FDI will flow into a particular loca tion so long as operations would be profitable. It is also important to note that FDI orig inate from diverse sources cutting across developed and developing countries alike For instance, with regard to FDI inflows to Ghana, the traditional sources had be en the UK, USA, Netherlands and the likes (developed countries) but recent trends show increasing FDI

PAGE 281

281 inflows from developing countries such as China, India, Malaysia, Libya and South Africa (GIPC, various years). With these diversified FDI flow sources, i t would not be surprising if rate of increase in volume grows astronomically, given the relatively stable, de mocratic political environment as well as conducive investment climate in Ghana (GIPC, various years). In effect, the magnitude of FDI impact is f urther amplified when one reasons that in the discussion of socioeconomic induced urban land expansion modeling, much of FDI impulses may be masked by the use of population and GDP (or GNP) per capita as explanatory variables as pertains in current urban l and expansion analysis. This study has suggested, in view of the unremarkable changes in these variables for Accra and Ghana during the 1991 2002 it is doubtful if they had any significant contribution to the rapidity of physical expansion that has been experienced by Accra during this period. However, population and income show up as strong and significant contributors of land cover expansion in the regression estimation in Chapter 4, possibly masking the true impact of FDI in urban land expansion, espec ially in the case of Accra. Consequently, these coefficients have been adjusted and implemented in IUSEEM with the results indicating a dramatic increase in the magn itude of the impact of the FDI urban l and e xpansion process (Table 5 3 through to Table 5 5 ). It can be concluded that in addition to the rapidity of urban physical expansion that can be brought on by rapid increases in the volume of FDI inflows, and the differential spatial impacts according to concentration levels, adjusting the population an d GDP per spatial expansion process in Accra in contemporary times. The adjustments have been based on the assumption that FDI might have increased incomes and contribut ed to the

PAGE 282

282 kind of purchasing power necessary to afford the quality houses in peri urban areas of Accra but the impact could have been masked by the use of population numbers and GDP per capita in regression estimation in this work and other related works s uch as Angel et al. (2011). For Accra, s omething fundamental and enduring might have dramatically changed around this time, which may be more related to external economic influences than local ones. In this regard, it may be useful to conside r remittances and FDI as Given that remittances affect the urban expansion process through mu ltiple pathways (Chapters 2; Chapter 4), it is conceivable that remittances could have significant impact on the urban expansion pr ocess in Accra). For example, the quality housing units suggest people with high incomes, remittances from UK and USA with strong currency virtually have windfall as these foreign currencies a re changed to the local currency equivalent. Also, the nature of housing units suggest affluence such as identified for urban sprawl in the USA for example, giving indications to the notion that some of the owners might be Ghanaians living abroad. Thus, the possible influence of remittances on the urban expansion proce ss in Accra cannot be written off entirely. However, the measurement of remittances is fraught with numerous difficulties as discussed in Chapter 4 such as unrecorded transfers, definition issues, numerous non reporting countries, among other things. Thes e notwithstanding, it seems that remittances should show some positive correlation with urban spatial expansion. To the contrary, however, results obtained in Chapter 4 of this dissertation showed a negatively signed co efficient but statistically signifi cant at 95 percent confidence level. Consequentl y, even though there are strong

PAGE 283

283 indications that remittances may influence urban spatial expansion the remittances coefficient obtained in Chapter 4 has not been considered in the analysis in Chapter 5 This is left for future empirical research. V arying FDI concentration levels From the foregoing, it is logical to expect that varying FDI concentration levels should influence the rate of FDI induced urban land expansion. Results from IUSEEM demonstrat e this char acteristic (Table 5.4; Table 5 6 ), using hypothetical FDI concentration levels for Accra, showing that low levels of FDI concentration (in percentage) corresponded to slow physical expansion rate and high FDI concentration levels corresponded to rapid rate of physical expansion. As foreign capital (FDI) enters the domestic economy of Ghana but over concentrated in Accra, it is reasonable to suggest that average incomes in Accra will increase relative to incomes in other urban areas in Ghana, at least until in migration smoothens this income peak. Consequently, demand for housing and land increases rapidly, translating into conversion of new lands in the absence of densification and effective land use controls, leading into rapid rate of physical expa nsion of the city. Thus, while demand for land i n Accra receives a boost from external economic inflows, demand for land in urban areas in the rest of Ghana avail able data, may not be as rapid as the inflow of FDI to be commensurate with the physical growth. The broader geographical implication is that urban areas receiving large proportions of FDI are more likely to experience rapid rate of physical expansi on, compared to urban areas receiving low to moderate amounts of FDI inflows. It is important to reiterate that all these tentative conclusions have been based on the estimation results using only the FDI coefficient as estimated in Chapter 4.

PAGE 284

284 However, the re are reasonable grounds to suggest that the use of the FDI coefficient alone may result in underestimation of the magnitude of FDI as explained above. Consequently for example, it could be misleading to attribute 0.215 coefficient in the estimated reg ression in Chapter 4 to GDP entirely. Similarly, the estimated population coefficient of 0.787 makes the variable overly important, contrary to the raw statistical evidence adduce d already. Further, the rapidity of change in built up area does not match the rate of change in per capita income and population growth as has been noted above. Hence the need to reconsider the apparent high values of these tra ditional urban land expansion drivers since they could be masking the potential economic influences o f external forces such as FDI and remittances In short, t he GDP coefficient in the case of Accra may not be as significant as the estimated regression model suggest s. The same can be said for the coefficient of population. However, GDP per capita and pop ulation coefficient s may be adjusted to reflect the overarching importance of FDI in the ur ban economy of Accra It is obvious that the magnitude s of impact for FDI and related activities become larger. Admittedly, the adjustment procedure is not a standar dized one and therefore not proven to be robust ( conclusive ) circumstances whereby population and income appear to explain very little of the rapidity of the physical expa nsion. What may be particularly contentious is the use of the 80% FDI concentration level for adjusting the GDP per capita and population coefficients. Definitely, this is worth further investigation. The limitations of the adjustment procedure notwithstan ding, it does not seem reasonable to attribute to

PAGE 285

285 population and income the estimated coefficients as resulted from the regression modeling exercise. One important characteristic of regression modeling is the provision for the random aspect of va riation in a phenomenon thereby allowing for the modeling of socioeconomic processes more appropriately. Thus, modelers are able to estimate the proportion of the variation in the response variable which could be explained by the estimated regression mode l. In the context of this dissertation, the regression model from which the coefficients were extracted for implementation in the IUSEEM explained 82% of the variation in built up area (Accra in this case). As a result the IUSEEM results do not account f ully for the variation in the urban built up area of Accra. Also, since IUSEEM purports to quantify the spatio temporal dynamics of the FDI urban land expansion relationship, errors may accumulate over time. It would therefore be desirable if the actual le vel of FDI induced urban land expansi on in Accra can be determined. Unfortunately, data available at this time would not allow for such analysis. However, a simple comparison of IUSEEM results with independent urban expansion data for Accra may yield usef ul insights in this regard. The IUSEEM Results Compared to Indep endent D ata The IUSEEM results for the varying FDI concentration levels can be compared with data obtained f rom Moller Jensen et al. (2005), based on the following simple calculation : It woul rate for the period 1985 1991 was 10sq.km per annum; and for 1991 2002, the rate of physical growth was 25sq.km per annum. Therefore, average annual increase in built up of Accra for the period 1985 2002 can be calculated as : (10 sq. km + 25 sq. km) / 2 = 17.5 sq. km per annum. This average annual built up area expansion rate will be compared with the

PAGE 286

286 average annual growth rate of Accr a estimated with the IUSEEM using the combined adjusted coefficients comprisin g FDI, population and GDP per capita. W ith 80% FDI concentration level and firstly using IUSEEM average compared to Moller Jens en et al. (2005). It would be recalled from above that the estimated annual average built up area for Accra between, 2001 2011 u sing FDI coefficient alone yielded 0.66sq.km, based on the IUSEEM. Compared to the calculated rate of annual growth rate of 17.5sq.km from Moller Jensen et al. (2005), the IUSEEM average was found to be 3.75%. This would imply that given that about 18sq.km of land is converted to urban uses in Accra, 0.675sq.km could be attributable to FDI, if 80 percent of all FDI inflows for the period 2000 2010 were concentrated in Accra. This comparison of the IUSEEM estimated growth rate with actual, independent data does not seem significant enough. However, it has been suggested already above that using the estimated FDI coefficient alone could lead to underestimation of the actual impact of FDI on the built up area expansion process, hence the need to adjust for th e population and GDP per capita for inclusion in order to reflect the full impact of FDI as has been demonstrated above. Therefore, to examine the full impact of FDI and related activities th e average estimated built up area within a year based on IUSEEM and using the sum of the adjusted coefficients can be compared with that of Moller Jensen et al. (2005). Table 5 10 presents the comparison of IUSEEM estimation results at v arying FDI concentration = Moller Jensen et al. (2005), this time with the combined coefficients of FDI, population and GDP per capita. At assumed 80 percent FDI concentration level in Accra, the IUSEEM estimated annual average is found to be 54.31% against data from Moller Jensen et al. (2005). This implies that, on the average, about 54 percent of land converted to urban uses in Accra

PAGE 287

287 per annum for the period 2001 2011 ma y be attributed to FDI and other economic activities likely influenced by FDI. Thus, at least about half of all land converted to urban uses in Accra per annum during 2001 2011 could have been induced by FDI inflows to Ghana during 2000 2010, with assumed 80 percent concentration (based on GIPC publications) of funds in Accra. Thus, even though contribution of FDI to urban spatial expansion in Accra would vary from year to year depending on the fluctuations in FDI inflow volumes to Ghana, the external funds inflow appear to be making some significant contribution to urban expansion in Accra in contemporary times. Further insight may be gained when t he cumulative built up area for Accra can also b e used for comparison Before comparing IUSEEM estimated built up are a using the estimated annual average from Moller Jensen et al. (2005), the minimum and maximum expected cumulative built up of Accra for 201 1 estimated by IUSEEM is computed, u sing the confidence interval for the estimated coefficients for FD I, population and GDP p er capita presented in Table 5 2 The results are shown in Table 5 11 The expecte d total built up area in 2011 estimated by IUSEEM (80 perc ent FDI concentration level) = 430.81 sq. km. The cumulative built area from the lower boundary is computed as followers: Sum med all lower bound values for the 3 varia bles (Ta ble 5 11 highlighted ) and implemented in IUSEEM, yielding 406.29 sq. k m. The upper bound is computed in a similar manner yielding 456.92 sq. km. This implies the lowest and highest cumulative built up area (expected) as induced by FDI up to 2011 (from 2001) would be 406.29 sq. km 456.29 sq. km. On average, the FDI induce d built up area would be between 7.12 sq. km 11.72 sq. km, implying that at least about 7sq km and at most about 12 sq. km land expansion in Accra annually would be

PAGE 288

288 attributable to FDI inflows ; with the c onfidence level of 95% since 2001 This outcome be comes more meaningful when compared with data by Moller Jensen et al. (2005). Table 5 1 2 displays the comparison of cumulative estimated built up areas for 2011 Jensen et al. (2005). (Ad justed coefficients have been combined). T he annual growth in built up area of Accra by Moller Jensen et al. (2005) can be sq. km (from 555 sq. km in 2000 and growing at annua l rate of 17.5s q.km). (Table 5 10 ). The corresponding estimation using IUS EEM is about 431sq. km (Table 5 10 ). In comparative terms, it can be calculated that the IUSEEM cumulative built up area comes up to 60% of the cumulative built up area obtained from the Moller Jen sen et al. (2005) calculation This 2011 could be attributable to FDI inflow s to Ghana directly and indirectly (through purchasing power enablement via population and income). The estimation using the confidence interval for IUSEEM comes to 57% (lower bound) and 64% ( upper bound ) Thus, when the cumulative estimated built up area estimated for Accra in 2011 from IUSEEM is compared to data obtained from Moller Jensen et al. (2005), the share o f built up area attributable to FDI would be in the neighborhood of 57% 64%. (It is assumed that these data have acceptable quality and accuracy. It is also important to note that the regression model from which the coefficients were ex tracted to implem ent in IUSEEM explained 82% of the variation in urban land cover. Hence, the IUSEEM cannot account for the 18 % of the variation in land expansion in Accra, but of reasonable explanatory power, given the accuracy of the processed remote sensing data impleme nted ).

PAGE 289

289 These notwithstanding, FDI and related economic activities seem to be making significant contribution to urban land expansion in Accra. This is not surprising considering the possible multiplier effects of FDI and the multiple pat hways through whic h its influences may be transmitted into th e urban landscape (Chapter 2; Chapter 4). This finding supports that of Seto (2005) for the Red River Delta in Vietnam and the Pearl River Delta in China the only quantitative study of FDI caused urban land expa nsion known to this investigator Consequently, in addition to adding to the literature on analysis of contemporary urban land expansion with the potential to affording a more comprehensive understanding, this study seems to be the fi rst to have modeled a nd quantified explicitly FDI induced urban built up area expansion over time for a specific city which can be replicated elsewhere in developing countries It is also pertinent to point out that i n both the single variable (FDI co efficient alone) and co mbined variables ( combined estimated coefficients: FDI, population and GDP per capita) cases, varying the FDI concentration levels resulted in marked differential impact on built up area expansion rates This implies that the rate of urban spatial expansio n may be influenced by the rate (FDI inflow volume per annum) of increase/decrease in the flow of FDI in the particular urban areas. From the forego ing discussion, it may be difficult to dispute the fact that FDI may be contributing more to Ac l expansion in recent time s especially given that population and income growth have been less than remarkable compared to the physical growth rate of Accra during the era of intensifying globalization Even though urban expansion researchers like Karen S eto make some attributions to FDI as a potentially important urban land expansion driver (Seto et al., 2011; Seto, 2005),

PAGE 290

290 mainstream urban expansion theoretical works and modeling have largely neglected to incorporate FDI. But this preliminary work which is the first detailed attempt to model and quantify such globally induced urban land expansion, should help draw attention to FDI as an emerging potent urban land expansion driver, especially in fast globalizing primate cities in developing economies. Per haps this will gain currency if it is demonstrated that FDI induced urban spatial e xpansion can be grounded in a standard urban economic theory such as the Thunen Alonso theoretical framework. FDI and the Thunen Alonso T h eoretical F ramework Basically, the Thunen Alonso theoretical framework (outlined in Chapter 4 and in this chapter ) persp ective of utility maximization; positing that land value would be determined by the economic benefits derived from the kind of usage to which the land is put. Typically, the Thunen Alonso theore tical framework has emphasized transportation (impro vement and cost), population, income and physical /geographical factors (which influence accessibil ity) as the main drivers of land values, particularly in the urban context, and therefore being implicated in the expansion of the urban extent (Angel et al., 2011; Seto et al., 2011; McDonald & McMillen, 2010; McGrath, 2005). For example, lands located in more accessible areas such as the CBD tend to be of high value attracting mostly commercial activities able to pay for the cost and leave room for profit for investors. Similarly, population growth puts pressure on demand for shelter and other infrastruct ure whose provision require the use of land. There is also interrelationships among these variables whereby, for example, improved transportation results in improved accessibility, thereby enhancing land values. However, it is arguable that even though t hese factors remain important in determining urban land values, their

PAGE 291

291 effects have been mitigated in many respects, as a result of technological advancements and economic globalization. Technological advancements in transportation and communication have ma de it cheaper to live in the su burbs even though the cost of transportation per se may not have bee n reduced. Physical geographical factors likewise do not pose serious limitation on what piece of land is suitable for residential development as wetlands get filled and elevated grounds are leveled using modern technology In short, traditional drivers of urban land values may not be entirely responsible for the dynamics of urban land demand in a given urban area since global economic forces may be exertin g significant influences, which must be recognized and adequately represented in standard urban economic theorization modeling and analysis. Essentially, external economic forces can drive urban land values in distant areas from their origination location s, as globalization and technological advancement continue to mitigate the magnitude of impa ct of the traditional factors influencing accessibili ty At the same time people are being purchasing power enabled by FDI, making domestic / national population an d income growth less of enduring determinants of urban land values. Consequently, urban land values, e specially in peri urban areas, may appreciate significantly and rapidly even as domestic / national population and income do not experience any appreciabl e growth as shown above for the case of Accra and Ghana (Table 5 7 ). Thus, in discussing urban land value determinants in the era of increasing economic globaliza tion, it may be useful to consider the fact that urban land values in a given location may sta rt increasing as a result of events happening in similar urban areas and countries located far awa y. For instance, currency appreciation in European markets relative to the USA dollar, if

PAGE 292

292 sustained for a considerable length of time, may trigger demand for housing and real estate in major urban areas in USA such as Manhattan in New York City (Sassen, 2006). Thus, deterioration in investment conditions in London could induce FDI inflows to New York City even though domestic economic conditions in the USA in general might not have changed in any significant way. In both cases, therefore, it is entirely possible for domestic urban land values to increase, not necessarily triggered by domestic economic conditions such as national economic growth, as urban land d emand gets driven by external (global) economic stimuli. However, in the case of advanced urban economies, this may not necessarily lead to conversion of new lands, largely due to effective growth management policies which w ould help ensure densification infilling, high rise apartmen t buildings and the likes For example, London is a typical primate city receiving considerable amounts of financial resources from foreign sources, but the existence of the Greenbelt is believed to have helped restrict the gr o wth of inner London In other cases, a country like USA may have a good number of virtually equally competitive urban areas such as New York, Los Angeles, Chicago, among others, which are attractive for foreign direct investments. Thus, inflow of FDI may be more geographically diversified than concentrated which would help mitigate the spatial impact of the external funds inflow other things being equal It is therefore not the ar gumen t being advanced in this dissertation that those FDI inflows automatic ally would have positive influence on the urban spatial expansion p rocess. The important point this dissertation advances is that FDI inflows would likely cause real estate property values, including land, to start appreciating in the domestic urban econom y. In advanced countries with functioning and effective urban growth

PAGE 293

293 management tools and systems, increases in land demand may be absorbed without necessarily causing a significant increase in the urban extent. However, in a primate city in a developing country, which usually lacks growth management policies, increased flow volumes of FDI may contribute significantly to rapid physical expansion of the receiving city, with the process likely to be characterize d by sprawl in the face of gradually increasing opportunity cost of peri urban lands. More formally, inflows of external f unds such as FDI when significantly concentrated in a large urban location would serve as the external economic stimulus (or exogenous shock) to the urban land market, generating increase in demand for land (which could be described as an curve ) (C hapter 2; Chapter 4). Agricultural land prices, howev er, remain relatively constant or decrease in real terms as peri urban agricultural production outside the country, thanks to international trade and technological advancement in transportation and communication. Consequently, i ndications are that FDI may be inducing urban spatial ex pansion in Accra, particularly accentuating the rate of physical expansion, and increasing in importance with intensification of economic globalization The foregoing analysis, to a large extent, is consistent with postulates of the Thunen Alonso theoreti cal framework when viewed from urban land value perspective as induced by external economic force of FDI. It is the argument in this chapter that explained within the context of t he classical urban land use, land value theory (Alonso, 1964). More formally, the increased FDI induced land demands will drive urban land values upwards (Figure 5 10 ), even as population and incomes grow at mu ch slower

PAGE 294

294 rates. In Figure 5 10 A (E 1 ) is t he initial equilibrium for urban land use value and agricultural land use value, which i s shown spatially in Figure 5. 10 B However, this equilibrium cannot be stable in the face of stagnant or declining agricultural land values whereas urban land use value s rise induced largely by high volumes of FDI inflows into the urban space economy. It is expected that high levels of FDI concentration in the city increases demand for land for urban uses such as housing and infrastructure, which causes urban land use value to rise. This can be considered as a right shift (autonomous shift ) of the urban land value curve, since FDI, which is the main stimulus in this case is externally generated and exogenous to the urban and national economies In this regard FDI inflow market, which makes it capable of inducing autonomous change in the urban lan d use valu e curve in Figure 5 10 A A new equilibrium (E 2 ) is established outward of E 1 This is reflected in the move ment of the 10 B But since agricultural uses in the peri urban areas is less profitable compared to urban land uses, this second equilibrium may not exist per se. Consequently, these lands may just be waiting to be converted t space. In a nutshell, the interactions of the urban land value curve and the agricultural land value curve results in the spatial ramif ication reflected in Figure 5 10 B ; where by t he urban lands in its path. The reasoning behind the shift in the urban frontier is that in the peri urban area of the city, the rise in urban land use value increases the opportunity cost of holding lands in natural and agricultural uses (Hanink, 1997). Peri urban lands get converted to urban

PAGE 295

295 uses since this is more profitable compared to farming, thereby increasing the urban extent. Specifically, relating to Accra, it can be argued that agricultural land values in the peri urban areas of the city have remained stagnant or even declined, (since farming is less profitable compared to urban uses) whereas urban land uses continue to receive m FDI concentration in Accra. Logicall y and theoretically, therefore, this may partly explain the d ocumented sprawl development as per (Angel et al., 2011; 2005; Aryeetey Attoh, 2010; Moller Jensen et al., 2005), indicating urban density decline as argued elsewhere in this dissertation of be ing characteristic of relative affluence that can be linked to global economic influences. Indeed, d ecline in urban density in Accra has also been confirmed by some studies, including Ang el et al. (2005). T he authors note that in 1985 the average density (persons/ sq.km ) was 14,120.39. In 2000 this figure had reduced to 8,102.64, representing a decline of 3.66% per annum and implying that the average Accra resident was occupying more land in 200 0 than he/she did in 1985. Generally, on the global scale, Ange l an d colleagues have asserted that The projected rate of increase in urban land cover will be higher than the rates of increase of the urban population because urban population densities can be expected to dec line. (Angel et al. 2011: 46). This may be a ttributable to other factors but there are indications that FDI may be playing a significant role substituting for country specific economic growth increases required to sustain the observed affluence in the character of the new real estate development. The important implication is that : even though economic growth usually occurs in tandem with urbanization (including urban spatial expansion), the primate, globalizing

PAGE 296

296 city may undergo unprecedented physical growth in the face of slow growth in the natio nal economy as well as low population growth Based on the analysis presented in this dissertation Accra may represent such a paradox, which can be resolved by appealing to external economic influences of globalization such as FDI. More importantly, the situation can be grounded in a refined form of the standard ec onomic theory of urban land use and land value as espoused by Von Thunen (1826, 1926 ) and Alonso (196 4), particularly building on its current extensions (Angel et al., 2011; 2005; McGrath, 2005) that models the expansion of the urban extent as a function of relevant socioeconomic variables. More specifically, Thunen Alonso theoretical framework, formally articulated in Chapter 4 of this dissertation may help explain how FDI might h ave contributed to the ra pid spatial expansion of Accra. It is the cardinal argument in this regard that in a developing country setting urban locations receiving disproportionate share of FDI will experience rapidly rising urban land values whereas agric ultural land values remain constant or decline thereby increasing the opportunity costs of holding lands in agricultural uses and natural fallow in the peri urban areas of the city Consequently, rapid land conversion will follow other things being equal, as t he calculations show in Table 5 3 above. Reinforcing tendencies may also be associated with FDI inflows whereby existing agglomerations attract more FDI and population who become purchasing power enabled, demanding housing and infrastructure construc tion ; the construction of which deepens the attractiveness of the ur ban location, generating further rounds of FDI attraction. Thus, once FDI starts concentrating disproportionately in an urban location, relative to other urban are as in the country, From a

PAGE 297

297 purely economic perspective the peri urban spatial expansion appears to be a win win situation for both peri urban agricultural land holders and developers. For land holders, increasing peri urb an land values incur high opportunity c ost if held in agricultural use; it would be more profitable to sell the land to developers or build on it (for renting, perhaps) For the developer, increasing urban land values exhibit spatial inequality in its dis tribution pattern, with higher land values close to inner locations in the city but expensive while peri urban lands are of high value and cheaper. Hence, from both the peri urban landowner and the developer perspective s the land is put to highest value a nd best use by converting from agricultural uses to urban uses To sum up, FDI flows into geographically restricted urban space increase urban land values, increase opportunity costs of peri urban agricultural lands, making land conversion to urban uses m ore profitable, especially in the peri urban areas of the city. However, the Thunen Alonso theoretical framework assumes the operation of free market in land. In the specific case of Accra, this would require the existence of a functioning peri urban lan d market through which land can be exchanged based on supply and demand for the postulates of the standard urban economic theory (the Thunen Alonso theoretical framework) to be properly analyzed. However, the peri urban land market in Accra and urban Afri ca in general, unl ike urban markets in advanced economies, is fraught with difficulties, necessitating some further elaboration. The Peri Urban Land Market in Accra u nder Globalization The discourse on the peri urban in Africa is beset with vexing question s, according to urban researchers bordering on definitions, laws as well as land markets Theodore Trefon has asserted that

PAGE 298

298 Peri urban areas are fringe areas because they lie both at the edge of the city and at the limit of the rural hinterland. These wor lds overlap and intermingle; making it difficult to establish where the city ends a nd where the rural place begins. (Trefon, 2009: 15). C omplexity has therefore bedeviled the precise/comprehensive definition for the peri urban as well (Locat elli & Nugen t, 2009). One simple definition for th e peri urban is quoted in Trefon, 2009: 19). This definition devoid of function ality. This can be complemented by a second definition bordering on the composition of the peri urban area T he peri urban zone can be broadly characterized as a mosaic of different land uses inhabited by communities of different economic status, in a st ate of rapid change with a lack of infrastructure a nd a deteriorating environment. (Birley & Lock, 1998, cited in Locatelli & Nugent, 2009: 19). and the of the peri urban area is made more explicit, implyi ng an area undergoing continuous change. The explanation of the formation of the peri urban has been based mainly on demographic factors and have often been carried out in developed countries (Trefon, 2009). However, Theodore Trefon has indicated that such studies could be drawn on to help understand peri urbanization in the Central African context (Trefon, 2009:20) and by extension to the Ghanaian context B riefly i n an Australian case study, Tania Fischer has identified and conceptualized four main grow th processes which could serve as a model to explain population growth in the peri urban area: (i) sub urbnisation (in migration from metro area to peri urban locations adjacen t to the metro boundary); (ii) counter urbanization ( in migration from through out the metro area to peri urban locations, both adjacent to the metro area and those more distant); (iii) population retention ( increased period of residence and reduced ra te of out migration); and (iv) centripetal

PAGE 299

299 migration (in migration from outlying ru ral areas) (Fischer 2003: 553, cited in Trefon, 2009: 20). In effect, Theodore Trefo n summarizes the situation thus As housing plots and agricultural space within the city become rare and ew migrants hailing from rural areas also tend to occupy land in th ese areas for the same reasons. (Trefon, 2009: 20). It is obvious from the foregoing that the focus is on internal dynamics within the country and within the urban area. But when this is a pplied to the case of Accra, it would suggest that population and income growth in the country enabled people to own the qual ity residential property widely held to be responsible for the rapid growth in the consumption of peri urban lands. However, this i s doubtful as the arguments and discussions advanced so far in this dissertation show It is therefore imperative that the influences of exte rnal economic forces originating outside the national system are considered in the discourse given the almost seam less transmission of global Even though cities have been in existence for hundreds of years in Africa (Rakodi 2007 ) widespread urbanism (urban lifestyles) can be considered to be largely a recent ph enomenon on the cont inent largely occurring in tandem with increasing economic globalization ( Kasanga & Woodm an, 2004) Essentially, therefore, urbanization and globalization forces have combined to contribute to the complexity of land ownership and land markets in Africa, subject ing urban spaces in Africa to competing claims as (Locatelli & Nugent, 2009; Rakodi, 2007 ) would put it even in the absence of smooth operation of free market mechanism in land According to A. Trefon: The complexities of land tenure in Africa hav e been studied in depth by the interdisciplinary network of social scientists: AFREFA (Association pour la Promotion des Recherches et des Etudes Foncieres en Africque) (Le Bris

PAGE 300

300 et al. 1991). They sugge st a model of land tenure which suits the peri urban c ontext relatively well. In their view, the totality of social relations is based on land and space. These social relations are primarily conditioned by politics, economics, institutions and the nature of infrastructure making land tenure a total social fa c t (Trefon, 2009: 20 21). F or most parts of Africa, during the pre colonial times, land ownership operated under customary laws, whereby communities, families and stools (chieftaincy) had vested interest in the land and could only be used and transferred but not sold outright ( Rakodi 2007; Gough and Yankson, 2000 ). Also, most land acquisition had had to do with the cultivation of the land for food or some cash crops (Payne, 1997, cited in Gough & Yankson, 2000 ). To acquire a piece of land for such an o peration, the interested party of the family owning the land; and in se (Gough & Yankson, 2000 ). In short, according to customary land tenure system in pre colonial Africa, land was not considered a commodity to be sold for indi vid ual (private) ownership (Rakodi, 2007 ). The advent of the Europeans on the African continent did not do much to change the customary land tenure system si nce according to Mabogunje (1992 cited in Gough & Yankson, 2000) I t suited the colonial administration to strengthen the myth of all land being under communal ownership, as this enabled them to increase the power of the local chiefs through whom they coul d rule indirectly. (p. 2486). Alongside the customary land tenure system, however, with the advent of colonization, European based land acquisition systems had been introduced in several parts of Africa, creating a dual system of land tenure (Mabogunje, 1992 cited in Gough & Yankson, 2000). The most relevant aspect for the discussion presently is the assertion operation of the free market economy in land for Africans wh

PAGE 301

301 (Mabog unje, 1992 cited in Gough & Yankson, 2000: 2486). Thus, free market areas in particular, have not been encouraged until recently. Concerns with the inefficacy of the traditional land that the customary system of land tenure might have served the needs of rural communities well but found wanting as people began to migrate to cities to eke out a living and wishing to ac quire lands i n their name (Mabogunje, 1992 ). This points to a dissatisfaction with the customary land tenure as it exists, with efforts aimed at some reforms. Again, increasing globalization and democratic dispensa tion which have allowed for private ownin g of real estate property have c ontributed to freeing lands for urban development. Consequently, in addition to responding to pressures of urbanization, most land markets in Africa are now opening up to more market forces alien to the land (Africa), especi ally during the era of intensifying economic globalization however discrete the transfers may be As a specific example for Ghana, even beyond the capital city of Accra, Kasanga and Woodman have asserted that the p eri urban land market in Kumasi, the sec ond largest city of Ghana, has been influenced by global economic forces in the forms of remittances and FDI, indicating that demand for land for residential purposes have been boosted by these ext ernal funds inflow (Kasanga & Woodm an, 2004). The case of p eri urban land market in Kumasi forces is not different from that of Accra however. Thus, peri urban lands in A ccra are no longer by and large, these lands have become commoditized, privatized to some extent, allowing for individual ownership (Gough & Yankson, 2000). As a result, peri urban land demand is increasingly being influenced by external economic forces

PAGE 302

302 including FDI and remittances. Consequently, u nlike most urban land markets in som e countries in central Africa, where economic globalization does not appear to be the major force directly driving urban land dem and the major urban land demand in Ghana, and in Accra in particular, are being significantly influenced by forces of economic globalization Indeed, it has been documented that foreign private companies are beginning to participate in real estate development in urban Accra (Grant, 2009 ; 2007; 2001 ) indicating the opportunity for private ownership of lands in Accra, even though the acquisition procedures may be fraught with difficulties. By and large, then, there seem to be some form of peri urban land market in Accra where lands can be freely bought and / or sold, even though it may not be what one would expect in developed cou ntries. It appears therefore that the requisite for the existence of free market in land as a requirement for the Thunen Alonso t heoretical framework can be found in pe ri urban Accra, to some extent. Trade in land does seem to exist. However, in popul ar explanation of urbanization in Ghana in general, analysts have ignored the potential influences of foreign economic influences. For example, Farvacque Vitkovic and colle a gues have noted that What is evident in the Ghanaian context is that urbanization has led to rapid increase in consumption of land in the past two decades. The factors that have influenced the increase of land values are a combination of the following: population growth, inefficient town planning, activities of the Central Government, and the monopoly of land holdings ( Farvacque Vitkovic et al., 2008: 12). The problem with this assertion is that impact of foreign economic forces on domestic land markets are masked and not given any role in the urbanization process (and in the urban s patial expansion process, by extension). Recognizing global economic influences is especially important in the situation where the peri urban spatial expansion

PAGE 303

303 a (Yeboah, 2003). For Accra, therefore, even though in migrants from rural hinterlands have contributed to the expansion process (from a demographic perspective) this aspect may not be as signifi cant as the demand driven from forces like FDI (from the perspective of spati al expansion) To help draw attention and encourage research in order to address this situation, FDI induced urban spatial expansion in Accra has been theorized via the Thunen Alonso theoretical framework and quantified using an integrated method IUSEEM f ormulated by this author rapid physical expansion has been induced, in part, by high volume s and sustained concentration of FDI, following intensifying economic globalization and economic liberali zation (SAPs) Two main deductions may be advanced: I f a globalizing primate city in a developing country experiences rapid rate of physical growth within a short time but population and income growth both remain low, FDI inflows might be contributing sign ificantly to the urban land expansion process such as is happening in Accra. R apid increases in the volume of FDI inflow into the urban economy of Accra is manifested in equally rapid physical expansion of the city. Thus, the rate of urban spatial expansi on seems to mimic the flow characteristics of FDI. Therefore, based on the presentation so far in this dissertation and this chapter in particular a theory of FDI induced urban spatial expansion is proposed thus : In a globalizing primate city located i n a developing country, the rate of physical expansion of the city is a function of the volume of annual FDI inflows ; during the expansion process peri urban land is put to its highest value and best use by converting from agricultural to urban use there by accelerating the rate of urban spatial expansion as FDI inflows increase over time In effect, t his theory would argue that the rapid rate of physical growth experienced by a fast globalizing city, located in a developing country, has more to do with i ncreasing

PAGE 304

304 volumes of FDI and its unequal geographical distribution favoring urban agglomerations, than resulting from population and income growth which have been consistently identified as the most important urban expansion drivers. Underlying a ssumption s A bsence of growth management and land use controls L ow growth rate (GDP/GNP per capita growth lags physical growth rate) L ow to moderate population growth rate (lags physical growth rate) U nequal distribution of FDI wi It is important to put on record that the theory bei ng proposed aims to build on the classical Thunen Alonso theoretical framework, in an attempt to incorporate influences of economic globalization on urban land values and urban spatial expansion from a geo graphical perspectiv e This theoretical extension will be particularly useful for the analysis and for a more comprehensive understanding of contemporary urban spatial expansion in fast globalizing primate cities in developing count ries. Given the above, t herefore, it can be argued that this study lays out both the theoretical foundations and methodological framework specifically suitable for theorizing and quantifying the emerging phenomenon of FDI induced urban spatial expansion which should, hopefully, h elp draw more attention to it. A full articulati on of the proposed hypothesis would require a better understanding of the geographical concepts and themes of spatial differentiation movement and human environment interaction as discussed above This wou ld require integration of concepts, themes and theory across economic geography, urban geography and urban economics as demonstrated in the literature reviewed above In particular, t he geographical fundamental theme o f human environment interaction provi des a framework to embed FDI induced urban spatial expansion firmly in the broader context of human induced environmental change which

PAGE 305

3 05 is widely recognized to be one of the formidable challenges faced by society (UN, 2005 ; IHDP, 2005 ) FDI and Human Induce d Environmental C hange FDI induced urban spatial expansion can be viewed as an instance of human induced environmental change playing out in the urban setting. Already, urbanization is increasingly being identified as one of the formidable aspects of glob al environmental change, posing one of the greatest challenges of the twenty firs t century (IHDP, 2005) and is projected to rise. For example, it has been projected that by the year 2030 about reas, majority of which will be expected to live in urban areas of the developing world (UN, 2003). It goes almost becoming urban dwellers then it is important that ad equate attention is paid to the biophysical environments that would contain such human populations and their settlements as well as the increased demand for environmental resources. Importantly, the loss of vegetation cover associated with urbanization (s patial expansion) affects the microclimate via alterations in ecological functions and integrity of the biophysical environment, among other things, which would feed into global environmental change, via land cover change and land degradation. It is widely acknowledged that economic growth is inextricably tied to the environment since the environment serves as the natural capital (input) as well as the global sink for wastes from the production processes and consumption ( Munasingh, 2002; England, 2000; Cost anza et al., 1997 for example ), in addition to its life supporting functions such as ensuring ecosystems health and functions. Costanza and colleagues have rightly asserted that regardless of the sophisticated nature of technological advancement basic ra w environmental

PAGE 306

306 resources serve as inputs for numerous productive activities, not to mention the critical life support functions the environment provides such as air purification, pollination, and carbon dioxide sequestration and the likes (Costanza et al. 1997). However, economic growth continue s to put these invaluable natural assets under stress. For example, in a cross country empirical study, using species diversity and income per capita as proxies for biodiversity and economic growth, Asafu Adjaye ( 2003), found that economic growth generally had a dverse effect on biodiversity. Of particular importance to this work is the loss of pristine and agricultural lands to urbanization and the development process in general at the expense of land conservation ; and the case of Florida, even in an advanced country will serve to illustrate more. According to Zwick and Carr (2006), economic growth and urbanization, coupled with population growth could have lasting neg ative impacts on conservation lands in Florid a More specifically, the study projects population of about 18 million (2006) to grow to about 36 million by the year 2060; correspondingly, urban development level occupy 5,998, 523 acres of land and is projected to increase to about 112 951 787 acres d uring the same period (Zwick & Carr, 2006: 1). However, land conservation is not projected to increase and rather the attendant urban expansion would occur at the expense of pristine an d agricultural lands (Zwick & Carr, 2006: 1). Located in a highly advanced country, this Florida study provides a valuable food for thought, adding to the doubts raised against the Environmental Kuznet Hypothesis, since increasing economic growth hardly translates into en h anced environmental protection ( land con servation in this case ). If rich countries cannot hold onto their pristine and agricultural lands in the face of increasing economic growth and urbanization, one can safely assume that it would be doubly

PAGE 307

307 difficult for developing countries, often lacking g rowth management policies and effective environmental quality controls. In the developing world, therefore, urbanization and urban spatial change have often resulted in environmental degradation, feeding into the larger problem of human induced environment al degradation. Human induced environmental change (degradation) is becoming widespread around the globe and the processes involved are complex. The United Nations International Strategy for Disaster Reduction (UNISDR) acknowledges that The types of human induced degradation are varied and include land misuse, soil erosion and loss, desertification, wild land fires, loss of biodiversity, deforestation, mangrove destruction, land, water and air ( http://www.unisdr.org/we/inform/terminology ) In many respects, therefore, environmental degradation can be viewed as a broad term which can be simply described as the decline in environmental quality to the extent that environmental resources contained therein may not be usable by humans and animals Examples include excess carbon dioxide in the atmosphere, air pollution, soil contamination as well as land degradation in general whereby land a nd its resources are rendered unsuitable for use by humans and animals. From this perspective, therefore land degradation can be considered a compone nt of environmental degradation. The problems associated with e nvironmental degradation appear to be furth er compounded by globalization; econ omic globalization to be more specific, with its tendency to drive increasing levels of economic activities and economic growth across the surface of the earth Rightly so, it has been observed that high rates of economi c growth, which inevitably drives increasing levels of urbanization, could put enormous strain on the ecological resources of the earth including urban lands and its biological

PAGE 308

308 resources ( Wackernagel, 2006 ; Costanza et al., 1997 ). Even though many analyst s extension of processes began long ago, including modernizati (Clapp & Dauvergne, 2005: 20), it is equally valid that the intensification of economic globalization in recent times have been unparalleled (Dicken s 2003) and could contribute in no small measure to global environmental change (Clapp & Dauvergne, 2005) In deed, in the opinion of many commentators and researchers, economic globalization is still picking up steam, despite the economic downturn in 2008 (UNCTAD, 2010); and FDI is likely to continue to be attracted into urban agglomerations, such as primate cities in developing countrie s. Sustainable development and sustainability S cience do ha ve appealing aspirations that promise to help solve the social economic environmental problems but still lack s a core theory of sustainability which would aid rational and consistent explanation, not to mention the fact that intensifying economic globaliza tion c From the foregoing, it is not too difficult to see that urban spatial expansion which almost invariably removes the vegetation cover entirely, if uncontrolled, could result in land degradation and feed into environment al degradation. Consequently, if FDI is proven to be a potent driver of contemporary urban land expansion, it would have wider of human induced environmental change in u rban areas. However, such FDI induced environmental impacts would be spatially differentiated, since FDI flows generally exhibit unequal geographical distribution among countries as well as within a country. For instance, among developing countries, China is noted to receive a greater chunk of

PAGE 309

309 the FDI pie (Chapter 1; Chapter 2). Similarly, within a country FDI distribution pattern usually favor urbanized areas wit h existing agglomerations making primate, urbanizing and g lobalizing cities such as Accra, attractive for FDI inflows ; and meriting theorizing and quantifying for a better understanding Consequently, insights in this chapter, even though from a case study, could have broader implications for research on human induced environmental change necess itating the replication and testing of the proposed theory elsewhere, especially in the developing world. Obviously, the importance of geographical research in this regard cannot be overemphasized. Summary So, c an FDI be implicated in the rapid rate of Acc especially during 1991 2011? The answer seems to be Essentially, Chapter 5 has focused on understand ing spatial size has expanded over time in response to increasing inflows o f FDI ; theorizing FDI induced urban s patial expansion and contextualizing the phenomenon in the broader human induced environmental change discourse In pursuant of these reasonabl e arguments were advanced as to whether or not FDI could be implicated in the rapid physical expansion of Accra particularly, increasing consumption of peri urban lands. Rapid u rban land expansion (usually characterized as sprawl) such as what is happening in Accra have been explained using population and income as explanatory variables. However, indications so f ar are that these variables may not adequately explain the rapidity of the spatial expansion that has occurred in Accra, especially, during 1991 2011 Consequently, FDI induced urban spatial expansion theoretical argument was advanced, drawing heavily on current extensions of the Alonso t heory in which the expansion of the urban extent is model ed as a function of socioeconomic variables Initial hypotheses developed were

PAGE 310

310 tested using the Integrated Urban Spatial Expan sion Estimation Method The IUSEEM is simply an allometric based model integrating regression and exponential growth modeling, to quantify the urban built up area over t ime as induced by FDI inflows. Essentially, the IUSEEM results show that increasing volumes of FDI inflows to Ghana (with 80 percent concentration in Accra) corresponded closely with the rapidity of rate of physical expansion of Accra during 1991 2011. It is estimated that for 2001 2011, FDI induced average annual increase in the built up area of Accra was about 9 sq. km assu ming 80 percent of all annual FDI inflows to Ghana during 2000 2010 concentrate d in Accra Comparison with results of anot her study (Moller Jensen et al., 2005), showed that for the period 2001 2011 the IUSEEM estimate d average annual built up area increas e constituted about 54 % of the total estimated by these authors. In terms of the cumulative built up are a in 2011 alone, the comparison showed that FDI induced built up area might have accounted for about 60 expansion (based on Moller Jensen et al., 2005) Tentatively, therefore, it can be argued that FDI induced urban spatial expansion may be taking place in Accra. The IUSEEM results further showed that varying the FDI c oncentration levels resulted in differential rate over time The important broader implication of this system may help explain, in part, differential spatial expansion rates in different urban locati ons. Specifically, urban locations within the country receiving high volumes of FDI inflow are likely to experience rapid physical growth rate in their urban extent. Based on the outcomes above an initial FDI induced urban spatial expansion theory has be en proposed as an extension to the Thunen Alonso theoretical framework :

PAGE 311

311 In a globalizing primate city located in a developing country, the rate of physical expansion of the city is a function of the volume of annual FDI inflows ; during the expansion proces s, peri urban land is put to its highest value and best use by converting from agricultural to urban use, thereby accelerating the rate of urban spatial expansion as FDI inflows increase over time The basic argument of the proposed theory or general hypo thesis is that the rapid rate of physical growth experienced by a fast globalizing primate city like Accra, located in a developing country, has more to do with increasing volumes of FDI and its unequal ge ographical distribution which favor urban agglomera tions, than resulting from population and income growth which have been consistently identified as the most important urban land expansion drivers (Angel et al., 2011; 2005; Seto et al., 2011) FDI inflows increase demand for land (resulting from purchasi ng power enablement from FDI and related activities) increase urban land values; with peri urban lands in agricultural use incurring rising opportunit y cost. It makes economic sense for the land holder to convert to urban use. At the same time, even tho ugh peri urban land values are rising, the lands are cheaper than those in inner city locations hence developers will find it more profitable t o build in the peri urban area. Thus, from purely economic perspective, both developers and landlords are better off converting peri urban agricultural lands to urban uses: utility in both instances are maximized and economic efficiency would have been achieved since the peri urban land is put to its highest value and best use. This, according to the proposed FDI i nduced urban spatial expansion, would explain why the primate globalizing city will keep expanding rapidly (spatially) into peri urban areas in response to increasing volumes of FDI inflows even as population growth has been slow and economic growth simil arly sluggish In general, t he theory suggests that urban researchers, particularly those interested in urban land expansion

PAGE 312

312 will do well to begin pay ing attention to FDI as a potent urban land expansion driver, beyond numbers and statistical correlations linking population and income to urban land expansion. Consequently, a cogent argument is advanced to the effect that urban land expansion theory and models should seek to incorporate economic influences flowing from global sources to better reflect curr ent realities of economic globalization. In view of the increasing importance of FDI with intensifying globalization, and its potential repercussions on the biophysical environment, the research theme pursued in this chapter will likely assume increas ing societal and scientific importance and this chapter in particular and the dissertation in general would serve as a pioneering effor t. Importantly, i t can be argued that the theoretical foundation and methodological framework for a systematic study of FDI induced urban spatial expansion have been laid out in this chapter of the dissertation. This should help draw attention from urban geographers, economic geographers and urban economist s as well as urban planners to develop and test hypotheses from the proposed theory aimed at testing the theory in other locations to increase understanding. Moreover, s ince FDI induced urban spatial expansion can be put in the broader context of human induced environmental change and associated problems such as land degradation; increased understanding in this area would be a welcoming development for researchers interested in understanding the socioeconomic aspects of global environmental change in general The importance of increased understanding in this area of r esearch has been underscored (NRA, 2010). However, the economy environment relationship dynamics are complex, eliciting different views from different quarters who have suggested disparate worldviews. Complexity notwithstanding, sustainable development as a concept seems to capture

PAGE 313

313 many of the c oncerns from the various camps. discipline has been noted to have bourgeoned since the formalization of the concept of sustainable development by t he World Commission on Envi ronment and Development (WCED) otherwise known as the Brundtland Report in 1987 (Bettencourt & Kaur, 2011) attesting to the increasing societal c oncern for effective solutions that will balance the social, economic and environmental imperatives. Land Cha nge Science Sustainability Science, which arguably is directly concerned with the study of land use change causes and consequences (including urban expansion). However, a core the ory has not emerged as yet. Of particular note is that the focus in land change science has usually been but generally fail to incorporate FDI influences, in th e urban setting. The economy environment relationship problems appear to be further complicated by intensifying economic globalization, especially in primate cities of developing countries. In view of this the efforts in this chapter which seek to integra te geographical concepts and fundamental themes with urban economic theory to help better understand the spatial imprints of economic globalization in the urban landscape can potentially contribute to Sustainability Science in general and Land Change Scien ce in particular In addition to the potential contribution to scientific enquiry, the results above and the foregoing discussion, if empirically supported in other urban areas may have important policy implications for urban policy, urban planning and urban environmental conservation For urban policy in gener al, policy makers may be focusing on population

PAGE 314

314 numbers, which may flock into the city from within the country whereas the main away, largely beyond the control of these policy makers. In effect, even though population influx into Accra may be controlled, this would likely do little to stop the sprawling of Accra and the associated loss of peri urban lands to urban development sinc e the underlying land conversion driver may be more related to FDI than anything else. Insights in this chapter may encourage urban policy makers to look beyond the urban and national systems. Secondly, land conversion from agricultural to urban uses g ive rise to increasing loss of vegetation and declining urban green spaces. Since vegetation cover forms a fundame ntal basis of species habitats ( Wackernagel, 2006; Defries et al. 2004; McKinney, 2002) it would not be too far fetched to imagine how the loss of peri urban lands could have negative impact on species habitats and ecosystems resulting in a general dec line of urban terrestrial biodiversity which should be a concern for the urban environmental conservationist. Thus, an important implica tion of pa rticular interest to this writer is the potential negative impact that unbridled FDI inflows could have on urban environmental conservation efforts. Finally, urban planners are concerned with all aspects of the city, making sustainable development the cent ral theme of contemporary urban planners (Campbell, 2003). Increasing volumes of FDI inf lows would complicate the traditional concerns with transportation, housing and the likes, making it worthwhile for urban planners to pay attention to increasing FDI i nflows into the urban economy. Further elaboration on these practical implications has been provided in Chapter 6

PAGE 315

315 Figure 5 1. S patio temporal growth of the city as a spatial /geographical system (A) City expands in all directions unconstrained (B) City expansion is constrained A T 1 T 2 T 4 T 5 T 3 City core B T 1 T 2 T 4 T 5 T 3 City core

PAGE 316

316 Figure 5 2 Steps in formulating the Integrated Urban Spatial Expansion Estimation Method (IUSEEM) struct.

PAGE 317

317 Figure 5 3 Spatial expansion of Accra: 1985 2002 Source: Moller Jensen et al. (2005)

PAGE 318

318 Figure 5 4 A section of Accra and a simplified diagrammatic scheme of recent internal structure Source: Aryeetey Attoh (2010).

PAGE 319

319 Figure 5 5. Sprawled residential developments in the peri urban areas of Accra, Ghana Source: Aryeetey Attoh (2010) Figure 5 6 Trends in FDI inflows to Ghana (US$) 1993 2007 construct base d on UN (2012) data.

PAGE 320

320 Figure 5 7. Growth in FDI projects in Ghana, (first half year), 2003 2007 Source: Adapted from GIPC (2008)

PAGE 321

321 Figure 5 8 Sectorial distribution of employment to be generated by FDI inflow in the 2nd quarter of 2011 (1s t April 30th June, 2011 ) Source: Adapted from GIPC (2011)

PAGE 322

322 Figure 5 9 Geographical distribution of FDI projects among the regions of Ghana, 2 nd quarter 2011 Source: GIPC (2011)

PAGE 323

323 Figur e 5 10. FDI induce d peri urban land conversion and loss of peri urban ecological resources in a ( hypothetical ) city ( A) Depicts the dynamics of urban and agricultural land use values in the face of increasing FDI inflows into the urban space economy. (B) Sh ows the translation of land use value dynamics in to peri urban land consumption in Source:

PAGE 324

324 Table 5 1. Spatial expansion of Accra and annual growth rate, 1985 2002 Label Year Total area Growth in period Yearly growth in period Urban 1985 216 Urban 1991 276 60 10 Urban 2002 555 279 25 Data s ource: Moller Jensen et al. (2005) Table 5 2 Portions of regression estimation results in Chapter 4 Variable CityPOP 20 00 Co eff .787 S.E .040 t 19.458 Sign ificance .000 Confidence interval Lower B .707 Upper B .867 GDP 1999 .215 .046 4.645 .000 .123 .307 FDI 1999 .087 .032 2.763 .007 .025 .150 REMIT 1999 .134 .041 3.245 .002 .217 .052 c onstant .959 .417 2.301 .023 1.786 .132 N=107 F=121 (.000) R 2 = 820 S.E =.238 Durbin Watson = 2.096 Note: Italicized variables and corresponding values were not implemented in IUSEEM Table 5 3 IUSEEM estimate d built up are a of Accra (2001 2011) : FDI coefficie nt alone with 80 percent FDI concentration Year 1999 2000 Annual FDI flow (current US$) 243700000 1 165900000 Relative Change in FDI flow (f) Adjusted FDI Co eff (c) c(1+ ) = exponent ( ) Input ( 0 ) Estimated Built area ( ). (km 2 ) Annual change ( km 2 ) 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 89300000 58900000 105400000 139270000 144970000 636000000 855400000 1220400000 1684740000 2527350000 0.319245 0.63357 0.75831 0.5675 0.42852 0.40513 1.609766 2.510053 4.007796 5.913172 9.370743 0.000696 0.000696 0.000696 0.000696 0.000696 0.000696 0.000696 0.000696 0.000696 0.000696 0.000696 0.000474 0.000255 0.000168 0.000301 0.000398 0.000414 0.001816 0.002443 0.003485 0. 004812 0.007218 328.00 328.16 328.24 328.29 328.39 328.52 328.66 329.26 330.06 331.21 332.81 328.16 328.24 328.29 328.39 328.52 328.66 329.26 330.06 331.21 332.81 335.22 0.16 0.08 0.06 0.10 0.13 0.14 0.60 0.81 1.15 1.60 2.41 Total change (2001 2011) 7.22 Ave change 0.66 FDI coef ficient = 0 .087 (80% = 0.0696). 1. This value will remain the base year (point of reference) for all IUS EEM computations in Chapter 5

PAGE 325

325 Table 5 4 IUSEEM sensitivity to v arying FDI concentrations ( FDI co efficient only ) : 2001 2011 FDI Con centration Level (%) 100 90 80 70 60 50 40 30 20 10 Cumulative Built up Area (km 2 ) 337.05 336.14 335.22 334.31 333.40 332.50 331.59 330.69 329.79 328.89 Total Built up Area (km 2 ) 9.05 8.14 7.22 6.31 5.40 4.50 3.59 2.69 1.79 0.89 Average Built up Area (km 2 ) 0.82 0.74 0.66 0.57 0.49 0.41 0.33 0.24 0.16 0.08 Table 5 5 IUSEEM estimate d built up are a of Accra (2001 2011) : FDI, GDP per capi t a and population coefficients with 80 percent FDI concentration Year 1999 2000 Annua l FDI flow (current US$) 243700000 1 165900000 Relative Change in FDI flow (f) Adjusted Sum of Co eff. (c) c*(1+f) = exponent (g) Input (B 0 ) Estimated Built area (B). (km 2 ) Annual change ( km 2 ) 2001 2002 2003 2004 2005 2006 2007 2008 20 09 2010 2011 89300000 58900000 105400000 139270000 144970000 636000000 855400000 1220400000 1684740000 2527350000 0.319245 0.63357 0.75831 0.5675 0.42852 0.40513 1.609766 2.510053 4.007796 5.913172 9.370743 0.008712 0.008712 0.008712 0.008712 0.0 08712 0.008712 0.008712 0.008712 0.008712 0.008712 0.008712 0.005931 0.003192 0.002106 0.003768 0.004979 0.005183 0.022736 0.03058 0.043628 0.060228 0.09035 328.00 329.95 331.01 331.70 332.96 334.62 336.36 344.09 354.77 370.59 393.60 329.95 331.01 331.70 332.96 334.62 336.36 344.09 354.77 370.59 393.60 430.81 1.95 1.05 0.70 1.25 1.66 1.74 7.73 10.68 15.82 23.00 37.21 Total change (2001 2011) 102.81 Ave change 9.35 Coefficients: FDI = 0.087, P opulation = 0.787, GDP = 0.215 1 = reference value Source:

PAGE 326

326 Table 5 6 IUSEEM results sensitivity to vary ing FDI concentrations (all coeffici ents) 2001 2011 FDI Con centration Level (%) 100 90 80 70 60 50 40 30 20 10 Cumulative Built up Area (km 2 ) 461.20 445.74 430.81 416.37 402.42 388.94 375.91 363.31 351.14 339.37 Total Built up Area (km 2 ) 133.20 117.74 102.81 88.37 74.42 60.94 47.91 35.31 23.14 11.37 Average Built up Area (km 2 ) 12.11 10.70 9.35 8.03 6.77 5.54 4.36 3.21 2.14 1 .03 Table 5 7 Five year averages for the regression expl anatory variables 2007 2003 2001 1997 1996 1992 1990 1986 1984 1980 855380000 636010000 144970000 139270000 136750000 382476000 89300000 165900000 243 700000 167400000 81800000 149620000 120000000 106500000 233000000 125000000 22500000 121400000 14800000 15000000 5000000 4700000 4300000 8760000 2000000 2400000 16300000 16263752 15600000 10512750 16.07990* 10.54788* US$) 1090.05 920.61 794.76 673.91 593.47 814.56 433.85 416.65 660.88 655.88 618.53 557.16 637.49 609.57 527.25 593.99 705.01 614.66 674.81 584.76 594.53 558.90 532.85 589.17 486.81 447.01 466.65 488.48 476.49 473.09 0.05* 0.30* opulation Gr owth Rate 2.41 2.42 2.43 2.44 2.44 2.43 2.41 2.37 2.35 2.36 2.42 2.38 2.51 2.64 2.75 2.83 2.85 2.72 2.75 2.69 2.69 2.80 2.98 2.78 3.45 3.51 3.31 2.92 2.39 3.12 0.14* 0.13* Notes: I talicized values = period average s is the r elative change with re ference to ave rage values in column 4 ( 1990 1986).

PAGE 327

327 Table 5 8 Geographical distribution of FDI projects among the regions of Ghana, 2 nd quarter of 2007 Region No. of Projects Percent of Total Asha nti 4 4.21 Brong Ahafo 1 1.05 Central 3 3.15 Eastern 4 4.21 Greater Accra 81 85.26 Northern 1 1.05 Volta 3 3.15 Western 2 2.10 Total 95 100.00 GIPC (2008 ) data. Table 5 9 Comparing IUSEEM results with Angel et al. and Moller Jensen et al. (FDI coefficient alone) varying FDI concentration levels FDI Con centration Level (%) 100 90 80 70 60 50 40 30 20 10 Average built up area IUSEEM Average as percent IUSEEM 0.82 0.74 0.66 0.57 0.49 0.41 0.33 0.24 0.16 0.08 Angel et al. (2011) 13.30 13.30 13.30 13.30 13.30 13.30 13.30 13.30 13.30 13.30 Moller Jensen et al. (2005)* 17.5 17.5 17.5 17.5 17.5 17.5 17.5 17.5 17.5 17.5 Angel et al. (2011) 6.20 5.58 4.95 4.32 3.70 3.08 2.46 1.84 1.23 0.61 Moller Jensen et al. (2005) 4.70 4.23 3.75 3. 28 2.81 2.34 1.87 1.40 0.93 0.46

PAGE 328

328 Table 5 10 Comparing IUSEEM results with Angel et al. and Moller Jens en et al. (all estimated coefficients ) varying FDI concentration levels FDI Con centration Level (%) 100 9 0 80 70 60 50 40 30 20 10 Average built up area IUSEEM Average as percent IUSEEM 12.11 10.70 9.35 8.03 6.77 5.54 4.36 3.21 2.14 1.03 Angel et al.(2011)* 13.30 13.30 13.30 13.30 13.30 13.30 13.30 13.30 13.30 13.30 Moller Jensen et al. ( 2005)* 17.5 17.5 17.5 17.5 17.5 17.5 17.5 17.5 17.5 17.5 Angel et al. (2011) 92.86 82.06 71.63 61.56 51.83 42.43 33.34 24.57 16.10 7.91 Moller Jensen et al. (2005) 70.40 62.21 54.31 46.67 39.29 32.16 25.28 18.63 12.20 6.00 Tab le 5 11 C onfidence interval for IUSEEM at the 80% FDI concentration: 2001 2011 Variable Pop GDP FDI Confidence interval Lower Bound 0.707 0.123 0.025 Upper Bound 0.867 0.307 0.15 Sum 80% of sum Decima l 0.855 0.684 0.00684* 1.324 1.0592 0.010592* Cumulative estimated Average estimated 406.29 7.12 456.92 11.72 Notes: *These values are used as the exponent in the IUSEEM to compute the lowest and highest expected cumulative as well as the a verage annual growth rate in built up area for 2001 2011.

PAGE 329

329 Table 5 12 IUSEEM results compared to independent data Year 1985 1991 2000 2001 2002 2011 Annual growth IUSEEM 80% FDI con centration 337.35 430.81 9.35 Moller Jensen et al. (2005) 216 276 555 712.50 17.50 Angel et al. (2011; 2005)* 129 328 473.93 13.27 IUSEEM results as percent Cumulative Upper Bound Lower Bound 60.46 1 64.13 57.02 90.90 96.41 85.73 Average Upper Bound Lower Bound 53. 41 2 66.97 40.67 70.45 88.34 53.65 Notes: *This column is provided for further information only. The built up are a for 2011 from Moller Jensen et al.(2005) was calculated based on estimated annual average growth of 17.50sq.km [(10sq.km+25sq.km)/2] since 20 02. 1 = 430.81/712.50*100 2 = 9.35/17.50*100

PAGE 330

330 CHAPTER 6 GENERAL CONCLUSIONS Summary This dissertation research set out to expl ore the theoretical underpinnings and methodological approach to understanding and quantifying the linkages between FDI and remittances on the one hand, an d rapid land conversion in the peri urban areas of Accra, Ghana The major outcomes are three fold : ( 1 ) The study proposes the FDI induced urban spatial expansion theory as an extension to the Thu nen Alonso theoretical framework from a geographical perspective ; aimed at explaining rapid rate of land conversion in the peri urban areas of primate, globalizing cities in developing countries as well as to help better account for realities of economic g lobalization in urban expansion theorizing and modeling ; (2) The integrated Urban Spatial Expansion Estimation Method (IUSEEM) is proposed as an appropriate method for quantifying FDI i nduced urban spatial expansion; and ( 3 ) The study shows that articulat ing FDI induced urban spatial expansion through integration of geographical concepts and themes and economic theory helps place the phenomenon in the broader context of human induced environmental change, an increasingly important societal concern. By thes e outcomes this dissertation has sought to lay down the conceptual and theoretical foundations, as well as the methodological framework for analyzing contemporary urban la nd expansion as induced by FDI, facilitates a theoretically grounded geographical ana lysis of urba n land expansion induced by FDI; with implications for urban planning, policy and environmental conservation planning. The premises upon which the dissertation is based are as follows. Globalization related economic forces (FDI and remittanc es) are assuming increasing importance,

PAGE 331

331 impacting expansion of built up areas, especially in primate cities in developing countries where these funds show tendencies to over concentrate. However for the most, part global economic forces are inadequately represented in urban land use theorizing and urban expansion modeling approaches, indi cating a less than fuller understanding. land, the potential negative consequences on urban ecologic al resources of rapid and uncontrolled ur ban land conversion has been underscored by many researchers. For example, Karen Seto and colleagues have st ated in no uncertain terms that irreversible human impacts on the global biosphere. It drives the loss of farmland, affects local climate, fragments habitats, and threatens biodiversity (Seto et al., 2011: 1). Since urban land conversion responds to global capital flows into urban spa ces, it is imperative that a more comprehensive understanding is sought in order to ensure sustainable urban development. In this regard, the overall goal of this dissertation has question number 7: How is the movement of people, goods, an d ideas transforming the world? (NRC, 2010: 75). urban land use (Alonso, 1964). P ropositions developed were tested using log transformed mult iple regression approach (OLS regression estimation method ), for a sample of 107 cities. FDI and remittances ; to gether with city population and GDP per capita explained 82 percent of the variatio n in urban built up area More importantly, FDI was found t o be positively correlated with urban built up area and stat istically significant at the 95 percent confidence level; a 10 percent increase in FD I corresponded to almost 1 percent (0.087 %) increase in the urban built up area.

PAGE 332

332 Apparently small in size, this FDI co efficient is non trivial considering that FDI can multiply dramatically within a very short period of time ; and that FDI affects urban expansion through multiple pa thways. Surprisingly, however, the coefficient for remittance was negatively signed contrary to the conceptual and theor etical postulates in this work as well as evidence fro m the literature Consequently, remittance was excluded in the subsequent analysis. To better understand the spatio temporal dynamics of the FDI urban la nd convers ion relationship, the Integrated Urban Spatial Expansion Estimation Method / Model (IUSEEM), was formulated to estimate the extent of FDI induced urban spatial expansion in Accra during 2001 2011 The IUSEEM draws on the Allometric growth principle (or the Allometric Law) and integrates regression analysis with exponential growth modeling. Formulated a s an exponential function, the A llometric law can be linking function to fo rm; and in this case linking economic functions to urban form through time. IUSEEM was implemented using processed remote sensing data for Accra; with preliminary results indicating that the method could be useful for estimating d conversion as induced by FDI. In particular, I USEEM results compared to data obtained from Moller Jensen et al. (2005), seem to confirm that FDI induced urban expansion exhibited non linear and exponential characteristics. According to the prelim inar y results about 54 percent of the annual average growth in built up area of Accra during 2001 2011 could have been induced by FDI inflows to Ghana (with 80% concentration in Accra ) For 2011 alone IUSEEM results represent ilt up area of Accra based on data from Moller Jensen et

PAGE 333

333 al. (2005). In all FDI may account for at least 5 0% of the built up area increase of Accra during 2001 2011. Even though effort in this dissertation research may be modest, the author envisages cont ributions in the areas of geographical theory construction, urban spatial expansion modeling and estimation methodological development as well as trans disciplinary research ; designed to better understand human induced urban environmental change in general Moreover, by seeking to draw attention to the increasing importance of FDI as a potent driver of urban land conversion in an era of unprecedented global economic integration, insights from this dissertation research may have practical implications for urba n planning and policy and could aid urban environmental conservation planning. Contributions of t his Dissertation The major contributions of this dissertation are in the areas of theory co nstruction, model building and methodological development In tegrated Geographical Theory Construction Clarion calls on geographers to focus on theory development have been numerous especially after the so called quantitative revolution in geography. So important was geographical the orizing to Harvey that he wrote W ithout theory we cannot hope for controlled, consistent, and ra tional explanation of events It seems to me, therefore, that theory construction on a broad and imaginative scale must be our first priority in the coming decade (Harvey, 1969: 486). Ind eed Harve y concluded his classical book Explanation in Geography by saying : 486). This, to a very large extent,

PAGE 334

334 has been the motivational message for this dissertation research ; attempting to link processes of economic globalization to peri urban environmental / land degradation through loss of natural areas in the peri urban areas of a ra pidly globalizing primate city in a developing country setting. A formal the ory in this area of increasing societal importance as have been noted by researchers ( for example, Seto et al., 2011; NRC, 2010) will increase our understanding about human induced environmental change in gener al The advent of economic globalization with attendant almost seamless flow of financial resources across locations, is contributing to the profound changes Planet Earth is going through. The complexity of the issues cannot be overemphasized. However, human geographical research, particularly in the sub disciplines of economic and urban geography, with its integrative nature, is well situated to contribute to solving some of these problems, especially those that border on the interface between the socio economic subsystem and the biophysical system. In seeking to contribute to this course it is important that one does not need to re invent the wheel, neither should one adhere to rigid disciplinary boundaries, hence the decision to build on the latest ex tensions of Alonso (1964) economic theory of urban land use as well as adopting an empirical law from biological sciences as the basis for the urban spatial expansion model proposed in this study. While being fully aware about the dangers of this c walk theories and concepts relevant to the view advanced in the work Indeed, throughout, the author disser tation is not new to geography. With reference to the use of economic concepts

PAGE 335

335 and theories for developing geographical theories, for example, geographers have had many luminary advocates (Harvey, 1969; Burton, 1963 ; cited in Harvey, 1969 just to mention a few). According to Harvey (1969): Central place theory provides just one example out of many to demonstrate how geographical theory may be derived from the basic postulates of economics. The existence of such postulates was undoubtedly an important ne cessary condition for the emergence of a theoretical human geography. (p. 119). Burton was even more e one role of an economic geographer is to refine and adapt avai la (Burton, 1963: 159; cited in Harvey 1969: 119). It is the fervent hope of this investigator that further refinement of the globalization relevant theoretical framework presented in this dissertation research will aid the geographically motivated theorization of contemporary spatial expa nsion in cities where external funds may be over concentrated, since the emergence of globally induced urban ex pansion can longer be ignored. Consequently, the study has proposed the theory of FDI induced urban spatial expansion : In a globalizing primat e city located in a developing country, the rate of physical expansion of the city is a function of the volume of annual FDI inflows; during the expansion process, peri urban land is put to its highest value and best use by converting from agricultural to urban use, thereby accelerating the rate of urban spatial expansion as FDI inflows increase over time. Urban Spatial Expansion Modeling and Estimation M ethodology Most modern spatially explicit urban expansion modeling approaches use very complex methods w hich are usually computationally intensive, and may place them beyond the means of the average curious mind that has an interest in modeling urban growth. Furthermore, many of these models are not linked to any particular economic

PAGE 336

336 theory which could affor d a deeper understanding of the underlying economic drivers which usually are the most important factors i nfluencing activities and decisions of urban residents (Pacione, 2009) In arguing for the adoption of modern approaches such as Cellular Automata fo r modeling urban spatial expansion advocates have often cited the inadequacy of traditional models including being static, linear and unrealistic assumptions (Batty, 2005), for example However, these so called modern models are yet to deliver on a lot o f the promises. It is argued in this study that urban spatial expansion can be modeled analytically, thereby simplifying the modeling process. Therefore, t he methodological ap proach suggested in this work t he IUSEEM) appears to have the potential to c ontribute to making urban spatial expansion modeling simple, analytically tractable, dynamic and non linear. Moreover, t he IUSEEM is f irmly grounded in the Thunen Alonso theoretical framework which is a positive development; given that most modern urban m odeling approaches lack standard e conomic theoretical grounding. The building of this model, it is hoped, would have demonstrated that with the ability to connect concepts and ideas across disciplines one does not necessarily have to limit his/her scope in developing spatially explicit urban expansion models to only computationally intensive and methodologically complex simulation methods In short, the IUSEEM has attempted to lay the foundations for building a more realistic, simple and theoretically i nformed model for monitoring the impact of FDI on peri urban nat ural areas over time, potentially useful for studying contemporary urban spatial expansion such as Accra in Ghana More importantly, however, it is hoped that the urban spatial expansion method/ model (IUSEEM) developed in this dissertation would be coupled with GIS to contribute to

PAGE 337

337 efforts aimed at making GIS analyses mo re dynamic and policy relevant. The need for dynamic GIS has been underscored b y researchers. In particular, Kavouras (2001) has lam ented that Most GIS are restricted to static snapshots of spatial data and have difficulties with the management of time series People are interested in change but current GIS remain unable to ana lyze changing data. Future GIS must be able to store, retrieve, analyze and pres ent time varying spatial data. (Kavouras, 2001 ). Commendable efforts are on going but the approach advanced in this dissertation research may help simplify and may speed up t he process as well largely due to its simplicity ; with specific application in FDI induced urban spatial expansion analysis An added advantage is that it would be grounded in standard economic theory of urban land use / land value. Thinking Across Geogra phical Scales and Disciplines Li to local / regional spatial and environmental changes as well as synthesizing ideas and concepts across disciplines seem to offe r a more relevant approach to studying contemporary urban la nd expansion, with the potential to helping better understand changes in the coupled socioeconomic biophysical environments such as FDI induced environmental degradation Geographers, in particular, have emphasized the need for such skills for contemporary problem solving especially in the area of human induced environmental change as global economic impulses impact local places (AAG, 2003). This study has sought to utilize ideas, concepts and theories from urban geography, economic geo graphy; and urban ec onomics; and to some extent, ecological economics as well as principles from biological science, to conceptualize, theorize and model and quantify contemporary urban spatial expansion as influenced by external economic forces FDI spe ci fically.

PAGE 338

338 C onceptual izing, theorizing, modeling and quantifying the urban spatial impacts resulting from global scale economic forces such as FDI flows necessitated the capability of thinking across geographical scales. This ability has, hopefully been demonstrated throughout the pages of this dissertation. Interd isciplinary research is widely recognized as an effective approach to solving societal problems especially those that result from human environment interactions. Consequently, geographers skilled in the integrative way of thinking demonstrated in this dissertation will be equipped to contribute to the broad academic community, to the specific discipline of geography as well as engaging in policy relevant research on issues bordering on human environment relationships In addition to the above contributions, it is hoped that insights contained within the pages of this dissertation would help inform policy in the practical realm, especially in the area of urban planning and policy as well as urban environmental conserva tion planning Implications of this Dissertation R esearch Urban Planning and P olicy Just as the case with geography, an all encompassing definition has so far eluded urban planning, at least t o an extent (Bayer et al., 2010; Hall & Tewdr Jones, 2010; Campb ell & Fainstein, 2003). However, elusive definition notwithstanding, many urban planners will share the assertion by Bayer and colleagues that One of the distinguishing characteristics of urban planners, no matter what their approach to planning, is their focus on the future No one can predict what the future will bring. But planners are in the business of attempting to anticipate and recommend strategies to help cities thrive as the future unfolds. (Bayer et al., 2010: 13).

PAGE 339

339 Michael Pacione even pu ts the t ask a littl rban Planning and urban policy are concerned with the management of (Pacione, 2009: 164)). It is precisely the focus o that it is anticip ated that research findings and arguments advanced in this work could be of some use to urban planners and urban policy makers who are concerned with the proper land use and environmental planning of the ever changing urban system. From a very simplified perspective, the main subject matter of this dissertation can be described as urban spatial expansion as induced by economic forces, usually operating at a larger scale but producing spatial change in localities in the urban setting. This work reveals that expansion such as population and incomes ( say, GDP per capita) which have been identified consistently by researchers as contributing significantly to urban land expansion (for example, Seto et al., 2011; Ange l et al. 2011 ) ; may be masking underlying economic factors which are not too obvious and most likely induced by global economic forces such as FDI. In this regard, th e conceptual framework (Chapter 2; Chapter 3) revealed that FDI and remittances, through job creation, may be a major source of the purchasing power afforded to city residents to enable them express housing and land, in the absence of densification, will tran slate into conversion of new agricultural/natural lands usually in th e peri urban areas of the city. Thus, the conceptual and theoretical analyses in this study should help draw att ention to the of population increase and GDP growth, which may prompt urban land use planners and environmental managers to look beyond the population numbers

PAGE 340

340 and GDP statistics; and to focus more attention on external economic forces like FDI, which can multiply so fast in a very sho rt time as has been show n in the pages of this dissertation In advanced countries, changing taste (affluence) for housing consumption may be a major factor of increasing urban la nd expansion sprawl (Carr & Zwick, 2007 ; Burchtel et al., 2005), with the affluence usually result ing from economic growth of the country. But the same argument may not be valid for a primate city in developing country such as Accra, with consistent concentration of about 80% of all FDI funds inflow to Ghana ; the spatial ramifications can be profound. In this regard FDI may well be the single most important source of increasing purchasing power to influence land demand for urban uses. Consequently, if economic globalization generates forces that appear to impact urban change in these primate cities u rban planners and urban policy makers would be better served if they paid heed to this emerging phenomenon of externally induced urban spatial expansion and its characteristic tendency to cause loss of natural areas in the peri urban areas. A formidable challenge urban planners may face in a rapidly globalizing primate city is the dilemma posed by market forces and national/urban policy. Liberalization and Free market dictates would seek to enforce the operation of market forces perhaps at the expense o f urban policy, state activities that seek to influence the distribution and operation of investment and consumption processes in The conflicting goals are not too difficult to see. Thus, it may not be easy to put policy to practice. Hence, the need for policy to be informed by verifiable research findings like this dissertation has sought to do. In a nutshell it is suggested that urban planners and urban policy makers of rapi dly

PAGE 341

341 globalizing cities in developing countries, especially in Sub Saharan Africa should watch trends of FDI inflows into their country and how much of it goes to the city they are planning and seeking to ll communities are c oncerned 17). The earlier those entrusted with this task start looking beyond the national boundaries the better for the city residents. For, i n a rapidly global izing world, with seamless flow of global financial resources into geographically restricted urban spaces, urban land expansion can be pretty quick, even though the overall economic health of the country may not be all that impressive. One way for urban p lanners and policy makers to be able to monitor the external influences on the urban land expansion process is to implemen t the IUSEEM, formally developed in Chapter 5 of this dissertation, which will give the estimate of urban land cover in any given year to afford a in response to FDI inflows IUSEEM will be easy to implement and provide analytically tractable results, as well as steeped in urban economic theory. A related suggestion is that urb an land use planners must begin to pay adequate attention to external economic influences in the advent of globalization. More specifically, and with particular relevance to primate cities in developing country setting; policy interventions that target ur ban population control / management in lieu of urban expansion should be augmented with effective FDI management policy otherwise the triggers associated wi th FDI as explained in this dissertation will likely undermine their efforts. In view of the fore going, this dissertation research suggests a comprehensive FDI management in order to ensure a more geographically diversified FDI distribution

PAGE 342

342 into smaller but economically viable urban centers. Of course, such an approach may come at the expense of bene fits of econo mic agglomerations, perhaps; but this makes it even more imperative that a more complete understanding in this area of research is sought for which several questions may be raised. For example, what levels of FDI can urban lands, likely associated with FDI overconcentration? What proportion of land demand can be attributed to FDI related activities? How best can FDI be managed to ensure a more geographically diversified distribution among urban centers without undermining the benefits of agglomeration and economic linkages? In short, for those the urban economy, it may be worthwhile to consider these questions. In addition to the foregoing FDI induced urban land expansion; especially the appropriation of peri urban natural lands into urban uses may pose a formidable challenge for urban environmental conservation planning Urban Environmental C onservation So far in this dissertation, it appears that rapidly increasing levels of FDI inflows, w hen significantly concentrated in an urban location, drive urban land value s up, increases the opportunity cost of holding lands in a gricultural and na tural uses in peri urban areas; thereby boosting rapid land conversion to urban uses. Consequently, FDI inflows into in developing countries, with the tendency to cau se urban land values to rise, thereby ca using agricultural and natural lands to incur increasing opportunity costs ; may not bode well for urban environmental conservation. Certainly, the debate about economic growth and development in relation to environmental degradation is still on going, gat hering momentum as our knowledge

PAGE 343

343 advances. However, there appear to be a general consensus among researchers and conservationists alike that land conservation is one effective means through which the environment could be preserved, underscoring the need f or protected areas even in the city T he International Union for the Conservation of Nature (IUCN) defines a protected area as A clearly defined geographical space, recognised, dedicated and managed, through legal or other effective means, to achieve the long term conservation of nature with associated ecosystem services and cultural values (Dudley 2008 : 8) Thus, strictly speaking the key function of a protected area would be to conserve a piece of land and whatever resources found therein, regardless of whether they are of immediate utility to humans or not. However, the basic economic fact of life is that resources are limited relative to needs to be satisfied making choice inevitable. In this regard, any society or community that sets the land apa rt to conserve nature would have to be content with foregoing an alternative preference which could have been satisfied if the land had been put to some other usage. This is the basic underlying concept of what economists call opportunity cost an altern ative foregone as a result of It seems logical and imperative therefore, for members of the society to demand to know, directly or way or the other would deny t hem some form of benefits. Consequently protected areas elsewhere have been called upon, directly or indirectly through perceived efficient allocation of resources, to account for their existence in terms of their economic ( usually in monetary terms)

PAGE 344

344 benefits to society and to justify continued financial support for their maintenance and operations. Whereas the foregoing would hold true for all protected ar eas in general, those in eloping countries such as Accra in Ghana may be particularly vulnerable to increasing inflows of FDI since these funds accentuate the urban land conversion process as has been demonstrated in the pages of this dissertation. The important lesson that can b e drawn from this insight is that conserving urban lands would aid environmental quality and enhance societal well being, broadly defined. However, such urban environmental conservation efforts will more likely be associated with higher opportunity costs to society (Chapter 5). It is in this regard that the finding in this research likelihood of FDI induced increasing urban land values would suggest that urban environmental conservation efforts are more likely to be successful if the high opportunity costs of urban lands can be offset in some way. Hence, this dissertation suggests that urban environmental conservation planners be cognizant of the fact that even though FDI would bring numerous benefits to the city, it may also undermine conservation ef forts as market dictates would favor the conversion of natural lands to urb an uses. In the light of this, it is suggested that urban conservation planners concerned with rapidly globalizing primate cities in developing countries such as Accra, who may hav e to deal with FDI imperatives, may want to look into adopting methods of mixed uses for urban protected areas as opposed to the Planning urban parks and protected areas for mixed uses would likely yield enormous benefits to society both in terms of ecosystems services provision ; and

PAGE 345

345 pecuniary incentives such as revenues, including enhanced real estate value, recreation, sports, special events hosting, among others. Perhaps, this is one secret for th of the New York Central Park, located within an area of extremely high land values? For example, the amenity values of parks to urban residential areas have been well documented ( Cho et al., 2008; Geoghegan, 2002; Irwin, 2002 ; Bolitzer & Netu sil, 2000 ). This could be generally interpreted as contentment among residents of areas near urban parks hence their willingness to pay more for the adj acent real estate property. In the ca se of the New York Central Park, for example, real estate propert y values adjacent to the park are considerably higher than areas not in close proximity to the park (Central Park Conservancy, 2006). It is heartening to note, therefore, that u rban residents are not averse to natural areas preservation after all with ur ban conservation efforts presumably rooted in history Accordin g to Christensen and colleagues Cities have historically been the dynamic engines of conservation, as well as the engines of economic growth and innovation. Some of the best environmental his tory in recent years has focused on urban environments, including Matthew Gandy's Concrete and Clay: Reworking Nature in New York City (MIT Press, 2002), Jared Orsi's Hazardous Metropolis: Flooding and Urban Ecology in Los Angeles (University of California Press, 2004), Matthew Klingle's Emerald City: An Environmental History of Seattle (Yale University Press, 2007), and Michael Rawson's Eden on the Charles: The Making of Boston ( Harvard University Press, 2010) (Christensen et al., 2012). Indeed, the conce rn for conservation of lands in and around urban centers seems to be gaining renewed vigor For instance, Peter Kareiva, the Chief scientist for Nature Conservancy, is recently reported to have issued a warning t o the members, thus : Conservation is facin g a crisis of irrelevance it is an enterprise that is not urgent to most people If conservation is to build the support it needs, it must energize young urban dwellers, who now make up most of the world. The best way to get city people to care about cons ervation is to do

PAGE 346

346 conservation where they live, so that nature is seen as relevan t and connected to modern life. ( Christensen et al., 2012). biological entities but for our own good (so called modern human society). To sum up, FDI will more likely continue to flow into urban agglomerations, increasing urban land values, with the tendency to make urban environmental conservation economically unattractive. Thus, real ities of economic globalization and free market tendencies cannot be ignored; neither can they be stopped. But it seems a s and natural areas conservation, especially, for ecosystems services and sustena nce of biological diversity, can urban environmental conservation minded people will find something insightful in the pages of this dissertation Suggestions for Fur ther R esearch FDI, Remittances and Cost of L ow Income Housing in Accra Household income characteristics and spatial mobility of city residents result in of housing concept s uggests that as a housing unit ages and deteriorates its relative price will decrease, and it will become available to lower income families (Johnston, 1979 ). From a spatial mobility perspective, this has often entailed the movement of high income househo lds from the inner areas of the city to the peri urban areas where new houses and subdivision s are often created. Middle income earners then move into houses previously occupied by the high income city residents. Theoretically, the end result of this pro cess is that affordable houses of reasonable quality are made available

PAGE 347

347 to lower income households near the center of the city (Johnston, 1979) One potential problem that may stop this filtering process would be that if the composition of housing demand shows a significant proportion related to external forces such as FDI and remittances thereby placing the newly constructed housing units above the means of the average worker, residing and earning incomes in the local economy. The average Ghanaian (resid ent in Ghana and earning income locally) may not be able to afford the urban areas of Accra. And many researchers have pointed out the fact that most of these houses are owned by Ghanaian migrants or resid connections or earn supplementary incomes from oversea (subject to possible positive currency effect). This has contributed to the high cost and low quality of low income housing in Accra compared to similar cities in Africa has been confirmed by (Konadu Agyemang, 2001). In short it is widely acknowledged that the ordinary low income Ghanaian household in Accra may not be positively affected by the recent boom in residential housing development in the city. The conceptual and theoretical analyses presented in this dissertation seem to suggest that a significant share of the housing and land demand in Accra are related to FDI and remittances. With the new houses being put up mainly in the peri urban areas, at least indirectly, it seems the external economic forces of FDI and remittances may be blocking the house filtering process, thereby contributing to rising cost of housing but poor quality of housing for low income residents. Many factors have been identif ied as contributing to this situation, with one cited as the increasing housing demand by expatriates (Grant & Yankson 2003), largely resonating with the conceptual and theoretical analyses i n this dissertation However,

PAGE 348

348 there is more to this since accor ding to the conceptual and theoretical analyses, expatriates housing demand is only one of the many pathways through which FDI influence the housing and land demand. Further research will add to our understanding in this area. One question t hat readily co mes to mind is: in what specific ways is FDI affecting housing cost and quality fo r low income residents of Accra ? And: what is the rate of change of housing demand in Accra in response to inflows of FDI? Answers to these questions hopefully, will add to our understanding of the potential impact of FDI on housing of low income households, which could inform housing policy formulation. IUSEEM, GIS and Empirical Research Another area of this dissertation research where further investigations could yield en ormous benefits to the urban modeling community and GIS researchers and practitioners is that fur ther refinement of the IUSEEM (Chapter 5) will be useful. Specifically, interested researchers may want to put IUSEEM to test in other similar cities like Accr a in Ghana. It would also be interesting if the analytical model developed in this dissertation would be coupled with GIS to formulate the (IUSEEM GIS) to help make GIS analysis of urban expansion and modeling more dynamic, theoretically informed and an alytically simple. Particularly, if the formula can be programmed into GIS and results simul taneously displayed spatially, it will be a good step towards the development of dynamic GIS for urban expansion modeling which is grounded in the standard urban l and use theory. Unlike simulation methods, the modeler can incorporate external economic variables directly into the model to examine the The importance to urban planners in primate cities of the developing countries, of knowing the external economic influences on urban spatial expansion cannot be over

PAGE 349

349 globalization forces. Thus, from a practical point of view IUSEEM GIS may be useful for urban land use planning and conservation of urban nature once one can monitor the rate s at which socio economic drivers impact the process of urban spatial expansion, as the IUSEEM advocated in this dis sertation purports to do. Expanding the Scope of this Research Modeling Related to the foregoing is the potential for the expansion of the modeling exercise undertaken in this dissertation in order to enhance its policy relevance. In particular, probabili ty modeling capable of predicting differential expansion rates in different localities in Accra will be useful when policy makers are attempting to protect sensitive ecological areas in and around Accra. In this regard, it seems combining IUSEEM Markov c hain modeling and Cellular Automata (CA) will likely yield an integrated model of high practical relevance. As has been demonstrated in this work, the IUSEEM is able tr ansition of parcels of land from agricultural to urban. This shortfall can be remedied by the Markov chain model But Markov chain models do not detail the spatial change process in a spatially explicit manner (Lambin, 1997 cited in Adhikari, 2011) This c an be best handled by spatially explicit dynamic model s such as CA (Adhikari, 2011) Thus, if successful, the integration of IUSEEM, Markov chain model and CA should yield a more realistic model of urban spatial expansion as induced by FDI with due consid eration given to external economic drivers of urban spatial change which are likely to assume increasing importance should current globalization rates be sustained into the future. More importantly, from policy perspective, such an approach is likely able to help protect ecologically sensitive areas but based on a scientific approach that

PAGE 350

350 size fits ay not be necessary. Thus, policy will be targeted towards specific localities that can be identifi ed using this integrated modeling approach without unnecessarily restricting development in less ecologically sensitive peri urban areas Scale of Data Capture This dissertation calls for research into appropriate data for urban expansion modeling since t he scales at which the traditional variables (population and income (GDP per capita) are measured may not adequately capture the potential influences of FDI on urban spatial expansion (as revealed by the conceptual and theoretical analyses presented in thi s work). On the one hand, the high significance attributed to population change as land use change driver may be over bloated, given the fact that the purchasing power enablement is not considered. On the other hand, GDP per capita measured at the nation al level would not capture increased incomes enjoyed by city residents, via the creation of FDI related jobs attracted by the city. The combined efficient (resulting from the regression estimation in Chapter 4 of this dissertation research) is small compared to population and GDP per capita. Further research into how to capture the data that will help adequately reflect the influences of FDI on urban spatial expansion in prim ate cities of developing countries, undergoing rapid globalization is likely to be useful Remittances and Urban Spatial Expansion Anecdotal evidence, literature review as well as theoretical and conceptual analyses in this dissertation all pointed to a p ositive correlation between remittances and urban land expansion. Surprisingly, however, remittance was found to b e

PAGE 351

351 negatively correlated with increase in the urban extent and not statistically significant at the 95% confidence level. What is even more pu zzling is that when remittance is included with city population, GDP per capita growth rate and FDI in the log transformed regression, the overall explanatory power of the model improved slightly, but remittance co efficient continued to be negative and no t statistically significant. When urban built area was regressed on FDI and remittances, the R 2 for this model was 0.13, the same as for the model in which urban built area was regressed on FDI alone. For the co efficients, remittance remained negative, not statistically significant at the 95% but FDI co efficient remained positive and statistically significant at the 95% confidence level. Further research on the remittances urban spatial expansion relationship, especially, employing larger samples, may c onfirm ( or yield results consistent with ) the literature and the conceptual and theoretical postulates advanced in this dissertation thereby aiding understanding Some Limitations of this R esearch Data Data for this research were obtained from secondary s ources which are deemed reliable and relevant for our work. However, use of secondary data entails many challenges and difficulties, with the potential for inaccuracy. As such the validity and accuracy of the numerical results presented in this dissertatio n are subject to validity and accuracy of the source documents and databases. In future work, the author intends collecting primary data to be implemented in the regression models constructed in this work which can then be compares with the results obtain ed using the secondary data. Having said this, it should be noted that using secondary data did not have any adverse consequences on the logical reasoning presented in this dissertation. Thus, for

PAGE 352

352 a large measure, this work is of substance to the research community and would potentially contribute to the literature. Scale of Data Capture The FDI, remittances and GDP (explanatory variables) data used for the regression estimation and analysis were captured at the national level, whereas the urban built up a rea (respondent variable) data were captured at the city / urban level. In other words, the magnitude of the independent variables measured for the whole country were attributed to the selected cities in the sample. Thus, with the three explanatory varia bles are captured at the national (country) scale whereas the response c ities in question. It should be noted, however, that since by nature regression analysis Also, in regard to Accra, there is documented evidence that the bulk of annu al FDI funds inflow to Ghana get concentrated in this capital city of Ghana, with estimates ranging between 75% spatial expansion as asserted in this dissertation research may not be entirely out of place. But further investigation is warranted. Scope The scope for this research is limited to large /primate cities in developing countries that are receiving disproportionately large amounts of external funds. Additionally, these cities usually lac k effective urban land use planning and environmental policies, thereby failing to control urban development, especially in housing / residential development. Particularly, the model, IUSEEM is not intended for

PAGE 353

353 modeling spatial expansion in cities in deve loped countries where external funds inflow are likely to be more geographically diversified. In this case, enough demand the urban expansion process. For example, FD I inflows to the USA is more likely to be geographically diversified by virtue of the existence of considerably large number of Ghana where such funds have consistently flow ed disproportionately into its urban space. IMPORTANT NOTE: This study does not purport to tow any ideological line of thinking. For example, Marxist theorists have sought to blame capitalism for urbanization and attendant crises. However, valid this may be, this author does not necessarily share this approach, and neither is capitalism endorsed. The aim is to analyze the emerging phenomenon of FDI induced urban spatial expansion, utilizing the theories and methods deemed appropriate and effective should t hey come from Marxists or Capitalists. I aim neither to condemn globalization nor endorse it. It is simply not the call of my dissertation.

PAGE 354

354 APPENDIX A DATA USED FOR REGRES SION ANALYSIS Table A 1. Extracted Data for the Sample of 120 cities and urban ar eas City Accra Country Ghana T2 2000 Built Up T2 (Ha) 32834 City Pop T2 2693371 Addis Ababa Ethiopia 2000 11865 2510904 Ahvaz Iran 2000 21945 1258713 Akashi Japan 2001 3572 294657 Alexandria Egypt 1999 18780 3378392 Algiers Algeria 2000 22913 36279 12 Anqing China 1999 3552 578216 Ansan Korea, R 2001 8864 993560 Astrakhan Russian F 1999 16153 594015 Aswan Egypt 2000 1579 310069 Bacolod Philippines 2000 3294 510321 Baku Azerbaijan 1999 9015 2067017 Bamako Mali 1999 12992 1239935 Bandung Indone sia 2001 15382 3671064 Bangkok Thailand 2002 102593 9761697 Banjul Gambia 2000 4949 447985 Beijing China 1999 157638 11866221 Budapest Hungary 2000 36977 2052781 Buenos Aires Argentina 2000 135722 11915543 Cairo Egypt 2000 56917 13083621 Caracas Ven ezuela 2001 15673 1308279 Casablanca Morocco 2001 11431 3004505 Castellon Spain 2000 8203 268712 Cebu Philippines 2002 6365 1524080 Changzhi China 2000 11500 928518 Chicago USA 2001 425126 8590032 Chinju Korea, R 2000 5208 342454 Chonan Korea, R 200 0 5898 424046 Cincinnati USA 1999 77225 1517716 Coimbatore India 1999 10479 1399225 Dhaka Bangladesh 1999 16563 9196964 Fukuoka Japan 2001 37203 2598370 Gorgan Iran 2001 10711 455061 Guadalajara Mexico 1999 40518 3669578 Guangzhou China 2000 64106 7 156071 Guaruja Guatemala City Brazil Guatemala 2002 2000 3664 18903 277993 1766093

PAGE 355

355 Table A 1 (Continued) City Country T2 Built Up T2 (Ha) City PopT2 Harare Zimbabwe 2000 24310 889421 Ho Chi Minh City Vietnam 1999 21033 4309449 Hong Kong China 2001 9763 5179089 Houston USA 1999 182374 3656247 Hyderabad India 2001 31754 5707677 Ibadan Nigeria 2000 29996 2421369 Ilheus Brazil 2001 4970 151509 Ipoh Malaysia 2001 21906 655200 Istanbul Turkey 2000 53269 8826758 Jaipur India 2000 14084 2779119 Jalna India 2000 1952 24304 Jequie Brazil 2001 3655 138216 Johannesburg S. Africa 2000 99279 4695165 Kampala Uganda 2001 20169 1761733 Kanpur India 2000 17668 2674116 Kigali Rwanda 1999 4502 354273 Kingston Jamaica 2002 11945 875605 Kolkata India 2000 48344 13170280 Kuala Lumpur Malaysia 2001 80529 5389624 Le Mans France 1999 7444 21206 4 Leipzig Germany 1999 19580 664696 Leshan China 2001 9944 966091 London UK 2000 185551 10028978 Los Angeles USA 2000 385089 13218754 Madrid Spain 2000 36979 45 88170 Malatya Turkey 2000 13747 852864 Manila Philippines 2002 63317 17335085 Marrakech Morocco 2000 16475 1722999 Medan Indonesia 2001 14559 2239596 Mexico City Mexico 2000 105853 17224096 Milano Italy 2001 63517 3708980 Minneapolis USA 2001 109971 2167795 Modesto USA 2000 16892 344540 Montevideo Uruguay 2000 35725 1838801 Moscow Russian F 2002 104632 91446 24 Mumbai India 2001 37090 16161758 Ndola Zambia 2002 5390 326119 Oktyabrsky Russian F 2001 9786 228371 Ouagadougou Burkinafaso 2001 13746 874623 Palembang Indonesia 2001 17504 1616527 Palermo Italy 2001 8087 833992 Paris France 2000 148208 9519527 Philadelphia USA 1999 233491 5238892

PAGE 356

356 Table A 1 (Continued) City Country T2 Built UpT2 (Ha) City PopT2 Pittsburgh USA 1999 47213 1185877 Port Sudan Sudan 2001 4104 209154 Pretoria S. Africa 2000 17004 659834 Puna India 1999 19149 4041868 Pusan Korea, R 2000 19686 3485359 Rajshahi Bangladesh 2000 2026 599525 Ribeirao Preto Brazil 2001 10034 512239 Saidpur Bangladesh 2001 759 233478 San Salvador El Salvador 1999 12862 2 022047 Sanaa Yemen 2000 15528 1474635 Santiago Chile 2000 43851 5337512 Sao Paulo Brazil 200 0 155418 15481476 Seoul Korea, R 2001 70614 14546082 Shanghai China 2001 116186 14133931 Sheffi eld UK 2002 1 5880 764213 Shimkent Kazakhstan 2000 1 4565 577753 Songkhla Thailand 2001 1899 244403 Springfield USA 2002 2 3770 427126 Sydney Australia 2002 75951 2754486 Tacoma USA 2000 44 526 697104 Tebessa Algeria 2001 7537 457364 Teheran Iran 2000 47170 7803538 Tel Aviv Israel 2000 34022 2610373 Thessaloniki Greece 2001 7144 857935 Tijuana Mexico 2000 20808 1174193 Tokyo Japan 2001 256425 29615263 Ulan Bator Mongolia 2001 12857 776538 Valledupar Colombia 2001 2704 288448 Victoria Canada 2000 10735 255055 Vijayawad a India 2000 6234 1117042 Warsaw Poland 2002 33217 2002178 Wien Austria 2001 39450 2118871 Yerevan Armenia 2000 41569 2063290 Yiyang China 1999 9562 1207164 Yulin China 2000 25001 3387078 Zhengzhou China 2001 78954 5133266 Zugdidi Georgia 20 00 4134 157008 Source: Angel, S., Parent, J., Civco, D. L. and Blei, A. M. (2 010). Atlas of Urban Expansion, Cambridge MA: Lincoln Institute of Land Policy. Available at: http://www.lincolninst.edu/subcenters/atlas urban expansion/urban national data tables.aspx Last accessed: December 14, 2011. The d 1: Spatial Metri

PAGE 357

357 Table A 2 Socio economic dataset used for regression estimation Country GDP99 a FDI99 b Remit99 b FDI99 b Ave* Remit99 b Ave* Ghana 405 243700000 31000000 243700000 31000000 Ethiopia 116 69980000 34000000 69980000 34000000 Iran 1659 35000000 508000000 11666667 169333333 Japan 34535 12308393000 1109000000 4102797816 369666667 Egypt 1389 1065300000 3235000000 355100000 1078333333 Algeria 1617 291700000 790000000 145850000 395000000 China 892 38753000000 4796100000 3875300000 479610000 Korea, R 10015 9333400000 666000000 933340000 66600000 Russian F 1331 3309430000 1292000000 1103143333 430666667 Egypt 1389 1065300000 3235000000 355100000 1078333333 Philippines 1000 1247000000 691 8000000 415666667 2306000000 Azerbaijan 568 510317000 54000000 510317000 54000000 Mali 285 2178012 86000000 2178012 86000 000 Thailand 1983 6102677500 1460000000 3051338836 730000000 Gambia 648 49480000 7000000 49480000 7000000 China 892 38753000000 4796100000 3875300000 479610000 Hungary 4792 3307673090 213000000 3307673094 213000000 Argentina 7764 23987695600 64000000 23987696390 64000000 Egypt 1389 1065300000 3235000000 355100000 355100000 Venezuela 4056 2890000000 17000000 2890000000 17000000 Morocco 1395 2651866 19380000 00 1325933 969000000 Spain 15468 18523396100 4313000000 2058155051 479222222 Philippines 1000 1247000000 6918000000 415666667 2306000000 China 892 38753000000 4796100000 3875300000 479610000 USA 32689 289443000000 2748000000 28944300000 274800000 Korea, Rep 10015 9333400000 666000000 933340000 66600000 Korea, Rep 10015 9333400000 666000000 933340000 66600000 USA 32689 289443000000 2748000000 28944300000 274800000 India 442 2168591 100 11124000000 240954562 1236000000 Bangladesh 324 179662976 1807000000 59887657 602333333 Japan 34535 12308393000 1109000000 4102797816 36966666 7 Iran 1659 35000000 508000000 11666667 169333333 Mexico 5371 13696200000 6649000000 3424050000 1662250000 China 893 38753000000 4796100000 3875300000 479610000 Brazil 3420 28576000000 1862000000 5715200000 372400000 Guatamala 1489 154600000 466000000 154600000 466000000 China 892 38753000000 4796100000 3875300000 479610000 USA 32689 289443000000 2748000000 28944300000 274800000 India 442 2168591100 11124000000 240954562 1236000000 Nigeria 294 1004916740 1301000000 1004916719 1301000000 Brazil 3420 28576000000 1862000000 5715200000 372400000 Malaysia 3612 3895 263230 1042000000 1947631579 521000000

PAGE 358

358 Table A 2 (Continued) Country GDP99 a FDI99 b Remit99 b FDI99 b Ave* Remit99 b Ave* Turkey 3817 783000000 4529000000 391500000 2264500000 India 442 2168591100 11124000000 240954562 1236000000 India 442 2168591100 11124000000 240954562 1236000000 Brazil 342 0 28576000000 1862000000 5715200000 372400000 S. Africa 3012 1503332480 327000000 751666227 163500000 Uganda 281 140200000 233000000 140200000 233000000 India 442 2168591100 11124000000 240954562 1236000000 Rwanda 258 1725717 5000000 1725717 5000000 Jamaica 3467 52 3700000 790000000 523700000 790000000 India 442 2168591100 11124000000 240954562 1236000000 Malaysia 3612 3895263230 1042000000 1947631579 521000 000 Germany 26131 55906673000 4329000000 55906671344 4329000000 France 24080 45986861000 9301000000 22993429694 4650500000 China 892 38753000000 4796100000 3875300000 479610000 UK 25598 89337012000 3400000000 44668505172 1700000000 USA 32689 289443000000 2748000000 28944300000 274800000 Spain 15468 185 23396100 4313000000 2058155051 479222222 Turkey 3817 783000000 4529000000 391500000 2264500000 Philippines 1000 1247000000 6918000000 415666667 230600000 0 Morocco 1395 2651866 1938000000 1325933 969000000 Mexico 5371 13696200000 6649000000 3424050000 1662250000 Italy 21055 6942860300 1999000000 3471430178 999500000 USA 3 2689 289443000000 2748000000 28944300000 274800000 USA 32689 289443000000 2748000000 28944300000 274800000 Russian F 1331 3309430000 1292000000 1103143333 430666667 India 442 2168591100 11124000000 240954562 1236000000 Russian F 1331 3309430000 1292000000 11 03143333 430666667 Burkinafaso 263 7925951 80000000 7925951 80000000 Italy 21055 6942860300 1999000000 3471430178 999500000 France 24080 45986861000 9301000000 22993429694 4650500000 USA 32689 289443000000 2748000000 28944300000 274800000 USA 32689 289443000000 2748000000 28944300000 2748 00000 Sudan 314 370800000 665000000 370800000 665000000 S. Africa 3012 1503332480 327000000 751666227 163500000 India 442 2168591100 11124000000 240954562 1236000000 Korea, R 10015 9333400000 666000000 933340000 66600000 Bangladesh 324 179662976 1807000000 59887657 60233333 3 Brazil 3420 28576000000 1862000000 5715200000 372400000 Bangladesh 324 179662976 1807000000 59887657 602333333 Brazil 3420 28576000000 1862000000 5715200000 372400000 El Salvador 2108 215900000 1387000000 215900000 1387000000 Korea, R 10015 9333400000 666000000 933340000 66600000

PAGE 359

359 Tab le A 2 (Continued) Country GDP99 a FDI99 b Remit99 b FDI99 b Ave* Remit99 b Ave* China 892 38753000000 4796100000 3875300000 479610000 UK 25598 89337012000 3400000000 44668505172 1700000000 Kazakhstan 1119 1587000000 64000000 1587000000 64000000 Thailand 1984 6102677500 1460000000 3051338836 7 30000000 USA 32689 289443000000 2748000000 28944300000 274800000 Australia 22607 3311038460 2029000000 3311038586 2029000000 USA 32689 289443000000 2748000000 28944300000 274800000 Algeria 1617 291700000 790000000 145850000 395000000 Iran 1659 35000000 508000000 11666667 16933 3333 Israel 18607 3210900000 424000000 3210900000 424000000 Greece 12897 567300000 2284000000 567300000 2284000000 Mexico 5371 13696200000 6649000000 3424050000 1662250000 Japan 34535 12308393000 1109000000 4102797816 369666667 Mongolia 442 30400000 7000000 30400000 7000000 Colombia 2478 1507907070 1312000000 1507907130 1312000000 India 442 2168591100 11124000000 240954562 1236000000 Poland 4360 727000 0000 825000000 7270000000 825000000 Austria 26416 3009014270 1526000000 3009014242 1526000000 Armenia 597 122040000 95000000 122040000 95000000 China 892 38753000000 4796100000 3875300000 479610000 China 892 38753000000 4796100000 3875300000 479610000 a a GDP at current prices http://data.un.org/Search.aspx?q=gdp+per+capita GDP per capita is rounded up to the nearest whole number b FDI. Foreign direct investme nt, net inflows (BoP, current prices US$, 1999). Available at: http://data.un.org/Search.aspx?q=foreign+direct+investmen t Remittances: Workers' remittances and compensation of emp loyees, received (current US$ 1999). Available at: http://data.un.org/Search.aspx?q=remittances *Ave = Values for FDI and remittances were averaged based on the number of cities within a cou ntry that are captured in the sample of107 cities.

PAGE 360

360 Definitions : GDP per capita, FDI and remittances are as defined by the World Bank. Source Available at: http://data.un.org/_Docs/WDI%20definitions.pdf The Undata is essentially an internet search engine which allows the user to access databases stored and managed by the United Nations by performing queries. The original sample of 120 cities (covering 58 countries) in Table A1 was reduced to 107 cities (covering 51 countries) which were used in the regression estimation in Chapter 4. Thus, 13 cities were not included. The reason is that FDI and / or remittances data for t he countries in which the cities were located were not available. It must be noted, for example, that some countries do not report remittances flow into their countries. The 13 cities removed from the original sample of cities are: St Catharines and Victor ia (Canada), Singapore (Singapore), Santiago (Chile), Ho Chi Minh City (Vietnam); Ndola and Harare (Zimbabwe); Montevideo (Uruguay); Kuwait City (Kuwait); Bandung, Medan and Palembang (Indonesia); and Sanaa (Yemen). The Built Up Area is the area occupied by built up pixels within the set of ( http://www.lincolninst.edu/subcenters/atlas urban expansion/area metrics.aspx )

PAGE 361

361 APPENDIX B SUPPLEMENTARY MAPS Figure B 1. Location Map of the sample of 120 cities Source: Angel, S., Parent, J., Civco, D. L. and Blei, A. M. (2010). Atlas of Urban Expansion, Cambridge MA: Lincoln Institute of Land Policy. Availab le at: http://www.lincolninst.edu/subcenters/atlas urban expansion/global sample cities.aspx Last accessed on December 14, 2011. Note: This map is provide d to give an indication of the geographical distribution of the sample of the 120 cities upon which the regression analysis in this dissertation is based. Of particular importance to this research is the wide regional coverage of the dataset spreading ac ross all the continents.

PAGE 362

362 Figure B 2 Map of Accra, Ghana, showing main Central Business Districts (CBDs ). Source: Adapted from Grant & Yankson (2003).

PAGE 363

363 Figure B 3. Map of Accra (and part of GAMA) showing points of interest, including the only international airport and the Export Processing Zone (EPZ) Source: Adapted from Grant & Yankson (2003).

PAGE 364

364 APPENDIX C GENERAL NOTES ON DAT A SUITABILITY Two main issues are discussed briefly below, i. e. the nature of remote sensing land cover change data; and the use of secondary statistical data in this dissertation. Firstly, u rban land cover change data derived from satellite remote sensing are essentially categorical, based on some classification scheme. According to Donnay and Unwi 206). Thus, the nominal scale of measurement would appear to make such data not amenable to manipulation by many statistical and math ematical methods, especially in urban modeling, since according to Donnay and Unwin: U rban modeling frequently supposes the existence of quantitative variables that are spatially continuous while planning makes use of discretization of space into zones whi ch are very different f rom those given by image pixels. (Donnay & Unwin, 2001: 206). However, the data are usually transformed to suit such purposes. Indeed some authors have claimed that the classification of satellite remote sensing is largely motiva ted by the possibility of data transformation for further analyses. For example, Donnay and Unwin (2001) assert that: A primary aim of classifications of land use/ land cover data derived from satellite remote sensing is to provide data for a large range of urban models by a transformation which changes their measurement scale from nominal categor ies into ratio/interval scales (p. 207). One way to quantify such data has been suggested as grouping pixels according to attributes and computing the area(s) of polygon (s) correspondin g to each attribute (Donnay & Unwin, 2001). By so doing the discrete and nominal land cover change data are transformed into spatially continuous and interval data which are useful for most

PAGE 365

365 urban modeling (Donnay & Unwin, 2001). Thus, the use of processed remote sensing land cover change data in Chapter 4 and Chapter 5 of this dissertation is consistent with accepted urban modeling practices as evidenced in the literature. Secondly, u se of secondary data in research could pose f ormidable challenges, especially bordering on misalignment of original objectives of those who collect the primary data and those who use such data, often for purposes other than the originally intended. In view of this the decision to use secondary dat a in this dissertation was carefully considered, particularly examining the aims and objectives for which the data were collected and what they could be used for. In this regard, it is gratifying to note that the authors have asserted that: The main thrus t of our work has been a concentrated effort to measure different aspects of the spatial structure of a large number of urban areas the world over in a rigorous and consistent manner, in a way that can make it possible for us as well as for others to s t udy them across time and space (Angel et al., 2010 : online ). With this statement, it seems reasonable to expect that the land cover change data used in this dissertation have been sourced from re liable and authoritative source Indeed, the authors enco urage the use of their data by stating that: dimensions of urban expansion and its key attributes in cities the world over. The data and images are available for free downloading, for s cholars, public officials, and planners, those engaged in international development, and concerned citizens. The global empirical evidence presented here is critical for an intelligent discussion of plans (Angel et al., 2010, online). With this coming from researchers associated with the Linco ln Institute of Land Policy, a well known reputable organization, one can be reasonably confident about the

PAGE 366

366 validity of the secondary data. In a nutshell, the goal s for which the land cover change data were collected are in alignment with those of this dissertation as discussed in the text. It must also be noted that the only additional dataset on GDP per capita, FDI and remittances were source d from United Nations databases for similar arguments as the preceding apply. In addition to the above, this writer has extensive background knowledge of Accra, the study area, having been born and lived in Ghana all his adult life and watched the spatio temporal evolution of Accra first hand. It is with this background knowledge that the writer became curious about the rapidity of urban land conversion in recent times and suspected that external economic forces of globalization, FDI in particular, may be implicated. Having ob expand spatially with time, and armed with a deep understanding of our cultural and traditional practices as well as the socio economic processes in Ghana, the author believes he is well equipped to be able to analyze and discuss the results and issues pertaining to this dissertation in their appropriate socio cultural and environmental contexts.

PAGE 367

367 LIST OF REFERENCES Acosta, P. A., N.R Baerg & F.S. (2009). Financial Development, Remittances, and Real Exchange Rate Appr eciation. Federal Reserve Bank of Atlanta Economic Review N o.1. Adewuyi, T. (2011). Land Degradation in the Peri Urban Area. The Case of Kaduna Metropolis, Nigeria. Saarbrucken. LAP LAMBERT Academic Pub lishing GmbH & Co Addison, E.K.Y. (2004). The Macro economic Impact of Remittances in Ghana. Research Department, Bank of Ghana. (A Presentation). September 3, 2004. Available at: http://pdfcast.org/pdf/the macroeconomic impact of remittances in ghana Accessed: February 20, 2011 Adhikari, S. (2011). People, Park a nd Suburbanizati on: A Spatio Temporal Analysis of Change in and a round Bannerghatta National Park, India U npublished PhD Dissertation University of Florida, Department of Geography. Ajayi, S. I. (Ed). (2006). Foreign Direct Investment in Sub Saharan Africa: Origins, Targets, Impact and Potential. Nairobi. African Economic Research Consortium. Pp. 2 4. Alberti, M., Marzluff, J. M., Shulenberger E., Bradley, G ., Ryan, C., & Zumbrunnen, C., (2003). Integrating Humans into Ecology: Opportunities and Challenges for Studying Urban Ecosystems. Bioscienc e, Vol. 53 No. 12. Alonso, W. (1964). Location and land use: Toward a general theory of land rent. Cambridge, MA: Harvard University Press. Amuedo Dorantes, C. & Pozo, S. (2006). Remittances as insurance: evidence from Mexican Migrants. Journal of Population Economics, Vol, 19 pp. 227 254 (2006). Angel, S., Paren t, J., Civco, D., Blei, A., & Potere, D. (2010). A Planet of Cities: Urban Land Cover Estimates and Projections for All Countries, 2000 2050. Lincoln Institute of Land Policy Working Paper. Available at: https://www.lincoln inst.edu/pubs/download thankyou.asp?doc_id=1171 Accessed May19, 2011. Angel S., J. Parent, D. L. Civco & A. M. Blei, (2010). Atlas of Urban Expansion. Cambridge MA: Lincoln Institute of Land Policy. Available at : http://www.lincolninst.edu/subcenters/atlas urban expansion/ Last accessed on: December 07, 2011. Angel, S., Sheppard, S.C ., & Civco, D.L. (2005). The Dynamics of Global Urban Expansion. World Bank, Transport and Dev elopment Department. Washington, D.C.

PAGE 368

368 Aniya Masamu, C. M. (2010). Spatial Temporal Analysis and Modeling of Urban Growth. Spatial Analysis and Geo simulation of urban growth using Cellular Automata, Remote sensing and GIS. Saarbrucken. LAP LAMBERT Academ ic Pu blishing GmBH & Co. KG Antle, J. M. & Heiderbrink, G. ( 1995 ). Environment and Development: Theory and International Evidence. Economic Development and Cultural Change Vol. 43 No. 3 pp. 603 625. Aryeetey, A. S. ( 2010 ) Africa South of the Sahara. In Johnson, D. L., Haarmann, V. Johnson M. L. and Clawson D. L. (Eds ). World Reg ional Geography 10 th Edition. New York. Prentice Hall. Pp. 400 420 Asafu Agyei, J. ( 2003 ) Biodiversity Loss and Economic Growth: A Cross Country Analysis. Contempo rary Economic Policy Vol. 21, pp. 173 185. Asso ciation of American Geographers (AAG). (2003). Global Change and Local Places. Estimating, Understanding, and Reducing Greenhouse Gases. (AAG GCLP Team). Cambridge. Cambridge University press. Chapters. 1 3. Baabereyir, A. (2009). Urban Environmental Problems in Ghana: A Case Study of Social and Environmental Injustice in Solid Waste M anagement in Accra and Sekondi Takorad i. Unpublished PhD T hesis. School of Geography, University of Nottin gham. Pp. 84 102. Available at: http://etheses.nottingham.ac.uk/847/1/Full_thesis_pdf_copy.pdf Last accessed: January 17, 2012 Bank of Ghana. (2009). Monitoring Foreign Private Capital Flows in Ghana 2008 Survey Findings & Policy Issues. Bank of Ghana Policy Briefs, June 2009. Available at www.bog.gov.gh Batty, M. (2005). Cities and Complexity. Understanding Cities with Cellular Automata, Agent Based Mo dels, and Fractals. Cambrige. MIT Press. Chapter 1. Bayer, M., Frank, N., & Valerius, J. (2010). Becoming an Urban Planner. (American Planning Association). Hoboken. John Wiley & Sons, Inc. Bettencourt, L.M.A. & Kaur, Jasleen. (2011). E volution and structure of sustainability science. PNAS 19540 19545, Dec6, 2011 Biology Online (2008). Deterministic Model. Available at: http://www.biology online.org/dictionary/ Deterministic_model Last accessed: February 4, 2012 Biology Online (2008). Stochastic Model. Available at: http://www.biology online.org/di ctionary/Stochastic_model Last acc essed: February 4, 2012. Black, M., (1962). Models and Metaphors. Ithaca, New York. P. 267

PAGE 369

369 Blue, S.A. (2004). State Policy, Economic Crisis, Gender, and Family Ties: Determinants of Family Remittances to Cuba. Economic Geography Vol. 80 No. 1, pp. 63 8 2 Bond, P. (2006). Looting Africa. The Economics of Exploitation in Africa. London. Zed Books. Bolitzer, B. & Netusil. N. R. ( 2000 ) The impact of open spaces on property values in Portland, Oregon. Journal of Environmental Management Vol. 59, pp. 185 193. Braimoh, A. K. & Onishi, T. (2007). Spatial determinants of urban land use change in Lagos, Nigeria. Land Use Policy Vol. 24, pp. 502 515. Washington, DC. The Worl d Bank. Brueckner, J. ( 1987 ) The Structure of U rban Equilibria: A Unified T reatment of the Muth Mills M odel. In Mills, E. (Ed.). (1987). Handbook of Regional and Urban Economics. New York: Elsevier. Pp. 821 845. Brueckner, J. K., & Fansler, D. A. (1983). The Economics of Urban Sprawl: Theory and Evidence on the Spatial Sizes of Cities. The Review of Economics and Statistics Vol. 65, No. 3, p p. 479 482. Burchell, R. W., Downs, A., McCann, B. & Mukherji S ( 2005 ) Sprawl Costs: Economic Impacts of Unchec ked Development. Washington and London. Island Press. Pp. 38 47. Burton, I. (1963). The quantitative revolution and theoretical geography. The Canadian Geographer, Vol.7, pp.151 162 Cadwallader, M.T. (1996, 1985). Analytical Urban Geography. Spatial Patte rns and Theories. Englewood C liffs, Prentice Hall, Inc. Pp. 24 50. Campbell, S. & Fainstein, S. S. (2003). Readings in Planning Theory. 2nd Edition. Malden. Blackwell Publishing Ltd. Carmody, P. (2010). Globalization in Africa: Recolonization or Renaiss ance? Boulder. Lynne Rienner Publishers, Inc. Carr, M. H. & Zwick P. D. (2007). Smart Land Use Analysis. The LUCIS Model. Redlands. ESRI Press. Pp.1 8. Central Intelligence Agency (CIA). (Various years ). The World Fact book. https://www.cia.gov/library/publications/the world factbook/geos/gh.htm l Last accessed: January 20, 2012.

PAGE 370

370 Central Park Conservancy. (2006). Annual Report, 2006. Available at: http://www.centralpark.com Accessed: Dec., 10, 2007. Chami, R., Barajas, A., Cosimano, T., Fullenkamp, C., Gapen, M. & Montiel, P. (2008). Macroeconomic Consequences of Remittances. IMF. Washington D.C. Pp. 21 30. Champ, P. A., Boyle, K .J. & Brown, T.C. (Eds.). (2003). A Primer on Nonmarket Valuation. Boston. Kruwe r Academic Publishers. Pp. 1 158. Chatterjee, S. & Hadi, A. S. (2006). Regression Analysis by Example. 4 th Edition. New Jersey. John Wiley & Sons, Inc. Cho, S. H., Poudyal, N.C. and Roberts, R.K. (2008). Spatial analysis of the amenity value of green open space. Ecological Economics Vol. 66, pp. 403 416. Chorley, R. J. and Haggett, P. (1970). Socioeconomic models in geography. London. Methuen and Co. Ltd. Christ ensen, J., McDonald, R. and Denning, C. (2012). Ecological Urbanism for the 21 st Century. The Chronicle Review January 22, 2012. Available at: http://chronicle.com/article/ Ecological Urbanism for the/130384/ Accessed: February 4, 2012 Clapp, J. and Dauvergne, P. (2005) Paths to a Green World. Cambridge. The MIT Press. Coffey, J. W. (1981). Geography: Towards a General Systems Approach. London & New York. Methuen. Cooper F. (2001). What is the Concept of Globalization Good For? An African African Affairs Vol. 100, pp. 189 213. Costanza, R., Cumberland, J., Daly, H., Goodland R. and Norgaard, R. (1997). An Introduction to Ecologic al Economics. Boca Raton. St. Lucie Press. Pp. 1 175. Croitoru, L. & Sarraf, M. (Eds.). (2010). The Cost of Environmental Degradation. Case Studies from the Middle East and North Africa. Washington, D.C. The World Bank. Cruz, W., Munasinghe M. and Warford J. ( 1996 ) Environmental Implications of Economic Policies. Environment. Vol. 38, No. 5. Daly, H.E. (1991). Steady State Economics. Washington, D.C. Island Press. Defries, R., Gregory, S., Asner, P. and Houghton, R. ( 2004 ) Ecosystems and Land Use Change. W ashington D. C. American Geophysical Union. Pp. 1 25. Denisia, V. (2010). Foreign Direct Investmen t Theories: An Overview of the M ain FDI Theories. European Journal of Interdisciplinary Studies Issue 3, December 2010.

PAGE 371

371 Dicken, P. (2003). Global Shift. R eshaping the Global Economic Map in the 21st Century. New York and London. The Guilford Press. Chs.1 3. Diko, J., and Tipple, A. G. (1992). Migrants build at home: Long distance housing development by Ghanaians in London. Cities November, 1992. Dudley, N. (Ed). (2008). Guidelines for Applying Protected Area Management Categories. Gland, Switzerland. IUCN. pp. 7 24. Available at: http://data.iucn.org/dbtw wpd/edocs/PAPS 016.pdf Last acces sed: Feb. 04, 2012. Egan, M.T. (2002). Grounded Theory Research and Theory Building. Advances in Developing Human Resources 4:277 295. England, R. W. ( 2000 ) Natural capital and the theory of economic growth. Ecological Economics, Vol. 34, pp. 425 431. Faber, D. (Ed ). ( 1998 ). The Struggle for Ecological Democracy: Environmental Justice Movements in the United States. New York. The Guilford Press. framework. Journal o f Population Economics Vol. 7, pp. 235 245. Farvacque Vitkovic C., Raghunath, M., Eghoff, C. and Boakye, C. (2008) Development of the Cities of Ghana. Challenges, Priorities and Tools. The World Bank, African Region Working Paper Series No. 110, p p. 1 19. Frank, A., Raper, J., and Cheylan, J P. (2001). Life and Motion of Socio Economic Units. London. Taylor and Francis Ch apters 1 4. Geoghegan, J. ( 2002 ) The Value of Open Spaces in Residential Land Use. Land Use Policy, Vol. 19, pp. 91 98. Georgescu Roegen, N. 1971). The Entropy Law and the Economic Process. Cambridge. Harvard University Press. Ghana Invest ment Promotion Centre (GIPC, 2011; v arious years). Quarterly Operational Reports. Available at: http://www.gipc.org.gh/publications_content.aspx?id=25 Last Accessed: 10/31/11. Ghana Inves tment Promotion Center (GIPC, 2009). Ghana Increases FDI By 160% to a record US$3.17 Billion in 2008. The GIPC Quarterly Update Vol. 4, No.4. January, 2009. Glick, R. (Ed). (1998). Managing Capital Flows and Exchange Rates: Perspectives from the Pacific Basin. Chapters 1 5. Goldblum, C., and Wong, T. (2000). Growth, crisis and spatial change: a study of haphazard urbanization in Jakarta, Ind onesia. Land Use Policy Vol. 1 pp. 29 37.

PAGE 372

372 Gough, K. V. & Yankson, P. W. K. (2000). Land Markets in African Cities: The case of Peri urban Accra, Ghana. Urban Studies Vol. 37, No. 13, 2485 2500 Grant, R. (2009). Globalizing City. The Urban and Economi c Transformation of Accra, Ghana. Syracuse. Syracuse University Press. Grant, R. (2007). Geographies of Investment: How Do the Wealthy Build New Houses in Accra, Ghana. Urban Forum Vol. 18, No. 1, January March 2007. Grant, R. (2001). Liberalization and F oreign Companies in Accra, Ghana. Environment and Planning A Vol. 33, pp. 997 1014. Grant, R. and Yankson, P. (2003). City Profile Accra. Cities Vol. 20, No.1, p p. 65 74 Grossman, G. M. and Krueger, A. B. ( 1991 ). Environmental Impacts of a North Americ an Free Trade Agreement. National Bureau of Economic R esearch Working Paper, No. 3914. Hall, P., and Tewdwr Jones, M. (2011). Urban and Regional Planning. 5 th Ed. New York. Routledge. Hanink, D. M. (1997). Principles and Applications of Economic Geography Economy, Policy, Environment. New York. John Wiley & Sons. pp. 128 129 Hanink, D.M. (1994). The International Economy: A Geographical Perspective. New York. John Wil ey & Sons. pp. 115 136. Harvey, D. (1985). The Urbanization of Capital. Studies in the History and Theory of Capitalist Urbanization. Baltimore. The Johns Hopkins University Press. Harvey, D. (1969). Explanation in Geography. London. Edward Arnold (Publishers) Ltd. Hoddinott J., (1994). A Model of Migration and Remittances Applied to Weste rn Kenya, Oxford Economic Papers Vol. 46. pp. 459 76. Hoogvelt, A. (1997). Globalization and the Post Colonial World. Baltimore. Johns Hopkins University Press. International Human Dimensions Project (IHDP). (2005). Urbanization and Global Environmental Change. (Science Plan). Report No. 15. Bonn: IHDP Secretariat. International Organization for Migration (IOM). (2011). Facts & Figures: Global Estimates and Trends. Available at: http://www.iom.int/jahia/Jahia/about migration/facts and figures/lang/en Last accessed on May 20, 2011. Irwin, Elena G. 2002. The Effects of Open Space on Residential Property Values. Land Economics Vol. 78 No 4, pp. 465 480.

PAGE 373

373 Jaccard, J. a nd Jacoby, J. (2010). Theory Construction and Model Building Skills. A Practical Guide for Social Scientists. New York and London. The Guilford Press. Johnson, D. L., Haarmann, V., Johnson, M. L., and Clawson, D. L. (Eds.). (2010). World Regional Ge ograp hy. 10 th Edition New York. Prentice Hall. Johnston, R. J. (2001). Geography And Geographers: Anglo American Human Geography since 1945. 5th editio n. London. Arnold. Pp. 127 139 Johnston, R. J., Gregory, D., Pratt, G., and Watts, M. (Eds). (2002). T he Dictionary of Human Geography. Oxford. Blackwell Publishers, Inc. Pp. 508 510. Kasanga, K. & Woodman, G. R. (2004). Local Law Making and Land Conversion in Kumasi, Asante. In: Woodman, G. R., Wanitzek, U., & Sippel, H. (Eds.). (2004). Local Land Law a nd Globalization. A Comparative Study of Peri urban Areas in Benin, Ghana and Tanzania. Munster. LIT VERLAG. Pp. 153 216. Kavouras, M. (2001). Understanding and Modeling Spatial Ch ange. In: Frank, A., Raper, J., and Cheylan, J P. (2001). Life and Motion o f Socio Economic Units. London. Taylor and Francis. Pp.49 59. Keivani, R. and Mattingly, M. (2007). The Interface of Globalization and Peripheral Land in the Cities of the South: Implications for Urban Governance and Local Economic Development. Internat ional Journal of Urban and Regional Research Vol. 31.2. Konadu Agyemang, K. (2001). Structural adjustment programs and housing affordability in Accra, Ghana. The Canadian Geographer 45, No. 4, pp. 528 544. Konadu Agyemang, K. ( 2001 ) The Political Eco nomy of Housing and Urban Development in Africa: Ghana's experience from colonial times to 1998. Westport. Praeger Publishers Lambin, E. F. & Meyfroidt, P. (2011). Global Land Use Change, Economic Globalization, and the Looming Land S carcity. Proceedings of the National Academy of Sciences (PNAS), Vol. 108, No. 9. Pp.3465 3472. Available at: http://www.pnas.org/content/108/9/3465.full Accessed: October 9, 2011. Lambin, E.F. (1997). Modeling and monitoring land cover change processes in tropical regions. Progress in Physical Geography, 21(3), 375 393. Latorre, M. C. (2008). Multinationals and Foreign Direct Investment. Main Theoretical Strands and Empirical Effects. Departamento de Economa Apli cada II Universidad Compl utense de Madrid. Available at: http://www.ucm.es/centros/cont/descargas/documento22427.pdf Accessed:January 27, 2011

PAGE 374

374 Lepers, E., Lambin, E. F., Janetos A. C., Defries, R., Achard, F., Ramankutty, N. and Scholes, R. J. (2005). A Synthesis of Information on Rapid Land Cover Change for the Period 1981 2000. Bioscience. Vol. 55, No. 2. pp. 115 121 Li, L., Sato, Y., and Zhu, H. (2003). Simulating spati al urban expansion based on a physical process. Landscape and Urban Planning 64 (2003) 67 76. Lim, G C. (2010 Globalization, spatial allocation of resources and spatial impacts: A conceptual framework. In: Richardson, H.W. and Chang Hee. C. Bae. (Eds.). 2010. Globalization and Urban Development. Berlin. Springer. Pp.1 29. Locatelli, F. & Nugent P. (eds.). (2009). African Cities. Competing Claims on Urban Space s Leiden. The Netherlands. Koninklijke Brill NV Luniak, M. (1995). The development of bir d communities in new housing estates in Warsaw. Memorabilia Zoologica Vol. 49, pp. 257 267. Lynham, S.A. (Ed.). (2002b). The General Method of Theory building in applied disciplines. Advances in in Developing Human Resources, Vol. 4 pp. 221 241. Mabogun je, A. (1997). The State of the Earth. Contemporary Geographic Perspectives. Oxford. Blackwell Publishers. Chs. 1 3 & 10. Malczewski, J., (1999). GIS and Multicriteria Decision Analysis. New York and Toronto. John Wiley & Sons. Pp. 1 66. Maimbo, S.M., a nd Ratha, D. Eds.(2005). Remittances: Development Impact and Future Prospects. Washington, D.C. The World Bank. Mallampally P., and Sauvant, K. P. 1999. Foreign Direct Investment in Developing Countries. Finance & Development, 35. March 1999. Mazzucato, V ., van den Boom B., and Nsowa Nuamah, N.N. (2008). Remittances in Ghana: Origin, Destination and Issues of Measurement. International Migration Journal Vol.46 (1) 2008. Marzluf, J.M. (2001). Worldwide urbanization and its effect on birds. In Marzluf, J.M ., R. and Brown R. D. (Eds.). (2001). Avian Ecology in an urbanizing world. Norwell. Kruwer. Pp.19 47. McDonald, J. F. and McMillen D. P. ( 20 11) Urban Economics and Real Estate. Theory and Policy. 2 nd edition. Place? John Wiley & Sons. McGrath, D. T. (2005). More evidence on the spatial scale of cities. Journal of Urban Economics Vol. 58 pp. 1 10. McKinney, M. L. (2002). Urbanization, Biodiversity and Conservation. Bioscience Vol. 52, no.10, pp. 883 890.

PAGE 375

375 Milllington, J. D.A., Perry, G. L. W., and R omero Calcerrada, R. (2007). Regression Techniques for Examining Land Use/Cover Change: A Case Study of a Mediterranean Landscape. Ecosystems Vol. 10 pp 562 578. Mills, E.S. and Tan, J.P. 1980. Comparison of urban population density functions in develop ed and developing countries, Urban Studies 17: 313 331. Moller Jensen, L., Kofie, R. Y. and Yankson, P.W.K. (2005). Large scale urban growth observations a hierarchical kernel approach based on image texture. Danish Journal of Geography 105(2): 39 47. Mmieh, Frederick and Owusu Frimpong, N. (2004). State Policies and the Challenges in Attracting Foreign Direct Investment: A Review of the Ghana Experience. Thunderbird International Business Review 46: 575 99. Mohapatra, S., Rather, D., and Silwa, A. (2 011). Outlook for Remittance Flows 2011 12. Recovery after the crisis, but risks lie ahead. Migration and Development Brief 13, Nov 8, 2010. Migration and Development Unit, World Bank. Moran, T. H. (2011). Foreign Direct Investment and Development. Launch ing a Second Generation of Policy Research. Washington, D.C. Peter G. Peterson Institute for International Economics. Munasingh, M. (2002). Introduction Elements of Environmental Macroeconomics. In Munasingh, M (Ed.). (2002). Macroeconomics and the En vironment. Northampton. Edward Elgar Publishing, Inc. Pp. xiii xliv. Muth, R. F. (1969). Cities and housing: The spatial pattern of urban residential land use. Chicago: University of Chicago Press. National Research Council. (2010). Understanding the C hanging Planet: Strategic Directions for the Geographical Sciences. Washington, D.C. National Academies Press. National Research Council (NRC). (1999). Global Environmental Change: Research Pathways For The Next Decade. Washington, D.C. National Academies Press. Pp. 293 375. Ndulu, B., Chakraborti L., Lijane, L., Ramachandran, V., and Wolgin, J. (2007). Challenges of African Growth: Opportunities, Constraints and Strategic Directions, Washington, D.C. World Bank. s Development. (Founding Document). Available at: http://www.nepad.org/ Accessed on May 18, 2011.

PAGE 376

376 Nordbeck, S. (1965). The Law of Allometric Growth. Discussion Papers, The Michigan Inter University Committee of Mathe matical Geography Classical Reprints: http://www personal.umich.edu/~copyrg ht/image/micmog.html. Accessed on October 30, 2008. O'Brien, K.L. and Leichenko, R. M., (2000). Double exposure: assessing the impacts of climate change within the context of econom ic globalization. Global Environmental Change 10 (2000) 221 232. Ofori, S.C. (2002). Regional Policy and Planning in Ghana. Making things happen at the territorial community. Hampshire. Ashgate Publishing Company. Pacione, M. 2009. Urban Geography. A Glob al Perspective. 3rd Edition. London and New York. Routledge. Pp. 164 180. Parnell, S., Simon, D., and Vogel, C. (2007). Global environmental change: conceptualizing the growing challenge for cities in poor countries. Area (2007) 393, 357 369. Ratha, Dili External Development Finance. Global Development Finance 2003. Ch.7. Washington, D.C. World Bank. Rakodi, C. (2007). Land for Housing in African Cities: Are Informal Delivery Systems I nstitutionally and Pro Poor? Global Urban Development Vol.3, No.1 Rees, W. E. 1999. Achieving Sustainability: Reform or Transformation? In Satterwaite, Earthscan Publishing Ltd. Pp. 28 40 Rees, W.E. (1997). Urban ecosystems: the human dimension. Urban Ecosystems 1, 63 75. Rees, W. E. (1996). Revisiting Carrying Capacity: Area Based Indicators of Sustainability. Population and Environment: A Journal of Interdisciplinary Studies. Vol. 17, No. 3, 1996. Rees, W.E. (1992). Ecological Footprints and appropriated carrying capacity: what urban economics leaves out. Environment and Urbanization, Vol. 4, No. 2, Oct. 1992. Rees, W.E. and Wackernagel, M. (1994). Ecological Footprints and appr opriated carrying capacity: measuring the natural capacity requirements of the human economy. In: Janssen, A., Hammer, M., Folke, C., and Costanza, R. (Eds). (1994). Investing in Natural Capital. Washington D.C., Island Press. Reid, N. (2002). The Restless Landscape. An Introduction to Human Geography. Dubuque. Kendall/Hunt Publishing Company. Pp. 109 146.

PAGE 377

377 Rogerson, P. A. (2005). Statistical Methods for Geography. Thousand Oaks. Sage Publications. Pp. 105 153. Rudestam, K. E and Newton, R. R. (2011). Surviving Your Dissertation. A Comprehensive Guide to Content and Process. Los Angeles. Sage Publications Sanderson, E.W., Jaiteh, M., Levy, M.A., Redford, K.H., W annebo, A. V., and Woolmer, G. ( 2002 ) The Human Footprint and t he Last of the Wild. Bioscience. Vol. 52 No. 10. 891 904. Sarraf, M. & Jorio, A. (2010). Land Degradation: The Case of Morocco. In: Croitoru, L. & Sarraf, M. (Eds.). (2010). The Cost of Environmental Degradation. Case Studies from the Middle East and Nor th Africa. Washington, D.C. The World Bank. Sassen, S. (2006). Cities in a World Economy. Thousand Oaks. Sage Publications. Pp. 1 43 Savageu, M. A. (1979). Allometric morphogenesis of complex systems: Derivation of the basic equations from first princip les. Proc. Natl. Acad. Sci Vol. 76, No.12, pp. 6023 6025. Seto, K.C., Fragkias, M., Guneralp, B., and Reilly, M. K. (2011). A Meta Analysis of Global Urban Land Expansion. PLoS ONE 6(8): e23777. Available at: http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0023777 Seto, K. C. (2005). Economies, Societies, and Landscapes in Transition: Examples from the Pearl River Delta, China, and the Red Ri ver Valley, Vietnam. In: Entwisle, B., and Stern, P. C., (Eds). (2005). Population, Land Use, and Environment: Research Directions. (NRC). Washington, D.C. National Academies Press. Pp. 193 216. Shestopaloff, Y. K. (2010). Properties and Interrelationship s of Polynomial, Exponential, Logarithmic and Power Functions with Applications to Modeling Natural Phenomena. Toronto and Coral Springs. AKVY Press. Short, R. J. and Yeong Hyun K. ( 1999 ) Globalization and the City. New York. Addison Wesley Longman Li mited. Pp. 3 22. Singh, R. J., Hacker, M., and Lee, K W. (2009). Determinants and Macroeconomic Impacts of Remittances in Sub Saharan Africa. IMF Working Paper, WP/09216. IMF African Department. Sophism, Y. (2006). Remittances Policy That Works: The Case o f Ghana. Available at: http://www.ghanaweb.com/GhanaHomePage/NewsArchive/artikel.php?ID=106354 Accessed : May 19, 2011

PAGE 378

378 Songsore, J. (2008). Environmental and Structur al Inequalities in Greater Accra. The Journal of the International Institute. University of Michigan. Available at: http://quod.lib.umich.edu/cgi/p/pod/dod idx?c=jii;id no=4750978.0016.105 Last accessed: January 25, 2012 Spray, S. L., and McGlothin, K. L. ( 2003 ) Lo ss of Biodiversity. New York. Rowman and Littlefield Publishers, Ltd. pp. 3 26. Stark, O., and Lucas, R. E.B. (1988). Migration, Remittances, and the Fami ly Economic Development and Cultural Change Vol. 36, pp. 465 81. Sutcliffe, R. (2001). 100 Ways of Seeing an Unequal World. London. Zed Books. Thomas, R.W. and Huggett R. J. ( 1980 ) Modeling in Geography: A Mathematical Approach. Totowa. Barnes and Nob le Books. Tipple, G.A., Korbe, D., Willis, K., and Garrod, G. (1998). Who is building what in urban Ghana? Cities Vol.15, No. 6. Pp 399 416. Trefon, T. (2009). Hinges and Fringes: Conceptualizing the peri urban in Central Africa. In: Locatelli, Franc es ca and Nugent Paul (Eds.). (2009). African Cities. Competing Claims on Urban Space. Leiden. Koninklijke Brill NV. Pp. 16 35. Turner, W. R., Nakamura, T., and Di netti, M. (2004). Global Urbanization and the Separation of Humans from Nature. Bioscience Vol. 54 No. 6. 585 590. United Nations. (2011) http://data.un.org/ Last Accessed: 10/03/2011. United Nations. 2003. Ecosystems and Human Well Being: A Fram ework for Assessment. I n Millennium Ecosystem Assessment (MA ). Washington D.C., Island Press. Pp. 1 25 UNCTAD. (2011). Global and Regional FDI Trends in 2010. Global Investment Trend Monitor, No. 5, 17 January 2011. UNCTAD. (2010). Investing in a Low Carbon Economy. World Investment Report 2010. New York. United Nations Publications. UNCTAD. (2007). Development and the next generation. World Investment Report, 2007. New York. United Nations Publications. UNDP. (2005). Investment in Development. A Practical Plan to Achieve the Millennium Development Goals. Sterling Earthscan. Available a t http://www.unmillenniumproject.org/reports/fullreport.htm Accessed : May 18, 2011. Vale T.R. & Vale, G.R. (1976). Suburban bird populations in west centr al California. Journal of Biogeography 3: 157 165.

PAGE 379

379 Vernon, R. (1966). International Investment and International Trade in the Product Life Cycle. Oxford Bulleting of Economics and Statistics 41: 255 267. Voltairenet. (2006). Sub Saharan economies are gr owing faster but are they really getting stronger ? Available at: http://www.voltairenet.org/article141702.html Accessed : May 17, 2011. Von Thunen, J. H. 1966 (1826). The isolated state, An En glish edition of Der isolierte Staat, trans. Carla M. Wartenberg; edited with an introduction by Peter Hall. Oxford/New York: Pergamum Press. Wackernagel, M. (2006). The Ecological Footprint of cities and regions: comparing resource availability with dem and. Environment and Urbanization Vol. 18 no. 1, pp. 103 112. Wackernagel, M., Ktzes, J., Moran, D., Goldfinger, S., and Thomas, M. (2006). The Ecological Footprint of cities and Regions: comparing resource availability with resource demand. Environme nt & Urbanization Vol. 18 (1): 103 112. Wackernagel, M., and Williams, R. E. (1995). Our Ecological Footprint: Reducing Human Impact on the Earth. British Columbia and Philadelphia. New Society Publishers. d Use. Available at http://www.sjsu.edu/faculty/watkins/thunen.htm Last accessed on: Oct., 15, 2011. Wilson, A. G. (1981). Geography and the Environment. System s Analytical Methods. New Yo rk. John Wiley & Sons Ltd. Winkler, J. A. (2011). Geographical Sciences versus geography. How and Why? Professional Geographer Vol. 63, No.3 Woodman, G. R., Wanitzek, U., & Sippel, H. (Eds.). (2004). Local Land Law and Globalization. A comparative study of peri urban areas in Benin, Ghana and Tanzania. Munster. LIT VERLAG. Woodman, G. R., Wanitzek, U., & Sippel, H. (2004). Relating Local Legal Activity to Global Influences: A Theoretical Survey. In: Woodman, G. R., Wanitzek, U., & Sippel, H. (Eds.). (2 004). Local Land Law and Globalization. A Comparative Study of Peri urban Areas in Benin, Ghana and Tanzania. Munster. LIT VERLAG. Pp. 1 79. World Bank. (2004). Global Development Financing 2004: Harnessing Cyclical Gains for Development. Washington, DC. World Bank. World Bank. ( 2004 ) Assessing the Economic V alue of Ecosystem Conservation. Environment Department Paper No. 101. Washington, D.C.

PAGE 380

380 World Bank ( 1997 ) Valuating the Economic Impacts of Urban Enviro nmental Problems: Asian cities Working Paper N o. 13. Washington, D.C. Yeboah, I. E. A. (2003). Demographic and Housing Aspects of Structural Adjustment and Emerging Urban Form in Accra, Ghana. Africa Today Vol. 50, No. 1, pp. 106 119. Yeboah, I. E.A (2000 ). Structural Adjustment and Emerging Urban Form in Accra, Ghana. Africa Today Vol. 47, No. 2. Pp. 61 89. Yeboah I. E. A. & Briggs, J. (2001). Structural Adjustment and the Contemporary African City. Area Vol. 33, No. 1, pp. 18 26 Zwi ck, P. D., & Carr, M. H. (2007 ) Smart Land Use Analysis. The LU CIS Model. Redlands. ESRI Press. Pp. 1 8. Zwick, P. D., & Carr, M. H. (2006). Florida 2060. A Population Distribution Scenario for the State of Florida. Gainesville. Geoplan Center, Univ ersity of Florida. Available at : http://www.1000friendsofflorida.org/PUBS/2060/Florida 2060 Report Final.pdf Last accessed: February 4, 2012.

PAGE 381

381 BIOGRAPHICAL SKETCH Kofi Adu Brempong was in Kumasi, Ghana, and currently married with two children. He is a broadly trained human geographer with interests in urban and economic geography. His doctoral research is designed to contribute to understanding the complex linkages among economic globalization, urb anization and environmental chan ge in less developed countries, with special reference to Accra, Ghana. His research approach, broadly contextualized in the human environment interactions geographical trad ition draws extensively on classical economic / geographical theories and models w hich he attempts to refine to reflect contemporary realities of economic globalization. Kofi received the MA degree in Geography from the University of Toledo, Ohio tial analytical sk ills ( GIS and quantitative geographical methods) and their applications to soc ioeconomic phenomena. He earned the BA (Honors) degree in Economics and Geography from KNUST, Kumasi Ghana in (1999), after which he taught Economics at K onongo Odumasi Seconda ry School, Ghana (1999 2000) ; and Geography at the Toase Secondary School, Ghana (2000 2003) Kofi served as a graduate teachin g assistant (TA) in geography at both the University of Toledo (2003 2005) and the University of Florida (2005 2010); and is loo king forward to teaching and research caree r in higher education.