A Reversed Collective Action Dilemma and the Challenge of Enhancing Bureaucratic Capacity in Botswana and Ghana

Material Information

A Reversed Collective Action Dilemma and the Challenge of Enhancing Bureaucratic Capacity in Botswana and Ghana
Pankani Lindberg, Winifred V
Place of Publication:
[Gainesville, Fla.]
University of Florida
Publication Date:
Physical Description:
1 online resource (242 p.)

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Political Science
Committee Chair:
Committee Co-Chair:
Committee Members:
Graduation Date:


Subjects / Keywords:
Agriculture ( jstor )
Cattle ( jstor )
Civil service ( jstor )
Collective action ( jstor )
Farmers ( jstor )
Farms ( jstor )
Mathematical dependent variables ( jstor )
Political power ( jstor )
Public administration ( jstor )
Women ( jstor )
Political Science -- Dissertations, Academic -- UF
africa -- agriculture -- botswana -- building -- bureaucratic -- civil -- comparative -- development -- economy -- ghana -- politics -- service -- state
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Political Science thesis, Ph.D.


African countries generally exhibit lower levels of administrative and state capacity compared to other regions of the developing world. Decades of civil service reforms across the continent have yielded little by way of enhanced bureaucratic capacity. These reforms have been focused on civil servants and the formal aspects of the state, not the informal institutions within which officials live and work. Using data from Botswana and Ghana, both considered strong states by regional standards, this dissertation finds that demands on individuals' resources constitute significant sources of financial strain and often result in individuals acting in ways that undermine administrative and, by extension, state capacity and that the most common forms of assistance sought reflect the need for basic safety net programs. Reforms have so far not dealt with these realities in a systematic way, and they should if we are to improve administrative capacity. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
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.
Thesis (Ph.D.)--University of Florida, 2015.
Co-adviser: SMITH,DANIEL A.
Statement of Responsibility:
by Winifred V Pankani Lindberg.

Record Information

Source Institution:
Rights Management:
Copyright Pankani Lindberg, Winifred V. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
LD1780 2015 ( lcc )


This item has the following downloads:

Full Text




© 2014 Winifred Pankani Lindberg


To all that I have met and am yet to meet


4 ACKNOWLEDGMENTS I owe the completion of this dissertation to many. Without the financial support from the would not have been able to pursue a PhD degree much more complete one. Th e pre dissertation (APPP). I will forever be grateful for their generous f inancial support. Administrative Studies (PAS), the Ministries of Agriculture in both Botswana and Ghana, and the Real Sector Division of the Ministry of Finance and Economic Planning in Ghana. My affiliation with PAS gave me access to the University Botswana community, one of the best libraries, and provided an office. The support I got from the Real Sector Division was invaluable, as was the hospitality and support I received from both Ministries of Agriculture and many field officers. Throughout my time in Florida, the staff at both the Center for African Studies and the department of Political Science have been invaluable in both my academic and professional devel opment. To Debbie Wallen, Suzanne Lawless Yanchisin, Patricia Root, Corrina Green and Mrs. Ike Akinyemi a heartfelt thank you. A heartfelt thank you to Mme Segolame Dintwa and Mma Wapula Noge, PAS, Ms. Benedicta Korsah, MoFA, Ghana, Roland Ofori, Ministry of Finance, Ghana whose watchful eyes and companionship embodied the very subject of this study. I would also like to acknowledge help I got from the staff of the library of the University of Botswana and the Botswana National Archives. I thank Mma Rebecca Lesego Kgosi, Mma Itumeleng Jacqueline Morapedi, Mr. Khukhutha, DESC Staff and the best research assistants for their hospitality and help. I could not have wished for a better boss than Mma Morapedi whom I


5 learnt so much from and thoroughly enjoyed worki ng for every single day. My Ministry of Agriculture colleagues in Ghana are too many to name, a collective thank you to each and every one of you. Special gratitude goes to my numerous informants for providing interviews: farmers, fishers, weavers, traders , community leaders, traditional practitioners, the emerging entrepreneurs, and officers in the public and private sectors. Dr. Alhassan Iddrisu, Ministry of Finance, Alhaji Osman, Ministry of Interior, Ghana. As we say in Ghana, listen for my gratitude in the dawn crows of the rooster. Drs. Edelgard Mahant, Dorothy Mpabanga, Phoebe Lostroh, and Pat Coy and Karin L. Tanquist, you will forever be associated with the evening whiff of the jacaranda tree. Thank you for a lifetime of funny memories and long dri ves. I thank Drs. Morapedi and Bongani Gumbo both from the history department of the University of Botswana for the many interviews they arranged for me. I cannot imagine pursuing a PhD anywhere but at the University of Florida. My heartfelt thanks to all many more. I am indebted to Magda Giurcanu and Jen Boylan for helping me with my data analysis. I owe a debt of gratitude to all my teachers. I am grateful for the men and women who be lieved in my abilities long before I recognized them in myself. Brian Moraski has always been took years ago but am only now just beginning to appreciate, to Bren da Chalfin, Terje Østebø, Alioune Sow and Luise White who patiently answered my questions and introduced me to literature I would otherwise not have read, I owe all of you and many others a real debt for contributing to my education. Without Denis Galvan a nd Steve Wooten I would never have explored political science, much less pursued a PhD in political science. Thank you. Finishing


6 this dissertation would not have been possible without the support of my professors and committee members; to Michael Bernard, a very concrete way to framing political and sociologica l problems. My final paper for the seminar became my proposal and eventually parts of this dissertation. Dr. Bernard helped me refine my research question. Long before I left for fieldwork, he insisted I finish my questionnaire. A month into my stay in Bot swana I experienced my first meltdown, and overwhelmed and alone, I called Dr. Tselaesele. I could not have asked for a better field became my road map. Dr. Renata Serra introduced me to the work of Jack Goody and development economists Arthur Lewis and Michael Lipton. A month into my stay in Botswana I was recruite d to help draft a report, and this is when I discovered the magnitude of Michael discovery about Arthur Lewis in Ghana. Working with Dr. Serra has been one of the great learning experiences of my life. Dr. Smith (Ben) has always skillfully pulled me back up when I got lost in the minutiae of the dissertation. I am indebted to Sarah McIntosh whose comments and edits gives coherence to my arguments. I have heard it s aid you think of water when the well is empty; Esther Obonyo thank you for everything. To Dr. Dan Smith, as we say up north, gratitude is the heart's memory. on me. My pr ofound gratitude to Dr. Serra whose detailed reading and thorough comments made


7 the process of writing and editing my dissertation one of the best seminars of my graduate experience. Dr. Serra asks much and gives even more and for this I would be forever i ndebted. myself short. For always giving your time so generously, for the understated encouragement when I doubted, for the confidence you had in me when I had none, for all the myriad little things that are impossible to describe, much more quantify, but made a difference during some of my bleakest moments and for seeing me through, thank you sounds so inadequate in expressing what I feel. To the Lindbergs and Ohlssons, the Akinyemis, Ma and Papa, and the Anyanfuls I owe a debt of gratitude for all the support and love. Staffan, my heartfelt thanks for the experiences of our youth. To Maylem, Akush and Dan together we have always pulled through. To both sides of Busunu Da boya alliance to whom I owe the gift of life and the burden of history and all the alliances forged in between thank you.


8 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ......................... 11 LIST OF FIGURES ................................ ................................ ................................ ....................... 14 ABSTRACT ................................ ................................ ................................ ................................ ... 15 CHAPTER 1 TRACKING CATTLE ................................ ................................ ................................ ........... 16 1.1 Intro duction ................................ ................................ ................................ ....................... 16 1.2 The Research Question ................................ ................................ ................................ ..... 21 1.3 Scope of the St udy ................................ ................................ ................................ ............ 25 1.3.1 The Art of Formalizing the Informal ................................ ................................ ...... 30 1.3.2 Situating the Dissertation within the Africanist Canon ................................ .......... 32 1.3.3 Informal Institutions and the Lack of Satisfactory Answers ................................ .. 33 1.4 Research Design and Case Selection ................................ ................................ ................ 34 1.4.1 W hy Study the Ministry of Agriculture? ................................ ................................ 34 1.4.2 The Road Less Traveled ................................ ................................ ......................... 35 1.4.3 Comparative Case Studies ................................ ................................ ...................... 36 1.4.4 Why Botswana and Ghana? ................................ ................................ .................... 37 1.4.5 Fieldwork ................................ ................................ ................................ ................ 39 1.5 Overview of the Dissertation ................................ ................................ ............................ 41 1.6 The Relevance of the Study ................................ ................................ .............................. 41 1.7 Conclusion ................................ ................................ ................................ ........................ 42 2 THEORIZING THE REVERSED COLLECT IVE ACTION DILEMMA ............................ 44 2.1 Introduction ................................ ................................ ................................ ....................... 44 2.2 Situating the Study within the Africanist Cannon ................................ ............................ 45 2.2.1 Sowing the Wind ................................ ................................ ................................ .... 48 2.2.2 Reaping t he Whirlwind ................................ ................................ ........................... 56 2.3 Affective Reciprocity ................................ ................................ ................................ ........ 57 2.3.1 A Reversed Collective Action Problem ................................ ................................ . 58 2.3.2 Institutional Persistence ................................ ................................ .......................... 59 2.4 Conclusion ................................ ................................ ................................ ........................ 70 3 STATE CAPACITY: COMPARING STATE OUTCOMES IN BOTSWANA AND GHANA ................................ ................................ ................................ ................................ .. 71 3.1 Introduction ................................ ................................ ................................ ....................... 71 3.2 The Quest For A More Capable State ................................ ................................ ............... 72


9 3.3 Rolling Back the State ................................ ................................ ................................ ...... 82 3.3 Capacit y to Do What? Selecting the Indicators ................................ .............................. 87 3.3.1. Capacity to Administrate ................................ ................................ ....................... 88 3.3.2 Comparing Capabilities _ Regional and Across Countries ................................ .... 90 3.4 Conclusion ................................ ................................ ................................ ........................ 92 4 NOTES ON METHODOLOGY ................................ ................................ ............................. 95 5 A TALE OF TWO CITIES ................................ ................................ ................................ ..... 99 5.1 Introduction ................................ ................................ ................................ ....................... 99 5.2 Overview of the D ata ................................ ................................ ................................ ...... 101 5.3 Extension officers: who are they? ................................ ................................ ................... 105 5.3 Portraits from The Sample ................................ ................................ .............................. 109 5.5 Dia gnostic Tests ................................ ................................ ................................ .............. 110 6 FROM FARMERS EXPECTATIONS TO ACTION ................................ .......................... 115 6.1 Introduction ................................ ................................ ................................ ..................... 115 6.2. Weak State Administrative Capacity ................................ ................................ ............. 116 6.2.1 Dependent Variable 1: Frequency of Contact ................................ ...................... 118 6.2.2 Dependent Variable 2: Favor Seeking ................................ ................................ .. 121 6.2.3. Dependent Variable 3: Got Favors ................................ ................................ ...... 124 6.2.4 Aggregating the Dependent Variables ................................ ................................ .. 125 6.2.5 Transforming the Variables for Binary Logistic Regression ............................... 126 6.3 The Independent Variables: Measuring Affectiv e Reciprocity ................................ ...... 127 6.3.1. Independent Variable 12: Embeddedness ................................ ........................... 127 6.3.2. Independent Variable 2: Fairness ................................ ................................ ........ 129 6.3.3. Independent Variable 3: Attitudes towards Officers ................................ ........... 133 6.3.4. Independent Variable 4: Request for Assistance ................................ ................. 134 6.3.5. The Control Variables ................................ ................................ ......................... 137 6.4. Data Analysis and Discussion of the Results ................................ ................................ 138 6.4.1. Linear Regression Analyses of Affective Reciprocity ................................ ........ 140 6.4.2. Negative Binominal Regression Analyses of Affective Reciprocity .................. 142 6.4.3. Logistic Regression Analysis of Affective Reciprocity ................................ ...... 145 6.5 Conclusion ................................ ................................ ................................ ...................... 148 7 EXTENSION OFFICERS AND AFFECTIVE RECIPROCITY ................................ ......... 163 7.1 Introduction ................................ ................................ ................................ ..................... 163 7 .2 Conceptualization and Measurement of the Dependent Variables ................................ . 164 7.2.1 Dependent Variable 1: Resource Diversion ................................ ......................... 164 7.2.2 Dependent Variable 2: Use of Official Time ................................ ....................... 168 7.3 Measuring Affective Reciprocity: Specifying the Independent Variable ....................... 169 7.3.1 Independent Variable 1: Embeddedness ................................ .............................. 170 7.3.2 Independent Variable 2: Fairness ................................ ................................ ......... 172 7.3.3 Independent Variable 3: Weberianess ................................ ................................ .. 176


10 7.3.4 Independent Variable 4: Requests for Assistance ................................ ................ 177 7.3.5 The Control Variables ................................ ................................ .......................... 180 7.4 The Data and Discussion of the Results ................................ ................................ ......... 1 81 7.4.1 Linear Regression Analyses of Affective Reciprocity ................................ ......... 182 7.4.2 Negative Binominal Analysis. of Affective Reciprocity ................................ ...... 185 7.4.3 Results of Logistic Regressions of Affective Reciproc ity ................................ ... 186 7.5 Conclusion ................................ ................................ ................................ ...................... 186 8 THE PARADOX OF THE ECONOMY OF AFFECTION ................................ ................. 197 APPENDIX: SUPPLEMENTAL FIGURES ................................ ................................ .............. 201 LIST OF REFERENCES ................................ ................................ ................................ ............. 227 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 242


11 LIST OF TABLES Table page 3 1 Regional Comparisons of Select Outputs of State Capacity ................................ .............. 93 3 2 Regional Comparisons of Select Outputs of State Capacity ................................ .............. 93 3 3 Select Outputs of Robust State Capacity ................................ ................................ ........... 93 3 4 Select Outputs of Robust State Capacity ................................ ................................ ........... 94 5 1 Distributions of Farmers by Region ................................ ................................ ................. 112 5 2 Distributions of Farmers by Gender and Country ................................ ............................ 112 5 3 Distributions of Farmers by Gender and Region ................................ ............................. 112 5 4 Distributions of Farmers by Age ................................ ................................ ...................... 112 5 5 Distributions of Farmers by Religious Affiliation ................................ ........................... 112 5 6 Distributions of Farmers by Settlement Type ................................ ................................ .. 112 5 7 Distributions of Farmers by Level Of Education ................................ ............................. 113 5 8 Distributions of Officers by Region ................................ ................................ ................. 113 5 9 Distributions of Officers by Gender ................................ ................................ ................ 113 5 10 Distributions of Officers by Age ................................ ................................ ...................... 113 5 11 Distributions of Officers by Level Of Education ................................ ............................. 113 5 12 Distributions of Officers by Religious Affiliation ................................ ........................... 113 5 13 Distributions of Officers by Settlement Type ................................ ................................ .. 114 5 14 Distributions of Officers by Marital Status ................................ ................................ ...... 114 5 15 Distributions of Officers by Marital Status ................................ ................................ ...... 114 5 16 Distributions of Officers by Number of Dependents ................................ ....................... 114 5 17 Distributions of Officers by Own Economy Compared to Last Year .............................. 114 5 18 Distributions of Officers by Motivation for Taking Civil Service Jobs .......................... 114 6 1 How often do you visit Ministry Officials_ Past Year N (%) ................................ .......... 151


12 6-2 Doing Official Business Outside Sanctioned Channels_ Past Year N (%) .....................151 6-3 Influence Peddling _Past Year N (%) ..............................................................................151 6-4 Sought Favors_ Past Year N (%) .....................................................................................151 6-5 Got Favors_ Past Year N (%) ..........................................................................................151 6-6 Summary Statistics of the Dependent Variable ...............................................................151 6-7 Correlation Matrix of the Dependent Variables ...............................................................152 6-8 Summary Statistics Binary Dependent Variables ............................................................152 6-9 ..................................................153 6-10 ......................................................153 6-11 ...................................................153 6-12 Embeddedness 4 _ No Respect for Officers N (%) .........................................................153 6-13 ...........................................153 6-14 Fairness 2_ Perceived as Wicked N (%) ..........................................................................154 6-15 Fairness 3_ Kabelo should not listen to his relatives N (%) ............................................154 6-1 Fairness 4_ Mpho Act in a Professional Manner N (%) .................................................154 6-17 Fairness 5 _Nothing wrong with Kabelo helping Mpho N (%) .......................................154 6-18 Attitudes Towards Officials 1_ Expect Immediate Attention N (%) ...............................154 6-19 Attitudes Towards Officials 2_Assist Me Regardless N (%) ..........................................155 6-2 Requests for Assistance 1_ Costs of _Healthcare/Tuition N/(%) ...................................155 6-21 Requests for Assistance 2_Cost of Weddings/Funerals N (%) ........................................155 6-22 Variables Counts ..............................................................................................................156 6-2 Linear Regression Analysis of Affective Reciprocity ....................................................157 6-24 Negative Binominal Analysis of Affective Reciprocity ..................................................158 6-25 Logit Analysis of Affective Reciprocity ..........................................................................159 6-26 Changes in Odds ..............................................................................................................160


13 7 1 Misapprop riating Resources: Last 3 Months N (%) ................................ ........................ 190 7 2 Favoring Trading: Last 3 Months N (%) ................................ ................................ ......... 190 7 3 Funding Favors: Last 3 Months N (%) ................................ ................................ ............ 190 7 4 Use of Official Time N (%) ................................ ................................ ............................. 190 7 5 Summary Statistics of the Dependent Variable ................................ ............................... 191 7 6 Embeddedness 1 Help defray the cost of healthcare/tui tion N (%) ................................ . 191 7 7 Embeddedness 2 Helping relatives and friends financially N (%) ................................ .. 191 7 8 Embeddedness 3 Contribute toward developing place of origin N (%) .......................... 191 7 9 Embeddedness 4 Obeying formal rules N (%) ................................ ................................ 191 7 10 ................................ .................. 191 7 1 Fairness 2 Perceived as Wicked N (%) ................................ ................................ ........... 192 7 12 Fairness 3 Kabelo should follow policy N (%) ................................ ................................ 192 7 13 Fairness 4 Mpho should follow the rules N (%) ................................ .............................. 192 7 14 Fairness 5 Nothing wrong with helping N (%) ................................ ................................ 193 7 15 Weberianess 1 Preferential treatments N (%) ................................ ................................ .. 193 7 16 Weberianess 2 I Degree of Professionalism N (%) ................................ ......................... 193 7 17 Financial contributions past year _Healthcare/education N (%) ................................ ..... 193 7 18 Financial contributions past year _ Job Search N (%) ................................ ..................... 194 7 19 Financial contributions past year_Wedding/funerals N (%) ................................ ............ 194 7 20 Financial Contributions past year _Religious Orgs N (%) ................................ .............. 194 7 21 Financial Contributions past year _ Navigating bureaucracies N (%) ............................. 194 7 22 Descriptive Statistics ................................ ................................ ................................ ........ 195 7 23 Linear Regression Analysis of Affective Reciprocity ................................ ..................... 195 7 24 The Negative Binominal Regressions of Affective Reciprocity ................................ ...... 196 7 25 The Logistic Regressions of Affective Reciprocity ................................ ......................... 196


14 LIST OF FIGURES Figure page 1-1 Affective Reciprocal Feedback Loop. ...............................................................................43 6-1 Specifying the Dependent Variable .................................................................................161 6-2 Sought Favors ..................................................................................................................161 6-3 Three Alternate Specification of Weakened Capacity .....................................................161 6-4 The Binary Indicators ......................................................................................................162 6-5 The Independent Variables ..............................................................................................162 6-6 Embeddedness..................................................................................................................162 6-7 Fairness ............................................................................................................................162 6-8 Attitudes towards Officials ..............................................................................................162 6-9 Requests for Assistance ...................................................................................................162


15 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requiremen ts for the Degree of Doctor of Philosophy A REVERSED COLLECTIVE ACTION DILEMMA AND THE CHALLENGE OF ENHANCING BUREAUCRATIC CAPACITY IN BOTSWANA AND GHANA By Winifred Pankani Lindberg May 2015 Major: Political Science African coun tries generally exhibit lower levels of administrative and state capacity compared to other regions of the developing world. Decades of civil service reforms across the continent have yielded little by way of enhanced bureaucratic capacity. These reforms h ave been focused on civil servants and the formal aspects of the state, not the informal institutions within which officials live and work. Using data from Botswana and Ghana, both considered strong states by regional standards, this dissertation finds th sources of financial strain and often result in individuals acting in ways that undermine administrative and, by extension, state capacity and that the most common forms of assistance sought refle ct the need for basic safety net programs. Reforms have so far not dealt with these realities in a systematic way, and they should if we are to improve administrative capacity.


16 CHAPTER 1 TRACKING CATTLE 1.1 Introduction Months of pestering my boss had fi nally paid off. The main headquarters of the Ministry been unusually quiet. All but a skeleton staff had been deployed around the country for different purposes. S ome of the officers had been dealing with yet another outbreak of foot and mouth flies, a necessary but tediously mind numbing task. Most of the staff from the ani mal department were covering the various animal related events aimed at boosting the production of less capital intensive forms of livestock farming. Due to this shortage of staff, I was on my way to one of the community crushes 1 just outside of Mochudi in a well equipped, four sky, I retreated into my own world, gradually tuning out the conversations of my colleagues. About an hour into our trip, our car suddenly lurched from side to side, flinging my fellow travelers and me around. Snapped out of my reverie, I noticed we were now on one of the familiar rutted roads leading to the more rural settlements, farms and cattle posts. In a moment our vehicle became engulfed in the fine, grayish dust I had come to ass ociate with the dust that often covered the garments of field officers and I, along with the others in our car, lost sight of the other vehicles in the convoy. Then, as suddenly as the bumping had started, it stopped. We had arrived. Assailed by flies and the smell of cattle and dung, I stumbled out of the car. After a 1 Multi purpose close fitting stalls for holding livestock for a variety of reasons, among them to exam ine, mark, help calves suckle or to administer veterinary treatment. Crushes protect both the livestock and humans while the animal is receiving some type of care.


17 great deal of hemming and hawing, equipment and workstations were successfully set up. The attending veterinary technician signaled for me to follow him into the community kraal as he explain was g oing to be his right hand woman. I was thrilled beyond measure. About twenty cows were already locked in the kraal when we arrived, and a group of cowhands jumped from atop their perches on the kraal into the enclosure as we entered the pen. With the young men restricting the cattle, the officer quickly examined the brands, skin, teeth, and hooves, firing off items as I furiously tried to keep up with my checklist, a task made a little difficult by the flies and gnats bent on exploring my face. Already, I w as desperate to switch places with the officer who, it seemed, had the easier job. After what seemed like an interminable hour, I finally got my turn to switch places, and no sooner had I stepped next to my first cow than I was swatted in the face with a dung soaked tail. The howls of laughter reminded me, once again, that I was the sideshow. Hours later my supervisor told me I could take a break. Covered in dung, I collapsed in a heap under the cooling comfort of one of the few trees, to the ribbing of fa rmers and colleagues. A poke in the rib and my sluggish attempt to rejoin my supervisor must have betrayed me. Nothing in my resume from my first three hours suggested I was qualified for my new assignment with the crush crew. The privilege of shepherding what must be the most obstinate cow that day into the narrow crush could only have been assigned to me for falling asleep on the job. Cattle have been, and continue to be, one of the most important factors of social and economic differentiation in Botswan a. The cattle sub sector currently accounts for 73% of the


18 total livestock population, the bulk of which is raised in open grazing areas on communal lands. Beef and its by products constitute about 75% of all meat exports. Given the importance of cattle in Tswana society, cattle rustling is a serious problem that effects both the average citizen as well as the cattle barons who form the core of the political and economic elite. Thus, ensuring the viability of this industry is crucial to both the country and its citizens. importance to the industry Botswana had no choice but to implement the mandate that all animal products be identifiable and traceable from birth to (Bowling et al 2008:288; Fanikiso 2009; Marumo and Monkhei 2009). 2 In 2001 Bot swana rolled out its Livestock Identification Traceback System (LITS) at no small cost in addition to existing cattle identification practices like ear notching, hot iron branding and ear tags. Botswana requires all cattle to be branded, registered and ind ividually identifiable by ear tags, one in each ear, have paper passports and be tracked through the centralized national brands and certificates directory, the most important component of the trace back system. Additionally, owners are required by law to brand their cattle within a month of their move from one ecological zone to another. Officials of the Veterinary Department are responsible for all branding. Once awarded, brands and certificates are valid for ten years and can be renewed up to a year befo re their expiration date (Moreki et al 2012). Roughly the size of Texas, Botswana has a population of only two million people. Kraals and cattle posts are scattered over vast distances, so farmers are clustered into groups or units 2 Regulation (EC) No 1760/2000 of the European Parliament and of the Council of 17 July 2000 establishing a system for the identification and registration of bovine animals and regarding the labeling of beef and beef products and repealing Council Regulation (EC) No 820/97. (Scroll all the way down for document).


19 that then share crushes . These crushes form the smallest epidemiological units where cattle are monitored for disease and also receive treatments including castration, deworming, dipping, and hoof trimming. Each crush is identifiable in LITS by its geographical coordinates, loca tion department uses this information in conjunction with branding data to monitor cattle from birth to . The actual tracking of the animal is done through the oral insertion of a ceramic bolus 3 using an esophageal balling gun. Each individu al reticulum bolus has a unique 16 digit alphanumerical code that allows the trace back system to track and monitor the animal. Inserted boluses are recovered and recycled when the animal is slaughtered and each time the bolus is reused, the recycled seque I was drenched in sweat by the time my supervisor reassigned me to the crush crew. I was beginning to appreciate how tedious each of these tasks were. Taking my place at one side of the crush as a cow was p room for the insertion of the balling gun. I was shocked at the resistance I encountered. Both my supervisor and partners yelled out tips that I did my best to follow, determined to p rove myself useful. Needless to say, I soon found myself digging for the tongue, and before long before I was relieved of this task too. The teasing was now an open free for all as I shifted from one task to another until at last I was called to insert the bolus. The ceramic bolus is placed in a balling gun, gullet. bureaucratic pr ocedures. By the end of the day I had a deeper understanding of some of these 3 A high quality electronic ceram ic transponder used to track, control and manage livestock herds from birth to table.


20 arguments. Driving cattle to crushes is a laborious, time consuming, task with no guarantees that cattle will receive care. Ministry officials frequently run out of boluses, equi pment sometimes malfunctions during campaigns, and scheduled visits get cancelled. Cattle sometimes get injured; some have even died after bolus insertion (Moreki et al 2012:928). But even when everything works as it should, not all cattle owners who usual ly travel great distances to their community crushes, receive services. Farmers, unhappy with the fact that they were not going to receive any service that day, wanted to know when the crew would be coming back to the area. None of the officers had any ans wers. For those looking to sell their cattle to abattoirs for premium prices, this setback was particularly galling after having spent a whole day waiting for service. For eleven years the government of Botswana pressed on with its implementation efforts and even though the program has been discontinued, its mixed record offers important lessons espousing local ownership of development programs came out, reformers have been touting the efficacy of stakeholder ownership of local projects. LITS had the support of all the important stakeholders including bureaucrats, commercial ranchers, and the general population; cattle ownership after all is a fundamental aspect o f the society. In a country with a well resourced, capable bureaucracy where the ownership of cattle is so woven into the fabric of society, the failure of a program like LITS 4 underscores the issues pursued in this study. Post project analyses show offici als failed to adequately monitor the program at all levels through its many different phases. 5 Drawing on and weaving together ethnographic and survey data, this study 4 5 Journal of Animal Science. (12): 925 933 Ndubo N. S., Ditshupo T. and Ntesang J. B.


21 explores the persistence of limited administrative capacity within African civil service s decades of reforms notwithstanding. 1.2 The Research Question In comparative analyses, the countries in sub Sahara Africa are often lumped together into an indistinguishable mass of struggling economies or uns political entities. According to Freed om House data much of the earlier gains in democratization in Mali, the Gambia, Guinea Bissau, Ethiopia, Burundi and the Central Africa Republic have now been rolled back. 6 That African countries have a long and distinguished history of poor economic perf ormance is well established. Of the 48 countries currently designated by the United Nations as least developed, 33 are in Africa. 7 Saharan Africa. 8 Between 2004 and 2008 Afri compared to the 4.5% average of some of the more advanced economies. Even accounting for year. And even with global re 9 In 6 7 Sub 2014, forty three y ears after its creation, the following countries were listed: Angola, Benin, Burkina Faso, Burundi, Central African Republic, Chad, Comoros, Democratic Republic of the Congo, Djibouti, Equatorial Guinea, Eritrea, Ethiopia, Gambia, Guinea, Guinea Bissau, Le sotho, Liberia, Madagascar, Malawi, Mali, Mauritania, Mozambique, Niger, Rwanda, São Tomé and Príncipe, Senegal, Sierra Leone, Somalia, Sudan, Tanzania, Togo, Uganda, Zambia. 8 According to the World Bank and IMF Angola, the Democratic Republic of the Cong o, Ethiopia, Ghana, fastest growing in 2014. 9


22 Russia. 10 Between 1960 and 2000 Mauritius, the Seychelles, Cape Verde, and Botswana; for example, exceede or 11 These countries, in turn , cannot be compared to Somalia, for whom the UNDP has no data due to political instability and whose ability to provide basic services for its citizens is almost nonexistent. These differences notwithstanding, African countries also share many similariti es. For example, pre colonial African polities ran the gamut from small decentralized acephalous societies to centralized kingdoms ( Goody 1980, Fortes and Evans Pritchards 1987, Rothchild and Chazan 1988, Boahen an d Falola 2004 ). 12 Correspondingly, the ability of pre colonial states 10 Sout h Africa was ranked 52nd just below the Southeast Asian average and above the emerging market economies of India and Russia by the World Bank in its 2012 Global Competitiveness Report (ibid 2012: 11). 11 Botswana, according to World Bank and IMF data, avera ged about 11% between 1970 and 1995, one of the index of .634, the average Motswana (a person from Botswana) had a GNI per capital income of $13,10 2, is expected to live up to about 53 years, and would have at the least 8.9 years of education mostly subsidized by the state. Ghana, ranked 135th overall with an index of .558, has a GNI per capital income of $1,684. The average Ghanaian, all things b eing equal, would live to age 65 years and could be expected to have at least 7 years of education. Kenya ranked 145th and with an index of .519, has a GNI per capital income of $1, 541, a life expectancy rate of about 58 years, and Kenyans are expected to have at least 7 years of education. Nigeria, whose recent economic boom has yet to trickle down to the proverbial man on the street, has an index of .471, is ranked 153 rd , and even though average Nigerians make more money ($2,102) than their Kenyan ( $1, 541) and Ghanaian ($1,684) counterparts, their life expectancy rate of 52 years is lower. They are also more likely to have less education (5 years) than their Kenyan and Ghanaian counterparts. Niger is ranked 186 th with an index of .304 and GNI per capital income of $701, a life expectancy rate of about 55.1 years and the expectation of having at least 1.4 years of education. 12 A council of elders instead of chiefs governed acephalous societies like the Tallensi of Ghana and the Neur of southern Sud an. While chiefs with almost absolute powers governed the centralized kingdoms of Old Ghana and the Songhai empires of West Africa and of Zimbabwe in Southern Africa. Mansa Musa of Mali and Sonni Ali of Songhai; for example, had absolute power that was at once executive, legislative, and judicial. Regimes in the region range from the enduring democracies of Botswana and Mauritius to the repressive authoritarian regimes of Eritrea. In between these two extremes are the plebiscitary one party systems like K enya, and competitive one party systems like Malawi, Seychelles and San Tome. Not to be outdone are the settler oligarchies of Namibia, the reluctant democracies of Benin, and the now consolidated democracies of Ghana, Namibia and Senegal. Meanwhile the Iv ory Coast and Mali struggle towards restoration of once stable democracies even as Mugabe continues to run Zimbabwe to the ground.


23 to broadcast power and govern territories under their control was greatly circumscribed by geography and technology. The colonial state, much more impressive in size and scale, was also limited in its ability to effectively and thoroughly penetrate and incorporate society into its ambit of power in any appreciable way beyond the colonial capitals ( Baldwin 1969, Collins 1970, Brett 1973, Boahen 198 7, Coquery Vidrovitch 1988, Young 1994, Hassan 2002, Austin 2010 Abernathy 2002; Austin 1987; Boahene 1987; Goody 1971; Hassan 2002; Young 1994). In time, the post colonial state would inherit these structural constraints and in short order would corrode w hat limited capacity was bequeathed by departing colonial administrators to Africanization policies. Once in full control the ruling elite became even more unresponsive to local demands. Unlike their colonial administrators who were accoun to governme nts and business interests back home in Europe, African elites were answerable to only a few local and international interests ( Amin 1974, Sklar 1985, Chabal 1999, Bayart 2007 Amin 1974; Bayart 2007; Chabal and Dalo z 1999; Sklar 1985). In the early days of independence, senior civil servants who fought hard to retain the Weberian character of the civil service were often viewed with suspicion, or were marginalized or coopted. A few who refused to cooperate were even forced into exile ( Apter 1955, Apter and Center for the Study of Democratic Institutions. 1962, Azarya and Chazan 1987, Boahen 1989 Ekeh 1975; Leonard 1977; Agger 1989; Ladipo 1993; Chabal 1999; Rose Ackerman 1999). Botswana is an exception in this respect. Sstarting from independence President Seretse Khama and continued by Quett Masire and to a lesser extent Festus Mogae, Africanization was gradually introduced and implemented over a two plus decade period. 13 The c ase of David N. 13 Seretse Khama (1921 1980) was the first President of Botswana from 1966 until his death in 1980. Khama is still very popu popular The Number One Ladies Detective Agency series. Khama was succeeded by Quett Masire who served for 18 years (1980 1998). Best known for his role in


24 Magang, recounted in his autobiography is particularly illuminating. Magang recounts how he expatriate policies. Today Magang is one of the richest, if the not the richest person in Botswana. 14 In contrast, Kwame Nkrumah, the first president of Ghana, not only banned political parties soon after independence, he also imprisoned political rivals, and forced opponents too powerful to be jailed, into exile and politicized the civil service. The political expediency of making civil service posts tools of cooptation and patronage and its effects on the service still endure. 15 od without taking into consideration the dependencies created by the earlier success of the Botswana discount the fact that there were multitudes of people s uch as David Magang in Ghana flooding the state with demands it was ill prepared to handle. Their exit into the private would eventually bleed the state of precious tax revenues because of its inability to collect tax from some of its most productive citiz ens now operating in the informal sector. These observed diversities among African countries highlight the different levels of developmental success and failure within the continent. Hence it is essential to study the region in relation to itself, rather t han to highlight the 2008) was a tec hnocrat who started his career in 1968 as Planning Officer and progressed to become Director of Economic Affairs, and then as the Permanent Secretary of the Ministry of Finance and Development Planning. See: See:, htt p:// McCall Smith/e/B001BOPZXG. 7 SUCCEEDS SERETSE KHAMA e/dg_e/dft_panel_e/festus_mogae_bio_e.htm 14 The Magic of Perseverance: The Autobiography of David Magang. Centre for Advanced Studies of African Society. 2008 . 15 2008). He started his career in public service in Botswana as a Planning Officer in 1968 and progressed to become Director of Economic Affairs, and then Permanent Secretary of the Ministry of Finance and Development Planning.


25 numerous ways in which Africa has fallen short of standards set by such countries as Argentina, Brazil, China, and India. 1.3 Scope of the Study This study differs from studies about instrumental disorder as argued by Chabal and Dal oz (1999), and also differs from studies about corruption, petty or grand, along the lines suggested by Rose Ackerman or Michael Johnston (Ackerman 1999; Johnston 2005). A focus on neopatrimonialism, 16 or clientelism 17 is also left out of the discussion here . This is a study about the challenges of enhancing administrative capacity in an affective reciprocal context. My central claim is that a key requirement of legal rational administration, the direct supervision of lower ranking officials by higher ranking ones, is made more difficult in reciprocal interactions being used as the de facto insurance model by millions of Africans. 18 Officials of all ranks devote significa nt portions of their time and resources towards helping relatives and friends while devoting less of their time and energy to the business of the state. By so doing they undermine their own efforts at creating the conditions necessary for economic developm ent. I argue that it is this phenomenon that has hampered decades of civil service reforms. 16 van de Walle 2001; Richard 1987 17 Lemarchand 1972, Eisenstadt and Lemarchand 1981, Roniger, Luis and Güne Ayata 1994, Piattoni 2001, Stokes 2005, Kitschelt and Wilkinson 2007 and Wilkinson 2007. Eisenstadt and Lemarchand 1981, Clapham 1982, Jackson and Rosberg 1982, Me dard 1982, Eisenstadt and Roniger 1984, Clapham 1985, Eisenstadt 1995, Clapham 1996a, Clapham 1996b, Clapham 1998a, b Kasfir 1976, Schmidt 1977, Jackson and Rosberg 1982, Nelson 1987, Bates 1990, Ellis 1996, Posner 2005 Lemarchand 1972, Schmidt 1974, 1977, Medard 1982, Roniger and Güne*s Ayata 1994, Weder 1995, Kitschelt 2000, Piattoni 2001, van de Walle 2003, Wantchekon 2003 18 The networks of support and communication amon (Hyden: 1980:8).


26 affection as the de facto welfare model for millions of Africans, a model i n which affective ties and interactions are geared mostly towards limiting risk and protecting and promoting the well being of the individual and the group. My research finds that most of the money spent in this milieu is directed towards defraying costs a ssociated with basic human needs, among them healthcare and tuition related expenses. Because the currency and premiums of this welfare system are paid in cash and in kind everyone, young and old and rich and poor, are able participate. I then explore the effects of this insurance model on the individual. The final step in the argument involves the examination of the impact of this affective insurance model on bureaucratic and state capacity. In a region where poverty is endemic and governments are unable to provide basic services, affective reciprocity takes on added meaning because its importance is reinforced daily. Many Africans face old age without pensions or other government subsidized safety nets because ibrant informal sector. 19 However, few experiences are as searing as seeing the once self sufficient well to do, bowed by age and ill health, dependent on the good graces of relatives and friends. In addition is the fact that high mortality rates underscore the fleeting nature of human existence and reinforce the need for affective reciprocity. In this environment, individualism is frowned upon and obligations to the group lionized. Affective premiums, whether paid in cash or in kind, are still costly, time consuming, and burdensome. Low wages, under employment, or unemployment, and general poverty only deepen the burden. The necessity of these extractions notwithstanding, they are sometimes 19 Saharan rkers work in the informal sector. championing inclusive growth across africa/post/recognizing africas informal sector 11645/


27 oth the individual and, depending on the kind of and level of infraction, are high for close relatives and associates as well. Lost or diminished opportunities, denial of, or withholding of future assistance, and stigmatization are a steep price very few, even the rich, can afford. To understand the persistence of a practice many find less than optimal, I suggest we recast the argument as a reversed Olsonian collective action dilemma where the provision of collective benefits constrains individual behavior and options. 20 The essence of this reversed dilemma is this: individuals are born into families not of their choosing, inheriting a cache of affective obligations and rights at birth. And although these networks may change with age and time, very few would ever fully escape their affective obligations. Unlike in an Olsonian collective action dilemma, individuals have to actively take steps towards opting out of affective interactions if they seek to be completely free. However, the people in this study are respect. By not exercising their exit option they also fail to act individually or collectively in achieving the best outcome for themselves and the rest of their fellow citizens. At the individual level t maximizing the collective benefits. At the group level, reforming the system to make it more transparent and efficient while giving individuals more discretion about wh o, when, and how to invest their affective currency to be cashed in as needed is a plausible goal. A third option would be a functioning, responsive state that provides services and social protection from taxes citizens pay. Instead, a sub optimal status q uo persists because it is a rational strategy where individuals can invest as much or as little as they choose, and as long as they are not perceived as gaming the 20 The Logic of Collective Action: Public Goods and the Theory of Groups . 1965.


28 system, they can enjoy protection in numbers. Since the incentives work similarly for everyo ne, the net effect is the persistence of affective reciprocal interaction and a perpetuation the system. But as long as everybody plays by the current rules, and as long as there are no credible formal alternatives, the rational strategy is that of continu ed engagement, even if barely. For civil servants, participating in affective interactions sometimes means redirecting resources aimed at improving the lives of the many towards the fulfillment of a few narrowly targeted obligations. Officers are well awar e that their continued engagement undermines national development goals and limits their personal economic prospects, but they choose to meet their affective duties anyway. They participate because there are no credible alternatives to protect them against sanctions. Neither are there any systems in place that could provide cover Affective interactions are multi level, simultaneous dyadic interactions in which people are either giving or receiving or doing both at the same time with any number of people at a time as in the following example in which two directors try to negotiate their way through a thorny situation. The Director of the De partment of Transportation (DOT) is scrambling to fill a request from the director of Veterinary Services who has requisitioned for a four pickup truck convoy for a two man field team to evaluate the final stages a project. The Nubian Ibex Goat Project (NI GP), aimed at combating the effects of climate change, is in its final stages. The Director of veterinary services calls the DOT director to request a personal favor, not for himself, but for one of his staff. The director wants a package sent to one of hi s officers. The package in question contains 10 packets of roofing sheets this officer had just purchased to be delivered by his colleagues. Unfortunately for the staff of the goat project, DOT has only two pickup trucks due


29 to a lack of spare parts. Fitti ng in the equipment and other supplies the substation needed was challenging enough; therefore, as much as he wanted to help he, the director of DOT, could not. Both directors know the Harmattan season was just starting, and the need for the roofing sheet s was only going to get worse. 21 The DOT director suggests sending half the shipment of sheets. They both know full well a partially roofed room during that particular time of the year was futile. Officers deal with competing claims from colleagues, friends and relatives daily. Best outcomes for individuals and the state often run counter to each other. On a personal level the best outcome for both directors is to send the roofing sheets along with the team at no personal cost to either director. The ministr y, after all, pays both the driver and field officers per diem in addition to other allowances any time they travel. The ministry of course loses out, and quite possibly will continue to do so going forward because a precedent has been set. The directors c ould also strike a compromise by shipping some of the roofing sheets. Again, the state bears the costs as before, and both men expect to increase their individual stock of social capital by showing willingness to compromise, costs notwithstanding. Or, the DOT enough to even push the point, the other for not being a team player. Yet as civil servants, perhaps the request should never have been raised at all. Both the well as the written reports, will be produced on time because the field team made the trip on time. And perhaps the farmers, depending on the results of the evaluat ion, could potentially benefit from the project as the region gets drier and at no extra cost due to project over run. 21


30 The second alternative, offering to take some of the roofing sheets, satisfies all but benefits none, and the wastage created could pote ntially undermine the project. Also, by sending just half the shipment NIGP needed, the DOT director could potentially be viewed as incompetent for failing to adequately plan for and properly supply the sub station. The Ministry of Agriculture proves once again that no matter how much donors are willing to contribute, 22 At issue is not officia ls misappropriating national resources, corruption after all is not unique to Africa and Africans. Officials the world over sometimes funnel national resources towards private ends. The issue for those seeking to enhance capacity are the structural conditi ons, the set of factors that make this kind of behavior possible on a wide scale that cannot easily be changed without fundamental disruptions to both society and economies. The existing sanctions regime and the lack of credible alternatives to participati ng in affective reciprocal relationships make s opting out extremely costly. 1.3.1 The Art of Formalizing the Informal Strictly speaking the logic of a collective action dilemma implies that the individual is always better off defecting Hardin 1968, Platt 1973, Hardin 1985, 1991, Rothstein 2005 . Officials may be able to improve capacity by limiting, or refusing to divert state resources towards particularistic ends, but to do this officials also need the state to p rovide them cover. As things stand there are no viable alternatives for hedging risk outside of the Affective Insurance system, so officers continue to pay their premiums. Pressure from family and friends makes it 22 W. I. Thomas and D. S. Thomas (1863 1947) T


31 extremely difficult for officials to break away from actions they know subvert institutional capacity, creating a dynamism and feedback loop (Figure 1 1 at the end of the chapter) that makes it extremely difficult to opt out or reform the whole system of favors, government ineffectiveness, and ine fficiency. As a result it is likely that by holding constant the levels of service provided by the state, higher levels of affective reciprocity should be negatively correlated with state capacity and vice versa. The current state of affairs endures for a number of structural reasons. A symbiotic relationship between affective reciprocity and threats to subsistence have created over time a plethora of hydra headed informal complementary institutions that are not just resistant to change, but also present c hallenges in terms of sequencing of change. Indeed, these arrangements can only survive within certain social and political conditions, what Chabal and Daloz have termed instrumental disorder Chabal 1999 . On one level, decision making in a context of the economy of affection is not structured with a view at arriving at a strategic outcome meant to benefit the public as a whole. On another level; however, decision making is opportunistic and ad hoc, and allows for change to or adherence to, for various reasons, the normal stability required in institut ionalized organizations, and thus may be a potential threat to the usual expectations for public institutions. These dispositions are resilient because they are flexible and can be fashioned towards new ends ( Bourdi eu 1977, Dryzek 1996, Offe 1996, Galvan 2004, Sil and Galvan 2007 . They are also unified in providing an individual a way to relate to and view the world. Thus, affective reciprocity driven interactions blur the boundaries between the state and society. Furthermore the preferences of the actors in an economy of affection are not as well understood or studied as much as they are in conventional mod els of publ ic choice decision -


32 making . This makes it a lot harder to argue for a realignment of preferences thr ough incentives such as a better social security regime. Better working conditions, higher wages and other incentives do not really get to, or fix, the problem as I see it for a number of reasons. For one, governments and their development partners have so far failed to create a social security system capable of producing the benefits of the economy of affection. Secondly, not even the more well to do African countries like Botswana can afford the range of critical services needed to free individuals from t he current economy of affection based social security model, nor are there any public/private partnerships aimed at this issue. Another reason is that neither civil servants, development practitioners, nor academics have found a way to scale up the benefit s of the economy of affection. Nor have they found a way to leverage the informal, but effective, sanctioning mechanisms of the economy of affection to work at the macro level. Rather what we have is a crystalized informal institution with a dynamism and p ath dependency of its own, one that is perverse from the point of view of improving public sector performance. Improving public sector performance from this perspective is not so much a question of incentives as it is about sanctions and the provision of c redible alternatives to affective reciprocity. 1.3.2 Situating the Dissertation within the Africanist Canon (discussed more fully in Chapter Three) aimed at addressing them have had mixed results. From Africanization, to centrally planned multi year development manifestos, to export substitution and neoliberal derived austerity programs, experts have struggled to create and sustain the conditions needed to improve the lives o f the majority of Africans. And even though progress has been made on many fronts in the study of African politics, especially in the areas of democratization, much remains to be done on the messier, more fundamental cluster of variables that make up some


33 1.3.3 Informal Institutions and the Lack of Satisfactory Answers The pre dissertation research that informs this project sought to highlight the positive aspects of the economy of affection. This phase of the work had anticipated a positive effect between affective reciprocity in resource poor bureaucracies and getting things done as official evelopmental patrimonialism perspective (Kelsall 2009:2). But after two summers of intense fieldwork from 2008 and 2009, I found that the data suggested otherwise. Affective reciprocity was negatively correlated to bureaucratic performance. This finding is not new. Robert Price, for example, found in his study of Ghanaian civil servants that the benefits, wealth and status that came with a career in the public sector created unsustainable expectations of reciprocity. Civil servants, regardless of rank and wage scale, were expected to support their relatives, creating enormous pressure and prodigious stress on officers and eventually hindering bureaucratic performance (1975). Similarly Albert Hirschman found that sharing dampened economic activity ( Hirschman 1958 ). Agreeing with Hirschman, Hoff and Sen also found that kin based mutual assistance systems tended to, on average, create dysfunctional institutions that hindered economic development (Ho ff and Sen 2005:4). These forms of extractions, critics charge, hinder the important aspect of affective reciprocity, the need to adequately separate the normative/ moral dimensions of affective reciprocity from its economic dimensions. Holding all other factors constant, relatives are more likely to meet demands they consider legitimate; a category of demands best described as basic needs primarily food and shelter and, increasingly, education related costs whether they be learning a trade or pursuing formal education. With this in mind it seems quite reasonable to expect economic factors rather than normative values drive much of


34 the interactions in economies of af fection. If this is true, we should observe differences between officers and farmers in; for example, how much time and resources they devote to helping relatives and friends and the type of assistance relatives request or the kinds of aid offers extend to those in need. Whether or not the officers and farmers we encounter in this study fulfill these expectations should tell us something about affective reciprocity and the ensuing reversed collective action dilemma. We now turn our attention to a discussio n of the research design , the case selection, methodology and my justifications for choosing a mixed methods approach to the question in section 1.4. The study is informed by a combination of ethnographic approaches. This multi method technique to test the reversed collective action dilemma thesis also allows for process tracing of the expected causal relations as well as generalizing its findings. 1.4 Research Design and Case Selection Measuring the full capacity of the modern state is fraught with aggre gation problems. Thus any attempt at mapping out and tracing key factors believed to shape its capabilities are best broken up into manageable units of analysis. To get around these difficulties I employ a combination research methodologies to better explo re of the patterns of affective reciprocity posited as detrimental to enhanced capacity in one department in one ministry. An outline of the research design and the rational for case selection adopted in the study are presented next in the sections. 1.4.1 Why Study the Ministry of Agriculture? From the English bread riots of 1817 to the violent street protests in Tunisia in 2010, we economic growth and social an d political stability. The importance of the agricultural sector to


35 African economies cannot be overemphasized. According to the World Bank, this sector ( Thompson 1966 ; Clark 1940 ; David Ricardo 1952, Kuznets 1966, Ranis 1988 ; Bates 1981 ; Leonard 1977, Bates 1983a, b, Chazan 1988, Mann, et al. 1989, Bevan, et al. 1993, Collier and Gunning 1999, Hassan 2002, Hassan 2002 ; MacDonald 1989, Mann, et al. 1989, Diao 2006 ; Lipton 1968, Lipton 1977, 1987, 1996, 2009 ) . Subsistence farmers, small scale fishermen, and livestock farmers produce about 80% of the staple foods and animal products consumed in the region. However, continent wide, agricultural productivi ty has been declining since the 1960s. practices, inequities in land distribution, and entrenched poverty. Civil strife and conflicts, natural disasters and ant iquated farming practices have only compounded the problem, and it is for these and other reasons, that I choose to focus on the extension officer and the subsistence farmer. To do this I focus on two policy goals of both ministries: (i) increasing farmers and (ii) increasing agricultural productivity within the backdrop of the larger context of the political economy of the region. 1.4.2 The Road Less Traveled The study departs from the well worn paths of contemporary area studies in a number of w ays. For one it eschews the high politics of the powerful for the less glamorous footpaths of the farmers, they are usually not involved in policy formulation (Scott 1997:130). For the extension officer, changing farming practices is as much about relationships as it is about carefully crafted, well implemented policy. A s street level bureaucrats, extension officers are a crucial link


36 between the citizen and the state. They also, depending on length of tenure, have a deep understanding of the communities within which they live and work and thus are repositories of knowled ge. Likewise, the subject of the study, the economy of affection, like the farmers and extension officers in the study, is not glamorous. Unlike corruption, described with such facility, d themselves to neat, crisp definitions up or down the proverbial ladder of abstraction at least not during these initial stages of inquiry ( Sartori 1970 ). Training our lens on the actual actions of farmers and extension officers; therefore, lays the groundwork for exploring the effects of affective reciprocity on administrative practices even as it contributes to the growing body of work firmly grounded in the concept. 1.4.3 Comparative Case Studies To more effectively explore the effects of the economy of affection and its patterns of reciprocity on administrative capacity I use a comparative case study approach. The cases were selected using a most similar systems design approach because they share important similarities as well as variations on the main variables of interest to us and thus allow for a systematic exploration of the patterns of affective reciprocity p osited as detrimental to enhanced capacity. Comparative case studies also have the added advantage of acting as a control variable even as it allows a range of variation in the main independent variables King, et al . 1994, Geddes 2003 ;) The cases do two things. First they are used test the hypothesis that demands made on sapping interactions we have come to associate wit h African civil service. Second, as critical cases, Botswana and Ghana provide important insights for theory building and refinement (Adcock 2006; Adcock and Collier 2001a, 2006; Adcock and Collier 2001b; Alexandrova 2000,


37 2009; Annez and Buckley 2008; Ben nett and Elman 2006; Capoccia and Keleman 2007; Collier 1993; Culpepper 2005; Eckstein 1975; Eisenhardt 1989; Flyvberg 2006; George and Bennett 2005; Gerring 2007; Greif 2006; Goertz 2006; Munck 2004a, b; Munck and Verkuilen 2002, 2005; Odell 2001; Pierson 2000, 2004; Przeworski and Teune 1970; Ragin 1987, 1989; Ragin, Rueschemeyer and Skocpol 1996; Sartori 1970, 1991, 1984; Seawright and Gerring 2008; Skocpol 1979; Thelen 1999; Thelen and Steinmo 1992; Walton 1992). Considered strong, capable states, the two countries underscore both the importance of informal institutions and the difficulty of enhancing bureaucratic capacity. They also provide a framework for the systematic exploration of the internal dynamics of informal institutions like the economy of affection and its 1.4.4 Why Botswana and Ghana? Botswana and Ghana form an interesting pair of countries to compare. In a continent characterized by one party dictatorship s, ethnic conflicts, military coups and state collapse, Botswana has been a model of democratic governance and, as such, is treated as an exception in comparative development analysis of Africa. The country also has many of the hallmarks of an institutiona lized modernized state. Its functioning rational bureaucracy, staffed for the most part by a technocratic cadre of civil servants sets it apart from other African countries (Acemoglu 2001, 2002, 2003; Dale 1995; Danevad 1993; Dissez 2004; Edge 1998; Fawcus and Tilbury 2000; Goldsmith 1999; Harvey and Lewis 1990; Holm and Molutsi 1989; Lange 2009; Leftwitch 2000; Leith 2004; Lewis 1993; Masire 2006; Otlhogile 1974; Picard 1985, 1987; Samatar 1999; Steadman 1993; Thumberg Hartland 1978; Vengroff 1977; Webner 2004). Yet not too long ago, Botswana was a neglected dusty protectorate useful to the British colonies. The British did not earnestly start developing the country until 1955. The government


38 of Botswana expended considerable resources in its quest to transform subsistence agriculture policies, focused mostly on arable fa rming and cattle production, have been well funded and land tenure reforms also set it apart from other African countries which, for the most part continue to ignore the problem. majority of African countries. At independence in 1957, Ghana seemed destined to live up to its image as the Black Star of Africa. Then the world's leading exp orter of cocoa, minerals and timber, Ghana had a well developed transport network, a relatively high per capita income, low national debt, considerable foreign currency reserves, and a comparatively advanced education system and work force. The prestige it enjoyed upon its independence, both within Africa and internationally, was unrivaled in the region (Berry 1975, 1993; Chazan 1982; Herbst 1993; Hill and Austin 1997; Austin 2005). However less than a decade after independence, Ghana experienced the first of significant deterioration, stagnation and short lived improvements. A study in squandered opportunity, Ghana became the poster child of much of what was wrong with Africa. Pa radoxically the very men from the barracks whose actions contributed in its agricultural sector. General Ignatius Kutu Acheampong, the sixth head of state of Ghan a, is agricultural sector. Yet, along with the charismatic Commissioner for Agriculture, Colonel Bernasko, General Acheampong championed Operation Feed Yourself (O FY). Decades later the


39 indelible mark on aspects of Ghanaian culture and social relations. Interviews with retired officials from the Ministry of Agriculture and the Ministry of Defense underscore the role of the regime in banning the importation of what were then called essential commodities: corned beef, polished white rice, granulated sugar and canned milk. Starting with OFY, according to those in the know, foo Perhaps it is not surprising that the Provisional National Defense Council (PNDC), the military junta that gove rned the country from 1982 to 1992, was an avid supporter of the farmer. Many of the upper and mid level officials who crafted and implemented these policies were themselves part of the OFY generation when food self sufficiency was the national policy. The ministry, under various able ministers during this era, relentlessly promoted a diversified agricultural sector, especially one removed from cocoa production. The 31st December Revolution Women's Movement under the leadership of Nana Konadu Agyeman Rawlin gs, then the first lady, promoted and supported cottage industries. With the increased attention drawn to road network, and increased investments in supp ort services. The jury, though, is still out on the 1.4.5 Fieldwork To explore the persistence of limited administrative capacity within the two departments, surveys and in depth interviews were conducted in two districts from two regions in each country. Fieldwork for the Botswana cases started in August 2010 through February 2011 and from May through December 2011 for the Ghanaian cases. As in Ghana the first two and half


40 months of fieldwork were spent in the capital Gaborone. During this period I worked as a research affiliate with the Department of Extension Services Coordination provided invaluable experiences and opportunities for data collection. Participants for the study were selected using a variety of techniques. First, farmers were randomly selected from purposively chosen regions and districts. Next, farming households were then selected using a combination of systematic sampling with probability proportional app roaches due to the different sizes of the farming communities surveyed. I worked closely with officials at the district level during this phase of the research. The selection of extension officers was slightly different. First, where possible, all extensi on officers were surveyed at the four sites of fieldwork because they are so few. Next, at the headquarters level three levels of personnel were systematically selected 23 to ensure a fair coverage of upper, middle and junior officers. 24 From this sample, pa rticipants were then randomly selected to fill out the survey to better explore some of the different dynamics hypothesized as important for bureaucratic effectiveness and efficient service delivery. In Botswana, pretesting of the questionnaires was done using a few staff from the Gaborone for the farmers. The questionnaire was not pretested in Ghana; instead I collaborated extensively with the Extension Services Di rectorate in tweaking the questionnaire to fit the Ghanaian case. 23 Even though we value the work of drivers, cleaners, security guards, messengers, trainees, typist, carpenters, and mechanics, they were systematically excluded from the sample. Depending on the size of the institution, only 19 to 36 per cent of the to tal staff was therefore considered eligible as potential respondents. 24 Even though we included public servants working for the Ministry of Local Government, employees working directly for local governments were excluded. In Zanzibar there are two types o f local government staff, those who are employed directly by the government and are considered part of the public service, and those who are employed directly by local governments; these employees often tend to be employed on short term basis and overall t end to be low skilled, as local government in Zanzibar is still in its developmental phase. This group of officials was excluded from the sample.


41 1.5 Overview of the Dissertation The rest of the dissertation is organized as follows: Chapter Two explores the effects of the governing techniques employed by pre colonial, colonial and th e post colonial elites on enhancing capacity. The few studies that have explicitly focused on informal institutions within the Africanist literature have tended to focus on different aspects of informal institutions which makes comparisons difficult. So Ch apter Two also focuses on the link between patterns of governing, the lack of basic welfare guarantees, and the resultant reversed collective action dilemma. The chapter ends with a discussion of the research designed to evaluate the hypothesis. Chapter Th ree offers a very brief overview of the state capacity literature with an eye on the definitional, operationalization and measurement problems within that corpus as they pertain civil services and bureaucrats more effective. Chapter Four deals with the methodological issues pertaining to this study. In Chapter Five, the first of the three empirical chapters I shift my attention to the survey data on Botswana and Ghana by introduc ing the survey data for the two cases. Chapters Six and Seven explore the answers advanced by the reversed collection action dilemma thesis to understand the differences in state capacity in Botswana and Ghana. The focus of analysis in Chapters Six and Sev en is the individual. Together these two chapters suggest living in a high turn increases the odds of engaging in affective reciprocal interactions even if the ind ividual in question does not entirely share these attitudes and who, by so doing, acts to reinforce the reversed collective action dilemma thus ensuring the undermining of state capacity. 1.6 The Relevance of the Study One factor out of many that might hel p deepen our understanding of political behavior is the role affective norms play in influencing state capacity as reflected in bureaucratic


42 performance. This study highlights the paradox of the economy of affection and its ability to create, on one hand, inefficiencies in resource allocation and policy implementation and, on the effects of informal norms of reciprocity, the study also provides a way of anal yzing the effects of these norms in a more holistic manner. The study also contributes directly to the state building literature advancing our understanding of political economy in Botswana, in Ghana in particular, and in Africa in general. Additionally, t he study provides some food for thought for policy makers and civil servants about what can or should be done to mitigate some of the reasons for affective reciprocity. 1.7 Conclusion This study investigates the critical impact of affective reciprocity o n administrative capacity with a focus on the agricultural sector in Botswana and Ghana. To accomplish this, the study uses three forms of causal inference. First, a large N cross national analysis is conducted to affirm that stronger affective norms are a ssociated with diminished state capacity. Second a small N historical comparison of the developmental trajectories of Botswana, a case of sustained progress over time, and the trajectories of Ghana, where state capacity gradually improves over time, are us ed to underscore the difficulties of improving administrative capacity even within


43 Figure 1 1 Affective Reciprocal Feedback Loop.


44 CHAPTER 2 THEORIZING THE REVERSED COLLECTIVE ACTION DILEMMA 2.1 Introduction On Thursday, April 5 th , 2012 President Bingu wa Mutharika of Malawi died of cardiac arrest in Blantyre, Malawi. He was 78 and in his second term as president. His death would be kept secret for 18 hours. An on going electricity crisis prevented Malawian do ctors from performing an autopsy or refrigerating the body, so officials had the body airlifted from South Africa. The president was buried about two weeks later in a white marble and granite mausoleum at his sprawling farm. 1 Findings from the Commission of Inquiry constituted in June 2012 to investigate circumstances surrounding the death reported that the government of Malawi spent $1.5 mil lion on the state funeral. 2 nature of the post colonial African state. While the decision to fly the body of the president to South Africa could well be driven by political c alculations as his various successors jockeyed for position , it also goes to show the functional weakness o f the Malawian state. If, after eight years in office the most powerful man in the country is left to rot due to an energy crisis, one can only imagine what happens to the proverbial man on the street. The acuteness of the post continent with the worst development record and the highest concentration of countries with heart wrenching tragedies , Africans have somehow always managed to pull through with the 1 vism9dWptMaQy.html 2 report death malawi president bingu wa mutharika


45 help of foreign aid but, more importantly, by drawing on its resilient informal institutions, especially the economy of affection. The posited effects of this economy of affection on bureaucr atic capacity is the subject of this chapter. The basic idea upon which my argument rests is simple. Due to a number of structural factors many Africans live close to the margin. With no safety nets from the state, the economy of affection has become the i nsurance of last resort. 3 Given its importance, violating the taken for granted mores undergirding affective reciprocity is actively discouraged, a reaction not old European peasant societies studied by Eric Wolf are cases in point (Thompson 1963; Wolf 1999; Moore 1966; Redfield 1960). The theoretical and conceptual frameworks, along with the relationships between and among the primary variables of interest adopted for the study, are discussed in this chapter and the next. One of the main tasks of the chapters is to place these concepts within their disciplinary context and, in so doing, to refine and operationalize the key indicators central to the analysis at hand. The chapter is divided into two main sections. In Section 2.2 I weave together the work of four scholars to support and advance my claim that the economy of affection is the by product of the governing techniques of elites in power seeking to maximize the ir power. In Section 2.3 I outline the contours of the reversed collective action dilemma and the dynamics through which it weakens capacity. 2.2 Situating the Study within the Africanist Cannon 3 ecology, climate conditions, geography, and unresponsive governments with a propensity for bad policies have conjoined over the years to create a situation in which the majority of Africans live with great deprivations.


46 If risk management is the catalyst for the economy of affec tion, its structuring properties are the crucible that not only reconfigures its failure. 4 How this process unfolded is the subject of the next section. Between 1980 and 1983 Goran Hyden published two books. The first, Beyond Ujamaa in Tanzania: Underdevelopment and an Uncaptured Peasantry , argued that the inability of the post colonial African state to fully capture or systematically incorporate the subsistence farmer into its orbit posed serious devel opmental challenges for the post colonial state. His second book, No Short Cuts to Progress , building off the first, elaborated his economy of affection thesis introduced in Beyond Ujamaa. According to Hyden, the low productivity of subsistence farming le ft little by way of surplus that could be reinvested into expanding or scaling up farm operations. Due to this low productivity, farming households also found it necessary to diversify household incomes, usually through migration in search of wage employme nt. But unlike in other parts of the world the move from the village did not sever the link between the rural and urban extensions of the household. Rather, the income from wage labor gave the subsistence farmer a degree of autonomy and an extra layer of p rotection over production, decisions rare in other parts of the world for similarly situated groups (Hyden 1980:9). These arrangements, once layered over the imperatives of a monetized economy, produced a particular kind of subsistence ethic Hyden calls th network of support, communications and interaction among structurally def ined groups 4 The moral underpinnings of the economy of affection stem from its ability to harken back to age old ideas about, durable, transposable di ( Logic, pp. 53 and 55; see also Outline, p p. 78 8; quoted from Gaunt and Kay 1997: 194) accounts for the proliferation of the practice.


47 layer of protection provided by the economy of affection allows peasants to elude state capture. A year after the publication of Beyond Ujamaa , the Un iversity of California Press issued Markets and States in Tropical Africa . Bates offered a different set of answers to s were due largely to the growth undermining policies pursued by self interested elites and that these polices, at their core, were predatory. Thus they were antithetical to growth due to rent seeking (Bates 1981:8). In a debate not unlike that between Jam es Scott and Samuel Popkin on the nature of peasant economies. 5 Mostly agreeing with Bates, Lemarchand argued that peasants made decisions from a rational calculation of a cost/benefit analysis and were as likely as any other group to make economic decisio ns based on during a particular crop season in order to take advantage of the higher prices the state was paying was as much of an economic decision as it was about the Hyden, Nelson Kasfir noted gave the subsistence farmer too much autonomy even as Gavin Williams wondered if the appropriation of peasant surplus through graft, corruption and nepotism was really peculiar to Africa (Kasfir 198 6: 33 58; Williams 1987 637). These debates within African studies coincided with the rise of neoclassical economic models and prescriptions, market driven efficiency, and the rational pursuit of carefully weighed preferences. While Tony Waters suggeste conclusions and Cliffe faults his deployment of a nondeterministic materialist methodology 5 For Scott we cannot fully appreciate why and when peasants choose to risk all by rebellion without understanding the strongly held social and cultural values g uaranteeing a right to subsistence. Samuel Popkin on the other hand minded economic actors who carefully weigh their options (Scott 1976; Popkin 1979).


48 (Waters 1992; Cliffe 1987). Moral economies of Scott and Hyden and their methodology full of thick descriptions were losing ground to parsimonious formal models. In time the pendulum Scholars of state formation such as Charles Tilly contend the basis of power in European state formation derived from control over territory ( Tilly 1990 ). In Africa, where land was abundant and populations small, control over people, rather than territory became the basis of power (Herbst 2000:36 39). How this came to be is the subject of the next section. 2.2.1 Sowing the Wind Pre colonial African rulers, faced with the costs and logistics of governing sparse populations scattered over vast areas, would settle on the strategic use of resources to consolidate power. By controlling communal labor, access to land a nd other reproductive resources, the ruling elites were able to woo and retain followers. 6 For the ruled, economic, political, and social mobility meant staying in the good graces of those in power (Boone 2003; Feierman 1990, 1995; Kitching 1980; Lonsdale 1986; Spear 1997; Vansina 1990). Yet rule over people, by its very nature, is political; it works in so far as the ruled consent to be ruled. 7 Europeans, faced with the 6 For example, control over women was used to attra ct young men into households and villages. Resources transferred into these households and villages in the form of paying a bride price and having children boosted the power of heads of households or village leaders. To gain access to land or cattle with w hich they could raise the necessary resources for marriage, young men needed the support of older men, be these men fathers or village headmen. 7 The Reverend Hepburn's 1895 memoir of pre colonial Botswana offers a window into the past via Chief however, was at hand. The people were gro wing weary of the chief who was virtually no chief, who never could be gotten to give a decision; from whom the injured could obtain no redress, and who left disputes to right themselves, or what was more likely, to become more complicated. In the end of A ugust 1872 the culminating point was 1985: 15 retu soon became a general one. For weeks long strings of people women with baskets on their heads, men some with bundles, and others driving pack oxen filed up the kloof. The stream flowed steadily on, until the town of Bamangwato proper became almost deserted, and only a few old men and servants remained with Sekhomi and


49 same conundrum of broadcasting power over vast territories with small scattered populat ions, few navigable rivers, and a harsh ecology, had little incentive to strengthen the tools of territorial control. Thus they adopted the governing strategies of their pre colonial counterparts. What started as a matter of mercantile expediency became in stitutionalized as colonial aspirations and the scope of colonial administrations grew. In time they would institutionalize rule over people through indirect rule. The unintended consequence of this policy was the creation of a bifurcated system with two competing moralities ( Ekeh 1975 ). African chiefs, backed by the coercive force of the European state, became increas ingly despotic. The delicate balance between the ruled and the ruling had changed. Under colonialism the primordial public morphed into something apart from the community. For example, public communal labor was now mobilized to implement public sanitation or agricultural policies based on European theories that were sufficiently different enough to be alien and somehow illegitimate. This by itself was not enough to cause rancor, but combined with colonial taxation policies and its administrating machinery, a failure to clearly communicate to locals the necessity of specific projects over time added to already existing and growing negatives views of colonial machines. The disaffection towards colonial administrations came to a head during World War II when th e imperatives of running war economies coarsened colonial domination even more, creating a backlash with far reaching implications. Peasants, who had silently chafed against; for example, the forced production of cash crops over staple crops, became vocal in their opposition. With less land and less time for subsistence farming and no money to buy food, family budgets were stretched thin. While it was common for young men to resent being forced into providing free labor, the relentless enforcement of labor and conscript quotas turned antipathy into outright


50 resistance (Akurang Parry 2000; Mbaye 2006; Browne 1966; Curtin 1975; Davidson 1994; Gutkind 1978; Inikori 1976; Inikori and Engerman 1992; Lovejoy 1982, 1989, 2011; Nzula et al 1979; Redfield 1960; Rodri guez 1997;). The heavier tax burdens imposed on locals did not help either. Not surprisingly the delicate balancing of interests between the rulers and ruled, especially in rural areas, was set awry (Davidson 1994:63). Political entrepreneurs, mostly west ern educated Africans taking advantage of the power vacuum, set about raising political awareness, stoking dissatisfaction and whipping up popular resentment against colonial administrations. The ruling elite, people were told, were united with colonial ad ministrators with their purpose being repression and plunder (Buah 1982; 2000). By the time the war was over the political sphere had a new set of actors: a fairly experienced, skilled organizer corps willing to take a more militant stance in their struggl e for self determination. Unlike the politicians of yore, these new men took up the grievances of labor and peasants. 8 These internal changes and alignments unfolded against the backdrop of the commodity driven boom created by post war European reconstruc tion. The initial excitement that had young men and women immigrating to cities in search of jobs was short lived. The high costs of urban life precluded them from participating in the transformations they saw all around them. 9 As 8 In Ghana rich coastal merchants like George Alfre d Grant and well to do cocoa farming families from the Asante, Eastern and Western regions financed political organizations like the United Gold Coast Conversation (UGCC) only to lose control to impatient young men like Kwame Nkrumah who had little patienc e with the gradualism of the traditional elites. J. B. Danquah, one of the doyens of Ghanaian politics, to be the General Secretary of UGCC, hired Nkrumah. The fallout between Danquah and Nkrumah over the direction of the independence movement is still an enduring and worrisome aspect of Ghanaian politics. 9 piah, was a Ghanaian politician and lawyer. David Magang, a lawyer turned businessman and politician, once one of the few of my engagement , the hierarchy of the Attorney General, Deputy Attorney General, Assistant Attorney General, Senior State Counsel; and State Counsel. I joined at the very base rwhelmingly white, naturally, as educated or skilled Batswana at


51 Bourdieu pointed out in h is Algerian studies, the quest for economic enhancement and modernity not only created an underclass, it also created hybrid category of people who were neither peasant nor proletarian (Bourdieu, 1958; Bourdieu et al., 1995). In this new uncertain world, m igrants latched onto ethnic connections for material and psychological support. Social practices ions in the young through boon and bane and ever more salient. Meanwhile the nascent educational systems quietly churned out a steady supply of graduates of all s orts who believed their education guaranteed them a role within the colonial apparatus. Disgruntled and disillusioned, it would not take much to set them off. In Ghana it was the Accra Riots of 1948. 10 As in Anouilh's Antigone order; it has march. For the educated and professional classes, precluded mostly by racial prejudice from enjoying the same privileges as European expatriat es, it was education that, instead of 12, 43). Believing the system rigged against them, many would abandon hopes of working for colonial administrations and/or for European firms so they could create their own private practices only to discover the private the time were a very rare species. From the Senior State Council upwards, only Mokama, the Attorney General, and ana in the entire hierarchy: Mokama, David Magang 2008:245. This was in 1969. 10 After a protest march by unarmed ex servicemen was broken up by police, several leaders of the group were left dead. Among those killed was Sergeant Adjetey, [1][2] who has since been memorialized in Accra. In January 1948, the Ga chief Nii Kwabena Bonne III had organized a boycott of all European imports in response to their inflated prices, [3][4] and the 28 February incident is considered "the straw that broke the camel's back", marking the beginning of the process of independence for the Gold Coast as Ghana the first African colony to achieve this. [5]


52 sector also had its own challenges. Colonial authorities favored European firms, the majority of which were Wes tern owned and operated, in the granting of contracts and access to labor and licenses (ibid; 12). In the fight to discredit the colonial system, workers were encouraged to strike. Sabotage 11 Exhorted to evade taxes, citizens happily obliged. 12 Hitherto frowned upon, insubordination towards authority, colonial and traditional, became the duty of every patriotic nationalist. 13 Nnam di Azikiwe, the editor of the renowned African Morning Post , a daily newspaper in Ghana, promoted the African nationalist agenda through the newspaper . 14 Lacking pre existing justifications on which to base their claims on the soon to be post colonial stat e, nationalist leaders took their case straight to the masses whom they portrayed as ill suited, and lacking the disposition needed to confront the demands of enlightened modern governance (Collins 1970:87). The tactics employed to discredit the colonial s ystem and cobble 11 This is still a popular Ghanaian saying and attitude. 12 Across the conti nent, rural Africans resisted taxation, land alienation, compulsory cultivation of cash crops, and 1898 occurred, and in 1900 the Asante of Ghan a revolted against direct taxation, compulsory labor, and the introduction of Western education. In Nigeria the Ekumeku rebellion lasted for 13 years (1893 1903). ( The Ekumeku Movement. Western Igbo Resistance to the British Conquest of Nigeria , 1883 1914 by C. Dhadike). The Ndebele Shona (Zimbabwe) chmurenga (rebellion) of 1896 97 was brought on when the Ndebele and Shona blamed Europeans for the decimation of their cattle by rinderpest, for drought and plagues of locusts, for stealing their land and cattl e, and for acts of rape and murder. The Herero revolt of 1904 was an ill fated attempt at stopping continued Herero people were killed in the 13 non mah launched his Positive Action campaign with the goal of fighting imperialism through nonviolence and the mass political education of citizens. The protests; however, devolved into riots as protestors turned into looters. When all was said and done, 29 p eople would lose their lives and another 237 would be injured. The colonial administration declared a state of emergency imprisonment. The sentence was reaffirmed by the High Court on appeal. 14 Nnamdi Azikiwe migrated to Ghana (then the Gold Coast) where he founded the African Morning Post . He would later become the first president of Nigeria (1963 to 1966).


53 together the buy in of key elites and important voting blocs created a dynamic where the to kith and kin crossed a hitherto invincible line ( Ekeh 1975, 103). 15 The morality of the private sphere had heretofore always hovered around margins of the civic public. But now primordial attachments together with colonial ideologies of legitimation now inverted, turned inward against the very system it had once sustained (ibid: 96). 16 Right or wrong as defined by the private sphere no longer extended to the public sphere. Ghana, then the first country in Africa to gain independence, was blazing a trail other African countries would follow. Religion and ethnicity had become political instruments in forging alliances. Northern Ghana, conquered, long before colonialism, in the words of the Indian born English man, Captain Robert Sutherland Rattray, by better armed, better clothed, familiar with the idea of kingship or chieftainship in our modern sense, in some cases conversant with the rudiments of Mohammedanism and accustomed (even if circumstances had not later compelled it) a patrilineal 1932 : xii ). In fact the Northern territories, as the area was known then, were largely left to their own devices so perhaps it is not sur prising that the first political parties were the Northern 15 The country was then divided into four admi nistrative regions: the Colony (coastal areas other than the Volta Region), Ashanti, Northern Territories and Trans Volta. 16 The creation and propagation of six ideologies: (1) the backwardness of the African past; (2) the lack of contributions by Africans to the building of Africa; (3) the idea that pre colonial Africans were locked in internecine inter tribal feuds; (4) the belief that European colonial rule, on balance, benefited Africans more than it harmed them; (5) the idea that on the whole, the admi nistrative costs of colonization to Europeans was much higher than any benefits gleaned, and last but not the least; (6) the double speak of European colonialism that at once defined the native as both noble/savage, moral/immoral and the westernized Africa only backhandedly justified their penetration into Africa but also justified to their fellow countrymen their continuing actions. In addition, and more to our point here, they also tried to persuade Africans to acc ept European rule as beneficial. These latter attempts aimed at colonized Africans are what I have called colonial ideologies. They were wrought jointly by the colonial administrators and their close collaborators in the colonial enterprise, the Christian 100).


54 was now at a critical juncture of her political history. Meanwhile in Botswana educated members of the r uling classes forged alliances with traditional elite. They also coopted enough of the non royal activists to form a ruling coalitions power more than 60 years party conflicts are intricately interwoven with the destabilizing effects of the Difaqane wars and the reconstruction strategies deployed, most notably, by the Tswana states of the Ngwaketse, Kwena, and Ngwato. 17 Kgari, chief of the Bangwato, started reconstruction at the group level by giving cattle plundered in war to commoners. Previously the spoils of war were the right of the chief to do with as he pleased. It was up to him to share or hoard. By sharing his war boot y, Kgari gained the admiration and loyalty of his followers. With the loyalty of his followers secured, Kgari and his sons turned their attention to eliminating his political rivals. His state building aspirations were to be continued by his sons Sekgoma I and Khama III. The current president of Botswana, Ian Khama, like his father before him, Sir Seretse Khama, and his great uncle, Tshekedi Khama, 18 Unlike in Ghana where the Western educated elite were able to discredit tr aditional elites, in Botswana the educated and traditional elite would be coopted into the ruling party, the Botswana Democratic Party. The Ghanaian and Batswana approaches to post independence realignment are political bookends in the range of approaches to consolidating power. Kenya presented a slightly different 17 Without getting into the debate about origins and charges of Zulu centricity, Mfecane/ Difaqane, a Sotho wording referring to wandering hordes of people (Tlou and Campbell: 2001: 146), is used here to describe the wars between and among the great Zulu and Sotho tribes as they fought each other for space and domination throughout Southern Africa, killing, capturing, and displacing hundreds of thousands of people across the sub continent. 18 text/classic/blpr.htm


55 model in which political opponents and very senior civil servants were either coopted into KANU, imprisoned or harassed into exile and in some instances assassinated (Throup and Hornsby, 1998). T he effect was an effectively emasculated opposition and a s one party system. In Tanzania, the choice to neutralize ethnicity, forge a common identity, and reorganize society fared no better in terms of economic development. The model also never gaine d traction among African leaders. In the long run the forging of narrow coalitions in Botswana, though qualitatively different, would also undermine the institutionalization of robust participatory democracy. One by one, African elites repackaged and used begotten colonial than the traditional elites, were more suited for the enlightened business of modern governance (Ekeh 1975: 100 104; Osaghae 2006:235 ). 19 The divide and rule tactics deployed to gain control over the state would be devastating for Africans, who emerged from these struggles as the ultimate losers. The case of Congo is illustrative. Less than a we ek after Congo was declare d independent from Belgium rule the army mutinied. Belgium stepped in to protect Belgian of office. Five months later he was dead. Congo has known no peace si nce. Seven years after independence the ethnic, cultural and religious tensions exploited to win elections backfired, plunging the country into a civil war. By the time the war was over three years later (1967 to 1970), about a million people were dead, ma ny of them from starvation. 19 The traditional elites, mostly chiefs and their courtiers were aligned with the colonial powers. Nkrumah, for example, effectively argued that these chiefs had sabotaged the wishes of the people for self government now by advocating a more tempered approach. The chiefs in the Ashanti region did not help matters much when they joined the dwindling ranks of UGCC in supporting the Coussey Report, commissioned by the colonial government to The process was very similar in Nigeria as well.


56 But unlike Ekeh and scholars who argue colonialism matters, I do not subscribe to the notion that the primordial public and its undergirding moralities evolved to fill the gaps created by colonialism. European colonialism was t oo brief and too limited in scope fundamentally change existing social structures. The colonial state, save for a few cases, lacked the capacity to reorder society as was the case in Eastern Europe during the consolidation phase of Soviet state building. T he levels of economic, social and political upheaval, bad as they were, were not nearly enough to make people seek that kind of exit from the state. I do however, agree with Ekeh and Osaghae rovide public goods and services the state is unable to provide (Osaghae 1999; 2003:6). I will argue throughout this study that the development, dominance, and continued relevance of affective morality is still very much about basic survival. The morality of the primordial public was already well developed by the time colonialism made an appearance on the continent which is why colonialism was unable to transform or bend it to new ends. Colonialism did not create the primordial public, though it did rarefy it, just as the post colonial state made it even more important for survival and economic, political, and social mobility. 2. 2.2 Reaping the Whirlwind While citizenshi p in the West evolved over a protracted, contentious, but negotiated period, discussion of citizenship in Africa only became salient during the lead up to, and throughout, the actual period in the struggle for autonomy. Except for Angola, Guinea Bissau, Mo zambique, South Africa, Namibia and Zimbabwe, Africans emerged from the independence struggle nominally as full fledged citizens without the moil or trauma of standing up to those in power who were as bent on self preservation as they were in Europe.


57 Unde country never took root. Rather, ethnicity and regionalism were the closest thing to a commonly shared marker. These attitudes would come to embody popular perceptions and a ttitudes towards the public sphere as well as to the role of government. The masses had been given a promissory note, a note for expanded government services and relief from much despised state extractions that they intended to cash in once Africans finall y took over. But instead of a friendly, capable state, the masses got an aloof, disinterested entity that elevated its own interests above all else. Having been educated to view the state as amoral and undeserving of sacrifice, citizens actively took to e vading taxes, plundering from the state the very resources needed to make possible the education and healthcare people clamored for. The fragile post colonial state could not handle the demands made on it. Efforts at rapid modernization undercut the nascen agrarian policies under colonialism. Rather than investing in the agricultural sector, many countries used the earnings from the foreign exchange to invest in ill conceived projects designed for show and that were meant to catapult their economie s into a new era of industrialization. In this context of multiple claims and meager resources it was not surprising that people drew even more heavily on affective norms of obligation. The injustice of skewed distribution made politics even more contenti ous. To deal with public demands, the state was forced either to co opt or crack down on various factions. 2.3 Affective Reciprocity By the mid 1970s, thirty eight countries were independent. South of the Sahara desert, the dist inct features of what we now associate with post colonial


58 Africa were beginning to take shape. 20 Politicians who had found it politically expedient to delegitimize the state came back to haunt post colonial efforts at state building and development. The pri 1975:91). A pattern of thinking and being that on one level emphasizes individua l achievement and self fulfillment and on the other level emphasizes group advancement and stature was becoming institutionalized. The economy of affection and its attendant code of conduct embedded in a complex moral ecology where each person, from the yo ungest to the oldest, has a role, though not codified, lends the institution a degree of fluidity that makes it resilient and malleable to new uses and challenges. Like any ideology or worldview the economy of affection has its own vested interests. Rawlsi an 21 in its propositions, the logic and moral underpinnings of the nothing could better justify their claims. With this in mind, I shall now turn my attention to discuss ing how access to vast public resources fed unsustainable patterns of reciprocity which, over time ultimately overwhelmed the state and by so doing laid the foundation for the ensuing reversed collective action dilemma. 2.3 . 1 A Reversed Collective Action P roblem Whether states are effective and capable depends largely on the existence complementary supporting institutions, both formal and informal. Not only are effective states better positioned 20 Ethnic polarization: the one party rule justified as imperative for national unity and integration, and the subsequent institutionalization of authoritari an rule, cooptation of the civil service and rising levels of corruption, and civil wars whipped up by the exploitation of ethnic, cultural, social, religious and other cleavages in the various African societies. 21 lass position or social status; nor does he know his fortune in the


59 to collect taxes than their less effective counterparts, they are also more likely to do a better job of redistributing revenues collected by providing, both in terms of quantity and quality, public goods. These services in turn reduce the pressure on individuals and groups to subsidize governmental functions. Resour ced conserved in this manner can then me invested into businesses or improving individual quality of life. Where this is lacking some civil servants meet their affective obligations by funneling state resources meant for improving the welfare of many will, instead, go towards the fulfillment of narrow personal obligations. The private sector often has to resort to bribery and corruption to survive or thrive, living the very poor to deal with these effects. And even though civil servants are very much aware of the aggregate effects of their actions they nonetheless continue to meet these obligations for a number of reason. For one, it is often very difficult to witness a certain level of misery and not help. Besides, for those willing to depart from or limit affective obligations, a lack of credible alternatives makes it more difficult to sustain withdrawal from these interactions, thus ensuring participation. This quagmire is what I call a reversed collective action dilemma (RCAD). What this predicament looks like, what its effects are, and how it is manifested are the subject of the next section. 2.3.2 Institutional Persistence By dint of their occupation, African civil servants of the 1950s and 60s enjoyed a standard of living hitherto unparalleled At that time even low level officials enjoyed the security of a monthly salary; positions also often came with subsidized housing, sometimes a car plus a driver, and low interest loans. 22 22 Until recently most middle level to upper level officials received, in addi tion to a steady income, housing, telephone, travel and car allowances, and trips abroad for conferences or education. They also get pensions, again, something the general population, back then and now, do not have. In Ghana, frontline extension officers a re called


60 much is given, 23 Early efforts to assist relatives and co ethnics soon became settled expectations and established practice, thereby institutionalized the economy of affection. 24 For the masses of uneducated relatives and co ethnics, each civil servant was a mini state through whom services could be , Weberian formalistic impersonality, a civil servant would forever be a white official because his African counterpart was too embedded in place and time. Colonial administrations could act without hatred or 1978: 225). His African counterpart could not be more different. The Weberian persona, the person inhabiting the position for the vast majority of Africans, was the human being. The very idea of Weberian bureaucracy (Du Gay 2008: 338) 25 on the market bec 23 Luke 12:48 24 During the early days of mass education many African civil servants contributed toward educating relatives with ana for example, the area then known as the Northern territories, present day Northern, Upper East and West Regions had one secondary school located in Tamale. It was not uncommon for students from East or West Gonja to seek help from the few officers who were Gonjas. As the number of the educated increased, so too did the number of students, especially the most poor, who hoped to be adopted by civil servants from their ethnic group. This is not to imply assistance was extended only to co ethnics. Many of t he retired agricultural officers I interviewed recounted how they and others benefited from the good will of civil servants who were not co ethnics but nonetheless invested in education by offering students accommodation so they did not have to go back to them secure admission into one institute or another, paying for tuition and school supplies or accommodating them Ghana, November 2011; Colonel Iddisah. Personal Interview. Airport Resi dential, Accra, Ghana, December 2011). Civil servants were also expected to help relatives and co ethnics find jobs, assist in any and all ways to ease dealing with the national bureaucracy. 25 A. L. Adu, one of the most respected authorities on the subject integrity, impartiality, efficiency, loyalty to the government of the day and devotion to duty are worth preserving


61 African official who took over the reigns of the service after independence. The conflation of person and office w ould, over time, undermine this fundamental ethos of Weberian bureaucracy. Then, as now, the locus of affective duty was the family. Even with increasing urbanization, decreasing birth rates 26 and changing family patterns the nuclear family central to affec tive responsibility, 27 yet the elasticity of what constitutes family allows for the incorporation of fictive he reasonable comfor clearly defined formal hierarchical structures, the bedrock of central planning and centralized decision making, (ii) management by rules that allow decisions to be made and executed in a consistent manner by all relevant parties, (iii) organization by functional specialty based on skills or training and the type of work officials are hired to do. And last but not least, (v) purposely impersonal, entry into the ser vice is based on technical qualifications. Civil servants were educated and trained to live up to the requirements of their formal job descriptions. Each office or position came with clearly specified responsibilities, and subordinate rights believed to b e imperative for that particular office (Coser, 1977:230 233; Hall 1963; Condren, 2006; Condren 2006). 26 continue well into 2050, with Africa becoming predominantly urban by 2030. Of the 28 mega cities, cities with 10 million or more inhabitants worldwide, two (Nigeria and Ethiopia) are in Africa. Also, recent census and survey data suggest that African fertility is falling more slowly than expected and projections are that most countries will see their population at least triple in 2010. According to the International Labor Organization (ILO), Africa has the third highest unemployment rate. How much weight to put on this finding is an open question given the fact that much of the African economy is still informal. Although when we take into account the number of countries currently experiencing drought in the Horn and in East Africa, and the destabilizing effects of the Islamic insurgencies in the Sahel, Chad overflow . With little or no services, relatives a nd friends are still the helpers of last resort for both urban migrants and rural relatives, and this help is especially true for young girls looking to make something of themselves. Often, when these girls get pregnant, they are compelled by the circumsta nces of their lives to send their children back to the village where older relatives assume care for these children. More often than not, these rural relatives make ends meet through the charity of others. It is often quite common for migrants to go back h ome after having contracted chronic diseases, or having been maimed. With no pensions or disability insurance of any kind, affective reciprocal insurance becomes the only, albeit very inadequate, means to moving on. 27 Note: The following is excerpted from Social Trends Institut kinship institutions, for the most part, ha ve been resistant to change, modern lifestyles notwithstanding. Individuals think of themselves as belonging merely to one social group; insofar as they lead their lives in a variety of social contexts, they are somehow induced to elaborate a personal synthesis of the inputs they receive in each of those social contexts, a synthesis that enhances their self consciousness as individual beings, and not merely as members of a social group, which permeates all the dimensions of their lives. As a result, if they keep their ancient loyalties and ways of life, it will increasingly be more because of personal decisions than because of necessity. Yet, what I would like to highlight in this context is that, whether people freely choose to keep ancient ways of life, or choose to break with them, in a certain sense they are equally modern, for in both cases their reason to do so is no longer tradition but a per


62 kin. This is especially true in urban areas. Yet, fictional as these ties may be, they are just as strong, if even a little tenuous and more horizon tal and therefore potentially equal compared to the older ties of affinity. That said the reason why they existence getting things done has not changed, reflecting the inherent paradox of affective reciprocity. 28 In a region where states are unable to provide basic safety nets for the majority of their citizens, affective reciprocity fills the gaps. Getting enough people to move beyond affective reciprocity requires concerted and coordinated effort on all fronts. 29 Yet each individual has no guarantee th at others will do what is required to fundamentally change the underlying sanctions regime undergirding the system. Given the importance of affective welfare protections, the price for reform is high, even for the more well to do. Depending on the egregio usness of the here. To the extent that a process of modernization is under way in African societies, tensions between traditional and modern institutions are necessarily going to take place. And one of the main loci of these tensions is family and consider things from the perspective, say, of economic prosper ity and adaptation to the logic of modern organizations, to the extent the latter seems not compatible with the logic of patronage. It is a resource if we consider things from the perspective of the support every individual requires in times of hardship wh ich are very 28 This paradoxical quality of affective reciprocity, especially those formed in urban areas are most likely to change if and when credible alternatives exist. 29 Anecdotal evidence from Ghana, Nigeria, South Africa and even Botswana s hows individuals are increasingly responsible for their own protection from criminal rackets. While the rich and middle classes can afford the services of private sector actors, ordinary folk depend on informal arrangements. Ghana and Nigeria are known for their vigilante justice. In Ghana the use of land guards to protect residential plots of land yet to be developed from encroachment and outright theft is now becoming institutionalized albeit informally. Families generally use young unemployed local you th. Usually friends of young men form a group, who then move from one property to the next to ensure no one is encroaching on land owned by a member of th eir family. These young men are and then. Nigerians have been doing this even longer . During fieldwork in Northern Ghana the same pattern was evident. Increasingly subsistence farmers in the Northern and Brong Ahafo regions in Ghana depend on local youth to protect local women from attacks by Fulani herdsmen during the rainy season. Young men, usually friends or relatives from close by villages, form protection bands to protect the mostly female groups of women who do almost all the farm related work between planting and harvesting. All because the police are either ineffective or are just not present, as in the case of many of the rural farming communities visited. Additionally, many of the new sub divisions in Accra for example have no running water, to cut down on costs, neighbors often coordinate water delivery from private sector actor s to cut down on cost. Again, this phenomenon has been going on in neighboring Nigeria, hence how Ghanaian entrepreneurs learnt the business. Nigeria is the incubator for all kinds of innovations good and bad for many of Ghanaians businessmen and women , and artists alike.


63 and his or her close associates. Transgressors stand to lose future favors and aid such as contributions towards the care of children, the elde rly, and those with chronic health issues. 30 Particularly galling to many were the wealthy who failed to sufficiently commensurate grieving relatives and others who felt that, because of the relationship between the deceased and these individuals, the death at least warranted dispatching emissaries to sympathize. These norms of reciprocity are even more important for the poor and less fortunate. 31 Moral judgments, Amitai Etzioni reminds us, do not exist outside of individual and group interests for it is onl (Etzioni 1988: xi). The reverse, I suggest, is also the case. Viewed from this perspective, continued minimum participation is not only rational, it is also the best strategy. Since the incentive structure works similarly for everyone, the net effect is stasis, even when participation does produce a sub optimal equilibrium by undermining both the personal and national accumulation needed for development. For civil servants, participation means draining resources from developmental efforts and investments aimed at improving the lives of the many towards the fulfillment of narrowly targeted obligati ons, the fulfillment of which often imposes many costs on officers and society at large. Participation as the dominant strategy produces a sub optimal equilibrium because it undermines both personal and national accumulation and development. This is what I (RCAD). 30 Usually by having the children of poorer relatives stay with, and be educated by, the more well to do 31 Many of those interviewed recounted stories about how specific individuals were punished because they did not donate money, or d id not attend important events like marriages and deaths. Follow up interviews with some of those punished in these ways revealed the public nature (sooner or later friends and neighbors find out), which lent this form of holding others to account particul arly painful, hence effective.


64 While both farmers and officers lamented the never ending requests for assistance from both relatives and friends, they nonetheless expended considerable amount of time, energy, and money meeting them . highest for the poor, but not for the reasons advanced by Olson. Recent scholarship on collective action dilemmas suggests the ability of groups to cooperate is key to understanding why some communities are able to overcome social traps and collective action inertia. Bardhan and Dayton Johnson for example found social heterogeneity weakened social norms and sanctions that would otherwise have resulted in the enforcement of cooperative behavior and collective agreements schemes corroborated Dayton Johnson where once again diversity was negatively correlated with cooperation ((Bardhan et al 2002; Dayton associations in Indon esia, residents were less likely contribute voluntary labor, materials and money towards irrigation projects, security arrangements, rice cooperatives, and local health centers (ibid: 2004). As did Khwaja, who also found high levels of heterogeneity to be negatively correlated with the maintenance of public goods in rural Pakistan (Khwaja 2009). Analogously, Miguel and Gugerty in their study of the determinants for school funding in Kenya, found that parents in ethnically diverse communities were less likel y to act collectively to raise funds in support of local schools ( Miguel and Gugerty 2005 ). What then, about homogenous groups makes them more likely to overcome collective action problems? Central to the above studies is the notion that homogenous groups are more likely to trust each other ( Alesina and Barro 2001 ; Alesina and La Ferrara 2004 ; Costa and Kahn,


65 2003; Easterly 1997; Evans, et al. 1993 ; Knack 1997 , Fafchamps 2001, 2002 ; Glaeser et al. 2000; Easterly 1997; Knack 1997; Knack and Keefer 1997; Keefer, 2008; Knack and Zak 2002; Uslaner 2002, 2003; Zak and Knack 2001) and draw on shared cultural cues, or experiences, and share information ( Barr 2000 ) to come up with workable strate gies. Not only are homogenous have interests in common or are from the same ethnic group are also more likely to have repeated interactions and, over time, this l eads to cooperative behavior ( Axelrod 1984; Ostrom and Ahn 2003; Ostrom 2003a, b ). According to theoretic models on games, in games with two players and in symmetric games, players, over a period of time, employ c ooperative strategies. Depending on context, cooperation is governed by frequency of interactions ( Aoki 1984, Hardin 1991, 1995 ; Marwell, 1988 ), and according to Habyarimana, behavioral evolution is; therefore, influenced by both the physical environment and interactions between other individuals ( Habyarimana 2009 ). Nevertheless, the empirical evidence suggests we might be well served by reminding ourselves that ethnic homogeneity is not necessarily a panacea for cooperation. Botswana, L esotho and Swaziland have Tswana ethnic majorities; 79% of the population in Botswana identify as Tswana, 99.7% in Lesotho and 97% in Swaziland also identify as such. 32 But few would argue Lesotho and Swaziland done as good a job of improving the lives of t heir citizens as Botswana. Even fewer would hold up Somalia even more homogenous than the Tswana trio as a model of ethnic homogeneity. Not only are all Somalis from the same Afar clan, they are also 32 See Goode 2008 for a contrary argument


66 overwhelmingly Sunni Muslims (99.9%), yet acting togethe r for much needed peace continues to elude to Somalis. 33 Mérove Gijsberts, Tom van der Meer and Jaco Dagevos, for example, found no clear correlation between diversity and cohesion; as living in an ethnically diverse setting was not necessarily negatively correlated to social cohesion. Findings by Western European scholars have generally been inconclusive, unlike Alesina and La Ferrara (2000) or Costa and Kahn (2003) whose studies were conducted in ethnically diverse American neighborhoods. Gustavsson and J ohrdahl for example suggest the relationship is tenuous 34 The findings of Western European scholars are mixed. Glennerster, Miguel and Rothenberg for example found that ethnic diversity in post civil war Sierra Leone had no effect on the provision for publ ic good that carries the possibility for multiple outcomes and specifications (Glennerster et al 2013). Ethnicity, I argue, is just one of the many identities individuals draw on. People from the same ethnic group are more likely to know people from their region, towns or villages. It is common practice for parents to attend the various rites of passage with their children and younger relatives. 35 Even in rural areas families travel back and forth for these events as a way of teaching younger generations ab out their relatives, past and current, their history, and their 33 Clan effects, though more subtle, remain strong enough to call into question some of these findings. Also, another major frustration with the exaggerated use of ethnicity to explai n almost everything in Africa has resulted in a failure to even tell us which level of ethnicity is more important. Ethnicity driven studies tend to focus on aggregates, the ethnic or sub ethnic group, but usually not on the clan which is the level at whic h we can hold a host of factors constant to better understand the effects of identity. My objections to the use of ethnicity as a catchall category are simple. Ethnicity can, and should, be leverage to understand certain phenomena, certain types/levels of mobilization (like local government elections and certain aspects of conflict and violence). But the use of the ethnic variable, for the most part, is used as though people from the same ethnic group have no other interests and that ties to the ethnic grou p are unchanging. 34 Gustavsson and Johrdahl for example suggest the relationship is tenuous. Gustavsson, Magnus, and Henrik Jordahl. 2008. "Inequality and Trust in Sweden: Some Inequalities Are More Harmful Than Others." Journal of Public Economics 92: 348 65. 35 From birth (out dooring ceremonies when children are named) to engagements, weddings, funerals.


67 culture. These rights of passage are important and deeply ingrained in many young people to the extent that even those living abroad often go back to their countries of origin to get married, often as a way to honor their parents or those who raised them. But for each of these events there are often friends in attendance who are not from the same ethnic group as the family they came to support. In melting pots like Ghana and Nigeria, these gue day be getting married and you would want your friends to attend the wedding and signal to the broader community that you are indeed a person well worth the effort, means that you attend. The point being made here is that people from the same ethnic group, or religious affiliation are more likely to socialize, or run into each other even if they do not know each other initially. They are also likely to meet at future event s, making it more likely that people from the same group will cooperate because of the possibility that word of non cooperation might spread and the defector might be socially sanctioned. It is quite plausible that some of the effects noted in the above s tudies have less to do with trust stemming from ethnicity than trust developed from frequent interaction. What matters network starts with the extended family of both parents. Subsequent networks are layered over the initial network. People within a particular network are more likely to socialize together and thus get to know each other quite a lot. Cooperation, then, might not be driven by ethnicity in and of itself, and it is quite possible that reputation and/or the way one is perceived within these social groups is just as important. In our case it is likely that protection from sanctions or negative threats rather than trust or positive cooperation is more important to overcoming this reversed collective action dilemma.


68 Individuals will defect if they have credible protection from sanctions. Individuals are also more likely to defect if their initial efforts at defection are hidden from the view of those ab le to exact punishment. The opposite is also true. As the initial cohort of defectors go unpunished a demonstration effect sets in. When certain groups achieve protection, other groups will recalculate their opposition. Additionally, for an individual to risk sanctions, the individual would like to know the probability of success. The literature on Eastern Europe posits that students took their cues from dissenters (dissenters took their cues from outside activists), workers took their cues from students, and party supporters from workers (Karklins and Petersen 1993). It would be a stretch to claim that all political behavior undermining development efforts in Africa can be sufficiently understood in this framework. It makes sense; however, as an important , albeit partial, explanation as to why so many even highly educated urbanized individuals continue to be active central agents in the reproduction of social and political informal institutions that undermine state capacity. The importance of values and pr eferences along this theme should not be forgotten as they do shape the rationality inherent in this kind of system. "Whatever happens to the individual happens to the whole group, and whatever happens to the whole group happens to the individual. The indi vidual can only say: 'I am, because we are; and since we are, therefore I am.' This is a cardinal point in the understanding of the African view of man" (Mbiti 1969:109; quoted in Lassiter 1999:3). This emphasis on the group is not surprising given the per ils that can be wrought should individualism triumph. But is the individual so inextricably bound in the group? According to Ghanaian philosopher Gyekye, it is


69 community ex pect individuals to take personally enhancing and socially responsible decisions and actions (emphasis mine, Gyekye 1988:31 32: quoted in Lassiter, 1999:4). Of course this creates tension between the individual and society. On one hand the individual is ex pected to put the welfare of the group above self. How these expectations and tensions are played out on the local or national level has direct impact on develo lives, officials, even when they are not from an area, eventually develop intimate knowledge of the needs of their communities. Also as members of their communities, officials generally fee l, even when this is not the case, that they have an obligation to help those in need of assistance. Furthermore, as members enmeshed in their communities, they often know how to get services to those in need and are more likely, when so inclined, to deliv er these services quickly and effectively. This view, in contrast to views from the Western development model camp, tend to about the effectiveness of conventional those who sought to reform Af egalitarian aspects of affective reciprocity and their ability to generate social networks across geographical, ethnic and political borders as a mechanism for unity and cooperation, reciprocal acts of sharing, or as important for cooperation and accountability (Bratton and Rothchild, 1992; Chazan 1988; Kimambo et al 2008; Manor 2008; Schatzberg 2002; Vansina 1990). There is no


70 doubt that the economy of affection has a positive effect on a local ized micro level. It should be noted; however, that these studies and examples were based on small, localized studies of mostly farmers or villagers. So far it has been impossible to scale up what little positive benefits have accrued from affective recipr ocity derived solutions. For example, each time an official dips into resources budgeted for some other end to fulfill a need for which the resources were not targeted, crucial resources are diverted from official developmental goals. Because these transfe rs are unofficial, it is usually left to the discretion of the officer to decide how much of the service in demand can be skimmed off the official stock, newly implemented checks and balances notwithstanding. More often than not, officials end up giving le ss than enough to deal with the problem effectively. In the end not only are resources wasted because not enough is allocated to deal with the problem, the diversion reduces the available stock of resources and its potential effects. Since there are no obj ective means of ensuring that those most in need receive this kind of informal help, there is no way of telling how effective or efficient these informal interventions are. This of course ensures a collective action dilemma at the macro level. 2.4 Conclus ion The set of explanations advanced in this study take, as their starting point, the idea that administrative capacities. Having highlighted the evolution of affec tive reciprocity and the dynamics by which it weakens capacity we now turn our attention to efforts at reforming and strengthening the state.


71 CHAPTER 3 STATE CAPACITY: COMPARING STATE OUTCOMES IN BOTSWANA AND GHANA 3.1 Introduction Karl Polanyi conten ds states provide and guarantee some of the most important institutions that enable the exchange of goods and services and accumulation to develop with some degree of predictability and security. Markets, capital and labor operate in contradistinction to t he powerful interests of cooperation, landlords, unions, peasant groups and their advocates. Where these tensions are not regulated grave social disruptions can result. States then not only enact policies aimed at breaking down resistance to the market, bu t also to counter some of the more negative impacts of the market through regulation and social policy ( Polanyi 1957 ). States also intervene directly by crafting and implementing economic policy, especially when the local capitalist class is unable to perform these functions. The state also sets the rules of the game within which politics plays out Rae 1971, Taagepera and Shugart 1989, Cox 1997 . 1 States, however, differ in their abilities to govern. For example, states like South Korea and Taiwan have been able to mobilize, guide, and sustain the conditions necessary for economic growth, and by so doing, dramatically improved the quality of life for a majority of their citizens. In Africa the absence of a big enough indigenous entrepreneurial class with the requisite resources to play a role functionally equivalent to that of the capitalist bo urgeoisie in the West continues to be a drag on the private sector. Hence the post colonial state, like its colonial antecedent has had to assume a more prominent role compared to other developing regions of the world. 1 Chinese Riots in Late 20th . 39, No. 2, pp. 231 242.


72 3.2 The Quest For A More Capable St ate ionships, extract better at these essential tasks ( Mann 1984, Migdal 1988 1984/1988: 1 32; 1993: 59 60; Migdal 1988: 4; S cott 1998:2; Thompson 1968). 2 However, the ability of a government to formulate, implement, monitor and evaluate policy is shaped by a number of factors. Among them the capabilities of government officials and the resources at their disposal needed to car ry out their duties. Attempts at comprehensive structural reform of the colonial administrations started in earnest only after the Second World War. Indeed, it was the very necessity of hastening transitions to independence that laid bare the problems of c olonial administrations. The consensus then was to reorient civil services towards national aspirations. of modernization theory derived prescriptions, post colonial civil service reforms sought to social backwardness. 3 If modernization was the end, the means for its attainment was the 2 functions of taxation, conscription, and prevention of rebellion. Having begun to think in these terms, I began t o see legibility as a central problem in statecraft. The premodern state was, in many crucial respects, partially blind; it knew precious little about its subjects, their wealth, their landholdings and yields, their location, their very identity. It lacked interventions were ofte n crude and self 3 Heady, for example, viewed development administration as different from traditional administration in that traditional administration was too focused on routine operations and efficiency and tended to emphasize the individual and centralized decision making. Development administration was viewed as relationship oriented,


73 adoption of development administration. Erad icating structural and administrative obstacles to planning, implementation and management was viewed as a prerequisite for modernizing bureaucracies. To this end, advocates of development administration focused on the right kind of training for staff and making available the resources deemed necessary for discharging their duties. The notion that given enough time and resources post colonial administrations, made anemic by decades of colonialism on the cheap, would operate like their metropolitan counterpa rts was never in doubt. failings, were not only the patient to be cured; in most countries they were also the only doctor in ( Rodman 1968:9) . Nonetheless, proponents of development administration were confident that with time they could turn these administrations into well trained, well equipped agencies run by a cosmopolitan workforce that was able to efficiently implement operations. The idea that given enough time and resources post colonial administrations, made anemic by decades of colonialism on the cheap, would operate like their metropolitan counterparts was never in doubt. However, the pathologies of the post colonial bureaucracy (poor leadership, seniority based authority and ritualism) 4 continued to whittle away at the nascent, but professionalized bureaucracies inherited after colonialism. These entrenched administrative weaknesse s led to some important forensic studies by scholars like Lucien Pye emphasizing the importance of role relationships and group performance. Development administration was also more encouraging of innovation ( Heady 1966, Riggs 1964; Montgomery and Siffin 1966). 4 A preoccupation with mundane operational routines, rules and regulations to the point of thwartin g an for an excellent discussion. http://wps.pea http://www.institutions bureaucracies


74 who came up with a diagnosis of an incomplete psychological adjustment to modernity and social change. What was needed, advocates like Hahn Bern suggested, was time. The problem was consi derable but not insurmoun (Hahn Been 1968; Pye 1963, 1966 ). Skeptics like Schaffer; however, underscored the paradoxical position of development administrators and the problems this created. On one hand it was difficult to divorce the management of the bureaucracy from its colonial origins and structure. On the other hand the very change reformers sought warranted technocratic approaches rather than the cadre of politicized officials rapid Afr icanization had imposed on civil services. It also did not help that many of these politicians lacked the education and experience their positions warranted ( Schaffer 1969 ). The fact that the mostly expatriate, technocratic, reforming class had m arkedly different interests and incentives than the politicians and political activists only compounded the problem. And so policies were tweaked, expectations adjusted, and timelines for improvements redrawn. Yet the promised improvements by organization s like the Ford Foundation, based on 5 As the problems mounted and the fortunes of the development administration wane d, its critics grew louder. Its dismal performance, it was argued, was due in no small part to its inherent inflexibility which made it particularly ill suited for the fluid African context ( Dresang 1973, Fry 1989, Davis 1996 Esman 2006 ) Botswana, no longer a protectorate of Britain or under threat of South African annexation, began its own interventionis t developmental program. Starting with the construction of a brand new capital, Gaborone the state also set about corralling the citizens in towns that 5 The Ford Foundation, the most important supporter of Comparative Administration and Development (CAG) annel CAG efforts into the


75 were for the most part well planned with an eye on governing and service delivery. Ghana meanwhile was u nraveling. Careening from one program to another in its multi year oriented development strategies became bogged down due to political inference. Gone were the technocrats and the resources budgeted fo r development hitherto 20% of the total investment budget. Kwame Nkrumah diverted ever more resources towards Pan Africanism, especially as France and Belgium did all they could to undermine the nascent regimes in Guinea and Congo ( Aryeetey 2000 ). In 1966 Nkrumah's authoritarian rule and a sagging economy led to his ouster while on a trip to China in 1966. He would live out the r est of his years in exile in Guinea as the personal friend of Ahmad Toure, the first president of Guinea. Affective reciprocity at its best since Busia, a univers ity professor, voted in 1969 fared no better. Three years after Busia won elections his regime was also kicked out of power by the military. Ghana had now entered a new s, political instability and economic decay. In the meantime, the government of Botswana and beginning its own interventionist developmental strategies. T hese policies focused mainly on the livestock industry that was dominated by the cattle owning elites who had privatized land and water rights and who also happened to be the ruling party. By 1975, a now well entrenched elite introduced the Tribal Grazing Land Policy (TGLP) aimed at managing grazing and water resources. Though progress was slow, the progr am will have deep, lasting effects. Ministry officials in Ghana fared no better, at least in the short term. Against a backdrop of grave and persistent food shortages, the military regime of Colonel I. K. Acheampong, the National Redemption Council (NRC 19 72 1975),


76 ceased power. The regime introduced Operation Feed Yourself (OFY), an ambitious program aimed at making Ghana food self democracy of one man, one vote, [was] meaningless when there [wa (Girner et al 1980:15). Unlike in Botswana, Operation Feed Yourself was championed by a group of charismatic young men and women more aligned with the masses, the most no of whom was Colonel Bernasko. Bernasko, as the Commi ssioner for Agriculture, traveled the length and breadth of the country tirelessly promoting the program and overseeing projects. Bernasko and the in communal work and cleanup exercises to improve health and sanitation in Cape Coast, as well as to improve agricultural projects nation enlisted to dig and maintain canals and, according to several farmers I interviewed, the much of the recent increase in food production in Ghana. The views of these farmers have been corroborated by ministry officials, official reports, as well as the views of several retired, high ranking ministry officials. I nspired by the behaviorist turn in political science, administrators called for the incorporation of decision making models from the field of economics into the planning and management of the public sector. African administrators knew first hand the dysfunction wrought by the differing interests of politicians and bureaucrats. How well these new adjustments would have worked is a matter of conjecture because the gales of the global debt


77 crisis of the late 1970s and devastating droughts and famines in Africa, combined with stagnant economies and political unrest, set reforms awry. 6 Conditions were about to get even more difficult for the subsistence farmer as Africa entered the 1980s. Later to could imagine what was in store. Large swaths in middle Africa were going through what would turn out to be two decades of intermittent drought and famine . As conditions continued to det eriorate, those who could afford to left for greener pastures, first to other African countries and, as those dried up, to the West, the Middle East, or anywhere but Africa. Ghana was sent reeling wh en the Nigerian Aliens Expulsion Orde r of 1983 went into effect. Within a two week period of that year, an estimated 1.2 million Ghanaians (a tenth of the population) had been forced back to the country they fled, deepening a growing crisis. In 1969, Ghana enacted the Aliens Compliance Order Nigeria, and hundreds of thousands of immigrants, mostly Nigerian, were expelled from the c ountry, Nigeria was now being paid back. 7 6 From the 1960s through the mid 1980s the countries in the middle of Africa from Mauritania in West Africa all the way to Ethiopia in the Horn of Africa experienced sever e droughts. The famine and dislocation stemming from the drought was massive. People in drought prone areas live under a continuous threat of hunger and malnutrition. The 1980s droughts forced 10 million farmers to abandon their land and over 1 million peo ple died. The impact of the The cost in lives would have been far greater but for the mobilization of food aid by the international community. Wo rldwide, the population affected by floods rose from 5.2 million in the 1960s to 15.4 million in the 1970s. The developing world is particularly affected; in India alone deaths from floods were 14 times greater in the 1980s than in the 1950s, and the area affected grew from 25 to 40 million hectares. 9_disasters.pdf drought.htm#intro See pages 34/35 for graphs continent simultaneously ; death estimates were 100,000 in the sparsely populated Sahelian countries of Mauritania, 1985 famine in Africa is scant, and disaster appears to be continuing as drought, locusts, and civil war plague certain http://www 7 The human tragedies of those days are still raw as I learnt during fieldwork in Northern Ghana where many Yoruba and Hausa traders and settled. Stories about parents separated for ever from chi ldren left in the care of local families to be sent for later when things calmed down, Ghanaian women who left with their husbands never to be


78 To add to the tumult, Chadians, fleeing from Chadian Libyan (1983 1987) conflict found their way into Ghana. By 1984, Ghana had a sizeable Chadian refugee population in addition to its own returnee and famine pr oblems. Many of the Chadian refugees; however, were absorbed into local Muslim communities just like the Ghanaians expelled from Nigeria. 8 Economies of affection, already spread thin, were strained once again as various community, religious, ethnic based a ssociations, and individuals pooled resources via social capital for refugees. For a handful of countries, the struggle by soldiers and civilians over the control of the state, would devolve into civil war (see Appendix 3). The onset of the Liberian civil war in 1989 brought more Ghanaians back home along with another group, Liberian refugees. While Botswana and Ghana were blazing new paths, public administration specialists were still struggling to find ways of speeding up development processes. By the mid 1980s administrators, inspired by the behaviorist turn in political science, called for the incorporation of human decision making models from economics into the planning and management of the public sector. African administrators knew first hand the dysf unction wrought by the differing interests of politicians and bureaucrats. How well these new adjustments would have worked is a matter of conjecture because the gales of the global debt crisis of the late 1970s, devastating droughts and famines in Africa combined with stagnant economies and political unrest set reforms awry. 9 heard from again as well as Ghanaians who appropriated alien businesses and property were told and retold from di strict to district. But I found very little official records to corroborate these stories. 8 Interview Sheikh Osman May 10 11 th , Newtown Mosque focus group, Accra Newtown. Sheikh Yakubu Musa and elders, Tamale Central Mosque. May 20 23, 2011. Tam ale, Northern Ghana. 9 From the 1960s through the mid 1980s the countries in the middle of Africa from Mauritania in West Africa all the way to Ethiopia in the Horn of Africa experienced severe droughts. The famine and dislocation stemming from the drought was massive. People in drought prone areas live under a continuous threat of hunger and malnutrition. The 1980s droughts forced 10 million farmers to abandon their land and over 1 million people died. The impact of the drought in Ethiopia was worsened by The cost in lives would have been far greater but for the mobilization of food aid by the international community.


79 While much of Africa was in an active state of stagnation Botswana was undergoing a construction boom. The population had started to coalesce around Gaborone, Francistown, Serowe/Pal apye, Selebi Phikew and the Maun. In 1982 the government of Botswana introduced the Arable Lands Development Program (ALDEP) 10 in an attempt to promote arable agriculture and to boost income and employment for rural communities. 11 Farmers were given subsidiz ed farm equipment and access to marketing facilities through the Botswana Marketing Board. Low income farmers in the poorest communities were also provided cultivators, harrows, ploughs, fencing materials, and draught power (mostly donkeys). As part of the implementation process extension services were revamped. Yet, ALDEP did not achieve its intended goals of increasing cereal production and food self sufficiency. Critics charge the program was hampered by a lack of clear guidelines and political will to i alternatives ; indifference to planned social change on the part of recipient communities, bureaucratic bottlenecks in the processing of applications. Post project analysis also uncovered general abuses like wiv es to posing as heads of female households in order to take advantage of the lower down Worldwide, the population affected by floods rose from 5.2 million in the 1960s to 15.4 million in the 1970s. The developing world is particularly affected; in India alone deaths from floods were 14 times greater in the 1980s than in the 1950s, and the area affected grew from 25 to 40 million hectares. 9_disasters.pdf drought.htm#intro See pages 34/35 for graphs http://www. continent simultaneously; death estimates were 100,000 in the sparsely populated Sahelia n countries of Mauritania, 1985 famine in Africa is scant, and disaster appears to be continuing as drought, locusts, and civil war plague certain cou 10 T he African Development Bank (ADB) and International Fund funded the program for African Development (IFAD). 11 Oluyele Akinkugbe, Joel Sentsho, (eds.), Eco nomic Development in Botswana: Facets, Policies, Problems and Prospects (Gaborone: Bay Publishing, 2005).


80 payments set aside for generally worse off female headed households (Makepe 2005; Picard handling of the program countered officers had little, if any, local knowledge which made them particularly ill suited for the project, a view still held by both the commercial farmers along the Kasani Pandamatenga corridor and a large segment of the subsi stence farming community. The Ghanaian Ministry of Agriculture fared no better. As in Botswana, poorly coordinated programs failed to meet national rural development goals. No among these projects were the World Bank supported Upper Region Agricultur al Development Project (URADEP), Volta Region Agricultural Development Project (VORADEP), and the Agricultural Services Rehabilitation Project (ASRP). Interviews with retired ministry official familiar with the nuts and bolts of the projects suggest a lack ed of an established and institutionalized extension system meant projects were implemented by different departments within the ministry with no overarching standards. The system worked well enough with donors participation, but once donors left, directors reallocated resources towards new projects, farmers were often paid late which in turn depressed supply in subsequent cropping seasons. 12 Which meant with the conclusion of the programs and the termination of external funding, the implementing department w as now left with the costs and logics of running a program they did not design. The failure of URADEP and VORADEP, excellent projects by all accounts suffered a similar fate in the hands of the nascent Department of Agricultural Extension Services. The 12 Journal of Agricultural Economics. 65:2. 383 405. See also Veronique Theriaulta, Renata Serra, and James A. Sterns on the effects of late payments and poorly functioning credit schemes


81 de partment at that point in time lacked the capacity to carry on by itself. The fate of the projects is particularly telling. The projects sought, among other things, to strengthen rural extension gricultural production and farm income. The programs were pioneering the Training and Visit (T&V) System through which officers were required to visit and organize regular workshops to demonstrate various techniques. Extension officers affiliated with the project were paid comparably higher salaries, Logistical supports provided under both URADEP and VORADEP were generally perceived as satisfactory. In interviews, two of the officers affiliated with the programs reported they and their been trained to do [and] even though the hours were long and hot, [they were successful be cause they] operated like farmers by doing most of the back breaking work early in the morning. [Usually] farmers would often pass them on their way to their farms and they would still be in the fields working when farmers were heading back to their villag with them they were able to teach by example. As much as they enjoyed farming, these officers also had to meet their goals because their project managers were constantly monitoring them. 13 For the rest of the country though, ye ars of cooptation and repression had resulted in a severely emasculated, but bloated civil service populated by an army of ill qualified, apathetic staff. Civil services all over the region were unable to perform the most basic of their administrative and managerial functions. Officials could hardly be expected to administer 13 Perso nal interview S.R.K Ashiaman, Retired Ministry of Agriculture official, 31 st October 2011. Accra.


82 development processes from their ill equipped offices. Many government offices did not have regular electricity or running water and with the infrastructural power of the state greatly weakened, repression and patronage became more important in ensuring a quiescent populace. With economies in free fall and livelihoods increasingly threatened, citizens took to the streets to voice to their displeasure ( Baldwin 1969, Helpman 1985 1969, Helpman 1985, Lipton 1991, Fei 1992, Mittelman 1997, Jalilian and Weiss 2000; Heidhues and Obare 2011 ). 3.3 Rol ling Back the State Faced with the prospect of social upheaval, African leaders approached the World Bank and the International Monetary Fund who took the opportunity to reshape African economies. Leaders could accept Structural Adjustment Programs designe d to, stabilize and liberalize economies, and the political arena or face the wrath of their citizens. Since their colonial days, the policy goal of many African countries was food self sufficiency. Under neoliberalism Africa would seek food security not s elf sufficiency, use market derived incentives, rather than price controls, to increase farmer productivity. Farm inputs and spare parts where now widely available due to private sector participation and reforms in import licensing. In Ghana, cocoa market ing was partially liberalized and investments in, and promotion of, improved varieties of staple crops were increased and widely publicized. In collaboration with the 31 st popularized sma ll scale food processing using very rudimentary technologies. The campaign, building off Operation Feed Yourself, got farmers to adopt improved varieties of staple crops also spawned a secondary cluster of technologies. For example large scale production of gari, a West African staple made from cassava tubers became a viable economic activity for many. Gari production use to be tedious undertaking: first cassava tubers are peeled, washed and grated by hand using large manual


83 aluminum or metal graters. This method of processing also posed grave health hazards. The large scale production of gari also created the need for commercial processing mills and other equipment, all produced locally. The quality of gari produced this way was a vast improvement on how i t was produced before. The new gari, though a little more expense, is fine grained, crisp consumed now by both rich and poor. Shea butter processing was also acti vely promoted. 14 Over post harvest waste for horticultural produce like tomatoes. Additionally feeder roads were resurfaced to facilitate the movement of food fr om the hinterland to the main urban centers. Concurrently, and also having adopted food security, Botswana set about modernizing and privatizing its agricultural sector. Lease and freehold tenure arrangements replaced customary land (use/ownership). 15 In 1 983 Botswana established an enclave of large scale commercial farms near Pandamatenga in the Chobe District. The area has some of the most fertile soils in the country, referred to as black cotton, and a relatively high rainfall, averaging 629 mm annually, the highest in the country. 16 The Chobe Land Board demarcated 25,000 hectares for the project, 50 commercial farms, each measuring 500 hectares with renewable leases after 15 years for an annual lease of P1.00 per hectare. The Land Board demarcated and all ocated the first 10 leasehold farms either individuals or syndicates. With leases as collateral for loans farmers sought credit. could then use the land they were leasing. The program was 14 34,3. 421 440 and Shea Butter Republic 15 Keith Hart, The Political Economy of West African Agriculture (Cambridge: Cambridge University Press, 1982), pp. 91 92. See also, I. Livingstone and H.W. Ord, Agricu ltural Economics for Tropical Africa (London: Heinemann, 1981), pp. 190 191. 16 Government of Botswana, An Impact Study of Pandamatenga (Gaborone: Ministry of Local Government and Lands, Applied Research, 1985). p. 27.


84 geared large scale production of sorghum, a staple of Botswana, inte rmixed with sunflower, pulses mostly cow peas and black eyed beans, and corn . But the program was plagued by problems from the start. First, farmers bought the wrong type of machinery for land clearing t asks like destumping only to realize they needed to double declined. Additionally, delays in loan processing and land preparation meant the raining season ca ught farmers unprepared as they were still clearing their fields. The boggy nature of the soil absence of a local repair centre, farmers abandoned the equipment and resorted to manual labour. Farmers alleged that report blamed the Ministry of Agriculture for poor planning, inadequate research on the local environment and failure to guide the pioneer farmers appropriately 17 As part of restructuring their economies, both Botswana and Ghana were now required to devolve and decentralize some a dministrative, fiscal, and planning and implementation functions to regional, district and local governments even though they lacked the administrative capacity for these functions. Yet the efficiency and effectiveness promised under decentralization never national study of 68 countries, the posited effects of decentralization is contingent on existing structural variables like GDP per capita, the degree of democrati cness at the time of implementation, as well as the type and nature of the functions being devolved. Countries willing to politically devolve power and open up the political space were found more likely to succeed 17 Government of Botswana, Pandamatenga Commercial Farms Rehabilitation Study, Final Report 1998 (Gaborone: Ministry of Agriculture, 1998), pp. 13 14.


85 : 252 ). Meanwhile layoffs in the public sector strained economies of affection and self help groups and associations. In Ghana some of these laid off civil servants took advantage of government policies to grow high value crops like cocoa, cotton, oil palm, cashew, pineapple, and vege s. 18 I f development administration was the handmaiden of modernization theory, the wisdom was that, viewed as rational self interested actors, bureaucrats, at the right wage and gi ven the right conditions could be induced to improve their performance. Along with a less hierarchical, more democratic organizational structure, the flexible and responsive management found in the private sector would be needed if bureaucratic inertia was to be countered (Heimer 1992; Grindle and Thomas 1991 ; Grindle 1997; Hughes 2003). Government programs were now expected to build consensus into programs, a process tes much of the problems associated with attempts at revamping the cotton sector were, for the most part, due to divergent interests and philosophies. domestic actors, supportive of a capillary presence of the cot rent seeking behavior or of unrealistic expectations, out of touch with international market cotton sector, dialogue among stakeholders was difficult and polarized, forcing the MRSC to 18 Personal interviews John Tsra kasu, Volta Regional Minister, Ministry of Agriculture, November 2011


86 spend considerable time and resources to find a sui priva rather than sheer lack of commitment were responsible for lack of implementatio In the final analysis, only a few African countries followed through with most of the more of these prescriptions. But overall in the smaller, leaner states, the economic boom that was expected to follow did not materialize ( Grindle 1997 : 3). Critics of neo liberalism charge that even under neoliberalism the fundam ental role of the state never changed. These functions were merely refocused towards allocating and disbursing funds to sectors deemed important to the neoliberal project (Held 1987; Teitell 1992; Pierson 1996). The scale of private sector implementation i magined by neoliberals never materialized, even in the more advanced economies. Africa never had a large, or enough of an institutionalized private sector, to assume this role in the first place. With hindsight we know that economic performance is shaped b y other institutional factors like the type of constitutional design, political culture, and previous Chalmers Johnson, together with others, highlighted the centrality of the state in creating an enabling environment in then corporativist Japan. Agreeing with him, Peter Evans underscored the nexus between policy formulation, implementation and collaborating between a cadre of technocratic bureaucrats and private sector actors. Where the two groups succeed in marrying private sector interests with those of the public the results are positive developmental outcomes (Evans 1995; Kohli 2004; Johnson 1982; Evans 1995).


87 study of 35 low and middle income countries. The authors found (i) meritocratic recruitment, (ii) compensation packages comparable to that of the private sector and (iii) predic , rewarding career paths were positively correlated to increased bureaucratic perf ormance, low corruption levels, increased job satisfaction and dampened short term self interested behaviors injurious to organizations. Additionally these factors interacted to attract and retain talent and thus increase productivity (Rauch and Evans 1995 ). Ades and Di Tella also found that rewarding long term careers through tenure and promotion coupled with paying wages comparable to those in the private sector not only lowered corruption and other rent seeking behavior among bureaucrats but also increas ed economic growth (Ades and Di Tella 1987). But by 1997 even the World Bank was coming to terms with the notion that its market led development efforts were not working. Shrinking government had proven not to be a panacea. After five decades of public se ctor reforms, a professionalized, accoun , responsive, and effective civil service had not materialized. Yet, the idea that better governmental performance was best attained through formal institutions was and is still gospel in Africa. 3.3 Capacity t o Do What? Selecting the Indicators Douglas North contends rules and regulation, fairly and consistently enforced sustains cooperation, reduces uncertainty and protection from the criminal elements rackets. Thus taken together create an enabling and stim ulating climate for economic accumulation, reinvestments and development (Acemoglu, Johnson, and Robinson 2001; Fearon and Laitin 2003; Fukuyama 2004; Hall and Jones 1999; Glaeser and Shleifer 2001; North and Weingast 198; Tarrow 1998; Tilly 1978; Slater, Nair and Smith 2014: Skocpol 1979). North also emphasized the two dimensional nature of institutions formal and informal rules and the fact they work together to form incentive structures (transaction costs) of society.


88 Yet public sector reforms, no m atter how well designed are implemented by people, thus the administrative context and organizational structures within officials operate and the context in which these functions take place matter. If the goal of capacity enhancing reforms is to increase o fficer productivity, then focusing on for the most part on the formal institutions of the state alone leaves a lot to be desired. 3.3.1. Capacity to Administrate Heredia and Schneider for example suggest administrative capacity is the byproduct of institut ionally shaped elite preferences for specific economic policies ( Heredia and Schneider 2003). Viewed from this perceptive, Doner, Ritchie and Slater argue the impressive gains by South East Asian countries was rooted in self preservation. Elites, the autho rs posit, had to deliver economic growth rates large enough to hold together the broad political coalitions they had stitched together to stay in power. Working with a reliable, competent technocratic bureaucracy was a means to an end (Doner et al. 2005). Taking a slightly different tack Skowronek for example contends expanded national administrative capacities are the byproduct of political struggles defined and mediated by the institutional arrangements rooted in a pre established state era. Skowronek co mes closest to both with the embedded autonomy argument stresses the importance of bureaucratic autonomy to enhanced administrative capacity (Carpenter (2001). Wh at is still unclear however is how this need for self preservation affects extension officers who neither have clout nor power. Judith Tendler suggest a way forward. Tendler links increased bureaucratic capacity to strong middle management structures (Tend ler 1997). Administrative capacity from the above is at once a core function of the state, clearly a


89 functions a la Margret Levi and Charles Tilly come to mind. Or as power, either over or through creating and sustaining the conditions that make the exercise of either dimensions of administrative capacity possible. Thus the particul ar policy mix elites choose becomes a function of their goal. Administrative capacity also looks different depending on the level at which it is viewed. Explored at the aggregate level is becomes a mechanism or means through which power is gained, sustaine d and welded as agued by Heredia and Schneider and Doner et al above. Judith Tendler suggests it comes about from the push and pull of mid level management. The common thread in the literature binding these seemingly disparate approaches to the concept i intrinsic 19 This dimension of state capacity is best captured by four Worldwide Governance Indicators . 20 The first Government Effectiveness 19 international environments, which must be present for political development to occur (Eisenstadt 1963), and as the new, articulated functions taken on by the political system over t ime to respond to a new range of problems (Almond 1965). In these conceptions, capacity as a trait, or as differentiated functions, changes or increases in response to societal interests expressed as demands, with the political system arbitrating neutrally in the process of 20 WGI Aggregation Methodology: Each of the six indicators is constructed by averaging out the underlying variables used in the construction of the indices in a three step process are cons tructed by averaging together data from the underlying. STEP 1: Individual questions from the underlying data sources are assigned to each of the six aggregate indicators. STEP 2: The questions from the individual data sources are first rescaled to ran ge from 0 to 1, with higher values corresponding to better outcomes. Where individual data source provide more than one question relating to a particular dimension of governance the rescaled scores are averaged. STEP 3: The unobserved components model is then used make the 0 1 rescaled data comparable across sources, as well as constructing weighted averages of the data from each source for each country. The unobserved components model assumes that the observed data from each source are a linear function o f the unobserved level of governance, plus an error term. This linear function is different for different data sources, and so corrects for the remaining non comparability of units of the rescaled data noted above. The resulting estimates of governance are a weighted average of the data from each source, with weights reflecting the pattern of correlation among data sources. The composite measures of governance generated by the UCM are in units of a standard normal distribution, with mean zero, standard dev iation of one, and running from approximately 2.5 to 2.5, with higher values corresponding to better governance. We also report the data in percentile rank term, ranging from 0 (lowest rank) to 100 (highest rank). The indicators are then reported in two ways: (1) in their standard normal units, ranging from approximately 2.5 to 2.5, and (2) in percentile rank terms from 0 to 100, with higher values corresponding to better outcomes. All


90 measures perceptions about the quality of public services, the civil service, its independence from political pressures, quality of policy formulation and implementation, and the cre dibility of the government's commitment to such policies. The second, Regulatory Quality captures perceptions of the ability of the government to formulate and implement sound policies and regulations that permit and promote private sector development. The third, Rule of Law measures the extent to which people have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violenc e. The fourth and last, Control of Corruption captures perceptions of the extent to which public power is exercised for private gain, including both petty and grand forms of corruption, as well as "capture" of the state by elites and private interests. Ho w Africa and Botswana and Ghana fare on these matrices is the subject of the next section. The results are reported in percentages from zero (worst) to 100 the best a country could possibly score on any one indicator. As before the s are at the end of the chapter. See Appendix C for more s. 3.3.2 Comparing Capabilities _ Regional and Across Countries Regionally sub Saharan Africa performs poorly on all four indicators, worst of all on Government Effectiveness , just four points above the low incom e regional mean of 23 as reported in 3.1. The trend is similar for the other three indicators: 30% for Regulatory Quality , which is seven points above the low income country mean for that indicator, 29% on the Rule of Law Index , 10 points above the l owest income regional mean and 30% on the Control of Corruption country scores are accompanied by standard errors. These standard er rors reflect the number of sources available for a country and the extent to which these sources agree with each other (with more sources and more agreement leading to smaller standard errors). These standard errors reflect the reality that governance is difficult to measure using any kind of data. In most measures of governance or the investment climate they are however left implicit or ignored altogether.


91 Index , 11 points above the low income mean. So far the evidence suggests Africa is indeed lagging on these measures. But having decried comparing Africa to other regions, I now take look at ho w the different African countries perform on the four indicators. Between 1996 and 2013 Botswana consistently outperformed Ghana on all four indicators, although the gap between their Rule of Law scores started narrowing in 2006, two years before Ghana bec ame a consolidated democracy. At which point we should expect voters prospective and retrospective accountability tests out (Lipset 1959:69 105; Przeworski and Limon gi 1997:155 83). The evidence however calls into question this version of democratic accountability as none of the other three indicators improve for Ghana, see 3.2. For example Ghana only made, at best modest improvements in Regulatory Quality post 2006. After holding steady at 52% from 2005 through 2007, it increases very modestly to 53% and then by a percentage point every year since until. Similarly efforts at curbing corruption seem to have been even more challenging. Circa 2006 Ghanaian officia ls seemed to have had their best year was 2000 at 58%, the previous gains disappear and officers seemed to struggle to gain ground; peaking again 2007 and then in 2009 and 2010. The Botswana story is even more telling. Starting from 2004 Botswana has strug gled to curb corruption for a number of reasons that may have had little to do with administrative also meant the replacement of locals by foreign workers, the major ity of whom were from other African countries because they were cheaper than expatriate stuff of European ancestry. It is quite plausible these officials brought with them behaviors usually not associated with Tswana civil service. It is also plausible tha t locals became financially overburdened as they assumed


92 financial responsibilities towards relatives affective by the pandemic, or any number of reasons. s 2003 seems to have been a waterma rk for Botswana. For a example young person living in African today, according to 3.2, will live in the region with the lowest GDP ($1.521 trillion), 533 billion dollars less than developing Europe and Central Asia the next lowest performing region after Africa. Holding constant the effects of inflation, costs of living, and differences in living standards between and within countries, a GNI per capita of 1,624 sets this person back by $710 than if this person lived in any of the countries in develop ing Europe & Central Asia the area with the next lowest GNI per capita. Finally as can be seen from 3. 4 our young person, just by being born and raised in Africa, has 73% chance of acquiring the most basic of literacy skills needed to communicate if male, 66% if female compared the next worse performer, Latin America & Caribbean, where 93% of male and 96% female citizens possess these basic literacy skills. Yet our young person is one of the lucky ones since about 56 children for every 1000 born in A frica die before their fifth birthday. Infant mortality rates are still high, 98 for 1000 live births and 86 for female infants. 3.4 Conclusion Taken together, chapters two and three provide the scope and scale of the research problem, the historical and sociological context within affective reciprocity, and the civil service reforms that have been pursued. The chapters also center the main hypothesis which states that affective obligations, on average, redirect resources away from national development go als towards narrow particularistic ends and, by so doing, undermine administrative and, by extension, state capacity. I assert as well that officials, who know better but who; nonetheless, participate in these reciprocal arrangements, do so due to an under appreciated collective action


93 dilemma. How the independent variable, affective reciprocity, is conceptualized, operationalized and measured along with other pertinent methodological issues is the subject of the next chapter. 3 1 Regional Comparisons of Select Outputs of State Capacity Region Government Effectiveness (WGI 2013) Regulatory Quality (WGI 2013) Rule of Law (WGI 2013) Control of Corruption (WGI 2013) North America 89 90 90 89 Europe and Central Asia 68 69 66 63 Latin Americ a & Caribbean 58 56 51 34 Upper middle income 52 49 49 57 East Asia and Pacific 49 47 56 53 Middle East and North Africa 44 44 44 45 South Asia 34 26 32 32 Lower middle income 33 32 34 35 Sub Saharan Africa 27 30 29 30 Low Income 23 23 19 19 Source : The Worldwide Governance Indicators (WGI) Project 3 2 Regional Comparisons of Select Outputs of State Capacity Year Rule of Law Government Effectiveness Control of Corruption Regulatory Quality Botswana Ghana Botswana Ghana Botswana Ghana Bot swana Ghana 1996 63 41 68 53 75 49 75 36 1998 68 36 71 50 78 54 74 38 2000 65 55 71 58 75 58 72 49 2002 65 52 73 52 75 47 75 35 2003 68 55 75 52 86 49 75 45 2004 73 55 73 52 80 50 72 41 2005 66 57 72 51 83 57 70 52 2006 62 59 70 59 78 59 66 52 200 7 61 60 73 55 79 59 65 52 2008 62 59 70 56 80 57 65 53 2009 59 62 68 54 79 60 66 54 2010 60 63 67 54 80 60 67 54 2011 60 62 67 54 80 59 69 55 2012 64 61 67 52 79 56 74 56 2013 63 61 62 51 79 56 73 56 Source: The Worldwide Governance Indicators (WGI) Project 3 3 Select Outputs of Robust State Capacity Gross National Income 2013 (*in $ Billions) GNI Per Capita 2013 (*in $ Billions) Purchasing Power Parity 2013 (GNI $Billions) Per Capita Gross Domestic Product Growth (2013) % Per Capita % Europe/Central Asia _Developing 1,929 7,086 3,711 13,633 4 3 Latin America & Caribbean 5,477 9,314 8,143 13,848 2 1 East Asia & Pacific 11,103 5,536 21,511 10,724 7 6


94 M iddle East & North Africa 1 2 South Asia 2,462 1,474 8,363 5,005 5 4 Lower middle income 5,296 2,068 15,289 5,970 5 4 Sub Saharan Africa 1,521 1,624 3,001 3,206 4 2 Low income 563 664 1,510 1,780 6 4 World 75,261 10,564 101,250 14,211 2 1 S ource: The Worldwide Governance Indicators (WGI) Project. * Atlas Method 3 4 Select Outputs of Robust State Capacity Primary Completion Rate Life Expectancy At Birth (2012) Infant Mortality Rate (2013) Infant Mortality Rate % By Age Group (Male 2012) % By Age Group (Female 2012) Per 1,000 Live births Per 1,000 Live births (Male 2012) Per 1,000 Live births (Female 2013) Europe & Central Asia 99 98 72 20 26 20 Latin Am & Caribbean 93 96 74 16 20 16 East Asia & Pacific 105 105 74 16 21 18 Middle East & North Afr. 98 92 71 21 27 24 South Asia 92 90 67 45 56 57 Lower middle income 93 91 66 44 61 57 Sub Saharan Africa 73 66 56 61 98 86 Low income 73 69 62 53 81 71 World 93 91 71 34 47 44 Source: The Worldwide Governance Indicators (WGI ) Project


95 CHAPTER 4 NOTES ON METHODOLOGY The thrust of methodological issues are dealt with consecutively in the empirical chapters primarily to enhance readability. That said, a short discussion about the choice to use three estimation strategies Ordinary Least Squares (OLS), Negative Binomial model (NBRM) and a Logistic Regression model is warranted. My data analysis strategy is also premised on the idea that the three behaviors socializing with officers when they should be working, conducting official ministry business outside of official channels and dabbling in influence peddling theorized as inimical to enhanced capacity gradually drains the state of its abilities. Hence the use an additive scale in the construction of the dependent varia bles. Additionally, because the dependent variables are categorical the question of intrinsic ordering becomes irrelevant. This makes the data amenable to the regression techniques utilized. The dependent variables are a checklist of the behaviors we are looking out for, which, once found, are to be checked off and tallied. While a simple count model like a Poisson Regression Model (PRM) would ordinarily be used, the technique is unsui in this case due to oversampling in the Ghanaian case where farmer s in the Northern Region were over sampled. A d ditionally Poisson regressions impose strict conditio ns on the data, making them unsui for the analysis at hand hence the use of the Negative Binomial model (NBRM) which is less rigid and thus more appropriate for the over dispersed data on the dependent variable. The three regression estimation strateg ies employed in making sense of the data compensate for the weaknesses of each model if used alone. Each instance of engaging in affective reciprocal behavior is counted as a unique event and even though standard Ordinary Least Squares (OLS) are sometimes used in analyzing count data, doing so often results in analyzing the data incorrectly with OLS regression resulting in inefficient, inconsistent, and


96 biased estimates. The Negative Binomial Regression Model thus provides an alternative means of exploring the data further. Instead a Negative Binomial Regression Model was used because the conditional variance required for the utilization of Poisson regression presupposes a conditional variance no greater than the conditional mean. First we need to run a Poisson distribution test to determine if the mean and variance of the dependent variable Capacity Sapping are equa l. 1 A quick check of the data shows the average officers has 11 opportunities to engage in capacity sapping behaviors, truncating the data at 11. The distribution of the data is as expected since we expect clustering of data points around the 10 20 range for count models Two core assumption we need to adopt for this estimation strategy would be that the distance between each point on these scales are equal and that the causal relationships between independent and dependent variables are linear, symmetric, and probabilistic. Those assumptions are probably too demanding but social science analyses regularly violate them, and as long as we are aware of the assumptions of the model and we are cautious in interpretation of the results, it remains a useful empiri cal strategy that can give us significant leverage on the hypotheses. An alternative conceptualization of how the dependent variables are operationalized would be to think of them as non linear count of actions taken by farmers that undermine state adminis trative capacity; that is each action is countered/ checked off as a wound/cut, though not mortal. Then again, we do not need to know at this point and at this crude level of analysis, what the fatal blow is. Thought of this way, the dependent variables su ggest the use of some type of 1 Poisson distribution(s) allows us to test for the discrete probability that given number of actions or events will occurring in a fixed or given period. That these events will also occur with a known average rate and independently of the time since the last event or occurrence.


97 the generalized linear model (McCullagh and Nelder 1989). Count regression models based on maximum likelihood procedures implementing conversions make the non linear count dependent variable linear (Elhai, et al. 2008, 131). T he simplest count model, the Poisson regression model (PRM), assumes a Poisson distribution (substantial positive skew and variance equal the mean). The problem though is PRM assumes variance cannot be greater than the mean, even though this is almost alw ays not the case. This has caused over dispersion (Elhai, et al. 2008, Long 1997, 219). So to get around the issue of over dispersion a negative binominal (NBRM) model is introduced. This also takes care of any non normal, heteroskedastic distributions in the dispersion parameter to capture unobserved heterogeneity in the model and/or temporal dependency (Cameron and Trivedi 1998; Long 1997, 236). The NBRM corrects some of the under prediction of zeros in the PRM by increasing the conditional variance but not by affecting the conditional means. It seems possible that each of these different models is reasonable for the analysis and the interpretation of relationships to the independent variables. In order to avoid making a choice between them a priori that possibly affects the results, the following analysis will be based on running both estimation strategies in parallel. If the two options present consistent results we can also be more certain that the findings are robust. Finally, the Got Favor indicator, the last of the dependent variables used in measuring the diversion of resources from the state, is the most restrictive of the three because it focuses only on farmers who report actually asking for, and getting help from extension officer(s) for a favor officer(s) to help them anyway. Because Got Favors is a dummy variable (yes/no) response category, logistic regression is more appropriate for the current a nalysis because the dependent


98 variable is dichotomous with two levels and the logit regression model does not assume that distances between categories of the dependent variable are equal (Long and Freese 2003, 151). The theoretical underpinnings of a logi t regression are that the odds ratio becomes the same regardless of the cut off selected for categorization (Agresti and Finlay 268 271). The efficiency of this technique lies in its ability to use information in the dichotomous variable in a regression. in addition to running OLS and NBRM estimations, we are also running standard logit reg ressions and interpret the results alongside each other. With this in mind we can now turn to the analysis of the data. The section sought to summarize the data collection and the methods used in the analyzing the data. We now turn our attention in Chapter 5 , the first of the three empirical chapters.


99 CHAPTER 5 A TALE OF TWO CITIES 5.1 Introduction Leaving Sir Seretse Khama Airport for the city, one is immediately struck by the tidiness and quiet of Gaborone, the capital of Botswana . main international airport is located 15km north in the outskirts of the capital. Gaborone, like much of Botswana, is clean. Its streets are swept daily by an army of well equipped uniformed cleaners. Trash is picked up on a regular basis and clean drinkin g water flows most of the time. Blackouts, while not unheard of, are not frequent. Littering is minimal, and starting around 2010, street hawking was banned. Even during rush hour Gaborone looks and feels empty. And even though the workhorse of the countr recognizable. Combis are required by law to be registered, road worthy, and to use designated stops and stations. Combi drivers a lmost always drop off and pick up passengers at designated stops, rarely cram their passengers in like sardines, and rarely break down in the middle of the street causing hours of traffic jams. International airport is some of the most upscale neighborhoods. Street hawkers and pedestrians weave through heavy traffic on each of the streets leading off the main airport access road. Youth, who should be in from an assortme windows calling out orders to hawkers or roadside sellers. Beggars on skateboards, in wheel

PAGE 100

100 chairs, or on foot compete with the mass of bodies and cars for use of the street. The tea ming assortment of people and vehicles slows down traffic even more as man and car do their best to get the better of traffic lights and conductors. These attempts to outwit and outsmart one another invariably brings traffic to long unavoidable fume filled jams that add to the frustration. spite of the best efforts of its mayors and teams of cleaners. Citizens litter and violate basic civic rules with impunity. But even in lamented culture of indiscipline, Ghanaian trotro drivers hold a special place. Renowned for their unruly behavior, it is not unusual to see them driving on the wrong side of the road towards oncoming traffic if traffic going in their di rection is not moving fast enough for them. Blaring horns, hurling insults and intimidating passengers and fellow road users, especially female drivers, and even the traffic police is par for the course. To say the least, the careless trotro drivers are th e cause of many a fatal accident. often right in the middle of the street, and seem to have no regard for standard driving conventions. Why have the many attem major streets failed? And why is Gaborone much cleaner than Accra? We take a closer look at the two countries in Chapter 5 starting the cross tabulations of the descriptive data in section 5.2; this data sketches out the main characteristics of the sample by region, gender, age, and religious affiliation, as well as settlement type, level of education and whether or not farmers use ministry of Agriculture services. The data on farmers is presented first, followed by the data about extension officers. Also the data for Botswana will be presented first in this and the next two empirical chapters. We perform the necessary diagnostic tests in

PAGE 101

101 section 5.3 in readiness for the data analysis that is presented in the two empirical Chapters to follow. 5.2 Overview of the Data Agriculture is still the main source of livelihood for the majority of rural sub Saharan Africans. These smallholder subsistence farms, usually no more than 2 hectares, produce up to 90% of the total agricultural output for the region. 2 economy shrunk from about 40% at independence in 1966 to 2.4% in 2010. 3 These trends are captured in the data. Of the 1,801 farmers surveyed for the study 524 farmers (29%) are from Botswana ( 5.1). Despite our best efforts to survey as many farmers as possible, we were only able to survey 124 from the Kasane/Pandamatenga corridor in the Chobe Region, Northern Botswana, 251 farmers from the greater Gantsi area and the r est were drawn from Charles Hill, and its environs: Ncojane, Makunda, Chobokwane, Tsootsha, New Xanagas. Even accounting for the differences in population, subsistence farming in particular, and agriculture in general. 4 Fifty five percent of the surveys f rom the Ghanaian case are from the West Gonja and the Savelugu Nanton districts in the Northern region where a total of 987 farmers were surveyed. 5 The rest of the 290 farmers were drawn from Techiman/Kintampo area. Surveys in Brong Ahafo 2 There are some large farms and plantations, particularly for rubber, oil palm and coconut and to a lesser extent, rice, maize and pineapples. 3 Statistics Botswana, External Trade Statistics 2010:1 4 Even with investment into the Pandamatenga Agricultural Hub by the Government of Botswana in partne rship with the African Development Bank, agricultural productivity increased only marginally, Gross Domestic Product had declined to 2.6% by 2000 and is still sliding downward. Government of Botswana, Management of Quelea Birds (Gaborone: Ministry of Agric ulture, July 2005). 5 selected because it has several features that make it very similar to both research sites in Chobe and Gantsi in Botswana. Thou gh not as nearly as arid, the area is sparsely populated, dry and depends on wells as sources of water. . Along with the Upper East and Upper West Regions the Northern Region is one of the poorest in the country. The within region disparities between urban and rural northerners is just as vast.

PAGE 102

102 were cut short du e to a combination of factors. Days into the surveys it started to rain, causing great damage to livelihoods as village after village started flooding. By the time the rains ended, one green mango fruit after another started splitting, in a matter of days a calcium deficiency had According to the Food and Agriculture Organization, women comprise about 43% of the agricultural labor force in developing countri es and about 50% in sub Sahara Africa. Even though women and children do most of the daily tasks in small hold farming; men own the majority of these small scale farms, 57% in Botswana and 71% in Ghana. The problem is exacerbated by land tenure regimes tha t disproportionately favor men; this is also reflected in the data too. In Botswana where each adult Motswana, regardless of gender, is entitled to three types of land land for homesteads, farming, and for cattle posts women owned 43% of the farms samp led compared to 24% in Ghana. Chiefs hold land in trust for the community in Ghana. Depending on the area and region, land could be leased up to 99 years, with population pressures leases now run from 50 to 75 years. These arrangements differ, but only sli ghtly, from region to region in Ghana. Land acquisition tends to be easier in rural communities. The data from the Northern region cases hint at these differences. The region is one of the least developed in the country, and still predominantly rural. Thes e factors work in favor of women in farm ownership, Though a small number of women owned their farmers, the majority were either widowed, or daughters farming on a or have access to better land. See 5.3. distributions, at first glance seem to depart from current demographic trends in the sub region.

PAGE 103

103 Young men and women continue to move to urban areas in search of better jobs, 19% of the farmers in the Botswana and 10% of the Ghanaian sample are 24 years old or younger. According to a 2012 African Developme nt Bank Report, this shows that Africa is experiencing an unprecedented growth of the youth population (AfDB 2012:4), a trend not reflected in the data. This is largely due to the fact that in 2004 the Government of Botswana rolled out its Young Farmers Fu nd (YFF) a program aimed at encouraging youth, especially educated youth, to participate in commercial agriculture. The program sought to encourage those between the ages of 18 and 35 by offering low interest loans of up to P500, 000, around USD $70, 000, to take up farming (Madisa and Assefa 2011; Williams 2012). Madisa and Assefa found that in 2009 alone the program funded 84 horticultural projects (ibid: 266). These programs, combined more common barriers to participation. Young men in Ghana, on the other hand, have limited options. Government programs encouraging youth participation in farming have not fared well mostly due to misperception from the farming public, bureaucratic and financial barriers. Young men and women often have to raise their own money to lease farmland. This practice was more common in Brong Ahafo where land tenure laws allow female ownership. Poorer older women, mostly widows who are unable to work their lands, lease them out. The women who owned farms either inherited them from their fathers and/or husbands or bought or leased them. Almost all the medium to large scale commercial farms were owned by individual who were 45 years and older. The in depth interview s show the farmers in the Botswana sample grew mostly grains and/or sunflowers or beef, were almost always white for whom farming or ranching was their main source livelihood and thus businesses. These farmers lived on homesteads compared to their Ghanaian

PAGE 104

104 counterparts, who, though fewer in number and scale of operations did not have their permanent homes on their farms. A small number of white farmers fleeing the racial troubles in South Africa and Zimbabwe were either working as foremen for other commerci al farmers or owned their much small commercial horticultural farms in the Kananse/Pandamatenga corridor. But even this sub group of farmers were distinct from their African counterparts in their approach to farming farming as business and hence lots of book keeping. Almost The Ghanaian farmers in this age category were more diverse, ranging from local elites who made their money from trading commercially, usually as distributors of; for example, soap, rice, sugar and/or agro chemicals, 5.4. Most o f these farmers made their money from cocoa farming and then diversified into the wholesale/distribution subsector. Participation in the commercial export oriented fruit or cash crop production is also viewed as a way to diversify. This group also includes politicians, civil servants (both retired and those still in active service) and consortiums or various groups of people getting together to invest in medium to large scale operations. These tend to be absentee owners, usually living abroad or in Accra. T he farmers in the sample mostly identify as Christians, 36% in Botswana and 58% in Ghana, with 64% self identify as Muslim in the Ghanaian, keeping in mind the oversampling from the Northern region. As expected the majority of the farmers are rural, 52% in Botswana and 80% in Ghana 5.5. The sample is mostly rural, 5.6, 52% in Botswana and 80% in Ghana. Farmers in Botswana on average have higher levels of education ( 5.7), 12% report not having any formal education compared to 47% the Ghanai ans. Again, the data is skewed due to the over sampling from northern Ghana, a historically marginalized area where Koranic schools are not just simply more easily accessible, but are flexible enough for a lifetime of learning. 17% have elementary educati on 43% secondary, 23% postsecondary, 4% have a university degree and

PAGE 105

105 1% a postgraduate degree in Botswana. Ten percent of the Ghanaian sample completed primary school, 6% secondary 26% have some kind of postsecondary, 11% report having a university degree. 5.3 Extension officers: who are they? Of the 308 extension officers surveyed, 93 are from Botswana and of these 8% are from Chobe and 6% from Gantsi, 5.8. The Ministry of Agriculture is headquartered Gaborone in Botswana and Accra in Ghana many of the questions and interviews administered from both places even though the respondents were sometimes from different districts in town for a meeting, conference or workshop. The rest of 215 officers are Ghanaian. A quarter of the Ghanaian officers were fr om the Northern Region, 20% from Brong Ahafo, 25% and from Accra and the rest of other parts of the country but interviewed from Ministry of Food and Agriculture, Accra. Even with more women entering the work force and performing the bulk of the day to day tasks associated with farming like pruning and weeding; agricultural extension is still predominantly male. As can be seen from 5.9 54% of the Botswana sample and 81% in the Ghanaian are men. There are twice as many women in Botswana (46%) as Ghana 19% taking into consideration population size. The sample also shows an aging work force, but more so in Ghana ( 5.10). The youngest cohorts 25 34 make up 31% and 25% of the sample from Botswana and Ghana respectively. The next groups of officers aged 35 to 44 years comprise 32% of the officers in Botswana and 21% in Ghana. Those 45 years and older makeup 37% of the sample from Botswana and 54%, more than half of the Ghanaian officers. The issue of an aging workforce and changing attitudes towards the training of officers and the usual disparities in worldviews between age groups were topics officials, especially at

PAGE 106

106 the district level, wanted addressed and almost always introduced on their own. To these these were the most pressing challenges facing th e ministry. Distilled to its essence the debate was about training generalists as opposed specialists. Both seasoned officers and farmers find younger extension officers ill informed. Commercial farmers in Botswana especially disdained these officials who knowledge of, but which farmers think they should now. 6 Asked to name the qualities farmers down in Botswana and Ghana respectively. Asked to elaborate had to say, ork, yet you see so many of the officers they send us better. Years ago officers enjoyed and had passion for the profession, they enjoyed village life. Now they are off to Kasane, dr in because he has gone to their help, but day in and day out, there they go. [B]ut if you have to see him and his is not t here trained to feel for the leg of the elephant, mind you, the leg only. 7 Another Ghanaian farmer had similar recommendations for the recruitment of extension officers: could see [a 6 Intervie nd 2010 7 Interview with Mr. Mojo December 1 st 2010, Kasane, Botswana

PAGE 107

107 particular] weed and could tell you what was coming this year the rains will be bad, or use grow beans for a [about a month or two] then hire a [ plough] to [plough the beans] back into the soil before planting. Now you send for an officer and he takes days or weeks to show up. Then he tells you he is a block farm officer, 8 or cocoa only office you seen any of the officers with dirty hands and hardworking tell the people who sent you that if they did this it would benefit a 9 . In adequate staff levels and backdoor hiring, which over time is being suggested would undermine the already weak capacity of the service have only contributed to the problem. The better educated a workforce, they better its ability to analyze , articulate and critically review policy and its implementation. Thus the educational levels of the sample are an indirect measure of formal bureaucratic capacity. Overall, both samples show a well educated officer corps. 5.11 offers a summary of th e educational qualifications of officials. One percent of officers in both countries completed primary, these officials use to be numerators but the position now requires more than primary education. Numerators collect vital data like food pricing and 8 The block farms program brings several individual farmers together in large groups in product ion area where they then receive extension services including credit in form of mechanization services, certified seed, and subsidized fertilizer if they qualify. The program is one of four programs aimed at getting farmers to increase adoption of modern i nputs, increase farmer productivity and incomes. The program also targets youth, in an attempt to get them to view agriculture and farming as a viable business option. 9 Interview with Mr. Amoako, May 17 th 2011 Ejisu, Ghana Interview with Paul Duah, Ibrahi m Awudu and Osei Akoto, May 17 th 2011 2 4pm, Ejisu, Ghana

PAGE 108

108 crop ping patterns. Ghana has discontinued the practice of employing people with just primary education as enumerators and is currently phasing them out of the ministry. Data collection is now being done by personnel with higher levels of education at the direc torate of Statistics, Research and Information (SCRID). Another 1% in Botswana and 4% in Ghana have high school diplomas. Three percent of the officers in Botswana and 25% of the Ghanaians have some type of vocation education; 5% of the Tswana officers and 21% of the Ghanaians have diplomas or certificates, 12% and 15% have a college degree in Botswana and Ghana respectively while another 10% have some type of post graduate degrees in Botswana compared to 5% in Ghana. While officers in Botswana are more li kely to have the cost of higher education covered by government programs, in Ghana where government aid is much lower or even non existent for certain classes of officers, officers invariably cover these costs by themselves. In cases where the government w as funding officers, they still had to finance their education while waiting for reimbursement. Ghanaian officers were also more likely to attend technical schools because it was cheaper, and required less of a sacrifice by families. Officers typically att end classes, usually a few times a week, while still working and thus collecting a paycheck. Therefore family members are not expected to, or asked to, sacrifice as much. Payoffs on investing in a diploma are quite high. A person can become a mid level off icial, which comes with perks like more travel opportunities, as well as the likelihood of being assigned to one of the many projects that extension departments are constantly trying out. This also opens additional career options and opportunities by way o f project/program related wages/fees. The officers, like the farmers are mostly self identify as Christian, 83% in Botswana and 66% in Ghana ( 5.12), also mostly rural 52% in Botswana and 80% in Ghana ( 5.13), and are either married or in relati onships 61% and 84% in Botswana and Ghana respectively. 5.14. 5.15

PAGE 109

109 shows 42% and 55% of the junior officers are married, 38% and 37 of mid level officials and 20% and 8% of the upper level officials are married in Botswana and Ghana respectivel y. The distributions for the number of dependents of officers are reported in 5.16. Of the 40 officers who answered the question in the Botswana sample 8% report not having any dependents, 23% have at least one, 38% two, 20% three and 14% four or mor e. Twenty two percent of the 63 Ghanaians who answered the question have no dependents, 24% have one, 16% report two, 17% three and 21% four or more. Asked to describe their economic situation as of the time of the survey 4% of the officers thought they we re doing much worse, 16% worse, 39% about the same 36 better and 4% much better than a year ago ( 5.17). Finally, asked if officers chose a career in the civil service because of better wages and conditions of service 7% answered yes, 93% said no in Botswana compared to 29% of the Ghanaians who chose a civil service career for the wages the conditions of service. Seventy one percent claim wages and services conditions did not factor into their decision to work for the government. The results are repo rted in 5.18. Having described the salient characteristics of the sample, we will now turn our attention to the issue of norms. The goal of the analysis below is to establish (i) if officers and farmers differ in their perceptions of RCAD norms and ( ii) if there are any country differences. The Botswana is presented first, followed by the data on Ghana for each of the indictors. 5.3 Portraits from The Sampl e I will offer a quick summary about the essential features of the respondents in the study. Our average officer in Botswana is male, at least 45 years old and iden tifies as a Christian. This officer is either married or in a relationship and from the qualitative data he is mostly likely in a

PAGE 110

110 relationship as marriage rates in Botswana are some of the lowest, even for a Southern African country. Women are increasingly opting out of marriage, choosing instead to retain autonomy over self and control over their children and resources. 10 Our officer is most likely a mid level official. His Ghanaian counterpart is also male, also 45 years and older, has a diploma, identifie s as a Christian, and is married or in a relationship, but unlike his colleague in Botswana, this officer is more likely to be a junior officer. The average farmer in the general sample is male and young, 25 to 34 years old. The average Batswana farmer is 25 to 34 years old, attended secondary school, is Christian, uses some of the services provided by the ministry and lives in a rural area, though what is considered rural in Botswana would be more of a town in Ghana. The average Ghanaian farmer is also 25 to 34 years old, has either no formal education or never graduated from primary school, and he may or may not use services provided by the ministry (50% did not). These are important factors to keep in mind as we turn our attention to the discussion of t he affective topography of these individuals. With this in mind we can now turn our attention to the diagnostic tests. 5.5 Diagnostic Tests Next, I perform the necessary diagnostic tests in this section. Given the oversampling from Northern Ghana we need t o conduct some non parametric tests. To this end I need to test for correlations to test for associations between the various measures of dependent variables. square tests are performed to test for the strength of the association where necessary. The Spearman correlation test is especially apt because it does not assume normality in the distribution of the data. The use of Ttests and Anova test the 10 Co/Botswana.html#b#ixzz2gVVIPKaR

PAGE 111

111 difference between groups means to more effectively communicate k ey differences between the two country samples. I test for the strength of the relationships using four types of tests: chi ratio chi squared ( X 2 ) statistic examines the likelihood that th e results occurred by chance. The significance levels are reported (* 0.10 ** ranges from 0 to 1, where I indicate a highly strong association between the variab V values that are less than .20 indicate a weak relationship. Values that lie between .20 and .49 indicate a moderate relationship, and those that are more than .50 indicate a strong relationship. Gamma tests for the relationship between ordi exact test has been included where the cases in the cross tabulation are few; usually it is when the cases a rejected because there is a relationship between the variables. Both samples were tested for heterogeneity, the findings are discussed in detail in the relevant chapte rs, using multiple checks given the differences in sample sizes. For example, of only 1of 42 officers in the sample are from Botswana, and the rest (179) are Ghanaian, a two sample ttest with unequal variances is performed to check whether there are any si gnificant differences, even though no assumption of equal variances for the two populations are made.

PAGE 112

112 5 1 Distributions of Farmers by Region Botswana N % Ghana N % Total N % Chobe 124 (7) 12 4 (7) Gantsi 400 (22) 400 (22) Brong Ahafo 336 (19) 336 (19) Northern 941 (52) 941(52) Total 524 (100) 1,277 (100) 1,801 (100) 5 2 Distributions of Farmers by Gender and Country Botswana N (%) Ghana N (%) Total N (%) Male 292 (57) 948 (76) 1240 (71) Female 222 (43) 296 (24) 518 (29) Total 514 (100) 1244 (100) 1758 (100) 5 3 Distributions of Farmers by Gender and Region Region Male N % Female N % Botswana Chobe 71(6) 52 (10) Gantsi 221 (22) 170 (33) Total 292 (100) 222 (100) Ghana Brong Ahafo 192 (15) 79 (15) Northern Region 756 (61) 217 (42) Total 1244 (100) 518 (100) 5 4 Distributions of Farmers by Age Botswana N % Ghana N % Total N % Younger than 24 years 99 (19) 132 (10) 231 (13) 25 34 years 176 (34) 485 (38) 661 (37) 35 44 years 115 (22) 339 (27) 454 (25) 45+ years 125 (25) 312(25) 437 (25) Total 515 (100) 1,268 (100) 1,783 (100) 5 5 Distributions of Farmers by Religious Affiliation Botswana N % Ghana N % Total N % Christian 185 (36) 403 (58) 588 (34) Muslim 2 (0) 786 (64) 788 (45) No 221 (43) 4 (0) 225 (13) Traditional African Believes 103 (20) 38 (3) 141 (8 ) Total 1,231 (100) 1,762 (100) 5 6 Distributions of Farmers by Settlement Type Botswana N % Ghana N % Total N % Rural 273 (52) 1008 (80) 1339 (74)

PAGE 113

113 Urban 251 (48) 247 (20) 462 (26) 5 7 Distribution s of Farmers by Level Of Education Botswana N % Ghana N % Total N % None 60 (12) 592 (47) 652 (37) Primary 85 (17) 130 (10) 85 (5) Secondary 216 (43) 72 (6) 346 (20) Post Sec 113 (23) 335 (26) 185 (11) University 20 (4) 136 (11) 355 (20) Postgrad 3 (1) 139 (8) Total 497 (100) 1,265 (100) 1,762 (100) 5 8 Distributions of Officers by Region Botswana N % Ghana N % Total N % Chobe 25 (8) 25 (8) Gantsi 17 (6 ) 17 (6) Gaborone 51 (17) 51 (17) Brong Ahafo 62 (20) 62 (20) Northern 77 (25) 77 (25) Accra 76 (25) 76 (25) Total 93 (100) 215 (100) 308 (100) 5 9 Distributions of Officers by Gender Botswana N % Ghana N % Total N % Male 50 (54) 172 (81) 222 (73) Female 43(46) 41 (19) 84 (27) Total 93 (100) 213 (100) 306 (100) 5 10 Distributions of Officers by Age Botswana N % Ghana N % Total N % 25 34 27 (31) 51 (25) 78 (27) 35 44 28 (32) 42 (21) 70 (24) 45yrs.+ 32 (37) 108 (54) 140 (49) Total 87 (100) 201 (100) 288 (100) 5 11 Distributions of Officers by Level Of Education Botswana N % Ghana N % Total N % No 0 0 0 Primary 4 (4) 5 (2) Secondary 4 (4) 11(5) 15 (5) Cert/Dip 3 (3) 73 (36) 76 (26) University 20 (4) 136 (11) 355 (20) Postgrad 3 (1) 139 (8) Total 497 (100) 1,265 (100) 1,762 (100) 5 12 Distributions of Officers by Religious Affiliation Botswana N % Ghana N % Total N %

PAGE 114

114 Christian 35 (83) 114 (66) 79 (60) Muslim 7 (17) 59 (34) 52 (39) No Affiliation Trad. African 1 (1) 1 (1) Total 42 (100) 174 (100) 132 (100) 5 13 Distributions of Officers by Settlement Type Botswana N % Ghana N % Total N % Rural 273 (52) 1008 (80) 1339 (74) Urban 251 (48) 247 (20) 462 (26) 5 14 Distributions of Officers by Marital Status Botswana N % Ghana N % Total N % Single 26 (29) 28 (14) 54 (19) Married/In relationship 55 (61) 163 (84) 218 (76) Widowed/Divorced 9 (10) 4 (2) 13 (5) Total 90 (100) 195 (100) 285 (100) 5 15 Distributions of Officers b y Marital Status Botswana N % Ghana No (%) Total No(%) Jnr Officer 31 (42) 103 (55) 134 (51) Mid Level 28 (38) 70 (37) 98 (37) Snr. Officer 15 (20) 15 (8) 30 (11) Total 74 (100) 188 (100) 262 (100) 5 16 Distributions of Officers by N umber of Dependents Botswana No (%) Ghana No (%) Total No (%) 0 3 (8) 14 (22) 17 (17) 1 9 (23) 15 (24) 24 (23) 2 15 (38) 10 (16) 25 (24) 3 8 (20) 11 (17) 19 (18) 4+ 5 (14) 13 (21) 18 (18) Total 40 (100) 63 (100) 103 100) 5 17 Distributions of Officers by Own Economy Compared to Last Year Much worse 9 (4) Worse 33 (16) Same 80 (39) Better 73 (36) Much better 8 (4) Total 203 (100) 5 18 Distributions of Officers by Motivation for Taking Civil Service Jobs Botswana No (%) Ghana No (%) Total No (%) Yes 3 (7) 37 (29) 40 (24) No 38 (93) 87 (71) 125 (76) Total 41 (100) 124 (100) 165 (100)

PAGE 115

115 CHAPTER 6 FROM FARMERS EXPECTATIONS TO A CTION 6.1 Introduction Building on the insights from the previous chapters we now, bo th in this chapter and the next, shift our attention to the individual. The task at hand is twofold. The first and more pressing is assessing the degree to which the hypothesized relationships between immersion in affective reciprocal interactions and thei r negative effects on state capacity hold at the micro level. 1 The second goal is to lay the groundwork for fleshing out the empirical manifestation of the reversed collective action dilemma. The individual, as the main mechanism through which affective re ciprocity is channeled, is the unit of analysis here and, again, in the following chapter. The empirical evaluation is conducted in several steps and includes various descriptive statistics and a series of regression techniques. Given the nature of the dat a many self reported survey questions instead, through the use of several alternative techniques, demonstrate that the evidence points in the same direction and that the res ults remain substantially the same regardless of estimation strategy. The chapter is divided into four broad sections. Sections 6.2 and 6.3 present the descriptive data on the dependent and independent variables. The results of the linear, negative binomin al, and logistic regression analyses are presented in section 6.4. The chapter ends with a discussion of the results and the concluding remarks in section 6.5. 1 aggregation of values o and five hold at the individual level (King 1997:6; Peters 44 45; Przeworski and Teune: 60; Robinson 1950: 351 57).

PAGE 116

116 6.2. Weak State Administrative Capacity The outcome of interest, Weak Administrative Capacity , consists of a constellation of behaviors inimical to the delivery of effective extension services in particular and government services in general. What weakened capacity looks like and how the state gets bled of its abilities usually starts with how we ap ply, operationalize, and aggregate disciplinary concepts. However, given the subject matter and the study participants, I take as my starting point the factors that both farmers and extension officers believe to be undermining agricultural productivity. Th e qualifications, skill sets and job descriptions of the officials in this study are the same in Botswana as they are in Ghana. Both extension service units have a mix of technical, professional and support staff, and the roles in each countr y require simi lar skill sets. Given the which they are evaluated to be clearly defined. We also expect officials to be responsive to their are to approximating these standards, the more we can assume farmers, on average, will be fairly satisfied with ministry services. In the absence of sophisticated indicators, affective reciprocity is measured indirectly using proxies. The final indicators used in the analysis were developed after months of discussions with both farmers and extension officers. Long before the surveys were administered, farmers were asked whether they understood what the agricultural policies of their governments were and how those policies related to them. The two most common answers that farmers gave in regards to these policies were: (i) increasing food production and (ii) making farming an economically viable busin ess. 2 2 Becoming self sufficient was the third mos t common reason given by male Ghanaians in both the Brong Ahafo

PAGE 117

117 Despite these defining characteristics, discussions about productivity and income included widen, will continue to grow until structures are r uined. Farmers in Botswana pointed to what they saw as the difference between extension officers and commercial farmers or ranchers. fail to check expensive equipment, tighten a loose bolt or screw because, first and foremost, supervisors and senior officials lacked these important skills themselves. Secondly supervisors did not properly supervise and monitor their subordinates, leading to a general culture of impunity. The key factors, agreed upon by both farmers and officers, converged around: (i) inadequate supervision and monitoring of civil servants and (ii) a generalized apathy stemming from low wages and a demotivated workforce. Ghana. 3 fire way to end up a hot man at 4 your own many afforded women, their Northern counterparts cited being able to educa te their children. FIX [Interviews from Bots Oct conducted during world food day events in Maun, Chobe interviews Dec 2010, Gantsi Jan/Feb 2012. Ghana interviews Anita Hotel workshop, FOA/WB cowpea workshop Tamale, May 2011. 3 A common adage in Ghana tha t was trending among younger during the course of the study together with staying either mind your own business or not my job. 4 Mr. Ohene, Department of Extensions Services, Government Fertilizer Subsidy Program, Ministry of Food and Agric ulture, Accra, Ghana.

PAGE 118

118 We now turn our attention to the construction, measurement, and aggregation of indices shown in Figure 6 1 which are used to test the hypothesis that, all else being equal, farmers who socialize with officials are more likely to seek and get favors and by so doi ng are also more likely to weaken administrative capacity. 6.2.1 Dependent Variable 1: Frequency of Contact The Frequency of Contact index is made up of three behaviors: (i) fraternizing with officials, (ii) transacting official business outside officiall y sanctioned channels and (iii) using influence peddling that is thought to lead to the personalization and eventually, a perversion of, the depersonalized Weberian bureaucrat. The first, the degree to which farmers fraternize with officers during their official 5 The second indicator measures the number of times farmers and officials transact state business outside officially sanctioned channels. This is measured on a zero to three scale 6 The third and final indicator measures soliciting an influential person for aid to solve a problem with any branch of the national bureaucracy, and this also runs from zero to three. 7 The cross tabulations are reported in 6 1. The average farmer has little contact with the ministry or its officials. Even in 51% of the farmers report not visiting the 5 Farmers were first asked to answer yes or no to: Have you ever used any services offered by the ministry of How often do you visit officials at the minist ry of agriculture in a year? Coded: [0] Never, [1] Annually, [2] Biannual, [3] Quarterly, [4] Monthly. 6 In the last 12 months, how often have you gone to the home of a government worker in your area to seek help, or talk about an issue or collect some of ficial document? Coded: [0] Never, [1] Only once, [2] A few times, [3] Many times 7 During the past year, how often have you contacted some influential person with more money and power than you who can speak or act on your behalf? Coded: [0] Never, [1] Onl y once, [2] A few times, 3 [Many times]

PAGE 119

119 ministry or its local affiliates in the last year, 22% do so at least once a year, 10% biannually, another 10% quarterly, and 7% visit monthly ( M = 1. 03, ± 1.39). Slightly fewer Ghanaians (49%) report no contact, but of the 51% who do, 22% do so at least once a year, 7% biannually, 10% report visiting quarterly a nd 13% monthly ( M= 1.21, ± 1.52). A chi square test conducted to assess differences in socialization patterns between the two samples showed no meaningful across sample differences. The Ghanaian farmers were .17 points more likely to socialize with ministr y officials during official working hours : 2 (6) = 28.65, p =000, a finding corroborated by a weak Cramér's V of .13 1 as well as a two sample t test comparison of unequal means t ( 2.25) = 956.07, p= .025. However, even though there are no striking differe nces between the farmers on the second indicator, Transacting Official Business Outside Official Sanctioned Channels, reported in 6 2, the frequency distributions offer interesting contrasts. The majority of farmers in both countries do not engage in the practice, 50% and 57% in Botswana and Ghana respectively. In Botswana, 17% report doing so a few times a year, 19% many times (1.04, ±1.20). In contrast, 80 farmers (6%) out of 1,271 resort to this strategy many times a year, 19% at least once a year, and 18% visit a few times in Ghana (.73, ±.96). A phenomena masked by the various variance tests: Pearson chi square test of heterogeneity, run to check whether the frequency distributions differ across the two groups of farmers, suggests no meaningful di fference, (3) =67.32, p= sample t test (unequal) t (5.27) = 802.83, p = .001. The distributions for the third indicator, Influence Peddling, presented in 6 3 shows the Ghanaian farmers were more likely (50%) to resort to influence peddling compared to their Botswana counterparts, 61% of whom report never having to resort to this strategy t ( 3.55) = 964.42, p= .004. Of the 39% of farmers in Botswana who did, 14% claim to have engaged in the

PAGE 120

120 behavior only once, 13% a few tim es, and another 13% have found it expedient to do so many times ( M =.77, ± 1.09). In Ghana, 16% of farmers report resorting to this strategy at least once, 21% a few times, and 13% many times in the past year ( M =.97, ± 1.11), 2 (3) = 25.7, p = .001, Cramér's V=.12. 1 The underlying assumption of these indicators is that frequent interactions between farmers and officials in an affective reciprocal context increases the odds of personalizing what should be a formal working relati onship between a bureaucrat and a user of state services. Once personalized, the possibility of benefiting from subsequent interactions increases for both parties. The practice then becomes intrac due to spillover effects brought on when others witnes s these behaviors and their consequences. This is not to imply that individuals are incapable of interacting in a formal, depersonalized manner in Africa. Indeed, the majority of social interactions are of this nature. The point being made here is that la cing and layering affective expectations with Weberian formalism creates hybrid attitudes that are neither fully personalized nor completely depersonalized. This introduces a degree of arbitrariness into interactions between individuals acting in their for mal capacity as civil servants and the general public. For example, an officer may decide to bend certain rules or procedures for one user and not another and may not even be required by those watching the interaction to justify the criteria by which the o fficial arrived at that particular decision. It is quite common in bureaucracies the world over for formal rules and procedures to be tweaked or changed in the course of implementation. While this is not unique to Africa, the magnitude and pervasiveness o f the phenomena complicates how both officials and ordinary citizens deal with the state. Officials grumble about the lack of transparency in; for example, the

PAGE 121

121 consistent application of formal requirements and rules governing hiring and firing, tenure and promotion decisions. They get as frustrated as other citizens when some formal rules and procedures they took pains to learn about and base their claims on are substituted changed or neglected and leave them with no recourse but to make adjustments. In dep th interviews conducted during the pre dissertation phase of the study showed officers favored the broad dissemination of clearly specified rules and processes of the ministry. These rules and regulations should be easy and clear enough for the general pub lic and disseminated through multiple media. Officers favor this course of action because it makes it possible for officials to discharge their duties without fear of retaliation. Botswana has effectively branded itself in this regard. Almost every office in each ministry, department and agency has multiple posters listing the seven pillars national vision . At this point we could very well be charged with mixing apples with oranges because the indicator, Frequency of Contact, lumped together visitors who might have legitimate reasons for their frequent visits and cont acts with ministry officials with visitors who may not. The index also conflates or fails to sufficiently differentiate amongst the myriad of reasons people visit the ministry and its officers. The crudeness of the indicator also made it incapable of separ ating out visitors who were potential saboteurs of enhanced state capacity. To correct these issues a second indicator, Sought Favors, is used to get around the limitations of the Frequency indicator. 6.2.2 Dependent Variable 2: Favor Seeking Seeking and granting favors, often labeled corruption or nepotism, has been blamed for into three categories. The first group usually had no official business to trans act with ministry officials and tended to include relatives and friends or general members of the public looking for

PAGE 122

122 directions to some other government building. The second group had business to transact with the ministry. Included in this category are jo b applicants following up on applications. Another group were people who have been recommended to particular officials for assistance they need in navigating the national bureaucracy. Also included in this group are citizens looking to sign up for one or p rogram or another. 8 A third cluster of visitors included retired or still active civil servants who also happen to be farmers, 9 commercial farmers in Botswana, and out growers in Ghana. 10 Other visitors included individuals and/or groups affiliated with the agro/agrochemical industries. 11 These people tend to have considerable business to transact with ministry officials. 12 8 hours on end waiting to be helped. Showing up, no matter how effective the tactic is, is viewed as communi cating to into acting. 9 These are civil servants who also happen to be farmers and may or may not be producing for the more lucrative e xport market, but are definitely producing for local commercial markets. 10 maize growers. Contract farming refers to long term supply agreements between farmers and agribusiness processing/marketing companies/buyers that bring mutual gains and normally include price and supply arrangements (date, quantity and quality). Contractual arrangements may be verbal or written and vary widely, depending on the countries, crops and companies concerned. Schemes usually entail a range of activities (servi ces) that secure access to produce as in kind input supply or on credit extension services, transport for produce, and credit guarantees (Paglietti, and Sabrie, 2012:1 footnote 1 Ghana_Outgrower%20schemes.pdf ) . The agricultural industrial complex covers everything from credit, to inputs, processing and increasingly private extension services. In Ghana many of the executives and upper level officials in these companies were expatriates, the majority of the people from these companies who paid frequent visits were locals. This third grow of visitors were the most frequent of visitors in Ghana. They tend to be everywhere: workshops and seminars (favorite past times of both ministry officials and the agro industry/agrochemical/catering and hospitality industries). South African companies provided these services in Botswana. But for a variety of reasons I ne ver had the opportunity of interacting with them as much as I did in Ghana even though I interviewed a few. 11 Small to medium scale processors of raw materials and intermediate products derived from the agricultural, forestry and fisheries sectors. It is one of the most crucial links between harvesting and to the table. They range from simple preservation (such as sun drying) and operations closely related to harvesting to the production. The farmers in the meat production subsector have as wide a variety of end uses and at different levels of complexity in the two groups and individuals whose business brings them into close contact with the ministry and its officials. 12 Farmers in Botswana were more likely to visit ministry offices for permits whether to move cattle (Gantsi farmers) or produce (farmers in Chobe) to markets in the bigger towns. The government was dealing with fruit fly infestation a t the time of fieldwork. This should not change the results though since Botswana does a much better job enforcing its permit regulations than Ghana. This is no doubt driven by a desire to maintain access to EU markets while Ghana has no such pressures on its exports. The most common being continued efforts at

PAGE 123

123 The indicator Sought Favors is constructed from a question asking farmers how helpful officers were in responding to requests for persona lized help and for favors that go beyond the actions of thes administrative capacity. To get around possible social desirability issues, several measures of data triangulation were employed. First, farmers were asked if they found extens ion officers helpful in extracting unofficial favors. 1 The results are reported in 6.4. Once again the differences between the two samples are slight, yet they are, perhaps, qualitatively suggestive. Adding together farmers who report finding extens ion officers very unhelpful and unhelpful, we can infer that about 19% of the farmers in Botswana and 26% in Ghana were unsuccessful in their attempts at extracting favors from officials and although. For example, while 39% of the farmers from Botswana rep orted officers were helpful, only 20% in Ghana shared this sentiment ( M =1.69, ±.46). However 35% of the Ghanaians found officers to be very getting a vouche M = 1.55, ± .50). Both the chi square test and t tests substantiate this finding: 2 (4) = 129.14 , p= .000, a moderately strong relationship quarantining foot and mouth diseased cows. They are also less likely to be kept waiting which is not surprising. They work in the sector, know how the system works and how to get things done. They h ave extensive contacts and/or colleagues who might be either in middle or upper management and so can send junior officers running around to do the biding of their bidding. Some have enough clout to get even the big men running around on their behest. Ghan aian farmers in Brong Ahafo, on average were more likely to stop by ministry offices to inquire about information about credit, fertilizers and other agricultural inputs, or seek informal advice, the underlying message was usually a desire to trade or supp ly these inputs. While the individuals who I came to identify as frequent visitors to the Accra office, on average had policy agendas, a sizable number also had trading ambitions. Whether or not these individuals are successful in getting an official to he lp them is another matter.

PAGE 124

124 t (1.29) = 742.17, p= .20, with the Botswana farmers on average just a little more (.14) likely to seek favors. But just because farmers seek favors does not mean they succeed in getting officials to grant their requests. The indicator, Got Favors, seeks to capture only farmers who not only services for which these farmers were neither qualified, nor were entitled. 6.2.3. Dependent Variable 3: Got Favors The extension officer/farmer ratio in many African countries is extremely high. The current estimates for Botswana is a single extension officer serves about 2000 farming househo lds, meaning officers can only visit a fraction of the farmers in their area . 13 But even within the small group of farmers who might be working with a particular officer only a few receive any kind of unofficial help. The third dependent variable is the m ost restrictive of all three. Farmers who never sought favors, or sought but failed to garner any are treated the same as farmers who had no contact whatsoever with the ministry and its officials. Given the fact that the majority of farmers do not use mini stry services, we do not expect a lot of farmers to report getting favors. The results are reported in 6.5. Most farmers report getting some kind of favor from officials, thus 69% in Botswana were more likely to receive favors than 55% of their Ghan aian counterparts, 69% (± 2.45), averaging 1.69 favors compared to 55% of their Ghanaian counterparts (±2.39) who received an average of 1.55 favors. A one way analysis of variance (ANOVA) found no statistically 13 one extension officer to about 4,000 farmers. The official figures from Ghana i s one Agricultural Extension Agent serves about 1,300 farmers, estimates from fieldwork calculations with extension officers suggest a much higher ratio of 4000 or more farming households to a single extension officer.

PAGE 125

125 significant differences between the two grou ps of farmers in their success rates for extracting favors from officials: ( F (1, 1692) =3.39, p = .07). A two sample t test with an unequal variances comparison of means also produced similar results: t (1.82) = 725.14, p= .07. on of coefficients, computed to assess strength of correlation between pairs of the three dependent variables reported in 6.6, shows the indicators are all highly and positively correlated. The shows a strong and positive correlation between Fa vor Seeking and Frequency of Contact ( r s = .81, n= 430 p = .00), Got Favors with Frequency of Contact ( r s = .67, n= 430, p = .00) and Favor Seeking ( r s = .90, n= 430, p = .00). Similarly the correlations for the Ghanaian case are strong and positive: Favor Se eking and Frequency of Contact ( r s = .79, n= 1264 p = .00), as is Got Favors with Favor Seeking ( r s = .64, n= 1264, p = .00) and Got Favors with Favor Seeking ( r s = .86, n= 1264, p = .00). The farmers are more similar than different. 6.2.4 Aggregating the De pendent Variables The three alternate specifications of the dependent variable, Weakened Capacity, are constructed in a three step process and summarized in Figure 6.2 below. Step One, Constructing the three underlying indicators of Frequency of Contact The indicators and their ranges are reported below. Next all the values are added up to form the zero to ten range. See appendix C for a more detailed discussion. Step Two, Sought Favors: Farmers who found officials either very helpful or helpful in aidin g them in getting favors for which they were not qualified, are assumed as not only attempting to, but also succeeding in, extracting favors from officials. These scores are doubled, or weighted twice to run from 0 to 20.

PAGE 126

126 Step Three, Got Favors : Farmers w ho succeeded in getting favors are coded 1. Everyone else is coded 0. The summary statistics and a correlation matrix of the resulting dependent variables are reported below. 6.2.5 Transforming the Variables for Binary Logistic Regression A final step in the construction of the dependent variables involves transforming the indicators into dichotomous variables for the logistic regression analysis. To this end all farmers who reported not: (i) fraternizing with officers during their official working hours, or (ii) visiting officials at home or outside officially sanctioned channels of transacting state business, or (iii) any branch of the national bureaucracy ar index. Likewise, for the Sought Favors indicator all farmers who reported finding officers unhelpful are coded [0], all others [1]. The same goes for the Got Favors indicator. 6.5 provides the s ummary statistics for these three alternative measures as binaries. In sum, even though we see high levels of capacity weakening behaviors in both countries and even though the differences are modest at best, one would have expected farmers in Botswana to visit the ministry and its local affiliates more than their Ghanaian counterparts. Taken together the three indicators not only lend additional support to the similarities between g challenges. The number and range of regulations Botswana imposes to protect its beef industry and the huge investments it is making towards the Pandamatenga Agricultural Hub Project would have us believe state officials would enforce these rules and regu lations in a more robust way. One would also have expected the Ghanaian farmers to participate more in favor seeking interactions. Yet

PAGE 127

127 the cross tabulations suggest this might be not be the case. These issues will be taken up in the concluding section of t he Chapter, but for now we turn our attention to the construction and scaling of the independent variables. 6.3 The Independent Variables: Measuring Affective Reciprocity Affective Reciprocity is measured on three dimensions using four indicators. The fi rst, Embeddedness, captures the idea that those so blessed ought to help, and the second, Fairness , describes the idea that fairness and justice, against which claims are made and adjudicated, should be context specific. The normative aspects of prevailing affective norms are measured using on a four point scale from 0 = never to 4 = strongly agree. The second dimension, Weberianness , measured by the third variable, Professionalism, l service. This variable is also runs from 0 = never to 4 = strongly agree. The fourth and last of the independent variables, Requests for Assistance, measures th e material aspects of affective reciprocity shown in Figure 6.5 and it, too, runs from 0 = never to 4 = strongly agree. 6.3.1. Independent Variable 12: Embeddedness The Embeddedness indicator, the idea that helping relatives and friends is not only the ri ght thing to do and failure to do so deserves our scorn, measures the degree to which individuals are subsumed in, and subscribe to, the norms of duty of affective reciprocity. As members of society, individuals are enmeshed in cross cutting networks. Whil e embeddedness along the lines suggested by Peter Evans has several positive implications for enhancing state capacity. In Africa, the inverse is usually the default, and the state and the public sphere are

PAGE 128

128 viewed as another, separate part of society . 14 The indicator is constructed from five questions shown in Figure 6 6 and runs from [0] 15 farmers who strongly disagree, or disagree to [4] those in strong agreement or agreement on societal perceptions about social prestige and respectability. The descriptive statistics for each of the underlying indicators are reported in s 6.7 through 6.12. The farmers shared similar views about attaining and sustaining social prestige and a good name. The majority (87%) of the farmers from Botswana agreed with the sent iment that helping others conferred prestige (M= 3.45, ± .98). The Ghanaian farmers were even more .000, a moderate effect size, Cramér's V=.36, The farmers al so had similar views on the attainment of social respectability with 88% of the farmers in Botswana and in Ghana (Botswana M =3.37, ± .94 and Ghana M =3.43, ± .63). A Chi square test also showed the two groups differed only moderately 2 (4)=33. 03, p= .000, t( 1.36) = 713.82, p = .17. The vast majority of farmers, 98% of the Ghanaians ( M= 3.43, ± .63, Median =Agree) and 94% of the Tswana ( M= 3.42, ± .73, Median = Strongly Agree) agree with the idea that fin 14 The degree to which state elites ar e enmeshed in social networks and other relations that put them in close proximity with other important segments of society. For Evans the degree to which state elites and key actors from civil society interact enhances the ability of the state to formulat e long term goals. 15 Respondents were asked to agree or disagree with the following statements: V17. If you want prestige and a good name you should help people V19. People who help their relatives and friends are highly respected in our society V20. Som eone who has contributed financially to the development of his or her village is highly respected in our society V22. Someone who has served his village, for example as a councilor, is highly respected in this society V23. In our society we do not respect government workers who stick to the law when dealing with people Coded [0] Strongly Disagree, [1] Disagree, [2] Neither, [3] Agree, [4] Strongly Agree

PAGE 129

129 social prestige and respectability ( 6.9), t( .29) = 821.833, p= .77 and X 2 (4)=233. 03, p= .000 and 2 (4)=33.03, p= reported in 6.10. The majority (61%) of Tswana farmers disagreed with the idea that their fellow citizens are contemptu ous of officers who adhere too closely to the rules ( M= 1.61, ± = 1.30, V ar = 1.70, [ p50 = 1 disagree]). Conversely 69% of the Ghanaian agree with the statement ( M= 2.58, ± = 1. 20, V ar = 1.44, [ p50 = 3 agree]). A chi square test assessing differences between the two groups of farmers suggests statistically significant differences: 2 (4)=233. 03, p = .000 with a strong effect, Cramér's V value.36 as did a two sample t test, unequal variances, t ( 14.56)= 889.15, p= .000, (CI [1.50 1.72]). To sum up, the Embeddedness indicator suggests that both sets of farmers share similar views about prestige and obligations to relatives and communities. Yet, just because people profess to hold certain values does not mean they actually abide by them. 6.3.2. Independent Variable 2: Fairness If it is indeed true that the dominant norm system is one of affective reciprocity, then we the government and the primary implementers of national policies. The Weberian civil servant is the antithesis of his a operating in an affective context are expected to exercise discretion. Yet, even within an affective context, fairness is a much prized virtue. The Fairness variable measures tensions be tween what is considered fair from just an affective reciprocal point of view and what is considered fair from just a civic/public or formal view. Also measured, in part, are when one is expected to help and when one can or should

PAGE 130

130 expect assistance within reason. The indicator is constructed from answers to a vignette about an to [4] strongly agree. 16 Farmers were asked to react to the vignette in which the que stions are aimed at tapping into the posited tension between local notions of fairness against the Weberian impersonal official. Because the average subsistence farmer is rural, poor, conservative, and has little or no formal education, and is effectively excluded from the formal job market, and because these farmers will have no pensions in their old age, we expect them to be more sympathetic towards the officer. I expect farmers (i) to advocate for Mpho so he will be allowed to stay in the city, (ii) to favor the use of social sanctions to bring about their desired outcome, and (iii) to be more inclined to suggest that Mr. Kabelo deal with this family dilemma in a traditional manner by adhering to family values of affective reciprocity. In addition, I exp ect farmers will not want either Mr. Kabelo or Mpho to (iv) go through the formal channels of the civil service, and 16 Respondents were asked to listen to the following story: Mr. Officer is being transferred from the ca pital city to a village. His children are attending one of the best schools in the country. Mr. Officer also has his mother living with him due to her poor health. Living with him in the city means she has access to the best hospitals in the country. Becau se of these reasons Mr. Officer does not want to go the same place. Now, please tell me if you strongly disagree, disagree, neither agree or disagree, agree or strongly agree with the following statements: Q30. Mr. help Mr. Officer Q31. Mr. Boss should not listen to these people he should follow the rules and transfer Mr. Officer Q32. Mr. Officer should write a letter to Mr. Boss in his official capacity as head of department and explain his situation in a professional manner Q33. There is nothing wrong with Mr. Boss helping Mr. Officer Coded: [0] Strongly Disagree, [1] Disagree, [2] Neither [3] Agree, [4] Stro ngly Agree

PAGE 131

131 finally, the expectation, instead, would be that (v) Mr. Kabelo, the older more powerful member of the family, ought to do all he can to as sist his nephew. The results are reported in s 6.12 through 6.16. Asked to strongly disagree or disagree, or strongly agree or agree with the statement that particula r situation and allow him to stay in the city, 12% of the farmers in the Botswana sample strongly disagreed, 19% disagree, and 4% had no opinion one way or another. But of the 66% who thought otherwise, 21% and 45% agree or strongly agree with the suggesti on that Mpho be considered ( M= 2.70, ±1. 48, [ p50 = 3 agree]). The Ghanaian farmers also wanted the uncle to help his nephew, although by a smaller margin, with 40% agreeing, and 17% who strongly agreed ( M= 2.21, ± = 1. 33, [ p50 = 3 agree]); 2 (4) = 178.48, p =. 000, Cramér's V.32 and t( 6.38) = 847.67, p =.000 respectively , a strong association. slightly more than half the farmers in the Botswana sample rejected the use of social pressure to get Kabelo to keep his nephew in the city. Even supporters (38%) were less enthused with this argument. The suggestion was not too popular with the Ghanaian farmers either (36%), a trend the statistical analysis of variance un derscores: 2 (4) = 160.33, p =.000. The effect size for this finding, Cramér's V, was moderate, .30, t ( 7.93) = 834.08, p= .000, with the Ghanaian farmers rejecting this course of action more emphatically ( M= 56, ± = .07). The farmers were none too supportive of c dilemma either, as evidenced by the 33% of the Botswana farmers either strongly disagreeing or disagreeing with the statement that the uncle should not listen to his relatives but rather should

PAGE 132

132 transfer Mpho ( M= 2.42, ±=1.33). The results were similar for the Ghanaians, 36% ( M= 2.50, ±=1.27). All told, farmers from both countries were very supportive (57% Botswana, 59% Ghana) of keeping nephew and mother in the city. The variance tests showed moderate differences : (4) =77.09, p t ( 1.13)= 874.05, p= .26. A sizable group of farmers in both countries abide by norms of affective reciprocity in daily life. We also found strong support among fa rmers for Mpho to handle his transfer in a professional manner by registering his concerns through the formal rules and procedures of the civil service. Respectively eighty six per cent and 77% favor this strategy in Botswana and in square test shows statistically significant differences between the farmers 2 (4)= 158.46, p = .000), a moderate association. The result from a two sample t test with unequal variances indicates a difference in means: t (9.1452)= 1774, p= .000). Asked to respond to the statement that there was nothing wrong with Mr. Kabelo helping Mpho stay in the city, 17% and 29% of farmers in Botswana and Ghana strongly disagreed/agreed with the notion, 11% had no opinion in Botswana (7% Ghana), and 72% strongly agreed/agreed in Botswana (64% in Ghana). Once again the difference in means is stati stically significant (4) = 72.1445, p= .000) a moderately weak effect (Cramér's V= .20), t (5.5986)= 1760, p= .000). The hypothesized tension between the formal rules of the civic public and the private spheres means formal rules and procedures of the civic public arena get sh ort changed. Individuals who score high (either strongly agreeing or agreeing) on the affective reciprocal scale should sympathize with the officer, thus are more likely (i) to think Mpho be allowed to

PAGE 133

133 boss, Mr. Kabelo, that transferring his nephew could potentially be viewed negatively. We also expect more farmers, compared to the extension officers to disagree with the suggestion that Mr. Kabelo, the uncle in question should follow ministry policy and transfer his nephew. 6.3.3 . Independent Variable 3: Attitudes towards Officers The third independent variable, Attitudes towards Officers, measures the idea that the ce functions and (ii) actions of officials who are the face of the government. Attitudes towards Officers, of the dominant norms of the civil service and its officials a re reported in s 6.17 and 6.18. 17 The farmers in the Botswana sample are almost evenly divided on the subject of with 40% who strongly disagree/agree co mpared with 43% who strongly agree/agree with the idea ( M= 2.07, ± = 1.32). Most of the Ghanaians (76%) in the sample strongly agree/agree ( M= 2.40, ± = 1.22). A two differ in m eaningful ways about their expectations. Differences between the two samples suggest the Ghanaian farmers are about five times more likely to have these expectations as compared to their Botswana counterparts, t ( 4.91)= 882.39, p= .000). A Pearson chi squ are test ascertained across sample differences showed a moderately weak relationship (4)= 84.44, p =.000, 17 Please tell me if you disagree, or have no opinion or agree with the following statements: of what I find them doing Q28. I exp ect government officers to help me with whatever business I have even if it might not be directly related to their job Coded: [1] Disagree, [2] Neither [3] Agree

PAGE 134

134 distinctions or job description. The results are presented in 6.18. Thirty eight percent of the farmers in Botswana strongly disa greed/disagreed with this notion, and 51% strongly agreed/agreed ( M= 2.33, ±1.37, p50 = 3); 29% of the Ghanaian farmers strongly disagree/disagree, and 59% strongly agreed/agreed ( M= 2.44, ± = 1.17, p50 = 3). The across sample variation is slight: t ( 1.51) = 833.26, p=. 13, M= .10). (4) = 78.60, p= .001, application of norms for material aspect of affective reciprocity. 6.3.4. Independent Variable 4: Request for Assistance From Alexis de Tocqueville to Max Weber, Almond and Verba to Samuel Huntington we are reminded that culture matters. In societies where thrift and investment, hard work and prosper (Almond and Verba 1963, Eckstein 1988, Huntington 1968, 1996/2001, Tocqueville 1835/40; Weber 1905). nd control the means of production and those who sell their labor for a wage. Most political scientists and economists argue that given the right conditions and the right policies and institutions, individuals, regardless of culture, can improve their stan dard of living. The next variable, Requests for Assistance, is used to test the material dimension of affective reciprocity. To get around potential social desirability problems we examine the actual behaviors of respondents (Phillips and Clancy 1972, Arno ld and Feldman 1981, Watson, et al. 1986, Fisher 1993, Chung

PAGE 135

135 and Monroe 2003, Lara et al 2004, Allyson and Jon 2010, Tabbach 2010, Dalton and Ortegren 2011, Gonzalez Ocantos, et al. 2012). We expect individuals in denser networks to be under considerable pressure to meet earned resources. Using a two step process, we first need to establish the fact that people are actually under pressure to engage in these capacity sapping behaviors. To d o this we need to get an idea of the demands others made on respondents. See Appendix for a list of the most requests for assistance. Figures 6 through 8 report on the more common requests reported in the surveys. The variable, Request for Assistance, is constructed from responses to self reported 8. The Requests for Assistance indicator is a composite of five variables. The index has a zero to six range [0, never to 5, weekly]. T he results reported in s 6.19 through 6.24 show only the most common 6.20 shows the summary statistics for the two most common requests: helping to defray healthcare and tuition related costs is the first, a nd contributing towards weddings and funerals is the second. Respectively, seventy six per cent and 69% of the farmers in Botswana report making financial contributions towards defraying health/education and wedding/funeral related costs. The figures are e ven higher for the Ghanaian sample, 83% and 88% respectively. The distributions reported for the rest of the s have no special ranking or order. The data on financial contributions towards education and health care needs is presented in 6.21a nd shows that 56% of the farmers from Botswana make this type of outlay a few times as do their Ghanaian counterparts (69%).

PAGE 136

136 Respectively, 14% and 10% of the farmers make these contributions on a monthly basis, and in both countries these contributions ar e made 3% bi weekly, and in Botswana and Ghana 3% and 2%, respectively, contribute on a weekly basis. The across sample variation is slight: t (.1921)= 816.07, p= .000. The results of the chi square test also indicate that the two groups are largely similar on the issue of making financial contributions to defray tuition and health related costs: (4) = 28.91, There are significant differences between the farmers on contributions for weddings and funerals as presented in 6.24. Unlike Ghana, in Botswana the use of death insurance is fairly well established. Insurance or not, relatives and friends still have to contribute lots of money towards funerals. 6.22 shows that 31% of the farmers in the Botswana sample and 12% from Ghana report not making these outlays. But of those in Botswana who did, 35% reported havin g to make these contributions a few times year, 21% monthly, 8% biweekly and 5% weekly. In Ghana 67% of farmers reported making these contributions a few times a year, 10% monthly, and 6% bi weekly and weekly respectively, (4) = 180.20, p= .000, Cramér's V. 32. A two sample t test of unequal variances tests; however, suggests a more modest variance t( 1.17) = 827.96, p= .24. Once again we find strong evidence for affective reciprocity. In fact 76% of all the requests for f inancial assistance in Botswana are directed towards defraying healthcare and tuition related costs with 69% defraying costs for weddings or funerals. The pressures are similar for the Ghanaian farmers with 83% of requests for assistance directed towards d efraying healthcare related costs, and 88% of all requests for financial assistance are directed towards funerals and weddings. Given how poor subsistence farmers are, these are staggering financial

PAGE 137

137 burdens. They also underscore the pull of affective recip rocity in both countries. We now turn our attention to the material dimensions of affective reciprocity. 6.3.5. The Control Variables The majority of the communities in the study are rural. Economic activities, much of which revolves around small scale s ubsistence farming, are geared towards meeting the day to day needs of farming households. Though women and children do most of the farm related work, their access and control over reproductive resources is greatly constrained (Denise Paulme 2013, CARE 201 0; Chaudhury, 2012, FAO 2011). Feminist scholars have long contended gender (Barrett and Morgenstern 1974; Johnson 1983; Kulik 2000; King and Mason 2001; Timberlake 200 5). Women are; therefore, more likely to be self employed in the unregulated informal sector and have little, if any, direct impact on effects we are testing for. As in many conservative societies men are under considerable pressure, as bread winners, to provide for relatives, including in more likely to engage in the types of interactions of interest to us. I therefore control for sex: female = 0; male = 1. With women working in the informal sector, unemployed and under employed males with lots of free time become natural partners for those seeking to divert state buying schemes at the very local level. For these reasons we also control for political affiliation (supporter of opposition party = 0; supports incumbent government = 1). e asked respondents to evaluate their personal economic situation relative to other citizens, the indicator

PAGE 138

138 Estimate net worth is coded (very poor = 0, very rich = 4; transformed into 0/1 for logit analysis). We also control for settlement type (urban = 1, rural = 0). 6.4. Data Analysis and Discussion of the Results The goal of this section is the empirical evaluation of the hypothesized effects of data, I analyz e it using three techniques; each of these estimation strategies has its limitations, especially if used alone. I report the results for all three empirical strategies rather than making a choice a priori. To the extent that the various approaches point in the same direction regardless of strategy should be indicative of a robust empirical relationship corroborating the general theory proposed in this dissertation. The first and perhaps more obvious is the use of an additive scale in the construction of th e three dependent variables. The effects of the economy of affection on administrative capacity are gradual, symmetric, and cumulative. Generally, if one accepts this characterization of the dependent variables, a linear regression is an appropriate way of estimating the effects posited here. This is the case even where the indices are restricted range wise as long as the hypothesized causal relationship between the independent and dependent variables is linear, symmetric, and probabilistic (Agresti and Fin lay 1997). But reliance on a linear regression estimation strategy alone, given the nature of the data, opens the study up to charges of inefficient, inconsistent, and biased estimates (Long and Freese 2006). treats affective reciprocal interactions as count outcomes of self reported normative and behavioral choices and actions of the individual farmers and officers surveyed. This second approach may be more appropriat e, at least on a theoretical level. I am well aware of the problematic nature of such self reported data (as discussed above) as well as the fact that this

PAGE 139

139 approach may portray a level of precision in the measures that, strictly speaking, is not there. Giv en the possibility that the data may well have a non linear relationship to other factors, a count regression framework in some ways addresses the nature of the underlying data more effectively than a linear regression model. A Negative Binominal Regressio n Model (NBRM) also offers a way out for two principal reasons. First, count outcomes are usually not normally distributed and thus violate some of the assumptions of linear regressions. A Poisson test showed variance greater than the means, confirming ove r dispersion in the data and justifying the use of the NBRM. Added to this are s tandard tests that also indicate a rejection of the hypotheses that the extra parameter alpha is equal to 0 for all three dependent variables: the Frequency of Contact indicato r ( N = 1682, alpha = .29, 2 (10) = 330.81, p= .000), Sought Favors ( N =1591, alpha = .75, 2 (10) = 220.61, p= .000) and Got Favors ( N =1591, alpha = 1.83, 2 (10) = 117.25, p= .000). Finally, the question of the actual precision of the measures built from the underlying survey data must be dealt with. One could argue that the relatively fine graded measures of the dependent variables resulting from the index/count constructions artificially create a degree of precision that does not exist empirically. The question of whether frequency of contact or the number of favors a farmer receives varies from 3 to 4 or 5 and what this information actually means is a valid question and critique. That said, the variation from a low of zero to traits counting upwards to 20 is probably still meaningful. The three dependent variables have ranges from zero to twenty: Frequency of Contact 0/10, Sought Favors 0/20 and Got Favors 0/1. The three of the four independent variables have a 0/20 Embeddedness, Fairness and Request for Assistance while the fourth, Attitudes towards Officials , has a 0/8 count ( 6.22). My intuition stems partly from spending months in the field with these farmers, and while the difference

PAGE 140

140 between two and three in of itself may not be very important, differences between having two or three traits is also meaningfully different from having ten or twelve. Some may find these issues important and problematic enough to argue that the only qualitatively important difference is between zero and some (traits), be those a few or many. If one considers this a devastating enough of a shortcoming, thus a decisive limitation of the findings, a way out is offered by an alternative conceptualization that distinguishes between farmers and officials who rarely engage in reciprocal relationships an d those who do. A third empirical strategy, logistic regression, is used to tease out these effects. This empirical evaluation of the suggested theory is the most stringent approach of the three. For this purpose, the dependent variables were transformed a s discussed above into binaries where zero The analysis and discussion of the results are organized as follows: the results for the linear regressions are presented first in the following section, these are followed by the results from the NBRM models. The findings from the logistic regressions are presented in the third section. The chapter concludes with a summary the findings, and a discussion of the findings for both C hapter 6 and 7 is presented in the concluding chapter. 6.4.1. Linear Regression Analyses of Affective Reciprocity The results of the linear regression analysis, conducted to estimate the effects of affective reciprocity on the administrative capacity rep orted in 6.23 suggests farmers who subscribe to the cluster of values subsumed in the indices Embeddedness (helping relatives and friends is not only the right thing to do; failure to do so deserves our scorn) and Fairness (being obliged to help and having the right to reciprocal assistance, within reason) are negatively correlated to the three dependent variables, but significantly so for only the Fairness index.

PAGE 141

141 Holding all other factors constant, a one unit increase on the Embeddedness index is ass ociated with a three point decrease in Frequency of Contact , an 11 point decrease in Sought Favor , and a six point decrease in Got Favors . Fairness on the other hand is significantly, but negatively correlated across all three models. A unit increase on th e Fairness scale not only triggers a seven point decrease in Frequency of Contacts , it also reduces the odds of engaging in any of favor seeking behaviors by 12 points, which, not surprisingly, translates into a five point loss in Got Favor . Attitudes towa rds Officials has no relationship to the dependent variables. Requests for Assistance, on the other hand, is robust across models and in the expected direction. Comparatively, holding all other factors at their means, a unit increase in Requests for Assist ance increases Frequency of Contact by .23 points, Sought Favor by .35 points, and Got Favor by .11 points. A unit increase in Age increases contact with officials by .16 points, .44 points for Sought Favors and .27 for Got Favors . As expected male farmer s are more likely .55 points to have some kind of contact with officials, increasing the odds for,seeking favors by .93 points and getting a favor by .40 points. Wealthy farmers are also significantly more likely to have contact with officials; a unit incr ease in wealth has a corresponding effect of .18 points on increased contact with officials, and a .22 point advantage on seeking favors, though both contacts and favor seeking behaviors are unlikely to translate into favors for the wealthy. And while Ghan aian farmers have a 41 point advantage over their Tswana counterparts when it comes to Frequency of Contact , it seems like these interactions do not translate into getting favors. Meanwhile supporters of any of the opposition parties in both countries lose out on Frequency of Contact by 50 points. Not supporting the party in government also sets farmers back .94 points when it comes to seeking favors and .37 points in Got Favors . A unit increase in remoteness decreases

PAGE 142

142 contact with officials by 16 points, s etting farmers back by .48 and .29 points with regards to seeking and getting favors. On one level this rather unsettling finding since the overwhelming number of farmers are rural, on another level it is to be expected, the average subsistence farmer in t he study hardly visit their local ministry of agriculture offices or even use their services for that matter. But that is not to say they do not seek informal assistance from other farmers known for their associations with extension officers. The claim tha t a set of normative values predisposes individuals to engage in affective reciprocal interaction that ultimately undermine state capacity has so far not held up. Rather, the be important factors in seeking and courting relationships that can ultimately be mined for favors if need be. I explore this dynamic further with a negative binominal regression analysis. 6.4.2. Negative Binominal Regression Analyses of Affective Recip rocity The results of the negative binomial regressions (NBRM) are presented in 6.24. All the variables used in the analyses are the same as those used in section 6.4.1. The only difference between the OLS models and the NBRM regressions is the trea tment of unit changes increases or decreases as counts rather than as intervals on an infinite scale. The substantive effects of the hypothesized effects teased out by the NBRM are also in some ways easier to interpret within a count regression framewo rk. To facilitate a direct understanding of these effects, I report the beta, robust standard errors, and levels of significance (first three columns under each of the dependent variables) in 6.24 in addition to the predicted percent changes in expec ted count on the dependent variables for each unit increase in the independent variables all in one . As expected, only the material dimension of the affective reciprocity Request for Assistance is significantly and positively correlated across model s. This lends additional support

PAGE 143

143 to the main hypothesis that economic pressures, rather than some inherent normative worldview drive much of the behaviors posited as inimical to enhanced administrative capacity. Fairness and Request for Assistance have sig nificant effects on all three of the dependent variables. Embeddedness has no effect on contact with officials, but even when farmers who subscribe to the ideas of affective reciprocity subsumed in the index do finally manage to interact with officials the y are statistically significantly, though negatively correlated to Sought Favors and Got Favors . And as before Attitudes towards officials fails to demonstrate any relationship to the dependent variables that is distinguishable from zero. Again, the claim that farmers still expect their relatives to give them preferential treatment in their dealings with bureaucracies they happen to work in is no longer plausible. Robert Price found this to be the case in Ghana, circa 1975. If it holds at all, it is usually at the very elite levels when the rich and powerful interact. That said, the results add to the plausibility of my argument, arguably strengthening the earlier findings discussed above and in depth interview data. The first of the independent variables, Embeddedness, running on a continuum of 0 to 20 (the counts are reported in 6.22) has statistically significant effects on two of the three dependent variables. The first, Sought Favors increase on t he Embeddedness index decreases the expected count on Sought Favors by 2 percent. This means that all else equal, a farmer who strongly subscribes to the values measured by the index (helping relatives and friends is not only the right thing to do, failure to do so deserves our scorn) is expected to have a 40% lower count on Sought Favors compared to a farmer who does not share these normative preferences at all. That is a substantively significant finding when it comes to taking actions that in effect thre aten administrative capacity.

PAGE 144

144 The same set of preferences captured by the index for Embeddedness has an even stronger effect on the third dependent variable Got Favors . Recalling that this third alternative specification of the outcome of interest (weakeni ng of state administrative capacity) is the most demanding and restrictive with its 0/1 range where [1] represents farmers who both sought, and succeeded in getting personalized assistance outside of the regular job of the extension officers and [0] is use d for everyone else. With a binominal outcome variable, the NBRM essentially turns into a logistic regression. Yet, the effect is substantial. The expected count on the dependent variable is a 4 percent increases for each unit increase on the Embeddedness index. This means that a farmer who holds strong norms of affective reciprocity is 80% percent less likely to have sought and gotten personalized favors from extension officers than a farmer who rejects such norms. The models also show that the Fairness in dex, the obligation to help and the right to correlated to all three dependent variables, albeit negatively as with Frequency of Contact (0/10 scale), Sought Favors (0/20) and Got Favor (0/1). Thus a unit increase on the Fairness index decreases the expected count on Frequency of Contact by 5 percent, meaning that all else being equal, a farmer who strongly subscribes to the values measured by the Fairness ind ex lowers contact with officials by 50%. Similarly a unit increase on the index corresponds to a 3 percent decrease in Sought Favors, 60% drop. A unit increase in Got Favor decreases the expected count in the indicator by 3 percent and translates into a lo ss of 60% in potential favors. Request for Assitance the board with a unit increase increasing Frequency of Contact by 160%, Sought Favors by another 160%, and Got Favors by 120% for individuals under financial pressure.

PAGE 145

145 Finally, we note that among the control variables, gender and being a supporter of the opposition party have consistent and substantive effects on the expected counts on the three dependent variables. The relationship s are in the direction expected. Men are much more likely to hold norms of reciprocity, seek out officers to get favors, and are more likely to get such favors than are women. These results corroborate the literature that speaks of reciprocal relationships are typically excluded from this form of favoritism to a greater extent than are men. Conversely, supporters of the opposition parties are much less likely than others to h old such norms and to seek and get favors. It is a finding that supports those who have argued that favoritism and clientelism in Africa to a large extent are distributed along party lines and may fill a function to (re)create political loyalties that can be called upon during elections but at other times as well. 6. 4. 3. Lo gistic Regression Analysis of Affective Reciprocity We now turn our attention to the results of the logit regression analyses reported in 6.25. The beta coefficients reported are o dds ratio transformed to show factor changes in the odds of farmers acting in ways that weaken administrative capacity. Stata regresses the exponential regression on the right side of each logistic equation then multiplies that by the other raised to the p ower of x to increase the odds by e in a multiplicative manner. This means that for a unit change in an independent variable, the odds of an outcome are expected to change by a factor of e . The odds ratio can be interpreted as follows: for a unit increas e in x k , the odds of a lower outcome compared with a higher outcome are changed by the factor exp ( k ), holding all other regressions have multiple interpret ations; probabilities, odds and logged odds each have

PAGE 146

146 advantages and disadvantages. I will primarily use probabilities because they have intuitive meaning but are non linear and non additive. 18 A variance inflation factors (VIF) multicollinearity test revea led no approximate linear relationship between the independent variables. The mean VIF for all three models was 1.15. Even though collinearity does not necessarily affect the ability of a regression equation to predict a response, it hinders our ability to effectively evaluate the impact of individual variables for the question at hand. Finally, because the variables used in the logit model are categorical and nonlinear, we cannot infer much directly from the coefficients. That said, a quick look at the (6 25) suggests statistically significant correlations across some of the indicators are consistent with the OLS and Negative Binominal regression models. We have so far established that the two measures of normative affective reciprocity have a substa ntive, but negative effect across all three models, but in an inconsistent manner. Economic, rather than normative concerns, still seem to be driving these interactions. While we failed to demonstrate any relationship between Embeddedness and Frequency of Contact and Sought Favors in Models 7 and 8, the indicator has a moderately statistically significant effect on getting favors, negatively so, as expected; the Fairness index is still statistically significant, but moderately so, and trends in the right d irection. Meanwhile, for the first time Attitudes towards Officials is significantly and positively correlated to contact with officials and engaging in favor seeking behaviors. That said, the hobnobbing does not seem to translate into favors. And as expec ted, farmers under financial strain brought about my Requests 18 This means that each factor has a varying impact on the dependent variable depending on th e values of the other independent variables (Pampel 2000, 18 19). The test for significance in logistic regression is similar to standard c an then be evaluated using the z distribution as long as the size of the sample is sufficiently large (Long 1997, 54; Pampel 2000, 30).

PAGE 147

147 for Assistance from relatives are significantly more likely to contact officials and engage in favor seeking interactions, even though the odds of these financially stressed farmers getting favo rs are small, but they are nonetheless the only group of farmers who get any favors in this most stringent test of the theory. While we failed to demonstrate any relationship between age and Frequency of Contact and Sought Favors , age is robustly correla ted to getting favors; in fact it is the only indicator robustly so. Being a man has a significant and positive effect on contact with officials and seeking favors, but the odds of getting favors are small. Conversely while Ghanaian farmers are significant ly more likely to socialize with officials and engage in favor seeking behaviors they are least like to get favors granted. In fact engaging in the previous activities is actually a handicap. Farmers who hold the values subsumed in the Embeddedness index are 2% less likely to socialize with those officers with, 1% less likely to also seek favors. It is therefore not surprising that they are 6% less likely to get favors. Once again Fairness is negatively, but strongly correlated with not socializing with of ficers showed by 4% not seeking favors, and 5% not getting favors by 3%. Similarly the cluster of expectations farmers have about extension officers measured by the Attitudes towards Officials index show the odds of Frequency of Contact increases by 10% fo r a unit increase, 11% for Sought Favors and 2% for Got Favors . Not surprisingly the odds of farmers contacting officials, seeking favors and whether or not they get favors increased by 28%, 28% and 4% respectively with each a unit increase in requests. A unit change in odds for the control variables have the corresponding effects of: Age 3%, 2% and 19%; Ghana 73% 60%, but the odds of Ghanaian farmers actually getting favors for all their

PAGE 148

148 hobnobbing is 24%. The odds for male farmers are: 69%, 79% and 30%; for a supporter of any of the opposition parties: .43, .44 and .18, for rural farmers: 11%, 22% and 14%. The model estimation is slightly less strong with a pseudo R square of .10 and a log likelihood of 740.96, but it is still indicative of a model with substantive interesting relationships. The third and final model estimating the effects of affective reciprocity on actually getting favors shows a different patter n. The model estimation is absolutely the strongest with a pseudo R square of .26 which explains a quarter of the variance, crude as the indicators are, along with a log likelihood of s help interpret the results. 6.5 Conclusion The theory shared by both farmers and extension officers that corruption and favoritism through personalized relations are aspects of African culture that hold farmers back and hinder the efforts of extension o fficers have not be been borne out. But the message of the regression analyses is more nuanced. On one level farmers who are deeply embedded in affective reciprocal relationships and interactions seemed to have little effect on resource diversion. This is the level captured by the regressions. The other dimension of the relationship is better explained by the institutional context. Men, rich or poor, are expected to support themselves, immediate family members, and the extended family if they can afford to, and if married, they are also expected to extend help, when needed, to relatives of spouses. In close knit rural society problems are shared; this means the effect of being embedded in community is a diffuse, multi phenomenal problem not captured in the r egressions. Secondly, capturing Resource Diversion is complicated as this example shows. The process by which affective reciprocity bleeds the state of its limited resources is gradual. The

PAGE 149

149 effects of this weakening capacity is initially felt by the poore st of the poor who need state services the most; they are also the very last group to experience increasing state abilities. These citizens, in turn, respond to the empirical realities of being left increasingly to their own devices by turning to relatives and friends. While looking, waiting and hoping for help the poor urban and rural as well as rural citizens create alternatives through incremental adjustments to their situation. Over time effective or efficient alternatives brought on by expediency cal cify, creating new paths of, and towards, dependencies, challenges and opportunities with a dynamism and this feedback loops of their own that become, for the most part new suboptimal outcome for both citizens and the state. For example, seasonal skirmishe s over stray cattle, water issues, and cattle rustling between rural Ghanaians in the northern reaches of the country and nomadic Fulani headsmen usually worsen during years of drought. Increasingly these clashes are becoming more frequent and deadlier eac h year, and in some smaller, more remote hamlets they are beginning to destabilize communities. As the northern regions dries up earlier and longer than in season past due to changes in climate we are told, Fulani herdsmen have pushed further south, inev itably clashing with the relatively wealthier, more populated Brong Ahafo region. Conflicts that had hitherto gotten little attention from the national press were now well publicized. According to interviews with some of the effected farmers in both regio ns, Fulani headsmen who had previously traveled in small groups far away from their homes in Mali and Niger now travelled in bands for protection. Believing the government was unable or unwilling to protect them, villagers took matters into their own hands by beating up headsmen who in turn found protection in numbers. The inability of the Ghanaian, Nigerian, and Malian governments to protect the most vulnerable of their citizens from militias in Niger and Mali and to ensure the

PAGE 150

150 livelihoods of Northern and increasingly, the Brong Ahafo region, caused some in these rural communities to force both groups to seek protection and solutions outside of formal channels. now. Some of these guards were from the security forces and used government resources in ways that had not been sanctioned by the government. The reasons for their actions had very little to do with their normative beliefs. One of the central claims of this di ssertation is that economic pressures, rather than cultural values, drive much of the dysfunction in African bureaucracies. Governments unable to generate the levels of economic development needed to provide a certain minimum level of protection against un foreseen contingencies have effectively shifted the burden of providing for the aged, infirmed and chronically ill onto families in a country where unemployment and poverty are extremely high due to, among other factors educational opportunities and basic welfare guarantees. African politicians, like their counterparts the world over, try to prevent discontent from growing by showering supporters with patronage which undermines agricultural productivity. This might mean that young men, once incorporated int o these networks, would have very little incentive to find another way of earning a living. As they grow older, with their connections and ability to deliver voters, they become local power houses in their own right, and as long as they can deliver voters, they would be courted by candidates, and even possibly by both parties. This in the long run undermines both the quality of democracy and the community.

PAGE 151

151 6 1 How often do you visit Ministry Officials_ Past Year N (%) p<.05 , Cramér's V= .13 6 2 Doing Offic ial Business Outside Sanctioned Channels_ Past Year N (%) Botswana Ghana Total Never 259 (50) 724 (57) 983 (55) Only once 73 (14) 240 (19) 313 (18) A few times 87 (17) 227 (18) 314 (18) Many times 98 (19) 80 (6) 178 (10) Total 517 (100) 1,271 (100) 1788 (100) p<.05 , Cramér's V = .19 6 3 Influence Peddling _Past Year N (%) Botswana Ghana Total Never 315 (61) 631 (50) 946 (53) Only once 70 (14) 203 (16) 273 (15) A few times 65 (13 ) 273 (21) 338 (19) Many times 66 (13) 163 (13) 229 (13) Total 516 (100) 1,270 (100) 1,786 (100) p<.05 6 4 Sought Favors_ Past Year N (%) Botswana Ghana Total Very Unhelpful 35 (8) 22 (2) 57 (3) Not Helpful 49 (11) 304 (24) 353 (21) Neither 49 (11) 247 (20) 296 (18) Helpful 170 (39) 250 (20) 420 (25) Very Helpful 127 (31) 441 (35) 568 (34) Total 430 (100) 1,270 (100) 1,786 (100) p<.05 6 5 Got Favors_ Past Year N (%) Favors Gotten Botswana Ghana Total No 133 (31) 573 (45) 706 (42) Yes 297 (69) 691 (55) 988 (58) Total 430 (100) 1264 (100) 1,694 (100) p<.05 6 6 Summary Statistics of the Dependent Variable Variable Min Max Frequency of Contact Visit Ministry 0 10 Sought Favors 0 20 Got Favors 0 10 Botswana Ghana Total Neve r 249 (51) 599 (49) 848 (49) Yearly 105 (22) 265 (21) 370 (22) 6months 49 (10) 87 (7) 136 (8) 3months 46 (10) 122 (10) 168 (10) Monthly 35 (7) 161 (13) 196 (11) Total 484 (100) 1234 (100) 1718 (100)

PAGE 152

152 6 7 Correlation Matrix of the Dependent Variables Frequency of Contact Favor Seeking Got Favors Botswana Frequency of Contact Spearman Correlation 1.00 Two sided p value N 522 Favor Seeking Spearman Correlation .81 1.00 Two sided p value .00*** N 430 430 Got Favors Spearman Correlation .67 .90 1.00 Two sided p value .00*** .00*** N 430 430 430 Ghana Frequency of Contact Spearman C orrelation 1 Two sided p value N 1277 Favor Seeking Spearman Correlation .80 1 Two sided p value .00*** N 1264 1264 Got Favors Spearman Correlation .65 .88 1 Two sided p value .00*** .00*** N 1264 1264 1264 *. Corre lation p<0.00 (2 tailed) 6 8 Summary Statistics Binary Dependent Variables Frequency of Contact Favor Seeking Got Favors Botswana Mean .74 .77 .54 Min .00 .00 .00 Max 1.00 1.00 1.00 SD .44 .42 .50 N 522 430 430 Ghana Mean .79 .8 0 .48 Min .00 .00 .00 Max 1.00 1.00 1.00 SD .41 .40 .50 N 1277 1264 1264 Total Mean .78 .79 .50 Min .00 .00 .00 Max 1.00 1.00 1.00 SD .42 .41 .50 N 1799 1694 1694

PAGE 153

153 6 Botsw ana Ghana Total Strongly Disagree 5 (1) 3 (0) 8 (0) Disagree 39 (8) 24 (2) 63 (4) Neither 20 (4) 8 (1) 28 (2) Agree 147 (28) 620 (49) 767 (43) Strongly Agree 306 (59) 618 (49) 924 (52) Total 517 (100) 1273 (100) 1790 (100) p<.05 Cramér's V = .36 6 Botswana Ghana Total Strongly Disagree 5 (1) 3 (0) 8 (0) Disagree 39 (8) 24 (2) 63(4) Neither 20 (4) 8 (0) 28 (2) Agree 147 (2 9) 620 (49) 767 (43) Strongly Agree 306 (59) 618 (49) 924 (52) Total 517 (100) 1273 (100) 1790 (100) p<.05 , Cramér's V = .24 6 11 Embeddedness 3 Botswana Ghana Total Strongly Disagree 4 (0) 2 (0) 6 (0) Disagr ee 5 (0) 12 (1) 17 (4) Neither 19 (4) 8 (1) 28 (2) Agree 202 (39) 599 (47) 801 (45) Strongly Agree 282 (55) 653 (51) 924 (52) Total 514 (100) 1274 (100) 1790 (100) p<.05, Cramér's V= .14 6 12 Embeddedness 4 _ No Respect for Officers N (%) Botswana Ghana Total Strongly Disagree 107 (21) 59 (5) 166 (9) Disagree 207 (40) 303 (24) 510 (28) Neither 35 (7) 37 (3) 72 (4) Agree 117 (23) 594 (47) 711 (40) Strongly Agree 51 (10) 281 (22) 332 (19) Total 517 (100) 1,274 (100) 1791 (100) p<.05 , Cramér's V= .36 6 Botswana Ghana Total Strongly Disagree 59 (12) 131 (10) 190 (11) Disagree 94 (19) 401 (32) 495 (28) Neither 18 (4) 21 (2) 39 (2) Ag ree 107 (21) 502 (40) 609 (34) Strongly Agree 229 (45) 217 (17) 446 (25) Total 507 (100) 1277 (100) 1, 779 (100) p<.05 , Cramér's V=.32

PAGE 154

154 6 14 Fairness 2_ Perceived as Wicked N (%) Botswana Ghana Total Strongly Di sagree 79 (16) 72 (6) 151 (9) Disagree 180 (36) 376 (30) 556 (31) Neither 46 (9) 17 (1) 63 (4) Agree 109 (22) 559 (44) 668 (38) Strongly Agree 80 (16) 248 (20) 328 (19) Total 494 (100) 1272 (100) 1766 (100) p<.05p= 6 15 Fairn ess 3_ Kabelo should not listen to his relatives N (%) Botswana Ghana Total Strongly Disagree 43 (9) 24 (2) 67 (4) Disagree 120 (24) 435 (34) 555 (31) Neither 55 (11) 60 (5) 115 (7) Agree 148 (30) 390 (31) 494 (28) S trongly Agree 133 (27) 361 (28) 494 (28) Total 499 (100) 1270 (100) 1769 (100) p<.05, 6 16 Fairness 4_ Mpho Act in a Professional Manner N (%) Botswana Ghana Total Strongly Disagree 10 (2) 44 (3 ) 54 (3) Disagree 35 (7) 196 (15) 231 (13) Neither 27 (5) 52 (4) 79 (4) Agree 162 (32) 668 (53) 830 (46) Strongly Agree 276 (54) 306 (24) 582 (33) Total 510 (1000 1266 (100) 1776 (100) p<.05, Cramér's V = .30 6 17 Fairness 5 _Nothing wrong w ith Kabelo helping Mpho N (%) Botswana Ghana Total Strongly Disagree 28 (6) 79 (6) 107 (6) Disagree 54 (11) 285 (23) 339 (19) Neither 57 (11) 83 (7) 140 (8) Agree 199 (40) 580 (46) 779 (44) Strongly Agree 164 (32) 233 (18) 397 (23) Total 502 (100) 1260 (100) 1762 (100) p<.05 , Cramér's V = .20 6 18 Attitudes Towards Officials 1_ Expect Immediate Attention N (%) Botswana Ghana Total Strongly Disagree 69 (13) 49 (4) 118 (7) Disagree 137 (27) 378 (30) 515 ( 29) Neither 87 (17) 119 (9) 206 (12) Agree 132 (26) 462 (36) 594 (33) Strongly Agree 89 (17) 36 (40) 350 (20) Total 514 (100) 261 (20) 1783 (100) p<.05,

PAGE 155

155 6 19 Attitudes Towards Officials 2_Assist Me Regardless N (%) Botswana G hana Total Strongly Disagree 44 (9) 60 (5) 104 (6) Disagree 149 (29) 305 (24) 454 (25) Neither 63 (12) 146 (11) 209 (12) Agree 112 (22) 523 (42) 650 (36) Strongly Agree 149 (29) 222 (17) 371 (21) Total 517 (100) 1271 (100) 1788 (100) p<.05, s V = .21 6 20 Requests for Assistance 1_ Costs of _Healthcare/Tuition N/(%) Country Most Common Requests Sample Total Request total Request % Mean Std. Dev. Vari ance Skew. Se (mean) Botswana Healthcare/Tuition 513 390 76 1.20 1.11 1.24 0.80 0.05 Weddings/Funerals 515 355 69 1.04 0.85 0.72 1.17 0.04 Ghana Healthcare/Tuition 1,271 1061 83 1.03 0.72 0.52 1.48 0.02 Weddings/Funerals 1,267 1120 88 1.27 0.94 0.89 1.48 0.03 To tal Healthcare/Tuition 1,784 1451 81 1.03 0.76 0.58 1.37 0.02 Weddings/Funerals 1,782 1475 83 1.25 1.00 0.99 1.20 0.02 6 21 Requests for Assistance 2_Cost of Weddings/Funerals N (%) Botswana Ghana Total Never 160 (31) 147 (12) 3007 (17) Few times a year 180 (35) 845 (67) 1025 (58) Monthly 108 (21) 129 (10) 237 (13) Every two weeks 42 (8) 74 (6) 116 (7) Weekly 25 (5) 72 (6) 97 (5) Total 515 (100) 1267 (100) 1782 (100) p<.05,

PAGE 156

156 6 22 Variables Counts Dimension s of the Main Variables Count Dependent Weakened Capacity 1. Frequency of Contact 0 10 2. Sought Favors 0 20 3. Got Favors 0 1 Independent Normative 1. Embeddedness 0 20 2. Fairness 0 20 Weberianes s 3. Attitudes Towards Officials 0 8 Material 4. Requests for Assistance 0 20 Controls Age 21 65 Country Dummy: Ghana_du 0 1 Sex: Male_du 0 1 Opposition Party Supporter 0 1 Estimate_ Net Wealth 0 1 Settleme nt Type_ Rural 0 1

PAGE 157

157 6 23 Linear Regression Analysis of Affective Reciprocity Linear Regression (OLS) (1) (2) (3) Frequency of Contact Sought Favor Got Favor Dimensions of the Independent Variables Normative Embeddedness .026 .106* .062* (.023) (.044) (.025) Fairness .070*** .119*** .054** (.015) (.030) (.017) Weberianess Attitudes towards officials .023 .003 .014 (.027) (.053) (.030) Material Request for Assistance .233*** .347*** .109*** (.017) (.033) (.020) Controls Age years .162*** .440*** .270*** (.044) (.089) (.051) Ghana Country du .413** .229 .135 (.133) (.269) (.152) Gender Male .546*** .928*** .401** (.117) (.229) (.128) Supports Opposition Party .501*** .935*** .365** (.112) (.226) (.128) Estimate of Farmer's Wealth .179*** .216* .017 (.054) (.107) (.062) Settlement Type rural .184 .480* .294* (.116) (.233) (.131) Constant 1.560*** 4.085*** 2.231*** (.391) (.770) (.433) N 1682 1591 1591 Adj. R sq .169 .122 .065 Robust standard errors in parentheses * p<.05, ** p<.01, *** p<.001

PAGE 158

158 6 24 Negative Binominal Analysis of Affective Reciprocity Model (4) Frequency of Con tact Model (5) Sought Favors Model (6) Got Favors b. RSE p>|z % b. RSE P>|z % b. RSE P>|z % Embeddedness .01 .01 .35 .8 .02 .01 .05 2 .04 .02 .01 4 Fairness .03 .01 .00 .5 .03 .01 .00 2 .03 .01 .00 3 Attitudes towards offi cials .02 .01 .09 2 .12 .01 .31 1 .00 .02 .97 .10 Request for Assist. .08 .01 .00 8 .07 .01 .00 8 .06 .01 .00 6 Age_ years .06 .01 .00 6 .09 .02 .00 9 .14 .03 .00 15 Ghana dummy .15 .05 .00 16 .06 .06 .37 6 .09 .09 .33 8 Gender Male .22 .05 .00 24 . 24 .06 .00 26 .28 .08 .00 32 Supports Opp. Party .19 .04 .00 18 .21 .05 .00 19 .18 .07 .01 17 Estimated Wealth .06 .02 .00 6 .03 .02 .15 3 .04 .04 .33 4 Settlement Type rural .07 .05 .15 7 .11 .06 .05 10 .17 .08 .04 15 Constant .51 .15 .00 1.29 .18 .00 .84 .28 .00 N 1682 1591 1591 Log pseudolikelihood 3481.57 4141.39 2861.45 AIC 3.61 5.22 4.15

PAGE 159

159 6 25 Logit Analysis of Affective Reciprocity Dimensions of the Independent Variables Model (7) Odds Ratio RSE Model (8) Odds Ratio RSE Model (9) Odds Ratio RSE Normative Embeddedness .015 .010 .061** (.027) (.030) (.023) Fairness .045** .052** .034* (.017) (.018) (.014) Weberianess Attitudes towards officials .096** .102** .015 (.031) (.032) (.027) Material Factors Request for Assistance .247** .247** .038 (.028) (.030) (.017) The Controls Age_ years .028 .015 .175*** (.055) (.056) (.045) Ghana_ Country dummy .546*** .471** .267* (.146) (.160) (.132) Gender_ Male_du .526*** .581** .260* (.138) (.145) (.116) Supports Opposition Party_ du .556*** .573*** .195 (.130) (.136) (.113) Estimate of Farmer's Wealth .191** .185* .125* (.069) (.072) (.053) Settlement Type_ rural du .134 .197 .145 (.139) (.147) (.123) Constant .856 .350 .891* (.460) (.514) (.389) N 1682 1591 1591 Pseudo R sq . 112 .108 .027

PAGE 160

160 6 26 Changes in Odds Dimensions of the Independent Variables Model (7) Frequency of Contact % Change in Odds 1 vs. 0 Model (8) Sought Favor % Change in Odds 1 vs. 0 Model (9) Got Favor % Change in Odds 1 vs. 0 No rmative Embeddedness 2 1 2 Fairness 4 5 4 Weberianess Attitudes towards officials 10 11 2 Material Factors Request for Assistance 28 11 4 The Controls Age_ years 19 2 3 Ghana_ Country du 24 60 73 Gender_ Male du 30 79 69 Supports Opposition Party_ du 18 44 43 Estimate of Farmer's Wealth 12 20 21 Settlement Type_ rural du 14 22 14 Constant N Pseudo R sq

PAGE 161

161 1. Frequency of Contact (i). Fraternizing with officers (ii). Transacting official business outside officially sanctioned channels (iii). Influence peddling Weak Capacity 2. Sought Favors 3. Got Favors Figure 6 1 Specifying the Dependent Variable Sought Favors Range 0 2 How helpful are the extension officers you know, or officers who work in your area when it comes to helping you, or someone you personally know with things that are not part of their official duties? For example an officer making calls to help you or someone you know find a quotation for corn or sorghum that you needed to buy or sell. For even trying to help you or someone you know buy fertilizer, or a sprayer at a cheaper price, or helped you get with a voucher for petrol or some other favor that has to do with agriculture, but which you did not qualify for, but the officer considered you. V. Unhelpful/Unhelpful = 0 V. Helpful/Helpful*2 Figure 6 2 Sought Favors 1. Frequency of Contact run s from 0 to 10 a). Frequent Interactions with Officers During Business Hours [0 4] 19 b). Visiting Officials at Home for Official Purposes [0 3] 20 3] 21 2. Soug ht Favors runs from 0 to 20 Weakened Capacity How helpful are the extension officers you have come into contact with when it comes to helping you with issues that are not part of their official duties? For example calling to find a quotation for you, or helping you buy sorghum, fertilizer, or a sprayer at a cheaper price, or helped you with a voucher. Coded: 0 = Very unhelpful, 1= Not helpful, 2 = Neither 3 = Helpful], 4 = Very Helpful Farmers who found officials either very helpful or he lpful in aiding them get help get favors for which they were not qualified for are assumed as not only attempting to, but also succeeded, in extracting favors from officials. These scores are doubled, or weighted twice to run from 0 to 20 3. Got Favors runs from 0 to 1 Farmers who succeeded in extracting favors are coded 1 all else coded 0 Figure 6 3 Three Alternate Specification of Weakened Capacity 19 Coded: 0 = Never, 1= Annually, 2 = Biannually, 3 = Quarterly 4 = Monthly 20 Coded: 0 = Never, 1 = Only once, 2 =A few times, 3 =Many times 21 Coded: 0 = Never, 1 = Only once, 2 = A few times, 3 = Many times

PAGE 162

162 Variable Range 0/1 Frequency of Contact 0= not contact 1= Contact Sought Favors_ tried to ex tract favors from official 0= Unhelpful 1= Helpful Got Favors_ succeeded in extracting favors 0= Failed to extract favors, 1 =Succeeded Figure 6 4 The Binary Indicators Three Possible Mechanisms 1. Normative prevailing ideas about affective norms a). Embeddedness_ those so blessed ought to help 0 Never 4 Strongly Agree b). Fairness _ context specific application of norms 0 Never 4 Strongly Agree 2. Weberianness c). Professionalism measures degree of impartiality and inter unit coordinat ion 0 Never 4 Strongly Agree 3. Material d). Requests for Assistance 0 Never 4 Strongly Agree Figure 6 5 The Independent Variables Normative Dimension of Affective Reciprocity Embeddedness runs 0 20 1. If you want prestige and a good name in o ur society you should help other 2. People who help relatives and friends when necessary are respected 3. You should help relatives and friends as needed 4. We respect people who contribute financially to the development of their towns 5. We do not re spect officials who stick to the law when dealing with the public Coded: Strongly Disagree [0] Strongly Agree [4] Figure 6 6 Embeddedness Fairness runs 0 20 Remind Mr. Kabelo society will view him as w icked if he does not help Mpho Mr. Kabelo should not listen to his relatives Mpho should formally appealing his transfer There is nothing wrong with Mr. Kabelo helping Mpho Coded: Strongly Disagree [0] Strongly Agree [4] Figure 6 7Fairness Weber ian vs. Affective Reciprocity Professionalism runs 0 8 to attend to me regardless of what I find them doing 2. I expect government officers to help me with whatever business I have even if it might not be directly related to their job Coded: Strongly Disagree [0] Strongly Agree [4] Figure 6 8 Attitudes towards Officials Material Dimension of Affective Reciprocity Requests for Assistance 0 20 1. Contributed financially towards healthcare, or school fees for a relative or friend 2. Contributed financially towards helping relative or friend find a job 3. Contributed financially towards religious organization/football club/mekgatho 4. Contributed financially towards helping a relative or friend navigate government bureaucracy 5. Contributed financially towards helping a relative or friend with wedding, funeral costs Coded: [0] Never [1] A few a year [2] Monthly [3], Every two weeks [4] Weekly Figure 6 9 Re quests for Assistance

PAGE 163

163 CHAPTER 7 EXTENSION OFFICERS AND AFFECTIVE RECIPROCITY 7. 1 Introduction As the last link in the implementation chain and the first in the feedback loop, extension officers exercise considerable discretion over the day to day impleme ntation of ministerial policy. At the local level, individual officers decide which communities to visit, when, and how frequently these visits will be made. They also decide which farmers to work more closely with and how much of their official working ho urs they allocate to the myriad communities under their charge. Extension officers tweak, abandon, or otherwise alter policies that higher level officials at headquarters assume have been are still in use, jettisoned or modified. Also, as members of the co mmunities that they serve and in which they live, officers mentor young adults, act as consultants on issues ranging from putting together business plans (a requirement for some ministry programs these days), to helping families with school choice, to navi gate government bureaucracies or to access services. They also start clubs and associations and serve as patrons in many more. In these ways officers perform more tasks than their formal roles or pay scale require. The above notwithstanding, extension offi cers have few, if any, serious opportunities to directly shape policy at the very top. Besides when they do such as during training workshops where their feedback is explicitly solicited, it was not always clear what their impact was. Bureaucratic capacity , primarily measured in quantifiable metrics such as goals met or outputs achieved due to some type of official intervention, exclude many of the intangible contributions civil servants make to rural society. The main task in this chapter is to test the hy pothesis that increased demands on officers resources are negatively correlated to enhanced administrative capacity. The chapter is divided into four broad sections. Sections 7.2 and 7.3 discuss the

PAGE 164

164 construction and scaling of the dependent, independent an d control variables. Then in section 7.4 the data analyses are provided followed by a discussion of the results and concluding remarks. Because the construction and scaling of the independent variables are the same as in Chapter 6 the discussion is kept in tentionally short to avoid repetition. 7 .2 Conceptualization and Measurement of the Dependent Variables between the four independent variables and enhanced adminis trative capacity and in so doing parse out the mechanism by which affective reciprocity erodes capacity. The two dependent variables, Resource Diversion and Use of Official Time, measure a cluster of behaviors deemed detrimental to enhanced capacity. 7.2.1 Dependent Variable 1: Resource Diversion The first dependent variable, Resource Diversion, which measures the deployment of the behaviors: (i) Misappropriati ng Resources , (ii) Trading Favors, and (iii) Funding Favors . The Misappropriating Resources indicator is created from responses to the question asking if officers had personally experienced or witnessed diversion of state resources as favors to relatives a nd friends. The indicator runs from [0] never, to [4] daily. 1 The second, Trading Favors, asked officers if they had personally experienced favor trading within their ranks and also has a [0] never to [4] daily. 2 The idea here is to avoid the social desi rability biases arising from attempts 1 During the last three months, how often have you personally seen your colleagues use ministry assets or resources like vehicles, fuel, photocopier, telephone, farm e quipment or implements for unofficial use like helping their relatives and friends? Coded: [0] Never, [1] A few times a year, [2] Monthly [3] Weekly [4] Daily 2 Over the course of the last three months, how often have you experienced or witnessed officers asking each other for favors like buying animals/produce from the village, and then arranging for a driver or vehicle to send it to the city for them? Or, officers stationed in the rural areas calling their colleagues in the city to check on documents or

PAGE 165

165 at possible political correctness on the part of officers. 3 It really does not matter who officers helped as long as the time and resources employed in fulfillment of these affective obligations belong to the ministry and its affiliates. The third and last indicator, Funding Favors, measures how favors are usually funded and also has a never zero to monthly four range . 4 Civil service salaries are generally low and in countries like Ghana, circa 1992, and currently in Nigeria for short but regular intervals, payments are sometimes still irregular. In Zimbabwe and the Democratic Republic of Congo the government pays salaries if and when s create a degree of financial distress that makes it highly probable that officers may need to find other ways of supplementing their incomes. Depending on the country, the level of institutional decay, elite capture, or the remoteness of the outlying age ncies from the capital, officials may, for example, have to purchase many of the basic supplies needed for the daily discharge of their routine duties. These investments are often discussed as investments in return for a portion of the fees charged to the public. In this way, civil service posts become prebends from which revenue can s ome other issue for them. Or any other kind of favor in general that has to do with their positions as offices? Coded: [0] Never, [1] A few times a year, [2] Monthly 3 Given the subject of the research and the tendency of survey respondents to give an answ er they deem to be more socially acceptable than may be the case, I triangulated questions, sources and approaches. See for example Alexander and Becker, 1978; Gerber 1996; Goldenberg1996; 2002; Gower and Nargundkar, 1991; Martin and Polivka, 1995; Gower a nd Nargundkar, 1991; Schober and Conrad 1997; Stettler 2000; 2001; Weber, Lavine, Huddy and Federico, 2014. Huddy and Feldman 2009; Mendelberg 2001; Sears and Henry 2005; Bassili 2003 Berinsky 1999; Gilens, Sniderman, and Kuklinski 1998; Kuklinski et al. 1 997. The survey literature also shows the race of an interviewer has an effect respondents answers (Anderson, Silver, and Abramson, 1988a, 1988b; Davis 1997a, 1997b; Finkel, Guterbock, and Borg 1991; Hyman 1954; Kinder and Sanders 1996; Lowery, Hardin, and Sinclair2001; Sinclair et al. 2005; Blanchard et al. 1994; Crandall and Eshelman 2003; Crandall, Eshelman, and Forbes 1997). Or responses are shaped by respondents self monitoring to conform with perceived situation. (Snyder 1974, 1979; Snyder and Gangestad 1986; Gangestad and Snyder 2000; Snyder 1974; Snyder and Gangestad 1986). 4 How often have you used your own money and resources to do a fa vor for a colleague in during the past 3 months? Coded: [0] Never, [1] A few times a year, [2] Monthly

PAGE 166

166 be extracted (Blundo 2004, 2006, 2009; Bryant, et al. 2007, Pierson and Skocpol 2007). The distributions for the three indicators used in the construction of Misappropriating R esources, the first dependent variable are reported in s 7.1 through 7.3. 7.1 shows that in the three months leading up to the study, 80% of the officers in the Botswana sample reported never personally witnessing colleagues use ministry resourc es such as vehicles, fuel, a photocopier, telephone, or farm equipment and implements to help relatives and friends. But of those who had, a single officer reported witnessing this a few times, another officer reported this to be a monthly occurrence. Two officers; however, noted favors like these were a weekly occurrence ( M = .61, ± 1.34). In contrast, significantly fewer Ghanaian (5%) reported never witnessing said behaviors, 11% witnessed these a few times, 12% monthly, 32% weekly, and 41% daily ( M = 2.94, ± 1.67). By wide margins the Ghanaian officers were significantly more likely to report this type of behavior: 2 (4) = 104.12, a two sample t test comparison of unequal means also confirms the pattern, t ( 10.03) = 59.89, p =.000. However, asked how often they had experienced or witnessed favor trading within their ranks, slightly more officers from the Botswana samp le report witnessing favor trading within their ranks compared to their Ghanaian colleagues ( 7.2). Forty three percent of the Tswana officials reported never witnessing this kind of behavior, 36% a few times a year, 10% monthly, and 12% had seen it w eekly ( M = .90, ± 1.01) . Correspondingly, 59% of the Ghanaians had never witnessed this behavior, 28% a few times a year, 7% monthly, 3% weekly, and another 3% witnessed this daily ( M = .63, ± .46). Again, the cross samples differences were slight : 2 (4) = 8.65, t (1.58 ) = 65.77, p =.12. Responses to the question are presented in 7.3.

PAGE 167

167 Fewer officers in Botswana (55%) than in Ghana (7%) report never using their own resources to fund favors. But of those in Botswana who do, 29% do so a few times a year, 5% monthly, 10% weekly, and 1% daily ( M = .76, ± .98). Compared to 52% of the Ghanaians who report spending their own resources on fun ding favors a few times a year, 26% monthly, 8% sample t test of unequal variance, t( 4.25) = 64.73, p= .000 also substantiate this finding. In sum, more Ghanaians use th eir own resources to fund favors (95%) and do so more frequently than officials in Botswana. A final step in the construction of the indicator involves aggregating the three variables, Misappropriating Resources , Favor Trading, and Funding Favors into the Resource Diversion index. The Resource Diversion index offers a way of triangulating the unofficial assistance farmers reported getting from ministry officials in Chapter 6. It also allows us to start filling out the contours of the reversed collective act ion dilemma. But, before moving on to discussing in the second dependent variable a quick note: not only are the Ghanaians in the sample more likely to use government resources for favors for relatives and friends, the frequency of these favors are also ve ry different from the patterns established in the Botswana sample. About a third of the officers in the Ghanaian sample report witnessing this type of behavior on a weekly basis, 41% on a daily basis. By wide margins the Ghanaians in the sample were also m ore likely to use their own funds and resources to finance or bring to bear these favors. We can glean from the above that the two environments not only differ on our three measures, we can also infer that the proliferation of favors in the Ghanaian case m ust present special challenges for differently situated actors.

PAGE 168

16 8 7.2.2 Dependent Variable 2: Use of Official Time The Use of Official Time variable measures the amount of time officers spend in pursuit of issues that have nothing to do with their official duties over the course of a typical working day. 5 The response categories run from never [0 minutes] to eight hours a day [8]. This is the most direct way of quantifying the amount of time extracted from the state while on the , when combined with the qualitative data, allows for approximating time use shown in 7.4. A one way analysis of variance found no statistically significant differences between the officers in their Use of Official Time ( F (1, 275) = .16, p = .69). Of ficers in Botswana spent an average of 1.96 hours on personal issues or fulfilling social obligations that had nothing to do with their official duties compared to their Ghanaian counterparts who spent 2.04 hours out of their eight hour working day in simi lar pursuits in addition to their one hour mandated break. Again there is very little across sample difference between the officials in how they use time: 2 ( 9) = 6.57, p =.68, Kendall's tau b value of .06 ± .05 as does an equality of variance test F Levene = .17 df (1, 275) p = .68. The next step in the construction of the dependent variables involves the creation of the dummy variables for the Logistic an alyses. To this end, all officers who reported (i) never diverting state resources towards particularistic ends, (ii) never participating in favor trading, and (iii) never using their own resources doing favors for colleagues or admitting to engaging in an y of these three behaviors a few times a year are coded zero [0]. These officials are assumed to have a less deleterious effect on administrative capacity compared to the officers in group one who have been reported to actively or more frequently engage in resource diversion, favor trading, or spending their own resources to fund favors at least on a monthly basis. The officers 5 How many hours of official time per day, do you think you spend on personal business or doing things for other people (social obligations) that have nothing to do with your official duties?

PAGE 169

169 who admitted engaging in any of these actions at least monthly are coded one [1]. The new indicator labeled Resource Diversion _du has a zero to one range. With our officers sorted into two groups, we can now turn our attention to constructing the dummy variable for the Use of Official Time indicator. Officers who reported never attending lock are coded zero, all others are coded one; the resulting Use of Official Time_du has a zero to one range. The summary statistics reported in 7.5 show that officers in Botswana average 2.26 points on the 8 point Resource Diversion index, spending on average 2 hours of the working day on personal issues that have nothing to do with their formal job roles. This is compared to the Ghanaians who averaged 4.93 points on Resource Diversion ends. Again, the differences between the two groups of officials are slight and more often than not they are not even statistically significant. We now turn our attention to the construction and scaling of the independent variables. The discussion of the c onstruction and scaling of the independent variables is kept intentionally short to avoid repetition of material already covered at length in Chapter 6. 7.3 Measuring Affective Reciprocity: Specifying the Independent Variable As in Chapter 6, affective r eciprocity is measured on three dimensions. The normative aspect of affective reciprocity is an index of three indicators: (i) Embeddedness and (ii) Fairness. Societal perceptions about the civil service, or put another way, the anti Weberian bureaucratic aspect of the service is measured by the Professionalism indicator. These three indicators are them. The second dimension of affective reciprocity focuses on the economic aspect of affective reciprocal interactions and is measured by Requests for Assistance. The indicator is a composite

PAGE 170

170 indices are constructed and scaled is the subject of the next section. 7.3.1 Independent Variable 1: Embeddedness As in Chapter 6 the Embeddedness indicator is constructed from answers to four of ficers have about the social expectations their fellow citizens. 6 The fourth question, though views of officials. The response categories for each of the four qu estions runs from 0 to 4 with [0] coded as never, to [4] strongly agree. Together the four indicators offer a window into adulthood. The variable also allows us to te st one of the hypotheses that the more a person believes others share his or her view of affective reciprocity the greater the chances will be of acting on those beliefs. The results are reported in s 7 6 through 7 9. Generally, officers believed thei r fellow citizens would expect our hypothetical director to help relatives and friends if and when they could. The officers in the Botswana ( 7 6) sample 6 Please answer yes or no to the following questions. What you think society expects of a 50 year old director in the civil service? 1. Society expects him/her to help relatives and friends financially with medica l costs and school fees when possible. 2. Society expects him/her to contribute to the development of their hometown. 3. Society expects him/her to always help relatives and friends financially. Coded [0] No, [1] Yes. Please tell me if you strongly disag ree, disagree, neither agree nor disagree, agree, or strongly agree with the following statements: Officers should obey government rules and regulations whether they like them or not. Coded: [0] Strongly Disagree, [1] Disagree, [2] Neither [3] Agree, [4 ] Strongly Agree

PAGE 171

171 were evenly divided on the question of whether our hypothetical director should be helping relati ves and friends defray tuition and healthcare related costs, with 50% believing society did indeed expect the director to do so. A clear majority of the Ghanaians (59%); however, believe society expected the director to help relatives and friends defray he althcare and tuition related expenses if and when possible (Botswana M = .5, ± .51, Ghana M = .41, ± .49). Again, the cross samples differences are slight t (1.03) = 66.67, p =.31), 2 (1) = 1.09, p= .30, Cramér's V = .08. By the same token, the majority of officers in 6 7 also believed their fellow citizens expected the director, when necessary, to extend financial assistance to relatives and friends, with the officers from Bot swana are 9% more likely to hold this view than the Ghanaians (Botswana M = .71 ± 46, Ghana M =.62, ± 49; t (1.15)= 71.87, p =.26; 2 (1) = 1.23, p =.27, Cramér's V = .08). 7 8 presents results to the question of whether officials believed their fellow citizens expected them to help develop their places of origin. By wide margins, (88%) of officers from both countries think society, indeed, expected those who have succeeded to invest in their home communities (Botswana M = .88, ± .33; Ghana M = .88, ± 3 3) compared to 12% who did not think so t (.07) = 68.22, ns. There was also no association between officers and the notion that our director ought to do his bit in the development of his hometown 2 (1) = 1.09, p= .30, Cramér's V = .08. The survey data mirrors information gleaned from in depth interviews. Officers in both countries, especially at the district and municipal levels, were more likely to report they were expected, and believed they had an obligation, to contribute their fair share towards improving the socio economic conditions of their home communities. The last question used in the construction of the Embeddedness index asked officers to strongly disagree, disagree, neither agree nor disagree, agree, or strongly agree with the

PAGE 172

172 7 9. The majority, 62% and 68% of officers from Botswana and Ghana respecti vely, support the statement that regardless of their personal feelings, officers were expected to follow government rules and regulations (Botswana M = 2.57, ± 1.52, Ghana M = 2.73, ± 1.50). Slightly more Ghanaians (6%) subscribe to this argument t ( 0.61) = 179, p=.54) , although twice as many of the officers in the Botswana sample had no opinion one way or other compared to 7% of the Ghanaians, 2 (4) = 3.06, p Once again the two groups of officials share similar views about what they perceived is expected of, and from, a successful civil servant at the pinnacle of his or her career. 7.3.2 Independent Variable 2: Fairness The hypothesized tension between formal and informal rules is tested with the Fairness index that is constructed from answers 7 to a vignette 8 about an officer seeking to avoid transfer 7 The full wording of the vignette: Mpho is being transferred from the city where he lives with his wife, children and ailing mother. His children attend one of the best schools in town, and his mother has access to the best doctors. Be cause of these reasons Mpho does Kabelo. Mpho and Mr. Kabelo both pray at the same place. Now, please tell me if strongly disagree, disagree, you neither agree or disagree, or you agree or strongly agree with the following statements: Mr. Kabelo should not listen to these people; he should follow the rules and transfer Mpho Mpho should write a let ter to Mr. Kabelo in his official capacity as the director and explain his situation in a professional manner. There is nothing wrong with Mr. Kabelo helping his nephew Coded: [0] Strongly Disagree, [1] Disagree, [2] Neithe r [3] Agree, [4] Strongly Agree 8 The use of a vignette addresses concerns about social desirability by drawing the focus away from the respondent and to depersonalize the questions while still effectively capturing norms held by an individual (Alexander and Becker, 1978; Gerber 1996; Go ldenberg1996; 2002; Gower and Nargundkar, 1991; Martin and Polivka, 1995; Gower and Nargundkar, 1991; Schober and Conrad 1997; Stettler 2000; 2001)

PAGE 173

173 from his current post in the city to a village. The index has a zero/ four range ([0] strongly disagree, [4] strongly agree). While the officers in the study are, on average, more educated than the average farmer sampled, they also have a shared identity. Civil servants, as residents of the villages in which they work, and w ith friends and relatives to whom they have, as they must, essential ties, cannot easily shed the bonds, values and outlook that they share with the community. We do not expect officials to defect en masse from the formal cultural values of the service of which they have become a part, but because of their close association to their village; nonetheless, we expect them, as actors who also have a foot in the world of economy of As dilemma range from a forceful endorsement of sticking to existing hiring and firing, and tenure and promotion policies to a more muted response to the statement that there was nothing wrong with an uncle doing all he could to help his nephew. The views of the officers are reported in s 7 10 through 7 14. transferred, to stay at his current post in the city so he could continue providing for his ailing mother, the majority of the officers, 52% and 60% in Botswana and Ghana respectively, disagreed (Botswana M = 1.81, ± 1.38, Ghana M = 1.73, ± 1.35) (See 7.10). It is also noteworthy that slightly more than a third of the officers, adding together officers who either strongly disagreed or agreed (38% and 36% in Botswana and Ghana), disagreed with their colleagues. A two sample t test, assuming unequal variance in means, underscores the acro ss sample similarities between the two groups of officers ( t (.33) = 66.62), p =.74. Also there is no statistically significant association between officers who either support the application of

PAGE 174

174 nuanced interpretation of policy. And by country of origin, both Batswana and Ghanaian officers equally, prefer following existing policy as shown by the results of a Pearson Chi square test of association 2 (4) = 2.15, p = .71, Fisher's exact .67. The Ghanaian officers; however, are even more emphatic with (70%) rejecting the idea that Mr. Kabelo, the director in question, should be pressurized informally into helping Mpho stay in the city (Botswana M = 1.82, ± 1.27, Ghana M= 1.39, ± 1.19). The responses from the Botswana sample, on the other hand, though generally favoring antagonistic position compared to 7% from Ghana. Once again, close to a third (28%) of the officers in the Botswana sample, favor the application of informal pressure on Mr. Kabelo if only to remind him that should anything happen to his sister, whom he would otherwise have to support, it would be he, no t his nephew who would come off looking very bad. Even worse, he the Ghanaians supported this course of action. That said, the number of officers who would rathe hand. Again, a comparison of means suggests no meaningful across sample differences: t( 1.98) = 58.18, p =.06. A chi square test of association also found no meaningful a ssociation between officers who support or do not support this course of action and the effects of/or on their country 2 (4) = 8.22, p =.08, p = .08; Fisher's exact .08. While officers generally agreed with the idea that the director, Mr. Kabelo, follow exi sting civil service tenure and promotion policies, officers in the Botswana sample were more muted in their support with 46% in favor of this course of action and 45% against.

PAGE 175

175 In contrast and by a wide margin, the Ghanaians are even more supportive of fol lowing existing policy with 65% supporting this course of action (Botswana M = 2.05, ±1.21, Ghana M = 1.55, ±1.24). The results of the various tests support this finding: t( 2.31) = 69.75, p =.02; 2 (3) = 6.55, p Many of the officers, 64% in Botswana and 75% in Ghana, agree with the suggestion that Mpho handle his transfer in a professional way by formally petitioning the decision to transfer him rather than rallying rel atives to pressure his uncle. The officers were also similar in their M = 1.5, ±1.25, Ghana M = 1.23, ±1.18; 2 are similar to those of the t test (t(1.21) = 65.07, p= .23. helping his nephew stay in the capital given his particular circumstance, the resulting answers reported in 7.13 show 54% of the officers in Botswana thought not, as did 48% of the officers in Ghana. Once again the within sample differences are small; for example, 38% of the officers in Botswana disagree with the statement while 40% agreed. The breakdown is similar in the Ghanaian case with 38% disagreeing and 34% agreeing (Botswana M = 2.31, ±1.14, Ghana M = 2.01, ±1.28). Overall there are no significant differences between the officers on the issue: x 2 (4) = 76.16, p = .14). In sum, the majority of officers disagree with the idea that (i) Mr. Kabelo bend the rules (iii) and are also strongly supportive of the idea t hat Mpho plead his case through the available official channels. But paradoxically, they are just supportive of the notion that there was nothing

PAGE 176

176 wrong with Mr. Kabelo helping Mpho and, at the same time, advise Kabelo to go through with the transfer. By an d large we find no evidence of officials jettisoning formal rules in favor of informal norms. Rather the officers, at least at this abstract level, argue in favor of following ministerial policy. 7.3.3 Independent Variable 3: Weberianess One of the critiq ues leveled at the ministries in Botswana and Ghana was that both were ineffective and inefficient mostly due, among other things, to inadequate coordination between and across units within the ministries. This gave officials too much autonomy and discreti on on certain issues but constrained them on others. The indicator, Professionalism, is constructed from ies fairly and in a consistent manner. Answer categories run from strongly disagree [0] to strongly agree [4]. The next question asked officers about the level of coordination between and across units in the Ministry. The indicator ranges from [0] strongly disagree to [4] strongly agree. The assumption here is that by focusing on the actual behavior we can, hopefully, minimize the potential social desirability problems. The results are reported in 7.15 to 7.16. The officers differed significantly on t he question of the professionalism of officers. The majority, 72% of the officers from Botswana, rejected the idea that officers are, on average, more likely to favor those they know in the day to day discharge of their official duties (Botswana M = 1.28, ± .85). Comparatively only to 42% of the officers from the Ghanaian sample felt this way ( Ghana M = 1.99, ± 1.45). The across sample variance is statistically significant: 2 (4) = 50.25, p sample, assuming unequal variances, t ( 5.12) = 249.86, p =.000. But mostly driven by the large number of officers in the Botswana sample who disagreed and adding together officers who chose not to express an opinion one way or another

PAGE 177

177 with those who agree, we have close a to a third of the sample thinking otherwise which hints at a possible polarized view of the issue. More than half the officers also rejected the idea that intra departmental/inter departmental and ministerial coordination was inadequate (Botswana M = 1.09, ± 1.02, Ghana M = 1.03, ± 1.03 ). The across sample differences are not statistically significant t (.44) = 295, p = .66), and the chi square test corroborates those of the ttest x 2 . Once again, the hypothesized relationship between the two values driven indicators of affective reciprocity and the Fairness index have so far not been borne out. Faced with moral quagmires, measured by the Embedded ness and Fairness indicators, the officers in the sample also show loyalty to the ministry and seem to take pride in defending themselves and their organization. What are we to make of this? With this question in mind, we turn our attention to the economic dimension of affective reciprocity. 7.3.4 Independent Variable 4: Requests for Assistance With states unable to provide basic services for many, relatives and friends are often left to supplement the needs of poorer relatives. The myriad demands on peopl adjudicated along an intuitive basic needs scale depending on who is being asked to assist and perceived as important and is; therefore, a legitima te reason for sacrificing, or adjusting for in commitments, any additional demands made on their resources creates additional financial stress and increases the pos sibility that an officer might be forced to look for other sources of supplementing his or her salary which increases the likelihood of the officer engaging in actions that could possibly undermine state capacity.

PAGE 178

178 The variable, Requests for Assistance 9 , i s constructed from responses to five questions index has a zero to five range [0, never 5 weekly], and because the construction and scaling of the economic dimensions of affective reciprocity have been discussed in Chapter 6, they need not be repeated here. The results are reported in s 7.17 through 19. See appendix for comprehensive . The officers differ significantly in frequency of financial contributions towards defraying the health care or tuition related expenses of relatives and friends ( t (5.69) = 51.42, p= .000). For example, 14% of officers in Botswana reported making monthly contributions, 21% biweekly, and 48% contributed weekly. This is in comparis on to their Ghanaian counterparts, the majority of whom (53%) make these contributions a few times a year, 31% monthly, 5% biweekly, and 6% make them weekly. Officers in Botswana made more weekly contributions (M = 3.86, ± 1.42) than their Ghanaian counter parts (Ghana M = 2.53, ± .90). A chi square test lends additional support to the finding that, generally, officers in Botswana contribute more towards helping relatives and friends defray healthcare and education related costs 2 (6) = 72.54, p= .000, Fisher 9 How often have you been asked to contribute financially towards: (i). Defraying medical and school relate d expenses (ii) Defraying job seeking related expenses (fare to travel back and forth and also once they get to where they are following up on the job; accommodation, living expenses, printing, money to buy and pay for forms, taking pictures, and use for tips) (iii). To religious groups/church/mosque (iv). Towards defraying costs associated with navigating government bureaucracy (fare to travel back and forth, and also once they get to where they are following up on the job; accommodation, living expense s, printing, money to buy and pay for forms, taking pictures, and use for tips) (v). Towards a wedding, funeral, naming ceremony

PAGE 179

179 One way of ensuring financial independence and easing and spreading the burdens of reciprocity for the rest of those in an affective network is by helping other family members find jobs by providing startup financing, or by raising money to fund those perceived as sui and worthy of the costs associated with migrating. Money raised this way is often spent on fares for In addition to covering living expenses during the job search, the funds are also used implicitly for brides and for tokens of appreciation. There are significant across sample differences between officials in their contributions towards funding job searches, though the Ghanaian officers contributed more. In frequency, 65% of officers in Ghana reported doing this a few times a year compared to 49% of officers in Botswana (Botswana M = 2.43, ±1.02, Ghana M = 1.85, ± .72), t e test were similar (4) = 32.78, Once again the across sample differences are slight t ( 2.80)= 169, p = .005 and square test (4) = 36.58, 7.19). The majority of officers in Botswana (61%) report not making financial contributions compared to 47% of their Ghanaian counterparts who do so a few times a year (Botswana M = 1.89, ±1.28, Ghana M= 2.50, ±1.14). A two sample t test, assuming unequal means, conducted to compare the officers fr om these two countries on their contributions towards religious bodies, showed no significant differences between officers t ( 2.70) = 106.82, p= .005 (Botswana M= 1.76, ± .88, Ghana M= 2.25, ±1.34) as did a chi square test 2 Along with helping a friend or relative look for a job, it is not uncommon for family and community members to make financial contributions towards the attainment of certain official documents like land registration, building permits, or for claiming benefits of a deceased relative

PAGE 180

180 (Botswana M= 1.92, ± 1.06, Ghana M =1.49, ±.92). A two sample t test with unequal variances conducted to compare the means of officers from both countries shows some cross samp le variations t (2.28)= 57.44, p = .02. The results of a chi square test also indicate the result was statistically significant, 2 Of the five indicators aggregated into the Request for Assistance, officers in Ghana reported contributing several times a year towards defraying (i) healthcare and education related costs, (ii) job see king related expenses, and (iii) weddings and funerals. The officers in the Botswana sample reported being asked to make financial contributions to help with (i) job seeking related expenses and to provide aid to (ii) religious bodies which made it harder for officers to exercise control over their finances. Besides giving us a sense of some of the financial strain officers might be under, it also gives us an idea of some of the more common problems individuals and officers grapple with. The also unde contributions towards defraying healthcare costs and/or school fees is a source of strain on families. As a final step we need to establish whether or not officers actually ac t on the demands made on them. 10 We do this by asking how often officers remitted money to relatives. 7.3.5 The Control Variables As in Chapter 6, we control for age: measured in years (from 24 to 65+), and gender: female = 0, male = 1. We likewise treat 10 Q60. How often, if at all, do you send or give money to help your relatives? Coded: [0] Never, [1] About once a year, [2] A bout every 6 months, [3] About every 3 months, [4] Monthly, [5] Weekly.

PAGE 181

181 three categories from junior officer = 1 to senior officers = 3 as a plausible factor in determining ability to potentially counter or opt out of the reversed collective action dilemma. High level officials are more likely to have more access to resources and; therefore, are more likely to an are lower level officials. The descriptive statistics of the independent variables and the controls are reported in 7.23 Our average officer is male, 43 years old and is a middle level bureaucrat with at least five years of post secondary educatio n. Our officer shares three of the seven factors society deems essential elements of social respectability, prestige, and success; these indicators are used to construct our Embeddedness scale. The average officer is; however, on the lower end on the Fairn ess scale. Five out of 20 times this officer finds himself having to choose between deploying affective reciprocity or following the formal rules and procedures of his job and profession. We also find the average officer is under considerable pressure to h elp relatives, friends, colleagues and community members and on average, is receiving three out of every five demands; the same goes for the 174 officers who reported sending money home to relatives and friends. With this in mind, we will move straight to interpreting the results. 7.4 The Data and Discussion of the Results Because many of the issues related to the analysis have been discussed in previous chapters and need not be repeated, I will present the various models without further discussion. The ana lysis and discussion of the results is organized as follows: the results for the linear regressions are presented first; these are followed by the results from the NBRM models. The findings from the logistic regressions are presented in the third section. The chapter concludes with a summary of the findings. In general, I expect officers who receive lots of requests for assistance to be positively correlated to Resource Diversion and Use of Official Time .

PAGE 182

182 7.4.1 Linear Regression Analyses of Affective Recip rocity The results of the two linear regressions analysis conducted to determine the effects of affective reciprocity on administrative capacity are presented in 7. 24. Two of the four specifications of the independent variables are statistically si gnificant to the dependent variable. Embeddedness index are negatively, but not significantly, correlated to Resource Diversion or Use of Official Time . The Fairness index is; however, negatively correlated to Resource Diversion but not to Use of Official Time. This means that officials who subscribe to the idea that norms should be context specific are paradoxically the very officers who end up attending to personal business while on th Professionalism, the index measuring impartiality and inter unit coordination, shows no relationship with either to the two specifications of the dependent variable. However, as expected, Requests for Assistance is substanti ally and statistically significant in Model 1, Resource Diversion, though it never approaches significance with Use of Official Time. This is not surprising since participant observation time use data differed markedly from ey question. Officers in both countries spent an average of 3 hours per day on personal matters at the headquarters level that had nothing to do with their official duties and 4 hours at the district and sub district levels. In Ghana whole days were spent either sitting around chatting, or officers would report for work, leave their jackets and/or other when upper managers were away. 11 In Botswana officials showe d up for work very early, but 11 Finding from pre dissertation research showed a similar pattern: data collected in 2008 from four ministries showed officials spent between 3 hours 15mins (the most) to1hour 15minutes of their official work time on personal matters that had nothing to do with their official duties.

PAGE 183

183 then spent between 15 minutes to half an hour on breakfast, and depending on what was to be done, they sat around and chatted while a few people did the work. As in Ghana there were whole units that had nothing to do precisely because monitoring was weak at all levels. Hastily convened meetings, inadequate planning, and a lack of a clear sense of what had to be done, when, and by whom were easily identified by officers as the reason for this state of affairs. In answering why t his was the case, the most common reason given by officers in both countries, , girl 12 The issue was not solved by detailed work plans either, since in practice they were usually not followed in Botswana. Model 1 shows Embeddedness , the index measuring the idea that those so blessed ought to help, and failing to do so deserves our scorn, has no effect on either dependent variable. A unit increase along the Fairness scale, the indicator measuring the idea that each person can and should expect help from others within reasonable limits, however is positively correlated to Use of Official Time but not Resource Diversion . A unit increase in Fairness reduced Use of Official Time by .151 units holding all else equal. As expected , the Professionalism index measuring level of coordination between and across units within the ministries, has no effect on both dependent variables. Requests for Assistance is positively correlated with Resource Diversion though not Use of Official Time. with a .278 unit increase in diverting resources from the state. 12 Focus group discussions, Alisa hotel, Kumasi Ghana 17 May 2011.

PAGE 184

184 of education and the Ghana dummy are significantly correlated with Resource Diversion. The more educated an officer, the less likely they are to engage in this behavior, with each additional year in school reducing reso urce diversion by .544 units. Living in Ghana makes it 2.649 units more likely to divert state resources. Again, this is not surprising given the fact that: first, men outnumber women 4:1 (172 men and 41 women), so while the sample is representative, agri cultural extension is still mostly a male dominated department in Ghana, at least during fieldwork. Second, as the clientelism literature shows, men are more likely to engage in the set of behaviors measured by the index. Ghanaian officers had, on average , four children, and Tswana officers had, on average, two children. More than half of the Ghanaian offices are 45 years and older compared to 37% of the officers in the Botswana sample. Of these 92% and 61% report being primarily responsible, financially, for their parents and other relatives. In one on one interviews 65% of the officers report having at least one parent plus a younger relative living with them. And while 65% of officers in Botswana reported making enough to support their families, only 20% of the Ghanaian officers reported making enough; not surprisingly 74% of the Ghanaian officers report having another income generating job or sideline. In Botswana the most common income generating activity for women was rental income. The second most com mon means of generating extra income, primarily by male officers cattle ownership. In Ghana the one number income generating activity for the female officers was petty trade which tied with farming and raising poultry/small stock. The second was food relat ed, common activity was producing tie/die fabric or honey, or baby food made from local nutritious grains. The responses from male officers about extra incom e opportunities were more

PAGE 185

185 One on one interviews revealed projects to be a catchall phrase for hassling. Older and younger officers were also more likely to be abse ntee farmers. With this in mind we now turn to the Negative Binominal analysis. 7.4.2 Negative Binominal Analysis. of Affective Reciprocity None of the independent variables show any correlation with the Use of Official Time indicator in 7.2 2 . Addit ionally the two normative dimensions of affective reciprocity, Embeddedness and Fairness , are not correlated to Model 3, Resource Diversion . However officers who reported they and their colleagues were professional and agreed with the statement that inter unit/departmental coordination could not be better were significantly and positively correlated with diverting state resources. This means all things equal, a unit increase on the eight point Professionalism scale translates into 56% increase in count of r esource diversion for the officers in Botswana and a 77% increase for the Ghanaian officers. As expected officers under financial stress are once again significantly correlated with diverting resources with each additional unit change causing a correspond ing 48% in expected count in resource diversion for officers in Botswana and 66% increase for the Ghanaians. Also, interactions, participation drops by 15% for each additional year of education, but increases by 71% for Ghanaian officers. We can now turn our attention to the two dependent variables, transformed earlier into discrete dichotomous variables, to estimate and evaluate the effects of affective reci procity on enhancing bureaucratic capacity.

PAGE 186

186 7.4.3 Results of Logistic Regressions of Affective Reciprocity A variance inflation factors (VIF) multicollinearity test revealed no approximate linear relationship between the independent variables. The mean VI F for all three models was 1.60. Perhaps, not surprisingly at this point, none of the two normative measures of affective reciprocity, Embeddedness and Fairness, were correlated in any way to Resource Diversion (model 3) or to Use of Official Time (model 4 ) which are not included because the model showed no correlation whatsoever between and across variables. Request for Assistance an d Professionalism are both positively, though moderately, correlated with resource diversion behaviors at the .05 significanc e level. Meanwhile the Ghanaian officers are robustly and positively correlated with resource diversion. Level of education still has strong dampening effect on misappropriating state resources. Officers ranking high on the Professionalism scale are on ave rage 2.5% more likely to undermine capacity through resource diversion, those under financial pressure measured by Request for Assistance 6%, while additional year of education reduced participation by 5% and being Ghanaian by 70%. 7.5 Conclusion As in C hapter 6 and as expected, the Ghanaian officers are more likely to engage in the behaviors that are posited as corrosive to enhanced capacity. The models suggest economic pressures, rather than some inherent normative commitment to family, relatives or one affective group, drive these interactions and also seem to drive affective reciprocity. However, the distributions across the individual indicators show about a third of all the officers consistently favor informality over formal civil service processes . For example, a third of the officers from both countries disagreed with their colleagues about questions that ask things such as what society would expect from a 50 year old civil servant who also happens to be a director. That

PAGE 187

187 most people know full well how much civil servants earn, suggests attitudes like these are measured by Request for Assistance performed consistently as expected. Not surprisingly Professionalis m is negatively correlated to Resource Diversion , even if moderately. The results of the regression analyses are supported by the qualitative interviews. Close to half the officers from both countries agreed with the statement that interdepartmental coordi nation was inadequate at best and left a lot to be desired. In depth interviews with older officers revealed their expressions of concern about current hiring practices. While entry into the civil service was done through a formal centralized hiring proces s in Botswana, in Ghana years they missed out on the experience of going through a formal hiring process the civil service exam, interviews, orientation and in tra and inter department rotations instead by hiring through back doors, the civil service has become a place for political appointees to while away time Accordi ng to my data, very few frontline officers, have such access. Rather, multiple in depth interviews with both departments of Human Resources, retired directors and other senior officials showed hiring based on networks and personal relationships not only he lps those in administrations simplify their search, as it often does outside Africa, but it allows those in positions of power to fulfill social obligations and to demand greater loyalty. Additional anecdotal evidence from two units at the Ministry of Fina nce and two metropolitan divisions in Ghana, and the Chobe Land Board seems to suggest, at least, at a certain level, that tapping talent through networks and personal relationships has had a positive effect.

PAGE 188

188 Officers in Botswana; however, were more concer ned with inadequate levels of manpower and dwindling resources. Compared to their counterparts in Ghana, officers in Botswana had computers, access to the internet even if slow, a fairly well stocked on site library with a well trained staff support staff who showed up for work and actually worked. These things along with frequent workshops and seminars all well catered would no doubt surprise their Ghanaian counterparts whose computers, when they worked, had to be shared. But this fretting of Tswana o fficers about the good old days were eye opening. Thus, while the weak morale in Ghana was due mostly to institutional factors like inadequate wages and supervision, and not having enough to do, officers in Botswana felt the system they knew and toiled har d for was changing and not for the better due mostly to new belt tightening measures imposed by an over dependence on a diamond economy. Officers in both countries generally agreed that some of the reasons why extension officers were not as productive as they could be were related to a lack of supporting institutions, among them subsidized tuition for their children (officials used to have these benefits), health reimbursement for cellphone expenses, and, surprisingly, counseling services to help them deal o take up farming as a 13 . Officers were surprisingly candid about who the hard workers were in their units and why it was sometimes a bad idea to include upper level officers on project s because they had too many things going on; 13 Post surveys final interviews in: Botswana, January February 2010 2011. Ghana, November and December 2011.

PAGE 189

189 officer often comp lained after a long day of debating farmers. Ghanaian officers were particularly candid about the dynamics and problems associated with the move towards delivering services to groups of farmers. Many believed the success policy initiatives were largely dep endent on focusing on a smaller group of farmers and more follow up with farmers. Indeed, extension officers were at their best when discussing the micro dynamics of their catchment areas. At the very local level officers, had detailed knowledge of which f armers needed which kinds of interventions, and why. The general impression I got was a more ionships, relationships fostered and anchored in economies of affection. The very institution that also made it just a little harder for officials to be agents of change.

PAGE 190

190 7 1 Misappropriating Resources: Last 3 Months N (%) Botswana Ghana Total Never 33 (80) 6 (5) 39 (22) A few times 1 (2) 14 (11) 15 (9) Monthly 1 (2) 16 (12) 17 (10) Weekly 2 (5) 43 (32) 45 (26) Daily 54 (41) 58 (33) Total 41 (100) 133 (100) 174 (100) *p<.05, Fisher's exact = .000 7 2 Favoring Trading: Last 3 Months N (%) Botswana Ghana Total Never 18 (43) 79 (59) 97 (55) A few times 15 (36) 38 (28) 53 (30) Monthly 4 (10) 9 (7) 13 (7) Weekly 5 (12) 4 (3) 9 (5) Daily 0 4 (3) 4 (3) Total 42 (100) 134 (100) 176 (100) * p<.05, Fisher's exact = .08 7 3 Funding Favors: Last 3 Months N (%) Botswana Ghana Total Never 23 (55) 9 (7) 32 (19) A few times 12 (29) 67 (52) 70 (46) Monthly 2 (5) 34 (26) 36 (21) Weekly 4 (10) 10 (8) 14 (8) Daily 1 (2) 9 (7) 10 (6) Total 42(100) 129(100) 171 (100) * p<0.05, 7 4 Use of Official Time N (%) Hours Per Working Day Botswana Ghana Total Never 0 6 9 8 About an hour 1 79 77 78 About two hours 2 5 4 4 About three hours 3 3 2 2 About four plus 4 6 9 8 Total 8 100 100 100

PAGE 191

191 7 5 Summary Statistics of the Dependent Variable Country Variable Obs. Mean Std. Dev. Se Mean Min Max Var Kurt. Skew Botswana Resource Diversion 42 2.26 2.08 .32 0 8 4.34 2.81 .78 Ghana Reso urce Diversion 137 4.93 2.20 .19 0 11 4.83 2.91 .18 Total Resource Diversion 179 4.31 2.45 .18 0 11 5.98 2.59 .14 Botswana Resource Diversion _du 42 .40 .50 .08 0 0 .25 1.15 .39 Ghana Resource Diversion _du 137 .88 .33 .03 0 0 .11 6.20 2.28 Total Resource Diversion _du 179 .77 .42 .03 0 0 .18 2.57 1.25 Botswana Official Time 78 2.06 2.05 .23 0 8 4.22 5.46 1.76 Ghana Official Time 199 1.94 2.23 .16 0 8 4.97 5.46 1.90 Total Official Time 277 1.97 2.18 .13 0 8 4.75 5.46 1.86 Botswana Official Time_du 78 .46 .50 .06 0 0 .25 1.02 .15 Ghana Official Time_du 199 .37 .48 .03 0 0 .23 1.31 .55 Total Official Time_du 277 .39 .49 .03 0 0 .24 1.19 .44 7 6 Embeddedness 1 Help defray the cost of healthcare/tui tion N (%) Botswana Ghana Total No 21(50) 56 (41) 102 (57) Yes 21(50) 81 (59) 77 (43) Total 42 (100) 137 (100) 179 (100) p<0.05, Cramér's V = .08 7 7 Embeddedness 2 Helping relatives and friends financially N (%) Botswana Gha na Total No 12 (29) 52 (38) 64 (36) Yes 30 (71) 85 (62) 115 (64) Total 42 (100) 137 (100) 179 (100) p<0.05, Cramér's V = .08 7 8 Embeddedness 3 Contribute toward developing place of origin N (%) Botswana Ghana Tot al No 5 (12) 17 (12) 22(12) Yes 37 (88) 121 (88) 158 (87) Total 42 (100) 138 (100) 180 (100) p<0.05, Fisher's exact = 1.00 7 9 Embeddedness 4 Obeying formal rules N (%) p<0.05, Fisher's exact = .56 7 10 Fairness 1 Kabelo should underst Botswana Ghana Total Botswana Ghana Total Strongly Disagree 8 (19) 22 (16) 30 (17) Disagree 2 (5) 12(9) 14 (8) Neither 6 (14) 10(7) 16 (9) Agree 10 (24) 32(23) 42 (23) Strongly Agree 16 (38) 63 (45) 79 (44) Total 42 (100) 139 (100) 181 (100)

PAGE 192

192 Strongly Disagree 8 (19) 24 (18) 32 (18) Disagree 14 (33) 57 (42) 71 (40) Neither 4 (10) 6 (4) 10 (6) Agree 10 (24) 32 (23) 42 (23) Strongly Agree 6 (14) 18 (13) 24 (13) Total 42 (1 00) 137 (100) 179 (100) p<0.05 7 11 Fairness 2 Perceived as Wicked N (%) Botswana Ghana Total Strongly Disagree 3 (8) 28 (20) 31 (18) Disagree 19 (49) 69 (50) 88 (50) Neither 6 (15) 10 (7) 16 (9) Agree 4 (10) 19 (14) 23 (13) St rongly Agree 7 (18) 11(8) 18 (10) Total 39 (100) 137 (100) 176 (100) p<0.05 , Fisher's exact = .08 7 12 Fairness 3 Kabelo should follow policy N (%) Botswana Ghana Total Strongly Disagree 2 (5) 7 (5) 9 (5) Disagree 17 (40) 36 (26) 53 (29) N either 4 (9) 4 (3) 8 (4) Agree 15 (36) 63 (46) 78 (44) Strongly Agree 4 (10) 26 (19) 30 (17) Total 42 (100) 136 (100) 178 (100) p<0.05 exact = .09 7 13 Fairness 4 Mpho should follow the rules N (%) Botswana G hana Total Strongly Disagree 1 (2) 4 (3) 5 (3) Disagree 10(24) 23 (17)(3) 33 (19) Neither 4 (10) 7 (5) 11(6) Agree 17 (41) 62 (46) 79 (44) Strongly Agree 10 (24) 40 (29) 50 (28) Total 42 (100) 136 (100) 178 (100) p<0.05 , Fish exact = .52

PAGE 193

193 7 14 Fairness 5 Nothing wrong with helping N (%) Botswana Ghana Total Strongly Disagree 0 14 (10) 14 (8) Disagree 16 (38) 52 (38) 68 (38) Neither 3 (7) 6 (4) 9 (5) Agree 17 (40) 47 (35) 64 (36) Strongly Agree 6 (14) 17 (13) 23 (13) Total 42 (100) 136 (100) 178 (100) p<0.05 , Fisher's exact = .18 7 15 Weberianess 1 Preferential treatments N (%) Botswana Ghana Total Strongly Disagree 10 (12) 41 (21) 51 (19) Disagree 50 (60) 40 (21) 90 (33) Neither 14(17) 27 (14) 41(15) Agree 8 (10) 45(24) 53 (19) Strongly Agree 1 (1) 38 (20) 39 (14) Total 83(100) 191 (100) 274 (100) p<0.05 , 7 16 Weberianess 2 I Degree of Professionalism N (%) Botswana Ghana Total Strongly Disagree 35 (40) 92 (4 4) 127 (43) Disagree 17 (19) 33 (16) 50 (17) Agree 29 (33) 69 (33) 98 (33) Strongly Agree 7 (8) 15 (7) 22 (7) Total 88 (100) 209 (100) 297 (100) p<0.05 , Cramér's V = 0.05 7 17 Financial contributions past year _Healthcare/education N (%) Botswana Ghana Total Never 6 (14) 7 (5) 13 (7) Few times/yr. 1 (2) 73 (53) 74 (41) Monthly 6 (14) 42 (31) 48 (27) Biweekly 9 (21) 7 (5) 16 (9) Weekly 20 (48) 8 (6) 28 (16) Total 42 (100) 137 (100) 179 (100) p<0.05 , Fisher's exact = .000

PAGE 194

194 7 18 Financial contributions past year _ Job Search N (%) Botswana Ghana Total Never 7 (17) 36 (27) 43 (25) Few times/yr. 17 (49) 85 (65) 102 (59) Monthly 13 (31) 7 (5) 20 (120 Biweekly 3 (7) 3 (2) Weekly 2 (5) 3 (2) 5 (3) Total 42 (100) 131 (100) 173 (100) p<0.05 , Fisher's exact = .000 7 19 Financial contributions past year_Wedding/funerals N (%) Botswana Ghana Total Never 22 (61) 20 (15) 42 (25) Few times/yr. 3 (8) 64 (47) 67 (39) Monthly 6 (17) 27 (20) 33 (19) Biweekly 3( 8) 11 (8) 14 (8) Weekly 2 (6) 13 (10) 15(9) Total 36 (100) 135 (100) 171 (100) p<0.05 exact = .000 7 20 Financial Contributions past year _Religious Orgs N (%) Botswana Ghana Total Never 18 (43) 49 (38) 67 (39) Few times/yr. 19 (4 5) 38 (29) 57 933) Monthly 3 (7) 22 (17) 25(15) Biweekly 1 (2) 4 (3) 5 (3) Weekly 1 (2) 17 (13) 18 (10) Total 42 (100) 130 (100) 172 (100) p<0.05 , Fisher's exact = .69 7 21 Financial Contributions past year _ Navigating bureaucracies N (%) Botswana Ghana Total Never 18 (46) 84 (69) 102 (63) Few times/yr. 10 (26) 26 (21) 36 (22) Monthly 8 (21) 6 (5) 14 (9) Biweekly 2 (5) 2 (2) 4 (2) Weekly 1 (3) 4 (3) 5 (3) Total 39 (100) 122 (100) 161 (100) p<0.05 ,

PAGE 195

195 7 22 Descriptive Statistics Summary Statistics_ Key Independent and Control Variables Main Independent Variables Mean Min Max Sd N Prevailing Affective Norms 2.72 0 7 2.69 308 Embeddedness 4.78 0 20 4.85 308 Fairness 2.17 0 7 1.77 308 Request for Assist ance 3.33 0 15 3.68 308 Control Variables Age 43 26 60 9.76 288 Male_du .73 0 1 .45 306 Level of Education 5.15 2 7 1.21 294 Religion 1.35 1 3 .49 174 Grade/rank of officer 1.77 1 3 .78 262 7 23 Linear Regression Analysis of Affe ctive Reciprocity Dimensions of Affective Reciprocity (1) Resource Diversion (2) Use of Official Time Normative Embeddedness .052 .038 (.081) (.116) Fairness .020 .151* (.041) (.062) Weberianes s Professionalism .106 .026 (.0 91) (.124) Material Requests for Assistance .278*** .104 (.055) (.081) Controls Age (years) .021 .038 (.0134) (.024) Level of education (0, 6) .544*** .181 (.121) (.163) Ghana_ du 2.649*** .627 (.372) (.523) Gender_ du (mal e) .584 .153 (.350) (.580) Grade/Rank .076 .177 (.259) (.298) Constant 3.213** .006 (1.129) (1.666) N 145 143 adj. R sq .573 .032 Robust standard errors in parentheses * p<0.05, ** p<0.01, *** p<0.001

PAGE 196

196 7 2 4 The Negative Binom inal Regressions of Affective Reciprocity (3) (4) Dimensions of Affective Reciprocity The Variables Resource Diversion % Change in Expected Count Use of Official Time % Change in Expected Count Values Fairness .081** 8.5 (.029) Professionalism .051* 1.6 (.021) Economic Requests for Assistance .060*** 6.5 (.012) The Controls Education .172*** 13.6 (.033) Ghana dummy .497*** 38.1 (.126) Constant 1.263*** .879 (.259) (.805) N 145 143 Pseudo R sq .153 .036 ln alpha .499 Alpha .607 .105 Robust standard errors in parentheses * p<0.05, ** p<0.01, *** p<0.001 7 2 5 The Logistic Regressions of Affective Reciprocity Dimensions of Affectiv e Reciprocity Model 5 Resource Diversion (RSE) Resource Diversion Odds Ratio (RSE) Predicted Probabilities at mean Weberianess Professionalism .494* 1.64 2.52 (.237) (.34) Economic Requests for Assistance .397* 1.42 5.63 ( .157) (.22) Controls Level of Education .606** .545 4.85 (.219) (.119) Ghana dummy 2.659*** 14.281 .71 (.639) (.119) N 145 Robust standard errors in parentheses * p<0.05, ** p<0.01, *** p<0.001

PAGE 197

197 CHAPTER 8 THE PARADOX OF THE ECON OMY OF AFFECTION This study has explored the persistence of diminishing administrative capacity in two Agricultural Extension Services departments within the Ministries of Agriculture in Botswana and in Ghana. I have sought to show that while both Botswan a and Ghana are fairly capable, strong, and s states they nonetheless continue to struggle with much of the routine business of administration and management. As critical case studies, the two countries impart important lessons and provide important i nsights for theory building and refinement. In a region where poverty is endemic, where many still earn their living in the informal sector and therefore face old age without pensions, and where governments are unable to provide basic services, the burden of care falls on relatives and friends. The importance of informal institutions like the economy of affection and its attendant supporting institutions is reinforced daily. To meet their affective obligations; however, agricultural officers often divert sc arce resources meant for benefiting many towards narrow, particularistic ends. At a very local level, these extractions are often better targeted and deployed in a timelier manner because local officials, on average, have better knowledge of the communitie s within which they live and work. On a national level, hiring based on informal networks and personal relationships not only helps those in administrations simplify their search (as it often does outside of Africa as well), but it allows those in position s of power to fulfill social obligations and to demand greater loyalty and commitment from beneficiaries. This can be an invaluable outcome for these resource poor bureaucracies. However, the effects of affective reciprocity at the national level, on avera ge, are negative. In addition, because of the manner in which these resources are deployed, they are usually not monitored or evaluated which means we have no way of ensuring that the evaluation

PAGE 198

198 of these local level micro processes will be effective. This is the primary reason that African countries struggle to administrate more effectively, a problem that occurs as a result of inadequate institutional support and limited human resources that have hampered the ability to track and collect the data need for this type of evaluation. These same issues that surround the lack of monitoring and evaluation currently plague most institutions in the civil services. The process by which affective reciprocity bleeds the state of its limited resources is gradual. The ef fects of this weakening capacity are initially felt by the poorest of the poor who need state services the most; they are also the very last group to experience increasing state abilities. These citizens, in turn, respond to the empirical realities of bein g left increasingly to their own devices by turning to relatives and friends thus creating alternatives and through incremental adjustments weaken state capacities. Over time these effective or efficient alternatives calcify, creating new paths of, and tow ards, dependencies, challenges and opportunities, and this feedback loop becomes the suboptimal outcome for both parties. For example, seasonal skirmishes over stray cattle, water issues, and cattle rustling between rural Ghanaians in the northern reaches of the country and nomadic Fulani headsmen usually worsen during years of drought. Increasingly these clashes are becoming more frequent and deadlier each year, and in some smaller, more remote hamlets they are beginning to destabilize communities. As the northern regions dried up due to changes in climatic and weather patterns, Fulani herdsmen have pushed further south, inevitably clashing with the relatively wealthier, more populated Brong Ahafo region. Conflicts that had hitherto gotten little attention from the national press were now well publicized. According to interviews with some of the effected farmers in both regions, Fulani headsmen who had previously traveled in small groups far away from their homes in Mali and Niger now travelled in bands for protection. Believing the

PAGE 199

199 government was unable or unwilling to protect them, villagers took matters into their own hands by beating up headsmen who, in turn, found protection in numbers. The inability of the Ghanaian, Nigerian, and Malian governments to protect the most vulnerable of their citizens from militias in Niger and Mali and to ensure the livelihoods of Northern and increasingly, the Brong Ahafo region, caused Ghanaians in rural communities to force both groups to seek protection and solutions ou tside of formal channels. Farming communities mobilized their own every now and then to assist relatives deal with the problem and used government resources in ways that had not been sanctioned by the government. The reasons for their actions had very little to do with their normative beliefs; they acted to help secure food supplies and to assist relatives in meeting their basic subsistence needs. As one officer put i do not need my brothers and sisters staying with me in the one bedroom my best friend shares One of th e central claims of this dissertation is that economic pressures, rather than cultural values, drive much of the dysfunction in African bureaucracies. Governments unable to generate the levels of economic development needed to provide a certain minimum lev el of protection against unforeseen contingencies have effectively shifted the burden of providing for the aged, infirm and chronically ill onto families in a country where unemployment is extremely high. For the officials and the farmers who are the focus of this study, providing basic financial assistance, the bulk of which are for defraying tuition and healthcare related costs that are essential in meeting the needs of family, friends, and community, is an immense economic hardship. This means officers d evote less of their time to the business of the country and by so

PAGE 200

200 doing, hinder their own efforts at creating the conditions necessary for development of the ministry. Officials feel burdened by these extractions, yet they are compelled to contribute, thus as the findings from the study show, the reversed collective action dilemma comes into play because they are very much aware of the effects of their actions on state capacity.

PAGE 201

201 APPENDIX A SUPPLEMENTAL FIGURES 0 20 40 60 80 100 Percentage Score WGI 2013 Rule of Law Rankings Botswana/Ghana ) Botswana Ghana 0 20 40 60 80 100 1 3 5 7 9 11 13 15 Perecentage Score Control of Corruption Botswana Ghana 0 20 40 60 80 100 Percentage Score WGI 2013 Government Effectiveness Botswana/Ghana Botswana Ghana 0 20 40 60 80 100 Percentage Score WGI 2013 Requlatory Quality Rankings Botswana/Ghana Botswana Ghana 0 5 10 Welfare Regime Basic Admin Social safety nets Prioritization Policy coordination Bertlesmann Stiftung's 2014 Transformation Index Ghana Botswana

PAGE 202

202 0 20 40 60 80 100 Percentage Score Sampling areas by region/location Botswana Ghana 0 20 40 60 80 100 Male Female Percentage Score Gender of respondent Botswana Ghana Country Welfare Regime Mgt. Perfor m. Basic Admin Social capital Social safety nets Prioriti zation Implement ation. Policy coordin ation. Angola 3 4 4 5 2 5 5 5 Benin 5 7 6 6 4 5 5 6 Botswana 7 8 8 7 7 8 7 9 Burkina Faso 3 5 6 6 3 6 5 5 Burundi 4 4 7 4 3 4 4 4 Cameroon 5 4 5 4 4 3 3 4 Cent. African Repub. 3 4 1 3 2 3 4 4 Chad 3 3 3 3 3 1 3 2 Congo, DR 1 3 2 4 1 3 3 2 Congo, Rep. 3 4 5 3 3 4 4 3 Côte d'Ivoire 2 5 6 3 2 6 5 3 Eritrea 2 1 3 3 2 1 1 3 Ethiopia 5 4 6 4 5 4 5 6 Ghana 6 7 7 7 5 7 8 7 Guinea 4 5 5 5 3 5 4 6 Kenya 4 5 5 6 3 5 6 4 Lesotho 4 5 6 5 3 6 5 3 Liberia 4 7 5 5 3 7 6 5 Madagascar 4 4 5 4 3 3 3 4 Malawi 4 6 6 4 3 7 6 5 Mali 3 4 4 7 3 4 4 4 Mauritania 4 4 6 4 4 4 4 3 Mauritius 7 8 9 7 8 8 8 8 Mozambique 4 6 5 5 3 7 6 6 Namibia 6 7 7 6 6 8 7 7 Niger 3 6 3 6 3 5 6 7 Nigeria 4 5 5 4 4 4 4 5 Rwanda 7 5 7 3 6 5 5 7 Senegal 5 7 6 7 4 6 6 7 Sierra Leone 3 6 6 4 3 6 6 4 Somalia 1 2 1 2 1 1 1 2 South Africa 7 7 7 6 6 8 7 6 South Sudan 3 4 2 3 2 3 3 3 Sudan 2 2 3 3 1 2 2 2 Tanzania 4 6 5 5 4 5 6 4 Togo 4 5 6 6 3 4 4 6 Uganda 5 6 7 6 4 7 7 7 Zambia 5 6 7 6 4 6 7 5 Zimbabwe 3 3 5 4 3 3 3 3

PAGE 203

203 0 20 40 60 80 100 <24yrs 25-34 35-44 45+ yrs. Percentage Score Age of farmer Botswana Ghana 0 20 40 60 80 100 Percentage Score Level of education Botswana Ghana 0 20 40 60 80 100 Percentage Score Religion Botswana Ghana 0 20 40 60 80 100 Percentage Score Marital Status Botswana Ghana

PAGE 204

204 0 20 40 60 80 100 Rural Urban Percentage Score Settlement type Bostwana Ghana 0 20 40 60 80 100 Percentage Score Sampling areas by region/location Botswana Ghana

PAGE 205

205 0 10 20 30 40 50 60 70 80 90 Never A few times a year Monthly 81 2 17 5 11 85 Percentage Score Misappropriating government resources Botswana Ghana 0 10 20 30 40 50 60 70 80 90 100 Never A few times a year Monthly 43 36 21 59 28 13 Percentage Score Favoring trading within officer ranks Botswana Ghana

PAGE 206

206 0 10 20 30 40 50 60 70 80 90 100 Never A few times a year Monthly 55 29 17 7 52 41 Percentage Score Self financing favors Botswana Ghana

PAGE 207

207 0 10 20 30 40 50 60 70 80 90 100 0 0.5 1 2 3 4 5 6 7 8 6 17 31 19 13 4 1 1 1 6 9 17 37 17 6 3 1 1 1 9 Percentage Score Use of official time_hours per working Day Botswana Ghana 0 10 20 30 40 50 60 70 80 90 100 Never About an hour About two hours About three hours More than 4 hours 6 79 5 3 6 9 77 4 2 9 Percentage Score Hours per working day Botswana Ghana

PAGE 208

208 0 10 20 30 40 50 60 70 80 90 100 Botswana Ghana 12 12 88 88 Percentage Score Officials internalization of societal norms No Yes 0 10 20 30 40 50 60 70 80 90 100 Botswana Ghana 29 38 71 62 Percentage Score Expect 50 year old director to finance family and friends No Yes

PAGE 209

209 0 10 20 30 40 50 60 70 80 90 100 Botswana Ghana 50 59 50 41 Percentage Score Expect 50 year old director to help pay bills and school fees No Yes 0 10 20 30 40 50 60 70 80 90 100 S. Disagree Disagree Neither Agree S. Agree 19 5 14 24 38 16 9 7 23 45 Percentage Score Civil servants should obey rules even if they don't like them Botswana Ghana

PAGE 210

210 0 10 20 30 40 50 60 70 80 90 100 S. Disagree Disagree Neither Agree S. Agree 19 33 10 24 14 18 42 4 23 13 Percentage Score Mr. Kabelo should let Mpho stay in the city Botswana Ghana 0 10 20 30 40 50 60 70 80 90 100 S. Disagree Disagree Agree S. Agree 10 36 50 5 19 46 29 5 Percentage Score Mr. Kabelo should transfer Mpho Botswana Ghana

PAGE 211

211 0 10 20 30 40 50 60 70 80 90 100 S. Disagree Disagree Neither Agree S. Agree 0 38 7 41 14 10 38 4 35 13 Percentage Score There is nothing wrong with Mr. Kabelo helping his nephew Botswana Ghana 0 10 20 30 40 50 60 70 80 90 100 S. Disagree Disagree Neither Agree S. Agree 8 49 15 10 18 20 50 7 14 8 Percentage Score Society will view Kabel as wicked if he does not help his nephew Botswana Ghana

PAGE 212

212 0 10 20 30 40 50 60 70 80 90 100 S. Disagree Disagree Agree S. Agree 24 41 33 2 29 46 22 3 Percentage Score Mpho should act professionally be formally pleading his case Botswana Ghana

PAGE 213

213 0 10 20 30 40 50 60 70 80 90 100 Hardly ever About half the time Almost always 20 51 29 42 35 23 Percentage Score Officials strive to implement policies fairly and consistently Botswana Ghana 0 10 20 30 40 50 60 70 80 90 100 S. Disagree Disagree Neither Agree S. Agree 12 60 17 10 1 22 21 14 24 20 Percentage Score Officers favor some people when implementing policies case by case Botswana Ghana

PAGE 214

214 0 10 20 30 40 50 60 70 80 90 100 Never Few times/yr. Monthly 14 2 83 5 53 42 Percentage Score How often are you asked to contribute for healthcare costs and school fees by friends and family Botswana Ghana 0 10 20 30 40 50 60 70 80 90 100 Never Few times/yr. Monthly 43 45 12 38 29 33 Percentage Score How often are you asked to make financial donations to your religious organization, hometown association, alma mater, or a group/club you belong to? Botswana Ghana

PAGE 215

215 0 10 20 30 40 50 60 70 80 90 100 Never Few times/yr. Monthly 61 8 31 15 47 38 Percentage Score How often are you asked to contribute to a wedding or funeral by freinds and family Botswana Ghana 0 10 20 30 40 50 60 70 80 90 100 Never Few times/yr. Monthly 17 41 43 28 65 8 Percentage Score How often are you asked to contribute financially towards a job search Botswana Ghana

PAGE 216

216 0 10 20 30 40 50 60 70 80 90 100 Never Few times/yr. Monthly 46 26 28 69 21 10 Percentage Score How often are you asked to help relative/friend navigate a government bureaucracy, or find a job, or gain admission to a school? Botswana Ghana 0 10 20 30 40 50 60 70 80 90 100 Never Yearly Every 6months Every 3months Monthly 0 17 48 19 17 5 11 5 19 60 Percentage Score How often do you send money to your relatives Botswana Ghana

PAGE 217

217 0 20 40 60 80 100 Male Female Percentage Score Gender of respondent Botswana Ghana 0 20 40 60 80 100 <24yrs 25-34 35-44 45+ yrs. Percentage Score Age of Officer Botswana Ghana

PAGE 218

218 0 20 40 60 80 100 Percentage Score Level of education Botswana Ghana 0 20 40 60 80 100 No Christian Muslim Trad.A. Percentage Score Religion Botswana Ghana

PAGE 219

219 0 20 40 60 80 100 Percentage Score Marital status Botswana Ghana 0 20 40 60 80 100 Upper Mgt Mid Mgt Jnr. Officer Percentage Score Officers rank Botswana Ghana

PAGE 220

220 0 10 20 30 40 50 60 70 80 90 100 27 18 55 29 36 35 Percentage Score Research Sites by Region_Officers' Surveys Botswana Ghana 0 20 40 60 80 100 Male Female 54 46 81 19 Percentage Score Gender of Officers Botswana Ghana

PAGE 221

221 0 10 20 30 40 50 60 70 80 90 100 25-34 35-44 45+ yrs. 31 32 37 25 21 54 Percentage Score Age of Officers Botswana Ghana 0 10 20 30 40 50 60 70 80 90 100 Primary Secondary Cert/Dip University Postgrad 4 4 20 40 31 1 5 66 21 7 Percentage Score Officers' Level of Education Botswana Ghana

PAGE 222

222 0 10 20 30 40 50 60 70 80 90 100 Christian Muslim Trad.A. 83 17 0 60 39 1 Percentage Score Officers' Religious Affiliation Botswana Ghana 0 10 20 30 40 50 60 70 80 90 100 Single Married/Relationship Widowed/Divorced 29 61 10 14 84 2 Percentage Score Officers' Marital Status Botswana Ghana

PAGE 223

223 0 10 20 30 40 50 60 70 80 90 100 Upper Level Mid Level Jnr. Officers 43 46 11 37 21 42 Percentage Score Officers rank Botswana Ghana 0 10 20 30 40 50 60 70 80 90 100 Male Female 57 43 76 24 Percentage Score Gender of Farmers Botswana Ghana

PAGE 224

224 0 10 20 30 40 50 60 70 80 90 100 <24yrs 25-34 35-44 45+ yrs. 19 34 22 24 10 38 27 25 Percentage Score Age of Farmers Botswana Ghana 0 10 20 30 40 50 60 70 80 90 100 No Primary Secondary Cert/Dip University Postgrad 12 17 44 23 4 1 47 0 10 6 27 11 Percentage Score Farmers' Level of Education Botswana Ghana

PAGE 225

225 0 10 20 30 40 50 60 70 80 90 100 No Christian Muslim Trad.A. 40 38 0 21 0 33 64 3 Percentage Score Farmers' Religious Affiliation Botswana Ghana 0 10 20 30 40 50 60 70 80 90 100 Single Married/Relationship 44 56 50 50 Percentage Score Farmers' Marital Status Botswana Ghana

PAGE 226

226 0 10 20 30 40 50 60 70 80 90 100 Rural Urban 52 48 80 20 Percentage Score Settlement Type Bostwana Ghana 0 20 40 60 80 100 Chobe Gantsi Brong-Ahafo Northern 24 76 23 77 Percentage Score Research Sites by Region_Farmers' Surveys Botswana Ghana

PAGE 227

227 LIST OF REFERENCES Abdi, Samatar. 1999. African Miracle: Sta te And Class Leadership And Colonial Legacy In Botswana Development. Abernethy, David B. 2002. The Dynamics of Global Dominance: European Overseas Empires, 1415 1980 : Yale University Press. Acemoglu, Daron, Johnson, Simon and Robinson, James A. 2001. "The Colonial Origins of Comparative Development: An Empirical Investigation." American Economic Review 91 1369 401. Acemoglu, D. And J. A. Robinson (2006). Economic Origins Of Dictatorship And Democracy. Cambridge; New York, Cambridge University Press. A cemoglu, D., S. Johnson, and J. A. Robinson. 2002. "Reversal of Fortune: Geography and Institutions in the Making of the Modern World Income Distribution." Quarterly Journal of Economics 117: 1231 94. Adcock, R., and D. Collier. 2001. "Measurement Valid ity: A Shared Standard for Qualitative and Quantitative Research." American Political Science Review 95: 529 46. Agresti, Alan, and Barbara Finlay. 1997. Statistical Methods for the Social Sciences . 3rd ed. Upper Saddle River, N.J.: Prentice Hall. Albrow, Martin. 1970. Bureaucracy , Key Concepts in Political Science. New York,: Praeger. Alesina, A. and La Ferrara, E. . 2000. "Participation in Heterogeneous Communities." Quarterly Journal of Economics 115: 847 904. Alesina, Alberto, and Robert J. Barro. 2001. Currency Unions , Hoover Institution Press Publication. Stanford, Calif.: Hoover Institution Press, Stanford University. Allen, Chris. 1999. "Warfare, Endemic Violence and State Collapse in Africa." Review of African Political Economy 81: 367 84. Alm ond, G. A. and J. S. Coleman.1960. The Politics of The Developing Areas. Princeton, N.J., Princeton University Press. Almond, G. A. And S. Verba (1989). The Civic Culture Revisited. Newbury Park, Calif., Sage Publications. Amin, Samir. 1974. Accumulation on a World Scale: A Critique of the Theory of Underdevelopment . New York,: Monthly Review Press. Aoki, M. 1984. The Cooperative Game Theory of the Firm . Oxford: Clarendon Press. Apter, D. E. (1965). The Politics Of Modernization. Chicago, University O f Chicago Press.

PAGE 228

228 Apter, D. E. (1971). Choice And The Politics Of Allocation; A Developmental Theory. New Haven, Yale University Press. Apter, D. E. (1987). Rethinking Development: Modernization, Dependency, And Postmodern Politics. Newbury Park, Calif., Sa ge Publications. Argyriades, Demetrios. 2006. "The Rise, Fall, and Rebirth of Comparative Administration: The Rediscovery of Culture." Public Administration Review 66: 281 84. Aryeetey, A. M. A. E. (2004). "Operationalizing Pro Poor Growth." A Joint Initi ative Of AFD, BMZ (GTZ, Kfw Development Bank), DFID, And The World Bank. Aryeetey, Andrew McKay and Ernest. 2004. "Operationalizing Pro Poor Growth." A Joint Initiative of AFD, BMZ, DFID, and the World Bank . Ashforth, Adam. 2005. Witchcraft, Violence, And Democracy In South Africa. Chicago: University Of Chicago Press. Austin, Gareth. 2010. "African Economic Development and Colonial Legacies." International Development Policy Series : 11 32. Austin, Gareth. 2010. "African Economic Development and Coloni al Legacies." International Development Policy Series : 11 32. Auyero, J. And Netlibrary Inc. (2001). Poor People's Politics Peronist Survival Networks And The Legacy Of Evita. Durham, Duke University Press. Axelrod, Robert M. 1984. The Evolution of Coope ration . New York: Basic Books. Ayittey, George B. N. 1992. Africa Betrayed . New York: St. Martin's Press. Baldwin, R.E. 1969. "The Case against Infant Industry Tariff Protection." Journal of Political Economy 77: 295 305. Bardhan, P. K. And C. Udry (20 00). Readings In Development Microeconomics. Cambridge, Mass., MIT Press. Bardhan, P., Maitreesh Ghatak and Alexander Karaivanov. 2002. "Inequality and Collective Action." Paper presented at the Conference on Inequality, Cooperation and Environmental Sus tainability and the NEUDC 2001 Conference, Santa Fe Institute and Boston University. Bardhan, Pranab and Ghatak, Maitreesh 2000. "Inequality, Market Imperfections, and Collective Action Problems." Public Economics 0004001, EconWPA. Barr, A. and Oduro, A. Sector." Working Paper Series . Barro, R. J. 1989. "Economic Growth in a Cross Section of Countries." University

PAGE 229

229 Bates, R. H. 2006. "Institutions and Development." Journal of African Eco nomies 15: 10 61. Bates, Robert H. 1981. Markets and States in Tropical Africa: The Political Basis of Agriculatural Policies , California Series on Social Choice and Political Economy. Berkeley: University of California Press. Bates, Robert H. 1998. An alytic Narratives . Princeton, N.J.: Princeton University Press. Bates, Robert H. 1981. Markets And States In Tropical Africa: The Political Basis Of Agricultural Policies. Berkeley: University Of California Press. Bawn, Kathleen. 1993. "The Logic Of Instit utional Preferences: German Electoral Law As A Social Choice Outcome." American Journal Of Political Science 37 (4): 965 89. Bayart, Jean François, Stephen Ellis, And Beatrice Hibou. 1999. The Criminalization Of The State In Africa. Bloomington: Indiana Un iversity Press. Behn, Robert D. . 1995. "The Big Questions of Public Management." Public Administration Review 55. Bekke, A. J. G. M., J. L. Perry, et al. (1996). Civil service systems in comparative perspective . Bloomington, Indiana University Press. B ello, W. F., S. Cunningham, Et Al. (1994). Dark Victory: The United States, Structural Adjustment, And Global Poverty. London Oakland Amsterdam, Pluto; Food First; TNI. Bendix, Reinhard. 1956. Work And Authority In Industry; Ideologies Of Management In The Course Of Industrialization. New York, Wiley. Berger, Peter L. and Luckmann, Thomas. 1966. The Social Construction of Reality: A Treatise in the Sociology of Knowledge . 1st ed. NY: Doubleday. Bevan, David, Paul Collier, Jan Willem Gunning, and Ian Gol din. 1993. "Government Policies and Agricultural Performance: Tanzania and Kenya." In Economic Reform, Trade and Agricultural Development . New York: St. Martin's Press; London: Macmillan Press in association with the OECD Development Centre. 19 47. Bhat tacharyya, S. 2009. "Root Causes of African Underdevelopment." Journal of African Economies 18: 745 80. Blau, Peter Michael, And W. Richard Scott. 2004. Formal Organizations: A Bloom, David E. and Jeffrey D. Sachs. 1998. "Geography, Demography, and Eco nomic Growth in Africa." Brookings Papers on Economic Activity, Economic Studies Program, The Brookings Institution 29: 207 96. Blundo, de Sardan, and Arifari. 2006. Everyday Corruption and the State . Translated by Susan Cox. 2006 ed: Zed Books London and New York.

PAGE 230

230 Blundo, de Sardan, and Arifari. 2006. Everyday Corruption and the State . Translated by Susan Cox. 2006 ed: Zed Books London and New York. Blundo, Giorgio and Le Meur, Pierre Yves. 2009. The Governance of Daily Life in Africa: Ethnographic Explor ations of Public and Collective Services , African Social Studies Series. Leiden ; Boston: Brill. Boahen, A. Adu. 1987. African Perspectives on Colonialism , The Johns Hopkins Symposia in Comparative History. Baltimore: Johns Hopkins University Press. Boi x, Charles, And Susan Carol Stokes. 2007. The Oxford Handbook Of Comparative Politics. Oxford; New York: Oxford University Press. Bourdieu, Pierre. 1977. Outline of a Theory of Practice . Cambridge; New York: Cambridge University Press. Bratton, M, And N Van De Walle. 1996. "Political Regimes And Regime Transitions In Africa, 1910 1994." ICPSR. Buah, F. K. 2005. Government in West Africa . Accra, Ghana: Readwide Publishers and FABS Callaghy, Thomas M., and John Ravenhill. 1993. Hemmed In : Responses to Africa's Economic Decline . New York: Columbia University Press. Campbell, John P. 1977. "On the Nature of Organizational Effectiveness." In New Perspectives on Organizational Effectiveness , ed. P. S. and Pennings Goodman, J. M. . San Francisco, CA: Jos sey Bass, Inc. Canache, Damarys, J. J. Mondak, and M. A. Seligson. 2001. "Meaning and Measurement in Cross National Research on Satisfaction with Democracy." In Public Opinion Quarterly . Vol. 65. 506 28. Chabal, P. and J. P. Daloz (1999). Africa works : disorder as political instrument . Bloomington. Indiana University Press. Chabal, Patrick and Daloz, Jean Pascal. 1999. Africa Works: Disorder as Political Instrument , African Issues. Bloomington: Indiana University Press. Chabal, Patrick. 1992. Power i n Africa: An Essay in Political Interpretation . New York: St. Martin's Press. Chabal, Patrick. 1986. Political Domination in Africa: Reflections on the Limits of Power , African Studies Series ; 50. Cambridge [Cambridgeshire] ; New York: Cambridge Universi ty Press. Chazan, Naomi. 1988. "Patterns of State Society Incorporation and Disengagement in Africa." In The Precarious Balance : State and Society in Africa , eds. Donald S. Rothchild and Naomi Chazan. Boulder: Westview Press. x, 357 p.

PAGE 231

231 Chazan, Naomi . 1983. An Anatomy of Ghanaian Politics: Managing Political Recession, 1969 1982 . Boulder, Colo.: Westview Press. Chibber, V. 2002. "Bureaucratic Rationality and the Developmental State Clapham, C. 1998a. "Degrees of Statehood." Review of Internationa l Studies 24: 143 57. Clapham, Christopher S. 1996b. Africa and the International System : The Politics of State Survival , Cambridge Studies in International Relations ; 50. Cambridge ; New York: Cambridge University Press. Clark, Colin. 1940. The Condit ions of Economic Progress . London,: Macmillan and co., limited. Clark, John F. 2005. "Petroleum Revenues and Political Development in the Congo Republic: The Democratic Experiment and Beyond." In Hamburg African Studies, , ed. Matthias and Mehler Based au, Andreas. Hamburg: Institute of African Affairs. 355 p. Collier, D. 1993. "The Comparative Method." In Political Science: The State of the Discipline Ii , ed. Ada Finifter: American Political Science Association. 105 19. Collier, Paul and Jan Will em Gunning. 1999. "Explaining African Economic Performance." Journal of Economic Literature 37: 64 111. Collier, Paul, and Jan Willem Gunning. 1999. "Why Has Africa Grown Slowly?". Journal of Economic Perspectives 13: 3 22. Converse, Jean M., And Howard Schuman. 1974. Conversations At Random: Survey Research As Interviewers See It . New York,: Wiley Cox, Gary W. 1997. Making Votes Count : Strategic Coordination in the World's Electoral Systems , Political Economy of Institutions and Decisions. Cambridge : Cambridge University Press. Crook, Richard C 1989. "Patrimonialism, Administrative Effectiveness and Economic Development in Côte D'ivoire." African Affairs 88: 205 28. Dahl, Robert A. . 1947. "The Science of Public Administration: Three Problems. ." Public Administration Review 7 1 11. Dahl, Robert Alan, Ian Shapiro, and Jos© Ant©þnio Cheibub. 2003. The Democracy Sourcebook . Cambridge, Mass.: MIT Press. David Ricardo, Piero Sraffa, Maurice Dobb, ed. 1952. The Works and Correspondence of David Ri cardo: Volume 8, Letters 1819 June 1821 . 11 vols, Works and Correspondence: Cambridge University Press.

PAGE 232

232 Davidson, Basil, and F. K. Buah. 1966. A History of West Africa to the Nineteenth Century. With F. K. Buah and the Advice of J. F. Ade Ajayi . [Rev. ] ed. Garden City, N.Y.,: Anchor Books. Davis, Charles R. 1996. Organization Theories and Public Administration . Westport, Conn.: Praeger. Dayton Johnson, Jeff. 2000. "Determinants of Collective Action on the Local Commons: A Model with Evidence from Me xico ". Journal of Development Economics 62(1), pages 181 208: 181 208. Diao, Xinshen & Hazell, Peter & Resnick, Danielle & Thurlow, James,. 2006. "The Role of Agriculture in Development: Implications for Sub Saharan Africa." DSGD Discussion Papers In ternational Food Policy Research Institute (IFPRI) . Dollar, David, Aart Kraay, Anthony Shorrocks, and Rolph van der Hoeven. 2004. "Growth Is Good for the Poor." 29 61. Dresang, Dennis L. 1973. "Entrepreneurialism and Development Administration." Administra tive Science Quarterly 18: 76 85. Donald J. Savoie (2006), What Is Wrong With The New Public Management, In Eric E. Otenyo, Nancy S. Lind (Ed.) Comparative Public Administration (Research In Public Policy Analysis And Management, Volume 15), Emerald Group Publishing Limited, Pp.593 602 Easterly, William and Levine, Ross. 1997. "Africa's Growth Tragedy: Policies and Ethnic Divisions." The Quarterly Journal of Economics 112: 1203 50. Easterly, William Russell. 2001. The Elusive Quest for Growth Economist s' Adventures and Misadventures in the Tropics . Cambridge, Mass.: MIT Press. Eisenstadt, S. N. 1995. Power, Trust, and Meaning : Essays in Sociological Theory and Analysis . Chicago: University of Chicago Press. Eisenstadt, S. N., and René Lemarchand. 1 981. Political Clientelism, Patronage, and Development , Contemporary Political Sociology ; V. 3. Beverly Hills: Sage Publications. Ekeh, Peter P. 1975. "Colonialism and the Two Publics in Africa: A Theoretical Statement." Comparative Studies in Society and History 17: 91 112. Ellis, Stephen. 1996. Africa Now : People, Policies, Institutions . Portsmouth, NH: Heinemann. Esman, Milton J. 2006. "Development Assistance in Public Administration: Requiem or Renewal." In Comparative Public Administration (Re search in Public Policy Analysis and Management) , ed. Nancy S. Lind Eric E. Otenyo. Vol. 15: Emerald Group Publishing Limited. 345 58.

PAGE 233

233 Esping Andersen, Gøsta. 1990. The Three Worlds of Welfare Capitalism . Cambridge: Polity. Evans, P. and J. Rauch. 1999. "Bureaucracy and Growth: A Cross National Analysis of the Effects of "Weberian" State Structures on Economic Growth." American Sociological Review 64: 748 65 Evans, Peter B. 1995. Embedded Autonomy: States and Industrial Transformation . Princeton, N. J.: Princeton University Press. Evans, Peter B. 1997. State Society Synergy: Government and Social Capital in Development . Berkeley, Calif: University of California at Berkeley. Evans, Peter B., Harold Karan Jacobson, and Robert D. Putnam. 1993. Double Edg ed Diplomacy : International Bargaining and Domestic Politics , Studies in International Political Economy. Berkeley: University of California Press. Fafchamps, M. and Minten, B. 2002. "Returns to Social Network Capital among Traders." Oxford Economic Paper s 54: 174 206. . 2001. "Social Capital and Agricultural Trade." American Journal of Agricultural Economics 83: 680 85. Fry, Brian R. 1989. Mastering Public Administration: From Max Weber to Dwight Waldo , Chatham House Series on Change in American Politi cs. Chatham, N.J.: Chatham House Publishers. Fukuyama, Francis. 1999. Social Capital And Civil Society. IMF Conference On Second Generation Reforms, Pp. 1.13. Galvan, Dennis Charles. 2004. The State Must Be Our Master Of Fire: How Peasants Craft Culturally Sustainable Development In Senegal. Berkeley: University Of California Press. Geertz, Clifford. 1973. The Interpretation of Cultures; Selected Essays . New York,: Basic Books. Geertz, Clifford. 1973. The Interpretation Of Cultures: Selected Essays. New Yor k: Basic Books. George, Alexander L. and Bennett, Andrew. 2005. Case Studies and Theory Development in the Social Sciences , Bcsia Studies in International Security. Cambridge, Mass.: MIT Press. Geschiere, Peter (1997). The Modernity Of Witchcraft: Politics And The Occult In Postcolonial Africa Geschiere, Peter (2011) "Autochthony, Citizenship, And Exclusion Paradoxes In The Politics Of Belonging In Africa And Europe," Indiana Journal Of Global Legal Studies : Vol. 18: Iss. 1, Article 14.

PAGE 234

234 Goertz, Gary. 200 6. Social Science Concepts : A User's Guide . Princeton: Princeton University Press. Goldsmith, Arthur A. . 1999. "Africa's Overgrown State Reconsidered: Bureaucracy and Economic Growth." World Politics Vol. 51: 520 46. Grindle, Merilee S. 1997. Getting G ood Government: Capacity Building In The Public Sectors Of Developing Countries. [Cambridge, MA]: Harvard Institute For International Development Distributed By Harvard University Press. Grindle, Merilee Serrill. 1997a. Getting Good Government : Capacity B uilding in the Public Sectors of Developing Countries , Harvard Studies in International Development. Cambridge, MA: Harvard Institute for International Development Harvard University : Distributed by Harvard University Press. Groves, Robert M. 2004. Survey Methodology , Wiley Series in Survey Methodology. Hoboken, NJ: J. Wiley. Groves, Robert M. 2004. Survey Methodology . Hoboken, NJ: J. Wiley. Gyimah Brempong, K., and Corley, M. E. 2005. "Civil Wars and Economic Growth in Sub Saharan Africa." Journal of Afri can Economies 14: 270 311. Habyarimana, James P.; Humphreys, Macartan; Posner, Daniel N. and Weinstein, Jeremy M. 2009. Coethnicity: Diversity and the Dilemmas of Collective Action , Russell Sage Foundation Series on Trust. New York: Russell Sage Foundation . Haddad, Lawrence, And John Maluccio. 2003. Trust, Membership In Groups And Household Welfare: Evidence From Kwazulu Natal, South Africa. Economic Development And Cultural Change 51(3): 573 601. Hadenius, Axel. 1992. Democracy and Development . Cambridge ; New York: Cambridge University Press. Hadenius, Axel. 1992. Democracy And Development. Cambridge; New York: Cambridge University Press. Hardin, Russell. 1991. "Acting Together, Contributing Together." Rationality and Society 3 365 80. . 1995. One for A ll . Princeton: Princeton University Press. Hazell, P., and A. Roell. 1983. "Rural Growth Linkages: Household Expenditure Patterns in Malaysia and Nigeria." In Secondary Rural Growth Linkages: Household Expenditure Patterns in Malaysia and Nigeria , ed Secon dary. Washington, D.C.,. Reprint, Reprint. Heady, Ferrel. 1966. Public Administration, a Comparative Perspective . Englewood Cliffs, N.J.,: Prentice Hall.

PAGE 235

235 Hecht, Robert. M. 1983. "The Ivory Coast Economic 'Miracle': What Benefits for Peasant Farmers?". The Journal of Modern African Studies 21: 25 53. Held, David, and Anthony G. McGrew. 2003. The Global Transformations Reader : An Introduction to the Globalization Debate . 2nd ed. Cambridge, UK Helmke, Gretchen, and Steven Levitsky. 2006. Informal Institutions and Democracy : Lessons from Latin America . Baltimore: Johns Hopkins University Press. Herbst, Jeffrey Ira. 2000a. States And Power In Africa: Comparative Lessons In Authority And Control. Princeton, N.J.: Princeton University Press. Huber, Evelyne, And J ohn D. Stephens. 2001. Development And Crisis Of The Welfare State : Parties And Policies In Global Markets . Chicago: The University Of Chicago Press. Hydén, Göran. 1980. Beyond Ujamaa In Tanzania: Underdevelopment And An Uncaptured Peasantry. London: Hein emann. Hyden, Göran. 1983. No Shortcuts To Progress: African Development Management In Perspective. Berkeley: University Of California Press. Isaksen, J. 1981. "Macroeconomic Management and Bureaucracy, the Case of Botswana." Scandinavian Institute of Afri can Studies . Jackson, Robert H., and Carl Gustav Rosberg. 1982a. Personal Rule in Black Africa : Prince, Autocrat, Prophet, Tyrant . Berkeley: University of California Press. . 1982b. Personal Rule in Black Africa : Prince, Autocrat, Prophet, Tyrant . Ber keley: University of California Press. Javeline, Debra. 1999. "Response Effects in Polite Cultures: A Test of Acquiescence in Kazakhstan." The Public Opinion Quarterly 63: 1 28. Johnson, Chalmers A. 1982. Miti and the Japanese Miracle : The Growth of Indus trial Policy, 1925 1975 . Stanford, Calif.: Stanford University Press. Joseph, Richard. 1987. Prebendalism and Democracy in Nigeria . Cambridge: Cambridge University Press. Kaplan, Berton H. 1968. "Notes on a Non Weberian Model of Bureaucracy: The Case of De velopment Bureaucracy." Administrative Science Quarterly 13: 471 83. Kasfir, Nelson. 1976. The Shrinking Political Arena: Participation and Ethnicity in African Politics with a Case Study of Uganda . Berkeley: University of California Press. Kaufmann, Danie l, and Aart Kraay. 2008. "Governance Indicators: Where Are We, Where Should We Be Going?". World Bank Research Observer 23 1: 1 30.

PAGE 236

236 Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi. 2004. "Governance Matters Iii: Governance Indicators for 1996, 1998, 20 00, and 2002." World Bank Economic Review 18 2: 253 87. Kaufmann, Daniel, Aart Kraay, Massimo Mastruzzi, and Susan Rose Ackerman. 2006. "Measuring Governance Using Cross Country Perceptions Data." In International Handbook on the Economics of Corruption . C heltenham, U.K. and Northampton, Mass.: Elgar. 52 104. Kaufmann, Daniel, Aart Kraay, and Pablo Zoido Lobaton. 2000. "Governance Matters from Measurement to Action." Finance and Development 37 2: 10 13. Kaufmann, Daniel, Kraay, Aart and Mastruzzi, Massimo. 2008. "Governance Matters Vii: Aggregate and Individual Governance Indicators 1996 2007." The World Bank, Policy Research Working Paper Series: 4654. Keefer, Philip, and Stephen Knack. 1997. "Why Don't Poor Countries Catch Up? A Cross National Test of an I nstitutional Explanation." Economic Inquiry 35: 590 602. Kitschelt, H. 2000. "Linkages between Citizens and Politicians in Democratic Polities." Comparative Political Studies 33: 845 79. Kitschelt, Herbert, and Steven Wilkinson. 2007. Patrons, Clients and Policies . Cambridge: Cambridge University Press. Knack, Stephen and Keefer, Stephen. 1997 "Does Social Capital Have an Economic Pay Off? A Cross Country Investigation." Quarterly Journal of Economics 112 1251 88. Kohli, Atul. 2004a. State Directed Developm ent : Political Power and Industrialization in the Global Periphery . Cambridge, UK ; New York: Cambridge University Press. Kraay, Aart, and Daniel Kaufmann. 2007. "Governance Indicators : Where Are We, Where Should We Be Going ?". The World Bank, Policy Re search Working Paper Series: 4370. Krieckhaus, Jonathan Tabor. 2006. Dictating Development : How Europe Shaped the Global Periphery . Pittsburgh: University of Pittsburgh Press. Lange, Matthew. 2005. "States and Development: An Introduction." In States and Development: Historical Antecedents of Stagnation and Advance , ed. Matthew Lange and Dietrich Rueschemeyer: Palgrave Macmillan Press. 3 25. Lemarchand, Rene. 1972. "Political Clientelism and Ethnicity in Tropical Africa: Competing Solidarities in Nation Bu ilding." The American Political Science Review 66: 68 90. Lewis, Peter. 1998. Africa : Dilemmas of Development and Change . Boulder, Colo.: Westview Press.

PAGE 237

237 Lipset, Seymour Martin. 1959. "Some Social Requisites of Democracy: Economic Development and Politica l Legitimacy." American Political Science Review 53: 69 105. Lipset, Seymour Martin, Reinhard Bendix, and University of California Berkeley. Institute of Industrial Relations. 1959. Social Mobility in Industrial Society . Berkeley,: University of California Press. Mahoney, J. and Goertz. 2006. "A Tale of Two Cultures: Contrasting Quantitaive and Qualitative Research." Political Analysis 14: 227 49. Mahoney, James and Rueschemeyer, Dietrich. 2003. Comparative Historical Analysis in the Social Sciences , Cambri dge Studies in Comparative Politics. Cambridge, U.K.; New York: Cambridge University Press. Mastruzzi, Massimo, Aart Kraay, and Daniel Kaufmann. 2005. "Governance Matters Iv: Governance Indicators for 1996 2004." The World Bank, Policy Research Working Pap er Series: 3630. Medard, Jean Francois. 1982. "The Underdeveloped State in Tropical Africa: Political Clientalism or Neo Patrimonialism?" In Private Patronage and Public Power: Political Clientelism in the Modern State , ed. Christopher S. Clapham. New York : St. Martin's Press. vi, 222 p. Miguel, Edward and Gugerty, Mary Kay 2005. "Ethnic Diversity, Social Sanctions, and Public Goods in Kenya." Journal of Public Economics 89: 2325 68. Milne, R.S. 1973. "Bureaucracy and Development Administration." Public Adm inistration 51: 411 26. Moraski, Bryon J., and William M. Reisinger. 2007. "Eroding Democracy: Federal Intervention in Russia's Gubernatorial Elections." Democratization 14: 603 21. Ndulu, O'Connell, Bates, Collier and Soludo, ed. 2008. The Political Eco nomy of Economic Growth in Africa, 1960 2000 . 2 vols. Vol. 2: Cambridge University Press. Nelson, Joan M. 1987. "Political Participation." In Understanding Political Development: An Analytic Study , eds. Myron Weiner, Samuel P. Huntington and Gabriel Abraha m Almond. Boston: Little, Brown. 103 59. Nordås, Hildegunn Kyvik. 2000. "The Role of Government in Growth and Income Distribution: The Case of Botswana." In Secondary The Role of Government in Growth and Income Distribution: The Case of Botswana , ed Second ary . Reprint, Reprint. North, Douglass C. 1981. Structure and Change in Economic History . 1st ed. New York: Norton. North, Douglass C. 1990a. "A Transaction Cost Theory of Politics." Journal of Theoretical Politics 2: 335 67.

PAGE 238

238 North, Douglass C., and Ba rry R. Weingast. 1989. "Constitutions and Commitment: The Evolution of Institutions Governing Public Choice in Seventeenth Century England." The Journal of Economic History 49: 803 32. North, Douglass Cecil. 1990b. Institutions, Institutional Change, and E conomic Performance : Cambridge University Press. Quarterly Journal of Economics 1: 139 76. O'Dwyer, Conor. 2004. "Runaway State Building: How Political Parties Shape States in Postcommun ist Eastern Europe." World Politics 56: 520 53. Kellogg Institute for International Studies at the University of Notre Dame Working Paper #222 . Ojo, O., and Oshikoya, T. 1995. "Determinants of Long Term Growth: Some African Results." Journal of African Economies 4: 163 91. Olson, Mancur. 1965. The Logic of Collective Action; Public Goods and the Theory of Groups , Harvard Economic Studies, V. 124. Cambridge, Mass.,: Harv ard University Press. Ostrom, E and Walker, J. 2003a. Trust and Reciprocity: Interdisciplinary Lessons from Experimental Research , Russell Sage Foundation Series on Trust. New York: Russell Sage Foundation. Ostrom, E and Walker, J. , ed. 2003b. Trust and R eciprocity . New York: Russel Sage. Ostrom, Elinor, and T. K. Ahn. 2003. Foundations of Social Capital , Critical Studies in Economic Institutions. Northhampton, MA: Edward Elgar Pub. Piattoni, Simona. 2001. Clientelism, Interests, and Democratic Representat ion: The European Experience in Historical and Comparative Perspective , Cambridge Studies in Comparative Politics. Cambridge ; New York: Cambridge University Press. Pierson, Christopher. 1996. The Modern State , Key Ideas. London; New York: Routledge. Posne r, Daniel N. 2005. Institutions and Ethnic Politics in Africa . Cambridge: Cambridge University Press. Price, Gregory N. 2003. "Economic Growth in a Cross Section of Nonindustrial Countries: Does Colonial Heritage Matter for Africa?". Review of Development Economics 7 3: 478 95. Price, Robert M. 1975. Society and Bureaucracy in Contemporary Ghana . Berkeley: University of California Press.

PAGE 239

239 Przeworski, Adam. 1991. Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America , Stu dies in Rationality and Social Change. Cambridge; New York: Cambridge University Press. Przeworski, Adam, and Fernando Limongi. 1993. "Political Regimes and Economic Growth." Journal of Economic Perspectives 7: 51 69. Rae, Douglas W. 1971. The Political Co nsequences of Electoral Laws . Rev. ed. New Haven: Yale University Press. Ragin, Charles C. 1987. The Comparative Method : Moving Beyond Qualitative and Quantitative Strategies . Berkeley: University of California Press. Rauch, James E. 1995. Choosing a Dict ator: Bureaucracy and Welfare in Less Developed Polities , Nber Working Paper Series. Cambridge, MA: National Bureau of Economic Research. Roniger, Luis, and Ay se Güne*s Ayata. 1994. Democracy, Clientelism, and Civil Society . Boulder, Colo.: L. Rienner Pub lishers. Rueschemeyer, Dietrich, and Theda Skocpol. 1996. States, Social Knowledge, and the Origins of Modern Social Policies . Princeton, N.J. New York: Princeton University Press/ Russell Sage Foundation. Sachs, J.D. and A.M Warner. 1997. "Sources of Slow Growth in African Economies." Journal of African Economies 6: 335 76. Schaffer, Bernard ed. 1969. The Deadlock in Development Administration . Edited by Colin Leys, Politics and Change in Developing Countries: Cambridge University Press. Schmidt, Steffen W . 1974. "Bureaucrats as Modernizing Brokers? Clientelism in Colombia." Comparative Politics 6 425 50. . 1977. Friends, Followers, and Factions: A Reader in Political Clientelism . Berkeley: University of California Press. Schuman, Howard, and Stanley Pre sser. 1981. Questions and Answers in Attitude Surveys : Experiments on Question Form, Wording, and Context . New York: Academic Press. Scott, James C. 1976. The Moral Economy Of The Peasant : Rebellion And Subsistence In Southeast Asia . New Haven: Yale Univ ersity Press. . 1998. Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed . New Haven, Conn.: Yale University Press. . 1985. Weapons of the Weak : Everyday Forms of Peasant Resistance . New Haven: Yale University Press.

PAGE 240

240 Shefter, Martin. 1994. Political Parties and the State : The American Historical Experience , Princeton Studies in American Politics. Princeton N.J.: Princeton University Press. Siffin, William J. 1976. "Two Decades of Public Administration in Developing Co untries." Public Administration Review 36: 61 71. Silberman, Bernard S. 1993. Cages of Reason: The Rise of the Rational State in France, Japan, the United States, and Great Britain . Chicago: University of Chicago Press. Simon, Herbert A. 1947. "A Comment o n 'the Science of Public Administration." Public Administration Review 7 Taagepera, Rein, and Matthew Soberg Shugart. 1989. Seats and Votes : The Effects and Determinants of Electoral Systems . New Haven: Yale University Press. Teal, Francis. 1990. "Why Ca n Mauritius Export Manufactures and Ghana Not?". The World Economy 22: 981 93. Thompson, E. P. 1966. The Making of the English Working Class . New York: Vintage Books. Tilly, Charles. 1990. Coercion, Capital, and European States, Ad 990 1990 , Studies in Soc ial Discontinuity. Cambridge, Mass., USA: B. Blackwell. Tilly, Charles 1985. "`War Making and State Making as Organized Crime'." In Bringing the State Back In , ed. Dietrich Rueschemeyer & Theda Skocpol Peter Evans. Cambridge: Cambridge University Press. va n de Walle, N. 2003. "Presidentialism and Clientelism in Africa's Emerging Party Systems." Journal of Modern African Studies 41: 297 321. van de Walle, Nicolas. 2001. African Economies and the Politics of Permanent Crisis, 1979 199 . Cambridge: Cambridge Un iversity Press. Veen, Roel van der. 2004. What Went Wrong with Africa: A Contemporary History . Young Amsterdam: KIT. Wantchekon, L. 2003. "Clientelism and Voting Behavior Evidence from a Field Experiment in Benin." World Politics 55: 399 422. Weber, Max. [1904 1911] 1968. Economy and Society . Edited by Guenter Roth and Claus Wittich. Wittich: New York: Bedminster. . 1968. Economy and Society: An Outline of Interpretive Sociology . New York: Bedminster Press. Weber, Max, Hans Heinrich Gerth, and C. Wrigh t Mills. 1958. From Max Weber : Essays in Sociology , A Galaxy Book. New York: Oxford University Press.

PAGE 241

241 Weder, Beatrice. 1995. "Legal Systems and Economic Performance: The Empirical Evidence." In Judicial Reform in Latin America and the Caribbean, World Ban k Technical Paper Number 280 , ed. Malcom Rowat et. al. Washington, D.C.: World Bank. Young, Crawford. 1994. The African Colonial State in Comparative Perspective . New Haven: Yale University Press.

PAGE 242

242 BIOGRAPHICAL SKETCH Winifred Pankani Lindberg holds an MA in International Studies and a Post Graduate Certificate in Non Profit Management from University of Oregon and MA in Political Science from the University of Florida. Her doctoral research carried out under the guidance of Drs. erra examines the challenges of enhancing administrative capacity instit utions produce positive outcomes at the very local levels but suboptimal outcomes at the micro level. She argues for a more robust approach to social safety as credible alternatives to affective reciprocity. Her dissertation, completed in 2015 employed a m ixed method approach to data collection and analysis in Botswana and Ghana after extended fieldwork in both countries were she worked as a research affiliate with the extension services divisions of both Ministries of Agriculture from August 2010 to Dec 20 11. She has consulted for the World Bank; local NGOs in Ghana, HN Consultants, Denmark and the Norwegian based International Law and Policy Institute.