<%BANNER%>

Evaluating Technology Business Incubator as a Tool of Government Intervention

Permanent Link: http://ufdc.ufl.edu/UFE0024989/00001

Material Information

Title: Evaluating Technology Business Incubator as a Tool of Government Intervention Public vs. Private
Physical Description: 1 online resource (222 p.)
Language: english
Creator: Jang, Yongseok
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: development, economic, entrepreneur, evaluation, incubator, innovative, planning, policy, program, public, technology, urban
Design, Construction, and Planning -- Dissertations, Academic -- UF
Genre: Design, Construction, and Planning Doctorate thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Mixed results from empirical studies concerning the effectiveness of Technology business incubators (TBI) raise questions of the role of local government as a legitimate sponsor of incubators. In response to the current debate on the role of government in incubator program,this study undertakes a comparative evaluation to assess the differences both in selection of inputs and in outcomes between public and private technology incubators. The primary question examined through this research is whether publicly funded technology business incubators could be substituted by private ones. The following questions guide this assessment. 1) What are the implications of public engagement in incubator program 2) Are there any significant differences between them in the level of performance? 3) What are the major differences that might lead incubators to succeed or fail? 4) How should public incubators be guided to promote better performance? Using case study methodology, this research studies exemplar incubators of different sponsorship types to observe roles of public engagement for different sponsorships. Using National Establishment Time Series (NETS) data base, this research then conducts two-group comparison to exam whether performances of TBIs differ by sponsorships, in terms of sales,wages and employment. By synthesizing findings of the case studies and quantitative findings of performance of client companies, this research concludes, that public resources may play important roles for operation of TBIs in the US. In regards to the differences in the inputs, the differences in the status and capacity of the resources provided to each sponsor may drive the differences in services and strategies. Based on the findings about primary performance, none of the two sponsorships dominates in terms of efficiency, while several differences are observed. These findings altogether imply that the old fashioned private-public framework might not be appropriate for this subject. An alternative perspective may be that the roles of public TBIs may need to be diversified. The public sector may find an active role in utilizing large-scale capital. It also means that the public sector may step back from operating tasks, as the private sector may respond to the market dynamics more effectively.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Yongseok Jang.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Silver, Christopher.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024989:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024989/00001

Material Information

Title: Evaluating Technology Business Incubator as a Tool of Government Intervention Public vs. Private
Physical Description: 1 online resource (222 p.)
Language: english
Creator: Jang, Yongseok
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: development, economic, entrepreneur, evaluation, incubator, innovative, planning, policy, program, public, technology, urban
Design, Construction, and Planning -- Dissertations, Academic -- UF
Genre: Design, Construction, and Planning Doctorate thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Mixed results from empirical studies concerning the effectiveness of Technology business incubators (TBI) raise questions of the role of local government as a legitimate sponsor of incubators. In response to the current debate on the role of government in incubator program,this study undertakes a comparative evaluation to assess the differences both in selection of inputs and in outcomes between public and private technology incubators. The primary question examined through this research is whether publicly funded technology business incubators could be substituted by private ones. The following questions guide this assessment. 1) What are the implications of public engagement in incubator program 2) Are there any significant differences between them in the level of performance? 3) What are the major differences that might lead incubators to succeed or fail? 4) How should public incubators be guided to promote better performance? Using case study methodology, this research studies exemplar incubators of different sponsorship types to observe roles of public engagement for different sponsorships. Using National Establishment Time Series (NETS) data base, this research then conducts two-group comparison to exam whether performances of TBIs differ by sponsorships, in terms of sales,wages and employment. By synthesizing findings of the case studies and quantitative findings of performance of client companies, this research concludes, that public resources may play important roles for operation of TBIs in the US. In regards to the differences in the inputs, the differences in the status and capacity of the resources provided to each sponsor may drive the differences in services and strategies. Based on the findings about primary performance, none of the two sponsorships dominates in terms of efficiency, while several differences are observed. These findings altogether imply that the old fashioned private-public framework might not be appropriate for this subject. An alternative perspective may be that the roles of public TBIs may need to be diversified. The public sector may find an active role in utilizing large-scale capital. It also means that the public sector may step back from operating tasks, as the private sector may respond to the market dynamics more effectively.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Yongseok Jang.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Silver, Christopher.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024989:00001


This item has the following downloads:


Full Text

PAGE 1

1 EVALUATING TECHNOLOGY BUSINESS INCU BATOR AS A TOOL OF GOVERNMENT INTERVENTION: PUBLIC VS PRIVATE By YONGSEOK JANG A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Yongseok Jang

PAGE 3

3 ACKNOWLEDGMENTS There are many people who have helped m e th rough the process of completing this work. First, I would like to thank my advisor, Ch ris Silver, for all of the help, support, and encouragement he offered for to me. I also th ank the members of my committee, Joseli Macedo, Ann Williamson, and Grant Thrall, for their mental support and considerate academic guidance. Their insight and guidance allowed me to create a product that is much larger and more in depth than I had originally anticipated. Whatever contribution this work makes to the understanding of the subject is largely due to this guidance. I could not forgive myse lf if I forget to express my special gratitude to Cameron For d, a founding director of Director of the University of Central Florida Center for Entrepreneurship and Innova tion for offering every possible resource and assistance to me to access and uti lize the NETS database. I also thank the Lowe Foundation that permitted my use of the NETS database a contribution that was critical and essential to the bulk of my research. I also acknow ledge and thank the incubator ma nagers for their sincere and honest response that upgrade my researchs qualitative completeness. I owe much of my achievement to my wife, Junglim. I deeply appreciate her constant support and endless patience. Five years of doc toral work and another three years before that here at the University would have been impossi ble if Junglim was not there. Not only did she come with me to a foreign country, but she did this with magnanimous goodwill. Grace and Timothy, my children, I thank you for loving and trusting your fath er. Finally, I send a special thanks to my parents who willing gave their cons tant love and never lost their trust of me regardless of what struggle I was fighting. Th anks to all of my family for this lifetime achievement. It is theirs, not mine.

PAGE 4

4 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 3LIST OF TABLES ...........................................................................................................................8LIST OF FIGURES .......................................................................................................................11ABSTRACT ...................................................................................................................... .............14 CHAP TER 1 INTRODUCTION .................................................................................................................. 16Background .................................................................................................................... .........16Technology Business Incubator ...................................................................................... 16Snapshot of Business Incubator of the US ......................................................................16Issues of TBI as an Economic Development Program .................................................... 17Direction of This Study ...................................................................................................18TBI as a Tool of Government Interv ention: Market Failure Approach .................................. 19High Mortality Rate of High-Tech Startups .................................................................... 19Origins of TBIs ................................................................................................................20General Roles of TBI .......................................................................................................22A Tool of Technology-Based Economic Development (TED) ....................................... 22Evaluating TBI in Policy Context ...........................................................................................24Levels of Policy ...............................................................................................................24Evaluation Study in Policy Context ................................................................................ 24Problems of TBI Evaluation in Policy Context ............................................................... 26Legitimacy of Government Intervention ......................................................................... 27Public Intervention in Technology Policy .......................................................................29Criticism on the Public TBIs ........................................................................................... 31Conclusion and Summary ....................................................................................................... 322 LITERATURE REVIEW .......................................................................................................35Introduction .................................................................................................................. ...........35Taxonomies of TBIs by Sponsorship .....................................................................................36Effectiveness of Technology Business Incubators .................................................................40Discussion .................................................................................................................... ...........513 RESEARCH APPROACH .....................................................................................................53Research Framework ..............................................................................................................53Research Question: How are They Different? ................................................................. 53Focuses of Cost-Reduction Strategy ...............................................................................55

PAGE 5

5 Focuses on the Intervention Mechanism .........................................................................57Focuses on Surrounding Resources ................................................................................. 59Conceptual Framework ...................................................................................................61Taxonomy of Incubators .................................................................................................. 63Hypotheses .................................................................................................................... ..........64How the Cases were Selected .................................................................................................64Methodologies: Case studie s and Group Comparisons .......................................................... 654 CASE STUDIES .....................................................................................................................71Introduction .................................................................................................................. ...........71Public Incubator ......................................................................................................................72Regional Info rmation ....................................................................................................... 72Incubator Characteristics .................................................................................................73Overview ...................................................................................................................... ...73Establishment ................................................................................................................. .73Motivation .......................................................................................................................74Goals and Objectives .......................................................................................................74Advantage of Moving to the Research Park ....................................................................74Business Model ...............................................................................................................75Selection of Clients ..........................................................................................................77Monitoring Characteristics .............................................................................................. 77Source of Knowledge and Talent ....................................................................................79Private Incubator .....................................................................................................................80Regional Info rmation ....................................................................................................... 80Incubator Characteristics .................................................................................................80Overview ...................................................................................................................... ...80Establishment ................................................................................................................. .81Motivation .......................................................................................................................81Goals and Objectives .......................................................................................................81Location Advantage .........................................................................................................82Business Model ...............................................................................................................82Selection of Clients ..........................................................................................................84Monitoring Characteristics .............................................................................................. 86Source of Knowledge and Talent ....................................................................................87Impact Study .................................................................................................................. .........88Industry Type ...................................................................................................................89Growth in Employment ................................................................................................... 90Growth in Sales ...............................................................................................................91Summaries and Discussion .............................................................................................. 94Conclusion and Discussion ..................................................................................................... 95Cost-reduction Strategies .................................................................................................95Intervention Mechanisms ................................................................................................96Surrounding Resources ....................................................................................................97Conclusion .................................................................................................................... ...98

PAGE 6

6 5 TWO-GROUP COMPARISON ...........................................................................................104Introduction .................................................................................................................. .........104Issues About Measurement of Growth .................................................................................104Description of Study Area: Rockvill e, Montgomery County. Maryland .............................106The State of Maryland ................................................................................................... 106Maryland economy and majo r technology industries ............................................107State tax incentive for NTBFs ................................................................................ 109Selection of Study Area a nd Regional Specification ....................................................110Selection of study area ...........................................................................................110Economic geography .............................................................................................. 111Employment structure ............................................................................................ 113Tax incentive programs .......................................................................................... 115Regional knowledge resource ................................................................................116Subject Sample ..............................................................................................................117Brief review of sample incubators .........................................................................117Client companies ....................................................................................................118Client companies and type of industries ................................................................ 119Impact Study .................................................................................................................. .......121Growth in Employment, Before and After Graduation ................................................. 122Employment growth, before and after graduation ................................................. 122Employment growth rates, before and after graduation ......................................... 125Annual employment growth rate, before and after graduation .............................. 127Changes of total number of jobs over time, before and after graduation ............... 129Growth in Wage, Before and After Graduation ............................................................ 131Methodological discussion .....................................................................................132Wage growth, before and after graduation .............................................................135Total wages, before and after graduation ............................................................... 140Changes of weekly wages over time, before and after graduation ......................... 142Growth in Sales .............................................................................................................146Sales growth before and after graduation ............................................................... 146Total sales, before and after graduation ................................................................. 152Changes of total sales over time before and after graduation ............................... 1546 DISCUSSION AND POLICY IMPLICATION ...................................................................202Discussion .................................................................................................................... .........202Policy Implications ...............................................................................................................207Input Side Aspects .........................................................................................................207Cost-reduction strategies ........................................................................................207Intervention mechanism .........................................................................................208Surrounding resources ............................................................................................209Efficiencies in Outcomes ............................................................................................... 209Other Implications ............................................................................................................ ....210Research Implications ................................................................................................... 210

PAGE 7

7 Implications for Public Finance ..................................................................................... 211Role of Upper Tier Government .................................................................................... 212 7 CONCLUSION .................................................................................................................... .213LIST OF REFERENCES .............................................................................................................215BIOGRAPHICAL SKETCH .......................................................................................................222

PAGE 8

8 LIST OF TABLES Table page 2-1 Realistic prospects of TBI ..................................................................................................52 3-1 Conceptual framework: m easurement of input ..................................................................70 3-2 Conceptual framework: m easurement of outcomes .......................................................... 70 4-1 Industry type of client com panies .................................................................................... 100 4-2 Number of jobs, 2001-2007 .............................................................................................100 4-3 Sales revenue, 2001-2007 ................................................................................................102 5-1 Major hiring industries of Maryland* .............................................................................157 5-2 Major technology indus tries of Maryland 2001-2005* ....................................................157 5-3 Summary of sample incubators ........................................................................................ 157 5-4 Regional description of incubator locations .................................................................... 160 5-5 Top three hiring industries* .............................................................................................161 5-6 Summary of major employer by county* ........................................................................162 5-7 Employment and wages of m ajor technology industry* .................................................163 5-8 Tax incentive of municipalities* ......................................................................................163 5-9 Area incubators and other innovation centers ................................................................. 164 5-10 Sample Client Companies ................................................................................................ 164 5-11 Client companies by type of industry, total .................................................................... 165 5-12 Client companies by ty pe of industry, m ajor only .......................................................... 166 5-13 Biotechnical research, noncommercial a nd m edical research in public incubators ........ 166 5-14 Descriptive Statistics, employ m ent growth, before Graduation .....................................167 5-15 Mann-Whitney Test results, employment growth, before graduation ............................ 167 5-16 Descriptive statistics, employ m ent growth, after graduation ...........................................168 5-17 Mann-Whitney Test results, employment growth, after graduation ................................ 168

PAGE 9

9 5-18 Descriptive Statistics, employme nt growth rates, before graduation ............................. 169 5-19 Mann-Whitney Test results, employm ent growth rate, before graduation ..................... 169 5-20 Descriptive statistics, employ m ent growth Rate, after graduation .................................170 5-21 Mann-Whitney Test results, employm ent growth rate, after graduation ........................ 171 5-22 Descriptive statistics, annual empl oym ent growth rate, before graduation .................... 171 5-23 Mann-Whitney Test results, annual em ploym ent growth rate, before graduation ......... 172 5-24 Descriptive statistics, annual empl oym ent growth rate, after graduation ....................... 172 5-25 Mann-Whitney Test results, annual em ploym ent growth rate, after graduation ............ 173 5-26 Descriptive statistics, wa ge growth, before graduation .................................................. 179 5-27 Mann-Whitney Test results, wage growth, before graduation ........................................ 179 5-28 Descriptive Statistics, Wage Growth, After Graduation .................................................180 5-29 Mann-Whitney Test results, wage growth, after graduation ........................................... 180 5-30 Descriptive statistics, wage growth rates, before graduation ..........................................181 5-31 Mann-Whitney Test results, wage growth rates, before graduation ............................... 181 5-32 Descriptive statistics, wage growth rates, after graduation .............................................182 5-33 Mann-Whitney Test results, wage growth rates, after graduation ..................................182 5-34 Descriptive statistics, annual wage growth rates, before graduation .............................. 183 5-35 Mann-Whitney Test results, annual wage growth rates, before graduation ................... 183 5-36 Descriptive statistics, annual wage growth rates, after graduation ................................. 184 5-37 Mann-Whitney Test results, annual wage growth rates, after graduation ...................... 184 5-38 Total wages, descriptive statistics, during the residency ..............................................185 5-39 Mann-Whitney Test results, total wage, before graduation ............................................ 185 5-40 Total wages, descriptive statistics, af ter graduation ....................................................... 186 5-41 Mann-Whitney Test results, total wages, after graduation ............................................. 186 5-42 Descriptive statistics, sa les growth, before graduation ................................................... 190

PAGE 10

10 5-43 Mann-Whitney Test results, sa les growth, before graduation ........................................190 5-44 Descriptive statistics, sales growth, after graduation ......................................................191 5-45 Mann-Whitney Test Results, sales growth, after graduation .......................................... 191 5-46 Sales growth rates, descript ive statistics, before graduation ........................................... 192 5-47 Mann-Whitney Test results, sale s growth rates, before graduation ................................ 192 5-48 Sales growth rates, descri ptive statistics, after graduation ............................................. 193 5-49 Mann-Whitney Test results, sale s growth rates, after graduation ................................... 193 5-50 Annual sales growth rates, desc riptive statistics, before graduation ..............................194 5-51 Mann-Whitney Test results, annual sa les growth rates, before graduation ..................... 194 5-52 Annual sales growth rates, desc riptive statistics, after graduation ................................. 195 5-53 Mann-Whitney Test results, annual sales growth rates, after graduation ....................... 195 5-54 Total sales, descriptive statistics, during the residency .................................................. 196 5-55 Mann-Whitney Test results, total sales, before graduation ............................................. 196 5-56 Total sales of the public and priv ate incubator, after graduation, 1990-2007 ................. 197 5-57 Mann-Whitney Test results, total sales, after graduation ............................................... 197

PAGE 11

11 LIST OF FIGURES Figure page 1-1 Snapshots of technology business incubator ..................................................................... 34 3-1 Generic model of incubation process ................................................................................ 68 3-2 Incubator as a cost-reduction strategy ............................................................................... 68 3-3 Intervention Mechanism of Incubators .............................................................................69 3-4 Surrounding Resources of TBIs ........................................................................................ 69 4-1 Total number of jobs over time, 2001-2007 ................................................................... 101 4-2 Number of jobs per company, 2001-2007 ...................................................................... 101 4-3 Number of c lient com pany, 2001-2007 ..........................................................................102 4-4 Total sales over time, 2001-2007 .................................................................................... 103 4-5 Total sales per company, 2001-2007 .............................................................................. 103 5-1 Geography of technology busin ess incubators, R ockville, MD ..................................... 158 5-2 Geography of County Busine ss Personal Property tax rate ............................................159 5-3 Numbers of client com panies by zip code ...................................................................... 164 5-4 Sample distribution by age..............................................................................................165 5-5 Frequency distribution, employ m ent growth, before graduation ....................................167 5-6 Frequency distributions, empl oym ent growth, after graduation ......................................168 5-7 Frequency distribut ions, em ployment growth rate, before graduation ........................... 169 5-8 Frequency distributions, employ m ent growth rate, after graduation ..............................170 5-9 Frequency distributions, annual empl oym ent growth rate, before graduation ...............171 5-10 Frequency distributions, annual empl oym ent growth rate, after graduation .................. 172 5-11 Total number of jobs, tenants, 1990-2007 ...................................................................... 173 5-12 Change in number of jobs by year, tenants, 1990-2007 ................................................. 174 5-13 Number of jobs pe r client, tenants, 1990-2007 ...............................................................174

PAGE 12

12 5-14 Number of tenants, 1990-2007 ....................................................................................... 175 5-15 Total number of jobs, graduates, 1990-2007 .................................................................. 175 5-16 Change in number of jobs by year, graduates, 1990-2007 .............................................176 5-17 Number of jobs pe r client, graduates, 1990-2007 ...........................................................176 5-18 Number of graduates, 1990-2007 ................................................................................... 177 5-19 Number of jobs and wage effect, private 1990-2007 ......................................................177 5-20 Number of jobs and seasonal chan ge of the estim ated weekly wage, 1990-2007 ..........178 5-21 Average weekly wage of NAICS 541710, 1990-2007 ................................................... 178 5-22 Frequency distributions, wa ge growth, before graduation ..............................................179 5-23 Frequency distributions, wage growth, after graduation ................................................180 5-24 Frequency distribut ions, wage growth rates, bef ore graduation ..................................... 181 5-25 Frequency distri bution, wage growth rate s, after graduation ......................................... 182 5-26 Frequency distributions, annual wage growth rates, before graduation .........................183 5-27 Frequency distribut ions, annual wage growth rates, after graduation ............................ 184 5-28 Frequency distributions, to tal wages, before graduation ................................................ 185 5-29 Frequency distributions, total wages, after graduation ...................................................186 5-30 Total wages over time, tenants, 1990-2007 .................................................................... 187 5-31 Changes of weekly wages, tenants, 1990-2007 .............................................................. 187 5-32 Weekly wages per tenant, 1990-2007 ............................................................................. 188 5-33 Weekly wages over time, graduates, 1990-2007 ............................................................ 188 5-34 Changes of weekly wage s over tim e, graduates, 1990-2007 ..........................................189 5-35 Weekly wages per graduates, 1990-2007 ....................................................................... 189 5-36 Frequency distribution, total wages, before graduation .................................................. 190 5-37 Frequency distributions, sales growth, after graduation .................................................191 5-38 Frequency distribution, sales growth rates, before graduation ....................................... 192

PAGE 13

13 5-39 Frequency distributions, sales growth rates, after graduation ......................................... 193 5-40 Frequency distributions, annual sa les growth rates, before graduation .......................... 194 5-41 Frequency distributions, annual sa les growth rates, after graduation .............................195 5-42 Frequency distributions, to tal sales, before graduation .................................................. 196 5-43 Frequency distributions, total sales, after graduation ..................................................... 197 5-44 Total sales over time, tenants, 1990-2007 ...................................................................... 198 5-45 Changes of total sales of tenants, 1990-2007 ..................................................................198 5-46 Total sales per tenant, 1990-2007 .................................................................................... 199 5-47 Number of tenants, 1990-2007 ....................................................................................... 199 5-48 Total sales over time, graduates, 1990-2007 .................................................................. 200 5-49 Changes of total sales over time, graduates, 1990-2007 ................................................. 200 5-50 Sales per graduates, 1990-2007 ...................................................................................... 201 5-51 Number of graduates, 1990-2007 ................................................................................... 201

PAGE 14

14 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EVALUATING TECHNOLOGY BUSINESS INCU BATOR AS A TOOL OF GOVERNMENT INTERVENTION: PUBLIC VS PRIVATE Yongseok Jang August 2009 Chair: Christopher Silver Major: Design, Constructi on, and Planning Doctorate Mixed results from empirical studies concer ning the effectiveness of Technology business incubators (TBI) raise questions of the role of local government as a legitimate sponsor of incubators. In response to the current debate on the role of government in incubator program, this study undertakes a compara tive evaluation to assess the differences both in selection of inputs and in outcomes between public a nd private technology incubators. The primary question examined through this research is whether publicly funded technology business incubators could be substitu ted by private ones. The following questions guide this assessment. 1) What are the implicat ions of public engagement in incubator program 2) Are there any significa nt differences between them in the level of performance? 3) What are the major differences that might lead incubato rs to succeed or fail? 4) How should public incubators be guided to prom ote better performance? Using case study methodology, this research stud ies exemplar incubators of different sponsorship types to observe roles of public en gagement for different sponsorships. Using National Establishment Time Series (NETS) data base, this research then conducts two-group comparison to exam whether performances of TBIs differ by sponsorships, in terms of sales, wages and employment.

PAGE 15

15 By synthesizing findings of the case studies and quantitative findings of performance of client companies, this research concludes, that public resources may play important roles for operation of TBIs in the US. In regards to the differences in the inputs, the differences in the status and capacity of the resources provided to each sponsor may drive the differences in services and strategies. Base d on the findings about primary performance, none of the two sponsorships dominates in terms of efficiency, while several differences are observed. These findings altogether imply that the old fashioned private-publ ic framework might not be appropriate for this subject. An alternative pers pective may be that the roles of public TBIs may need to be diversified. The public sector may find an active role in utiliz ing large-scale capital. It also means that the public sector may step b ack from operating tasks, as the private sector may respond to the market dynamics more effectively.

PAGE 16

16 CHAPTER 1 INTRODUCTION Background Technology Business Incubator Technology business incubators (T BI) have gained a wide attention as a potential strategy of local econom ic development planning that aims to nurture technology startups as engines of economic prosperity, due to following reasons. Young firms tend to show high mortality rate. Particularly high-tech start-up s are more intensely experienci ng that problem because of the combination of their capital-intensi ve nature and their slow profit return. In other words, a bulk of capital is required to be inve sted during the pre-production stag e, also known as a stage of research and development (R&D), and it tends to take a long time to make up that initial investment. The local capital lenders, however, complain about di fficulties with investing large scale capital to such a risky item. From local go vernments perspective, this risky situation is considered as impediment that hinders the pote ntial prosperity of the local economy (David N. Allen & Richard McCluskey, 1990; Bllingtoft & Ulhi, 2005; Colombo & Delmastro, 2002; R. Smilor & Gill, 1986). To response this situati on, TBIs nurture targeted technology startups during their fledgling period wh en they are most fragile (Hamdani, 2006). Government intervention is initially justif ied in that these young potential companies suffer from poor survival rates (Eisinger, 1988). In this respect, business incubators include physical infrastructure as an essential provision to offset fi xed costs. Yet primary role of the incubator is to provide client with monitoring and busin ess assistance (Hackett & Dilts, 2004a). Snapshot of Business Incubator of the US Prim arily business incubators were the in strument of urban renewal and community development. Then in the 1970s, the incubator industry became a tool for commercialization of

PAGE 17

17 university research and technologi es (Hamdani, 2006). The m ovement to establish innovation centers was initiated by support of US national Science Foundati on. After this, universities became more enthusiastic in operating inc ubator facility as a means of technology commercialization (Hamdani, 2006). In this peri od, one of the major issues arising from activities related with technology development was controlling spillover effect. Knowledge spillover refers a production of non-appropr iable knowledge. Once knowledge generation activity is regarded as non-appropr iable, firms are discouraged to invest in innovative activities, such as R&D. This occurs due to failures in protection of outcomes of research and technology generated by innovating agents, such as New Technology Based Firm (NTBF) (Kaiser, 2002). The Bayh-Dole Act in 1980 was an approach by congress to reduce such risk of engaging in research activity and commerciali zing the outputs of federally f unded research (Hamdani, 2006) This enabled business incubators to serve as to ol of urban revitaliza tion, a means of technology transfer and commercialization. The majority of business incubators are ma naged by non-profit organizations. There are also for-profit incubators that target creation of returns to shareholders. Technology business incubators comprises of about 40 percent of all incubators. The remaining 54 percent of incubators serves business from all industries. This type is commonly called mixed-use. In terms of geography, about half of all incubato rs operate in urban areas (Knopp, 2007). Thirtyone percent of North American bus iness incubators are publicly spons ored. Incubators served by government entities cosntitute 21 percen t of the total. NBIA also reports 8 percent of incubators as being sponsored by more than one sponsor (Knopp, 2007). Issues of TBI as an Economic Development Program Econom ic development policy geared toward promoting employment growth through the innovative economy often requires strong justification for two reasons. First, policies to target

PAGE 18

18 employment growth and innovation are expens ive (Bartik, 1990). Also, job growth and innovation can take place without external he lp. For these charac teristics, government intervention for economic developm ent is justified only when mar ket failure is explicitly perceived (i.e., a situation when the market ec onomy alone fails to maximize efficiency). Similarly, a careful approach is always suggested when a government program is likely to distort market mechanisms, since public programs tend to be less efficient than private actors. In this sense, delicate choice of action is critical to make a public program successful (Bartik, 1990; Gruber, 2005). Therefore, a good policy evaluation is generally an answer to the question; why should government act the way it does, unless it is to be a pure economic analysis. Revisiting the primary inquiry of policy study, this study particularly investigates the relationship between governmen t intervention and the efficiency of technology business incubators (TBI) in the US context. Efficiency is indicated by the performance levels of tenant and graduate firms, and they are comp ared by the types of sponsorship. Direction of This Study To f ulfill the primary goal of this study that examines different level of performance compared to sponsorship, a cross-program compar ative evaluation has been utilized. The crossprogram evaluation method is defined as a fram ework that compares program performance of multiple groups by analyzing level of performance of incubatees divided by the types of sponsorship. Most of the existing studies empl oy a quasi-experimental method that compares the level of performance between a control group, incubatees and a compared group, nonincubatees, to shows the effectiveness of pr ograms. The framework of existing evaluation studies as control group vs comparable group compares incubatees a nd independent startups, the framework of this study is control group A vs control group B that compares incubatees of group A and incubatees of group B.

PAGE 19

19 The remainder of this chapter discusses the rationale for this study. The characteristics of TBI are explored as a tool of government interven tion. Discussion about issues that have gained attention from existing studies will serve as a conceptual platform for the discussion of the research question of this study. The following secti on explains the structure of this research and addresses sub-questions that re late to the primary question th at is, the relationship between sponsorship and program efficiency. TBI as a Tool of Government Inte rvention: Market Failure Approach Business incubators (BI) are wi dely accepted as a tool of governm ent intervention, due to the popular consensus that the success of small bus iness requires external help and the capital market does not provide a patient investme nt (Jenssen & Havnes, 2002; Rice, 1992). As previously noted, government intervention is a wa y in which public authority comes into play where market mechanism fails (Hackett & Dilt s, 2004a; Jenssen & Havnes, 2002). The term, market failure refers to a situation in whic h aggregated economic ex change of independent market agents end up failing to maximize the u tility of public resource s (Eisinger, 1988). In order to remedy such situation, public resources ar e often invested to ad just the imbalance of supply and demand in a way to optimize utility. In the case of high-tech markets, a major symptom of market failure is a high rate of early drop out of high-tech startups which otherwise could bring economic prosperity to a region. High Mortality Rate of High-Tech Startups A high rate of early dropout am ong small firms generally indicates a lack of managerial skills and access to capital, instead of being a sp ecial character of high-t ech startups. Eisinger attributes this problem to rati onal, risk-averse behavior and so me imperfections in the market (Eisinger, 1988). In particular, high-tech star t-ups are more intensely experiencing the problem because of the combination of thei r capital-intensive nature and thei r slow profit return. In other

PAGE 20

20 words, a bulk of capital is required to be invested during the pre-productio n stage, also known as a stage of research and development (R&D), a nd it tends to take a long time to make up that initial investment. The local cap ital lenders, however, complain about difficulties with investing large scale capital to such a ri sky item. Not surprisingly, venture companies are typically young and new to local bankers. Unknown young companie s are a type of company least favored by bankers. Conceptually speaking, capital dema nd exceeds the capacity of capital supply. From local governments perspective, this ris ky situation is considered as an impediments that hinders the potential prosperity of th e local economy (David N. Allen & Richard McCluskey, 1990; Bllingtoft & Ulhi, 2005; Colombo & Delmastro, 2002; R. Smilor & Gill, 1986). In general, small businesses are well known as job generators (Stam, Suddle, Hessels, & Stel, 2007). High-tech startups ar e usually expected to have even greater potential to generate new jobs and become innovative leaders in lo cal and regional economies (Malecki, 1991). Potential positive externality is assumed as probable, including increased income. Government intervention is justified in th at these young potential companies suffer from poor survival rates (Eisinger, 1988). A government intervenes usin g public resources to remedy the market failure and to build a patient, nurtur ing environment in which those young firms can survive and grow (Allen & Rahman, 1985; Malecki, 19 91). In this given situation, TBIs undertake particular roles to reach the goal of bolstering local economies a nd channeling government resources to help the potential economic agents. Origins of TBIs In general, TBIs are an effort to enhance innovation and entrepreneurship to respond the challenge. Historically, university-affiliated inn ovation centers came as a first type, developed by National Science Foundation (NSF) in 1973 (Cam pbell & Allen, 1987). They started as an experimental program to enhance entrepreneurship education, development of new technologies

PAGE 21

21 in existing companies, and the establishment and nurturing of new businesses" (Campbell and Allen 1987, p.180-181). Although the NSF has no direct role in the development of incubators, the model is still being adopted and replicated. Small business incubators coordinated by university-sponsored Master of Business Administration (MBA) teams provide managerial se rvices and advice, rental space, and office supplies. The service also includes networki ng opportunities, access to local venture capital, and links to research activities that are only avai lable in a university setting. The Technology Park (TP) or Science Park (SP) is the most popular models. Technology business incubators are, in many cases, services included in an SP package (Tamsy, 2007). In summary, as an intervention tool, business incubation services provide a package of real estate -based services, managerial and strategic advisement, technical assistance networks, and new, early-stage capitalization mechanisms. (Campbell & Allen, 1987; Sherman, 1999) In this way, SBI programs aim to achieve efficiency as well as a high ra te of survival among young firms. Like some public services, such as public health care, business incubator programs are provided in coordinated public-priva te partnerships. Partnership is a way to combine advanced management skills from the private sector with the risk bearing capacity and resource of the public sector (Eisinger 1988, p.22). Business incubators are often affiliated with universities to promote the effective management of public investment in technology. Phelps and Brockman (1992) found that most states limit their own roles as investors and let lo cal institutions, universities, or colleges operate the incubators. Possible models of TBIs include partne rships with public i nvestment and private management, and vice versa (Paul Westhead & Batstone, 1998)

PAGE 22

22 General Roles of TBI TBIs are ex ternal actors that work with the ma rket, firms, and ventur e capitals with an aim to help young companies survive and succeed (Ri ce, 1992). Mian puts emphasis on the role of networking business, capital and social i nputs (Mian, 1996). Soetano and Geehuizen (2005) also view incubators as intermediary agents and state that for TBIs favorable conditions are created in which firms and non-corporate institu tions interact (Soetanto & Geenhuizen, 2005). Second, TBIs play a role in coordinating actions in positive directions or in discouraging any negative outcomes to be produced to enhance form ation of young firms, which also means that TBIs are a tool to encourage creations of university based spin-offs (Lfsten & Lindelf, 2001, 2005; Tamsy, 2007). Third, TBis change circumstance to encour age all participants to collaborate and eventually to succeed (Sherman, 1999). Improving access to venture capital is considered as an important service that incubators provide. Thro ugh this service, each party finds an available provider and customer. In addition, Smilor and Gill (1986) finds roles to leverage talent, to accelerate the development of new firms, and to commercialize advanced technology. The tasks of the TBI reflect its demand-side ap proach, mostly real estate based programs, including the provision of affordable rent, shar ed office space and office supplies, managerial advice and legal consultation, and fostering networking between fi rms within the facility and within the local economy (Lofsten & Lindelof 2005; Lfsten & Lindelf, 2002; Mian, 1996). Provision of essential resources and services re flect progressive activism, which stimulates positive externalities and deter any unnecessary wast e due to possible outcomes of trial and error. A Tool of Technology-Based Ec onomic Development (T ED) As one program of small business oriented pol icies, the TBI intends not only to remedy, but also to lead local economic growth as a desirable strategy of local economic development

PAGE 23

23 (LED) (Esisinger 1988). Programs driven by TE D policy are legitimately supported since TED is accepted a mode of establishing linkages between technology and innovation and new industrial and economic growth (Phillips, 2003) Economic competitiveness is likely to be achieved through innovative activities (Schumpeter, 1934; Pa ul Westhead & Batstone, 1998). According to Schumpeter (1934), long term econo mic development is achieved through leapfrog innovation in advanced technology (Phillips, 2003; Schumpeter, 1934, 1939, 1942). Relying on Schumpeter, Phillips (2003) conten ds that industry learns to ad apt to continuously challenging market conditions by implementing new innovations in to products and processes, especially in a market situation in which technology plays a cr ucial role. Appreciating the value of knowledge, the regional economic development practitioners chose to direct their efforts towards building knowledge-based economic structure for territorial innovation (Malecki, 1997). In an effort to achieve this paramount goal, a variety of ne w concepts were introdu ced, including: innovative milieu (Aydalot, 1986), the science park (Lfsten & Lindelf, 2003b), and new industrial spaces (Soetanto & Geenhuizen, 2005). Incubators are to uted as the most effective engines that invigorates the competitiveness of th e local economy (Phillips, 2003). Phillips (2003) noted, competitiveness is the a bility of an industry to produce qualitatively differentiated products and respond quickly to mark et changes. Rothwell and Zegveld find, in concert, that an innovative economy is an engine for a country to leverage their economic competitiveness in the international economy (Rot hwell & Zegveld, 1981). In other words, hightech firms are found driving leap frog i nnovations based on unconventional technical approaches. In short, new public intervention oriented toward technol ogy-based firms (NTBF) are a popular choice since it is a strategic choice of LED (Colombo & Delmastro, 2002).

PAGE 24

24 Evaluating TBI in Policy Context Levels of Policy An evaluation study requires a thorough unde rstanding the scope in which the subject policy is structured. Hughes (1991) contends that a resear ch fra mework for both policy evaluation and analysis should systematically co nsider the organizatio ns structure of the agencies involved, the internal pressures of co mpeting policy objectives and the feasibility of designing (Hughes, 1991, p. 911). These aspects are often defined by a programs range and scope, which are establishe d by the policy agency. Therefore, misguided evaluation study may only provide mistuned implications. For instance, evaluation of one program may have impact on nationa l policy, but the implication would be fairly limited. Rather, such a program evaluation study may change program guideline that local government should follow. This is a reason that understanding the level of policy is critical. According to Hughes (1991), there are four broad levels in the policy field. National polic y is the first level, which is established by central government. Then, the second level is strat egies involving broad approaches to development. Diverse priorities an d sectoral balance are examples of strategies. The third level is policies that provide conceptual targets of the given str ategy. Forth level is programs. The program is a set of detailed action to achieve the targets of a corresponding policy with more practical measurem ent. At this step, allocation of financial and staff resources are coordinated based on the agencys strategy an d priorities. Conceptually speaking, TBI is a program which is structured in a strategy th at targets building an innovative economy. Evaluation Study in Policy Context Another aspect of an evaluation study for policy is tha t it should reflects policy context, rather than remain as providing management im plication, which can be relatively neutral about

PAGE 25

25 the political context that has in fact a greater influence on te chnology policy. Here, a primary concern of the evaluation study is to test whether a program work s or not. Quasi-experimental method is effective in showing the difference be tween a group under control of policy, and the other group that is free fr om policy influence. Although effectiveness is a one of the importa nt issues of evaluation study, a program evaluation only highlights effectiveness would be limited in serving the needs of public agency that should make a wise choice among many options. In other words, effectiveness is important, but not the only concern of public agency. Th ey also concern discri minating which program serves the stated goal better th an alternative options. An agency may be misdirected if an evaluation study disregards this political economy by only highli ghting the efficiency of one program. In this sense, following four critical issues should be addresse d for an evaluation study to be well structured in a policy context. when the government should intervene in the economy how the government may intervene the effect of inte rvention on the economy why government chooses to intervene in the way it does In regards to the above mentioned critical four issues, Hughes (1991) abstracted and reformulated these issues in a more conceptu al manner. According to Hughes, a desirable evaluation study should include th e following four aspects. the ability to compare actual performance against target performance; the ability to compare similar departments and programs; the ability to highlight key i ssues and areas of interest; the identification of trends over time and the development of specific norms or targets (Jackson & Palmer, 1989) Hughes continues to sugges t that the study framework should include performance measures which provide ex-post evaluation but with discrimination among effects (Hughes

PAGE 26

26 1991, p. 911). The framework should thus provide guidance on policy amendments that will increase the effectiveness of policies (Hughes 1991, p. 911) In short, a meaningful evaluation can be produced when program objectives and al location of resources, usually categorized in programs, are addressed in relations to each other. Problems of TBI Evaluation in Policy Context Most incub ator studies in the existing lit erature test the effectiveness of TBI by investigating the differences in performance between control group, the tenant firms and comparable group, non-tenant firms. This framework, named program vs non-program, has the potential to describe the actua l target performance. Here, the target performance is the effectiveness of the incubator program, indicated by the level of performance of tenant firms. Also most of these studies coordinates rela tively well between pr ogram objectives and performance measurement. Recall that policies are implemented under th e pressure of competing objectives and alternative programs. The existing framework, wh ich only highlights the fruits of the program, is likely to fail to provide refe rential insight about which program performs better among given active options. As previously mentioned, pr ogram vs. non-program framework only measures the effectiveness of a policy, which, for the purpo se of this discussion, is TED. According to Hughes, this framework is useful only for addressi ng the third level of LED. Pertinent issues included in the fourth level of LED, including programs that involve financial and staff resources, still remain unaddressed since different program s are not compared within the framework. Consequently, if an agency so lely relies on this framework, differentiating the different efficiency of different programs would be near ly impossible. Resour ce allocation and staffing strategies are likely to become arbitrary.

PAGE 27

27 Legitimacy of Governme nt Intervention In rega rds to the incubator issue, theorist a nd empirical researchers have focused on issues centering around the legitimacy of a policy program and the potential effects of local resources on developing an innovative econom y. Any type of public resour ce allocation, aiming to give economic agents incentives, involves the interven tion of public agencies, venture capitalist, and higher education institutions (HEI). These interventions aim to provide environmental infrastructure that nurtures econom ic agents who are less likely to survive and grow if there is no external help (Flynn, 1993). Fr om the public agents perspective, this intervention provides valuable resources that the private agents coul d not have accessed otherwise (Flynn, 1993). In this way, the utility of those resource s is expected to be maximized. The question here is the degree to which govern ment could intervene in the market as the invisible hand, which is believed to be the be st agent for maximizing the utility of limited resources. Is it legitimate for a government to determine the candidacy of market agents? How does a government know who is going to be a winne r? Does it even work? (Flynn, 1993) These are the questions asked by market advocates wh o are suspicious of the role of government intervention in general, and of intervention in incubators. Pessimism on direct intervention of public sector has been deeply rooted in neo-liberal economic perspective that uncriti cally champions market capitalis m as a single principle that determines who should participate and who should survive, and to what degree. It further supports non-interventionism that asserts laissez-faire and free ma rket as preconditions that are necessary for economies to function productively. The underling philosophy of this position is that market price is determined by equilibrium between demand and supply. Influenced by this perspec tive, government has long been generally considered as an ineffective producer. This perspective has also been evolving as skepticism on public

PAGE 28

28 organization. According to this perspectiv e, government is less knowledgeable (i.e. less effective) in information acquisition than market agents. Measures of supply and demand based on government decisions fails to reflect actual market conditions for lack of accurate information. Government may not know where to increase inve stments and where to withdraw investments. Also, neo-liberalism contends that government tends to be sluggish in responding quickly to changing market conditions. In contrast to market agents, who ar e mostly interested in making profits, government works with multiple stakehol ders who do not share a common goal. Driven by multifaceted goals and norms government actions may guide the economy in an undesirable direction. Finally, this perspec tive introduces size of public ag ent being large as one of the factors that hamper effective perf ormance. It claims that such a large organization also hinders prompt decision-making. Because of this, govern ment intervention and regulation often cause a distortion of market mechanism. In sum, conven tional economic theory is deeply suspicious of the roles of government in the economy. It states that government more of ten than not fails as economic agents and only has a negative impact on economy. This implication stems from such perception that government is less knowledgeable, less adaptable, and less efficient than market agents. Based upon empirical study, the public sector does not necessarily show significant difference in efficiency comp ared to private agents (Atk inson & Halvorsen, 1986; Borger, Kerstens, W.Moesen, & Vanneste, 1994; Levie, 1993). Some indica te economies of scale as an affective variable, rather than the ownership (Christoffersen, Paldam, & urtz, 2007). Theories also provide rationale behind such empirical findings. Focusing on economic development in conventional economic theories, government inte rvention is conditionally justified when it offsets market failure. Market failure is conceived when private market fails to achieve

PAGE 29

29 economic efficiency. In such a situation, net social benefits do not re sult. When a negative externality occurs, such as toxic chemicals releas es into a river from a factory, a government can legitimately intervene to mitigate the problem. When a positive externality is conceived, but market fails to realize the benefits, a government can legitimately intervene to maximize utility. In other words, when a good is considered as non-rival and non-excludab le, intervention by the public sector is justified in that it may enha nce productivity and efficiency in making business out of such goods (Musgrave & Muscrave, 1973). Fortunately, there seems to be a sound consensus to technology business incubator is an appropriate form of public intervention. In this context, supporters and critics bot h agree that an incubation program is only justified when it makes an increase in net be nefits in social welfare(Flynn, 1993). More specifically, due to the difficulty in measuring inc ubators impact on net social welfare, indirect measurement is used to evaluate program performa nce. Growth rate in employment, sales, and survival rates are examples of indirect measuremen ts. They may be regarded as direct in terms of the impact of incubation service, but cons idered indirect because they only arbitrarily indicate the degree of im pact on net welfare, which is the u ltimate target the program aims to achieve. The focus of this argument is that it is not conclusive whether technology incubators generate a positive impact on net welfare. Public Intervention in Technology Policy Generally speaking, governm ent intervention in technology policy has diverged from two modern economic theories. This primary theoreti cal diversion typically emulates one that has been applied to the discussion of roles of govern ment in industrialization (Heijs, 2003). On one hand, broad support for non-interventionism, laiss ez-faire (Smith, 1776), has been in place. Although, a role of government has been attribut ed as a force to stimulate and expedite industrialization.

PAGE 30

30 The importance of this dispute lies in the fact that government intervention always implies allocation of limited financial resources and strategic approaches to maximize utilities of the resources. Furthermore, justification of a certain financial po licy depends on the effectiveness of a program. It seems that genera l principles of evaluation criteria apply to this par ticular policy. In the same line of reasoning, the incubator pr ogram, as a particular program of LED, only could be justified when net social welfare increases (Heijs, 2003), while the difficulty of calculating net social welfare is another issue. As Geor ghiou (Georghiou, 2002) notes public intervention is justified when additionality is observed. Additionality is defined as achievement or improvement attributed to policy e ffect. Input side additi onality is reflected by an increase in R&D investment. This input si de additionality is, however, only a necessary condition that is expected to draw innovation, not necessarily a sufficient condition that guarantees actual return of pub lic investment. Observation of the output side additionality, which is technological and commercial additiona lity, would mean the effectiveness of the intervention. Lalkaka (2003) observes potentia l positive aspects of TBIs as following. Creation of jobs at reason able net $ public subsidy Taxes paid by corporations & workers per net $ Reduces gestation and costs of entering market Enhances chances of su ccess ( 2 to 4 times) Income, sales, exports generated for community Disadvantaged regions and groups empowered Client satisfaction at services, costs saved, time Sponsor satisfaction at return on investment Promote climate for innovation & entrepreneurship In short, government financing in technology deve lopment is one of the issues in which the same economic theory may be applied in the following different ways. First TED implies a government intervention on the private market. As justified by market failure theory, public intervention in incubator busine ss appears to be inevitable (Allen & Rahman, 1985; Bartik, 1990;

PAGE 31

31 Malecki, 1991). Second, it entails public investment, using ta x dollars. Therefore, a subsequent question is which way can one maximize utility using limited resources, i.e., tax dollars. Third, program effectiveness is one of the most importa nt evaluation criteria. The bottom line is that inviting public agents to the private market is jus tified when such an intervention first creates an incentive for private agents to participate in th e game more aggressively and when it generates a solid benefit to the communities which bear the cost for the intervention. Criticism on the Public TBIs Critics of incubation program s severely hammer the legitimacy of the program, as a waste of money and as a failure in maki ng significant differences in its reci pients level of performance. Incubator programs are perceived as an unproductiv e policy approach that is driven and fed by political delusion (Tamsy, 2007). Based on the sa me logic applied to criticisms of general technology policies, Tamsy (2007) calls for a withdrawal of public intervention because technology incubators are not as e ffective in creating pos itive externalities as expected. Instead, she suggests roles of private management wit hout public funds. She reports that German technology incubators fail to promote a sustaina ble high-tech economy be cause the number of sponsored firms is too small and even decreasing In addition, the potential of the incubated firms appear to be poor since most of them do not show satisfactory levels of technological sophistication (Tamsy, 2007). In this case, about 20 pe rcent of the companies were not even newly founded; they were an average of two years old when they were incepted to the facilities (Tamsy, 2007). Also, the incubator did not appear to provide potential entrepreneurs with motivation (Tamsy, 2007). Nearly all of the fo unders indicated that they would have started their business even if there was no incubation service available. The report indicated also that the availability of incubation serv ices does not even attr act quality firms from outside. It may be attributed to the immobility of firm founders, wh ich is usually associated with local resources

PAGE 32

32 and knowledge that the founders have already ac quired from previous work experience (Tamsy, 2007). Empirically speaking, however, no significant di fference is substantia ted between types of incubator sponsorship. There are private incubators, which are for-p rofit entities. As evidenced by the NBIA report (2003), public, not-for profit incubators we re not outperformed. Can we conclude, then, that there is a justification fo r dismissing public intervention? Such findings imply that sponsorship may not be the one that so lely determines the eff ectiveness and efficiency of incubator program. The bottom line is that fail ure in observing significant differences of what is inside may enhance the credib ility of comparative evaluation. It may be true that the performance of in cubation programs failed to make a difference compared to those companies without incubator services. Also, it is true that the incubators are under governmental influence. Nevertheless, th ere is still not enough empi rical and analytical evidence that the performance of public incubators is poorer than private ones. For now, the only legitimate implication is incubator programs may be an illusion, regardless of type of sponsorship. They are in a dichotomous world where only two groups of firms exist. One is under the influence of an incubato r and the other is outside the ga rden. Critics are mistakenly inviting private incubators to their world as saviors. However, they are more like inexperienced aliens than saviors. Conclusion and Summary The purpose of this research study is to inve stigate the relationship between sponsorship and the perform ance of TBIs. Th e efficiency of incubators by sponsorship will be compared in multiple dimensions that are prevalently employed in diverse incubator evaluation studies. A major strategy of this study is to minimize the chan ces of making biased assumptions is to utilize diverse methodologies so that each of them can address shortcomings of each other. Absolute

PAGE 33

33 numbers and figures will be still used as refe rences even though their potential drawbacks are fully acknowledged. Absolute values are sti ll important for policy agencies as the most commonly used tool for illustrating program effe ctiveness and efficiency, due to its intuitive nature, which effectively highlights subtle diffe rences that may not be outstanding by absolute figures only.

PAGE 34

34 Figure 1-1. Snapshots of tec hnology business incubator Urban areas Rural areas Suburban Mixed-use Technology businesses Servic/othr Manufactrng 1,115 191 120 0 200 400 600 800 1,000 1,200 United StatesMexico Canada

PAGE 35

35 CHAPTER 2 LITERATURE REVIEW Introduction As evaluation makes significant progress in incubator studies, majo r debates are formed around several topics. A major approach for discussing effectivene ss of TBIs is to measure the productivity of the output compared to the input of a policy program. Studies focusing on the attributes of input analyze different arrays of services which different incubators provide. The effectiveness of inputs and outcome s have played an important role in the discussion of whether incubators are actually good enough to serve their original purpos e: delivering economic prosperity to local or regional ec onomies. Literature considering i nput attributes is important for establishing a theoretical platform for the jus tification of this study s primary independent variable, the type of sponsorship. This literature review discusses the implicati ons of existing studies for each topic in order to enrich understanding of operati ng mechanism of TBIs. This chapte r is structured in two parts. The first section briefly expl ores existing studies concer ning the relationship between sponsorship and the performance of the incubator. The second section is then dedicated to the different approaches and arguments of evaluatio n issues. This section also explores the subtopics related to the effectiven ess of incubator services. Th e studies are generally in favor of the conventional expectations which favor the effectives of incubato r services, but several points are suggested as either unc lear or questionable. Existing studies are explored as samples of each approach. Limitations and utilities will be briefly men tioned to build a methodological platform of this study.

PAGE 36

36 Taxonomies of TBIs by Sponsorship Several authors discussed th e taxonom ies of the incubators. Mian (1994) questions whether sponsorship generates any difference in incubator performance. He conducted an extensive exploration of procedur al characteristics of universit y sponsored technology incubators (USTI). Although his study does not compare the leve l of performance, the work is one of a few studies which conducted a comparative analysis over different program sponsorships. Three state university-sponsored and three private university-sponsor ed facilities are compared. Thriteen Key dimensions are explored to measure practices and performance. They include their origins and objectives, organizational desi gn, governance and policy guidance, tenant performance review procedures, funding sources, targeted technologies, te nant entrepreneurs personal traits, impact on tenant firms, strategi c operational policies, se rvices, and their valueadded services for the client firms. Extensiv e interviews were conducted and survey research was supplemented. Survey data was collected from 47 respondents out of 150 contacts during 1991 and updated near the end of 1993. The conclusions were unexpected. There seems to be no significant difference in performance regardless of public or private university sponsorship. The author thus infers that regardless of the types of sponsorship, USTI in the US context has almost the same implication to the local or regional community. The conclusion is notable because it finds that almost all USTI articulate not-for-profit objectives in accordance, but both sponsorships are found to still be effective in drawing private inve stment. This conclusion still, however, appears to be premature or incomplete since it only implies a positive consequence of the community activities motivated by the in cubator, but consequences in enhancing performance of startups itself has not been empirically established. Another shortcoming of this study is that the study excluded for-prof it incubators, which might impose some different directions in their program, thus it may affect the performance of

PAGE 37

37 the startups. Since all USTIs are categorized as non-profit, the conclusion in favor of indifference might be biased from the beginning by the misleading classification. The inference drawn by this study, although seemingly attempting to tackle the public/private partnership issue, appears to suffer from its overgeneralization of the classification of USTI s sponsorships. For the purpose of making a more distinctive comp arison on the effect of private and public sponsorship, alternatively, a comparison between for-profit and non-profit would be more relevant. Since the given findings reject the mentioned di fference, this study might be an anomaly of the general theory that admits different effects of different s ponsorships. Enriching empirical study is still meaningful for testi ng the theory in applic ation to incubation res earch, if it conducts an investigation of the relationshi p between such partnership and th e level of performance. Mian revisited this study in his later works and drew a positive implication on incubation performance. These works will be discussed later. Sherman (1999) presented an additional study on the same subject of Mian. Dealing with the sponsorship issue, Sherman conducted a perf ormance based evaluatio n: a pro incubation study that affirms positive changes in the graduate s performance. The primary purpose of this study was to examine the extent to which business incubators are effective in improving survival rates of startup businesses. Three methods of study were used. Using a quasi-experimental method, the study compares outside firms (com parable group) with the clients (treated group)commencing the use of incubators between 1991 and 1996. From a national sample, a total of 49 incubation programs provided a list of clients and graduate companies. Finally, the sample consisted of 126 companies who returned the questionnaires. The macro-economic model was also used to calculate the economic im pact of the incubators Twenty-three firms

PAGE 38

38 were chosen from 4 incubators. It included 2 technology incubators and 2 mixed-use incubators. The third methodology is called sta ke holder analysis, primaril y relying on survey research methodology that measured the stake holders per ceived value of the business incubator. This study contributed to enriching suppor tive insights by illust rating the positive economic impact of the TBI program. Alt hough it widened the horizon of evaluation study by using three methodological applicati ons, the study largely reiterates the points that have already been made by existing supportive studies. It successfully reassures th e effectiveness of the incubator, but is still unable to measure the influence of different sponsorships. The major difference between the above two re presentative works dealing with taxonomies of incubation programs is that first one focuse d on the differences of input by investigating service provisions, whereas the other work viewed the same issue from multiple angles. In response to debates about sponsorship, Tamsy (2007) questions the legitimacy of public investment in incubation programs. Thereby, sh e suggests the withdrawal of public intervention and allowing private actors to deal with the issues. Tamsy claims that the incubator program is created based on a political foundation, but that the actual effect of the program is either exaggerated or not empirically substantiated. She reviewed empirical studies and highlighted the in effectiveness of the progr am. Empirical studies conducted in major three countries, the United St ates, New Zealand, and Germany, are explored as evidences of the negligible effectiveness. According to Tamsys review, literature converges on the point that the incubation progr am is either ineffective or the effectiveness is offset by the operating costs of the program. Employment gene rating power is questionable. Survival and growth rate is only marginally improved by the incubation program. Typical optimism supported by major organizations, such as the NB IA are criticized as misguiding reports, based

PAGE 39

39 on either dismissible figures that do not base any statistical signi ficance, or the significance is misleadingly represented. The conclusion is that the public role, in especially the technology business incubation program, should withdraw its financial commitment since this public expenditure fails to serve public interest. Suggest ions are made for calling for the role of private agents to handle this issue. These reports are couched in a logical structure that public intervention on the technology business incubation program is only justified when the investment, through the program, serves to effectively deliver the desired goal. It introduces a series of ar ticles that shed suspicious light on the effectiveness of business incubators. The conclusion inv iting private agents as major providers, however, appears to be premature for tw o reasons. First, although the report stands on a fairly logical structure that focuses on the e ffectiveness of the incubation program, it is not comprehensive enough. It does not account for the significance of the market failure phenomenon, which is the primary reason that the priv ate party is not motivated to play a role in this venue. Obviously, as long as a risky image of technology startups pe rsists, expecting private actors involvement is naive. Second, the report does not provide any empirical evidence that private agents do their jo b better than public agen ts. There is no theoretical or empirical ground to support the roles of private ag ents, especially when only the e ffectiveness of a program is focused on. In addition, the report does not pr ovide any reason for its distrust causing the arguments made in favor of the effectiveness of the incubation program. In sum, the report suffers from an unbalanced perspective, seemi ngly because the report does not stand on its own empirical research, whereas it was a valuable at tempt that calls forth deliberation on roles of public and private agents in incubation program.

PAGE 40

40 Effectiveness of Technology Business Incubators The definition of effectiveness for the exis ting studies refers to whether an incubator could enhance the performance of the startups. Since the perf ormance of the startups is considered as a consequence of incubator service, the effectiveness of incubators is examined using indirect measurements. The performances of startup companies are regarded as proxies that may reflect how well an incubator does its job. Overall, the findings appear to be mixed. In accordance with the affirmative works in the previous section, the potential bene fits of incubation services ar e supported, but its validity is questioned. However, the studies mostly provid e empirical evidences in favor of improved survival rate, growth rate, and the innovative activity in terms of patent generation. Indicators include the survival rate, the growth rate in terms of employment size and sales revenue, and the innovative activities of the fi rms that received incubator se rvices. Higher survival rate implies that incubators have helped the firms sur vive the fledgling period that is hazardous for young high-tech firms. Growth rate in employ ment and sales revenue indicates incubators contribute to the local and re gional economy, primarily serving the political needs that seek research backups for continuous revenue s upport for the incubation program. Innovative activities, often indicated by the number of patents made by the st artup firms, are considered as an important criterion that indicates effectiveness of an incubator. The following studies are the examples of individual atte ntion on each subtopics of th e effectiveness issue. Mian (1996) analyzed value-added contributions of UTBI to tenant firms. The typical incubator services and university related services were analyzed using the same sample that he himself used for his previously mentioned work. The author found a positive impact of UTBI on the performance of tenant firms. But as it focused only on the physical supplies and infrastructures, it failed in appreciating the va lue of intangible value-added services, such as

PAGE 41

41 psychological supports, managerial intervention, a nd miscellaneous activities including seminars, meetings, and mentoring with local entreprene ur activists. This limitation may open up a possibility for a future research on the rela tionship between tenant performance and those attributes unaccounted for. Also, the survey -based measurement may suggest another bias related to the positions of each participant. Service recipients are usually appreciative of the providers. Although cross-program comparison is absent in this study, this study not only contributes to enriching general understanding of incubators, but also provides a set of bench marks from which policy implication could be draw n for future development of incubation. Mians emphasis on the roles of universi ties on incubation performance continues and even becomes explicit in his later work: A ssessing and managing the university technology business incubator: an integrative framework (M ian, 1997). In this work, using the same data set employed in the previous works, an in tegrative framework is employed to conduct a comparative evaluation study. The study compares and evaluates the level of performance of different programs that shared the similar core objectives, while the comparison was not based on the different sponsorships. Three perfor mance dimensions are analyzed: (1) program sustainability and growth; (2) tenant firms surv ival and growth; and (3) contributions to the sponsoring universitys mission. Ge neral growth in revenue and employment was again found as being positive affectedduring the incubation peri ods. A general positive impact was observed in all dimensions tested together in the integrative methodological framework. Colombo and Delmastro (2002) conducted anothe r comparative evaluati on with an aim to answer the same question: whether TBI contributes to the formation of NTBFs and their growth. A sample of 45 Italian independent NTBFs was compared with a similar group that is consisted of a matched sample of 45 similar off-incubato r firms. The result appears to confirm the

PAGE 42

42 conventional optimism for the incubation program. Incubated firms are found to have been positively influenced by the program in terms of post-entry growth based on the number of employees. Although innovative activit ies are only marginally differe nt, the report indicates that incubated firms are more cooperative with other un its located in the facility and university than off-incubator firms. Output measures also in dicates that the on-inc ubator firms show better survival and growth rate. Siegel and Westhead (2003) tested a subissue: the common belief which expects a positive influence on innovative activities relate d to research and development, instead of conducting comprehensive evaluation. Using th e conventional quasi-experimental research methods, this study compares the re search performance of firms located in the university science park, and firms that are located outside of scienc e parks. In conclusion, this study reaffirms the positive influence of the incubation program with stating slightly higher research productivity from park-located firms than those firms located outside the park In terms of its methodological shortcoming, using data collected in 1992, this study indicates its limitedness due to its count-oriented analysis. While consid ering the number of patents, it may dismiss the different qualitative significance of individual patent s. It suggests as a remedy to conduct a productivity-o riented analysis, measuring the actual size of revenue, created by technology innovation. Among the sub-issues of effectiveness, another interesting study was done by Westhead and Storey (1994) on the conditions related to the geographic location of the firms. The critical feature they suggested was the collaboration with universities ba sed on the theory that assumes linkages between knowledge and pr oductivity. For instance Appold (1991) finds a tendency that industrial research la boratories cluster in large urban areas with a good university.

PAGE 43

43 Lofsten and Lindelof (2002) conducted research on the difference of SP firms and out-ofPark firms in establishing such links, especi ally for those firms called New Technology-Based Firms (NTBFs) in Sweden. Differences are found in three areas: sa les, employment, and profitability. Generally speaking, the study rea ffirms the positive impacts of SP on NTBFs in sales and employment, as found in their previous work (Lfsten & Lindelf, 2001), but was not able to make sure its role in increasing profitability. Exte nding the previous study, the major purpose of this study was to identify any added-va lue of SPs to NTBF, distinguished by other firms outside SP. The research compared th e level of the formal linkages bridging among onsite-NTBFs and those for off-park-NTBFs. Their findings showed that firms located in SPs are more effective in establishing formal linkages with universities than those outside SPs. However, the NTBF in SPs were found to not have so mu ch different in channe ling the advantage to making profit. Thus, effectiveness of incuba tor programs on improving profitability remains unanswered, in this study. In their later study, al so looking at the question of whether incubation service is effective in stimu lating innovative activity(Lfste n & Lindelf, 2003b), they found no significant difference in profitability between the NTBFs on and off park firms. A question raised by this study concerns the gap between the high level of collaboration among firms and high profitability. Recalling the theoretical foundati on of the incubator program, innovation is one of the important goals for university affiliated incubators to deliver. Based on the agglomeration effect, firms tend to be better off when they are located closely to each other so that they can easily collaborate. Technology based economic deve lopment is a strategy that aims to build a nurturing circumstance for many companies to find an attractive place to work together in the targeted neighborhood, eventually for them to c onstruct a leap frog development of technology.

PAGE 44

44 Two hypothetical reasons support the argument that incubators would generate the agglomeration effect. First, firms would take advantage by locating closel y with other firms in SPs since they can easily work together. Sec ond, incubation facility serves as a node that transforms pure knowledge and technology genera ted by affiliated universities to commercial products; young and small companies take the role (P. Westhead & Storey, 1994) In regards to the first hypot hesis, collaboration among the onsite NTBF is a critical assumption. While the previous study by Lofst en and Lindelof (2002) found a higher formal linkage among on-site-firms, the level of collabora tion may not have been very high. It may be due to weak informal linkage, through which t acit knowledge is transf erred. One explanation is that on-site-NTBFs have little motivation to co llaborate due to the hete rogeneous character of the tenants (Lfsten & Lindelf, 2005). The wi de range of specialty deters development of collaborative relationships. The agglomeration effect may be only found for those with shared interests, while this proposition is still empirically untested. Sec ond, in regards to the spill-over effect of universities, intera ction between on-site-NTBFs and the universities are usually found to be generally low, while the level of interaction is slightly higher than the relation with off-siteNTBFs (Macdonald, 1987b; Massey, Quintas, & Wi eld, 1992). The fact that NTBFs have only limited access to the academic resources ma y be one of the reasons for this. However, this unexpected low level of interaction does not appear to hinder growth of onsite-firms. As indicated by previous studies, sales and employment growth are indeed found higher from on-park NTBFs than outside firms. Backed by Link and Scott (2006), the NTBFs located in the university affiliated SPs shows a bette r survival and growth rate, as proved in other studies. They found that the number of universitie s formally affiliate with the park is positively linked to faster growth. Also, co mpany location being closer to th e park is found to be effective

PAGE 45

45 for firm productivity, echoing the existing st udies (Audretsch & Feldman, 1996; Feldman & Audretsch, 1999; Henderson, Kuncoro, & Turner, 1995) This finding should not be interpreted as university managed parks being more effectiv e, since they also found that parks managed by private organizations drive a faster growth rate th an the park operated by uni versities. One of the findings of this study to note here is that technology focused incuba tors show a fast er growth rate, especially in the information technology industry, than parks without a technology focus. The other interesting finding is that slower growth is reported from SPs than with incubators, when compared to the SPs without incubators. Also university ownership, whet her the university is private or public, does not show any statisti cal significance in influencing on the firm performance. Reaffirming that park formati on does increase efficiency of innovative activity, the study concludes that closer proximity, private operation, and a technology-focused strategy are three elements that make a difference for the firm performance. As noted by Link and Scott (2006), the agglomer ation effect may be found on a site if one serves a narrower target businesses, such as the bio-tech industry or information technology, so that the companies can find a stronger incentive to collaborate. The other reason that high-tech incubators show a weaker agglom eration effect may be the charac teristic of their market being global rather local (Lfsten & Lindelf, 2003a). Obviously, world-wide, long distance trade range dismisses the significance of physical di stance from one central market, which was assumed to affect location pattern for certain traditional industries. The conventional location theory rests on the notion that di stance based variations in industrial locations are affected based on the characteristics of products and foods for each industry. In this logical structure, delivery cost, determined by distance, was an influential fact or that affects the deci sions of a location of a company. This variable may not be a strong dete rminant to the high-tech industry as it was to

PAGE 46

46 the traditional industries (Mar kusen, 1986). Also, a formal li nkage, found more with on site NTBFs, may not have such a meaningful signifi cance since the trust based collaborations tend to work better through informal relations, which is used as channels for exchanging tacit knowledge. These findings all look for an explan ation of location strategys effect on NTBF from the agglomeration effect and those from collaborating with other firms. The following study on the other hand focuses on a positive effect that companies expect by locating in the incubator facilities. The next study highlights the significance of collaboration with universities, as opposed to collaboration with other tenant firms. It also confirms that access to new customers has little significance for startups of a high-tech industry. Fergusons primary hypothesis is that value-added services w ould have a positive effect on incubators provided to tenant firms (Ferguson & Olofsson, 2004). To test this hypothesis, he conducted research comparing on and off park NTBFs to show possible difference. Among 66 firms that were studied, 30 on park firms and 36 off park firms were grouped together to be compared. One interesting finding of this res earch was that location advantage occured only when cooperative relationships between universitie s and firms are establis hed. Other advantages expected from locating in SPs, such as better access to new customers, were found to be little associated with changing firms performance in re gard to growth both in employment and sales. However, this research falls short in addressing the university services in detail; whether firms productivity is affected by resear ch collaboration with faculty me mbers or access to lab facility remains unspecified. Again, this research does not take into consideration one important causal factor for agglomeration effects: cooperation among the tenant firms. While the focus is different, it would be appropriate if the study looked into how cooperation among tenant firms was

PAGE 47

47 significant in connection with othe r variables. In addition, th is finding conflicts with other research that suggests that coope rative relationships between univers ities and tenant firms are, in reality, rarely found. This may imply that the small amount of incubators which contribute to the growth of NTBFs is attributed to the diffi culties for incubators to maintain cooperative relationships with universities. In general, this study reaffirms the effec tiveness of incubator services in improving survival rate of young firms, while it reserves its position on the effectiven ess of incubators in improving growth in sales and employment. In ot her words, incubators are effectively helping young firms to penetrate the hazardous fledgling period that usually suffocates most of the NTBFs, but meanwhile it may not necessarily deli ver prosperity to those firms. Additionally, this research interestingly points out that the n eeds of on-park-NTBFs appear to be met, even though the range of needs are wider th an those of off-park-NTBFs. The most recent study by Squicciarini (2008) employs econometric modeling methods as a way to calculate the marginal effect of incubation programs on tenants innovativ e activities. The comparative framework is the same as those typically used in the other studies, comparing science park firms and out-of park firms. Fo rty-eight science park firms were observed to measure the influence of SPs on their patenting activities, using a data set collected between 2002 and 2003. The study tests a primary hypothesis that NTBFs in a successful park with accomplishing innovation-supporting tasks would be better off in patenting performance after joining the SPs than the matching firms outside SPs. The study compared the temporal dynamics of tenant firms and outside firms to show if there is any difference created by SPs. It indicates that tenants demonstrate relativ ely better performance, keeping a higher patenting rate during their life cycle after they register their first patent. Based on th e findings, the analysis suggests a

PAGE 48

48 positive impact of science parks on innovative performa nce of tenant companies. In general, this study reaffirms the optimistic view that incuba tion programs help tenant firms perform. Another report conducted by the National Bu siness Incubation Asso ciation (NBIA, 2003) was a comprehensive investigation that attempted to figure out a set of service items to improve tenant outcomes. Primary and secondary outcome s were regressed with the diverse services provided by incubators categorized as the best-in-class. Interestingly, no predictive relations were found to be significant as factors that yielded better performance. Only adjacency to a major institution, such as university or research centers, is found as a common aspect shared by those best-performing incubators. Based on th e regression analyses, none of the incubator services and assistances were found to be signi ficant in enhancing primary outcomes, i.e. growth in employment and sales. Incubato r assistance, however, was found to somewhat positively contribute to firms attracting extern al research grants and acquiring intellectual property. The report attributes this low predictive power to perhaps a small number of incubator samples. Besides, the indivi dualized needs of clients were assumed as one of the possible reasons that one big trend has not emerged significantly. This report is an interesting attempt in that it focuses on the relations hip between a list of assistances of incubators and the level of perfor mance. However, the major drawback of this report lies in, as indicated by aut hor, the lack of consideration of other factors, in cluding a larger context, such as location effect. Also, factors of firm resources, such as CEOs experience or firm sizes, are largely disrespected, which are pe rhaps one of the most important factors that predetermine the quality of outcomes. Another interesting remark on the lower predictive relationship between incubator se rvices and primary outcomes is that primary outcomes may be way beyond the incubators reach and it may have more of an impact in a larger context. The

PAGE 49

49 role of the incubator may be redefined as a leveraging agent, but with more limited goals, focused on secondary outcomes such as helping firms to acquire more intellectual properties, which is a precursor to the real business outcomes. This rede finition may serve as a topical platform on which more realistic evaluati on criteria could be conceptualized. Especially regarding other factors th at might affect business performance of client forms, Westhead (1997) conducted a comparative evaluation using the same framework: program vs non program. However, the subject of this study is science parks in the UK, not incubators. One hundred eighty-three SP compan ies participated in interviews, and were compared with 101 non SP organizations. A uni que contribution of this study is that performance was compared within groups with sim ilar characteristics: in dustry, ownership type of the organization, age, and location of the organi zation. Geographic variation is also taken into consideration. It also uniquely pays attention on ownership type. The conclusion was, as stated by the author, somewhat unexpected in that he fa ils to support the positive impact of science parks on the performance of tenant firms in improving investments in R&D and improving degrees of technology diffusion. There was no st atistically significant di fference in the number of patents or applications betw een science parks and off-park firms. Performance in software copyright and application also does not show any improvement from client firms. A number of new products and services were virtua lly observed at the same level. Another study that failed to find program effectiveness, by Chan et al (2005), conducted multiple case studies and ranked 9 value-added services of SP, using the business development data of six technology startups in Hong Kong SP. To capture the incubator effect throughout different stages of venture development proc esses, this study suggests new formation on the evaluative framework. In-depth interviews follo wed and analyzed the development processes of

PAGE 50

50 participating companies and the impact of incuba tors on the tenant performance. This report provided insightful illustrations wh ich highlight the contribution of incubators: the different roles for different stages of firm development process. However, it also remains as an anecdotal, descriptive study, illu strating how firms are motivated, even de tailed with educational experience during the founders life time that might have mo tivated firm foundation. It concludes with presenting the ranked significance of added-value incubator services. It also subconsciously studies the impacts of the founders upbringing to be a business fron tier, the relevance of which is questioned. Subject-o riented assessment methodology is also open to criticism in the sense that beneficiaries of public services tend to feel obliged to be appreciati ve rather than to be objective. Also, no comparative reference has been found to have further biased service assessments. Based on subject-orient assessm ent methodology, free or affordable rental provision is viewed as the most significant contributing factor. An interesting trait of this study is in that it differentiates incubator services from other programs, such as subsidy-oriented, or tax incentive programs. Another in teresting point is that the c ontribution of incubators in encouraging networking and cluste ring is found to be insignifican t, which is also empirically backed by Acs et al (1992; Acs, Audretsch, & Feld man, 1994). It is indeed interesting that this finding opens the question on the incubators role as a technological hub, which is one of the fundamental philosophies just ifying public investment thr ough the creation of incubation programs (Colombo & Delmastro, 2002). This findi ng is a reaffirmation of Macdonald (1987a), who questioned the cluster effect and the roles of the incubator as an information hub based on the same premise. The overall stance of Chan et al (2005), however, is rather reserved rather than advocating or challenging, in that it balances some contributive role of the incubator, while criticizing existing perspectives.

PAGE 51

51 Discussion The questio n that remains to be answered in whether or not TBIs work. TBIs may have a moderate impact on job generation and on the multip lier effect than do incubators dedicated for general small businesses. Direct impact on empl oyment tends to be limited since the operation of technology businesses does not rely on many employees. Indirect employment growth, however, may be fairly expected. Even in terms of sales growth, policy makers should note that some businesses take a long time to start making profits: the magnitude of the programs impact may vary by business sectors. Biotech startups are examples of businesses that move slowly. Finally, TBIs do help survival and growth of young high-tech startups, but usually those incubators in metropolitan areas te nd to succeed better in an a gglomeration economy. All of the findings discussed in this secti on are summarized in the Table 2-1

PAGE 52

52 Table 2-1. Realistic prospects of TBI Expected Evidenced Job creation At reasonable rate Unlikely Taxes By corporations & workers Maybe by corporation Gestation and costs of entering market Reduces Maybe, vary by industry Chance of success Enhances In terms of sales Income, sales, exports Generated for community N o evidence about income Empowerment Disadvantaged regions and groups Unlikely, agglomeration advantage is essential and sk ill-dependent Client satisfaction At services, costs saved, time Likely but nothing to do with business outcomes Sponsor satisfaction At return on investment Likely if it is for-profit, Public sector may be disappointed at poor employment creation Climate for innovation & entrepreneurship Will produce May create positive image to a region

PAGE 53

53 CHAPTER 3 RESEARCH APPROACH Research Framework Research Question: How are They Different? After acknowledging the inconclu sive characteristics of the TBIs, the prim ary question of this study is how different forms of sponsorships actually affect the efficiency of TBIs. The research questions of this study responds to the existing pessim ism on public sector that public intervention often results in lowe ring program efficiency. Why sponsorship is important? More often than not, sponsorship tends to create va riation in organizations objectives and services accordingly. Allen and Rahman (1985) found th at services critical to small business development such as, business taxes, marketi ng, health and benefit packages, accounting and computing and information service are less likely to be available in incubators created by the public sectors than in those created by the private sectors. Also, they found that private sector incubators are more active in providing gene ral business services (Allen & Rahman, 1985). These findings reaffirm the implication of different sponsorship, but they stil l need to be updated. Although the research is tw enty years old, theoretica lly speaking, sponsorship is still regarded as a critical factor that changes the characteristics of organizations. Potential difference may be found from service inputs and outcomes. Phillips (2003) contends that publicly-sponsored incubators are commonly concerne d with increasing the number of jobs in the local and regional community, primarily aiming to making positive community economic impacts. While empirical studi es fail to verify the widely accepted belief that small business incubators are job creators (Campbell & Allen, 1987), public sector TBIs prefer working with industries th at have a high potential as job generators, accordingly, rather than industries likely to make higher sales revenue (Allen & Rahm an, 1985). This is an example

PAGE 54

54 of a case that organizations objective drives the operation of program. Alternatively, the private sector tends to target industries that are more likely to generate higher sales revenues to for-profit incubators. The bottom line is that not enough studies have e ngaged in this discussion either reviewing differences of input aspects driven by sponsor ships or productivities of outcomes drawn by them; hence, it is premature to make a conclusion. In addition, such observational studies tend to pay attention overtly on the se rvice inputs provided by TBIs, while th ere are lots of ot her factors that affect business performance of the client compan ies. Consequently, they failed to address why some services are provided in some incubators. The only assumption is that the service profile is a result of choice decided by the managers call. This position does not offer any logic that explains different managerial styles of different sponsorship. The only explanation is that internal differences, such as given objectives, are th e only factors that make di fference. In reality, many other aspects are different. As a consequence, most research treats incubator services as autonomous and independent, this may mislead all of the outcomes as fruits of incubator process. This perspective is illustrated in Figure 3-1. In this model, incubator r ecruits a group of potential companies that have different outfits. After th ey are incubated, they become more desirable drawn by the characteristics of incubator. So, if incubator drives the client companies to become red, the client companies becomes close to it. Fo r instance, private incubators are usually limited in making large investment. They might also have difficulties earning respect from local innovators community while government agencies easily collaborate with them for their authoritative image. Most of all, as stated ear lier, not many studies investigated difference in outcomes generated by TBIs with different sponsors.

PAGE 55

55 An alternative approach of this study to study input side aspect is to look at factors that make TBI works, rather than focusing only on the individual services provisions of TBIs, and examine models that suit constr aints and capacities given to TBIs of each sponsorship. As briefly introduced, there are thr ee groups of factors that make TBIs work. Following section offers short reviews on the three perspectives that suggest differe nt factors as ingredients of success of innovative economies. In short, this study questi ons how TBIs with different sponsorship are different, in terms of inputs and outcomes. Focuses of Cost-Reduction Strategy The f irst perspective is related with its cost reduction strategy. Recall that the role of an incubator is two fold, while they are both common strategies to cut the cost. First, an incubator contains the cost of failure by providing supplies, such as infrastructure, at an affordable rate (Hackett & Dilts, 2004b; Lalkaka, 2002). Second, an incubator is an agent to utilize resources and services to help client firms thrive. Therefore, the basic stra tegy of an incubation program is for incubators to bear the cost, otherwise NTBFs suffer in their beginning era (Knopp, 2007). The primary rationale of this perspective on incubation business is that locating in the incubator facility may reduce production costs for NTBF in the fledgling period. This perspective explains roles of the incubator as a means to offset market failure of technology related industries. The original version of neoclassic econo mic theory is characterized as a conceptual framework that assumes economies to be static. Under the paradigm of the conventional economic perspective, microeconomic s provides predictive in sight, and individual participants are presumably able to make rational decisions, either based on calculation or experience. Utility maximization is achieved through optimization of cost and production. Cost reduction is one strategy through wh ich rational individuals strive toward achieving the ultimate goal, maximizing profitability. On e of the critical fact ors that determine cost effectiveness is

PAGE 56

56 location. Rooted in market failure theory, tr aditional location theory is a good reference for understanding this strategy. Traditional location theory, developed by L sch (1954) and Weber ( 1929), views location decisions as consequences of r easonable decision making processes, using cost related on spatial variation, based on the delivery cost between major market and plant, and production cost determined by resource availability. The major ma rket is where most consumers are aggregated. Production cost depends on the localized availabil ity of resources. Location is determined as where the business maximizes the excess of gross profits over gross oper ating costs (Oakey & Cooper, 1989). For high-tech industry, however, traditional lo cation theory loses it s relevance. The significance of consumer location is largely diminished since they are rather well dispersed, thus, central market assumptions may be inappropriate (Oakey & Cooper, 1989). Localized customers cannot be a proposition for high-t ech industry; many customers are often from aboard (Oakey & Cooper, 1989). Also, distance-determined deliv ery cost loses the ea rlier relevance since products of high-tech industry tend to show high transportability due to its small size and high profitability (Oakey & Cooper, 1989). Under the paradigm of conventional economic theory, location theory, in application to incubator research, is therefore transformed as a conceptual framework that focuses on costreduction advantages expected from moving in a facility that make available a series of unique provisions from which client firms can reduce ne t production cost. Therefore, the focus is on whether incubator services can exclusively channel those affordable services and resources, which means otherwise NTBFs expects those servi ces that they are less likely to obtain. Couched in the paradigm of ne o-classical economic location theo ry, researchers particularly

PAGE 57

57 focus on cost or availability of premises (Sternberg, 1990), and service provisions with necessary business skills for NTBFs (Collinson & Gregson, 2003; Flynn, 1993). This perspective is underlying th e rationale of the r eal-east model of incubators. Based on this perspective, a critical task of incubators is necessarily providing a series of affordable infrastructure. Office space is offered as low as, or at the competitive level in comparison to the market rate, as a way of reducing fixed costs. Incubators also provide office supplies, furniture, shared conference rooms and even clerical services. From an incubators perspective, these services are provided to maximize cost reduction. From the client s viewpoint, they can expect advantages from moving in a spec ific facility, which is why inc ubators were used for this study, as opposed to a location. As illustrated in the Fi gure 3-2 the major task of TBI is to provide client companies with affordable services exclusively. Of course, there are counterarguments on this view. Empirical studies state that NTBFs are willing to pay even higher rent if they can gain something more valuable by moving to a particular area (Saxenian, 1985). This includes the image of a business with respect to scientific expertise (Tamsy, 2007), when a location has such image. Moving into a university incubator affiliated with a large research university is one of the cases. There is an argument suggesting that cost became a signi ficant factor in the second stage, not for the companies in first stage, which are the subject of the incubation program (Schmenner, Huber, & Cook, 1987). Focuses on the Intervention Mechanism One chronic problem of NTBFs, in regards to in cubator issues, is the old issue of liability of newness, alluded by Stinchcombe (1965), whic h refers to a genera l tendency of high mortality of young companies. In reality, high-t ech industry indeed reports a high rate of early drop out. As discussed earlier, to reduce mortality rates of NTBFs is among the many reason TBIs intervene in the private market. Based on the perspective of ecology theory, such

PAGE 58

58 challenges are much more complicated and adve rse due to a greater popul ation density of the real world market (Carroll & De lacroix, 1982). This view, empirically substantiated later by Smilor and Feeser (1991), finds population as a chal lenging factor for new co mpanies. It asserts that the intervention mechanism spoils young firms and, thereby, only results in weak firms that are less likely to survive in their later phase due to the challenges of a large number of competitors. In other words, spoiled firms will no t be adaptive to handle such higher uncertainty generated by collective actions of a large number of compet itors. TBIs, under hypothetical circumstance, only twist the natural market pro cess when selection is determined by sponsored organization with deliberate care (Flynn, 1993). In a real world situation, higher population density implies greater uncertainty, which w ill easily suffocate ill-muscled incubated organizations. The artificial selection bias of the interv ention mechanism means th e entire process of incubation is not limited as a literal reference of the selection process. In the incubator system, this intervention mechanism, however, can take pl ace through three sub-seque nces, as a series of nurturing processes of incubators. First, the initial intervention occu rs through the selection process. Incubators intervene in markets by se lecting participants thr ough an admission process (R. G. Phillips, 2002) through which TBIs select most potential profiles. The rate of admission may supplement the intensity of the admissi on process. Intervention occurs through a monitoring process during their residency. The cas e studies, in the follo wing chapter, provide comparable observations influenced by differe nt organizational objectives imposed by the sponsorships. Monitoring criteria and rigorous ness may affect the directi on and magnitude of client performance. Through monitor processes, the inc ubator provides formal and informal advice to

PAGE 59

59 encourage firms to make progress. Conceptu ally speaking, monitoring processes serve as channels through which external information, guidance, and managerial pre ssure are offered. In this sense, it is reasonable to assume that more frequent monitoring may have a stronger impact, either stimulating or deterring the progress of clie nts, although generally it is assumed to be a stimulant. The type of monitor process may refl ect the intensity of such procedures. A formal monitor that files progress records may legitimatel y imply that incubator managers put a greater effort into conducting monitoring tasks than othe r managers who do not. Besides the frequency and types of monitoring methods, the number of graduation criteria a nd graduation rate may reflect the intensity of interventi on and of incubators on clients. Lastly, graduation criteria, as an intervention process, may have an impact on firm performances with the same reason for the monitoring process. As illustrated in the Figur e 3-3, the TBIs are agents intervene market process through selecting, monito ring, and providing guidance, including imposing graduation criteria. Focuses on Surrounding Resources Finally, based on the perspective of the evolut ion ary theory of economics, the case studies discuss strategies to making use of surrounding resources that leverage innovative activities to emerge and spin. In terms of evolutionary th eory, economic growth is the result of the coevolution of technologies, firms, industry structures, supporting institutions, and governmental systems (Nelson, 2006). Evolutionary theory high lights the role of non-market institutions, such as universities, research orga nizations, and government program s. Also, the labor market, education system, financial institutions, and regu latory structures altoge ther shape the dynamism of innovative economies (Nelson, 2006). Especially for the companies at the first ph ase, by the time second stage expansion is reached, infrastructure is the most influentia l factor on organizational survival, along with

PAGE 60

60 sponsoring organizations and availability of regional R&D capital (Flynn, 1993; Nelson, 2006). Surrounding resources include the availability of research labo ratories, area wide technical expertise (Carlton, 1983), the availa bility of skilled labor, or ex isting businesses where informal social interaction could be formulated a nd operated (Collinson & Gregson, 2003). More importantly, in terms of structural persp ective, innovative economy emerges where the synergistic effect of those res ources is likely to be generated. Metropolitan areas, where the agglomeration effect is expecte d, or a university town might be strong candidates. A technology business incubator may take advantage of such neighborhood because such areas allow fluent exchange of tacit knowledge. For this reason, th e agglomeration economy is often ranked as the most important success factor (A betti, 2004). It is not surprising that NTBFs even prefer paying higher rent in exchange for locating in a metr opolitan area over saving rent for locating in an unattractive area in a business sense, such as a rural area (Tamsy, 2007). Another focal point under this paradigm is th at surrounding resources al so serve as sources to cut redundant expenditure. In regards to cost saving, one must s ee the strategies of incubators to cut costs related to human resources. NTBFs are operated, more often than not, by one or two technology or business experts. Sometimes, an NTBF enters the incubator only with a business idea that needs supportive technology. Obvi ously, businesses lacking managerial/technical expertise are highly likely to fail. Also, availabili ty and cost of technical labor is one of the most important factors in location decisions for NTBFs (Premus, 1982). Many times, technological developm ent itself is an issue for NTBFs to solve. For such cases, one of the important roles of incubators is to introduce new tec hnology or license existing technology. NTBFs take cost advantage from using already develope d technologies and knowhow by being a client of an incubator. Access to a university is important in this sense. It

PAGE 61

61 includes access to library facilities, research faci lities, student recruiting as research associates, and a positive image (Premus, 1982). Ferguson also supports univers ity-associated image benefits as having positive impacts on grow th and success (Ferguson & Olofsson, 2004). The Figure 3-4 illustrates surrounding factors of technology busine ss incubators. For this reason, community economic developers are often suggested to pay attention in addressing contextual and linkage issues (NBIA, 2003). The case studies focus on strategies that attempt to maximize external utilities, as its unit of analysis is the incubator rather than regional economic structure. One incubator might not be located in such an attractive area. This circumstance itself may hinder success of such an incubator. How could the incubator offset such disadvantage? Do different sponsorships have anything to do w ith such strategy? Conceptual Framework This study com pares the input side aspects and the outcomes of the programs. Three groups of factors are examined as measurements of the input side measurement. The first group of factors is those related to cost -saving strategies. The second group of factors is related to an intervention mechanism that includes the select ion criteria, the thoroughne ss of the monitoring activities, and the polic ies of graduation. The third group of factors includes strategies for establishing linkages with extern al resources; such as universi ties, research facilities, and governments. The major focus of the investigation of these factors is to highlight the different approaches that each incubator employs to make the program work, rather than showing the quantitative differences in it. Two reasons jus tify this rationale. First, quantifying individual services may not be appropriate since the size of th e comparable sample is too small. It is very hard to find purely private for-profit and purely p ublic non-profit technology incubators. Also, it is even harder to find two of th em in the same location at which the influence of external factors can be controlled. Second, the quan titative aspects of individual se rvices, such as the number of

PAGE 62

62 social meetings and the number of fax machin es, are by and large predetermined by incubator policy and capacity. A recent trend in the incuba tor business has been to provide cubicles which emulate the garage model because it is believed that this architectural style may enhance collaboration among client companies. On the other hand, in many cases public incubator programs work on the large-scale in frastructure, such as wet labs and large conference rooms. Presumably, such provision is made available beca use the public sector is able to utilize largescale investment; which is also an option that private agents are likely to emulate. In this sense, comparing quantitative differences predetermine d by different sponsorship may not generate meaningful inferences. Rather, an appropriate approach is to investigate the alternative strategies for each incubator in or der to respond to the challenges that arise from the constraints that different sponsorships have predetermined. In regards to the outcomes of incubators, the performance of tenant and graduate firms is a widely accepted indicator. Among many variables, this study examines the primary outcomes, such as employment, wages and sales revenue ; rather than secondary outcomes such as intellectual property or financial improvement. W ithin this framework, incubator efficiency is measured from multiple angles so as to reflect the different objectives; the significance of these angles was elaborated upon in the previous sectio n. The first approach is to measure indicators in aggregated forms by the types of sponsorshi p. The aggregated number of jobs, wages, and sales revenues will indicate the capacity for spon sorship. Aggregation may confuse the variation in capacity that may have been predetermined by the types of industry and the size of firms, which may have existed in effect even before th e firms were incorporated into the programs. Nevertheless, this method can still be said to be justified, based on the premise that the level of performance of an incubator measured by aggreg ated business performan ce is a result that is

PAGE 63

63 affected by a particular selection strategy of the incubator. In ot her words, when a research study measures the performance of a profile of client co mpanies, all of the stra tegies and choices that the group of client companies employed could be considered to be dependent on the incubators choice. Table 3-2 summarizes the con ceptual framework of this study. Taxonomy of Incubators Only four major types of s ponsorship of TBIs are obser ved from a mong a handful of studies (Kuratko & LaFollette, 1987; R. Smilor, 1987; Temali & Campbell, 1984) that used taxonomies to classify incubators based on their financial sponsorship. Although they appear to be useful as a template for this research, modification seems to be unavoidable, for data availability. The four types of sponsorship-based taxonomies refl ect degrees of involvement of private and public agents: Publicly-sponsored, Nonprofit-sponsored, University-sponsored, and Privately-sponsored. These classifications, however, may not be re levant for more recent incubation programs. Typical type of public incubator serves as non-profit organization. Recently more popular type of public incubator is a public -private partnership. Also, many non-profit organizations are founded by government but operation is delegated to an organization dedicated to this business. Such organizations are independent from gove rnment in operation of fund, but still their management is driven to deliver the pub lic goals and objectives imposed by founding government. Even for-profit private organizatio ns enter into the program as main operating organization. This type of partnership seems to be a result of public sectors expectation of higher efficiency from delegating private for-profit organization as a program manager. In sum, financial influence of founding sponsorsh ip is in reality pretty much scrambled by such partnering strategies, thus lost magnitude of implications that it used to have. This research alternatively investigates two in cubators of two radical extrem es of sponsorship: private for-

PAGE 64

64 profit vs. public non-pro fit. Hopefully, investigating thes e two radical examples would produce insightful inferences on the influe nce of different sponsorships. Hypotheses The m ajor hypothesis of this study is that the input side aspects and outcome efficiency may vary between different types of sponsorsh ip. The hypothesis statement is as follows: H: Government intervention may have an impact on the efficiency of TBI To investigate the input side differences, sub-hypotheses ar e drawn based on the following three theories: h1: For the limited options allowed for private incubators, typical co st-reduction strategies, such as affordable rent, may be limited. h2 : With higher flexibility, private incubators may prefer informal intervention activities. h3: Private incubators may have more difficu lty taking advantage of local sources of knowledge, capital, and instituti ons than public incubators. How the Cases were Selected As previously m entioned, this study selected incubators from two ex treme cases: a private for-profit incubator and a public non-profit incuba tor. To be a private for-profit incubator, requires being founded and managed by a private agent without any financial help from the public sector. Private non-prof it incubators are excluded, accord ingly. To be a public non-profit incubator, an incubator should be owned a nd founded by the government. In addition, the government agent drives most of the operations of the incubator. Also, TBIs founded by the government, but managed by a private agent, are excluded. Obviously, both incubators should serve NTBFs as opposed to mixed industries. Based on the above criteria, to investigate the input side aspects, one for-profit incubator, and one public non-profit incubator were recruited from two different locations in Florida. Due to the limited number of client companies and da ta availability, quantit ative research was done

PAGE 65

65 with sample client companies from two other in cubators which were both located in Rockville, Maryland. These TBIs also meet the criteria stated above. One adva ntage expected from analyzing incubators located in the same city is that factors related to economic geography are almost perfectly controlled. Methodologies: Case studies and Group Comparisons As discussed previously, the prob lem then aris es as to how we address the efficiency issue. The case study method is relevant to explain how the different type s of incubators are similar or different from each other. Indeed, the case st udy method may be the only eligible approach for the given conditions of incubators, as long as a b it of exaggeration is tolerated. In addition, the case study method is effective when the number of subjects is small (Yin, 1994). Methodologically speaking, measuri ng the correlation association be tween elements of incubator service provisions and client pe rformance may be quite unrealistic. As indicated by the NBIA research team (2003), predictive relationshi ps between service pr actice and performance outcome may be difficult to observe. They infer that such a low level of predictive relationship may be due to the individualized needs profile of technology businesses. Incubator services are offered as part of a packet of services, and the services are given in concert with surrounding conditions that usually define the business atmos phere of the area. In this type of case, experimentation on the relationship between se rvice practice and results may distort the particular holistic dynamics that are created by concerted contextual c onditions. These special characteristics may reinforce the relevance of case study methodology as an approach for incubator study (Yin, 1994). To illustrate the possible differences im posed by sponsorship, the method should be referred to as a comparative case method. For th is research, the comparative case study method is employed to achieve multiple goals. As mentioned, the first goal is to demonstrate differences

PAGE 66

66 or similarities. In this sense, the main task of the case study is to discu ss why such services and managerial fashions are chosen, and it would be appropriate to s how these items. The aim is to discuss the meaning of using them as strategies to meet the individually defined objectives and the overall assigned mission; not to list names of se rvices or to count the number of them. The method also explores the goals and objectives imposed by different types of sponsorship to explain how services and manageri al strategies function to achiev e them. Along the same line of reasoning, strategies to maximize the utilities of given resources are explored and explained based on the different theories of innovation proc ess. The ultimate goal of the case studies in this research study is to pr esent logical inferences on th e predictive consequences of operational/organizational characteristics that ar e observed from each incubator; rather than present stochastic inferences base d on repetitive patterns. This leads into the discussion in the next section of the methodologi cal constraints of incubation study which actually add to the relevance of the case study method for incubator study. Two group comparison methodologies were em ployed to compare the primary outcomes produced from two incubators. This study investig ates both the business outcomes of the client companies, and the aggregation of these to the inc ubator. Data are gathered using three sources. Incubator managers were contacted and asked to provide list of client co mpanies. Supplemental information about the clients is gathered from the websites of the companies and incubators. Finally, business record of em ployment and sales were provided by National Establishment Time-Series (NETS) database. Using SPSS and supplemental applications, the Wilcoxon-Mann-Whitney tests (which use nonparametric statistics) were employed to test the statistical significance of the business outcomes of client companies, and also to cove r the wide distribution. This method ranks the

PAGE 67

67 data to test the hypothesis that two samples, with two different sizes, come from the same population; assuming they are similar in nature, when normality assumption is badly violated. Another advantage of this test is that the number of samples of the two groups need not be equal, which is the case in this study. I ndicators related to growth in employment, wages, and sales are explored and tested. Detailed test hypotheses and test st atistics are elaborated in the section that presents the findings. In the following sections of case studies and group comparison, four subject incubators are named by codes. Capital letters will be combined to indicate sections and sponsorship. C will indicate Case study and G will be used to indicate Group compar ison. Then P and G respectively indicate Private a nd Government (instead of Public to avoid confusion with p of private). The private incubator in the case study will be called, hence, CP. CG indicates the Government (public) incubator in the case study. GP is the code of the private incubator in the group comparison, by the same logic. Finally, GG is given to the Governme nt (public) incubator, a subject in the group study.

PAGE 68

68 Figure 3-1. Generic model of incubation process Figure 3-2. Incubator as a cost-reduction strategy Candidate Pool Outcomes Technology Business Incubator Office space/ supplies/furniture/ clerical services Shared conference room/ on-site laboratories Service provisions Business skills Expensive Office space/ supplies/furniture/ clerical services/Shared conference room/ onsite laboratories Candidate Pool Outcomes Technology Business Incubator Mission Statement/ Objectives/ Managerial guidance/

PAGE 69

69 Figure 3-3. Intervention Mechanism of Incubators Figure 3-4. Surrounding Resources of TBIs Candidate Pool Outcomes Public-sector Proactivism Agglomeration advantage V enture Capital Skilled Labo r Research University Business Network Expensive Office space/ supplies/furniture/ clerical services/Shared conference room/ onsite laboratories Technology Business Incubator Manager Selection Proces s Monitoring Proce ss Guidance activity Candidate Pool Outcomes Technology Business Incubator Manager SelectionProcess Monitoring Process Guidance activity Expensive Office space/ supplies/furniture/ clerical services/Shared conference room/ onsite laboratories

PAGE 70

70 Table 3-1. Conceptual framework: measurement of input Concepts Categories Indicators Different Strategies Making the Incubator Work Cost-reduction Strategy Business model Rent Service fee Intervention Mechanism Selection process Monitoring activities Graduation policies Surrounding Resources Access to universities Access to research facilities Collaboration with innovative communities Collaboration with government Table 3-2. Conceptual framework: measurement of outcomes Concepts Variables Indicators Efficiency in creations of jobs, wages, and sales revenue Number of Jobs Growth in Number of Jobs Employment Growth Rates Annual Employment Growth Rate Changes of Number of Jobs over Time Estimated Wages Growth in Wages Wage Growth Rates Annual Employment Growth Rate Total wages Changes of Wages over Time Sales revenue Growth in Sales Sales Growth Rates Annual Sales Growth Rate Total Sales Changes of Sales over Time

PAGE 71

71 CHAPTER 4 CASE STUDIES Introduction As indicated by existing studies different m otivations behind different sponsorship may not necessarily mean different service provi sions (D.N. Allen & R. McCluskey, 1990). Nevertheless, different sponsorship means differ ences in the profile of available sources and capabilities. For instance, the publ ic sector might be more effici ent in helping firms to connect to existing networks or the public sector might be capable of fo rmulating large-scale funding to start a large-scale incubator. No research has yet investigated the possible differences in modes of different sponsorships in addition to the rela tions with outcomes affected by such differences (Hackett & Dilts, 2004b). This ch apter presents two case studies of two radically different models using three theoretical models about incubator and innovative economies previously discussed. The first model is a private for-prof it incubator. This section illustrates how this model works with given conditions that are preconstrained by its sponsorship, being a private manager. Second, it analyzes the working model of a public not-for-profit incubator. This incubator is founded and governed by a local government agency. The same issues are explored in terms of input measurement. First, in term s of a cost-reduction stra tegy, the case studies discuss issues related to st rategies of minimizing production cost (thus maximizing costreduction effects). Reduction of production cost is expected to be obtained by moving in the incubators. Second, differences between intervention mechanisms are analyzed. Finally, based on evolutionary economic theory, the case stud ies discuss strategies of making use of surrounding resources that hinge on innovative activities. In terms of outcome measurement, the effectiveness of each sponsorship is evaluated by examining direct outcomes, i.e., business performance of their client companies: creation of number of jobs and sales revenue. In sum,

PAGE 72

72 this case study is owed to the perspective of evolutionary theory of economies that sees economic or noneconomic institutions continuing to evolve (Nelson, 2006). This study also vows the presumption of the perspective in that it also expects to observe evolving patterns of private and public incubators while they s eek maximization of given resources and face challenges. The goal of this case study is to provide the public sector with lessons so that they can facilitate the program with a wise approach. To achieve this goal, case studies attempt to present a comparative perspective by illuminating differences between the models of incubators managed by two radically different types of s ponsorship, rather than focusing on individual service provisions. The underlying assumption of this case study approach is that there might be optimal models that suit those different models constrained by the availability of resources and capability of utilizing those given to each sponsor. The ultimate goal of this research is, again, to present policy insights with which local economic development agents can make alternative directions that realize the differences in cap acities and limitations observed from two radical models in this study. The case studies however, would be limited because it presents relatively weak quantitative support to show how such diff erences generate results in different outcomes, in terms of business performances of client comp anies. It is due to such a small number of sample populations, being not large enough to gene rate statistical inferences. The case studies will try its best to capture the quantitative magnitude generated by each model by providing descriptive statistics. Hopefull y, this weakness could be offset with an intensive quantitative analysis provided in the next chap ter, while the study area is different. Public Incubator Regional Information Population: 284,539

PAGE 73

73 Median household income: $36,441 Average household income last yr ($) (2000): $49,336 Unemployment rate (2000): 8.2% Pct. pers. 25+ yrs. old with a bachelor s or graduate/prof. degree (2000): 36.7% Incubator Characteristics Governm ent owned and managed incubator Fostering economic growth Targeting employment growth Overview Incubator CG is owned and m anaged by County Research and Development Authority. The incubator is an effort to stimulate the ec onomic growth of the region through providing a nurturing environment for technology business st artups. It takes adva ntage of the abundant facilities and human resources available within the technology park in which the incubator is located, as well as of the serving area. The pa rk offers office, lab, and manufacturing spaces, including hands-on business support. Currently, th ere are over 45 organizations located in the park, employing 1,700 people. As of 2008, the incuba tor has four resident clients, and some companies have moved or graduated. The incubato r also helps virtual clients. Some companies were also helped even before the incubator was officially formed. Establishment The County R&D Authority held and m anaged th e Research Park before foundation of the Incubator CG. When it was a typical research park, without an inc ubator on site, the park wanted an incubator in the park but did not have funding, and thus needed partners to help operate it. A partner did finally show up. Florida Agricultur al and Mechanical University Small Business Development Center (SBDC) was trying to move out of its previous location and seeking a new site. The Park suggested leasing SBDC 4,000 square feet of space to use as a technology incubator. The Park was able to make that o ffer because it solved the mortgage issue for the

PAGE 74

74 building that was slates to be used for the incubator, so there was no severe cost imposed to enable this partnership. The Park hopes for future expansion of the site and for more partners to join the program. Motivation The m otivation of this incubato r is rooted in the economic structure of the region. The small communitys desire to diversify the empl oyment base is rooted on the economies of the region, that is, mainly dependent on the unive rsities and state government. The CRDA and the Research Park utilize university research to create jobs, and diversify the economy by operating the incubator as an agent that diversifie s the regions economic/employment base. Goals and Objectives As a public incubator, this incubator aim s to booster the local economy of the Capitol region. It is clearly dedicated to serving the local communit y, supported by survey results proving that more than 80% of the tenants of the research park come from the community. To meet this end, Incubator CG ta rgets assisting technology-oriented startups and small companies during the critical phase of startup development as a means of increasing their level of income and employment opportunity. As a strategy to achie ve these goals, the inc ubator provides clients with resources, hands-on assistance a nd a set of business-related services. Advantage of Moving to the Research Park The Research Park and the incubator were both located in the region even before the incubators move to the Park. Moving to the Park, however, provided a set of new advantages for the incub ator. The Park supports the incubator by providing space, including furniture and miscellaneous supplies. A county authority, County Research and Development Authority develops and manages the park. The Park and th e SBDC collaborate to help the firms recruit employees.

PAGE 75

75 Business Model The business m odel of Incubator CG is along the lines of a typical real estate model. It provides the clients with office space, rent for th e facilities, office supplies, and miscellaneous provisions, including furniture, etc. As a technology incubator, Incubator CG also consults client companies to give them appropriate business guidance, positioning the managers as business assistants. A Board of Governances is appointed by the county government to operate the incubator at the Park. The Board includes busin ess and community leaders. Management is handled by the onsite staff of SBDC. The primary task of SDBC is to consult with the client companies. The member list of Board of Governan ce and the on site staff is comprised of agents from multiple organizations that enhance collab oration in an regional innovator community. Local government agents hold the responsibility of management. Th e position of chairman of the Board of Governance has been held by a person from local bank, the vice president from local community college, other members from the univers ity and representatives of a community of local companies. Two major characteristics of the real estate business model are inher ited by Incubator CG. First, Incubator CG receives rent and fees as the return on the service package, rather than taking an equity stake. Second, the incubator serves clie nts as an external assistant. Services are provided as part of a package of multiple business sources. As mentioned, it covers services from physical infrastructure, such as office sp aces, to business guidance, including networking opportunities. Typically, public incubators using the real estate model tend to focus on providing affordable rent as a way to help cut the initia l cost for young companies. On this real estateoriented model, the incubator tends to stay outsi de of business and provides formalized executive services that apply to all clients. The mon itoring process is periodical and mandatory, based on given criteria established by Board of Governance. Progress is evaluated based on the

PAGE 76

76 milestones created by the same process. There are set tasks that client companies should do each year, which are also formally required. Incubator CG, however, goes beyond this model. In addition to flexible leases for spaces, there are one-on-one business consulting, mentoring, education development resources, networking, grant assistance, and marketing servi ces all available for client companies, which combines the traditional models with strengths ob served in private incubators. It seems to be a combination of the typical model and new incubati on trends. Due to its re latively short history, it is irrelevant to conclude this new models consequence; however, it seems to be a very interesting experiment The following is the list of generic services that are available for clients (with some flexibility). List of services Business Development & Consulting Business Workshops & Seminars Legal & Financial Services Networking opportunities with establ ished industries and universities Clerical Support/Receptionist Video conferencing capabilities Access to LCD projector and laptop Website links to Innovation Park and SBDC at FAMU homepages Short term/Flexible leases Marketing and Public Relations Furnished offices & Conference space High speed internet and email access Mail room Free parking Access to competitive grant money Besides the above services, th e incubator provides training and workshops to businesses upon their admission. Topics of th e training and workshop include start a busine ss, business plans, marketing, financial projections, and additional training based on the initial assessment. These programs are to mean to help startups through training in necessary skills and

PAGE 77

77 knowledge for successful business development. Unless they are exempted, all clients are required to complete the core training workshop requirement. Other than these basic programs, SBDC requires clients to meet with a busine ss counselor from SBDC to discuss business progress every month. Selection of Clients Client s election is determined primarily based on business feasibility. Incubator CG requires a feasibility proposal as part of app lication process. Basi cally, the proposal should describe the business idea. The incubator dete rmines the business feasibility by evaluating the definition and information on the market, barrie rs to start-up and mitigating strategies, and potential actions and steps to begin and operate the businesses. The company should team up with persons who are knowledgeable about the busine ss, as well as strongly motivated. To fulfill the mission to foster the economic growth of the region, the clients must show potential to make a significant and sustained contribution to loca l economies and eventually to create regional community wealth. It should indicate the quality and quantity of employment possibilities. The proposal also should include re venue creation possibilities, cap ital investment options and predictions for return of the investment, including strategies to reach the possible market share of the business. While the incubator provides accesse s to venture capital, the incubator also sees an ability to attract external sources of investment and the availability of critical financial resources to support business operations through their first y ear. Finally, the applican t needs to address the reasons it needs to be in the incubator, and how the Incubator CG genera tes synergy by serving the business. Monitoring Characteristics The m onitoring process varies by the differe nt philosophies of incubators on business progress. Some incubators conduct formal processe s to collect evidence that work as indicators

PAGE 78

78 of business progress. Formal processes have two aims. First, they allow the formation of tracking data that form the basis for evaluation studies (NBIA website). Second, a formal process allows standardized analysis on the progr ess of client firms. Informal monitoring, on the other hand, is preferred by incubators that believe business progress is less determined by standardized measurement. Such incubators prefer to keep on their eyes on the business progress by maintaining informal relations with their clients. The type of progress monitoring may relate to the type of sponsorship. Public incubators may have a strong motivation to choose more rigorous and formal monitoring processes, as they are often required to submit progress reports to the authority supervising their us e of public subsidy. Incubator CG seems to use a style resembling both models. It has formal and periodical monitoring processes through which the incubator rigorously reviews the business progress of client firms. In terms of formality, the inc ubator requires tenants to submit a brief quarterly report on their progress and the SDBC reviews th em. However, the incubator seems to be flexible on its definition of busin ess progress. Rather, the incubator monitors progress based on self-defined needs and goals by individual c lients and the SBDC (Manager, interview). According to the incubator packet and suppl emental explanation by the executive director of the park, first year clients are supposed to work with SBDC to develop a realistic business plan by the end of sixth month af ter the admission. The progress milestone for those clients is based on the progress of this work. Failure to de velop a successful business plan at this review could result in a decision to release the business. At the end of the first year, the Advisory Group reviews business progress to determine whether the incubator sh ould continue support the company. Based on the review, the Advisory Group may decide to allow a compa ny to graduate, or to terminate further participation. In case

PAGE 79

79 a company seems to need more time in the facilit y, they can also increase rents to market rates. Possible decisions at this point also include simple extension, without changing anything. The Advisory Group may take any othe r action if necessary, as observ ed in the annual review. Another unique factor of this incubator is that they look for actions that may affect business development indirectly, rather than m easuring the business outp uts, such as revenue sales and amount of external investment. They check on participation in training opportunities and workshops, which thus requires the clients to participate. Th e incubator also looks for the attendance of clients on mandatory and optional business conferences, social events, etc, as such activities are provided to help the clients achieve the goal of the company. The rationale for monitoring these activities, rather than focusi ng on outcomes, such as revenue and employment, is that the companies, in th eir early stages, should focus on building a solid foundation for the business. An interesting finding to note is that this is a public incubator, under political pressure to show a tangible contribution to local employ ment, and they dont push client companies to hire employees until they are mature enough. As mentioned previously, public incubators usually care about the employment generated by client companies; this incubator is relatively flexible on that issue when it monitors the business progress of client firms. However, the em ployment numbers are stil l a critical factor for graduation. Graduating companies also report revenues and obtained capital. This information is collected but not with an aim to disqualify the company for graduation. This information may be used as a basis for program improvement. It is questionable whether this requirement acts as a stimulant to the business progress of client companies. Source of Knowledge and Talent In addition to the infrastructure provision, TBIs often provide s ources of technology and licenses. Som e incubators requir e clients to enter into bindin g contracts to use the technology

PAGE 80

80 and licenses developed by associat ed universities, while most inc ubators provide connections to the resources of these propertie s. Incubator CG does not own a ny technologies. Instead, they maintain cooperative relations with technology transfer offices at area universities. Florida State University, Florida Agricultural and Mechanical University and Tallahassee Community College are members of the governing board for the park, and along with representatives of the municipal government. Local Economic Development Council and patent attorneys, they are sources for information about innovation. Private Incubator Regional Information Location : Mid size metropolit an area of central FL Population: 589,959 Median household income: $40,649 Average household income last yr ($) (2000) $56,753 Unemployment rate (2000): 3.6% Pct. pers. 25+ yrs. old with a bachelor s or graduate/prof. degree (2000) 24.6% Incubator Characteristics For-prof it private incubator Founded by an entrepreneur Dedicated to serving businesses w ith an Internet Business Model Practice business consulting servic es as active equity partner Serving only two companies pe r year to launch and fund Overview Incubator CP is a private, for-profit technology business inc ubator, established in 2001. Located in m idwest Florida, it is directed and ma naged by one serial entrepreneur with help from three other entrepreneurs. The incubator is dedi cated to serving businesses related to internet, new media and IT software industries. As of August 2007, this inc ubator has graduated 22 companies and it currently serves four tenant co mpanies. One unique factor distinguishing the

PAGE 81

81 business model of Incubator CP from the usual mode l most of incubators is that the Incubator CP recoups its costs in the long term by taki ng benefits from the client companies. Establishment W ith some financial engagement from a ngel investors, it was founded by a serial entrepreneur who established and sold two star tup companies prior to his current career. Interestingly, Incubator CP was funded mostly by its founder, without any financial support from the public sector. The founder pays most of the cost, so that Incubator CP can be an equity holder in its client companies. The founder e nvisioned an investment of $500,000 to make this business model possible. When it was starting he was successful in fundi ng about half of the money by himself. Harvey Vengroff, a regional entrepreneur, agreed to help the founder by providing contacts of investors a nd space in downtown Sarasota. Motivation Incubator CP was m otivated to fill a busine ss vacuum in surrounding counties, which lack high-tech startup companies that will drive the development of regional economies of the area by providing economic dynamism. It meant a lack existe d of gatherings of smart people, not a lack of smart people. Incubator CP was established to bring talented people toge ther and to contribute to the development of local and economies by making essential reso urces available. Goals and Objectives This incubator strives to achieve busine ss success through working on three business guidelines to create prof it out of the incubation bu sinesses. First, it ai ms to produce profitable companies. Second, it plays as an investor, no t just an assisting organization. Third, and ultimately, the incubator particularly targets fast-g rowing tech companies in Florida. From these three business guidelines, one sees Incubator CP s mission clearly: it is driven by commercial interest, rather than by public interest. C onsequently, one can also note that expanding

PAGE 82

82 employment opportunity is not considered to be of any significance, wh ereas profitability and practicability are explicitly emphasized. Location Advantage Out of m any other possible areas for developm ent, the founder chose Region P. The major reason for this choice was his observation of the presence of opportunity untapped by any other supplier. In his view, Region P is an area renowned for its re tirees. The founder observed this dense group of retired individuals as a place wher e talented professionals and abundant capital came along hand in hand. It was his prior busin ess experience that made him value abundant availability of capital as an important resource for businesses success. He fell short of capital when his first business was starti ng in Gainesville, a college town that is rich in young collegeage talent, but poor in capital avai lability. In comparison to Gaines ville, Sarasota appeared to be a better place for high-tech start ups in terms of capita l availability. Although there is no research university in the area, he believed growing business communities were sources of talent. Also, the absence of any existing competitor was another reason Incubator CP was welcomed by the local business leaders. Business Model A m ajor characteristic of the business model of Incubator CP that is unique from the general model of business incubato rs is that it makes available an optional payment plan. Tenant companies can choose either the above monthly plan, or return equity stakes in exchange for customized business services later on. It m eans paying fees when they succeed. Based on the real estate-oriented model, the most popular model is that incubators receive some amount of rent in return for the physical infrastructure a nd monthly fees for executiv e services. This trend may have been inherited from the original dire ction of old incubators serving general small business, which aimed to cut the fixed costs requi red for starting small businesses in their early

PAGE 83

83 phase. Conceptually, the real estate provision of the incubator implies an alteration of market supply by supporting the initial investment, usually by helping with the cost for office space and supplies (Eisinger, 1988; R. G. Phillips, 2002). The incubator provides the affordable real estate as if it is a chunk of subsidy, i.e. the indirect public supply that offset s cost of indivi dual startups. Although the strategy of taking an equity stake is also observed at other incubators, it is by nature difficult for private incuba tors to practice, in that the incubator will need enough capital on hand to run companies until they actually return profit. For this reas on, this strategy might possibly work for incubators that receive strong financial backing from foundations, universities, or the public sector. A reason that the model of Incubator CP appears to go against common business sense is that private agents largely tend to avoid risks (Eisinger, 1988). The reasons that they are not happy to invest in technology startups are threefold. First, start ups are usually new, which means they are unknown to local bankers. Capitalists and bankers tend to not to trust strangers. Second, technology businesses demonstrate high failure rate s when they are young. Therefore, chances of returning benefits to investors seem to be very low. Third, even when technology businesses hit a jackpot, it takes a large am ount of investment, and takes long time to commercialize technologies. Incubator CPs mode l is a challenge to this comm on sense. It runs based on the principal of long term banking on a risky business, betting that a case with very low chances of winning will be successful. The reason it works fo r this incubator may be that the incubator becomes the partner of the client companies, so client companies are guided by the expertise provided by the core group of Incubator CP, a gr oup that has experience and startup knowhow. In this relationship, the risk of failure is shared between incubator and startups, which may end

PAGE 84

84 up in keeping entry very selective and thus lowering the failure rate in the end. The next section discusses, in detail, the rationa les for the selection process. One of Incubator CPs reason for taking an e quity stake or a serv ice fee, instead of receiving a rent fee, is that the founder re alized that this rent fee is likely to burden startups rather than help them. He even stated that affordable rent is less lik ely to save costs for technology businesses. Instead he believes that focusing on critical aspects of technology startups is more critical for early stage NTBFs. He cited tec hnology development, funding, and recruiting people as examples of critical support he found ove rlooked by most publicly funded incubators. He even believes outrageous rents instead kill them trying to find ways to pay the rent (the founder, interview). He believes typi cal incubator service provi sions, such as clerical services and office services, do not add any value. The Incubator CP tries to not to create a onesize fits all approach. Services are provided as they are needed. This principle applies to office space. Basically, the Incubator prefer s a garage model, in which startups can easily intermingle, over the corporate model that doles out la rge office space, regardless of necessity. Selection of Clients Making a quality profile is a cr itical process that enhances the general perf orm ance of an incubator. In this sense, selec tion criteria not only re flect the rigorousness of the process, but are also filtering devices by which incubators realize the objectives of the organization. For instance, if an incubator is dedicated to improving loca l employment opportunity, th e tenant profiles are likely to be comprised of labor intensiv e businesses that hire more people. One unique character of Incubato r CP, in regards to selection or process, is that it is relatively flexible about the location of the clie nt companies. Usually, incubators require the client to move in so the incubator can work with them closely. Also, client companies prefer to move in with the expectation of saving costs. In cubator CP also prefers companies to be working

PAGE 85

85 under its roof, but it is open to the choice of the client companie s at the same time. One reason for this philosophy is that it does not have a large office facility (the founder, interview). In case client companies need a space, Incubator CP finds one for them from the local real-estate stock (The founder, interview). Incubator CP finds a cheap or free space from someone they know who has extra space. Usually, they receive it as a donation or fo r low cost through their local connections. Nevertheless, the founder expressed his wish to receive more support from the public sector or universities so Incubator CP ma y put their client in the same location. While Incubator CP usually finds such spaces easily, the founder described di fficulties in getting the public sector to commit. One reason he suggested for these difficulties is the pessimistic perspective held by the public s ector about incubator businesses. The other reason he indicated was an overly complicated process and too many restrictions. The other possible situation is one in which the client companies prefer to work remotely. In this case, Incubator CP maintains a close co llaborative relationship by keeping each other informed through online communication. The Incuba tor is developing onli ne conference devices to serve such clients remotely. Another characteristic of Incubator CP is th at it has very narrow selection criteria, as briefly mentioned in the previous section. Generally, TBIs have a relatively wide definition of the target industry, such as h igh-tech, or just technology-or iented industry. Bio-tech industry and light-manufacturing industry are examples of the narrow definitions. Consequently, TBIs tend to focus on chances of success, instead of looking for chances for collaboration based on specific knowledge or skills required to succeed in a specified field. As a result, this process combined with incuba tor managers limited cap ability and understanding, may allow incubators to misguide client startups thus lowering its chances for ultimate success.

PAGE 86

86 Instead, Incubator CP looks for applicants who show possibiliti es for collaboration with the incubator. Obviously, one good ju dgment criteria would be wh ether or not the business is aligned with the knowledge and business expertise of Incubator CP. In this sense, "toughness" in admission standards may not be the right concept perhaps it is more about synergy (The founder, interview). The admission rate is low, not because the selection process is picky with an aim to disqualifying applicants, but because th e incubator selectively seeks applicants that they can help among many quality applicants. In short, Incubato r CP chooses a company that it can help. The limitedness of its re source is a part of the reason fo r rejecting applicants, but the primary reason is a mismatch betw een applicant and incubator. As discussed, Incubator CPs extra cautiousness in recruiting seems to be due to its role as a business partner or investor. This position makes them believe that the success of the incubator hinges on the success of the client company. As a result, it must target businesses based on very narrow definition. In any case, the capacity of the incubator is about two new companies a year. Therefore, the admission rate is very low, about 1%, according to the manager. It should be noted that the number of admission cr iteria on its list is not greater than that of other incubators, on average. Some incubators have more than 10 selection criteria. The selection criteria for Incubator CP are as following: commitment to the business a viable business model which will enable th e company to grow; a good product or service that will create a demand Monitoring Characteristics Serving client com panies as a business partner or investor makes the incubator a part of the businesses, rather than an external agent th at conducts periodic monitoring. Therefore, monitoring is rather informal but frequent or on a daily basis, rath er than formal and occasional.

PAGE 87

87 As a for-profit incubator, Incuba tor CP explicitly focuses on re venue growth and the amount of external investment attracted by client companies as indicators of the progress of its client firms. Hence, Incubator CP is less concerned with incr easing the number of employee than incubators operated by public sector might be. The underl ying philosophy behind overlooking the number of employees in its monitoring criteria states that the number of employee does not lead growth, but revenue growth causes it. Graduation policy is not very different from the rationale of monitoring guidelines. The client companies leave the facility when th e company starts making revenue income and operating by itself. Therefore, indicators of pe rformance may vary company by company in that sense. The incubator helps clients with customized services to meet thei r individualized needs for progress. A relatively small servi ce capacity, at two clients a year may be the reason that this incubator allows customized service. In a nutshell, Incubator CP focuses on whether a business is ready to run independently. Incubator CP even suggests outsourci ng if a business seems not to be mature enough to hire full-time employees, but still needs them. Another reason for their discouraging policy on em ployment is that the Incubator CP itself supports labor demand by providing technical su pport. Incubator CP provides the technical services and assistance that client companies requ ire. For instance, the incubator supports credit card payment system when it is needed. The in cubator clearly wants its clients to focus on product and sales. Source of Knowledge and Talent Resources for knowledge and talent are essential elem ents that make the TBI model work. Incubator CP seems to be in need of support from universities, for use of laboratories, libraries, and human resources, even though the manager noted the rich supply of tale nted business people. Region P does not have a major research unive rsity, only local universities and community

PAGE 88

88 colleges that do not have research facilities. Neither do they have entrepreneurial/business programs which might provide a collaborative plat form. Student resources are utilized, but in a very marginal fashion. Student interns work for the client companies, but this participation is limited due to difficulties with scheduling. The relationships between incubator and universities are still developing, according to the founder (the founder, interview) In regards to sources for technology and licenses, the incuba tor seeks out opportunities from major universities within the state. The University of Florid a and University of Central Flor ida are the supplemental sources for technologies. But the incubator does not hol d any formal relationship with any research center or any university. Impact Study This section presents business profile snaps hots o f client companies that worked with Incubator CP and Incubator CG. As briefly men tioned in the introduction, this section is meant to provide an idea of how the company has perfor med within the two incubators. Unfortunately, any statistical inference is unlikely to be drawn here, as th e sample size is too small to conduct statistical analyses. Therefore, this section shou ld be regarded as providi ng descriptive analyses, which give an idea of the business performance serv ice of the two different types of incubators. Again, the data used in this section is taken from the NETS database. The number of samples depends on the availability of NETS database. According to the manager and the website of In cubator CP, this private for-profit incubator is currently serving six client companies, and has helped about 18 companies. Three graduate companies have already acquired been by anothe r company. Out of 24 companies, NETS data only provides data for five companies: three gradua tes, one current client and one formal client company, all acquired in 2006. The data on the acquired company was available up to year 2007, the last year the company repor ted the data. A partial reason for this small number is that

PAGE 89

89 the subjects are too young to be re gistered in the database. Thr ee of those companies stayed in the same city in which they were incubated, a nd are currently active in business. There is no evidence for the location of the acquired company. According to the manager, the company was acquired by a company located in Boston. The othe r graduate company was one that was helped by this incubator remotely from another region. Th is company was found from the list of client companies of another incubator in the same state. For Incubator CG, founded in 2006 and recently expanded, four companies were available in the NETS database out of a total of seven c lient companies currently served. Since Incubator CG is relatively young, this secti on includes all of the business performance available in the NETS data, regardless of the residency or post-graduate period. The following section summarizes nine sample companies and th e type of industries they represent. Industry Type According to SIC code and supplem ental descri ptions of NETS data the eight businesses serve technology related bus inesses. One note here is that th e clients of Incubator CP, a private for-profit incubator, seem to have narrow variati on in terms of industrial type. For Incubator CP, two graduate companies (CP1G03, CP1G03) fall into same SIC category, Computer software development, in different industry groups. The other two businesses work in different industries; company CP1T01 for Magnetic and optical recording media, and company CP1G04 for Electrical appliances, televi sion and radio. However, they ar e both media-specialized, which is publicized as one of specialties of the Incubator CP. It is premature to draw conclusions only based on the data of four companies, but it ma y indicate the reason that the manager found active collaboration among client companies. The next section provides grow th in employment and sales in a descriptive manner. For external variables not be ing controlled, the growth of wage is not compared in this section. The

PAGE 90

90 study period is between 2001, the earliest year of both incubators, to 2007 the latest year of available data. Growth in Employment Table 4-2 presents sum mary statistics of the number of jobs created by the client companies of the both incubators. In terms of growth, Incubator CG seemed to create one more job than the Incubator CP. The average number of jobs created by Incubator CG was 2, while only 1 job was created by Incubator CP. It may mean that an average of 1 additional job was created by Incubator CP between 2001 and 2007, and 2 jobs by Incubator CG between 2006 and 2007. The median number of jobs of the Incubator CP is also greater with 2 than that of the Incubator CG, of 0. But the number does not account for the employment size of the base year, which is the first year of the businesses. The growth rate indicates the percentage of the created jobs in comparison to the number of jobs of the first year of individual business. It is calculated by divi ding the difference in the number of jobs between the first year and the last year by the numb er of jobs reported in the first year. Incubator CP achieved 40% employment grow th, while it reached 86% for Incubator CG. 200% was the greatest employment growth achieved by a client company of Incubator CP. A client company of the Incubator CG has achie ved 133% employment growth as well. Summary statistics may mislead, however, in that Incu bator CP has only one company increased 6 more jobs with five other companies that did not create any employment growth. As discussed in the previous chapters, growth and growth rate du ring the study period are a good reflection of the magnitude of economic contribution, but it may be misleading as to the actual level of performance, in that it disreg ards the length of businesses. Annual growth rate is included to account for th e length of business. Annual growth rate is calculated by dividing growth rate by the leng th of business. The client companies of the

PAGE 91

91 Incubator CP seemed to have achieved about average 10% growth in employment, while Incubator CG achieved about 43% em ployment growth, every year. The last column, Average em ployment, provides insight in to the employment fluctuation during the study period. Since growth and growth ra te are calculated based on the number of jobs recorded in two critical time point s, the first year and the last y ear of the businesses, the Table 42 may not capture jobs that are disappeared in the middle of study period. To calculate Average employment the mean value of the number of jobs each year for a business are first produced. The final value 5.98 resulted as average of the values for Incubator CP. One inference is that client companies of the Incubator CP mainta ined on average 5.98 employees through the study period. Using the same logic, client companies of Incubator CG hired average 2.75 employees during the same frame. Figure 4-1 demonstrates changes in the total num ber of jobs over time for both incubators. Over all years, Incubator CP, the for-profit priv ate incubator, hired more people than Incubator CG. This pattern still appears even in the number of jobs per company, illustrated in the Figure 4-1. Figure 4-3 shows the number of client compan ies that contributed to the total number of jobs each year for each incubator. According to the Figure 4-2, when the aggregated number of jobs is divided by the number of companies, the maximum number of jobs is seven for Incubator CP in year 2007, while a client company hired four people as a maximum number of jobs in the same year in Incubator CG. Note that the Incubator CG started in year 2006, so the number of clients before 2006 should be considered only as experimental values. Growth in Sales Table 4-3 presents sum maries of sales growth. One difference of this table from the Table 4-2 is that sales of each year can be added together, while employment can not be added up

PAGE 92

92 every year. Therefore, in Table 4-3, total sales are reviewed in the Total column. The values of the column are the summation of sales reported by client companies during their lifetime. Another difference is that the average sales may not have significance for reviewing sales since sales fluctuation is captured in the summation of sa les. Hence, the column is excluded in Table 4-3. In the Total column, the row sum indicates the entire size of sales revenue reported during the study period. Since the value demons trates quantitative magnitude of economic contribution, it is the most popular indicator used in the impact study. The sales revenue of the Incubator CP, at $19,725,800, is much greater than that of In cubator CP, at $1,993,200. Also, the average sales revenue for Incubator CP, at $3,945,160, is about five times greater than the average sales of the Incubato r CG, of $498,300. Same patterns are observed from the median numbers of jobs of both groups. However, the gr owth rate tells a somewhat different story. Client companies of Incubator CP did not only fail to grow, but also reported negative growth. In comparison to the first year, the sale s of client companies of Incubator CP appeared to have declined; each business seemed to lose an average of -$63,000. However, one should note that the negative growth resulted because th ere was one company that made huge loss. The amount reported as minimum, of -$313,800, should be attributed to the nega tive average. Also, two other companies reported moderate negativ e growths in the group. Although the total sales were relatively small, the client companies of the Incubator CG produced average $191,133 sales during the same period. The median sales grow th of the Incubator CG, of $153,400, is also greater than -$38,100 median sales gr owth of the Incubator CP. None of the client companies in Incubator CG reported negative sales growth.

PAGE 93

93 While Incubator CG produced less sales reve nue, the group has achieved a greater growth rate, with 88%, than the client companies of Incubator CP, which achieved only 4% growth. The maximum growth rate of the Incubator CG reached 133%, while the maximum sales growth of Incubator CP is only 47%. With a minimal dec line in sales growth, the client companies of Incubator CP still demonstrated an average 6 pe rcent annual growth. Again, the median growth rate of sales for the Incubator CP is -0.051, while it is 1.000 for th e Incubator CG. Incubator CG reported also greater annual grow th rate of the client companies with 44 percent than the Incubator CP that reported only 0.061. Figure 4-4 demonstrates total sales over time for both incubators. Again, over the comparable time period, between 2006 and 2007, the total sales of the Incubator CP dominated Incubator CG. In the year 2007, five client co mpanies of the private Incubator CP produced $5,258,800 together, while in Incubator CG, onl y $1,150,800 was produced by three companies. Figure 4-5 demonstrates total sales per company. It is also calculat ed by dividing the total sales by the number of companies. The same pattern exists in Figure 4-5; a for-profit private incubator dominated in terms of sales per company over all time periods. The maximum sales revenue was $1,051,760 in year 2007 for Incubator CP, and $383,600 for Incubator CG in the same year. An interesting observation in Figur e 4-5 is that Incubator CP reported negative growth from 2004 to 2005. In the year 2004, th e client companies produced $3,603,800 as sales revenue, but this number dropped to $3,562,500 in the following year. The range of the decline reached -$41,300. Bouncing back in the following year seems to owe to new members join. The sales revenue that the new member br ought in was $1,638,100, which nearly matches the range of increase of $1,576,300. It e xplains the fact that the periodical in crease of sales per company was only about $100 between 2005 and 2006.

PAGE 94

94 Summaries and Discussion Although the analyses of this section are done with a sm all sample, the pattern shown here is that the private incubator tends to be better in profiling in general, an d the public incubator is either competitive or better in enhancing busin ess performances. The followings are possible inferences; when reading, one may keep in mind th e biases of a small sample and a relatively short study period. In terms of the number of jobs, the privat e incubator hired more people, but a higher growth rate was continuously observed in the public incubator. It affirms that private incubator may have accepted companies that hired more people, but the companies may not increase the number of employment as much as the companies in the public incubator. Between 2006 and 2007, the average number of jobs in the private incubator was great er than those of the public incubator. However, the growth rate was comp etitive between both incubators. One reason that the private incubator hires less than the public inc ubator may be that the companies of the private incubator are less encouraged or less motivated or even discouraged by the incubator. Another reason for lower growth may be that it is related with industr ial characteristics. A possible inference seems to be that it is due to privat e incubators less rigorous motivation in employment expansion, according to the interview with the founder of Incubator CP. According to the founder of Incubator CP, he stresses only production and sales. He continued that he suggests clients wait to hire new employees until they can independently afford them. Does the private incubator cause fast growth? It may not. The same pattern is observed. The private incubator demonstrated more sales re venue, but it also showed slow growth. Total sales revenue as summarized in the Table 4-3 is not eligible since the total sales, alone itself, is biased by the longer study period entered in the calcula tion of the total sales for the Incubator CP. But even between 2006 and 2007 in the Figure 4-5, the sale revenue of the private incubator

PAGE 95

95 is about 4 times greater than that of the public incubator. Sales per company of the private incubator are also three times greater than that of the Incubator CG. The larger size of the client companies may be the reason that sales growth is sluggish for the private incubator. The public incubator demonstrated fast grow th rates, but the size of the client companies seem to be relatively small. Age is not an explanatory factor in this case, since th e slope of the first two years of Incubator CP, observed in 2001 and 2002, is not greater than those of Incubator CG in 2006 and 2007. In conclusion, the clients of the private incu bator might be driven to be more cautious when they accept client companies for their reso urce constraints. As results, they hire more people, and they produce more sales. But as discussed, it does not mean that the private incubator is better than public incubator in e nhancing business performan ce, in that the public incubator showed be tter growth in both employment and sa les. An implication might be that without the public incubator, small companies mi ght have little chance to receive incubator services. The other implication is that the public sector may be better producer, at least in the incubation business. Should the public sector withdraw support from incubation businesses? This section found little evidence in favor of such argument. Conclusion and Discussion Cost-reduction Strategies A m ajor difference between the two types of inc ubators is observed in aspects related to cost-reduction strategies. Although there seems to be an upgrade, the public incubator, Incubator CG, followed the existing model that emphasizes the significance of cutting fixed costs. Therefore, affordable office space, office suppl ies, furniture and miscellaneous supplies are introduced as major service provisions in the pack et, website and in the in terview. Incubator CP, on the other hand, clearly developed an alternative method for cutting costs. Instead of focusing

PAGE 96

96 on fixed cost saving, it suggests providing clie nts with sources of funding, recruiting, and technology in order to shorten the lag time to independence. This approach is also justified by the fact that technology businesses are not very sensitive to fixed costs. Instead, as a strategy to save costs Incubator CP chose not to expect direct returns for providing services, but instead to take an equity stake after client companies beco me successful so that th ey can save on redundant costs in the early stages. This strategy also seems adaptable to the public incubator, in that abundant capital backup is necessary to sustain the operation of the incuba tor itself while it waits profits to return. Intervention Mechanisms The sam e principle seems to be possibly extend ed in reviewing processes of selection and monitoring. Following the traditiona l model, public Incubator CG focuses on the feasibility of the business, with broader defini tions of target businesses, while the private for-profit incubator takes a more cautious approach, thus accepting business plans with which the expertise of the incubator may possibly create s ynergy. The weakness of the model of Incubator CG may be that the approach does not consider whether the ex pertise of managing staff matches the business needs of the proposed plans. Obviously, the majo r tasks of incubator include more than just providing spaces and supplies. This approach, however, seems to offer chances for small businesses that have the potential to succeed, while matching tactics provides a relatively slim chance that the business idea and incubating expe rtise match. Also the strategy of the public incubator may be relevant to one of the organizations objectives, that is, diversification of the economic base, in terms that it opens chances to businesses that propose a ny feasible plans which allow the regional economic profile s to become more diversified. On the other hand, the private Incubator CP s approach, focusing on whether the company matches the expertise of incuba tor, seems to bring about a cl oser collaborative relationship

PAGE 97

97 between incubator and client companies. Based on the result of the im pact study, this matching approach seems to have caused its selection profile to be comprised of companies that are readily successful; the client companies produced more employment and greater sales revenue, although they are sluggish in growth. In terms of monitoring processes, Incuba tor CG employed a formalized approach. It requires periodical progress reports that should be entered by means of official forms. Incubator CP uses a more informal monitoring process. An in teresting trait of public Incubator CG is that it does not necessarily check the growth of empl oyment. It may be due to sharing the same philosophy with the private for-profit Incuba tor CP, that employment will follow when companies are successful enough. Surrounding Resources In utilizing its surrounding resources, Incubator CG takes advantage of its status of being supported by the public sector. Th e Research Park, the host of Inc ubator CG, is a m ajor resource of knowledge and technology by itself. Also, the incubator and Research Park receive external resources from members of the governing board, representing the regional innovative community. Consequently, the Research Park provides a pl atform within which local government, business community, and academic community effectively inte rmingle. Although it is not stated in any document or interview, this supportive infrastruc ture and collaborative community are privileges that public incubator could easily enjoy. The first and most serious challenge that Incubator CP faces is the passive local government. Complicated and numerous regulations ma y be a result of its limited perspective. According to the founder of Incubator CP, local and regional government seem unmotivated to devote resources to incubator businesses. Lacking a major research university is another problem. Although the founder saw talented regional businesspeople as a cr itical advantage, having no

PAGE 98

98 major research university may be a major bias of the program that cannot be offset by the innovative business groups alone. As result, Incubator CP seems to be struggling to connect clients with sources of resear ch and development. He hims elf travels to find sources of knowledge and licenses in other c ounties. It is, however, relativ ely long distance collaboration for an incubator. The founder also expressed regret that the government is reluctant to provide a place for his office. One possible economic appro ach for the public sector in developing incubators in an area is simply to make local real estate available for the private incubator to use as office space. In this way, the public sector takes advantage of the expertise of a private incubator, and the incubator can more easily provide office space for its clients. Conclusion In conclusion, the m ajor difference between publ ic and private incubators seems to lie in several factors related to its s ponsorship. First, the capacity to utilize resources seems to drive different styles of management and operation. In cubator CG was able to take advantage of moving into the Research Park, because the Pa rk was governed by a public sector that can manage large-scale real estate and capital subsidy. Another trait is that th e public sector holds an advantageous position that can authoritativel y formulate platforms within which the local innovative community can effectively collaborate. Second, the different di rection of the private incubator is also driven by its relatively limited capacity to utilize resources. For limited capital stock, the incubator becomes cautious about recruiting clients; this business-mindedness motivates the incubator to rigorously examine the potential of applicants based on possibilities for collaboration, rather than feasibility alone. Consequently, only a limited number of companies are served, which leads to a closer collaborative relationship between incubator and client. Due to limited resources, the private incu bator needs to seek out office space and sources of knowledge from external agents. Luckily, thus far the private incuba tor in question does not

PAGE 99

99 have difficulties in searching for office space, bu t it does not seem to be easy to get attention from government and universities.

PAGE 100

100 Table 4-1. Industry type of client companies Incubator Company Code Industry Industry Group SIC4 Incubator CP CP1T01 Magnetic and optical recording media Misc. Electrical Equipment & Supplies 3695 CP1G01 Computer software development Computer and Data Processing Services 7371 CP1G03 Computer software development Personnel Supply Services 7361 CP1G04 Electrical appliances, television and radio Electrical Goods 5064 CP1G05 Design services Miscellaneous Business Services 7389 Incubator CG CG1T01 Air and gas compressors General Industrial Machinery 3563 CG1T02 Computer related services, nec Computer and Data Processing Services 7379 CG1T03 Scientific consulting Services, Not elsewhere classified 8999 CG1T04 Commercial physical research Research and Testing Services 8731 Table 4-2. Number of jobs, 2001-2007 Age Growth Growth Rate Annual Growth Rate Average Employment Incubator CP Average 4.4 10.4000.100 5.98 Stdv. 1.8 30.8940.224 1.80 Minimum 2.0 00.0000.000 4.00 Maximum 7.0 62.0000.500 8.40 Median 00.0000.000 6.00 Incubator CG Average 1.8 20.8610.431 2.75 Stdv. 0.5 20.5550.277 1.55 Minimum 1.0 10.2500.125 1.00 Maximum 2.0 41.3330.667 4.50 Median 21.0000.500 2.75

PAGE 101

101 Total Number of Jobs over Time, 2001-20070 5 10 15 20 25 30 35 40 2001200220032004200520062007 Private Public Figure 4-1. Total number of jobs over time, 2001-2007 Number of Jobs per Company, 2001-20070 1 2 3 4 5 6 7 8 2001200220032004200520062007 Private Public Figure 4-2. Number of jobs per company, 2001-2007

PAGE 102

102 The Number of Client Companies, 2000-20070 1 2 3 4 5 6 2001200220032004200520062007 Private Public Figure 4-3. Number of client company, 2001-2007 Table 4-3. Sales revenue, 2001-2007 Total Age Growth Growth Rate Annual Growth Rate Incubator CP Average $3,945,1604.4-$63,0000.040 0.061 Stdv. 2,701,6291.8158,5610.249 0.104 Minimum $1,113,5002.0-$313,800-0.136 -0.014 Maximum $8,313,8007.0$124,6000.479 0.216 Median $3,238,100-$38,100-0.051 -0.007 Sum $19,725,800 Incubator CG Average $498,3001.8$191,1330.880 0.440 Stdv. 485,606.3840.5$114,7510.524 0.262 Minimum $75,0001.0$100,0000.305 0.153

PAGE 103

103 Total Sales over Time, 2001-2007$0 $1,000 $2,000 $3,000 $4,000 $5,000 $6,000 2001200220032004200520062007 Thousands Private Public Figure 4-4. Total sales over time, 2001-2007 Total Sales per Company, 2001-2007$0 $200 $400 $600 $800 $1,000 $1,200 2001200220032004200520062007 Thousands Private Public Figure 4-5. Total sales per company, 2001-2007

PAGE 104

104 CHAPTER 5 TWO-GROUP COMPARISON Introduction While m easuring firm growth is a daunting task because of its complexity, employment is one of the most popular variables for policy make rs to use since they operate their programs under strong political pressure that demands better working circ umstances of a served region. Sales is another variable that is popular as an indicator of growth (Davidsson, Leona, & Naldi, 2005). In this chapter, the fi rst section explores characteristics of Montgomery County, Maryland to provide the contextu al understanding of the study area throu gh reviewing economic geography of the region and the surrounding area. The economic structure, resources for innovative economies, and tax ince ntives are also major focal poi nts. Descriptive studies on sample client companies will follow. The performances of the private and public incubators are compared in multiple dimensions of employment and sales, using traditional approach of economic impact study and studi es of program evaluation. Issues About Measurement of Growth The goal of this chapter is to investigate the primary question of whether TBIs with different sponsorships result in different outcomes as indicated by the business performance of client companies. Using the dominant growth concept (i.e., the change -in-amount perspectives (Davidsson et al., 2005)), sales and employment are chosen as tw o dependent variables of this investigation. According to Sh ane and S. Venkataraman (2000), firm growth can be achieved by introducing a new product or service, but they are outside the scope of this research. Davidsson and Leona (2005) include assets, physical output, market share and profits as candidates for growth indicators. Sales may not be applicable, however, since NTBFs in the biotech industry may have long development times. Thus, such companies might not able to report any sales or

PAGE 105

105 revenues for a long time. In this case, a better in dicator of growth might be the total changes in assets, such as knowledge of the product, patents, intellectual pr operties, license, and recruiting activities of quality researchers (Davidsson et al ., 2005). In addition, the existing literature has demonstrated an increasing consensus for sales as a preferable choice, a nd it is also a popular indicator that young entrepreneurs themselves use (Barkham, Gudgin, Hart, & Hanvey, 1996). Furthermore, sales precede changes of other asp ects, such as employment (Flamholtz, 1986), as mentioned by the manager of Incubator CP discusse d in the previous chapter. One justification of this researchs employment of sales as an indi cator is that the public se ctor wants to observe a positive impact in employment from Incubator programs. Even a conclusion that some industries may not generate any sales growth may still contribute because it disenchants government agents and political masses who ar e blinded by the fantas y of technology. Instead, an industry like the biotech i ndustry may be able to receive renewed support that makes longterm funding and institutional suppo rt available. Similar justifi cation applies to employment as another indicator of growth. Different types of business growth comprise a multi-dimensional phenomenon that can be investigated by different determinants and co rresponding approaches (Davidsson et al., 2005). To investigate the treatment effect of TBIs on these specified dependent variables, independent variables include many other factors that are la rgely considered as determinants of business growth in the existing literature. Variations in geogra phic characteristics are one of the strongest factors that may determine scal es of business performance. In this study, however, geographic variation is controlled beforeha nd by choosing subject samples from one region so that the characteristics constrained by the macro economi cs of geography are relatively homogeneous.

PAGE 106

106 As briefly mentioned, when measuring bus iness growth, one should consider the complicated dynamics of other related factors in that growth, including the outcomes of dynamics of internal and external factors. Most previous evaluation studi es, however, appear to have disregarded the significance of external constraints imposed by different macro economic conditions that vary throughout di fferent regions (Phillips, 2003) This may be due to the oversimplified concept of the treatment effect as a simple change in quantity in a given period of time that is affected by incubators. As a way to minimize the impact of regional va riation, this study invest igates TBIs located in close proximity within one state. Although it is impossible to control regional influences perfectly, regional descriptions may provide a robust understanding of th e spatial and economic background of the area underpinning business perform ance of client companies. This study will cover a wide range of issues related to locati on, from basic demographics to the tax structure related to businesses. It incl udes population characteristics, m acro economic indicators, and tax structures of different municipality levels. Another reason to choose one region to study is to avoid selection bias that might otherwise occur when incubators are pooled together in a group. Followed by a description of each study area, th e selection criteria of the subject samples will be briefly discussed in the followings sections. Characteristics of the sample clients will cover, age, size, and type of i ndustries. Then, this study will di scuss factors supposedly affecting firm performance in order to formulate hypotheses of multiple regression analyses, and finally, to present the results of this study. Description of Study Area: Rockville, Montgomery County. Maryland The State of Maryland This study investigates public and private incu bators located in Ma ryland. Therefore, characteristics of Maryland influence the entire sample population. In other words, variation

PAGE 107

107 regarding state policy and geography became minor issues in regards to the sample populations since the populations are exposed to very similar settings. Fact ors related to state policy and geography can be, in this sense, regarded as constant variables, rather than explanatory variables. Maryland is a background for the entire sample population, thereby either nurturing or hindering the business development of client firms. This section reviews the factor s and characteristics of Maryland that might affect business performance. It includes a review of general economic factors such as labor statistics, state tax incentive provisions, a nd research and education system. Maryland economy and major technology industries Marylands em ployment is comprised of about 22 percent of businesses in Trade, Transportation, and Utilities. Twenty percent is served by a Professional and Business Services cluster. The Education and Health Service industr y contributes seventeen percent, which reflects the fact that the state is host to Johns Hopkins University, a world renewed medical science university. These industries are followed by leisure and hospitalit y, retail trade, and health care and social assistance. Table 51 reports the major industries of Maryland and their levels of employment in comparison to the total employment of the US. Using the base area of the US, the location quotients are either higher than 1 or near to 1, implying that the major industries of Maryland make up a larger share of Maryland em ployment than they do for the country as a whole. The fact that the indus tries of professional and business services, education and health services, health care and social assistance are included as comp etitive industries of Maryland implies that the area has relativ ely strong potential for the development of technology businesses in comparison to other regions of the state. Median household income of Maryland, based on the 2000 Census, was $52,868, which placed the state near the top nationa lly. This higher level of income is one of the factors that is positively associated with the growth of technol ogy businesses. Another factor that makes a

PAGE 108

108 region attractive for the growth of technology businesses is the proportion of work force in management occupations to total work for ce. For Maryland, approximately 40% of the population that is 16 years old or older holds mana gement occupations. This is about average in comparison with national statistics. The availabi lity of leading businesses is another critical factor that facilitates the fo rmulation of new businesses. Montgomery, Howard Counties and Baltimore together are home to twelve Fortune 1000 companies. 1 Currently, Maryland has been tr ansforming itself into a leadi ng technology-oriented state, while it retains traditional strength in manufacturing, shi pping, financial services, and government contracting. Information tec hnology, telecommunicati ons and aerospace and defense are the target industri es of the new economy of Mary land. Table 5-1 summarizes the major technology industries of Maryland. 2 Moderate growth in employment has been observe d in all four industries. While the total amount of employees on the payroll of the aeros pace industry slightly de clined in the year 2002, growth in payroll reflects sim ilar moderate growth. Total pa yroll employment of aerospace companies comprises 2.6% of Marylands total pa yroll, but it contribut es 4.8% of the total payroll of state. The average weekly pay of aerospace employees, $1,426, is about 80 percent greater than the average of a ll industries of Maryland ($782). The information technology industry of Maryland appears to have experienced the global decline of the dotcom industry that occurred in the ear ly 2000s. Employment in the IT industry declined by about 5.6% in year 2001 2002. A sl ow turnaround started in 2003 and added about 730 jobs to industry payrolls by 2005. Als o, the average weekly wage in 2005, was $1,426, which was about 80 percent higher than that of all Maryland industries. 1 www.ChooseMaryland.org 2 www.Choosemaryland.org

PAGE 109

109 The healthcare industry contri butes one in 10 jobs to Ma rylands economy. The industry has continuously reported growth since 2001 wi thout interruption. The city of Baltimore and Montgomery County supported this i ndustry; in fact, 4 out of 10 j obs were created within these regions. State tax incentive for NTBFs As a m eans of incentive for investment in early stage biotech companies, Maryland provides income tax credits for investors in qualified biot echnology companies. Biotechnology Investment Tax Credit is equal to 50% of an eligible investment. As in other states, Maryland also provides Enterprise Zone Tax Credits. Businesses located in a Maryland Enterprise Zone may be elig ible to participate in this credit program. Income tax credits and real property tax credits are offered in retu rn for job creation and investments. A one-time $1,000 credit per ne w employee is the general credit. For businesses that conduct research a nd development, Maryland provides R&D tax credits, accounting for three percent of eligible R&D expenses through the Basic R&D Tax Credit program. The Growth R&D Tax Credit gives back 10 percent of total R&D expenses that exceed average expenditures in R&D over the last four years. 3 The state created a fund, called The Mary land Venture Fund, which makes direct investments in technology and life science firm s and indirect investments in venture capital funds. Software, communications and IT security receive about 60 percent of the fund, and the rest is invested in life science companies. The businesses will pay 8.25% of the State Corporate Income Tax for the year 2009. 3The City of Baltimore

PAGE 110

110 There are 11 universitie s of in the state university system called the University System of Maryland, and one private major re search university, Johns Hopki ns University in Maryland. They have created more than 250 research centers and science technology institutions. In the year 2004, the National Science Fo undation report named Maryland as the fourth highest state for overall R&D spending. It was a significant cha nge from the previous year when the state was ranked in 9th place; this shift resulted from a 41% increase in total R&D performance. The actual amount of money increased by nearly 4 billion dollars. Also, Ma ryland is now the second strongest state in terms of R&D intensity. In 2003, the proportion of R&D to state GDP was 6.3, whereas the national average wa s 2.4 (OMalley, Brown, & Edgerl ey, 2007). In the 2006 NSF report, Maryland was ranked in 4th place in R&D expenditures at universities and colleges and 3rd place in federal government R&D spendi ng at universities and colleges (Maryland Department of Business and Economic Developmen t, 2009). Maryland was also ranked in 7th place in terms of the number of venture capital deals in 2007, of 99 deals, and for total capital investment (Maryland Department of Busine ss and Economic Development, 2009). In sum, Maryland encourages research and development by providing institutiona l and capital support. Selection of Study Area and Regional Specification Selection of study area The geographic focus of this study is Montgom ery County. In this section, four counties and one city of m id-central Ma ryland are explored to highlight the characteristics of Montgomery County by providing regional context surrounding the subject area. The four counties and one city are Howard County, Fr ederick County, Montgomery County, Ann Arundel County, and Baltimore. One private for-profit in cubator and one public non-profit incubator are examined; both incubators are located in Ro ckville City, Montgomery County. Table 5-3 summarizes the sponsorships, incubator code and lo cations of the sample incubators. To avoid

PAGE 111

111 confusion, the public incubator is indicated by character G, as it is a government funded incubator as well. Economic geography Map 5.1 illustrates the geography of M ontgom ery County, Maryland. The map also presents the population of munici palities of the surrounding area. The purple circles, varying in size, represent the size of population. The area se rved by the two subject incubators is Rockville, the municipality boundary of which is colored red. As shown in Figure 5-1, Rockville is a mid-si ze city in terms of population. Surrounding counties also have cities with similar populati ons. The population of Howard County seems to be similar but its low density means that the population is dispersed rather than concentrated. The most populated and concentrat ed areas appear to be the ci ties of Prince Georges County located between Anne Arundel and Montgomery Coun ties. The Rockville area, bridging the city of Frederick and Prince Georges County, seem s to be good for incubator business. This regional context is presented using census 2000 data, as summarized in Table 5-2. Table 5-4 summarizes regional characterist ics in terms of demographics, economy, housing, and residential services. As shown on the map, Rockville is the largest city in Mongomery County. Annapolis is the smallest city in the region. The difference between Rockville and the largest city, Columbia, is about 50,000 people; th e difference between Columbia, the largest city and the smallest c ity, Annapolis is about 150 percent, based on the population of Annapolis. In terms of population density, Howard County may have a dispersed population, indicating relatively lower density. In contrast, Annapolis reports the highest population density with the least population, which signals a high concentra tion within a specific area. Such a high population density may imply uneven development of the county. Or it could be interpreted positively, as an agglomeration of industry, a fact that make s the area attractive to

PAGE 112

112 technology businesses, and precedes the formulation of an agglomeration of innovative businesses. Availability of a highly educated work fo rce is a factor that may facilitate an agglomeration of technology business. Rockville ( 53%) is, along with Columbia (59%), a city of which the majority of residents are highly educ ated, in terms of the proportion of people with college degrees or higher. Alt hough Frederick reports the lowest pe rcentage in this category, the city is between the top 25%30% if compared to 1,273 other places in the USA. Rockville shows an extraordinarily high ra te (31%) of foreign born citizen s, while the maximum foreign born population of the three other cities is only 13%. One should be cautious when interpreting this statistic. Richard Florida may see a large foreign born population as a positive signal of a creative economy because of its high diversity (F lorida, 2003), but it also means a high potential for social struggles or social instability. Neve rtheless, diversity is usually positively associated with innovative clustering (Florida, 2003). In terms of income, Columbia in Howard County has the highest with a $71,524 median household income in the year 2000. Rockville takes second place. Annapolis and Frederick do not differ significantly in this category. This pa ttern repeats in any inco me related categories. Columbia businesses appear to have b een very active in 2000. The number of establishments is nearly double th at of Frederick, which takes second place in this category. When it is adjusted by the population, the number of establishments pe r 1000 people within the four cities becomes si milar, about 35. Single family housing is dominant in Rockville and Columbia, which might indicate the economic prosperity of the cities. Both lead in the number of single family detached homes and homeownerships rates. The other two cities have relatively low proportions of single family

PAGE 113

113 detached homes, either in comparison to other cities in this group, or to the rest of the nation. There seems to be no significant difference in th e level of housing value among the four cities in other dimensions. The four cities have similar levels of employment, between 65% and 75%. Also, the percentage of employment in sales and related occupations is similar, at about 22%. No significant difference is observed in relation to other indicators of employment. In conclusion, the four cities can be considered to be very similar in several dimensions. Annapolis may have a smaller population, but the size is offset by a re latively higher income level and a balanced economy. The next section specifically discusse s the economies of the subject counties. Employment structure To provide a general understanding of the econo m ies of the subject counties and city, Table 5-5 summarizes the top three industry categories in terms of employment and wages of the subject areas. The three industrie s, using NAICS three digit aggr egation, are selected based on the percentage of employment. The data is based on the 2007 Quarterly Census of Employment and Wages data, using Maryland as the base area In this section, the city of Baltimore is included in the discussion since some client companies practice their business in the city. The three major employing industries of Maryla nd, Trade, Transportation, and Utilities, Professional and Business Services, and Education and Health Services appear many times in the list of each county. The major ity of Montgomery employment is comprised of Professional and Business Services, which appear th roughout the entire list. Trade, Transportation, and Utilities is the largest employing industry of Frederick and Anne Arundel Counties. Education and Health Services provides the majority of employm ent opportunities in Baltimore. Most location quotients of these industries also affirmatively in dicate that the job share of the industry for the county is larger than it is in the base area, Maryland.

PAGE 114

114 To provide a more detailed perspective, major employers of the counties are summarized in Table 5-6. The lists of employers are sorted by the type of industries to show the aggregated number of employment by industry sectors. 4 The information in this table is based on the data arranged in November 2008. The lists of major employers for each county are available on the website of the Department of Business and Economic Development. Montgomery Countys major employer seems to be the federal government and companies in the healthcare industry. F ourteen businesses in these i ndustry categories employ around 58,553 people altogether. In Howard County, the number of employees hired by businesses in professional services is outstandi ng. But one business in educati on and one in information hire about 2,000 people each. There is no such outst anding observation in the profile of Frederick County. Employers in all industrie s hire about several thousands people. As observed in Table 5-5, Education and Healthcare appear to be majo r employment producers of Baltimore. Thirteen employers of these two industry categories empl oy 76,426, which is three times greater than the total number of employees, 19,169 people, hired by 19 major employers in Frederick County. In Anne Arundel, the federal government is the major employer. It employs 31,425 people. Interestingly, subject counties possess a la rge cluster of major technology industries as explored previously. Overall, 28.8 percent of the entire Aerospace industry in 2003, 37 percent of Information technology in 2005, and 30.1 percent of all Bioscience businesses were clustered in Montgomery County. A cluster of Healthca re industry businesses is observed within Baltimore, which reaffirms previous findings. The city has the largest share, of 22.9 percent in 2005, of all industry clusters, in terms of number of establishm ents. Table 5-5 summarizes the employment and wages of these industries. 4 The source does not indicate criteria of the collection.

PAGE 115

115 As illustrated, a large cluster of Bioscience businesses affirms that Montgomery County has a very competitive Bioscience industry. Th e employment change of the industry between 2001 and 2003 was more than 10 times greater than Anne Arundels change in the same period. The average wage of the area, however, remained at the same level of Howard County. The competitiveness of Anne Arundels economy seems to be leveraged by the Aerospace industry. The employment change, of 449, between 2001 and 2003, was nearly nine times greater than what Baltimore achieved. Anne Arundel also reported the highest average weekly wage in the industry in the year 2003, $1,611. However, Montgomery County suffered a majo r loss of jobs within the Information technology industry in the period from 2001 to 2005. It lost 4,296 jobs from the sector. Never the less, Montgomery County sti ll reports the highest average weekly wage. Howard County took second place in terms of level of wage, but only 164 jobs were gained, whereas other counties gained about 500 jobs. Montgomery County may offset the deficit caus ed by severe job loss in the Information technology cluster by gaining jobs in the Healthcare industry cluster. Although the county reported the second highest job increase, with 4,847, the average weekly wage was the highest among the group. In the period from 2001 to 2005, Howard County and Anne Arundel County only gained less than 2000 jobs in the Healthcare industry. Th e Biotech industry and advanced technology (mostly Information technology) are two of the four targeted i ndustries of Frederick County, according to an announcement from the De partment of Economic Development, but any of those areas can outperform comparable counties in any dimension. Tax incentive programs Table 5-8 provides a summary of tax incen tives for businesses that are offered by municipalities. No significant di fference is observed in the Local Personal Incom e Tax Rate.

PAGE 116

116 Montgomery has the highest income tax, with 3.2 0%, and Anne Arundel has the lowest, of 2.5%. Baltimore is again ranked as the second highest. As shown in Figure 5-2, zero dollars of Frederick County is an extraordin arily low amount in comparison to the whole vicinity. Caroll County and Howard County are cl assified in the upper 25% in this category. However, Frederick County provides no tax exemption program whatsoever whereas all other municipalities allowed 100 percent of Manuf acturing/R&D Machinery and Equipment and Manufacturing/R&D Inventory to be exempted. Baltimore has the largest Enterprise Zones. There are some Enterprise Zones in Montgomery County, too, but they do not seem to affect incubator clients as demons trated in Figure 5-3. Figure 5-3 illustrates the numb er of client companies by graduated colors. The map uses zip code area files to indicate the location of co mpanies instead of geocoding the actual location. Dark red means that more companies aggregat ed in the corresponding zip code area. As indicated by the symbology, the closer one gets to Rockville, the more client companies are present. As mentioned before, Enterprise Zone s located in southern Montgomery do not affect companies located within Rockville. In contrast, Enterprise Zones in Baltimore cover nearly half of the entire city. Regional knowledge resource TBIs located in the Baltim ore-Washington area may enjoy the knowledge and talent produced by 65 accredited institutions of higher education. However, within the territory of Anne Arundel and Frederick Counties, there are only two 4-year institutions in each county, whereas Baltimore hosts 15 of them. Ho ward County and Mont gomery County have good sources of knowledge and talents. Howard County has about 15 HEI. The number of institutions in Montgomery County may not be as high as that of Howard County, but Montgomery County has several major research universities in it, su ch as Johns Hopkins

PAGE 117

117 University, Loyola College, the University of Phoenix, and the University of Maryland University College. More often than not, the av ailability of knowledge resources emerges as the strongest determinant that enhances the performance of TBIs. Business incubators do not aim to create knowledge or talent by themselves, but the availability of such institutions within the vici nity affect the creation of both codified knowledge and tacit knowledge because client s of incubators can share and transfer their own knowledge. Montgomery County seems to have abundant resour ces that may provide platforms for regional innovative agents to interact each other. The county has 6 incubators, one technology center and one affiliated university, Technology Park. Bal timore has two business incubators at three locations. In addition, one research-oriented in dustrial park at Johns Hopkins University supports the regional innovative system of the area, along with two research and business parks affiliated with the same university. Howard a nd Anne Arundel Counties may currently be in the process of developing such a system. There is only one business incuba tor in Frederick County but it is located in tw o locations. The innovative activitie s of Frederick County, however, are supported by six technology centers and parks.5 Table 5-9 presents the su mmary of this section. Subject Sample Brief review of sample incubators Rockville p rovides an idealistic setting for comparative studies, as two treated groups, private for-profit and public incubators, are available within the same municipal boundary. The 5 The data provided in this section is based on th e following source. Incuba tors are inclusive all business incubators, regardless of technology or mixed-use. Technology and Science centers are selected based on the name of facility for that include any terms indi cate technology or science orientation. Specialty of the faci lities is not in considerati on. Source: Brief Economic Facts, Maryland Department of Business and Economic Development (http://www.choosemaryland.org/regionsandcounties/randcindex.html)

PAGE 118

118 client sample selected from these treated groups is composed of 10 client companies from private for-profit incubators and 36 companies from public incubators. Based on an NBIA report (2003), a commonality of the private Incubator GP and the private incubator CP, reviewed in the case study secti on, is that they dont have established requirements for graduation. They both emphasize the sophisticatedly defi ned maturity of the individual situation of client companies. The major difference between GP and CP is that GP was established by an organiza tion that utilized large scale investment, which presumably provided large scale research faci lities necessary for bioscience re search. The NBIA list includes an approved animal facility, an animal treatment room, a low temperature repository, a clean room, and warehouse space (2003). Another diffe rence is that GP is surrounded by abundant resources that may encourage innovative activities, whereas CP may struggle to bridge external resources with its clients. The availability of resources is benefi cial in itself, and institutional efforts from the local community and government seem to be aggressive enough for those agents and organizations to be collabora tive. Incubator GP, the public in cubator of Rockville seems to enjoy such institutional and governmental support. Another similarity between the two public incubators is that CG and two of the subsidia ry incubators within group GP are very young. They are still in the process of development. However, the reason that GP was chosen in this section is that one incubator is mature enough to provide many client samples. Also, Rockville seems to be a more nurturing setting for an innovative economy in that there are plenty of organizations of research and development, and the local government seems to be very supportive. Client companies There are several issues that ne ed to be m entioned in regard s to the dataset. Multiple sources were used to create the final sample. The incubator website was the primary source that

PAGE 119

119 provided basic information of incuba tors and the list of client companies. One issue is that some incubators provide only a sample client list. Some incubators do not provide a list of graduates on their website. There was no way to make sure each website list was complete. Issues regarding the trickine ss of acquiring complete lists of clients often include the fact that incubator managers use different standards to define client companies. Some managers use a very broad definition, including affiliates as well, but some oppose the idea. W ith a list of client companies provided on their website, creati ng a final subset depends on the availability of the necessary data in the NETS database. Client companies th at moved to other states are also excluded from the final subset. Table 5-10 summarizes the client sample. The proportion of client companies to the orig inal population seems to be large enough to represent the population. The aver age age of the client firms is 7 years old. The oldest company is 18 years old, and the youngest one is 1 year ol d as of 2007. Figure 5-4 illustrates the sample distribution by age. Client companies and type of industries Using an NAICS code to provide a general pers pective, Table 5-9 lists the m ajor industries of client companies and decryptions. The major industry, for both groups of clients, seems to be concentrated in Research and De velopment in the Physical, Engineering, and Life Sciences. The portfolio of private incubator clients seems to be more focused, especially considering the minor type of industries; they ar e three companies in the Manufacturing industry (325412, 325414, and 334519), one company in Management consulting services (541618) and one company in the Technical consulting services (541690). The three companies in Manufacturing seem to be involved in biological or pharmaceutical production pr ocesses, while the industry focus of clients of the public incubator range s from Building inspection to Manufacturing related to Photographic production.

PAGE 120

120 Table 5-12 presents detailed descriptions of the client companies using SIC 8 digit classification. Industries of three client compan ies of the private incubator are classified as Biological Research, and the other three are clas sified as Biological Research, Commercial. Client companies in the first group appear to be ol der but produced fewer sale s in their first year. One consideration that needs to be taken into account is the start year of e ach business. For these companies, the first year occurred in the early 1990s, so a simple comparison of the early 1990s dollar with todays dollar may be misleading in terms of actual significance of the amount of money earned. As observed in Table 5-9, the largest concentrat ion is in one specific industry: Biological research. The public incubator seems to focus on Commercial Physical Re search. In the same industries, the number of first year employees an d amount of sales generally look greater in the table of the private incubator, while further investigation is require d to draw a convincing implication. An interesting point of this portfolio is the two lines of reco rds at the bottom of the public incubator table. Four client companies classified as Medical research-noncommercial, and Medical research look differently. The numbe r of first year employ ment is nearly three times that of any other comparab le company. Also, first year sales are outstanding. Since the summary of the two companies may cause c onfusion, Table 5-11 provides disaggregated information. The new information shows that there are huge gaps betw een the two companies within each industrial class. Ho wever, no generalizable implicati on can be drawn at this point. Ages of companies are quite similar. One may also question how size affects the scales of sales. An in-depth study or any kind of further investig ation review might be required to provide more information.

PAGE 121

121 As of 2007, the NETS data has discontinued the list of eight companies among 36 client companies of the public incubator. Their last ye ars are reportedly before 2007, the last year that the NETS database made this information availabl e. The description of NETS data informs that such records indicate the deaths of companies. Based on the information available on the webs ite about length of residency, the client companies of the private incuba tor took approximately 5.9 years to develop. A shorter period of time has been committed by the companies incubated in the public sector, with an average length 2.3 years. One interesting finding is that the number of tenants drastically star ted doubling around year 2000 for both incubators. The average number of re sident companies is counted for each year. This section explored figures that provide description of the sample dataset, not the analytical findings. The next section of the paper offers an impact study that compares the performance of the client companies, in terms of j ob creation, estimated wages, and sales. Impact Study As stated in the m ethodology section, the ev aluative studies of this research employ a nonparametric method because of its small sample size. A generous confidence level of 0.2 is used. As such, a generous range of significance is excused (R. G. Phillips, 2002). The analysis in this section investigates both absolute and relative figures to acknowledge the magnitude of economic impact actually contributed by the incuba tor, and the efficiencies of the incubator program measured from the unit of analysis, the client companies. A list of client companies is obtained from the subject incubators as well as re lated information such as year of admission and graduation. NETS database pr ovides information about employment and sales. Wages are estimated using the Quarterly Census of Em ployment and Wages Publication Changes by multiplying the number of jobs created each year by each company.

PAGE 122

122 A major purpose of this section is to presen t compare business performance of the private and public incubators, in regards to three busine ss aspects: employment, wages, and sales. Three sub-hypotheses of this research will be tested throughout the chapter. h4: TBIs in the public sector generate greater jobs than TBIs in the private sector. h5: TBIs in the private sector report greater wages than TBIs in the public sector. h6: TBIs in the private sector report greater sales revenues than TBIs in the public sector. Growth in Employment, Before and After Graduation This section reviews hiring ac tiv ities reported in the busine ss performance of the client companies for two periods. The first period is wh en they were being helped under the influence of the two incubators. The second period is afte r graduation until the latest year the data is available. The major purpose is to compare th e business performance in terms of employment and to find out if there is any difference be tween the two groups ove r the two periods. The subject of the analysis is the number of jobs of the companies during the two study periods, available in the NETS database. The major hypothe sis of this chapter is hypothesis 4 of this research. Each subsection will present tests for hypothesis 4. h4: TBIs in the public sector generate grea ter jobs than TBIs in the private sector. Employment growth, before and after graduation The first m easurement is employment growth while the companies are under the direct influence of the incubators. The employment gr owth of the post-graduation period is compared. The primary objective of this analysis is to highlight business progre ss achieved through the program by comparing the same business performa nce in the period that the companies became independent of the program. Diff erences between the two groups are also investigated within the same periods. In this presentation, employment growth is the difference between the number of jobs at the year of graduation a nd also at the year of admission.

PAGE 123

123 Figure 5-5 demonstrates the fr equency distribution of empl oyment growth achieved by individual client companies of both groups. Two ex treme outliers, one at -45, and the other at 75, are excluded in the histogram. Also, the observati ons are widely distribute d, with a fairly large standard deviation of 13.85, and the Figure is al so somewhat positively skewed. With the given conditions that the normality assumption is severely violated, a nonparametric method appears to be appropriate. In Table 5-14, two outliers are observed, one from the maximu m of Incubator GP, the other from the minimum of Incubator GG. Both are excluded in the frequenc y distribution. Because the distributions are skewed and have a wide rang e with a couple of outliers, the median seems to be more relevant as a summary measurement. The average number of jobs, 11.22, of Incubator GP is highly influenced by the wide range with several outlier s, and thus may not be an appropriate summary indicator. Among the 15 client companies of the private incubator, six client companies are excluded because they have only one year of employment records for which measuring growth is not appropriate. For the companies that stayed only one year in the incubator, the same reasoning is applied for ex clusion. The median of the employment growth observed from the private incubator is 1, while it is 0 for the public group. To test the significance of the difference, a Mann-Whitney test is employed. Based on Table 5-15, the employment growth of the private incubator seems to be higher, with a higher mean rank 27.39, but the difference does not appear significant, w ith a p-value of .218, greater than the relatively generous confidence level of 0.2. The test hypothesis is as follows. h4a: There is no difference in employment gr owth between the client companies of Incubator GP and the client companies of Incubator GG, before and after graduation. H0 : F1 F2 = 0 H1 : F1 F2 0

PAGE 124

124 Table 5-16 summarizes the descriptive statis tics of employment gr owth reported by the client companies of both groups after they graduated. Again only nine companies are included from Incubator GP. This time, six companies ar e excluded because they are currently tenant companies of Incubator GP. As illustrated in Figure 5-6, the distribution is slightly skewed to the left, and Incubator GP has some extreme outliers. After the client companies have graduated, a slig ht increase in number of jobs seems to be created by the public incubator, based on the greater mean rank of 23.08. Again, the average number of jobs appears to be inflated by the outliers of both incubato rs. Employment growth after graduation is the difference between the number of jobs at the latest year and the number of jobs in the year of graduati on. It is to measure how much employment growth the client companies have achieved after they became indepe ndent from both incubators up to the point of this investigation. After graduation, the same patte rn observed from the previous period seems to have continued. However, the medians, at 0, are not only the same, but also the test results, with a p-value of .923, are not significan t, based on the results of th e Mann-Whitney test summarized in Table 5-17. The only conclusi on is that the difference between the two groups seems to be negligible. Employment growth indicates th e quantity of jobs created by the two incubators during the two study periods. Growth in the number of jobs however, should not be interpreted as an indicator of the level of performance of both incubators, in that the number of jobs is mostly sensitive to the length of residency as well as the age of the program. Al so, this analysis is effective in showing increase and decrease, but lim ited in appreciating the implications of such changes since the measuring framework disregards initial size of employment. For instance, it

PAGE 125

125 does not distinguish differences between two addi tional employees from a company that started with four people and a company with an initial staff of one. Employment growth rates, before and after graduation To account for such lim itedness, this section presents an analysis of growth rates of employment. To calculate the employment growth rate, two values are used. The first is the number of jobs created or lost at the end of the periods. Th e second is the number of jobs reported at the base year, which is the year of admission for the residency and the year of graduation for the analysis of post-graduation. Th erefore, the difference between the number of jobs at the year of graduation and the number of jobs at the y ear of admission is divided by the number of jobs at the year of admission. The growth rate of the post-graduation period, using the same formula, is calculated by dividing the di fference between the number of jobs in 2007 and the number of jobs at the year of graduation by the number of jobs at the year of graduation. For instance, if a company started the business with three people, and four j obs are reported at the year of graduation, the difference, one, is the number of jobs th e company has created; thus one is divided by four to produce the growth rate The employment growth rate of this company during the residency is 0.25, that is, 25 percent. The major differe nce of this presentation from the previous one is that this presents the percentage of the crea ted/lost number of jobs, whereas only the number of jobs was presented in the prev ious analysis. Following is the test hypothesis. h4b: There is no difference in employment growth rate between the client companies of Incubator GP and the client companies of Incubator GG, before and after graduation. H0 : F1 F2 = 0 H1 : F1 F2 0

PAGE 126

126 Table 5-18 presents the descriptive statistics on the employment growth rate of the client companies of the two groups. Figure 5-7 demons trates the frequency distribution of the employment growth rate that is pos itively skewed with some outliers. While they were in the incubator, the median employment growth rate of the clients of the private incubator was 25 percent, s lightly greater than that of clie nts of the public incubator, with a median of 0. Again, the 45 percent average employment growth rate of Incubator GG is inflated by the outliers. Based on Table 5-19, the test also finds that th e difference is not statistically significant, based on the p-value of .249 slightly above the acceptable level of 0.20. One outlier exists in the public Incubator GG, as illustrated in Figure 5-8. Distributions of both incubators have a long tail on the right hand of the curve, with extreme outliers in Incubator GG. Based on the descriptive statistics in Tabl e 5-20, after graduation, the median growth rate of the private incubator dropped to 0 percent, while the median growth rate of the public incubator remained the same. One company that started with only one job ended up creating 17 additional jobs, thus reporting 1700 percent employment growth. It also inflated the average growth rate of Incubator GG to 84 percent. In both periods, however, the Mann-Whitney te st does not support significant differences, presented in Table. 5-21. The P-value, 0.82, is ex tremely higher than the acceptable level of 0.20. Therefore, the difference does not appear to be statistically significant. Employment growth rates ar e presented to account for the difference of the initial employment size between the two groups. In both periods, both groups do not appear to be significantly different. One cannot, however, link this finding to a conclu sion, as the level of performance of both incubators may or may not be different, in the sense that periodical growth

PAGE 127

127 rates are also sensitive to the number of year s through which companies continuously hired and fired. The next table presents annual growth rate s that account for the length of the residency and post-graduation period. Annual employment growth rate, before and after graduation As m entioned, the first section was effective in presenting quantitative aspects of the employment impact of both incubators, but was limited in capturing the given size of employment. The second section provided, acco rdingly, the proportional employment growth, that is, the employment growth rate. The proportional employment growth is insightful in that it accounts for the default employment size, thus distinguishing differe nt implications of the same number of employment growth. However, the analysis does not account for the number of years the companies were in business. This section provides annual growth rates, by dividing the proportional employment growth by the number of y ears. The numbers of years is the length of residency and years of business af ter graduation, respectively. This indicator effectively captures marginal changes affected by the program. For this reason, the annual employment growth rate is the most popular indicator in the area of the eval uation studies of economic development. It is also widely called simply em ployment growth rate. The te st hypothesis is as follows. h4c: There is no difference in annual employme nt growth rate between the client companies of Incubator GP and the client co mpanies of Incubator GG, before and after graduation. H0 : F1 F2 = 0 H1 : F1 F2 0 In Figure 5-9, the distribution of the annual em ployment growth rate of both groups has long tails on the right side. Also, several outli ers are observed from both incubators. Table 5-22 summarizes the descriptive statistics of the two groups. The distributi on is slightly skewed to the

PAGE 128

128 left and has some outliers. The maximum annual em ployment growth rate is 200 percent of one client in Incubator GG. Incubator GP shows a slightly greater median, 0.05 greater than Incubator GG, of 0. ---According to SIC code and supplemental descriptions of NETS data. The minimal difference between the two groups is confirmed by the Mann-Whitney test, summarized in Table 5-23. Although a slightly higher mean rank is observed from Incubator GP, at 26.61, the p-value .305 is still grea ter than the acceptable level of 0.20. Based on the frequency distribut ion of annual employment gr owth in the post-graduation period, the two groups seem to show very sim ilar patterns observed fr om the analysis of residency, as illustrated in Figure 5-10. Although the curve shows relatively even distribution, with a moderate positive skew, several outliers ca ll for use of the median, rather than the mean. Most of the distribution is con centrated around 0. In Table 524, the medians of both groups are also 0, which indicates that the annual employment growth of both groups is relatively marginal. The maximum annual employment growth rate of the private incubator is 340 percent. In Table 5-22, the difference between means is very small, and in Table 5-25, test results support no difference between the two groups with the p-value of 0.794, which is greater than 0.20. Although the difference is not statistically signi ficant, the private inc ubator seems to have achieved larger employment growth while the cl ients were in the incubators. After they graduated, the annual growth rate of the graduate s of the private incubato r, again, dropped to 10 percent a year, whereas the same value of 17 per cent is observed from the graduates of the public incubator in both periods. Is there any difference in job creation be tween the two incubators? None of the measurements turned out to be statistically significant. The fo llowing section visualizes the

PAGE 129

129 employment impact created by the subject incuba tors, through measuring growth in employment and sales over time. Changes of total number of jobs over time, before and after graduation Prim arily, the employment impact means the de gree to which a program contributes to the creation of new jobs in a region. Suppose an incubator started a program a couple of years before research begins, and a program manger wa nts to know how many jobs it has created thus far. The investigator counts the number of jobs in the client company report at the moment of investigation. Changes of total number of jobs, before graduation. The summation of the number of jobs is the total number of jobs that the program has created in the meantime. Using same two study periods, before and after gr aduation, Figure 5-11 illustrates changes in the number of jobs generated by tenant companies, by showing jobs reported each year. Both groups seem to have increased over time. The line of tenants of the pu blic incubator shows fast growth in the late 90s until its peak in 2004 and it showed a slowdown at 2003, 4 years thereafter. The number of employment reported by tenants of the private in cubator seemed to have stayed between 50 and 100, roughly. After year 1999, public incubator appeared to have provided more employment opportunities for the region than the private incubator. Figure 5-12 presents job gains and loss each ye ar reported by the tenant companies of both incubators. An interesting findi ng is that the tenants of the pr ivate incubator tend to rather maintain given number of jobs, than actively hi ring many people. With some fluctuation, on the other hand, client companies of the public incubator seemed to be rigorous with hiring activities, while they lost some time, th rough their life time. Over all pe riod, the number of jobs of the private incubator seemed to be larger, but they might have be en a bit cautious with expanding employment opportunity.

PAGE 130

130 Number of jobs per client, before graduation. Again, the fact that the total number of jobs of each ear is sensitive to the number of clients should be remembered. Therefore, the average number of jobs should be counted toge ther, in order to better illustrate incubator performance. Figure 5-13 demonstrates the number of jobs per client for each year. The primary purpose of this presentation is to show the amou nt of jobs per company for each year. To produce Figure 5-13, the total number of jobs in each year is divided by the number of tenant companies of the corresponding y ear during their residency. Using this method, Figure 5-13 shows the average employment numbers for each year. In the previous section, the public incubator seemed to have provided more number of jobs than the private incubator. But in Figure 5-13, when the number of companies joining the hiring activity in the same year is also accounted for, the story becomes different. As a result, the Figure 5-13 shows that average number of jobs created by the pr ivate incubator seemed to dominat e nearly all periods. Figure 514 provides the number of companies resi ded in both incubators each year. Changes of total number of jobs, after graduation. The Figure 5-15 presents the same indicator, total number of jobs, by companies after they graduated both incubators. Graduate companies of the public incubator, collectivel y, provided more jobs after year 2002 than the graduate companies of the private incubator. Again, the number of jobs created by the graduates of the private incubator did not show inte nse fluctuation in th is presentation. Any interpretation of the period ical change in the total number of jobs here should not disregard implications of the magnitude of these changes. To illustrate the change in the number of jobs over time, Figure 5-16 shows the amount by which jobs increased and decreased each year, which is illustration of slopes of the curves in Figure 5-14 and Figure 5-15.

PAGE 131

131 For the job gain and losing of graduate compan ies, Figure 5-16 presents changes of them over time. The same pattern observed from th e Figure 5-12 reappears in the Figure 5-16 that public incubator seemed to have been rigorous in hiring and firing, while the client companies of the private incubator was rather passive in such activities. The number of jobs per company, after graduation. Same pattern of the Figure 5-13 appears in the Figure 5-17 that presents the num ber of jobs per graduate companies, although the period of analysis is shorter than the previous one. For all pe riods, graduate companies of the private incubator have hired more people than th e graduate companies of the public incubator. Severe job loss observed from 1995 to 1996 peri ods is because a major employer has got of business at year 1995. The company, presumab ly going out of business in 1995, died with 95 jobs. As a reference, Figure 5-17 illustrate s the number of graduates of each year. Growth in Wage, Before and After Graduation This section investigates estim ated wages po ssibly paid by the client companies of two periods. Again, first period is when they were bei ng helped under the roof of the two incubators. Second period is after graduation unt il the latest year the data is available. The major purpose is to compare the business performan ce in terms of wage creation and to find out if there is any difference between two groups, over the two periods. The subject of the analysis is the estimated wage calculated by multiplying number of employm ent of the companies, available in NETS database, by the average weekly wages, of the Quarterly Census of Employment and Wages Publication Changes, during the two study periods. The major hypothesis of this chapter is hypothesis 5 of this research. h5: TBIs in the private sector re port greater wage creation than TBIs in the public sector.

PAGE 132

132 Methodological discussion Counting employm ent numbers is a meaningful investigation, in that it effectively quantifies the extent to which the incubators cont ribute to the local econo my by offering jobs to the region. However, counting the total number of jobs created by a client company during its residency may capture only the jobs available in the moment of investigation, whereas the company may have created more jobs at some point of its residency. Such displaced jobs are not accounted in the traditional methods of LED since they dont appear in the moment of investigation (Hughes, 1991). From the perspective of the area, the region could have benefitted even from such a temporary availa bility of employment opportunity. Responding to this challenge, this chapter reviews the estimated weekly wage effect of employment as a proxy indicator to capture the employment impact of the programs. Three underpinning reasons of this supp lemental approach are as follows First, the amount of wages do reflect the changes in the num bers of jobs, whereas counting employment figures may inflate the actual impact, in that the sum of jobs doe s not correspond to the number of employees. Consider adding up the numbers of jobs through three years, in order to not to miss any jobs displaced in the middle of the period. The company reported 8 jobs in the first year, 11 jobs in the second year and 10 jobs in the third year. To overcome a dr awback that ends up concluding that only 2 jobs were created based on a simple subtraction of number of jobs, an investigator adopted a new approach of adding up every job that appeared during the study period and may instead have concluded that the sum of all jobs equals 29 jobs, which is where inflation occurs from double counting number of jobs that existed from the previous years. The second reason to investigate weekly wages, in addition to summaries of the number of jobs, is that wages do not assume the prolongation of the impact. Back to the example: the actual number of employees benefited by the company may be less than 29, since the members that

PAGE 133

133 hold the primary 8 jobs were unlikely to change. Until the number of jobs occupied by these primary members is assured, an investigator is unable to figure out the actual hiring impact created by this company. But is there any da ta that informs us about the replacement of members of each establishment? Concluding that a company created 29 jobs does not only inflate the actual impact, but also elongates their implication; 19 jobs of 29 jobs do not even exist any longer. Measuring wages re flects the actual employment nu mbers, even if the wages are paid to the same people. Also, the primary assu mption of wage is that its effect is instant, although its impact may actually co ntinue as it circulates wea lth in the service area. The third reason to investigate the weekly wage is that the wage al so allows appreciation of different powers of wealth distribution in di fferent industries, wher eas merely counting the number of jobs neglects the different values among jobs. Measuring wages allows this difference to show up. As different industries pa y different wages, the same number of jobs may return different wages. Figure 5-19 demonstrates a comparison be tween traditional methods that sums up of the number of employees, and an accounting of employment through the wage effect. Using employment data of the Private group as an example, the green dotted line represents the sequential changes of total week ly wages of each period over time. The dotted line is created through multiplying the number of jobs of each year by the average weekly wages of each year for each industry. Values of the Y ax is are the actual amount of dollars of the total average estimated weekly wage paid by the client companies of the private incubator. The green solid line shows the change of the simple summa tion of all employees over the same period. To make both patterns, not the actual values, show up in the same bracket, the solid line is artificially created by multiplyi ng the actual number of total jobs by 1000. Thus the number in

PAGE 134

134 the Y axis should not be taken as the actual number of jobs. Although inflated, the pattern should remain same. As demonstrated, the wage effect makes the se quential patterns of lines different from the pattern illustrated by the solid line. As it passes the year 2001, the sl ope of the dotted line explosively escalates. It means th at either businesses paying high wages joined in this period, or that the level of wages of overa ll industries increased in this period. The number of companies in each year is controlled in this case, as it is same for both lines. In either case, one inference is that the dotted line accounts for wage effects, wher eas the solid line, the number of jobs, nearly ignores differences in wages over industries a nd over different seasons. One can argue that gauging sales may meet the same objective. But this position is open to the criticism that it does not distinguish the industries th at produce wealth and industrie s that distribute wealth. Using employment and wage data of one particular sample company, Figure 5-20 demonstrates the effects of seasonal wage change The solid line and dotted line represent the same variables as in Figure 5-19. The values fo r the actual number of employment in Y axis are 1000 times inflated to show the pattern on the Figure, and thus should not be interpreted as actual numbers of employment. As hypothesized, the seasonal wage fluctua tion changes the pattern, sharpening some peaks at the year 2000 and the year 2004. In Figure 5-11, sequential changes in the average wage in this particular industry support this inference, showing fast increases in two years, the year 2000 and 2004 respectively. Also, while th e number of jobs between 1994 and 2000 either maintained its level or decreased, the total weekly wage line inclined or remained the same in this period as result of constant wage increas es, demonstrated in Figure 5-21. In the year 2002, wages dropped drastically, according to Figure 5-20, thereby resulting in a decrease in the total

PAGE 135

135 weekly wage of the year, even when maintain ing the same or similar number of jobs, as demonstrated in Figure 5-19. Accounting for the fact that the company has been in the same industry (titled Research and Development in th e Physical, Engineering, and Life Sciences based on NAICS 6 digit aggrega tion) without changing at any time during the study period, these differences can be interpreted as the results of the wage market effect within an industry. Wage growth, before and after graduation Grow th in wage. With all these accounted, this is a supplemental i nvestigation to estimate the wages that might have been paid for the j obs created by the companies. Wage growth means how much the dollar amount has in creased or decreased during the periods of the residency and post-graduation. By subtracting the total amount of weekly wages at the year of admission from the total amount of weekly wages at the year of graduation, the wage growth during the residency is calculated. In both pe riods, the wage increase in the private incubator was more intensive than that in the public incuba tor. The test hypothesis is as follows. H5a: There is no difference in wage growth betw een the client companies of Incubator GP and the client companies of Incubato r GG, before and after graduation. H0 : F1 F2 = 0 H1 : F1 F2 0 As demonstrated in Figure 5-22, the distribution of wage growth is moderately skewed to the right, if an extreme outlier of the public incubator is ex cluded. Six companies out of 15 clients of the private incubator ar e also excluded for the reason st ated in the previous section. The greatest sales growth achieved by the cl ients of the private incubator is $21,000, and $15,856 for the public incubator. On e interesting observation is that a client company of Incubator GG lost $53,265 while it was in the in cubator. The median value of the private

PAGE 136

136 incubator, $1,946, is a lot greater than the median value of the public incubator, 0, based on the descriptive statistics pr esented in Table 5-26. The difference appears to be significant based on the results of the Mann-Whitney test that returns a p-value, 0.106, which is below the acce ptable level of 0.20 of this study. With the combination of higher mean and median, the wage growth of the private incubator seems to be greater than that of the public incubator. In terms of average, the sales of one client company of the private incubator appear to have increased by $5,026 during the residency, while the wage increase of the public incubator was only $863. Figure 5-23 demonstrates the distribution of frequency of the wage growth achieved after graduation. Distributions of the both groups have long tails on the right-hand side, with a very wide range. As illustrated, there are several extreme outliers in Incubator GP. Wage increase after graduation for the private incubator was $60,839 on average, and $6,885 for the public incubator. There may be a couple of extreme outliers in th e private incubator. The maximum wage growth was $477,840, which is even out of the scope of Figure 5.23. For the very wide range of frequency distribution of both the public and the private incubator, a nonparametric method seems to be inevitable for the test statistics For the difference of the wage increase after graduation, however, it is not sign ificantly different, with a p-va lue of 0.853, much greater than 0.20. Wage growth indicates the amount of wage that has increased or decreased, but it does not distinguish levels of wages at the beginning year. Obviousl y, this implication would be interpreted differently if the value of the star ting point were different. So, the next section

PAGE 137

137 presents the proportional growth of the weekly wage, accounting for the different initial size of the weekly wage. Wage growth rates. Wage growth rates are calculated by dividing the wage increase/decrease by the wage of the base year. For instance, one company paid $30,000 in weekly wages in the year of admission, and becam e able to pay $40,000 in the year of graduation. The wage increase is $10,000. By dividing $10,0 00 by $30,000, the wage growth rate is about 0.33. The test hypothesis to compare differences in wage growth rates is as follows. H5b: There is no difference in wage growth rates between the client companies of Incubator GP and the client companies of Incubator GG, before and after graduation. H0 : F1 F2 = 0 H1 : F1 F2 0 Figure 5-24 demonstrates the fr equency distribution of wage growth rates achieved during the residency. One extreme outlier exists in the public incubator, a client that achieved 616 percent growth in wages. Most in the public in cubator achieved a wage growth rate between 0 and 0.5. Also, the median of the public incubator is zero. The private incubator seems to have achieved a 53 percent wage growth during the residency on average. The maximum growth rate is 180 percent for this group, as summarized in Table 5-30. A very small difference is observed from average and median, showing a greater growth rate from Incubator GP. In terms of average, there is only a 20 percent difference between the two groups. The client companies of Incubator GP seem to have increased by about 20 percent more than the clients of Incubator GG. The median value of Incubator GP is only 0.09, and it is 0 for Incubator GG. Nevertheless, the Mann-Whitn ey test strongly supports that they are

PAGE 138

138 significantly different, with a p-value of 0.065, si gnificantly less than 0.20, the acceptable level of this study. Table 5-31 presen ts the results of the test. Figure 5-25 demonstrates the fr equency distribution of wage growth rates recorded after the clients graduated from the incubators. Both groups have a very wide range of distribution. The maximum wage growth of Incubator GP is 4 97 percent, and one company increased its total weekly wage by 1690 percent in Incubator GG, according to the summaries in Table 5-32. This growth rate was slightly increased to 71 percent after the companies graduated, while the growth rate of the public incubator decreased to about 103 pe rcent. In spite of some gaps observed, the Mann-Whitney test does support di fference, with a p-value of 0.522, summarized in Table 5-33. Wage growth is an effective indicator for showing how much wages increase or decrease during the periods of interest. Howe ver, in terms of efficiency, it is sensitive to the number of years of the residency and postgraduation period. The longer th e period gets, the higher the wage growth would be likely to return. Annual wage growth rate. The annual wage growth rate is alternatively presented to measure yearly change in wages. It is calcula ted by dividing the growth rate by the years of periods, the residency and post-graduation. One extreme outlier is found in Incubator GG, a company that achieved about 200 percent annual wage growth in wages. However the median of the group, zero, is below the median of the ot her group of Incubator GP, 0.20, since Incubator GG seems to have companies that reported ne gative annual wage growth. The descriptive statistics are summarized in Table 5-34. On aver age, total weekly wages of the clients of Incubator GP seem to have increased by 14 perc ent a year, whereas 12 percent growth has been

PAGE 139

139 achieved by the clients of Incubator GG. The te st hypothesis to compare differences in wage growth rates is as follows. H5c: There is no difference in annual wage growth rates between the client companies of Incubator GP and the client companies of Incubator GG, before and after graduation. H0 : F1 F2 = 0 H1 : F1 F2 0 In Table 5-34, a small gap between the two groups became even narrower, which implies a longer length of residency of Incubator GP. The median of the incubator is only 2 percent, whereas it is 0 percent for In cubator GG. Two percent is also the difference of the average annual growth of the two groups. Nevertheless, the difference seems to be significant enough, based on the test results summari zed in Table 5-35. The p-value, 0.120, is significantly smaller than the acceptable level of 0.20 of this study. One implication that can be drawn here is that the clients of the private Incubator GP seem to have shown better wage growth than the clients of Incubator GG, at least while they were in the incubator. Figure 5-27 demonstrates the fr equency distribution of annual wage growth rates reported after graduation. It has a long tail on the righthand side, with several extreme outliers in Incubator GG. The distribution has a slightly wider range than th at in the previous period. Two outliers are observed from the clients of Incuba tor GG. The maximum value of Incubator GG is 3.398, which indicates that a company achie ved 340 percent annual wage growth. Both groups have companies that ended up decreas ing the total weekly wa ges. On average, however, both groups achieved posit ive growth. Clients of Incubato r GP seem to have increased total weekly wages by 17 percent a year, and 22 percent was increase d by the clients of Incubator GG after they graduated from the incubators.

PAGE 140

140 Such differences, however, do not appear sign ificant, based on the test statistics, summarized in Table 5-37. The p-value, 0.618, is signi ficantly higher than the acceptable level of 0.20 of this study. The two groups do not appear different. In summary, both incubator companies increased weekly wages. One interesting finding is that the private incuba tor showed a higher wage growth ra te, without statistical significance, when the companies were in the incubator. After graduating from the incubator, clients of the public incubator tended to grow better. This pa ttern is supported by th e greater annual wage growth rate as well. Total wages, before and after graduation Although it is useful, m easuring growth rate using several time points may mislead in appreciating the comprehensive co ntributions that the client comp anies have made to the serving area. As discussed before, wages reported in the middle of the gauging point remain unaccounted for. It is due to the rationality of the traditional method of evaluation studies for LED that measures the difference between before a nd after. The method is not bad in the sense that it marginalizes temporary achievement appeared in the middle of study period and disappeared at the moment of inve stigation. In this paradigm, therefore, such an observation is not regarded as a consequence of a program. In addition, subtracting last wages with the wages at the first year is only a va lid method when linear growth is generally accepted. In the application of this paradigm to evaluation of a technology incubator, it may be misleading to marginalize the significance of wage fluctu ation in between since those wages indeed contributed to the economies of the serving area. To take those unaccounted wages into consideration, total wages are reviewed in the next section. Total wages, before graduation. Table 5-38 summarizes the to tal wages created by client companies of the public and private incubators Total wages create d during residency is

PAGE 141

141 produced by adding up all the wage s paid by client companies of a group. Again, the number of years and season are not taken into consideration in this analysis The Figure 5-28 demonstrates frequency distribution of the total week ly wages, paid during the residency. The total amount of dollars paid as wages by grou p Public seem to be as much as twice the wages estimated by the private incubator. As the number of client is greater for the public incubator, the total estimated wage of the group Public with $1,411,512, is about two times greater than the group Private, of $749,621, Table 5-19 provides total estimated wages of their post-graduation period. In terms of total weekly wage, public incuba tor appeared to have paid more than the private incubator. After converting total into average, the order is reversed. The client companies of the private incubator have pa id average $49,974.73 as weekly wage while they were in the incubator, while the average weekly wage of the client companies of the public incubator was only $39,208.67. The difference seems to be also insignificant for p-value, 0.950, higher than 0.20. One should be cautious in interpreting this findi ng in that total amount of wages is always sensitive to the number of companies and the leng thy of residency. The only implication can be drawn this statistics is that the there is differen ce in terms of magnitude of economic impact that clients of two incubator have contributed to while they were in the incubators. Total wages, after graduation. The Figure 5-29 shows frequency distribution of total wages, reported after graduation. In compared to the total wage obser ved during the residency, the gap has become narrower when the client co mpanies have graduated. The weekly wage client companies of the private incubator ha ve reached $2,575,678, while the counter part paid slightly more than that, with $2,748,056. Both groups have companies paid zero as weekly wage.

PAGE 142

142 It may be due to no employment in the study pe riod. In Table 5-40, descriptive statistics are provided. The weekly wage per company then increa sed to $286,186.44 after they graduated for the private incubator. It is about five times incr ease from the previous period. The median of the Incubator GP, of $32,361, is, although, slightly smaller than that of the Incuba tor GG, of $48,340. Perhaps, it is because wide range of distributi on and extreme outliers. Table 5-41 presents test results that compared the two groups. With very large p-value 0.629, there is no evidence to support the difference of the two groups in terms of the total wages they paid after they graduated the incubators. In sum, in terms of total dollar amount paid as weekly wage, public incubator demonstrated greater contribution in either period. However, Incubator GP produced greater dollars paid as weekly wage for a co mpany, based on the average, although differences are not significant by Mann-Whitney test. Changes of weekly wages over time, before and after graduation Weekly w ages over time, before graduation. The next presentation is the aggregation of weekly wages reported each year, the most popular milestones of impact studies. The purpose of this presentation is to visualize changes of th e aggregated estimated weekly wage paid by the incubators over time. The aggregated weekly wages indicate differences in the amount of weekly wages that each group has produced each year However, it should not be interpreted as the public incubator performed better in helpi ng firms to produce more since Figure 5-30 only reports lump sum weekly wage records of the tw o groups, which disregards the number of tenant companies. Figure 5-30 demonstrates seasonal change of th e weekly wages, reported while they were in the incubator. After 1999, the year with a dr amatic increase, public incubator dominantly paid more wages than the private incubator. The domination continues until 2005, even after 2003

PAGE 143

143 when wage started dropping. Presumably, within this period, such large wages paid by the companies of the public incubator could have cont ributed to an increase of tax revenue. Tenant companies of the private incubator appear to ha ve maintained a similar level of wages creation through the study period, regardle ss of the number of the contri buting companies. Figure 5-31 presents changes of the weekly wages of tenant companies over time. It presents increases and decreases of weekly wages each year based on the record of the previous year. Weekly wage of the public incubator seems to be more fluctuating than that of the private incubator. The only period that the public incu bator maintained positive growth wages in two consecutive years is between 2001 and 2003. After that period, the public incubator continuously reported negative wage growth. Aggregated weekly wage provides a seasona l perspective of the performance of both incubators. As discussed, a policy maker could easily find out that the public incubator collectively paid more weekly wages than the private incubator from 1999 through 2005. However as it is lump sum dollar amount of weekly wages, it does not tell the number of client companies contributing to the aggregation. The Figure 5-32 provides weekly wage per company to account for the number of tenant companies. Weekly wages per company, before graduation. According to Figure 5-32, the previous hypothesis, the aggregated weekly wage affect ed by the number of tenants was valid. The dominancy of the public incubator over virtually all period does not appear Figure 5-32. Rather, the tenant companies of private incubator seemed to pay more wages than the tenants of the public incubator. Average weekly wage of the year 2001 of the priv ate incubator reached $28,000, while it was only about $12,776 for the public incubator. However, after 2000, the

PAGE 144

144 wage of the public incubator maintained over $1 0,000, but the weekly wage of private incubator dropped below $10,000 at year 2007. Weekly wages over time, after graduation The study period of this section is from the year of graduation to the last year that the data of each company are available. For most of the record, the last year with the most recent availabi lity of NETS data is 2007. But the last years of some companies were prior to 2007 because they might have stopped their businesses. The last years of such companies are the year the last employment data was available, because estimated wages are multiplied by the number of jobs. In the Figure 5-33, both incuba tors showed fast growth af ter 2001. Public incubators domination again appeared afte r 2001, but the weekly wages of the private incubator also increases at similar phase. Maybe such a radical increase is due to the explosive increase of the number of client companies incepted to the public incubator starting from the early 2000s. The number jumped from two in 1998 to 17 in 2000, wh ereas there were only two tenant companies at the same period in the private incubator. Figure 5-34 presents changes of weekly wages the graduate companies have made from the previous year. As illustrated, both incubators recorded positive growth after 2001, although there were some fluctuations. Weekly wages per company, after graduation. The increase of weekly wage of the public incubator, observed in the Figure 5-34, se ems to be the result of the increase of the number of contributing graduate companies. In the Figure 5-35, the weekly wage per graduate company of the public incubato r stayed about $20,000 from 2001 to 2007. In the same period, the graduate companies of the private incubator explosively increased. Th e highest wage for the private incubator was $ 110,319 at year 2006. On the other hand, the maximum weekly wage of

PAGE 145

145 the public incubator was record ed, with $27,663, at year of 2003 and the second highest wage was $22,674 at year 2006. Recapping this section, although the greater total wage should not necessarily be interpreted as a positive growth in that it is also sensitive to change of number of employee, it is a good indicator that demonstrates degrees of co ntribution to the local economies. Also, in reality, given program capacity that determines si ze of economic contribution is hard to change as is in hypothetical situation. First higher wage imply positive contribution to improving level of income. Second, higher wage also means that companies have increased local tax revenue in that the wage eventually returns as income ta x. Third, higher wage is likely to stimulate circulation of wealth within the area. One possibl e application is that industry with higher wage may have greater multiplier effect than indus tries with lower wages. It may leads to formulations of local capital stocks that are potential stimulation for further development. Two limitations to this method exist, however. Firs t, this method relies on the estimated average weekly wage, so there may be differences in th e actual values. Second, by using the estimated average weekly wage, this method does not account for differences between occupations within a company. Obviously, the wage of CEO will be diffe rent than that of a researcher. Fortunately, the subject of this research being young, sma ll companies would hopefully render this bias negligible. In sum, in terms of total dollar amount paid as weekly wage, public incubator demonstrated greater contribution in either period. Both incubator companies increased weekly wages positively. One interesting finding is th at the private incubator showed higher wage growth rate, with statistical significance, when they are in the incubator. After graduated the

PAGE 146

146 incubator, clients of the public in cubator tend to grow better. This pattern is reassured by greater annual wage growth rate as well. Growth in Sales This sec tion reviews the sale s record and business performa nce of the client companies during the same two periods. Again, the first period was when they were being helped by the two incubators. The second period was after gra duation until the latest year that data was available. The major purpose is to compare the business performance in terms of sales and to find out if there is any difference between the two groups. The subject of the analysis is the sales record of all the companies during the two study periods, available in th e NETS database. This section tests the sixth hypothe sis of this research. h6: TBIs in the private sector re port greater sales revenues than TBIs in the public sector. Sales growth before and after graduation Grow th in sales. Emulating previous chapters, the first section compares growth in sales between two periods. Sales increases while the companies are under the di rect influence of the incubators are compared with that of the postgraduation period, in order to see if there was a difference between the programs. Again, sales growth means the difference between the amount of dollars at the year of gradua tion and also at the year of admi ssion. As illustrated in Figure 536, the sales growth of all the clients in both gr oups had a wide range of distribution with a couple of outliers. The test hypothesis is as following. h6a: There is no difference in sales growth betw een the client companies of Incubator GP and the client companies of Incubato r GG, before and after graduation. H0 : F1 F2 = 0 H1 : F1 F2 0

PAGE 147

147 Table 5-42 summarizes descriptive statistics The client companies of the private incubator appeared to have gain ed an average of $1,389,270 in sales growth during the residency. In the same period, client companies of the pub lic incubator have lost an average of $36,550. Based on the average, during the re sidency, Incubator GP showed pos itive growth in sales, while Incubator GG ended up losing sales. Due to extreme outliers, the median may summarize the distribution more appropriately than the means. The median of Incubator GP, of $159,800, is greater than the median of Inc ubator GG, of zero. Also the maximum growth achieved by clients of Incubator GP seems to be about four times greater than the maximum growth of clients of Incubator GG, while a client of Incubato r GG seems to have lost $7,815,000 when it was graduating the incubator. The difference of th e two groups seems to be obvious and is supported by the Mann-Whitney test summarized in the Table 5-43 that genera ted a p-value of 0.180, below the 0.20 confidence level of this study. Sales growth after graduation is the difference between the sales record of the latest year and the amount of dollars in the y ear of graduation. It is to meas ure how much growth the client companies have achieved after they become independent from both incubato rs up to the point of investigation. The positive growth of the gr oup Public changed after they graduated the incubator. Also, average sa les of the group Private also dropped to $257,521 while the group still maintained positive growth from $1,389,270 in the previous period. After graduation, again, client companies of the pub lic incubator reported a loss in terms of average sales. The loss was about $339,741 per company. However, with a given distribution illustrated in Figure 5-37 that has an extreme outlier in Incubator GG, the negative average may mislead the actual distribution of the growth rate. Figure 5-37 shows that many client s of Incubator GG also demonstr ated a positive sales growth,

PAGE 148

148 while a number of clients indeed ended up repor ting negative growth in sales. The negative average seems to be a result of the huge loss reco rded by one u client. Based on the median of Incubator GP, zero, is greater than the median of Incubator GG, of -$12,650, still the same conclusion is reached, as the clients of Incubato r GP produced more sales growth than Incubator GG in general, in the post-graduation period. Ta ble 5-44 presents descript ive statistics of the sales growth. Based on the MannWhitney test resulted in 0.798, a larger p-value than 0.20, the two groups are not statis tically significant. Sales growth implies the amount of dollars one client company would have made during the two periods. It might also refer to the amoun t of tax revenue each client might have paid. The difference is indicated only from the peri od of residency by the Mann-Whitney test. Although the same test revealed th at both incubators are not signi ficantly different in the sales growth achieved after graduation, their economic impact on the region may be different because the amount of dollars created differs. However, sales growth should not be interpreted as an indicator of the level of performance of both incubators because the amount of sales created are sensitive to the total sales, sales of the base year, and finally, the length of residency. Sales growth rates. The next analysis presents compar isons of the growth rates of sales that account for variations in the initial size of sales; growth rates proportionally increase or decrease from the sales of the base year. For instance, when a company raised its sales $10,000 a year and the companys sales in the first ye ar was $40,000, the growth rate was 25%. Whereas, for a company with initial sales of $50,000, an in creased growth rate of $10,000 is only 20%. The implication of the amount of the money is di fferent, while the created amount of sales is the same. The growth rate is essential to account fo r variations in the firs t year of sales, which

PAGE 149

149 results in some changes in the patterns observe d from the analysis of growth. The test hypothesis of this section follows. h6b: There is no difference in sales growth rates between the client companies of Incubator GP and the client companies of Incubator GG, before and after graduation. H0 : F1 F2 = 0 H1 : F1 F2 0 Figure 5-38 demonstrates that the distribution of Incubator GP has one extreme outlier of 19.64. Even without the outlier, the distribution is skewed to the right which indicates a violation of the normality assumption. The Mann-Wh itney test is again the appropriate method. Based on the descriptive statistics of the sa les growth rates, presented in Table 5-46, Incubator GP seems to report a higher growth rate, with an aver age of 344 percent, while it is only 49 percent for Incubator GG. The higher av erage of Incubator GP is, however, affected by the extreme outlier. A greater median for In cubator GP, of 1.476, nevertheless, still supports higher growth rates. Mann-Whitn ey test results, presented in Table 5-45, support the difference with a p-value of 0.123 which is smaller than the 0.20 level of conf idence for this study. During the residency, the companies of the private incubator achieved an explosive average sales increase of 340% (although is the incr ease is inflated to large extent). However, the remarkable phase of growth slowed down after they graduated the incubator; they ended up reporting negative growth. After they graduated, the sales growth rate of Incubator GP fell to about an average 4%, based on the summary of Ta ble 5-48. The client companies of Incubator GG achieved about 49.1 % in sales increase during the residency. But after graduation, the sales growth rate doubled to about 107.5%, in average. Figure 5-39 demonstrates the frequency of the distribution of the sales growth rates achieved by clients after they graduated. Two extreme

PAGE 150

150 outliers are observed from the clients of Inc ubator GG. The doubled sales growth rate is possibly a result of the calculation including these two extreme outliers. The negative median of Incubator GG, -0.034, pr esented in Table 5-48, affirms an inflated average of Incubator GG. The median of 0 for Incubator GP indicates slightly higher sales growth rates for Incubator GP than for Incubato r GG. However, the difference is not supported by the Mann-Whitney test with a p-value, 0.712 being greater than the acceptable level of 0.20. Tests supported the differences as significan t for the sales growth reported during the residency. One fair implication drawn from this analysis is that the client companies of the private incubator saw larger sales growth duri ng the residency. One cannot conclude, however, the level of performance of both incubators may be different since period ical growth rates are also sensitive to the number of years that companies accumulate dollar amounts. The longer they stayed in the incubator, the larger growth ra tes they experienced. The next section presents annual growth rates that indivi dual companies have achieved. Annual sales growth rate. Annual sales growth rate is an alyzed to account for the length of residency and post graduation period. It indicates the prop ortional growth achieved every year. The test hypothesis is as following. h6c: There is no difference in annual sales grow th rates between the client companies of Incubator GP and the client companies of Incubator GG, before and after graduation. H0 : F1 F2 = 0 H1 : F1 F2 0 Figure 5-40 demonstrates a wide range of distribution that is skewed to the right. According to descriptive statis tics in Table 5-48, the maximum of Incubator GP is 245%, 194% is the maximum for Incubator GG.

PAGE 151

151 In Table 5-50, it is shown that the client comp anies of the private incubator achieved about 57% in sales growth annually while they were in the incubator. The growth rate of the client companies from the public incubator during thei r residency seem to be lower than the counterpart, with only 17.4% a year. The difference in annual growth rate of sales during the residency between the two groups seems insignifi cant according to the Mann-Whitney test; a pvalue of 0.242 being greater than the 0.20 confiden ce level of this study, as presented in the Table 5-51. After graduation, however, the cl ients of Incubator GP repor ted negative growth. The sales seem to have decreased by 4% every year. The clients of Incubator GP, who grew by 17% a year during the residency, seem to have rec overed the deficit by maki ng 23% in annual growth after graduation. Figure 5-41 shows the frequency of the distribution of the annual sales growth. The public incubator, Incubator GG, has several extreme outliers. The 23% annual sales growth of Incubator GG seems to be perhap s inflated by the extreme outliers. Based on the descriptive statistics presente d in Table 5-52, a client of Incubator GG achieved 412% in sales growth annually, while the maximum annual sales growth of Incubator GP remained at only 4.5%. The average of Incu bator GG being inflated a nd having a client lost 24% annually, the median of Incubator GG, of -.0.013 is smaller than the median of Incubator GP being zero. Although the client companies of the private incubator suffered from negative growth after graduation, and the counterpart enjoyed a positiv e annual growth after graduation, the difference does not appear to be significant according to th e result of the Mann-Whitney test generating a pvalue of .798, above 0.20, as presented in Table 5-53.

PAGE 152

152 Three dimensions of growth rate are explored thus far. Sales growth indicates the extent to which both client groups have contributed to the local economy by producing sales. Changes in the total amount of sales effectively illustra te the quantified impli cation of the economic contribution. During the residency, sales growth achieved by Incubator GP turned out to be greater than Incubator GG. But it may be due to a longer length of residency. Unfortunately, the sales growth of Incubator GG in the post-graduation period, turned out to be negative, but it was due to sensitivity to the aggr egated dollar amount. If the su m of total sales of individual companies came out negative, affected by the bi g loss of a particular company, it could have made the growth negative for total sales growth The growth changes from negative to positive in the following analyses, which presented the av erage growth rate and annual growth rate of each company because they showed proportional growth regardless of the in itial sales size. The last table presented an annual grow th rate of sales in order for the analysis to account for number of business years. During the residency, sales growth rates of Incuba tor GP again turned out to be superior to Incubator GG. Although the Mann-Whitney test failed to support this, the same pattern is observed from an analysis of annual growth rates during the same period. In the post-graduation period, Incubator GP still maintained a superior position, although the rate s are not statistically significant. One implication may be drawn as the hypothesis of this section. An assumed higher sales growth being observed from a private incuba tor is moderately supported. The difference is explicit especially while they were in the facility. It became less explicit after they graduated the incubators. Total sales, before and after graduation Total sales, before grad uation The primary concern of th is section is to quantify economic contributions made by the sales of the gr oups of client companies. The aggregation of

PAGE 153

153 sales refers to the summation of all sales reported every year during each period. The total sales are the aggregation of all sales for both groups. Again, the number of years and seasons are not taken into consideration in this analysis. Figure 5-42 demonstrates the frequency of the distribution of total sales reporte d before graduation. The distribut ion displays a long tail to the right with several extreme outliers. Table 5-54 presents a summary of descriptiv e statistics of the total sales for both incubators during the study peri ods. The total amount of dollars created as sales by Incubator GP seem to be twice as much as the sales repor ted by the group Private. The maximum record of sales of the group Public, with $51,685,000, is also nearly two times greater than the maximum sales of the group Private, w ith $28,779,070. Among the clients of Incubator GG, however, five companies failed to make sales while they were in the facility. Two companies among the five were categorized as research rela ted businesses, which usually take a long time to make a profit. The Mann-Whitney test, in Table 5-55, does not fi nd the difference to be significant either, by returning a probability of 0.591, greater than the 0.20 confidence level of this study. Total sales, after graduation Figure 5-38 demonstrates the frequency of the distribution of total sales reported after graduation. Again, the distribution displays a long tail to the right with several extreme outliers in both Incubator GP and Incubator GG. Table 5-40 provides the summary of descriptive statistics of total sales of the companys post-graduation period. The total sales generated by Incubator GG, of $237,721,364, are twice as much as Incubator GP has produced, $106,178,359, even in the period of post-graduation. This time, both groups reported companies with zer o sales after graduation since both groups had companies go out of business without making any sales after graduation.

PAGE 154

154 Obviously, bigger total sales are the result of a large number of client companies in both periods. This is supported by the large average total sa les of Incubator GP, being $11,797,595.44, while Incubator GGs sales dropped to $6,603,371.22. The medians are however very similar. This difference is due to a wide range of distribution and extreme outliers for both incubators. The median of the Incubato r GP, of $3,908,000 is sligh tly greater than that of the Incubator GGs $2,603,007. Thus, the Mann-Whitney, in Table 5-57, test does not find the difference to be significant either, by retu rning a probability of 0.561, greater than the 0.20 confidence level of this study. Changes of total sales over time before and after graduation Total sales over time, b efore graduation Figure 5-44 demonstrates the amount of sales dollars produced by tenant companies of both incubators. The total numbers of sales are summations of sales records by all tenant companie s that were residents of each incubator at the corresponding period. A fast sales increase is observed between 1999 and 2003 from the tenant companies of the public incubator. At a glance, the public incuba tor seems to have created a strong impact on the region by increasing outstanding sales in virtua lly all periods of time since 1998. Again, it should have contributed to an in crease of tax revenue. Tenant co mpanies of the private incubator appear to have maintained a similar level of sales creation through the study period. The group reported $11,919,100 as maximum sales in 1999, whereas it was $52,604,200 for the counterpart in 2003. Figure 5-45 presents changes of the to tal sales of tenants over time. It presents increases and decreases of sales each year based on the record of the previous year. Tenants of the public incubator showed a bumpy sales record between 2001 and 2004. Sales per company, before graduation. In Figure 5-46, the tota l sales are divided by the number of tenant companies of each year. As de monstrated, the tenant companies of the private

PAGE 155

155 incubator produced more sales in general un til 2002, which is a contrasting finding from previous observations. According to Figure5-46, the pr evious hypothesis, th e total sales affect ed by the number of tenants seems to be right. Recall that the total sales of the public incubator after 2000 was greater than the counterpart; th e average sales record of the public incubator after 2000 is relatively low in Figure5-45. It means that the steep increase observed from total sales was possibly caused by the increase of the number of client companies. Figure5-47 evidences this hypothesis, showing an explosiv e increase of the tenant comp any after 2000 in the public incubator. As a result, the average sales of th e group Public become rather constant through the study period after a dramatic in crease in 1993. In addition, averag e sales of the tenants of the group Private show a fast increase in 2000, which di d not explicitly stand out in the previous section. Changes of total sales over time, after graduation. Again, the study period of this section is from the year of graduation to the last year that the data of each company are available. Total sales are summations of all sales records reported within the period defined by the described method. Figure 5-48 dem onstrates total sales created by the graduate companies of the private and public incubators. Recall, betw een 2001 and 2003, the total sales of tenant companies of the public incubator skyrocketed as illustrated. A similar pattern is observed in Figure 5-48, even from the graduate companies. Total sales of the group Public increased from 2001 through 2004, which also dominates the sale of graduate companies of the pr ivate incubator. Such a radical increase is due to the explosiv e increase of the number of client companies incepted to the public incubator starting from th e early 2000s. The number jumped from two in 1998 to 17 in 2000, whereas there were only two tenant companies at the same period in the

PAGE 156

156 private incubator. The maximum number of tenants was only six in 2004, according to Figure 547. As mentioned previously, the aggregated dol lar amount, however, should not be interpreted as an indicator of efficiency. Sales records of individual client comp anies are embedded, but do not stand out from the angels of these presenta tions. To observe average sales per company, Figure 5-49 presents changes of sales the graduate companies have made from the previous year. For the group Public, it shows a continuous increase from 2002 through 2004. After the sharp drop in 2005, negative growth continued, while the group Private achieved moderate growth during the same period after two years of consecutive losses between 2002 and 2003. Sales per company, after graduation. As discussed, the total sales does not account for the number of companies. Figure 5-50 demonstrat es total sales per company. Again, the amount of dollars in the Y axis represents the amount of sales a graduate company made in the corresponding period. Two-year co nsecutive loss re-appears in Figure 5-50 from the line of the group Private in 2002 and 2003. Nevertheless, the y early sales total of graduate companies of the private incubator seem to be larger (most of the time) than yearly sales created by the graduate companies of the public incubator. Figure 5-51 provides the number of graduate companies that has created sales each year. Again, such an explosive increase in the number of graduate companies of the public incubator after 2002 seems to be the reason for th e expansion of total sales, whereas average sales were actually not larger than that of the group Private. However, it is arguable that larger capacity of the public incubator might be a unique strength of the incubator founded by the public sector. It is closely associated with large investments that private agents can hardly utilize.

PAGE 157

157 Table 5-1. Major hiring industries of Maryland* Industry Statewide Employment Share Location Quotient Trade, Transportation, and Utilities 22.65%0.98 Professional and Business services 19.02%1.21 Education and Health services 17.22%1.13 Leisure and Hospitality 11.23%0.96 Retail trade 14.45%1.06 Health care and Social assistance 14.52%1.09 Bureau of Labor Statistics Table 5-2. Major technology i ndustries of Maryland 2001-2005* Industries Employment Payroll ($) Average Weekly Wage ($) 2001 2002 2003 2005 2001 2002 2003 2005 Aerospace 62,659 63,049 63,616 Na 4,475,479,587 4,558,917,880 4,717,201,657 NA 1,426 Information Technology 96,222 90,856 88,334 89,564 NA 6,667,648,817 NA 7,216,116,727 1,549 Healthcare 233,838 241,677 248,507 257,750 8,763,528,547 NA NA 11,232,295,480 838 Bioscience 33,080 35,669 36,964 NA 2,188,718,204 2,280,394,146 2,455,501,963 NA 1,277 *www.ChooseMaryland.org Table 5-3 Summary of sample incubators Sponsorship Incubato r code City County Private GP Rockville Montgomery Public GG Rockville Montgomery

PAGE 158

158 Figure 5-1. Geography of technology business incubators, Rockville, MD

PAGE 159

159 Figure 5-2. Geography of County Busi ness Personal Property tax rate

PAGE 160

160 Table 5-4. Regional descripti on of incubator locations Frederick (Frederick) Rockville (Montgomery) Annapolis (Anne Arundel) Columbia (Howard) Total population (2000) 52,693 47,257 35,806 88,391 Population density (2000) 2,580.831 3,514.394 5,321.224 3,206.972 Pct. pers. 25+ yrs. old with a bachelors or graduate/prof. degree (2000) 29.9% 52.9% 38.7% 59% Pct. foreign born population (2000) 7.3% 31% 9.7% 13.2% Median household income last yr ($) (2000) $47,700 $68,074 $49,243 $71,524 Average household income last yr ($) (2000) $57,064 $82,211 $63,586 $83,876 Median household income last yr ($) (2000) $47,700 $68,074 $49,243 $71,524 Average household income last yr ($) (2000) $57,064 $82,211 $63,586 $83,876 Total number of establishments (2003) 1,764 N/A 1,367 3,033 Number of establishments per 1,000 population (2003) 33 N/A 38 34 Median year structure built (2000) 1979 1967 1968 1980 Pct. housing units in single-family detached homes (2000) 34.6% 58.9% 39.6% 41.8% Homeownership rate (2000) 55.9% 67.9% 51.9% 66% Average value of specified owneroccupied housing units ($) (2000) $154,194 $230,582 $218,519 $202,120 Pct. pop. 16 years old and over who are employed 69.8% 64.9% 67.8% 74.7% Pct. persons 16+ years old employed in mgmt. occ. (incl. farms) (2000) 39.2% 55.6% 42.9% 62.3% Pct. persons 16+ years old employed in sales and related occ. (2000) 25.9% 22% 24.9% 21.9% Pct. persons 16+ years old employed in production occupations (2000) 9.8% 4.8% 6.9% 3.5%

PAGE 161

161 Table 5-4. Continued Frederick (Frederick) Rockville (Montgomery) Annapolis (Anne Arundel) Columbia (Howard) Pct. employment in establishments with 10-49 employees (2003) 29.5% N/A 33.8% 25% Number of commercial banks per 1,000 population (2003) 0.6 N/A 0.6 0.4 Table 5-5. Top three hiring industries* Industries Percentage of Employment Number of Employment Location Quotient Maryland Trade, Transportation, and Utilities 22.56%472,497NA Professional and Business Services 19.02%396,619NA Education and Health Services 17.22%359,183NA Montgomery Professional and Business Services 27.12%103,191 1.43NAICS 541 Professional and Technical Services 16.5%62,799 1.55Trade, Transportation, and Utilities 16.45%62,639 0.73Howard Trade, Transportation, and Utilities 27.18%35,374 1.20Professional and Business Services 24.64%32,057 1.30NAICS 54 Professional and technical services 16.37%21,302 1.54Frederick Trade, Transportation, and Utilities 21.63%17,402 0.95Professional and Business Services 17.29%13,915 0.91NAICS 44-45 Retail trade 16.23%13,059 1.12Baltimore City Education and Health Services 34.38591,727 2.00NAICS 62 Health care and social assistance 25.12%67,010 1.73Professional and Business Services 15.89%42,400 0.84Anne Arundel Trade, Transportation, and Utilities 28%54,941 1.24Professional and Business Services 17.65%34,634 0.93NAICS 44-45 Retail trade 16.91%33,176 1.17 *Quarterly Census of Employment and Wa ges, Bureau of Labor Statistics

PAGE 162

162 Table 5-6. Summary of major employer by county* Type of Industry Number of Employer Total Employee Montgomery Federal government 8 38,881 Healthcare 6 19,672Sum 14 58,553 Howard Education 1 2,500Healthcare 5 3,618Information 1 2,028Manufacture 4 1,951Professional services 8 10,124Sum 19 20,221 Frederick Federal government 1 4,436Finance and insurance 6 5,103Healthcare 3 3,344Manufacturing 6 2,143Professional services 3 4,143Sum 19 19,169 Baltimore City Educational services 5 3,4604Healthcare 8 41,822Sum 13 76,426 Anne Arundel Federal government 2 31,425Manufacturing 1 8,000Professional services 6 6,090Retail trade 7 6,517Sum 16 52,032 *Department of Business and Economic Development

PAGE 163

163 Table 5-7. Employment and wage s of major technology industry* Baltimore Howard MontgomeryFrederick Anne Arundel Bioscience Empl. Change (2001-2003) 4845231285168 90 Average Weekly Wage (2003) 9521,4921,426989 1,322 Aerospace Empl. Change (2001-2003) 5632622433 449Average Weekly Wage (2003) 1,2181,5151,4251,038 1,611 Information technology Empl. Change (2002-2005) 509164-4,296262 564Average Weekly Wage (2005) 1,4641,6461,7511,333 1,531 Healthcare Empl. Change (2001-2005) 5,3711,2334,847698 1,911Average Weekly Wage (2005) 890771919784 834 Career and Workforce Information, Maryland Department of Labor Table 5-8 Tax incentiv e of municipalities* Maryland Department of Business and Economic Development ** per $100 of assessed value *** Brief Economic Facts by County Anne Arundel Baltimore City Frederick Howard Montgomery Local Personal Income Tax Rate 2.56%3.05%2.96%3.20% 3.20% County Business Personal Property Tax Rate ** $2.20$5.67$0.00$2.535 $1.652 Manufacturing/R&D Machinery and Equipment Exemption 100%100%NA100% 100% Manufacturing/R&D Inventory Exemption 100%100%NA100% 100% Corporate Income Tax*** None None None None None Acreage in State Enterprise Zones 021,61500 627

PAGE 164

164 Figure 5-3. Numbers of clie nt companies by zip code Table 5-9. Area incubators and other innovation centers Anne Arundel Baltimore City Frederick Howard Montgomery Incubators (locations) 1 2(3)1(2)1 6Technology Park & Centers 0 360 2 Table 5-10. Sample Client Companies Incubator code Location (City/County) Company Final Sample Size Proportion GP Rockville/ Montgomery County tenant 12 6 50% graduate 15 9 60% total 27 15 55.6% GG Multiple locations/ Montgomery County tenant 28 20 71.4% graduate 22 16 72.7% total 50 36 72%

PAGE 165

165 Figure 5-4. Sample distribution by age Table 5-11. Client companies by type of industry, total NAICS Code Number of Incubator Description Private Public Total 541710 7 15 22 Research and Development in the Physical, Engineering, and Life Sciences 517910 3 3 Other Telecommunications, Cellular and Other Wireless Telecommunications, Satellite Telecommunications 423450 2 2 Medical, Dental, and Hospital E quipment and Supplies Merchant Wholesalers 541511 2 2 Custom Computer Programming Services, Computer Programming Services 541350 1 2 3 Building Inspection Services 325412 1 1 Pharmaceutical Preparation Manufacturing 325414 1 1 Biological Product (except Diagnostic) Manufacturing 334519 1 1 Other Measuring a nd Controlling Device Manufacturing 339112 1 1 Surgical and Medical Instrument Manufacturing 325992 1 1 Photographic Film, Paper, Plate, and Chemical Manufacturing 541810 1 1 Advertising Agencies 334611 1 1 2 Software Reproducing 541512 1 1 Computer Systems Design Services 443120 1 1 Computer and Software Stores 541519 1 1 Other Computer Related Services 621111 1 1 2 Offices of Physicians 621112 0 Offices of Physicians, Mental Health Specialists" 621511 1 1 Medical Laboratories 621999 1 1 Health and Allied Services, NEC (e xcept blood and organ banks, medical artists, medical photography, and childbirth preparation classes) 541990 0 All Other Professional, Scientific, and Technical Services 541618 1 1 2 Other Management Consulting Services 541690 1 1 Other Scientific and Technical Consulting Services 532490 0 Medical Equipment Rental and Leasing NA 1 1 NA SUM 15 36 51

PAGE 166

166 Table 5-12. Client companies by type of industry, major only Private SIC8 Description Cnt. Avr. Age Range First Yr. Avr. First Yr. Employee Avr.First Yr. Sales 87310100 Biological Research 3 131989-1993 13 $691,533.3 87310102 Biotechnical Research, Commercial 3 52000-2003 10.3 $1,170,700.0 Public SIC8 87310000 Commercial Physical Research 3 71999-2001 5.7 $465,566.7 87310102 Biotechnical Research, Commercial 3 52000-2005 6.3 $387,100.0 87339902 Research Institute 3 4.661999-2003 7.7 $626,733.3 48999901 Data Communication Services 2 6.51999-2000 7.5 $770,600.0 73710301 Computer Software Development 2 9.51991-2004 5.5 $501,450.0 87330102 Biotechnical Research, Noncommercial 2 7.51998-2001 17.5 $2,316,200.0 87330103 Medical Research 2 3.51999-2003 34.0 $2,888,400.0 Table 5-13. Biotechnical researc h, noncommercial and medical re search in public incubators SIC 8 First Year Employee First Year Sales 87330102 Client 1 25$3,750,000.0 Client 2 10$882,400.0 Average 17.5$2,316,200.0 87330103 Client 1 65$5,525,000.0 Client 2 3$251,800.0 Average 34$2,888,400.0

PAGE 167

167 Frequency Distribution, Employment Growth, Before Graduation0 5 10 15 20 25 5 3 1 1 3 5 7 9 11 13 15 17 1 9 21 Private Public Figure 5-5. Frequency dist ribution, employment growth, before graduation Table 5-14. Descriptive Statistics, em ployment growth, before Graduation Employment Growth Achieved During the Residency Incubator GP Incubator GG COUNT 9 36 MAX. 74 21 MIN. -1 -45 STDEV. 24.05 9.37 MEDIAN 1 0 MEAN 11.22 1.06 Table 5-15. Mann-Whitney Test results, employment growth, before graduation Mean Rank* Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 27.39 122.5 788.5 .218 Public 21.90 *Mean rank is the sum of the ranks divided by the number of cases for each group

PAGE 168

168 Frequency Distribution, Employment Growth, After Graduation0 5 10 15 20 25 30 10 5051015202530354045505560657075 Private Public Figure 5-6 Frequency dist ributions, employment gr owth, after graduation Table 5-16. Descriptive statistics, em ployment growth, after graduation Employment Growth Achieved After Graduation Incubator GP Incubator GG COUNT 9 36 MAX. 73 33 MIN. -5 -10 STDEV. 8.48 8.56 MEDIAN 0 0 MEAN 10.89 3.22 Table 5-17. Mann-Whitney Test results, employment growth, after graduation Mean Rank* Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 22.67 159.0 204.0 .923 Public 23.08

PAGE 169

169 Frequency Distribution of Employment Growth Rate, Before Graduation0 5 10 15 20 25 30 1012345678910111213 Private Public Figure 5-7. Frequency distri butions, employment growth rate, before graduation Table 5-18. Descriptive Statis tics, employment growth ra tes, before graduation Employment Growth Rates Achieved During the Residency Incubator GP Incubator GG Count 9 36 MAX. 12.33 6 MIN. -0.06 -1 STDEV. 4.03 1.35 MEDIAN 0.25 0 MEAN 1.641 0.453 Table 5-19. Mann-Whitney Test results, empl oyment growth rate, before graduation Mean Rank* Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 27.06 125.5 791.5 .249 Public 21.99

PAGE 170

170 Frequency Distribution of Employment Growth Rate, After Graduation0 5 10 15 20 25 30 10123456789101112131415161718 Private Public Figure 5-8. Frequency distri butions, employment growth rate, after graduation Table 5-20. Descriptive statistics, employment growth Rate, after graduation Employment Growth Rate, After Graduation Incubator GP Incubator GG Count 9 36 MAX. 3.31 17 MIN. -0.26 -0.5 STDEV. 1.11 3.08 MEDIAN 0 0 MEAN 0.381 0.848

PAGE 171

171 Frequency Distribution of Annual Employment Growth Rate, Before Graduation0 5 10 15 20 25 30 0.500.511.522.5 Private Public Figure 5-9. Frequency distribut ions, annual employment growth rate, before graduation Table 5-21. Mann-Whitney Test results, empl oyment growth rate, after graduation Mean Rank* MannWhitney U Wilcoxon W Sig.(2tailed) Private 22.22 155.0 200.0 .820 Public 23.19 Table 5-22. Descriptive statistics, annual em ployment growth rate, before graduation Annual Employment Growth Rate During the Residency Incubator GP Incubator GG Count 9 36 MAX. 1.55 2 MIN. -0.01 -0.33 STDEV. 0.52 0.47 Median 0.05 0 Mean 0.26 0.17

PAGE 172

172 Table 5-23. Mann-Whitney Test results, annual employment growth rate, before graduation Mean Rank* Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 26.61 129.5 795.5 .305 Public 22.10 Frequency Distribution of Annual Employment Growth Rate, After Graduation0 5 10 15 20 25 30 0.500.511.522.533.54 Private Public Figure 5-10. Frequency distri butions, annual employment grow th rate, after graduation Table 5-24. Descriptive statistics, annual employment growth rate, after graduation Annual Employment Growth Rate Achieved During the Residency Incubator GP Incubator GG Count 9 36 MAX. 0.83 3.4 MIN. -0.02 -0.13 STDEV. 0.27 0.63 Median 0 0 Mean 0.1 0.17

PAGE 173

173 Table 5-25. Mann-Whitney Test results, annual employment growth rate, after graduation Mean Rank* Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 22.11 154.000 199.000 .794 Public 23.22 Total Number of Jobs, Tenants, 1990-20070 50 100 150 200 2501990 1991 19 9 2 19 9 3 1 9 94 19 9 5 1996 19 97 19 9 8 1 9 99 2 0 00 2001 2002 20 0 3 2 0 04 2 0 05 2006 2007 Private Public Figure 5-11. Total number of jobs, tenants, 1990-2007

PAGE 174

174 Change in Number of Jobs by Year, Tenants, 1990-2007-150 -100 -50 0 50 100 1501 990 199 1 1 992 19 93 1994 19 95 199 6 1 997 199 8 1 999 200 0 2 001 20 02 2003 20 04 2005 20 06 200 7 Private Public Figure 5-12. Change in number of jobs by year, tenants, 1990-2007 Number of Jobs per Client, Tenants, 1990-20070.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.01 9 90 19 9 1 1992 19 93 1994 19 9 5 1 9 96 1 9 97 1998 19 99 2000 2 0 01 20 0 2 2 0 03 20 0 4 2 0 05 20 06 2007 Private Public Figure 5-13. Number of jobs per client, tenants, 1990-2007

PAGE 175

175 Number of Tenants, 1990-2007 0 5 10 15 20 251 9 90 1991 1 9 92 1993 19 9 4 19 9 5 19 9 6 1 9 97 1998 1 9 99 2000 20 0 1 20 02 20 0 3 2 0 04 2005 2 0 06 2007 Private Public Figure 5-14. Number of tenants, 1990-2007 Total Number of Jobs, Graduates, 1990-20070 50 100 150 200 250 300 350 400 450 5001990 1991 19 9 2 19 9 3 1 9 94 19 9 5 1996 19 97 19 9 8 1 9 99 2 0 00 2001 2002 20 0 3 2 0 04 2 0 05 2006 2007 Private Public Figure 5-15. Total number of jobs, graduates, 1990-2007

PAGE 176

176 Change in Number of Jobs by Year, Graduates, 1990-2007-150 -100 -50 0 50 100 150 2001990 1 99 1 1992 1 9 93 1 994 1 9 95 1 99 6 1997 1 99 8 1999 2 00 0 2001 2 0 02 2 003 2 0 04 2 005 2 0 06 2 00 7 Private Public Figure 5-16. Change in number of jobs by year, graduates, 1990-2007 Number of Jobs per Client, Graduates, 1990-20070.0 20.0 40.0 60.0 80.0 100.0 120.0 140.019 90 1992 19 94 1996 1998 20 0 0 2002 20 0 4 20 0 6 Private Public Figure 5-17. Number of jobs per client, graduates, 1990-2007

PAGE 177

177 Number of Graduates, 1990-20070 5 10 15 20 25 301 9 90 19 91 1 9 92 1 9 93 1994 1 9 95 1996 1 9 97 1998 1999 20 0 0 2001 20 0 2 20 0 3 20 0 4 20 0 5 20 06 20 0 7 Private Public Figure 5-18. Number of graduates, 1990-2007 Number of Jobs and Estmtd.Total Weekly Wage, Private 1990-20070 20 40 60 80 100 1201990 1992 199 4 199 6 1998 200 0 200 2 2004 200 6Thousands Wage Number of Jobs*1000 Figure 5-19. Number of jobs a nd wage effect, private 1990-2007

PAGE 178

178 Number of Jobs and and Seasonal Change of the Estmtd. Weekly Wage, 1990-20070 5 10 15 20 25 30 3519 90 19 9 2 19 9 4 1996 1998 2000 2 0 02 2 0 04 20 06Thousands Estimated Wage Number of Jobs *1000 Figure 5-20. Number of jobs and seasonal change of the estimated weekly wage, 1990-2007 Average Weekly Wage, NAICS 541710, 1990-20070 200 400 600 800 1000 1200 1400 1600 18001 9 90 19 9 2 19 94 1996 19 9 8 2000 2002 2 0 04 20 0 6 Average Weekly Wage, NAICS 541710, 1990-2007 Figure 5-21. Average weekly wage of NAICS 541710, 1990-2007

PAGE 179

179 Frequency Distribution of Wage Growth, Before Graduation0 2 4 6 8 10 12 14 16 18 55000 4 5 0 0 0 35000 25000 15000 5000 500 0 1 5 0 0 0 25000 Private Public Figure 5-22. Frequency distributions wage growth, before graduation Table 5-26. Descriptive statistics, wage growth, before graduation Wage Growth Achieved During the Residency Incubator GP Incubator GG COUNT 9 36 MAX. $21,000 $15,856 MIN. $0 $-53,265 STDEV. $7,120 $11,035 MEDIAN $1,946 $0 MEAN $5,026 $863 Table 5-27. Mann-Whitney Test results wage growth, before graduation Mean Rank Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 29.28 105.5 771.5 .106 Public 21.43

PAGE 180

180 Frequency Distribution of Wage Growth, After Graduation0 5 10 15 20 25 1 5000 500 0 5000 15000 25000 350 0 0 450 0 0 5 5000 65000 75000 85000 More Private Pubic Figure 5-23 Frequency di stributions, wage grow th, after graduation Table 5-28. Descriptive Statistics Wage Growth, After Graduation Wage Growth Achieved After Graduation Incubator GP Incubator GG COUNT 9 36 MAX. $477,840 $56,232 MIN. $-10,007 $-10,568 STDEV. $158,430 $13,807 MEDIAN $350 $1,318.5 MEAN $60,839 $6,885 Table 5-29. Mann-Whitney Test results wage growth, after graduation Mean Rank Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 22.28 155.5 200.5 .853 Public 23.18

PAGE 181

181 Frequency Distribution of Wage Growth Rate, Before Graduation0 5 10 15 20 25 1 0. 5 0 0. 5 1 1. 5 2 2. 5 3 3. 5 4 4. 5 5 5. 5 6 6. 5 7 7. 5 Private Public Figure 5-24. Frequency distributions, wage growth rates, before graduation Table 5-30. Descriptive statistics, wage growth rates, before graduation Wage Growth Rate Achieved During the Residency Incubator GP Incubator GG COUNT 9 36 MAX. 1.801 6.165 MIN. 0 -1 STDEV. 0.732 1.122 MEDIAN 0.093 0 MEAN 0.535 0.338 Table 5-31. Mann-Whitney Test results, wa ge growth rates, before graduation Mean Rank Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 30.06 98.5 764.5 .065 Public 21.24

PAGE 182

182 Frequency Distribution of Wage Growth Rate, After Graduation0 2 4 6 8 10 12 14 1 0.5 0 0 5 1 1.5 2 2 5 3 3.5 4 4 5 5 Private Public Figure 5-25. Frequency distribution, wage growth ra tes, after graduation Table 5-32. Descriptive statistics, wa ge growth rates, after graduation Wage Growth Rate After Graduation Incubator GP Incubator GG COUNT 9 36 MAX. 4.975 16.988 MIN. -0.601 -0.614 STDEV. 1.731 3.178 MEDIAN 0.012 0.087 MEAN 0.710 1.039 Table 5-33. Mann-Whitney Test results, wa ge growth rates, after graduation Mean Rank Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 20.50 139.500184.500 .522 Public 23.63

PAGE 183

183 Frequency Distribution of Annual Wage Growth Rate, Before Graduation0 5 10 15 20 25 0.5 0.300.250.50.7511.251.51.7522.25 Private Public Figure 5-26. Frequency distributions, annua l wage growth rates, before graduation Table 5-34. Descriptive statistics, annual wage growth rates, before graduation Annual Wage Growth Rate During the Residency Incubator GP Incubator GG COUNT 9 36 MAX. 0.860 2.055 MIN. 0 -0.333 STDEV. 0.276 0.395 MEDIAN 0.020 0 MEAN 0.141 0.126 Table 5-35. Mann-Whitney Test results, annual wage growth rates, before graduation Mean Rank Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 28.94 108.500 774.500 .120 Public 21.51

PAGE 184

184 Frequency Distribution of Annual Wage Growth Rate, After Graduation0 5 10 15 20 25 0.500.511.522.533.54 Private Public Figure 5-27. Frequency distributions, annual wage growth rates, after graduation Table 5-36. Descriptive statistics, annual wage growth rates, after graduation Annual Wage Growth Rate After Graduation Incubator GP Incubator GG COUNT 9 36 MAX. 1.244 3.398 MIN. -0.046 -0.123 STDEV. 0.412 0.041 MEDIAN 0.012 0.666 MEAN 0.171 0.221 Table 5-37. Mann-Whitney Test results, annual wage growth rates, after graduation Mean Rank Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 21.06 144.500189.500 .618 Public 23.49

PAGE 185

185 Frequency Distribution of Total Wage Created During the Residency0 1 2 3 4 5 6 7 8 9 100 2 0 0 0 0 40000 60000 8 0 0 0 0 100000 120000 140 0 0 0 160000 180000 2 0 0 0 0 0 220000 240000 2 6 0 0 0 0 280000 300000 Private Public Figure 5-28. Frequency distributions, total wages, before graduation Table 5-38. Total wages, descriptiv e statistics, duri ng the residency Total Wages Produced During the Residency Private Public COUNT $749,621 $1,411,512 MAX. $346,782 $409,065 MIN. $0 $0 STDEV. $91,353.17 $79,198.06 MEDIAN $14,731 $16,806 MEAN $49,974.73 $39,208.67 Three sources are used to produce this table. First, year of admission and graduation was necessary and they were obtained from the in cubators. Second, average weekly wage was prepared for each industry. This data is ava ilable as Quarterly Census of Employment and Wages, in Bureau of Labor Statistics. Thir d NETS database provides number of employee for each year. Table 5-39 Mann-Whitney Test results total wage, before graduation Mean Rank Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 26.20 267.000933.000 .950 Public 25.92

PAGE 186

186 Frequency Distribution of Total Wage Created After Graduation0 1 2 3 4 5 6 70 20000 40000 6 0000 80000 100000 120000 1 4 0 0 0 0 160 0 0 0 180000 200000 2 20000 2 4 0 0 0 0 26000 0 280000 300000 Private Public Figure 5-29 Frequency distributions, total wages, after graduation Table 5-40. Total wages, descrip tive statistics, after graduation Total Wages Produced After Graduation Private Public COUNT $2,575,678 $2,748,056 MAX. $1,823,901 $425,724 MIN. $0 $0 STDEV. $588,209.93 $102,680.78 MEDIAN $32,361 $48,340 MEAN $286,186.44 $76,334.89 Table 5-41. Mann-Whitney Test results total wages, after graduation Mean Rank Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 24.89 145.000 811.000 .629 Public 22.53

PAGE 187

187 Weekly Wages over Time, 1990-20070 50 100 150 200 250 300 3501990 19 91 19 9 2 1 9 93 19 9 4 1 9 95 19 9 6 1997 19 98 1999 20 0 0 20 0 1 2 0 02 20 0 3 2 0 04 2005 2006 20 0 7Thousands Private Public Figure 5-30. Total wages over time, tenants, 1990-2007 Changes of Weekly Wages, Tenants, 1990-2007-200 -150 -100 -50 0 50 100 150 200 2501991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007Thousands Private Public Figure 5-31. Changes of w eekly wages, tenants, 1990-2007

PAGE 188

188 Weekly Wages per Tenants, 1990-20070 5000 10000 15000 20000 25000 300001 9 90 19 9 2 1 9 94 1 9 96 19 9 8 2 0 00 20 0 2 2 0 04 20 0 6 Private Public Figure 5-32. Weekly wages per tenant, 1990-2007 Weekly Wages over Time, Graduates, 1990-20070 100 200 300 400 500 600 7001990 199 1 199 2 199 3 199 4 199 5 1996 1 99 7 199 8 199 9 2000 2 00 1 200 2 2 00 3 200 4 200 5 200 6 2 00 7Thousands Private Public Figure 5-33. Weekly wages over time, graduates, 1990-2007

PAGE 189

189 Changes of Weekly Wages over Time, Graduates, 1990-2007-150 -100 -50 0 50 100 150 200 250 3001990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007Thousands Private Public Figure 5-34. Changes of weekly wages over time, graduates, 1990-2007 Weekly Wages per Graduates, 1990-20070 20 40 60 80 100 1201 9 90 199 1 1 99 2 1993 1994 199 5 1 9 96 1 9 97 1 99 8 1 9 99 2000 200 1 2 0 02 2 0 03 2004 2 00 5 2 0 06 200 7Thousands Private Public Figure 5-35. Weekly wa ges per graduates, 1990-2007

PAGE 190

190 Frequency Distribution of Sales Growth, Before Graduation0 5 10 15 20 $8 000 ,000 $4,000,000 $1 000 ,00 0 $4 00 ,0 00 $100, 0 00 $1 00 ,000 $400,000 $7 00, 0 00 $1,000,000 $2,1 00 ,000 $6 000 ,00 0 Private Public Figure 5-36. Frequency distribution, total wage s, before graduation Table 5-42. Descriptive statistics, sales growth, before graduation Sales Growth Achieved During the Residency Incubator GP Incubator GG Count 3 36 MAX. $8,326,100 $2,100,000 MIN. -$540,000 -$7,815,000 STDEV. $2,774,295 $1,423,676 Median $159,800 $0 Mean $1,389,270 -$36,550 Table 5-43. Mann-Whitney Test results, sales growth, before graduation Mean Rank* Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 28.22 115.000781.000 .180 Public 21.69

PAGE 191

191 Frequency Distribution of Sales Growth, After Graduation0 5 10 15 20 25$ 2 8, 0 0 0,0 00 $25,000,000 $22,000,000 $ 1 9, 0 0 0,0 00 $ 16 0 0 0,0 00 $ 1 3,000, 0 00 $ 1 0,000, 0 00 $7 ,000, 00 0 $4 ,000, 00 0 $1,000 000 $2 ,000 ,00 0 Private Public Figure 5-37. Frequency distributions, sales growth, after graduation Table 5-44. Descriptive statistics sales growth, after graduation Sales Growth Achieved During the Residency Incubator GP Incubator GG Count 9 36 MAX. $2,745,400 $3,300,000 MIN. -$136,000 -$24,830,000 STDEV. $935,505 $4,311,543 Median $0 -$12,650 Mean $257,521 -$339,741 Table 5-45. Mann-Whitney Test Results sales growth, after graduation Mean Rank* Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 22.00 153.000 198.000 .798 Public 23.25

PAGE 192

192 Frequency Distribution of Sales Growth Rate, Before Graduation0 5 10 15 20 25 101234567891011121314151617181920 Private Public Figure 5-38. Frequency distribution, sales growth ra tes, before graduation Table 5-46. Sales growth rates, descript ive statistics, before graduation Sales Growth Rate Achieved During the Residency Incubator GP Incubator GG Count 9 36 MAX. 19.642 5.832 MIN. -0.200 -1.000 STDEV. 6.333 1.356 Median 1.476 0 Mean 3.441 0.491 Table 5-47. Mann-Whitney Test results, sa les growth rates, before graduation Mean Rank Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 29.00 108.000774.000 .123 Public 21.50

PAGE 193

193 Frequency Distribution of Sales Growth Rate, After Graduation0 5 10 15 20 25 30 10123456789101112131415161718 Private Public Figure 5-39. Frequency distributions, sales growth rates, after graduation Table 5-48. Sales growth rates, descript ive statistics, after graduation Sales Growth Rate Achieved During the Residency Incubator GP Incubator GG Count 9 36 MAX. 0.314 17.545 MIN. -0.123 -0.973 STDEV. 0.128 3.994 Median 0 -0.034 Mean 0.004 1.075 Table 5-49. Mann-Whitney Test results, sa les growth rates, after graduation Mean Rank Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 24.44 149.000 815.000 .712 Public 22.64

PAGE 194

194 Frequency Distribution of Annual Sales Growth Rate, Before Graduation0 5 10 15 20 25 0.500.50.511.522.533.5 Private Public Figure 5-40 Frequency distributions, annual sales growth rates, before graduation Table 5-50. Annual sales growth rates, descriptive statistics, befo re graduation Sales Growth Rate Achieved During the Residency Incubator GP Incubator GG Count 9 36 MAX. 2.455 1.944 MIN. -0.040 -0.333 STDEV. 0.853 0.467 Median 0.123 0 Mean 0.571 0.174 Table 5-51.Mann-Whitney Test results, annual sales growth rates, before graduation Mean Rank* Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 27.56 121.000 787.000 .242 Public 21.86

PAGE 195

195 Frequency Distribution of Annual Sales Growth Rate, After Graduation0 5 10 15 20 25 30 0. 2 5 0 0. 25 0.5 0.7 5 1 1.2 5 1.5 1.7 5 2 2.2 5 2.5 2.7 5 3 3.2 5 3.5 3. 75 4 4.2 5 Private Public Figure 5-41. Frequency distri butions, annual sales growth rates, after graduation Table 5-52. Annual sales growth rates, desc riptive statistics, af ter graduation Sales Growth Rate Achieved During the Residency Incubator GP Incubator GG Count 9 36 MAX. 0.045 4.125 MIN. -0.059 -0.243 STDEV. 0.032 0.902 Median 0 -0.013 Mean -0.004 0.235 Table 5-53. Mann-Whitney Test results, annual sales growth rates, after graduation Mean Rank* Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 24.00 153.000819.000 .798 Public 22.75

PAGE 196

196 Freqeuncy Distribution, Total Sales, Before Graduation0 5 10 15 20 25 300 5000000 10000 0 0 0 1 5000000 200 0 0 000 25000000 3 0000000 350 0 0 000 40000000 4 5000000 50000000 55000000 Private Public Figure 5-42. Frequency distributio ns, total sales, before graduation Table 5-54. Total sales, descriptive statistics, during the residency Total Sales Produced During the Residency Private Public Total $106,178,359 $237,721,364 MAX. $28,779,070 $51,685,000 MIN. $80,000 $0 STDEV. $8,638,161.14 $11,040,462.82 MEDIAN $1,458,900 $1,365,850 AVERAGE $5,715,693.13 $4,689,955.56 Table 5-55. Mann-Whitney Test results total sales, before graduation Mean Rank Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 27.73 244.000910.000 .591 Public 25.28

PAGE 197

197 Frequency Distribution of Total Sales, After Graduation0 5 10 15 20 250 5000000 1 0 0 0 0 0 0 0 15000000 20000000 25000000 30000000 35000000 400 0 000 0 45000000 50000000 5 5 0 0 0 0 0 0 60000000 65000000 7 0 0 0 0 0 0 0 75000000 Private Public Figure 5-43. Frequency distributio ns, total sales, after graduation Table 5-56.Total sales of the public and pr ivate incubator, after graduation, 1990-2007 Total Sales Produced After Graduation Private Public Total $106,178,359 $237,721,364 MAX. $73,169,707 $52,585,000 MIN. $0 $0 STDEV. $23,310,820.96 $11,045,933.91 MEDIAN $3,908,000 $2,603,007 AVERAGE $11,797,595.44 $6,603,371.22 Table 5-57. Mann-Whitney Test results total sales, after graduation Mean Rank Mann-Whitney U Wilcoxon W Sig.(2tailed) Private 25.28 141.500807.500 .561 Public 22.43

PAGE 198

198 Total Sales over Time, Tenants, 1990-2007$0 $10,000 $20,000 $30,000 $40,000 $50,000 $60,0001 9 9 0 1 9 9 2 1 9 9 4 1 9 96 1 9 98 2000 2002 2004 2006Thousands Private Public Figure 5-44. Total sales over time, tenants, 1990-2007 Changes of Total Sales of Tenants, 1990-2007-$40,000 -$30,000 -$20,000 -$10,000 $0 $10,000 $20,000 $30,000 $40,0001990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006Thousands Private Public Figure 5-45. Changes of tota l sales of tenants, 1990-2007

PAGE 199

199 Total Sales per Tenants, 1990-2007$0 $1,000 $2,000 $3,000 $4,000 $5,000 $6,000 $7,00019 9 0 19 9 2 1994 1996 1998 2 0 00 2 0 02 20 0 4 20 0 6Thousands Private Public Figure 5-46.Total sales per tenant, 1990-2007 Number of Tenants, 1990-2007 0 5 10 15 20 251990 19 91 1 992 1993 19 94 1 995 1996 1 997 1998 199 9 2 000 2001 2 002 2 003 2004 2 005 2006 200 7 Private Public Figure 5-47. Number of tenants, 1990-2007

PAGE 200

200 Total Sales over Time, Graduates, 1990-2007$0 $10,000 $20,000 $30,000 $40,000 $50,000 $60,0001 990 1 992 1994 1 996 1 998 2 0 00 2 0 02 2 004 2 006Thousands Private Public Figure 5-48. Total sales ove r time, graduates, 1990-2007 Changes of Total Sales of Graduates, 1990-2007-$10,000 -$5,000 $0 $5,000 $10,000 $15,000 $20,000 $25,000 $30,000 $35,000 $40,000 199019921994199619982000200220042006 Thousands Private Public Figure 5-49. Changes of total sales over time, graduates, 1990-2007

PAGE 201

201 Total Sales per Graduates, 1990-2007$0 $2,000 $4,000 $6,000 $8,000 $10,000 $12,0001 9 90 1 9 9 2 19 9 4 19 9 6 19 9 8 2 0 0 0 20 0 2 20 0 4 20 0 6Thousands Private Public Figure 5-50. Sales per graduates, 1990-2007 Number of Graduates, 1990-20070 5 10 15 20 25 301 9 90 19 91 1 9 92 1 9 93 1994 1 9 95 1996 1 9 97 1998 1999 20 0 0 2001 20 0 2 20 0 3 20 0 4 20 0 5 20 06 20 0 7 Private Public Figure 5-51. Number of graduates, 1990-2007

PAGE 202

202 CHAPTER 6 DISCUSSION AND POLICY IMPLICATION Discussion Due to serious m ethodological difficulties, mo st of the evaluation studies on LED have remained anecdotal or descriptive, rather than scientific. Evaluation studies are still important since the information produced by these studies is used as a source to influence the status of funding and to set up lists of the necessary amendments. This ev aluative study is an attempt to enrich our understanding of incubators that are operated un der different sponsorships. Using a comparative framework, this study provides scientif ic insight into the co mparisons of the input side aspect and the outcomes which are potentially affected by the different circumstances constrained by sponsorships. Ca se studies on the input side as pects focused on three groups of factors that make TBIs work. To investigate the differences between outcomes, this study conducted a comparison using business data from th e client groups of two incubators, in which the external correlations with geographic factors were inherently controlled. The major goal of this research is to investigate whether ther e is an impact of government intervention on the efficiency of TBI. This study utilized the case st udy methodology to study the differences between the input side aspects of the incuba tors sponsored by di fferent organizations; specifically, public non-prof it incubators were compared with private for-profit incubators. One private incubator and one public incubator were recruited and studied in order to examine the possible differences in the input side aspects. The first hypothesis was that differences may be observed in the cost-reduction strategies employed by the incubators. The service provision and managerial strategies of both incubators seemed to be largely predefin ed, along with the cap acities and conditions of the sponsorship. Different directions were chosen, in regards to cost-reduction strategies in accordance with the

PAGE 203

203 different s economies of scale constrained by sponso rship. The private incubator, working with a tighter budget and a smaller infrastructure, sugg ests providing clients with sources of funding, recruiting, and technology in order to either save costs or to shorten the lag time for independence. The public incubator, on the other hand, provided large-scale real estate property as a source of cost savings. The difference, wh ich lies in the scale of the capital given to the founders, leads to another difference in manageri al approach. The strategy of the private incubator to recoup the investment by taking an equity stake after the client is successful may actually reduce the initial costs for the client: they can save money on rent which they would otherwise have to pay. In addition, these saving s can be utilized as investments in sales and production. By requiring rents a nd fees, the old fashioned appro ach of the public incubator can reduce expenditures by clients who would othe rwise have to pay the market rate. The second hypothesis was whether the thoroug hness of the intervention process may be associated with sponsorship; which can be de termined by looking at the selection process, monitoring activities, an d graduation policies. As briefly mentioned above, for the private incubator, the selection proce ss implied the choosing of busine ss partners; while the process seems to be to choose program partners for the public incubator. C onsequently, the private incubator is strongly motivated to realize the pot ential of working together to achieve success. The public incubator is also concerned with th e potential of being successful. Thoroughness can be witnessed in the standardized selection criteria and process. When the private incubator intervenes in business activities with the aim of helping the client to gr ow and be successful, it employs informal daily monitoring activities; rather than the formal, periodical ones that can be observed in the public incubator. This differe nce seems also to be due to the different characteristics of the two sponsors. First of all, the private incuba tor does not have any

PAGE 204

204 motivation to conduct formal monitoring processe s; while the formal monitoring process is required for the public incubator to produce objective evidence to secure continuous support and funding from the public sector. Also, formal monitoring is used as a reference by the incubators to judge the progress of their c lient businesses. Second, the role of the private incubator as a business partner seems to be a factor that motiv ates the managers to wo rk closely with their clients. Since they work together as if they were in the same company, the periodical monitoring process may become redundant. On the other ha nd, the public incubator defines itself as an external assistant that provides corporate st yle services for its clients and takes on the responsibility of managing the facilities. Since the incubator managers are not involved in the daily operations of their client businesses, periodical monitoring is necessary for them to learn what is going on. Finally, the monitoring proces ses and graduation policie s of both incubators reflect the organizations goals. The private incubator places less emphasis on job creation, but stresses sales growth. The pub lic incubator, on the other han d, is very concerned with job creation; while it also agrees with the premise that increased employment will follow sales growth. The third focus of the case study was to examine the different ways that the incubators take advantage of surrounding resources; such as universities, local cap ital, and other institutions. Hypothetically, this study assumed that the privat e incubator might have difficulty with getting support from those resources; an assumption which turned out to be corre ct. Geographically, the private incubator complained abou t the lack of research univers ities in the ar ea. Although the incubator continuously reached out for help, it wa s not readily available. Consequently, lab facilities and student labor were also in high demand. Among these issues, a major difficulty for the private incubator seems to be working with a public sector that is less motivated about

PAGE 205

205 encouraging technology incubator businesses as a tool of local development. The government seems to be preoccupied with the development of phys ical infrastructure as they believe that this helps a larger number of people. The public incu bator obviously seems to have an advantage in this respect: it is itself a government agency, and therefore it is very easy to acquire the support of the public sector. Also, the lo cation of the public incubator in a research park allows greater access to abundant research facilities and human re sources. Within this setting, the park, the incubator, the innovative local community, local universities, a nd the local government are all encouraged to work together. Based on the assumption that the public sector is more concerned w ith job creation than the private sector, the fourth hypot hesis is that the public sector would generate more jobs than he private sector. None of the measurements of the impact on employment turned out to be statistically significant, however. The premis e that the public inc ubator produces more employment opportunities is not in question. Rather, it seems that the clients of private incubators tend to hire more employees, regardless of whether they show better employment growth or not. Also, the clients of the public incubator are more exte nsively engaged in the activities of hiring and fi ring than the public incubator itself. Another interesting finding is that the total number of empl oyees hired by the tenants of the private incubator tends to remain stable. On the other hand, the total number of client comp anies of the public incubator seemed to have fluctuated throughout their life time. This may mean that the clients of th e private incubator are more reluctant to hire or fire employees than the clients of the public incubator. Over the entire period, the number of jobs produced by the privat e incubator seemed to be larger, even though they might have been more cautious with expanding their employment opportunities.

PAGE 206

206 The fifth hypothesis was that th e private incubator may be be tter at wage creation. Both incubator companies showed positive increases in weekly wages. However, the private incubator tended to be better at increasing wages; while clients of the pub lic incubator tended to pay higher wages to only a few people, resulting in a moderate overall increase. A nother interesting finding is that the private incubator showed higher wage growth rate, without st atistical significance, when clients were in the incubator. However, af ter graduating from the incubator, clients of the public incubator tended to expand faster; a pa ttern which was accompanied by greater annual wage growth rate. Although higher wages overa ll should not necessarily be interpreted as positive growth, in that it is also sensitive to changes in the number of employees; it is a good indicator of the degree of contribut ion to the local economy. Also, in reality, it is hard to change program capacities given to sponsors, as in a hypo thetical situation. Firs t, higher wages imply a positive contribution to improving the level of in come. Second, higher wages also mean that companies have increased local tax revenues, in th at the wage eventually returns as income tax. Third, higher wages are likely to stimulate the circulation of wea lth within the local area. One possible application is that indus tries with higher wages may have a greater multiplier effect than industries with lower wages. This may lead to the formulation of local capital stocks that are another potential source of stimulation for further development. The last hypothesis is whether the incubator in the private s ector would produce more sales revenue. This was only moderately supported. The difference was explicit, especially while the clients were in the facility. It became less ex plicit after they graduated from the incubators. Greater sales growth during the residency was obser ved from clients of th e private incubator. Changes in the total amount of sales effectiv ely illustrate the quantif ied implication of the economic contribution. Better sales growth, in terms of the number of dollars, reappeared in the

PAGE 207

207 analysis of growth rate that accounted for the sales revenue of the initial year. During the residency, sales growth rates of the Incubator P on ce again turned out to be superior that of the Incubator G. In summary, clients of compan ies produce more sales than clients of public incubators; and they also grew faster than the c lients of the public incubator while they were in the incubator. However, this analysis does not account for the length of the residency. It may imply, nevertheless, that the private incubator is unlikely to release clie nts until their success has been fully confirmed. The growth pattern durin g the post-graduation peri od did not appear to significantly differ in any measure. Policy Implications Input Side Aspects Cost-reduction strategies In regards to the input side aspect, the private sector can complain about the lack of physical infrastructure, especially sp ace for offices and laboratories. The public sector, on the other hand, might develop a program that connect s this demand with lo cal supply, including surplus real estate and research facilities. The public sector should, however, be patient while waiting for the benefits to be realized; as the profits from technology businesses are often produced over the long term. The public sector might also improve access to venture capital. One way seems to be making public funding available for private venture capita lists, either with very low interest rates, or with a very long period of return, as a means of sharing the risk of investment of early stages. The last suggestion is to copy the business model observed in the private incubator: recouping benefits in the long term. Intuitively, matching skills and expertise is very important for the success of an incubator. The public sect or may consider refining the selection process to include this principle. Basicall y, client companies can receive the service in the early stages

PAGE 208

208 nearly free of charge. This strategy may improve the motivation of manage rs since it ties the destiny of the incubator to the su ccess of the clients. This may al so offset the chronic problem of the public sector being accused of being a less effici ent provider; while improving the autonomy of the incubator program by making profits from client businesses. Finally, governments becoming an equity holder would mitigate political resistance to the program, in that the program would generate profits for the investor, namely the government. If the public sector incubator could promote the intellectual pr operty developed by universities or other third parties, this could also be an effective method for the commerciali zation of technology. As a matter of fact, this strategy is better suited for the public incubator in th at it requires a great amount of money in reserve. If that money is available, and a public incubator is not present in the area, that money could help a private incubator to stabilize dur ing the early stages of development until the incubator makes a profit later on. Intervention mechanism One of the important findings is that the publ ic sector realizes that em ployment growth follows sales growth. Seemingly, the public sector does not push client co mpanies to hire more people or increase the number of jobs as evidence of progress or proof of gr aduation. It is likely that many public sector incubators have already moved on to become a hybrid type of incubator. These alternatives are employed to integrate ma nagerial and technologic al expertise from the public sector. However, a private agent may feel redundant while working with the public sector because of the limitations. During the pilot study, some interviewees complained about the difficulties of working with public managers that request formal reports in a bureaucratic manner. To a certain degree, political pressure that requi res proof of progress or investment is the reason that these processes are very fo rmal and strict. One suggestion is again that the public sector may have to stay back and be run autonomous ly. Again, for government, becoming an equity

PAGE 209

209 holder seems to be an answer here in that it binds the destinies of managers and clients together; while also partially freeing the managers from pub lic subsidies that are an essential condition for the managers to be more responsive. Surrounding resources In addition, the old premise that stresses the im portance of regional networks of innovation is reaffirmed. The subject private incubator seems to be isolated from this type of network, and therefore is having a hard time establishing itse lf, although it is a member of Florida Business Incubator Association. On the other hand, th e public incubator is surrounded by abundant resources. In the case of Maryland, the private incubator is seemingly in cluded as a member of the local community of i nnovators. In any case, the local or regional community should make an effort to be a source of knowle dge and institutional support wh ich all of the members could easily access. The public sector agent may find a leading role by making the provision of these networking activities one of their res ponsibilities. Efficiencies in Outcomes Another interesting observation of this study is that the initia l scale of businesses m ay have a strong correlation with the rate of growth. As widely accepted, bigger companies tend to be sluggish. Should private incubators then be more supportive of the smaller companies since they tend to grow faster? This may be a political de cision. Serving larger companies, including the NTBFs, would obviously generate a greater imp act in terms of quantit y, although with slower growth. If the public sector want s to increase the magnitude of the economic impact, larger firms may serve to achieve this goal. On the other hand, such a policy would result in reducing the chances for smaller companies to receive incubato r services, since private incubators do not have special motivation to serve small firms only for thei r fast marginal growth when the gross return is relatively small. If the public sector wanted to demonstrate fast gr owth (obviously political

PAGE 210

210 pressure is likely to force them to prove fast gr owth as evidence of support) so that they decided to offer more chances to smaller companies am ong the NTBF; the public sector incubator would have a smaller impact on the economy of the loca l area. One question is whether serving micro companies is in the public interest. If this is not the case, the public s ector should consider an expansion of the TBIs to enlarge the service cap acity. One attractive approach is increasing the number of branches which are networked togeth er in order to serve different technology segments, rather than building a larger site that can host many clients at one location. Other Implications Research Implications Most of the findings of this study reaffir m the findings of existing studies (e.g., this study observed more efficient sales increase from privat e agents). One challenging finding of this study, however, is the validity of the existing para digm for the research framework. The existing paradigm for the study framework is justified unde r an assumption that measuring efficiency is a valid approach for evaluative study; i.e., compar ative superiority of uni t productivity indicates betterment in performance. However, the assump tion of this approach is only valid when the two subject agents, for instance, private incubato rs and public incubators, have the equivalent chance to start a technology busines s incubator when none are curren tly available in a market. One question is whether researchers can validly assu me that private agents could indeed jump in the technology incubator business if public agents withdraw from the program. Realistically, this is very unlikely based on the large amount of capital that is required to establish and operate a business. For this reason, a private incubator may be more productive, but this finding may not necessarily justify a withdrawal by the public sector due to the relatively small likelihood for a private technology incubator to take over the business. In this sense, coun ting only on efficiency may dismiss the realistic implication that a private ag ent is less likely to engage in the business.

PAGE 211

211 One alternative perspective is to revisit the implication of the magnitude of impact rather than focusing only on efficiency. This research incorporated investigatio ns of the magnitude of impact by measuring growth of sales, wages, and employment. The magnitude of impact is measured in terms of quantity rather than un it change, which is usually employed to indicate efficiency. One justification for the use of this a pproach is that it considers the reality that the capital capacity of the public s ector cannot be easily obtained by private agents. Moreover, this study indicated some inconsistency between the measurement of magnitude and the measurement of efficiency, which may imply th at counting only on proportional growth may not fully deliver the actual meaning of a technology business incubato r to a region. In other words, when public agencies show stronger magnitude, even if they show relatively poor efficiency, the public sector may still have reason to continue th e technology business incubator program. If so, evaluative research also has reason to revis it the implication of the programs impact. Implications for Public Finance Another finding of this research, which rein f orces an existing perspective, is that technology incubators are not very good producers of jobs, but they are likely to increase local tax revenue. From the perspective of public fina nce, one concern derived from this finding is identifying who should then bear the cost to ope rate this costly program Obviously, local tax payers ideally should not burden the cost since th e benefit of the program to the region seems to be limited in terms of job generation. Poor job ge neration is also listed as a reason why the local government of one of the subject technology business incubators of this study was not very motivated. Also, the benefit retu rns in the long run further justif y shifting the cost to the upper tier government rather than the local government because upper tier governments are usually more efficient in terms of utilizing long term investments. However, this approach may be undesirable if a state or nati onal government takes initiative for the program. Thus, good

PAGE 212

212 coordination is required between the local and upper tier governments in terms of cost and planning. Role of Upper Tier Government Another reason to stress active roles for regional or state govern m ents is that the success of the TBI program hinges on the quality and quantity of surrounding resources, which may vary over regions. The local government may be motiv ated to host a technology business incubator when large amounts of financial support are pr oposed from a central government. In this situation, as competition among local governments b ecomes stronger, regional suitability may be disregarded. In addition to act ing in the role of financial s upporter, the central government should actively coordinate the program in orde r to deploy programs to regions that have abundant resources such that a TB I program is likely to prosper.

PAGE 213

213 CHAPTER 7 CONCLUSION This research was undertaken to respond to a policy question: whet her the public sector should withdraw financial support from the tech nology business incubators. Critics have argued that the TBIs in the public sector are not as efficient as the pr ivate agents. However, little is known about the structure and operation of these two different types, or the actual impact of any differences. This study has tried to fill the gap in knowledge. To answer the question, this research attempted to observe diffe rences in (1) the input side as pect of incubators by different sponsors, and (2) the outcomes of the two types of incubators. The differences in the input si de aspect were well observed, as hypothesized. In regards to the differences in the input side aspect, the differe nces in the status and capacity of the resources provided to each sponsor may drive the differences in services and strategies. Private incubators are likely to suffer from limited resources and cap acities, but are also likel y to respond to these challenges with a more flexible approach. Anothe r factor that creates these differences is the limited sources of capital which forces private incubators to narrow dow n the range of target businesses to only those matching th eir skills and expertise. Based on the findings about primary perfor mance outcomes observed from every possible angle, none of the type of the two sponsorships dominates in terms of efficiency, while several differences are observed. The only hypothesis reaffirmed by this research is that private incubators may be better at produ cing sales revenue. However, publ ic incubators seem to offset the productivity issue with la rger capacities. Neither does the evidence suppor t the simple argument that the public sector s hould withdraw from the TBIs, or regard public TBIs as being superior to the private ones. Rather, the findings imply that the old fashioned private-public framework might not be appropr iate for this subject.

PAGE 214

214 An alternative perspective based on the findings is that the roles of TBIs may need to be diversified. The public sector may find an active role in utilizing large-scale capita l. Obviously, a large amount of capital is useful in securing an incubators self-re liance. It also means that the public sector may step back from operating task s for the TBIs, as the private sector may respond to the market dynamics more effectively. The only shortcoming of this research is th at it shies away from investigating the correlation between inputs and outco mes, thus making for relatively poor predictability. This is largely due to methodological limita tions, and the fact that the nu mber of cases was very low, while the number of variab les was considerable.

PAGE 215

215 LIST OF REFERENCES Abetti, P. A. (2004) Governm ent-supported incubators in the helsinki region, finland: infrastructure, results, and best practices. Journal of Technology Transfer, 29, 19-40. Acs, Z. J., Audretsch, D. B., & Feldman, M. P. (1992) Real effects of academic research: comment. American Economic Review 81, 363-367. Acs, Z. J., Audretsch, D. B., & Feldman, M. P. (1994) R&D spillovers and innovative activity. Managerial and Decision Economics 15, 131-138. Allen, D. N., & McCluskey, R. (1990) Structure, policy, service and performance in the business incubator industry. Journal of Entrepreneurship Theory and Practice 14, 61 Allen, D. N., & McCluskey, R. (1990) Structur e, policy, services and performance in the business incubator industry. Entrepre neurship Theory and Practice, 15, 2, 61-77. Allen, D. N., & Rahman, S. (1985) Small busin ess incubators: a posit ive environment for entrepreneurship. Journal of Small Business Management, 23, 3, 12-22. Appold, S. J. (1991) The location process of industrial rese arch laboratories. The Annals of Regional Science, 25, 131-144. Atkinson, S. E., & Halvorsen, R. (1986) The relativ e efficiency of public and private firms in a regulated environment: the case of US. electric utilities. Journal of Public Economics 29, 281-294. Audretsch, D. B., & Feldman, M. P. (1996) R& D spillovers and the geography of innovation and production. The American Economic Review 86 3, 630. Aydalot, P. (1986). Milieux Innovateurs en Europe Paris: GREMI. Barkham, R., Gudgin, G., Ha rt, M., & Hanvey, E. (1996) Determinants of Small Firm Growth: an Inter-Regional Study in th e United Kingdom 1986-90 (Vol. 12). Gateshead, Tyne and Wear, UK: Athenaeum Press. Bartik, T. J. (1990) The market failure appr oach to regional economic development policy Economic Development Quarterly 4 4, 361-370. Bllingtoft, A., & Ulhi, J. P. (2005) Th e networked business incubatorleveraging entrepreneurial agency? Journal of Business Venturing 20, 265-290. Borger, B. D., Kerstens, K., W.Moesen, & Vann este, J. (1994) Explai ning differences in productive efficiency: an applicat ion to Belgian municipalities. Public Choice 80, 339-358. Campbell, C., & Allen, D. N. (1987) The sma ll business incubator industry: micro-level economic development. Economic Development Quarterly, 1 2,. 178-191

PAGE 216

216 Carlton, D. W. (1983) The location and employ ment choices of new firms: an econometric model with discrete and c ontinuous endogenous variables. The Review of Economics and Statistics 65 3, 440-449 Carroll, G. R., & Delacroix, J. (1982) Organizational mortality in the newspaper industries of argentina and ireland: an ecological approach. Administrative Science Quarterly 27, 2, 169-198 Christoffersen, H., Paldam, M., & Urtz, A. H. W. (2007) Public versus private production and economies of scale. Public Choice 130, 311-328. Collinson, S., & Gregson, G. (2003) Knowledge networks for new technology-based firms: an international comparison of local entrepreneurship promotion. R&D Management 33, 2. 189-208 Colombo, M. G., & Delmastro, M. (2002) How e ffective are technology incubators? Evidence from Italy, Research Policy 31, 1103-1122. Davidsson, P., Leona, A., & Naldi, L. (2005) Re search on small firm growth: a review. Paper presented at the Proceedings Europe an Institute of Small Business. Eisinger, P. K. (1988) The Rise of the Entrepreneuria l State: State and Local Economic Development Policy in the United States Madison, Wisconsin: The University of Wisconsin Press. Feldman, M. P., & Audretsch, D. B. (1999) Innovation in cities: sc ience-based diversity, specialization and lo calized competition. European Economic Review 43, 409. Ferguson, R., & Olofsson, C. (2004) Science park s and the development of NTBFs location, survival and growth, The Journal of Technology Transfer 21 1, 5-17. Flamholtz, E. G. (1986) Managing the Transition from an Entrep reneurship to a Professionally Managed Firm San Francisco, CA: Jossey-Bass. Florida, R. (2003). The flight of the creative class: the new global competition for talent, New York, NY : Harper Business Flynn, D. M. (1993) A critical e xploration of sponsorship, infrastructure, and new organizations. Small Business Economics 5, 129 -156. Georghiou, L. (2002) Impact and additionality of innovation policy, Paper presented at the Contributions to a Six Countries Programme Conference, Brussels. Gruber, J. (2005) Public finance and public policy New York, NY: Worth Publishers. Hackett, S. M., & Dilts, D. M. (2004a) A real options-driven theory of business incubation. Journal of Technology Transfer 29, 1, 41-54.

PAGE 217

217 Hackett, S. M., & Dilts, D. M. (2004b) A syst ematic review of busin ess incubation research. Journal of Technology Transfer 29, 1, 55-82. Hamdani, D. (2006) Conceptualizing and Measuring Business Incubation Ottawa, Canada: Minister of Industry. Heijs, J. (2003) Freerider behavi our and the public finance of R&D activities in enterprises: the case of the Spanish low interest credits for R&D. Research Policy 32, 3, 445-461. Henderson, V., Kuncoro, A., & Turner, M. ( 1995) Industrial devel opment in cities. The Journal of Political Economy 103, 5, 1067-1090 Hughes, J. T. (1991) Evaluation of local economic development :a challenge for policy research. Urban Studies, 28, 6, 909-918. Jackson, P., & Palmer, B. (1989) First Steps in Measuring Perform ance in the Public Sector.A Management Guide. London: Public Finance Foundation. Jenssen, J. I., & Havnes, P. A. (2002) Public intervention in the entrepre neurial process: a study based on three norwegian cases. International Journal of Entrepreneurial Behavior & Research 8, 3, 173-187. Kaiser, U. (2002) Measuring knowledge spillovers in manufacturing and services: an empirical assessment of alternative approaches. Research Policy 31, 1, 125-144. Knopp, L. (2007) 2006 State of the Business Incubation Industry Athens, Ohio. Kuratko, D. F., & LaFollette, W. R. (1987) Small business incubators for local economic development. Economic Development Review, 5, 1, 2, Lalkaka, R. (2002) Technology business incubators to help build an innovation-based economy, Journal of Change Management 3, 2, 167-176. Lalkaka, R. (2003). Role of technology busine ss incubators in SADC countries. Unpublished Presentation slides. Levie, J. (1993) Can government nurture young gr owing firms? Qualitative evidence from a three nation study. Frontiers of Entrepreneurship Research 198-211. Link, A. N., & Scott, J. T. (2006) U.S. university research parks. Journal of Productivity Analysis 25 1-2, 43-55. Lofsten, H., & Lindelof, P. (2005) R&D netw orks and product innovation patternsacademic and non-academic new technology-based firms on Science Parks, Technovation 25, 9, 1025-1037. Lfsten, H., & Lindelf, P. (2001) Science parks in Sweden industrial renewal and development? R and D Management 31, 3, 309-322.

PAGE 218

218 Lfsten, H., & Lindelf, P. (2002) Science pa rks and the growth of new technology-based firmsacademic-industry links, innovation and markets, Research Policy 31, 6, 859-876. Lfsten, H., & Lindelf, P. (2003a) Determinants for an entrepreneurial milieu: Science Parks and business policy in growing firms. Technovation, 23, 1, 51-64. Lfsten, H., & Lindelf, P. (2003b) Science pa rk location and new technology-based firms in Sweden implications for strategy and performance, Small Business Economics 20(3), 245-258. Lfsten, H., & Lindelf, P. (2005) R&D netw orks and product innovati on patternsacademic and non-academic new technology-based firms on Science Parks Technovation, 25, 9, 1025-1037. Lsch, A. (1954) The Economics of Location New Haven. Connecticut: Yale University Press. Macdonald, S. (1987) British science parks: re flections on the politics of high-technology. R&D Management, 17, 1, 25-38. Malecki, E. J. (1991) Technology and Economic Development: the Dynamics of Local, Regional and National Change London: Longman Scientific and Technical. Malecki, E. J. (1997) Technology and Economic Development: the Dynamics of Local, Regional and National Change (2nd ed.). New York, NY: Longm an Scientific & Technical. Markusen, A. R. (1986) High tech America: the What, How, Where, and Why of the Sunrise Industries. Boston, MA: Allen & Unwin. Maryland De partment of Business and Economic Development. Massey, D., Quintas, P., & Wield, D. (1992) High Tech Fantasies: Science Parks in Society London.: Routledge. Mian, S. A. (1996) Assessing value-added contributions of univers ity technology business incubators to tenant firms. Research Policy 25, 325-335. Mian, S. A. (1997) Assessing and managing the university technology business incubator: an integrative framework. Journal of Business Venturing 12, 251-285. Musgrave, R. A., & Muscrave, P. B. (1973) Public Finance in Theory and Practice New York: McGraw-Hill, Inc. NBIA. (2003) A national Benchmarking Analysis of Technology Business Incubator Performance and Practices Report to the Technology Administration U.S. Department of Commerce Nelson, R. R. (2006) Economic Development from the Pers pective of Evolutionary Economic Theory. Norway: The Other Canon Foundation.

PAGE 219

219 OMalley, M., Brown, A. G., & Edgerley, D. W. (2007) Economic Pulse: Maryland Department of Business Economic Development. Oakey, R. P., & Cooper, S. Y. (1989) High tech nology industry, agglomera tion and the potential for peripherally sited small firms. Regional Studies, 23, 4, 347-360. Phelps, P. B., & Brockman, P. R. (1992) Science and Technology Progr ams in the States: 1992. Alexandria, VA: Advanced Development Distribution Phillips, R. (2003) Evaluating Technology-Based Economic Development: Gauging the Impact of Publicly-Funded Programs and Policies. Lewiston, NY: The Edwin Mellen Press, Ltd. Phillips, R. G. (2002) Technology business incu bators: how effective as technology transfer mechanisms? Technology in Society 24, 299-316. Premus, R. (1982) Location of High Technology Firms and Regional Economic Development Washington, D.C: Joint Economic Committee of Congress, Government Printing Office. Rice, M. P. (1992) Intervention Mechanisms Used to Influence the Critical Success Factors of New Ventures: an Exploratory Study Unpublished dissertation, Rensselaer Polytechnic Institute, Troy, NY. Rothwell, R., & Zegveld, W. (1981) Industrial Innovation and Public Policy: Preparing for the 1980s and the 1990s. Westport: Greenwood Press. Saxenian, A. (1985) Silicon Valley and Route 128: re gional prototypes or hi storic exceptions? In M. Castells (Ed.), High Technology, Space, and Society Beverly Hills and London.: Sage Publications. Schmenner, R. W., Huber, J. C., & Cook, R. L. (1987) Geographic differences and the location of new manufacturing facilities. Journal of Urban Economics 21, 1, 83-104 Schumpeter, J. A. (1934) The Theory of Economic Development. New York, NY: Oxford University Press. Schumpeter, J. A. (1939) Business Cycles New York, NY: McGraw-Hill. Schumpeter, J. A. (1942) Capitalism, Socialism, and Democracy. New York, NY: Harper & Row. Shane, S., & Venkataraman, S. (2000) The promise of entrepreneurship as a field of research. Academy of Management Review, Vol. 25, 1, 217-226. Sherman, H. D. (1999) Assessing the intervention effectiveness of business incubation programs on new business start-ups. Journal of Developmental Entrepreneurship Fall.

PAGE 220

220 Siegel, D. S., Westhead, P., & Wright, M. ( 2003) Assessing the impact of university science parks on research productivity: exploratory firm-level evidence from the United Kingdom. International Journal of Industrial Organization 21, 9, 1357-1369. Smilor, R. (1987). Managing the incubator system : critical success factors to accelerate new company development. IEEE Transactions on Engineering Management EM 34, 4, 146156. Smilor, R., & Gill, M. (1986) The New Business Incubator: Li nking Talent, Technology and Know-How. Lexington, MA: Lexington Books. Smilor, R. W., & Feeser, H. R. (1991) Chaos and the entrepreneurial process: patterns and policy implications in technology research. Journal of Business Venturing 6, 3, 165-172. Smith, A. (1776) The Wealth of Nations. New York, NY:P.F. Collier & son, Soetanto, D. P., & Geenhuizen, M. v. (2005) Technology incubators as nodes in knowledge networks. Paper presented at the 45th C ongress of the European Regional Science Association. Squicciarini, M. (2008) Science Parks' tenants versus out-of-Park firm s: who innovates more? A duration model. Journal of Technology Transfer 33, 1, 45-71. Stam, E., Suddle, K., Hessels, S. J. A., & Stel, A. J. V. (2007) High growth entrepreneurs, public policies and economic growth. Jena Economic Research Papers NO. 2007-019. Sternberg, R. (1990) The impact of innovation centers on small technology-based firms: the example of the federal republic of Germany. Small Business Economics 2, 2, 105-118 Stinchcombe, A. L. (1965) Social Structure and Organizations. In J. G. March (Ed.), Handbook of Organizations Chicago, IL: Rand McNally. Tamsy, C. (2007) Rethinking technology-orient ed business incubators: developing a robust policy instrument for entrepreeurship, innovation, and regional development? Growth and Change 38, 3, 460-473. Temali, M., & Campbell, C. (1984) Business Incubator Profiles: a National Survey. Minneapolis: University of Minnesota, Hubert H. Humphrey Institute of Public Affairs. The National Business Incubation Association. (2003) A National Benchmarking Analysis of Technology Business Incubator Performance And Practices Report To The Technology Administration : U.S. Department of Commerce Walls & Associates. National Establishment Time-Series (NETS) Database: Database Description. Unpub lished manuscript. Weber, A. (1929) Theory of The Location of Industries. Chicago, IL: University of Chicago Press.

PAGE 221

221 Westhead, P. (1997) R&D 'inputs' and 'outputs' of technology-based firms located on and off Science Parks. R&D Management 27, 1, 45-62. Westhead, P., & Batstone, S. ( 1998) Independent technology-based firms: the perceived benefits of a science park Location. Urban Studies 35. Westhead, P., & Storey, D. J. (1994) An Assessment of Firms Locate d in And off Science Parks in UK. London.: HMSO. Yin, R. K. (1994) Case Study Research: Design and Methods. Thousand Oaks, CA: SAGE.

PAGE 222

222 BIOGRAPHICAL SKETCH Yongseok Jang started h is academic career with earning Bachelor of Arts in. political science at Inha University in Korea. During his undergra duate work, he was honored two consecutive times. After he finished his milita ry service in Korean Army, he resumed his academic career in politi cal science at University of Flor ida since 2001. Starting doctoral work in urban and regional planning at year 2004, his new life has continued his dedication to social issues, by specializing economic development (20042009). He has been married for 8 years and he is a father of one daughter; Grace age 6 and one son; Timothy, age 3