The Right Stuff

  • 720 views
Uploaded on

 

More in: Business , Technology
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Be the first to comment
    Be the first to like this
No Downloads

Views

Total Views
720
On Slideshare
0
From Embeds
0
Number of Embeds
1

Actions

Shares
Downloads
13
Comments
0
Likes
0

Embeds 0

No embeds

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
    No notes for slide
  • Admission highly competitive – Hu Jintao, the current Chinese president is a Tsinghua alum
  • First let me talk about where I place this in my line research and where I see my research trajectory going in the near and mid-term. I might be a bit pre-mature in trying to put this together but I think it’s good to start iterating on it and thinking about it. But I don’t want to dwell on this slide for too long.I see myself as falling broadly under the rubric of an empirically focused strategy / economics of innovation guy. I basically have three strands to my research. The first is work on health policy which I started at Duke working with Frank Sloan and this line has been on pause for a while. The middle and main line of research is really focused on technology-based entrepreneurship and examining labor issues and entrepreneurial career concerns. There is an initial paper co-authored with David Hsu published in RP examining changes over time in MIT alumni entrepreneurs. Two other papers in process examine serial entrepreneurs and I’m collaborating with Josh Lerner and Antoinette Schoar on a project VC and entrepreneurial finance. The paper that I’m going to talk about today falls in this line and examines tech entrep. in a setting overseas as the title alluded to.Finally, a third line of my research which I would like to eventually tie back into this second line, is on environmental strategy and firm self-regulation. I started this work while at Duke collaborating with Mike Lenox and see great potential in exploring how to tie work on policies for sustainability and energy entrepreneurship back into my more general work on entrepreneurs. In general I am very interested in the constraints and drivers of commercialization of technology at the individual level in terms of work experience and training and at the broader policy and regulatory level. In future work would like to focus more on the energy and biomedical industries since they are tech-based and highly influenced by policy and regulation. As you’ll see my current work spans any single industry.

Transcript

  • 1. Who has ‘The Right Stuff’? Human Capital, Entrepreneurship and Institutional Change in China Charles E. Eesley cee@stanford.edu Stanford University Department of Management Science & Engineering Washington University Oct. 7th, 2009 (with support of a Kauffman Foundation Dissertation Fellowship, the Tsinghua Univ. Alumni Association, and the MIT Entrepreneurship Center)
  • 2. Strategy / Economics of Innovation and Technological Change High Tech Entrepreneurship (selection and performance) Current Research Labor Economics Institutional (individual/micro- Variation level) (macro-level) Charles Eesley
  • 3. Motivation 19892004 China 29% growth vs. US 1% (State Statistics Bureau) ―High-impact or high- growth entrepreneurs … in particular, help drive growth in productivity and living standards and, thus, are of special interest and importance.‖ – Robert Litan, Vice President, Research and Policy Ewing Marion Kauffman Foundation (Chinese State Statistics Bureau) • Critical for their role in creating new markets, technologies and products, … process of ―creative destruction‖ (King and Levine, 1993a; 1993b; Djankov et al. 2002; Klapper et al., 2007) • Acs, Z. J., and D. B. Audretsch, 1988; Christensen; Henderson, Utterback, etc. • How might we encourage highly talent individuals to found firms? Charles Eesley - The Right Stuff
  • 4. Results Proportion Becoming Entrepreneurs by Year (Unobserved Human Capital) 0.10 0.09 0.08 0.07 0.06 Proportion 0.05 Low human capital 0.04 High human capital 0.03 0.02 0.01 0.00 Founding Year Charles Eesley - The Right Stuff
  • 5. Overview 1. Theory 2. Hypotheses 3. Two reforms 4. Empirical context – novel survey data 5. Results 6. Conclusion and Implications 7. Robustness checks (US data)
  • 6. Institutional Level Theories • Linked with long-run economic growth (Acemoglu, 2002; 2005; Johnson, McMillan & Woodruff; Porta, 1998) • Transition,idea common to all usages property „institution‟ is “The only institutional reform, of the term rights, financial liberalization, liquidity constraints that of some sort of establishment of relative permanence of a (Walder, 2003; Gans, Hsu &social sort.” (Hughes, 1936:Woodruff; Baumol, 1990; distinctly Stern 2002; Johnson, McMillan & 180) Thornton, 1999; Nee, 1996; Katila& Chen, 2009; Katila& Shane, 2005) • Neo-institutionalism (Meyer & Rowan, 1977; DiMaggio & Powell, 1983) Individual Level Theories • Training and prior experience (Hsu, 2007; Sorensen, 2007; Baumol, 2004; Groysberg, et al. 2007; Burton et al. 2002; Phillips, 2002; Dobrev& Barnett, 2005, Gompers et al, 2005;Simons & Roberts, 2007; Baron et al. 1996; Klepper& Sleeper, 2002; Eisenhardt&Schoonhoven, 1990) • Networks(Stuart and Ding, 2006; Katila et al., 2008) • Human capital, preferences (Zucker, Darby & Brewer, 1998; Lazear 2004; Irigoyen 2002, Nanda 2008; Amit et al. 1995; Amit et al. 1990; Knight 1921; Lucas 1978; Jovanovic 1982; Holmes and Schmitz 1990; Jovanovic and Nyarko 1996; Sorensen 2004) Charles Eesley - The Right Stuff
  • 7. Institutional Level Very little on ability/human capital (Banerjee and Munshi, 2004) Individual Level Charles Eesley - The Right Stuff
  • 8. Focus on Human Capital as a Driver of Selection •Talent and labor skills- Roy Model (1951) Selection on Potential Earnings •Identification requires exclusion restrictions, exogenous shocks, or instruments affecting returns or skill in one sector only •Jack-of-all trades •Liquidity Constraints – Evans and Jovanovic (1989) •Assets are endogenous, static models, typically poorly measured •Buera (2008) Liquidity Constraints •Psychological Traits – Kihlstrom and Laffont (1979), Dunn and Holtz-Eakin (1996), Stajkovic et al (2000) •Evidence on risk-aversion or tastes for independence is limited Talent Psychological Traits and Labor Demographic Factors •Network factors (Stuart and Ding, 2006, Skills Ruef, Aldrich, & Carter, 2003, Nicolaouand Birley, 2003) •Demographic factors (McClelland, 1961, Blau& Duncan, 1967 Dunn & Holtz-Eakin, 2000; Roberts, 1991) Network Factors Charles Eesley - The Right Stuff
  • 9. Model Roy Model as extended by Borjas (1987) Labor markets 0 and 1, respectively. Log earnings in the wage sector: w0 = μ0 + ε0 ε0 ~ N (0, ς02) The wage sector earnings would be the following: w1 = μ1 + ε1 ε1 ~ N (0, ς12) Assume that the cost of becoming an entrepreneur is C π = C/w0 The correlation between entrepreneur and wage worker earnings is ς01 = cov(ς0, ς1). A worker will choose entrepreneurship if Charles Eesley - The Right Stuff
  • 10. Model Define the indicator variable I, equal to 1 if this selection condition is satisfied, 0 otherwise. Define ν = ε1 − ε0. The probability that a randomly chosen worker from the wage sector chooses to entrepreneurship is equal to: Φ (·) is the CDF of the standard normal and z = (μ0 − μ1 + π) /ςν Calculate expectation of earnings in wage and entrep. sectors for those who chose the other sector. Selection conditions: THREE CASES: Define Q0 = E (ε0|I = 1), Q1 = E (ε1|I = 1) Charles Eesley - The Right Stuff
  • 11. Model Case 1: Positive hierarchical sorting: • Entrepreneurs are positively selected from the wage sector distribution and are also above the mean of the entrepreneurship distribution: Q0> 0, Q1> 0. This will be true iff • First, ς1/ ς0> 1 implies that entrepreneurship has a higher ‘return to skill’ than the wage sector. Second, ρ > ς0 /ς1, implies that the correlation between the skills valued in the wage sector and in entrepreneurship is sufficiently high Case 2: Negative hierarchical sorting • In this case, entrepreneurs are negatively selected from the wage sector distribution and are also below the average of the entrepreneur distribution: Q0< 0, Q1< 0. This will be true iff Charles Eesley - The Right Stuff
  • 12. Model Case 3: ‘Reverse’ sorting Where Q0 < 0, Q1> 0, that is, entrepreneurs are selected from the lower tail of the wage sector distribution but arrive in the upper tail of the entrepreneurship distribution. This can only occur if A fourth case? Note that there is not a fourth case where Q0> 0, Q1< 0. This would only happen if an individual from the top of the wage sector distribution joined the bottom tail of the entrepreneurship distribution. Charles Eesley - The Right Stuff
  • 13. Conventional View Cost to Start a Business Net Returns to Entrepreneurship Initial institutional environment Institutional environment #2 Higher Human Capital Increase is among those of relatively lower ability (Nanda, 2008)
  • 14. Much Less Focus on Returns Entrepreneurial Returns (net of entry $) Wage empl. income Institutional environment #2 Initial institutional environment Higher Human Capital
  • 15. Selection on Talent and Expected Earnings Payoff in high Payoff in low variance sector variance sector Individuals form expectations on talent / earnings in each sector (Roy, 1951) Charles Eesley - The Right Stuff
  • 16. 2. Hypotheses H1: An institutional change reducing barriers to growth will increase entrepreneurship among individuals located relatively higher in the talent distribution. H2: Individuals who show evidence of higher talent in their wage employment careers will experience higher returns to talent (in entrepreneurship) after an institutional change lowering barriers to growth. H3: Individuals who show evidence of higher talent in their wage employment careers will start firms that have higher performance. Charles Eesley - The Right Stuff
  • 17. 3. Ideal Experiment Exogenous increase (decrease) in σ1 Returns to talent in entrepreneurship relative to wage Ideal data is on the t employment relative variance of wage and entrepreneurial income Roy Model (Appendix A) 1988 – barriers to 1999 – barriers to entry growth t Charles Eesley - The Right Stuff
  • 18. Reform Easing Constraints on Growth March 1999 An amendment of Article 11 of the Constitution - Officially ending discriminatory practices against private firms …places private businesses on an equal footing with the public sector by changing the original clause "the private economy is a supplement to public ownership" to "the non-public sector, including individual and private businesses, as an important component of the socialist market economy" (China Daily, March 16, 1999). Kellee S. Tsai, ―Capitalists without a Class: Political Diversity Among Private Entrepreneurs in China, Comparative Political Studies, 38:9, 2005. - Privateproperty - R&D policy (expansion of tax incentives, incubators, science parks) 18
  • 19. Reform Yingyi Qian, current SEM Dean at Tsinghua University, former economics faculty at Stanford and UC Berkeley • …places private businesses on an equal footing with the public sector • The Jiangsu Provincial Government adopted a new policy to give private enterprises equal treatment as state-owned and collective enterprises… (People's Daily, April 9, 1999). Qian, Yingyi. "The Process of China's Market Transition (1978-1998): The Evolutionary, Historical, and Comparative Perspectives." Journal of Institutional and Theoretical Economics, March 2000, 156(1), pp. 151-171. Charles Eesley - The Right Stuff
  • 20. Reform Yingqiu Liu, a Senior Research Fellow and Professor of Economics at the Chinese Academy of Social Sciences (CASS) Since 1999: • Zeng Peiyan, minister at the State Development Planning Commission: ―[We will] eliminate all restrictive and discriminatory regulations that are not friendly towards private investment and private economic development in taxes, land use, business start-ups, and import and export. In the area of stock listings, private enterprise should enjoy equal opportunity which was enjoyed by the state-owned enterprises.‖ • A large number of provincial governments that have issued documents that support and promote private enterprises in order to help them develop and grow rapidly. Liu, Y. Development of Private Entrepreneurship in China: Process, Problems and Countermeasures. Working paper.
  • 21. Entrepreneur‘s Perspective • One pair of founders had very high human capital with one being a lawyer and an MBA and the other having a Ph.D. Founded in 2003 and said that: I then spent 20 years in the Bay area in life sciences companies. In the mid- 1990’s I came back to China to survey biotech companies in China and found that the environment was not ready yet. …spent an entire year just looking for the right office space…each product must be registered and approved by the government. It’s an expensive and time consuming procedure. She eventually found space for the company’s first store in a children’s museum which was perfect since they were selling toy bears aimed at children. This also allowed them to “hide” from government inspectors. Charles Eesley - The Right Stuff
  • 22. 4. Context Alumni survey Tsinghua University • 30,000 mailed • 3,000 surveys • 10% r.r. • Growing tradition: Stanford GSB, Chicago, HBS • Disadvantages: Biased towards tech., other response bias? • Advantages: Defined‗at risk‘ set, first abroad, detailed work historydevelopfounding ―Additionally, there is a strong need to and data sets to study data, less biased political factors affect entrepreneurship.‖ - how economic and by govt. concerns (Klapper, Amit, Guillén, & Quesada, 2008) Hsu, D.H., Roberts, E.B., Eesley, Charles. 2007. Entrepreneurs from Technology-Based Universities: 22 Evidence from MIT. Research Policy, 36: 768–788.
  • 23. Survey Questions 8. How many companies have you founded (not including State-Owned Enterprises (SOEs))? ____ 9. How many companies have you privatized or bought? ____ Please answer the remaining following questions about your first company (Company A) 2. Will spending money on research and development be a major priority for this new business? Yes No 3. Were the products and services to be provided by your new business available on the market 3 years ago? No Yes, but only a few Yes, but not too many Yes, almost everywhere Charles Eesley - The Right Stuff
  • 24. Coefficient on Year Fixed Effect 3 2.5 2 1.5 1 Coefficient 0.5 Dep. Var. = income from start- up 0 Dep. Var.= employees -0.5 -1 -1.5 -2 Charles Eesley - The Right Stuff
  • 25. Reform Indep. Vars. log(revenue) log(employees) POST-1999 1.262** (0.744) 0.510** (0.309) Master’s degree 0.370 (0.348) 0.083 (0.164) Doctorate degree -0.342 (0.631) 0.200 (0.305) Privatized 1.407** (0.690) 1.409*** (0.315) Bought -1.212 (0.751) 0.087 (0.349) Firm Age 0.460*** (0.085) 0.278*** (0.034) Communist Party 0.128 (0.346) 0.039 (0.163) Overseas 0.710 (0.449) 0.328* (0.198) FamilyEconomic Status -0.078 (0.179) -0.018 (0.083) Constant 2.544** (1.208) 1.420** (0.518) Obs. 195 267 ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. Charles Eesley - The Right Stuff
  • 26. Charles Eesley - The Right Stuff
  • 27. Distribution of Observable Human Capital Highest Degree Earned 60 50 40 Percent 30 20 10 0 Bachelor's Master's Doctorate Histogram of GPA Rank 35 30 25 20 15 10 5 0 First (Highest) Second Quartile Third Quartile Fourth (Lowest) Quartile Quartile
  • 28. 5. Methods Differences-in-differences estimation Cox Hazard Rate Model (robust to logit) Prob (founding a firm = 1) = F(α + 1’θi + 2’ θi*POST + 3’POSTi + ’Xi + τt+ ηj + φa+εi) • Dependent variable: First start-up founded • Subjects start being “at risk” of founding a firm at the time of their graduation • Θi = human capital measures • POST = 1 if individual was at risk for founding between 2000 and 2007. • Xi = Set of controls academic dept., region, education, work history, job type, Communist party, overseas educ. or work, family economic status. • Include (τ + η + φ) grad. year, region and Bachelor’s academic dept. fixed effects • (Acemoglu& Finkelstein, JPE 2008) 28 Proportional Hazards Test
  • 29. Coefficients on Year Fixed Effects Hazard Rate Coefficients for Year-by-Year (controls for region, education, department) Interactions 1.7 1 0.9 1.5 0.8 1.3 Unobs. 0.7 Talent 0.6 1.1 (Income 0.5 Year fixed Residual) 0.4 effects 0.9 Years 0.3 0.7 Education 0.2 0.1 0.5 0 2000 1996 1997 1998 1999 2002 2003 2004 Coefficients on Educ. Interaction with Year Fixed Effects 0.2 0.25 0.2 0.1 0.1 0.15 0.05 -0.05 -0.15 Master's Degree -0.25 ** ** -0.35 -0.45 ** Charles Eesley - The Right Stuff
  • 30. Reduced Barriers to Growth: Observable Measures Dependent Variable = Start-up founded (subjects start being at risk upon graduation) Note: reported coefficients are hazard ratios Master’s degree 0.444*** (0.121) Master’s x POST 1.771* (0.581) PhD degree 1.131 (0.630) PhD x POST 0.889 (0.549) Highest salary 0.771** Student leader 0.718** (pre-founding) (0.079) (0.097) Salary x POST 1.225* Leader x POST 1.336* (0.147) (0.209) Log(yrs. work exp) 0.832*** Parents’ educ. 0.240*** (0.022) (above median) (0.073) Parent edu. x Log(yr. work exp.) 0.999 POST 4.575*** x POST (0.027) (1.588) Promoted 0.216*** High GPA 0.350*** (0.112) (0.111) Promoted x POST 3.361** GPA x POST 1.811* (1.953) (0.651) Years 2000-07 0.000*** 0.033*** 0.060*** 0.012*** Years 2000-07 0.007*** 0.027*** 0.020*** (POST) (0.000) (0.001) (0.019) (0.007) (POST) (0.004) (0.007) (0.006) N=1,821; 308 foundings; 44,248 total years at risk; ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.
  • 31. Drawbacks to Observable Measures • Could be correlated with family wealth • Consistent with an opportunity costs story • Subject to shifts in who gets a graduate degree, gets promoted, etc. • Difficult to test changes in the shape of the distribution Ideally want some continuous underlying measure of talent Charles Eesley - The Right Stuff
  • 32. Unobserved Talent: Income Regression Specification for the ordered logit regression is as follows: Yi = Φ(α + γXi +τt+ ηj + φa + εi) Individual i ; 2:1 match on grad. yr. and salary year Xit = controls including job type (academia, business, government), tenure, Communist party, overseas, family wealth, and education Dependent variable: salary (6 categories) τt+ ηj + φa = year + region + Bachelor’s department Use εi in the entrepreneurship regressions Andersson, F., M. Freedman, J.C. Haltiwanger, J. Lane, K.L. Shaw. 2006. Reaching for the stars: Who pays for talent in innovative industries? NBER Working Paper No. 12435 Gibbons, R. et al., 2005 JLE Charles Eesley - The Right Stuff
  • 33. 5. Quantile Regression Model • Many of the attractive properties of OLS or mean regression • Advantage of allowing changes in the shape of the entire conditional distribution to be examined (Koenker& Basset, 1982; Koenker, Hallock 2001) • Dependent variable: income regression residual • Bootstrap method (with 100 repetitions) is used to generate standard errors (Horowitz 2001, Rogers 1992) 33
  • 34. Quantile Regression Panel A Dependent variable = income residuals (N=595) Percentiles 10 25 50 75 80 Founded in 1978-89 1.558*** 0.776*** 0.125** -0.494*** -0.680*** (0.506) (0.271) (0.065) (0.187) (0.246) Founded in 1990-99 0.769*** 0.496*** 0.009 -0.210 -0.233 (0.196) (0.161) (0.072) (0.259) (0.381) Founded in 2000-07 0.501*** 0.519*** 0.382*** 0.722*** 0.688*** (0.215) (0.174) (0.136) (0.194) (0.151) Panel B Entrepreneurs only (N=132) Percentiles 10 25 50 75 90 Ln(profit) 0.215 0.093 0.195 0.242** 0.246** (0.295) (0.234) (0.155) (0.123) (0.126) Observations 132 132 132 132 132 Pseudo R-squared 0.461 0.360 0.293 0.418 0.601 Controls for firm age, registered capital, privatized, bought, and salary. Dependent variable is the residual from the income regression in Table 6. Bootstrapped standard errors (100 repetitions); Standard errors are in parentheses. * p< 0.10, ** p< 0.05, *** p< 0.01 34
  • 35. Individual Fixed Effects Dependent Variable = Start-up founded (subjects start being at risk upon graduation) Note: reported coefficients are hazard ratios Indiv. F.E. 90th quartile 0.651 (0.203) Indiv. F.E. 90th * POST 1.293* (0.206) Indiv. F.E. 10th quartile 0.976 (0.479) Indiv. F.E. 10th * POST 1.130 (0.810) Indiv. F.E. > median 0.624** (0.144) Indiv. F.E. median*POST 7.329*** (2.165) N=1800; Controls not shown; ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. Charles Eesley - The Right Stuff
  • 36. Hypotheses H1: An institutional change reducing barriers to growth will increase entrepreneurship among individuals located relatively higher in the talent distribution. H2: Individuals who show evidence of higher talent in their wage employment careers will experience higher returns to talent (in entrepreneurship) after an institutional change lowering barriers to growth. H3: Individuals who show evidence of higher talent in their wage employment careers will start firms that have higher performance. Charles Eesley - The Right Stuff
  • 37. Shift in Returns to Talent in Entrep. Log(income from start-up) Log(profit margin) Independent Variables (7-5) (7-6) (7-7) (7-8) (7-1) (7-4) Master’s degree -1.037 0.090 -0.449 -0.269 (0.905) (0.928) (0.276) (0.340) POST x Master’s degree 1.783* 0.163 0.236 -0.111 (1.039) (1.054) (0.345) (0.412) High GPA -1.191 -1.417 -0.523 (0.737) (0.903) (0.589) POST x High GPA 2.348*** 2.376** 0.866 (0.887) (1.042) (0.595) Income residual -3.239*** -- -- (0.901) POST x income residual 2.624** -- -- (0.987) Promoted -1.123 -0.900 (1.182) (0.611) POST x Promoted 5.803*** 1.751** (1.678) (0.768) POST-1999 founding year -1.329 -0.139 1.027 -6.443*** -0.301 -2.200** (0.977) (0.686) (0.819) (1.861) (0.455) (0.937) Standard errors are robust. The results are Tobit models, but are robust to a Poisson specification (as well as not taking the natural log). ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. All models include controls for year and region fixed effects as well as revenues, capital, employees, firm age, overseas returnee, and bach. graduation year. 37
  • 38. Robustness checks 1. Returns to talent in wage employment 2. MIT Data – 1998-2000 dotcom boom 3. Increased legitimacy 4. Alternative talent measures 5. Universities and research institutes began to invest and had better information on underlying ability levels 6. Demography of the funders/lead investor community itself might have changed and preferences towards higher ability individuals or technically skilled entrepreneurs (dotcom) 7. The nature of the economic opportunities or competition changed as China liberalized and the available opportunities required those with higher ability levels. Or payoff to skills, demand for innovation/R&D in the economy increased 8. Labor market, Increased college student enrollment, flood of graduate students with limited wage job opportunities 9. Specification
  • 39. Dotcom Boom (MIT Alumni Data) Dependent Variable = Start-up founded (subjects start being at risk upon graduation) Independent vars. Note: reported coefficients are hazard ratios Software firms Software firms only, only EE&CS All Grads, All firms only, all grads grads Master’s degree 0.281*** 0.767 1.226 (0.120) (0.204) (0.098) Doctorate Degree 0.249** 0.733 1.261 (0.153) (0.333) (0.127) Master’s x Years 98-00 7.466** 1.936* 0.936 (5.844) (0.741) (0.152) Doctorate x Years 98-00 5.560* 1.446 0.905 (5.683) (0.849) (0.185) Non-U.S. citizen 1.866* 1.172 0.825 (0.634) (0.292) (0.078) Gender (male=1) 5.814* 3.305*** 1.495 (5.934) (1.302) (0.169) Years 1998-2000 0.001*** 0.002*** 0.011 (0.001) (0.001) (0.002) Obs. 3,266 18,896 19,188 Note: Grad. years 1980 and after; 52 failures; 44,525 total years at risk; ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. All models include controls for Bachelor‟s graduation year (age), Bachelor‟s Major (academic department). Charles Eesley - The Right Stuff
  • 40. 6. Conclusion and Implications Hypothesis: Supported? Decr. barriers to growth = increases entrep. by H1 YES those higher in talent distribution Talent = higher performance in H2 YES entrepreneurship • Institutional Level • Second margin at which the institutional environment affects entrepreneurship • Different market failures affecting diff. individuals and start-ups, at different stages in market development Social Welfare Analysis • ΔSW IndividualΔPS = ΔCS + Level Δ CS •= Δp(Innovation)policies to encourage high ability entrepreneurs (need more Suggestive of + Competition analysis here) Δ PS •= I(profith-profitl) –characteristics associated with performance/innovation Human capital C(prev. organization) Induce greater investments in certain types of human capital • Strategy • Large firm new ventures Cofounder / Investor Choice Analysis • Understanding the drivers of high growth entrepreneurship Increased probability of survival or growth • Institutional environment can shift competition and possibly the types of start- Extra equity or salary to obtain cofounder (amt. required affected by inst. ups being created change) 40
  • 41. Strategy / Economics of Innovation and Technological Change High Tech Entrepreneurship (selection and technical skills and science and •Interact performance) engineering / industry context •Regional level policy experiments •Comparison with US data Current Research Labor Economics Institutional (individual/micro- Variation level) (macro-level) Charles Eesley
  • 42. Research Trajectory •Eesley, Charles. 2009. Who has „The Right Stuff‟? Human Capital, Entrepreneurship and Institutional Change in China. Winner SASE Student Paper Award, June 2009. • Eesley, Charles, Lenox, Michael. 2009. Secondary Stakeholder Actions and the Selection of Firm Targets. Working paper. •Eesley, Charles; Roberts, E.B. 2009. Cutting Your Teeth: Learning from (One or More) Rare Experiences. Under review. •Rockart, Scott & Eesley, Charles. 2009. Prestige and Collaboration Patterns. Work in progress. Entrepreneurial Firms •Roberts, E.B.; Eesley, Charles. Entrepreneurial Impact: The Role of - Selection MIT. Kauffman Foundation, Level 2009. Institutional Feb. • Lenox, M. and Eesley, Charles. 2009. Private Environmental (Stakeholders) - Strategy / Idea Strategy Activism and the Selection and Response of Firm Targets. Journal - Outcomes •Hsu, D.H., Roberts, E.B., Eesley, Charles. 2009. Entrepreneurial - Rate and Direction of Economics Management and Strategy, 18(1), Jan. issue. of Innovative Activity Ventures from Technology-Based Universities: Evidence from MIT. Established Firms Work in progress. • Eesley, Charles; Lenox, Michael. 2006. Firm Responses to - Selection Individual Level Secondary Stakeholder Action. Strategic Management Journal, - Strategy / Idea •Hsu, D.H., Roberts, E.B., Eesley, Charles. 2007. Entrepreneurs 27(8):765-781. - Outcomes from Technology-Based Universities. Research Policy, 36: 768– 788. • Sloan, Frank A.; Eesley, Charles. 2007. Implementing a Public Subsidy for Vaccines. in Pharmaceutical Innovation: Incentives, Competition, and Cost-Benefit Analysis in International Perspective. edited by Frank A. Sloan and Chee-Ruey Hsieh. New York: Cambridge University Press. Right Stuff Charles Eesley - The
  • 43. Directions for Future Work Focus on research designs disentangling causal mechanisms in commercializing innovation Field Experiments – US or abroad Natural Experiments - Work well for institutional and policy questions • Add advisors, Strategic choices • Education/Training Instrumental Variables • Upgradeto find instruments in strategic contexts - Difficult technology Individual Level • High Tech Entrepreneurship • Joe-the-Plumber < Joe-the-bio/nano-tech-founder • Commercializing Innovation Different Institutional Environments • MIT and Tsinghua Data • India and other countries
  • 44. Thank you! Chuck Eesley eesley@mit.edu
  • 45. Performance Log(profit Pr(IPR Independent margin) Log(employees) important) Variables (5-1) (5-3) (5-5) Promoted -0.118 (0.228) 0.380* (0.211) 0.042 (0.966) Log(work exp.) 0.986 (1.230) -2.447** (1.094) 0.668 (4.723) Years of Education 0.004 (0.086) 0.152* (0.082) 0.850** (0.407) Talent (income residual) 0.410** (0.180) -0.12 (0.159) 1.535* (0.837) Prior salary -0.442** (0.170) 0.173 (0.150) -0.231 (0.727) Overseas 0.731** (0.359) 0.128 (0.305) 2.120 (1.672) High GPA 0.308 (0.232) -0.321 (0.213) -0.848 (0.914) Worked in R&D - - 1.639* (0.925) N=150; Standard errors are in parentheses. * p< 0.10, ** p< 0.05, *** p< 0.01 Charles Eesley - The Right Stuff
  • 46. Index of Backup Slides • Future Work • Quotes • Identification • Types of Talent • Theory contribution • Web survey • Constructs/Measures • Four Levels of Social Analysis • Response Bias • Robustness checks • Boundary Conditions – Returns in wage emp. • Opportunity Costs – U.S. data • Descriptive statistics – Placebo regression – # jobs, industries – Legitimacy • Implications – Parallel Trends • China policy reforms – Logit • Tsinghua history – Macroeconomic data • Roy Model – Proportional Hazards Charles Eesley - The Right Stuff
  • 47. Theoretical contribution • Economics of technological innovation and entrepreneurship • Rational expectations, profit max. – not necessarily • Institutional theory • Talent theory of bubbles or technology progress • Theory of the direction of innovation (non-SCOT) – artifact – larger scale story • Large vs. small firms tech waves of creative destruction • Industry differences – more room for strategically influencing perceptions of payoffs – VC herding behavior • Strategic influencing directions (e.g. Eyeonics, YouTube, vaccines - less competition) Charles Eesley - The Right Stuff
  • 48. Reform Easing Constraints on Entry 1988 Reform officially recognizing private businesses with more than 8 employees (Xu and Zhao, 2008) Lower barriers to entry  Lower quality entrepreneurs • Nanda, 2008 • Xu and Zhao, 2008 In the years following the 1988 reduction in barriers to entry individuals located relatively lower in the talent distribution were more likely to found firms. Charles Eesley - The Right Stuff
  • 49. Methods Cox Hazard Rate Model (robust to logit) Prob (founding a firm = 1) = F(α + 1’θi + ’Xi + τt+ ηj + φa+εi) • Dependent variable: First start-up founded • Subjects start being “at risk” of founding a firm at the time of their graduation • Θi = human capital measures • Xi = Set of controls academic dept., region, education, work history, job type, comm. party, overseas educ. or work, family economic status. • Include (τ + η + φ) grad. year, region and Bachelor’s academic dept. fixed effects Charles Eesley - The Right Stuff
  • 50. Reduced Barriers to Entry: 1988 - 1999 Dependent Variable = Start-up founded (subjects start being at risk upon graduation) Note: reported coefficients are hazard ratios; coefficients below 1.0 represent a Independent Variables decreased likelihood of entrepreneurship (N=1,540) Master’s degree 0.675* 0.562** (0.158) (0.152) Doctorate degree 0.344** 0.641 (0.166) (0.323) Low work exper. (0-10 yrs.) 1.333 1.500 (0.351) (0.442) High work exper. (>30 yrs.) 0.060*** - (0.039) - Promoted 0.689 0.686 (0.251) (0.270) High GPA (above median) 0.685* 1.288 (0.150) (0.325) Last Salary (Pre-founding) 0.667*** 0.694*** (0.065) (0.071) Communist Party member 0.760 0.744 0.807 0.77 0.774 0.776 (0.167) (0.166) (0.176) (0.169) (0.173) (0.185) Controls for academic dept., region, education, work history, job type, comm. party, overseas educ. or work, family economic status, graduation year, major, and region fixed effects. Note: 102 foundings; 30,716 total years at risk; ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. All models include controls for Bachelor‟s graduation year (age), Bachelor‟s Major 51 (academic department), and region fixed effects.
  • 51. Income Regression Ordered Logit Dependent variable = salary Independent Variables category (1-6) Master’s degree 0.466*** (0.051) Doctorate degree 0.905*** (0.051) Work exper. (0-10 yrs.) 0.938*** (0.054) Work exper. (10-30 yrs.) 1.065*** (0.047) Work exper. (>30 yrs.) 0.500*** (0.055) Gender (male=1) 0.498*** (0.058) GPA quartile (1st = top) -0.199*** (0.046) GPA quartile (3rd) 0.514*** (0.046) GPA quartile (4th = bottom) 0.047 (0.045) Overseas exper. 0.407*** (0.043) Academia -0.943*** Bach. Dept. and Year Effects (0.051) included. Business 0.335*** Standard errors are in (0.052) parentheses. * p< 0.10, ** p< Government -0.326*** 0.05, *** p< 0.01 (0.044) Pseudo R2 0.148 Number of observations 561 52
  • 52. First Foundings by Year 35 30 Number of Start-ups 25 20 15 10 5 0 First Founding Year
  • 53. Industry Breakdown for First Firms NUMBER OF INDUSTRY FIRMS % AEROSPACE 3 0.90 ARCHITECTURE 13 3.88 BIOTECH AND DRUGS 7 1.09 CHEMICALS 8 2.39 CONSUMER PRODUCTS 17 5.07 ELECTRIC 12 3.58 ELECTRONICS 69 20.60 ENERGY 14 4.18 FINANCE 10 2.99 INTERNET 33 9.85 LAW, ACCOUNTING 22 6.57 MACHINERY 19 5.67 MANAGEMENT 21 6.27 MATERIALS 13 3.88 MED DEVICES 4 1.19 OTHER MFG 16 4.78 PUBLISHING 11 3.28 SOFTWARE 34 10.15 TELECOM 9 2.69 TOTAL 335 100 Charles Eesley - The Right Stuff
  • 54. Income Regression Residuals Founders only Matched sample
  • 55. Acknowledgements • Funding - Ewing Marion Kauffman Foundation Dissertation Fellowship, Charles Zhang, founder and CEO of Sohu.com, and the MIT Entrepreneurship Center • MIT – Professors Diane Burton, Fiona Murray, Edward B. Roberts – Yanbo Wang (PhD student) – Profs. Yasheng Huang and Elena Obukhova – Undergraduate Research Assistants • Stephanie Yaung, Jennifer Sim, Celia Chen • Tsinghua University – Prof. Delin Yang – Li Zhihua (Tsinghua Alumni Association) Charles Eesley - The Right Stuff
  • 56. Percentage of "At-Risk" Individuals Becoming Entrepreneurs (By Education Level) 4 3.5 3 2.5 Percentage (%) 2 Graduate Degree Bachelor's Only 1.5 1 0.5 0 Charles Eesley - The Right Stuff
  • 57. Returns to Talent in Wage Empl. Ordered Logit Dep. Var = salary (6 bands) Master’s degree 0.693*** Non-entrepreneurs only (0.137) Robust to high salary and last salary Master’s x POST -0.228 as well as negative binomial (0.157) specifications. Doctorate degree 1.211*** Controls for bachelor‟s graduation (0.237) year, age, gender, overseas, comm. Ph.D. x POST -0.782*** Party, and family economic status. (0.243) High GPA 0.384** (0.170) High GPA x POST -0.292 (0.193) Ln(yrs. tenure) -0.028*** -0.033*** (0.005) (0.008) Business 0.179 0.24 (0.149) (0.238) N=3,276 job spells; Robust Government -0.883*** -1.054*** standard errors (clustered (0.160) (0.244) at individual level) are in Academia -0.646*** -0.659*** (0.165) (0.248) parentheses. * p< 0.10, ** Years 2000-07 (POST) 1.760*** 1.348*** p< 0.05, *** p< 0.01 (0.157) (0.177) Charles Eesley - The Right Stuff
  • 58. Robustness checks 1. Returns to Talent in wage employment 2. MIT Data – 1998-2000 dotcom boom 3. Increased legitimacy 4. Alternative talent measures 5. Universities and research institutes began to invest and had better information on underlying ability levels 6. Demography of the funders/lead investor community itself might have changed and preferences towards higher ability individuals or technically skilled entrepreneurs (dotcom) 7. The nature of the economic opportunities or competition changed as China liberalized and the available opportunities required those with higher ability levels. Or payoff to skills, demand for innovation/R&D in the economy increased 8. Labor market, Increased college student enrollment, flood of graduate students with limited wage job opportunities 9. Specification
  • 59. Robustness – legitimacy high status = Ever job gov., Comm. Party, Ph.D. Independent Variables Dependent Variable = Year start-up founded (subjects start being at risk upon graduation) Note: reported coefficients are hazard ratios (N = 1,910) (1) (2) Years Education - 0.537*** (0.072) Education x Post-1999 - 1.737*** (0.259) High Status (Gov., Ph.D. and Comm. Party) 1.442 (0.401) 1.269 (0.341) Status x Post-1999 0.590* (0.187) 0.725 (0.224) Post-1999 dummy 0.055*** (0.014) 0.000*** (0.000) 1991-1999 dummy 0.071*** (0.011) 0.064*** (0.010) pair-wise correlations High Status Years of Education 0.101 High GPA 0.080 Promoted 0.040 Ln(work exp.) 0.066 Parents’ Education 0.046 Back N= 626; Standard errors are in parentheses. * p< 0.10, ** p< 0.05, *** p< 0.01
  • 60. Placebo Regression Dependent Variable = Start-up founded (subjects start being at risk upon graduation) Note: reported coefficients are hazard ratios Years of Education 0.630*** (0.086) Educ. x POST 1.082 (0.205) GPA High 0.726 (0.231) GPA x POST 0.716 (0.317) Promoted 0.218** (0.131) Promoted x POST 7.714** (7.210) Ln(workexp) 0.454*** (0.116) Work exp. x POST 0.367*** (0.119) Income residual 0.717 (0.179) Residual x POST 1.333 (0.450) Years 1997-1999 (POST) 0.160 (0.521) 0.690 (0.256) 0.083*** (0.079) 4.362* (3.346) 0.887 (0.349) Years 1990-1996 0.206*** (0.065) 0.307*** (0.098) 0.216*** (0.070) 0.258*** (0.097) 0.479* (0.182) Controls for academic dept., region, education, work history, job type, comm. party, overseas educ. or work, family economic status, graduation year, major, and region fixed effects. N=1,821; Note: 119 failures; 20,541 total years at risk; ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. Charles Eesley - The Right Stuff
  • 61. Identification Concerns: Tests: • 1999 reform lowered entry • MIT Data barriers • Placebo regression • Reform had broader effects • Direct test of returns to • More continuous changes talent in entrep. • Reform did not increase • Incr. in foundings controlling returns to entrepreneurship for macro-economic vars. • Simultaneous changes with • Returns in wage empl. similar results on growth • Qualitative data barriers Charles Eesley - The Right Stuff
  • 62. Boundary Conditions 1) Types of skills and talent necessary to overcome entry barriers are not largely orthogonal to those useful for firm growth 2) Feedback loops are weak: that is, increases in the number of entrepreneurial firms do not strongly increase competition or create significantly better wage employment opportunities 3) Initial relationship between talent and returns to entrepreneurship is not one where primarily those at the top of the talent distribution become entrepreneurs 4) Sufficient variation in the distribution of talent (by measures relevant for wage and entrepreneurial payoffs) in the sample at risk for entrepreneurship Charles Eesley - The Right Stuff Back
  • 63. Institutional Initial institutional Human Capital environment #2 environment Market Ability / Higher Non-Market Ability (Gov. connections, bureaucracy nav., wealth) The increase is among those of relatively higher ability. Back
  • 64. Kaplan-Meier curves Charles Eesley - The Right Stuff
  • 65. Kaplan-Meier curves Charles Eesley - The Right Stuff
  • 66. Kaplan-Meier curves Charles Eesley - The Right Stuff BackBackup
  • 67. Tests and graphs based on the Schoenfeld residuals Charles Eesley - The Right Stuff
  • 68. Histogram of Number of Jobs 60 50 40 Percent 30 20 10 0 0 1 2 3 4 5 Number of Jobs Reported Charles Eesley - The Right Stuff
  • 69. Proportion of Foundings with a Direct Match to the Industry and Founder's Bachelor's Degree 0.45 0.4 0.35 0.3 Proportion 0.25 0.2 0.15 0.1 0.05 0 Before 1999 After 1999 * Differences are statistically significant (p<0.05; t = 1.759) Charles Eesley - The Right Stuff
  • 70. Human Capital and Opp. Costs wages = a+ *schooling high-HC people can either have high a or high wages=a+ 1*HC+( 2-c)*entrep+ 3*HC*entrep Charles Eesley - The Right Stuff
  • 71. Human Capital and Opportunity Costs (incr. opportunity costs, decreased entrepreneurship) $ Wage empl. income Entrepreneurial Returns Higher Human Capital Charles Eesley - The Right Stuff
  • 72. Human Capital and Opportunity Costs (Incr. opportunity costs, decreased entrepreneurship) Entrepreneurial $ Returns Wage empl. income Higher Human Capital Charles Eesley - The Right Stuff
  • 73. Human Capital and Opportunity Costs (incr. entrep. returns to talent) $ Wage empl. income Entrepreneurial Returns Higher Human Capital Charles Eesley - The Right Stuff
  • 74. Robustness: Logit Dep. Var = 1 if founded a firm Indep. variables Graduation Years < 2002 Talent (income resid.) -0.400 -0.643 (0.460) (0.521) Talent x POST 1.099** 1.454*** (0.475) (0.544) Years of Education -0.402 -0.974 (0.311) (0.723) Educ. x POST 0.320 0.424 (0.319) (0.739) Promoted -0.398 0.678 (0.900) (1.718) Promoted x POST 0.294 -0.873 (0.957) (1.796) POST -4.722*** -9.964* -4.742*** -11.388 (0.943) (5.544) (0.915) (13.029) Observations 451 1277 1277 451 Controls for academic dept., region, education, work history, job type, comm. party, overseas educ. or work, family economic status, graduation year, major, and region fixed effects. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.
  • 75. Constructs and Measures • Human Capital/Talent – education, parents’ education, GPA, work experience, income, income residuals, student leader • Variance of returns – categories of wage income, income from start-up, profit • Barriers to Growth • Status/Legitimacy – PhD, government Charles Eesley - The Right Stuff
  • 76. Appendix A Comparison of Key Demographic Characteristics by Survey Wave Variable Responded before Responded during/after Aug. t-stat for equal Aug. 2007 2007 means (N=2,667) (N=299) Age 49.3 54.1 -4.216** Age (founders only) 38.4 37.4 0.602 Bachelor’s Graduation Yr 1980.9 1977.4 3.777** Bach. Grad yr (founders only) 1991.6 1993.2 0.941 Years of Education 17.2 17.0 2.381** Entrepreneur parents 0.09 0.12 -0.713 Entrepreneur 0.29 0.40 -2.168** Privatized 0.10 0.05 1.392 First start-up founded 2000.3 2001.1 -0.661 Tech only 0.28 0.29 0.757 Business only 0.10 0.09 0.235 Gender 0.88 0.90 0.901 Family economic status 3.75 3.85 -1.871* High Salary 3.21 2.93 3.351** Avg. Tenure 6.94 8.01 -2.045* Overseas work exp. 0.26 0.26 -0.126 Number of positions 2.39 2.26 -2.012* High government 0.03 0.03 -0.239 Low government 0.18 0.17 0.617 Last job academia 0.19 0.19 -0.051 Ever job academia 0.32 0.27 2.323** Last job business 0.62 0.61 0.348 Student Leader 0.61 0.57 0.874 GPA Rank 2.28 2.58 -2.661** Bach. Grad Yr. 10th percentile 1954 1953 -- Bach. Grad Yr. 25th percentile 1965 1961 -- Bach. Grad Yr. 50th percentile 1986 1979 -- Bach. Grad Yr. 75th percentile 1996 1993 -- Bach. Grad Yr. 90th percentile 2001 2001 -- 77 ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. Back
  • 77. Charles Eesley - The Right Stuff
  • 78. Benchmarking Tsinghua Data Categories Tsinghua CHNS NBS HH survey NBS HH survey Sample Urban Rural and Urban Urban – self- Urban – employed non.Entrep. Male 0.89 0.53 0.56 0.50 Age 50.13 41.45 36.2 37.2 Married 0.88 0.98 83.4 84.2 Years of Education 17.1 9.1 9.2 9.4 Household Size 3.40 3.9 -- -- Self-employed 0.26 0.14 (4% in 1999) -- (0.8% in 1999) Experienced a 0.13 -- 0.26 0.19 layoff Father’s Educ. 4.11 -- 5.4 5.2 Mother’s Educ. 4.89 -- 6.0 5.9 Parent Self-Empl. 0.08 -- 0.06 0.05 Comm. Party 0.62 -- 0.05 0.18
  • 79. Work History Attributes Non-Founders Founders t-stat (n=2152) (n=670) Variable Mean Age 52.8 42.9 12.388*** Entrepreneur Parents 0.036 0.031 0.614 Privatized 0 0.260 - Male 0.879 0.933 -3.931*** Family Econ. Status (category) 3.834 3.639 4.348*** Recent Salary (category) 2.317 2.045 3.686*** Avg. Tenure 8.074 4.813 7.554*** Overseas 0.126 0.212 -5.525*** Number of Positions 2.115 3.109 -18.605 Ever Job High Government 0.036 0.042 -0.713 Ever Job Low Government 0.242 0.176 3.574*** Last Job Academia 0.166 0.081 5.475*** Ever Job Academia 0.233 0.207 1.402* Last Job Business 0.399 0.687 -13.451*** Master’s 0.402 0.550 7.187*** Ph.D. 0.101 0.115 1.066 Student leader 0.674 0.903 -5.314***
  • 80. Percentage of "At-Risk" Alumni Body Becoming First-Time Entrepreneurs By Year 1.4 1.2 1 Percentage (%) 0.8 0.6 0.4 0.2 0 Charles Eesley - The Right Stuff
  • 81. Four Levels of Social Analysis Williamson (2000) • Social Embeddedness – Culture, Religion (Economic Historians, Anthropologists, Sociologists) • Institutional Environment – “Rules of the Game”, Constitutions, judiciary, politics (Poli. Sci., Inst. Economics) • Governance – Contracts, Theory of the firm (Contract/Info Econ, TCE) • Resource Allocation – Incentive aligment, Quantities and Prices (Neoclassical Economists)
  • 82. Negative Binomial Regressions on Macroeconomic Data Dependent variable = number of firm Independent Variables foundings (1959-2007) R&D to GDP ratio (t-1) -0.143 GDP per capita (in RMB, t-1) (0.723) Shanghai Stock Exchange Market Cap (t-1) 3.43E-04*** 3.30E-04*** Domestic Patents Issued (t-1) (8.59E-05) (8.63E-05) Post-1999 dummy -2.52E-05** -2.36e-05** Post-1988 dummy (1.16E-05) (1.20E-05) Constant 5.01E-06 -1.01e-05** -9.48E-06* Log likelihood (3.60E-06) (4.82E-06) (5.04E-06) Num. obs. 4.326*** 3.856*** 2.901*** Pseudo R2 (0.652) (0.530) (0.590) Data merged from the State Statistics Bureau, Chinese Statistical Yearbooks ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. 83
  • 83. Strategic and Competitive Implications of Different Institutional Shifts Lower Barriers to Entry Lower Barriers to Growth Market/Commercializat Increase in high and low ability Unclear predictions on whether ion Talent is Orthogonal entrepreneurs increase is among high or low ability to ‘Bureaucratic’ Talent No prediction on the type of Easier to overcome the opportunity firms founded costs for entrepreneurship Possibly greatest increase in Increase in high growth firms marginal firms which were not More high-growth entrepreneurial profitable with the previously opportunities supports venture high cost of entry capital Increased competition (possibly lower profit margins?) Market/Commercializat Relative increase in low ability Increase in high ability entrepreneurs ion Talent is NOT entrepreneurs Easier to overcome the opportunity Orthogonal to Those who cannot maintain costs for entrepreneurship wage employment can overcome Easier to recruit talented co- ‘Bureaucratic’ Talent the barriers to entry founders (a large component is Relative increase in low growth More high-growth entrepreneurial common between the two) firms opportunities supports venture Increased competition, smaller, capital less profitable firms Increase in high-growth firms Increase in innovative firms (higher returns can support higher risk of an innovation strategy) Charles Eesley - The Right Stuff
  • 84. China’s S&T Policy Reform 1978-85 1986-97 1998-2006 •Deng Xiao •1992 Tour Ping reform •Private ownership •Adoption of •Bankruptcy law report (1975) (1999) medium and Long for SOEs (1986) •Open door Term S&T •National Natural •Promotion of VC/PE Strategic Plan policy (1978) Science Fndtn. (1998) (2006) •National Key (1986) Tech. R&D •CAS Knowledge •>7 emp. Innov. Prog. (1998) Program permitted (1984) •Innov. Fund for Tech. Torch Program •Univ. reform (1988) SMEs (1999) (1985) •Stock Exchange •SE Zones (1990) •Join WTO (2001) created (1980) •1997
  • 85. Tsinghua Univ. •Established in Beijing in 1911 •1952 reorganized Soviet style •1966-1976 Battlefield during Cultural Revolution •1978 restored departments in sciences, economics and management, and humanities •1984 – First Graduate school in China created at Tsinghua •1998 – Tsinghua Science Park established 86 Charles Eesley - The Right Stuff
  • 86. Model Roy Model as extended by Borjas (1987) Labor markets 0 and 1, respectively. Log earnings in the wage sector: w0 = μ0 + ε0 ε0 ~ N (0, ς02) The wage sector earnings would be the following if everyone from the wage sector were to migrate to entrepreneurship (ignoring general equilibrium effects): w1 = μ1 + ε1 ε1 ~ N (0, ς12) . Assume that the cost of becoming an entrepreneur is C π = C/w0 π is constant, meaning that C is directly proportional to w0 The correlation between entrepreneur and wage worker earnings is ς01 = cov(ς0, ς1). A worker will choose entrepreneurship if Charles Eesley - The Right Stuff
  • 87. Entrepreneur‘s Perspective • One pair of founders had very high human capital with one being a lawyer and an MBA and the other having a Ph.D. Founded in 2003 and said that: I then spent 20 years in the Bay area in life sciences companies. In the mid- 1990’s I came back to China to survey biotech companies in China and found that the environment was not ready yet. …spent an entire year just looking for the right office space…each product must be registered and approved by the government. It’s an expensive and time consuming procedure. She eventually found space for the company’s first store in a children’s museum which was perfect since they were selling toy bears aimed at children. This also allowed them to “hide” from government inspectors. Charles Eesley - The Right Stuff
  • 88. Entrepreneur‟s Perspective on Pre-period (4/5) • Environment was costly, not prohibitive for entrepreneurship. One entrepreneur reported that she: …spent an entire year just looking for the right office space…each product must be registered and approved by the government. It’s an expensive and time consuming procedure. She eventually found space for the company’s first store in a children’s museum which was perfect since they were selling toy bears aimed at children. This also allowed them to “hide” from government inspectors. • Individuals managed to find ways to raise funding At that time there was not much VC investment in China (1997) so they raised the first round of money among their family and divided up the shares according to how much each of them could raise from family. – JX Charles Eesley - The Right Stuff
  • 89. Interview Quotes In the 1990s … government was giving them less support. Many universities created “university –run” enterprises and were basically selling off the periphery of campus.Even high schools had school—run enterprises that were considered acceptable. There is a Tsinghua name to many of them. - GC – 2nd generation investor 90 Charles Eesley - The Right Stuff
  • 90. Interview Quotes, cont. Strategy-making is different in China than in the US. In China you pick up a person who has resources, then you choose the appropriate field to enter then you come up with the strategic goals, then the vision. In the US he feels that it is the opposite. - WW – Sea Corp. Charles Eesley - The Right Stuff
  • 91. Percentage of Businessmen/Businesswomen Becoming Entrepreneurs 3.00% 2.50% Percentage 2.00% Last Job w as in Business 1.50% Ever had a job in 1.00% Business 0.50% 0.00% 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 Founding Year Charles Eesley - The Right Stuff
  • 92. Percentage of Government Officials Becoming Entrepreneurs 12.00% 10.00% Percentage 8.00% High level 6.00% government official position 4.00% Low level government 2.00% official 0.00% 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 Founding Year Charles Eesley - The Right Stuff
  • 93. Why China? • Tech-based entrepreneurship in developing countries rarely appears in academic literature (Lu 1997, 2000; Puga and Trefler, 2005) • Vernon‘s (1966) product-cycle model • 19892004 China 29% vs. US 1% (State Statistics Bureau) 19782004 # employed in private business up 300X • Policies and institutions changing rapidly (Cull & Xu, 2006; Nee, 1998; 1992; 1996; Peng & Heath, 1996; Steinfeld, 2007) Charles Eesley - The Right Stuff
  • 94. Context Charles Eesley - The Right Stuff
  • 95. Entrepreneurship and Institutions (Gollin, JPE 2002)
  • 96. Research Trajectory Science and Tech.-based Entrepreneurship Hsu, D.H., Roberts, E.B., Eesley, Charles. 2007. Entrepreneurs from Technology-Based Universities. Research Policy Roberts, Edward B., Hsu, David , Eesley, Charles. Entrepreneurial Ventures from Technology-Based Universities: An Empirical First Look. Eesley, Charles. Roberts, Edward B. The Second Time Around?: Serial Entrepreneurs from MIT. Eesley, Charles; Roberts, Edward. Cutting Your Teeth: Learning from (One or More) Rare Experiences. Hsu, D.H., Roberts, E.B., Eesley, Charles. 2007. Experienced Founders and Early Financing. (in progress) Eesley, Charles. Who Has the Right Stuff? (in progress) Charles Eesley - The Right Stuff
  • 97. Future Work •Eesley, Charles. 2008. Who has „The Right Stuff‟? Human Capital, Entrepreneurship and Institutional Change in China. Winner SASE Student Paper Award, June 2008. Individuals •Eesley, Charles; Roberts, E.B.Firms Cutting Your Teeth: Learning Entrepreneurial 2008. from (One or More) Rare Experiences. Under review. - Selection •Hsu, D.H., Roberts,- E.B., Eesley, Charles. 2007. Entrepreneurs Outcomes from Technology-Based Universities. Research Policy, 36: 768– Institutional Level Strategy 788. - Rate and DirectionEesley, Charles, Lenox, Michael. 2008. Secondary Stakeholder • of Innovative Activity Actions and the Selection of Firm Targets. Working paper. Stakeholders •Lenox, M. and Eesley, Charles. 2009. Private Environmental Established Firms Activism and the Selection and Response of Firm Targets. Journal - Selection of Economics Management and Strategy, 18(1), Jan. issue. - Outcomes •Eesley, Charles; Lenox, Michael. 2006. Firm Responses to Level Institutional Secondary Stakeholder Action. Strategic Management Journal, 27(8):765-781. - The Right Stuff Charles Eesley
  • 98. Hypotheses H1: An institutional change reducing barriers to growth will increase entrepreneurship among individuals located relatively higher in the talent distribution. H2: Individuals who show evidence of higher talent in their wage employment careers will experience higher returns to talent (in entrepreneurship) after an institutional change lowering barriers to growth. H3: Individuals who show evidence of higher talent in their wage employment careers will start firms that have higher performance. Charles Eesley - The Right Stuff