Entrepreneurship in the macro-economic context Roy Thurik Erasmus School of Economics and Montpellier Business School IMES (Innovation, Management, Entrepreneurship and Sustainability) Conference, May 30, 2019, University of Economics, Prague

Economics of entrepreneurship What is entrepreneurship? Psychology v economics Nascents v incumbents Engagement levels Orientation Causes? Empirics Jack of all trades Occupational choice Consequences? Effect measure Remuneration Persons Firms Industries Economies

Economics of entrepreneurship What is it? Economics v psychology Incumbents v nascents Engagement levels Entrepreneurial orientation Causes? Empirics Occupational choice, jack of all trades Biology Consequences? Effect measures: remuneration New effect measures: health, happiness, satisfaction Persons Firms Industries Economies

Economics of entrepreneurship What is it? # firms # nascents Turbulence Intrapreneurship Causes? Regulations - Micro reforms Education Culture Consequences? Growth Recovery from recession Persons Firms Industries Economies

Economics of entrepreneurship Persons Firms Industries Economies Interplay with level of economic development Interplay with growth / unemployment Interplay with the cycle

Total entrepreneurial activity and GDP/capita

Stylized fact Business ownership per workforce per capita income

Stylized fact Business ownership per workforce per capita income

Stylized fact Business ownership per workforce per capita income

What will I do? Entrepreneurship and the cycle Ph.D. Köllinger and A.R. Thurik (2012), Review of Economics and Statistics, 94(4), 1143-1156. Thurik, A.R. (2014), Entrepreneurship and the business cycle, IZA World of Labor, 90, (doi: 10.15185/izawol.90). Erken, H., P. Donselaar and A.R. Thurik (2018), Total factor productivity and the role of entrepreneurship, Journal of Technology Transfer, 23(6), 1493-1521. Thurik, A.R., M.A. Carree, A. van Stel and D.B. Audretsch (2008), Does self-employment reduce unemployment? Journal of Business Venturing, 23(6), 673-686. Cima, J., J. Costa, C. Pinto and T. Pinto (2017), The dynamics of self-employment and unemployment during the great recession, International Review of Entrepreneurship, 15(3), 367-374. Parker, S.C. (2018), The Economics of Entrepreneurship, 2nd edition, Cambridge University Press, Sections 3.3 and 6.5.

What do we want to know? Is there a cyclical effect in entrepreneurship? Is there co-variation between entrepreneurship and the business cycle? Pro- or counter-cyclical? Pre- or post-cyclical? Can this co-variation be interpreted as causality? What are the economic mechanisms? What can policy do?

Would this make sense?

Practical background Banking crisis (cleaning up) Global aspect (international coordination) Keynesian remedies Country budget crisis (monetary and fiscal issues) Budget austerity (politics) Schumpeterian remedies? Micro reforms Freeing up labor and product markets H1: comes from an assumption of the models of B&G and C&F. The point of these papers was to argue that agency costs are counter-cyclical. H2: Real business cycle model of Rampini has entrepreneurial activity as an endogenous variable. Extent of entrepreneurial activity depends on external productivity shocks (which are procyclical and increase the share of entrepreneurs) and agency costs (which are counter-cyclical and negative influences the availability of financing for entrepreneurs during recessions). Hence, Rampini concludes that entrepreneurship is pro-cyclical even if entrepreneurs have access to financial intermediaries and optimize their behavior over the cycle.

What does the literature say about entrepreneurial activity and the cycle? Hypothesis 1: ‘Entrepreneurial activity is constant over the business cycle’ Bernanke and Gertler (AER, 1989) Carlstrom and Fuerst (AER, 1997) Hypothesis 2: ‘Entrepreneurship is pro-cyclical’ Rampini (JME, 2004) Congregado, Golpe and Parker (IZA, 2009) Golpe (Huelva, 2009) No a priori Frankish, Roberts, Storey (London, 2009) Bonnet and Renou-Maissant (E&P, 2000) H1: comes from an assumption of the models of B&G and C&F. The point of these papers was to argue that agency costs are counter-cyclical. H2: Real business cycle model of Rampini has entrepreneurial activity as an endogenous variable. Extent of entrepreneurial activity depends on external productivity shocks (which are procyclical and increase the share of entrepreneurs) and agency costs (which are counter-cyclical and negative influences the availability of financing for entrepreneurs during recessions). Hence, Rampini concludes that entrepreneurship is pro-cyclical even if entrepreneurs have access to financial intermediaries and optimize their behavior over the cycle.

What does the literature say about entrepreneurial activity and the cycle? Entrepreneurs create new firms displacing old ones Entrepreneurs absorb risks Entrepreneurship is a way out of unemployment Entrepreneurship is a financial vehicle Entrepreneurship as a cyclical indicator H1: comes from an assumption of the models of B&G and C&F. The point of these papers was to argue that agency costs are counter-cyclical. H2: Real business cycle model of Rampini has entrepreneurial activity as an endogenous variable. Extent of entrepreneurial activity depends on external productivity shocks (which are procyclical and increase the share of entrepreneurs) and agency costs (which are counter-cyclical and negative influences the availability of financing for entrepreneurs during recessions). Hence, Rampini concludes that entrepreneurship is pro-cyclical even if entrepreneurs have access to financial intermediaries and optimize their behavior over the cycle.

Data 22 OECD countries, annual data for 1972-2007 Sources: Variables: Compendia 2007.1 (business ownership rate) See van Stel (IEMJ, 2005) OECD (GDP and unemployment rate) Variables: Business ownership rate Self-employed and incorporate business owners In % of labor force (employed + unemployed) Harmonized across countries and over time GDP in national currencies and constant prices (2000) Standardized unemployment rates GEM data 2001-2006: robustness check

Decomposition of trend and cycle Business cycle is defined as GDP deviations from long term growth trend Decomposition using the Hodrick-Prescott filter (JMCB, 1997) with a smoothing parameter of 6.25 for annual data (Ravn and Uhlig, RES, 2002) Business ownership cycle ……idem Unemployment cycle …..idem HP filter is the standard method for de-trending macroeconomic data.

Average deviations from trend across 22 countries Entrepreneurship was “leading the way” out of the oil crisis recession (1975-1980). Also, it was leading the long cyclical upturn in the mid 1990’s that culminated in the burst of the burst of the high-tech bubble in 2000 Finally, it also led the recovery from the recession after 2001.

Average deviations from trend across 22 countries Entrepreneurship was “leading the way” out of the oil crisis recession (1975-1980). Also, it was leading the long cyclical upturn in the mid 1990’s that culminated in the burst of the burst of the high-tech bubble in 2000 Finally, it also led the recovery from the recession after 2001.

Two relationships ? GDPt Et ? GDPt Et

Two relationships <0 GDPt Et+1,2 >0 GDPt+1,2 Et

Two relationships GDPt Et+1,2 GDPt+1,2 Et <0 >0 Granger causality test business-ownership cycle  GDP cycle: significant >95% Granger causality test GDP cycle  business ownership cycle: significant >99% (based on VAR including cycles of business ownership, GDP and unemployment)  Entrepreneurship is an early indicator of the cycle in most countries and years

Playing with lags Et Et+n GDPt =0 GDPt+n

Playing with lags Et Et+n GDPt Et Et+n GDPt >0 <0 GDPt+n GDPt+n

Just playing with data but … There is much evidence that entrepreneurship leads to economic growth spillovers and innovation competition newness through turbulence variation and learning share of self-employed Entrepreneurs during the boom ‘what is high must come down’ Opportunity costs of entrepreneurship Entrepreneurs during the recession Opportunity costs and prices of production factors are low Low risk awareness No new hardware; no sunk costs Incumbents are busy downsizing

Conclusion 1 Entrepreneurship leads the way out of recessions in most cases Rejects earlier theoretical hypotheses Using two indicators and large data sets New empirical finding Remaining heterogeneity across countries and time Why Rampini (2004) doesn’t hold Some agents are risk seeking during recessions Entrepreneurship is a career choice that doesn’t pay Overconfidence leads to an information structure where lenders may be better informed (≠ principle agent structure)

Conclusion 2 Policies for encouraging entrepreneurship during the crisis Do not raise unemployment benefits Koellinger and Minniti (EL, 2009) Bureaucracy and market entry regulations Grilo and Thurik (ICC, 2008) Van Stel et al. (SBE, 2007) Financial support (e.g. tax based venture capital funds) Jobbers, freelancers, etc Patents and students Governments are not alone

Conclusion 3 Three links between entrepreneurship and the economy: Explanations of economic growth Keynesian, Romerian and Schumpeterian Entrepreneurial economy Audretsch and Thurik (ICC, 2001) Entrepreneurship and the cycle We need integrated models!

Conclusion 3 Is the last ‘great recession’ different? Country heterogeneity and world cycles Entrepreneurship has many faces! Data and analyses are inclusive policy effects Which policy measures influence entrepreneurship? Is there a cyclical effect of these measures?

Entrepreneurship in the macro-economic context Roy Thurik Erasmus School of Economics and Montpellier Business School IMES (Innovation, Management, Entrepreneurship and Sustainability) Conference, May 30, 2019, University of Economics, Prague