The Economic Impact of Entrepreneurship: Comparing International Datasets

• Published date: 01 March 2014
• Date: 01 March 2014
• Author: Minjie Zhang,Sofia Johan,Douglas Cumming
• DOI: http://doi.org/10.1111/corg.12058

The Economic Impact of Entrepreneurship:

Comparing International Datasets

Douglas Cumming*, Sof‌ia Johan, and Minjie Zhang

ABSTRACT

Manuscript Type: Empirical

Research Question/Issue: What is the impact of entrepreneurship on GDP/capita, unemployment, exports/GDP, and

patents per population across countries? Is the impact of entrepreneurship mitigated by legal and cultural differences across

countries? Do different international datasets provide different answers to these questions? We empirically compare the

impact of entrepreneurship on GDP/capita, unemployment, exports/GDP, and patents per population across countries by

examining three datasets from the World Bank, the OECD, and Compendia.

Research Findings/Insights: Based on a comprehensive sample of all available countries and years, with the World Bank

data being the most comprehensive, we f‌ind entrepreneurship has a signif‌icantly positive impact on GDP/capita, exports/

GDP, and patents per population, and a negative impact on unemployment. Inferences from the Compendia data are very

consistent. By contrast, inferences from the OECD data are not supportive of any of these propositions.

Theoretical/Academic Implications: Our f‌indings point to institutional and cultural impediments to the effectiveness of

entrepreneurship. Most notably, the impact of entrepreneurship is signif‌icantly mitigated by excessively strong creditor

rights that limit entrepreneurial risk-taking. Furthermore, the data indicate that cultural attitudes associated with low

risk-taking limit the effectiveness of entrepreneurship. By contrast, the impact of entrepreneurship on exports/GDP does

not appear to be directly tied to costs of exporting, which is perhaps best explained by the new economy goods and services

created by entrepreneurs that depend less on such costs. For some subsets of the data we f‌ind evidence consistent with the

view that top tier venture capital funds enhance the impact of entrepreneurship on GDP/capita. Finally, our results show

how different def‌initions of new business entry matter for empirical analysis of entrepreneurship across countries.

Practitioner/Policy Implications: The data highlight the importance of access to f‌inance without downside costs so that

entrepreneurs are encouraged to take risk. Further, the data highlight institutional differences in risk attitudes that more

generally inhibit risk-taking and thereby limit the effectiveness of entrepreneurship. Moreover, the data highlight a central

role for careful measurement of entrepreneurial activities and for inclusion of as many countries and years as possible in

order to effectively analyze the impact of entrepreneurship.

Keywords: Corporate Governance, Law and Entrepreneurship, Unemployment, Patents, Venture Capital

INTRODUCTION

Neoclassical studies of growth have focused on the role

of capital and labor in stimulating economic develop-

ment (Solow, 1956, 1957). In more recent years, however, it

has become increasingly recognized that entrepreneurship

is a key driver of economic growth (e.g., Acs, Audretsch, &

Strom, 2009; Acs, Desai, & Hessels, 2008; Audretsch, 2007a,

2007b; Audretsch & Acs, 1988; Baumol, 1990; Chavis,

Klapper, & Love, 2011; Fairlie & Chatterji, 2013; Klapper &

Love, 2011; Marcotte, 2012; McMullen, 2011; Naude, 2010;

Stam & Wennberg, 2009; Thurik, Carree, van Stel, &

Audretsch, 2008). For this reason, government bodies

around the world have sought ways to stimulate growth

through entrepreneurship,including access to entrepreneur-

ial f‌inance (Bonini, Alkan, & Salvi, 2012; Fossen, 2011;

Nahata, 2008; Wang & Wang, 2012) and appropriate gover-

nance and legal protections for contracts, shareholders, and

creditors (e.g., Acemoglu & Johnson, 2005; Armour &

Cumming, 2006, 2008; Fan & White, 2003; Klapper, Laeven,

& Rajan, 2006). For instance, many government bodies

around the world have implemented direct investment pro-

grams to f‌inance entrepreneurs through incubation centers

*Address for correspondence: Douglas Cumming, Schulich School of Business, York

University,4700 Keele St, Toronto,ON, Canada M3J 1P3. Tel: +1-416-736-2100 X 77942;

E-mail: douglas.cumming@gmail.com

162

Corporate Governance: An International Review, 2014, 22(2): 162–178

© 2014 John Wiley & Sons Ltd

doi:10.1111/corg.12058

and government venture capital funds (Cumming & Fischer,

2012). Governments have likewise implemented tax policies

to stimulate entrepreneurial activity (World Bank, 2004).

Prior work is highly consistent with the view that these

policies are important for stimulating entrepreneurship

(Keuschnigg & Nielsen, 2003, 2004).

In this paper, we seek not to assess and measure the

determinants of entrepreneurial activity, but rather, the

effect of entrepreneurial activity through newly established

f‌irms on economic growth. Our approach differs from

national studies of the effect of entrepreneurship on growth

(e.g., Audretsch & Keilbach, 2004, 2007) in two primary

ways. First, we examine newly available international data,

which enables one to assess whether or not there are legal

and institutional differences across countries thatenhance or

impede the effect of entrepreneurship on growth. Second,

we examine three different datasets that cover the same sta-

tistics but with different measurements and with different

countries. To this end, we explore whether or not the rela-

tionship between entrepreneurship and growth is contin-

gent on the use of different data, different countries, and

different time periods. We contribute to the theoretical

entrepreneurship literature by examining the economic

implications of entrepreneurship in reference to legal and

institutional barriers, and by providing international evi-

dence in reference to different datasets that measure entre-

preneurship in different ways.

In our analyses, we examine three datasets over the years

2004–2011 that offer different measures of new business

start-ups, and each with different coverage in terms of years

and countries. The f‌irst and largest entrepreneurship dataset

is that from the World Bank. The World Bank sample com-

prises information on business start-ups for 125 countries.

The second sample is fromthe OECD, which covers 24 coun-

tries from Western and Eastern Europe as well as Brazil,

Canada and the US. The third sample is from Compendia,

which covers 11 Western European Countries, Canada, and

the US.

The data highlight a number of interesting features about

the impact of business start-ups on GDP/capita, exports/

GDP, unemployment, and patents per population. First,

there are signif‌icant differences across the data in terms of

both the summary statistics and the regression analyses. In

the summary statistics, the OECD data show a negative

effect of business start-ups on GDP/capita, exports/GDP,

and patents per population, and a positive effect of business

start-ups on unemployment. In the regression analyses,

these effects are statistically insignif‌icant when we control

for other things being equal. Put differently, the implication

from the OECD data is that business start-ups are either

harmful to the economy, or the data are incomplete or

improperly measured for the purpose of assessing the eco-

nomic impact of business start-ups. Because both the World

Bank and Compendia data offer the exact opposite infer-

ences – i.e., that business start-ups have a positive economic

impact – we are inclined to believe that the OECD data that

we examine herein are not perfectly appropriate for assess-

ing the economic effects of business start-ups.

Second, from the World Bank and the Compendia data, we

infer that business start-ups positively impact GDP/capita,

exports/GDP, and patents per population, and negatively

impact unemployment. These f‌indings are statistically and

economically signif‌icant in regression analyses controlling

for other things being equal. In the course of our analyses,

we discovered that many of these country-year level control

variables are highly correlated, and hence we acknowledge

at the outset that the economic signif‌icance of our f‌indings

are highly contingent on the regression specif‌ications. Nev-

ertheless, in a wide range of plausible specif‌ications, our

f‌indings are highly statistically signif‌icant. Moreover, even

in our most conservative estimates, the data indicate that the

economic signif‌icance of our results is very pronounced.

Based on the most complete World Bank data, we show that

a 1 percent increase in new business start-ups in one year

improves GDP/capita in the subsequent year by approxi-

mately 0.24 percent relative to the mean values, reduces

unemployment by 0.13 percent, increases exports/GDP by

0.03 percent, and increases patents per population by 0.29

percent. Put differently, a one standard deviation increase in

business density gives rise to a 38.44 percent increase in

GDP/capita relative to mean values, a 20.02 percent reduc-

tion in unemployment, a 5.33 percent increase in exports/

GDP, and a 51.99 percent increase in patents per population.

Third, we show that the effect of entrepreneurship on

economic outcomes is mitigated by legal and institutional

barriers to risk-taking, but enhanced by investors that stimu-

late risk-taking. In particular, countries with stronger

creditor rights make costs of borrowing relatively higher

for entrepreneurs, thereby reducing entrepreneurial risk-

taking. Similarly, countries with higher uncertainty avoid-

ance indices, or less cultural acceptance of risk, similarly

have a smaller impact of entrepreneurship. Therefore, our

study has shown how legal, institutional, and cultural bar-

riers to risk-taking inf‌luence the overall economic impact of

entrepreneurship. Entrepreneurs are more likely to set up

new entities and in turn have more positive impact on eco-

nomic growth only if the benef‌its of forming new ventures

outweigh the related costs arising from such legal, institu-

tional, and cultural barriers. In addition, we f‌ind evidence

consistent with the view that top-tier venture capital funds

enhance the impact of entrepreneurship. Note, by contrast,

that we do not f‌ind similar evidence of other legal and

institutional differences across countries; for example, the

physical cost of exports is not statistically related to the

propensity of start-ups to stimulate exports.

This paper is organized as follows. The next section dis-

cusses the prior literature and hypotheses. Thereafter we

present the data, summary statistics, and regression evi-

dence. The last section provides concluding remarks and

discusses the limitations of the paper as well as future

research directions.

RELATED LITERATURE AND HYPOTHESES

Prior research is consistent with the view that entrepreneur-

ial activities in newly established f‌irms will have a pro-

nounced effect on economic growth for the following

reasons. First, start-ups have a direct effect in facilitating the

development of new capacities. The importance of entrepre-

neurs was f‌irst fully recognized by Schumpeter (1911/1934,

1939, 1942). Schumpeter described the entrepreneurial

THE ECONOMIC IMPACT OF ENTREPRENEURSHIP 163

Volume 22 Number 2 March 2014© 2014 John Wiley & Sons Ltd