Options
Explaining differences in the structure of early-stage entrepreneurial activity across countries
Type
conference paper
Date Issued
2010-10-01
Author(s)
Arenius, Pia
Abstract
In the Global Entrepreneurship Monitor (GEM) project, total early -stage entrepreneurial activity (TEA) is frequently used as main indicator of entrepreneurial activity in a country. TEA consists of two components: nascent entrepreneurs (NE) and new business owners (NB). These two components can be considered as subsequent stages in the entrepreneurial process. While it is clear that not every nascent entrepreneur will develop an operational business, a somewhat similar ratio between the numbers of nascent entrepreneurs and new business owners could be expected for countries with a similar institutional background and level of development (Wennekers et al 2005; Carree et al. 2002). However, the ratio between these two components varies substantially. In the 2005 GEM-study, the NB/NE ratio ranged from .3 in France (i.e. only .3 NB for every NE) to 2.5 in Brazil (Arenius & Ehrstedt, 2008). There are also large differences between countries with a similar level of entrepreneurial activity and economic development.
There are at least two reasons for differences in the NB/NE ratio between countries: It is reasonable to assume that different institutional framework condition influence the process of business creation and lead to unequal completion ratios of nascent entrepreneurs. Results from PSED-type studies show that across countries different shares of nascent businesses become operational over time (Parker & Belghitar 2006). GEM data report significantly higher levels of early-stage entrepreneurship in developing compared to developed economies than do the World Bank business entry data. The magnitude of the difference between the datasets across countries is related to the local institutional and environmental conditions for entrepreneurs (Acs et al. 2008).
However, given the substantial differences between countries with a somewhat similar institutional setting, there is likely also another explanation: Due to language, translation, and cultural differences people might respond differently to GEM screening questions. Reynolds (2009) has stressed the significance of screening item effects in estimating the nascent entrepreneurship rate. He did this for the U.S. but acknowledges that there might be major differences between countries just because of translation differences. Comparing Australia and the U.S., Davidsson & Reynolds (2009) find a substantially higher participation rate in the U.S. which might partly be due to a higher "willingness to be identified as a nascent entrepreneur" at an early stage, i.e. social desirability. This is confirmed by the fact that U.S. nascent entrepreneurs on average
Our aim is to disentangle these effects by using data from the GEM and PSED-type studies from different countries. Based on 2001-2007 pooled GEM data, we calculate NE and NB rates, excluding all countries with less than 4000 respondents. We run OLS models to explain the NB/NE-ratio and use the acceptance of entrepreneurship as a career choice (as indicators for social desirability) and language, as explanatory variables, while controlling for different sets of macroeconomic and institutional factors. For comparative reasons we do this analysis also for the established business /new business (EB/NB) ratio across countries.
Additionally, we compare gestation activities and business birth rates of nascent entrepreneur cohorts across countries that conducted PSED-type studies. PSED-type studies use the same screening questions for nascent entrepreneurs as the GEM project.
Our results contribute to the understanding of the entrepreneurial process and have implications for research and the methodological set-up of the GEM project. We question the widespread use of the TEA rate as overarching indicator and recommend to use either the NE or NB measure instead. Our results call for more studies of what GEM actually measures and internationally comparative studies on the wording of GEM screening questions.
There are at least two reasons for differences in the NB/NE ratio between countries: It is reasonable to assume that different institutional framework condition influence the process of business creation and lead to unequal completion ratios of nascent entrepreneurs. Results from PSED-type studies show that across countries different shares of nascent businesses become operational over time (Parker & Belghitar 2006). GEM data report significantly higher levels of early-stage entrepreneurship in developing compared to developed economies than do the World Bank business entry data. The magnitude of the difference between the datasets across countries is related to the local institutional and environmental conditions for entrepreneurs (Acs et al. 2008).
However, given the substantial differences between countries with a somewhat similar institutional setting, there is likely also another explanation: Due to language, translation, and cultural differences people might respond differently to GEM screening questions. Reynolds (2009) has stressed the significance of screening item effects in estimating the nascent entrepreneurship rate. He did this for the U.S. but acknowledges that there might be major differences between countries just because of translation differences. Comparing Australia and the U.S., Davidsson & Reynolds (2009) find a substantially higher participation rate in the U.S. which might partly be due to a higher "willingness to be identified as a nascent entrepreneur" at an early stage, i.e. social desirability. This is confirmed by the fact that U.S. nascent entrepreneurs on average
Our aim is to disentangle these effects by using data from the GEM and PSED-type studies from different countries. Based on 2001-2007 pooled GEM data, we calculate NE and NB rates, excluding all countries with less than 4000 respondents. We run OLS models to explain the NB/NE-ratio and use the acceptance of entrepreneurship as a career choice (as indicators for social desirability) and language, as explanatory variables, while controlling for different sets of macroeconomic and institutional factors. For comparative reasons we do this analysis also for the established business /new business (EB/NB) ratio across countries.
Additionally, we compare gestation activities and business birth rates of nascent entrepreneur cohorts across countries that conducted PSED-type studies. PSED-type studies use the same screening questions for nascent entrepreneurs as the GEM project.
Our results contribute to the understanding of the entrepreneurial process and have implications for research and the methodological set-up of the GEM project. We question the widespread use of the TEA rate as overarching indicator and recommend to use either the NE or NB measure instead. Our results call for more studies of what GEM actually measures and internationally comparative studies on the wording of GEM screening questions.
Language
English
Keywords
Global Entrepreneurship Monitor (GEM)
Panel Study of Entrepreneurial Dynamics (PSED)
entrepreneurial process
HSG Classification
contribution to scientific community
Refereed
No
Publisher
Imperial College
Publisher place
London
Event Title
4th Global Entrepreneurship Research Conference
Event Location
London
Event Date
01.-02.10.2010
Subject(s)
Division(s)
Eprints ID
201675