ABSTRACT
This paper implements a multiple linear regression model on panel data of a sample of 64 countries over the period (2002–2015), in order to examine the influence of Entrepreneurship on the growth of output per worker. Our estimation strategy allows for providing a framework that accounts for some of the conflicting findings on the entrepreneurship–growth nexus. We account for the endogeneity of entrepreneurship by using a valid instrument, andfor the level of development and its interaction with entrepreneurship, we also account for the size of the informal economy as a moderating variable of the entrepreneurship–growth nexus. We find that the effect of entrepreneurship on growth is positive and significant, but the analysis fails to support that the relation varies by level of development and the size of the informal economy. Policy implications point to fostering entrepreneurship and designing policies that take into consideration the macroeconomic environment of the country.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 See Gould and Ruffin (Citation1993) and Plosser (Citation1992) for more details.
2 For more information, please see https://www.gemconsortium.org/.
3 GEM adopted this classification from the World Economic forum (WEF), to which we refer for the classifications of countries for the years before 2008, since GEM started to use this classification in 2008.
4 We have implemented a Hausman test in order to compare between fixed and random effect specifications, and the null that random effect model is preferred was rejected at 1%. Moreover, a Wald test for adding time-fixed effects indicated that we do not need to include them in our model, since we failed to reject the null that the coefficients for all year dummies are jointly equal to zero.
5 This might be caused by the high variability of this factor for our sample of countries, which was discussed in Section 3.
6 The reported graphs in and are from the OLS estimation only, since the graphs from 2SLS are very similar and show the same behavior of the examined relationship.
7 The results still insignificant and behaves equally considering each level of economic development separately.
8 The results are shown in . This is a heteroskedastic robust version of the Durbin–Wu–Hausman test (Durbin Citation1954; Hausman Citation1978; Wu, Citation1974), which is reported when using robust VCE in the estimation.