Steve Jobs or No Jobs? Entrepreneurial activity and performance among Danish college dropouts and graduates

When searching for the keywords entrepreneurship and education, there are many studies that deal with the impact of entrepreneurship education. In our discussion, we would like to focus on education in general.

Discussions have also addressed whether one can learn to become an entrepreneur or whether one is born with it, i.e., nature versus nurture. Recent twin studies suggest that both nature and nurture matter (Nicolaou and Shane 2010; Lindquist et al. 2015).

Our distinction is related to that between opportunity and necessity entrepreneurship (Reynolds et al. 2002; Block and Wagner 2010). “Opportunity” student entrepreneurs may decide to leave college without graduating to focus on their venture, while poor employment prospects may drive dropouts into “necessity” student entrepreneurship. Note, however, that Block and Wagner (2010) relate the distinction of opportunity versus necessity entrepreneurship to whether or not the prior employment spell was ended voluntarily. Extending a similar definition to student entrepreneurship would limit the group of necessity student entrepreneurs to those who were expulsed from their university. Fossen and Büttner (2013) alternatively define necessity entrepreneurship as self-employment out of unemployment, which does not capture the situation of student entrepreneurs well.

Individuals are registered as Danes if they are born in Denmark and have at least one parent who is Danish.

Confirming these reservations, in the health care-related programs far more firms are started by graduates than by dropouts, and almost exclusively these firms enter in the health care industry.

Labor market returns of individuals who attended two years of college compared to four years of college have been found to be significantly lower (Kane and Rouse 1995). This resonates with Braguinsky et al.’s (2012) finding that entrepreneurs who do not make use of their education earn less (see above). It is also supported by the results in Table 8.

Around 3 % of the science and engineering graduates enter into entrepreneurship. This is about half of the fraction of start-up owners observed by Åstebro et al. (2012) for their comparable sample of the US science and engineering graduates.

To rule out that these results are driven by the nature of our sample, we re-estimated Model 1 for two more selective samples. First, we excluded all students who are enrolled in a professional bachelor program. Graduates from these programs do not normally enter a Master program, but the respective programs are primarily designed to prepare students for a direct labor market entry. The second alternative sample only includes graduates from programs in science, engineering and business. Results for both alternative samples (available from the authors) are very similar to those reported above.

Following Cameron and Trivedi (2009, ch. 6) we re-estimated Model 2 using a latent-variable model in the first stage to account for the binary character of the Dropout variable. Doing so also yielded a non-significant coefficient for Dropout. The same result was obtained using a bivariate probit model as an alternative for Model 3. Detailed results are available from the authors.

To improve the precision of the estimates, all observations are included in the model but time-invariant variables will be dropped when running the fixed effects model.

Note that the above results only capture income effects in the first 3 years after leaving college. A large number of individuals exit from entrepreneurship, both in our sample and more generally (cf., e.g., Åstebro and Thompson (2011, fn 10) for data on the USA and Canada). This suggests that even if entrepreneurship allowed dropouts to avoid their labor market disadvantage, this would in many cases only delay the punishment.