Author: Bram Timmermans

Click here for the first draft of the paper which will be presented during the DRUID winter conference 2009. The presentation can be view here.

A paper I recently started on is one that looks at the effect of diversity in new start-ups and the effect this diverse composition has on their survival. Researchers in the field of entrepreneurship have recognized that entrepreneurial activity is more a collective than an individual activity. As a result, studies on entrepreneurial teams started to emerge, including those on diversity in entrepreneurial teams. However, these studies look at team diversity cross-sectionally thereby neglecting the dynamic characteristic of teams, i.e. the potential change of team composition, which leads to a possible change in diversity. Whether or not this results in an increase or decrease of diversity will depend on the personal characteristics of the members that leave or enter the venture. This study will adopt a dynamic approach looking at the degree or diversity in the start of these ventures, identify if and how this diversity changes over time and say something about the consequence of this diversity on firm performance. This diversity will be measured not only based on the founding team but including another important source of human capital, i.e. the employees.

The literature is rather ambiguous on the effect of team diversity on team performance. New ventures face complex and non-routine problems, which are better solved by heterogeneous teams arguing for the positive effect of diversity. On the other hand, diverse teams increase the likelihood of conflict. The problem-solving dimension has a more cognitive nature while conflict occurs in demographic diverse groups. I argue in this paper that diversity in achieved characteristics increases the likelihood of firm survival while diversity in ascribed characteristics is negative for firm survival.
The empirical analysis of this paper will build on the Danish Dataset for Labor Market Research (IDA). IDA is suitable for this analysis as its longitudinal character shows firm dynamics and the employment history of all individuals residing in Denmark from 1980 and onwards. In addition, personal and firm characteristics can be identified. Given this structure it is possible to identify the composition of new ventures and the change in this composition in the following years

From this dataset I will identify all new start-ups in the private sector in the period 1995-2000, including the ascribed and achieved characteristics of the individuals connected to the new venture. An entropy measure will calculate the diversity based on these characteristics. These values will be used to (1) identify the composition of new ventures at founding by employing a structural event analysis, and (2) test the effect of diversity on firm performance. The last analysis will identify the effect of a diverse composition in the first year. This analysis will then be extended to a more dynamic approach looking at the change in diversity due to the hiring of new and firing/resigning of existing members.

Some tentative analyses have indicated that the initial composition of new ventures in Denmark tends to be homogeneous. Results on the effect of diversity, both on the ascribed and achieved characteristics, on firm survival show the same ambiguity that can be observed in the existing diversity literature. Further fine-tuning of the diversity measure might solve this problem.