Entrepreneurial Exit

Karl Wennberg, Stockholm School of Economics
Visiting researcher ('07-'08), SCANCOR, Stanford University

karl.wennberg@hhs.se

Entrepreneurial exit has often been conceptualized as 'failure' as a fundamental performance measure of new organizations. Yet, recent studies show clear indications that exit from entrepreneurship is theoretically different from failure (Bates, 2005; Headd, 2003). Except for a study by Gimeno and colleagues (1997), there is however a lack of theoretical models to explain how and under what conditions performance is distinct from exit.

In this research I address the literature on exit in entrepreneurship and organization theory by drawing on the aspiration-level performance literature (Greve, 2003). I use arguments of risk avoidance and aspiration-level performance to hypothesize how entrepreneurs' social comparison affects their choice to exit or persist in entrepreneurship. Matched employee-employer databases are used to identify entrepreneurs' social comparison groups based on their similar others - individuals with comparable social background and education, and how reference to these groups form their aspiration levels.

Strategy research have tended to approximate firms' social comparison based on salience and ease of observation (Baum et al., 2005) whereas the organizational behavior research have tended to use more empirically salient measures of social comparison groups (Goodman et al., 2007). In this research I use the theories of socialization that suggests that peer groups are formed during late adolescence and early adulthood, often in a school setting (Ryan, 2000). I propose that the performance of individual entrepreneurs with comparable education (e.g. high school graduates, mechanical engineers; sociologists, or business graduates from the same school and cohort of study) should be used as the social aspiration level.

To test my theoretical model, I use detailed matched employee-employer databases from Sweden that includes all residents and all active firms between 1989 and 2002. I take advantage of the detailed information on schooling in the individual-level database and information on economic activities and profitability in the firm-level database. My sample consists of three cohorts of entrepreneurs starting firms in the knowledge-intensive sectors (N= 10,320), i.e. high-tech manufacturing and skilled services. I follow these individuals until they leave entrepreneurship (i.e. close their firm and exit from self-employment), or until 2002. The hypotheses are tested by the estimation of censored regression models.