Schedule - Parallel Session 2 - Applications: Business and Enterpreneurship

Entrepreneurship is considered to be an engine for labour market stabilization, for structural change, and for economic growth (Audretsch & Fritsch 1994). Entrepreneurship is also crucial in providing the competitive market entry forces that prevent excess profits, supporting efficient market outcomes (Audretsch, Keilbach & Lehmann 2006). Therefore, it is crucial to understand the determinants of entrepreneurs’ decisions to enter self-employment in the first place. One factor that has been widely discussed as a crucial determinant of the self-employment decision is the individual’s attitude towards risk (Kihlstrom and Laffont 1979; Bellante and Link 1981; Barsky et al. 1997; Cramer et al. 2002; Fairlie 2002; Lazear 2005; Caliendo, Fossen & Kritikos 2014; Skriabikova et al. 2014). Since studies comparing the risk attitudes reported by entrepreneurs to those reported by other individuals generally report higher values for the entrepreneurs, the usual conjecture is that a positive attitude towards risk is a prerequisite for self-employment. The empirical evidence that we provide in this study, however, suggests that the usual unidirectional conjecture is not rich enough to describe observed data. Our analysis provides evidence for a bidirectional causality where, on the one hand, individuals’ attitudes towards risk affect their self-employment decisions and, on the other hand, experiencing self-employment affects the individuals’ risk attitudes. Using a large general population panel, we show that individuals with higher reported risk attitudes are more likely to enter self-employment. Additionally, we find that these individuals’ reported risks attitudes are even higher after experiencing self-employment. Hence, we can document a substantial positive effect of self-employment on the willingness to take risks that goes significantly beyond the self-selection bias of the risk-seekers into self-employment. We confirm the robustness of our results with various model specifications and show that all our central findings hold throughout. Our study adds to the mounting evidence that attitudes towards risk may be affected by environmental factors or individual experiences (Bowles 1998; Heaton & Lucas 2000; Guiso & Paiella 2008; Malmendier & Nagel 2011, Loewenstein, Weber, & Hsee 2001). The central contribution of our study is to show that the self-employment experience is amongst those environmental factors that have a sizeable and positive impact on the development of individuals’ risk-taking attitudes.

Empirical research on entrepreneurship has identified two types of strategies to reduce uncertainty, namely, prediction- and control-based strategies (Sarasvathy, 2001). Prediction-based strategies focus on estimating unknowns via sampling methods and control-based strategies focus on shaping unknowns via pro-active behavior. Both prediction- and control-based approaches aim to reduce uncertainty by providing information to refine prior beliefs. Yet they presuppose different cognitive models of the situation at hand and different degrees of involvement by the decision maker. With regard to cognition, predictive strategies may be interpreted as producing reliable information about current market trends, whereas control-based strategies may be seen as first hand evidence of the chances of transforming customers’ preferences. As far as involvement is concerned, predictive strategies are passive in nature and their outcomes are relatively independent of the behavior of the entrepreneur. Control-based strategies in contrast, presuppose an active involvement of the entrepreneur and yield results that heavily depend on her efforts. By eliciting distinctive feelings of confidence, the information provided by these strategies may affect the willingness to engage in entrepreneurial action. Experimental evidence on betting behavior provides support for this hypothesis. Based on Ellsberg’s urn as a model of uncertainty, Kuechle et al. (forthcoming) model prediction as random sampling from the urn and control as inserting marbles of an exogenously given color. Using a between subjects design that randomly assigns individuals to the treatments, they find that control-based methods to reduce uncertainty lead to a higher proportion of betting behavior after a favorable outcome compared to predictive methods, results that revert in the presence of unfavorable outcomes. This experiment does not measure any preference for the method to acquire information. It is possible that the results are based on decisions of individuals who intrinsically like and dislike the method they were assigned to. If that is the case, these results would underestimate the difference between prediction- and control-based strategies. The goal of our research is to extend this experiment by allowing subjects to decide how much of their endowment they are willing to invest to acquire information before considering whether to bet. Although subjects will be randomly assigned to the two treatments, their willingness to pay will provide a measure of their intrinsic preferences for the procedures to reduce uncertainty. If such preferences exist, namely if the amount that the subjects are willing to pay on average in each treatment is significantly different, we should expect stronger differences in betting behavior between the prediction and the control treatments after receiving information.