Beyond Risk Propensity-The Influence of Evaluation Period and Information Relevance on Risk Taking Behavior

INTRODUCTION Entrepreneurs face and respond to risks every day (Sarasvathy et al., 1998). Entrepreneurial researchers have long debated whether entrepreneurs are different than others in the way they perceive and respond to risks. For some scholars, the essence of being an entrepreneur is taking risks and functioning in less structured environments in which uncertainties are welcomed rather than feared (Cantillon, 1755 [1979]; Knight, 1921). Others argue that entrepreneurs are no different than others in the way they perceive and take risks (Brockhaus, 1980; Palich and Bagby, 1995). Still other researchers maintain that entrepreneurs might be more risk averse than the general population (Xu and Ruef, 2004). One reason for such diversity might be the different risk conceptualizations that researchers adopt (Sitkin and Pablo, 1992). Two general approaches to entrepreneurial risk taking can be distinguished from previous literature. In some accounts, entrepreneurial risk taking behaviors (the decision-making behaviors in risky context) are considered relative stable behaviors that stem from innate, dispositional traits such as risk taking propensity (Stewart and Roth, 2001). Accordingly, entrepreneurs and non-entrepreneurs might differ in their general dispositions towards risk and such differences tend to persist across a variety of context and environments. Alternatively, risk behaviors can be conceptualized as the output of a context-specific judgment process. From this perspective, individual's risk behaviors largely result from contextual and environmental factors that influence how they view, frame, and solve the problem at hand (Kahneman and Tversky, 1979; Tversky and Kahneman, 1991). Accordingly, influencing how decision makers view problems should alter their behaviors, making them either more or less likely to engage in risky behaviors. Whereas individuals' actual risk behavior might be a combination of dispositional risk propensity and environmental factors, any study investigating the risk taking behaviors of entrepreneurs vs. non-entrepreneurs has to control for risk taking propensity and simultaneously evaluate the influence of decision making context. A growing body of research suggests that investment time horizon (i.e., the choice of evaluation period) is an important contextual factor that influences risk framing and risk behavior (cf., Behartzi and Thaler, 1999; Gneezy and Potters, 1997; Gneezy et al., 2003; Klos et al., 2005). The longer time horizon decision makers adopt, the longer the period that they choose to evaluate their decision outcome, and the less likely that they will experience loss. Consider an investment option that offers equal chances to win $200 or to lose $100. The probability of losing money after one independent trial is 0.5, but goes down to 0.25 for two independent trials, to 0.18 after 10 independent trials and to 0.0077 for 50 independent trials (Klos et al., 2005). In this way, the risky business of investment becomes less risky after more independent trials over a longer evaluation period. Another stream of research also emphasizes the importance of information relevance in influencing risk behaviors (Rothman and Schwarz, 1998). Relevant information tends to elicit systematic processing of available information whereas irrelevant information fosters heuristic processing of information (Crano and Prislin, 2006). Systematic information processing might reduce risk taking behavior because decision makers have an opportunity to evaluate arguments in favor and against a decision. If decision makers tend to be generally risk averse, they might weigh losses more heavily than gains and thus take less risk. The choice of evaluation period and the relevance of information are of particular importance in entrepreneurs' work environments. Entrepreneurs operate in highly chaotic, complex and unpredictable environments where they make time sensitive decisions despite a lack of relevant and available information (Aldrich and Martinez, 2001). …