The earliest cited definition of entrepreneur is generally Cantillion who wrote circa 1700 (Carland, Hoy, Boulton & Carland, 1984). Among the characteristics which Cantillion cited was risk bearing (Kilby, 1971). Mill (1848), who was credited with bringing the term 'entrepreneur' into general use (Schumpeter, 1934), also believed that the key difference between entrepreneurs and managers was risk bearing. Risk bearing or risk taking behavior continued to be a major aspect of entrepreneurship until modern times (Carland, et. al., 1984). Brockhaus (1980) cast doubt on the traditional perspective with an empirical study which showed no difference between the risk taking characteristics of entrepreneurs, managers and the general public. The issue is far from settled. There have been a number of empirical studies of risk taking and the results are frequently contradictory. Are differences in findings related to the instruments used? This paper will present an empirical examination of risk taking propensity and test two popular measures of risk taking to identify sources of conflict in the literature.
CONFLICTING RESULTS IN RISK TAKING PROPENSITY
There have been a number of empirical studies of risk taking behavior; however, these studies have not produced uniform findings. In his recent book, Stewart (1996) examined risk taking studies. The following exhibit is drawn from his review with additions of newer studies appearing in 1995. The table shows the studies, the number of respondents, the instruments employed to measure risk, and the major findings of the research in each study.
As the exhibit shows, the empirical studies employed a wide range of risk taking instruments, however, the JPI and the CDQ were the most frequent instruments in use. The literature review shows more support for higher propensity in risk taking by entrepreneurs. Despite this history, the findings by Brockhaus (1980) are the most widely cited. This is primarily a result of the stature of the journal in which that work appeared. In addition, in those studies in which no differences in risk taking were identified, the CDQ was most often the instrument of choice for risk measurement.
THE BROCKHAUS STUDY
In a frequently cited Academy of Management Journal article, Brockhaus (1980) examined the risk taking propensity of entrepreneurs employing the Wallach and Kogan (1959, 1961) Choice Dilemma Questionnaire (CDQ). He defined an entrepreneur as one who was a major owner and manager of a business venture who was not employed elsewhere. He examined three groups of individuals: business owners who had initiated their business ventures within three months prior to the study (N=31); managers who had changed positions in their companies within three months prior to the study (N=31); and, managers who had changed employers within three months prior to the study (N=31). Finally, he compared the scores of the participants to the normative data reported by Kogan and Wallach (1964).
Brockhaus (1980) reasoned that using entrepreneurs near the beginning point of their ventures would include those that would ultimately fail, therefore eliminating bias from examining only successful entrepreneurs. Pairing these entrepreneurs with managers who had also undergone recent changes in their professions ensured that more stable individuals who might be less risk taking were not compared to the entrepreneurs.
Employing the Analysis of Variance procedure, Brockhaus (1980) found no statistical differences in risk taking propensity among the three groups. Using a chi-square test with a .25 confidence level, he found no difference between the respondents' scores and the normative Kogan-Wallach data. Brockhaus concluded that the distribution of risk taking propensities of entrepreneurs is the same as the general population.
THE MASTERS AND MEIER STUDY
Masters and Meier (1988) using the Kogan-Wallach Choice Dilemma Questionnaire replicated the Brockhaus study. …