Sex Differences in Worker Quitting
Viscusi, W. Kip
Although women quit more both overall and within major occupational groups than do men, this observation is not particularly informative due to the substantial heterogeneity of worker characteristics and job characteristics. Analysis of a sample of almost 6,000 male and female workers suggests that sex differences in quitting have been overdrawn in many previous discussions. Female quit behavior differs from that of males by more than the addition of a sex-specific intercept term. For example, women are more likely to quit work in hazardous industries due to the likely greater uncertainty regarding their appropriateness to such jobs. Unlike their male counterparts, better educated women are more likely to quit their jobs, perhaps because of the greater uncertainties associated with jobs traditionally held by men. Conventional notions regarding female quitting are reflected by the lower stabilizing effect of age on their quit rates. Certainly the most important single difference is that female employees are more likely to have no more than a year of experience and within this low experience category they display greater quit rates. The source of the TENURE1 difference is not clear since it reflects specific human capital investments, learning about job characteristics that alters the position's attractiveness, as well as periodic labor force attachments other than those reflected through work on a part-time basis (since inclusion of this influence did not substantially affect the results). After the initial year of work, male and female quit rates are roughly identical. Almost the entire predicted male-female quit difference and half of the actual difference can be explained by differences in their jobs and regional economic conditions. If women had the same job characteristics and the same percentage with more than one year of experience at the firm, their predicted quit rate would be below that for men and their mean quit rate for the sample would be equal to that of men after adjusting for these influences. Indeed, women display greater stability than they would if characterized by the coefficients in the male quit equation. Coupled with the almost identical response of each group's quit rates to additional wage payments, these findings suggest that the overall quit rates resulting from somewhat different behavior leads to turnover rates more similar than earlier studies have suggested. As a consequence, sex differences in wages and unemployment should not be so readily attributed to turnover-related differences in the behavior of male and female employees.