Voluntary turnover and alternative job opportunities

Models of turnover specify important roles for both general labor market conditions and perceptions. There is consistent support for the role of general labor market conditions, but evidence on labor market perceptions is mixed. However, no empirical study has included both types of variables. Using a national sample of young adults, both types of measures were found to influence voluntary turnover, either directly or through other factors. However, the two constructs are not necessarily closely linked. For example, despite an intention to quit (based partly on perceived ease of movement), an employee may stay with the organization because general labor market conditions result in a generally low level of alternative job openings. Alternative Opportunities 3 Turnover and Alternative Job Opportunities A recurrent theme in the turnover literature is that the availability of alternative jobs influences turnover intentions and behavior. For example, March and Simon stated that: "Under nearly all conditions the most accurate single predictor of labor turnover is the state of the economy When jobs are plentiful, voluntary movement is high; when jobs are scarce, voluntary turnover is small" (1958, p. 100) . This view is consistent with the economic literature: ".. .when labor markets are tight (jobs are more plentiful relative to job seekers) one would expect the quit rate to be higher than when labor markets are loose (few jobs are available and many are laid off)...One ineasure of tightness is the unemployment rate" (Ehrenberg & Smith, 1982, p. 285). March and Simon (1958), however, further argued that general labor market conditions influenced voluntary turnover through perceived ease of movement, which interacted with perceived desirability of movement to influence turnover. Their model suggests that certain factors (e.g. dissatisfaction) may "push" the employee to look for alternative employment, while other factors (e.g. the perception of attractive alternative job opportunities) may "pull" the employee to consider alternative employment. A subsequent model by Mobley, Griffeth, Hand, and Meglino (1979) similarly hypothesizes that "economic-labor market" factors (e.g. unemployment, vacancy rates) influence turnover indirectly through "labor market perceptions." Perhaps because different types of employees are thought to face different tl~es of labor ~arkets, Alternative Opportunities 4 occupational and personal characteristics (e.g. aptitude, tenure) are also included as determinants of labor market perceptions. Finally, these latter perceptions are specified to interact with desirability of movement perceptions to influence intention to leave which, in turn, has a main effect on turnover. Other models, however, question whether the translation of intention into voluntary turnover behavior is so direct. For example, steers and Mowday (1981) and Michaels and Spector (1982) have argued that an intention to quit is more likely to result in voluntary turnover when labor market conditions are such that alternative jobs are more generally available. As Michaels and Spector suggest: "If a person intends to quit a job, he or she most likely would quit when another job became available" (p. 58). Similarly, Muchinsky and Morrow (1980, p. 276) argue that "when the Economic Opportunity valve is 'open' (alternative means of employment are readily available)", the relation between individual factors and turnover will be stronger than when the valve is closed. Thus, in contrast to Mobley et al. (1979a), these models hypothesize an interaction between general labor market conditions and intention in influencing turnover. Another potential deviation from the Mobley et al. (1979a) model is the possibility that general labor market conditions influence voluntary turnover directly, in addition to their effects through labor market perceptions and intentions. The argument is that most workers "do not quit on the basis of probabilities estimated from alternatives available; they quit on the basis of certainties represented by jobs already offered" (Hulin, Roznowski, & Hachiya, 1985, p. 244). The Alternative Opportunities 5 probability of an alternative offer is linked to general labor market conditions (Hulin et al.). An implication is that although a person may perceive ease of movement to be low, an attractive alternative job offer may nevertheless later arise that results in turnover, consistent with Granovetter's (1974) finding that job offers are often unexpected and unsolicited. On the other hand, because most people do not quit one job without first lining up another (Mattila, 1974), high perceived ease of movement without an alternative offer may fail to result in turnover. Finally, although general labor market conditions influence the probability of receiving an alternative job offer, the "specific mix of skills and experiences of the person in question" are at least equally important (Hulin et al., 1985, p. 239). At the extreme, one could think of a separate labor market existing for each person. Thus, although general labor market conditions should influence perceived ease of movement, the magnitude of the relation is limited to the extent that perceived ease of movement also reflects idiosyncratic differences in individual labor markets that stern from variations in skills, abilities, experience, and so on. Empirical Evidence Given that no study has included measures of both general labor market conditions and labor market perceptions, we do not know the nature of the interplay between the two in determining turnover. As a consequence, the hypotheses concerning labor market conditions described above (i.e. an indirect effect through perceptions, a direct effect, and the interaction with intention) have not been tested. Alternative Opportunities 6 Economic time series research demonstrates that more quits occur under tight labor market conditions (Eagly, 1966; Armknecht & Early, 1972; Parsons, 1977). However, because these results are based on aggregate level data (e.g. annual national turnover), it does not necessarily follow that a comparable relation with labor market conditions exists for individual level turnover data. (See Hammond, 1973 and Roberts, Hulin & Rousseau, 1978 for a discussion of aggregation issues.) Recent empirical work has, however, found support for the idea that general labor market conditions influence individual level turnover as well. For example, a meta-analysis by Carsten and Spector (1987) found that correlations between job satisfaction and turnover tend to be higher when unemployment is lower, suggesting an interaction such that job dissatisfaction is more likely to translate into turnover when the unemployment rate is low. Similarly, in direct studies of individual workers, Youngblood, Baysinger, and Mobley (1985) and Gerhart (1987) have found evidence of an interaction between the unemployment rate and job satisfaction (as well as a main effect for the unemployment rate) in predicting voluntary turnover. In contrast, there has not been consistent support for a relation between labor market perceptions and individual turnover. Bluedorn (1982), for example, concluded that there was evidence for a main effect of perceived ease of movement on voluntary turnover, but a lack of evidence for an interactive effect with job satisfaction. A metaanalysis by steel and Griffeth (1989) reported a correlation of .13 between perceived employment opportunity and turnover, also supportive Alternative Opportunities 7 of a main effect. The review by Hulin et al. (1985), however, questioned even the existence of a main effect. They noted that although zero-order correlations between perceived ease of movement and turnover are sometimes statistically significant, the relation rarely holds up in multivariate models. Like Bluedorn, Hulin et al. also found little support for an interaction between job satisfaction and perceived ease of movement. In summary, evidence suggests a main effect for general labor market conditions in both the economic and psychological literatures. The hypothesized interaction between individual factors (e.g., job satisfaction) and general labor market conditions in the psychological literature has also ~eceived tentative empirical support. However, the evidence on the role of ease of movement perceptions is more ambiguous. Because no study has included measures of both general labor market conditions and perceived ease of movement, it is not clear why the measures yield different results or how they operate vis-a-vis one another in the turnover process. Model and Hypotheses The purpose of the present study is to provide the first test of a voluntary turnover model that incorporates measures of both general labor market conditions and perceived ease of movement. Measures of general ability and experience are also included to help control for individual variations in opportunities. As a point of departure, the solid lines in Figure 1 represent a model consistent with ideas expressed by March and Simon (1958) and Mobley et al. (1979a). Voluntary turnover is a function of job Alternative Opportunities 8 satisfaction (or perceived desirability of movement) and perceived ease of movement (or "labor market perceptions"). In addition to these main effects, an interaction is specified such that low job satisfaction is most likely to be translated into actual movement when perceived ease of movement (e.g. Mobley et al.) or "economic opportunity" (e.g. Muchinsky & Morrow, 1980) is high. Moreover, Mobley et al. specify that the most immediate precursor to turnover is intention to leave/stay, which mediates the effects of other factors on turnover. -------------------------Insert Figure 1 about here -------------------------Of particular interest in the present study are the factors that affect perceived ease of movement. In their discussion, March and Simon (1958) included level of business activi