پیش بینی شبکه و کار از سن بازنشستگی کارکنان

In a study of the likely causes of retirement decisions, 197 older employees of a state government and their spouses were surveyed as they were getting ready to retire. Data also were collected from the employees' personnel records. Holding finances, gender, and health constant, a set of work characteristics (especially being tired of working) and a set of nonwork characteristics (personal characteristics and expected retirement activities) that were conceptualized as potentially inducing employees to retire each provided a unique contribution of about 10% of the variance to expected retirement age.

The average age of the American population has been steadily increasing in recent years and is projected to continue doing so for some time (Fullerton, 1991). There have been dramatic increases in the number of workers retiring “younger” in recent years (Feldman, 1994), resulting in a decreasing work force participation rate by people over the age of 55 (Sum & Fogg, 1990). Although The data were collected with the funding of a grant from the Andrus Foundation. Further tangible assistance was provided by a sabbatical leave and an associated grant from the Faculty Research and Creative Endeavor Committee of Central Michigan University. The authors appreciate the help of those agencies. We also thank Sonja Faulkner and Papia Ghosh for their assistance on this project. A previous version of this paper was presented at a meeting of the Midwestern Psychological Association. Address correspondence and reprint requests to Terry A. Beehr, Program in Industrial/ Organizational Psychology, Life-Span Development Research Center, 101 Sloan Hall, Central Michigan University, Mt. Pleasant, MI 48859.

This study sought to determine whether work environment characteristics could act to push people out of the workplace and into retirement and whether expectations of retirement activities and situations could entice them to retire at earlier ages. It appears that both work-related and non-work-related factors can influence retirement decisions. This basic idea, that factors in both work and nonwork categories can lead to retirement decisions, is strikingly paralleled in models of another work withdrawal behavior, absenteeism (e.g., Steers and Rhodes, 1978). Furthermore, the combination of these two life domains has a greater influence than either one alone, i.e., they predict unique variance in the ages at which people decide to retireBecause finances have been a consistent predictor of retirement decisions in previous research, wealth was held constant in the hierarchical regression analyses. As expected, wealth did predict retirement age. However, health, which has probably been the second most consistent predictor (though much weaker than finances; Talaga and Beehr, 1989; Taylor and Shore, 1995), and gender did not predict unique variance in expected retirement age in the three models. The entire Model 1 regression analyses showed that a larger number of the nonwork variables (five of eight) predicted retirement age than work-related variables (one of seven). In other words, more elements of life outside the workplace were predictors of the retirement criterion than attitudes and perceptions about the workplace itself. This might suggest that employees contemplating retirement are forward-looking, that is, they are pulled toward retirement more by what they believe awaits in the future after retirement than pushed away from work by thoughts of the current workplace. This conclusion must be tempered, however, by the fact that the effect size, or percentage of variance accounted for by the set of work-related variable(s) in Model 2 of Table 2 was almost as large as the effect size of the set of nonwork variables in Model 3 of Table 2 (.08 versus .09).