Absenteeism and Performance as Predictors of Voluntary Turnover

Abstract

Turnover is one of the more widely studied organizational behavior phenomena. Both the consequences and antecedents of turnover have received considerable attention in the literature. For example, Dalton, Krackhardt and Porter (1981) and Johnston and Futrell (1989) are among the many who have examined the positive and negative consequences of turnover. Cotton and Tuttle's (1986) meta-analysis represents a good summary of the literature on the antecedents of turnover. In addition, a variety of models of turnover have been proposed, including those of Mobley, Griffeth, Hand and Meglino (1979), Muchinsky and Morrow (1980), Porter and Steers (1973), Price (1977), Steers and Mowday (1981), and Lee and Mitchell (1994). Despite the large body of literature regarding turnover, relatively little research exists on the relationship of turnover to other behavioral outcomes, most notably absenteeism. Recent meta-analyses have provided evaluative overviews of the relationships between job performance and turnover (Bycio et al., 1990; Williams and Livingstone, 1994) and between job performance and absenteeism (Bycio, 1992). However, few empirical studies on the relationship between turnover and absenteeism have been conducted. From a manager's viewpoint, several issues are fundamental to understanding employee turnover. First, managers should be aware of what their organization's (or subunit's) turnover rate is. They should also have some basis for comparison in order to give this figure some meaning. For example, a twenty percent turnover rate can be viewed either positively or negatively depending upon its comparison to industry norms, previous years' rates, ease of employee replacement, etc. Cost issues are also important. Nonexempt employee turnover costs may be minimal while exempt employee turnover expenses can be substantial as these expenditures entail separation costs (e.g., exit interviews), replacement costs (e.g., testing, interviewing, background checks), and training costs (Cascio, 1991). Second, managers need to separate voluntary from involuntary turnover. Many companies today are experiencing extremely high rates of turnover, but it is involuntary turnover brought on by downsizing, restructuring and reengineering. Involuntary turnover, does not represent a choice on the part of the individual. It is a management decision based on strategy, a result of poor employee performance, or perhaps a decision forced upon the individual by society (e.g., mandatory retirement) or by circumstance (e.g., ill health). Managers need to gain a better understanding of voluntary turnover or turnover that is entirely within the discretion of the individual employee if they are to truly understand the magnitude and nature of turnover rates. A third fundamental issue that managers need to understand is who is leaving. Is the organization losing its high performers or low performers, experienced employees or recent hires? Finally, a question of interest to managers and academics alike is whether there is any connection between absenteeism and turnover. Can absenteeism be used as an early warning system for predicting who is likely to voluntarily leave the organization? This paper seeks to examine the last two of these fundamental issues by examining whether a relationship exists among absenteeism, as measured by the use of sick leave, performance, and voluntary turnover within the context of a state agency. A few authors have theorized on the relationship between turnover and absenteeism. Mobley (1982), for example, argues that there are some conditions under which one would not expect a relationship between the two, other conditions where one would expect to see a relationship, and still others where a relationship would be expected but not observed due to constraining factors. An example of the first case is the situation in which turnover is attributable to a positive pull factor (e.g., an attractive job offer) rather than dissatisfaction with the current situation. …