Withdrawal Behavior: Turnover

In its simplest form, turnover refers to whether an employee stays or leaves. Refinements in the measurement and definition of turnover have led researchers to consider the voluntariness, avoidability, and functionality of turnover. Voluntary turnover refers to situations in which employees have an opportunity to remain with their employer but choose to leave. Involuntary turnover refers to situations where employees do not have a choice concerning continued employment, such as employer-initiated termination. Turnover has been conceptualized as avoidable when the reasons for leaving are related to the organization, including low pay and long hours, and unavoidable when reasons are not work related, such as spouse relocation and family demands. Finally, turnover that benefits the organization, such as when high performers stay and low performers leave, has been termed functional turnover, whereas the opposite pattern is thought to reflect dysfunctional turnover. Although these three dimensions are defined by their extremes, they are perhaps better captured along a continuum in practice.

Key Models and Critical Antecedents of Turnover

Content Models

Content models attempt to specify antecedent variables that help predict employee turnover. At least 50 different predictors have been examined in previous research including demographic variables, work attitudes, intentions and cognitions related to quitting, and job search activities. Meta-analysis procedures have been used to summarize the strength of the relationships that have been found in previous research between these antecedents and turnover behaviors. In general, these results reveal modest effects. In absolute terms, few relationships exceed .30, and most are less than .20.

As a class, demographic variables such as education, marital status, and age are only weakly related to turnover. For other demographic variables such as race, gender, and cognitive ability the average correlation is near 0. There are two notable exceptions to this general pattern of findings. Turnover tends to be greater for employees with children and employees with longer tenure, but these effects are fairly small.

Work attitudes have a long history of research in the turnover literature. In particular, low to moderate negative relationships have been found between job satisfaction and turnover, and between organizational commitment and turnover. Weaker associations have been found for specific facets of satisfaction, such as satisfaction with a supervisor, coworkers, pay, promotion opportunities, and work content. Perceived organizational support (POS), which involves employee beliefs about how much their organization values and supports them, has been found to correlate negatively with turnover intentions and actual turnover. Research on organizational justice supports the contention that perceived unfairness is associated with greater levels of employee turnover. In particular, justice dimensions (i.e., procedural, distributive, interactional) have shown negative relationships with intentions to quit and turnover, suggesting that employees who feel they are treated fairly in terms of outcome allocations, procedures, and interpersonal treatment are more likely to remain with the organization.

Perhaps the strongest predictor of turnover is the employee’s reported intentions to quit. Similarly, withdrawal cognitions and thinking of quitting are positively related to actual turnover, as are job search intentions and behaviors. As a class, cognitions and intentions related to quitting along with job search activities represent the best predictors of turnover to date. Although also positively related to turnover, perceptions of available alternatives and comparisons of these alternatives with one’s present job tend to show somewhat weaker associations.

Other variables that have been studied in relation to turnover include actual pay, stress, lateness, absenteeism, and job performance. Weak, positive effects have been found for stress-related variables, lateness, and absenteeism. Weak, negative effects have been found for pay and job performance. Overall, content models of turnover help identify what variables predict quitting behavior on the part of employees. Process models, which are reviewed next, conceptually organize these variables to help understand how employees make these decisions.

Process Models

One of the earliest process models emphasized the importance of two factors in explaining turnover behavior: the perceived desirability of movement and the perceived ease of movement. Put forth in the late 1950s, the basic idea of the model is that employees will quit when they are unhappy with their jobs and when they feel that there are alternative employment opportunities available. Thus, desirability of movement has been assessed through measures of job satisfaction, whereas ease of movement has been captured through some measure of alternative job opportunities. Empirical tests of this two-factor approach have revealed fairly weak relationships with turnover. As a result, many of the later models have expanded the network of antecedent variables and have offered new perspectives on the causal sequencing of these dimensions. These approaches have been called intermediate linkage models and reflect the growing complexity of turnover models that have dominated the literature since the 1970s. The logic of this perspective is that job dissatisfaction prompts employees to think about quitting and evaluate the costs and benefits of searching for another job. At that point employees may develop intentions to search for another job, actually carry out the search, and evaluate the alternatives that are found. If these alternatives are more favorable than the current job, employees will develop intentions to quit, followed by actual turnover. Variations on this basic model have appeared in the literature, and as a class, the intermediate linkage models have received modest empirical support. Meta-analysis data revealed the strongest support for a reduced linkage model that included only the linkages between dissatisfaction, withdrawal cognitions, and turnover.

Given the generally low predictive power of earlier models, new approaches to the study of turnover processes have emerged since the 1990s. The unfolding model of employee turnover represents one of these recent advancements and draws from basic research in decision making and social psychology. The approach suggests that multiple decision paths underlie turnover, and some of these do not involve much thought or planning on the part of the employee. The model introduces the notion of shocks to the system, or a particular jarring event that initiates deliberate judgments by employees concerning their employment relationship. Examples include receiving unsolicited job offers, experiencing a merger, and spouse relocation. These shocks are said to prompt different script-driven actions, which may or may not lead to a search for alternatives, lower job satisfaction, and actual quitting. The model is more complex than previous approaches and emphasizes the timing and sequencing of events to explain when, why, and how employees leave. In particular, four paths are specified by the model to explain the different ways in which employees leave their jobs. The first path is characterized by a shock to the system in combination with a plan for leaving. For example, employees may accept a position with full knowledge that they will quit once some period of time passes or they earn a specific amount of money. Once the shock occurs (e.g., three months have passed), the employee quits. It is important to note that traditional antecedents of turnover such as dissatisfaction and search for alternatives do not enter the process. The second path also involves a shock to the system but does not include a specific plan or script concerning alternative jobs. For example, some employees may quit spontaneously after receiving a harsh review, learning of an ill family member, or being given new, unwanted job responsibilities. The third path in the model is more gradual in that the shock leads to minor dissatisfaction and a search for alternatives. As an example, employees who receive unsolicited job offers (i.e., a shock to the system) may begin to contemplate their satisfaction relative to the new alternatives, and eventually quit after thoughtful deliberation. The fourth and final path is characterized by dissatisfaction compounded over time, which may or may not be coupled with an active search for alternatives. For example, employees may decide that after several years of work, the job falls short of their expectations. In response, some employees will quit without an alternative in hand, whereas others will leave on finding an alternative that is more attractive. Note that this final path is most closely associated with the tradi-tional approach of the intermediate linkages model that dissatisfaction leads to intentions to leave and actually leaving. There is some empirical support for the paths outlined in the unfolding model, although additional research is certainly needed to refine the model and replicate these effects.

It is also important to note the existence of a wide range of economic models that help explain turnover. These approaches emphasize labor market conditions such as the availability of alternatives or the supply and demand of labor. Because these models tend to focus on factors unrelated to the individual, they are not reviewed here.

New Directions

Over time, approaches to the study of turnover have become more complex. The unfolding model of turnover is one approach that pushes turnover theory and research in new directions. Research has also shown that simple correlational designs can oversimplify relationships between antecedent variables and turnover. Studies that account for interactions among antecedents reveal that simple additive relationships may be insufficient for explaining turnover behavior. In addition, researchers have found support for nonlinear effects in some instances.

Another avenue of inquiry that holds promise concerns the study of those factors that promote retention of employees. Although on the surface one might expect that the same set of factors would be responsible for both turnover and retention, researchers have shown that there may be subtle but important differences in the processes underlying these phenomena. Researchers have proposed the construct of job embeddedness to capture the broad set of factors that cause individuals to remain with their employer. The concept builds on earlier turnover models that included both work and nonwork factors to explain why employees quit. The components of job embeddedness include the following:

  • Links to others within and outside the organization
  • Employees’ perceptions of fit with their work and community
  • Assessments of what employees would have to sacrifice if they left their jobs

Empirical evidence shows that job embeddedness explains incremental variance in both turnover intentions and actual turnover beyond work attitudes and job alternatives.

Finally, there is recent research that synthesizes research on both turnover and attachment processes and proposes a set of eight motivational forces that can help explain intentions and decisions related to quitting or staying with an organization. These forces and a description of each are in the following list:

  1. Affective: Reflects the emotional responses toward the organization; emotional comfort motivates attachment, while discomfort motivates quitting.
  2. Calculative: Reflects a rational assessment of the probability of attaining goals within the organization in the future; favorable assessments motivate attachment, whereas unfavorable assessments motivate quitting.
  3. Contractual: Reflects a response to norms of reciprocity based on psychological contracts; fulfillment of obligations motivates staying, whereas perceived violations motivate quitting.
  4. Behavioral: Reflects the actual and perceived costs of quitting; higher costs motivate staying, whereas lower costs motivate quitting.
  5. Alternative: Reflects self-efficacy toward finding an alternative opportunity; low self-efficacy motivates staying, whereas high self-efficacy motivates quitting.
  6. Normative: Reflects motivation to comply with perceived expectations of others concerning turnover decisions; high motivation to comply with expectations to remain motivates staying, whereas high motivation to comply with expectations to leave motivates quitting.
  7. Moral and ethical: Reflects a desire to maintain consistency between values and beliefs about turnover; values that emphasize persistence motivate staying, whereas values that emphasize change or variety motivate quitting.
  8. Constituent: Reflects the degree to which employees feel attachment toward others within the organization; high attachment to constituents motivates staying, while low attachment motivates quitting.

Overall, the model of motivational forces provides a parsimonious taxonomy of the array of factors that lead employees to quit. Further, they also fill a gap in the literature by specifying the conceptual processes that link antecedent variables such as work attitudes with turnover and retention outcomes. Like the unfolding model, the motivational forces also offer an opportunity to account for changing relationships among variables over time. Empirical research that tests the value of these proposed advantages is needed.

Practical Recommendations

Given the costs often associated with turnover, a number of suggestions for reducing and managing turnover have been suggested in the academic and practitioner literatures. Most follow from empirical research, but many of the techniques themselves have yet to be subjected to rigorous empirical evaluation.

  • Identify the nature and extent of turnover: Find out where turnover is greatest and identify the extent to which it is functional versus dysfunctional; ensure that retention efforts are targeting the right people.
  • Enrich the job: Look for ways to improve the job itself; consider ways to restructure the job to reduce monotony and improve employee motivation.
  • Train supervisors: Many employees leave supervisors, not the job; provide supervisors with continuous training and updating in management skills; use feedback from direct reports, peers, supervisors, and other constituents to help supervisors gain perspective.
  • Promote fairness in the workplace: Communicate difficult decisions with respect and concern for those affected; allow employees to participate in decisions that will affect their jobs; enforce rules and procedures consistently across employees.
  • Monitoremployee attitudes: Implement regular employee opinion surveys and act on the findings; consider including measures of job satisfaction, organizational commitment, and organizational fairness.
  • Provide incentives to stay: Tie rewards to tenure; consider awarding perks that enhance embeddedness in the organization and community. Offer realistic previews of the job: Employees may quit because of a mismatch between their expectations and the reality of what the job involves; provide a reasonable perspective on both the positive and negative aspects of the job.
  • Attend to nonworkneeds: Consider employees’ needs outside the work domain; some companies have responded by providing services and programs such as flexible work arrangements, on-site child care, fitness facilities, tuition reimbursement, concierge services, and eldercare benefits.


  1. Griffeth, R. W., Hom, P. W., & Gaertner, S. (2000). A meta-analysis of antecedents and correlates of employee turnover: Update, moderator tests, and research implications for the next millennium. Journal of Management, 26, 463-488.
  2. Hom, P. W., Caranikas-Walker, F., Prussia, G. E., & Griffeth, R. W. (1992). A meta-analytical structural equations analysis of a model of employee turnover. Journal of Applied Psychology, 78, 890-909.
  3. Lee, T. W., & Mitchell, T. R. (1994). An alternative approach: The unfolding model of voluntary employee turnover. Academy of Management Review, 19, 51-89.
  4. Maertz, C. P., Jr., & Campion, M. A. (1998). 25 years of voluntary turnover research: A review and critique. International Review of Industrial and Organizational Psychology, 13, 49-81.
  5. Maertz, C. P., Jr., & Griffeth, R. W. (2004). Eight motivational forces and voluntary turnover: A theoretical synthesis with implications for research. Journal of Management, 30, 667-683.
  6. Mitchell, T. R., Holtom, B. C., Lee, T. W., Sablynski, C. J., & Erez, M. (2001). Why people stay: Using job embeddedness to predict voluntary turnover. Academy of Management Journal, 44, 1102-1121.
  7. Mobley, W. H. (1977). Intermediate linkages in the relationship between job satisfaction and employee turnover. Journal of Applied Psychology, 62, 237-240.
  8. Trevor, C. (2001). Interactions among actual ease-of-movement determinants and job satisfaction in the prediction of voluntary turnover. Academy of Management Journal, 44, 621-638.

See also: