Absence of conflict of interest.
Citation
Cahill, K. E., James, J. B., & Pitt-Catsouphes, M. (2015). The impact of a randomly assigned time and place management initiative on work and retirement expectations. Work, Aging and Retirement, 1(4), 350-368.
Highlights
- The study examined the impact of a time and place management (TPM) initiative at a medical provider on retirement expectations among workers ages 50 and older.
- The study was a randomized control trial and used a series of surveys and administrative data to estimate the impact.
- The study found that, when surveyed 12 months after randomization, employees in the treatment group were statistically significantly more likely to expect to directly exit their job as opposed to staying in their current job within the next five years.
- The quality of causal evidence presented in this report is moderate because the study was a randomized controlled trial with high attrition, but the authors did demonstrate that the treatment and control groups were similar before the intervention. This means we are somewhat confident that the estimated effects are attributable to TPM, but other factors might also have contributed.
Intervention Examined
Time and Placement Management (TPM)
Features of the Intervention
TPM began in 2012 at ModernMedicine, a regional medical provider. The program involved an online training and self-assessment about the impacts of schedule changes, formal discussions between employees and managers about their work schedules, and encouraging managers to implement schedule changes such as allowing employees to occasionally leave early or work remotely. The program was randomly assigned, in which some work units received TPM and some received no TPM.
Features of the Study
The study used a randomized control trial to examine the impact of TPM on retirement and bridge job expectations. The study analyzed 768 employees and managers who completed at least one survey, were ages 50 and older, worked at least 30 hours per week, and had been at ModernMedicine five or more years. In all, 292 of these people were randomly assigned to receive TPM, 244 were assigned to receive no TPM, and 232 were not randomized because these units could not be split because of ties with managers in other work units. The authors used a statistical model to compare the outcomes of treatment and control group members.
Findings
Employment
- The study found that when surveyed 12 months after randomization, employees in the treatment group were statistically significantly more likely to expect to directly exit their job as opposed to staying in their current job within the next five years. There were no significant differences at other time points and no differences for expecting to move to bridge employment.
Considerations for Interpreting the Findings
The study was based on a randomized controlled trial with high attrition at the final follow-up survey and unknown attrition at the other survey periods. The authors demonstrated that the treatment and control groups were similar in age, gender, race, ethnicity, household income, and education before the intervention. Everyone in the study had been employed at ModernMedicine for at least five years and were full-time (30-plus hours) employees.
Since employees could drop out of the survey if they left ModernMedicine, non-response bias could affect the share of employees who expect to leave within the next five years. Expectations about retirement and bridge jobs might not reflect actual retirement and job transition decisions. Further, the study focused on retirement expectation but the measure appears to be a subset of those who expect to be out of labor force, not total number who expect any bridge job.
Causal Evidence Rating
The quality of causal evidence presented in this report is moderate because the study included a randomized controlled trial with high attrition, but the authors did demonstrate that the treatment and control groups were similar before the intervention. This means we are somewhat confident that the estimated effects are attributable to TPM, but other factors might also have contributed to the findings.