A Longitudinal Study of the Buffering Effect of Volunteering on the Relationship Between Unemployment and Depression for Older Adults

Background/ Purpose:  Research from both cross-sectional and longitudinal studies conducted at both aggregate and individual levels has indicated that unemployment negatively influences individuals’ mental health. Among the long-term unemployed workers in the US, the proportion of those age fifty and over is disproportionally large. Meanwhile, empirical evidence has demonstrated that volunteering is positively associated with older adults’ psychological well-being. Although some previous research has examined volunteering as a path to reemployment, little research has tested the buffering effect of volunteering on the relationship between unemployment and depressive symptoms for older adults. This study aims to fill that gap.

Methods: A longitudinal design using the first difference approach was adopted to examine the effect of change in the interaction of unemployment and volunteering status on change in depressive symptoms (CESD). Two waves (2010, 2012) of panel data from the Health and Retirement Study were used with a sample of 9,421 older adults age 50-85. Control variables included change in self-reported health, change in difficulties in activities of daily living, change in total household income, and change in marital status. Any time invariant variable would be cancelled out using this approach. Altogether 7 models were created to examine the influence of different combinations of unemployment and volunteering status.

Results: Model 1 showed that there was a significant increase in depressive symptoms for those who became unemployed in wave 2 and who were not volunteering in either wave. Models 2-4 showed a nonsignificant change in depressive symptoms when people became unemployed in wave 2, having volunteered in either wave or both waves, indicating there was a buffering effect of volunteering compared to the result from Model 1. Models 5-6 tested the effect of change in volunteer status on depressive symptoms for those who reported being unemployed in both waves. Results showed that simply changing from volunteering to not volunteering or vice versa was not associated with change in depressive symptoms. However, using a dichotomous variable indicating whether a person had volunteered over 100 hours during the past year or not, Model 7 demonstrated that those who didn’t volunteer in wave 1 but volunteered over 100 hours in wave 2 had a significant decrease in depressive symptoms.

Conclusions and Implications: Results from multiple models in this study demonstrated that there was a buffering effect of volunteering on the impact of unemployment on depressive symptoms for unemployed older adults. Although reemployment has been found to be a protective factor for mental health, a recent workplace audit study showed that long-term unemployed applicants experience discrimination in the process of job application and are much less likely to be interviewed compared to a short-term unemployed applicant. When unemployed workers are age 50 or over, they experience another layer of age discrimination. Therefore, alternative interventions for mental health problems of these long-term unemployed older adults are urgently needed. The findings of this study provide insights to inform future intervention development around creating volunteering opportunities for these long-term unemployed older adults in order to enhance their mental health.