Previous work for the Department of Defense (DoD) examined the impact of military service on the subsequent labor market outcomes of military retirees and their families. Major questions that arose out of that research included:
- Do differences in spousal outcomes reflect differences in the experiences of military spouses, or different divorce and marriage patterns of military personnel relative to their civilian counterparts?
- To what extent do differences in labor market outcomes reflect differences in initial and final geographic location, and hence differences in demand-side factors?
- How do labor market outcomes for retirees vary by race and ethnicity?
As a result, the first part of the new work will involve examining the dynamics of marriage for military personnel, the second the impact of geographic location and the third ethnic and racial differences in the earnings and outcomes of military personnel.
Summary of Results from QRMC Research
The Effect of Military Service on the Labor Market Outcomes of Military Families
In this report, we used data from the National Longitudinal Survey of Youth (1979) to examine the impact of being in the military on long-term labor market outcomes and additional investments in education and training made by military personnel and their spouses. We find substantial wage and income differences in labor market outcomeds between military and nonmilitary families. Individuals who serve in the military less than four years typically do worse in the labor market than their civilian counterparts. By contrast, military personnel who are in the military for four or more years typically work more hours and (in the case of those in the military between four and twenty years) more weeks than their civilian counterparts. This result holds, although the size of the effect is reduced, even after taking into account the appropriate comparison group and including labor supply and labor demand variables. Being in the military has a strong positive effect on additional investments in training, and longer-term military personnel tend to be more educated. The impact of longer-term service on asset accumulation relative to civilians is not significant. Spouses of personnel who are with the military less than twenty years are, in general, less likely to be employed, but those who work, work more hours per week than spouses of comparable civilian personnel.
The importance of ensuring that military oucomes are compared to an appropriate group of comparisons is very clear from this study. Simple comparisons of mean differences, without adjusting for initial selection, or for differences in labor demand and supply conditions substantially misstate the effects of military service. The persistently strong and important effects of the AFQT score on outcomes serve to further underscore this difference.
A number of major questions result from this investigation. First, how do the experiences of career military personnel evolve as they move out of the transition from military to civilian service? Do they establish career paths that are similar to or differnent from those of their civilian counterparts? Second, do differences in spousal outcomes reflect differences in the experiences of military spouses, or different divorce marriage, and childbearing patterns of military personnel relative to their civilian counterparts? Future research will examine this possibility by investigating differences in the timing of first marriage, the timing and number of births, and the timing of any subsequent divorce as well as remarriage among military personnel.
A third question is the extent to which differences in labor market outcomes reflect differences in initial and final geographic location, and hence differences in demand-side factors. There is ample evidence that demand-side factors can be extremely important in determining the labor market outcomes of workers. This study simply controlled for these characteristics by including local area unemployment rates, which are not particularly reliable at the county level. Finally, the analysis presented in this study did not differentiate between outcomes for individuals of different races and ethnicities. Future research will examine the impact of race, ethnicity, and locational choice on the labor market outcome of military members.
Mobility and the Military
This study set out to address three questions: whether former military personnel were more or less likely to move than their civilian counterparts; whether former military personnel were more or less likely to move to more economically vibrant local areas than civilians; and whether the impact of mobility on earnings for military personnel was different from that of their civilian counterparts.
The results from this analysis suggested that over a 25-year period, former military personnel are indeed much more likely to be living in a different geographic location than they were in 1979. This result is very large and robust, and much greater for personnel with more than 20 years' service than those with between four and nineteen years. The answer to the second question is that personnel with less than twenty years of service end up in areas with roughly the same level of economic activity as their civilian counterparts; those with more than twenty years are significantly more likely to go to areas with less economic activity. Finally, the impact of mobility on earnings is markedly different between civilians and military. Not surprisingly, given the substantial differences in mobility between the two groups, the marginal civilian who moves earns a greater premium than the former military personnel with less than twenty years service. This is consistent with the idea that the latter have more knowledge about the benefits of moving as a result of substantial networks and first-hand experience as well as lower costs of moving. This fact, combined with the much higher rates of mobility for former military personnel, suggests that they have maximized the economic returns to mobility.
The very interesting result that former personnel with more than twenty years experience actually have an earnings discount as a result of mobility is consistent with the notion that they have not changed locations in search of work. It is entirely possible that their military pension and benefits cause them to locate near military bases to consume base-related leisure activities, or to other locations which offer other types of leisure activities.
The Effect of Military Service on the Demographic Outcomes of Military Families
In this background paper we examined the impact of military service on different sets of demographic outcomes: marriage, divorce, and the timing of first birth. Because we expected different compensation to have an impact on each of these outcomes, we also constructed a predicted value of lifetime earnings and included that as an explanatory factor.
The findings on marriage rates were consistent with earlier literature. Military personnel are substantially more likely to get married than their civilian counterparts, but higher expected compensation delays the timing. Divorce rates are higher, but there is no evidence that compensation affects these rates. And while there is no empirical basis to support the notion that military personnel have children at an earlier age than civilians, higher expected earnings at age 40 act to delay fertility; higher expected earnings at age 25 accelerate the decision to have a child.
Racial Differences in the Earnings and Outcomes of Military Personnel
In this report we examined the impact of being in the military on long-term labor market outcomes and additional investments in education and training made by military personnel by race.
Overall, and consistent with earlier literature, Blacks who are in the military do better in the civilian labor market after they exit the military then do their counterparts who have remained outside the military. Mean employment probabilities are higher, as are earnings, wages, hours and weeks workd. Generally these results hold, although the size of the effect is reduced, even after taking into account the appropriate comparison group and including labor supply and labor demand variables. Being in the military also has a strong positive effect on additional investments such as training: military personnel are significantly more likely than civilians to receive training.
Of course, these results reflect only the initial stages of the transition to the civilian labor market of the oldest members of the NLS97 cohort. Further research will be necessary to determine whether the long-run earnings trajectories of Black military retirees exceed those of Black civilians, and the extent to which any differences reflect differences in industry employment or geographic location.