skip to main content
Article Podcast Report Summary Quick Look Video Newsfeed triangle plus sign dropdown arrow Case Study All Search Facebook LinkedIn YouTube
Thomas M. GeraghtyGerald E. CoxJared M. HuffRachel TownsleyLauren MaloneJacklyn Kambic
Download full report

The Department of Defense (DOD) is considering moving to a single salary system (SSS) that would eliminate the basic allowances for housing (BAH) and subsistence (BAS) and increase basic pay to compensate servicemembers. The 13th Quadrennial Review of Military Compensation (QRMC) is studying this issue, and the QRMC’s director asked CNA in 2018 to identify and prioritize potential second- and third-order effects of moving to an SSS. This report considers one of the potential effects identified in that work: changes in servicemembers’ retention driven by changes in their marriage behavior. It analyzes the effects that a move to an SSS is likely to have on the percentage of servicemembers who are married, and it examines the changes in retention rates and force size that may be induced by any changes in marriage behavior. Overall, we find that these effects are likely to be small.

Our approach includes a review of the literature on the relationships between compensation, marital status, and retention; computation of pay changes under different SSS implementation scenarios; estimation of the effect of marital status on retention using personnel data; and development of a model that can forecast marriage rates and force size over time.

We consider three SSS implementation scenarios: (1) a “full compensation” scenario in which basic pay is increased to fully offset the loss of BAH, BAS, and the associated tax advantage (to members without dependents); (2) a “partial compensation” scenario in which the increase in basic pay is reduced so as to maintain cost neutrality to the federal government; and (3) a “partial compensation with housing rents” scenario, in which servicemembers living in military-provided housing are assessed rents to counteract the large pay increases going to servicemembers not currently receiving BAH under the first two SSS implementation scenarios. Under the full and partial compensation scenarios, non-BAH recipients receive regular military compensation (RMC) increases, while married BAH recipients see their pay reduced, with the largest pay reductions for married junior enlisted. Scenario 3, partial compensation with housing rents, eliminates much of this differential treatment between servicemembers who do and do not receive BAH. In general, among BAH recipients, married servicemembers receive somewhat larger pay reductions than single members, although the difference tends to be small.

Our literature review shows that servicemembers in the current environment, both enlisted and officers, are more likely to marry, and tend to marry earlier, than comparable civilians. With respect to the relationship between compensation and marriage, the literature supports a “marriage bar” hypothesis, in which higher levels of income are linked to higher marriage rates, but only for incomes up to a certain level (usually defined as a local community median level of income). The literature also provides evidence that marriage positively affects military retention, with the strongest such effects for men who are early in their military careers.

Our statistical analysis of Defense Manpower Data Center data confirms the literature’s findings on marriage and military retention, with the largest positive effects for male enlisted, somewhat smaller effects for male officers, and no effect for most female enlisted (with the exception of those in the Army). For female officers, our analysis finds that being married has a negative effect on retention, which is consistent with findings in previous CNA studies.

Our force inventory modeling analysis, however, suggests that SSS implementation is likely to have only small effects on the percentage of the force that is married and on retention and force size. Reasons include the following:

  • The nature of pay changes under an SSS means that some servicemembers may receive pay increases that offset the effect of pay reductions received by others.
  • The effects of compensation on marriage behavior, and of marriage behavior on retention, do not affect all servicemembers equally strongly (male, junior enlisted are the most affected).
  • The effects of compensation on marriage behavior and of marriage behavior on retention, when combined, result in a smaller overall effect on retention than might be anticipated by considering the magnitude of either of the individual effects in isolation.

We note, however, that our work is not a comprehensive analysis of the potential effects of an SSS on military retention. We focus here on retention effects induced by changes in marriage behavior. There may be additional important retention effects that are beyond the scope of this analysis. Overall, however, our study suggests that the effects of an SSS on military marriage rates, and the effect of changes in marriage behavior on military retention, are likely to be small. Therefore, there is little need for policy-makers to be concerned about these effects when considering a change to an SSS.

Download full report

DISTRIBUTION STATEMENT A. Approved for public release: distribution unlimited.

Specific Authority. To protect information not specifically included in the above reasons and discussions but which requires protection in accordance with valid documented authority such as Executive Orders, classification guidelines, DoD or DoD-component regulatory documents. 4/27/2020


  • Pages: 74
  • Document Number: DRM-2020-U-026047-Final
  • Publication Date: 4/27/2020
Back to Marine Corps and Defense Workforce Program