The Relationship Between Colocation and Reenlistment in the Marine Corps

In this annotated briefing, we examine the relationship between colocation and reenlistment in the Marine Corps. This analysis is part of a larger CNA project titled “The Effects of Personnel Policy Changes on Budgets and Manpower Inventories” sponsored by the Office of the Assistant Secretary of the Navy (Financial Management and Comptroller) (ASN(FM&C)).

In the larger project, we aim to identify and explore ways to reduce personnel costs while maintaining or even improving retention. The impetus for the project is that military personnel costs, including those in the Department of the Navy (DON), are quite large and are always under scrutiny. DON must continuously strive to make the personnel system more effective and efficient.

Additionally, DON has pushed to increase the female share of accessions and inventory. This has implications for retention and manning. Historically, female servicemembers in DON have not retained as well as men. In addition, pregnancy and operational deferment, along with higher female limited duty rates, can make manning more challenging.

The research agenda for this project revolves around four issues related to the cost of DON uniformed personnel. (These four main issues are briefly described in the next slide.) In particular, the sponsor is interested in how costs may change as the female share of the uniformed personnel in DON increases; three of the four issues directly relate to the female share of the active component inventory.

These are the four main issues to be addressed in the larger project. 

In this annotated briefing, we address issue number three. Colocation in the USMC is when two servicemembers who are married to each other (also known as a dual-military marriage) are assigned to units no further than 50 miles apart. A dual military couple is not colocated if the units to which they are assigned are greater than 50 miles apart.

Our analysis is motivated by the possibility that, as the share of women in the services increases, there may be an increase in dual-military marriages, as well as an increase in the demand for colocation. Therefore, it is increasingly important to examine the third question above—that is, to know whether the decision of servicemembers in dual-military marriages to reenlist may be affected by colocation. In this annotated briefing, we specifically examine the following aspects of this question:

• How prevalent is colocation among enlisted Marine-Marine couples?
• How does colocation of enlisted Marine-Marine couples affect reenlistment decisions?

The remaining three issues in the larger study are addressed in other research documents. Our approach is to use Marine Corps personnel records to identify Marines who are married to other service personnel, including other Marines. We focus on enlisted Marines because the Marine Corps’ enlisted force is the largest in the Marine Corps’ active component and has the greatest share of women (greater than the female share of commissioned officers and of warrant officers). In addition, we limit our analysis to Marine-Marine couples, since we only information on spouses’ unit location for Marines who are married to other Marines. Per Marine Corps policy, we define an enlisted Marine-Marine couple as being colocated if the two Marines’ assigned locations are within 50 miles of each other. For 6 percent of the sample, we were not able to calculate a mileage distance, since we did not have the latitudes and longitudes for at least one of the Marines’ units. The most common reason for this was that the unit address was an Air/Army Post Office (APO) or Fleet Post Office (FPO), neither of which has zip codes indicating the exact location of a unit. In other cases the unit was located in a foreign country to which zip codes do not apply or the unit zip code was simply not listed on the Marine’s personnel file. In these cases, we compared the three-digit geo-location codes of the Marine and his or her spouse. The couple was determined to be colocated if the geolocation codes matched.

Using CNA’s decision file, we determined the marital and colocation status of Marines at the time of their reenlistment decisions. We then estimated the relationship between colocation and the probability of reenlistment. We considered the costs and benefits of colocation compared with other reenlistment incentives.

This annotated briefing presents our enlisted Marine-Marine results. An earlier report presented the enlisted Navy-Navy results.

Our estimates suggest that colocation is associated with a higher estimated propensity to reenlist for certain groups only—namely, Zone A women, Zone B men, and Zone C women. Specifically, at the Zone A reenlistment decision, colocated female Marines are 3.7 percentage points (ppts) more likely to reenlist than their noncolocated counterparts (40 vice 36.3 percent). The estimated effect for men is insignificant: colocated and noncolocated Zone A male Marines reenlist at different but statistically indistinguishable rates (50.9 vice 47.8 percent).

At Zone B, our estimates suggest that colocated men are 5.2 ppts more likely to reenlist than their noncolocated counterparts (81.5 vice 76.3 percent), while we find no statistically significant difference in the estimated probability of reenlisting for colocated vice noncolocated women (70.6 vice 69.2 percent). At Zone C, the difference in the estimated probability of reenlisting for colocated vice noncolocated women is 5.0 ppts (87.8 vice 82.7 percent). For men, the difference is 0.9 ppts (91 vice 90 percent), but it is not statistically significant.

We also compared male and female Marines in the different marital categories to their respective single counterparts.

For married male Marines, we find that they are more likely to reenlist than their single male counterparts, regardless of whether they are married to a civilian spouse, a colocated Marine spouse, a noncolocated Marine spouse, or an other-service spouse.

For female Marines, however, the differences between single and married reenlistment rates are less consistent across the various spouse types. Among married female Marines, only those who are married to other Marines and are colocated are more likely than their single counterparts to reenlist. Single female Marines have higher estimated reenlistment rates than female Marines with civilian spouses and spouses in other services (statistical significance varies somewhat by zone). Single female Marines have higher estimated reenlistment rates than female Marines with noncolocated Marine spouses in Zones A and C, and although noncolocated female Marines have higher estimated reenlistment rates than their single counterparts in Zone B, the difference is not statistically significant.

The overall single v. married comparisons for female Marines suggests that, on average, female Marines may have a harder time maintaining their service career while married to a non-Marine or a Marine with whom they are not colocated than male Marines do.

Finally, we report an unexpected result: at the Zone A decision point, single female Marines have a notably higher estimated probability of reenlisting than single male Marines (40 percent v. 25 percent). Single Marines make up the majority of male Marines and a plurality of female Marines at the Zone A decision point, so this finding summarizes the reenlistment behavior of a significant portion of first-term decision-eligible Marines.

To put our analysis in context, we present some basic statistics on the number of Marine-Marine enlisted marriages as of September 2017. Using the Marine Corps’ enlisted personnel records from the Total Force Data Warehouse (TFDW) for the quarter ending September 2017, we found 6,403 enlisted Marines who had military spouses. Among these 6,403 Marines, about 64 percent, or 4,132, were married to other active component (AC) enlisted Marines. The remaining 2,271 Marines were either married to non-Marine servicemembers (we expect that most fall in this category), had an invalid spouse SSN, were married to a Marine officer, or were married to a Marine who recently left the service and the dependent relationship and spouse SSN had not yet been updated on their file. For these reasons we are ultimately able to identify a Marine spouse for only 4,132 of the 6,403 enlisted Marines with military spouses, allowing us to identify 2,066 enlisted Marine-Marine marriages in September 2017.

We were able to identify the assignment locations for each spouse (and thus the distance between them) in nearly all of the 2,066 enlisted Marine-Marine couples. Specifically, of the 4,132 AC enlisted Marines who we identified as being married to other AC enlisted Marines, we were able to identify zip codes and therefore compute distance for 94 percent of them. Colocation for the remaining 6 percent was defined by whether the two Marines had the same three-digit geo-location code.

We repeated this process for all quarters from FY05 through FY17 to identify all enlisted Marines married to other enlisted Marines in this time period. We then calculated the distance between assignment locations for the spouses and merged this information into a file containing each Marine’s decision history.

Our final sample includes all Marines who made reenlistment decisions some time between FY05 and FY17. Each enlisted Marine is coded as being in one of the following colocation categories, which are a combination of marital and colocation statuses:

1. Married to another Marine and assigned to a location within 50 miles of spouse’s assignment (colocated)
2. Married to another Marine and assigned to a location more than 50 miles away from spouse’s assignment (not colocated)
3. Married to a non-Marine Corps military spouse (colocation unknown)
4. Married to a civilian spouse
5. Single

The data are organized as a 13-year collection of snapshots of decisions (i.e., a cross-sectional dataset) rather than as observations that follow each Marine over time (i.e., a longitudinal dataset). At the time of Zone A decisions, roughly 3.5 percent of our sample is colocated compared to 3.8 percent at the time of Zone B and Zone C decisions. In Zone A, the majority of decision-making Marines are single, whereas in Zone B and C decisions the majority have civilian spouses. Overall, our sample consists of about 257,000 Zone A decisions, 63,000 Zone B decisions, and 32,000 Zone C decisions.