Although the retreat from marriage has affected Americans of all backgrounds, racial and ethnic differences in marriage rates have persisted or even widened over the past few decades. In 1960, the black-white marriage divide was negligible: a mere 8 percent of black adults ages 25 and up, and 9 percent of their white counterparts, had never been married. By 2012, those numbers had shot up to 36 percent for African Americans and 16 percent for whites.

But there is one major U.S. context where, in two scholars’ words, “marriage rates bear an anachronistic resemblance to those of the 1950s era”: the military. Bucking the trend toward delaying or forgoing marriage, military service members continue to marry at young ages and at high rates. What is more, at least as of 1999, the racial marriage divide evident among civilians was absent in the military. Jennifer Lundquist, a sociologist and demographer at the University of Massachusetts-Amherst, made that discovery in a fascinating 2004 study with implications beyond the military context.

Her primary source of data was the National Longitudinal Survey of Youth 1979. Using the first several waves of the annual survey, she assessed how young adults’ likelihood of marrying in each year varied by their race and whether they were civilians or military enlistees. (Thus the findings tell us about early marriage, not about survey participants’ likelihood of ever marrying.) The results of her logistic regression event history analysis were striking:

Notwithstanding various controls for socioeconomic status—educational attainment of both respondent and parent, AFQT scores, employment status, and childhood family structure—as well as attitudinal measures, race continues to be one of the strongest predictors of remaining single in the civilian world but means little in the military world.

Next, in a multivariate propensity score matching analysis, Lundquist “matched” black and white members of the military with civilians who—by all available measures—closely resembled them, and then compared the marital outcome of each military member with that of his or her paired civilian(s). This analysis reveals, as far as is possible, what difference being in the military (as opposed to other background factors) makes to an individual’s likelihood of marrying. Here, it turned out that both white and black enlistees were significantly more likely to marry over the course of the survey than matched civilians, but the magnitude of the military-civilian difference was greater for African Americans.

How the military fosters marriage among young adults is not a great mystery. For Americans with no college degree, the military provides higher wages, greater job security, and more generous employment benefits (health insurance, subsidized childcare, etc.) than most available jobs in the civilian world. Though serving in the military also means facing stresses that could undermine relationships, soldiers’ financial and job security certainly make it easier to commit to marriage.

Being married also gives service members access to certain perks that their unmarried counterparts do not enjoy. For instance, Lundquist writes,

Marriage is a way to leave barrack life during the first term of enlistment. Living in the barracks, though not necessarily required, is free. Marriage enables the enlistee to move off base with a housing allowance. Furthermore, military members with dependents (which include children and/or a spouse) receive a higher housing allowance than members without dependents when there is not housing available on base. While it is unlikely that soldiers marry solely for the extra allowance, it may influence family formation decisions.

Service members who are cohabiting with a romantic partner, by contrast, are treated as single. Whereas some lower-middle-class couples in the civilian world receive more generous government benefits and tax credits if they live together without marrying, military couples are generally better off if they marry.

In a more recent paper that Lundquist coauthored with Zhun Xu, based on interviews with members of the Army, the two report that the expectation of deployment and frequent moves also spur couples to marry for the sake of stability amid major life changes.

Being married also gives service members access to certain perks that their unmarried counterparts do not enjoy.

Finally, the military may informally promote marriage and pro-family values, Lundquist writes. This could both make service members more likely to marry and attract more family-oriented civilians to join the military. Lt. Colonel John Gregory, a professor of Chinese at the United States Military Academy who has spent many years in the Army, told me he believes that “the secure financial environment is the most significant factor” in the military’s high marriage rates, but he agrees that marriage is also directly encouraged in the armed forces.

The more difficult question is why serving in the military has greater effects on the marriage rates of African Americans, such that there is no racial gap in marriage rates in the military. Lundquist suggests in her 2004 study and a separate study on divorce that the answer may partly lie in how black service members suffer less discrimination and segregation than black civilians. Military housing, for instance, is racially integrated, which (among other things) equalizes access to schools, hospitals, stores, public transportation, and other resources important to families. Plus, Lundquist notes, “unlike in the general population, similarly educated whites and blacks in the military differ little in terms of pay scale and promotion.” For such reasons, two sociologist-veterans have called the Army “an organization unmatched in its level of racial integration…[and] its broad record of black achievement.”

Lt. Colonel Gregory says “there is zero tolerance for overt racism in the military,” and confirms that it “is certainly much less segregated than many parts of our society.” At West Point’s elite elementary school, for example, “everyone from the general’s children to the lowest-ranking soldiers’ children can all be found in the same class and have the same teachers.” He adds, “The same could be said of our health care services as well as all of our community centers.” There are few civilian communities where similar conditions prevail.

The military’s atmosphere of relative racial equality may help explain African Americans’ ongoing overrepresentation in the armed forces, and it could indirectly increase black soldiers’ likelihood of marrying. W. Bradford Wilcox and Nicholas Wolfinger document in Soul Mates: Religion, Sex, Love, and Marriage among African Americans and Latinos that “the experience of discrimination can engender feelings of anger, depression, distress, and hopelessness,” and “lead men and women to lose faith in society and conventional morality, turn away from education and gainful employment, [or] pick up drugs or alcohol.” In minimizing racial discrimination, perhaps the military fosters greater hope, trust, and life satisfaction among black service members, factors that spill over to benefit their relationships.

At the end of her 2004 paper, Lundquist looked at more recent (but also more limited) data from 1999 to investigate whether her findings remained relevant. Judging by the percentage of black and white civilians and military service members who had never been married at various ages, little had changed since the early 1980s: Americans in the military in 1999 were more likely to have married than those outside it, and civilians’ black-white marriage divide was absent in the military. In addition, a 2012 paper by other scholars showed that in 2002-2005, among both whites and blacks, men in the military were more likely to have ever married than civilians similar to them in age, employment status, and educational attainment, and the military-civilian gap was wider for African Americans. So it seems reasonable to assume that Lundquist’s conclusions still hold.

Although we could not and should not attempt to replicate in the civilian world all of the factors behind the military’s high marriage rates, Lundquist believes certain aspects of the military model should instruct policy-makers on how to promote marriage most effectively. “On the basis of the military example, marriage is widespread in part due to stable, decent-paying jobs and the availability of health care and education benefits to family members,” she argues. “Given growing class inequality, precarious underemployment, and long-term unemployment, perhaps there are some aspects of the military employment model that could be extended to all U.S. youth.”