The following piece was co-authored by R Street Associate Fellow Neil Gilbert.


Traditional marriage is in big trouble in the United States. Between 1960 and 2011, the share of white adults 18 and older who were married declined by 25 percent, while the declines for Hispanic and black adults were 35 percent and 50 percent respectively.

This trend has profoundly negative consequences for adults, children, and society as a whole. Because a stable marriage is one of the best means of avoiding poverty, many children born and/or raised out of wedlock are economically disadvantaged. In addition, single parents tend to have less time, energy, and other resources for fostering their children’s development. And they are more likely to work at low-wage jobs with irregular hours. The negative effects run in both directions: The psychological stress of caring for a child alone can create barriers to professional success.

Some blame the decline in marriage on decaying moral values and loosening social norms abetted by the media. They yearn for a return to what they remember as the moral foundations of marriage. Programs designed to move us in this direction during the administration of George W. Bush had disappointing results.

Others, looking at the same data, see the unfortunate byproduct of a welcome increase in social tolerance and sexual freedom. Still others see the failure to marry as chiefly an economic decision. The decline in male employment and wages has reduced the number of “marriageable males.” Hence proposals to increase the earnings of low- and middle-income workers through expanded job training, a stronger role for labor unions, a higher minimum wage, more career-oriented high schools, and more effective public schooling generally. But translating these proposals, too, into effective programs has proven problematic.

What can be done?

There is one step that can be taken to make marriage more attractive to low- and moderate-income families that seems to be working in other countries with extensive social welfare benefits and high levels of cohabitation: reduce the marriage penalties embedded in our means-tested welfare programs. The growth in these programs has been dramatic: Federal spending on the nine largest means-tested programs (adjusted for inflation) climbed almost 750 percent between 1968 and 2011, while the U.S. population grew by only 56 percent. Ending the marriage penalty in these programs, then, could have far-reaching effects.

Marriage & the welfare state

 As far back as 1997, a Congressional Budget Office report noted that a disincentive to marry is created by the potential loss of means-tested benefits received by low-income mothers—food stamps, housing assistance, Medicaid. If marriage to an employed man would increase the family income enough to make the mother and children ineligible for these benefits, she might choose to remain unmarried. Gene Steuerle of the Urban Institute notes wryly, “Cohabiting or not getting married has become the tax shelter of the poor.”

Eligibility for the major means-tested benefits is determined by the income of either the “family unit” (individuals living together related by blood, marriage, or adoption) or the “economic unit” (individuals living together who share resources, for example, to purchase food). The family unit is used to calculate eligibility for child-care vouchers, housing-choice vouchers, Medicaid, and, in all but a few states, Temporary Assistance for Needy Families (TANF). For these programs, the income of a cohabiting adult who is not the children’s biological parent is disregarded. Eligibility for the Supplemental Nutrition Assistance Program (SNAP), the Special Supple-mental Nutrition Program for Women, Infants and Children, school meals, and the Low Income Home Energy Assistance Program is based on the economic unit.

For cohabiting couples, the incentives are clear: The way to maximize benefits is to avoid getting married or to fail to disclose parental status and financial sharing to authorities.

Measuring marriage penalties

Most safety-net benefits have steep phase-out rates or even “cliffs” as income rises. The recent expansion of safety-net programs to more middle-income households has extended marriage penalties to more Americans. Depending on the state where they live, a family of four making as much as $92,150 may be eligible for Medicaid; a similar family making $71,436 may qualify for child care assistance; a family earning $53,267 may claim the Earned Income Tax Credit; and a family making up to $31,008 may receive SNAP.

acking all the marriage penalties is difficult, as they are usually inadvertent and can appear in the most unexpected places. For example, the Affordable Care Act subsidizes health-insurance premiums for households making between 133 percent and 400 percent of the poverty line. For each additional dollar earned, households are required to contribute a higher percentage of their income.

For a family of three earning between $59,370 and $79,160, the ACA subsidy is a flat 9.5 percent of household income. If the couple is unmarried, each partner qualifies separately for the subsidy, even if they live together. But if they are married, their combined income reduces their subsidy and may disqualify them altogether. If one spouse earns a lot more than the other, they may lose as much as $3,486 in subsidies.

We used the Urban Institute’s newly updated Net Income Change Calculator database of national and state tax and social-welfare rules for all 50 states and the District of Columbia to estimate the effects of marriage on the incomes of seven hypothetical couples with two children. Our couples had earnings of $10,000, $20,000, $40,000, or $50,000, with several variations in the division of earnings between the partners; and we placed each scenario in states at the 10th, 50th, and 90th percentile of the distribution in marriage penalties and bonuses under current law. For each, we first ran the numbers with the couple cohabiting, one partner not a biological parent of the children present, and reportedly not sharing resources. We then ran the numbers for the same couple, only this time married and sharing resources, and compared the results to see whether marriage produced an increase (bonus) or decrease (penalty) in the couple’s income. Overall, we found that:

– Marriage penalties for those earning $10,000 range between $1,824 and $5,016.

– Marriage penalties for those earning $20,000 range between $1,784 and $8,902.

– Marriage penalties for those earning $40,000 range between $5,544 and $13,248.

– Marriage penalties for those earning $50,000 range between $6,960 and $14,148.

Ending marriage penalties

The simplest way to end marriage penalties in the welfare system would be to treat all cohabiting couples—whether or not they are biological parents—as economic units for all means-tested benefits. The default rule could be adopted that all couples who cohabit for longer than a year are automatically treated as economic units. Those who live together without sharing expenses could apply for an exemption and submit evidence demonstrating their separate finances. This would reduce the marriage penalty for cohabiters—but raise the cost of cohabitation and perhaps discourage even more biological parents from trying to form a family.

In Europe, where as many as a quarter of couples are unmarried cohabiters, benefits frequently are based on the household unit, rather than the family or economic unit. But there is also a recognition that not all cohabitation is marriage-like. Consequently, European agencies often impose further requirements on cohabiters seeking to qualify for benefits, such as joint bank accounts or shared mortgage loans. The French look for evidence that a couple is living a “common life.”

To enforce these rules, some countries, such as the Netherlands, use home visits. That sounds a lot like the discredited “man-in-the-house” rules of pre-1965 welfare programs and could enmesh U.S. social agencies in endless arguments about household economics. We think this approach is unlikely to gain widespread support in the United States.

Less politically contentious, but much more expensive, would be to minimize marriage penalties by extending to means-tested welfare programs the tax code’s approach to marital income. Married couples may file their taxes jointly, and the tax code rewards many of them for doing so—or at least softens the impact of the code’s progressively higher rates for higher incomes. Upon marriage, means-tested social-welfare benefits could be phased out gradually to reflect society’s interest in stable family arrangements.

This is not a novel idea. Marriage penalties under the old pre-welfare-reform Aid to Families with Dependent Children (AFDC) were enough of a concern that, before the enactment of TANF, at least three states obtained AFDC waivers designed to reduce the loss of welfare benefits due to marriage. These reforms were termed “wedfare” or “bridefare.” Nothing much came of these initial attempts to soften the marriage penalty because, with the enactment of the 1996 federal welfare reform, caseloads fell across the country and the issue became moot.

With the expansion of means-tested programs, it’s time to reconsider such efforts. It’s impossible to predict precisely how much more a marriage-neutral or marriage-friendly system would cost, but strengthening marriage would, over the long run, lift many families out of poverty.

Featured Publications