Sixty years ago, in 1963, 94% of American children were born to married mothers. Today, the figure is only 60 percent. This decline signals a fundamental disruption in the long-standing stability of the traditional family, the foremost institution shaping each generation of children. Using the Census Bureau’s American Community Survey, I find that in 2021, 40% of 17-year-olds with native-born parents had spent some part of their childhood without one or both biological parents.
Because policy focuses unduly on economic outcomes and problems, we tend to view this trend through the prism of its impact on poverty and upward mobility. But even in economically secure families, the increase in family disruption has had untold emotional, psychological, and social consequences.
If income or mental or physical health had fallen 40% over 60 years, it would be the central issue driving our policies and politics. For the same reason, the decline in married parenting should torment policymakers today. One way that policy can potentially reverse this trend is by addressing the ways that our safety net has weakened marital childbearing and childrearing. But we need ambitious proposals to tackle such an entrenched problem.
Reforming the Earned Income Tax Credit (EITC) presents such an opportunity, and one that could also address issues of family affordability and work-family balance. To encourage marriage and marital childrearing and to provide working- and middle-class families with more options in combining work and parenting, EITC eligibility should be tied to individual earnings rather than the earnings of the tax unit. This change would not affect single beneficiaries, but it would markedly alter the benefits available to married couples.
Under current law, there are four benefit schedules for married couples, depending on whether they have no children, one, two, or three or more children. The income of the couple is compared against the appropriate benefit schedule, which first increases with income, then levels off, then declines for higher-income couples. The reform proposed here would let each spouse of a married couple use their own individual income to determine their own EITC benefit, using the existing benefit schedules for married couples.
At the most basic level, this would allow any couple to marry who would do so but for the loss of EITC benefits. But not only would this marriage penalty be eliminated; there would also be a sizable marriage bonus. That should affect the behavior of a larger group of couples.
Consider an unmarried couple, in which one partner with two children makes $20,000 per year and the other (childless) partner makes $30,000 per year. Under current policy, while they remain unmarried, the parent gets a $6,604 EITC benefit, so their combined income is $56,604. If they get married, their EITC drops to $1,991 because the couple’s income is relatively high and the benefit phases out. They are worse off by $4,613. Under the reform proposed here, after marrying, the parent continues to receive an EITC of $6,604, but now the childless partner also gets a $6,203 benefit, based on their $30,000 earnings and the two children. That leaves the couple better off by $6,203—a net swing of $10,816.
The marriage bonus would be even more powerful if both partners were previously getting an EITC. Consider unmarried partners, each with one child, making $20,000 and $30,000. Their combined EITCs, both based on the schedule for a single child, come to $6,637 (under current policy and the reform). Under current policy, if they marry, their EITC falls to $1,991. But under the reform, if they marry, they both get to move to the EITC schedule for two children. Their combined EITC nearly doubles to $12,807.
For families already headed by a married couple, the reformed EITC would, at a minimum, serve as an effective tax cut, increasing their ability to afford necessities, pay bills, or invest in their children. But it would help married couples in other ways as well.
For one, the proposed reform would help already-married parents to reduce their combined work if they wish to have one parent spend more time caring for the children. Consider a married couple with two children where, as above, one parent makes $20,000 and the other makes $30,000. Under the reformed EITC, if the first parent cut their work in half, so that their earnings dropped to $10,000, the couple would still receive a combined EITC of $10,213. So their combined income (with the credit) would be $50,213—not much less than the $51,991 they get under current policy working their existing hours. If the first parent was previously working 52 weeks for 40 hours a week at $10 an hour, they could drop down to 20 hours a week or 26 weeks of full-time work. Or they could take a more flexible job, such as in the gig economy, that works better with their parenting schedule.
Or perhaps the couple would like to have a third child. Under current policy, if both parents continued to work the same amount, their EITC amount would increase from $1,991 to $2,816 if they moved from the schedule for two children to the schedule for three or more children—an increase of just $825. Under the reform, the amount would rise from $12,807 to $14,458—by $1,651, or twice as much as today. And apart from the increased boost in moving from two kids to three, their EITC would be over five times as large as under current policy, which may be the more important consideration.
My colleague at the American Enterprise Institute, Bodi Yang, estimated the cost of this proposal at $395 billion over 10 years. (To avoid subsidizing upper-income families with one high-earning spouse and one who does a nominal amount of work, the new EITC would be restricted to people in tax units with under $100,000 in combined income.) That estimate assumes that no one changes their marital, fertility, or work decisions.
Of course, the entire point of the proposal is to influence those decisions. To the extent that it is successful in doing so, the cost would rise. But that increase would reflect people taking advantage of the expanded opportunity to do the things the reform incentivizes. Higher outlays would come from more marriage, more marital fertility, and more spouses altering their working decisions to optimize under expanded choices. (Some would presumably move from one-worker to two-worker, while others would move from two to one.)
This proposal could go further in incentivizing marriage. It could, for instance, eliminate the EITC for single childless adults. Providing EITC benefits for single adults disincentivizes marriage because it reduces the return to getting or staying married. The number of single childless beneficiaries is relatively small, as are the benefits they receive. Research suggests that these benefits do not incentivize work very much for this subpopulation. This addition to the proposal would reduce the 10-year cost to $370 billion.
Alternatively, the EITC schedules for married parents could be reformed along the lines of my 2021 proposal. That reform envisioned giving all married couples a single schedule, regardless of how many children they have. The super-schedule would follow the current-policy schedule for a married couple with three or more children, even for parents who only have one or two children, up to the maximum EITC amount available under current policy ($7,430). This maximum amount would be available for couples with incomes nearly 50% higher than under current policy before phasing out for those with incomes above this threshold.
This version of the reform would have several effects. It would incentivize marriage, disincentivize childlessness, and incentivize marital childbearing. Most notably, the reform would incentivize people to have and raise their first child in marriage rather than as single parents, since the increase in the EITC over current policy would be largest for a family with one child. If these marriages are sticky, then subsequent births will also occur within marriage. By tying this super-schedule to individual earnings, which my 2021 proposal did not envision, the incentives would be even stronger.
Imagine an unmarried couple expecting their first child and making $20,000 and $30,000. Under current policy, if they remain unmarried, their combined income will be higher than if they marry by the $3,995 EITC benefit the mother would receive. Under the reforms proposed here, including the super-schedule for married parents, both parents would receive a $7,430 EITC if they marry, for a combined $14,860, but the mother would still only get $3,995 if they do not, leaving them better off by $10,865 if they wed. Without the super-schedule, they both get a $3,995 EITC and are better off by only $3,995 versus not marrying.
Adding the super-schedule to the proposal would increase the 10-year cost to $767 billion. Adding it and eliminating the single childless EITC would cost $741 billion.
An EITC with a marriage bonus would counter, at least partly, the loss of benefits a couple might see from marrying, such as from food stamps or housing subsidies. The marriage penalties in these programs are why reform must go further than simply removing the EITC marriage penalty; an EITC marriage bonus is necessary to counteract the more powerful marriage disincentives in cash and noncash benefit programs. The point is not to make the tax code pro-marriage but to make the combined effects of tax and transfer policy less anti-marriage. The proposal here is much more preferable to policies that directly attempt to reduce marriage penalties in safety net programs. Those policies—which phase-out benefits more slowly as income rises, allow couples to ignore spousal income in determining eligibility, or tie benefits to individual rather than family income—are expensive, increase dependency, and risk luring more people off of work.
Reducing the number of children being raised without one of their biological parents should be at the top of the nation’s priorities, and it is time to throw everything at the problem. Reforming the EITC along the lines described here should be part of that effort. It would help already-married working- and middle-class families while leaving the safety net undiminished for single-parent families, even as it clearly rewards marriage.
Correction: This article has been updated. The original version used cost estimates that were understated. The tax model used data on tax unit earnings, splitting those earnings between spouses in a tax unit based on a simple rule. Using data that include the actual individual earnings of spouses, the tax model produced the significantly higher costs in the current version. The author thanks Max Ghenis, CEO of PolicyEngine, for discovering the error, and Angela Rachidi, of AEI, for expressing her skepticism of the original results.