The Welfare Cliff: How a $1 Raise Costs a Black Family $10,000

Let me tell you about a woman I will call Denise, though her real name does not matter because there are two million women exactly like her in America today. Denise is a certified nursing assistant in Memphis, Tennessee. She earns $14.50 an hour, which comes to roughly $30,160 a year. She has two children, ages four and seven. She receives SNAP benefits worth $5,400 annually. She receives a Section 8 housing voucher worth $8,760 annually. Her children are enrolled in Medicaid, which provides healthcare coverage that, if purchased on the private market, would cost approximately $6,200 per year. She receives a childcare subsidy through the Child Care and Development Fund worth approximately $7,800 annually. Her total compensation — wages plus benefits — is approximately $58,320. She is surviving. Barely, but surviving.

Denise’s supervisor offers her a promotion to shift lead. The raise is $2.50 an hour — an additional $5,200 per year, bringing her gross income to $35,360. She should take it. Any rational person would take it. Except Denise has done the math, because poor people are not stupid, whatever the comfortable classes may believe, and the math tells her something that should make every American citizen furious: that $5,200 raise will cost her over $12,000 in lost benefits. Her new income pushes her above the eligibility threshold for the childcare subsidy. It reduces her SNAP benefits by $1,800. It triggers an increase in her Section 8 copayment. She does not lose Medicaid immediately — the Affordable Care Act provided a coverage gap buffer — but she is now close enough to the threshold that any overtime, any additional hours, any Christmas bonus could push her children off their health insurance.

Denise turns down the promotion. She is not lazy. She is not lacking ambition. She is making the most economically rational decision available to her within a system that punishes the precise behavior it claims to encourage. And the tragedy — the unspeakable, bipartisan, decades-long tragedy — is that Denise’s situation is not a bug in the system. It is a feature. It has been documented, studied, quantified, and reported on by the Congressional Budget Office, the Urban Institute, the Government Accountability Office, and a dozen academic institutions. Everyone who makes policy in Washington knows about the welfare cliff. And no one fixes it. Because fixing it would require both parties to abandon positions that are more politically useful than they are morally defensible.

The Math of the Trap

The Congressional Budget Office published a landmark report in 2015, updated in subsequent analyses through 2022, documenting the effective marginal tax rates faced by low-income families as their earnings increase. The findings are staggering: a single parent with one child earning between $15,000 and $25,000 per year faces an effective marginal tax rate of approximately 60 percent. That means for every additional dollar earned, 60 cents is lost to reduced benefits and increased taxes. For families in certain income ranges in certain states, the effective marginal tax rate exceeds 80 percent. In some documented cases, it exceeds 100 percent — meaning a family literally loses more in benefits than it gains in earnings.

Let me say that again, because the number deserves to be absorbed rather than skimmed. There are income ranges in the United States of America where a working parent — a person doing exactly what society says they should do, getting up in the morning, putting on a uniform, going to a job — keeps less than twenty cents of every additional dollar they earn. There are ranges where they keep nothing. There are ranges where earning more money makes them poorer. This is not an opinion. This is the published finding of the nonpartisan Congressional Budget Office, and it has been confirmed by every serious analysis conducted in the past two decades.

The Urban Institute’s 2021 analysis of benefit phase-outs across all fifty states mapped the specific income thresholds where families experience the cliff. Their findings are a cartography of cruelty. In Pennsylvania, a single mother with two children earning $29,000 per year who receives a raise to $36,000 loses a net total of approximately $9,500 in combined SNAP, childcare, and housing benefits — a net loss of $2,500 for working harder. In Colorado, the cliff is steeper: the same family profile loses over $14,000 in benefits when crossing from $28,000 to $38,000 in earned income, creating a net loss of $4,000. In Connecticut, the loss can exceed $21,000.

The programs that create the cliff are individually well-intentioned. SNAP — the Supplemental Nutrition Assistance Program, formerly food stamps — provides food assistance to 42 million Americans, of whom approximately 26 percent are Black. Section 8 housing vouchers serve 2.3 million households, of which approximately 48 percent are headed by Black tenants. Medicaid covers 83 million Americans, with Black Americans constituting approximately 20 percent of enrollees despite being 13 percent of the population. The Child Care and Development Fund provides subsidies to approximately 1.4 million children per month.

Each program, taken alone, makes sense. Taken together, they form a labyrinth with invisible walls. Because each program has its own eligibility threshold, its own income calculation method, its own phase-out schedule — or, more often, its own cliff, where benefits do not phase out gradually but vanish entirely the moment income crosses a line — the cumulative effect is that a family can be receiving $25,000 or more in combined benefits and lose all of it within an income range of $8,000 to $12,000. The programs were never designed to work together. They were designed by separate committees, administered by separate agencies, funded by separate appropriations, and governed by separate rules. The result is a system that, in its totality, does something that no individual program intended: it makes upward mobility economically irrational.

The Marriage Penalty Nobody Mentions

The welfare cliff has a companion that is even less discussed, because discussing it requires acknowledging that government policy has consequences for family formation, and that is a conversation that makes both parties deeply uncomfortable. It is the marriage penalty embedded in virtually every means-tested benefit program in America.

Consider two parents, each earning $18,000 per year. Unmarried and living apart — at least on paper — each qualifies as a single-parent household for benefit purposes. Together, their combined benefits might total $30,000 or more in SNAP, Medicaid, housing assistance, and childcare subsidies. If they marry, their combined income of $36,000 is assessed as a single household, and they lose the majority of those benefits. The Urban Institute calculated that for families in the $20,000 to $40,000 income range, marriage can result in a net loss of $10,000 to $20,000 annually in combined benefits.

This is not a theory about why Black marriage rates have declined. It is documented mathematics. When the government constructs a system in which a man and a woman are financially better off pretending they are not a family than legally becoming one, some portion of them will respond to the incentive. Not all. Not most. But enough to bend the curve. And the curve has been bending for sixty years. In 1960, 61 percent of Black adults were married. By 2020, that number had fallen to 30 percent — the lowest marriage rate of any demographic group in America.

I am not arguing that the marriage penalty is the sole cause of declining Black marriage rates. The causes are multiple and complex — mass incarceration, economic dislocation, cultural shifts, the legacy of discriminatory policies that predate the welfare state. But I am arguing, with the full weight of the documented evidence, that a system which financially penalizes marriage in the precise income range where Black families are concentrated is not a neutral factor. It is an active force. And the refusal to acknowledge it, the insistence on treating family structure as a subject too sensitive for policy analysis, is a form of negligence that is measured in broken homes and fatherless children.

The Rational Refusal of a Promotion

The academic literature on the welfare cliff contains a phrase that should haunt every policymaker who reads it: the rational reluctance to earn more. Researchers at the Federal Reserve Bank of Atlanta built an interactive tool — the Career Ladder Identifier and Financial Forecaster, or CLIFF — that allows users to input their state, family size, income, and current benefits, and then model what happens when their income increases. The results, across every state and every family configuration, tell the same story: there are vast income ranges where working more, earning more, or advancing in a career produces a net loss in family resources.

A 2019 study published in the National Tax Journal found that among single mothers receiving multiple benefits, the effective marginal tax rate in the phase-out range averaged 65 percent, with peaks exceeding 95 percent at specific income thresholds. The study’s authors wrote, with the restrained language of academic economics, that these rates “create significant disincentives to labor force participation and earnings growth among low-income families.”

“Significant disincentives.” Let me translate that from academic into English. It means that we have built a system where a mother in Detroit who is offered a two-dollar raise has to sit down with a calculator and figure out whether earning $4,160 more per year will cause her to lose $8,000 in housing assistance, $3,600 in food assistance, and $6,000 in childcare subsidies. It means that the rational decision — the decision that any economist, any banker, any senator would make if faced with the same mathematics — is to say no. To refuse the raise. To stay exactly where you are. Because the American safety net is not a ladder. It is a floor with a trapdoor, and the trapdoor opens the moment you try to climb.

The Bipartisan Failure

Here is where I am going to make everyone angry, because the welfare cliff is not a Republican problem or a Democratic problem. It is a bipartisan failure of moral imagination, and both parties are complicit in its perpetuation for reasons that have everything to do with political convenience and nothing to do with the lives of poor families.

Democrats, broadly, refuse to acknowledge the cliff because doing so would require admitting that the programs they built and defend have structural flaws that produce perverse outcomes. The Democratic position, reduced to its essence, is that these programs must be protected and expanded, and that any criticism of their design is an attack on the vulnerable people they serve. This is a rhetorical strategy, not a policy analysis. You can love the intention of a program and still demand that it be designed competently. You can support feeding hungry children and still insist that the mechanism for doing so not punish their mothers for earning a raise. But the Democratic coalition has made the defense of existing program structures a litmus test of moral seriousness, and the result is that the cliff remains untouchable.

Republicans, broadly, want to reduce spending on means-tested programs, which they frame as promoting dependency. The Republican solution to the welfare cliff is typically to lower the cliff by lowering the benefits — to reduce SNAP allotments, to impose work requirements, to tighten eligibility. This does not fix the cliff. It simply moves it to a lower income level, where it traps even poorer families. The Republican position treats the cliff as evidence that the programs themselves are the problem, when in fact the cliff is evidence that the design of the programs is the problem — a distinction that requires more nuance than a campaign slogan permits.

The result is a fifty-year stalemate in which the people who suffer the consequences of bad policy design are the people with the least political power to change it. And because the welfare cliff disproportionately affects Black families — who are disproportionately represented in the income ranges where the cliff is steepest — it functions, in practice, as a mechanism for the perpetuation of Black poverty, regardless of the intention behind any individual program.

The One Program That Gets It Right

There is one major federal program that does not create a cliff, and its success is the clearest evidence that the cliff is a design choice, not an inevitability. The Earned Income Tax Credit, or EITC, was enacted in 1975 and has been expanded multiple times by presidents of both parties. Unlike SNAP, Section 8, and Medicaid, the EITC does not vanish at an income threshold. It phases in gradually, reaches a maximum at a moderate income level, and then phases out gradually as income continues to rise. A family never faces a cliff. A family never loses more in EITC reduction than it gains in additional earnings. The incentive to earn more is preserved at every income level.

The results speak for themselves. The EITC is the single most effective anti-poverty program in the United States, lifting approximately 5.6 million people above the poverty line annually, including approximately 3 million children. Research by Hoynes and Patel (2018) found that a $1,000 increase in the EITC reduces the share of families in poverty by 7.4 percent. The program has been shown to increase labor force participation, particularly among single mothers — the precise population most affected by the welfare cliff in other programs. It improves birth outcomes, educational achievement, and long-term earnings for children in families that receive it.

The EITC works because it was designed with a simple principle: never make earning more money punitive. The gradual phase-out ensures that a family always keeps a meaningful portion of every additional dollar earned. There is no cliff. There is a gentle slope. And on that gentle slope, people climb.

If we know how to design a program that does not trap people, and we have known since 1975, then the continued existence of the welfare cliff in every other major benefit program is not ignorance. It is choice. It is the choice to maintain a system that is easier to administer than it is to live under. And the people who bear the cost of that administrative convenience are overwhelmingly Black, overwhelmingly female, and overwhelmingly invisible to the people who make the rules.

What Reform Actually Looks Like

The solutions are not mysterious. They are documented, modeled, and ready for implementation. What they lack is political will.

Universal benefit phase-outs. Every means-tested program should replace income cliffs with gradual phase-outs, modeled on the EITC structure. For every additional dollar earned, benefits should be reduced by no more than 20 to 30 cents. This ensures that work always pays, that promotions always increase net family income, and that the rational decision is always to earn more rather than to stay in place. The Urban Institute has modeled this approach and found that it would cost approximately $15 billion annually in additional benefit expenditures — a fraction of what is spent on programs that trap their recipients.

Coordinated benefit calculations. Currently, each program calculates eligibility independently, which means a family can face simultaneous cliffs in multiple programs at the same income point. A coordinated system would ensure that total benefit reductions across all programs never exceed a combined effective marginal tax rate of 50 percent. This is not a radical proposal. It is basic arithmetic applied to a system that currently ignores arithmetic in favor of bureaucratic silos.

Elimination of the marriage penalty. Benefit calculations should be restructured so that two parents in a household receive at least as much in combined benefits as they would as two separate single-parent households. The current system, which calculates benefits based on household income and thereby penalizes family formation, should be replaced with a system that is at worst marriage-neutral and at best marriage-positive. The cost of this reform has been estimated by the Brookings Institution at approximately $8 billion annually — less than the annual cost of the social pathologies that family dissolution produces.

Expanded EITC. The program that works should be made larger. Proposals to expand the EITC to childless workers, to increase the maximum credit, and to extend the phase-out range have been supported by economists across the ideological spectrum, from the American Enterprise Institute to the Center on Budget and Policy Priorities. A 2019 proposal by the National Academy of Sciences estimated that a 40 percent expansion of the EITC would reduce child poverty by 25 percent.

These reforms are not utopian. They are not expensive by the standards of federal spending. They do not require the elimination of any program that feeds a hungry child or houses a homeless family. They require only the willingness to redesign systems so that they do what they claim to do: help people move from poverty to self-sufficiency rather than trapping them in the space between.

A Trap With a Compassionate Name

I want to return to Denise, the nursing assistant in Memphis who turned down a promotion because the math told her she could not afford to succeed. Denise is not a statistic. She is a person making a rational decision within an irrational system. She is a mother who calculated the cost of ambition and found that it exceeded the cost of stagnation. She did not choose poverty. She chose the option that kept her children fed, housed, and insured, because the system presented her with a choice between progress and survival and forced her to pick one.

There are two million women like Denise in America right now, and the overwhelming majority of them are Black. They are not lazy. They are not gaming the system. They are responding to incentives that were designed by people who will never face them, administered by agencies that do not coordinate with each other, and defended by politicians who have never had to choose between a raise and their children’s health insurance.

“Anyone who has ever struggled with poverty knows how extremely expensive it is to be poor.” — James Baldwin, “Fifth Avenue, Uptown” (1960)

Baldwin wrote those words sixty-six years ago, and they remain the most precise description of the welfare cliff ever committed to paper. It is expensive to be poor. It is expensive in ways that the non-poor cannot imagine and do not care to learn. And the welfare cliff is the mechanism by which that expense is maintained — the invisible wall that ensures that poverty, once entered, is nearly impossible to exit through the normal channels of work and advancement that are available to everyone else.

A system that was designed to help but punishes progress is not compassion. It is a trap with a government seal. And the fact that it has been documented, studied, quantified, and reported on for decades without being fixed tells you everything you need to know about whose interests the system actually serves. It does not serve the families trapped on the cliff. It serves the bureaucracies that administer the programs, the politicians who campaign on defending them, and the ideologues who use them as evidence for whatever argument they were already making.

The families on the cliff do not need defenders. They do not need advocates. They do not need another study. They need the cliff to be dismantled, replaced with the gradual slopes that we already know how to build, and reconstructed into a system that rewards every step upward rather than punishing the attempt. We have the knowledge. We have the models. We have the money. What we lack is the political courage to tell both parties that their sacred positions are less important than a mother’s ability to say yes to a promotion without calculating whether it will cost her children their dinner.

That is not a policy failure. That is a moral one. And until we name it as such, Denise and the two million women like her will continue to do the only rational thing the system allows: stay exactly where they are, on the edge of a cliff that their government built and their government refuses to remove.