Let me tell you about a woman I will call Denise. Her real name does not matter. 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. That comes to about $30,160 a year. She has two children, ages four and seven.
She receives SNAP benefits. These were once called food stamps. They are worth $5,400 a year. She gets a Section 8 housing voucher. This federal subsidy pays part of her rent. It is worth $8,760 a year. Her children are on Medicaid. This provides healthcare coverage. It would cost about $6,200 per year on the private market. She gets a childcare subsidy through the Child Care and Development Fund. It is worth about $7,800 a year. Her total pay — wages plus benefits — is about $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. That is an extra $5,200 per year. It would bring her gross income to $35,360. She should take it. Any rational person would take it.
Except Denise has done the math. Poor people are not stupid. The math tells her something that should make every American furious. That $5,200 raise will cost her over $12,000 in lost benefits.
Her new income pushes her above the limit for the childcare subsidy. It cuts her SNAP benefits by $1,800. It triggers an increase in her Section 8 copayment. She does not lose Medicaid right away. But she is now close enough to the limit. Any overtime or extra hours 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 rational decision available to her. The system punishes the very behavior it claims to encourage.
The Math of the Trap
The Congressional Budget Office published a landmark report in 2015. It was updated through 2022. It documents the effective marginal tax rates faced by low-income families. These rates measure the percentage of each extra dollar earned that disappears. It disappears through lost benefits and higher taxes. 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 about 60 percent (CBO, Publication No. 50923, 2015). That means for every extra dollar earned, 60 cents vanishes.
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. This means a family literally loses more in benefits than it gains in earnings (CBO, “Marginal Federal Tax Rates on Labor Income,” 2019).
There are income ranges in the United States where a working parent keeps less than twenty cents of every extra dollar earned. There are ranges where earning more money makes them poorer.
The Urban Institute’s 2021 analysis mapped benefit phase-outs across all fifty states. They identified the specific thresholds where families experience the cliff (Maag et al., Income Volatility and the Safety Net, Urban Institute, 2021). Their findings are a map of cruelty.
- Pennsylvania. A single mother with two children earning $29,000 gets a raise to $36,000. She loses about $9,500 in combined benefits. She has a net loss of $2,500 for working harder.
- Colorado. The same family loses over $14,000 in benefits when crossing from $28,000 to $38,000. This creates a net loss of $4,000.
- Connecticut. The loss can exceed $21,000.
Each program makes sense on its own. SNAP, Section 8, Medicaid, and childcare subsidies each have a goal. Taken together, they form a labyrinth with invisible walls. Each program has its own eligibility limit. Each has its own income calculation method. Each has its own phase-out schedule. More often, each has its own cliff. Benefits vanish entirely the moment income crosses a line. The total effect is devastating. A family can be receiving $25,000 or more in combined benefits. They can lose all of it within an income range of $8,000 to $12,000.
The programs were never designed to work together. They were created by different committees. They are run by different agencies. They are funded by different budgets. The result is a system that does something no single program intended. It makes upward mobility economically irrational.
The Marriage Penalty Nobody Mentions
The welfare cliff has a partner that is even less discussed. Talking about it means admitting government policy affects family formation. That partner is the marriage penalty in every means-tested benefit program in America. “Means-tested” means income-limited.
Consider two parents. Each earns $18,000 per year. Unmarried and living apart, each qualifies as a single-parent household for benefits. Together, their combined benefits total $30,000 or more. This includes SNAP, Medicaid, housing assistance, and childcare subsidies. If they marry, their combined income of $36,000 is assessed as a single household. They lose most of those benefits (Thomas & Sawhill, The Future of Children, 15(2), 2005; Carasso & Steuerle, The Future of Children, 15(2), 2005).
The Urban Institute calculated the cost. For families in the $20,000 to $40,000 income range, marriage can cause a net loss of $10,000 to $20,000 every year in combined benefits.
Black Representation in Safety Net Programs vs. Population
USDA FNS, 2023; HUD, 2023; CMS, 2023
This is not a theory about why Black marriage rates have declined. It is documented mathematics. The government built a system where a man and a woman are financially better off pretending they are not a family. Some portion of them will respond to that incentive. 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. This is the lowest marriage rate of any group in America (U.S. Census Bureau, America’s Families and Living Arrangements — 2020, 2021).
The Strongest Counterargument — and Why the Data Defeats It
“The welfare cliff is not the primary driver of Black poverty or family dissolution. Cultural factors, mass incarceration, and employment discrimination are far more significant.”
Three data points show the cliff’s independent power. First — the CBO documented effective marginal tax rates of 60 to 100%+ in the precise income ranges where Black families are concentrated. These are not theoretical. They are published federal findings (CBO, 2015). Second — the EITC is the one program designed without a cliff. It lifts 5.6 million people above the poverty line every year. It increases work among single mothers. It improves birth outcomes, education, and long-term earnings. If design did not matter, the EITC would not outperform every other program (Hoynes & Patel, 2018). Third — no one is arguing the cliff is the sole cause. But a system that makes marriage cost $10,000 to $20,000 is not a neutral factor. It is an active force. It is documented, quantified, and ignored by both parties (Urban Institute, 2021; HUD, 2023).
How Old Is Your Body, Really?
The same data-driven rigor behind this article powers the Real Bio Age assessment. It measures your biological age across 12 health domains with peer-reviewed science.
Try 10 Free Bio Age Questions →Top 5 Solutions That Are Already Working
1. Finland Basic Income Experiment (Finland). Finland ran the world’s first nationwide basic income experiment from 2017 to 2018. Two thousand unemployed citizens received 560 euros per month with no conditions. They kept the money even if they found work. This is the exact opposite of a welfare cliff. The results were striking. Recipients did not stop looking for jobs. They reported much better health and lower stress. Only 17% experienced high stress versus 25% in the control group. The experiment proved something. When you remove the punishment for earning more, people do not become lazy. They become healthier and more stable (Finnish Ministry of Social Affairs and Health, 2020; University of Helsinki, 2020).
2. GiveDirectly Universal Basic Income Study (Kenya). In rural Kenya, GiveDirectly launched the world’s largest UBI trial in 2017. About 23,000 people across 195 villages receive $22.50 per month unconditionally. Recipients did not reduce their work. They invested and became more entrepreneurial. Hunger rates fell from 68% to 57%. The lesson for the American welfare cliff is direct. Unconditional support does not create dependency. Conditional cliffs create dependency (GiveDirectly, 2023; NPR, 2023; J-PAL Evaluation, 2023).
3. Year Up Workforce Development Program (United States). Year Up takes low-income young adults aged 18 to 29. It puts them through six months of technical training. Then they do a six-month corporate internship. A controlled trial showed graduates earned 30% more than their peers seven years later. That is an increase of $8,251 per year. The program has served over 36,000 students. This matters for the welfare cliff. Year Up launches people past the benefit phase-out zone entirely. Graduates leap to incomes where the trap no longer applies (PACE Evaluation, Abt Associates/MDRC, 2022; What Works Clearinghouse, 2023).
4. Cleveland Evergreen Cooperatives (United States). In Cleveland, Ohio, a network of worker-owned cooperatives exists. They are anchored to the purchasing power of major hospitals and universities. Workers earn over $28,000 a year. They build a $65,000 ownership stake over seven years. They get profit-sharing and free medical benefits. In 2023, the cooperatives distributed $1.5 million in profits to employee-owners. Worker ownership changes the cliff equation. When you own a share of the business, your wealth grows through equity. This bypasses the benefit limits designed around hourly pay (Shelterforce, 2021; US News, 2016; Triple Pundit, 2025).
5. M-Pesa Mobile Money (Kenya). M-Pesa turned basic mobile phones into banking tools. Research in Science found it lifted 194,000 households out of extreme poverty. It shifted 185,000 women from subsistence farming into business. Financial inclusion in Kenya rose from 26.7% in 2006 to 82.9% in 2019. The parallel to the American welfare cliff is instructive. M-Pesa did not phase out as users earned more. It did not punish savings. It did not penalize growth. It simply provided a financial tool. That frictionless design is why it reached 96% of Kenyan households (Suri and Jack, Science, 2016; Harvard Business School Case Study).
The Bottom Line
The numbers tell a story that no political narrative can override.
- 60 to 100%+ — effective marginal tax rates faced by low-income families in benefit phase-out ranges (CBO, 2015)
- $10,000 to $20,000 — annual net loss from marriage for families in the $20K to $40K income range (Urban Institute; Brookings)
- 48% — share of Section 8 tenants who are Black, vs. 13% of the population (HUD, 2023)
- 5.6 million — people lifted above poverty every year by the EITC, the one program designed without a cliff (Hoynes & Patel, 2018)
- 50 years — the length of time we have known how to design programs that do not trap people, and chosen not to
The welfare cliff is not a bug in the system. It is a feature. It is documented by the CBO, the Urban Institute, and a dozen academic institutions. Everyone who makes policy in Washington knows about it. No one fixes it. Fixing it would require both parties to abandon politically useful positions.
The families trapped on the edge are far more often Black and female. They are invisible to the people who make the rules. They are not lacking ambition. They are making rational decisions within an irrational system. Every year we fail to redesign that system is another year of mothers turning down promotions. It is another year of couples refusing to marry. It is another year of children growing up in poverty their parents could have escaped. They could escape if the government had not made escaping it a worse financial decision than staying.