Authored by Vladimir Snurenco via the American Institute for Economic Research (AIER),
Benefit cliffs occur when earning slightly more—through a raise or extra hours—leaves people worse off because they lose government benefits. Many economists believe these cliffs can be “smoothed out” through better program design. But this view overlooks a critical factor: human psychology. To truly address benefit cliffs, policymakers must move beyond formulas and graphs and consider how people actually think, feel, and behave. In some cases, benefit cliffs may not just be hard to eliminate—they may be impossible to fully remove.
The reason why so many experts and economists worry about benefit cliffs is that we want to have a welfare system that’s not just a safety net, but a springboard. Modern-day welfare in America includes over 80 programs such as Head Start, Medicare, Medicaid, housing vouchers, food stamps, and the like. Ideally, these programs shouldn’t just catch people when they fall—they should help launch them forward. The great tragedy is when they instead work more like quicksand and hold people back. While benefit cliffs are typically analyzed as purely mathematical problems, a full understanding requires recognizing powerful psychological realities.
Benefit Cliffs: A Visual and Emotional Drop
For many people, the term ”benefit cliff” sounds abstract or technical—something you’d find in a policy paper or economic model. But the reality it describes is deeply personal and practical. The word ”cliff” was chosen for two reasons. First, it captures what happens visually on a graph: the income line suddenly dives downward when a person earns a bit more money but loses more in benefits than they gain in wages. Second, it reflects how that moment feels—like standing on the edge of a steep drop. The fear of falling, of suddenly losing vital support, can drive people to act in ways that may seem irrational but are emotionally understandable: turning down job offers, quitting after a pay raise, or avoiding overtime altogether. These decisions, though rooted in self-preservation, often end up blocking real paths out of poverty. Matt Paprocki, president of the Illinois Policy Institute, puts it bluntly: 70 percent of people who begin receiving welfare benefits stay on them for life—eventually dying while still dependent on the system.
The Mathematical Perspective: Economists’ View
So, if benefit cliffs are real and if they actually trap people, can they be fully eliminated? Many experts believe so, but I wouldn’t be so optimistic. There are two ways to understand benefit cliffs: mathematically and psychologically.
First, let’s have a look at a technical, mathematical definition of benefit cliffs. According to a report by the Institute for Research on Poverty at the University of Wisconsin-Madison, benefit cliffs occur when an income increase triggers a loss of public benefits equal to or greater than the additional earnings. Economists refer to this as a marginal tax rate exceeding 100 percent—earning an extra $100 results in losing more than $100 in benefits.
The Georgia Center for Opportunity created an interactive tool that models how income changes affect eligibility for various state and federal benefits (see BenefitCliff.org for details). Below you can see a graph from their report, “Disincentives for Work and Marriage in Georgia’s Welfare System.” Every dip you see in this graph is a benefit cliff.
This graph models the statewide average scenario of a single mom, age 30, with a 10-year-old girl and a 2-year-old boy.
It shows that as her income increases from $20,000 to $30,000, she experiences several abrupt drops in her budget as housing vouchers, subsidized childcare, and medical insurance are reduced or completely taken away.
Psychological Reality: Humans vs. Models
I find it striking, however, that experts can pat themselves on the back and celebrate the elimination of benefit cliffs as soon as someone merely breaks even and starts earning just slightly more than they lose. For example, let’s say our single mother works overtime and earns an additional $100. If this causes her to lose $75 in reduced food stamps, many economists would declare that she is now $25 better off, and the benefit cliff successfully eliminated.
That conclusion might make sense to economists, but not to our subject—a 30-year-old single mother of two. From her point of view, she just worked extra hours and lost most of the payoff.
The disconnect between how our subject views the situation, and how economists understand it, can be attributed to Rational Choice Theory, a model through which many economists see the world. Rational Choice Theory assumes that people always make logical, rational choices to maximize their personal gain. But this model fails to incorporate the realities of human behavior.
Consider the well-known Ultimatum Game: one person proposes how to divide $100, and the other person decides whether to accept or reject the offer. Rational Choice Theory predicts that the responder should accept even $1—after all, $1 is better than nothing. Yet in practice, people often reject offers they consider unfair, even if it means walking away with nothing. Fairness, dignity, and emotion matter. These are powerful factors that classical economic models often ignore.
Yes, our single mom might technically be $25 ahead, but I have never met a mother who works overtime, makes $100, learns she’s lost $75, and feels satisfied. A person who feels like they’ve lost most of their paycheck is not likely to stay motivated, even if they break even or come out a few dollars ahead. Experts who insist that this counts as a win are living in a theoretical world, not the one the rest of us inhabit. They celebrate technical “successes” like smoothing out the benefit cliff, but for the single mom, there’s very little to celebrate.
To see the situation through the eyes of an economist, consider the following quote from “Disincentives for Work and Marriage in Georgia’s Welfare System”:
“One principle that cannot be violated is that a family must always be better off from earning more money. Mathematically, this is easy to explain. The marginal benefit from earnings and subsidies must always be more with increased earnings. Graphically, it is easy to show because the line combining earnings and subsidies must always have a positive slope. If at any time the slope is negative, then there is a welfare cliff.”
Notice the technical language and emphasis on graphs. Yes, the slope may be positive. Yes, the math checks out. But while economists see a line bending upward, our single mother still feels like she’s falling. And that feeling matters. As one mother relying on government assistance explained to me, “It feels like no matter how much harder I try, I’m always being pulled backward.”
Loss Aversion and Real-World Choices
Let’s use a different lens and look at benefit cliffs from a psychological point of view. Loss aversion—a well-documented behavioral response—tells us that people feel the pain of loss two to three times more strongly than the pleasure of an equivalent gain.
How would you feel if you found a $100 bill on the ground, and then 15 minutes later got a flat tire that cost exactly $100 to fix? According to Rational Choice Theory, you should feel neutral—the gain and the loss cancel each other out. But in real life, you’d likely feel upset. That nail ruined your tire and whatever fleeting plans you might have had for that extra $100.
This psychological lens helps us better understand benefit cliffs. If people perceive losses two or three times more intensely than gains, even gradual reductions in benefits can feel like real setbacks. Losing benefits doesn’t motivate work—it demoralizes.
And things get even more complicated when you consider that many recipients don’t have PhDs or the skills to build graphs, interpret benefit curves, or navigate a maze of programs. Ordinary people have no idea whether their personal income graph has a slightly positive or slightly negative slope. (Have you ever seen anyone, when offered a raise, say: “Wait a minute! Let me first build my income graph and see how the raise would affect its slope?”)
What people do know—intuitively and from lived experience—is that welfare programs are complex and uncoordinated. Even experts struggle to make sense of them. So it’s no surprise that the mathematical explanations of benefit cliffs we often see in reports and policy papers have little to do with how people actually operate and make decisions—a disconnect that is as striking as it is overlooked.
The Savings Trap: Surveillance and Psychology
Just as benefit cliffs discourage work, the structure of modern welfare programs also discourages saving—creating a parallel psychological and bureaucratic trap. Experts often argue that modern welfare no longer disincentivizes saving because many asset tests that determine benefit eligibility have been relaxed or removed. But even when not actively penalized for holding assets, applicants are still routinely required to report any change in assets—how many bank accounts they have, how much cash they keep, what kind of car they drive, whether they’ve sold anything, even how much they were paid for donating blood. In Michigan, for example, recipients must report changes in income or assets within 10 days of them occurring through the MI Bridges portal online. These rules create an atmosphere of surveillance. From a psychological perspective, it is clear why many low-income families avoid formal saving—not because they don’t want to build an emergency fund or earn interest, but because they fear losing what little help they have. This locks them out of the banking system, deters wealth-building, and keeps them stuck in financial precarity. In this sense, the system is not just unhelpful—it’s actively disincentivizing the very behaviors that help people climb out of poverty.
Toward a Truly Empowering Welfare System
In the end, as long as benefit cliffs are treated primarily as a math problem, we will continue to miss the mark. Benefit cliffs are not just technical design flaws—they are psychological traps. Once government assistance is extended, the experience of losing it—even gradually—feels like a punishment, not progress. This emotional reality is stubborn, deeply human, and nearly impossible to erase through policy tweaks alone. Yet many experts continue to celebrate easy fixes: a graph with an upward slope, a marginal tax rate below 100 percent, a rule relaxed. But these celebrations ring hollow when they fail to account for how loss aversion distorts perception and demoralizes effort. As long as the system makes people feel like they’re falling—even when the math says they’re rising—the cliff remains. A truly effective welfare system must begin not with a chart, but with a sober recognition of the profound difficulty of the task. Only then can we begin designing programs that truly help the poor—without hurting or trapping them along the way.
