Do People Really Use Benefits to Buy Drugs?
Vouchers and cash transfers lead to different spending patterns.
- By
- August 04, 2026
- CBR - Economics
The One Big Beautiful Bill Act of 2025—while slashing funding for food, healthcare, and higher-education subsidies—expanded the Child Tax Credit to put more cash into the pockets of low-income citizens. In theory, any money refunded through the CTC can be put toward groceries, medical bills, college tuition, and other needs, and yet various critics have expressed concern over the years that it won’t be. In 2021, then-Senator Joe Manchin (Democrat of West Virginia) helped kill a Biden-era expansion of the CTC, reportedly concerned that poor parents would use their cash refund to buy drugs.
His qualms were not entirely unfounded, suggests research by Icahn School of Medicine at Mount Sinai’s Anna Chorniy, MIT’s Amy Finkelstein, and Chicago Booth’s Matthew J. Notowidigdo. Among some people, cash transfers can lead to higher spending on “temptation goods” such as liquor and lottery tickets, the study finds. However, it also suggests that this tendency can be counteracted by coupling cash transfers with in-kind ones tied to targeted necessities such as food or housing.
When it comes to supporting the poor, American policymakers have a history of preferring in-kind transfers to cash. The Clinton administration’s welfare overhaul in 1996 largely did away with money transfers to able-bodied people younger than 65.
In theory, cash allows benefit recipients to optimize their spending. Over several decades, economists have sought to understand and explain why, then, the policy preference for restricted benefits.
Chorniy, Finkelstein, and Notowidigdo add to the body of research by exploring the notion that policymakers act on a paternalistic impulse, concerned that cash will be spent inappropriately. The study involved a head-to-head comparison of the impact of cash and in-kind benefits among the same individuals using 20 years of data for half a million low-income people in South Carolina who received both Supplemental Security Income and Supplemental Nutrition Assistance Program benefits. SSI provides cash assistance (and often access to Medicaid, a federal healthcare program) to people who are elderly or disabled. SNAP, by contrast, offers vouchers to be spent on food.
The researchers obtained records from the state and Medicaid that allowed them to track the benefits granted to each person and tie them to those beneficiaries’ spending on medical care in the days following the receipt of each type of assistance. (SSI usually pays out on the first day of each month, while the timing of monthly SNAP benefits in South Carolina during the period studied varied by recipient.)
The spending differences that resulted from SSI and SNAP payments were dramatic. After the SSI program made its monthly cash disbursements, recipient spending on emergency-room visits to treat symptoms of drug use and alcoholism (used as a proxy for excessive drug and alcohol consumption) jumped by 20–30 percent, on average. By contrast, payments of SNAP benefits resulted in no such increase.
Cash and food aid led to different behaviors
SSI benefits also led to increased consumption of newly prescribed drugs, which the researchers used as a stand-in for goods that were neither temptation related nor covered by SNAP. In the week after people received SSI payments, their spending on new fills of prescription drugs increased by about 20–40 percent. SNAP benefit distributions resulted in no such additional purchases.
These patterns reflect human psychology more than economic theory, Notowidigdo says. While SNAP recipients have to spend the vouchers on food, essentially the benefits function like cash added to the budget and should free up money for other purchases. Therefore SNAP should increase spending on temptation goods in the same way that SSI does.
Instead, the researchers write, the results suggest that recipients treat SNAP benefits differently. In their mental accounting, cash and in-kind benefits are not interchangeable. That dovetails with existing evidence, from University of Washington’s Justine Hastings and Harvard’s Jesse M. Shapiro, that the marginal propensity to consume groceries is higher out of SNAP benefits than cash—an indication that SNAP may make recipients more likely to spend freed-up cash on food.
Consistent with that observation, “we actually find that payments of SNAP benefits reduced nutrition-sensitive conditions that led to emergency room visits,” says Notowidigdo, citing hypoglycemia (low blood sugar) and hypertension as two examples.
One way of discouraging beneficiaries from buying temptation goods is through “sin” taxes that make such purchases more expensive. But retail taxes are blunt instruments that raise prices for everyone, not just recipients of government assistance. Also, sin taxes don’t capture purchases of illegal drugs.
The researchers find through modeling that it would be more efficient to structure benefits so that they include both cash and in-kind transfers. “In-kind transfers can have better targeting properties,” says Notowidigdo. By including both kinds of assistance, policymakers can encourage wise spending without imposing or raising taxes or knowing exactly who is more likely to spend cash on temptation goods.
- Anna Chorniy, Amy Finkelstein, and Matthew Notowidigdo, “Paternalistic Social Assistance: Evidence and Implications from Cash vs. In-Kind Transfers,” Working paper, November 2025.
- Justine Hastings and Jesse M. Shapiro, “How Are SNAP Benefits Spent? Evidence from a Retail Panel,” American Economic Review, December 2018.
More from Chicago Booth Review
More from Chicago Booth
Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.