Being poor costs money. Not in the metaphorical sense, but as a set of measurable price differences: the same groceries cost more at the store you can walk to, the same car costs more to finance, the same $100 costs more to borrow, and the same checking account charges you for running out of money. Researchers call it the poverty premium, and every component of it is documented in federal data.
The premium is not a moral failure on either side. It is what happens when prices track risk, volume, and access, and a household lacks all three.
The groceries
A household without a car shops where it can reach. Prices there run higher for identical items.
Researchers led by Melissa Laska measured this directly in the Minneapolis and St. Paul area, publishing in the International Journal of Environmental Research and Public Health in 2017. They matched staple items in small food stores against the nearest supermarket carrying the same product. Every item except white bread cost more in the smaller store. Cheerios ran 54 percent higher, bananas 53 percent, and white rice 50 percent.
Earlier USDA Economic Research Service work found smaller gaps against convenience stores specifically, with milk about 5 percent higher, ready-to-eat cereal about 25 percent, and bread about 10 percent. The same agency added a caution worth repeating: convenience stores account for only 2 to 3 percent of food spending among low-income consumers, so the premium is real but not the whole grocery bill.
Both studies are old enough that the exact percentages should be treated as illustrative. The direction has not been contradicted.
The bank account
The FDIC’s National Survey of Unbanked and Underbanked Households found 4.2 percent of U.S. households unbanked in 2023, about 5.6 million households, the lowest rate since the survey began in 2009. Another 14.2 percent, roughly 19 million households, were underbanked. Among unbanked households, 66.2 percent relied entirely on cash.
The reason households give is the point. In the 2021 survey, the most cited main reason for not having an account was not having enough money to meet minimum balance requirements, at 21.7 percent. The product requires a cushion, and the households most in need of the product are the ones without one.
Cash reliance carries its own charges. Check cashing takes a percentage. Money orders cost a fee per transaction. Neither builds a record a lender will later accept.
The overdraft
Having an account does not remove the penalty for running out of money. Bankrate’s 2025 checking account fee study, based on 245 institutions surveyed in June and July of that year, put the average overdraft fee at $26.77, charged by 94 percent of accounts examined. The average non-sufficient funds fee was $16.82.
The Consumer Financial Protection Bureau found the cost concentrates sharply. Consumers paying more than ten overdraft fees a year account for nearly three quarters of all overdraft fees, and those frequent overdrafters paid about $380 across a year. The bureau also reported that ninety percent of frequent overdrafters typically held no more than a few hundred dollars in their accounts at the end of any given day.
A fee of $26.77 on a $12 shortfall is not priced to the cost of covering the shortfall. It is priced to the fact that the customer has no alternative at that moment.
The loan
Short-term credit prices the same way. The Consumer Financial Protection Bureau reported in 2016 that the median storefront payday loan carried a fee of $15 per $100 borrowed on a median 14-day term and a median loan size of $350, producing an annual percentage rate of 391 percent.
The APR looks absurd because the term is short, and defenders of the product say so. The bureau’s own data answers that defense. Its 2014 analysis of more than 12 million storefront loans found that over 80 percent were rolled over or followed by another loan within 14 days, that half of all loans sat in a sequence at least ten loans long, and that 22 percent of new loans were renewed six or more times. A borrower renewing six times has paid more in fees than the amount originally borrowed.
Two-week pricing applied to a months-long borrowing pattern is the mechanism, and the pattern is the norm rather than the exception.
The food budget that does not stretch
Federal food assistance carries a version of the same timing problem. The USDA’s Thrifty Food Plan priced a month of groceries for its reference family of four at $1,024.90 in July 2026. The maximum monthly SNAP allotment for a household of four in the 48 contiguous states for fiscal year 2026 is $994.
The benefit is set once a year from the June cost of the plan and then held fixed. Food prices keep moving. By July the government’s own estimate of what the basket costs had run about $31 a month ahead of the maximum benefit built to buy it.
Why the premium persists
Each price above is defensible in isolation. A small store buys in smaller volume and pays more per unit. A lender facing higher default rates charges more. A bank setting a minimum balance is describing what the account costs to service.
Stacked together, they mean a household earning less pays more for the same goods and credit, which leaves less to escape with. Fight For A Living Wage, a nonpartisan grassroots 501(c)(3), frames the broader problem as affordability across housing, healthcare, childcare, food, transport, education and retirement rather than any single wage figure. The poverty premium is the version of that argument visible at the cash register.
Anyone wanting to check these figures can do it from the source. The FDIC publishes the unbanked survey, the Consumer Financial Protection Bureau publishes the overdraft and payday research, and the USDA publishes the Thrifty Food Plan cost every month. The numbers are not contested. What gets contested is whether a price that tracks risk is therefore the right price, and that argument belongs in public rather than in a fee schedule.
