Licensed Payday Lender Vacancies and Energy Assistance Participation: Evidence from Wisconsin

• In Event: Financial Fragility and the Price of Credit: Payday Lending, Safety Nets, and Mortgage Costs

Abstract

Background

The State of Wisconsin has limitless interest rates on payday loans and upholds one of the highest limits on the maximum amount one can borrow (mode limit: $500, WI: up to $1,500). Neighboring states Illinois and Minnesota impose 36% APR interest rate caps on payday loans. Contrarily, the average APR for payday loans in Wisconsin during 2025 was 599.51%. Despite these favorable conditions for payday lenders, the total licensed payday lending branches operating statewide have decreased from 423 branches in 2011 to just 48 branches currently. So, what do economically vulnerable consumers do once payday loans are eliminated from their financing option set? Since consumers often borrow payday loans to cover housing, food, and utilities, public assistance is a plausible alternative.

I examine how the decrease in payday lending operations in Wisconsin has affected participation in state-provided energy assistance programs through a two-way fixed effects differences-in-differences design exploiting cross-county by fiscal year variation in payday lender exits (i.e., vacancies). I define a “vacancy” as when all licensed payday lending branches located in a county have ceased operations.

I focus on Wisconsin given its persistently laxed payday lending regulations and stable eligibility requirements for medical, food, and energy assistance, none of which have changed since 2011. Wisconsin has not expanded Medicaid but has maintained socioeconomically inclusive eligibility criteria for food and energy assistance. Both programs have no asset limits, and both implement income limits at maxima allowed under federal law. For energy assistance, the income limit is 60% of the state’s median income which lies between 200% and 250% FPL. By holding policy parameters constant, I can pick up sterile effects of AFS market forces on consumer decisions.

I employ unique administrative data on licensed payday lender branch establishments in Wisconsin since 2011, and I employ public-use administrative data on energy assistance participation statewide by county during FY 2011-2023 from the Direct Benefits Caseload Comparisons. I employ data on county characteristics from the U.S. Census Bureau and the Bureau of Labor Statistics. Constructed outcome variables include number of households paid heating versus non-heating (electric) assistance. I estimate effects from linear fixed effect models incorporating county and fiscal year fixed effects with counties’ time-variant characteristics as covariates. I then estimate event study models to assess how overall effects fare dynamically. Preliminary evidence reveals that on average, payday lender vacancies increase benefit receipt for both heating and non-heating (electric) assistance. However, these effects are delayed.

Policy Significance

Wisconsin has inclusive food and energy assistance policies and laxed payday lending regulations, a rarity in credit-welfare state paradigms. Energy assistance participation plausibly increases post-vacancy, albeit delayed. These results are conditional on Wisconsin maintaining socioeconomically inclusive safety net parameters for energy assistance. This socioeconomic inclusivity is threatened amidst proposals to eliminate federal funding for LIHEAP. Furthermore, Wisconsin lawmakers have attempted to pass laws that would effectively eliminate payday loans. If policymakers simultaneously restrict AFS credit and safety net supports, then especially near-poor and moderate-income households may have trouble covering their short-term shocks.

Author

• Melody Harvey, University of Wisconsin-Madison
  Presenting Author