The Liquidity Gap: Payday Lending and the 36% War
The Liquidity Gap: Payday Lending and the 36% War
Somewhere in America, roughly every second of the business day, someone borrows a few hundred dollars at an annualized rate near 400% — and four out of five of those loans will be rolled into another. Payday lending is the market's answer to a question the mainstream system declines to answer: who lends $400 to a person the score can't see? The war over that answer has a number — 36% — and a map: twenty states have effectively banned the product, thirty still host it, and a wave of substitutes is now competing for the same desperate moment. This report is the full economics of the liquidity gap.
In this report
The core thesis
Payday lending is best understood not as a credit product but as a toll booth on the liquidity gap — the space between an unexpected $400 expense and a household with no buffer, no scoreable file, and no mainstream option. The product's defenders are right that the demand is real: the gap exists, and something fills it. Its critics are right that the product is engineered for repetition: a single balloon payment due in two weeks from a borrower who couldn't cover $400 this paycheck is structurally unlikely to be covered from the next one either — which is why roughly 80% of payday loans are rolled over or re-borrowed, and why the business model's profits live in the rollover, not the loan.
Our thesis ties this report to the three that precede it in our liquidity series: payday, overdraft, and the alternatives now crowding in (earned wage access, bank small-dollar products) are all prices on the same underlying condition — the illiquidity of tens of millions of households, compounded by the credit invisibility that locks them out of cheaper money. The 36% war is therefore necessary but not sufficient: caps close the toll booth, but only credit-file construction and cheaper substitutes close the gap. The endgame worth watching isn't regulatory — it's substitutional: the moment the $400 emergency is routinely met by a 28% credit-union loan, a flat-fee bank product, or a wage advance, the 400% product dies of competition in the thirty states where law never touched it.
The rate cap closes the toll booth. Only cheaper money closes the gap — and for the first time in the product's history, cheaper money is actually showing up.
The machine: how payday economics actually work
The unit mechanics: an average loan around $350–$400, a fee of $15–$20 per $100, full balloon repayment in about two weeks, secured by a post-dated check or — overwhelmingly now — direct debit access to the borrower's bank account. Annualized, the typical structure runs ~391% APR, ranging by state law from 140% to over 660%. The industry's standard defense — APR misleads on a two-week product; it's "just" $15 per $100 — meets an empirical wall: the CFPB's foundational research found 80% of loans rolled over or followed by another within weeks, meaning the modal borrower experiences the product not as a two-week bridge but as a months-long rotation, paying the fee again each cycle. A borrower rolling a $400 loan through a year pays more in fees than the principal — several times over in the highest-rate states. Aggregate scale: in the thirty permissive states, borrowers took over 20 million loans (~$8.6 billion) in a recent year and drained roughly $2.4 billion in fees — storefront alone, before online volume. Note also the plumbing: repayment via bank-account debit means a failed payday debit cascades into the NSF and overdraft fees we've already tallied — the two toll booths are queued on the same road, often charging the same driver in the same week.
Who borrows — and why the score is the story
The borrower profile maps almost perfectly onto the populations this series keeps finding at the system's edge: usage concentrates among adults in their 20s–40s in modest-wage, variable-hour work; Black and Hispanic adults use payday products at roughly three times the rate of white adults (about 10–11% versus 3% in recent survey years); and the near-universal precondition is a thin, damaged, or absent credit file — the ~32 million unscoreable adults plus the deep-subprime tier just above them. This is the analytical heart of the matter: payday lending doesn't underwrite creditworthiness, it underwrites access to your paycheck — the bank-account debit is the collateral, which is why no credit file is needed and why no credit file gets built: payday lenders generally don't furnish positive history to the bureaus. The product is thus a closed loop with the invisibility problem — it requires no file, builds no file, and leaves the borrower exactly as unscoreable after ten loans as before the first, minus the fees. Any serious exit from the loop runs through the file-construction ladder we mapped in the reporting pipeline and the invisibility report.
36%: how a number became a battle line
Why 36%? Genealogy: it descends from early-twentieth-century small-loan model laws, was enshrined federally by the Military Lending Act — Congress capped loans to service members at 36% in 2007 after the Pentagon concluded payday debt was a readiness problem — and has since become the consensus ceiling advocated for everyone else: the rate at which small-dollar lending can be honest but not predatory. The map today: roughly 20 states plus DC cap around 36% or lower (several — New York, New Jersey, Massachusetts, Vermont, Arkansas — much lower), a list that has grown steadily by legislation and, tellingly, by ballot measure: when voters decide directly, caps win overwhelmingly regardless of the state's partisan lean (Nebraska's 2020 initiative passed with over 80%). Since 2005, no state has newly authorized storefront payday lending — the map only moves one direction, just slowly. Federally, the story is quieter: the CFPB's ability-to-repay rule was gutted before taking effect (its payment-withdrawal provisions survived into 2025), and a national 36% cap — extending the MLA to everyone — is perennially introduced and perennially stalled. The battle line, like everything in our EWA and scam-liability coverage, runs through the states.
What caps actually do
The supply effect is unambiguous: at 36%, the traditional storefront model doesn't pencil, and lenders leave — Oregon's licenses fell from 346 to 82 within about a year of its cap. The contested question is the demand side: where does the $400 need go? The evidence, honestly summarized: cap advocates point to post-cap surveys showing former borrowers coping via alternatives (payment plans, family, credit unions, cutting expenses) and reporting better financial outcomes; skeptics point to migration — online lenders, tribal-sovereignty models, and out-of-state structures that caps reach imperfectly — and to the access argument: the riskiest borrower loses her worst option without automatically gaining a better one. Both effects are real; the balance of research suggests the debt-trap harm avoided exceeds the access harm imposed, which is why the map keeps moving and why even the industry has largely stopped contesting 36% for the smallest loans, retreating instead to fee-loaded installment structures in the $1,000–$10,000 range — the next front, as state installment-rate surveys document. The deepest cap critique, though, isn't the industry's — it's ours from the thesis: a cap without substitutes is subtraction, not solution. Which brings us to the substitution wave.
The substitution wave
| Substitute | Typical cost | What it solves / limits |
|---|---|---|
| Credit union PALs | Capped at 28% APR | Regulated, installment-structured, credit-building potential; requires membership, modest scale so far |
| Bank small-dollar loans | Flat fees — commonly ~$5 per $100, repaid in installments | Several major banks now offer them to existing customers; a fraction of payday cost, underwritten on account history — cash-flow underwriting in production |
| Earned wage access | Free-to-modest fees | Solves timing gaps against wages already earned; doesn't solve expense > income, and heavy use has its own loop risk |
| Employer/nonprofit programs | Low to zero | Real where offered; coverage is spotty |
| File construction | — | The structural exit: a scoreable file converts the next emergency from a 400% problem into a mainstream-credit problem |
The strategic read: for the first time, the payday moment has genuine competition inside the mainstream system. Bank small-dollar products — priced at a twentieth of payday cost and underwritten on deposit-account behavior rather than credit files — are the most important entrant, because they attack the product where caps can't reach: in the thirty permissive states, on price. The wildcard is whether these programs scale beyond existing customers in good standing, or remain a perk for the almost-fine while the deep gap stays tolled.
Scenarios and what we're watching
| Scenario | Shape of the world | Signposts |
|---|---|---|
| Base case — the slow squeeze | The cap map adds a state or two per cycle; storefront volume keeps shrinking; industry migrates to high-rate installment and online; substitutes grow but don't dominate | New cap enactments/ballots; installment-lending rate fights; storefront license counts |
| Bull case (for borrowers) — substitution wins | Bank small-dollar and EWA scale into the payday moment on price; cash-flow underwriting reaches the unscored; payday shrinks even in permissive states; the fee drain collapses | Bank small-dollar origination volumes; credit-union PAL growth; payday volume in non-cap states |
| Bear case — the gap re-tolls | Consumer strain rises faster than substitutes scale; online/high-rate installment absorbs demand at payday-like effective costs; the toll booth just changes signage | Online high-cost volume; installment APR survey trends; delinquency at the subprime edge per the cycle report |
What we're watching: the next ballot measures (voter caps are undefeated); the installment-lending front in state legislatures; bank small-dollar scale (the substitution thesis's proof point); whether any payday-adjacent product begins furnishing positive credit data (the loop-breaker); and the federal 36% bill's perennial reintroduction. The liquidity gap is the most regressive market in American finance — but for the first time in its century of history, the exits are being built faster than the toll booths. Which one wins the next decade is, in large part, what this research desk exists to track.
Frequently asked questions
$15–$20 per $100 for a two-week balloon annualizes to ~391%+ APR. The industry calls APR misleading for short terms; the CFPB's finding that 80% of loans roll over means heavy users experience the annualized rate quite literally, cycle after cycle.
Roughly 20 states plus DC cap around 36% or lower or otherwise bar high-cost payday lending — with recent additions by legislation and ballot (Nebraska 2020, Illinois and Hawaii 2021, New Mexico 2023, Minnesota 2024). Elsewhere, single-payment APRs run ~140% to 660%+.
Storefront supply largely exits (Oregon: 346 licenses → 82). The debate is demand-side: advocates document borrowers shifting to cheaper coping; skeptics document online/tribal migration and lost access. The evidence tilts toward net benefit — which is why the map only moves one way.
Credit-union PALs (28% cap), bank small-dollar installment loans at flat fees, earned wage access for timing gaps, employer and nonprofit programs — and structurally, building a scoreable credit file, since payday dependence is largely an invisibility symptom.
Key takeaways
- Payday is a toll booth on the liquidity gap: real demand, met by a product engineered for rollover.
- The economics live in repetition — 80% of loans roll, draining $2.4B/year in fees from the poorest borrowers.
- The product requires no credit file and builds none — a closed loop with credit invisibility.
- 36% is the century-old consensus ceiling; twenty states enforce it, voters never reject it, and no state has expanded payday since 2005.
- Caps close toll booths; only substitutes close the gap — and bank small-dollar products are the first real price competition in the product's history.
This report is for general information only and does not constitute financial or legal advice. Figures are drawn from publicly reported sources, principally CFPB, CRL, and NCLC research, and state laws change over time.