There is a phrase bankers use when the marketing team is not in the room: "the overdraft economy." It refers, pleasantly enough, to the portion of a bank's revenue generated by people accidentally spending money they don't have. In 2025, that portion came to roughly $8 billion in the United States — down from its pre-CFPB peak, but still, by any honest measure, a staggering number.
Overdraft fees are the quiet tax of modern banking. They are charged, almost without exception, on customers who can least afford them. They are triggered by amounts that are often smaller than the fee itself — a $4 coffee that costs $39 by Wednesday morning. And they are, in the aggregate, the single most profitable consumer product on the retail banking menu.
This essay is not a moral argument. It is a structural one. We want to walk you through how the $167 national household average actually gets built — mechanically, line by line — and then explain why a different product, earned wage access, operates on an entirely different set of rails.
Part I · The mathHow the average gets built.
The most striking thing about the overdraft economy is not the total dollars. It is the concentration. The CFPB's 2024 deep-dive found that 8% of customers incur 80% of all overdraft fees. The median household in that 8% pays roughly ten overdraft events a year — which, at $35 a pop, lands right around the national average.
These are not "power users" of overdraft. They are people whose paychecks arrive on Friday and whose rent hits Thursday. The product they are charged for — temporarily spending $14 more than they have — is one that a bank's own systems could resolve, at a cost of essentially zero, by simply waiting twenty-four hours.
An overdraft fee is charged when your bank decides to cover a transaction that exceeds your balance. An NSF fee is charged when your bank declines it. Either way, the fee is roughly $35. You pay for the favor or you pay for the refusal. This is not an accident of design.
Part II · The alternativeWhy earned wage access is not a payday loan.
A pay advance — or, in regulatory language, earned wage access — is a fundamentally different product. It is not credit. It does not accrue interest. It does not have an APR. You are not borrowing. You are, by the most literal definition, withdrawing money you have already worked for but have not yet received.
The CFPB recognized this distinction in its 2020 advisory opinion, which drew a bright line between earned wage access models and traditional payday lending. The short version: if a product has no recourse against the user — no debt collection, no credit reporting, no finance charge — it is not a loan.
If a product has no finance charge, no recourse, and no interest — it is not a loan. It is a payroll-timing adjustment.
This matters, because the comparison most people reach for when they hear "pay advance" is the wrong one. Payday loans average 391% APR. Earned wage access has no APR at all, because it has no interest. The economic equivalent is closer to a company letting you clock out at 3pm on Friday instead of 5pm — a timing shift, not a financial product.
Part III · The questionThe one question that determines whether an advance is helping you.
We've spent a lot of time with our team asking customers — and potential customers — a single question: if you took an advance this week, what would have happened otherwise?
The answers cluster into roughly three groups:
- "I would have overdrafted." In our data, this is about 60% of the advances Kronos issues. The math is stark: a $35 overdraft fee on a $14 shortfall is a 250% effective rate. A $0 advance is... $0.
- "I would have used a credit card." Roughly 25% of advances. Credit card cash advance APRs average 29%, with no grace period. A no-fee advance paid back on next deposit is, again, a structural improvement.
- "I would have gone without." The remaining 15%. This is the group where advances genuinely expand access — people who couldn't get credit at any price, who now have a way to smooth a lumpy income.
The one question the customer should ask themselves is the same one: what would have happened otherwise? If the answer is "I'd have overdrafted," or "I'd have swiped credit," the advance is a cost reduction. If the answer is "I'd have done nothing," the advance is a new decision — and decisions should be made carefully.
A hard rule we wrote down early
When we were designing Kronos, we made one rule we were not willing to negotiate: no advance would ever cost more, in any dimension, than not taking it. That meant:
- No interest, ever. Not a "low" rate. Not a "promotional" rate. Zero.
- No tip. The tip model in some earned wage access apps is, in our opinion, interest wearing a funny hat.
- No penalty for late repayment. If your deposit doesn't land, the advance waits.
- No credit reporting. An advance is not credit, and we will not pretend otherwise.
These rules cost money. Specifically, they cost Kronos several revenue streams that are standard in the category. We think that's fine. If an advance only makes economic sense when it is priced like a short-term credit product, it shouldn't exist as a banking feature.
Part IV · In conclusionThe real fix is not a better fee.
Overdraft fees are not an accident. They are a design choice made by banks that, in most cases, could afford to make a different one. And the alternatives — earned wage access, real-time low-balance alerts, simple cushion features — are cheap and well-understood.
The quiet violence of the overdraft fee is that it is charged, almost invisibly, on people who are already having a bad week. The alternative is not a better-priced fee. It is a financial product that simply refuses to charge one at all.
That is what we've tried to build, and we'll keep writing about how it's going.