A Bank Told Its Own Customers It Had a Better Account. Almost Nobody Switched.

In a field experiment across five UK banks, researchers picked out savers who were sitting on a strictly worse deal than one their own bank already offered. Same provider. No new application, no credit check, no paperwork to chase down from somewhere else. All the bank had to do was tell them. So, for the purposes of the experiment, it did.

The switch took about 15 minutes. The average saver who moved would have picked up roughly $190 a year, according to the study. And still, across every version of the disclosure the researchers tested, from a plain notice to a pointed comparison, switching stayed rare. Most people who were shown, directly, that their own bank had something better for them left their money exactly where it was.

That finding comes from Paul Adams, Stefan Hunt, Christopher Palmer, and Redis Zaliauskas, in a randomized-controlled trial covering 124,000 savings-account holders, published in the Journal of Financial Economics in 2021 after circulating as an NBER working paper in 2019. It is one of the largest field experiments ever run on financial disclosure, and it is worth sitting with, because a rate comparison site's entire premise rests on the opposite assumption: that showing someone a gap is most of the work.

What this result rules out

Start with the excuses that would normally explain away a null result like this.

It is not friction. Fifteen minutes is not a meaningful barrier, and the paperwork requirement was close to zero since the better account sat at the customer's existing bank.

It is not information. The bank did not bury the better rate in fine print. In the strongest treatment arm, it told the customer directly.

It is not the size of the prize. $190 a year is not life-changing money, but it is not nothing either, and it recurs every year the money stays parked in the wrong account. Over five years, at a stable balance, that is close to $1,000 walked past for no cost at all to claim it.

Once friction, information, and stakes are all accounted for and the gap still doesn't close, what is left is the saver's own expectation of what will happen if they act.

The part that is not really about laziness

The researchers' own explanation is pessimistic beliefs: many savers assumed, going in, that switching would not be worth the trouble, or that an offer this easy had to come with a catch. That sounds like a personality trait. It is closer to a rational shortcut.

For a saver who has already concluded that reading a disclosure will not change the outcome, whether because every "better rate" offer looks like a trick, or because the real-world gain seems likely to be smaller than advertised, not reading it isn't laziness. It's the correct use of attention, given what that saver already believes going in. The problem was never that people can't do fifteen minutes of paperwork. It's that a lot of people have already priced in that the paperwork won't be worth it, often before they've seen a single number.

That distinction matters because it points to a different fix. A false prior cannot be out-formatted. A cleaner rate table does not touch it.

This is not a one-country result

The same pattern shows up somewhere with a completely different obstacle removed. Fernando Cirelli and Arna Olafsson studied transaction-level data from a major Icelandic bank serving roughly a third of that country's population, looking at how households split money across accounts that were identical in risk and maturity but paid different rates, with transfers that were instant and free. No fifteen-minute wait. No application. Just a phone screen and a tap.

Households still barely moved money toward the higher-paying option, with one exception: the wealthiest depositors, who were roughly ten times more responsive than everyone else and earned about two percentage points more on their liquid balances as a result. Two very different countries, two very different frictions removed (information in one case, transfer cost in the other), and the same basic non-response from most depositors either way. That is not a coincidence easily waved off as a fluke of one study's design.

Banks are not surprised by this

There is a reason this pattern persists at scale, and it is not that banks haven't noticed. Xu Lu and Lingxuan Wu, in research that won the 2026 Ieke van den Burg Prize for work on systemic risk, measured depositor inattention directly using millions of US bank accounts, by comparing how quickly people move money that arrives on a predictable schedule versus money that shows up unexpectedly. The lag on the unexpected money is the inattention.

Banks with more inattentive depositors, by that measure, set lower rates, pass through less of any Federal Reserve rate move, and see less money leave even when their rates fall further behind. In other words, the gap between what a bank could pay and what it does pay isn't an oversight on either side. It's priced. A bank that has a good read on how inattentive its depositor base is has a good read on how much it can underpay that base without losing it, and disclosure duties, marketing spend, and rate-setting are all shaped around that estimate.

Where SwitchWize has gotten this half right

SwitchWize's own reporting has made a version of this argument before, and it is worth naming directly rather than around. A July 2026 piece on this site, The Hidden Cost of Inertia, argued that the gap between mega-bank profits and what they pay depositors is a friction problem, and that friction problems get solved once someone makes switching easy enough.

That is half of the finding above, not the whole of it. Friction is real, and reducing it helps at the margin. But the Icelandic data says friction is not the load-bearing constraint, because removing nearly all of it still leaves most households in place. A rate table is a disclosure. The Bank Gap Index, at its core, is a disclosure: a number showing a reader what they are giving up by staying put. Both are necessary. Neither is sufficient, and treating a clean number as the finish line, rather than the start, is the mistake worth naming plainly.

What might actually move the number

None of this means disclosure is worthless. The UK study's own design shows some versions of disclosure worked a little better than others, just not nearly well enough. What the belief-based explanation points toward instead is intervention aimed at the belief itself, not just the information gap:

• Address the specific doubt before the number, not after it. If the working assumption is "this has to be a trick," a rate comparison that leads with proof of legitimacy (FDIC insurance status, a named source, a verifiable rate as of a specific date) before the dollar figure is answering the actual objection, not the one a rate table assumes.
• Make the act of switching visible and time-stamped, not just the opportunity. A friction-by-friction account of what an actual switch involved, start to finish, does more to correct a pessimistic prior than a bigger, bolder version of the same static number.
• Treat a one-time comparison as the weaker half of the job. A pre-committed trigger, decided in a calm moment rather than re-litigated every time a new rate appears, removes the need to re-evaluate the same pessimistic belief on every visit.

Those are directions, not a finished feature. What changes SwitchWize will actually make in response to this finding are being scoped separately, and this piece will link to that decision once it ships rather than promise it here.

The number that stayed on the table

Go back to the study's own strongest case: a bank telling its own customer, directly, that it had something better for them. Fifteen minutes. No catch to check for. $190 a year, every year it went unclaimed. Most people still left it exactly where it was.

That result is not really about that one bank, or the ones running this site's own numbers. It is a reminder that the number was never the missing piece. What was missing was a reason to believe the number was real, and worth fifteen minutes of anyone's afternoon, and that is a harder thing to build than a better table.

This article is educational and is not personalized financial, tax, or legal advice. Verify current rates, terms, and FDIC insurance status directly with any institution before acting. Full methodology and sources below.

Related reading: Even with free, instant transfers, most people still don't move their money, The Hidden Cost of Inertia, The Loyalty Tax, and run your own numbers on the Bank Gap Index.