Interchange and free checking

By Felix Salmon
June 17, 2010

Interchange fees are a cross-subsidy too: this time it's merchants who help pay for the checking accounts of the rich. In fact, they do more than pay for their checking accounts, they pay them a nice tax-free income, when the rich people accept debit rewards cards.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Why do most people hate their bank? Because their relationship is based on the lie of “free checking”, and a relationship based on a lie is always going to be a dysfunctional relationship. Checking is never free, but in recent years banks have been able to conjure the illusion of free through a system of regressive cross-subsidies, where the poor pay massive overdraft fees and thereby allow the rich to pay nothing.

Interchange fees are a cross-subsidy too: this time it’s merchants who help pay for the checking accounts of the rich. In fact, they do more than pay for their checking accounts, they pay them a nice tax-free income, when the rich people accept debit rewards cards.

With the federal government finally cracking down on overdraft fees, and with the Durbin amendment threatening interchange fees to boot, the fiction of free checking looks as though it’s reaching the end of its natural life:

More than half of all checking accounts are currently unprofitable, according to a report issued last month by Celent, a unit of Marsh & McLennan Cos. It costs most banks between $250 and $300 a year to maintain one of the roughly 200 million checking accounts, according to industry estimates.

Checking accounts pay zero interest these days, but even being able to borrow money for free from depositors isn’t worth $300 per year to a bank. If a checking account has $1,000 in it on average, and the bank can lend that money out at 7%, net of defaults, then it’s making $70 a year on the account, which isn’t enough to cover its costs.

The natural answer, here, is to restart charging monthly fees on modest-balance checking accounts — and to shed few tears when your low-balance customers leave:

The offers of free checking without any minimum balance requirements attracted a new wave of low-income customers, who previously went to check-cashing stores. Some consumer advocates have warned that the elimination of free checking could drive some of those customers out of the banking system.

From the banks’ perspective, though, many of those customers aren’t profitable.

All of which provides some important background in understanding the stance of Patrick Adams, the CEO of St Louis Community Credit Union. Adams’s customers are relatively poor: his credit union is designated a Community Development Financial Institution, which targets the African-American community in St Louis. And he’s dead-set against any regulation of interchange fees, which provide an important source of income for his institution; he’s written three blogs on the subject, here, here, and here.

Adams, unlike Harriet May, was willing to provide me with concrete numbers:

We have 25,000 debit card holders with a 50/50 split on debit vs signature. We had 3.2 million debit card transactions in 2009 totaling $892,490 in debit interchange income or an average per transaction of just under 28 cents per. At an average interchange rate of 1.3%, our average member debit transaction is for 21.40. Our all in expense (including fraud) is $521,000.

We project that a 50% reduction in interchange would cost us $446,000 of top-line revenue. Something has to give if we lose that revenue. A $20 per year annual fee works, but we don’t want to fee our members.

Essentially, St Louis Community CU is getting about $35 of top-line revenue per year, per debit card. If that revenue disappears, it hurts the credit union’s finances. And so Adams is railing against interchange regulation:

Here’s another surefire lock of a bet. You will be more frustrated than ever. Your costs at the bank will be up. Your costs at the retailer will be up. You will be confused as to which retailers accept your debit card and which ones don’t. You will have no clue what the minimums and maximums of your debit card activity will be because there will be no consistency among retailers.

As a result, you will carry more cash and more checks… And, what about this double-dip possibility? You’ll use more checks at the check-out counter and the retailer will charge you a processing fee for doing it. (See, their handling of checks and cash are more expensive than debit cards.) You’ll pay for that, as well.

If this legislation is passed, I will mark my calendar to re-visit this issue a year after enactment. If I am wrong, I will eat the biggest piece of humble pie ever, including a public apology to everyone – starting with Senator Durbin. I must tell you that I’m extremely confident that an apology won’t be forthcoming.

I’ll take Adams’s bet. Yes, the costs of a checking account will be more transparent and visible to consumers. But costs at the retailer will not rise, since the retailer’s costs will have fallen. There will be no confusion about which retailers accept which debit cards, and debit-card minimums and maximums will be a non-issue. People will not carry more cash, and they certainly won’t carry more checks. And Adams will owe a public apology to Durbin.

Because some of Adams’s predictions are a bit vague, let’s drill down to the one where it’s easy to get hard data from his own credit union. Let’s look at the number of checks that Adams processes per checking account on the day that interchange regulations are signed into law (assuming that happens), and one year later. If it’s risen, I’ll be proved wrong, and will happily donate $100 of my own money to his credit union. If it has fallen, I’d like him to write a $100 check to my own credit union, Lower East Side People’s. Deal?

The fact is that if the biggest and most efficient banks in the world can’t make money from providing checking accounts to low- and moderate-income individuals, then it’s not going to be easy for credit unions to do so either, especially because low-income individuals tend to be high-touch customers who do a lot of their transactions at the teller window. Credit unions like St Louis Community and LES People’s pay for the cost of providing those checking accounts in many different ways, including, when we’re fortunate, getting large donations from the federal government.

In recent years, I can well believe that St Louis Community has been very happy to see its interchange income rise, especially from its signature debit cards. Signature debit is a low-security, high-fraud product, but it’s also very lucrative for banks: it’s so lucrative, in fact, that many of them strongly encourage their customers to use them by offering back cash or other rewards. They wouldn’t do that if their interchange fee revenue was eaten up in fraud costs. At places like St Louis Community, the customers don’t have the option of choosing a reward card, so all that money flows straight to the credit union instead.

St Louis Community is keen on extracting other fees from its customers, too: it’s urging all of them to opt in to its “Overdraft Privilege Service”, in which customers have the privilege of paying $15 when their debit card transaction brings their checking-account balance below zero. Yes, that’s lower than most bank overdraft fees — but the fact is that for most people, in most circumstances, they’d be better off simply seeing their debit or ATM transaction denied if they don’t have the money in their account. This service might be “a life-line when emergencies happen”, to quote the website, but the overwhelming majority of overdraft fees are not racked up in emergency situations, and indeed the customer never has the option whether or not to pay the fee: the money is debited and the fee is charged automatically.

The economics of low- and moderate-income credit unions are not easy, and certainly the loss of interchange fee revenue is going to make it harder to find those $300 of profits per checking account needed to cover costs. But it’s wrong to simply try to extract ever-rising fees from merchants as card usage rises and people write fewer and fewer checks which need, by law, to be redeemed at par. The fact is that if one of Adams’s customers wants to pay for an item, then PIN debit is by far the cheapest way of doing it, from the credit union’s perspective — there’s no cash handling at the teller or the ATM, there’s no check processing, the money just automagically leaves one account and enters another. Debit cards are a great way of bringing credit unions’ costs down. That should be enough: they shouldn’t be a major profit center in their own right.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

I am not sure all of the subsidies in the system work against the small/low-income bank/credit customers; higher balance customers do have more debit transactional activity, generate more revenue for the bank/cu and provide higher balances to lend to others (at 7% net of losses, though, Felix? Really? Not at many of my clients.). While some customers do reap reward benefits and consume some of that revenue that is still the exception on the debit card product. Instead, I suspect that smaller institutions actually see some subsidization of the revenue from their higher-balance-and-transaction customers to the benefit of low-income-low-balance customers.

Where does this change lead, then? Well, banks (especially large ones) are pretty good at determining household profitability and culling out those customers that are unprofitable. Because more will become unprofitable under the recent regulatory changes (and pending ones) you can expect banks will be working to cull more customers than before. Less sophisticated institutions (community banks and most credit unions), in addition to seeing existing relationship profitability suffer already, will take some of those customers in and add additional money-losing relationships.

Because they will lack the ability to identify money losing relationships and/or the will to cull those relationships these customers will drag down earnings and, on the margin, drive some small insitutions to their graves. Over time these pressures will continue and the data analysis and decision support technology advantages large institutions have will concentrate and smaller institutions will be further disadvantaged from where they are today.

The big guys are getting what they wanted, it seems: overall profit pressures from these changes are real for everyone but will be proportionally higher on smaller institutions.

Posted by TRKAdvisors | Report as abusive


That’s an interesting scenario you posit, but I’m skeptical. Sure, higher balance customers generate more transactional activity, but they’re also less likely to generate truly lucrative overdraft fees. As you point out, they don’t even make 7% net of fees on the balances – I suspect Felix was deliberately using an overestimate there to make the point that even the best-run lending operation with the lowest defaults can’t turn a profit on a $1k no-fee checking account. (If it’s true of 7%, it’s certainly true of, say, 4%.) So where do these banks scare up $300 a year in revenue per account – perhaps more, because they lack economies of scale enjoyed by their larger competitors? If it happens, it’s coming from cross-selling other financial services to these more affluent customers. But I suspect it does not. At least, that it hasn’t in recent years. And that’s why we’ve seen both credit unions and national banks aggressively pursue lower-tier customers.

But here’s the real problem. Who are the customers who are currently worth pursuing but will be rendered “unprofitable” by the new regulations, and whom you expect to move from sophisticated big banks to smaller ones? Presumably, they’re the ones currently paying overdraft fees and generating tens of billions in revenue that’s about to disappear. Then, since they don’t generate as much in transactional fees, and are more likely to use costly teller services, they’ll be dropped by the big banks.

But that’s not necessarily a bad thing. Banks can still make money of such customers by the simple expedient of charging them for the services they use. Big banks may not find it worth their while. But small banks haven’t been able to levy such fees because big banks have been competing with them for low-income customers. If the big banks withdraw from the game, they’ll abandon the playing field to smaller institutions, which will be able to impose more rational and transparent fee structures, which will have the added benefit of being more fair.

Potentially, almost everyone wins. Small institutions recoup some of their customers and enlarge their deposit base. Low-income customers will pay more in explicit fees, but in exchange, they’ll gain predictability since they’ll know the costs upfront, and may well end up paying less overall. The big losers here will be big banks, which are losing a huge revenue stream, and wealthy customers, who have been cross subsidized. But I can live with that.

So why do Patrick Adams and other small bank executives oppose the changes? Because most bankers hate change. If they have a profitable model, they generally prefer to milk it for all its worth. If the last few years have demonstrated anything, it’s that neither self-interest nor market pressure necessarily produces models of financial services that actually serve the long-term interests of the financial institutions themselves. They’d rather chase the easy profit today, and the chance of outsize returns, than the possibility of steady revenue for the indefinite future. This is just one more example.

Posted by Cynic | Report as abusive

Wait, banks don’t lend out my deposits. They make a loan and then find a way to meet reserve requirements.

Posted by petertemplar | Report as abusive

How can it possibly cost a bank $300 to provide my checking account? It gets deposists from my employer via ACH, which scales and hence costs roughly $0. It sends money hither and thither, also via ACH, still $0. It generates a PDF statement and emails me once a month, which scales, and costs $0. It covers debits via interchange, which as we know has negative cost.

Where are the $30/month costs?

Posted by wcw | Report as abusive

To wow: Executive Comp (plus, increasingly, compliance). Goldman’s checking accounts cost $1.2 million each in expenses.

Cynic: Thanks for the thinking, that’s why I like this venue.

Posted by TRKAdvisors | Report as abusive

Hi Felix,

This is your best article yet on the economics of debit / credit and interchange fees and the relitive impact on merchants and lower vs upper income consumers.

You will lose the bet with Adams though. At most small and medium sized banks (like mine) the interchange fees and interest income of the top 20% of customers reduce the cost of checking to zero for all our customers (who don’t overdraw their accoutns.)

If a hard working low income consumer earns $20,000 a year and spends every penny via a debit card paying the 1.3% interchange quoted in your article it is safe to assume that the merchants inflated the cost of goods and serices to the tune of $260… so that’s the “hidden cost” of their free checking account.

Now compare that to the person making 5 times as much… well they paid $1,300 for the same checking account.

Community banks like mine unquestionably benifit the little guy at the expence of the big guy. The best part is that the big guy doesen’t even grip about it because almost all our big fish started out as little fish at one point and they were glad to get their free checking accounts from us then.

Keep up the great writing but please don’t advocate for the death of free checking… the working poor at your credit union can’t afford “not-free” checking!

Posted by y2kurtus | Report as abusive

Felix – like healthcare, the REAL problem is that this is an industry where the economic actor does not have the proper incentives.

Historically, banks were an important part of the economy in that they separated good loans from bad loans. However, with the advent of standardization (credit scores, Fannie/Freddie), this role has greatly diminished. So, we have to ask: What do most banks do today? What is their business model?

In short, it is just a place where we store our cash. It’s like a parking lot. We choose our parking lots based the easy of access, cost, service, and security. However, we don’t choose our banks this way.

Sure, under the current conditions the rich may be subsidizing the poor. However, if we removed the arcane system of hidden fees, I expect to see the rise of Walmart style banking which will be cheap for both rich and poor. A checking account should not cost $300 a year, but if you have as many bank store fronts as Starbucks, I guess it does. At the moment, Chase can offer free checking so there is no point for consumers to move to a cheap online bank that understands operational efficiency.

If the bill is properly constructed and passed, I don’t expect it to be the end of free checking. But I do suspect that your corner bank, with the granite lobby, air conditioned offices, and friendly, and well-paid management to go away.

Posted by Brutus123 | Report as abusive

I’ve just moved to the US from the UK.

Ok in the UK we have some pretty egregious overdraft charges from time to time but basic services are far better than the US and almost always free (free ATM at any bank not just your own, online instant transfers to any domestic payee account, direct debit for almost every possible payment, more security thanks to chip and pin etc).

Is it just US institutions are much more inefficient? I laugh when I see people over here using checks, nobody uses them back home, and the chase online banking is laughable.

Posted by vk9141 | Report as abusive


Executive comp and compliance are static costs; I assume they are not increased by an additional small free checking account. Wow is getting at the marginal costs of an additional account. I have seen free checking account that require direct deposit and charge for counter transactions; it’s hard for me to see how the marginal cost of that account can approach $300/year.

Posted by MattJ | Report as abusive

The dark side of “free checking” as it now stands includes, among other indignities, having to pay a major bank (depending on which one) $5-$15 in immediate fees to them just for cashing any of their account holder’s checks. Can’t see them reversing this even if they start gouging their own customers for monthly maintenance.

Having a “free” checking account with most major banks these days is just the launch pad for a whole raft of other potential back-end costs. On top of which you’re supposed to feel sorry for them not making enough money to cover their gambling debts. You pay them to make better excuses, but they keep on recycling the same old tired ones.

Most American banks are like bad brothels into which go only the most desperate of customers. If they cared about customer service any one of the majors could, at any time, break ranks and start operating absolutely transparently, such that customers might make fully advised choices at all times, not just after getting hit on the head with supplementary fees. Chances of this actually happening? Not terrific.

Posted by HBC | Report as abusive

Further reading: check out Mike Konczal’s reaction to this. Key quote: “The rich pay more or less the same in fees as the poor, but the poor are more likely to pay them.” so-called-death-of-free-checking-12856/

Posted by BryceCovert | Report as abusive

I have to agree with wow, the number of $300 a year per checking account is vastly inflated. Use the example that Felix cites in the article– a small (no economies of scale) credit union with 25,000 debit cards has cost of $521,000 a year. That’s $20.84 per card. Checking accounts are similar.

My calculations show the industry is saying checking accounts cost them $60 billion a year. I just don’t believe it. I’ll bet the branch network is included in that.

The incremental cost of adding checking accounts to already existing infrastructure is probably on the order of $20 a year. At that price, using Felix’s numbers, they’re making $50 a year per account.

Felix, why don’t you ask them for an accounting of their numbers.

Posted by bff426 | Report as abusive

This seems like a problem with a fairly easy solution — since we’re all better off when households have access to basic financial markets, the government should simply subsidize checking under certain circumstances.

Posted by PunditusMaximus | Report as abusive