Interchange and the Price of Gas

With all the attention being paid to the soaring cost of gasoline, there's one price component that could easily be lowered that hasn't gotten a lot of attention: interchange. Americans pay about 10 cents/gallon or $2/tank (20 gallon) in interchange fees. Compare this with the estimated benefit of drilling in ANWAR--3.5 cents per gallon by 2027. Tremendous political energy that has been spent over drilling in ANWAR. If we're really interested in reducing gasoline prices, then reforming the interchange fee system presents an option of immediate savings that would far outstrip ANWAR. Of course, the maximum savings would be less than from a federal gas tax (18.4 cents/gallon), but a gas tax holiday isn't permanent. (I'm putting aside the question of whether we should want pricing that encourages fossil fuel consumption.)

Fortunately, the good folks at Visa are looking out for the American consumer. And out of concern for American drivers, Visa has adjusted its interchange rate schedule. Unfortunately, there's more than meets the eye to Visa's apparent benevolence and desire to "Help Ease Pain at the Pump." The devil, of course, like so many things in the card industry, is in the details.

When you go to the gas station and you use a Visa card, the gas station will pay its acquirer bank a merchant discount fee that consists of the interchange fee plus some thin profit margin for the acquirer. For simplicity's sake, let's just say that the merchant pays interchange and ignore the acquirer's margin.

So currently, if you fill up your tank and pay at either a service station or an automatic fuel dispenser with a Visa card, the interchange fee will be is a five or ten cent flat fee plus between 1.43% and 2.10% of the purchase price (depending on rewards level). So on a $50 tank of gas, you're paying your credit card issuer that's between $.80 and $1.15. I'm going to use the $1.15 for illustrative purposes.

Let's say you can buy 12 gallons of gas for $50, so gas is $4.17/gallon (close to the national average). If you weren't paying the interchange fee, you'd be getting 12 gallons for $48.85. That means gas would be $4.07/gallon. In other words, you're paying 10 cents a gallon to your credit card issuer. And that's before any interest or late fees, etc. For smaller purchases, the credit card tax on gasoline gets higher because of the flat fee. And if the price of gas/gallon is higher, then interchange represents an even greater cost per gallon.

[For debit, Visa's interchange fee is .70% plus 17 cent flat fee, but debit card transactions at service stations often result in debit card holds--much large notional charges--that can limit the cardholder's available funds for a few days. I'm going to leave the debit card hold issue aside for now.]

So what has Visa done in response to the "Pain at the Pump"?

Visa instituted a new interchange fee for all gas stations of 1.5% 1.15% of transaction price plus 25 cents, regardless of the level of rewards on the card. Because of the increased flat fee, consumers making high dollar amount transactions will pay less, but consumers with small transactions will pay more. Where is the dividing line? It depends on whether you were using an exclusive Visa Signature Preferred Card (2.10% + 10 cents) or a mass market rewards card like Visa Signature or Traditional Rewards (both 1.65% + 10 cents) or a non-rewards card (1.50% + 5 cents). If you are wealthy enough to land a Visa Signature Preferred Card, this means the cost of interchange to the gas station (and thus ultimately to you) will be less on transactions over $15.79. If you have a regular rewards card, then the savings kick in at $30. And if you don't have a rewards card at all, then the new rates are actually more expensive for the station for all transactions under $57. [Correction 7.2.08]

Does this new structure actually represent a lowering of interchange costs for gas stations? It depends on the card mix and the purchase prices. If everyone in town drives a Hummer and has a Visa Signature Preferred Card (Greenwich?), then this is a great program. But if most people have plain vanilla cards and aren't letting their tanks get bone dry, then that $57 threshold is pretty high; I rarely exceed $57 for a single gas purchase with my thirsty minivan. That'd mean that gas stations would be paying even more in interchange and the price to consumers would go up even more.

Surely Visa looked at card usage data and concluded that it would actually be more profitable to lower percentage fees while increasing the flat fee. Visa is a for-profit company, after all, and there's no tax write-off to helping consumers. So it is easy to see a scenario in which Visa gets the PR boost for trying to help consumers, while actually sticking it to them.

Curiously, the Visa move also undercuts Visa's litigation and lobbying position, that interchange is just an interbank fee. By announcing its "lowering" of interchange rates as a move to help consumers, Visa implicitly acknowledged that interchange contributes to consumer costs. (The best empirical evidence that interchange is passed on to consumers also comes from the gas station industry.) Interchange is not just a Main Street vs. Wall Street issue. It is also a consumer issue.

All of this points at the need to think about ways of reforming the interchange system. One possibility is that it will be done through the courts. Another possibility is that it will be done legislatively. Interchange week will continue with a post about the Credit Card Fair Fee Act.

Comments

[Don Lloyd]

Adam,

Unless I'm missing something, every credit card customer of every variety and stripe, and every cash customer pays exactly the same price per gallon for a given grade of gasoline and a given service level at a given station at a given time.

This would mean that the recovery of fees by the station operator can only be accomplished in the same general way as any overhead cost, such as rent.

However, with a high level of competition, I would expect that the fees will largely impact the profit and loss of the station operator and that the consumer impact of the fees will be mostly seen in the number of stations that can stay in business.

While I'm pretty sure that it is not legal for a station operator to give discounts for the more profitable cash sales at the pump, I'm surprised to not have yet seen smart loyalty cards filled by cash, even if they couldn't be discounted at the pump either.

Regards, Don

[Dan]

Allow merchants to pass card interchange costs onto customers in the form of a surcharge. That's our answer!

Every merchant knows it is against card network rules to allow merchants to pass costs of accepting credit cards to their customers (e.g., discount for cash). Would customers choose lower cost payment types if they knew how much the actual cost of interchange was? Or, if they had to pay for it, such as in the form of an allowable surcharge?

As you mentioned, savvy merchants know that some transactions and card types cost merchants more and some cost less. Examples include PIN-debit cards and Rewards cards (PIN cards cost less; Rewards Cards cost merchants more).

Under this scenario, customers can save themselves money by choosing cards and payment forms that costs them less. Merchants can serve their customers better by accepting all forms of payment - knowing that their customers will choose the payment credit and debit best for the customer.

Regulators - from Congress, to the FTC, to the Justice Department, can keep out of the regulating pricing business. After all, once given the information, consumers are pretty good at figuring out what the best value is for themselves.

Of course, the credit card industry would probably prefer that consumers be left in the dark on what that card is costing the merchant. And, there would be a substantial investment in upgrading POS systems to become intelligent enough to know how to pass back interchange to cardholders at the point of sale.

The current interchange system in the United States may be legal, but is clearly corrupt. The debate on how to fix it should include solutions that empower consumers to choose which is most cost-effective, and most valuable to themselves, without the burden of government intervention.

What's your opinion?

[Patches]

I have seen local gas stations advertise lower prices for gas if the customer pays with cash. As every small business owner knows (retail anyway), it takes some time to process the payments thru that system…. and then there are those fees. So is it illegal to offer cheaper gas if the customer pays in cash? It would be hard to imagine so. If a store gets the same amount for each gallon no matter the payment method but the store profits less from check card or credit card sales, why wouldn’t the store just say “Chunk it” when it comes to gas sales? They would make up the difference from the loss of convenience by the long lines of people wanting cheaper gas, even if it is just a few cents. Maybe revert to the good ole’ days of “service at the pump”? eeeehhhh? That would be nice.

I think in this case, if people see that they can get cheaper gas by paying in cash, who would want to use their check card or credit card for gas? I mean besides the people who drive Hummers. Convenience to me in this case would take a back seat. I can use my check card for other things, maybe be like my dad and keep a money clip. Do they still make those? Seems like a win win for the consumer and small business. Maybe get a little exercise by walking into the store. What a concept! It could be a whole new weight loss fad.

[Adam Levitin]

Don is right--everyone pays the same price regardless of type of card. Card network rules prohibit merchants from price discrimination among cards. This means that the marginal cost of a rewards card is either lumped by the merchant or passed on to all consumers in the form of higher prices (or both). The empirical evidence is that both happens, although the ratio of cost absorbed by the merchant and passed on to the consumer will depend on how competitive the merchant's local market is.

I'm with Dan 100%--I think surcharging is the way to go and have written quite a bit about this. I'm planning a post on the subject later this week.

Patches is rightly confused--gas stations, like all merchants, are allowed by federal law to offer discounts for cash, and some do (the gas station industry is the only one where cash discounts are at all common). But merchants are prohibited by card network rules from surcharging for credit, even though those are mathematical (but not economic) equivalents (100-2=96+2, but they appear as bottles half-full and half-empty to consumers, which have different psychological effects on consumption). Merchants don't really like cash, though--it has lots of security risks. They want to take credit cards, but avoid the rewards cards, and the Cash Discount Act of 1981 doesn't help at all with this. And merchants want interchange rates to generally be subject to market pressure (currently Visa can raise rates without losing market share).

[Howard Strong]

Many gas stations in California offer discounts for cash payment. The ones I've seen lately are around eight cents a gallon.

Apparently, the discount varies with something because I have seen stations where the cash-credut differental varies on an almost daily basis.

According to an odious commercial played over the TV at the pumps system at United Oil outlets the poor retailer makes only twelve cents per gallon if you pay cash, three cents per gallon if you use a credit card.

[geeniusatwrok]

the local Citgo gives a 5-cent discount for cash or debit card. I use my Amex Blue Cash anyway because the 5% reward beats the 5-cent discount at current prices. I have a Visa Signature Preferred Card but I never use it; I just stick to the one card and pay it off every month. I didn't know the Visa Signature Preferred Card was so shiny and exclusive, nor did I know that the merchant has to pay a higher fee for rewards cards. Sheesh.

[Adam Levitin]

Geeniusatwork--If you're wealthy enough to get a Visa Signature Preferred Card, you probably don't need the 5% reward. (Don't confuse Visa Signature Cards with Visa Signature Preferred Cards, however.) For what it's worth, the 5% rewards on Amex Blue Cash (a mass market card) aren't really 5%. It's 5% after the first $6,500. So it's only a good deal if you charge a lot each year. http://www.mymoneyblog.com/archives/2008/03/american-express-blue-cash-quick-review-up-to-5-cash-back.html.

Debit, or at least PIN debit (not signature debit) is much cheaper for merchants and probably the preferred payment method for most.

[Chris Phillips]

While I agree with your overall point (and with your math) the new rate is actually 1.15% plus $.25 (not 1.5% as you wrote). I think that's just a typo in your post.

[geeniusatwrok]

Adam: I'm not wealthy, but I have good credit. Maybe it is not Preferred; I'd have to dig it out and look at it as I don't honestly remember. It's black, though :) And I know about the Amex Blue Cash deal only working after $6500. I charge nearly everything to that card so it's not hard to get the spend past that. as I pay the balance monthly, even a paltry overall 1.5-2% average return beats a stick in the eye.

ps: typekey is apparently broken.

[GaDawg]

Concluding that judges or elected officials are the only way to solve this leaves out the role of competition and co-opetition.

Since April 2007, MasterCard has had a cap of 0.95 per transaction on gas purchases. The rate is 1.9% capped at 0.95. In effect, any ticket price over $50 and the gas station begins paying a lower effective rate.

So competition between two of the card networks is already having an effect.

Interchange rates for both Visa and MasterCard are available for the public.

In addition, gas merchants are also free to work directly with banks to get more advantageous pricing through alternative models such as co-branding arrangements, as in The Citi Shell Card, the Chase BP card.

Competition and creative working arrangements will resolve this issue long before the pols or judges make their mark.

[Adam Levitin]

GaDawg--this issue has been around for several years now and there hasn't been a market fix. How long do we have to wait before we admit that the market isn't working. I'd love to see a market solution, but market solutions don't happen in concentrated, anticompetitive markets with high barriers to entry.

The example you give of MC and Visa competing price down only holds true on purchases over $50 or so. But lots of gas station purchases (which aren't just fuel) are for smaller amounts. And, as I showed in my post, Visa has _raised_ the price on those transactions. So how exactly is this market working?

We shouldn't expect competition to lower interchange, because networks, which set interchange are competing for issuer banks far more than they are competing for merchant acceptance. If MC lowers its interchange too much, then banks will just switch to issuing Visa, and vice-versa. Network competition for issuers pushes interchange up.

Cobranding, etc.--yes, those are options for larger merchants, like major gasoline chains. But how many Ma and Pa stores or even mid-cap companies can cobrand? And cobranding doesn't give the merchant a huge cost savings. There's a reason Wal-Mart would rather have a bank than co-brand.

[anon]

you are assuming that merchants will pass the reduction to consumer. what is the guarantee of that happenning? given the state retail sector is in, they are more likely to report higher earnings than pass it on to the consumers. look at australian regulations, prices did not go down.

[Anand Goel]

It was noted in the original post that the acquirer makes a thin margin per transaction, and majority of the card acceptance cost comes from Interchange. That's true for large national merchants, but for the average small business about a third of the cost of credit/debit processing comes from the acquirer. According to research by First Annapolis, merchants with $250K - $500K in annual processing volume had an effective rate of 2.99%. 65% of this rate was Interchange, 3% was assessment fees from Visa/MC associations, and 31% was the acquirer gross margin.

Keep in mind that the acquirer shares this margin with its bank partners, resellers, etc. Although the top 7 acquirers in the U.S. process over 90% of the
credit/debit transactions, there are literally thousands of resellers working for them. And often times these resellers will have resellers under them...thus, making it easier to understand how an acquirer could generate such a hefty gross margin.

[PeterD]

Adam, I have read your recent paper in a good amount of detail. It has a lot of excellent information about the state of play in the area of payment systems in the US. However, I am going to have to disagree with your assessment that a loss of a cardholder or merchant by Visa which is subsequently gained by Discover creates a positive welfare externality that offsets a negative welfare externality and thus is welfare-neutral. Marginal welfare in general, increases with network size. A gain by Visa is more beneficial to welfare than a gain to Discover under traditional economic analysis. Each of the networks compete to provide maximum social welfare, with the largest holding the greatest advantage at any given time. Hence I think you're being too hasty dismissing the networks' main economic justification for interchange on this basis.

[Adam Levitin]

PeterD--the network effects literature really doesn't get into the question of competing networks. It's a topic that certainly deserves fuller theoretical treatment than it has received to date, and without addressing that question, I don't think the existing network effects literature is particularly informative about whether and how to regulate interchange fees.

That being said, I stand by the argument in my paper--under a traditional network effects welfare analysis, the marginal benefit to network participants of additional network participants diminishes as the network grows in size. The difference of an additional million network participants matters much more (in terms of positive network externalities) when a network has 5 million participants than when it has 500 million. This means that welfare is maximized if the marginal transaction goes to the smallest network (Discover), rather than to the largest (Visa).

If it were the opposite, and a gain by Visa would be more beneficial to welfare (whose welfare?) than a gain by Discover, it would turn into an argument for monopoly as the most efficient outcome--every transaction should go to Visa. And if you're really going to argue that, then I think the conversation has to change into one of whether such a monopoly should be in private hands or run by the government, as a way to check extraction of monopoly profits. I think that's a reasonable conversation to have--payment systems have many features of public goods--but the politics of credit card regulation isn't there yet.

[PeterD]

Adam, you say "under a traditional network effects welfare analysis, the marginal benefit to network participants of additional network participants diminishes as the network grows in size". Do you mean marginal benefit to each individual existing network participant? Then this statement may be true of networks in practice but the gross marginal benefit (which includes total marginal benefits to both the existing and additional participants) increases with network size under traditional analysis. There has been some debate in the payment systems context about whether as a network matures, network effects diminish because of substitution effects. In any case, your claim that there is an equal offset of welfare benefits as participants abandon one network for another is not supported by the literature.

Hope this is of some use to your thinking.

[Adam Levitin]

PeterD--of course my claim is not supported by the literature because the network economics literature never attempts to address the question of competing network products. The existing literature neither supports nor contradicts my claim. I accept your point that I would benefit much more from joining a larger network than a smaller one, but again, this points to monopoly as the outcome, and arguing for the efficiency of monopoly is a very strange position and ignores the benefits to the participants of all networks of having competing network products that spur innovation and reduce price.

As for the question of diminishing network effects with mature networks, I think there is a good theoretical argument that is the case. A major problem for a two-sided network is getting past the chicken and the egg problem of recruiting both types of network participants simultaneously. Unfortunately, with all of the network economics, we are just playing in the modeling sandbox and don't have empirical data to support any of the arguments.

anon--yes, I am assuming that merchants will pass on at least part of the cost reduction to customers, at least in competitive sectors. Retail is one of the most competitive areas of the economy, so I would think that it would be one of the areas in which we would see the greatest pass-thru of savings to consumers. It is important to note, however, that this might not appear in terms of a price reduction, but more likely in a slower rate of increase in prices. This is what happened in Australia--prices didn't drop, but inflation (measured by costs of goods) slowed. That's economically equivalent to a price drop. In any case, actually correlating the impact of the Australian regulations on prices is near impossible--there are so many factors that influence merchants' prices, and interchange is but one. Moreover, we would expect a much smaller impact in Australia than in the US because interchange rates in Australia were far lower than in the US to begin with, so the potential savings would be much smaller.

There is some empirical evidence for the pass-thru occurring from surveys of gasoline station prices in the 1980s, when some gas stations had one price for cash and credit, and others had a lower cash price and a higher credit price. The unified price was higher than the cash price and lower than the credit price--that means gasoline merchants were passing on some of the savings and pocketing some. This matches what we might intuit--the retail gasoline industry is generally fairly competitive, but there are lots of local monopolies (the only gas station in a small town or for 30 miles on an interstate).

Anand Goel makes a good point--the ratio of interchange to merchant discount changes significantly by merchant size and industry. Interchange rates never really top 3%, but for a small on-line porn website (or gambling websites when they were legal in the US), the merchant discount fee can hit 15%. Smaller merchants are generally riskier for the acquirer, so the merchant discount spread will be heftier. A merchant with $250-$500K in annual card sales presents only a small pillow for a set-off for an acquirer--maybe not enough to even cover all legal fees if the acquirer has to litigate with the merchant (say in bankruptcy). All that being said, acquiring contracts are typically short-term (1 year or less) and there is a lot of attrition in acquiring portfolios because there is vigorous acquiring competition. And even if a small merchant is paying 10% for accepting cards, that merchant would much rather pay 8.5% than 10% and consumers would likely see some of that 1.5% cost savings.

[TakeFive]

"I am going to have to disagree with your assessment that a loss of a cardholder or merchant by Visa which is subsequently gained by Discover ..."

I'm going to add my disagreement. A lot of retailers (more so the lower end ones I think) do not accept Discover card. My auto mechanic told me the other day that the bite by Discover is really painful (he didn't give me the exact percentage though).

Last night at the local Mex resturant (average plate: $5.95), my Discover was waved off - visa & MC only.

I also a little suprised there isn't a fee charged to the consumer for cash-back on debit transactions. The checkers at the store ALWAYS ask if you want cash back, so I fugured they were making a pretty good vig on it.

Hope you write more on this subject.

[PeterD]

Adam : "arguing for the efficiency of monopoly .... ignores the benefits to the participants of all networks of having competing network products that spur innovation and reduce price."

This, I agree with. Class economic analysis alone ignores many important forces impacting social welfare.