Home Depot Spends More on Interchange than on Health Care

I was bowled over by a figure in The Home Depot's presentation at the Chicago Fed’s 2009 Payment Systems Conference this week:  The Home Depot paid more in interchange than for employee health care last year.  That’s astounding.  Interchange is The Home Depot’s third largest operating cost.  And this is from a company that gets comparatively low interchange rates just by being large.  Interchange is costing large, sophisticated merchants more than health care.  And the value it gives is questionable:  The Home Depot's interchange costs have risen 16% in recent years, while purchase volume has increased 10%.  [COMMENTS NOW OPEN.]

I’ve previously argued that making interchange rates competitive would result in savings for consumers.  But as The Home Depot pointed out, reduced interchange rates could benefit consumers in other ways:  cutting interchange costs in half (as happened in Australia after the Reserve Bank of Australia mandated at-cost weighted average and allowed surcharging), would put another 5 full-time employees on the floor of every Home Depot store.  Not only would that make customers’ shopping experience a lot better (and faster), but it would also generate an awful lot of jobs.

Just to play this out on the back of the envelope:  $48 billion in interchange last year, according to the Merchants' Payments Coalition.  If interchange dropped from around 1.8% to .5% on average (.5% is what it is in Australia; MasterCard recently agreed to .3% for cross-border interchange in Europe), that's around a 70% drop.  That'd translate into roughly $34 billion in savings.  Bureau of Labor Statistics estimates that average total cost of employment for an employer of an employee working 50 weeks a year, 40 hours a week at roughly $60,000.  So the cost savings would be equivalent to about 560,000 average jobs.  That's nothing to sniff at, especially these days. 

I don’t want to harp too much on interchange as an employment issue.  If businesses paid lower taxes, they could hire more employees too.  But it is worth underscoring that interchange is a lot like a tax, just not one subject to political control or which is for the public benefit.  And lowering interchange could have the same stimulus effect of a tax break, but without an impact on the federal budget.  Dealing with the interchange issue would be a nice economic stimulus package. 

Comments

[Sharon Gamsin]

I was at the Chicago Fed’s conference too. There was a lot of lively debate on this issue, and Adam is only presenting one side. Lower interchange rates may help merchants, but can hurt consumers. A representative of the Reserve Bank of Australia spoke at the conference, and when he was asked if merchants had lowered prices when the government cut interchange fees, he said he doesn't have any evidence of that happening, and that we would have to "assume they did." What we do know for sure is that consumers were hurt by the RBA's regulation of interchange – through higher prices and fewer benefits. In the first five years following the RBA’s regulations, the fees consumers pay for credit cards rose by 22% for standard cards, and as much as 77% for rewards cards, while the value of those rewards programs fell 23%. So merchants benefited, but consumers were hurt. A report by London-based CRA international discusses the impact of interchange regulation in Australia: http://tinyurl.com/rb585g. Another factor Adam overlooks is the protection from credit risk merchants receive from payment card acceptance, which in today's economic climate far exceeds average interchange fees. Issuing banks absorb this risk because the merchant is guarantee payment, whether the cardholder pays their bills or not. In 2006, the average interchange fee was 1.8%, and the average credit loss, as a percentage of transaction volume was 2.41%. In 2008, credit losses rose to 3.35% and are expected to be much higher this year. But interchange the average interchange fee remains unchanged since 2006, providing merchants even greater value.

Sharon Gamsin
MasterCard Worldwide

[steve]

As Ms. Gamsin clearly pointed out, the Card Companies are going to get the consumers one way or the other, so just relax and enjoy it.

[Fraud Guy]

Of course, her statements conflate a few different concepts.

First, the interest rate cap is in Australia only. Fraud costs are passed on to merchants in many situations (as much as the issuers can do, especially internet.

Credit risk would not be an issue if customers paid in cash instead of cards, and the reason credit risk is so high is that, in order to increase their fee and interest stream, many banks target customers who will keep high balances on their cards, which make them more likely to default in an environment like the current one.

Also, the interchange is not the total revenue stream of the issuers--IIRC, it's only about 30%, the rest coming from cardholders, so to reverse the number, about 5.5% revenue of sales volume vs. 2.41% credit risk, or 3.35%, is still a good value to banks.

And, this is for credit, not debit cards, where the credit risk for banks is much smaller, the fees are only marginally lower, and which represent a majority of card use in the US.

They are not hurting on cards, not by a long shot.

[siva123]

rightly said hair loss is a common problem and even difficult to deal with. Vitamins are good if taken care to prevent hair loss. even there are other methods for hair restoration, these days different techniques are being used for hair loss.

[Buddha]

Adam's post about Home Depot, and this debate more generally, seems to confuse the economic and legal incidence of interchange fees. Which, by the way, is another reason for an analogy between interchange and taxes as the same confusion often arises in tax settings. If I have my way, this part of the overall issue will be a classic econ 101 question in a few years (although the whole issue is, admittedly, a bit more complex than the standard econ 101 tax question).

Please humor me if this comment is a bit long....

Merchants often claim, "Interchange fees screw us because they cost us $48 billion a year!!" In the next breath, they say, "And all consumers pay higher prices as a result." The former statement reflects the legal incidence of interchange fees – who legally pays the fee – while the latter is more reflective of the economic incidence of the fees. This distinction between legal and economic incidence essentially blows up the cute back-of-the-envelope calculation that Adam performs in this post and also calls into question some of the basic claims of damages made by merchants. Let me explain why. But first, a lengthy digression...

If one wanted to be really precise, one would note that the true legal incidence of interchange fees is on acquirers who pay the fee to issuers. The card networks often make this argument, but everyone immediately recognizes that interchange fees don’t screw acquirers, who simply increase the merchant discount to pass those costs along to helpless, hapless merchants. After all, we don't see the acquirer community desperately lobbying Congress. Why is that? Think about it, as it's not as obvious as it might seem.

Back to the main body of this comment... Suppose that you’re a merchant who has exclusively credit card payments and suppose that your interchange costs associated with those payments are $48 billion per year (for simplicity, I’ll use "interchange fee" as synonymous with "merchant discount"). You know that you’re facing those costs, so you build them into your prices and end up charging your customers $48 billion more. Your profits then reflect a $48 billion hit to your costs, but your revenues are $48 billion higher.

A second (basically irrelevant) digression, but interesting from a "framing" point of view...

If you receive a mix of payments as a merchant, you’ll set prices that reflect that mix – essentially a weighted average of your transaction costs - unless you're an idiot. Notably, you may lose money on any single card sale in a narrow accounting sense, but you’d still end up recouping the $48 billion cost through your pricing. For example, with a 50-50 cash-card mix you’ll charge a price that reflects that you pay interchange on 50% of your transactions. You’ll lose money on each single card transaction – because your price doesn’t reflect the full interchange fee – but you’ll make money on each individual cash transaction – because your price is higher than your cash cost. Those two results will imply that you recoup all of your card costs in aggregate. Oh, by the way, if you try to build the entire interchange fee into your price as a 50-50 merchant – something that merchants often claim that they do, to really get across how consumers get completely screwed – then I, another 50-50 merchant, will price the way that I just described and will steal all of your business. Of course, things get trickier if you’re a merchant with market power, but let’s not go there. The merchants certainly don’t want to.

Back to the crux of this comment...

Now suppose that interchange is cut to zero. Whoopee!!!! Your costs have fallen by $48 billion! You can hire 560K more workers! Now there’s a stimulus package!

Not so fast. Just like all of the merchants claim to be, you’re in a highly competitive market, so you have to pass some of those savings along to your customers. If you’re competitive enough, you pass them all along so that you end up charging your customers $48 billion less. In the end, your costs have fallen by $48 billion, but your revenue also falls by $48 billion, leaving no change in your overall profits. Whoops, can’t hire those 560K workers after all. Sorry guys.

The key is that the economic incidence of the interchange fee, as with a tax, depends on the ability of the party facing the legal incidence of the fees to pass them along. If merchants largely pass those fees along to consumers (as they implicitly claim to, because "all consumers pay higher prices"), then that mitigates any damages to merchants from high interchange fees. Similarly, any decline in interchange fees won’t boost the bottom line of merchants (or allow them to hire new workers – again, sorry guys) if they pass those savings through in prices.

Where, then, is the deadweight-loss (DWL) associated with high interchange fees? Someone must be getting screwed by this whole arrangement! But who?

Before answering that question, note that Adam’s calculation basically assumes that the entire $48 billion is a DWL (attributable to merchant damages?). That is, the $48 billion savings is entirely a social savings that essentially creates 560K jobs that didn’t exist before. However, even if the entire $48 billion were borne by merchants, it could mainly reflect a transfer from merchants to banks – or perhaps cardholders – rather than a true DWL. In an extreme case, the 560K more workers at Home Depot could reflect 560K fewer workers at Citigroup. I don’t know about you, but I’m not willing to make a value judgment about which job or worker is more valued, notwithstanding the mess that the financial firms have created. (You really need to be careful with these simple errors, Adam, especially if you’re computing provocative numbers and, even more, if you’re selling yourself as a law and econ guy.)

Back to the substantive question...

The possible DWL involves two channels. First, by raising prices, merchants may sell less. This is a social loss (i.e., a deadweight loss) as things that members of society could have otherwise had are not available. I implicitly assumed that quantities/sales volumes do not change when merchants lower their prices to reflect their $48 billion savings (perceptive readers may have caught this). If the higher prices from high interchange fees yield lower quantities, then that would be a DWL (and would reflect harm to merchants and their shareholders to the extent that this reflects lost profits.) (BTW, lost profits through this channel would seem to be very difficult to compute for the purposes of damage calculation. That’s likely why everyone relies on simplistic "legal incidence" calculations to compute such damages. Oh yeah, the latter are also bigger - likely way, way bigger - as well. That's certainly more attractive when you're a plantiffs' lawyer. Not that you are, Adam, but who knows about your readership.)

Second, high interchange fees can distort payment patterns. However, this is a bit more subtle. High interchange fees support rewards for card use – basically, it gives issuers a revenue stream to fund rewards. An excessive interchange fee can yield excessive rewards, thereby inducing excessive card use. (Why use cash when my AmEx card gives me 5% cash back at grocery stores? Good deal! Oh wait, AmEx doesn't have an interchange fee. Does it?) As a result, a DWL can arise due to overuse of certain payment methods – as a society, we expend too many real resources to perform transactions on an inefficient payment method. This is rarely voiced as an objection to high interchange fees, but it really reflects the core distortion that one should worry about (and focuses attention on the core social issue – which payment method is, in fact, more efficient?).

Incidentally, this second effect also involves a transfer between types of consumers that one might find troubling. In particular, the high rewards benefit card users, partly if not entirely or more (depends on how competitive issuers are, which determines how much of the interchange fee is passed back to card users) offsetting the higher retail prices. At the same time, cash users, who aren’t getting those rewards or any other benefits from cards, end up paying higher prices. This isn’t an economic efficiency issue, as it reflects a transfer not a DWL, but it could (should?) be a core policy question. Notably, by lowering interchange fees, you might not only induce a DWL due to a distortion in payment patterns in the opposite direction (i.e., too few card payments – would that be desirable? interchange fees have been a powerful mechanism to encourage growth in use of electronic payment cards), but you would also reverse this transfer from card users to cash users, in the process making card users worse off.

I apologize for all of that blather, but this issue of economic incidence is a point of confusion that has always bothered me, and Adam’s post starkly illustrates the logical error. Basically, the merchants can’t have it both ways – they get screwed by higher costs while consumers get screwed by higher prices. At best, if you use that logic, you end up double counting the harm. At worst, you end up misallocating the harm (bully for you, if you're a plantiffs' attorney) and obfuscating the real issues. There can be real damages caused by high interchange fees, as well as distributional effects, but they are much more subtle than the usual shibboleths tossed around by merchants (and their attorneys).

Two last points (as if I need to make any more).

First, if you watched the discussion during the mark-up of the Conyers bill in the last Congress, you saw confusion over exactly this issue. Who is the Conyers bill supposed to help? Merchants? Consumers? Are merchants required to pass their savings through? Are they required to retain them? Beyond not understanding how the heck the Conyers bill would work as a practical matter (does anyone?), members of Congress were horrifically confused over exactly this issue.

Second, I’m not a lawyer. So there may be some legal standard for "damages" that I don’t know about and don’t understand. But unless one is a plantiffs’ lawyer looking for creative ways to impute "damages," one should be very careful about where one identifies them. (Turning over rocks... "Are they here?"... "At the very least, will the judge believe that they are?"...)

[Adam Levitin]

I seem to have gotten under Buddha's skin. And yet a lot of Buddha’s rant was directed at points that were nowhere to be found in my post. If Buddha's true point is that "There can be real damages cause by high interchange fees, as well as distributional effects," I'm all in agreement. Was my post the paragon of subtlety? Of course not. It's a blog post for crying out loud. I actually address a lot of the points that Buddha raises in my academic work.

First, to address what is nominally Buddha's chief beef: that I confuse legal and economic incidence. Or put another way, either merchants pay for interchange or consumers, but it can't be both. Well, actually it is both. Of course the $48BN is only paid once, but part is paid by merchants and part by consumers. There isn't 100% pass-thru (we don't know the level and it varies by merchant). Merchants eat some of it and pass some along. Both merchants and consumers are seeing some of their surplus redistributed to banks and to high-cost card users.

Second, even if interchange dropped to zero, Buddha's simply wrong to assume that competition means that it all gets passed along to consumers in the form of lower prices. Not all competition is price-competition. There are some industries that compete on service. Thus, Wal-Mart will pass it along in lower prices, but Home Depot might look to improve service (which means more employment). So for a Wal-Mart, the cost-savings might be offset by lower revenue (but it might also gain market share, at another merchants, expense). But for a Home Depot, there would be employment gains. Thus, I don’t think Buddha’s economic wisdom has it quite right. By equating competition with price competition, Buddha has missed something important.

(Also, fwiw, I never assume $48BN in savings—I was basing my numbers off of $34BN in savings, and admittedly rough calculation, but it shows that Buddha is arguing against a strawman when claiming that I presume the $48BN to be all deadweight loss.)

Now Buddha is right that more jobs at Home Depot might mean fewer at some banks. But we'd also have to ask whether the banks would simply cut dividends or operate more efficiently (fewer private jets--I don't assume a perfectly competitive banking market) rather than cut jobs. And even if banks did cut jobs, if bank employees are paid more than Home Depot employees, there'd be a net employment gain. In any case, the snarky comment about my calculation skills was hardly called for, especially as I have, in my academic work, argued that interchange does create deadweight loss.

My arguments have been four-fold about interchange and associated network rules. First, interchange has a highly regressive distributional effect. It create a wealth transfer from non-card users and to high-interchange rewards card users and banks (45% of interchange is rewards, 55% goes to the banks). I'm bothered by that. Buddha isn’t and seems to think the regressive distributional effect is offset by the efficiency gains of encouraging electronic payment systems over legacy paper payment systems. There are some efficiency gains, there, but Buddha's argument proves too much because interchange encourages the use of suboptimal electronic payment systems. We should encourage electronic payment methods (EPM). But not all EPM are equal. We should want to encourage the cheapest and safest--PIN debit, not credit cards and not signature debit. Interchange encourages use of EPM, but it encourages the use of the wrong EPM. As for the doomsday scenario that we might encourage cash usage (heavens fofend!), cash actually makes a lot of sense for some small transactions, and I have no sympathy for "making card users worse off." No one is forced to use a card--everyone can use cash. Not everyone can get a card, much less a high-interchange rewards card.

So this brings in my second issue with interchange—it has encouraged the use of suboptimal electronic payment systems. My third issue is that interchange encourages consumer overleverage. Buddha rightly observes that interchange encourages the use of electronic payment systems, including credit cards as payment devices. But the law of large numbers says that this will lead to more use of credit cards as a borrowing device and to more people mired in leverage. Interchange also allows card issuers to have weaker underwriting standards, which means that more vulnerable consumers are also going to be using credit cards.

Finally, I have argued that interchange does create deadweight loss by leading to higher prices and thus fewer sales. I think this is a very fair assumption; it is basic price theory of demand, and it is a strange assumption for Buddha to have loosened. (fwiw, I have assumed that merchants haven’t been able to compensate for this with higher margins.)

So now a few final points:

First, Amex doesn’t have “interchange” per se, but it has an “issuer discount fee” for third-party issuers. And, like interchange, merchants can’t bargain over it, although it is a core component of the merchant discount fee. Amex is a somewhat different case than MC/Visa, but it's a sideshow on the interchange fight.

Second, for all of the cries for subtlety, a lot of the Buddha analysis wasn’t very subtle. For example, the 50/50 credit-cash mix. Nice for an illustration, but in practice meaningless. The cash/credit mix varies over time and is more complex. It is 3 or 4 tiers of regular credit cards, small business cards, signature and PIN debit cards, checks (including with ACH conversion) and cash that might be in the mix. The merchant doesn’t know the interchange fee on a particular card. So the merchant isn’t going to be able to price optimally. They might come up with a proxy based on past consumption patterns, but it will be rough. And the idea that competition will limit merchants’ pass-thru depends on the level of competition. There are a lot situations with less than perfect competition, either because of local monopolies or high search costs. Am I really going to leave CVS and go across the street to Walgreens over a potential 1% savings in the cost of shampoo due to a lower interchange pass-thru? This raises some important questions of whether lower interchange would mean more merchant surplus or more consumer surplus. I suspect it would be some of both, and I'm fine with that.

Third, I never have sold myself as a “law and econ guy.” Economic analysis is just one tool in the box, and, like a lot of the other tools, it is often indeterminate and easy to manipulate based on the assumptions made.

[Lao-Tse]

Snarky blog you've got here. It seems obvious that Adam let his dramatic side run away with his original post, Buddha went thermonuclear on a central issue or two (with a great deal of fallout in his wake on ancillary topics), and Adam ratcheted back to a far more defensible position. But in the end, Buddha still wins. The meat of the original post is the fact that $48B (or $34B) is a really, really big number with some nice macroeconomic implications of reform. But that only really matters to the extent that the number is DWL (i.e., wasted) as opposed to a transfer between economic agents.

A core defect of law professors playing economist is the distinct lack of humility that they bring to the table. Going for bombast at the expense of accuracy in an effort to get linked by TPM or HuffPo accomplishes little of value to the general welfare. Law professors at esteemed institutions should be above it.

[Adam Levitin]

Lao-Tse has now emerged and claimed to call the debate and to throw in a rather cowardly personal attack from behind a pseudonym. I'm going to disagree with Lao-Tse's assessment of my dialogue with Buddha. As for Lao-Tse's indictment of law professors engaged in economic analysis (present company included), my thoughts are more complicated.

(1) Who wins. Lao-Tse argues that it only matters if there is DWL, not a transfer. I disagree. Distributional issues matter. Economics doesn't have a lot to say about distributional issues (and to the extent it does, most economists ignore the issues because they do not neatly fit into social science models), and that is a major failing of economics. There clearly is a wealth transfer involved in interchange, and it is seriously regressive. On this basis alone, we should be concerned about interchange. But even accepting a world in which distributional concerns were irrelevant, there is also DWL involved. There's no question that merchants raise prices because of the cost of interchange (I've shown this empirically) and that this means (basic econ again) less sales. Not hard to find the DWL there. What portion of interchange is DWL is hard to say, and it will vary by merchant.

Given the inevitability of DWL and also the serious distributional issues involved, it's just wrong to say "Buddha wins." Buddha's correct that there are some nuances I didn't address, but my core point--that interchange is costing the US economy some jobs still stands. My 560K number was a prefaced as a "back of the envelope" calculation, and I don't think that I need to defend its precision for my core point to be correct.

(2) Both Buddha and Lao-Tse have laid into me (none too kindly) about lack of nuance and bombast. I think there's a lot of misunderstanding about blog posts that is reflected in their complaints.

For starters, I'll agree with the statement that when law professors play economist they often don't bring a lot of humility to the table. But frankly the same can be said of many economists, especially when they play with legal issues. But let's not tar everyone with too broad a brush.

The fundamental problem here is that both Buddha and Lao-Tse fail to understand that there are different standards for blog posts than for scholarly publications. If you think my scholarly work is insufficiently nuanced, fine, but that's a very different complaint than an attack on the nuance in my blog posts. It's unreasonable to expect blog posts that are designed to convey one simple point per post to carry a lot of nuance. I happy to accept lack of nuance as a response to any particular blog post, but that's just the nature of the genre.

In any case, a lot of what was in Buddha's comment was not in response to anything particular to my post, but as to broader interchange issues. And, for what it's worth, though, nuance can often be an excuse to fail to address an issue squarely. Indeed, one might argue that this is a key symptom of the Supreme Court's recent antitrust jurisprudence--if anything could possibly be efficiency producing, then that nuance will trump all countervailing considerations.

Finally, a word about why I blog. I blog because I think it is a way to connect with a broader public about issues I care about and about which I have some expertise. There is a very unfair and regressive hidden private tax with interchange and I think people should know about the issue. Scholarly works like law review articles aren't the way to make that happen. If my blog posts get picked up by third parties, great. If not, they are still out there on the Slips in an accessible form for the general public, and I always leave comments open and am happy to engage with commentators who differ with my presentation of issues. That means that there is greater information available to the public with a cross-check on accuracy, and that is certainly a gain to public welfare.

It is also something that is extremely becoming to a law professor at any institution. Being a law professor is not just being a detached social scientist of some sort, but also being part of a living world of advocacy, policy, and law reform. (This is not to say that there isn't an important role for pure theory.) The beauty of legal scholarship is that it is for the immediate world, not just the ivory tower, but that the analysis of worldly issues can take place in the atmosphere of ivory tower detachment. I advocate only causes I believe in based on my scholarship, not on remuneration, and I consider it a core part of my professional obligation to disseminate the fundamental insights of my scholarship to the larger policy and law reform world, and blogging is an important avenue for connecting scholarship with the public.

[Buddha]

Gosh, Adam, I didn't mean to get you so worked up, and I do (honestly) apologize for being so snarky in my post. I do find this overall issue to be one which is highly misrepresented - on both sides!!! And it really bothers me when people are either selective or wrong (or at least logically incorrect or imprecise) in their arguments. Again, I apologize for parts of my post.

But not for the overall motivation. Because while I agree that the calculation of 560K new jobs is just suggestive, it's also intellectually problematic and, quite frankly, could be dangerous. If I ever hear in a Congressional hearing that Professor Adam Levitin has computed that 560K jobs could be created by cutting interchange to zero (or whatever interchange level you actually posit in your post – see below), I'll come hunt you down and will call you out for posting that garbage.

Visa and MasterCard often run "regressions" where they relate economic growth across countries to the diffusion of credit cards. That's complete garbage. And it's also very dangerous. Because those who do not really consider the issues (and I will acknowledge that you do so in your articles, even if I don't always agree with you), not to mention those who aren't well informed (I believe that I've spanned the entire set of interested parties in Congress) will be misled by that analysis.

So I appreciate your blog. I acknowledge that your analysis is generally (but not always – agh, can’t help myself) sound, even if you have a particular point of view that often seems to cloud your analysis. And I again apologize for being a snippy little weasel in my previous post. (I’m not sure why you’re yelling at this Lao Tse person or writing about why you blog, but on some level, you gotta feel a bit flattered that anyone cares enough about your blog to criticize it to this extent. And what is this "anonymous" complaint? That's the nature of the web. Who the heck is Fraud Guy anyway?)

Regarding your response to my comment, I don’t think that you really "nailed" me on anything. On a few specific points:

(1) I apologize for pulling the wrong number ($48 billion instead of $34 billion - or whatever) out of your post. If online comment sloppiness is a logical fallacy (let’s call it "straw man"), then I am so very, very guilty.

(2) I don’t actually claim that I’m not bothered by the distributional effects of interchange fees. In fact, I think that I state that those distributional effects are one of the real concerns associated with interchange fees (with payment distortions being the other – just as you claim in your EPS discussion – although most of the discussion about this issue doesn’t focus on this crucial issue). My problem is that I don’t know how to evaluate those distributional effects. Sure, it sucks that the poor unbanked guy doesn’t get the rewards that I get from my AmEx card, but do we mess with the payment system to help that guy? Or do we use the standard tax system, which may be less distortionary, to transfer resources back to him? Or do we try to get him his own bank account and debit card (perhaps not a credit card, I’ll admit, especially if I’m feeling a bit paternalistic), so that he’s also reaping the benefits (sic – I concede) of cards? I don’t know the answer to this question, but I do know that when large retailers suddenly claim that they’re specifically concerned about the welfare of poor people who pay with cash or, even better from a PR perspective, food stamps, it’s a bit difficult to swallow.

(3) BTW, I don’t really care about AmEx vs. Visa/MC. I was just making a silly point about the fact that it’s not really interchange that we worry about. In fact, Visa/MC could do the same thing that they currently do without interchange. That’s all I was implying, albeit snarkily.

(4) The 560K Home Depot vs. 560K Citigroup workers wasn’t supposed to be taken literally. Suppose it was 560K Home Depot vs. 40K Citigroup workers? Who am I to say which is more valued? As I'm sure you recognized, the exact equality was just intended to make the comparison simple.

(5) The 50-50 mix also wasn’t supposed to be taken literally. I was just making two points. First, as a merchant, you should price based on your expected transaction costs (100% cards and 50-50 cards-cash being the simplest illustrative cases). If you don’t, you’ll probably be out of business pretty soon, as you’re systematically screwing up in your pricing. If you do, you’ll recoup your costs on average (and may get screwed – or make out well - if unexpected things happen, but that’s the nature of uncertainty.) The second point is that you’ll then lose money on each card transaction, even if you break even on average. Among other reasons, including the framing of this debate, this latter point is interesting because various proposals involve telling consumers how much they pay for each card transaction. But this gets to the fundamental economic/legal incidence issue – when you face a mix of payments, how do you communicate how much you are raising your price for a particular transaction? You may claim that the consumer is paying 2% more. Or that you’re absorbing costs are 2% higher. But both can’t be simultaneously true.

(6) Which gets to a key point, whether I assume that pass-through is 100%. Now, I implicitly assume that demand is completely inelastic, although I acknowledge this point and discuss what may happen if it’s violated in general terms (i.e., merchants may, in fact, be harmed, although it’s nowhere near the "legal incidence" calculation). Alternatively, merchants are perfectly competitive with constant marginal costs – that will give you the same basic price effects and welfare results for merchants. I suppose that we both need to be more precise in our assumptions about costs and competition, but when you pulled out 560K new workers, you were making some really, really, really strong assumptions.

(7) Last, regarding your point about merchants adjusting their employment margin rather than their price margin, I think that you need to be careful to distinguish the employment effects due to lower output from ancillary employment effects. That is, when you argue that employment may fall, it could mean that lower output (see point 6) requires lower labor inputs and, thus, employment falls. If you’re arguing that merchants view staffing as a characteristic of vertical differentiation (rather than a drop in staffing required for a lower level of output), then you need to be a bit more specific about who you have in mind.

In the end, I’ve enjoyed thinking about this stuff, prompted by your blog. You have a particular ax to grind (ergo, your blog) and, I suppose, so do I (although, to be honest, I’m not quite sure what it is). If you were writing a populist blog in support of the card networks, I’d also go after you. It’s just that you’re writing a populist blog in support of the merchants (and ostensibly consumers), and their arguments are often illogical, if not wrong. I do apologize, again, for the snarky tone of my comment. But I’ll also keep an eye on you.

Cheers!

[Buddha]

Here's a further logical problem that's been bothering me about this post: Suppose that one were to concede that employment would increase if interchange fees were decreased. In your telling, this employment increase would occur rather than a fall in prices (which, incidentally, would explain why one might not empirically find lower prices when interchange is cut - very convenient from the merchants' point of view as they could point to almost anything, say more or larger stores, as evidence of the beneficial effects of lower interchange fees.) If that's the story, however, are prices really higher because of interchange fees? That's the common story that merchants tell about the effects of interchange fees (esp. the higher prices that stick it to those on food stamps). But if lower interchange fees wouldn't yield lower prices, perhaps yielding higher employment instead, why would higher interchange fees yield higher prices, rather than simply lower employment? Please explain.

Again, if you get to the core point of my comments, one needs to be careful about double counting any harm, notwithstanding whether that harm actually reflects a net social harm.

[Joe F]

Interchange rate may be the straw that broke the average Americans back(among other fees). The lions share goes to a home mortgage. It seems to me that every one that wants a home should qualify for a 150k 3% home mortgage on there primary residence as a government mandate. Now this would be a stimulus.