« What real world experiences should an economist have? | Main | A modest proposal »
Anti-market paper of the month
I introduce a new series of posts, titled as above, just to keep you all on your toes. And by the way, I've long wondered if ATM surcharges aren't taking advantage of a consumer intransitivity of indifference...you don't mind losing the first fifty cents but...
We estimate a structural model of the market for automatic teller machines (ATMs) in order to evaluate the implications of regulating ATM surcharges on ATM entry and consumer and producer surplus. We estimate the model using data on firm and consumer locations, and identify the parameters of the model by exploiting a source of local quasi-experimental variation, that the state of Iowa banned ATM surcharges during our sample period while the state of Minnesota did not. We develop new econometric methods that allow us to estimate the parameters of equilibrium models without computing equilibria. Monte Carlo evidence shows that the estimator performs well. We find that a ban on ATM surcharges reduces ATM entry by about 12 percent, increases consumer welfare by about 35 percent and lowers producer profits by about 20 percent. Total welfare remains about the same under regimes that permit or prohibit ATM surcharges and is about 17 percent lower than the surplus maximizing level. This paper can help shed light on the theoretically ambiguous implications of free entry on consumer and producer welfare for differentiated products industries in general and ATMs in particular.
The core intuition is that a given ATM often has monopoly power ex post, once you are there and need the money. Lower fees mean fewer machines but that still might be better than facing the mark-up. Here is the paper, whack it down if you can.
Addendum: Don Boudreaux offers commentary and some whacks.
Posted by Tyler Cowen on September 8, 2006 at 07:09 AM in Economics | Permalink
Comments
Well, let's ban supranormal markups on chocolate bars, soft drinks, etc., all sold by variety stores, too. They all are sold under conditions of situational monopolies.
For repeat purchases, like candy, pop, and ATM usage, if one outlet rips me off in the situational monopoly, I don't use them again unless it is to my benefit to use them. And sometimes it is -- e.g. when I'd rather pay an extra 50 cents than walk or drive an additional block or two.
In other words, there is something fatally flawed in their measures of consumer surplus if those measures don't include convenience and don't make allowances for the value consumers receive from increased choice sets. I know some people (Bob Frank?) think that increasing consumer choice is a bad thing, but I like having more options to having fewer.
In our town we have options: we can use an ATM at a bank and pay nothing extra, or we can use an ATM at a convenience store and pay a bit extra. This seems ideal to me. We get the extra choice, but we don't have to pay more if we're willing to go to our bank.
Posted by: EclectEcon at Sep 8, 2006 8:28:53 AM
Do I use ATMs that charge me a fee? NO. Am I in favor of banning ATMs that charge me a fee? NO. If people choose to use them, then that is their choice. Do you ban Starbucks because their coffee is three times what I can coffee for at the 7-11 two blocks away? The perception of quality, convenience and whether the price is worth it is up to the consumer.
Posted by: Murphy at Sep 8, 2006 9:22:39 AM
"intransitivity of indifference"
What the heck does that mean???
Posted by: Matt at Sep 8, 2006 9:40:48 AM
""intransitivity of indifference"
What the heck does that mean???"
I assume something close to the whole color paradox. (Where you can follow a trail of slightly changing shades of color, and any two neighboring shades look the same, but the ends look quite different.)
Fot example, let one be indifferent between a cost of $0 and a cost of $.50, and indifferent between a cost of $.50 and a cost of $1, but not indifferent between a cost of $0 and a cost of $1. Indifferences is not an equivalence relation; it's not transitive.
Posted by: John Thacker at Sep 8, 2006 10:05:47 AM
The reason banks charge fees for the use of ATMs is that if you're not our customer it simply isn't free for us to give you money. If you want to be a total pedant, a bank that isn't yours is actually extending you credit by letting you draw from one of its ATMs, they have to then draw those funds from your bank. IIRC it's usually processed as an ACH transaction, and it doesn't cost a ton to do but charging people fees for things is generally how banks stay in business.
Posted by: Timothy at Sep 8, 2006 10:06:56 AM
The reason banks charge fees for the use of ATMs is that if you're not our customer it simply isn't free for us to give you money.
Oh come on. These fees are way in excess of costs. How much do you estimate it costs for banks to provide this service? A penny a transaction plus overnight interest charges on the withdrawal of maybe another two or three cents? Go ahead and charge, but don't pretend the fees aren't hugely profitable.
it doesn't cost a ton to do but charging people fees for things is generally how banks stay in business.
You mean all that interest they collect is irrelevant? Seems doubtful. Service charges are about 25% of net interest income here.
Posted by: bernard Yomtov at Sep 8, 2006 10:49:19 AM
Compensating differentials do seem very important; More convenient ATMs have higher surcharges, while the least convenient (the bank itself) has the lowest surcharge of zero. An interesting question is who benefits from the gov't capping ATM fees? If poor people are more likely to avoid ATM fees entirely by going to their bank, capping the fees won't help them at all.
Posted by: Brian at Sep 8, 2006 11:26:56 AM
If total welfare under the two regimes is "about the same," I'll take the one with more freedom, thanks.
Posted by: Noah Yetter at Sep 8, 2006 12:08:57 PM
I can tell you from personal experience that banning surcharges in a GREAT idea. I am from Canada but now live in Sweden, surcharges are rampant in Canada and there are none for cash withdrawal in Sweden (I think what happened was that in the 80s the state made it clear to the banks that they would legislate if the banks could not come up with a fee-free system).
Canada has the most ATMs per capita in the world, and the reason is because the banks and companies specialized in putting ATMs in bars are making their easiest money from surcharges. In Sweden there are ‘only’ a lot of ATMs (banks want their customers to use them and their customers want banks to provide them).
In Canada I would always find an ATM immediately but if I wanted to take my money out without a change I would have to find one of my own bank’s ATMs. In Sweden there are fewer ATMs but I can use any bank machine. The result is that there is very little difference when it comes to convenience and I do not pay 5+ % on my money every time I take out 20 bucks (those service fees really added up to a lot when I was a poor student even though I tried to avoid paying them). I am a much happy customer in the Swedish ATM market (and the banks seem to be ‘surviving’).
But wait, how could this be? If customers really would rather have fewer ATMs and no fees why hasn’t the banking market in Canada produced this result? I must be wrong, it must be the case that what I really want, i.e. what I would choose in a freer market (i.e. a non-Swedish market), is more ATMs and more fees. Or is it that banks have cartel ‘like’ tendencies and that once they see the kinds of profits they can make if they ALL charge these fees they have no reason to compete against each other on fees? Want to here another cool thing about ATM fees, evidence from Canada shows that introduction of fees resulted in movement from small banks to big banks because they have more ATMs and customers have a better chance of finding a no fee ATM with a big bank. Markets a just awesome ALL THE TIME!
I am not an economist and I know empirical evidence is not really that impressive in your circles. I have full confidence that someone out there will be able to develop a model showing me why I do not want what I think I want, but for now I am going to enjoy the blissful ignorance of spending a few ‘saved’ dollars on a Friday night beer.
Posted by: Aaron at Sep 8, 2006 12:52:06 PM
You guys still use cash?
The only thing I do with an ATM is deposit funds from my wife's small business. . .
Posted by: Matthew Cromer at Sep 8, 2006 2:06:35 PM
Aaron, I don't get the sarcasm. Note the heading of the post "anti-market paper of the month".
I'm not saying this is necessarily the case, but one possibility is that the banks will make up for lost ATM fees with other kinds of fees.
I'm sure credit card companies also like ATM fees. I have good access to feeless ATMs, but in rare cases when I don't, I'll buy with a credit card rather than pay cash.
Posted by: Jeffrey at Sep 8, 2006 2:44:49 PM
John,
Perfect explanation. Thank you very much!
Posted by: Matt at Sep 8, 2006 2:48:51 PM
The whole point of being in business is to exercise some form of monopoly power so you can make greater than commodity profits.
Posted by: Steve Sailer at Sep 8, 2006 2:54:31 PM
The result is that there is very little difference when it comes to convenience and I do not pay 5+ % on my money every time I take out 20 bucks (those service fees really added up to a lot when I was a poor student even though I tried to avoid paying them). I am a much happy customer in the Swedish ATM market (and the banks seem to be ‘surviving’).
But wait, how could this be? If customers really would rather have fewer ATMs and no fees why hasn’t the banking market in Canada produced this result? I must be wrong, it must be the case that what I really want, i.e. what I would choose in a freer market (i.e. a non-Swedish market), is more ATMs and more fees. Or is it that banks have cartel ‘like’ tendencies and that once they see the kinds of profits they can make if they ALL charge these fees they have no reason to compete against each other on fees? Want to here another cool thing about ATM fees, evidence from Canada shows that introduction of fees resulted in movement from small banks to big banks because they have more ATMs and customers have a better chance of finding a no fee ATM with a big bank. Markets a just awesome ALL THE TIME!
You are conflating your preferences with the market en-masse.
I am not an economist
You are right there.
Posted by: Varangy at Sep 8, 2006 3:11:47 PM
Fee-less ATM's subsidize small banks at the expense of large banks.
Personally I am sick of people trying to exploit sunk costs by businesses and trying to take a bite out of their incremental prices that they use to justify the initial up front investment. The opportunity cost that results from fewer new investments in areas covered by such anti-market behaviour may be invisible, but it is still very real, and in the long run hurts those who are supposedly being helped by these anti-market behaviours.
In my opinion this is the real reason why rust belt states are having a hard time right now while right to work states are doing just fine, or at least much better. It is because when choosing to build a new factory with a large immovable upfront fixed cost, it is foolhardy to locate that factory where you know that coercive unions and other antimarket forces will conspire to rip you off and turn an otherwise great investment into a loust return or a money losing proposition.
Posted by: happyjuggler0 at Sep 8, 2006 3:33:44 PM
Credit unions are now pooling resources so you can withdraw money from each others ATM's without a fee or a greatly reduced fee. So we will soon have the Swedish and Canadian system operating side by side.
Posted by: joan at Sep 8, 2006 4:35:02 PM
Varangy,
"You are conflating your preferences with the market en-masse."
No I am not.
Rational self-interested agents do not want to pay more for the same service, i.e. rational agents choosing between the Canadian model and the Swedish model would choose the Swedish model. That is the point of my post. The only significant difference between the ATM service in Canada and Sweden is that in Sweden I do not pay surcharges.
Consumers prefer a market were cartel building is prohibited so that companies providing essential services cannot take out monopoly rents. Having your money in a bank is virtually unavoidable in a modern state while admittedly people do not need to use cash as much as they do. But given the continued importance of cash in modern economies I think it is safe to say that banks are exploiting the situation.
"'I am not an economist' You are right there."
If you are an economist and you did not recognize the above points in my post you are in trouble
Posted by: Aaron at Sep 8, 2006 4:49:15 PM
Aaron,
Varangy was being rude, but you miss the point. Not everyone shares your tastes. Others might be willing to pay more to have more ATM's. The level of convenience is the same for you because you are comparing the number of ATM's from your network in Canada to all ATM's in Sweden.
But some might like the possibility of cheap ATM's (same bank) combined with expensive ATM's (other banks) that are slightly more convenient. Poor students might not care about searching a minute and walking an extra block, but a successful professional might find it useful to save walking and search time by withdrawing at any ATM he finds. In that case the latter person would strongly prefer the Canadian situation to the Swedish one.
Posted by: ghost at Sep 8, 2006 4:57:04 PM
Ghost,
Of course the bigger the difference in access to ATMs between non-regulated and regulated ATM markets the greater the value of fee-based ATMs. But the point is that the difference is not that significant because I have access to all the bank's ATMs and that the banks together do provide very good coverage in my regulated market. All I am noting is that the empirical facts do not support the claim that fee-based ATM service is driven by customer demand for coverage. I can of course be wrong about how I am interpreting the empirical facts, but I am not missing the fact that consumers have varying preferences for different levels of coverage.
Posted by: aaron at Sep 8, 2006 5:48:07 PM
Hoo Boy - this is gonna get confusing - Aaron stole my handle, so I'll change mine.
Anyways, I am from Canada, and I have been to Sweden - funny coincidence.
I also happen to be an economist.
Three points:
1. When I use an ATM from another bank, I know the fees will double, so I just take out at least double what I normally would at my own bank to get some bulk buying discounts on my cash holdings. I respond by increasing my cash holdings when I know the transactions fees are higher (and cut the greedy bank out of a penny or two of interest in the interim).
2. I have not responded to the market by going to a big bank; instead, it's the smaller community trust banks that offer the best deals.
3. As an example of how different the Swedish and the Canadian banking experiences are, check out the JAK "zero interest" bank.
http://www.jak.se/
Posted by: AaronB at Sep 8, 2006 6:14:45 PM
For your next Anti-Market paper of the week, I suggest the one presented in this blog entry. http://www.thisplaceis.com/archives/26
U.C. Berkeley’s
Transportation
Seminar Series
September 8, 2006
4:00 to 5:00 p.m. in 240 Bechtel Hall
Donald Shoup, Ph.D.
Professor, Department of Urban Planning, UCLA
The High Cost of Free Parking
About 87 percent of all trips in the U.S. are made by personal motor vehicles, and drivers park free for 99 percent of these trips.
If drivers don’t pay for parking, who does? Everyone does, even if they don’t drive. ...
Donald Shoup will explain how faulty data from the Institute of Transportation Engineers helped get us into this mess, and how we can get out of it.
Posted by: Charlie Quidnunc at Sep 8, 2006 6:54:25 PM
"If customers really would rather have fewer ATMs and no fees why hasn’t the banking market in Canada produced this result?"
I think you may be missing a point about the nature of market equilibrium. The equillibrium price point doesn't deliver the maximum consumer welfare in a many consumer/many producer scenario. Instead the welfare benefit from the trade (which is equal to most the consumer would pay minus the marginal cost for the producer once the fixed costs have been paid) is split between producer profits and consumer welfare increases. Consumers would want to obtain the entire benefit in consumer welfare and leave the producer only enough to cover costs - this occurs when you only have one consumer, analogous to how the producer tries to get the entire benefit when it has no competitors. Hence the ideal situation for the consumer won't be the multi-producer/consumer equilibrium value should be the value that optimizes the total benefit (refer to a microecon textbook for how this is signficant in the context for efficiency, redistribution, and welfare).
What the paper appears to be getting at is that they think the market is behaving slightly like a monopoly since they'd expect a decline in total welfare if they were shifting from the multi-producer/consumer equilibrium point by banning fees, but it ended up approximately the same. On the other hand, it is hard to quantify the benefits of the higher number of ATMs for the consumer and how the costs will scale for increased ATM usage, so the details of the methodology would be very important to the outcome. I'd like to look at the model, but I don't have an NBER subscription, 'cause I'm not an economist either.
Posted by: MattXIV at Sep 8, 2006 7:02:33 PM
As I recall, ATM service charges went up over the years -- I never saw a $6 charge until rather recently -- even as the cost of computers and networks dropped rapidly.
The commenters should next explain how the enormous Ticketmaster surcharges are another triumph of the free market.
Posted by: Steve Sailer at Sep 8, 2006 7:31:49 PM
What's to whack down? Of course there are situations where actual market outcomes deviate from the perfect competition model. As long as no one raises some non-sequitur argument such like "Free markets occasionally deviate from the perfect competition model, therefore some people should be entitled to make allocative decisions concerning the belongings of others without prior consent," I don't see the problem. Oh, wait...
Posted by: James at Sep 8, 2006 9:56:10 PM
In San Franciso the strip clubs have ATM machines at the front door - loaded with $1s. Now that is a situational monopoly.
Posted by: Alex Tabarrok at Sep 8, 2006 11:25:32 PM
