Sunday, August 14, 2005
Some Notes On The Economics Of The Used Car Market
by Tom Bozzo
At the Columnist Manifesto, Oscar offers his theory of the new versus used car purchases. To excerpt what strike me as the salient points:
First, let's answer the question of whether Oscar is right, or mostly right. Economic theory offers an an answer of yes, though with some potentially significant qualifications that I'll get to later. Ivy-educated lawprof Oscar has basically rediscovered one of the central results of the economics of asymmetric information, and specifically "adverse selection" — where quality is unobservable on one side of a transaction. This serves as an illustration both of the potential minor mischief that could be done with a time machine, were such a thing not to violate the known laws of physics, as well as Brian Leiter's periodically renewed critiques of economics' most prominent prize.
Suppose we set the way-back machine 37 years. Oscar is, I'd gather, between third and fourth grades or thereabouts, and I'm a precocious diaper-filler of five months. A couple of early middle-aged dudes show up with a paper that takes the essential ideas from Oscar's post and spices it up with enough math to impress the economics profession of the time, which we could give a title like "The Market for ‘Lemons’: Quality Uncertainty and the Market Mechanism." We could get this paper published in a top journal, say the Quarterly Journal of Economics.
Young Oscar and I would then be shoo-ins for the Johnny B. Clark Medal for achievement by economists under the age of 10. Following the research program to its logical conclusion, we could have been in a position to split the economics pseudo-Nobel with Joseph Stiglitz and Michael Spence, when the committee decided to honor the founders of information economics.
This probably would have irritated George Akerlof, who published a paper of that title in the QJE 35 years ago this month and went on to take 1/3 of the 2001 Bank of Sweden Prize in Economic Sciences with Spence and Stiglitz (*).
At the usual gross oversimplification risk, a main result from Akerlof's paper was a demonstration of how the absence of (or missing market for) credible quality information could lead to a market "equilibrium" in which the price buyers are willing to pay, given the prospect of buying a "lemon," is sufficiently low that only sellers of "lemons" are willing to supply their cars to the used car market. So, as Oscar suggests, there are low prices but no "deals" in the sense that one gets what one pays for.
As I said, there are significant qualifications. To offer a personal anecdote, two cars ago I bought a 1998 BMW M3 coupe that was 6-1/2 months old and had less than 1,000 miles on the odometer when I took delivery. I paid about $5,000 less than sticker price for a car that would not normally be discounted much as a new car. And it was a cream puff! What gives?
1. The segment of the market matters. In the realms of the woefully undertaxed, it's not uncommon for the fashion-minded to acquire and quickly unload whatever car is the flavor of the month. Mercedes CLS 55 today, BMW M5 tomorrow: nice work if you can get it. Similarly, in the luxury market segments, buyers can afford to indulge "princess and the pea" syndrome, and unload cars not for mechanical failures but because they turn out not to suit them so much. Plus, it seems like a lot of high-end sports cars get sold off due to divorce or downward financial mobility. Bottom line is these create a source of supply for lightly used non-lemons. Those motives may be too costly for someone to pursue in the econobox market.
2. Dealer reputation matters. The dealer that sold me the old M3 is the only show in town for Audi, BMW, Mercedes-Benz, Porsche and Saab. They do aspire to repeat business, which discourages them from knowingly putting lemons on the lot. When I got the M3, as a single guy with no immediate prospects for marriage, I had told myself that if I was still single and bitter in 2001, I was going to have my mid-life crisis early by getting myself a 911. As it happens, I wasn't and didn't, but had they pissed me off by sticking me with a lemon, they'd have risked locking themselves out of the eventual sales of Suzanne's current car, the probable successors to our current cars, etc. This is also why related-party transactions can offer higher-quality cars: the close friend or relative may care enough about the relationship not to risk it by trying to sneak a lemon past the goalie. This does not really alter Oscar's conclusion in that a dealer with a valuable reputation will be able to translate the reputation into higher prices.
3. The biggest recent innovation in the used car market, manufacturer certification of used cars, can be seen as an effort to provide a credible (and valuable) signal of quality. The promise of the more rigorous certification programs is a more extensive inspection than the $30 checkup from one's favorite mechanic (plus, items flagged in the inspection are supposed to be corrected, or the car can't be certified), backed up with a sometimes substantial warranty extension from the manufacturer. Since certified used cars can bear a premium of a couple thousand dollars over an equivalent non-certified car — which includes relatively useless dealer certifications, which are often little more informative than a "cream puff" sign in the windshield — this also doesn't alter Oscar's conclusion that you get what you pay for. Also, a few unscrupulous dealers have apparently taken the certified used car premium as an inducement to fraud, particularly in the certification programs of non-luxury brands.
4. Cars are, generally, highly reliable and have relatively long warranties. So while getting a lemon (or at least a car with Issues in need of Sorting Out) may be inconvenient, the financial burden of a late-model lemon is likelier than not to fall upon the manufacturer.
So I think Oscar has the economics more-or-less right. A used car can be a rational purchase choice, though not likely because it constitutes an exceptional bargain over a comparable new car: the low price is, in effect, a risk discount. Caveat emptor.
See also Isaac at Armchair Capitalists for a related take.
--------------------------
(*) Akerlof was not a recipient of the John Bates Clark Medal, given to a hard-charging economist under 40, but both Spence and Stiglitz were.
An open response to my buddy Oscar's "open letter" on the economics of car buying
At the Columnist Manifesto, Oscar offers his theory of the new versus used car purchases. To excerpt what strike me as the salient points:
...if you can afford to buy a new car for cash (without financing it), there's no point in buying a used car. Used cars, while cheaper than new cars, are almost never a better value, and are usually a worse value.Oscar also apologized for a long and humorless post — though I, for one, laughed heartily at the "2003 Cilantro." I've heard enough economist humor not to make representations about the answer to follow.
You and I are virtually never going to have better information about the value of a used car than the used car dealer...
With amateur private car sellers, the issue is one of mutual ignorance. Neither party has really good information about what the car is worth.
I think that there really is no new car premium. If cars lose disproportionate "value" when driven off the lot, it may be because this market is likely to have a disproportionate number of lemons, and the potential buyers' fear of buying a lemon pushes down the price.
First, let's answer the question of whether Oscar is right, or mostly right. Economic theory offers an an answer of yes, though with some potentially significant qualifications that I'll get to later. Ivy-educated lawprof Oscar has basically rediscovered one of the central results of the economics of asymmetric information, and specifically "adverse selection" — where quality is unobservable on one side of a transaction. This serves as an illustration both of the potential minor mischief that could be done with a time machine, were such a thing not to violate the known laws of physics, as well as Brian Leiter's periodically renewed critiques of economics' most prominent prize.
Suppose we set the way-back machine 37 years. Oscar is, I'd gather, between third and fourth grades or thereabouts, and I'm a precocious diaper-filler of five months. A couple of early middle-aged dudes show up with a paper that takes the essential ideas from Oscar's post and spices it up with enough math to impress the economics profession of the time, which we could give a title like "The Market for ‘Lemons’: Quality Uncertainty and the Market Mechanism." We could get this paper published in a top journal, say the Quarterly Journal of Economics.
Young Oscar and I would then be shoo-ins for the Johnny B. Clark Medal for achievement by economists under the age of 10. Following the research program to its logical conclusion, we could have been in a position to split the economics pseudo-Nobel with Joseph Stiglitz and Michael Spence, when the committee decided to honor the founders of information economics.
This probably would have irritated George Akerlof, who published a paper of that title in the QJE 35 years ago this month and went on to take 1/3 of the 2001 Bank of Sweden Prize in Economic Sciences with Spence and Stiglitz (*).
At the usual gross oversimplification risk, a main result from Akerlof's paper was a demonstration of how the absence of (or missing market for) credible quality information could lead to a market "equilibrium" in which the price buyers are willing to pay, given the prospect of buying a "lemon," is sufficiently low that only sellers of "lemons" are willing to supply their cars to the used car market. So, as Oscar suggests, there are low prices but no "deals" in the sense that one gets what one pays for.
As I said, there are significant qualifications. To offer a personal anecdote, two cars ago I bought a 1998 BMW M3 coupe that was 6-1/2 months old and had less than 1,000 miles on the odometer when I took delivery. I paid about $5,000 less than sticker price for a car that would not normally be discounted much as a new car. And it was a cream puff! What gives?
1. The segment of the market matters. In the realms of the woefully undertaxed, it's not uncommon for the fashion-minded to acquire and quickly unload whatever car is the flavor of the month. Mercedes CLS 55 today, BMW M5 tomorrow: nice work if you can get it. Similarly, in the luxury market segments, buyers can afford to indulge "princess and the pea" syndrome, and unload cars not for mechanical failures but because they turn out not to suit them so much. Plus, it seems like a lot of high-end sports cars get sold off due to divorce or downward financial mobility. Bottom line is these create a source of supply for lightly used non-lemons. Those motives may be too costly for someone to pursue in the econobox market.
2. Dealer reputation matters. The dealer that sold me the old M3 is the only show in town for Audi, BMW, Mercedes-Benz, Porsche and Saab. They do aspire to repeat business, which discourages them from knowingly putting lemons on the lot. When I got the M3, as a single guy with no immediate prospects for marriage, I had told myself that if I was still single and bitter in 2001, I was going to have my mid-life crisis early by getting myself a 911. As it happens, I wasn't and didn't, but had they pissed me off by sticking me with a lemon, they'd have risked locking themselves out of the eventual sales of Suzanne's current car, the probable successors to our current cars, etc. This is also why related-party transactions can offer higher-quality cars: the close friend or relative may care enough about the relationship not to risk it by trying to sneak a lemon past the goalie. This does not really alter Oscar's conclusion in that a dealer with a valuable reputation will be able to translate the reputation into higher prices.
3. The biggest recent innovation in the used car market, manufacturer certification of used cars, can be seen as an effort to provide a credible (and valuable) signal of quality. The promise of the more rigorous certification programs is a more extensive inspection than the $30 checkup from one's favorite mechanic (plus, items flagged in the inspection are supposed to be corrected, or the car can't be certified), backed up with a sometimes substantial warranty extension from the manufacturer. Since certified used cars can bear a premium of a couple thousand dollars over an equivalent non-certified car — which includes relatively useless dealer certifications, which are often little more informative than a "cream puff" sign in the windshield — this also doesn't alter Oscar's conclusion that you get what you pay for. Also, a few unscrupulous dealers have apparently taken the certified used car premium as an inducement to fraud, particularly in the certification programs of non-luxury brands.
4. Cars are, generally, highly reliable and have relatively long warranties. So while getting a lemon (or at least a car with Issues in need of Sorting Out) may be inconvenient, the financial burden of a late-model lemon is likelier than not to fall upon the manufacturer.
So I think Oscar has the economics more-or-less right. A used car can be a rational purchase choice, though not likely because it constitutes an exceptional bargain over a comparable new car: the low price is, in effect, a risk discount. Caveat emptor.
See also Isaac at Armchair Capitalists for a related take.
--------------------------
(*) Akerlof was not a recipient of the John Bates Clark Medal, given to a hard-charging economist under 40, but both Spence and Stiglitz were.
Comments:
<< Home
I'm glad my economics was more or less right, because my spelling was bad (I went back and corrected "amateur").
Funny you should keep mentioning "cream puffs," since I was just at the State Fair.
Funny you should keep mentioning "cream puffs," since I was just at the State Fair.
I think this discussion ignores the mindset of people who lease.
Lots of cars come off lease with unattractive terms for purchase. In addition, the leasee has usually committed to a monthly payment and would rather sign-up for a new lease for a new car rather than purchase the car.
Meanwhile the owner of the vehicle has assumed they will sell the vehicle at wholesale and has little incentive (or means) to convince the current owner to purchase.
So, the informational asymmetry for cars coming off lease is less likely to work against a used car buyer.
Lots of cars come off lease with unattractive terms for purchase. In addition, the leasee has usually committed to a monthly payment and would rather sign-up for a new lease for a new car rather than purchase the car.
Meanwhile the owner of the vehicle has assumed they will sell the vehicle at wholesale and has little incentive (or means) to convince the current owner to purchase.
So, the informational asymmetry for cars coming off lease is less likely to work against a used car buyer.
# posted by 
: Sat Aug 20, 09:47:00 PM
: Sat Aug 20, 09:47:00 PM
Anonymous: You're right that off-lease cars are an important factor in the market, and are largely neglected in this discussion (it's implicit in 'market segments are different' due to the greater prevalence of leasing in the luxury market).
I think the fundamental issue there is that many lessees lease because they don't want to keep their cars long-term, and so many leased cars hit the market only because the lease contract expired, and not because of any quality-related reason. (Though you might expect that purchase options would be exercised mostly on 'cream puffs.') That can be offset in part by lessees not having the same incentives as owners to keep their cars well-maintained, contractual wear-and-tear provisions notwithstanding. A motive for the included-maintenance programs for some luxury brands was actually to improve the quality of off-lease cars.
I think the fundamental issue there is that many lessees lease because they don't want to keep their cars long-term, and so many leased cars hit the market only because the lease contract expired, and not because of any quality-related reason. (Though you might expect that purchase options would be exercised mostly on 'cream puffs.') That can be offset in part by lessees not having the same incentives as owners to keep their cars well-maintained, contractual wear-and-tear provisions notwithstanding. A motive for the included-maintenance programs for some luxury brands was actually to improve the quality of off-lease cars.
...if you can afford to buy a new car for cash (without financing it), there's no point in buying a used car. Used cars, while cheaper than new cars, are almost never a better value, and are usually a worse value.
This is surely false. Used cars are a far better deal, as NPR's "The Car Guys" point out.
Of course, you have to have someone inspect the car to get around the information asymmetry problem. But it can be done quite easily.
This is surely false. Used cars are a far better deal, as NPR's "The Car Guys" point out.
Of course, you have to have someone inspect the car to get around the information asymmetry problem. But it can be done quite easily.
# posted by : Mon Aug 22, 07:54:00 AM
: Mon Aug 22, 07:54:00 AM
Anon 7:54: As you might note from the conclusion of the post, I don't disagree with the proposition that used cars can be an appropriate choice after everything is taken into account.
Since Oscar is a lawprof (progressive civil libertarian division) and not an economist, not to mention a friend, I gave him a pass on some of the exceptions I could have taken with the way he phrased his proposition. Rather, I reframed his question as a matter of whether you would expect not to "get what you pay for." In that context, I could take exception with Click & Clack via your characterization. You have to be precise about what you mean by "better deal," and be ready to explain what market failure allows the "better deal" to persist on the margin.
Since Oscar is a lawprof (progressive civil libertarian division) and not an economist, not to mention a friend, I gave him a pass on some of the exceptions I could have taken with the way he phrased his proposition. Rather, I reframed his question as a matter of whether you would expect not to "get what you pay for." In that context, I could take exception with Click & Clack via your characterization. You have to be precise about what you mean by "better deal," and be ready to explain what market failure allows the "better deal" to persist on the margin.
<< Home
