Is That New Car Smell Lemony?

In a Fox News piece, University of Maryland economist John Lott questions the notion that the sales price of a car plunges as soon as you drive it off the lot:

Many economists explain this drop as occurring because the people who are trying to resell their cars quickly are typically doing so to get rid of "lemons." Even if your virtually new car isn't a lemon, people who want to buy your car can't be sure, so they aren't willing to pay as much as your virtually new non-lemon car is really worth….Nice story—except it's wrong.

Lott notes that the seller of a used car can reassure the buyer by having the car's condition professionally certified, by offering a money-back guarantee, or by transferring the manufacturer's warranty. Such techniques, he argues, should enable the seller of an almost-new car to get close to the original price. Sure enough, Lott's analysis of used car prices in the Philadelphia area found that "used cars with only a few thousand miles on them sell for almost the same price as when new." He reports that "the certified used car price was on average just three percent less than the new car MSRP," compared to the 25 percent drop in value estimated by University of Chicago economist Steven Levitt and New York Times writer Stephen Dubner in their bestseller Freakonomics. 

Lott, whose new book is called Freedomnomics, has sparred with Levitt before, suing him for defamation over his disparaging comments about Lott's research on right-to-carry gun laws and crime. In January a federal judge dismissed the part of his suit based on a passage in Freakonomics that concludes, "Regardless of whether the data were faked, Lott's admittedly intriguing hypothesis doesn't seem to be true," since "when other scholars have tried to replicate his results, they found that right-to-carry laws simply don't bring down crime." The Chronicle of Higher Education reports that Levitt recently agreed to settle the rest of the suit by sending a "letter of clarification" to a third economist, John B. McCall, with whom he corresponded in 2005 about Lott's work on a special 2001 issue of The Journal of Law & Economics devoted to gun issues. Levitt erroneously told McCall that the issue "was not a peer-refereed edition of the Journal" and asserted that Lott had "put in only work that supported him." Levitt neglected to mention that he'd been invited to participate and declined.

The settlement makes Levitt look sloppy, but the lawsuit makes Lott look touchy. I'm not sure who comes out ahead, except those who enjoy academic sniping as a spectator sport.

Here is Dan Polsby's 1998 review of of Lott's book More Guns, Less Crime; the reason interview that Michael Lynch and I did with Lott in 1999; Robert Ehrlich's 2001 critique of Lott's work; and Lott's response. Fans of Lottiana, of course, won't want to miss Julian Sanchez's 2003 unveiling of Mary Rosh, an episode also mentioned in Freakonomics.

Comments

[Jonathan Hohensee]

Why legalized abortion leads to family breakdown, which creates more crime

? Why affirmative action in police departments leads to higher crime rates

? How women's suffrage led to a massive increase in the size of government

· Why women become more conservative when they get married and more

liberal when they get divorced

? How secret ballots reduce voter participation

? Why state-owned companies and government agencies are much more likely to engage in unfair predation than are private firms

? Why the controversial assertions made in the trendy book Freakonomics are almost entirely wrong

He's on our water polo leauge, but he spends all his time in the shallow end.

[highnumber]

I'm not sure who comes out ahead, except those who enjoy academic sniping as a spectator sport.

Hey! That's us! Woo-hoo!

[Geotpf]

I'll have to say, Freakonomics doesn't seem to have a precise political standpoint one way or the other, while the description of Freedomnomics makes it sounds like Michelle Malkin or Rush Limbaugh wrote it.

[Brian E]

Now hold on a second. The price for a certified used car doesn't necessarily have anything to do with the price that the person who purchased the car originally sold it far. As a matter of fact there are certified cars which can sell for more than the price of a new example of the same car because the manufacturer adds an additional warranty on the certified car.

The fact that dealers are selling the low-mileage used models for nearly the same as new cars is in fact evidence that the owners who sold those cars to the dealer took a substantial loss on the car; if they didn't, the dealer price for the used models would far exceed the new price.

The Blue Book is completely worthless as a source of information here. Every used car ad I've ever seen has advertised the price of the car as being "less than blue book". Even if private party transaction prices are close to the dealer price, most people are unwilling to deal with the hassle of selling a car private party.

[Brandybuck]

This book sounds interesting. I liked Levitt's book, but he drew too many conclusions from arbitrary correlations to my liking.

Here's the template for the typical Freakonomics chapter: find two seemingly unrelated sets of data, find a correlation between them, make no attempt to find an external source of the correlation, then pronounce a cause. For example, take a look at driving habits from around the world, and voting records. There is a correlation! Pretending that one causes the other, we can now definitively say that driving on the right side of the road causes people to vote for Bush!

[Brian Courts]

Here's the template for the typical Freakonomics chapter: find two seemingly unrelated sets of data, find a correlation between them, make no attempt to find an external source of the correlation, then pronounce a cause.

First, I have no feeling one way or the other about the ultimate accuracy of the conclusions Levitt draws and I really don't care at any personal level whether they are right or wrong. Certainly I expect there to be much legitimate criticism, but the criticism above is wildly inaccurate. Of course the statistical and other research details are necessarily left out of a book intended for a popular audience, but his published papers and other cited sources do indeed consider all those things brandybuck says he doesn't.

[MikeP]

The reality is that the price of a used car drops when it's driven off the dealer's lot not because a quick resale may indicate a lemon, but rather because the reputation of the seller changes dramatically when it is driven off the lot.

When you repatriate the car to the dealer's lot as a certified used car, its price quite correctly returns to the neighborhood of the new car price. But as Brian E notes, that in no way implies the original buyer gets that price.

[Mary Rosh]

But as Brian E notes, that in no way implies the original buyer gets that price.

However, when a frog jumps off a cold popsicle, it then causes the nearby fishing pole to hit the window pane. This conclusively proves that sale prices and peanut butter sandwich consistency are the root of dropping crime rates.

[Mo]

Many economists explain this drop as occurring because the people who are trying to resell their cars quickly are typically doing so to get rid of "lemons." Even if your virtually new car isn't a lemon, people who want to buy your car can't be sure, so they aren't willing to pay as much as your virtually new non-lemon car is really worth....Nice story-except it's wrong.

Actually, doesn't Lott's own research prove exactly that point. Once you get the car certified, guaranteed or add a money-back guarantee, you remove that uncertainty and (surprise, surprise) the price gets closer to the actual value of the car. Methinks Lott is trying to be clever and different by proving conventional wisdom.

[Other Matt]

The reality is that the price of a used car drops when it's driven off the dealer's lot not because a quick resale may indicate a lemon, but rather because the reputation of the seller changes dramatically when it is driven off the lot.

Actually, the ONLY reason that the price drops is that we've bought into the value drop simply by having it titled prior to purchase. There's no rationale for it, just what dealers have trained us to believe.

[Paul]

who want to buy your car can't be sure, so they aren't willing to pay as much as your virtually new non-lemon car is really worth....Nice story-except it's wrong.

Methinks I'm going to have to read further. Price: It's not what you (or Lott, or Levitt) says it is, it's what the market will bear. If I *can't* get anything more than 75% of the original sale price on my "virtually new" car with only a thousand miles on it, then the car is only worth 75% of the new price. Period.

And Brian E nails it. The used car won't possibly go for less than what the dealer paid the prior owner.

Try this: Buy a new car, drive it for a thousand miles-- maybe two-- then go back to the dealer and try to sell it back to them within two or three percent of the original price. See how fast you get laughed off the lot.

Certified vehicles on the dealer lot carry certain intangibles which raise their market value. Like, you have someone to go after if your certified used car breaks down a week after purchasing. Try to go after the guy with the mullet who sold it to you a week before from his trailer house who swore on a stack of bibles that it was just fine. There are clear differences. One entity (the car dealer) has much to lose. The guy with the mullet? Not so much.

[Brian Courts]

Actually, the ONLY reason that the price drops is that we've bought into the value drop simply by having it titled prior to purchase. There's no rationale for it, just what dealers have trained us to believe.

Whether you come down on the Levitt or Lott side, this market and mind manipulation stuff is nonsense. Clearly there are several potentially valid rationales for it as demonstrated by the comments above. And, to borrow a phrase, "I agree with MikeP." 🙂

[jb]

Paul has it. The fact that you can blame someone for your certified used car breaking, and that the law can make that blame stick, raises the value hugely.

In fact, I'd argue that that accounts for most of the value of a new car as well.

[Max]

This phenomenon isn't special to cars - you can see it with all types of goods. Peruse ebay and compare the price of "new" and "like new" goods. The "new" ebay price is very close to the mainstream retail price. "Like new" is much cheaper. The price difference is basically an insurance policy that the buyer demands, since he's taking a risk.

[madpad]

The first red flag should have been this statement:

Many economists explain this drop as occurring because the people who are trying to resell their cars quickly are typically doing so to get rid of "lemons."

On it's face, this is a ridiculous statement. In some 25 years of hearing that "a car drops (pick your number)% the minute you drive it off the lot," I've never once heard anyone explain it as some sort of 'lemon perception.'

The sole driving force in every iteration on this story is simply that the car has gone from new to used. That's it. I've never heard of anyone - economist, used car salesman or hapless buyer - ever bring up the lemon notion.

On top of that, Lott has always appeared more interested in agenda-driven economic analysis that actually producing solid economic research.

Lott's a wanker.

[Equality 7-2521]

madpad,

Umm... the "lemon" thesis is rather famous. Like it shows up in college economics textbooks all the time.

[jkii]

A pristine used car with just 0.1 mile has to sell for about 8% (CA, sales tax varies) less than the same new car. Sales tax makes this so. All in favor of a national sales tax, raise your paw.

[Matt J]

Nerd fight! Nerd fight!

[madpad]

Umm... the "lemon" thesis is rather famous. Like it shows up in college economics textbooks all the time.

It's been a while since I had economics but I don't remember that being in there. That, by the way, is not remarkable as there is a lot I've unlearned or relearned since then.

I will say that the notion that many people try to sell their cars quickly soon after buying them because they're trying to get rid of a lemon is silly and defies logic and my personal experience.

In a country where 10 year car warrantees and lemon laws are commonplace, not to mention a country where everyone and their grandmother sues at the drop of a hat, I find it hard to believe this is a big motivator.

While it's been many years since I took economics, I had a career in banking and currently have a career in marketing. Based on experience, I would say that buyers remorse, changes in income and just plain old stupidity are much more prevalent reasons than the lemon hypothesis.

[Brandybuck]

Of course the statistical and other research details are necessarily left out of a book intended for a popular audience, but his published papers and other cited sources do indeed consider all those things brandybuck says he doesn't.

I don't expect to see the statistics, but I would like to see a discussion of the factors studied. To the casual reader, it certainly looks like he's handwaving correlation into causation.

For example, his claim that abortion reduces crime. This is an extraordinary statement, which is of course, why his book is so fun to read. But immediately after he warns us that correlation does not mean causation, he starts throwing more correlation at us. He looks at different states, which shores up his correlation, but it doesn't demonstrate causation. Is there a third factor than can explain both? Could it be that Planned Parenthood also provides parenting advice? Or maybe states more likely to legalize abortion are also more likely to adopt other policies that would reduce crime? This is little to no discussion of other factors in the book.

[mediageek]

*Rolls 20 sided die*

I hereby summon Tim Lambert!

[Franklin Harris]

Freakonomics, which I've read, strikes me as a lot of interesting and plausible theories that need more research. Freedomnomics, which, admittedly, I've not read, strikes me as a book by someone who knows what conclusions he wants to reach and some questionable data that points in the chosen directions.

[Paul]

All in favor of a national sales tax, raise your paw.

Paw raised, with extreme prejudice.

Flat tax: Sucks. Full of problems.

National Sales Tax: A nearly flawless concept.

Brandybuck:

This is an extraordinary statement, which is of course, why his book is so fun to read. But immediately after he warns us that correlation does not mean causation, he starts throwing more correlation at us.

Bingo. And this little subject has been debated before on hit and run.

[Mary Rosh]

I just wanted to point out that I, personally, bought a used car from Professor Lott (who is dead sexy in person) for 3% off the new price.

[Other Matt]

Flat tax: Sucks. Full of problems.

Not to threadjack, but bullshit on that.

National Sales Tax: A nearly flawless concept.

Again, not to threadjack, but bullshit on that too. Why don't you just go for a VAT while you're at it?

[Mary Rosh]

How many more times can I post before people start noticing that Reason's writers didn't think Lott's sockpuppetry worth noting?

[Lamar]

Lott can fingerpuppet himself all day, why would Reason care?

[jf]

Fake Mary Rosh,

You must not have read the original post above.

Also, I'm not a fan of Lott based on what I know of him, but it's amazing how many people feel free to characterize his book without actually reading the thing. I have no intention of reading it, but then again I have no intention of reading Freakonomics either, mostly because economics is still in the pseudoscience stage as far as I'm concerned.

[Tim Lambert]

...What?!!....Who summoned me?!.....

[brian]

madpad

Umm... the "lemon" thesis is rather famous. Like it shows up in college economics textbooks all the time.

It's been a while since I had economics but I don't remember that being in there. That, by the way, is not remarkable as there is a lot I've unlearned or relearned since then.

I will say that the notion that many people try to sell their cars quickly soon after buying them because they're trying to get rid of a lemon is silly and defies logic and my personal experience.

"The Market for Lemons" is a paper that came out in 1970, which led to the authors (Akerlof et al) earning the Nobel prize in 2001.
I'm not sure when you took econ, so its entirely possible that it wasn't incorporated into the curriculum until later.

Also, I think it makes perfect sense that someone might turn around and sell their car to get rid of a lemon. If you buy a car from a dealer which turns out to be a lemon (or you get into an accident), won't you want to sell it to some unsuspecting sob to try and get your money back? That's were the problem appears--the buyers are suspicious that you might be doing that and therefore are willing to pay less, so the sellers of good cars are hurt and the sellers of lemons unfairly gain.

Basically, the lemon-sellers are getting subsidized by the legitimate sellers.

The sole driving force in every iteration on this story is simply that the car has gone from new to used.

Just to clarify, it being "used" is what gives potential buyers caution. With certainty of the car's non-lemony-status, a new car on the lot and a car just driven off the lot are of the same value--it's the same car. The price drops because of uncertainty of the car's status. The seller's motives for selling signal that the car might be a lemon.

Sorry, long post...

[Rimfax]

Mary Rosh,

Just because Lott is a touchy wanker (trust me I have the data to prove it...right here...somewhere) doesn't mean that his conclusions are false or that his research is flawed. He's a very easy target for ad hominem attacks, but that doesn't invalidate his research.

However, if you are making the point that he is ethically challenged, then yes, he seriously crossed the line with "Mary Rosh". So far, despite energetic attempts to find them, no one has found a comparable violation in his research.

[brian]

Other Matt

Flat tax: Sucks. Full of problems.

Not to threadjack, but bullshit on that.

National Sales Tax: A nearly flawless concept.

Again, not to threadjack, but bullshit on that too. Why don't you just go for a VAT while you're at it?

Agreed. The main benefit of a consumption tax is its ability to tax old capital, but this benefit erodes because people will spend their capital before the tax goes into effect.

Also, the biggest benefit people seem to think a consumption tax will have is that it doesn't discourage working like an income tax does. The problem is that this is completely false. People don't work for money--they work for the stuff money can buy. You can tax income, meaning an hour of work can buy less stuff, causing people (for the most part) to work less. OR you can tax consumption, meaning each hour of work can buy less stuff, likewise causing people (for the most part) to work less. Both have the same problem, contrary to popular belief.

[madpad]

brian,

I'm not suggesting that NO ONE sells their brand new car because they percieve it to be a lemon. Nor do I suggest that there's no market for lemons.

I merely assert that it's probably not as common a reason as others and is unlikely to be creating that much of an impact on the overall economy of selling low-mileage used cars.

Also, the paper you refer to predates lemon laws, long warrantees and a host of creative financing and credit issues that I'm suggesting may currently drive the issue.

[brian]

madpad

Ok. No disagreement here.

[ChrisO]

I have noticed that used cars tend to sell for a higher percentage of their original price nowadays than was the case 20 or 30 years ago. Heck, even the police department isn't "just giving them away" anymore like in the Blues Brothers.

A lot of this, I believe, is due to cars being better built now and designed to last for 200k miles or so. In the old days, 100k miles was asking a lot from a car, other than maybe a Mercedes or one of those unkillable Plymouth Valiants. In other words, 20 years ago, it didn't take more than a few months of use for a car to noticeably deteriorate from its showroom condition, so the drop-off in value was quite rational.

[Son of a!]

In other words, 20 years ago, it didn't take more than a few months of use for a car to noticeably deteriorate from its showroom condition, so the drop-off in value was quite rational.

Exactly. Between this and madpad's last comment, I'd say what we're seeing is that "everybody knows" cars fall off in value because of information that's hopelessly out of date.

An auto-enthusiast board I frequent hypothesizes that one of the big reasons car prices fall these days is that some manufacturers love to goose sales figures by selling lots of cars to rental fleets. This both pushes up the supply of used cars and lowers the average used car quality (because people tend to treat rentals like shit).

[Tom Walls]

Lots of authors have pen names; what's the big deal?

Mary Rosh is an anagram for Arm Horsy.

[madpad]

Lots of authors have pen names; what's the big deal?

Few of them use their 'pen name' to post glowing praise about themselves. Besides, who cares? Lott's still a wanker.

And it's also an anagram for Marry Hos

[Diesel Cars]

The things given are unanimous and needs to be appreciated by everyone.
=================
Mitchell
Diesel Cars