Levitt on Cash for Clunkers

by Tom Bozzo

Steven Levitt wonders aloud at the Freakonomics blog whether it could possibly be true that the CARS a/k/a Cash for Clunkers is about to run out of money, given the level of new vehicle sales in the U.S.:

The program went into place on July 24th. One week later, the program was said to be out of money.

In 2006, before the current ills of the automakers, the average number of new cars sold in a week in the United States was 125,000.

So if you believe the numbers, sales involving clunkers as trade-ins last week represented more than two times the weekly sales of new vehicles when the industry was healthy.

You have to go well into the comments thread before someone points out that pre-crisis, the number of new automobiles sold in the U.S. was more like 300,000/week — say 16 million sales/year divided by 52 weeks. As Spencer showed yesterday, second quarter sales were running around 9.5 million at a seasonally adjusted annual rate; by advanced math, that’s about 182,500 weekly sales. The difference in July vs. Q2 sales was about 1.7 million vehicles at SAAR, or 142,000/month. That would be among the greener shoots were it attributable to general economic conditions. So CARS actually could fund about two weeks’ worth of the difference between the Q2 sales rate and the pre-crisis rate. Though maybe given the recent track record of the Chicago school, being off by only a factor of 2.5 is not a bad performance.

A little more below the fold.

Consistent with reports that SUVs and pickups are getting traded in mainly on bread-and-butter cars and small crossovers, a number of high-volume models had very good months. For instance, Toyota’s Big Three (Camry, Corolla, and Prius) sold 23,413 more units in July than in June, a 39.4 percent increase. Prof. Hamilton’s auto-sales graph shows perhaps a modest June-July sales bump in the pre-crisis data. Honda, Ford, and Hyundai among others also had relatively good months by recent standards. While one of Levitt’s commenters boasts of trading a seldom-used pickup on a BMW Z4, lousy sales results from luxury makes suggest this is not a major problem (even though the $45,000 price ceiling is base price and hence relatively permissive); BMW’s sales report affirmatively cited a lack of CARS activity. [1]

So picking over the sales reports, it doesn’t seem hard to come up with 100,000 sales or thereabouts that could have been juiced by CARS, and the bottom line is that if that’s the case (and reported fuel economy improvement statistics are based off of the first 80,000 applications), then it’s perfectly possible that the program could run out of money quickly, or very quickly given a sufficient backlog of unprocessed applications.

Obviously the ordering of the program start and promulgation of the regulations is not a high point in government, but a look at the results suggests that some of the tea-partying reaction is unwarranted. Based on the reported fuel consumption difference of about 38 percent for the initial wave of applications, savings in fuel and avoided external costs from burning the fuel are on the order of $1200/year if the clunker was driven a typical amount (12,000 miles/year) with fuel at $2.50 and external costs of $1.50/gal. This is not a bad return on a $4,000 investment.

[1] Moreover, the $45,750 base price of even a “stripper” Z4 3.0 renders the model ineligible for CARS.