A few days ago, the government came out with their list of the top 10 vehicles being purchased under the Cash for Clunkers program. As you can see, they’re all cars except for the Ford Escape.

.

When this list was reported, the big deal was that most of the cars were produced by foreign companies (although most of them are built here in the United States). But what struck me was that there was only one SUV on the list and no pick-ups at all. I thought we were a country that loved our trucks and SUVs. Well, it turns out we are. The method the government used treated different versions of a vehicle (based on engine size and chassis) as completely different cars, so their list is biased in favor of vehicles that only come in one version. The Ford Escape comes in 6 different versions when you consider 2WD and 4WD models, hybrid versions and different engine sizes. SUVs and pick-ups tend to come in multiple versions, so most of those vehicles did not make the list. The Honda Fit, on the other hand, comes in 3 different varieties but they all have the same engine and chassis, so they were all listed as one vehicle, and it made the list. Edmunds grouped different versions of the same vehicle together and, voila, we have a completely different list.

On this new list, you have 2 trucks and 2 SUVs, much different than the previous list that had only 1 SUV. The Ford Escape rockets to from #5 to #1. The Honda Civic drops four places.

Personally, I think this is the right way to aggregate the data. The difference in MPG for the 2WD and 4WD Escape models is 2-3 MPG. For the Jeep Patriot, there’s only a 1 MPG difference. For a Hyundai Santa Fe, the 2WD and 4WD models somehow get the same MPG (not sure how the heck that works out, but that’s what it says). Putting together the 2WD and 4WD versions, when they’re basically the same car in terms of MPG, seems like the proper way to do this. 

This example illustrates that the way one aggregates data is important, and a small change can result in drastically different results. We calculate inflation rates in a variety of ways (CPI, PPI, GDP Deflator, etc.), each aggregating the changes in prices of individual goods differently. They all have their advantages and disadvantages and none of them is perfect. What’s the perfect measure? Ideally, “THE inflation rate” would be the increase in the amount of money the average person would have to have to be just as well off this year as they were last year as a result of changes in all the prices of available goods from last year to this year. Measuring price changes of individual goods is easy, but weighing them together can be tricky, and weighing changes in well-being as a result of those price changes is even harder – as you may know if you’ve taken some economics classes.

I understand the need for different inflation rate measures because of the millions of different prices in our economy, the difficulty in aggregating them, and the different audiences that might be concerned with specific types of goods. But something like a Cash for Clunkers Top 10 list should be simple to create. Yet the government can’t even define a car the way most people would.

Let’s hope they manage health care much more smoothly than they managed the Cash for Clunkers program.