If you lean against a bright red convertible parked outside a car dealership on a beautiful sunny day, it's easy to imagine yourself driving the car—on the open road, cool wind blowing in your hair, and without a care in the world. But research suggests you might be better off walking away from the dealership before you make a decision you may come to regret.
People are substantially more likely to buy a convertible on an unseasonably warm day, according to a study by Devin G. Pope, Chicago Booth associate professor, and his colleagues. When shopping for cars on a sunny day, people make the mistake of projecting their current tastes onto their future ones, much like we do when grocery shopping on an empty stomach. When you’re hungry, you stock up on more cans of soup than you’re ever likely to need. Similarly, when you’re at a dealership on a sunny day, a convertible looks especially appetizing. The mistake of assuming that the way we feel today is how we’ll feel in the future—a mistake psychologists call projection bias—has been studied before, but never with a decision as consequential as purchasing a car, and rarely with such statistical rigor.
Using data from more than 40 million car transactions from 2001 to 2008, Pope and his colleagues—Meghan R. Busse of Northwestern, Jorge Silva-Risso of UC Riverside, and Pope’s brother Jaren C. Pope of Brigham Young—tracked the relationship between car purchases and the weather in the specific area in which each purchase occurred. Their weather data included the temperature on a given day, as well as other factors including rainfall and cloud cover. Controlling for the time of year, the researchers determine how convertibles typically sell on a particular day, then look at what happens when the weather is colder or warmer than usual.
A rational consumer should be equally likely to buy a convertible on a sunny day as on a rainy one. But for every 20-degree increase in a day’s temperature, dealers sold 8.5 percent more convertibles than average. This was as true in November as it was in May, although the warmest months of the summer proved an exception, as temperatures then were already so high that a day that was 20 degrees warmer than usual made driving a convertible less attractive, rather than more so. A clear sky increased sales by about the same amount as a 16-degree increase in temperature.
The researchers conducted several additional analyses to ensure that the differences they found couldn’t be easily explained by another factor. For example, they tested for and ruled out the possibility that consumers simply shift their purchasing behavior from cold days to warm days, and they rule out the possibility that the results are driven by the ability to test drive a convertible on a warm day. They also find the same behavior holds true during chilly weather—people are more likely to buy 4x4 vehicles when the weather is abnormally cold, and particularly after a snowstorm.
But letting the temperature drive car purchases is a mistake. Pope and his colleagues find that convertibles purchased in nice weather, rather than on an average day, were 1.3 percent more likely to be traded in within a year.
The researchers suggest that consumers might benefit from a “cooling-off period” following their purchase, when they would have the opportunity to return the vehicle to the dealer at no additional cost. Even if a consumer doesn’t cool off immediately after buying a shiny new convertible, perhaps the weather will.
Meghan R. Busse, Devin G. Pope, Jaren C. Pope, and Jorge Silva-Risso, “Projection Bias in the Car and Housing Markets,” NBER working paper, July 2012.
George Loewenstein, Ted O'Donoghue, and Matthew Rabin, “Projection bias in predicting future utility,” The Quarterly Journal of Economics, 2003.