Recently I spoke with David Rapson, Associate Professor in the Department of Economics and Co-Director of the Davis Energy Economics Program (DEEP) at the University of California at Davis (UC Davis) about a study he co-authored with colleagues from UC Davis, Yale and MIT called, “Attribute Substitution in Household Vehicle Portfolios.” Essentially, the team revealed an unintended consequence of tighter fuel standards as reported in CityLab: When a two-car household replaces one of its vehicles, a household that already owns a fuel efficient car tends to buy one that has less for its second car. This decision making erodes more than 60% of the fuel savings that first car should have yielded. Following are a couple of highlights from our discussion. You can listen or download the podcast below or listen to it in ITunes.
“Well, the main result is that there actually does seem to be a very strong relationship between the attributes of a car that you have when you buy a new car, the car that you already have in your portfolio, and the car that you end up buying. And so we have a thought experiment in mind that forms the basis for our empirical approach in this project. So when you think about it, for two-car households in an ideal world we would want them to flip a coin and drop one of those cars, just sell it or retire it. They’re then with a car that we call the “kept car,” and our research question of interest is how do the attributes of that kept car influence the choice of the next car that’s bought?
And in particular, we’re curious about how the fuel economy of the kept car affects the fuel economy of the bought car. And what we find is that it actually has a very strong effect. If you were to cause a household to buy a fuel efficient vehicle today, then the next time they replace a vehicle, they’re going to choose a less fuel efficient car. And there are lots of reasons why this might be the case, but the extent to which that shows up very clearly in the data, is the most surprising thing to me. It does appear that households really are viewing the cars that they own as a portfolio. So they may, for example, want to have one car that is fuel efficient and then another car that has more power or comfort or safety or seats, and those attributes that I just named are negatively correlated with fuel economy. If you care about both fuel economy and all those other things, then if you force a household to buy one car that’s more fuel efficient, they might then go and buy a less fuel efficient car next time, and that’s in fact what we find.”
“This is a belief that I and many energy economists have held for quite some time, and this study that we’re discussing today is just one other reason why we do, but I think taking a step back and looking at the incentives that fuel economy standards produce in the transportation economy is something that’s worth doing. If your goal is to reduce gasoline usage, or say reduce greenhouse gases, then fuel economy standards are going to produce some incentives that act in the direction of achieving that goal, but there are going to be unintentional incentives that accompany the way that it’s set up, that act in the opposite direction. I think it’s really important if we care about carbon emissions, carbon abatement, and the cost effectiveness of climate change mitigation policies, then we really should be concerned about the unintended consequences of this type of policy. And I can describe to you just in very plain English, a few of the things that are going on there with fuel economy standards.
The way they’re set up, fuel economy standards require a vehicle fleet-weighted average fuel economy for each manufacturer. So let’s say that the standard has set an average of 25 miles per gallon, then say Ford would have to sell cars the weighted average fuel economy of which exceeded that threshold. What that essentially does is it makes it more costly for them to sell gas guzzlers, which causes them to raise the price of that a little bit, so it’s like taxing gas guzzlers. And, it’s like subsidizing gas sippers or fuel efficient cars, because in order to comply with that subsidy they need to get the right mix. And on the face of it that sounds quite reasonable, to tax gas guzzles and subsidize gas sippers.
But when you actually think about it, if our goal is to reduce gasoline demand, we don’t want to subsidize cars that consume gasoline, we want to have fewer cars on the road that are consuming gasoline, so we don’t want to subsidize any of them even if they’re fuel efficient. So that’s one area where the policy is going to be imperfect. Another is that it’s putting households, putting consumers into cars that are less expensive to drive a mile. I think many of your listeners might be familiar with the term the “rebound effect.” This is a common feature of energy efficiency interventions where they reduce the cost of producing whatever the energy consuming durable is doing, so in this case we’re talking about transportation.
A fuel efficient car costs less to drive a mile and what happens when something costs less? We consume more of it, and so that incentive is operating in the opposite direction to what we want. We would like people to be driving less if our goal is to reduce the amount of gasoline that’s consumed. Our best read of the literature is that the rebound effect is large, it’s probably somewhere between 20 and 60 percent of the anticipated savings from making a car more efficient on average is going to be eroded by all of the follow on rebound effect incentives that are created. ”