Fuel economy requirements and household driving

Mark Hoekstra, Jonathan Meer, Steve Puller, Jeremy West 15 July 2015



Effects of regulating fuel efficiency on household driving

Reducing pollution from motor vehicles is of great importance to mitigating climate change. Yet, economists’ preferred policy – directly taxing transportation fuels – faces political barriers in the US. Instead, the primary policy is to directly regulate fuel efficiency via the Corporate Average Fuel Economy (CAFE) standards. Recently, these standards increased dramatically, with a 2025 target of 46.2 miles per gallon for new light-duty vehicles.

However, a well-known phenomenon, often called the ‘rebound effect’, suggests that drivers will undo some or all of these increases by responding to the increased efficiency with more driving. After all, the effective cost of each mile driven is now lower. Indeed, an expectation that 10% of the gains from increased fuel economy will be clawed back is built into government estimates of the effects of the new Corporate Fuel standards (NHTSA 2012). Yet, the magnitude of this effect is a contentious empirical question, with estimates ranging from 5% to 30% depending on the empirical approach and the length of time studied (Gillingham et al. 2013).

This question has mostly been investigated using variation in fuel prices rather than variation in fuel economy (see, for example, Greening et al. 2000). This is for good reason – while drivers have little control over fuel prices, they choose their own vehicles, and do so in part on the basis of how much they expect to drive. As a result, while both fuel prices and fuel economy change the cost of driving, it is much easier to establish a causal link between fuel prices and driving than between fuel economy and driving. 

However, there is an important difference between changes in fuel prices and changes in fuel economy. When gas prices change, vehicles – and thus the benefits from driving – stay the same, at least in the short run. But when government policy induces consumers to buy more efficient vehicles, other vehicle characteristics change as well. This reflects the fact that the bundle of characteristics that make up a car involves trade-offs – greater efficiency almost invariably comes at the expense of performance, size, and comfort. As a result, increasing fuel economy reduces the underlying desire to drive, whereas a short-run change in fuel prices does not. Thus, while the existing literature addresses the behavioural responses of policies such as gasoline taxes, the projected rebound effects of the Corporate Fuel standards are better measured by directly examining how fuel economy affects driving. However, to do so one must avoid the pitfalls associated with people choosing their own vehicles, and instead exploit as-good-as-random variation in vehicle choice. 

How fuel economy affects driving: New evidence

In recent research (West et al. 2015), we examine the rebound effect using administrative household data from Texas around the time of Cash for Clunkers (CfC) stimulus programme, which ran for eight weeks beginning in July of 2009. This programme incentivised households to purchase more fuel-efficient vehicles by providing a rebate of $3,500 or $4,500 towards the purchase of a new car if they traded in a functioning, insured, and previously registered vehicle with a fuel economy of 18 miles per gallon (MPG) or less; this old vehicle, the ‘clunker’, was then scrapped. This 18 MPG eligibility cut-off provides a quasi-random source of variation. Barely eligible households were incentivised to buy more efficient vehicles than barely ineligible households. However, the barely eligible and barely ineligible households were otherwise very similar in their demographics, previous driving behaviour, and other characteristics. Thus the barely ineligible households serve as a valid ‘control group’ for what the driving behaviour of the barely eligible would have been if they had not been induced to buy more fuel-efficient cars. To account for any shifts in the composition of buyers due to the timing of purchase, we focus on all new car buyers who purchased between July 2009 and June 2010, as previous research showed that the programme accelerated purchases by no more than one year (Hoekstra et al. 2015).  

  • As expected, Cash for Clunkers induced households to purchase more fuel-efficient vehicles.

Figure 1 shows the average fuel economy for new cars purchased between July 2009 and June of 2010 as a function of the mileage of the household’s ‘clunker’, which we designate in our data as the household’s oldest vehicle. Note that the average fuel economy for barely eligible households deviates from the trend and is higher than that for barely ineligible households. The discontinuity around the cut-off in Figure 1 may seem small, but we include all new vehicle purchases over a relatively long period to alleviate concerns about the types of households that chose to buy a car around the time of Cash for Clunkers. Because the number of vehicles purchased under the programme was small relative to the total number purchased in Texas over the entire year, the differences appear small, but are in fact statistically significant and non-trivial in size.

Figure 1. Average fuel economy for new cars purchased between July 2009 and June 2020

One might expect that this increase in fuel economy would induce more driving because the price of driving each mile has fallen. However, improvements in fuel economy are often accompanied by other changes in vehicle characteristics. We show that the more fuel-efficient vehicles purchased by the eligible households were cheaper, smaller, and less powerful.

  • Figure 2 shows that vehicles purchased by barely eligible households were less expensive than those purchased by barely ineligible households; similar patterns hold for vehicle characteristics that proxy for comfort, size, and performance.

This suggests that given current technological limitations and the cost of fuel-saving technologies such as hybrids, households respond to fuel economy restrictions by purchasing vehicles that are more fuel efficient, but are less desirable along other dimensions.

Figure 2. Average manufacturer suggested retail price for purchased vehicles 

What, then, is the net effect on driving? On one hand, the lower price per mile increases demand for driving, but on the other hand, the vehicles are smaller, less luxurious, have lower performance, and are potentially more dangerous in the event of an accident. These changes in vehicle characteristics can affect the demand for driving through several channels. First, more comfortable and higher-performing vehicles are simply easier to spend more time in. Second, passengers in heavier vehicles experience lower fatality rates in the event of an accident (Anderson and Auffhammer 2014).

  • We find that households did not respond to owning more fuel-efficient, downsized vehicles by driving more.

This finding is illustrated in Figure 3; the barely eligible households do not drive more miles in the year after purchasing a new vehicle. Thus, we find no evidence of a rebound effect in response to improved fuel economy.

Figure 3. Annual household vehicle miles travelled (VMT)

Concluding remarks

This research has implications for energy policy. Energy efficiency standards are used to reduce energy consumption from a variety of consumer durables, such as vehicles and household appliances. A common concern of economists is that a rebound effect will undermine reductions in energy usage. However, fuel-efficiency regulations like the Corporate Average Fuel Economy change not only the price-per-mile of driving, but also other characteristics of vehicles. Estimates of the traditional rebound effect, using changes in this price, ignore the fact that more fuel-efficient vehicles also have characteristics that impact the desirability of driving. Our results give policymakers some cause for optimism, as they suggest that second-best strategies such as the Corporate Fuel Economy used to combat the negative externalities associated with gasoline consumption are more effective than previously thought.


Anderson, M and M Auffhammer (2014), “Pounds that kill: The external costs of vehicle weight”, Review of Economic Studies, 81(2):535–571.

Gillingham, K, M J Kotchen, D S Rapson, and G Wagner (2013), “Energy policy: The rebound effect is overplayed”, Nature, 493:475–476, January 2013.

Hoekstra, M, S L Puller, and J West (2014), “Cash for Corollas: When stimulus reduces spending”, NBER Working Paper No. 20349. July 2014.

Greening, L A, D L Greene, and C Difiglio (2010), “Energy efficiency and consumption – the rebound effect – a survey”, Energy Policy, 28(6-7):389–401, June.

NHTSA (2012), Corporate average fuel economy for MY 2017-MY 2025 passenger cars and light trucks. Technical report, U.S. Department of Transportation: National Highway Traffic Safety Administration.

West, J, M Hoekstra, J Meer, and S Puller (2015), “Vehicle Miles (Not) Traveled: Why Fuel Economic Requirements Don’t Increase Household Driving.” NBER Working Paper No. 21194. May.



Topics:  Energy Environment

Tags:  energy policy, fuel economy standards, fuel efficiency, household driving, Cash for clunkers

Associate Professor of Economics, Texas A&M University

Assistant Professor in the Department of Economics, Texas A&M University

Associate Professor of Economics, Texas A&M University.

Post-doctoral Scholar, MIT