Working Paper: NBER ID: w18583
Authors: Hunt Allcott; Nathan Wozny
Abstract: It is often asserted that consumers undervalue future gasoline costs relative to purchase prices when they choose between automobiles, or equivalently that they have high "implied discount rates" for these future energy costs. We show how this can be tested by measuring whether relative prices of vehicles with different fuel economy ratings fully adjust to time series variation in gasoline price forecasts. We then test the model using a detailed dataset based on 86 million transactions at auto dealerships and wholesale auctions between 1999 and 2008. Over our base sample, vehicle prices move as if consumers are indifferent between one dollar in discounted future gas costs and only 76 cents in vehicle purchase price. We document how endogenous market shares and utilization, measurement error, and different gasoline price forecasts can affect the results, and we show how to address these issues empirically. We also provide unique empirical evidence of sticky information: vehicle markets respond to changes in gasoline prices with up to a six month delay.
Keywords: Gasoline Prices; Fuel Economy; Energy Paradox
JEL Codes: D03; D12; L62; Q41
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Consumers undervalue gasoline costs (D11) | Vehicle prices do not fully adjust to gasoline price forecasts (Q47) |
| Consumers undervalue gasoline costs (D11) | Vehicle purchase decisions are not fully adjusted (D19) |
| Gasoline price forecasts (Q47) | Vehicle prices respond with a delay (E30) |
| Measurement errors in gasoline price forecasts (Q47) | Significant biases in estimating consumer responses (D12) |
| Consumer behavior (D19) | Implied discount rate for future gas costs is just under 15 percent (H43) |
| Consumers undervalue gasoline costs (older vehicles) (D11) | Less sensitivity to gasoline price changes (D11) |
| Consumers undervalue gasoline costs (newer vehicles) (D11) | More sensitivity to gasoline price changes (L90) |