Adv Behav Econ pdf

201
T I M E D I S C O U N T I N G
FIELD STUDIES
Some researchers have estimated discount rates by identifying real-world behav-
iors that involve trade-offs between the near future and more distant future. Early
studies of this type examined consumers’ choices among different models of 
electrical appliances, which presented purchasers with a trade-off between the
immediate purchase price and the long-term costs of running the appliance (as de-
termined by its energy efficiency). In these studies, the discount rates implied by
consumers’ choices vastly exceeded market interest rates and differed substan-
tially across product categories. The implicit discount rate was 17 to 20% for air
conditioners (Hausman 1979); 102% for gas water heaters, 138% for freezers,
243% for electric water heaters (Ruderman, Levine, and McMahon 1987); and 45
to 300% for refrigerators, depending on assumptions made about the cost of elec-
tricity (Gately 1980).
33
Another set of studies imputes discount rates from wage-risk trade-offs, in
which individuals decide whether to accept a riskier job with a higher salary.
Such decisions involve a trade-off between quality of life and expected length of
life. The more that future utility is discounted, the less important is length of life,
making risky but high-paying jobs more attractive. From such trade-offs, Viscusi
and Moore (1989) concluded that workers’ implicit discount rate with respect to
future life years was approximately 11%. Later, using different econometric ap-
proaches with the same data set, Moore and Viscusi (1990a) estimated the dis-
count rates to be around 2%, and Moore and Viscusi (1990b) concluded that the
discount rate was somewhere between 1 and 14%. Dreyfus and Viscusi (1995) ap-
plied a similar approach to auto-safety decisions and estimated discount rates
ranging from 11 to 17%.
In the macroeconomics literature, researchers have imputed discount rates 
by estimating structural models of life-cycle–saving behavior. For instance,
Lawrence (1991) used Euler equations to estimate household time preferences
across different socioeconomic groups. She estimated the discount rate of median-
income households to be between 4 and 13% depending on the specification. 
Carroll (1997) criticizes Euler equation estimation on the grounds that most
households tend to engage mainly in “buffer-stock” saving early in their lives—
they save primarily to be prepared for emergencies—and only conduct “retire-
ment” saving later on. Recent papers have estimated rich, calibrated, stochastic
models in which households conduct buffer-stock saving early in life and retire-
ment saving later in life. Using this approach, Carroll and Samwick (1997) report
33
These findings illustrate how people seem to ignore intertemporal arbitrage. As Hausman (1979)
noted, it does not make sense for anyone with positive savings to discount future energy savings at
rates higher than the market interest rate. One possible explanation for these results is that people are
liquidity constrained. Consistent with such an account, Hausman found that the discount rate varied
markedly with income—it was 39% for households with under $10,000 of income, but just 8.9% for
households earning between $25,000 and $35,000. Conflicting with this finding, however, a study by
Houston (1983, p. 245) that presented individuals with a decision of whether to purchase a hypotheti-
cal “energy-saving” device, found that income “played no statistically significant role in explaining
the level of discount rate.”

---