Discounting Energy Efficiency

Often, when actors do not behave as one would logically expect them to behave, it is common to label this behavior as either shortsighted or irrational. The story of appliance discounting is another case where the models were inadequate to estimate and capture rational behavior.

Discounting

Discounting provides an estimate of what the future amount is worth in present terms. This allows comparisons between projects and investments over different time periods. In the case of appliances, the question is whether or not the difference in savings outweighs the difference in the current price of the appliance.

The standard, discrete equation to find the present value is:

     PV = FV/(1 + r)^t

where the Present Value equals the Future Value times the discount factor 1/(1 + r) t where r is the discount rate. Add up the benefits from each time period (the calculated present value of the savings for each year), and it will give the present value of the future energy savings. The present value of future savings can be compared to the price difference to determine whether it's worth spending more in the present to get the future benefit or not.

Early Measurements

Hausman (1979) looked at the purchase of air conditioners, the third most common appliance after refrigerators and washing machines, and it accounts for an average of 11.6% total energy use. Hausman estimated a mean discount rate of 26.4%. This rate is higher than what is normally used in estimating the value of future payoffs. When correlated with income, the discount rate ranged from 89% at low income to 5.1% at high income.

Gately (1980) did a similar study with refrigerators, comparing sets of nearly identical models varying only in energy efficiency and initial price. Gately estimated from 45-130% at a low electricity cost to 120-200% at a high energy cost. It seemed odd that companies would sell two models that are so similar when the choice is so obvious. Gately also brings up the role of market failures in creating a situation where one strictly inferior item can still be sold.

Later, Dreyfus and Vicusi (1995) looked at fuel efficiency in cars. They estimated 11%, 13%, and 17% discount rates for fuel efficiency in different models. While this was high, the rates were similar to the financing rates of cars at the time: 12.6% for new cars and 15.1% for used cars. They conclude that consumers cared enough about fuel efficiency for their behavior to be affected through altering market payoffs.

Hausman had several suggestions to address the perceived gap between discount rates and expected savings. Taxes could be used to charge for inefficiency, changing the up-front cost. The inverse, incentives or rebates, could be used as well. Governments could also simply regulate efficiency, dictating how efficient any given appliance has to be, or set what technology to use. Education could be used to inform buyers about the potential costs and benefits. Utilities could also lease the most efficient appliances so that their customers will have the best technology available to them without as much risk and investment as buying appliances on their own.

Evaluation

The mean discount rate from this survey was 22.5%, which is not particularly high compared to similar studies like the ones above. Of the respondents that seemed able to conceptualize the implications of the provided tradeoff, respondent based their decisions on the expected cost and expected benefit. Respondents attempted to factor in unwritten costs and this reflected upon their discount rates. Those who expected smaller costs had lower discount rates while those who expected higher unexpected or uncounted costs had higher discount rates. Houston found square footage, family size, age of house, and intention to conserve energy influenced the discount rate. Unlike some of the studies, there did not seem to be a correlation with income.

The majority of those who did not account for costs and benefits in making their decision responded with "don't know" or "uncertain". Those who didn't know were more likely to have lower income, larger families, smaller homes, and less interest or experience with energy efficiency.

The observed thought processes of those with and without skills to compare present and future costs were rational. Those that had the ability used it to come to a certain measure. Those who didn't admitted to their inability to make a reasoned decision. Moreover, those who did weigh the costs and benefits assumed that there were going to be unplanned costs involved with the adaptation of the new efficiency measure and took that into account when making their decision.

Payoffs

Dubin, Miedema, and Chandran (1986) set out to determine the price effects of energy efficiency. It was hypothesized that the efficiency gains from new technology would make energy effectively cheaper and increase use. Their study supports this hypothesis.

The study also looked at the overall efficiency gains from using these technologies. They found that realized conservation was significantly lower than expected. Cooling savings were as much as 13% below engineering estimates while heating was 8-12% below. It seems that while new technology increases efficiency, the gains in efficiency lower the price of electricity. Consequently, consumers use more of it than anticipated. By increasing their use, the gains from efficiency decrease and are partially eaten up by additional use.

Metcalf and Hassett (1999) turned their attention to the "energy paradox" of why discount rates for home improvement have been consistently high. The purpose of their study was to determine whether past estimates of discounting were overestimated because the efficiency benefits were overestimated.

Metcalf and Hassett looked at attic insulation and measured median returns of 9.7%. Over 75% of participants realized savings under 13.5%. The returns range from 2% to 25% discount over the range of -3% to +3% change in energy costs. These rates were not as significant as engineering estimates, which may report savings as high as 50%.

The adjusted estimated discount rate was then much more reasonable in comparison to the savings accrued. Not only that, but those who did not invest would not have saved money by making the investment, giving more evidence that decisions were rational. It would appear that individuals can make reasonable estimates as to the overall savings they may attain, translating reported costs into reality.

Conclusions

Later on, after this energy paradox had been widely documented, it was found that some people were able to make reasonable and consistent choices between present and future values. These individuals had the skills necessary to make these tradeoff decisions. However, not everyone did. Among those who did not, they were for the most part willing to state their lack of ability to make a decision. A few of them made a choice anyway despite not properly taking costs and benefits into account, and while this may be 'irrational' it is no better than making a guess.

It was also discovered that individuals who were making cost and benefit comparisons were also factoring in unmentioned costs. They expected the efficiencies to be lower than was reported. This observation led smoothly into the next phase of the story.

Eventually, studies compared the efficiency savings estimated by engineers to the reality of how people used these products. It was found that the savings were not necessarily as good as advertised. This is partially because more efficient energy use means you can do more with your energy, leading to people using their devices more than they normally would have, and thus resulting in expending more energy than if they'd remained with the same behavior.

The other finding was that savings were simply not as good as they were estimated to be. Whether this is due to behavior, or various factors in the house, or any other mechanism that made the technology less efficient than they are theoretically capable of, the realized savings were much lower than the original engineering estimates. When the actual savings were used to calculate discount rates, the rates were far more reasonable and it was seen that individuals made reasonable choices on whether or not something will in the end save them money.

It turns out that individuals were not making irrational decisions but were doing the best they could. And they accounted for factors intuitively that made a difference, and helped them make good decisions. Everyone knows that the gas mileage posted in a car window is not what you're going to see at the pump every time. We are able to make estimates based on what we know about our own behaviors, our experience with estimates that are provided to us in stores, and an understanding of random factors. We know that nothing is going to be as good as advertised, and act accordingly.

Of course, this does not mean that there is no place for incentives to induce people to switch to better technology. It just shows that people are smarter than findings may assume, and there are likely sensible reasons to explain their behavior. We should take more care in when we try to induce people to make savings investments and consider what the true impact will be, and whether or not it will in the end have been worth it for the individual or society. It is far too easy to dismiss what is observed as weaknesses of mental processes but looking deeper can provide very useful insights and allow us to have a better, and more effective, understanding of how people truly behave.

At any rate, it is clear that the estimates of engineers are no substitute for testing out how well technology does in the field. It may look good on paper, but it does not really reflect what is going on. We should neither systematically overestimate nor underestimate the repercussions.

References

Dreyfus, Mark K. and Vicusi, W. K. 1995. Rates of Time Preference and Consumer Valuations of Automobile Safety and Fuel Efficiency. Journal of Law and Economics, Vol. 38, No. 1, pp. 79-105

Gately, Dermot. 1980. Individual Discount Rates and the Purchase and Utilization of Energy-Using Durables: Comment. The Bell Journal of Economics, Vol. 11, No. 1, pp. 373-374

Goodin, Robert E. 1982, Discounting Discounting. Journal of Public Policy, Vol. 2, No. 1, pp. 53-71.

Hausman, Jerry A. 1979. Individual Discount Rates and the Purchase and Utilization of Energy-Using Durables. The Bell Journal of Economics, Vol. 10, No. 1, pp. 33-54

Houston, Douglas A. 1983. Implicit Discount Rates and the Purchase of Untried, Energy-Saving Durable Goods. The Journal of Consumer Research, Vol. 10, No. 2, pp. 236-246