Grantee Research Project Results
Final Report: Understanding Observed Differences in Time-Preference Rates
EPA Grant Number: R827931Title: Understanding Observed Differences in Time-Preference Rates
Investigators: Gregory, Robin , Slovic, Paul , Knetsch, Jack , Lichtenstein, Sarah , Finucane, Melissa L. , Peters, Ellen , Arvai, Joseph , Burns, Katie
Institution: Decision Science Research Institute Inc. , The Ohio State University , University of Oregon
EPA Project Officer: Chung, Serena
Project Period: September 30, 1999 through September 29, 2001 (Extended to March 31, 2003)
Project Amount: $228,463
RFA: Decision-Making and Valuation for Environmental Policy (1999) RFA Text | Recipients Lists
Research Category: Environmental Justice
Objective:
Previous work by a variety of researchers has shown that the practice of using a single discount rate is not supported by behavioral (descriptive) research on how people typically make intertemporal choices. The main objective of this research project has been to improve understanding of the reasons people may discount future gains or future losses, as well as different types of future outcomes, at different rates. Studies focused on the investigation of three topics: (1) examining differences in intertemporal choices over gains and over losses; (2) understanding reasons different types of goods might be discounted at different rates; and (3) exploring the influence of affect and emotion on choices over time. The results of the experiments provide additional evidence of context-based variations in intertemporal choices that, if confirmed by later research, could influence the analysis of initiatives to protect the natural environment and public health and could contribute to a reconsideration of the conventional discounting practices of economic cost-benefit and policy analyses.
Summary/Accomplishments (Outputs/Outcomes):
The first year of the project focused on the development of experiments that tested different behavioral hypotheses regarding the construction of discount rates in response to various cues and signals derived from the problem statement and context. The design of these experiments involved problematic issues in that previous research on intertemporal choices had mixed, and often mixed-up, choices involving the four different possible combinations of gains and losses in the present and future. Therefore, we developed experiments that evaluated each of these combinations separately and were able to make comparative assessments of the implications for implied discount rates, as discussed further below.
The second and (partially funded) third years of the research have focused on interpreting the Year 1 and Year 2 experiments, designing and running targeted new experiments, and disseminating results through presentations and papers. The three guiding questions for these experiments have been: (1) whether there are systematic differences in expressed time preferences, depending on whether changes are framed as gains or as losses; (2) whether there are systematic differences in expressed time preferences for different types of goods (e.g., environmental, health, or financial), or for goods within the same domain but with different cognitively or affectively significant characteristics; and (3) how these questions might be relevant to economic and environmental policies followed by federal and/or state agencies. During the final year of the project (aided by a supplementary funding allocation), much of the attention of the investigators has been applied to thinking about implications of the experimental results for policies regarding intertemporal choices and for standard assumptions about pricing and consumer behavior made by economists and policy analysts.
Several of the more interesting results of the project remain exploratory, in that additional experimentation will be required before the findings are considered robust. Nevertheless, the research has made progress in identifying contextual reasons that lie behind expressed variations in time-preference rates and in examining the implications of these preferences for public policy. Three of the more important findings are described here; more detail is provided in the cited papers.
Intertemporal Choices Over Gains and Over Losses
The four combinations of present and future gains and losses, summarized in Figure 1 shown below, distinguish between gains and losses in the present (e.g., paying money to purchase a good versus receiving money for the sale of the good) and gains and losses in the future (e.g., having more or less of some good, item, or activity). Most studies of intertemporal preferences have asked Quadrant 2 questions, comparing near-term to long-term gains (or, in some cases from Quadrant 4, questions that compare near-term to long-term losses). This common practice can be expressed in the form of a choice between receiving X now or X + Y later: "Which would you prefer: $100 today or $150 1 year from today?"
Figure 1. The Four Combinations of Present and Future Gains and Losses
The more familiar valuation options, however, do not involve a choice between gains (or between losses) but rather ask for an individual's maximum willingness to pay (WTP) (see Quadrant 1), in which money is given up in the present to secure a future gain, or ask for the individual's minimum willingness to accept (WTA) compensation (see Quadrant 3), in which money is received in the present to accept or offset a future loss. Surprisingly, these WTP and WTA quadrants rarely have been used as the basis for empirical tests of discount rates. Only these frames, however, depict real exchanges involving tradeoffs between current and future consumption, with Quadrant 1 exchanges showing the present value of a future gain in terms of the amount an individual is willing to give up now to obtain it, and Quadrant 3 exchanges showing the present value of a future loss in terms of the amount demanded now to accept it.
The experiments conducted as part of this project therefore focused on eliciting Quadrant 1 (WTP) and Quadrant 3 (WTA) responses. A Quadrant 3 version of one type of question posed to participants is shown below:
Public authorities have found that concentrating health and safety expenditures in particular regions for a year at a time is a more effective means for reducing many accidents requiring hospitalization among adults. This means that accidents will go down in some areas, where efforts are concentrated, but they will go up in other areas where efforts have been reduced for that year. Assume that you live in a region where expenditures will be increased this year: adults will experience about 100 fewer serious accidents this year, but there will be more accidents 3 years from now when lower health and safety expenditures are made in your region. How many additional accidents in 3 years would just balance out the reduction of 100 accidents to adults anticipated this year?
It would be an even exchange if there were ______ additional accidents in 2004.
A robust finding of these experiments is that individuals use lower rates to discount the value of future losses than they use to discount the value of future gains. This is shown to be true over a wide variety of contexts (involving health and consumer choices, financial decisions, and environmental options) and, in the experiments conducted as part of this project, it is shown for the more appropriate WTP (Quadrant 1) and WTA (see Quadrant 3) measures.
Type and Characteristics of Goods
There is some experimental evidence that people may use different time-preference rates to discount the future value of different types of outcomes, such as those involving health or environmental outcomes, as opposed to financial. These results, however, have been inconclusive, and there is little research on understanding the reasons people might (or might not) vary the rate of discount for different goods.
The experiments conducted as part of this research project explored whether the important concern might be, not the type of good (e.g., environmental as opposed to financial), but rather the characteristics of the good, and whether these characteristics induce a consistent difference in participants? responses. For example, we anticipated that some environmental effects (e.g., potentially adverse impacts on an endangered species) might elicit a far stronger emotional reaction than others (e.g., improvements in litter control at a local park). Thus, different conclusions about the domain sensitivity of expressed time preferences could be reached, depending on the specific characteristics of the selected environmental, health, financial, or consumer item.
Several experiments were conducted testing this hypothesis, including questions asking participants to make pair-wise choices between options (involving receipt of different amounts of a good over various time periods) and other questions that asked subjects to evaluate streams of changes in both gains and losses over time. With both designs, results showed that discount rates varied according to the characteristics of the goods?a more fundamental level of meaning?rather than strictly with their type. Thus, the results of this research effort do not support earlier statements comparing discount rates among different classes or types of goods; such comparisons are likely to be misleading unless the specific characteristics of the goods are taken into account.
The Role of Affect and Emotion on Intertemporal Choices
Recent research in decision-making points to the importance of emotions and affect?including both the feeling states that people experience, such as happiness/sadness or calmness/arousal, and the qualities associated with a stimulus or event, such as its perceived goodness or badness?as key elements in how individuals form judgments and make decisions. Little research has been conducted, however, on how affect may influence choices involving intertemporal options. We believe that differences in affective responses might help to explain the somewhat surprising results noted above (i.e., experiments contrasting the type of good and the characteristics of goods).
An intuitively attractive hypothesis is that higher affect items?those for which people care most intensely?might exhibit a lower discount rate as compared to an otherwise similar but lower affect item, consistent with peoples' preferences for the good degrading less over time. If this is true, then some affectively important intertemporal exchanges?such as tradeoffs over time that involve declines in the health of children or the destruction of environmentally important habitats?may be subject to little degrading of values over time. As compared to the financial transactions that form the basis for much of conventional thinking about discount rates, these high-affect decision contexts might be relatively less sensitive to considerations of timing and relatively more sensitive to other contextual considerations, such as fairness, equity, or altruism. Time preferences, therefore, would not disappear necessarily, but other considerations would become more relevant and important as part of the process of constructing an appropriate intertemporal allocation.
Exploratory experiments conducted as part of this project support this perspective. Participants in these between-subject comparisons were asked two versions of each of two questions, parallel in construction except that one involved a higher affect item and the other a lower affect item. For example, one low-affect question concerned a program that, over time, would improve habitat for common birds, whereas the parallel high-affect question concerned a program that over time would improve the habitat for rare birds. Another low-affect question concerned audit investigations of a company yielding financial savings, as compared to a higher affect question concerning audit investigations of toxic emissions yielding the saving of lives. Affective responses were tested independently using five-point "intensity of feeling" scales.
We tested the following three hypotheses:
1. Future changes typically would matter more (as evidenced by a lower discount rate) for items higher in affect.
2. Differences due to affect would be more important for future allocations of a good than for the current year.
3. Future receipt of an item would matter less for losses than for gains; in other words, the future, positive value of gains would degrade more quickly than the future, negative value of losses. All three hypotheses were confirmed by the following results.
Across the three topics, we regard the findings showing lower discount rates for future losses than for future gains as robust. By contrast, the findings regarding characteristics of goods and the role of affect on implied discount rates are more exploratory. A natural response from the standpoint of an agency such as the U.S. Environmental Protection Agency would be: Where does this leave resource managers and other decision makers? There are at least three options, as discussed in the papers developed as part of this project. One option, of course, is to ignore people's expressed rates of time preference and carry on with the task of making choices according to established theory. The conventional use of a single discount rate provides a familiar mechanism for consistent decision making, and the domain-sensitive allocations implied by individuals may be wrong. A second option is to look to multiple rates, recognize that no single rate will do the job, and pay more attention to the attributes of choices so that discount rates can be linked to these characteristics. A third option is to sharply limit the use of discount rates as a key consideration in many situations characterized by streams of benefits and costs over time, following the logic that discount rates are highly variable and that other ethical or psychological considerations may be (relatively) far more important.
Although many questions remain, the results of this project contribute to, and help to refine, ongoing debates about the choice of discount rates among the broad community of economists, policy analysts, and behavioral researchers in psychology and the decision sciences.
Journal Articles on this Report : 1 Displayed | Download in RIS Format
Other project views: | All 11 publications | 5 publications in selected types | All 1 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Peters E, Slovic P, Gregory R. The role of affect in the WTA/WTP disparity. Journal of Behavioral Decision Making 2003;16(4):309-330. |
R827931 (Final) |
Exit |
Supplemental Keywords:
time preferences, discounting, decision making, cost benefit analysis., RFA, Economic, Social, & Behavioral Science Research Program, decision-making, Economics & Decision Making, intertemporal choices, alternative compensation, policy analysis, social psychology, surveys, discounting future gains, time-preference rates, community involvement, decision analysis, economic benefits, public issues, risk management, economic incentives, environmental values, preference formation, socioeconomics, cost/benefit analysis, environmental policy, long-term consequences, endowment effect, psychological attitudes, interviews, public policy, benefits assessment, econometric analysisProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.