Abstract
In the loss domain, both practical and ethical considerations rule out the systematic use of an incentive-compatible procedure involving real losses. The experimental study presented here aims at investigating whether some easier-to-implement procedure could be adequately used. For that purpose, the subjects’ degree of risk aversion is compared across three payment conditions: a real-losses condition based on a random-lottery (incentive-compatible) system, which serves as a benchmark, and two challengers, namely a “losses-from-an-initial-endowment” procedure and a hypothetical-losses condition. As a by-product, our experimental design also allows us to investigate the impact of monetary incentives in the gain domain. The main result is twofold: no significant difference arises between the three payment conditions in the loss domain, while real and hypothetical choices significantly differ in the gain domain. Our results suggest that the use of monetary incentives may be more crucial in the gain domain than in the loss domain.
Similar content being viewed by others
Notes
To counter this difficulty, Schoemaker (1990) introduced an ingenious procedure. At the beginning of the experiment, the subjects were informed that 7 of them would be selected at random at the end of the experiment and asked to play a loss lottery (selected at random too) for real. But at the end of the experiment, after playing the lottery for real, they were told that they would not actually lose any money. Obviously, while involving genuine monetary incentives (since the subjects did expect to lose real money when making their choices), such an experimental design remains ethically acceptable (since the subjects did not actually incur any real loss). But the price to pay is rather high. First, it is based on deceiving the subjects, which conflicts with the norms of honesty and trust that prevail in economics (see Bardsley 2000 for instance; it is worth pointing out Bardsley’s proposal of a new experimental design, called the Conditional Information Lottery, that “offers all the benefits of deception without actually deceiving anyone”, p. 215). Second, it may dampen the credibility of further experiments involving losses and prevent the subjects from trusting the experimenter. Third, it may be hard to replicate: the subjects may not believe the experimenter who tells them that they may lose for real at the end of the experiment, which prevents the payment scheme from being efficient.
Thaler and Johnson (1990) first introduced a small probability of losing for real, but their subject pool (based on voluntary participation) was strongly biased toward risk seeking. So they modified their experimental design to make the experiment more attractive for risk averse people: (1) the probability of losing money was made very small (0.04); (2) the subjects were told that they would be allowed to take part in another experiment after the current one so they could recover their losses in case they previously lost money; (3) the subjects were told that those who would have lost money at the end of those two experiments would be given a bonus. It was then possible to recruit risk averse subjects, but the experiment no longer actually involved real losses.
Under both RDU and PT, risk attitude is meant to depend on two components, namely attitude toward consequences (encapsulated in utility) and attitude toward probabilities (captured through probability weighting).
According to the mechanism described above and in Appendix B.
The main advantage of this strategy is that it prevents between-tasks violations of stochastic dominance; its main drawback is that it may induce potential order effects in the subjects’ answers. Pilot sessions convinced us that the advantage was large enough to counterbalance the drawback.
Naturally, running the real session before the covered one as we did could have also affected behavior in the covered session. In particular, those subjects who incurred a loss in the real session could have been tempted to try to make up for this loss in the covered session. In actuality, we found no significant difference in behavior in the covered session between those 17 subjects who experienced a loss in the real session and those 29 subjects (including 9 subjects with a zero gain) who experienced a gain. Risk attitudes in the covered session were also regressed on real payments to study whether subjects took gains and losses from the second session into account in the last session tasks. For both gains and losses, the regression coefficients associated with these payments were never significantly different from zero.
For instance, the expected reward for the entire experiment was about €17. The expected loss in the second session was about €3.8, but it was counterbalanced by an expected gain of about €5.8 (the expected reward was thus about €2). For the 17 subjects who actually lost some real money at the end of the second session, the average loss was about €8 (2 subjects incurred the maximum possible loss of €20). At the end of the whole experiment, only 2 subjects actually lost some money from their own pocket (less than €5).
Similarly, the subjects were not told anything about the future use of the money they might lose, except that we would not put it in our pocket. At the end of the experiment, the subjects who had lost some money were asked to choose a charity to which we could send their money.
To be more specific, the subjects were told that a single choice situation (i.e. a table as in Appendix A) would be selected at random first, and that a single choice task (i.e. a line in the selected table) would then be selected at random and played out for real.
Note that we chose to provide the subjects with the initial endowment at the beginning of the session. Giving it some time (a few days or weeks) before the experiment, as in Bosch-Domenech and Silvestre (2010) for instance, would have certainly been more appropriate, since our strategy could be suspected of enhancing the prospect-theory-with-memory effect. But our idea was to introduce a simple payment scheme that could be easily replicated and used in any experiment involving losses. The problem with more sophisticated strategies is that they may induce the subjects to quit between the time they are given the endowment and the experimental session itself. Moreover, if no prospect-theory-with-memory effect arose with our basic procedure, one could reasonably expect that no such effect would occur when using a more sophisticated one.
This does not hold if Holt and Laury’s findings reflect for losses.
Lottery (0, 25%; 20, 75%) offers the same specificities as the one that induced a specific behavior in the loss domain—it offers both the worst and best possible consequences as well as, consequently, the largest spread between consequences. The conditions under which Holt and Laury (2002)’s above-mentioned result is likely to apply are thus satisfied. However, the other three lotteries do not share these peculiarities, so they are outside Holt and Laury’s framework.
Note that we did not intend to estimate the entire prospect theory model here. As we did not collect any information about the subjects’ behavior when faced with mixed prospects, we could not make any inference about loss aversion in our experiment. As a result, our estimations were restricted to either gains or losses.
In particular, the losses-from-an-initial-endowment procedure may induce a large cost if the subjects happen not to lose much money from the endowment.
References
Abdellaoui, M. (2000). Parameter-free elicitation of utilities and probability weighting functions. Management Science, 46, 1497–1512.
Abdellaoui, M., Bleichrodt, H., & Paraschiv, C. (2007). Measuring loss aversion under prospect theory: a parameter-free approach. Management Science, 53, 1659–1674.
Abdellaoui, M., Bleichrodt, H., & l’Haridon, O. (2008). A tractable method to measure utility and loss aversion in prospect theory. Journal of Risk and Uncertainty, 36, 245–266.
Abdellaoui, M., Baillon, A., Placido, L., & Wakker P. P. (2010a). The rich domain of uncertainty. American Economic Review, forthcoming.
Abdellaoui, M., l’Haridon, O., & Zank, H. (2010). Separating curvature and elevation: a parametric weighting function. Journal of Risk and Uncertainty, 41, 39–65.
Arkes, H. R., & Blumer, C. (1985). The psychology of sunk costs. Organizational Behavior and Human Decision Processes, 35, 124–140.
Arkes, H. R., Herren, L. T., & Isen, A. M. (1988). The role of potential loss in the influence of affect on risk-taking behavior. Organizational Behavior and Human Decision Processes, 42, 181–193.
Baltussen, G., Post, T., van den Assem, M., & Wakker, P. (2008). The effects of random lottery incentive schemes: Evidence from a dynamic risky choice experiment. Working Paper, Erasmus University Rotterdam.
Bardsley, N. (2000). Control without deception: individual behaviour in free-riding experiments revisited. Experimental Economics, 3, 215–240.
Battalio, R. C., Kagel, J., & Jiranyakul, K. (1990). Testing between alternative models of choice under uncertainty: some initial results. Journal of Risk and Uncertainty, 3(1), 25–50.
Beattie, J., & Loomes, G. (1997). The impact of incentives upon risky choice experiments. Journal of Risk and Uncertainty, 14, 155–168.
Becker, G. M., DeGroot, M. H., & Marschak, J. (1964). Measuring utility by a single-response sequential method. Behavioral Science, 9, 226–232.
Birnbaum, M. H. (1999). How to show that 9 > 221: collect judgments in a between-subjects design. Psychological Methods, 4(3), 243–249.
Bosch-Domenech, A., & Silvestre, J. (2010). Averting risk in the face of large losses: Bernoulli vs. Tversky and Kahneman. Economics Letters, 107(2), 180–182.
Bostic, R., Herrnstein, R. J., & Luce, R. D. (1990). The effect on the preference reversal phenomenon of using choice indifferences. Journal of Economic Behavior and Organization, 13, 193–212.
Bruhin, A., Fehr-Duda, H., & Epper, T. (2010). Risk and rationality: uncovering heterogeneity in probability distortion. Econometrica, 78(4), 1375–1412.
Camerer, C. F., & Hogarth, R. M. (1999). The effects of financial incentives in experiments: a review and capital-labor-production framework. Journal of Risk and Uncertainty, 19(1), 7–42.
Chateauneuf, A., & Cohen, M. (1994). Risk seeking with diminishing marginal utility in a non-expected utility model. Journal of Risk and Uncertainty, 9, 77–91.
Clark, J. (2002). House money effects in public good experiments. Experimental Economics, 5(3), 223–231.
Cox, J. C., & Grether, D. M. (1996). The preference reversal phenomenon: response mode, markets and incentives. Economic Theory, 7(3), 381–405.
Cubitt, R. P., Starmer, C., & Sugden, R. (1998). On the validity of the random lottery incentive system. Experimental Economics, 1, 115–131.
Davis, D. D., & Holt, C. A. (1993). Experimental economics. Princeton: Princeton University Press.
Etchart-Vincent, N. (2004). Is probability weighting sensitive to the magnitude of consequences? An experimental investigation on losses. Journal of Risk and Uncertainty, 28(3), 217–235.
Etchart-Vincent, N. (2009). Probability weighting and the ‘level’ and ‘spacing’ of outcomes: an experimental study over losses. Journal of Risk and Uncertainty, 39(1), 45–63.
Fennema, H., & van Assen, M. (1999). Measuring the utility of losses by means of the trade-off method. Journal of Risk and Uncertainty, 17, 277–295.
Gärling, T., & Romanus, J. (1997). Integration and segregation of prior outcomes in risky decisions. Scandinavian Journal of Psychology, 38(4), 289–296.
Gibbons, R. (1997). Incentives and careers in organizations. In D. Kreps & K. Wallis (Eds.), Advances in economic theory and econometrics, vol. II. Cambridge: Cambridge University Press.
Goldstein, W., & Einhorn, H. (1987). Expression theory and the preference reversal phenomena. Psychological Review, 94, 236–254.
Harrison, G. W. (1994). Expected utility theory and the experimentalists. Empirical Economics, 19, 223–253.
Harrison, G. W. (2006). Hypothetical bias over uncertain outcomes. In J. A. List (Ed.), Using experimental methods in environmental and resource economics. Northampton: Elgar.
Harrison, G. W. (2007). House money effects in public good experiments: comment. Experimental Economics, 10(4), 429–437.
Harrison, G. W., & Rutström, E. E. (2008). Risk aversion in the laboratory. In J. C. Cox & G. W. Harrison (Eds.), Research in experimental economics, 12. Bingley: Emerald.
Harrison, G. W., Johnson, E., McInnes, M. M., & Rutström, E. E. (2005). Risk aversion and incentive effects: comment. The American Economic Review, 95(3), 897–901.
Hertwig, R., & Ortmann, A. (2001). Experimental practices in economics: a methodological challenge for psychologists? Behavioral and Brain Sciences, 24, 383–451.
Hertwig, R., & Ortmann, A. (2002). Economists’ and psychologists’ experimental practices: How they differ, why they differ, and how they could converge. In I. Brocas & J. D. Carillo (Eds.), The psychology of economic decisions. New York: Oxford University Press.
Holt, C. A. (1986). Preference reversals and the independence axiom. The American Economic Review, 76, 508–515.
Holt, C. A., & Laury, S. K. (2002). Risk aversion and incentive effects. The American Economic Review, 92, 1644–1655.
Isen, A. M., & Patrick, R. (1983). The effect of positive feelings and risk taking: when the chips are down. Organizational Behavior and Human Performance, 31, 194–202.
Keasey, K., & Moon, P. (1996). Gambling with the house money in capital expenditure decisions: an experimental analysis. Economics Letters, 50, 105–110.
Keren, G., & Raaijmakers, J. (1988). On between-subjects versus within-subjects comparisons in testing utility theory. Organizational Behavior and Human Decision Processes, 41, 233–247.
Lazear, E. (2000). Performance, pay and productivity. The American Economic Review, 90(5), 1346–1361.
Lee, J. (2008). The effect of the background risk in a simple chance improving model. Journal of Risk and Uncertainty, 36, 19–41.
Mason, C. F., Shogren, J. F., Settle, C., & List, J. A. (2005). Investigating risky choices over losses using experimental data. Journal of Risk and Uncertainty, 31(2), 187–215.
Read, D. (2005). Monetary incentives, what are they good for? Journal of Economic Methodology, 12(2), 265–276.
Romanus, J., Hassing, L., & Gärling, T. (1996). A loss-sensitivity explanation of integration of prior outcomes in risky decisions. Acta Psychologica, 93, 173–183.
Schoemaker, P. (1990). Are risk-attitudes related across domains and response modes? Management Science, 36, 1451–1463.
Smith, V. L. (1976). Experimental economics induced value theory. The American Economic Review, 66, 274–279.
Smith, V. L., & Levin, I. P. (1996). Need for cognition and choice framing effects. Journal of Behavioral Decision Making, 9, 283–290.
Starmer, C., & Sugden, R. (1991). Does the random-lottery incentive system elicit true preferences? An experimental investigation. The American Economic Review, 81(4), 971–978.
Stott, H. P. (2006). Cumulative prospect theory’s functional menagerie. Journal of Risk and Uncertainty, 32, 101–130.
Thaler, R. H., & Johnson, E. J. (1990). Gambling with the house money and trying to break even: the effects of prior outcomes on risky choice. Management Science, 36(6), 643–660.
Tversky, A., & Kahneman, D. (1992). Advances in prospect theory: cumulative representation of uncertainty. Journal of Risk and Uncertainty, 5, 297–323.
Weber, M., & Zuchel, H. (2001). How do prior outcomes affect risk attitude? Comparing escalation of commitment and the house-money effect. Decision Analysis, 2(1), 30–43.
Acknowledgement
We are grateful to Mohammed Abdellaoui, Aurélien Baillon, Han Bleichrodt and Peter Wakker for their insightful comments on the paper. We also thank the editor as well as an anonymous referee for their highly valuable comments and suggestions. The authors acknowledge the support of Research Grant ANR05-BLAN-0345 from the French National Research Agency.
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix A: A typical choice situation
Appendix B: The recoding mechanism used for building each gain lottery
To allow the testing of Assumption A1.2., the gain gambles had to be built from the loss ones using the following mechanism. Suppose that we use a losses-from-an-initial-endowment procedure with an initial endowment of A > 0. Now, consider a subject who is tempted to deal with covered losses as if they were real gains by integrating the initial endowment A into the subsequent losses she may undergo. As a result, when faced with the covered losses prospect P−= (X, p; Y, 1-p) [with −A ≤ X < Y ≤ 0], she may behave as if she were actually confronting the positive prospect P+ = (X + A, p; Y + A, 1-p) [with 0 ≤ X + A < Y + A ≤ A.]. This “as if” (AI) Lottery P+ is denoted P +AI . Now, let us denote CE− the certainty equivalent of the subject for P−, and CE +AI her “as if” certainty equivalent for P +AI . CE +AI is thus given by CE +AI = CE− + A. Note that the first formula assumes that the initial endowment is completely integrated into subsequent losses. If the integration process is only partial, CE +AI will be given by CE +AI = CE− + B, with 0 < B < A.
From an empirical point of view, we obviously cannot observe whether or not a subject integrates her initial endowment into subsequent losses (be it partially or completely). Nevertheless, we can in fact determine whether she does or not (and test Assumption A1.2.), by comparing her behavior in the covered losses condition, after recoding these losses as gains by integrating the initial endowment into subsequent losses, and her behavior in the corresponding real gains condition. More specifically, all we have to do is compare her “as if” certainty equivalent for P +AI , namely CE +AI , with her certainty equivalent when facing the same lottery P+ but under a “real gains” (R) condition, denoted CE +R .
To permit such a comparison, each of the 11 gain prospects P+ has to be built as the positive counterpart of one of the 11 initial loss prospects P− = (X, p; Y, 1-p), using the simple formula P+ = (X + A, p; Y + A, 1-p) where A is the initial endowment.
Rights and permissions
About this article
Cite this article
Etchart-Vincent, N., l’Haridon, O. Monetary incentives in the loss domain and behavior toward risk: An experimental comparison of three reward schemes including real losses. J Risk Uncertain 42, 61–83 (2011). https://doi.org/10.1007/s11166-010-9110-0
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11166-010-9110-0