Skip to main content
SpringerLink
Log in

Deterrence, expected cost, uncertainty and voting: Experimental evidence

  • Published:
Journal of Risk and Uncertainty Aims and scope Submit manuscript

Abstract

We conduct laboratory experiments to investigate the effects of deterrence mechanisms under controlled conditions. The effect of the expected cost of punishment of an individual’s decision to engage in a proscribed activity and the effect of uncertainty on an individual’s decision to commit a violation are very difficult to isolate in field data. We use a roadway speeding framing and find that (a) individuals respond considerably to increases in the expected cost of speeding, (b) uncertainty about the enforcement regime yields a significant reduction in violations committed, and (c) people are much more likely to speed when the punishment regime for which they voted is implemented. Our results have important implications for a behavioral theory of deterrence under uncertainty.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Similar content being viewed by others

Notes

  1. For simplicity, we shall use the term “deterrence mechanism” as a catch-all phrase for all enforcement tools.

  2. For example, Viscusi and Evans (2006) examine the differences between behavioral (anterior) probabilities and posterior assessments of the risk associated with an event after an individual receives additional information about the likelihood of an event, finding that anterior and posterior probabilities differ considerably. In another analysis, Viscusi and Hakes (2003) examine the use of risk ratings in determining the self-assessed versus actual survival probabilities at different ages. Again, these probabilities differ from one another quite a bit.

  3. The concept of uncertainty can take several forms in this analysis. For example, individuals could be unaware of the probability and fine, just the fine, or just the probability of apprehension. Savage (1954) claims that uncertainty may be treated similarly to risk; however Ellsberg (1961) conducts a series of experiments to show that risk and uncertainty are, in fact, different notions. In follow up work by Halevy (2007), a different form of uncertainty (compound lotteries) is determined. In the context of our experiment, implementing the uncertainty from Halevy (2007) allows the probability and fines for several enforcement regimes to be known, but the enforcement regime that is implemented is only probabilistically known. Thus, as seen in Halevy (2007), individuals are uncertain about the expected cost of committing a proscribed activity because they must reduce compound lotteries into simple probabilities. This form of uncertainty provides an environment that permits participants in the experiment to determine their preferred enforcement regime. See the discussion in Section 2 for more detail.

  4. Lazear (2004) also discusses the effect of police enforcement on speeding by noting that when few police are available to enforce speeding, it is best to announce their locations ex ante. Alternatively, when many police are present, it is best to not announce their locations. This detail does not play a significant role in our analysis, as police budgets or availability are not a direct concern in this analysis.

  5. Becker (1968) noted that enforcement efforts and sanctions are substitutes. For example, it is often the case that decreases in the number of police officers on highways and increases in speeding fines occur simultaneously.

  6. In addition, the authors use probabilities that are not on the linear portion (probabilities ranging between 0.2 and 0.8) of the S-curve in Wu and Gonzalez (1996). Wu and Gonzalez (1996) noted that individuals did not accurately perceive the probability of an event when it is either very unlikely (close to zero percent) or very likely (close to 100%).

  7. Note that in Bar-Ilan and Sacerdote (2004) the identification comes from the fact that fines are changing. In Ihlanfeldt (2003), it is argued that lower transportation costs reduce the transportation costs of committing a crime and, thus, it assumed that committing a crime is now effectively cheaper. That said, it is not clear that the individual committing the crime will be certain of the probability of being apprehended or fined when they travel to a different city to commit a crime.

  8. See Hutchinson and Yates (2007) for further details.

  9. Other very notable early work in this area also includes Von Neumann and Morgenstern (1947) and Savage (1954).

  10. But see Fox and Tversky (1995), who find that ambiguity aversion is reduced or eliminated in a between-subjects design.

  11. See Halevy (2007) for more detail. We thank Yoram Halevy for valuable discussions on this topic.

  12. In a certain sense, if one considers the inability to reduce compound lotteries as reflecting a weaker form of uncertainty, our experimental results might be considered to be a lower bound on the effects of uncertainty.

  13. Our design does not explicitly take into account the guilt, shame or stigma that might result from being caught committing an infraction. In a sense, this could be considered to be a part of the fine, but one that is costless to administer. Although we see no way to test for this in the data, it seems reasonable that people may have an aversion to simply being caught, which would lead to less speeding than otherwise.

  14. An implicit assumption of this experimental design is that speeding beyond the posted speed limit has a negative externality, as discussed in Ashenfelter and Greenstone (2004) and DeAngelo and Hansen (2009), in that it can increase the number of fatalities on the roadway. There are, however, opponents to the belief that increases in speed augment fatalities (see Lave (1985)).

  15. Of course, people may very well not be risk neutral, so that some risk-averse people might choose not to speed in periods 1–20, while some risk-seeking people might choose to speed in periods 21–30. Nevertheless, if an individual’s per se attitude towards risk does not change over the course of an experimental session, risk preferences may not affect our within-subject analysis.

  16. We chose to expose people to regimes before initiating voting, as it seemed that one should have some experience with some actual regimes before beginning to vote; this sets up the possibility of order effects. Nevertheless, we can test for trends over time in the decision to speed over each 10-period block, by regime and treatment. We find a significant effect in only one of the 12 tests (periods 21–30 in the first treatment with the regime with the smaller fine), as might be expected for 12 comparisons and a 5% significance level.

  17. It is true that this creates another layer of compounding, so that we do not in fact have a pure distinction between risk and uncertainty. Nevertheless, this aspect of compounding seems considerably easier, and in any case, we have a distinction induced by the greater complexity of the lotteries involving voting.

  18. We can use this test because we have within-subject data on the various blocks and regimes. The logic of this test is that if people are behaving randomly, we should expect as many people to speed more frequently in regime 1 as there are people who speed more frequently in regime 2. If these numbers differ a great deal, this indicates that behavior is not random.

  19. In this paper, we round off each p-value to the third decimal place. All tests are two-tailed, unless otherwise indicated.

  20. Note that these results extend what is observed in the tax compliance literature. That is to say, we do not observe increased compliance with the law when the preferred regime is instituted (as seen in Alm et al. (1993); Alm et al. (1999), and Feld and Tyran (2002)).

  21. We also note that there is no significant difference between either the speeding rates across the first 10 periods of settings 1 and 2 (61.6% versus 57.9%), or across periods 11–20 of setting 1 and periods 21–30 of setting 2; this is reassuring since they are identical decisions for each comparison. The Wilcoxon rank sum test (See Siegel and Castellan 1988) gives Z = 0.08 for the first comparison and Z = −1.05, neither close to statistical significance.

  22. Note that the speeding rates in block 3 in setting 2 are not significantly different than those in block 2 of setting 1.

  23. In addition to the results already presented, in block 3 of setting 1, 61.0% of the votes are in favor of regime 1 (with a detection probability of 3/5 and a potential fine of 0.833 units as opposed to a detection probability of 4/5 and a potential fine of 0.625 units with regime 2).

  24. In both Tables 3 and 4, the random-effects assumption that the unobserved individual effect is not correlated with any of the explanatory variables is violated when voting behavior is included in the regressions. To check for the robustness of the regression results, we have therefore performed both logit regressions and OLS regressions with fixed effects. The results are qualitatively nearly identical, with no serious changes in the significance levels of any of the explanatory variables. Since the random-effects regressions are appropriate for some of the columns, we choose to report these results in all cases.

  25. Having been caught (or not having been caught) speeding in the last period has no significant effect in our regressions. In principle, there is no reason to expect such an effect, as the outcome from speeding in the past period should have no effect on the actual expected cost of speeding. In fact, while there is no effect in the aggregate (people sped 80.2% of the time after having been caught speeding in the last period, compared to 76.6% of the time when they sped in the last period and were not caught), there is a considerable effect on individuals. In fact, the difference in these two rates exceeds 25 percentage points for 39% of the population. However, these differences go in opposite directions, effectively eliminating any aggregate effect. While it might seem natural that people are less likely to engage in a certain behavior after having a bad outcome (even though there is nothing to actually be learned), many people also seem to believe in the “law of averages.”

  26. The precise figures are 48.9%, 58.6%, and 37.9% for the 87 participants in blocks 1, 2, and 3, respectively, in setting 1, as well as 68.4%, 44.7%, and 68.4% for the 38 participants in blocks 1, 2, and 3, respectively, in setting 2.

  27. We report the results of probit regression models in Appendix B.

  28. From a theoretical standpoint, Halevy and Feltkamp (2005) discuss hedging in an environment with uncertainty.

  29. In the roadway speeding environment it is true that knowledge of the fines is available to all potential violators. However, most individuals are not accurately informed about these fines. Moreover, judges often allow violators to plead a lesser charge.

  30. For example, law enforcement agencies can implement uncertainty over the probability of being caught by setting up enforcement “blitzes” that randomly change. Eeckhout et al. (2010) discuss random crackdowns by law enforcement and determine that this could be part of an optimal policing strategy, specifically for speeding.

  31. As noted in Sunstein et al. (2000), a low probability of apprehension could suggest stealthiness on the part of the perpetrator. This claim is not refuted in our analysis; however, it should be noted that a lower probability of apprehension could appear to encourage stealthiness because the expected cost is perceived to be lower (relative to a higher probability/lower fine combination that has equivalent expected costs).

References

  • Alm, J., Jackson, B. R., & McKee, M. (1993). Fiscal exchange, collective decision institutions and tax compliance. Journal of Economic Behavior and Organization, 22, 285–303.

    Article  Google Scholar 

  • Alm, J., McClelland, G. H., & Schulze, W. D. (1999). Changing the social norm of tax compliance. Kyklos, 52, 141–171.

    Article  Google Scholar 

  • Anderson, L. R., & Stafford, S. L. (2003). Punishment in a regulatory setting: Experimental evidence from the VCM. Journal of Regulatory Economics, 24(1), 91–110.

    Article  Google Scholar 

  • Ashenfelter, O., & Greenstone, M. (2004). Using mandated speed limits to measure the value of a statistical life. Journal of Political Economy, 112(S1), S226–S267.

    Article  Google Scholar 

  • Baker, T., Harel, A., & Kugler, T. (2004). The virtues of uncertainty in law: an experimental approach. Iowa Law Review, 89, 1–43.

    Google Scholar 

  • Bar-Ilan, A., & Sacerdote, B. (2004). Response of criminals and noncriminals to fines. Journal of Law and Economics, 47(1), 1–17.

    Article  Google Scholar 

  • Bebchuk, L. A., & Kaplow, L. (1992). Optimal sanctions when individuals are imperfectly informed about the probability of apprehension. NBER Working Paper 4079.

  • Becker, G. S. (1968). Crime and punishment: an economic approach. Journal of Political Economy, 76(2), 169–217.

    Article  Google Scholar 

  • Block, M., & Garety, V. (1995). Some experimental evidence on differences between student and prisoner reactions to monetary policy and risk. Journal of Legal Studies, 24(1), 123–138.

    Article  Google Scholar 

  • Charness, G., & Levin, D. (2005). When optimal choices feel wrong: a laboratory study of Bayesian updating, complexity, and affect. American Economic Review, 95(4), 1300–1309.

    Article  Google Scholar 

  • Charness, G., & Levin, D. (2009). The origin of the winner’s curse: a laboratory study. American Economic Journal – Micro, 1(1), 207–236.

    Article  Google Scholar 

  • DeAngelo, G., & Hansen, B. (2009). Life and death in the fast lane: Police enforcement and roadway safety. Working Paper.

  • Di Tella, R., & Schargrodsky, E. (2004). Do police reduce crime? Estimates using the allocation of police forces after a terrorist attack. American Economic Review, 94(1), 115–133.

    Article  Google Scholar 

  • Eeckhout, J., Persico, N., & Todd, P. (2010). A theory of optimal random crackdowns. American Economic Review, 100(3), 1104–35.

    Article  Google Scholar 

  • Ehrlich, I. (1996). Crime, punishment, and the market for offenses. Journal of Economic Perspectives, 10(1), 43–67.

    Article  Google Scholar 

  • Eliaz, K., & Frechette, G. (2008). Don’t put all your eggs in one basket!: an experimental study of false diversification. Mimeo.

  • Ellsberg, D. (1961). Risk, ambiguity, and the savage axioms. Quarterly Journal of Economics, 75, 643–659.

    Google Scholar 

  • Feld, L. P., & Tyran, J. R. (2002). Tax evasion and voting: an experimental analysis. Kyklos, 55, 197–222.

    Article  Google Scholar 

  • Fischbacher, U. (2007). z-Tree: Zurich toolbox for ready-made economic experiments. Experimental Economics, 10(2), 171–178.

    Article  Google Scholar 

  • Fox, C., & Tversky, A. (1995). Ambiguity aversion and comparative ignorance. Quarterly Journal of Economics, 110(3), 585–603.

    Article  Google Scholar 

  • Garoupa, N. (2003). Behavioral economic analysis of crime: a critical review. European Journal of Law and Economics, 15(1), 5–15.

    Article  Google Scholar 

  • Greiner, B. (2004). An online recruitment system for economic experiments. In K. Kremer, & V. Macho (Eds.), Forschung und wissenschaftliches Rechnen. GWDG Bericht 63, Ges. für Wiss. Datenverarbeitung, Göttingen, 79–93.

  • Halevy, Y. (2007). Ellsberg revisited: an experimental study. Econometrica, 75(2), 503–536.

    Article  Google Scholar 

  • Halevy, Y., & Feltkamp, V. (2005). A Bayesian approach to uncertainty aversion. Review of Economic Studies, 72, 449–466.

    Article  Google Scholar 

  • Harel, A., & Segal, U. (1999). Criminal law and behavioral law and economics: observations on the neglected role of uncertainty in deterring crime. American Law and Economics Review, 1, 276–312.

    Google Scholar 

  • Heath, C., & Tversky, A. (1991). Preference and belief: ambiguity and competence in choice under uncertainty. Journal of Risk and Uncertainty, 4, 5–28.

    Google Scholar 

  • Hutchinson, K. P., & Yates, A. J. (2007). Crime on the court: a correction. Journal of Political Economy, 115, 515–519.

    Article  Google Scholar 

  • Ihlanfeldt, K. R. (2003). Rail transit and neighborhood crime: the case of Atlanta, Georgia. Southern Economic Journal, 70(2), 273–294.

    Article  Google Scholar 

  • Jolls, C., Sunstein, C. R., & Thaler, R. H. (1998). A behavioral approach to law and economics. Stanford Law Review, 50, 1471–1550.

    Article  Google Scholar 

  • Knight, F. (1921). Risk, uncertainty, and profit. Boston and New York: Houghton Mifflin.

  • Lave, C. (1985). Speeding, coordination, and the 55 MPH limit. American Economic Review, 75(5), 1159–1164.

    Google Scholar 

  • Lazear, E. P. (2004). Speeding, tax fraud, and teaching to the test. NBER Working Paper 10932.

  • McCormick, R. E., & Tollison, R. D. (1984). Crime on the court. Journal of Political Economy, 92(2), 223–35.

    Article  Google Scholar 

  • Mookherjee, D. (1997). The economics of enforcement. In A. Bose, M. Rakshit, & A. Sinha (Eds.), Issues in economic theory and policy, essays in honor of Tapas Majumdar (pp. 202–249). New Delhi: Oxford University Press.

    Google Scholar 

  • Mookherjee, D., & Png, I. P. L. (1994). Marginal deterrence in enforcement of law. Journal of Political Economy, 102, 1039–66.

    Article  Google Scholar 

  • Polinsky, A. M., & Shavell, S. (1979). The optimal tradeoff between the probability and magnitude of fines. American Economic Review, 69(5), 880–91.

    Google Scholar 

  • Posner, R. A. (1985). An economic theory of the criminal law. Columbia Law Review, 85, 1193–1231.

    Article  Google Scholar 

  • Rubinstein, A. (2002). Irrational diversification in multiple decision problems. European Economic Review, 46, 1369–1378.

    Article  Google Scholar 

  • Sah, R. K. (1991). Social osmosis and patterns of crime. Journal of Political Economy, 99(6), 1272–1295.

    Article  Google Scholar 

  • Savage, L. J. (1954). The foundations of statistics. New York: John Wiley.

  • Siegel, S., & Castellan, N. (1988). Nonparametric statistics for the behavioral sciences. Boston: McGraw-Hill.

    Google Scholar 

  • Stigler, G. J. (1970). The optimum enforcement of laws. Journal of Political Economy, 78(3), 526–36.

    Article  Google Scholar 

  • Sunstein, C. R. (1997). Behavioral analysis of law. University of Chicago Law Review, 64, 1175–94.

    Article  Google Scholar 

  • Sunstein, C. R., Schkade, D. A., & Kahneman, D. (2000). Do people want optimal deterrence? Journal of Legal Studies, 29(1), 237–53.

    Article  Google Scholar 

  • Viscusi, W. K. (1986). The risks and rewards of criminal activity: a comprehensive test of criminal deterrence. Journal of Labor Economics, 4(3), 317–40.

    Article  Google Scholar 

  • Viscusi, W. K., & Evans, W. N. (2006). Behavioral probabilities. Journal of Risk and Uncertainty, 32, 5–15.

    Article  Google Scholar 

  • Viscusi, W. K., & Hakes, J. K. (2003). Risk ratings that do not measure probabilities. Journal of Risk Research, 6(1), 23–43.

    Article  Google Scholar 

  • Von Neumann, J., & Morgenstern, O. (1947). The theory of games and economic behavior. Princeton, NJ: Princeton University Press.

  • Wu, G., & Gonzalez, R. (1996). Curvature of the probability weighting function. Management Science, 42, 1676–1690.

    Article  Google Scholar 

Download references

Acknowledgements

We would like to thank Ted Frech, Yoram Halevy and Joel Sobel for helpful comments. We would also like to thank participants of the Canadian Experimental and Behavioral Economics Research Group at the Canadian Economics Association 2009, the 2009 EWEBE meetings in Barcelona, and seminar participants at Vanderbilt University Law School. DeAngelo acknowledges financial support from the Department of Economics at UCSB. Charness acknowledges support from the National Science Foundation and the Dean of Social Sciences at UCSB.

Author information

Authors and Affiliations

  1. 110 8th Street, Troy, NY, 12180, USA

    Gregory DeAngelo

  2. Department of Economics, Rensselaer Polytechnic Institute, 3405 Sage Laboratory, Troy, NY, 12180, USA

    Gregory DeAngelo

  3. Department of Economics, University of California, Santa Barbara, CA, 93106, USA

    Gary Charness

Authors
  1. Gregory DeAngelo
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Gary Charness
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Gregory DeAngelo.

Appendices

Appendix A—Sample experimental instructions

1.1 Experiment instructions—Setting 1

General Rules: No talking. No use of cell phones. No looking at neighbor’s screens.

You are about to participate in an experiment on decision-making carried out by a researcher from UCSB. During this session, you can earn money. The amount of your earnings depends on your decisions and on the decisions of the other participants in this session. The session consists of 30 rounds. Your earnings will be determined by randomly choosing one payout from rounds 1–10, 11–20, and 21–30. The payout mechanism is explained in detail below. Your earnings will be paid to you in cash in private.

Your decisions are anonymous and confidential.

The experiment is split into 2 separate sections. The first section consists of 10 rounds and the second consists of 20 rounds. Instructions for the first 10 rounds are provided below. Instructions for the last 20 rounds will be given upon completion of the first 10 rounds.

Rounds 1–10: You are attempting to travel to the same destination 10 separate times—much like a commuter travels to work every day—and will receive a payout for each separate trip. You will have the option to speed or not when traveling. There will be two potential enforcement regimes—which you will be told before deciding whether to speed or not—and the probability that either regime will be instituted is equally likely. An enforcement regime is both a probability of getting caught when speeding and a corresponding fine, denoted f, which will be imposed if you are caught speeding. An individual that speeds and arrives at the destination without being apprehended will receive a payout of $1.00 per round. An individual that speeds and is apprehended for speeding receives a payout of $1.00-f per round, where f is the fine that is imposed by the enforcement regime that is instituted. Note that 0 < f < 1 always. Lastly, an individual that obeys the law and does not speed receives a payout of $0.60 per round.

There are two potential enforcement regimes:

  1. Regime A:

    Chance of getting caught = 1/3, Fine = $0.90

  2. Regime B:

    Chance of getting caught = 2/3, Fine = $0.45

Regime A and Regime B are equally likely—i.e. there is a 50% chance that either regime is randomly chosen.

1.2 Examples

Situation 1: Regime A is randomly selected. There are three possible payouts:

  1. Payout 1:

    Participant does not speed and receives a payout of $0.60.

  2. Payout 2:

    Participant speeds, is not caught and receives a payout of $1.00.

  3. Payout 3:

    Participant speeds, is caught and receives a payout of $0.10 (= $1.00−$0.90).

Situation 2: Regime B is randomly selected. There are three possible payouts:

  1. Payout 1:

    Participant does not speed and receives a payout of $0.60.

  2. Payout 2:

    Participant speeds, is not caught and receives a payout of $1.00.

  3. Payout 3:

    Participant speeds, is caught and receives a payout of $0.55 (= $1.00−$0.45).

Payouts: The payout that you will receive for rounds 1–10 will be determined by randomly choosing one of the 10 rounds and then multiplying the payout of that round by 5. Consider the following possible payouts:

  1. Example Payout 1:

    Suppose that round 5 is randomly chosen as the payout round and that you decide to not speed in round 5 and so your payout for round 5 is $0.60. Therefore, your payout for rounds 1–10 is $3.00 (=5*$0.60).

  2. Example Payout 2:

    Suppose that round 3 is randomly chosen as the payout round and that you decide to speed in round 3 and were not caught and so your payout for round 3 is $1.00. Therefore, your payout for rounds 1–10 is $5.00 (=5*$1.00).

  3. Example Payout 3:

    Suppose that round 7 is randomly chosen as the payout round. In round 7 you decided to speed, Regime A is instituted and you are caught speeding. Your payout for round 7 is $0.10. Therefore, your payout for rounds 1–10 is $0.50 (=5*$0.10).

  4. Example Payout 4:

    Suppose that round 9 is randomly chosen as the payout round. In round 9 you decided to speed, Regime B is instituted and you are caught speeding. Your payout for round 9 is $0.55. Therefore, your payout for rounds 1–10 is $2.25 (=5*$0.55).

Total Payouts: Your total payout is determined by multiplying a randomly chosen round in each block (1–10, 11–20, 21–30) by 5 and then adding a $5.00 show up payment. The payout is explained mathematically below:

$$ Total{ }Payout = (Payou{t_{{1 - 10}}} + Payou{t_{{11 - 20}}} + Payou{t_{{21 - 30}}})*{5} + \$ {5} $$

Rounds 11–20: As in the first ten rounds, you are attempting to travel to the same destination 10 separate times—much like a commuter travels to work every day—and will receive a payout for each separate trip. You will have the option to speed or not when traveling. In each of these rounds you and the other members of the experiment will now have two decisions to make. First, you will vote on an enforcement regime and the regime receiving the majority vote will be posted so that everyone is aware of the enforcement regime. After observing the winning regime, you will then be asked to decide whether you will speed or not. As in the first 10 rounds, an individual that speeds and arrives at the destination without being apprehended will receive a payout of $1.00 per round. An individual that speeds and is apprehended for speeding receives a payout of $1.00-f per round, where f is the fine that is imposed by the enforcement regime that is instituted. Note that 0 < f < 1 always. Lastly, an individual that obeys the law and does not speed receives a payout of $0.60 per round.

Rounds 21–30: As in rounds 11–20, in each of these rounds you and the other members of the experiment will have two decisions to make. First, you will vote on an enforcement regime and the regime receiving the majority vote will be posted so that everyone is aware of the enforcement regime. However, the regimes are different.

  1. Regime A:

    Chance of getting caught = 3/5, Fine = $0.833

  2. Regime B:

    Chance of getting caught = 4/5, Fine = $0.625

Whichever regime (Regime A or Regime B) that receives the most votes will be implemented and posted. After observing the winning regime, you will then be asked to decide whether you will speed or not. As in the first 10 rounds, an individual that speeds and arrives at the destination without being apprehended will receive a payout of $1.00 per round. An individual that speeds and is apprehended for speeding receives a payout of $1.00-f per round, where f is the fine that is imposed by the enforcement regime that is instituted. Note that 0 < f < 1 always. Lastly, an individual that obeys the law and does not speed receives a payout of $0.60 per round.

1.3 Examples

Situation 1: Regime A is implemented because it receives the most votes. There are three possible payouts:

  1. Payout 1:

    Participant does not speed and receives a payout of $0.60.

  2. Payout 2:

    Participant speeds, is not caught and receives a payout of $1.00.

  3. Payout 3:

    Participant speeds, is caught and receives a payout of $0.167 (= $1.00−$0.833).

Situation 2: Regime B is implemented because it receives the most votes. There are three possible payouts:

  1. Payout 1:

    Participant does not speed and receives a payout of $0.60.

  2. Payout 2:

    Participant speeds, is not caught and receives a payout of $1.00.

  3. Payout 3:

    Participant speeds, is caught and receives a payout of $0.375 (= $1.00−$0.625).

Payouts: The payout that you will receive for rounds 21–30 will be determined by randomly choosing one of the 10 rounds and then multiplying the payout of that round by 5. Consider the following possible payouts:

  1. Example Payout 1:

    Suppose that round 25 is randomly chosen as the payout round and that you decide to not speed in round 25 and so your payout for round 5 is $0.60. Therefore, your payout for rounds 21–30 is $3.00 (=5*$0.60).

  2. Example Payout 2:

    Suppose that round 23 is randomly chosen as the payout round and that you decide to speed in round 3 and were not caught and so your payout for round 23 is $1.00. Therefore, your payout for rounds 21–30 is $5.00 (=5*$1.00).

  3. Example Payout 3:

    Suppose that round 27 is randomly chosen as the payout round. In round 27 you decided to speed, Regime A is instituted and you are caught speeding. Your payout for round 7 is $0.167. Therefore, your payout for rounds 21–30 is $0.835 (=5*$0.167).

  4. Example Payout 4:

    Suppose that round 29 is randomly chosen as the payout round. In round 29 you decided to speed, Regime B is instituted and you are caught speeding. Your payout for round 9 is $0.375. Therefore, your payout for rounds 21–30 is $1.875 (=5*$0.375).

Total Payouts: Your total payout is determined by summing your payouts for all thirty rounds, dividing this total in half, and then adding a $5.00 show up payment. The payout is mathematically explained below:

$$ Total{ }Payout = (Payou{t_{{1 - 10}}} + Payou{t_{{11 - 20}}} + Payou{t_{{21 - 30}}})*{5} + \$ {5} $$

Note that the minimum total payout that you could receive is $6.84, which could be obtained by getting caught speeding in Regime A in all rounds. The maximum payout is $20.00 = ($5.00 + $15.00), which could be obtained by speeding and not getting caught in all rounds.

1.4 Experiment instructions—setting 2

General Rules: No talking. No use of cell phones. No looking at neighbor’s screens.

You are about to participate in an experiment on decision-making carried out by a researcher from the University of California at Santa Barbara. During this session, you can earn money. The amount of your earnings depends on your decisions and on the decisions of the other participants in this session. The session consists of 30 rounds. Your earnings will be determined by randomly choosing one payout from rounds 1–10, one payout from rounds 11–20, and one payout from rounds 21–30. The payout mechanism is explained in detail below. Your earnings will be paid to you in cash in private.

Your decisions are anonymous and confidential.

The experiment is split into 3 separate sections. Each section consists of 10 rounds. Instructions for each set of 10 rounds are provided below.

Rounds 1–10: You are attempting to travel to the same destination 10 separate times—much like a commuter travels to work every day—and will receive a payout for each separate trip. You will have the option to speed or not when traveling. There will be two potential enforcement regimes—which you will be told before deciding whether to speed or not—and the probability that either regime will be instituted is equally likely. An enforcement regime is both a probability of getting caught when speeding and a corresponding fine, denoted f, which will be imposed if you are caught speeding. An individual that speeds and arrives at the destination without being apprehended will receive a payout of $1.00 per round. An individual that speeds and is apprehended for speeding receives a payout of $1.00 - f per round, where f is the fine that is imposed by the enforcement regime that is instituted. Note that the fine is always positive and less than $1.00. Lastly, an individual that obeys the law and does not speed receives a payout of $0.60 per round.

1.5 Examples

There are two potential enforcement regimes:

  1. Regime A:

    Chance of getting caught if speeding = 1/3, Fine = $0.90

  2. Regime B:

    Chance of getting caught if speeding = 2/3, Fine = $0.45

    Regime A and Regime B are equally likely—i.e. there is a 50% chance that either regime is randomly chosen.

Situation 1: Regime A is randomly selected. There are three possible payouts:

  1. Payout 1:

    Participant does not speed and receives a payout of $0.60.

  2. Payout 2:

    Participant speeds, is not caught and receives a payout of $1.00.

  3. Payout 3:

    Participant speeds, is caught and receives a payout of $0.10 (= $1.00−$0.90).

Situation 2: Regime B is randomly selected. There are three possible payouts:

  1. Payout 1:

    Participant does not speed and receives a payout of $0.60.

  2. Payout 2:

    Participant speeds, is not caught and receives a payout of $1.00.

  3. Payout 3:

    Participant speeds, is caught and receives a payout of $0.55 (= $1.00−$0.45).

Payouts: The payout that you will receive for rounds 1–10 will be determined by randomly choosing one of the 10 rounds and then multiplying the payout of that round by 10. Consider the following possible payouts:

  1. Example Payout 1:

    Suppose that round 5 is randomly chosen as the payout round and that you had decided to not speed in round 5 and so your payout for round 5 is $0.60. Therefore, your payout for rounds 1–10 is $6.00 (=10 × $0.60).

  2. Example Payout 2:

    Suppose that round 3 is randomly chosen as the payout round and that you had decided to speed in round 3 and were not caught and so your payout for round 3 is $1.00. Therefore, your payout for rounds 1–10 is $10.00 (=10 × $1.00).

  3. Example Payout 3:

    Suppose that round 7 is randomly chosen as the payout round. In round 7 you had decided to speed, Regime A is instituted and you are caught speeding. Your payout for round 7 is $0.10. Therefore, your payout for rounds 1–10 is $1.00 (=10 × $0.10).

  4. Example Payout 4:

    Suppose that round 9 is randomly chosen as the payout round. In round 9 you had decided to speed, Regime B is instituted and you are caught speeding. Your payout for round 9 is $0.55. Therefore, your payout for rounds 1–10 is $5.50 (=10 × $0.55).

Total Payouts: Your total payout is determined by summing your payouts for all thirty rounds, dividing this total in half, and then adding a $5.00 show up payment. The payout is mathematically explained below:

$$ Total{ }Payout = (Payou{t_{{1 - 10}}} + Payou{t_{{11 - 20}}} + Payou{t_{{21 - 30}}})*{5} + \$ {5} $$

Note that the minimum total payout that you could receive is $6.50 = ($5.00 + $1.50), which could be obtained by getting caught speeding in Regime A in all rounds. The maximum payout is $20.00 = ($5.00 + $15.00), which could be obtained by speeding and not getting caught in all rounds.

Rounds 11–20: As in the first ten rounds, you are attempting to travel to the same destination 10 separate times—much like a commuter travels to work every day—and will receive a payout for each separate trip. You will have the option to speed or not when traveling. The main difference between the rounds 1–10 and rounds 11–20 is that you will be told which regime is in place before deciding whether to speed or not. In other words, you will know with certainty the enforcement regime that is in place. Recall that an enforcement regime is both a probability of getting caught when speeding and a corresponding fine, denoted f, which will be imposed if you are caught speeding. An individual that speeds and arrives at the destination without being apprehended will receive a payout of $1.00 per round. An individual that speeds and is apprehended for speeding receives a payout of $1.00—f per round, where f is the fine that is imposed by the enforcement regime that is instituted. Note that 0 < f < 1 always. Lastly, an individual that obeys the law and does not speed receives a payout of $0.60 per round.

Rounds 21–30: In each of the last 20 rounds you and the other members of the experiment will now have two decisions to make. First, you will vote on an enforcement regime and the regime receiving the majority vote will be posted so that everyone is aware of the enforcement regime. After observing the winning regime you will then be asked to decide whether you will speed or not. As in the first 10 rounds, an individual that speeds and arrives at the destination without being apprehended will receive a payout of $1.00 per round. An individual that speeds and is apprehended for speeding receives a payout of $1.00—f per round, where f is the fine that is imposed by the enforcement regime that is instituted. Note that 0 < f < 1 always. Lastly, an individual that obeys the law and does not speed receives a payout of $0.60 per round.

Situation 1: Suppose 7 out of 11 people vote in favor of Regime A and so Regime A is instituted. An individual who chooses to speed has a one in three chance of being caught speeding and paying a fine of $0.90. There are three possible payouts that a participant could receive:

  1. Payout 1:

    Participant does not speed and receives a payout of $0.60.

  2. Payout 2:

    Participant speeds, is not caught and receives a payout of $1.00.

  3. Payout 3:

    Participant speeds, is caught and receives a payout of $0.10 (= $1.00−$0.90).

Situation 2: Suppose 6 out of 11 people vote in favor of Regime B and so Regime B is instituted. An individual who chooses to speed has a two in three chance of being caught speeding and paying a fine of $0.45. There are three possible payouts that a participant could receive:

  1. Payout 1:

    Participant does not speed and receives a payout of $0.60.

  2. Payout 2:

    Participant speeds, is not caught and receives a payout of $1.00.

  3. Payout 3:

    Participant speeds, is caught and receives a payout of $0.55 (= $1.00−$0.45).

Payouts: The payout that you will receive for rounds 21–30 will be determined by randomly choosing one of the 10 rounds and then multiplying the payout of that round by 5. Consider the following possible payouts:

  1. Example Payout 1:

    Suppose that round 25 is randomly chosen as the payout round and that you had decided to not speed in round 25 and so your payout for round 5 is $0.60. Therefore, your payout for rounds 1–10 is $6.00 (=10 × $0.60).

  2. Example Payout 2:

    Suppose that round 23 is randomly chosen as the payout round and that you had decided to speed in round 23 and were not caught and so your payout for round 3 is $1.00. Therefore, your payout for rounds 1–10 is $10.00 (=10 × $1.00).

  3. Example Payout 3:

    Suppose that round 27 is randomly chosen as the payout round. In round 27 you had decided to speed, Regime A received the majority vote and you are caught speeding. Your payout for round 7 is $0.10. Therefore, your payout for rounds 1–10 is $1.00 (=10 × $0.10).

  4. Example Payout 4:

    Suppose that round 29 is randomly chosen as the payout round. In round 29 you decided to speed, Regime B received the majority vote and you are caught speeding. Your payout for round 9 is $0.55. Therefore, your payout for rounds 1–10 is $5.50 (=10 × $0.55).

Total Payouts: Your total payout is determined by summing your payouts for all thirty rounds, dividing this total in half, and then adding a $5.00 show up payment. The payout is mathematically explained below:

$$ Total{ }Payout = (Payou{t_{{1 - 10}}} + Payou{t_{{11 - 20}}} + Payou{t_{{21 - 30}}})*{5} + \$ {5} $$

Note that the minimum total payout that you could receive is $6.50 = ($5.00 + $1.50), which could be obtained by getting caught speeding in Regime A in all rounds. The maximum payout is $20.00 = ($5.00 + $15.00), which could be obtained by speeding and not getting caught in all rounds.

Appendix B

To examine potential determinants of consistency in the choices that individuals make when determining whether or not to speed, we run probit regressions for the decision to speed on whether or not the individual was caught in previous periods, was not caught in previous periods, did not speed in previous periods, and which regime that the individual faced. We report elasticities in the table below.

Table 5 Probit regressions for determinants of speeding
Full size table

Rights and permissions

Reprints and permissions

About this article

Cite this article

DeAngelo, G., Charness, G. Deterrence, expected cost, uncertainty and voting: Experimental evidence. J Risk Uncertain 44, 73–100 (2012). https://doi.org/10.1007/s11166-011-9131-3

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11166-011-9131-3

Keywords

JEL classifications

Search

Navigation