Skip to main content
SpringerLink
Log in

No fight, no loss: underinvestment in experimental contest games

  • Original Paper
  • Published:
Economics of Governance Aims and scope Submit manuscript

Abstract

This paper reports a series of laboratory experiments intended to identify conditions that attenuate the overdissipation of rents typical of experimental contest games. We examine the influences on contestant behavior of the observability and timing of preceding bids, allocation rules for the situation when no bids occur (random prize allocation vs. prize loss) and matching protocol for repeated contests involving pairs of bidders. Our results show that the simultaneous presence of three factors (simultaneous bids, random prize allocation if no bids occur and fixed matching) allows contestants to coordinate to realize efficient outcomes (underbidding). However, the absence of one of these factors causes overbidding to return. From the perspective of theoretical prediction, the decision to allocate the prize even when no bids occur (no fight, no loss) should be irrelevant. However, this allocation decision may strongly influence behaviour (by encouraging submission of efficient and minimal bids) if combined with features that encourage collusion (fixed matching and symmetry).

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

Similar content being viewed by others

Notes

  1. We must be careful here when speaking of collusive/efficient outcomes. In price competition, collusive pricing (to achieve prices that maximize payoffs to bidders) is ultimately inefficient for society because it occurs at the expense of the lower consumer. However, collusion and efficiency move together in the contest literature, because they imply lower levels of costly bids and thus smaller sunk costs.

  2. If an experimental subject finds bidding zero unnatural, then the bias towards positive bids might cause overdissipation.

  3. In those treatments with stranger matching, the 20 subjects were split into two sections of 10. Subjects were then randomly matched within each section in each round (see Sect. 2.3 for details).

  4. Oechssler (2013) found that under social preferences, when a stage game with a unique equilibrium is finitely repeated, the equilibrium of the (finitely) repeated game may not merely repeat that of the stage game. However, this result is not helpful in our case because it requires a social preference for introducing non-linearity in players’ utility functions. Two standard approaches exist to social preferences in contests. The first, following Sheremeta (2013), is concern for relative payoffs (Mago et al. 2014), which relies on a linear utility function that is frequently used in evolutionary contest theory (Leininger 2003; Hehenkamp et al. 2004; Riechmann 2007). The second approach is that of Fehr and Schmidt (1999), whereby preferences become linear if agents seek to maximize expected utility (see Herrmann and Orzen 2008—Appendix B).

  5. Note that, by running basic games SIM, SEQ_Y and SEQ_N in the same session for each scenario we avoid the possibility of session effects in comparisons among scenarios with regards to both the prize rule in the case of no bids (RP/NP) and also the matching protocol (partners/strangers).

  6. A copy of the instructions, translated into English, can be found in the “Appendix”.

  7. Regarding the notation used in Sect. 2, we chose.

  8. The differences are significant at the 10 % level for the SEQ_Y_RP treatment, the 1 % level for the SEQ_N_RP treatment, and the 5 % level for the SIM_NP, SEQ_Y_NP and SEQ_N_NP treatments. We do not report the differences with the stranger matching treatments because they have to be performed at the section level, reducing the number of independent observations to two. To better assess the treatment effects, the reader is referred to the next section, which presents and discusses results from panel data regressions.

  9. End-game effects are well known in experimental economics.

  10. We rely on a panel data analysis due to the dynamic nature of our data (each treatment involves a panel of 20 subjects who interact for 30 rounds). We correct the standard errors by clustering at the level of independent observations: pairs for partner matching and sections for stranger matching.

  11. For the stranger protocol, the coefficient of the variable SEQ_Y is negative and marginally significant at the 10 % level.

  12. Note that, in treatment SIM, either player in each pair can be the first (or second) mover, since they simultaneously choose their expenditure levels. Hence, in this case we include all the observations in the regression. This explains why the SIM treatments have twice the number of observations as the other treatments.

  13. To our knowledge, the few experimental papers that deal with sequential contest games do not include econometric analysis of the determinants of individual behaviour.

  14. In some cases, such as for larger groups, this variable is not significant (Savikhin and Sheremeta 2013). The reason might be the large cognitive load needed to form beliefs about the behaviour of others when the group is large rather than small.

  15. We abstract from the time trend parameter, which in most cases is insignificant and has a very small estimated value.

  16. In some contexts one immediate policy recommendation to avoid inefficient investments (prevent bids) could be to directly assign the prize to one contestant. However, in many situations (e.g., a local government/authority offering a public contract), no party can be excluded from participation in the contest. Therefore, no party can be excluded from investing effort, including non-productive effort, to win the contest. Moreover, once contestants have invested, they are expected to be treated fairly by the authority (i.e., their chances of success should depend on their effort). In such a case, the authority can still decide what to do if no one invests (in this case, either by not giving the prize/contract or by randomly allocating it, hence all parties are treated equally). Our results suggest that this decision matters, and, to avoid inefficiencies, it may be rational to allocate the prize even when contest fails to produce a winner.

  17. These instructions were read aloud by the researcher at the beginning of the session, and participants were also provided with a printed copy that was available to them throughout the whole experiment.

References

  • Abbink K (2012) Laboratory experiments on conflict. In: Garfinkel MR, Skaperdas S (eds) The Oxford handbook of the economics of peace and conflict. Oxford University Press, Oxford

    Google Scholar 

  • Abbink K, Brandts J, Herrmann B, Orzen H (2010) Inter-group conflict and intra-group punishment in an experimental contest game. Am Econ Rev 100:420–447

    Article  Google Scholar 

  • Arce D, Bakshi S, Croson R, Eckel C, Fatas E, Kass M (2011) Counterterrorism strategies in the lab. Public Choice 149(3–4):465–478

    Article  Google Scholar 

  • Bertrand J (1883) Review of theorie mathematique de la richesse sociale and recherches sur les principes mathematicque de la theoire des richesse. Journal des Savants 67:499–508

  • Cason TN, Sheremeta RM, Zhang J (2012) Communication and efficiency in competitive coordination games. Games Econ Behav 76:26–43

    Article  Google Scholar 

  • Cooper R, DeJong DV, Forsythe R, Ross T (1990) Selection criteria in coordination games: some experimental results. Am Econ Rev 80(1):218–233

    Google Scholar 

  • Dechenaux E, Kovenock D, Sheremeta RM (2012) A survey of experimental research on contests, all-pay auctions, and tournaments, WZB Discussion paper, No. SP II 2012–109 Mimeo

  • Dufwenberg M, Gneezy U (2000) Price competition and market concentration: an experimental study. Int J Indus Organ 18:7–22

    Article  Google Scholar 

  • Faravelli M, Stanca L (2012) When less is more: rationing and rent dissipation in stochastic contests. Games Econ Behav 74:170–183

    Article  Google Scholar 

  • Fatas E, Haruvy E, Morales AJ (2014) A psychological reexamination of the Bertrand paradox. South Econ J 80:948–967

    Article  Google Scholar 

  • Fischbacher U (2007) Z-Tree: Zurich toolbox for ready-made economic experiments. Exp Econ 10:171–178

    Article  Google Scholar 

  • Fehr E, Schmidt KM (1999) A theory of fairness, competition, and cooperation. Q J Econ 114(3):817–868

    Article  Google Scholar 

  • Fonseca MA (2009) An experimental investigation of asymmetric contests. Int J Indus Organ 27:582–591

    Article  Google Scholar 

  • Hehenkamp B, Leininger W, Possajenikov A (2004) Evolutionary equilibrium in Tullock contests: spite and overdissipation. Eur J Pol Econ 20:1045–1057

    Article  Google Scholar 

  • Herrmann B, Orzen H (2008) The appearance of homo rivalis: social preferences and the nature of rent-seeking. CeDEx Discussion Paper 2008–2010, University of Nottingham

  • Kimbrough EO, Sheremeta RM, Shields T (2011) Resolving conflicts by a random device. Chapman University, ESI Working Paper

  • Konrad KA (2009) Strategy and dynamics in contests. Oxford University Press, New York

    Google Scholar 

  • Leininger W (2003) On evolutionarily stable behavior in contests. Econ Gov 4:177–186

    Article  Google Scholar 

  • Lim W, Matros A, Turocy TL (2014) Bounded rationality and group size in Tullock contests: experimental evidence. J Econ Behav Organ 99:155–167

    Article  Google Scholar 

  • Lugovskyy V, Tucker S, Puzuello D (2010) An experimental investigation of overdissipation in the all-pay auction. Eur Econ Rev 54:974–997

    Article  Google Scholar 

  • Mago SD, Savikhin AC, Sheremeta RM (2014) Facing your opponents: social identification and information feedback in contests. J Confl Resolut. doi:10.1177/0022002714545333

  • Oechssler J (2013) Finitely repeated games with social preferences. Exp Econ 16:222–231

    Article  Google Scholar 

  • Potters J, de Vries C, van Winden F (1998) An experimental examination of rational rent-seeking. Eur J Pol Econ 14:783–800

    Article  Google Scholar 

  • Riechmann T (2007) An analysis of rent-seeking games with relative-payoff maximizers. Public Choice 133:147–155

    Article  Google Scholar 

  • Savikhin AC, Sheremeta RM (2013) Simultaneous decision-making in competitive and cooperative environments. Econ Inq 51:1311–1323

    Article  Google Scholar 

  • Schmitt P, Shupp R, Swope K (2004) Multi-period rent-seeking contests with carryover: theory and experimental evidence. Econ Gov 5:187–211

    Article  Google Scholar 

  • Sheremeta RM (2010) Experimental comparison of multi-stage and one-stage contests. Games Econ Behav 68:731–747

    Article  Google Scholar 

  • Sheremeta RM (2011) Contest design: an experimental investigation. Econ Inq 49:573–590

    Article  Google Scholar 

  • Sheremeta RM (2013) Overbidding and heterogeneous behavior in contest experiments. J Econ Surv 27(3):491–514

    Article  Google Scholar 

  • Shupp R, Sheremeta RM, Schmidt D, Walker J (2013) Resource allocation contests: experimental evidence. J Econ Psychol 39:257–267

    Article  Google Scholar 

  • Tullock G (1980) Efficient rent-seeking. In: Buchanan J, Tollison R, Tullock G (eds) Towards a theory of the rent-seeking society. Texas A&M University Press, College Station, pp 97–112

    Google Scholar 

  • Weber RA, Camerer CF, Knez M (2004) Timing and virtual observability in ultimatum bargaining and “weak link” coordination games. Exp Econ 7:25–48

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. University of Essex, Colchester, UK

    Sara Godoy

  2. Facultad de Ciencias Económicas y Empresariales, Universidad de Málaga, Plaza El Ejido s/n, 29013, Málaga, Spain

    Miguel A. Meléndez-Jiménez & Antonio J. Morales

Authors
  1. Sara Godoy
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Miguel A. Meléndez-Jiménez
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Antonio J. Morales
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Antonio J. Morales.

Additional information

The authors would like to thank editor Amihai Glazer and two anonymous referees for comments and suggestions that helped us to substantially improve the paper. The authors gratefully acknowledge financial support from the Spanish Ministry of Economy and Competitiveness through Project ECO2011-26996, and from the Regional Government of Andalusia through Projects SEJ2009-4794 and SEJ2011-8065.

Appendix: Experimental instructions

Appendix: Experimental instructions

1.1 General instructionsFootnote 17—common to treatments SIM, SEQ_Y and SEQ_N—of RP scenarios under partner matching

This experiment studies how individuals make decisions in certain contexts. Instructions are simple and, if you follow them carefully, you will confidentially receive a sum of cash at the end of the experiment. All payments to participants will remain confidential. You can ask questions at any time by raising your hand. Other than such questions, any kind of communication between participants is forbidden, and subject to immediate exclusion from the experiment.

  1. 1.

    The experiment consists of 30 independent rounds (your outcome in one round does not affect that in other rounds). During the experiment you will receive information via your computer screen. This information is relevant and you should pay attention to it.

  2. 2.

    In each and every round you form part of the same group of two participants. The composition of your group is randomly determined at the beginning of the experiment and remains constant throughout. At no time will you know the identity of the other member of your group. Each group is independent and no group affects the others.

  3. 3.

    Each group comprises two participants: a green participant and an orange participant. Your colour will be chosen automatically by the computer at the beginning of the experiment and will remain constant throughout. The initial screen will inform you of your colour.

  4. 4.

    At the beginning of each round, you will receive an endowment of 100 ECU. Your only decision involves how many balls of your colour (orange or green) you want to buy at a price of 1 ECU per ball.

  5. 5.

    In each round there will be a raffle for a prize of 100 ECU between each pair. The computer will place all the balls bought by the two participants inside an urn. The computer will then randomly pick a ball, and the winner will be determined by the colour of the drawn ball: If the ball is green, the green participant wins 100 ECU and vice versa if the ball is orange.

  6. 6.

    If neither participant in a pair buys any balls in a round (i.e., there are no balls in the urn), then the colour selection is decided by a virtual coin toss, where each colour has a 50 % probability of being selected.

  7. 7.

    Your benefit in a round will be given by subtracting the cost of the balls you bought from your initial endowment from any prize won.

  8. 8.

    At the end of each round, you will be informed of: your decision (how many balls you bought), the decision of the other participant (how many balls he/she bought), the colour of the ball picked by the computer and your payoff for that round. A table will also be made available with information on past rounds.

  9. 9.

    At the end of the experiment, you will confidentially receive a sum equaling your accumulated payoffs over the 30 rounds (applying a conversion rate of 180 ECU = 1€).

1.2 Treatment specific instructions (displayed on the initial computer screen of the participants)

  • TREATMENT SIM:

    Your colour is _____. In each round, each participant in your group will simultaneously decide how many balls of his/her colour to buy for the 100 ECU raffle prize.

  • TREATMENT SEQ_Y:

    Your colour is _____. In each round, the green participant in your group will first choose the number of green balls to buy for the 100 ECU raffle prize. Once the green participant has made his decision, the orange participant will be informed of how many green balls the green participant has bought. The orange participant will then decide how many orange balls to buy for the raffle.

  • TREATMENT SEQ_N:

    Your colour is _____. In each round, the green participant from your group will first choose the number of green balls to buy for the 100 ECU raffle prize. Once the green participant has made her decision, the orange participant will decide how many orange balls to buy for the raffle, but will not be informed of the decision made by the green participant.

1.3 NP scenarios

In the treatments corresponding to the NP scenarios, point 6 of the general instructions is replaced by:

  1. 6.

    If neither participant in a pair buys any balls in a round (there are no balls in the urn), then there is no raffle and no one wins the prize. Note that this implies you will not win the prize unless you buy balls.

1.4 Stranger matching protocol

In the treatments corresponding to the stranger matching protocol, points 2 and 3 of the instructions are replaced by:

  • In each and every round you form part of the same group of 10 participants. The composition of your group is randomly determined at the beginning of the experiment and remains constant throughout. You will not know the identity of the other members of your group at any time. The groups are independent and no group affects the others.

  • Each group of 10 participants is subdivided into two further groups: five green participants and five orange participants. Your colour will be chosen automatically by the computer at the beginning of the experiment and will remain unchanged throughout the experiment. You will be informed of your colour on the initial screen.

  • At the beginning of each round, participants will be randomly paired with a different colour participant from their group. Each pair will thus comprise a green participant and an orange participant.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Godoy, S., Meléndez-Jiménez, M.A. & Morales, A.J. No fight, no loss: underinvestment in experimental contest games. Econ Gov 16, 53–72 (2015). https://doi.org/10.1007/s10101-014-0150-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10101-014-0150-9

Keywords

JEL Classification

Search

Navigation