Abstract
We study a group formation game. Players with different strengths form groups before expending effort to win a prize. The prize has the nature of the reward for outdoing in competition such as holding a dominant position among players or being recognized as a dominant status. So, it has the nature of public goods within a winning group (group-specific public goods). In open membership game, we find that a single player stays alone and the others form a group together in equilibrium. The stand-alone player can be anyone except for the first and second strongest players in the contest. However, strong (Nash) equilibrium predicts that the weakest player is isolated. Similarly, we find that in exclusive membership game, every structure can emerge in equilibrium but the weakest player is isolated in the strong equilibrium.
Similar content being viewed by others
Notes
Beginning with the seminal works of Tullock (1967, 1980), Krueger (1974), and Becker (1983), research on the contest theory has been increasing rapidly, and the literature on contests is now enormous. The following are some studies on the contest literature: Nitzan (1994), Szymanski (2003), Garfinkel and Stergios (2007), Corchón (2007), Konrad (2009), Connelly et al. (2014), Chowdhury and Gürtler (2015), Congleton and Hillman (2015), Dechenaux et al. (2015), Mealem and Nitzan (2016), Vojnović (2016), Sheremeta (2018), Corchón and Marco (2018), and Fu and Wu (2019).
A group’s performance is defined as the sum of the group members’ performances and its contest success function (CSF) is imperfectly discriminating in these studies while the followings adopt different ways of aggregating group members’ performances or/and deterministic CSF: Lee (2012), Kolmar and Rommeswinkel (2013), Chowdhury et al. (2013), Barbieri et al. (2014), Chowdhury and Topolyan (2016a, 2016b), Chowdhury et al. (2016), Lee (2017), and Lee and Song (2019).
For example, suppose that \(I=\left\{ 1, 2, 3, 4\right\}\), \(g_1 =1\), \(g_2 =2\), \(g_3 =2\), and \(g_4 = 3\). Then, in the open membership game, three groups are formed and a group structure \(G=\left\{ G_1 , G_2, G_3\right\}\) is derived where \(G_1 =\left\{ 1\right\}\), \(G_2 =\left\{ 2, 3\right\}\), and \(G_3 =\left\{ 4\right\}\).
By Assumption 2, when \(K=1\), every player gets zero payoff and thus does not exert any effort. In the next section, we show that the grand coalition cannot be an equilibrium outcome.
Following Epstein and Mealem (2009), our model can be extended to the case of low complementarity among members’ performances and a convex cost function. We leave it for our future works and appreciate the anonymous referee on this precious comment.
In Remark 1, we formally show that this case cannot be an equilibrium outcome. See the proof.
We can also have (10) by adding share functions for the K groups in Cornes and Hartley (2005), where the share function for \(G_k\) is \(s_k (X) \equiv 1-\frac{X}{B_{1k}}\) and \(\sum _{k=1}^K s_k (X)=1\) with \(X=\sum _{l=1}^K X_l\). Then, from equilibrium value of X (\(X^*\)), \(X_k^*\) is given as \(s_k (X^*) X^*\), \(e_{1k}^*\) as \(\frac{X_k^* }{a_{1k}}\), \(p_k^*\) as \(s_k (X^*)\), and \(\pi _{1k}^*\) as \(v_{1k} s_k (x^*)^2\). The conditions in (7) are equivalent to the ones for \(s_k (X^*)>0\) for all \(k=1, \ldots , K\). We thank the anonymous referee for suggesting the share function approach to our analysis.
Note that \(B_{11} = \max \left\{ B_{11}, B_{12}, \ldots , B_{1K}\right\} =a_1 v_1\).
See the proof of Lemma 5 in Appendix A.
We can show this as well by the share function approach of Cornes and Hartley (2005). Since the share function for \(G_1\) (\(s_1 (\cdot )\)) monotonically increases as the number of groups participating in the contest decreases, player 1k in \(G_k\) can be better off by giving up his own group and joining to \(G_1\) as long as his movement does not result in the grand coalition. Thus, \(K=2\) should be at equilibrium. We formally present this in Appendix B, where our main results (Proposition 1, 2, and 5) are derived from the share function approach. We deeply thank the referee for sharing the outstanding proofs for Lemmas A, B, and C in the appendix.
The detailed proof is presented in Appendix A.
We appreciate the associate editor’s and the referee’s insightful comments on Proposition 1 in connection with the herding behavior of our players. Especially, the suggestion for extending our model to a sequential move game will be an important topic for our future work.
The strong equilibrium is the Coalition-Proof Nash one (Bernheim et al. 1987).
For example, suppose that \(I=\left\{ 1, 2, 3, 4\right\}\), \(g^1 =\left\{ 1\right\}\), \(g^2 =\left\{ 1, 2, 3\right\}\), \(g^3 =\left\{ 1, 2, 3\right\}\), and \(g^4 =\left\{ 1, 2, 3, 4\right\}\). In game \(\Gamma\), the strict unanimity is required to form a group and thus the group structure \(G^{(\gamma )}=\left\{ G_1 , G_2 , G_3, G_4\right\}\) is derived where \(G_1 =\left\{ 1\right\}\), \(G_2 =\left\{ 2\right\}\), \(G_3 =\left\{ 3\right\}\), and \(G_4 =\left\{ 4\right\}\). On the contrary, in game \(\Delta\), the group structure \(G^{(\delta )}=\left\{ G_1 , G_2, G_3\right\}\) is derived where \(G_1 =\left\{ 1\right\}\), \(G_2 =\left\{ 2, 3\right\}\), and \(G_3 =\left\{ 4\right\}\).
For instance, suppose that \(G=\left\{ G_1 , G_2, G_3 \right\}\), where \(G_1 =\left\{ 1, 2, 7, 8, \ldots , n \right\}\), \(G_2 =\left\{ 3, 4\right\}\), and \(G_3 =\left\{ 5, 6\right\}\). This implies that, in the first stage, all the players \(i \in I \backslash \left\{ 3, 4, 5, 6\right\}\) choose \(g^i =\left\{ 1, 2, 7, 8, \ldots , n\right\}\), player 3 and 4 choose \(g^3 =g^4 =\left\{ 3, 4\right\}\), and player 5 and 6 choose \(g^5 =g^6 =\left\{ 5, 6\right\}\). From Lemma 2, player 6 will be better off if he can join \(G_1\). What happens if he changes his decision from \(g^6 =\left\{ 5, 6\right\}\) to \(g_d^6 =\left\{ 1, 2, 6, 7, 8, \ldots , n\right\}\) for the purpose of joining \(G_1\)? Player 6’s deviation results in the new group structure \(G^{'}=\left\{ G_1 , G_2 , G_3, G_4 \right\}\), where \(G_1 =\left\{ 1, 2, 7, 8, \ldots , n \right\}\), \(G_2 =\left\{ 3, 4\right\}\), \(G_3 =\left\{ 5\right\}\), and \(G_4 =\left\{ 6\right\}\). That is, player 6’s deviation just brings the break of his former group, and it makes him worse off than under G.
For instance, suppose that, in the first stage of the game, player 1 chooses \(g^1 =\left\{ 1, 2\right\}\) and all the players \(i\in I \backslash \left\{ 1\right\}\) choose \(g^i =\left\{ 2, 3, \ldots , n\right\}\). Then, a group structure \(G^{(\gamma )} =\left\{ G_1 , G_2 \right\}\) is determined, where \(G_1 =\left\{ 1\right\}\) and \(G_2 =\left\{ 2, 3, \ldots , n \right\}\). Given the group structure G, if player 2 changes his decision \(g^2\) to \(g_d^2 =\left\{ 1, 2\right\}\), then a new group structure \(G^{(\gamma )^{'}}=\left\{ G_1 , G_2 , \cdots , G_{n-1}\right\}\) is chosen, where \(G_1 =\left\{ 1, 2\right\}\) and \(G_k =\left\{ k+1 \right\}\) for \(k=2, 3, \ldots , n-1\).
We are grateful for the anonymous referee’s justification for assuming \(g^{i}=\left\{ i\right\}\) whenever \(|g_k |=1\) in another way.
We thank the anonymous referee for clarifying this point.
The detailed analysis can be provided upon request.
We again thank the associate editor for suggesting this sequential game and providing intuitive guess for its solutions.
References
Aumann RJ (1959) Acceptable points in general cooperative $n$-person games. In: Tucker A, Luce R (eds) Contributions to the theory of games, vol 4. Princeton University Press, Princeton, pp 287–324
Baik KH (1993) Effort levels in contests: the public-good prize case. Econ Lett 41(4):363–67
Baik KH (1994) Effort levels in contests with two asymmetric players. S Econ J 61(2):367–378
Baik KH (2008) Contests with group-specific public-good prizes. Soc Choice Welf 30(1):103–117
Baik KH (2016) Endogenous group formation in contests: unobservable sharing rules. J Econ Manag Strategy 25(2):400–419
Baik KH, Lee S (2001) Strategic groups and rent dissipation. Econ Inq 39(4):672–684
Baik KH, Shogren JF (1995) Competitive-share group formation in rent-seeking contests. Public Choice 83(1):113–126
Barbieri S, Malueg DA, Topolyan I (2014) Best-shot group contests with complete information. Econ Theor 57:603–640
Becker GS (1983) A theory of competition among pressure groups for political influences. Q J Econ 98(3):371–400
Bergstrom T, Blume L, Varian H (1986) On the private provision of public goods. J Public Econ 29(1):25–49
Bernheim BD, Peleg B, Whinston MD (1987) Coalition-proof Nash equilibria I. Concepts. J Econ Theory 42:1–12
Bloch F (1996) Sequential formation of coalitions in games with externalities and fixed payoff division. Games Econ Behav 14(1):90–123
Bloch F (2012) Endogenous formation of alliances in conflicts. In: Garfinkel MR, Skaperdas S (eds) The Oxford handbook of the economics of peace and conflict. Oxford University Press, New York, pp 473–502
Bloch F, Sánchez-Pagés S, Soubeyran R (2006) When does universal peace prevail? secession and group formation in conflict. Econ Gov 7(1):3–29
Chowdhury S, Gürtler O (2015) Sabotage in contests: a survey. Public Choice 164(1):135–155
Chowdhury S, Topolyan I (2016) The attack-and-defense group contests: best-shot versus weakest-link. Econ Inq 54:548–557
Chowdhury S, Topolyan I (2016) Best-shot versus weakest-link in political lobbying: an application of group all-pay auction. Soc Choice Welf 47(4):959–971
Chowdhury S, Lee D, Sheremeta R (2013) Top guns may not fire: best-shot group contests with group-specific public good prizes. J Econ Behav Organ 92:94–103
Chowdhury S, Lee D, Topolyan I (2016) The max-min group contest: weakest-link (group) all-pay auction. S Econ J 83(1):105–125
Congleton RD, Hillman AL (2015) Companion to the political economy of rent seeking. Edward Elgar, Cheltenham
Connelly BL, Laszlo Tihanyi T, Crook R, Ashley Gangloff K (2014) Tournament theory: thirty years of contests and competitions. J Manag 40(1):16–47
Corchón LC (2007) The theory of contests: a survey. Rev Econ Design 11(2):69–100
Corchón LC, Serena M (2018) Contest theory. In: Corchón LC, Marini MA (eds) Handbook of game theory and industrial organization, vol 2. Edward Elgar, Cheltenham
Cornes R, Hartley R (2005) Asymmetric contests with general technologies. Econ Theor 26:923–946
d’Aspremont C, Jacquemin A, Gabszewicz JJ, Weymark JA (1983) On the stability of collusive price leadership. Can J Econ 16(1):17–25
Dechenaux E, Kovenock D, Sheremeta RM (2015) A survey of experimental research on contests, all-pay auctions and tournaments. Exp Econ 18(4):609–669
Epstein G, Mealem Y (2009) Group specific public goods, orchestration of interest groups with free riding. Public Choice 139(3):357–369
Esteban J, Sákovics J (2003) Olson vs. coase: coalitional worth in conflict. Theor Decis 55(4):339–357
Flamand S, Orestis T (2015) Prize-sharing rules in collective rent-seeking. In: Congleton RD, Hillman AL (eds) Companion to the political economy of rent seeking. Edward Elgar, Cheltenham, pp 92–112
Fu Q, Zenan W (2019) Contests: theory and topics. Oxford Research Encyclopedias of Economics and Finance
Garfinkel MR (2004) Stable alliance formation in distributional conflict. Eur J Polit Econ 20(4):829–852
Garfinkel MR, Skaperdas S (2007) Economics of conflict: an overview. In: Sandler T, Hartley K (eds) Handbook of defense economics, vol 2. North Holland, Amsterdam, pp 649–709
Hart S, Kurz M (1983) Endogenous formation of coalitions. Econometrica 51(4):1047–1064
Herbst L, Konrad KA, Morath F (2015) Endogenous group formation in experimental contests. Eur Econ Rev 74:163–189
Hiller T (2017) Friends and enemies: a model of signed network formation. Theor Econ 12(3):1057–1087
Katz E, Nitzan S, Rosenberg J (1990) Rent-seeking for pure public goods. Public Choice 65:49–60
Kolmar M (2013) Group conflicts. Where do we stand? Mimeo, New York
Kolmar M, Rommeswinkel H (2013) Contests with group-specific public goods and complementarities in efforts. J Econ Behav Organ 89:9–22
Konishi H, Chen-Yu P (2019) Sequential formation of alliances in survival contests. Int J Econ Theory. https://doi.org/10.1111/ijet.12246
Konishi H, Chen-Yu P (2019) Endogenous alliances in survival contests. Working paper
Konrad KA (2009) Strategy and dynamics in contests. Oxford University Press, New York
Konrad KA, Morath F (2018) To deter or to moderate? Alliance formation in contest with incomplete information. Econ Inq 56(3):1447–1463
Krueger AO (1974) The political economy of the rent-seeking society. Am Econ Rev 64(3):291–303
Lee D (2012) Weakest-link contests with group-specific public good prizes. Eur J Polit Econ 28:238–248
Lee D (2015) Group contests and technologies. Econ Bull 35:2427–2438
Lee D (2017) Average-opinion group contest. Econ Bull 37(1):79–93
Lee D, Song J (2019) Optimal team contests to induce more efforts. J Sports Econ 20(3):448–476
Mealem Y, Nitzan S (2016) Discrimination in contests: a survey. Rev Econ Des 20(2):145–172
Nitzan S (1991) Collective rent dissipation. Econ J 101(409):1522–1534
Nitzan S (1994) Modelling rent-seeking contests. Eur J Polit Econ 10(1):41–60
Noh SJ (2002) Resource distribution and stable alliances with endogenous sharing rules. Eur J Polit Econ 18(1):129–151
Ray D, Vohra R (1999) A theory of endogenous coalition structures. Games Econ Behav 26(2):286–336
Rubinstein A (1982) Perfect equilibrium in a bargaining model. Econometrica 50(1):97–109
Sánchez-Pagés S (2007) Endogenous coalition formation in contests. Rev Econ Des 11:139–163
Selten R (1973) A simple model of imperfect competition, where 4 are few and 6 are many. Int J Game Theory 2(1):141–201
Sheremeta RM (2018) Behavior in Group contests: a review of experimental research. J Econ Surv 32:683–704
Skaperdas S (1998) On the formation of alliance in conflict and contests. Public Choice 96(1):25–42
Stein WE (2002) Asymmetric rent-seeking with more than two contestants. Public Choice 113(3):325–336
Szymanski S (2003) The economic design of sporting contests. J Econ Lit 41(4):1137–1187
Tan G, Wang R (2010) Coalition formation in the presence of continuing conflict. Int J Game Theory 39(1):273–299
Tullock G (1967) The welfare cost of tariffs, monopolies and theft. West Econ J 5(3):224–232
Tullock G (1980) Efficient rent seeking. In: Buchanan JM, Tollison RD, Tullock G (eds) Toward a theory of the rent-seeking society. Texas A&M University Press, College Station, pp 97–112
Vojnović M (2016) Contest theory: incentive mechanisms and ranking methods. Cambridge University Press, Cambridge
Yi S-S (1998) Endogenous formation of joint ventures with efficiency gains. RAND J Econ 29(3):610–631
Yi S-S, Shin H (2000) Endogenous formation of research coalitions with spillovers. Int J Ind Organ 18(2):229–256
Acknowledgements
We thank Wooyoung Lim, Paan Jindapon, Jongwha Lee, Kyung Hwan Baik, Chulyoung Kim, Byung-Cheol (BC) Kim, Bob Hammond, Paul Pecorino, Mark Schneider, and seminar participants at the Sungkyunkwan University, the University of Alabama, and the conferences, and the associate editor and the anonymous referees of the journal for their helpful comments and suggestions. Dongryul Lee was supported by the Sungshin Women’s University Research Grant of 2018. Dongryul Lee is also grateful for the generous hospitality of the University of Alabama during his visit.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
We declare that we have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix A: Mathematical Proofs
Proof of Remark 1
(a) First, suppose that \(K=2\), i.e., a group structure \(G=\left\{ G_1 , G_2 \right\}\) is at equilibrium. Then, in the second stage of the game, the two groups will be active by Stein (2002). Stein (2002) showed that at least two players will be active in the contest among any \(N \ge 2\) players having different composite strengths. This implies that the two groups \(G_1\) and \(G_2\), formed in the first stage of our game, will be active in the second stage regardless of the strongest players’ composite-strength levels in each group. Now, suppose that \(K \ge 3\), i.e, a group structure \(G=\left\{ G_1 , G_2 , \cdots , G_K \right\}\) is formed in the first stage. We claim that all K groups should be active in the second stage if the group structure G is at equilibrium by showing that its transposition is true.
Given the group structure G, suppose that there exists at least one group inactive in the second stage. Then, according to Proposition 1 of Stein (2002), we know that the composite-strength level of the strongest player in the inactive group is lower than those of the strongest players in the active groups. This means that every group member in the inactive group has an incentive to deviate from his group and move to one of the active groups in the first stage, and thus the group structure G cannot be an equilibrium outcome (see Lemma 4 and 5 for details). Therefore, if the group structure \(G=\left\{ G_1 , G_2 , \ldots , G_K \right\}\) is an equilibrium outcome, all the K groups should be active in the second stage.
(b) Lemma 1 shows the effort levels chosen by the players of each group in the second stage. The conditions for \(e_{1k}^*\) for all \(k=1, \ldots , K\) to be strictly positive are given as (7). \(\square\)
Proof of Lemma 5
Given the group structure \(G=\left\{ G_1 , \ldots , G_K\right\}\) where \(K > 2\), the payoff for player 1k in \(G_k\) is \(v_{1k} \left( 1-\frac{K-1}{K} \frac{H}{B_{1k}} \right) ^2\). If player 1k departs from his own group and joins \(G_1\), he obtains payoff \(v_{1k} \left( 1-\frac{K-2}{K-1} \frac{H^{'}}{B_{11}} \right)\) where \(H^{'} = \frac{K-1}{\sum _{l=1, \ne k}^K B_{1l}^{-1}}\). Comparing these two payoffs, we show that player 1k is better off when he deviates from his own group and moves to \(G_1\) as follows:
Thus, the group structure \(G=\left\{ G_1 , \ldots , G_K\right\}\) with \(K > 2\) cannot be an equilibrium outcome. In addition to Lemma 3, this implies that \(K=2\) should be held at equilibrium, i.e., only two groups can be formed at equilibrium. \(\square\)
Proof of Proposition 1
In the body paragraph following right after Lemma 5, we have shown that \(G=\left\{ G_1 , G_2 \right\}\) where \(G_2 =\left\{ 2 \right\}\) is not an equilibrium structure, for player 1 has an incentive to deviate from \(G_1\) and join \(G_2\).
Now we consider the group structure \(G=\left\{ G_1 , G_2 \right\}\) where \(G_2 =\left\{ i \right\}\) for \(i=3, \ldots , n\). We first show that player 1 (in \(G_1\)) has no interests to deviate from his group. Given the group structure, player 1 gets his payoff \(v_1 \left( \frac{a_1 v_1}{a_1 v_1 + a_i v_i} \right) ^2\). If he departs from \(G_1\) to \(G_2\), he gets his payoff \(v_1 \left( \frac{a_1 v_1}{a_1 v_1 + a_2 v_2} \right) ^2\), which is less than the payoff obtained from staying at \(G_1\). Thus, player 1 has no incentives to deviate from \(G_1\) and join \(G_2\). What about deviating from \(G_1\) and staying alone? If he does so, the group structure changes to \(G^{'}=\left\{ G_1 , G_2 , G_3 \right\}\) where \(G_1 =\left\{ 2, 3, \ldots , i-1, i+1, \ldots , n\right\}\), \(G_2=\left\{ i\right\}\), and \(G_3 =\left\{ 1 \right\}\). Given the new group structure, player 1 obtains his payoff \(v_1 \left( \frac{(a_1 v_1 ) (a_2 v_2 ) + (a_1 v_1 ) (a_i v_i ) - (a_2 v_2 ) (a_i v_i ) }{(a_1 v_1 ) (a_2 v_2 ) + (a_1 v_1 ) (a_i v_i ) + (a_2 v_2 ) (a_i v_i )} \right) ^2\), which is less than the payoff \(v_1 \left( \frac{a_1 v_1}{a_1 v_1 + a_i v_i} \right) ^2\) obtained from staying at \(G_1\). So, player 1 doesn’t have any incentive to become a singleton or join \(G_2\) by deviating from \(G_1\).
Now consider deviation incentives of the other group members in \(G_1\). Given the group structure G, player \(j (\ne 1)\) in \(G_1\) has his payoff \(v_j \left( \frac{a_1 v_1}{a_1 v_1 + a_i v_i} \right)\). If he deviates from his group and joins \(G_2\), he gets either \(v_j \left( \frac{a_i v_i}{a_1 v_1 + a_i v_i} \right)\) in case of \(a_j v_j < a_i v_i\) or \(v_j \left( \frac{a_j v_j}{a_1 v_1 + a_j v_j} \right) ^2\) in case of \(a_j v_j > a_i v_i\). Obviously, player j’s payoff obtained by staying at \(G_1\) is greater than these payoffs obtained by joining \(G_2\), and he thus has no incentives to depart from \(G_1\) to \(G_2\). Does he then have an incentive to deviate from \(G_1\) and become a singleton? No, he doesn’t, because we have already known that belonging to \(G_1\) is better than staying alone in Lemma 5. See the proof of Lemma 5. Hence, any player \(j \in G_1\) doesn’t have the deviation incentives.
Last, does player i in \(G_2\) have an incentive to join \(G_1\)? No, because he gets nothing under the grand coalition, while he gets payoff \(v_i \left( \frac{a_i v_i}{a_1 v_1 + a_i v_i} \right) ^2 ~(>0)\) from staying alone as a singleton.
In sum, every player has no incentives to deviate from the group structure G in the first stage of the game, given the other players’ decisions unchanged, and the group structure G is an equilibrium group structure appearing at the SPE. \(\square\)
Proof of Proposition 2
Consider an equilibrium group structure \(G=\left\{ G_1 , G_2 \right\}\) where \(G_1 =I \backslash \left\{ i\right\}\) and \(G_2 =\left\{ i\right\}\) for \(i \in \left\{ 3, 4, \ldots , n-1 \right\}\). For this to be the strong equilibrium group structure, no group of players should be able to collectively deviate in a way that they are all better off. Suppose that a group of players, consisting of player 1, player 2, \(\ldots\), player \(i-1\), jointly deviate from \(G_1\) and join \(G_2\). Then, the group structure is changed to \(G^{'}=\left\{ G_1 , G_2 \right\}\) where \(G_1 =\left\{ i+1, i+2, \ldots , n\right\}\) and \(G_2 =\left\{ 1, 2, \ldots , i\right\}\). Given the new group structure \(G^{'}\), the seceders obtain the following higher payoffs than those obtained under G:
So, any equilibrium group structure \(G=\left\{ G_1 , G_2 \right\}\), where \(G_1 =I \backslash \left\{ i\right\}\) and \(G_2 =\left\{ i\right\}\) for \(i \in \left\{ 3, 4, \ldots , n-1 \right\}\), cannot be the strong equilibrium structure.
Then, what about the equilibrium group structure \(G^* =\left\{ G_1 , G_2 \right\}\) where \(G_1 =I \backslash \left\{ n\right\}\) and \(G_2 =\left\{ n\right\}\)? Given the group structure \(G^*\), neither any individual player nor any group of players do have the incentives to deviate from the group structure, and thus the group structure \(G^*\) is the unique strong equilibrium outcome. \(\square\)
Proof of Proposition 3
Suppose that \(G=\left\{ G_1, \ldots , G_K \right\}\), where \(K \ge 2\) and \(|G_k |\ge 2\) for all \(k=1, \ldots , K\). We show that any player in \(G_k\) has no incentives to deviate from his group. For convenience, we denote players in \(G_k\) by player 1k, 2k, \(\ldots\) , \(n_k k\), where \(n_k =|G_k|\), i.e., \(G_k =\left\{ 1k, 2k, \ldots , n_k k\right\}\). First, consider the deviation incentive for player 1k (the strongest player in \(G_k\)). From Lemma 1, given the group structure G, player 1k obtains the following payoff:
If player 1k changes his decision \(g^{1k}\) to \(g_d^{1k} (\ne g^{1k})\), given the other players’ decisions unchanged, his group \(G_k\) collapses and all the players in \(G_k\) become singletons. That is, the new group structure \(G^{'}=\left\{ G_1 , \ldots , G_{k-1}, \left\{ 1k\right\} , \left\{ 2k\right\} , \ldots , \left\{ n_k k\right\} , G_{k+1}, \ldots , G_{K} \right\}\) appears. Given the new group structure \(G^{'}\), player 1k obtains the following payoff:
From conditions in (7), we have the following prerequisites for all the groups under the new group structure \(G^{'}\) to be active:
Summing up the conditions for each players \(2k, 3k, \ldots , n_k k\), we have
We now compare player 1k’s payoffs, \(\pi _{1k}\) and \(\pi _{1k}^d\):
From \((*)\), we see \(\pi _{1k}-\pi _{1k}^d >0\), which implies that player 1k in \(G_k\) does not have the deviation incentive.
Next, consider the deviation incentive for player \(jk (\ne 1k)\) in \(G_k\). Given the group structure G, player jk has his payoff
If player jk changes his decision \(g^{jk}\) to \(g_d^{jk} (\ne g^{jk})\), given the other players’ decisions unchanged, the new group structure \(G^{'}=\left\{ G_1 , \ldots , G_{k-1}, \left\{ 1k\right\} , \left\{ 2k\right\} , \ldots , \left\{ n_k k\right\} , G_{k+1}, \cdots , G_{K} \right\}\) appears. Under the new group structure \(G^{'}\), he has the following payoff:
We now compare player jk’s payoffs, \(\pi _{jk}\) and \(\pi _{jk}^d\):
Thus, we see \(\pi _{jk}-\pi _{jk}^d >0\), which implies that player \(jk (\ne 1k)\) in \(G_k\) does not have the deviation incentive.
To conclude, no players in \(G_k\) have an incentive to deviate from the group structure G. \(\square\)
Proof of Proposition 5
For all possible equilibrium group structures, we keep to eliminate vulnerable ones to collective deviations of any conceivable group of players. First, consider the group structure \(G=\left\{ G_1 , \ldots , G_K \right\}\) where \(K\ge 3\). As before, we assume that player 1 belongs to \(G_1\), i.e., \(1\in G_1\). Given the group structure G, payoffs for players i in \(G_1\) and those for players j in \(G_k\) are as follows, respectively:
and
\(\square\)
Now suppose that players in \(G_1\) and players in \(G_k\) jointly change their decisions to combine the two groups, \(G_1\) and \(G_k\), given the other players’ decisions unchanged. Namely, players i in \(G_1\) change their decisions \(g^i =G_1\) to \(g_d^i =G_1 \cup G_k\), and player j in \(G_k\) change their decisions \(g^j =G_k\) to \(g_d^j =G_k \cup G_1\). Then, the new group structure \(G^{'}=\left\{ G_1^{'}, G_2, \ldots , G_{k-1}, G_{k+1}, \ldots , G_K \right\}\) is chosen, where \(G_1^{'} =G_1 (\in G) \cup G_k (\in G)\). Note that \(G_1\) and \(G_k\) in G are merged and the other groups in G remain the same. Given the new group structure \(G^{'}\), the payoffs for the deviators (players i in \(G_1\) and players j in \(G_k\) under the group structure G) are as follows:
and
Comparing the deviating players’ payoffs above, we have:
-
\(\pi _{11}^d - \pi _{11}=v_{11} (p_{1d}+p_1 )(p_{1d}-p_1 )=v_{11} (p_{1d}+p_1 )B_{11}^{-1} \times \frac{\sum _{l=1}^K B_{1l}^{-1}-(K-1) B_{1k}^{-1}}{\sum _{l=1}^K B_{1l}^{-1} \sum _{l=1, \ne k}^K B_{1l}^{-1}}>0\) by conditions in (7)
-
\(\pi _i^d - \pi _i=v_i (p_{1d}-p_1)>0\) for \(i\in G_1 \backslash \left\{ 11\right\}\)
-
\(\pi _{1k}^d -\pi _{1k}=v_{1k} (p_{1d}-p_{k}^2)>v_{1k} (p_{1d}-p_k)=v_{1k} \times \frac{(K-1)B_{1k}^{-1} \sum _{l=1, \ne k}^K B_{1l}^{-1}-(K-2) B_{11}^{-1} \sum _{l=1}^K B_{1l}^{-1}}{\sum _{l=1}^K B_{1l}^{-1} \sum _{l=1, \ne k}^K B_{1l}^{-1}}\) \(> v_{1k} \times \frac{(K-1)B_{1k}^{-1} \sum _{l=1, \ne k}^K B_{1l}^{-1}-(K-2) B_{1k}^{-1} \sum _{l=1}^K B_{1l}^{-1}}{\sum _{l=1}^K B_{1l}^{-1} \sum _{l=1, \ne k}^K B_{1l}^{-1}}=v_{1k} B_{1k}^{-1} \times \frac{(K-1) \sum _{l=1, \ne k}^K B_{1l}^{-1}-(K-2) \sum _{l=1}^K B_{1l}^{-1}}{\sum _{l=1}^K B_{1l}^{-1} \sum _{l=1, \ne k}^K B_{1l}^{-1}}\) \(=v_{1k} B_{1k}^{-1} \times \frac{\sum _{l=1}^K B_{1l}^{-1}-(K-1) B_{1k}^{-1}}{\sum _{l=1}^K B_{1l}^{-1} \sum _{l=1, \ne k}^K B_{1l}^{-1}}>0\) by conditions in (7)
-
\(\pi _j^d -\pi _j = v_j (p_{1d}-p_k ) >0\) \(j\in G_k \backslash \left\{ 1k\right\} .\)
All the deviators are better off under the new group structure \(G^{'}\), which means that, under the group structure G, the players in \(G_1\) and the players in \(G_k\) have an incentive to collectively deviate from the group structure G. So, every group structure \(G=\left\{ G_1 , \ldots , G_K \right\}\) with \(K\ge 3\) is defenceless against the collective deviation of some players.
Second, consider the group structure G with \(K=2\), i.e., \(G=\left\{ G_1 , G_2 \right\}\). Let \(G_2 =\left\{ 12, 22, \ldots , n_{2}2\right\}\), where \(n_2 =|G_2 |\ge 2\). Given the group structure G, payoffs for the players in \(G_1\) and those in \(G_2\) are as follows:
and
Now suppose that the players in \(G_1\) and the players, except for player \(n_{2}2\), in \(G_2\) jointly change their decisions to form a new group, given player \(n_2 2\)’s decision unchanged. Namely, players i in \(G_1\) change their decisions \(g^i =G_1\) to \(g_d^i =G_1 \cup G_2 \backslash \left\{ n_2 2\right\}\), and players j, other than player \(n_2 2\), in \(G_2\) change their decisions \(g^j =G_2\) to \(g_d^j =G_2 \backslash \left\{ n_2 2\right\} \cup G_1\). Then, the new group structure \(G^{'}=\left\{ G_1^{'}, G_2^{'}\right\}\) appears, where \(G_1^{'}=I\backslash \left\{ n_2 2\right\}\) and \(G_2^{'}=\left\{ n_2 2\right\}\). Given the new group structure \(G^{'}\), the payoffs of the deviators are as follows:
and
Comparing the deviating players’ payoffs above, we have:
-
\(\pi _{11}^d - \pi _{11} >0\)
-
\(\pi _i^d - \pi _i >0\) for \(i\in G_1 \backslash \left\{ 11\right\}\)
-
\(\pi _{12}^d -\pi _{12}>0\)
-
\(\pi _j^d -\pi _j >0\) \(j\in G_2 \backslash \left\{ 12, n_2 2\right\} .\)
All the deviators are better off under the new group structure \(G^{'}\), which means that, under the group structure G, the players in \(G_1\) and the players in \(G_2\), other than the weakest player in \(G_2\), have an incentive to collectively deviate from the group structure G. So, every group structure \(G=\left\{ G_1 , G_2 \right\}\) with \(|G_2|\ge 2\) is vulnerable to the collective deviation of the players.
Finally, consider the group structure \(G=\left\{ G_1 , G_2 \right\}\) with \(|G_2 |=1\). Specifically, suppose that \(G_1 =I \backslash \left\{ j\right\}\) and \(G_2 =\left\{ j\right\}\) where \(j\ne n\). Given the group structure G, players have the following payoffs:
and
Now suppose that the players, other than player n, in \(G_1\) and player j jointly change their decisions to organize a new group, given player n’s decisions unchanged. In other words, players i, other than player n, in \(G_1\) change their decisions \(g^i =I\backslash \left\{ j\right\}\) to \(g_d^i=I\backslash \left\{ n\right\}\), and player j changes his decision \(g^j =\left\{ j\right\}\) to \(g_d^j =I\backslash \left\{ n\right\}\). Then, the new group structure \(G^{'}=\left\{ G_1^{'}, G_2^{'}\right\}\) appears, where \(G_1^{'}=I\backslash \left\{ n\right\}\) and \(G_2^{'}=\left\{ n\right\}\). Given the new group structure \(G^{'}\), the payoffs of the deviators are as follows:
and
Comparing the deviating players’ payoffs above, we have:
-
\(\pi _{11}^d - \pi _{11} >0\)
-
\(\pi _i^d - \pi _i >0\) for \(i\in G_1 \backslash \left\{ 11, n\right\}\)
-
\(\pi _j^d -\pi _j >0\).
All the deviators are better off under the new group structure \(G^{'}\), which means that, under the group structure G, all the players other than player n (the lowest-composite-strength player in the contest) have an incentive to collectively deviate from the group structure G. So, every group structure \(G=\left\{ G_1 , G_2 \right\}\) with \(G_2 =\left\{ j (\ne n)\right\}\) is not immune to the collective deviation of the players.
Therefore, the unique group structure \(G=\left\{ G_1 , G_2 \right\}\), where \(G_1 =I\backslash \left\{ n\right\}\) and \(G_2 =\left\{ n\right\}\), is unsusceptible to every conceivable collective deviation by players. So, it satisfies the strong equilibrium property.
Appendix B: The share function approach to get the main results
Given a group structure \(G=\left\{ G_1, \ldots , G_K \right\}\) where \(K\ge 3\), let \(X=\sum _{k=1}^{K} X_{k}\). From (8), we then have the winning probability of a group \(G_k \in G\) or its share function, \(p_k\): \(p_k = 1-\frac{X}{B_{1k}}\). By summing up all K groups’ share functions, i.e., \(\sum _{k=1}^K p_k =1\), we have (10):
We denote \(p_k\) and X in equilibrium in the second stage by \(p_k (K)\) and \(X^* (K)\). Since \(p_k (K)=1-\frac{X^* (K)}{B_{1k}}\), we have the following lemma.
Lemma A
\(p_k (K)>p_l (K)\) for \(B_{1k}>B_{1l}\) and \(k\ne l\).
Denoting \(\pi _{i}\) for \(i\in G_k\) in the second-stage equilibrium by \(\pi _{i} (K)\), we have:
Suppose that \({\widetilde{K}}<K\). Given a group structure \(G=\left\{ G_1, \ldots , G_{{\widetilde{K}}}\right\}\), we then have equilibrium payoffs for players in a group \(G_k \in G\):
The payoff differences for players in \(G_k\) between the K-group structure (\(\left\{ G_1, \ldots , G_K \right\}\)) and the \({\widetilde{K}}\)-group structure (\(\left\{ G_1, \ldots , G_{{\widetilde{K}}}\right\}\)) are as follows:
and
We thus have the following lemma.
Lemma B
For \({\widetilde{K}}<K\) and \(i \in G_k\), \(\pi _i ({\widetilde{K}})>\pi _i (K) \Leftrightarrow p_k ({\widetilde{K}})>p_k (K) \Leftrightarrow X^* ({\widetilde{K}})<X^* (K)\).
Lemma B means, if \(X^* ({\widetilde{K}})<X^* (K)\), players in a group \(G_k\) are better off when the number of groups changes from K to \({\widetilde{K}}\), i.e., it decreases.
Now suppose that \(K\ge 3\) groups are formed in the first stage, i.e., \(G=\left\{ G_1 , G_2 , \ldots , G_K \right\}\), and that groups \(1, 2, \ldots , K\) are listed in descending order by the composite strengths of their strongest (active) players, thus \(1\in G_1\). Note that player 1 is the strongest among our players.
Lemma C
When \(K\ge 3\), the active player in \(G_2\), player 12, can be better off by joining in \(G_1\).
Proof
-
(a)
Suppose that \(|G_2|=1\) and player 12 moves to \(G_1\). The number of groups then changes from K to \(K-1\), and we have:
$$\begin{aligned} X^* (K-1) - X^* (K)=\frac{K-2}{\sum _{k=1, \ne 2}^K 1/B_k} - \frac{K-1}{\sum _{k=1}^K 1/B_k}<0. \end{aligned}$$From Lemmas A and B, \(p_1 (K-1)>p_1 (K)>p_2 (K)\). Thus, \(\pi _{12} (K) = v_{12} \left( p_2 (K)\right) ^2 < v_{12} \left( p_1 (K-1)\right) =\text {the payoff of player 12 when joining in } G_1\).
-
(b)
Suppose that \(|G_2|\ge 2\) and player 12 moves to \(G_1\). Then, although there are still K groups, \(p_1 (K)\) increases because player 12’s deviation results in the decrease of \(B_{12}\) and hence the decrease of \(X^* (K)\). Remind that \(X^* (K) =\frac{K-1}{\sum _{k=1}^K 1/B_{1k}}\) and \(p_k (K)=1-\frac{X^* (K)}{B_{1k}}\). Thus, \(\pi _{12} (K)=v_{12} \left( p_2 (K)\right) ^2< v_{12} \left( p_1 (K)\right) <v_{12}\cdot \text {increased }p_1 (K)=\text {the payoff of player 12 when joining in }G_1\). \(\square\)
By Lemma C, we see that \(K=2\) in equilibrium, that is, \(G=\left\{ G_1 , G_2\right\}\). By the same argument as in (b) in the proof of Lemma C, we also know that the active player in \(G_2\) has an incentive to join in \(G_1\) when \(|G_2|\ge 2\). Therefore, \(G_1 =I\backslash \left\{ i\right\}\) and \(G_2 =\left\{ i\right\}\) in equilibrium. By the way, if \(i=2\), player 1 can be better off by moving to \(G_2\). So, an equilibrium group structure should be as follows.
Proposition A
A SPE group structure is \(G=\left\{ G_1 , G_2 \right\}\) where \(G_1 =I\backslash \left\{ i\right\}\), \(G_2 =\left\{ i\right\}\), and \(i\in I\backslash \left\{ 1, 2\right\}\).
Under a SPE group structure described above, the payoffs of players in \(G_1\) are
where \(p_1 (2)=1-\frac{X^* (2)}{a_1 v_1}\) and \(X^* (2)=\frac{1}{1/{a_1 v_1}+1/{a_i v_i}}\). When \(i=n\), \(X^* (2)\) is minimized and thus \(p_1 (2)\) is maximized. This implies the players’ payoffs in \(G_1\) are maximized when \(G_2=\left\{ n\right\}\), which means that players \(1, 2, \ldots , n-1\) have a collective incentive to stick together and isolate player n.
Proposition B
The strong equilibrium group structure is \(G=\left\{ G_1 , G_2 \right\}\) where \(G_1 =I\backslash \left\{ n\right\}\), \(G_2 =\left\{ n\right\}\).
In exclusive membership game \(\Gamma\), the unanimity condition is required for organizing a group, and it remove a player’s unilateral deviation incentive from his group. Thus, other than the grand coalition, every group structure \(G=\left\{ G_1, \ldots , G_K \right\}\) with \(K\ge 2\) can be the SPE group structure. However, there exists a unique strong equilibrium group structure.
Proposition C
In exclusive membership game \(\Gamma\), the strong equilibrium group structure is \(G=\left\{ G_1 , G_2 \right\}\) where \(G_1 =I\backslash \left\{ n\right\}\), \(G_2 =\left\{ n\right\}\).
Proof
-
(a)
Consider a SPE group structure \(G=\left\{ G_1 , G_2 , \ldots , G_K \right\}\) with \(K\ge 3\) and check a collective incentive of all the players in \(G_2\) to join in \(G_1\). From the same argument as in (a) in the proof of Lemma C, we can see that the players in \(G_2\) obtain higher payoffs by collectively joining in \(G_1\) because \(X^* (K-1)<X^* (K) \Rightarrow p_1 (K-1)>p_1 (K)>p_2 (K)\). So, they have such a collective deviation incentive. In addition, since \(p_1 (K-1)>p_1 (K)\), every player in \(G_1\) can be better off by allowing \(G_2\) to join in his group. In sum, all the players in both \(G_1\) and \(G_2\) get higher payoffs when \(G_1\) and \(G_2\) merge. We hence find \(K=2\) in strong equilibrium.
-
(b)
Consider a SPE group structure \(G=\left\{ G_1 , G_2 \right\}\) with \(|G_2|\ge 2\). By the same argument as in (b) in the proof of Lemma C, we can see that \(|G_2|-1\) strongest players in \(G_2\) (the players except for the weakest one in \(G_2\)) have a collective incentive to join in \(G_1\). Also, by allowing them to join in \(G_1\), all the players in \(G_1\) can be better off due to the increase of \(p_1 (2)\). In sum, all the players in \(G_1\) and the strongest \(|G_2|-1\) players in \(G_2\) have a collective incentive to merge. That is, we find that \(|G_2|=1\) in strong equilibrium.
-
(c)
Consider a SPE group structure \(G=\left\{ G_1 , G_2 \right\}\) with \(G_2 =\left\{ i \right\}\) and \(i \in I\backslash \left\{ 1\right\}\). By the same argument as in Proposition B, payoffs of players in \(G_1\) are maximized when \(i=n\), which implies that players \(1, 2, \ldots , n-1\) have a collective incentive to move together and isolate player n. Finally, we have \(G_2 =\left\{ n\right\}\) in strong equilibrium. \(\square\)
Rights and permissions
About this article
Cite this article
Lee, D., Kim, P. Group formation in a dominance-seeking contest. Soc Choice Welf 58, 39–68 (2022). https://doi.org/10.1007/s00355-021-01346-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00355-021-01346-7