Skip to main content
Log in

Private provision of discrete public goods: the correlated cost case

  • Original Paper
  • Published:
Social Choice and Welfare Aims and scope Submit manuscript

Abstract

This paper explores the effects of correlated costs between players in an incomplete information game in the context of the private provision of discrete public goods. In such an incomplete information game, without correlation between players, equilibria in cut-point strategies (in other words, monotone strategies) always exist. A cut-point strategy prescribes to contribute a positive amount up to a certain cost level (‘the cut-point’) and to contribute zero above that critical level. However, when the players’ costs of contributing are correlated, an equilibrium in cut-point strategies may fail to exist because new incentives to free ride arise. A sufficient condition for the existence of an equilibrium in cut-point strategies is provided, and equilibria in non-monotone strategies are studied. The case without correlation is analyzed as a special case.

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

Access this article

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

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6

Similar content being viewed by others

Notes

  1. With the general distribution function, the main arguments are repeated in the "Appendix". Although it is important to include the general distribution function, a particular distribution function and its parameters are necessary to determine the nature of the equilibria. Therefore, we study normal distributions.

  2. This single cut-point strategy is monotonic in player type, and such strategies are called monotone strategies. In Sects. 2, 3, and, 4, we concentrate on monotone strategies. In Sect. 6, we seek equilibria in non-monotone strategies in which there are multiple cut-points.

  3. Without correlation (α → ∞), the left-hand side of (11) is not a function of ci, and no new incentive will arise. This finding explains why the validity of cut-point strategies has not been discussed in previous studies.

  4. Necessity is considered in the "Appendix".

  5. The order of the arguments is as follows. First, we confirm the existence of equilibria in non-monotone strategies under certain parameters, and second, we check whether there is an equilibrium in single cut-point strategies under these parameters.

  6. For example, in a complete information game (G = 10, M = 2, k =1, c1=3, and c2=5), (p1, p2)=(1,0), \( \left( {\frac{7}{10},\frac{1}{2}} \right) \), (0,1) are equilibria. pi is player i’s probability of contributing.

References

  • Carlsson H, van Damme E (1993) Global games and equilibrium selection. Econometrica 61:989–1018

    Article  Google Scholar 

  • Goldstein I, Pausner A (2005) Demand-deposit contracts and the probability of bank runs. J Financ 60:1293–1327

    Article  Google Scholar 

  • Gradstein M (1994) Efficient provision of a discrete public good. Int Econ Rev 35:877–897

    Article  Google Scholar 

  • Gradstein M, Nitzan S (1990) Binary participation and incremental provision. Soc Choice Welf 7:171–192

    Article  Google Scholar 

  • Larry K, In HL, Robin M (2007) A global game with strategic substitutes and complements. Games Econ Behav 60:155–175

    Article  Google Scholar 

  • McBride M (2006) Discrete public goods under threshold uncertainty. J Public Econ 90:1181–1199

    Article  Google Scholar 

  • McCarty N, Meirowitz A (2007) Political game theory: an introduction. Cambridge University Press, New York

    Book  Google Scholar 

  • Menezes F, Monteiro P, Temimi A (2001) Private provision of discrete public goods with incomplete information. J Math Econ 35:493–514

    Article  Google Scholar 

  • Morris S, Shin HS (1998) Unique equilibrium in a model of self-fulfilling currency attacks. Am Econ Rev 88:587–597

    Google Scholar 

  • Morris S, Shin HS (2004) Coordination risk and the price of debt. Eur Econ Rev 48:133–153

    Article  Google Scholar 

  • Nitzan S, Romano RE (1990) Private provision of a discrete public good with uncertain cost. J Public Econ 42:357–370

    Article  Google Scholar 

  • Palfrey TR, Rosenthal H (1984) Participation and the provision of discrete public goods: a strategic analysis. J Public Econ 24:171–193

    Article  Google Scholar 

  • Palfrey TR, Rosenthal H (1988) Private incentives in social dilemmas: the effects of incomplete information and altruism. J Public Econ 35:309–332

    Article  Google Scholar 

  • Seade J (1980) On the effects of entry. Econometrica 48:479–489

    Article  Google Scholar 

Download references

Acknowledgements

For helpful comments and suggestions on an earlier draft, I am grateful to Jun-ichi Itaya, Noritaka Kudoh, Daisuke Oyama, Atushi Kajii, Takashi Ui, Ryusuke Shinohara, and Hideki Konishi. Similarly, I thank the participants at the Hakodate conference, Onuma International Seminar House (2011); the contract theory workshop at Hokkaido University (2011); the summer workshop on economic theory at Hokkaido University (2011); the 68th Congress of the International Institute of Public Finance; and the 2012 Japanese Economic Association autumn meeting. Moreover, I thank the associate editor and referees for their suggestions, especially on Sect. 6. All remaining errors are, of course, mine.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Shingo Yamazaki.

Appendix

Appendix

1.1 Proof of Proposition 1

When (12) is satisfied, the slope of \( G{ \Pr }(n = k - 1{\mid }c_{i} ) \) with respect to \( c_{i} \) is less than 1. Thus, \( \widehat{{c_{i} }} \) is a unique solution to (11). Player \( i \), whose disutility is \( \widehat{{c_{i} }} \), is indifferent between contributing and not contributing. From (8), player \( i \)’s best action is to contribute at a sufficiently small \( c_{i} \) and to not contribute at a sufficiently large \( c_{i} \). Thus, around \( \widehat{{c_{i} }} \), player \( i \) changes his/her action from contributing to not contributing. The cut-point strategy is the best response to other players’ cut-point strategies, and an equilibrium in the single cut-point strategies exists. Therefore, it is sufficient to ensure that the maximum slope of \( G{ \Pr }(n = k - 1{\mid }c_{i} ) \) is below 1.

1.2 Necessary and sufficient condition for the existence of an equilibrium in cut-point strategies

Here, we study a more detailed condition of Proposition 1. An equilibrium fails to exist in the single cut-point strategies if and only if (11) has multiple solutions. In this case, there is at least one intersection with the slope of the left-hand side of (11) is greater than or equal to 1. Therefore, an equilibrium in the single cut-point strategies exists if and only if the slope of the left-hand side of (11) is below 1 at any intersection \( \widehat{{c_{i} }} \). The following is a necessary and sufficient condition for the existence of equilibrium in the single cut-point strategies:

$$ \begin{aligned} \left. {\frac{{\partial H(c_{i} , \hat{c})}}{{\partial c_{i} }}} \right|_{{at c_{i} = \widehat{{c_{i} }}}} = & \frac{{H\left( {\widehat{{c_{i} }}, \hat{c}} \right)\left\{ {k - 1 - (M - 1) \varPhi \left( {\gamma \left( {\hat{c} - \frac{{\alpha \theta + \beta \widehat{{c_{i} }}}}{\alpha + \beta }} \right)} \right)} \right\}\frac{ - \,\gamma \beta }{\alpha + \beta }}}{{\left\{ {\varPhi \left( {\gamma \left( {\hat{c} - \frac{{\alpha \theta + \beta \widehat{{c_{i} }}}}{\alpha + \beta }} \right)} \right)} \right\}\left\{ {1 - \varPhi \left( {\gamma \left( {\hat{c} - \frac{{\alpha \theta + \beta \widehat{{c_{i} }}}}{\alpha + \beta }} \right)} \right)} \right\}}} \\ & \phi \left( {\gamma \left( {\hat{c} - \frac{{\alpha \theta + \beta \widehat{{c_{i} }}}}{\alpha + \beta }} \right)} \right) < 1\,for\,any\,\widehat{{c_{i} }}, \\ & where H(c_{i} ,\hat{c}) \equiv G\left( {\begin{array}{*{20}c} {M - 1} \\ {k - 1} \\ \end{array} } \right)\left\{ {\varPhi \left( {\gamma \left( {\hat{c} - \frac{{\alpha \theta + \beta c_{i} }}{\alpha + \beta }} \right)} \right)} \right\}^{k - 1}\\ & \left\{ {1 - \varPhi \left( {\gamma \left( {\hat{c} - \frac{{\alpha \theta + \beta c_{i} }}{\alpha + \beta }} \right)} \right)} \right\}^{M - k} . \\ \end{aligned} $$
(29)

In addition to the arguments in Proposition 1, the following results are found. When \( \theta \) is sufficiently large or small, \( H\left( {\widehat{{c_{i} }}, \hat{c}} \right)/\left[ {\left\{ {\varPhi \left( {} \right)} \right\}\left\{ {1 - \varPhi \left( {} \right)} \right\}}] \right. \) becomes close to 0. Therefore, when \( \theta \) is sufficiently large or small, an equilibrium in the single cut-point strategies exists.

Moreover, there is a unique equilibrium if and only if the slope of the left-hand side of (15) is below 1 at any intersection \( \hat{c}^{*} \). A necessary and sufficient condition for the uniqueness of the equilibrium is:

$$ \begin{aligned}\left. {\frac{{\partial H\left( {\hat{c},\hat{c}} \right)}}{{\partial \hat{c}}}} \right|_{{at\,\hat{c} = \hat{c}^{*} }} &= \frac{{H\left( {\hat{c}^{*} ,\hat{c}^{*} } \right)\left\{ {k - 1 - (M - 1)\varPhi \left( {\gamma \alpha \frac{{\hat{c}^{*} - \theta }}{\alpha + \beta }} \right)} \right\}\frac{ - \,\gamma \beta }{\alpha + \beta }}}{{\left\{ {\varPhi \left( {\gamma \alpha \frac{{\hat{c}^{*} - \theta }}{\alpha + \beta }} \right)} \right\}\left\{ {1 - \varPhi \left( {\gamma \alpha \frac{{\hat{c}^{*} - \theta }}{\alpha + \beta }} \right)} \right\}}}\phi \left( {\gamma \alpha \frac{{\hat{c}^{*} - \theta }}{\alpha + \beta }} \right)\\ &\quad < 1\,for\,any\,\hat{c}^{*} .\end{aligned} $$
(30)

1.3 Proof of Proposition 2

If (15) has a unique solution, this solution gives a unique equilibrium. Therefore, similar to Proposition 1, it is sufficient to ensure that the maximum slope of the left-hand side of (15) is below 1.

1.4 With general distribution

The arguments for the single cut-point strategy are repeated here for the case with a general distribution function. As in former sections, symmetric Bayesian Nash equilibria are studied. Disutility \( c_{i} \) consists of the common term \( c \) and the characteristic term \( d_{i} \):

$$ c_{i} = c + d_{i} , $$
(31)

where \( c \) is common across players and follows a continuous distribution. \( d_{i} \) is independent across players and follows a continuous distribution. From these expressions, a posterior distribution of \( c_{j} \) for player \( i \) whose disutility is \( c_{i} \) is acquired, and its probability function is defined as:

$$ {\text{P}}(c_{j} {\mid }c_{i} ). $$
(32)

Player \( i \)’s net expected utility from contributing is:

$$ \begin{aligned} & {\text{L}}(\hat{c},c_{i} ) - c_{i} , \\ & where\,{\text{L}}(\hat{c},c_{i} ) \equiv {\text{G}}\left( {\begin{array}{*{20}c} {{\text{M}} - 1} \\ {{\text{k}} - 1} \\ \end{array} } \right)\{ {\text{P}}(\hat{c}{\mid }c_{i} )\}^{k - 1} \{ 1 - {\text{P}}(\hat{c}{\mid }c_{i} )\}^{M - k} . \\ \end{aligned} $$
(33)

At the cut-point for player \( i\left( {\widehat{{c_{i} }}} \right) \), the following is satisfied:

$$ {\text{L}}\left( {\hat{c},\widehat{{c_{i} }}} \right) = \widehat{{c_{i} }}. $$
(34)

We can restate the condition regarding the existence of an equilibrium in the single cut-point strategies as:

$$ \left. {\frac{{\partial {\text{L}}\left( {\hat{c},c_{i} } \right)}}{{\partial c_{i} }}} \right|_{{at\,c_{i} = \widehat{{c_{i} }}}} = \frac{{{\text{L}}\left( {\hat{c},\widehat{{c_{i} }}} \right)\left\{ {k - 1 - (M - 1){\text{P}}\left( {\hat{c}{\mid }\widehat{{c_{i} }}} \right)} \right\}}}{{\left\{ {{\text{P}}(\hat{c}{\mid }\widehat{{c_{i} }})} \right\}\left\{ {1 - {\text{P}}(\hat{c}{\mid }\widehat{{c_{i} }})} \right\}}}\left. {\frac{{\partial {\text{P}}(\hat{c}{\mid }c_{i} )}}{{\partial c_{i} }}} \right|_{{at\,c_{i} = \widehat{{c_{i} }}}} < 1\,for\,any\,\widehat{{c_{i} }}. $$
(35)

This condition is necessary and sufficient for the existence of an equilibrium.

Symmetric equilibria are found from:

$$ {\text{L}}\left( {\hat{c}^{*} ,\hat{c}^{*} } \right) = \hat{c}^{*} . $$
(36)

\( \hat{c}^{*} \) is an equilibrium cut-point. The condition that ensures the uniqueness of the equilibrium is:

$$ \left. {\frac{{\partial {\text{L}}(\hat{c},\hat{c})}}{{\partial \hat{c}}}} \right|_{{at\,\hat{c} = \hat{c}^{*} }} = \frac{{{\text{L}}(\hat{c},\hat{c})\left\{ {k - 1 - (M - 1){\text{P}}(\hat{c}{\mid }\hat{c})} \right\}}}{{\left\{ {{\text{P}}(\hat{c}{\mid }\hat{c})} \right\}\left\{ {1 - {\text{P}}(\hat{c}{\mid }\hat{c})} \right\}}}\left. {\frac{{\partial {\text{P}}(\hat{c}{\mid }\hat{c})}}{{\partial \hat{c}}}} \right|_{{at\,\hat{c} = \hat{c}^{*} }} < 1\,for\,any\,\hat{c}^{*} . $$
(37)

This condition is necessary and sufficient for the uniqueness of the equilibrium.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Yamazaki, S. Private provision of discrete public goods: the correlated cost case. Soc Choice Welf 52, 477–496 (2019). https://doi.org/10.1007/s00355-018-1153-1

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00355-018-1153-1

Navigation