Skip to main content
SpringerLink
Log in

Community dynamics in the lab

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

Abstract

This paper studies the dynamics of community formation when members differ substantially in their returns from voluntary local public good provision. Laboratory experiments are conducted to examine how agents relocate in response to both community provision and membership composition, as well as how the growth and stability of communities are dictated by moving costs and crowding. When the public good is congestible, such that returns are lower for larger populations, I find communities are characterized by instability, cyclical fluctuations in local provision, and a dynamic in which low demanders continually chase high demanders through locations. When congestion is eliminated, subjects with different returns do sometimes co-exist. Yet chronic, inefficient movement persists, suggesting that instability is driven by intrinsic preferences for community composition, as well as by sensitivity to congestion.

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

Similar content being viewed by others

Notes

  1. See, for instance, Hudson (1980)’s application of the invasion-succession model to urban revitalization.

  2. Florida and Melander (2007).

  3. Hudson (1988) expands the invasion-succession model to include this second wave.

  4. For instance, Cole (1987) describes the displacement of artists from New York City neighborhoods and Ley (2003) describes the movement of artists from gentrifying neighborhoods of Canadian cities.

  5. Likewise, trends similar to the cyclical pattern of neighborhood composition can be observed in other economic contexts. Karni and Schmeidler (1990) present a model of cyclical fashion trends in which consumers cycle through types of goods. In their model, members of a lower social class receive increasing utility from a good as more people consume it, while members of a higher social class receive utility that increases with the number of other high status users but decreases with the number low status users. Pesendorfer (1995) presents a model in which a monopolist introduces a redesigned good at a higher price as soon as the previous, discounted version becomes sufficiently popular. More desirable “high types” care more about matching with other high types than low types do and use the costly purchase of the new good as a means of coordinating with similar types.

  6. At a broad level, the inability of a central authority to observe individual values for a public good is at the heart of the preference revelation problem, which results in under provision (Musgrave 1939; Samuelson 1954). In the specific context of linear public goods experiments, Ledyard (1995) suggests that heterogeneous returns have a negative effect on contributions (unless information is incomplete and participants only interact once). Fellner et al. (2011) find that, on average, subjects contribute less when individual marginal returns cannot be linked to individual behavior and less still if they are unaware of the distribution of returns. In a public goods game in which participants are aware of each player’s returns, Reuben and Riedl (2013) find that a punishment mechanism is used to enforce contributions proportional to returns.

  7. For instance, Tiebout specifically references the desire of residents to have “nice” neighbors (Tiebout 1956, p. 418).

  8. Throughout U.S. history, the ubiquity of this concern may be seen in limitations on the mobility of the poor in federal legislation, ranging from the Articles of Confederation, which excluded “paupers” from those who had the right to move freely between states, to the Personal Responsibility and Work Opportunities Reconciliation Act of 1996, which prevented newcomers from receiving welfare benefits beyond what they had been receiving previously for up to a year following their move (Donahue 1997).

  9. In the typical public goods experiment, subjects receive an endowment in each period and, without discussion, choose how much to keep for private consumption and how much to contribute anonymously to the group. The total amount contributed is multiplied by a factor less than 1 and each subject receives this amount in addition to the portion of his endowment he kept for himself.

  10. Experimental studies that have found sustained cooperation in groups formed of previously cooperative subjects include: Gunnthorsdottir et al. 2007; Rigdon et al. 2007; Gunnthorsdottir et al. 2010; Yang et al. 2007; Cabrera et al. 2012; Burlando and Guala 2005; Ones and Putterman 2007; and Gächter and Thöni 2005. An exception is Ockenfels and Weimann (1999).

  11. The number of agents in the population is in line with the previous work on endogenous group formation reviewed in the previous section, which typically includes 9 or 12 agents. The limited number of locations used here is more restrictive than in these previous studies, which typically allow all participants to be in groups by themselves. However, all six locations are rarely used in this experiment (less than 1 % of observations) and thus the ability to form more communities likely would not have a large effect on the outcomes. Given the monetary incentives to form fewer groups, the availability of more locations could lead to more extreme results, in the form of lower efficiency and less stability.

  12. The ratio of High and Low Types is held constant across all sessions and was chosen to be a (near) equal split, consistent with public goods experiments in fixed, heterogeneous groups. Populations that are homogeneous or that have a more extreme ratio of types might experience greater stability, as the incentives for “chasing” would be muted. However, this is not a question addressed by the current investigation, which focuses on populations with heterogeneous returns.

  13. One CPG session was dropped because of a technical error, which caused no data to be recorded after the 27th period. One pilot CPG session was conducted that lasted 30 periods. Neither of these are included in the results reported in this paper, but are included in the data set. All of the pure sessions were conducted at Caltech, while the congestion sessions were conducted at both Caltech and Harvard. Every figure and table is reproduced with Caltech-only data in an appendix, available on request, which reveals no substantive changes or loss of significance.

  14. The number of sessions, or populations of interacting participants, in each treatment compares favorably with other work on public goods games with endogenous group formation (e.g. Charness and Yang (2014); Ahn et al. 2008, 2009); Gunnthorsdottir et al. 2007) and Page et al. (2005) have three to seven sessions of interacting subjects per treatment).

  15. Through pilot experiments it was determined that subjects who were provided with the full history never looked back more than two to three periods while choosing their location. Thus the available history was restricted to three periods to reduce clutter and confusion.

  16. Taking each subject as the unit of observation, the difference in contributions between the High and Low Types is significant at less than the 0.01 level in both conditions using Wilcoxon rank-sum tests (\(Z=6.39\) for CPG and \(Z= 5.91\) for PPG). The differences remain significant at less than the 0.01 level if the session is taken as the unit of observation (\(Z= 3.49\) and \(Z=3.4\).)

  17. See for instance: Isaac et al. (1984).

  18. The difference in movement is significant if the unit of observation is the subject (\(Z=-5.101\), \(p<0.01\)) or the session (\(Z=-2.252\), \(p=0.024\)).

  19. The average community size experienced by participants in the PPG condition is significantly higher if the subject is taken as the unit of observation (\(Z=9.43\)) or if the session is the unit of observation (\(Z=3.4\)). The average community sizes of 4 (CPG) or 7 (PPG) group members indicates that the groups formed are, on average, of similar size to those usually formed by experimenters in VCM experiments.

  20. If either the individual or the session is taken as the unit of observation, \(p<0.01\).

  21. This difference is significant at \(p<0.05\) (\(Z=-1.977\)).

  22. Neither the sign nor significance of any estimates change if session-level random effects are included.

  23. The decrease in contribution over time is significant for Low Types with individual fixed effects in both the CPG (\(p<0.01\)) and PPG (\(p=0.06\)) sessions.

  24. These patterns are broadly consistent with the findings of Isaac et al. (1994) that participants with low MPCR (in homogeneous groups) respond positively to group size.

  25. The difference between types in the provision of the community entered is significant at \(p<0.01\) in both the CPG and PPG sessions in Wilcoxon rank-sum tests, taking the individual subject’s average outcome over the session as the unit of observation. If session-level random effects are included the PPG sessions are different only at the \(p=0.03\) level and, if the session is taken as the unit of observation, \(p=0.14\). The difference remains significant at \(p<0.01\) in the CPG sessions.

  26. The individual subject is the unit of observation and the variable considered is the difference in the size of the community entered vs. exited (not including the individual moving). The difference between types is significant at the \(p<0.01\) level in the CPG (\(\mathrm{Z}=-5.36\)) and PPG (\(\mathrm{Z}=-3.38\)) conditions. The difference is significant at the \(p<0.01\) in both conditions if session-level random effects are included. If the session-type is the unit of observation, the difference is significant at the \(p<0.01\) level in the CPG condition (\(\mathrm{Z}=-3.57\)) and at the \(p=0.07\) level in the PPG condition (\(\mathrm{Z}=-1.81\)).

  27. At the individual level, the average contribution made by a participant in a high entry fee location is 11.7, compared to 5.8 in low fee locations. The effect of entry fee type on average contribution is significant at the \(p<0.01\) level (with each individual as the unit of observation and accounting for session effects) and holds even controlling for type. At the community level, average provision is 31 in populated high fee locations compared to 17.2 in other populated locations (\(p<0.01\) with session-level controls).

References

  • Ahn TK, Issac RM, Salmon T (2008) Endogenous group formation. J Public Econ Theory 10(2):171–194

    Article  Google Scholar 

  • Ahn TK, Issac RM, Salmon T (2009) Coming and going: experiments on endogenous group size for excludable public goods. J Public Econ 93:336–351

    Article  Google Scholar 

  • Andreoni J (1988) Why free ride? Strategies and learning in public goods experiments. J Public Econ 37:291–304

    Article  Google Scholar 

  • Bohnet I, Kübler D (2005) Compensating the cooperators: is sorting in the prisoner’s dilemma possible? J Econ Behav Organ 56(1):61–76

  • Brekke KA, Hauge KE, Lind JT, Nyborg K (2011) Playing with the good guys: a public good game with endogenous group formation. J Public Econ 95(9):1111–1118

    Article  Google Scholar 

  • Burlando R, Guala F (2005) Heterogeneous agents in public goods experiments. Exp Econ 8(1):35–54

    Article  Google Scholar 

  • Burlando R, Hey JD (1997) Do Anglo-Saxons free-ride more? J Public Econ 64:41–60

    Article  Google Scholar 

  • Cabrera S, Fatas E, Lacomba J, Neugebauer T (2012) Splitting leagues: promotion and demotion in contribution-based regrouping experiments. Exp Econ 16(3):1–16

    Google Scholar 

  • Chaudhuri A (2011) Sustaining cooperation in laboratory public goods experiments: a selective survey of the literature. Exp Econ 14(1):47–83

    Article  Google Scholar 

  • Charness G, Yang C-L (2014) Starting small toward voluntary formation of efficient large groups in public goods provision. J Econ Behav Organ 102:119–132

    Article  Google Scholar 

  • Cinyabugma M, Page T, Putterman L (2005) Cooperation under the threat of expulsion in a public goods experiment. J Public Econ 89(8):1421–1435

    Article  Google Scholar 

  • Cole DB (1987) Artists and urban redevelopment. Geogr Rev 77(4):391–407

    Article  Google Scholar 

  • Croson R (1996) Partners and strangers revisited. Econ Lett 53:25–32

    Article  Google Scholar 

  • Donahue J (1997) Tiebout? Or not tiebout? The market metaphor and America’s devolution debate. J Econ Perspect 11(4):73–81

    Article  Google Scholar 

  • Ehrhart K-M, Keser C (1999) Mobility and cooperation: on the run. Working Paper

  • Fisher J, Isaac RM, Schatzberg JW, Walker JM (1995) Heterogeneous demand for public goods: effects on the voluntary contributions mechanism. Public Choice 85:249–266

    Article  Google Scholar 

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

    Article  Google Scholar 

  • Fischbacher U, Gächter S, Fehr E (2001) Are people conditionally cooperative? Evidence from a public goods experiment. Econ Lett 71:397–404

    Article  Google Scholar 

  • Fischbacher U, Schudy S, Teyssier S (2014) Heterogeneous reactions to heterogeneity in returns from public goods. Social Choice Welf 43:195–217

    Article  Google Scholar 

  • Fellner G, Iida Y, Kröger S, Seki E (2011) Heterogeneous productivity in voluntary public good provision: an experimental analysis. IZA Discussion Paper, 5556

  • Florida R, Melander C (2007) There goes the metro: How and Why Bohemians, Artists, and Gays Affect Regional Housing Values. J Econ Geogr 10(2):167–188

    Article  Google Scholar 

  • Gächter S, Thöni C (2005) Social learning and voluntary cooperation among like-minded people. J Eur Econ Assoc 3(2–3):303–314

    Article  Google Scholar 

  • Goeree JK, Holt CA, Laury SK (2002) Private costs and public benefits: unraveling the effects of altruism and noisy behavior. J Public Econ 83:255–276

    Article  Google Scholar 

  • Gunnthorsdottir A, Houser D, McCabe K (2007) Disposition, history and contributions in public goods experiments. J Econ Behav Organ 62:304–315

    Article  Google Scholar 

  • Gunnthorsdottir A, Vragov R, Seifert S, McCabe K (2010) Near-efficient equilibria in contribution-based competitive grouping. J Public Econ 94:987–994

    Article  Google Scholar 

  • Gürerk Ö, Irlenbusch B, Rockenbach B (2006) The competitive advantage of sanctioning institutions. Science 312:108–111

    Article  Google Scholar 

  • Hamilton BW (1975) Zoning and property taxation in a system of local governments. Urban Stud 12:205–211

    Article  Google Scholar 

  • Hudson JR (1980) Revitalization of inner-city neighborhoods: an ecological approach. Urban Aff Rev 15(4):397–408

    Article  Google Scholar 

  • Hudson JR (1988) Pioneer artists and middle class settlers: lifestyle variables and neighborhood succession. J Family Econ Issues 9(3):250–267

    Article  Google Scholar 

  • Isaac RM, Walker JM, Thomas S (1984) Divergent evidence on free riding: an experimental examination of some possible explanations. Public Choice 43(2):113–149

    Article  Google Scholar 

  • Isaac RM, Walker J, Williams AW (1994) Group size and the voluntary provision of public goods: experimental evidence utilizing large groups. J Public Econ 54:1–36

    Article  Google Scholar 

  • Karni E, Schmeidler D (1990) Fixed preferences and changing tastes. Am Econ Rev 20:262–267

    Google Scholar 

  • Ledyard J (1995) Public Goods: A Survey of Experimental Research. In: Roth AE, Kagel J (eds) The Handbook of Experimental Economics. Princeton University Press, Princeton

  • Ley D (2003) Artists, aestheticisation and the field of gentrification. Urban Stud 40:2527–2544

    Article  Google Scholar 

  • Musgrave RA (1939) The voluntary exchange theory of public economy. Q J Econ 53:213–237

    Article  Google Scholar 

  • Ockenfels A, Weimann J (1999) Types and patterns: an experimental east-west-german comparison of cooperation and solidarity. J Public Econ 71:275–287

    Article  Google Scholar 

  • Ones U, Putterman L (2007) The ecology of collective action: a public goods and sanctions experiment with controlled group formation. J Econ Behav Organ 62:495–521

    Article  Google Scholar 

  • Ostrom E (2000) Collective action and the evolution of social norms. J Econ Perspect 14(3):137–158

    Article  Google Scholar 

  • Page T, Putterman L, Unel B (2005) Voluntary association in public goods experiments: reciprocity, mimicry and efficiency. Econ J, Royal Econ Soc 115(506):1032–1053

    Google Scholar 

  • Pesendorfer W (1995) Design innovations and fashion cycles. Am Econ Rev 85(4):771–792

    Google Scholar 

  • Reuben E, Riedl A (2013) Enforcement of contribution norms in public good games with heterogeneous populations. Games Econ Behav 77:122–137

    Article  Google Scholar 

  • Rigdon M, McCabe K, Smith V (2007) Sustaining cooperation in trust games. Econ J 117:991–1007

    Article  Google Scholar 

  • Robbett A (2014) Local institutions and the dynamics of community sorting. Am Econ J: Microecon 6:1–22

    Google Scholar 

  • Samuelson PA (1954) The pure theory of public expenditure. Rev Econ Stat 36:387–389

    Article  Google Scholar 

  • Tiebout CM (1956) A pure theory of local expenditures. J Political Econ 64:416–424

    Article  Google Scholar 

  • Wilson JD (1998) Imperfect solutions to the musical suburbs problem. In: Pines D, Sadka E, Zilcha I (eds) Topics in public economics: theoretical and applied analysis. Cambridge University Press, Cambridge, UK

    Google Scholar 

  • Yang C-L, Yu C-SJ, Yue I-T (2007) The rise of cooperation in correlated matching prisoners dilemma: an experiment. Exp Econ 10:3–20

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Department of Economics, Middlebury College, Middlebury, VT, 05753, USA

    Andrea Robbett

Authors
  1. Andrea Robbett
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Andrea Robbett.

Additional information

I wish to thank Charles Plott, Leeat Yariv, Peter Matthews, Rod Kiewiet, John Ledyard, and Jean-Laurent Rosenthal, as well as two anonymous referees and the editor for their helpful comments, Meghan McKeown for lab assistance, and the Caltech Social Science Experimental Laboratory and the Harvard Decision Science Laboratory for their hospitality. Funding was provided by the Caltech Laboratory for Experimental Economics and Political Science and by Middlebury College.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Robbett, A. Community dynamics in the lab. Soc Choice Welf 46, 543–568 (2016). https://doi.org/10.1007/s00355-015-0928-x

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00355-015-0928-x

Keywords

Search

Navigation