The Effects of Neighborhood Democracy on Cooperation: A Laboratory Study

ABSTRACT:

What effect does participation in neighborhood governments have on residents’ willingness to contribute to the provision of public goods? This is an important question to ask given the growth of homeowners associations (HOAs) in the United States. HOAs provide a medium through which citizens can practice democratic skills. The social capital and efficacy built through this participation may increase an individual’s willingness to use his or her own resources to help provide for a community public good. In this article the author hypothesizes that such participation increases the probability that individuals donate to public goods provision. Using an experimental design where subjects in treatment groups participate in a mock HOA meeting and then play an iterated public goods game, the study finds evidence that participation in HOA meetings improves cooperative behavior. This finding suggests that HOA participation has some effect on an individual’s propensity to contribute to the provision of a public good.

Notes

The original design of this study included a field experiment component where residents of the same neighborhood or homeowners association would play a public goods game at a community center during the weekend. Attempts to conduct these public goods experiments in the general population of homeowners in a medium-sized city in Florida were unsuccessful due to poor turnout. I received permission to perform this experiment using a public park area during a neighborhood association potluck weekend dinner. The experiment needed 8–10 subjects, and I randomly canvassed the neighborhood, distributing flyers advertising the experiment several days prior, and offering a $20 show-up fee. Of approximately 35 flyers distributed, only 2 individuals came to the experiment site. I attempted this experiment in one of the most civically-active neighborhoods in the city, and therefore concluded that either the show-up fee was not large enough to entice residents to participate, and/or a string of recent door-to-door scams led residents to distrust the merits of the activity. This original plan was abandoned due to low turnout in the most civically active neighborhood in Leon County. If these individuals were not willing to participate, other citizens in neighborhoods in Leon County would be even less likely to participate. I also did not engage in on-site recruitment, because that would have led to a biased sample and not allowed random selection into the experiment, which would hinder internal and external validity. The number of 8–10 subjects was recommended for the field experiment, because turnout was predicted to be low given the number of door-to-door scams during the study period. Twelve subjects is a more representative number for the experiment, because that mirrors the average turnout of citizens to HOA meetings as observed during a related project in Leon County. Therefore, twelve subjects were included in each treatment and control group of the actual laboratory experiment.

There is some debate as to whether or not students in the lab behave as adults, or even if adults in the lab behave as they do in field settings. Benz and Meier () help to address this concern in an experiment using college students that bears consideration for this study. They state, “While we find evidence that pro-social behavior is more accentuated in the lab, the data show that pro-social behavior in experiments is correlated with behavior in the field” (p. 268).

No demographic characteristics of the subjects were compiled other than whether or not they knew other subjects in the experiment due to constraints by the institutional review board (IRB). Very few subjects had friends or acquaintances in the experiment; therefore, I am confident there are no extraneous variables affecting donation behavior.

Treating the overly cooperative groups as outliers and removing them from the analysis provides significance for my hypothesis that Treatment Group 1 differs from the control group. However, I leave them in to increase statistical power and to provide a more rigorous test to reject the null hypothesis of no difference.

Homo economicus is a term that refers to an individual engaging in rational, utility-maximizing behavior.

Additional information

Notes on contributors

Daniel S. Scheller

Daniel S. Scheller is an Assistant Professor in the Master of Public Administration program at the University of Texas at El Paso. Prior to that, he was an Assistant Professor in the Department of Government at New Mexico State University and the Department of Political Science at Western Kentucky University. He received his PhD in political science from Florida State University. His research interests are in neighborhood governance, housing policy, and local economic development.