Game Responses as Survey Instruments: Measuring the Constituents of Social Capital

Download data and study materials from OSF

This experiment was fielded as part of a TESS telephone survey. Data and materials for all the studies included on this survey is available here.

Principal investigators:

Daniel Houser

George Mason University



Jason Shachat

Xiamen University



Dr. Jonathan W. Leland

National Science Foundation


Sample size: 1008

Field period: 10/2003


During the 1990s several social scientists argued that differences in economic success and political stability across countries and regions stem, to an important extent, from differences in their levels of social capital. While many studies have demonstrated the importance of social capital, further progress in understanding what social capital is, much less how it can be developed, requires better measures of its constituents. Games developed in the experimental economics literature can distinguish between trust, trustworthiness and altruism, all potential attributes of social capital. Consequently, our primary objectives here are: (i) to use these games directly as survey instruments in order to measure these dimensions of social capital; (ii) to use these measures to help better define the meaning of social capital, and (iii) to investigate whether these measures correlate with various demographic characteristics. This final objective is a first step towards better understanding why trusting relationships succeed and, sometimes, fail.


(i) Demographic characteristics might influence propensities to be trusting, trustworthy and altruistic.
(ii) Trust and related concepts might contribute to Internet adoption.

Experimental Manipulations

Subjects in telephone surveys were asked how they would respond when in various positions in either a trust or dictator game. Different subjects were placed in different roles in the trust game (either first or second mover.) The amounts of money involved (e.g., amounts received and amounts available to send back as second mover in the trust game, or amount available to send in a dictator game) were manipulated. Each subject reported how he or she would play in a trust game, and also a dictator game that was connected to their position in the trust game. Hence, both between and within analyses are possible with our data.


Amounts of money sent in each condition.

Summary of Results

Our data work continues. Our initial findings are related to the different possible meanings of the word "trust," an issue that our survey based design has forced us to explore. In common usage, the word trust concerns expectations that another person will behave in an honest and benevolent manner. People may, however, exhibit benevolent behavior for different reasons. It may be in a person's personal self-interest to do so. In this context, laws can create circumstances where honesty and benevolence (or at least not malevolence) is the best policy. The prospect of future interaction can also induce benevolent behavior. Alternatively, a person might choose to be good and/or honest because he receives psychic utility from doing so. A fourth possibility is that people behave honestly and benevolently in the expectation that such behavior will be reciprocated. Responses to the WVS question do not allow us to discriminate between these different definitions nor motivations for trust. Doing so, instead, requires much more precise descriptions of the context in which behaviors and preferences will be revealed - descriptions we now consider.


Our conclusions remain preliminary. We know that trust as measured by survey responses has produced an intriguing array of econometric results regarding the impact of trust on economic prosperity. The faith we put in these results is, on the other hand, undermined by questions regarding what the measure of trust derived from survey responses is measuring. The primary message of our early work is that if we are precise and careful in defining which connotation of trust we have in mind, clear hypothesis regarding implications for economic growth and crisp measurement approaches for exploring these hypotheses follow.


Leland, J., Houser D. and Shachat, J., 2004. "Measuring Trust and Trustworthiness." Mimeo.

Leland, Jonathan, Daniel Houser, and Jason Shachat. 2005. "Measuring Trust and Trustworthiness." In Trust and Entrepreneurship: A West–East Perspective edited by Hans-Hermann Hohmann and Friederike Welter. Cheltenham and Lyme: Edward Elgar.