(Page 2 of 55) - The Voluntary Provision of "Bi-Public Goods": Public Interest Groups and the Collective Action Problem, Theory and Tests authored by Jankowski, Richard.

ABSTRACT

Mancur Olson in “The Logic of Collective Action” presents an analysis of the freerider

problem faced by groups that provide a public good. His analysis is explicitly limited to the
collective action problem in economic interest groups, such as unions and business associations.
Olson specifically notes that his analysis does not apply to philanthropic organizations. In recent
years, Walker (1991) and others have documented a substantial growth in what are called public
interest groups, such as Greenpeace, Amnesty International and Common Cause. These public
interest groups are examples of Olson’s philanthropic groups because their goal is helping others
or society in general, rather than solely benefitting the members of the group.

I propose an explanation of how public interest groups solve their collective action

problem by modifying Olson’s analysis. First, I modify the underlying rational, self-interested
axiom by incorporating weak altruistic motives to individuals. Second, I analyze the game as
one of incomplete information. Third, benefit provided is not the traditional public good, rather
it is a “bi-public good.”

I incorporate these three features into a formal model of voluntary contributions to the

provision of a public good. I show that a Bayes-Nash equilibrium exists which provides for
positive contributions to the public good. My solution is different from that of Ainsworth and
Sened (1993) who analyze the collective action problem of interest groups by modeling it as a
signaling game with an entrepreneur. Their model is based upon Palfrey and Rosenthal’s (1984,
1988) analysis of the voluntary provision of a discrete, public good. However, Ainsworth and
Sened’s analysis applies only where there are no competing interest groups (a rarity in the
political process). My analysis explicitly models the provision of a “bi-public good” when there
are competing interest groups over a good that is simultaneously good and bad for different
groups.

In the second half of the analysis, I test the theoretical model. I use the 2004 GSS

analysis of altruism as the instrument for my tests. I find strong support for the hypothesis.
Membership in economic groups is not correlated with altruism, i.e., selective benefits are
sufficient. But, membership in groups advocating social or political causes is correlated with
altruistic motivation. In addition, contributing of money to social and political causes is
correlated with altruistic behavior.