We examine the behavior of not-for-profits using utility maximization. This
leads to testable hypotheses regarding the costs of agency associated with
the activities of not-for-profits. Our tests are similar to those previous
ly employed regarding not-for-profit objectives, but our interpretation is
different. Our empirical test uses data on environmental groups. Unlike pre
vious studies biased toward a finding of service maximization, we account f
or endogenous explanatory variables. A weak statement of our results is tha
t managers of environmental groups derive positive marginal utility from ex
penditures for purposes other than providing services. There may also be em
pirical support for a stronger conclusion.