Theory: Commitment problems make establishing long-term relationships
between members of Congress and organized interests extremely difficul
t. Hypothesis: Political action committees whose contribution behavior
is unfavorable to an election winner are not forced to pay for their
mistakes in the subsequent election to get back into the representativ
e's good graces. In other words, strong punishment strategies are not
employed to enforce what we label the campaign contribution contract.
Methods: Because theory suggests that campaign contributions involve d
ecisions about whether to make a donation and how much to give, a gene
ral econometric framework allowing these two decisions to be either in
dependent or correlated is utilized. This model subsumes both Tobit an
d the combination of probit and ordinary least squares estimation. A l
arge scale data set designed to uncover any evidence that elected repr
esentatives commit themselves ex ante to punish groups for opportunist
ic behavior is employed. Covering 1977-86, this data set consists of r
oughly 115,000 campaign contribution decisions made by large corporate
, labor, and trade political action committees concerning long-time me
mbers of Congress. Ancillary information is incorporated for model spe
cification purposes. Results: The willingness of legislators to punish
is marginal, suggesting that evidence for credible commitment is weak
. We need to reconsider whether legislators and group leaders can poss
ibly forge long-term relationships. Additional thought must also be gi
ven to the motivations for campaign contributions and to what a reason
able enforcement mechanism for commitment might look like.