ECONOMY

michael cutrone & nolan mC carty 193
wrong decision. If the disagreement policy is the correct one, bicameralism would
do better since it requires two majorities to overturn the disagreement policy. Thus,
superiority of bicameralism would depend entirely on which outcome is designated
as the default. Of course, if they had this information ex ante, they wouldn’t need to
vote!
Another problem for bicameralism is that it will never be the ex post best decision
rule. Under unicameral majority rule, all legislators would agree ex post that it is best
to implement the majority’s preference. However, under bicameralism, all legislators
would like to reverse any decision that disagreed with the majority of all votes cast. 20
While the arguments based on the Condorcet jury models suggest that the requirement
of concurrent majorities is unlikely to aggregate information better than simple
majority rule, bicameralism may still affect the incentives of legislators to acquire
information and develop legislative expertise. If information conveys a legislative
advantage, bicameralism might induce inter-chamber competition in information
acquisition. On the other hand, if information is a public good, bicameralism might
induce more free riding. Rogers (1998) develops a game-theoretic model of the
decision of each chamber to become informed about the consequences of pending
legislation. His model predicts that the chamber with the lower information costs
generally obtains a first mover advantage as the second chamber free rides off its
information. However, again it is not clear why the free-riding incentives across
chambers would exceed those within chambers. The same competition to obtain
agenda control by acquiring superior information would presumably exist within a
unicameral legislature.
6 Conclusions
.............................................................................
In this chapter, we have considered a number of theoretical arguments in favor of
bicameralism as an organizing principal for modern legislatures. Regardless of the
theoretical framework or the collective action problem to be solved, we find that
both the positive and normative arguments in favor of bicameralism tend to be weak
and underdeveloped. Most of the effects of bicameralism are due primarily to quite
distinct institutional choices, such as malapportionment and super-majoritarianism,
which correlate empirically with bicameralism. There seems to be no logical reason
why the benefits of these institutions—like the protection of minority rights and the
²⁰ A critical objection to this analysis is that the legislators were non-strategic. In the bicameral
context, strategic legislators should condition their vote based on their beliefs in states of the world
where they would provide the tie-breaking vote in their chamber given their expectations of the outcome
in the other body. In a fully strategic model, Austen-Smith and Banks (1996) show that, for a given
legislature, priors, and signal quality, there is an optimal q-rule which fully aggregates all information
and makes the optimal decision given this information. Thus, a unicameral legislature using the optimal
voting rule would do at least as well as any bicameral arrangement.