Agenda Setters

Formally, agenda setters are actors who control which alternatives (in addition to the status quo) other players may vote for and (if there is more than one) the order in which these votes will take place. Depending on the extent of their control, and contingent on the preferences and behavior of the other players, agenda setters may possess agenda power. In political science, the identity, role, and power of agenda setters have been investigated in at least two scenarios: legislative voting and approval of agency budgets, either by (local) public referenda or by executive branch “sponsors.” We illustrate the role of agenda setters with examples from the second.

William Niskanen argued that senior bureaucrats have goals that are functions of their agency's budget: therefore, they seek to maximize their budgets. They are in a strong position to do so because the bureau is a monopoly supplier of its output (such as national defense or education); officials have an informational advantage over their sponsors (they know more about the agency's internal costs, and direct monitoring of the quantity/quality of its output is difficult); and, crucially, the bureaucrat can make a take-it-or-leave-it budget proposal. The bureaucrat here is an agenda setter with a binary agenda presenting a choice between his or her preferred budget proposal and some other budget reduced to some inappropriate or unacceptable level to politicians.

Thomas Romer and Howard Rosenthal posited a model (which they later tested) based on this logic. The empirical case in point is the Oregon school budget process. In that state, each school has a maximum budget fixed by law but can expand the upper limit by a referendum at an annual meeting. If the new budget fails to be approved by at least 50% of voters, then the budget reverts to that of the previous year. Consider Figure 1, which displays the utility as a function of budget level for the median voter who will be decisive in the electorate for the referendum. The curve is centered at Bmed, which is the median voter's preferred level of school spending. When the reversion point is Brev1, the voter—and thus the electorate—will accept anything in the reflection of Brev1 about Bmed. So, for example, they will approve anything up to Bprop1. Notice though, that an [Page 14]agenda setter does best in terms of maximizing his or her budget when the reversion level is most threatening. When the reversion level is higher, such as at, for example, Brev2, the reflection allows less expansion: only up to Bprop2 will be approved. Romer and Rosenthal find empirical evidence in Oregon that is consistent with their theoretical model.

Figure 1 Agenda setter's power is contingent on reversion point

It is possible to reframe the Romer and Rosenthal model for any bureaucrat-sponsor relationship, although evidence is more difficult to come by. In general, weakening the ability of the agenda setter to make take-it-or-leave-it offers significantly reduces the agenda setter's power relative to the sponsor.