Thomas Romer

We use a policy change that occurred in Oregon in the late 1980s to re‐visit the budget‐maximizing agenda setter theory of local public expenditure. Prior to 1987, Oregon school districts held operating levy elections with an exogenous, often zero or very low, spending reversion. From 1987 through 1990, districts experienced a “safety net” regime where the reversion was at least the previous year's nominal spending. We find that the “safety net” sharply limited the agenda setter's ability to use the reversion as a threat to obtain voter approval of relatively large expenditures.

This article analyzes election data that permit simple tests of rational voting and agenda setting. The voting test pertains to aggregate election results. The prediction is that when voters have single-peaked preferences, there will be more opposition to the second of two budget proposals that are voted on simultaneously. Unlike the standard binary choice setting, not all voters have weakly undominated voting strategies, but the game among the voters can be solved simply by iterative application of weak dominance. The agenda-setting prediction tested is that agenda setters should make one proposal rather than two when given the option. The data come from Oregon school district financial elections from 1980–83, years in which the rules for these elections were abruptly changed.