Fed governments now use empirical measures to push policy goals down to local areas—what we call performance federalism. We argue this approach distorts democratic accountability in lower-level elections.

Study Context:

This research analyzes voter behavior during school tax referenda in the U.S., where a state provided a widely publicized indicator of district performance.

We show voters cannot reliably infer educational quality from these indicators, leading to unintended consequences. A federal signal indicating poor district outcomes actually increases failure rates for needed tax levies—a surprisingly robust effect that hits hardest in poorer communities.

Findings:

• Voter decisions appear influenced by misleading federal information • Poorer districts face significantly higher levy failure risks with negative signals • The effect holds across multiple identification strategies and mechanism tests • This indicates voters may lack the tools to properly evaluate performance data

These results suggest policy interventions using external metrics risk undermining local democratic decision-making processes.