💡 These findings raise troubling questions for democracies facing rising polarization. If citizens evaluate public services through partisan lenses, elected and unelected government officials are held less accountable and face weaker incentives to deliver outcomes.

Results from CBS data, based on classification model.

Across both datasets, we find supportive results, suggesting that citizens’ evaluations of public services (education, health, policing, and transportation) are colored by their partisan match/mismatch with the governing coalition.

To compensate for the lack of political indicators in CBS dataset, we applied a novel Machine-Learning model to classify partisanship (two-bloc voting) based on demographic and geographic information (achieving ~80% accuracy).

The second (CBS) dataset addresses the limitations of self-selection bias in opt-in panels and the ‘cheerleading effect’ associated with political surveys.

Our analysis is based on two datasets:
⚙️ A large seven-wave survey political study (~10,000 participants).
⚙️ An administrative survey dataset by the Israeli Central Bureau of Statistics (~14,000 participants).

Using Israel’s two full government changes in 2021 and 2022 as a test case, we conduct two pre-registered studies to examine the shift in partisan evaluations of government services based on affiliation with the governing coalition.

We demonstrate that in a polarized environment, partisan bias is not limited to citizens’ general views and evaluations of macro-policy outcomes but also extends to everyday public services like education, health, policing, and transportation.

Absolutely

@poppublicsphere.bsky.social