Research
Publications
Distributive Politics in Opposition: Electoral Targeting of Social Assistance in Istanbul
Forthcoming, European Journal of Political Economy.
Abstract
Research on distributive politics overwhelmingly examines actors who control the state apparatus. Far less is known about how opposition parties governing at the subnational level allocate resources under their direct control. I study Istanbul, where the main opposition Republican People’s Party (CHP) won the mayoralty in 2019 after twenty-five years of Justice and Development Party (AKP) dominance and expanded the municipality’s means-tested cash assistance program, the Social Support Istanbulkart, from roughly 176,000 to 224,000 recipient households. Using neighborhood-level data on aid distribution, electoral outcomes, and socioeconomic indicators, I find that allocation follows an inverted U in CHP vote share, peaking in electorally marginal neighborhoods and declining in both parties’ strongholds. The pattern replicates within a contiguous border-pair design and is strongest where need-based explanations are weakest. The targeting was followed by modest gains: a one-standard-deviation rise in aid recipients per resident is associated with a 0.36 percentage-point increase in opposition vote share in the 2023 presidential election, in a city whose annulled March 2019 mayoral race was decided by 0.16 percentage points. Electoral targeting in distributive allocation is thus not confined to national incumbents: an opposition party that lacks the machinery of contingent exchange can pursue it instead through the geographic calibration of a programmatic instrument.
Working Papers
Governing to Win: The Electoral Returns to Opposition Service Provision under Authoritarianism
Using geospatial data on Istanbul’s subsidized municipal restaurants (Kent Lokantası) and a quasi-experimental design, I show that, in a competitive authoritarian regime, an opposition party that governs a major metropolitan municipality can convert programmatic service provision into electoral support.
Work in Progress
An empire-wide digitization of Ottoman provincial yearbooks that uses large language models to transcribe and structure district-level administrative records from the Ottoman Turkish originals, building a panel of late Ottoman state capacity to trace the origins of divergence among the empire’s successor states.