Research

Work in Progress

• “Should I Mail or Should I Go: Insights From a One-Time All-Postal Runoff Election” (with V. Lindlacher)
  [Latest Version]
  Abstract: We investigate how reducing information costs due to forced experimentation with postal voting affects voting behavior. Leveraging a natural experiment during the Bavarian 2020 Mayoral Elections, we employ an event study design. We find a transitory increase in total turnout and a persistent substitution from in-person to postal voting. Notably, municipalities with a higher turnout in the past show a larger effect. Investigating the distribution of the information costs shows an age gradient with the highest information costs for the oldest age group. The conservative governing party (CSU) gains persistently from higher postal turnout and other conservative parties’ in-person voters.

• “Election Administration Failure Leaves Persistent In-Person Turnout Losses Despite Postal Adaptation”
  [Latest Version]
  Abstract: This paper studies whether visible election administration failure affects participation beyond the disrupted election itself. I examine Berlin’s 2021 elections, during which many precincts experienced ballot shortages, unlawful polling-place closures, and multi-hour queues—disruptions later validated by a Federal Constitutional Court annulment. Using precinct-level administrative data from eight elections between 2016 and 2025 and an event-study difference-in-differences design, I find persistent losses concentrated in in-person turnout, the voting mode directly affected by the breakdown. In affected precincts, in-person turnout remains about 2 percentage points lower in later regular elections. Postal voting rises by about 1 percentage point, but only partly offsets this decline, leaving total turnout persistently lower. The findings show that visible election administration failure can leave durable downstream effects on participation even when alternative voting modes remain available.

• “Environmental Shocks and Political Assimilation: Evidence from Movers and Flood Exposure in North Carolina”
  Abstract: This paper examines how major floods reshape partisan assimilation among internal migrants. I link the universe of North Carolina voter-registration records (2009-2024) to high-resolution flood maps and track approximately 170,000 one-time cross-county movers. A dual-clock stacked event study - aligned to both relocation and Hurricane Florence (2018) - shows that flood exposure differentially changes the dynamics of partisan assimilation. Republican movers who had already settled in flooded destinations display a sustained acceleration in alignment with local partisan majorities, whereas Democratic movers exhibit only a short-lived adjustment followed by partial reversion or shifts into Unaffiliated registration. Pre-move exposure has no comparable effect, indicating that post-move shocks act through the destination context rather than selection. The results suggest that environmental disasters alter the path - rather than the level - of political assimilation, operating through a combination of identity updating and social signaling mechanisms.

Policy Papers

• “Faktoren von Familiengründung, Kinderlosigkeit und Kinderreichtum in Ostdeutschland”
  (with K. Heisig and T. Scheurer), 2023, ifo Dresden Studie 89
• “Was Ersteltern von Personen ohne Kinder in Ostdeutschland unterscheidet”
  (with K. Heisig), 2023, ifo Dresden berichtet 31
• “Faktoren der Kinderlosigkeit in Ostdeutschland”
  (with K. Heisig and T. Scheurer), 2022, ifo Dresden berichtet 29