The unemployment impact of the 1993 US midwest flood: A quasi-experimental structural break point analysis

Yu Xiao, Edward Feser

Research output: Contribution to journalArticlepeer-review

Abstract

Most studies of the labor market impact of natural disasters report findings drawn from the observation of one or a few geographic areas. In this study, we provide large-sample evidence of the unemployment impact of disasters based on an empirical assessment of 320 counties affected by the 1993 Midwest flood, the costliest flood in the US history. We undertake a disaster economic impact analysis that combines quasi-experimental comparison group methods with econometric structural break point techniques. The combination of research designs offers advantages over the implementation of either impact evaluation design alone. We find that the flood caused unemployment rates to spike in heavily damaged counties but that the effects dissipated quickly. Unemployment rates dropped to below pre-flood levels shortly after the floodwaters receded. Temporary unemployment rate impacts in lightly damaged counties are detectable but were modest. Our findings are consistent with other macroeconomic studies which suggest that natural disasters in the United States have few persistent long-term negative effects on the economic trajectories of affected counties.

Original languageEnglish (US)
Pages (from-to)93-113
Number of pages21
JournalEnvironmental Hazards
Volume13
Issue number2
DOIs
StatePublished - Apr 3 2014

Keywords

  • economic impact
  • natural disasters
  • quasi-experimental control group research designs
  • structural break point methods
  • time-series methods

ASJC Scopus subject areas

  • Global and Planetary Change
  • Geography, Planning and Development
  • Development
  • General Environmental Science
  • Sociology and Political Science

Fingerprint

Dive into the research topics of 'The unemployment impact of the 1993 US midwest flood: A quasi-experimental structural break point analysis'. Together they form a unique fingerprint.

Cite this