A comparison of employment growth and stability before and after the fort worth tornado

Bradley T. Ewing, Jamie B. Kruse, Mark A. Thompson

Research output: Contribution to journalArticlepeer-review

39 Scopus citations


This study examined the time series pattern of employment growth and stability in Fort Worth, Texas taking into account the March 28, 2000 tornado. The tornado is treated econometrically as an intervention and both the mean and conditional variance of employment growth were estimated. Overall, this regional labor market experienced a decline in the employment growth rate following the tornado. Among the sectors that exhibited differences in employment dynamics between the pre- and post-tornado periods, the mining sector experienced a significant increase in employment growth following the tornado while the service andwholesale, retail trade sectors experienced significant declines in employment growth in the post-tornado period. The manufacturing, service, and wholesale, retail trade sectors were characterized by greater stability (i.e., a lower level of employment growth volatility) in the post-tornado period than in the pre-tornado period. Interestingly, in several sectors, no differences in the time series dynamics of employment growth were detected between the pre- and post-tornado periods. These sectors included construction, finance, insurance, real estate, government, and transportation and public utilities.

Original languageEnglish (US)
Pages (from-to)83-91
Number of pages9
JournalEnvironmental Hazards
Issue number2
StatePublished - Jan 1 2003
Externally publishedYes


  • Employment growth
  • Fort Worth
  • Intervention analysis
  • Labor market risk
  • Tornadoes

ASJC Scopus subject areas

  • Global and Planetary Change
  • Geography, Planning and Development
  • Development
  • Environmental Science(all)
  • Sociology and Political Science


Dive into the research topics of 'A comparison of employment growth and stability before and after the fort worth tornado'. Together they form a unique fingerprint.

Cite this