Spring 2001
Inquiry
, Vol.38, No.1 (Spring 2001): 6-21
Peter J. Cunningham, Paul B. Ginsburg
sing data from the 1996-97 Community Tracking Study household survey, this study examines variations in uninsurance rates
across communities in the United States. Specifically, regression-based decomposition is used to identify factors that account for
high rates of uninsurance in some communities. Differences in explained rates between "high uninsurance" and "low uninsurance"
communities are the result of differences in the racial/ethnic composition and socioeconomic status of the population (33%),
differences in employment characteristics (26%), and state Medicaid eligibility requirements (12.7%). Although higher costs are
associated with a higher likelihood that individuals are uninsured, high-cost communities tend to have lower rates of uninsurance as
a result of other factors. Despite the large number of identifiable factors included in the analysis, there is still a substantial amount
of unexplained regional variation in uninsurance rates.
Free access to this article is available at the Inquiry
Web site.