Affiliations: [a] Department of Urban Studies, University of Akron,
Gallucci Hall, Akron, OH 44325, USA | [b] Department of Urban Studies, Massachusetts Institute
of Technology, USA | [c] Council of State Policy and Planning Agencies,
Washington, DC, USA
Note: [1] Research in this study was funded by the U.S. Economic Development
Administration (EDA). The authors would like to thank James Souby and
Barbara Dyer of the Council of State Policy and Planning Agencies and David
Geddes of EDA for their support on this project. Special thanks are extended
to Celeste Long of the USDA Economic Research Service and Dick Sampson of
the Department of Employment Services for providing us with the USEEM and
ES202 data. We very much appreciate Governor Terry Branstad's help in
facilitating the project. Several eminent scholars with expertise on this
topic reviewed our paper: Sammis White, Urban Research Center, University of
Wisconsin–Milwaukee; Jim Miller, Economist, Economic Research Service, USDA;
Bruce D. Phillips, Director of Database Management, U.S. Small Business
Administration; and Roger Vaughan, Poger Vaughan and Associates. We are
deeply indebted for their assistance. We, of course, bare sole
responsibility for errors in and interpretation of the findings.
Abstract: Researchers conducting economic base, targeted industry, business survival, or
new business start-up research must be able to enumerate all firms in local
economies. Dun and Bradstreet market identifier files (and their derivative
public use database, U.S. Enterprise and Establishment Microdata file), state
unemployment insurance administrative files (ES202), and direct enumeration
(e.g. telephone Yellow Pages, etc) are becoming more widely used for
enumeration, but little is known about their comparative representativeness.
Using rural Iowa as a case study, the three databases were compared. Results
showed that each database was biased in its own way and therefore not strictly
comparable to the others. The study concluded that the databases would need to
be supplemented if high levels of accuracy are required.
