Company births, deaths and marriages: Flaws in age fields in business microdata

The gazelle is, by definition, rare. Born in Massachusetts Institute of Technology’s Center for the Study of Neighborhood and Regional Change, David Birch’s concept of the gazelle captured the imagination of researchers because he claimed that the lion’s share of new jobs in the United States were created by this single species of establishment: relatively small (less than 100 staff), relatively uncommon, making up just 4 per cent of all firms in the United States and relatively young (less than 4 years old) (Birch 1979; Birch and Medoff 1994). The taxonomy of gazelle companies has expanded considerably since Birch introduced the concept, but the focus on gazelles in the literature has remained on a narrowly defined group of firms associated with rapid employment growth. Dispute has emerged over the thrust of Birch’s thesis—that a small group of companies creates the majority of jobs—and this article tackles one of the sources of dispute, the integrity of age variables in Birch’s base data source, Dun and Bradstreet databases.