This study examines the variability of business growth rates in the U.S. private sector from 1976 onwards. Our central finding is a large secular decline in the cross sectional dispersion of business growth rates and in the average magnitude of business volatility. Put differently, employment has become much more stable at the level of individual businesses. This finding prevails for every major industry group in the U.S. economy, and it holds whether we define “businesses” in terms of firms or establishments.
The average volatility of firm-level employment growth rates has declined by more than 40% since 1982. This result stands in sharp contrast to previous findings of rising volatility for publicly traded firms. We confirm the volatility rise among publicly traded firms using new data, but we show that its impact is overwhelmed by a strong volatility decline among privately held firms. Media perceptions of deteriorations in employment stability place too much weight on developments for publicly traded firms. Privately held firms, hitherto little studied in this context, account for more than two-thirds of U.S. businesses employment.
The pattern of rising volatility among publicly traded firms and falling volatility among privately held firms holds in every major industry group. Employment shifts toward older businesses account for more than a quarter of the volatility decline among privately held firms. Higher volatility among more recently listed firms account for most of the volatility rise among publicly traded firms. This finding, and several other results discussed in our study, point to a major shift in the economic selection process governing entry into the set of publicly traded firms.
Steven J. Davis is a visiting scholar at AEI and a professor of international business and economics at the University of Chicago Graduate School of Business. John Haltiwanger is a professor of economics at the University of Maryland. Ron Jarmin and Javier Miranda work at the Center for Economic Studies, Bureau of the Census.