Job Creation and Destruction: The Dominance of Manufacturing

Article excerpt

ESTIMATES OF GROSS JOB CREATION and destruction (gross flows) give a deeper perspective on the ebb and flow of labor markets in a market economy than do the headline-grabbing announcements of net employment growth. Gross flow data give insight into the uniformity of employment growth across different parts of the economy. The path of total employment may be the total of many industries with similar growth experiences or of many industries with extremely diverse experiences; overall employment growth may be the result of lots of job creation canceling lots of job destruction or only a little of each.

In addition, the mix between job creation and destruction can and does vary dramatically over the business and seasonal cycles in the economy. Considerable attention has been devoted recently to the behavior of gross flows in the labor market (Blanchard and Diamond, 1990; Davis and Haltiwanger, 1990, 1992; Ritter, 1993), and stylized facts from these descriptive analyses have begun to generate theoretical research (Mortensen and Pissarides, 1993). Little attention, however, has been devoted to the question of whether these facts characterize all parts of the economy or only particular segments. This paper addresses that question using the method for measuring gross flows developed in Ritter (1993). It examines gross job creation and job destruction in three broad sectors: goods production, trade, and service production excluding trade.

The main conclusion is that job creation and destruction behave much differently in the goods-producing sector than in the rest of the economy. Manufacturing and other goods-producing industries, which make up only a quarter of private nonfarm payrolls, contribute disproportionately to changes in overall job creation and destruction, particularly during recessions. Given systematic differences between goods- and service-producing sectors, it is misleading to draw sweeping conclusions (that is, "stylized facts") about the economy from aggregate gross flows (Blanchard and Diamond, 1990; Ritter, 1993) or from manufacturing gross flows (Davis and Haltiwanger, 1990, 1992). Anderson and Meyer (1994), studying labor turnover, also concluded that manufacturing was "atypical in a large number of dimensions."

In addition, the dynamics of job creation and destruction in manufacturing appear to have changed during the most recent recession. Combined with the declining share of goods production in overall employment, this suggests that the dynamics of job creation and destruction for the economy as a whole may be substantially different in the future.


The raw data used to construct gross job creation and destruction are monthly employment levels in several hundred industries in the private nonfarm sector of the economy. The payroll or establishment survey, on which the employment data are based, currently covers more than 370,000 establishments, including all firms with more than 250 employees and a subset of smaller firms. These data are benchmarked annually using yet more comprehensive information. The survey excludes agricultural workers, unpaid family workers, domestic workers in private homes, and self-employed persons. To focus on job creation and destruction driven primarily by market forces, the data used for this paper also exclude government workers, though the survey includes them.(1)

The details of constructing job creation and destruction series (and caveats about them) are described in Ritter (1993), but the main idea is as follows. First, the breadth of coverage is defined by the set of industries for which continuous employment data are available since 1972. The 1972 start date was chosen because, for a large fraction of industries outside manufacturing, disaggregated employment data are not available for earlier years. Thus, the data cover a comprehensive cross-section of the nonfarm business sector. In January 1972, employment was 58. …