Job Flows in Dutch Manufacturing, 1979–1993** Empirical Evidence and Theoretical Implications

In the past decade the number of studies in labour market dynamics has proliferated. Especially the papers by Davis and Haltiwanger ~1990, 1992! on job creation and job destruction in the USA were followed by a host of studies on job flows in other countries. Examples include Burda and Wyplosz ~1994! for a number of European countries, Leonard and Van Audenrode ~1993! for Belgium and Blanchflower and Burgess for the UK. In this paper, we will describe and characterize the flows of job creation and job destruction in the Dutch manufacturing sector and we will discuss the implications of our results for a number of currently developed economic theories on job and firm turnover. For extensive treatment of these recent theories, see Jovanovic ~1982!, Davis and Haltiwanger ~1990!, Caballero and Hammour ~1994!, Burda and Wyplosz ~1994!, and Mortensen and Pissardes ~1994!. Most empirical studies of job flows rely on longitudinal plant-level employment observations. For The Netherlands, we have three potential sources. The one used by Hamermesh et al. ~1994! is based on a data set of Dutch firms in all economic sectors. Its major drawback, however, is that it is limited to only one year, so no account can be taken of entry and exit of firms. The second source is an annual survey by the Dutch Chambers of Commerce including firms of all sectors for the period 1987–1992. In this data set, only firms with more than 50 employees are present on a full sample basis; cf. Van der Hoeven and Verhoeven ~1994!. The source we use is based on a panel of firm-level data in the manufacturing sector collected by The Netherlands’ Central Bureau of Statistics over the period 1978– 1993. Our main findings are summarized in the following points. ~1! Firms, even defined in a narrowly defined class ~two-digit SBI, the Dutch SIC system!, are