EMPLOYMENT AND TURNOVER IN UK MANUFACTURING INDUSTRIES, 1963–82‡
| Date | 01 May 1989 |
| Published date | 01 May 1989 |
| DOI | http://doi.org/10.1111/j.1468-0084.1989.mp51002004.x |
| Author | Simon M. Burgess |
OXFORD BULLETIN OF ECONOMICS AND STATISTICS, 51,2(1989)
0305-9049 $3.00
EMPLOYMENT AND TURNOVER IN UK
MANUFACTURING INDUSTRIES, 1963-82e
Simon M. Burgess
I. INTRODUCTION
It is well known that employment in UK manufacturing has been declining for
many years now, the fall since the end of 1979 being especially sharp This
has been accompanied by a secular fall in turnover: even in 1980-81, separa-
tions were lower than in any previous recession. However, the differences in
employment and turnover experiences between the industries making up
manufacturing are as interesting as the similarities. This paper seeks to
explain these differences and forms a companion to two others focusing on
the aggregate time series (Burgess, 1988; Burgess and Nickell, 1987).
Over the period 1963-82, employment in some industries such as FD
(food, drink and tobacco), CC (chemicals) and PP (paper, publishing and
printing) fell at only half the aggregate rate. Others, including MM (metal
manufacture), TX (textiles) and CF (clothing and footwear) lost jobs at twice
the aggregate rate. There is considerable divergence even over the final
thirteen quarters. One of the aims of this paper is to explain these differences
in employment performance. This is the subject of Section II. The rate of job
loss depends on movements in equilibrium employment and the speed of
adjustment to reach the new equilibrium. We shall be particularly concerned
with the relative importance of these factors in determining employment
performance Differences in the speed of adjustment depend, in turn, on
differences in adjustment costs and quit rates. Some cross-section evidence is
presented, which explains, for example, why employment in the engineering
and metal industries responds so sluggishly to shocks.
Section III tackles the issue of turnover. The manufacturing industries
exhibit widely differing separations rates - ranging from 3.3 percent per
month in FD to 1.5 percent per month in VE (vehicles). The declining trend
in turnover evident at the aggregate level, is repeated for each individual
industry. The evidence for aggregate separations suggests that most separa-
tions are quits, and that the fall in separations is accounted for by a decline in
am grateful to Steve Nickel! and the Editors for helpful suggestions, and to the Depart-
ment of Employment and the Leverhulme Trust for financial support.
163
Fo s
0M N
MGTk)I
BE' N
EN N Pp wCC ii
8M w
VE N
l In fact more than accounted for, because layoffs have risen. See Burgess and Nickell (1987).
164 BULLETIN
-1.2 -1.0 -0.8 -0.5 -0.4 -0.2 0.0
Emp'oyment Growth
Fig. 1. Employment and turnover
quits.' If this also applies to the individual industries, then most of the differ-
ences in separations will be due to differences in quits. A variety of models
have been proposed for turnover: search theories (see Parsons, 1973;
Mortensen, 1986), specific human capital theories (Mincer and Jovanovic,
1981; Parsons, 1986) or union 'voice' theories (Freeman, 1980), and these
are considered below. Cross-section evidence on average turnover rates is
presented, focusing on skill and demographic factors. A time series analysis
of the decline in separations is provided for a number of industries.
The method used in Sections II and III is to consider all the constituent
industries of manufacturing at quite a broad level, and then examine four
industries in more detail. While considering data availability and importance,
it was decided to select industries representative of the diversity of employ-
ment and turnover experience. Figure 1 plots the average engagement rate
against the average rate of change of employment, and the four chosen
industries, FD, EN (engineering), VE and LC (CF, clothing and footwear plus
LF leather, fur etc.), provide a varied mix of job loss and turnover rates.
Section IV summarizes the arguments and results.
II. EMPLOYMENT ADJUSTMENT AND THE PERSISTENCE
OF DISEQUILIB1UUM
The rate of change of employment depends upon two factors: the current
stock disequilibrium in employment and the speed of adjustment to clear that
To continue reading
Request your trial