STRUCTURAL UNEMPLOYMENT: A REPLY
| Date | 01 February 1988 |
| Author | Bertrand Kan,Richard Jackman |
| Published date | 01 February 1988 |
| DOI | http://doi.org/10.1111/j.1468-0084.1988.mp50001006.x |
OXFORD BULLETIN OF ECONOMICS AND STATISTICS, 50, 1(1988)
0 305-9049 $3.00
S'IRUCTURAL UNEMPLOYMENT: A REPLY
Richard Jackman and Bertrand Kan
Adrian Wood's very careful paper raises a number of serious questions, but
as he offers no empirical work in support of his arguments it is hard to gauge
their possible quantitative importance. Nonetheless the conclusion of my
paper with Stephen Roper that there is no evidence of any increase in struc-
tural imbalance in the UK economy is becoming increasingly difficult to
accept. In this reply we suggest that the unemployment-vacancy mismatch,
while a good index of structural imbalance over our data period, has in recent
years started to behave in a way which is both out of line with its own past
behaviour and inconsistent with the theoretical model of the Jackman/Roper
paper.For reasons both of space and data availability we confine ourselves to
regional imbalance during the last ten years. Table 1 compares the coefficient
of variation of regional unemployment and vacancy rates with those of other
possible measures of regional economic conditions. While there has been no
increase in the coefficient of variation of regional unemployment or vacancy
rates towards the end of the period (consistent with our claim of no increase
in regional imbalance), there has been an upward trend since 1982 in the
coefficient of variation of non-manual wages, and a very sharp jump in 1985
and again in 1986 in the coefficient of variation of house prices.
To interpret the behaviour of these variables, it is instructive to look at the
correlations between them. The top part of Table 2 shows the correlation
across regions between a number of variables and the unemployment rate.
The (inverse) correlation between unemployment and non-manual earnings
and, especially, house prices has become more pronounced over the period.
The positive correlation between manual earnings and unemployment has
decreased. Most surprisingly, perhaps, the inverse relationship between
unemployment and vacancies has diminished and by the end of the period
disappeared. (By 1986, vacancy rates in both Wales and in the Northern
Region of England were higher than in the South East). The idea that the dif-
ferent regions can be represented by different points on a common u/v curve,
which appeared to be valid over the period 1964-84 (Jackman and Roper,
1987, Table 9), would appear difficult to reconcile with this more recent
observation.
The second part of Table 2 shows a sharp decline, particularly since 1984,
in the correlation between vacancies and the other variables. This gives
further support to the view that vacancies are becoming an increasingly
inaccurate measure of relative job opportunities.
83
To continue reading
Request your trial