A QUARTERLY MODEL OF THE LABOUR MARKET IN INTERWAR BRITAIN
| Author | T. J. Hatton |
| DOI | http://doi.org/10.1111/j.1468-0084.1988.mp50001001.x |
| Published date | 01 February 1988 |
| Date | 01 February 1988 |
OXFORD BULLETIN
of
ECONOMICS and STATISTICS
Volume 50 February 1988 No. 1
OXFORD BULLETIN OF ECONOMICS AND STATISTICS, 50, 1(1988)
0305-9049 $3.00
A QUARTERLY MODEL OF THE LABOUR
MARKET IN INTERWAR BRITAIN
T J. Hatton
I.
Until a few years ago the conventional system of belief about the working of
the interwar labour market was largely informed by the influential theories of
Keynes and the detailed empirical treatment of Beveridge. As the latter put it:
'The central problem of unemployment in Britain between the wars was not
what it had appeared to be before the first World War. It was not a problem of
cyclical fluctuation reducing demand for a time, or of disorganization of the
labour market wasting men's lives in drifting and waiting. It was a problem of
general and persistent weakness of demand for labour' (1944, p. 89). More
recently alternative explanations have been suggested which place the
emphasis firmly on the supply side. The implication is that it was the supply
of labour, rather than demand which contracted giving rise to mass un-
employment. This view has been put forward most forcibly by Benjamin and
Kochin who argued that 'the army of the unemployed standing watch at the
publication of the General Theory was largely a volunteer army' (1979, p.
474). This, it was argued, arose because of the effects on individual incentives
of the unemployment insurance system which provided generous rates of
benefit and was liberally administered. Hence the British economy was, on
average, fully employed even though measured rates of unemployment
exceeded 10 per cent for most of the time.
These may be regarded as extreme positions, and on the whole too sim-
plistic to adequately represent the majority of views which have been
expressed by economic historians on the subject. However, it is probably fair
to say that these two views represent the alternative approaches which have
been taken in recent empirical work. Perhaps a third view, which has been
favoured by many economic historians is that to a large extent unemployment
1
2BULLETIN
in interwar Britain was 'structural'. In his book, Aldcroft argues that as much
as half of all unemployment in the 1930's can be classified as structural
(1984, p. 29). This is ascribed chiefly to the decline of the traditional nine-
teenth century staple industries. Also in this vein, Booth and Glynn have
argued that 'high regional unemployment differentials arose because of the
geographical concentration of these ailing industries, the location and estab-
lishment of growth industries elsewhere and the immobility of the industrial
workforce'(1975, p.618).
In this paper a model of the labour market is estimated which attempts to
improve on the existing empirical literature in three ways. First, instead of
relying on 18 or 19 annual observations, a previously unexploited quarterly
data set is used which provides a richer basis on which to test models of
labour supply and demand. Second, the specification of models used in previ-
ous applied work have been seriously deficient and we derive a model which
is more faithful to both theory and the stylized facts of the interwar labour
market. Third, the model provides scope for testing alternative interpreta-
tions within this general framework. The remainder of the paper is set out as
follows. The next section examines the patterns exhibited by the quarterly
data and then summarizes some criticisms of previous empirical approaches
to the labour market.' The following three sections in turn specify and
estimate models for employment, the labour force and wage change respect-
ively. In the concluding section the implications of these results, taken
together are analysed.
II.
Despite the recent interest in empirical work on the interwar British labour
market, most studies have eschewed the use of quarterly data. This is a major
drawback since one can employ at most 19 observations on an annual basis.
Once one gets beyond very simple models with two or three variables, the
number of degrees of freedom becomes rather small and hypothesis tests are
increasingly open to doubt. This is all the more surprising since a set of
quarterly data on a consistent basis was published by the Ministry of Labour
and provides a readily available source upon which to build an appropriate
data base. The data begin only in 1924, and while it might be possible to
extend the series backwards on a consistent basis, this is not necessarily desir-
able. As a result of postwar adjustments, particularly the sharp reduction in
hours, the early 1920's saw labour market behaviour which was quite differ-
ent from the rest of the period (see Broadberry, 1986, p. 470). If the focus is
on the era of high unemployment and wage stability, then starting in 1924 is
not inappropriate.
The key variables are graphed in Figure 1. The upper panel plots the
insured labour force and employment for Great Britain (i.e. excluding
'The discussion of labour market structure and empirical work relating to it is expanded in
full in a separate survey paper (Hatton, 1986) and only the broad outhne is given here.
To continue reading
Request your trial