zyxwvu
Scottish Journal of Political Economy Vol. XXI, No. 3, November 1974
zyxwv
zyxwvu
REGIONAL POLICY A N D THE
SCOTTISH ECONOMY*
B A R R YMOORE
AND
J O H NR H O D E S
I
INTRODUCTION
In our earlier analysis (Moore and Rhodes, 1973a) we estimated the effects
of regional policy in the 1960’s on the level of employment in the Development Areas as a whole. If our estimate of the overall regional policy effect
of 200,000 to 250,000jobs were broadly correct it ought to be possible to
identify the effects of regional policy in individual Development Areas.
In this paper we carry out the same analysis for Scotland alone in order to
see if this is so and also because the performance of the Scottish economy
and the impact of regional policy on Scotland are matters of considerable
importance in their own right. We extend the methodology of the earlier
paper with a view to establishing the passive policy ‘alternative position’
with more conviction. We also show that most of the employment gain
resulting from regional policy in Scotland arose from the increased inflow
of immigrant firms rather than from additional employment in long established indigenous firms. This analysis of immigrant firms is also important
in that it provides further confirmation that the estimate of the regional
policy effect is of the right order of magnitude. The results of this work on
Scotland are encouragingly consistent with, and greatly reinforce our results
for the Development Areas as a whole.
We go on to show the contribution of regional policy in the context of
what would have been required to ‘solve’ the problem of the overall imbalance between the growth in labour demand and labour supply which
has been the root cause of Scotland’s economic problem for so long. We
conclude that the achievements of regional policy fall a long way short of
what a full solution to the problem would have required in the decade
1960-70.
We next explain why it is that an effective regional policy has apparently
made so little impact on the Scottish/U.K. unemployment differential. The
reason is that the registered unemployment differential is only one manifestation of the overall disequilibrium (shortfall in demand for labour) in
* We gratefully acknowledge the help of Mr. W. A. H. Godley, Director of the
Department of Applied Economics, Cambridge, Mr. R. S. Howard of the Department
of Trade and Industry, and Dr. R. G. L. McCrone of the Scottish Office in providing
most useful comments on an earlier draft of this paper. This paper is part of the
D.A.E. research programme of the Effects of Economic Policy which is financed by
the Social Science Research Council.
215
z
216
BARRY MOORE AND JOHN RHODES
the Scottish labour market and that when regional policy works to increase
the demand for labour in Scotland it is matched by increased labour supply
from sources other than the registered unemployed (i.e. from the previously
‘non-active’population of working age groups and from people who would
otherwise have migrated to other areas).
We conclude with a more speculative comment on prospects for the
Scottish economy in the light of our analysis of the post-war period and the
likely employment associated with the development of North Sea oil.
zyxwvut
zyx
I1
THEEFFECTIVENESS
OF REGIONAL
POLICY
IN SCOTLAND
1) The Size of the Policy E#ect
In our earlier paper (Moore and Rhodes, 1973a) we estimated that between
1963 and 1970 regional policy had created between 200,000 and 250,000
jobs in the Development Areas of the United Kingdom. The basic technique
used in that analysis was to calculate an expected employment series for
the manufacturing sector in Development Areas on the assumption that
each industry in these areas experienced the same annual growth rate as
its national counterpart. Comparison of this series with the actual employment series gave a U.K. /Development Area differential growth effect which
allowed for the main differencesin industrial structure between the Development Areas and the U.K. By comparing the actual employment series with
the expected series we were able to establish a passive policy ‘alternative
position’ which indicated what would have been expected of employment
in manufacturing in the absence of the more active regional policies of the
1960’s. In this paper we first present the results of applying this same technique to Scotland. Other evidence in then presented to support our estimates
of the effects of regional policy in Scotland. This evidence is based on an
analysis of the post-war employment changes in indigenous firms 0.e.
those established in Scotland before 1945) and of the opening of new
establishmentsin Scotland after 1945.
In Fig. 1 (a) we show actual and expected employment in manufacturing industry in Scotland for the period 1950-1971.’ For Scotland alone
actual employment falls relative to expected employment in the period of
passive regional policy up to 1960. (This contrasts with the position for
the Development Areas as a whole where actual and expected employment
were found to move closely together.) This means that for Scotland the
passive alternative position is not so conveniently established as that for
the Development Areas as a whole.2 However, in Fig. 1 (b) we show that
it is still possible to establish a convincing passive policy alternative
position on the basis of the downward trend in the series of actual minus
zyxwv
zyxwvutsrqpon
zy
Shipbuilding and metal manufacturing are excluded and treated separately.
For details of the strength and type of policies pursued in post-war years see
Moore and Rhodes (1973a).
1
2
zyxwvut
zyx
REGIONAL POLICY A N D T H E S C O T T I S H E C O N O M Y
217
expected employment in the passive policy period. The downward movement in this series, established by fitting a time trend in the period of passive
policy between 1952 and 1960, indicates that even after making allowance
for broad differences in industrial structure the growth of manufacturing
employment in Scotland was significantly lower than in the U.K. as a whole.
This trend is projected into the active policy period of the 1960's and
provides a guide as to what would have happened to the series in the
absence of the stronger regional policy of the 1960's. At first sight this
may appear to be a dangerous thing to do but we explain below that there
are convincing reasons for accepting the continuation of t h i s trend as the
alternativeposition.
zyx
zyxwvut
FIG.1 (a). Actual and expected employment in manufacturing industries
(excluding Shipbuilding and Metal Manufacture).
In Fig. 1 (c) the regional policy effect is then estimated as the difference
between the trend values and the actual values of this series of actual
minus expected (A - E ) employment.
On this basis, the strengthening of regional policy in the 1960's created
about 60,000 additional manufacturing jobs in Scotland by the end of the
1960's. After making allowance for the effects of regional policy on the
steel and shipbuilding industries, and for the indirect effects on the nonmanufacturing sector, the overall effect of strengthening regional policy
in the 1960's was somewhere between 70,000 and 80,000 jobs by 1971.
The estimated regional policy effect, as measured in this way, varies
with the overall pressure of demand in the U.K. economy. This is apparent
from the cyclical movements in the A-E series, which coincide with the
U.K. cycle. Ideally, therefore, the size of the regional policy effect should
218
zyxwvu
zyxwv
zy
B A R R Y MOORE A N D JOHN RHODES
be estimated at a constant pressure of demand. To measure the effect as at
1971 underestimates the true position by possibly as much as 10,000 jobs
because the overall pressure of demand in that year was substantially lower
than that prevailing in the passive policy period on which 'the alternative
position' is based. The year 1970 would be a preferable terminal year in
that the overall pressure of demand was not too far different from that
prevailing in the 1950's.
Actual minus expected
zy
%
The
.'passive' policy alternative
*. position
50L,;5,'
'
'
55
'
'
'
'
'
zyxw
60
'
'
'
'
65
'
'
'
'
'
RG.1 (b). Actual minus expected employment in manufacturing industries.
-1oL
FIG.1 (c). T h e regional policy effect (derived from Fig. 1 (b).
'
'
71
zy
zyxwvut
zyx
219
REGIONAL POLICY AND THE SCOTTISH ECONOMY
zy
2) Evidence to Support the Identification of the Size of the Regional
Policy Eflect
To support our contention that the trend fitted to the A-E series (Fig. 1 (b))
satisfactorily identifies the passive policy alternative position we distinguish
between employment change in the indigenous sector (defined as firms
established in Scotland prior to 1945) and employment change in firms
newly established in Scotland after 1945. What we aim to show is that
because the greater part of the observed regional policy effect can be
attributed to an acceleration in the inflow of immigrant iirms into Scotland
zyxw
zyxwvut
90r
P
0
gc
zyxwvutsrqpo
zyxwvutsrqp
50
-
40
-
30-
:
20-
10-
-10
-20
-30
-40
-50
-
1960
/ A .
1
'
'
I
I
1971
-
-
-
....
*. *.
FIG.2 (a). The difference between actual and expected employment in the indigenous
manufacturing sector (excluding Shipbuilding and Metal Manufacturing).
it follows that employment in indigenous firms continued to fall relative to
the U.K. even in the active policy period and that while policy had a
beneficial effect on employment in this sector the downward movement
in employment continued. We. conclude, therefore, that in the absence of
a strengthening of regional policy actual employment would have continued
to fall relative to expected employment (i.e. the A-E series would have
continued to decline in the 1960's).
We were able to show this by using hitherto unpublished data on
employment in immigrant manufacturing firms obtained from the Department of Trade and Industry in Scotland. The data consisted of the annual
cumulative employment of surviving immigrant manufacturing establishments from 1945-1972 together with an industry breakdown at the S.I.C.
order number level of employment in these establishments for the year
zyxwvu
zyx
zyxwv
zy
zyx
220
BARRY M O O R E AND JOHN RHODES
1963. This enabled us to construct an actual and structurally adjusted
expected employment series for the indigenous sector alone.s The difference
between the actual and expected indigenous series is plotted in Fig. 2 (a).
The continued downward movement in the indigenous A-E series in the
active policy period is c~nfirmed.~
Nevertheless when a passive policy
Employment
(000s)
60
FIG.2 (b). The difference between actual and expected employment in the immigrant
manufacturing sector (excluding Shipbuilding and Metal Manufacture).
3 We were able to obtain actual annual employment in indigenous establishments
for the period 1946-1972by subtracting annual employment in immigrant establishments from total manufacturing employment each year. Similarly the industrial
breakdown of indigenous employment as at 1963 was calculated by subtracting
immigrant employment by industry as at 1963 from total employment by industry
in that year. We then calculated an expected employment series for this indigenous
sector, using 1963 as the base year, showing what would have happened to total
indigenous employment had each indigenous industry experienced the same annual
growth rate as its national counterpart. This method may have a small element of
bias in that the expected series is based on national industry growth rates which
include the opening of new firms. The effect of this is to make the A-E series fall
slightly more rapidly throughout its length than it would had the standardisation
been based on the national ‘indigenous’ sector. However, this element of bias does
not preclude a satisfactory identification of a regional policy effect because active
and passive policy periods are affected alike.
4 The underlying causes of this decline ar6 not yet fully understood. One possible
explanation is that the decline in shipbuilding output has affected the growth of a
large number of firms in other industries in West Central Scotland where, historically,
the growth of manufacturing industry was heavily dependent on shipbuilding
(McCrone, 1973). A second possible reason is that, after our structural adjustment at
the S.I.C. Order Number level, important structural differences may remain in the
case of Scotland at the Minimum List Heading level (MacKay, 1968; Randall, 1973).
However, whilst this observation may be perfectly valid, it is common to both the
active and passive policy periods and is therefore unlikely to invalidate our comparison of Scotland’s relative employment performance in the active and passive
policy periods.
REGIONAL POLICY AND THE SCOTTISH ECONOMY
zy
221
alternative position is established, by fitting a linear trend in the period
1952 to 1963, there is evidence of a positive regional policy effect in this
sector which is entirely consistent in timing with the introduction of regional
policy designed to stimulate employment in indigenous firms, i.e. the
differentialinvestment incentives introduced in 1963 were available, for the
first time, to all firms in Development Areas rather than being restricted
to new immigrant firms.
In Fig. 2 (b) we show the difference between actual and expected
employment in the immigrant manufacturing sector. This was obtained by
subtracting the indigenous A-E series from the total A-E series. The alternative passive policy position is again established by fitting a linear trend
between 1952 and 1960. The period chosen for establishing the alternative
position differs from that in the indigenous analysis because policies
designed to influence the location of new factories (e.g. the I.D.C. policy
and the Local Employment Acts) were strengthened before policies
designed to influence the indigenous sector were introduced.
The upward movement in the immigrant series reflects a natural rate
of new factory opening in Scotland which could be expected in the absence
of an active regional policy. The upward movement in this non-policy
period is also determined by the subsequent build up of employment in
factories opened in previous years after the deduction of any closures. It
should be noted that the A-E series must be higher in 1963 than in 1945
because employment in this sector is, by definition, zero at the beginning
of the series in 1945. However, this does not preclude the identification of
a regional policy effect which can be estimated by comparing the rate of
growth of employment in new factories in the active policy period relative
to the period of passive policy.
However, before this can be done we need to be sure that the acceleration in the A-E immigrant series after 1961 is not caused by the natural
employment build up of firms established prior to the strengthening of
regional policy but rather by a marked increase in the opening of new
establishments which coincides closely with the strengthening of policy. The
number of new openings established in each year from 1945 to 1972 is
shown in Fig. 3.
The most remarkable aspect of this series is the very close association
of increased openings in Scotland with a more active regional economic
policy. In the periods of substantially strengthened regional policy immediately after the war and in the 1960's the number of openings in Scotland
increased sharply relative to openings in the period of passive regional
policy in the 1950's. In the immediate post-war period the two most
important instruments of Government regional policy were the factory
building programme (including the conversion of Government munitions
factories) and restrictions on industrial expansion around London and
Birmingham.'
The period of the 1950's was one of a low rate of openings in Scotland.
This coincides closely with a phase of passive regional policy in which few
zy
z
zyxwv
zyxw
See footnote 5 on page 222.
15
'
zyxwvuts
zyxwv
zyx
zyxw
zyxwvutsrqpon
222
BARRY MOORE A N D JOHN RHODES
Government factories were built and the I.D.C. policy was very much in
abeyance.
The sharp upturn in the number of openings at the beginning of the
1960's coincides with an acceleration of the Government factory building
programme which had been relatively dormant throughout the 1950's and
with a more vigorous use of the I.D.C. policy.8 Industrial openings in
Scotland continued to increase after 1963 following the very substantial
strengthening of regional policy that occurred in that year and which has
continued throughout the 1960's.'
50
r
1945
50
55
60
65
70
1972
FIG. 3. Number of establishments opened in Scotland in each year 1945-1972.
Source: Department of Trade and Industry, Scotland.
The patterns of industrial openings in Scotland as shown in Fig. 3 are
strongly confirmed by data on inter-regional movement of manufacturing
industry collected by the Department of Trade and Industry over the period
Government expenditure on factory building at constant prices (1963) between
1946 and 1950 was about double that of the 1950's. The control of industrial building
was exercised initially by means of Building Licences and later under the Town and
Country Planning Act (1947) by requiring firms to obtain an Industrial Development
Certificate for new building covering a floor space exceeding 5,000 square feet. The
evidence that is available on the operation of this policy suggests that the policy was
pursued with considerable stringency. In the Greater London Area the effect of the
restrictions was to limit new industrial building to 5.3 per cent. of the U.K. total
between 1945 and 1948, a share much below that of its working population of 22
per cent. (Board of Trade, 1948.)
6 Employment associated with refusals of 1.D.C.s as a ratio of employment
associated with refusals plus approvals in the South East and Midlands increased
sharply from an average of 4.9 per cent. for the period 1953 to 1957 to just under
14 per cent. in 1958 and 1959 and to an annual average of about 25 per cent. for
the period 1962 to 1968.
7 For details of the variations in the strength and type of policy pursued in this
period (Moore and Rhodes, op. cit., 1973a).
zy
zyxwv
REGIONAL POLICY A N D THE SCOTTISH ECONOMY
223
1945 to 1971.8 As Table I shows, this data confirms the view that the
amount of industrial movement, not only to Scotland but also to the other
Development Areas, is very much higher in the periods of active regional
policy than in the period of passive regional policy and as we show below
it also confirms our identification of a regional policy effect. Moreover
industrial movement to non-Development Areas such as the Midlands and
the South East was not responsive to policy strength, supporting the view
that the unique Development Area pattern of movement was a direct consequence of changes in regional policy.
zyxwvut
Table I
NUMBER
OF MOVES
TO SELECTED h G I O N S IN PERIODS OF ACTIVE
AND
PASSIVE REGIONAL
POLICY. 1945 TO 1971
I
Number of Moves
Active
Passive
Active
Active
Movement to
1965-1971
1945-1951
1952-1959
1960-1965
\
Scotland
Development Areas
(excluding Scotland)
-
77
50
133
140
368
149
323
486
71
Midlands
43
53
68
South East
21
48
35
I
zy
z
15
The upturn in industrial movement to Scotland after 1960 which we argue
was brought about by a strengthening of regional policy supports the view
that the acceleration in employment in immigrant firms in the 1960’s was
mainly caused by the introduction of new establishments after 1960 rather
than by the further expansions of immigrant firms established before 1960.
The consistency of the evidence so far presented and the evidence
presented below on employment associated with new openings convinces
us that the overall regional policy effect as measured in Fig. 1 is broadly
correct. The evidence of the continuous relative decline of employment in
indigenous industry in combination with the evidence on the opening of
new establishments suggests that the passive policy alternative position has
been satisfactorily identified and that regional policy has worked primarily
(though not exclusively) to encourage new industrial movement into
Scotland.
Evidence derived from a number of company surveys (Cameron and
,
8 A first analysis of this data was published in 1968 by R. S. Howard. For the
period from 1945 to 1965 the data refers to moves surviving as at 1966 whereas for
the final period 1966-1971 the data refers to moves surviving as at 1971. It is thus
not comparable with the series shown in Fig. 3 which shows openings in Scotland
gross of subsequent closures and also unlike the data in Table I covers openings
where the ikm or enterprise already had another establishment in the same industry
which had already qualified as a ‘move’. A ‘move’ is defined as the opening in a
new location of a new manufacturing establishment (employing more than 10 persons)
which in some sense could be said to have originated in another area.
224
zyxwvuts
zyxwvu
zyx
zy
BARRY M O O R E AND JOHN R H O D E S
Reid, 1966, and House of Commons, 1973), which suggests the importance
of labour availability in the location decision of b s , have led some to the
view that the variations in the number of moves to Scotland and the other
Development Areas are largely a consequence of this factor. We would not
disagree that the availability of labour (at a price which the firm is willing
to pay and of the requisite skill) is a most relevant factor in the choice of
location. It is of most importance in periods when the labour market is
tightening and is therefore a factor influencing cyclical variations in movement. It is difficult to accept, however, that labour availability was
responsible for the change of trend in movement which took place in the
1960's. If anything labour was more easily available outside Development
Areas in the 1960's than in the 19503, particularly from 1967 onwards, so
that there was less pressure on firms to consider relocation in order to
take advantage of the unemployed labour resources of the Development
Areas.
I11
THESIZEOF THE REGIONAL
POLICY
EFFECT
IN THE IMMIGRANT
AND INDIGENOUS
SECTORS
In this section we aim to disentangle the effects of policy on employment
in Scottish immigrant and indigenous firms. This is useful for two main
reasons. Firstly, a separate analysis should allow confirmation of the total
policy effect which we estimated in section I1 and thereby provide further
support for our identification of the passive policy alternative position.
Secondly, it is desirable to find out to what extent regional policy has encouraged the growth of employment in the indigenous sector as opposed to
bringing in new activity from outside Scotland. For example, in formulating
regional policy it is necessary to bear in mind that there may be a limit
to the volume of potentially mobile firms and the number which Scotland
can reasonably hope to attract. On the other hand, if it is very expensive
in Exchequer terms to revitalise the indigenous sector this may set further
limits on the potential achievementsof regional policy.
(1) The Immigrant Sector
We adopt two alternative approaches to estimating the number of new
jobs generated in immigrant firms as a direct consequence of regional policy.
The first approach makes use of the actual minus expected employment
series for the immigrant sector presented in Fig. 2 (b). This shows that an
acceleration of this series occurs soon after 1960, and coincides with the
transition from the period of passive regional policy in the 1950's to the
much stronger regional policy of the 1960's. This acceleration as confirmed
by the Howard data is mainly the result of an increase in the number of
new firms opening in Scotland after 1960. The effect of the strengthening
of regional policy is estimated by fitting a linear trend to the immigrant
zyxwvut
zy
REGIONAL POLICY AND THE SCOTTISH ECONOMY
225
series in the passive policy period 1952 to 1960. This trend projected forward
gives a guide to the alternative position in the absence of the stronger policy
of the 1960’s. In 1971 the difference between the expected and the actual
series suggests a regional policy effect of between 36,000 and 38,000.
This estimate is broadly conl3rmed by results of the second analytical
approach which makes use of the Howard movement data.QThe results of
this second approach are shown in Table 11. The volume of employment
associated with movement between 1952 and 1959 (the period of passive
policy) is used as a guide as to what would have happened in later years
had regional policy remained passive. This amounted to about 2,400 jobs
per annum and is adopted as our passive policy ‘alternative position’ for
subsequent periods of active policy. The actual volume of employment
associated with moves in the later period substantially exceeded that in
the passive policy period. For moves which took place between 1960-65,
employment generated by 1971 amounted to 39,000 jobs, an annual average
of 6,500 jobs (see row 2 of Table 11). These particular moves had survived
for an average period of eight years and were thus of a similar age to those
moves in the passive policy period where employment was measured at
1966. The policy contribution is then estimated as the difference between
actual employment generated and that indicated by the passive policy
alternative position. This amounted to 4,100 per annum.
Table I1
EMPLOYMENT
ASSOCIATED
WITH THE ESTABLISHMENT
OF NEWFACTORIES
IN
PERIODS OF ACTIVE AND PASSIVE REGIONAL
POLICY
Jobs per annum
Employment at 1966 in moves taking
place between 1952 and 1959
Employment at 1971 in moves taking
place between 1960 and 1965
Employment at 1971 in moves taking
place between 1966 and 1971
Total policy effect 1960-1971
Policy contribution
per annum
2,400
0
6,500
4,100
2,900
1,800
35,000
The alternative position is reduced to 1,100 jobs because the average age of thesc
new openings was only 3 years.
For moves taking place between 1966 and 1971 the associated employ.
ment is only available at 1971 and thus openings taking place in this period
have had considerably less time to mature (in terms of employment growth:
or close down. The alternative position based on ten year old moves i5
therefore not appropriate without further adjustment. This adjustment ie
made by using the ratio of employment in policy induced moves to employ
9 For this method we used unpublished data on employment as at 1966 associated
with the annual surviving ‘openings’ in Scotland over the period 1945 to 1960 and
employment as at 1971 for new factory openings which survived over the period
iann tn 1971
zy
226
zyxwvu
zyxwvu
zyx
zyxwvu
BARRY MOORE AND JOHN RHODES
ment in all moves as at 1971 for moves occumng in the period 1960-1965.
The policy contribution is therefore estimated at 10,800 jobs, an annual
average of 1,800jobs.
Thus for the whole period 1960-1971we estimate that using this method
employment generated in immigrant firms as a direct consequence of the
strengthening of regional policy was about 35,000 jobs. This estimate will
be increased to the extent that this employment generation had indirect
consequences on employment in other sectors of the Scottish economy.
Our two approaches, based on alternative sources of data, give encouragingly consistent results. The regional policy effect in the immigrant sector
is estimated between 34,000 and 38,000 jobs as at 1971 (excluding any
indirect effects).
(2) The Indigenous Sector
An analysis of employment in indigenous manufacturing firms shows
that although the downward trend continues throughout the period of more
active regional policy there is a significant upward shift in the trend after
1963. The broken line in Fig. 2 (a) shows the trend based on observations over the period 1953 to 1963 and projected forward to 1971. The
upward shift in the series after 1963 is significant in that it takes place at
the same time as Government financial assistance to the indigenous sector
was very substantially increased.lO Prior to this the financial inducements
available to the indigenous manufacturing sector were relatively insignificant and the other main instruments of regional policy, namely I.D.C.'s and
Government factory building, were aimed primarily at encouraging the
opening of new establishments in the Development Areas which would
otherwise have been located elsewhere.
Insofar as the relative improvement in this indigenous series in the
period 1963 to 1971 is largely the result of the effects of differential regional
financial assistance then the effect of those policies can be estimated as the
difference between the linear trend [our measure of the alternative position
derived mainly from the experience of the 1950's and early 196O'sl and
the actual series. This would suggest a policy effect on the indigenous sector
of a minimum of 12,000 as at 1971, before any indirect effects are taken
into account and before an adjustment is made for the very low pressure
of demand prevailing in 1971.11
zyxw
10 Regionally differentiated investment incentives for manufacturing industries
were first introduced in British Development Areas in 1963.
11 This estimate is biased downwards as the effects of variations in the pressure
of demand in the economy as a whole significantly influence the cyclical movement
of this indigenous employment series. Ideally in comparing the impact of regional
policy over time the pressure of demand should be approximately equal as there is
evidence to show that on the cyclical downturn the areas of high unemployment
suffer disproportionately compared with the economy as a whole [even after allowing
for differences in industrial structure between regions]. When the direct regional
policy effect in the indigenous sector is measured at a constant pressure of demand
the figure rises to 20,000 jobs or more.
zyx
zyxw
REGIONAL POLICY AND THE SCOTTISH ECONOMY
227
It is conceivable, although we think unlikely, that what we have identified as the regional policy effect on the indigenous sector could have been
partially caused by non-policy factors. We have already cast doubt on the
idea that labour availability is a significant cause, because labour was more'
easily available in non-Development Areas in the active policy period.
Another possible cause is that the shipbuilding industry (which could be
expected to influence the growth of other manufacturing firms which supplied it) declined less rapidly after 1960 than in the 1950's). If this was in
fact the cause, however, we would expect its influence to be more gradual
than is indicated in our analysis, and in any case the better employment
performance of the shipbuilding industry in the latter period may itself be
partially due to the strengthening of regional policy.la One further factor
which could be relevant is the multiplier effect generated by firms newly
established in Scotland which would stimulate activity in the indigenous
sector. However, as we have attempted to show the increase in the number
of newly established firms in the 1960's (and therefore any consequential
indirect effects) was also largely a consequence of regional policy.
A Summary of the Eflects ofRegional Policy
Table 3 summarises our estimates of the impact of regional policy OD
employment in the Scottish economy over the period 1960-1971. The direct
effects of policy have been adjusted to take account of multiplier effects
outside the manufacturing sector (using a multiplier of 1.2) and for the small
effect on metal manufacturing and shipbuilding.
Table I11
THEEMPLOYMENT
EFFECTS
OF ~ G I O N A LECONOMIC
POLICYON THE
zyxwvuts
zyxwvuts
INDIGENOUS AND IMMIGRANT SECTORS OF SCOTLAND 1960-1971
Employment generated by
regional policy 1960-1971
I
Direct effects
Total including
indirect effects
Immigrant Sector
34-38,OOO
4145,000
Total
46-53,OOO
59-68,000
Sector
The separate estimates for shipbuilding and metal manufacturing industries are
included in this figure.
After allowing for indirect effects the total amount of employment generated by regional policy between 1960 and 1971 is shown as 59-68,000 jobs.
However, if we make allowance for the low pressure of demand in 1971
12 It is also likely, however, that the reconstruction of the shipbuilding industry
following the Geddes Report also had an influence on the performance of this industry
in the 1960's.
228
zyxwvut
zyxwvu
zyx
zy
zyxw
zy
BARRY MOORE A N D JOHN R H O D B S
the final estimate is increased to 70,000-78,000 jobs. This overall effect
appears at first sight to be encouragingly large but has fallen far short of
what would have been required to fully correct the unemployment disparity
between Scotland and the more prosperous regions of the U.K. In the next
section we make some broad estimates of what would have been required
of regional policy if it were to have ‘solved’ the problems of the Scottish
economy in the 1960’s.
The Regional Policy Eflect in Scotland in relation to what would have been
required in rhe 1960’s to solve the Problem
The answer to the question ‘What would have been required of regional
policy to correct the economic imbalance which has persisted throughout
the 1960’s?’ is somewhat arbitrary. The concept of economic imbalance is
multi-dimensional embracing differencesin a number of economic indicators
such as output per head, average earnings, unemployment rates, activity
rates, migration flows, etc. Furthermore, it may not be possible or desirable
for the Government to aim to solve regional problems within the space of a
few years, particularly if regional policy continues to rely heavily on the
amount of mobile industry it can direct to Development Areas.
We have chosen to estimate what would have been required to provide
a solution in terms of the number of jobs it would be necessary to create
to achieve the followingobjectives :
(a) the equalisation of unemployment rates between Scotland and the
South East and Midlands,
(b) the equalisation of activity rates between Scotland and the South
East and Midlands, and
(c) the elimination of net-outward migration.
If these three conditions were met other features of economic disparity,
such as average earnings, might well be reduced to acceptable levels.
It is estimated that to have achieved these objectives over the period
1960 to 1971 would have required the creation of at least 230,000 jobs (or
approximately 21,000 jobs per annum for eleven yearsY3 in addition to
the 7,000 jobs per annum which the active regional policy of the 1960’s
is estimated to have contributed.
There is no presumption here that it was possible or even desirable to
attempt such a full solution to Scotland‘s economic problems within a
l3 The major part of this total arises from the diminution of net-outward migration
from Scotland which amounted to 360,000 persons over the period 1960/61 to 1970/71.
If it is assumed that one-third of this migrant population was made u p of employees
then to achieve the migration objective alone would have required the creation of
120,000 jobs. The equalisation of unemployment rates would have required a further
74,000 jobs with the remainder [36,000] coming from the equalisation of activity rates.
To the extent that the generation of new jobs by regional policy accelerates job
losses in Scottish firms, more than 230,000 jobs would have been required to meet
the three criteria.
REGIONAL POLICY AND THE SCOTTISH ECONOMY
zy
229
decade. The figures here serve only to illustrate that even with an active
regional policy as carried out in the 1960's there was still a very substantial
shortfall in the demand for labour relative to its supply. It is likely that
regional policy would have had to have been three or four times more
effective if unemployment differentials were to have been completely eliminated in the 1960's. In the next section we discuss in more detail how this
shortfall in the demand for labour has been matched by adjustments on
the labour supply side and why it is that a reasonably successful regional
policy so far had only a small effect on the unemployment differential.
IV
EMPLOYMENT,
UNEMPLOYMENT AND LABOUR
SUPPLY
In this section we undertake further analysis to show why a regional policy
which has generated some 70,000 to 80,000 jobs in Scotland has had only
a small effect on the relatively high level of unemployment in Scotland.
To do this we examine relative movements in total employment, registered
unemployment, the population of working age and the population of
working age who are not working or seeking work (i.e. non-active). These
are presented in Fig. 4 in such a way as to show how far changes in Scotland
have differed from changes in the U . K . between 1951 and 1971.
The employment and unemployment series (Fig. 4 (a) and 4 (b)) are
constructed by first calculating standardised series which tells us what
would have happened to employment and unemployment in Scotland had
the year to year movements starting with the base year of 1951 been proportionately the same as in Great Britain. These standardised series are
then subtracted from the actual series to give employment and unemployment in Scotland relative to the movement that would have occurred had
they followed the national movement^.'^
If we compare the behaviour of relative movements in employment and
unemployment two features stand out. Firstly, the decline in employment
in Scotland relative to that in the U . K . was very substantial and over the
whole period represented a relative loss of Scottish employment opportunities of 200,000 jobs. This decline was particularly marked between 1952
and 1964 but thereafter there was a noticeable improvement and employment in Scotland grew at about the national rate.
The second outstanding feature is the remarkable stability of relative
unemployment over much of the period. This stability is particularly surprising in view of the sharp decline in employment relative to the U . K .
How can we reconcile the behaviour of these two series? The fall in
l4 As an illustration of this analysis and what it shows, consider the employment
series in Fig. 4. Whenever this series is falling it implies an increasing cumulative loss
of employment opportunities in Scotland relative to the U.K. Thus between 1951
and 1964 the shortfall in employment opportunities had accumulated to about 200,000
jobs. Thereafter the series ceases to decline implying a change in employment similar
to that in the U.K., with no further losses or gains in employment compared with
the U.K. position.
z
230
zyxwv
zyx
zyxwv
BARRY MOORE AND JOHN RHODES
employment arises primarily on the demand side from a relative decline
in employment opportunities available rather than as a result of a shortfall
in labour supply [although the demand for labour will be influenced to
some extent by population changes rising for example from inward and
outward migration]. It follows therefore that given the stability of relative
unemployment that there must have been a relative decline in the supply
of labour in Scotland.
000s
Employment
-250
Unempbyment
zyxwvu
zy
231
REGIONAL POLICY A N D THE SCOTTISH ECONOMY
ooos
'T
Population of working age groups
0 0 0 s 100
Non-Active Population in working age groups
-150
zyx
zyxwvutsrqponm
i
,
,
1951
"
.
,
.
1955
-
.
.
'
.
1960
(4
"
'
.
zyxw
'
1965
.
.
'
.
I
1971
FIG.4. Employment, Unemployment, Labour Supply : Cumulative difference
in Scotland relative to the U.K. 1951-1971.
Changes in the labour supply can occur from two main sources, firstly
from changes in the number of persons in the working age groups and
secondly from changes in the number of persons in the working age groups
who are not registered as seeking employment (i.e. the non-active). This
232
zyx
zyxwv
B A R R Y MOORE AND JOHN RHODES
non-active part of the working population is derived by deducting employment and unemployment from the total working population defined as all
males between the ages of 15 and 65 and all females between the ages of
15 and 60. Fig. 4 (c) shows the difference between the actual population of
working age groups in Scotland and what would have happened to it had
it changed in line with the U.K. from the base year of 1951. The population
of working age groups has fallen relative to the U.K. by about 300,000 over
the period 1951 to 1971. This is largely the consequence of net outward
migration from Scotland.
Fig 4 (d) shows the relative Scottish/U.K. position with respect to the
non-active part of the working population. It shows that up to 1960 the
non-active part of the Scottish working population was growing relative
to that in the U.K. but thereafter there was a continuous relative fall as
increasing numbers enter the active labour force.
Three distinct periods can be distinguished from Fig. 4. From 1952
to 1958 the relative loss in employment opportunities was largely matched
by a relative increase in the number of non-active and hence relative
unemployment was broadly unchanged. In the next period, 1958 to 1966,
although relative employment continued to decline the upward trend in the
relative number of non-active was reversed and began to decline. However,
unemployment in Scotland relative to the U.K. remained stable (apart from
a cyclical movement) because of the acceleration in the decline of Scotland’s
population of working age groups relative to that of the U.K. which was
sufficient to offset both the loss in employment opportunities and the decline
in the relative number of non-active persons. In the final period from 1966
to 1971 when relative employment had ceased to decline (largely as a consequence of the effect of regional economic policy) the relative decline in
the working population began to slow down and for the first time there is
a marked improvement in relative unemployment.
This brief analysis highlights two important points. First if an effective
regional policy was to significantly reduce unemployment in Scotland
relative to the U.K. sufficient employment opportunities would have to
be generated to compensate firstly for any reduction in the very high rate
of net outward migration and secondly for any reductions in the relative
numbers of non-active persons. Secondly, an encouragingly successful
regional policy, which we have estimated to have contributed 70-80,000
additional jobs, is not necessarily reflected in a corresponding reduction in
unemployment and this is why movements in relative unemployment are a
poor guide as to the effectivenessof regional policy.
CONCLUSION
This paper on the impact of regional policy on the Scottish economy has
confirmed the results and extended the analytical approach adopted in our
earlier work on evaluating the effects of regional policy in the Development
Areas as a whole. Further development of this approach is being undertaken,
zyxwvut
zy
REGIONAL POLICY AND THE SCOTTISH ECONOMY
233
notably on the analysis of time-lags associated with the establishment of
new plants in Development Areas and on the problem of assessing the
impact of variations in the overall pressure of demand in measuring the
regional policy effect.
But of perhaps more significance, this analysis has highlighted the
progress of the Scottish economy and how much remains to be done before
Scotland ceases to be a Development Area. The main conclusions are :
zy
zyxw
(1) The strengthening of regional policy in the 1960's had generated some
70,000 to 80,000 new jobs in Scotland by 1971, rather less than a third
of what would have been required to produce a rough balance in the
labour market.
(2) Nearly three-quarters of this employment came from the establishment
of new factories, largely by firms operating from other regions.
(3) The effectof regional policy on the great mass of Scottish manufacturing
industry which we have termed the indigenous sector is relatively small
in terms of additional employment and this large sector continued to
decline, relative to the same industries nationally, even in the active
regional policy period and even when the shipbuilding and metal manufacturing industries are excluded.
(4) The employment gain derived from regional policy had affected relative
unemployment levels in Scotland to only a small extent. Instead, the
relative number of non-active persons declined and net outward migration was lower than it would otherwise have been. It appears that in the
case of Scotland these two latter aspects of serious imbalance in the
labour market have to be at least partially solved by regional policy
before a very substantial impact on unemployment rates will be realised.
Whilst comments on the future prospects for the Scottish economy must
inevitably be more speculative than the analysis of the past some clear
pointers have emerged, Firstly, whilst the achievements of regional policy
have been substantial they have not been sufficiently large as to contemplate
the abolition or even a substantial weakening of regional policy within the
next five to ten years. Further employment opportunities will almost
certainly be lost, relative to the English position. These will arise as a result
of Scotland's industrial structure in which coal mining, shipbuilding and
particularly metal manufacturing are likely to register further job losses
in the next decade. It will also arise from the continued relative decline of
other indigenous manufacturing industries, not only because of a ' poor '
industrial structure, but also an inability to compete adequately in national
and international markets. Scotland needs an extra 10,000 to 20,000 additional jobs per annum on top of the 6,000 to 7,000 new jobs per annum
which regional policy of the 1960's has helped create, if we are to see an
end to Scotland's economic problems by 1980.
The development of North Sea oil will be a very helpful new factor in
providing employment opportunities. But in relation to the size of Scotland's
234
zyxwvuts
zyx
BARRY MOORE A N D JOHN RHODES
zy
zyxw
zyxw
zyxwvut
problem the employment opportunities generated by North Sea oil may
well be limited to 2,000-5,000 new jobs each year over the decade (i.e.
20,000-50,000 in t0tal).l5 As in the case of regional policy, part of this
employment gain will reduce the numbers of non-active persons and net
outward migration rather than have a large direct effect on the unemployment rate.
Even with this beneficial effect on employment of North Sea oil the
question of devising a more powerful and effective regional policy for
West Central Scotland should be urgently considered.16 As we have
explained elsewhere, the Central Government should not be over-cautious
about the Exchequer costs of regional policy. In conditions where new
resources are brought to use to match the effects of regional policy expenditures on aggregate demand the resource cost of regional policy is genuinely
zero." Where the need is so great as in West Central Scotland and where
further policy action is in t h i s real sense very cheap or even costless, it ii
surely preferableto have more of it rather than less.
Department ofApplied Economics, Cambridge
l 5 We are assuming here that Scotland will not become wholly independent from
the U.K.and therefore will not be able to solve her economic problems by exporting
huge quantities of oil to the rest of the U.K.and abroad.
16 At the heart of the Scottish problem is the Special Development Area covering
most of West Central Scotland. For a more detailed analysis of t h i s sub-region see
3. M. Randall (1973) and the Report of the West Central Scotland Planning Group
(April 1974).
l7 B. C. Moore and J. Rhodes (1973b). Over the period 1960 to 1971 the total
Exchequer cost amounted to about f325m. which, on the basis of our estimates of
the jobs created by regional policy, gives an Exchequer cost per job generated of
about f5,000.
Table IV
EXCHEQUER
COST OF REGIONAL
ASSISTANCE
TO INDUSTRY
ANNUAL
AVERAGES
SMN
1960/61to 1964/65
1965/66to 1967/68
1968/69to 1970/71
I
2
14
8
11
1
21
30
SCOTLAND-
zyxw
1 1
Recoverable or
partially recover- Investment Grants,
able items such as Free Depreciation
loans and Govern- and operational
Grants
ment factories
Year
IN
Labour
subsidies
Total
excluding
recoverable
-
2
5
26
48
78
zyx
zyxwvutsr
REGIONAL POLICY AND THE SCOTTISH ECONOMY
235
REFERENCES
BOARD
OF TRADE
(1948). The Distribution of Industry, Cmnd. 7540, HMSO 1948.
CAMERON,
G. C. and REID,G. L. (1966). Scottish Economic Planning and the Attraction of Industry, University of Glasgow, Social and Economic Studies, Occasional Papers Number 6, 1966.
REPORTO F THE WORK OF THE WEST CENTRAL SCOTLAND PLANNING GROUP(April 1974).
HOUSEOF COMMONS
85-1 (1974). Evidence presented to the Trade and Industry
Sub-committee by the Department of Trade and Industry on the results of
their Inquiry into Location Attitudes and Experience.
HOWARD,R. S. (1968). The Movement of Manufacturing Industry in the UK 1945
to 1965, Board of Trade, HMSO, 1968.
MACKAY,
D. I. (1968). Industrial Structure and Regional Economic Growth, Scottish
Journal of Political Economy, June 1968.
MCCRONE,R. G. L. (1973). The Localisation of Economic Activity in the United
Kingdom, Urban Studies, 1973.
MOORE,B. C. and RHODES,J. (1973a). Evaluating the Effects of British Regional
Economic Policy, Economic Journal, March, 1973.
MOORE,B. C. and RHODES,J. (1973b). ‘The Economic and Exchequer Implications of
Regional Policy’, Memorandum to the Expenditure Committee (Trade and
Industry Sub-committee), House of Commons Paper 42XV1, 1972/73.
RANDALL,
J. M. (1973). Shift/Share Analysis as a Guide to the Employment Performance of West Central Scotland, Scottish Journal of Political Economy, February
1973.
zyx