-
https://www.cpes.org.uk/om/files/original/e4f68443da3693d3a92b9fd3871a171e.pdf
e63ce9e36dbbd80d315bda891ffa7e04
Dublin Core
The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/.
Title
A name given to the resource
Economic Policy Review Volume 3
Text
A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text.
Text
Any textual data included in the document
CHAPTER 6
MANUFACTURING EXPORT SHARES AND COST COMPETITIVENESS OF ADVANCED INDUSTRIAL COUNTRIES
by Martin Fetherston, Barry Moore and John Rhodes
Introduction
In recent months there has been considerable public discussion on the efficacy of devaluation as a policy instrument, capable of rectifying the UK's chronic tendency to trade deficit by assisting UK industries in improving their shares of home and world markets. It is generally agreed that the Marshaii-Lerner conditions hold in the long run-i.e. that an effective devaluation' will eventually leave the balance of trade in a better position than it would otherwise have been for a given level of output.
The main points at issue are: first, given the strongly adverse trends in UK trade, how large an effective devaluation would be required to achieve a satisfactory balance of trade at reasonably full employment; second, whether any once-for-all devaluation could be sufficient in the long run, or whether the devaluation must be progressive or repeated; and third, whether devaluation on the scale and frequency required is at all feasible in terms of its effects on inflation and the distribution of income.
As a contribution to discussion of the first two issues, this study examines manufacturing exports of the six largest trading nations over the last twenty years. It will be argued on the basis of this evidence that the UK has failed to secure enough effective devaluation, up to the present, to prevent a continuation of the long-run decline in the UK 's manufacturing export share and that further continuing real devaluation will be necessary if the decline in export share is to be halted. Even on fast world trade growth assumptions and allowing for the contribution to the balance of payments of North Sea oil. it will be crucial to hold the UK share of manufactured exports at about its present level if there is to be a return to full employment in the 1980s.
The main steps in the analysis are: (a) for the UK and each of five major trading competi-
tors, to examine changes in shares of manufactured exports and changes in cost competitiveness over the past twenty years: (b) using what we believe to be uncontroversial views about the elasticity of exports with respect to relative costs and prices, to examine how far changes in export shares were determined by changes in relative costs and how far there remained unexplained and persistent residual trend changes in these shares:
1An effective devaluation is here defined as a decline in the exchange rate in excess of relative increases in unit labour costs.
62
(c) then to consider on the same basis what scale of changes in cost competitiveness (relative to those which actually occurred) would have been necessary to offset the unexplained trend changes in each country's export share and thus hold each country's export share constant;
(d) on optimistic and pessimistic assumptions about the growth in world trade in manufactured goods up to 1985, and assuming relative cost competitiveness fixed at average 1976 levels. to predict export shares (and therefor;: export volumes) for each country and confront them with past rates of growth in exports;
(e) (for the UK only) to consider what would need to happen to net exports cf manufactures and relative costs for reasonably full employment and a satisfactory balance of trade to be achieved in the 1980s.
Changes in export shares and relative costs The UK share of the volume of world2manufacturing exports fell by over 50% in the two decades from 1956 to 1976 (see Table 6.1 ). This is part of a longer-term decline whereby the export share fell from about 33 ~;.; in 1899 to 8∑3% in 1976. Whatever the effects of more flexible exchange rates since 1967, they have not been sufficient to halt this decline in the UK share. In the two decades from 1956 to 1976 four countries, West Germany, France, Italy and Japan, increased their combined share of world trade by almost 20 percentage points from 27∑2 to 46∑8%- half of this increase accruing to Japan. Smaller industrial countries taken together lifted their share of world manufactured exports by almost 3 percentage points in the same period. On the other hand the share of the USA fell by over 40.% (13 percentage points) and the UK share fell by over half (9 percentage points).
Indices of relative unit labour costs for the six main industrial exporting countries are shown in Table 6.2. For each country, indices of the level of unit labour costs in manufacturing industry in dollars were calculated using OECD indices of manufacturing earnings in dollars and dividing these by indices of trends in productivity. Indices of relative unit labour costs were then constructed by dividing each country's index of dollar unit costs by an index of the weighted average of the unit costs of the other five countries- the weights in each case being determined by the export shares of each country at the base date of 1970.
2Jn this article the coverage of <:iata on world exports is restricted to developed market economies (with one or two exceptions equivalent to the group of OECD member countries).
Manufacturing export shares and cost competitiveness ofadvanced industrial countries
Table 6∑1
Percentage shares of the volume of manufactured exports from industrial countries (selected years, ! 956-76)
USA
UK West Germany France Italy Japan
1956 30∑0 17∑8 14∑4
1960 23∑1 15∑3 17∑8
1965 20∑2 12∑2 17∑6
1970 17∑4
9∑7 18∑1
1975 18∑7
8∑6 !7∑1
1976 17∑2
8∑3 17∑8
*Includes Canada, Belgium, Luxemburg, Netherlands, Swit-
zerland, Sweden. Source: Monthly Bulletin of Statistics, United Nations.
6∑0 8∑4 7∑7 7∑9 8∑4 8∑4
2∑6 4∑2 4∑2 5∑4 6∑0 8∑7 6∑5 10∑7 6∑6 !2∑5 6∑5 14∑1
Smaller* Industrial Countries
25∑0 25∑8 27∑5 29∑7 28∑0 27∑8
Table 6∑2 Indices of relative unit labour costs for six large exporting countries (selected years, 1956-76)
UK
1956 1960 1965 1970 1975 1976
96∑4 105∑5 105∑2 97∑0 97∑0 90∑4
Source: OECD Main Economic Indicators.
USA
123∑0 128∑4 105∑0 97∑9 63∑0 68∑1
West Germany
67∑9 78∑6 92∑1 98∑6 111∑5 Ill∑!
France
150∑0 118∑7 113∑4 101∑4 120∑0 l 18∑4
Italy
95∑0 88∑8 101∑5 98∑8 112∑9 103∑3
JGJ.pan
102∑9 89∑7 90∑2 107∑9 136∑4
~40∑2
The most striking points to emerge from this calculation are the large and persistent rise in measured relative costs in Germany, and the enormous decline in measured relative costs in the USA since I960. Movements in the other four countries fall between these two extremes. Relative costs in Japan were falling between I956 and I965, but increased sharply thereafter. Relative costs in the UK deteriorated gradually up to I967 and subsequently showed only a moderate tendency to improve, in spite of a substantial nominal devaluation of sterling; most of the sterling depreciation merely compensated for the persistently lower growth of productivity and higher inflation in the UK than in other countries.
Now turning back to the export shares shown in Table 6. I, we find trends in startling contrast to those in relative costs. In West Germany there has been a large and persistent increase in relative costs, yet her share of the volume of exports, after rising at the beginning of the period, remained constant between 1960 and I976. Thus a 40 ;;; rise in relative costs over I6 years was associated with West Germany's export share increasing at the expense of other industrial countries. In contrast the USA was unable to prevent her export share falling by over 40%, in spite of an almost 50% reduction in measured relative costs.
Explanation of changes in export shares in terms of cyclical effects, changes in relative cost competitiveness, and residual trends To identify more precisely how much of the movem~nt of export shares can be explained by changes in relative costs and to measure the magnitude of any persistent residual trend changes, a statistical model was postulated in which export shares are a function of the volume of world ~de itself (to isolate cyclical
fluctuations), appropriately lagged indices of relative costs, linear trends, and random disturbance terms.
If each country maintained its share of world exports through the trade cycle, there would obviously be no need to adjust export shares for cyclical effects. But since it is likely in practice that the response of individual countries' exports to fluctuations in trade differs, it is necessary to make some allowance for this as a first step to isolating the underlying trends.
A large number of statistical experiments were carried out to ascertain the response of export shares in the six large industrial countries to cyclical variations in world trade. These experiments comprised specifications, some in which 'the elasticities with regard to relative costs were freely estimated, others in which they were imposed; a variety of time periods and Jag structures were employed in each case. The values for cyclical elasticities finally selected were: UK -0∑2, USA +O∑I, France-0∑2, Italy -0∑45, West Germany 0, Japan +0∑15, the 'Rest' +O∑J.I These values are small enough, except for ltaly, to mean that cyclical factors are a small component of changes in export
shares. In this study effective devaluation is measured in
terms of relative unit labour costs in a common currency, because this takes account not only of changes in exchange rates but also of differences in the rates of increase in money earnings and productivity growth. When it comes to measuring elasticities most other work has been conducted in terms of changes in relative prices rather than costs. Relative export prices tend to move less than relative costs for two reasons. Firstly, some part of a fall in relative unit labour costs may be reflected in increased export profits rather
'The weighted sum of these elasticities, using 1970 export shares as weights, is zero-see Appendix to this chapter.
63
Economic Policy Review
Table 6∑3 Long-run price elasticities of demand for exports estimated in previous studies
Country
Number of estimates
Range of estimates
'Best' point estimate
USA (manufactured exports) Japan (manufactures) France (total exports) West Germany (total exports) UK (manufactures) Italy (total exports)
10 --0∑56 to ~ 2∑62 12 -0∑35 to -4∑14 6 -- 1∑06 to - 2∑27 6 ~0∑65 to --1∑88
3 -- 1∑00 to ~ 3∑00 7 -- 0∑03 to 1∑96
Source: Price Elasttclfles in International Trade, by R. M. Stern, J. Francis and B. Schumacher, for the Trade Policy Research Centre, Macmillan, 1976.
--1∑24 -1∑77 -1∑31 -1∑11 -2∑00 --0∑93
than in lower export prices.1 But even if there were no increase in export profits, relative prices still move less than relative unit labour costs, because of the effect of changes in the cost of imports on export prices. Price elasticities are therefore likely to be higher than cost elasticities.
The first step in establishing a plausible range of values for relative cost elasticities was to examine the r.!sults of other studies of export price elasticities in the six large countries. It was fortunate that we could draw on a recent summary of earlier studies commissioned by the Trade Policy Research Centre, on which Table 6.3 is based.
Altogether evidence from 44 studies is available for the countries with which we are concerned. The range of elasticities shown for each country is wide, but the authors give 'best' estimates which average - 1∑4, varying from -0∑93 for Italy to -2∑00 tor the UK. For the reasons given above, relative prices probably only change about half as much as relative costs. It is therefore to be expected that relative cost elasticities ~ would be roughly half the magnitude of price elasticities, and, on the evidence of earlier studies, would vary between -0∑5 and - 1∑00.
The next stage was to estimate relative cost elasticities from our own annual data for each of the six countries, whilst imposing preferred values for cyclical elasticities. Estimates of relative cost elasticities were made separately for six countries, over three distinct time periods (1956-76, 1956-73 and 1960-76) and using a short and long geometric lag formulation. With the exception of Germany the estimated elasticities were all of the'right sign, clustering at the higher end of the expected range around - 1∑0, and statistically significant. The sensitivity to the time period selected was small, but the longer lag structure resulted in rather higher relative cost elasticities (see Table 6∑4). The preferred estimate of relative cost elasticities for all c.:>untries is - 1∑0. But because of some variation between countries, and particularly in view of the statistically poor results for Germany, we carried out the further analyses discussed below using three alternative values for cost elasticities: --0∑5, -- 1∑0 and 1∑5.
Cyclical and relative cost elasticities were also freely estimated simultaneously. On the whole the results
lAt times relative cost indices and relative export price indices diverge sharply. This implies subsiamial changes in profit margins. A case in point is that of Japan in 1975/6 when, although relative unit labour costs increased, relative export prices fell. However, these changes in profit margins are unlikely to persist for very long.
were more variable. Nevertheless the results were distinctly encouraging, in that there was conclusive evidence of significant and strong elasticity effects cJnsistent with the results of other studies shown in Table 6.3. Of the 18 freely estimated relative cost coefficients shown in the Appendix Table 6A I, 13 were near to -1∑0.
Table 6∑5 shows the manufactured export shares of the six large trading naticns, progressively adjusted fvr cyclical variations in world trade about trend and f.:>r variations in lagged relative cost changes relative to the base year 1970.
It is clear from the final columns of the table that there are considerable unexplained trends remaining in the export shares of several countries, after making adjustments for variations in world trade and for changes iri relative cost competitiveness. These unexplained residual trends are summarised in Table 6∑6 for the preferred relative cost elasticity of -1∑0. The residual trend shares are not highly sensitive to the assumed value of the relative cost elasticity within the plausible range, - 0∑5 to 1∑5.
Table 6∑4 Estimated relative unit labour cost elasticities for six industrial countries - imposed trade elasticities
Shorter lag
Longer Ia
1956-76 1956-76 1956-73 1960-76
UK USA France Italy W. Germany Japan
-0∑71 (4∑5) -0∑80 (8∑3) --0∑64 (3∑2) -0∑58 ( 1∑5) -+0∑23 (0∑8) -0∑73 (7∑4)
-1∑10 (6∑5) --0∑98 (9∑0) -1∑10 (3∑1) -- 1∑3 ( 1∑6) --0∑2 (0∑4) -0∑83 (7∑1)
-1∑10 (5∑0) -0∑86 (4∑4)
-1∑70 (4∑2)
~2∑3
(2∑3) -0∑44
(0∑8) --0∑98
(6∑1)
-1∑37 (5∑0) --1∑00 (6∑8) -0∑44 ( 1∑0) -l∑l (1∑3) -0∑3 ( 1∑4) -1∑00 (5∑5)
Notes: F1gures 111 parentheses refer tot ratiOs.
The shorter lag involves a decline in the weight of earlier years at the rate of-60~;, per year; the longer lag involves a 30 ~'~decline per year.
Most striking amongst these ~::~idual trends are the consistent upward movements of the adjusted export share of West Germany and Japan at the expense of the UK and the USA. These are extremely persistent
64
Manufacturing export shares and cost competitiveness of advanced industrial countries
Table 6∑5 Manufactured export shares of six large countries adjusted for variations in world trade and relative costs
Actual share
Share alter adjusting
Adju~tment for variatfor variation IOn m in world trade world
trade
Adjustment for relative costs
Elasticity
-0∑5 -1∑0 -1∑5
Export share after all adjustments
Elasticity
-0∑5 -1∑0 -1∑5
UK
1956 17∑82
+0∑40
18∑22
0∑59 1∑38 2∑34 17∑63 16∑84 15∑88
1960 15∑32
+0∑03
15∑35
0∑02 0∑20 0∑55 15∑33 15∑15 14∑80
1965 12∑24
-0∑10
12∑14 -0∑26 -0∑48 -0∑65 12∑40 12∑62 12∑79
1970 9∑65
+0∑08
9∑73 -0∑11 -0∑22 -0∑32 9∑84 9∑95 10∑05
1975 8∑63
-0∑09
8∑54 0∑05 0∑15 0∑30 8∑49 8∑39 8∑24
1976 8∑27
--0∑07
8∑20 0∑17 0∑40 0∑68 8∑03 7∑80 7∑52
USA
1956 30∑02 1960 23∑13 1965 20∑23 1970 17∑35 1975 18∑70 1976 17∑17
-0∑32 -0∑03 +0∑09 -0∑06 +0∑10 +0∑08
29∑70 23∑10 20∑32 17∑29 18∑80 17∑25
-2∑61 -2∑68 -1∑28 -0∑55
2∑37 2∑48
-4∑99 -5∑35 -2∑55 -!∑II
4∑52 4∑69
-6∑10 -7∑96 -3∑82 -1∑66
6∑46 6∑65
32∑31 25∑78 21-60 17∑84 16∑43 14∑77
34∑69 28∑45 22∑87 18∑40 14∑28 12∑56
36∑80 31∑06 24∑14 18∑95 12∑34 10∑60
West Germany 1956 14∑39
0
14∑39
2∑80 5∑18 7∑16 11∑59 9∑21 7∑23
1960 17∑80
0
17∑80
2∑76 5∑24 7∑42 15∑04 12∑56 10∑38
1965 17∑58
0
17∑58
1∑18 2∑34 3∑47 16∑40 15∑24 14∑11
1970 18∑12
0
18∑12
0∑61 1∑20 1∑80 17∑51 16∑92 16∑32
1975 17∑10 0 17∑10 -0∑81 -1∑54 -2∑19 17∑91 18∑64 19∑29
1976 17∑77 0 17∑77 ~0∑79 -1∑47 -2∑04 18∑56 19∑24 19∑81
France
1956 6∑03 1960 8∑39 1965 7∑75 1970 7∑95 1975 8∑44 1976 8∑42
+0∑14 +0∑01 -0∑06
+0∑06 -0∑08 --0∑08
6∑17 -1∑11 -2∑31 -3∑60 7∑28 8∑40 -0∑76 -1∑47 -2-12 9∑16 7∑69 -0∑50 -1∑00 -1∑50 8∑19 8∑01 -0∑36 -0∑72 -1∑10 8∑37 8∑36 --0∑38 -0∑71 -1∑01 8∑74 8∑34 -0∑40 -0∑76 -1∑07 8∑74
8∑48 9∑77 9∑87 10∑52 8∑69 9∑19 8∑73 9∑11 9∑07 9∑37 9∑10 9∑41
rta1y
1956 2∑58
+0∑13
2∑71 0∑13 0∑28 0∑46 2∑58 2∑43 2∑25
1960 4∑18
+0∑02
4∑20 0∑23 0∑49 0∑77 3∑97 3∑71 3∑43
1965 6∑00
-0∑11
5∑89 0∑07 0∑17 0∑29 5∑82 5∑72 5∑60
1970 6∑55
:o∑tl
6∑66 0∑07 0∑15 0∑24 6∑59 6∑51 6∑42
1975 6∑61
-0∑15
6∑46 -0∑12 -0∑19 -0∑23 6∑58 6∑65 6∑69
1976 6∑50
- 0∑13
6∑37 -0∑08 -0∑11 -0∑10 6∑45 6∑48 6∑47
Japan
1956 4∑19 1960 5∑37 1965 8∑70 1970 10∑71 1975 12∑51 1976 14∑06
-0∑07
0 -1-0∑05
-0∑06 -'-0∑10 _! 0∑09
4∑12 5∑37 8∑75 10∑65 12∑61 14∑15
and Sizeable residual trends in export shares after allowing for the effects of significant and sometimes
large changes in relative costs. These figures imply that progressive effective devaluation of sterling on a large s~ale would have been necessary to preserve the UK
share of world trade, if it is assumed, for the moment,
that the trends themselves are not influenced by changes in relative costs.
In the case of Germany it is worth emphasising that,
under the assumptions of the analysis, her share of world trade would have risen by 10 percentage points
between 1956 and 1976 had she not been willing to let relative∑ costs in dollar terms rise by 64%. This large
rise in relative costs prevented Germany's actual
share of world manufactured exports from growing
0∑21 0∑44 0∑78 0∑36 -1∑11
0∑47 0∑90 1∑52 0∑71 -2∑22
0∑74 1∑36 2∑21 1∑06 -3∑34
3∑91 4∑93 7∑97 10∑29 13∑72
3∑65 4∑47 7∑23 9∑94 14∑83
3∑38 4∑01 6∑54 9∑59 15∑95
---1∑37 - 2∑73 --4∑10 15∑52 16∑88 18∑25
more rapidly. Japan's relative costs on the other hand
did not rise fast enough to prevent her export share
rising by a large amount and it is this aggressive Japan-
ese stance more than anything else which has been the
cause of the rapid deterioration in UK and US export shares.
Whatever the causal factors lying behind these
strong residual trends, it is clear that the amount of
effective devaluation required at any one point in time
must take account of them and that repeated effective
devaluation will be needed as long as they persist.
This would only cease to be the case if the residual t~ends were eventually reversed by successive devalua-
tions and/or by some other kind of intervention, such
as industrial policy or protection.
65
Economic Policy Review
Table 6∑6 Changes in manufactured export shares for which there is no adequate explanation in terms of relative costs and world trade (Relative cost elasticity assumed equal to - 1∑0)
UK
USA W. Germany France
Italy
Japan
1956-1960 1960-1965 1965-1970 1970-1975 1970-1976
-1∑69 (-IQ-0) -2∑53 ( -16∑7) -2∑67 (-29∑1) -1∑56 ( -15∑7) -2∑15 ( -21∑6)
-6∑24 ( -18∑0) -5∑58 (-19∑6) -4∑47 ( -19∑5) -4∑12 (-22∑4) -5∑84 ( -31∑7)
+3∑35 ( +36∑3) +2∑68 ( +21∑3) +1∑68 ( +11∑0) +1∑72 ( + 10∑2) +2∑32 ( + 13∑7)
+1∑39 ( + 16∑4) -1∑18 ( -12∑0)
+0∑04 ( +0∑5) +0∑34 ( +3∑9) +0∑37 ( +4∑2)
+1∑28 ( +52∑7) +2∑01 ( +54∑2) +0∑79 ( + 13∑8) +0∑14 ( +2∑2) -0∑17 ( -2∑6)
+0∑82 ( +22∑5) +2∑76 (+61∑7) +2∑71 ( +37∑5) +4∑89 ( +49∑2) +6∑94 ( +69∑8)
Total
1956-76
-9∑04 ( -53∑7)
-22∑13 ( -63∑8)
+10∑03
+0∑62
( + 108∑9) ( + 7∑3)
+4∑05
+13∑23
( + 166∑7) ( +362∑5)
Notes: Figures not in parentheses refer to changes in the share in terms of percentage points per period. Figures in parentheses refer to percentage changes in the share during the period.
Table 6∑7 Changes in relative costs necessary to maintain constant shares through selected periods (per cent)
1960-65 Required Actual
1965-70 Required Actual
UK USA France Italy W. Germany Japan
-22∑2 -23∑5 -9∑1 +57∑9
+9∑5 +64∑0
-2∑6 -12∑5 -1∑6 +10∑0 +10∑9
+1∑3
-25∑5 -14∑8 +0∑2
+4∑0 +3∑2 +37∑6
-5∑5 -0∑7 -2∑4 --4∑9 +0∑1 +11∑8
Note: Jtequired and actual changes in relative costs are both meas_ured as a ratio of end-period lagged relative costs to
initial (unlagged) relative costs.
1970-76 Required Actual
-16∑8 -26∑5 +18∑0
+3∑1 +8∑4 +59∑9
-2∑9 - 25∑7 +11∑4
+3∑8 +IQ-6 +21∑8
The scale of changes in cost competitiveness which would hne been required to maintain cm;s:ant n;anu∑ factoring export shares This section shows the amount of effective devaluation that would have been needed in each country to hold export shares constant through three selected time periods (including the amount necessary to offset cyclical movements). The required effective devaluations are compared with what actually took place.
Table 6.7 shows the required and actual changes in relative costs for each country and period. The results are most revealing. In all three periods the UK failed to secure anything like sufficient effective d~valuation to maintain her export share. Even the large nominal devaluations which took place between 1970 and 1976 were far short of that required t:l reduce relative costs by enough to hold the 1970 export share. The USA, with a similarly declining export share, also failed to halt the decline in the 1960s, but managed a very large effective devaluation in the early 1970s. West Germany was remarkable in the sense that the actual increase in relative costs in each period was very close to that required to maintain her export share constant. Japan, by contrast, held relative costs well below the level required to maintain her export share constant, even though relative costs increased by as much .as 21∑8% in the latter p::riod.
The general picture to emerge, therefore, is of a trading system dominated by strong long-run trends in export shares whose effects were reduced but not
reversed by effective devaluations and revaluations. This result goes some way to explaining the 'Kaldor paradox' that ex post the value cf net exports appeared to respond perversely to effective devaluations. I The reason is that relative cost changes, although moving in the right direction, have not been large or frequent enough to reverse the strong underlying trends in export shares.
Prospects for growth in world trade and UK manufactured exports up to 1980 and beyond In this section the prospects for export growth of each country up to 1980 and beyond are examined on the assumption that cost competitiveness remains at 1976 levels, given an optimistic and a pessimistic assumption about the growth of exports of OECD countries as a whole.
In setting a relatively optimistic and relatively pess!mistic assumption for growth of trade, regard must be had to past relationships between trade and output, as well as to what is thought possible by OECD countries in the future (see Table 6.~\.
Relative costs for each country would remain at their 1976 level if:
(i) productivity growth in each were to be the same,
(ii) the growth in earnings in each were to be the same,
'N. Kaldor, 'Economic crisis: exports and the sinking pound', The Times, 9 November 1976.
66
Manufacturing export shares and cost competitiveness ofadvanced industrial countries
Table 6∑8 Growth in OECD countries 1963-1976
Percentage change in
1963/4-1973/4 Average
1975 1976 1977 forecast*
GNP Industrial production Manufactured exports
5∑0
--1∑2
5∑0
3∑75
5∑7
--- 7∑8
8∑5
5∑0
10∑7
-4∑5
9∑5
6∑5
*Forecast by OECD, December 1976.
Table 6∑9 Expected growth in the volume of manufactured exports 1916-85 if world trade grows by 9,% per annum and 11% per annum respectively and if all countries' relative costs are fixed at 1976 levels, and remain unchanged until 1985
Growth rates in export volumes
(%per year)
Actual
Projected 1976-85
1963-1976 A B
UK USA France Italy W. Germany Japan The Rest Total
5∑8 4∑64 6∑14
7∑5 3∑60 5.64
10∑1 7∑51 9∑05
10∑1
11∑61
12∑71
9∑2
10∑42
12∑41
13∑5
14∑56
16∑93
9∑7 8∑27 10∑41
9∑3 9∑00 11∑00
(A) world manufactured exports assumed to grow at 9% per annum.
(8) world manufactured exports assumed to grow at II /o per
annum.
Actual 1976
8∑3 17∑2 8∑4 6∑5 17∑8 14∑1 27∑8 100
Export shares (%of total)
Projected 1976-85 AB
5∑7 10∑3 7∑4 8∑0 19∑9 22∑0 26∑2 100
5∑5 11∑0 7∑2 7∑5 19∑9 22∑5 26∑5 100
(iii) there was no further movement in exchange rates,
(iv) adverse movements in any one or two of these were matched by compensating favourable movements in the other(s).
If the UK fails to keep unit costs in line with those in other countries, because of faster inflation or lower productivity growth than in other countries, further devaluation of sterling relative to other currencies will be needed for the UK to achieve even the export growth' averaging 5 to 6% per annum projected in the table.
If the UK is to maintain its 1976 share of manufactured exports to 1980 and beyond, a further fall in UK relative costs of the order of 30 ;;, will be required to raise the growth of manufactured exports from the projected figure of 6% per annum to somewhere be-
tween 9% and II '% per annum. This compares with
a 15% reduction in actual relative costs achieved between 1965 and 1976. In the absence of a sig_nificant improvement in the trend of UK inflation relative to other countries, this could only come about by further effective depreciation of the exchange rate.
It would be hoped that such a substantial acceleration in the growth of export volumes, if sustained over a number of years, would itself lead to increases in investment, innovation and productivity, which would reduce unit costs in such a way that repeated devaluations would not be required. At this point what we have termed the decline in the residual trend share would have been halted.
As far as West Germany and Japan are concerned, their high output growth rate has been and will continue to be dependent on rapid export growth to pay for rapidly growing imports. With unemployment at historically high levels in all countries, there will be
great reluctance on the part of countries like Japan
and Germany to allow their currencies to appreciate
and thereby limit their rate of growth of manufactured
exports.
So far as the future growth of output is concerned,
the medium term strategy adopted by the OECD Coun-
cil at Ministerial level in June 1975 suggested that 'an
annual growth rate (in GNP) for member countries
collectively of 5% per annum or somewhat more
during the five years 1976-1980 was feasible'. Events
since then may have reduced the growth prospects in
some member countries, but by no means in all. On
this basis the optimistic growth assumption for indus-
trial production might be in the region of 6% a year
and the pessimistic assumption around the 4% mark,
the latter implying that the OECD countries in general
may not return to full employment by 1980.
Growth assumptions for industrial output can be
transformed into assumptions for manufactured
exports using the following relationship between
OECD industrial production and world manufac-
tured exports ( 1963-1976):
In X= =5∑4+0∑054t+0∑85 In Y
(I 0∑8) (9∑9) (7∑2)
R2 0∑997
where X is the volume of world exports of manufac-
tures
Y is the volume of industrial production in
OECD countries
t=time
figures in parentheses are t ratios.
This suggests growth in manufactured exports of
OECD countries of II% a year on the optimistic GNP
growth assumption, and 9% a year on the pessimistic
assumption.
Table 6.9 shows the growth of exports of each
country which would be expected if relative unit
67
Economic Policy Review
labour costs were held constant for all countries at their 1976 level over the period to 1985, on both assumptions about the growth in world trade.
Impncations for the UK manufacturing sector 1976-81 and beyond The analysis so far has been concerned with manufactured e~ports. It suggests th~t. given quite rapid growth m world trade and with the 1976 level of cost competitiveness held, these may grow by an average of about 6% per annum between 1976 and 1985. The main reason for this low figure is the persistent adverse trend in UK exports which remains after making allowance for changes in relative costs. We believe that these trends will not be easily reversed.
It remains to consider the prospects for the UK manufacturing sector as a whole. For this we use a projection of the main CEPG model, conditional on cost competitiveness being held constant with the basic balance of payments moving into a modest surplus of about £I ,OOOm per annum (at 1975 prices) after including the full expected benefits of North Sea oil. This may not be what actually happens, but it is a sensible set of conditions on which to assess the prospects for UK manufacturing industry.
Table 6.10 sets out the main trends in UK manufacturing industry over the past 15 years and confronts these with the conditional projections for the next decade. The growth in the volume of manufactured exports has been accelerating, although as we have seen, it lags lamentably below those of many trading partners. The acceleration is no doubt due in part to such relative cost reductions as have been achieved in recent years. The acceleration ot export growth is expected to continue in the next two years as the effects ot recent sterling depreciation work themselves through, giving a growth rate approaching 7% per annum between 1976 and 1981. But with cost competitiveness held constant at its 1976 level, the growth of exports after 1981 slows down in response to the adverse residual trends. In the early 1980s the projected growth rate is down to 5∑3 ~~ per annum.
Table 6∑10 The UK manufacturing sector
The rapidly increasing contribution of North Sea oil to the balance-of-payments position will necessitate increased growth of domestic demand, if a large balance-of-payments surplus is to be avoided. One of the important unknowns is how quickly the UK manufacturing sector could respond to a rapid increase in both export and domestic demand. Table 6.10 assumes that the r~sponse would be large and could take place before existing spare production capacity was exhausted. Hence an increase in manufacturing production of 5∑8% per annum ( 1976-81) is shown. Given that cost competitiveness is held at its 19761evel, such unprecedented growth rates in output are associated with projected growth rates for import volumes of 14∑3 ~-;; over the same period. Although the model indicates that the demand conditions necessary for such a radical transformation in UK growth rates will exist, it is in our view likely that supply constraints will prevent such rapid growth from taking place.
The main reasons for this can be seen from the experience of the UK manufacturing sector between 1970 and 1976, as shown in Table 6.1 0. The annual growth rate in manufacturing production averaged only 0∑63 ~~ per annum throughout this period. But this does not mean that there are huge amounts of spare capacity. Manufactured imports, growing at over 9% per annum, were taking a rapidly increasing share of the UK market for manufactured goods. The rate of industt ial closures was high and the rate at which new factories opened declined sharply relative to the 1960s. Throughout the period manufacturing investment in real terms failed to exceed its 1970
level. Six years of such an experience is not the ideal situ-
ation from which to 01 ganise a quick response to a rapid increase in demand of an unprecedented magnitude. Manufacturing investment is expected to rise in 1977 by perhaps 15-18 ~~ in real terms, but increases approaching this magnitude were necessary in 1959/60 and 1964/5 merely to sustain the growth rate of 3% per annum through the cycle. Continuing growth of this magnitude is unlikely to be sustained by the capi-
Annual average growth rates per period in manufacturing
(% per annum)
Production Employment
Productivity
Import Export volumes volumes
Actual
1960-65 1965-70 1970-76
Projected 1976-81 1981-85
3∑23 0∑33 2∑71 -0∑52 0∑63 -2∑21
2∑88 3∑26 2∑92
A8A8
5∑8 +2∑7 +1∑7 3∑0 4∑0
2∑7
-0∑3 -1∑2 3∑0
4∑0
6∑2 9∑4 9∑1
14∑2 9∑7
3∑84 5∑83 6∑04
6∑82 5∑29
Annual average
level of net exports
+2108 +2004 -1-1641
+583 3646
Average ' import share* per
period
8∑0 10∑4 16∑3
21∑2 28∑4
*The import share is defined here as M where DP is
DP~X+M
th~ value o~ gr?ss output of the manufacturing sector at 1970 pnces. The md1cat~r may be biased by changes in the ratio of fimshed to sem1-fimshed manufactured imports relative to the 1970 base.
(A) assumes productivity growth of 3 ~~ per annum.
(B) assumes productivity growth of 4";, per annum.
68
-
Manufacturing export shares and cost competitiveness of advanced industrial countries
tal goods industries. Moreover it should not be for-
gotten that in many capital-intensive manufacturing
Industries it takes 2-3 years for new industrial plant
to be installed and the average time for entirely new
plants to reach full producLion capacity is about seven
years. Given these kinds of supply constraint, and
notwithstanding the possible achievements of the
national industrial strategy, it is difficult to see how
an output growth rate approaching 6% per annum
could be sustained over the next five years, and how
such a growth rate could be associated with only a
14% growth rate in the volume of manufactured
imports,
.
But assuming the projections in Table 6.10 were
feasible from the supply side point of view, the longterm prospects for the UK manufacturing sector remain unfavourable. When the growth in North Sea oil revenues begins to slow down in the early 1980s,, an increasing contribution to the balance of trade will be required from manufacturing, if growth is to be sustained. But by this time net exports of manufactured goods are expected, under the conditions of the projection, to be in a sizeable deficit in excess of £3,000m per annum and deteriorating rapidly. A reversal of this trend would require further effective devaluation or other measures, such as protection, to allow growth to take place and thereby avoid a reduction in living standards as oil revenues decline.
APPENDIX
The form of equation used to analyse and project export shares is
In(!) =ao,+alit+a21 ln(:J:*) +a31(RCL,
(!)
where X, is exports of manufactures of the ith coun-
try, in 1970 dollar terms
W is total OECD exports of manufactures,
i.e. J.:X,= W
W* is the trend value of W, obtained from
fitting a log tr~nd to the period 1956-76.
RCL 1, 1=A In RC11 -!-(I-A)RCL 1, 1.. I where A is the geometric lag parameter on relative
unit labour costs, RC
t is a time trend
The relative unit labour cost indices are constructed
from the formula
In RC1=ln ULC,-1: f3ii ln ULCJ
i4oj
(2)
where ULCJ is the jth country's unit labour costs, expressed in a common currency and f3J 1 is the share ofj's exports in total OECD exports (excluding 'the
rest' and country i) in 1970. (Since no information is
readily available for the rest's unit labour costs, this
composite country is assumed to have constant relative
costs throughout.) The elasticity of the ith country's export share with
+respect to the world cycle is thus a 1,; note that this is
equivalent to an elasticity of (I a1,) for the level of exports. The elasticity of the share with respect to
relative costs is a 31 . The regressions referred to in the text involved either freely estimating all four coefficients in the above equation or imposing an a priori
value of one or other of a 11 and a 31 and estimating the three remaining coefficients. For the chosen values
of a 11 and a 31, the residual shares for the past shown in tables 6.5 and 6.6 are obtained as
~Jexp [In ( -a1 1 In(::*) -a3; (RCL,
J-In RC;, r 70)
The projected trend shares for the future can then be obtained by estimating the coefficients ao; and ali, cJnditional on the assumed values for a 1, and a3;, and using these estimates to project the trend. _Projected
a:::tual shares can then be obtained by adjusting for the assumed future levels of total OECD exports in relation to trend and of relative costs.
The export shares must, for consistency, sum to unity in any particular year, both for actual shares and after adjustment for cyclical factors and/or relative costs. Formally this imposes constraints on the admissible values of the sets of coefficients au, a 2, and a3 1, but with a logarithmic equation of the fm m shown above these constraints can in general only be exactly enforced for one year. Thus it may be shown that in any given year, the condition that ensures that the sum of cyclical adjustments to shares is zero is
_;'(, f
1:1
a?; -
.-V =0
r' t
(3)
and unless the shares never change, or the a 21 are all zero, then this condition cannot be satisfied for every year. The preferred elasticities given in the text do obey this condition for the base year, 1970.
T.he corresponding condition for the relative costs term is more complex, as it must take account of the construction of the relative cost indices: an increase in i's unit labour costs will increase i's relative costs but will reduce j's relative costs to an extent which depends on i's share∑ in the trade of j's competitors. The requirement here is that any change in unit labour costs in a given year should generate changes in relative costs with consequent changes in shares, the sum of which is zero. This condition is in fact a set of multiple conditions, namely
E, a3; f3ii XWu =0 for each country j
t
(4)
where f3J 1 is defined in (2) (/311 =-1)
The extent to which these conditions are not ful-
filled thus depends on the extent to which shares are
unequal and to which they change as compared with
the base year, even if it is assumed (as in the text) that
a31 is the same for each country (excluding the rest). Where any such discrepancies arise due to the non-
fulfilment of these conditions, then the relevant ad-
justed shares have to be scaled in order that they
should sum to unity (generally these further adjust-
ments are fairly small). A similar procedure is followed
when t:1e residual shares are projected by a trend into
the future.
69
Economic Policy Review
Table 6Al World trade and relative unit labour cost elasticities with all parameters freely estimated
UK USA
France Italy W. Germany Japan
1956-76
World
Relative
trade
cost
-0∑28 (2∑7)
+0∑07 (0∑5) -0∑29 ( 1∑4) -0∑56 (2∑0) +0∑01 (0.1) -0∑09 (0∑5)
-1∑14 (6∑0) -0∑99 (8∑9) -1∑0 (3∑0) --1∑59 ( 1∑6) -0∑21 (0∑4) -0∑76 (6∑0)
1956-73
World
Relative
trade
cost
~-0∑18
(l∑t) +0∑06 (0∑3) +1∑25 (4∑1) -0∑27 (0∑9) -0∑9 (6∑5) +0∑34 ( 1∑0)
-~ 1∑02 (2∑8)
~0∑9
(3∑4) -4∑1 (7∑9) -2∑6 (2∑2) --0∑84 (2∑5) ~ 1∑1 (4∑0)
1960-76
World
Relative
trade
cost
-0∑22 ( 1∑7)
~0∑01
(0∑1)
~-~D-21
( 1∑1) --0∑57 ( 1∑7)
-~--0∑18
(1∑7) -0∑5 (2∑3)
∑∑∑1∑47 (5∑0)
-~0∑95
(5∑8) --0∑44 ( 1∑0) - 1∑25 (1∑3) -0∑12 (0∑6) .. 1∑1 (7∑0)
70
Dublin Core
The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/.
Title
A name given to the resource
Chapter 6
Manufacturing Export Shares and Cost Competitiveness of Advanced Industrial Countries
Creator
An entity primarily responsible for making the resource
Martin Fetherston
Barry Moore
John Rhodes
Source
A related resource from which the described resource is derived
Economic Policy Review Volume 3, pages 62 - 70
Date
A point or period of time associated with an event in the lifecycle of the resource
March 1977