Jump to ContentJump to Main Navigation
British Economic Growth 1856-1973The Post-War Period in Historical Perspective$

R. C. O. Matthews, C. H. Feinstein, and J. Odling-Smee

Print publication date: 1982

Print ISBN-13: 9780198284536

Published to Oxford Scholarship Online: November 2003

DOI: 10.1093/0198284535.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 20 June 2018

Output, Inputs, and Productivity by Sector

Output, Inputs, and Productivity by Sector

Chapter:
(p.217) Chapter Eight Output, Inputs, and Productivity by Sector
Source:
British Economic Growth 1856-1973
Author(s):

R. C. O. Matthews

C. H. Feinstein (Contributor Webpage)

J. C. Odling‐Smee

Publisher:
Oxford University Press
DOI:10.1093/0198284535.003.0008

Abstract and Keywords

The increase in output growth came almost entirely from industrial sectors between 1873–1913 and 1924–37, and from non‐industrial sectors between 1924–37 and 1951–73. The U‐shaped pattern in the rate of growth of TFP was general, but its timing and magnitude varied between sectors. The decline between 1856–73 and 1873 –1913 was largest in agriculture, mining, and construction; there was also a fall in manufacturing. An exception to the general rise in TFP growth between twentieth century peacetime periods was the absolute fall in commerce in the interwar period compared with 1873–1913; this reflected concealed unemployment. The shifts in TFI across World War II toward manufacturing, mining, and agriculture, and away from commerce, were accompanied by opposite trends in the rate of growth of TFP, suggesting diminishing returns.

Keywords:   agriculture, commerce, growth by sector, manufacturing, mining, structural change

The sectoral measures contained in this chapter are consistent with output‐side estimates of GDP but not necessarily with expenditure‐side or income‐side estimates. Growth in output is measured by single indicators, not double deflation. Labor input is measured in man‐hours. The capital stock is measured gross. Sectoral distributive shares are used as the weights in calculating TFI. Nine major sectors are studied: agriculture, forestry, fishing; mining, quarrying; manufacturing; contruction; gas, electricity, water; transport, communications; commerce; public and professional services; and dwellings.

The Course of Structural Change

Long‐run changes in the relative importance of the nine major sectors are reviewed before looking at the sectoral growth rates for output, inputs, and productivity. The main findings are in Table 8.1, which shows each sector's share in inputs and output at benchmark years. The largest sectors are manufacturing and commerce, together with agriculture at the beginning of the period and public and professional services at the end. Agriculture's share in inputs fell at a fairly steady proportional rate (except for a rise in 1937–51), but the main absolute fall was over by 1914. Its share in output changed little after 1924, a reflection of its above‐average growth of productivity. Manufacturing's share in labor input rose across World War II but otherwise had no upward trend; it was the only sector not to show a rise matching agriculture's loss before 1914. However, output and capital stock regularly grew faster in manufacturing than in the economy as a whole. The net change in the share of commerce in inputs over the entire period 1856–1973 was small. It rose steeply before 1914 and also rose in the interwar and postwar periods, but it fell steeply across World War II. Productivity in commerce, as measured, grew less than in the economy (p.218) as a whole, and its share in output at constant prices had a downward trend. The proportion of the labor force in public and professional services rose in all periods but especially across World War II (measures of output in this sector are largely conventional). There was some tendency to convergence over time in sectoral capital‐labor ratios.

The Growth of Output, Inputs, and TFP in the Major Sectors of the Economy

Sectoral growth rates are given in Table 8.3. These data lead to six general findings.

  • (1) The increase in the rate of growth of output between 1873–1913 and 1924–37 came almost entirely from industrial sectors, and that between 1924–37 and 1951–73 almost entirely from nonindustrial sectors.

  • (2) The most prominent single change in the rate of growth of TFI was common to all sectors, viz. the fall in the postwar period compared with previous peacetime periods.

  • (3) The U‐shaped pattern in the rate of growth of TFP was general, but the timing and magnitude of the changes varied between sectors. The decline between 1856–73 and 1873–1913 was largest in agriculture, mining, and construction; there was also a fall in manufacturing. TFP grew more rapidly in 1924–37 than in 1873–1913 in every sector with the important exception of commerce (where it fell absolutely, reflecting concealed unemployment). It grew more rapidly in 1951–73 than in 1924–37 in every sector, the improvement being most pronounced in nonindustrial sectors in 1951–64 and in industrial sectors in 1964–73.

  • (4) The period 1937–51 stands apart. The shifts in TFI toward manufacturing, mining, and agriculture, and away from commerce, were accompanied by opposite trends in the rate of growth of TFP, suggesting a tendency to diminishing returns. There was a unique large absolute fall in TFP in construction, a sector where (in contrast to commerce) TFP was apparently adversely affected by the move to a full employment economy.

  • (5) The ranking of the nine sectors with respect to the rate of growth of TFP was in general more uniform between periods than their ranking with respect to the rates of growth of TFI and output. Differences between sectors in the rate of growth of TFP increased after World War I.

  • (6) Differences between periods in sectoral rates of growth of TFI did not closely match those in output. There was some tendency for (p.219) variations between periods in sectoral rates of growth of TFI to be in the opposite direction from those in TFP.

The Growth of Output, Inputs, and TFP in Subsectors of the Economy

Rates of growth of output, inputs, and productivity can also be calculated (usually from 1924 only) for 21 subsectors: 13 in manufacturing, 3 in commerce, and 5 in transport and communications. There were, naturally, expanding and declining subsectors. However, differences between periods in the rate of growth of TFP in the respective sectors were in general pervasive through the constituent subsectors. Thus, the distributive trades, insurance, banking, and finance, and miscellaneous services all experienced absolute declines in TFP in the interwar period. In the distributive trades this happened also, though less markedly, before 1914, when the increase in numbers employed was particularly large. A cross‐industry correlation for the 13 industries in manufacturing is undertaken in Appendix J , with broadly similar results to those obtained by previous writers. The same manufacturing industries (mainly metal‐using ones) tended to have high rates of growth of TFI in the periods 1924–37, 1937–51, and 1951–64, but not in 1964–73.

In Chapter 7 we considered the changes in TFI and TFP in the economy as a whole, and movements in aggregate output were reviewed in Chapter 2. In this chapter we examine the corresponding findings for the main sectors of the economy.

The measurement of output in the individual sectors is conceptually the same as that for the whole economy, but statistically, sectoral output can only be measured from output data. To maintain consistency with these sectoral estimates, the estimates for GDP given in this chapter are also the output series, not the average of output, expenditure, and income series used in Chapters 2 and 7. The change in output in all individual sectors and industries except agriculture is measured by single indicators rather than the conceptually preferable method of double deflation. Some implications of this and other aspects of the procedure are considered more fully in Appendix I.

(p.220) The method of measuring the growth of TFP in individual sectors is the same as that used for the whole economy. The labor input is the number of man‐hours worked per year, unadjusted for changes in the quality of labor. The approximate order of magnitude of such adjustments at the aggregate level was shown in the last section of the previous chapter, but though one may reasonably assume that similar improvements in labor quality occurred in most sectors, sector‐specific information is lacking. The capital stock is measured gross of depreciation (but after allowing for retirements) because data on the net capital stock are not available by industry for the postwar years. For 1856–73 and 1873–1913 there are no data on inventories by sector, and the capital stock estimates for these periods therefore cover only fixed assets. The weights used to combine the growth rates of labor and capital are the shares of labor and capital in the total income of the particular sector in the first year of the period.1

Our calculations of TFI and TFP are made for nine major sectors of the economy: agriculture, forestry, fishing; mining, quarrying; manufacturing; construction; gas, electricity, water; transport, communications; commerce; public and professional services; and dwellings. However, for two of these sectors — public and professional services, and the ownership of dwellings — output and TFP are not very meaningful concepts, and they are included only for completeness.2

Before considering the sectoral growth rates, we look first at the changes over time that they brought about in the relative importance of sectors in the economy.

The Course of Structural Change

Changes in the Relative Importance of the Major Sectors

Table 8.1 shows the changes in the relative importance of the nine major sectors over the years 1856–1973 based on output at constant prices, output at current prices, labor input, and capital input at current prices. Table 8.2 gives supplementary information on shares in the labor force within services, on a more disaggregated basis.

Movements in industries' shares in the total may follow different courses according to the variable under discussion — a point that has often led to discussions at cross purposes.*

In the case of output, changes in relative prices are liable to lead to divergences between the trends in shares at constant and current prices (p.221) (and the divergences would probably be even greater if the same constant price weights were used for all periods). This is most apparent in manufacturing and in public and professional services. Between 1951 and 1973 manufacturing's current‐price share fell and its constant‐price share rose, both by significant amounts. In the interwar period its current‐price share rose slightly, but its constant‐price share rose substantially. In public and professional services the downward trend found in the constant‐price share in both peacetime periods is replaced by an upward trend in the current‐price share.

In general movements in labor‐force shares correspond more closely to those in current‐price output shares than to those in constant‐price output shares. This is to be expected. Since wages are the major element in costs, the value of industries' output at current prices is likely to exhibit similar trends to their labor input. These will not correspond to constant‐price trends if the rate of growth of productivity differs between industries. However, in some sectors there are systematic differences between the levels of the shares in current‐price output and the shares in labor. The extreme example is dwellings, which has no labor force at all, but around a quarter of the entire capital stock. The other most prominent example is utilities, whose share in the labor force has always been much smaller than its share in the capital stock, and hence also smaller than its share in output. In general the sectors that had below‐average or above‐average capital‐labor ratios at the beginning of the period continued to do so 117 years later. However, there was some tendency to converge: most sectors were nearer the mean in 1973 than in 1856.3

Trends in Individual Sectors

Let us now look briefly at the experience of the various sectors in turn to see how their relative importance altered over time.

The series on manufacturing is of special interest and may be considered first. Its course is curious. The proportion of the labor force in manufacturing was more or less constant from 1861 until 1911.4 There was a sharp rise across World War I to 1920 (not shown in Table 8.1), but this was temporary and was not maintained in the face of the ensuing depression.5 During the rest of the interwar period there was little trend, and the level was about the same as in the nineteenth century. Then across World War II and in the early postwar years there was another steep rise. A peak was reached in 1960, and there was then a decline.

What emerges, therefore, is that the rise from 1937 to 1960 should not be regarded as a continuation of a long‐standing trend, nor should (p.222)

Table 8.1 Sectoral Shares of Output, Labor, and Capital in the Whole Economy, 1856–1973

(Percent)

1856–1913

1913–1937

1937–1973

Sector and variable

1856

1873

1913

1913

1924

1937

1937

1951

1964

1973

Agriculture, forestry, fishing:

Output, constant prices

18.4%

13.5%

6.4%

5.3%

4.9%

4.4%

4.2%

4.5%

4.4%

4.4%

Output, current prices

4.9

3.5

3.5

5.5

3.4

3.0

Labor

29.6

21.4

11.5

(9.2)

8.1

5.8

5.3

5.6

4.1

2.9

Capital

19.1

15.4

6.2

(9.2)

8.2

6.1

5.7

5.4

4.4

3.9

Mining, quarrying:

Output, constant prices

4.6

6.1

6.4

6.6

6.1

4.3

5.8

4.3

2.8

1.6

Output, current prices

6.1

3.7

3.7

3.5

2.5

1.5

Labor

3.6

4.1

6.5

(5.8)

6.2

4.0

4.1

3.7

2.6

1.5

Capital

1.6

2.2

2.1

(2.0)

2.2

1.9

1.8

1.3

1.5

1.2

Manufacturing:

Output, constant prices

22.2

24.6

26.6

29.0

30.9

34.8

29.5

34.6

37.0

38.2

Output, current prices

30.9

31.3

29.5

35.7

33.6

30.1

Labor

32.5

33.5

32.1

(32.5)

32.9

32.9

30.4

35.1

36.1

34.7

  Capitala

12.9

14.6

18.5

(20.4)

20.3

19.0

19.5

23.8

25.5

22.8

Construction:

Output, constant prices

3.2

3.8

2.9

2.7

4.2

5.6

7.9

5.6

6.5

6.0

Output, current prices

4.2

5.0

5.1

5.4

7.1

7.8

Labor

4.0

4.8

4.9

(4.9)

4.5

5.9

6.0

6.9

8.5

8.7

  Capitala

(0.8)

(0.8)

(0.6)

(0.6)

0.6

0.6

0.6

0.7

1.0

1.1

Gas, electricity, water:

Output, constant prices

0.3

0.5

1.9

1.6

2.2

3.4

1.3

2.2

2.9

3.6

Output, current prices

2.2

2.7

2.7

2.0

3.2

3.2

Labor

0.1

0.2

0.6

(0.7)

1.1

1.3

1.3

1.6

1.7

1.4

Capital

1.7

2.7

4.8

(4.4)

4.5

7.0

7.8

6.8

8.4

8.6

Transport, communications:

Output, constant prices

6.5

7.6

10.5

9.0

10.3

9.5

7.2

8.5

8.1

8.6

Output, current prices

10.3

9.7

9.6

8.9

8.4

8.9

Labor

4.1

5.6

7.8

(8.2)

8.6

7.7

7.6

8.1

7.5

7.4

Capital

18.5

22.6

23.4

(20.5)

19.5

16.5

15.8

12.6

10.9

9.9

Commerce:

Output, constant prices

23.7

25.5

27.2

33.3

27.7

25.9

30.4

24.2

25.2

26.1

Output, current prices

27.7

29.4

30.4

23.7

25.5

27.0

Labor

20.8

24.7

28.5

(31.1)

29.2

32.6

35.4

23.4

24.5

25.4

Capital

12.4

11.2

9.8

(12.8)

12.7

12.0

11.0

9.7

10.5

11.4

Public and professional services:

Output, constant prices

8.6

7.9

10.2

11.8

12.8

11.4

12.8

15.5

12.9

12.4

Output, current prices

12.8

13.1

13.9

14.8

15.6

18.0

Labor

5.3

5.7

8.1

(7.6)

9.4

9.8

9.9

15.6

15.0

18.0

Capital

6.5

6.2

11.1

(9.7)

9.5

10.0

11.4

10.8

11.2

13.2

Dwellings:

Output, constant prices

12.5

10.5

7.9

4.5

4.8

4.6

4.2

3.9

3.5

3.5

Output, current prices

4.8

5.6

5.6

3.0

4.3

5.8

Capital

26.5

24.3

23.5

(20.4)

22.5

26.9

26.4

28.9

26.6

27.9

Adjustment for financial services:b

Output, constant prices

0.0

0.0

0.0

−3.8

−3.9

−3.9

−3.3

−3.3

−3.3

−4.4

Output, current prices

−3.9

−4.0

−4.0

−2.5

−3.6

−5.3

Note: Figures in parentheses are rough estimates, either to provide a link across breaks in the data, as explained later in this note, or to fill gaps in the data, as explained in note a.

There are major breaks in all series in 1913 and 1937, and so comparisons should not be made between sets of figures in different periods. Two breaks are common to all series: (1) Southern Ireland is included in the 1856–1913 figures, excluded in the 1913–37 and 1937–73 figures; (2) the industrial classification is roughly consistent with the 1948 Standard Industrial Classification in the 1856–1913 and 1913–37 figures, but is based on the 1958 SIC in the 1937–73 figures.

Apart from these, the main breaks are as follows.

Output at constant prices: different price bases were used within periods; 1907 prices were used for 1856–1913, 1924 prices for 1913–37, and 1958 prices for 1937–73. The adjustment for financial services (see note b) was assumed to be zero in 1856–1913.

Labor: based on estimates of man‐years up to 1913 (first figure) and man‐hours from 1924. The second 1913 figure has been roughly linked to the 1924 figure to produce an approximate estimate, shown in parentheses, on a man‐hours' basis.

Capital: based on the estimates of gross fixed capital only through 1920, on gross fixed capital and inventories thereafter, all at current prices. The second 1913 figure has been roughly linked to the 1924 figure to produce an approximate estimate, shown in parentheses, on a total capital basis. There is also a break in 1937 resulting from inconsistencies between the statistical series for earlier and later years.

A number of minor breaks resulting from statistical inconsistencies have been removed in all series by splicing within periods, and thereby concentrated in 1913 and 1937.

(a) Separate estimates for capital in manufacturing and construction in 1856–1920 do not exist. The share of capital in construction in 1913 was therefore assumed to be 0.6%, and the stock was assumed to have grown from 1856 to 1873 and from 1873 to 1913 at the same rate as output of construction; the share of manufacturing was found as a residual.

(b) This adjustment is necessary to exclude the services of the financial sector to other sectors, because statistical deficiencies prevent its being allocated. The output of other sectors is therefore overestimated.

(p.223) (p.224)

Table 8.2 Shares of Service Industries in the Total Labor Input, 1856–1973

(Percent)

Commerce

Public and professional services

Year

Distributive trades

Insurance, banking, and finance

Miscellaneous services

Armed forcesa

Civil public administration

Professional and scientific services

1856

6.3%

0.1%

14.4%

1.9%

1.1%

2.3%

1873

7.8

0.3

16.6

1.7

1.2

2.8

1913

12.2

1.1

15.2

2.0

2.1

4.0

1913

(12.6)

(1.1)

(17.4)

(2.0)

(2.0)

(3.6)

1924

12.4

1.9

14.9

1.8

3.2

4.4

1937

14.3

2.0

16.3

1.7

3.5

4.6

1937

15.2

2.0

18.2

1.7

3.4

4.8

1951

11.2

1.8

10.4

3.5

5.5

6.6

1964

13.2

2.3

9.0

1.6

5.1

8.2

1973

12.4

3.3

9.7

1.4

6.1

10.5

Note: On the breaks in comparability at 1913 and 1937, see Table 8.1.

(a) Includes those stationed abroad and excludes members of the U.S. and other forces stationed in Britain.

the falling‐off after that be regarded as a reversal of such a trend. The rise from 1937 to 1960 was an altogether exceptional phenomenon by historical standards. The data as they stand suggest that since the beginning of the nineteenth century, every substantial increase in the proportion of the labor force in manufacturing in Britain has occurred in or after a period of war (including possibly 1811–21), though we should not wish to press this as a causal relationship.6

The conclusion that in peacetime periods the proportion of the labor force in manufacturing was roughly constant up to World War II calls for one further comment. In the early nineteenth century, and indeed in the later nineteenth century too, a significant proportion of those in manufacturing were domestic workers or workers in very small workshops rather than in factories. From the point of view of supply conditions, though not from the point of view of the structure of demand, it might be more relevant to consider the proportion engaged in “modern manufacturing industry,” i.e. in factories. Taking the year 1841, the earliest with reasonably usable data, we may deduct, from the 2,700,000 engaged in manufacturing in Great Britain, about 600,000 in clothing (exclusively a non‐factory industry, and much the largest one) and 250,000 hand‐loom weavers in textiles. This reduces the number to 1,850,000, 22 percent of the labor force.7 In 1924 the total number of manufacturing out‐workers, most of them in clothing, amounted to only 0.16 percent of the total in employment; thus their (p.225) exclusion would not alter significantly the proportion engaged in manufacturing. So whereas the overall proportion remained about constant between 1841 and 1924, the proportion in factories increased substantially. Moreover, the above adjustment understates the increase, because there were shifts from non‐factory to factory employment in other manufacturing industries besides clothing and textiles.

The trend in manufacturing's share in output at constant prices was different from its share in the labor force: as already noted, its share in output at constant prices had a persistent tendency to rise in peacetime, as well as across wars, despite the absence of a similar trend in its share of the labor force. This reflects partly the tendency for productivity to rise faster in manufacturing than in the economy as a whole, and partly the rise in manufacturing's share in the national capital stock in certain phases, namely in the period before 1914 and across World War II.*

Trends in other major sectors may be more briefly treated. The proportion of the labor force in agriculture plunged down throughout the period. The main decline took place before 1914. After that the proportion declined at an even faster rate, as a percentage of itself, but the proportion itself had become so much smaller that such declines were much less significant in absolute terms than they had been earlier. The fall in agriculture's share in the capital stock followed a broadly similar pattern. In the period 1873–1913 its share of the capital stock actually fell faster, proportionately, than its share in the labor force. The decline in agriculture's place in the economy in this period was not merely the result of normal structural change and the elimination of rural surplus population; it was due also to foreign competition and the consequent agricultural depression. Agriculture's share in output at constant prices changed little after 1924, despite the continuing diminution of its share in factor inputs — a reflection of its good growth in productivity.

The other main declining sector, mining, differed from agriculture in that up to World War I its share of the labor force and output were increasing rapidly; the increase was particularly great in the first decade of the twentieth century.8 Between 1924 and 1973 its share in the labor force declined absolutely by almost as much as agriculture's, (p.226) and proportionally it declined more. In contrast to agriculture, its share in output fell as much as its share in the labor force.

Like mining, transport and communications absorbed a rapidly increasing proportion of the labor force in the nineteenth century. After World War I its share in the labor force remained about constant, and its share in the capital stock declined.

Public and professional services was the great growth sector. It is made up of three principal components, the armed forces, public administration, and professional and scientific services, the trends in which were rather different. The armed forces' share in the labor force was much higher in 1951, where there was still conscription, than it was in 1937, but in the course of the postwar period it fell to a lower level than in earlier peacetime periods. Public administration increased its share in the labor force chiefly in two large steps across the wars. Within the postwar period it first fell, then rose. Professional and scientific services, of which health and education are the largest elements, increased its share in the labor force in all periods, peacetime as well as across the wars. Its increase in the postwar period was larger, absolutely and proportionally, than in earlier periods. In the sector as a whole, output at current prices is a more significant statistic than output at constant prices on account of the conventional basis on which the latter has to be measured. Between 1937 and 1951 (though not in the postwar period) the rise in the sector's share in output at current prices was much smaller than the rise in its share in labor input, owing to a very substantial fall in relative wages.

Trends in the share in the economy of commerce did not follow any simple pattern of rise or fall. That sector too is made up of rather heterogeneous elements. The substantial rise in the sector's share in the labor force up till World War I was due almost entirely to the distributive trades. There was no net change in the share of miscellaneous services before 1914,* and though there was a rapid proportional increase in insurance, banking, and finance, the absolute numbers involved were small. The distributive trades continued to give employment to an increasing proportion of the labor force in the interwar period. The proportion in miscellaneous services also increased in that period, but not enough to offset the fall across World War I. Two‐thirds of the enormous fall in commerce's share in the labor force between 1937 and 1951 came from miscellaneous services, one‐third from the distributive trades. Within the postwar period the proportion of the labor force in the distributive trades first rose (till 1964), then fell; the proportion in (p.227) miscellaneous services first fell (till 1960), then rose. The substantial rise in insurance, banking, and finance, which was particularly rapid in the late postwar period, resembled what happened in the other predominantly white‐collar subsector, professional services. It accounted for most of the increase between 1951 and 1973 in the commerce sector taken as a whole.

Focusing attention on shares in the labor force, we may thus summarize as follows trends over the three periods 1856–1913, 1913–37, and 1937–73 (thus including the war periods with the subsequent peacetime periods). Between 1856 and 1913 the enormous decline in agriculture's share permitted increases in the share of all other sectors, except manufacturing; the largest increase was in the distributive trades. Between 1913 and 1937 no sector was releasing labor on anything like the same scale as agriculture had done before 1913. The continuing decline in agriculture, and now also a decline in mining, were matched by fairly general increases in the shares of other sectors, the largest gains (but on a much smaller scale than before 1913) being in the service industries. Between 1937 and 1973 there was a decline in the share of miscellaneous services and, to a lesser extent, the distributive trades, of an amount far greater than the change in any sector's share between 1913 and 1937; the decline in their share of total labor input over those 36 years was larger than the decline in agriculture's share in the 40 years between 1873 and 1913 (11.3 percentage points, compared with 9.9 points). Substantial declines also continued in the shares of agriculture and mining. The declines in these three sectors were matched, to rather more than half their extent, by the increase in the share of public and professional services. There was also, for the first time in over 100 years, a significant increase in the proportion in manufacturing. The main part of this increase, and of the big decline in miscellaneous services and the distributive trades, had already taken place by 1951. Between 1951 and 1973 the principal net changes were that declines in agriculture and mining were matched by increases in the service industries, a pattern resembling in some respects that of the period before 1914.

The Growth of Output, Inputs, and Total Factor Productivity in the Major Sectors of the Economy

We now turn from the long‐term trends in the relative importance of the major sectors to the underlying growth rates of their output, inputs, and TFP. The data for the nine sectors for all periods except 1913–21 (for which the inputs by sector are not available) are set out in Table 8.3. The postwar period is divided into two parts, 1951–64 and (p.228)

Table 8.3 Growth of Output, Labor, Capital, Total Factor Input, Total Factor Productivity, and Weighted Total Factor Productivity by Sector, 1856–1973

(Annual percentage growth rates)

Variable and sector

1856–1873

1873–1913

1924–1937

1937–1951

1951–1964

1964–1973

Output:

Agriculture, forestry, fishing

0.2%

−0.1%

1.4%

1.7%

2.6%

2.5%

Mining, quarrying

3.6

1.9

−0.4

−1.0

−0.7

−3.3

Manufacturing

2.6

2.0

3.2

2.5

3.2

3.0

Construction

3.1

1.1

4.6

−1.2

3.8

1.8

Gas, electricity, water

5.5

5.1

5.8

4.2

5.1

5.2

Transport, communications

2.9

2.7

1.5

2.4

2.2

3.5

Commerce

(2.4)

(2.0)

1.7

−0.2

3.0

3.0

Public and professional services

(1.4)

(2.5)

(1.3)

(2.7)

(1.5)

(2.2)

Ownership of dwellings

1.0

1.1

1.9

0.8

1.8

2.6

GDP

2.0

1.8

2.3

1.5

2.8

2.7

Labor:a

Agriculture, forestry, fishing

−1.0

−0.6

−1.1

0.4

−2.4

−4.8

Mining, quarrying

1.7

2.1

−2.0

−0.6

−2.7

−7.2

Manufacturing

1.1

0.8

1.4

1.0

0.2

−1.6

Construction

2.0

1.0

3.5

1.1

1.6

−0.8

Gas, electricity, water

4.7

3.3

3.1

1.4

0.5

−3.4

Transport, communications

2.0

2.1

0.6

0.5

−0.7

−1.3

Commerce

1.9

1.3

2.3

−3.2

0.3

−0.7

Public and professional services

1.3

1.8

1.8

3.4

−0.4

0.9

GDP

0.9

0.9

1.5

0.1

0.0

−1.1

Capital:b

Agriculture, forestry, fishing

0.3

−1.0

−0.1

0.8

1.5

3.3

Mining, quarrying

3.3

2.0

0.5

−0.4

3.3

1.6

Manufacturing

3.2

2.6

1.0

2.9

3.3

3.3

Construction

(3.1)c

(1.1)c

1.6

2.4

6.2

6.4

Gas, electricity, water

4.9

3.6

4.8

0.9

4.0

4.1

Transport, communications

3.2

2.0

0.2

−0.4

1.1

2.7

Commerce

1.6

1.6

2.0

1.2

3.7

6.0

Public and professional services

2.1

3.1

2.3

1.2

2.9

5.8

Ownership of dwellings

1.5

1.9

3.4

1.0

2.3

3.2

GDP

2.0

2.0

1.8

1.1

2.8

3.9

TFI:

Agriculture, forestry, fishing

−0.7%

−0.5%

−0.7%

0.5%

−0.9%

−1.0%

Mining, quarrying

2.2

2.0

−1.6

−0.6

−2.2

−5.5

Manufacturing

1.7

1.4

1.3

1.6

1.2

−0.1

Construction

(2.3)

(1.0)

3.3

1.3

2.0

0.2

Gas, electricity, water

4.9

3.5

4.0

1.1

1.8

0.9

Transport, communications

2.4

2.0

0.5

0.2

−0.2

−0.2

Commerce

1.8

1.5

2.2

−1.8

1.3

1.2

Public and professional services

(1.4)

(2.4)

(1.9)

(2.7)

(0.5)

(2.3)

Ownership of dwellings

1.5

1.9

3.4

1.0

2.3

3.2

GDP

1.4

1.4

1.6

0.4

0.7

0.3

TFP:

Agriculture, forestry, fishing

0.9

0.4

2.1

1.2

3.5

3.5

Mining, quarrying

1.4

−0.1

1.2

−0.4

1.5

2.2

Manufacturing

0.9

0.6

1.9

0.9

2.0

3.1

Construction

(0.8)

(0.1)

1.3

−2.5

1.8

1.6

Gas, electricity, water

0.6

1.6

1.8

3.1

3.3

4.3

Transport, communications

0.5

0.7

1.0

2.2

2.4

3.7

Commerce

(0.6)

(0.5)

−0.5

1.6

1.7

1.8

    GDPd

0.64

0.45

0.70

1.09

2.04

2.42

TFP weighted by sector's share in GDP:

Agriculture, forestry, fishing

0.16

0.06

0.10

0.04

0.19

0.12

Mining, quarrying

0.06

−0.01

0.08

−0.02

0.05

0.06

Manufacturing

0.20

0.16

0.58

0.29

0.70

1.07

Construction

0.03

0.00

0.05

−0.12

0.10

0.11

Gas, electricity, water

0.00

0.01

0.04

0.08

0.07

0.14

Transport, communications

0.03

0.05

0.10

0.21

0.21

0.31

Commerce

(0.15)

(0.14)

−0.12

0.47

0.42

0.45

Total of above

0.63%

0.41%

0.83%

0.95%

1.74%

2.26%

    Total intra‐industryd

(0.55)

(0.35)

0.60

0.84

1.88

2.19

Residual

(0.09)

(0.10)

0.10

0.25

0.16

0.23

    GDPd

0.64%

0.45%

0.70%

1.09%

2.04%

2.42%

Note: The GDP figures for output differ from those in Table 7.2 because they are based on output‐side estimates of GDP at factor cost to make them comparable to the sectoral output estimates, and not on the (geometric) mean of expenditure, income, and output estimates. The estimates of TFP are similarly affected. Output‐based estimates are used in the rest of this chapter. The less reliable estimates are shown in parentheses. The problems with the output and weighted TFP estimates for commerce in 1856–1937 and with the output, TFI, and weighted TFP estimates for public and professional services are described in note 2 (pp. 646–47). The basis for the estimates of the capital stock in construction in 1856–1913 is explained in note c, below.

(a) For 1856–1913, man‐years; for other periods, man‐hours. See also pp. 211–12, 230.

(b) Gross fixed capital, 1856–73 and 1873–1913; gross capital stock plus inventories, other periods.

(c) Capital stock in construction, 1856–73 and 1873–1913, assumed to grow at the same rate as output of construction.

(d) Includes public and professional services, ownership of dwellings, and an allowance for financial services. See also note 2 to this chapter.

(p.229) (p.230) 1964–73. Comparable figures relating to the five postwar cycles are given in Appendix K. Labor input is measured in man‐years for 1856–1913 and in man‐hours thereafter. The quality of the labor supplied in the man‐years or man‐hours is thus included in TFP, not in TFI.*

The first four sections of Table 8.3 contain the sectoral data for the growth rates of output, labor, capital, and total factor input (TFHI or TFII) for each of the nine sectors. The growth rates of total factor productivity (TFHP or TFYP) are given in the fifth section for the seven sectors for which it is a meaningful concept.9 The last section shows the same data weighted by the sector's share in total output; the weighted figures indicate the contributions of the sector to the overall growth of TFP, and their variations between periods therefore include the effects of changes in the weights of the sectors. The sum of weighted sectoral growth rates in TFP (intra‐industry TFP growth) is less than the growth rate of TFP in the economy because of a structural element that will be discussed in Chapter 9.

We comment first on output, input, and TFP in the peacetime periods. Trends across World War II have certain special characteristics and will be discussed separately.

Output

The annual percentage rates of growth of total output in successive peacetime periods were 2.0, 1.8, 2.3, 2.8, and 2.7. This pattern of fall, rise, and fractional fall was reproduced in only one sector, that sector, curiously enough, being agriculture. Manufacturing differed in having no increase in the rate of growth between 1924–37 and 1951–64. The unreliable pre‐1914 figures for the other largest sector, commerce, for what they are worth, indicate a lower growth rate in 1924–37 than before 1914, contrary to the aggregate pattern. The rate of growth of output in mining declined between each successive peacetime period. In utilities the rate was high and steady.

Comparing the sources of the changes in the overall rate of growth of output between successive periods, we note that the fall between 1856–73 and 1873–1913 was common to most sectors. The increase between 1873–1913 and 1924–37 arose almost exclusively from the industrial (p.231) sectors (other than mining), viz. manufacturing, utilities, and construction, and from agriculture — a finding rather different from the normal stereotype of the interwar period. On the other hand, between 1924–37 and 1951–64 the increase came entirely from the nonindustrial sectors; output in manufacturing grew at about the same rate in the two periods, and mining, construction, and utilities showed a falling‐off in the rate of growth. Between 1951–64 and 1964–73 neither the largest industrial sector, manufacturing, nor the largest nonindustrial sector, commerce, showed any significant change in the rate of growth of output, nor did agriculture and utilities. A falling‐off in construction and a further falling‐off in mining were about balanced by the increased rates in transport and communications and in public and professional services.

Inputs

There was no significant difference between 1856–73 and 1873–1913 in the rate of growth of labor, capital, and TFI in the economy as a whole. There was some increase between the two periods in the rate of growth of the labor force in mining, and some slowing down in the percentage rate of decline in the agricultural labor force (hence a fortiori in its absolute rate of decline); this was offset by a slight slowdown in the rate of increase of the labor force in the two largest sectors, manufacturing and commerce.

Labor input grew more rapidly in 1924–37 than in 1873–1913 and capital input less rapidly (considerably less rapidly if dwellings are excluded). TFI grew more rapidly, by a small margin. Sectoral trends were mixed. In the largest sectors, manufacturing and commerce, and also in agriculture, differences between the two periods in the growth rates of TFI did not conform to the differences noted above in the growth rates of output. In manufacturing, more rapid growth of labor input in 1924–37 than in 1873–1913 was outweighed by much less rapid growth in capital input, so that growth in TFI was slightly lower, despite the much faster growth of output. In commerce, on the other hand, where output grew less rapidly in 1924–37 than in 1873–1913, both labor input and capital input grew more rapidly.

Comparisons involving the postwar period show rather more uniformity in input trends. The rate of growth of labor input was lower in 1951–64 than in 1924–37 in every sector, and the rate of growth of capital input was higher in every sector except dwellings and utilities. The trend in labor outweighed the trend in capital, and the rate of growth of TFI fell in every sector without exception.

(p.232) Comparing the two parts of the postwar period, we find that the rate of growth of labor input was lower in 1964–73 than in 1951–64 in every sector except public and professional services. Higher rates of growth of capital in 1964–73 than in 1951–64 were also fairly general, but there were important exceptions in utilities and manufacturing (with little or no change) and mining (which fell). The rate of growth of TFI was either lower or unchanged in 1964–73 compared with 1951–64 in all sectors except public and professional services. These net differences between 1951–64 and 1964–73 were the result in some cases of a long‐swing type of movement within the five cycles of the postwar period. This was most pronounced in utilities and construction, where the rate of growth of TFI was at a peak in the 1960–64 cycle, and thereafter fell steeply. In manufacturing a similar pattern appeared, in less pronounced form and on the side of capital only, the rate of increase in capital being higher in the 1960–64 and 1964–68 cycles than either before or after.

Reference may also be made to the trends in the capital‐labor ratio. One would expect the ratio to rise in the course of economic growth. And so it did in the economy as a whole and in every sector, in both parts of the postwar period and, to a much lesser extent, in 1873–1913 and (except for commerce) in 1856–73. In the interwar period, however, the rate of growth of the total capital stock excluding dwellings (1.3 percent) was actually less than the rate of growth of the labor input. The decline in the capital‐labor ratio in that period was particularly great in manufacturing.

Total Factor Productivity

The rate of growth of TFP in the economy as a whole followed the U‐shaped pattern discussed in the previous section. The downward phase of the U, as it appears in Table 8.3, is confined to the decline between the periods 1856–73 and 1873–1913. Sectoral data cannot be shown for the period 1913–24, which marks the low point for the economy as a whole in the rates of growth of TFQP and TFYP.

The U‐shaped pattern is found fairly generally across sectors. It occurs in agriculture, mining, manufacturing, and construction. It occurs in commerce, but with the low point in 1924–37 instead of in 1873–1913. Transport and utilities lack the downward phase of the U and have instead a rising rate of TFP growth between each period. There was, however, in the various phases a considerable variety between sectors in the magnitude of the movements, as well as some variety in their timing.

(p.233) Between 1856–73 and 1873–1913 there was a particularly large decline in the rate of TFP growth in agriculture, at that time one of the three largest sectors. Even larger were the declines in construction, where TFP barely grew at all between 1873 and 1913, and in mining, where it fell absolutely.10

The movements within the period 1873–1913 are considered in Appendix L. Until 1899 the main sources of slowing down in the rate of growth of TFP were agriculture and mining. After 1899 there was significant slowing down in manufacturing and also in construction (affected by the long swing in building). In a number of sectors there was a retardation in 1873–82, made good in the 1880's and 1890's. In manufacturing there was a similar retardation in the 1880's, made good in the 1890's.

TFP grew more rapidly in the interwar period than in 1873–1913 in every sector, with the important exception of commerce, where it fell absolutely. This is compatible with the generally held view that labor crowded into the distributive trades and miscellaneous services in the interwar period in face of the general unemployment. The fall in TFP was thus in a sense a measure of underemployment rather than an indication of an adverse shift in the production function. The improvement in commerce was, as may be seen from the final section of Table 8.3, much the largest single sectoral source of improvement in TFP growth between 1924–37 and 1951–64.* Allowance for the abnormal factors affecting the distributive trades and miscellaneous services in the interwar period would raise the overall figure for TFP growth in that period and make it more substantially in excess of the very low figure for 1873–1913, while leaving it much below the postwar rate. It would underline the pervasiveness of the upward trend in TFP growth in the interwar period compared with 1873–1913.

Productivity trends within the interwar period show considerable divergences between sectors (see Appendix M for a more detailed account). TFP in agriculture, mining, construction, and transport did substantially better in 1924–29, commerce and manufacturing in 1929–37 (Table M.2). The outcome for TFP in the economy as a whole was a better performance in 1924–29 than in 1929–37 (Table M.1).

Pervasive improvement continued in the postwar period. The rate of TFP growth was more rapid in 1951–64 than in 1924–37 in every sector. (p.234)

Table 8.4 Growth of Total Factor Input and Total Factor Productivity: Excess of Rates in Manufacturing Over Commerce, 1856–1973

(Annual percentage growth rates)

Period

TFI

TFP

1856–1873

−0.1%

0.3%

1873–1913

−0.1

0.1

1924–1937

−0.9

2.4

1937–1951

3.4

−0.7

1951–1964

−0.1

0.3

1964–1973

−1.3

1.3

Note: Labor is measured in man‐years for 1856–1913, in man‐hours for 1924–73.

This includes the industrial sectors where, as already noted, output grew less rapidly in 1951–64 than in 1924–37. However, the extent of the difference in TFP growth between the periods was significantly less in the industrial sectors (except utilities) than in the nonindustrial sectors. Most important, the rate of growth of TFP in manufacturing in 1951–64 scarcely exceeded what it had been in 1924–37.*

These contrasts between sectors were to a large extent reversed between 1951–64 and 1964–73. The rate of growth of TFP in 1964–73 was substantially higher than in 1951–64 in manufacturing, mining, and utilities, but not in commerce or agriculture.

Between the two parts of the postwar period, the opposite trends in TFI growth and TFP growth in the two largest sectors, manufacturing and commerce, thus largely canceled out. In manufacturing TFI growth fell to an unusual extent,11 and TFP growth rose to an unusual extent; in commerce there was no clear trend in either. In both, therefore, there was little trend change in the rate of growth of output. Some similar tendency to offset between TFI and TFP is observed also in the postwar period in some other sectors (agriculture, mining, utilities), where likewise a divergence from GDP‐average in the trend of TFI growth was accompanied by a divergence in the opposite direction in the trend of TFP growth.

The tendency to offsetting sectoral movements over a larger period in the rates of TFI growth and TFP growth is brought out in relation to the two largest sectors in Table 8.4, which shows the differences between (p.235) manufacturing and commerce in the rates of TFI growth and TFP growth. The change from one period to the next is almost always in opposite directions in the two columns, i.e. a change between successive periods in the extent of difference between the rates of growth of TFI in the two sectors was always (except for 1873–1913) accompanied by a change in the opposite direction in the extent of the difference between them in the rate of growth of TFP.

This opposition was most pronounced of all in the trans–World War II period, to which we now turn.

The Trans–World War II Period: 1937–1951

The pattern of growth in output, inputs, and TFP in 1937–51 was, not surprisingly, a good deal more uneven across the economy than in peacetime periods.

Total output grew less rapidly than in earlier peacetime periods. The fastest growth was in public and professional services (which includes defense) and in manufacturing; output in commerce fell. TFI in manufacturing grew fairly rapidly. TFI also grew in agriculture — the only period when this happened. TFI in mining declined at a much less rapid pace than in either 1924–37 or 1951–73. Agriculture and mining were both sectors where for strategic or balance‐of‐payments reasons (or both) the growth of inputs was encouraged. Most other sectors experienced relatively slow growth of TFI by comparison with other periods.

The overall rate of growth of TFP was slightly higher than in the interwar period; this was due chiefly to a structural change (discussed in Chapter 9), rather than to improved performance within industries. Intra‐industry improvement in TFP growth compared with 1924–37 came chiefly in commerce, which shed the surplus labor it had acquired in the depression. On the other hand, the three sectors where TFI grew more rapidly (or declined less rapidly) than in other periods, namely agriculture, mining, and manufacturing, experienced much slower growth of TFP in the transwar period than in the peacetime periods. This suggests that the policy of pushing resources into these sectors in the transwar period may have brought them up against diminishing returns. In fact, of the seven sectors shown in the last section of Table 8.3, there was only one, construction, in which the changes between the interwar and transwar periods in the growth of TFI and the growth of TFP were not in opposite directions. We shall revert in Chapter 9 to the question of whether a more general inference of diminishing returns is justified by the opposite trends in TFI and TFP in other periods.

(p.236) The trend in productivity in construction is unique and calls for special comment. The fall in productivity in 1937–51 was so great that notwithstanding a historically high growth rate during the postwar period, the absolute level of productivity in construction prevailing before World War II was not regained till near the end of the 1960's. (This is true whether it is measured as labor productivity or as TFP.) The absence of any net rise in productivity in construction between 1937 and the late 1960's, which has no parallel in any other sector, is the more remarkable in that technical advances in building materials and methods were more rapid than ever before (see, on this point and others that follow, Zweig 1951; Carter 1958; M. Bowley 1966, 1967; Richardson & Aldcroft 1968).

There were a number of special reasons for the fall in productivity in construction across World War II.* A major general reason, however, appears to have been the effect of full employment. Whereas in the distributive trades and miscellaneous services the move to full employment raised productivity, by drawing away surplus labor, the net effect in construction was in the opposite direction. High pressure of general demand can have unfavorable or favorable effects on productivity. There appear to be four reasons why the unfavorable effects predominated in the case of construction:

  1. (1) Temporary holdups due to bad weather, delays in supplies, or intervals between jobs are more of a feature in construction than in other industries. When there was extensive general unemployment, such holdups led to labor being laid off, but under full employment this was discouraged by competition for labor.

  2. (2) The industry always had more than its share of low‐quality workers, and this became more pronounced under full employment, both because of a loss of men to other industries and because when there was substantial unemployment the lowest‐quality workers had tended not to be in employment at all.

  3. (3) Holdups due to delays in supplies tended to become more frequent when the pressure of demand was higher.

  4. (4) Possibly, the threat of unemployment and the risk of failure were more important in construction than in other industries in maintaining the incentives of workers and employers.

It may be noted that though these aspects of full employment tended to lower productivity in construction, they did not necessarily represent (p.237) any absolute loss to the economy as a whole. Rather they reflected a tendency to diminishing returns to total production from increasing pressure of demand.12

Table 8.5 Growth of Total Factor Productivity by Sector, 1856–1913 and 1924–1973

(Annual percentage growth rates)

Sector

1856–1913

1924–1973

Gas, water, electricity

1.3%

3.3%

Agriculture, forestry, fishing

0.6

2.6

Transport, communications

0.6

2.2

Manufacturing

0.7

1.8

Mining, quarrying

0.4

1.2

Commerce

(0.6)

1.2

Construction

(0.3)a

0.4

Note: Labor is measured in man‐years for 1856–1913, in man‐hours for 1924–73.

(a) Based on the assumption that the capital stock in construction grew at the same rate as output. The rate of growth of labor productivity was 0.35%.

The statistics suggest that the very long‐run trend in construction, as in other sectors, was for the rate of productivity growth to rise from World War I onward. However, this trend was subject to a very substantial offset across World War II on account of the transition to a full‐employment economy, and it took a long time for the general upward trend to productivity to make that good.13

To conclude our discussion of TFP by sector, Table 8.5 shows rates of growth over the two long periods, each of about half a century, before and after World War I.

In every sector TFP grew more rapidly in 1924–73 than in 1856–1913. The sectors are ranked in Table 8.5 by their rate of growth of TFP in 1924–73. The 1856–1913 ranking is similar, to the extent that utilities come at the top and construction at the bottom. However, the differences between the remaining sectors in 1856–1913 are small and are well within the margin of statistical error (large for that period). From the data available to us, it thus appears that there was fanning‐out between sectors in TFP in 1924–73 that had not occurred before World War I.

There is much less stability between the two periods in the way the sectors ranked in respect of growth of TFI and output, and except for utilities, their ranking in TFI growth does not bear any regular relationship to their ranking in TFP growth. This suggests that the relative stability of the ranking in TFP growth was the result of technical or other supply‐side characteristics of the sectors.

(p.238) The Growth of Output, Inputs, and Total Factor Productivity in Subsectors of the Economy

Some of the sectors discussed in the last section are large or heterogeneous, and it is interesting to consider the movements in productivity in industries within such sectors. Two sectors, manufacturing and commerce, accounted for about 60 percent of GDP. Sectors as large as these two are not likely to be as homogeneous, in the sense of producing related types of goods or services in a distinctive way, as the smaller sectors. We therefore consider the growth of productivity in their component industries in this section. We shall also consider some of the industries in the transport and communications sector, partly because this sector is also a fairly large one, but also because it is made up of both declining and growth industries.

Discussion is confined almost exclusively to periods after World War I for lack of sufficiently disaggregated data on earlier periods. There is also a problem with the interwar capital stock data.14

Manufacturing

The manufacturing sector is divided into 13 industry groups that correspond, with one or two minor adjustments, to the Standard Industrial Classification (SIC).* Most of them are still fairly heterogeneous, but less so than when aggregated. The largest industry groups are food, drink, tobacco; mechanical engineering and shipbuilding; and textiles. Chemicals, the other metal‐using industries, and paper, printing, publishing are also important (Table 8.6). The rates of growth of output, labor, capital, and TFI in the 13 industry groups over four post‐1924 periods are shown in Table 8.7. As in the last section, labor is measured in man‐hours, and capital is gross of depreciation (net of retirements); and the weights used to combine the growth rates of labor and capital are their shares in total income in the first years of each period.15 Table 8.7 also shows the rates of growth of TFP and TFP weighted by the industry's share in manufacturing.

As before, we first consider the interwar and postwar periods, leaving the wartime period for separate treatment; and again we divide the postwar period into its earlier and later parts, 1951–64 and 1964–73.

There was, naturally, diversity in experience between industries. But insofar as there were differences between periods in the rates of growth of inputs and productivity in manufacturing as a whole, they (p.239)

Table 8.6 Shares of Value‐Added in Manufacturing, 1924–1973

(Percent)

Industry

1924

1937

1951

1964

1973

Food, drink, tobacco

15.0%

13.9%

9.5%

11.0%

10.8%

Chemicals

5.8

6.3

7.2

8.8

7.6

Iron and steel

6.2

7.4

7.5

6.6

5.0

Electrical engineering

3.6

5.5

7.4

9.2

10.0

Mechanical engineering and shipbuilding

11.5

11.8

15.4

15.8

16.9

Vehicles

6.2

9.3

8.7

10.8

10.7

Other metal industries

6.6

7.7

7.6

7.9

8.7

Textiles

16.0

10.5

12.6

7.2

5.9

Clothing

10.3

8.5

4.9

3.8

3.5

Bricks, pottery, glass, cement

3.4

4.5

4.2

4.3

4.4

Timber, furniture

3.8

3.7

3.4

2.9

3.7

Paper, printing, publishing

7.8

7.8

7.8

7.8

8.5

Leather and other manufacturing

3.8

3.1

3.8

4.0

4.3

Total manufacturing

100.0%

100.0%

100.0%

100.0%

100.0%

Note: The classification by industry in 1924 and 1937 is according to the 1948 SIC and that in 1951 and 1964 is according to the 1958 SIC. If the 1951 estimates were reclassified onto the 1948 SIC the most significant changes in shares would occur in food, drink, tobacco (+0.6), mechanical engineering and shipbuilding (−1.8), vehicles (+1.0), other metals (+1.1), and textiles (−0.6).

were pervasive. That is to say, they were the result of similar differences in the great majority of the 13 industries distinguished, rather than the result of extremely strong movements in a few.

In manufacturing as a whole output grew at almost the same rate in each of the three periods, 1924–37, 1951–64, and 1964–73. The rate of growth of labor was lower in 1951–64 than in 1924–37 (all 13 groups). The rate of growth of capital was higher in 1951–64 than in 1924–37 (11 groups).16 As a result of these opposite movements in labor and capital, the rate of growth of TFHI was about the same in 1924–37 and 1951–64 in manufacturing, and so was the rate of growth of TFHP.

The rate of growth of labor was lower in 1964–73 than in 1951–64 (12 groups). The rate of growth of capital was on average the same in 1964–73 as in 1951–64; hence the rate of growth of TFHI was lower in 1964–73 than in 1951–64 (10 groups). The rate of growth of TFHP was higher in 1964–73 than in 1951–64 (11 groups); the two exceptions, vehicles and clothing, were both industries where there had been a falling off in the rate of growth of output. However, there were other industries, such as iron and steel, where the rate of TFHP growth increased despite a substantial fall in the rate of growth of output.

The pervasiveness of these trends suggests strongly that the causal forces at work were of a general nature rather than particular innovations or other special factors affecting individual industries. (p.240)

Table 8.7 Growth of Output, Labor, Capital, Total Factor Input, Total Factor Productivity, and Weighted Total Factor Productivity in Manufacturing, 1924–1973

(Annual percentage growth rates)

Industry

1924–1937

1937–1951

1951–1964

1964–1973

1924–1937

1937–1951

1951–1964

1964–1973

1924–1937

1937–1951

1951–1964

1964–1973

OUTPUT

LABOR

CAPITAL

Food, drink, tobacco

2.8%

1.7%

2.6%

2.7%

1.8%

0.2%

0.1%

−1.2%

0.7%

1.6%

3.7%

4.0%

Chemicals

3.1

4.8

5.8

6.2

1.7

3.9

0.3

−1.1

1.7

4.0

5.5

5.0

Iron and steel

3.0

1.7

2.5

−0.2

1.0

0.6

0.0

−3.1

0.6

0.1

4.8

2.0

Electrical engineering

6.2

5.6

6.0

5.7

5.4

3.1

3.1

−1.1

2.4

6.3

5.0

3.5

Mechanical engineering and ship‐building

1.8

3.8

2.4

3.2

1.1

3.0

0.5

−1.6

0.7

6.6

4.0

3.7

Vehicles

6.3

3.6

4.9

0.7

3.2

3.7

1.0

−1.8

3.1

8.4

4.6

0.8

Other metal industries

4.5

2.8

2.0

0.9

2.3

1.8

0.6

−1.2

1.8

6.6

3.3

4.3

Textiles

1.6

0.2

0.1

2.9

−0.3

−1.2

−2.8

−3.5

−0.4

−0.1

−2.6

2.1

Clothing

2.1

−1.7

2.2

1.9

0.4

−1.6

−1.6

−2.4

2.1

2.2

0.5

2.0

Bricks, pottery, glass, cement

4.6

2.4

3.4

3.4

2.8

0.3

0.2

−2.0

0.1

1.0

4.9

5.0

Timber, furniture

4.8

−0.2

2.2

3.5

2.0

−0.4

−0.6

0.3

1.9

2.2

2.8

4.8

Paper, printing, publishing

2.8

2.6

4.1

2.7

2.0

−0.2

1.2

−1.4

2.1

2.1

3.1

3.3

Leather and other manufacturing

4.3

3.0

3.2

4.5

1.5

1.2

0.3

0.1

2.0

0.9

2.8

5.5

Total manufacturing

3.2%

2.5%

3.2%

3.0%

1.4%

1.0%

0.2%

−1.6%

0.9%

2.9%

3.3%

3.3%

TFHI

TFHP

WEIGHTED TFHPa

Food, drink, tobacco

1.3%

0.8%

1.8%

1.1%

1.5%

0.9%

0.8%

1.6%

0.22%

0.12%

0.07%

0.17%

Chemicals

1.7

3.9

2.6

1.8

1.4

0.9

3.2

4.4

0.08

0.05

0.23

0.39

Iron and steel

1.0

0.4

2.0

−1.7

2.0

1.3

0.5

1.5

0.13

0.10

0.04

0.10

Electrical engineering

4.2

4.3

3.7

0.2

2.0

1.3

2.3

5.5

0.08

0.07

0.17

0.51

Mechanical engineering and ship‐building

1.1

3.6

1.4

−0.2

0.7

0.2

1.0

3.4

0.08

0.02

0.15

0.54

Vehicles

3.2

4.9

1.8

−1.3

3.1

−1.3

3.1

2.0

0.19

−0.12

0.27

0.21

Other metal industries

2.2

3.4

1.5

0.3

2.3

−0.6

0.5

0.6

0.15

−0.04

0.04

0.05

Textiles

−0.3

−1.0

−2.7

−1.9

1.9

1.2

2.8

4.8

0.30

0.13

0.35

0.34

Clothing

0.7

−1.0

−1.2

−1.4

1.4

−0.7

3.4

3.3

0.15

−0.06

0.17

0.12

Bricks, pottery, glass, cement

2.1

0.5

1.7

0.5

2.5

1.9

1.7

2.9

0.09

0.08

0.07

0.12

Timber, furniture

2.0

−0.1

0.1

1.1

2.9

−0.1

2.1

2.4

0.11

0.00

0.07

0.07

Paper, printing, publishing

2.0

0.5

2.0

0.0

0.8

2.1

2.1

2.7

0.06

0.16

0.17

0.21

Leather and other manufacturing

1.6

1.2

1.2

1.6

2.7

1.8

2.0

2.9

0.10

0.06

0.07

0.11

Residual

0.2

0.3

0.1

0.2

Total manufacturing

1.3%

1.6%

1.2%

−0.1%

1.9%

0.9%

2.0%

3.1%

1.9%

0.9%

2.0%

3.1%

(a) The weights are the shares in total manufacturing output, as in Table 8.6, in the first year of each period.

(p.241)

(p.242) In addition to considering the extent to which individual industries conformed to sector‐wide trends, one may also analyze the relationships between the growth rates of inputs, output, and productivity in the individual industries. Correlations between and within periods for these and certain other variables are given and further discussed in Appendix J in relation to the periods 1924–37, 1951–64, and 1964–73. The following points emerge.17

As might be expected, there was a positive correlation between periods in the rate of growth of output by industry, reflecting changes in industrial structure. At the present level of aggregation the correlation is only fair, and in fact there were only two industries (electrical engineering and bricks, pottery, etc.) where output in all three periods grew faster than the average for manufacturing as a whole. Within each period there was a strong positive correlation between the rate of growth of output and the rate of growth of TFHP, as has been found in many other such studies. There was also a strong positive correlation between the output and TFHI rates. On the other hand, the growth rates of TFHI and TFHP were not correlated with each other at all in the postwar period, and were only weakly correlated in the interwar period.

In general, not surprisingly, the two parts of the postwar period showed more resemblance to each other than to the prewar period. This is especially notable in the rate of growth of TFHP, where there was no correlation at all between interwar and postwar but a reasonable correlation between the two parts of the postwar period. However, the interwar period and 1951–64 were more alike than the two parts of the postwar period in one respect: there was a strong tendency for the same industry groups to experience a high rate of growth of labor input, and hence of TFHI, in both periods. (To a considerable extent, these were the same industries that had rapid growth in TFHI in 1937–51; see below). However, there was virtually no such tendency between 1951–64 and 1964–73. Thus the correlation in output growth between the first pair of periods was achieved mainly through correlation in TFHI growth and in the second pair of periods mainly through correlation in TFHP growth.

The contributions of the several industries to growth in TFHP in manufacturing as a whole, shown in the last section of Table 8.7, are, of course, affected by changes in their weights over time. Three industry groups (chemicals, electrical engineering, and mechanical engineering, shipbuilding) between them account for almost the whole of (p.243) the increase in the rate of growth of TFHP in manufacturing between the first and last periods (1924–37 and 1964–73). Both the weights of these industries and their TFP growth rates increased. By contrast, the contribution made by textiles and clothing to the total remained about unchanged because the decline in their weights offset the substantial increases in their TFHP growth rate. A further contrast is provided by vehicles, whose weight rose but whose TFHP growth rate fell.

The trans–World War II period was marked by rapid growth in the output of and inputs into those industry groups that were essential for the war (metal‐using industries and chemicals), and that were important in the postwar investment and export booms (broadly the same industries); other groups grew much less rapidly. The contrast was particularly noticeable on the input side: the “war” industries experienced very rapid input growth, involving both labor and capital, and TFHI growth was more rapid in 1937–51 in these industries than in any of the peacetime periods. The metal‐using industries continued to have above‐average growth in TFHI in 1951–64 but ceased to do so in 1964–73. The other industry groups generally experienced slower growth of TFHI in 1937–51 than in the peacetime periods. This was the consequence of the wartime direction of labor policies and of the concentration of government‐financed investment during the war years in chemicals and the engineering and allied industries.18

Growth in TFHP in most manufacturing industries was slow (sometimes negative) in 1937–51 compared with the two peacetime periods; further, some of the rapidly growing engineering and metal‐using industries experienced the slowest growth.

Commerce

After manufacturing, commerce is the largest of the nine major sectors. It would be interesting to examine the growth of output and inputs in each of the three component industry groups — distributive trades; insurance, banking, and finance; and miscellaneous services — but unfortunately we do not have sufficiently reliable capital‐stock data for each industry group to calculate the growth rates of TFI and hence TFP separately. Output, labor and labor productivity growth rates are available, however. (Table 8.8).

Of the three groups, distributive trades is the largest (Table 8.9). It includes both wholesale and retail trades that are similar in size. The insurance, banking, and finance group is fairly large in terms of output and capital,19 though not in terms of labor input. The most important (p.244)

Table 8.8 Growth of Output, Labor, and Labor Productivity in Commerce, 1856–1973

(Annual percentage growth rates)

Variable and sector

1856–1873

1873–1913

1924–1937

1937–1951

1951–1964

1964–1973

Output:

Distributive trades

2.7%

2.0%

1.9%

−0.2%

2.9%

2.8%

Insurance, banking, finance

1.6

0.6

4.1

5.2

Miscellaneous services

1.6

−1.4

2.3

1.5

Total commerce

(2.4)

(2.0)

1.7

−0.2

3.0

3.0

Labor:

Distributive trades

2.1

2.1

2.6

−2.3

1.2

−1.9

Insurance, banking, finance

7.6

4.3

1.8

−0.8

2.0

2.3

Miscellaneous services

1.7

0.7

2.1

−4.5

−1.1

0.0

Total commerce

1.9

1.3

2.3

−3.2

0.3

−0.7

Labor productivity:

Distributive trades

0.6

−0.1

−0.7

2.1

1.7

4.7

Insurance, banking, finance

−0.2

1.4

2.1

2.9

Miscellaneous services

−0.5

3.1

3.4

1.5

Total commerce

(0.5)

(0.7)

−0.6

3.0

2.7

3.7

Note: Labor is measured in man‐years for 1856–1913, in man‐hours for 1924–73.

Table 8.9 Shares of Gross Domestic Product at Current Prices in Commerce, 1924–1973

(Percent)

Sector

1924

1937

1951

1964

1973

Distributive trades

13.5%

13.7%

11.5%

11.5%

10.6%

Insurance, banking, and finance

7.1

8.3

5.3

6.3

8.7

Miscellaneous services

7.1

7.4

6.9

7.7

7.7

Note: The classification by sector in 1924 and 1937 is according to the 1948 SIC and that in 1951 and 1964 is according to the 1958 SIC. Reclassifying the estimates for 1951 onto the 1948 SIC would alter the distributive trades' share in GDP by −0.4 and the miscellaneous services' share by −0.6.

part of miscellaneous services until World War II was private domestic service, but in the postwar period both catering, hotels, and the like and vehicle repair shops and garages were relatively more important.

In the interwar period employment in each of the three industry groups grew rapidly, faster than output, so that labor productivity fell, suggesting diminishing returns and the underemployment of cheap labor. The same conclusion would probably apply if we could measure the growth of TFHP as well as the growth of labor productivity: TFHI in the sector as a whole grew at a rate of 2.2 percent, only slightly below the rate of growth of labor; and TFHP growth was probably negative in (p.245) each industry. It is interesting that in 1873–1913 the distributive trades also experienced a fall in labor productivity when labor was growing rapidly. This provides further evidence of their capacity to absorb excess labor.

The demands of the wartime and reconstruction economies reversed this trend. Between 1937 and 1951 labor left the commerce sector, and especially miscellaneous services, where the rundown in private domestic service, which had low productivity, was most marked.* The result was that productivity rose again to a level in 1951 that was higher than it had been in 1924. In the postwar period labor moved into the distributive trades in 1951–64 and out of them in 1964–73. Insurance, banking, and finance was one of the few industry groups showing a substantial increase in the labor input in 1964–73, greater even than in 1951–64, when it was also high. In the whole sector there was throughout the postwar period a rapid growth in output and by past standards in labor productivity. As has already been noted, this made an important contribution to the faster growth of output and productivity achieved in the postwar period compared with the interwar period.

Transport and Communications

Finally, we consider the growth of output and inputs in the industry groups in the transport and communications sector (Tables 8.10, 8.11). This sector is sharply divided into expanding and declining industries. The two historically largest groups, railways and water transport, declined relative to the economy as a whole. The contribution to GDP of the other groups, road transport, posts and telecommunications, and air and other transport and storage, increased.

There is a marked contrast in both output and input between the declining industries and the growth industries in all four periods shown in Table 8.11.20 In railways and water transport output grew only slowly, and inputs declined or grew only very slightly, whereas in the rest of the sector output and both inputs grew at rates above the average for the economy as a whole. The increase in output growth between 1924–37 and 1951–64 was brought about by a shift in the (p.246)

Table 8.10 Shares of Gross Domestic Product at Current Prices in Transport and Communications, 1938–1970

(Percent)

Sector

1938

1958

1970

Railways

2.9%

1.4%

1.2%

Road transport

2.5

2.7

2.4

Water transport

2.3

1.7

1.3

Posts and telecommunications

1.4

1.7

2.5

Other (including air and storage)

0.1

0.6

1.2

Note: The classification by sector in 1938 is according to the 1948 SIC and that in 1958 is according to the 1958 SIC. An additional 0.1% of GDP is included in transport and communications, probably mostly in road transport, under the 1948 SIC as compared with the 1958 SIC.

Table 8.11 Growth of Output, Labor, Capital, Total Factor Input, and Total Factor Productivity in Transport and Communications, 1924–1973

(Annual percentage growth rates)

Variable and sector

1924–1937

1937–1951

1951–1964

1964–1973

Output:

Railways, water

0.2%

0.4%

0.3%

1.3%

Roads, posts, other

4.0

4.9

3.5

4.6

Total transport and communications

1.5

2.4

2.2

3.5

Labor:

Railways, water

−1.0

−0.1

−2.1

−5.5

Roads, posts, other

2.8

1.1

0.2

0.5

Total transport and communications

0.6

0.5

−0.7

−1.3

Capital:

Railways, water

−0.1

−0.9

0.2

0.7

Roads, posts, other

2.2

2.3

4.3

6.4

Total transport and communications

0.2

−0.4

1.1

2.7

TFHI:a

Railways, water

−0.8

−0.3

−1.4

−3.8

Roads, posts, other

2.7

1.5

1.4

2.2

Total transport and communications

0.5

0.2

−0.2

−0.2

TFHP:

Railways, water

1.0

0.7

1.7

5.1

Roads, post, other

1.3

3.4

2.1

2.4

Total transport and communications

1.0

2.2

2.4

3.7

Source: Basic statistics (Appendix N) and sources quoted there; Deakin & Seward 1969: Table A.1; communication from Deakin.

(a) The same weights for labor and capital were used for the subtotals as for total transport and communications.

(p.247) structure of the sector toward the rapidly growing industries and not by any significant increase in intra‐industry output growth (at a more disaggregated level, of course, there was some increase in output growth, e.g. in water transport). The increase in TFHP growth between the interwar and postwar periods, noted earlier in the chapter, occurred among both declining industries and expanding industries. In 1964–73 the growth of TFP in the declining industries actually exceeded the growth in the expanding ones, as a result of a drastic reduction in labor input, along with some increases in their rate of growth of output.

Notes:

(*) For example, in discussions about “de‐industrialization,” conclusions based on trends in shares in the labor force have often been used as if they betokened trends in shares in output at constant prices.

(*) It is striking that the fall in the share of manufacturing in output at current prices between 1951 (the peak was actually in 1955) and 1973 was substantially greater than the fall in its share in either labor input or capital input. This necessarily implies a fall in the wage rate or the profit rate (or both) in manufacturing relative to the rest of the economy. In fact the fall was entirely on the side of the profit rate. The decline in the relative profit rate in manufacturing was discussed in Chapter 6.

(*) Miscellaneous services includes domestic service and also catering, hotel service, entertainment and sport, hairdressing, laundries, and photography.

(*) However, the use of the man‐years' measure for 1856–1913 has the same effect (disregarding differences between sectors in changes in hours) as would result from taking a man‐hours' measure adjusted for the hours‐offset element in the change in the quality of an hour of labor. This is because the hours‐offset in this period is assumed to have been 100% (see Table 4.7).

() This was also true of the further deceleration within the period 1873–1913. See Appendix L and Table L.2.

(*) These conclusions appear to hold for those parts of the commerce sector, namely the distributive trades and insurance banking, and finance, where employment is not widely used as an indication of output. (Table 8.8 shows that labor productivity growth in the component sectors changed in the same way between periods as TFP growth in total commerce.) The interperiod differences are probably not therefore a statistical illusion.

(*) This needs to be qualified by reference to the possibility discussed in Appendix I, namely that some of the large productivity growth attributed statistically to the service industries and agriculture in the postwar period may have really had its origin in improvements in the quality of intermediate manufactured inputs.

(*) In particular the proportion of building done on contract increased relative to speculative building, a change that is thought to have made planning more difficult and reduced incentives to efficiency.

(*) See Appendix N for a precise definition of our industries in terms of the 1948 and 1958 Standard Industrial Classifications.

(*) In 1937 miscellaneous services employed 3,059,000 persons, of whom about half were in private domestic service. By 1951 only 417,000 were in private domestic service out of a total of 1,945,000 (1948 SIC).

() Moreover, since private cars are not considered part of the road transport industry, the effects of the switch of passenger traffic from the railways to the roads is not fully reflected in the growth of road transport.

(1.) The justification for this procedure is discussed in Appendix I.

(2.) Because output in public and professional services is measured almost entirely by a weighted employment index, improvements in labor productivity resulting either from an increased capital input or from increases in TFP are not reflected in the output index. It follows that it is not possible to estimate TFP growth as the residual output growth, since not only the TFP element but even the effects of the capital input are by definition excluded from output growth.

A separate problem arises in the measurement of TFI in public and professional services. A large proportion of the capital in this sector—schools, hospitals, roads—does not earn any profit. Actual distributive shares in the sector would therefore underweight capital input. Instead, as an approximation, overall distributive shares in GDP are used in the calculation of TFI in this sector in Table 8.3 (though not, for the sake of consistency with aggregates, in the figures for the total intra‐industry change in TFP in the last section of the table).

Output in the ownership‐of‐dwellings sector consists of the services provided by the capital input. The growth in TFP is therefore the decrease in the measured capital‐output ratio (negative in the periods shown in Table 8.3).

Though the figures for output, TFI, and TFP in these sectors are thus not very meaningful, they necessarily enter the aggregates for the economy as a whole. (p.647)

Table to Note 4 Distribution of the Labor Force in Great Britain, 1801–1861

(Percent of the total occupied population)

Census year

Agriculture, forestry, fishing

Manufacture, mining, industry

Trade and transport

Domestic and personal

All other sectors

1801

35.9%

29.7%

11.2%

11.5%

11.8%

1811

33.0

30.2

11.6

11.8

13.3

1821

28.4

38.4

12.1

12.7

8.5

1831

24.6

40.8

12.4

12.6

9.5

1841

22.2

40.5

14.2

14.5

8.5

1851

21.7

42.9

15.8

13.0

6.7

1861

18.7

43.6

16.6

14.3

6.9

We therefore include the figures for output and TFI growth in these sectors for completeness in Table 8.3, and their TFP growth is included in the total intraindustry growth of TFP in the last section of the table. The reservations relating to these data should, however, be borne in mind. It should also be noted that for 1856–1913 the estimation of output in some components of commerce (finance, catering, domestic and miscellaneous services) is based largely on employment indexes, and so the measure of TFP contains a larger arbitrary element in 1856–73 and 1873–1913 than in later periods.

(3.) These results are affected in some cases by changes over time in output‐mix within sectors. In utilities the most capital‐intensive industry, water, constituted a much lower proportion of the sector at the end than at the beginning.

(4.) Satisfactory data for years before 1861 do not exist. However, Deane and Cole (1962: 142) have made rough estimates (including retired persons and the unemployed) for Great Britain from the censuses, as shown in the accompanying table. These estimates suggest that the proportion of the labor force in manufacturing rose very steeply up to 1831, but not very much after that. But even in 1801 it was quite high—far higher than in the typical underdeveloped country of the present day. See Clark 1957: Chapter 9 for comparative data on many countries.

(5.) Falls in 1920–21 and 1929–31 reflected unemployment, which affected manufacturing more than most other sectors. But the census figure for 1931 is also substantially lower than that for 1921, showing that workers had in the meanwhile left the sector.

(6.) The early‐19th‐century data shown in the Table to Note 4 are extremely uncertain, and we cannot be at all confident about the relative magnitude of the increases in the decades between 1801 and 1841.

(7.) The total figure is from Deane & Cole 1962: Table 31; the others are from Booth 1886: Appendixes B.1–B.2; and Matthews 1954: 146, note 1. The effects of excluding domestic workers from the 1841 labor force are even more striking in the case of the UK because of the widespread cottage spinning and weaving industry in Ireland. But since this had largely disappeared by 1881 (Booth 1886: (p.648) 345–46), a more representative picture of changes in the second half of the century is provided by the figures for Great Britain.

(8.) Clark (1957: 500–501) points out that the proportion of the British labor force in mining between 1911 and 1921 was the highest of any country ever, apart from one or two exceptional cases, like South Africa.

(9.) See note 2, above. The pre‐1914 data for commerce are also very arbitrary as measures of output and productivity.

(10.) The decline of 0.2 of a percentage point in the rate of growth of TFP for the economy as a whole between 1856–73 and 1873–1913, shown in Table 8.3, is to be regarded as an underestimate of the unfavorable trend. The decline shown in Table 7.4, last column, is 0.6. Half of the discrepancy arises from a difference in 1856–73 between the growth rates shown by the compromise estimate of GDP (used in Table 7.4) and by the output‐side estimate of GDP (used in Table 8.3 for consistency with the sectoral figures). The other half of the discrepancy arises because Table 7.4 deducts from TFP growth the effects of the significant increase estimated (Table 4.7) to have taken place between 1856–73 and 1873–1913 in the rate of improvement in the quality of labor from sources other than intensity of work (an increase that is implicitly allowed for in Table 8.3; see the footnote to p. 230).

(11.) This fall would probably be still more marked if changes in the degree of utilization of capital were treated as part of changes in TFI, since there is some evidence that the trend rate of increase in the degree of utilization declined in the course of the postwar period (see Chapter 5, especially Table 5.8).

(12.) They did, however, have rather more significance than a random shift in the relative productivity of different industries, since (as we saw in Chapter 5) the rise in the relative cost of building across World War II substantially increased the proportion of national income that had to be saved in order to achieve a given amount of real capital accumulation.

(13.) Productivity in construction was also lower than prewar in the short‐lived boom of 1919–20 (M. Bowley 1967: 128). By 1924, however, productivity appears to have made a substantial recovery.

(14.) Feinstein revises his 1965 sectoral figures for the interwar capital stock, in one case substantially, in his forthcoming work, but he has not prepared revised estimates for subsectors. The old and new estimates of the annual percentage growth rates of the gross capital stock in the relevant sectors are:

Sector

Old

New

Manufacturing

0.9%

1.0%

Commerce

1.3

2.0

Transport and communications

0.2

0.2

In manufacturing, where the difference is small, we use the old figures for subsectors (hence a small discrepancy between Tables 8.3 and 8.7). We also use the old figures for subsectors in transport and communications, where there is no change. For commerce, no data for subsectors are available on either basis, so we confine our attention to labor productivity.

(p.649)

(15.) If weights based on the assumption of a constant net marginal profit rate are used instead, the estimates of TFHP growth in 1924–37 are hardly affected. See Appendix I.

(16.) One of the exceptions, textiles, would disappear on an alternative basis of calculation. See next note.

(17.) The capital‐stock data are not adjusted for the degree of utilization, and so the rate of growth of utilized capital will differ from our estimate of the rate of capital growth to an extent dependent on the change in the degree of utilization. Textiles is probably the only industry where a serious discrepancy between the growth of capital in use and the growth of capital in place existed. The data in Table 5.8, Chapter 5, suggest that there was about 30% more excess capacity in 1951 than in 1964. If allowance is made for this, the 1951–64 growth rates for textiles in Table 8.7 would be –0.3% for capital, –1.7% for TFI, and 1.8% for TFP. Because these figures differ considerably from those in the table, we have omitted textiles from the correlations in Appendix J. However, it should be borne in mind that minor errors may arise from this source in other industries.

(18.) See Dean 1964: 333, Table 1. A considerable margin of error must be attached to the transwar capital‐stock figures, and to the assumption that privately financed investment during the war years had the same industrial distribution as government‐financed investment (ibid., p. 334). Cambridge University Department of Applied Economics 1974 makes a different assumption from Dean's about the proportion of assets that were acquired during the war by the engineering and metal‐using industries and not converted back to peacetime uses. The authors assume much less conversion, and on this assumption lower rates of growth of capital in electrical engineering, mechanical engineering and shipbuilding, vehicles, and other metal industries between 1937 and 1951 would be obtained; they would be in the range of 5.0%–5.5%.

(19.) The IBF category includes the privately owned real estate industry, which controls a very large capital stock. Strictly, of course, this capital should be classified according to the industry that uses it (e.g. distributive trades in the case of shops and warehouses, miscellaneous services in the case of restaurants), and it is because our data do not enable us to do this that we are unable to disaggregate the capital of the whole commerce sector. An additional problem is that buildings may “move” from one industry to another without the change being recorded; a shop may become a bank, or a restaurant a shop.

(20.) The data in Table 8.11 are unsatisfactory in two respects. First, it is not possible to estimate the shares of labor and capital in each industry, and so the shares for the whole sector were used to weight the growth rates of labor and capital in calculating TFI growth for each industry. Second, the output of the part of the road goods industry that is not owned by road haulage contractors is included in our output series, but the vehicles and drivers that produce the output are not in the capital and labor series. Furthermore, as already noted, roads are not included in the capital stock. If allowance were made for these points, it is unlikely that the conclusions in the text would need to be altered very much.