National income accounting and sectoral rates of return on UK risk-bearing capital, 1855-1914 Paper number 03/02 A J Arnold* University of Exeter S McCartney University of Essex Abstract The series in Matthews et al. (in 1982) represent perhaps the most important indicators of the returns to risk capital in the British economy during the period concerned, although suffering from a number of problems concerning their specification, accessibility and reliability. In this paper, a set of redefined, extended and updated series are constructed from published macro-economic data sources which appear to overcome the problems with the earlier series yet result in levels of returns to risk-bearing capital that are much higher than in previous studies. Further research into the basis of macro-economic estimation may therefore be needed before reliable conclusions can be drawn about the profitability of UK commercial sectors, particularly during the important period from the middle of the nineteenth century until the start of the First World War. Keyword : rates of returns, British industry, historical macro-economic estimation _____________________________________________________________________ *Contact Author: Tony Arnold, Professor of Accounting and Business History, School of Business and Economics, University of Exeter, Rennes Drive, Exeter EX4 4PU, Exeter, Devon, England. Email: [email protected]
28
Embed
National income accounting and sectoral rates of return on ... · National income accounting and sectoral rates of return on UK risk-bearing capital, 1855-1914 The rate of return
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
National income accounting and sectoral rates of returnon UK risk-bearing capital, 1855-1914
Paper number 03/02
A J Arnold*University of Exeter
S McCartneyUniversity of Essex
Abstract
The series in Matthews et al. (in 1982) represent perhaps the most importantindicators of the returns to risk capital in the British economy during the periodconcerned, although suffering from a number of problems concerning theirspecification, accessibility and reliability. In this paper, a set of redefined, extendedand updated series are constructed from published macro-economic data sourceswhich appear to overcome the problems with the earlier series yet result in levels ofreturns to risk-bearing capital that are much higher than in previous studies. Furtherresearch into the basis of macro-economic estimation may therefore be needed beforereliable conclusions can be drawn about the profitability of UK commercial sectors,particularly during the important period from the middle of the nineteenth centuryuntil the start of the First World War.
Keyword : rates of returns, British industry, historical macro-economic estimation
_____________________________________________________________________*Contact Author: Tony Arnold, Professor of Accounting and Business History, School ofBusiness and Economics, University of Exeter, Rennes Drive, Exeter EX4 4PU, Exeter,Devon, England. Email: [email protected]
2
ISSN 1473 2904
3
National income accounting and sectoral rates of return on UK risk-bearing capital, 1855-1914
The rate of return on UK risk-bearing capital before 1940, particularly as invested in
commercial firms, is something 'we still know very little about' but which provides a
potentially important input into a number of important and unresolved debates
concerning (alleged) UK economic decline during the second half of the nineteenth
century.1
Despite the importance of these debates, the amount of relevant data that is available
is relatively limited, particularly at a macro-economic level. The series of gross and
net profit rates in the industrial sector for the period 1855-1975 as shown in Matthews
et al. (in 1982) accordingly represent an important indicator of the returns to risk
capital in the British economy, particularly during the period from 1855 to 1914,
when there are few alternative sources of data.
During this period, when the yield on (almost risk-less) consols was between 2.5 and
3.5 per cent, the Matthews et al. series appears to show gross returns of 12-16 per cent
and net returns on industrial capital of 15-20 per cent for much of the period to 1914.2
In this context ‘gross profit’ means profit after all costs apart from depreciation have
been recognised.3 The gross profit rate accordingly excludes cumulative depreciation
from its definition of ‘capital’, whereas the net profit rate shows profits after (annual)
depreciation and capital after the deduction of cumulative depreciation. In principle,
given the lower numerators and denominators used in the net profit rate calculation,
either of the net and the gross profit rates could be the greater; ‘whether the net
measure is higher or lower than the gross measure depends on the length of life of the
capital stock, its age distribution, the rate of growth and so forth’.4 In practice,
however, at least until 1964 the net rate was always the higher of the two.
1 Capie and Webber, Profits and profitability p. 1; see for example, Edelstein, 'Realized rates of return' andOverseas investment; Davis and Huttenback, Mammon.2 See Matthews et al., British economic growth, pp. 184-86; Fig 12.1, p. 346; Mitchell and Deane, Abstract, p. 455.In three years in the period 1865-72 the net returns were in excess of 20%.3 This should not be confused with the standard accounting definition of ‘gross profit’ as the profit after only thedirect costs of sale have been recognised.4 See Matthews et al., British economic growth, p. 186.
4
There are, however, a number of difficulties with the series concerning its limited
specification and the accessibility and reliability of some of the data used in its
preparation. The purpose of this paper is accordingly to examine the series in more
detail, with particular reference to its areas of difficulty and to then put forward and
discuss a modified and updated version of the series showing the rates of return on
UK risk-bearing capital in a number of sectors in the period 1855 to 1914, insofar as
they can be derived from published macro-economic data sources.
I
The Matthews et al. series is presented essentially as a set of graphs, although its
authors also define a number of key years (1856, 1873 and 1913 for the period 1855
to 1914) for which precise figures are given. For these years, underlying data is also
provided or can be calculated or inferred. The series does not appear to have been
published elsewhere.
Although the process is unsatisfactory, the graphs can be read against their x and y
axes to provide a roughly approximated series. The gross series, as set out in Table 1
(col. 17), yields an approximate simple mean across the period 1856-1914 of 13.0 per
cent and the net series, as shown in Table 2 (col. 17), provides a simple mean of 17.6
per cent across the same period. 5 These figures should not, however, be taken as
entirely reliable because the graphs cannot be read with any real accuracy and because
there is evidence (from the data given for the key years) that the net profit graph is not
accurately drawn and that the real returns may be slightly lower than the graphs
suggest.6
Far more precise and reliable figures could be obtained if the series could be recreated
from the original sources used by Matthews et al., although the determination of this
requires the analysis of the various definitions and the connection of the identities
involved with a number of (rather loosely specified) source documents. The profit rate
5 Matthews et al., British economic growth, pp. 184-86, 346.6 The observations from the graph suggest net profit rates of 1856 17.25%, 1873 20% and 1913 17.25% as againstthe rates given for those years of 15.6%, 17.3% and 14.9% respectively.
5
series depend upon three main definitions, of national income, of profit and finally of
capital, each of which needs to be consistent with the others.
As far as national income is concerned, Matthews et al. start with GNP and then
exclude the incomes whose capital, in the foreign, farming and rental sectors, will also
be excluded. The result, described as non-farm, non-rent GDP (NF NR GDP), consists
of GNP less (foreign income + non-farm rent + total farm income).
In the key years this produces (in £m):
1856 1873 1913
Gross net product 667 1172 2343
less foreign income (15) (52) (200)
less non-farm rent (32) (61) (150)
less total farm income (149) (172) (142)
Non-farm, non-rent GDP 471 887 1,8517
Two versions of profit were examined, gross and net. In order to arrive at the gross
profit, the defined version of national income was dichotomised into its overall labour
and property components and the latter, total property income, was then sub-divided
into four constituents, incomes from abroad, rental incomes, farm property incomes
and profits. At the gross level, these components were defined as:
a) income from abroad: net property income from abroad.
b) rent: from dwellings, publicly owned (central and local authority) non-trading
property (inc offices and schools) but NOT agricultural or business premises rents.
c) farm property income: rent of farm land and buildings plus farm profits (usually the
property income of self-employed farmers).
d) profits: all other property income, i.e.
i) the gross trading profits of companies
ii) the gross trading surpluses of the public sector
iii) rent paid by trading concerns and public authorities
7 Matthews et al., British economic growth, in principle use either the estimate of GDP 'most appropriate to theparticular context' or the geometric mean of the measures available. In this particular case the exact basis used isnot apparent; for 1856 for example, Matthews et al. appear (e.g. from p. 164) to use a figure (£667m) which differsfrom the figures set out in Feinstein, National income and in Mitchell, British historical, Tables 4 and 5a for theincome-based and compromise estimates of GNP.
6
iv) the property component of income from non-agricultural self-employment.8
In the second (net) version of the series, 'gross' is converted to 'net' by deducting
depreciation from d) i) and ii).
It should be noted that profits, as defined above, are the residual category such that
any estimation errors in arriving at other components of property incomes will affect
the figures attached to profits.9 There is no way of knowing whether or not this is a
significant consideration. Further, the data to fit these definitions is not readily
available and Matthews et al. accordingly arrive at their measures indirectly, as can be
seen from an examination of data provided throughout chapter 6 of Matthews et al.
relating to the key years 1856, 1873 and 1913.
Gross profits thus equal total property income (foreign income + rent + gross trading
8 See Matthews et al., British economic growth, Table 1 and footnote 26, pp. 164, 185. Annual figures for foreignincome, rent and the gross trading profits of companies are given in Mitchell, British historical, Table NA4.9 The reliability levels for the various components of total property income were defined as +/- 15 to 25% for1855-89 and 1914 and +/- 5 to 15% for 1890-1913; Matthews et al., British economic growth, p. 613.10 From Feinstein, National income.11 These calculations yield proportions of NF NR GDP of 35.6, 38.1 and 33.8% (+/- 0.1%) as in Matthews et al.,British economic growth, Table 6.8 p. 178.
7
The relatively complex route taken in Matthews et al. can, however, be simplified by
using gross trading profits rather than total property incomes as the base, which yields
gross profits as follows, for the key years (in £m):
1856 1873 1913
Gross trading profit 222 424 734
plus commercial rents 15 20 57
less (non-farm) employers’ own imputed wages (27) (52) (126)
less farmers’ incomes (42) (55) (39)
Gross profits (as in Matthews et al. above) 168 337 626
The definitions of gross income used in Matthews et al. are carefully constructed and
correspond closely to the economists’ usual concept of profit, although two of the
components, commercial rents and (non-farm) employers’ own imputed wages, have
been derived from private working papers, rather than from published sources.12
In the case of the third item, capital stock, which converts profit from an absolute
measure to a rate of return, its definition as gross (and then net) reproducible capital
stock at current prices, is relatively straightforward. The collection of the appropriate
data was, however, far more problematic, as the measurement of capital stock is
‘subject to more conceptual and statistical difficulties than the measurement of
income or profits’.13 A good deal of the data was taken from Feinstein (1972),
described as ‘a statistical companion volume’ to Matthews et al., although the tables
relating to investment and capital stock in Feinstein did carry warnings of the
likelihood of higher than usual error rates.14 By 1976, additional investigations had
confirmed the wisdom of these warnings; estimates derived by ‘extrapolating
backwards from 1920 [had] produced an estimate for c. 1860 which is very much too
high’ and Tables 43, 44 and 46 in F72 were accordingly to be ‘regarded as
withdrawn’.15
12 Commercial rents consist of rent (for which Mitchell, British historical, Table NA4, pp. 828-29 provides theseries), less farm rent (ibid., Agr Table 15) less non-farm residential rent (for which there appears to be nopublished series). Non-farm employers’ own imputed wages consist of total employers’ self-wage (for which thereappears to be no published series) less farmers’ self-wage. The latter in turn consists of farm incomes (Feinstein,National income, Table 23) less farmers’ gross profits (for which there appears to be no published series).13 Matthews et al., British economic growth, p.18514 Feinstein, National income. The reliability levels were defined as +/- more than 25% for 1855-1919; Matthewset al., British economic growth, p. 616.15 Feinstein, Statistical tables, p. viii.
8
More work on the period 1760-1860 ‘provided for the first time a reasonably reliable
foundation for estimates of the capital stock in the mid-nineteenth century’, but
exposed an ‘enormous discrepancy' with the earlier figures, particularly concerning
industrial plant and equipment, ships and vehicles and led to a substantial reduction in
the capital stock figures for 1860 from £3380m to £2310m.16 By the time final figures
covering the period 1750-1920 had been published in 1988,17 there had been a series
of ‘major revisions to the statistics relating to investment and the capital stock before
1914’, although not all the sources were published data series.18 The capital stock
figures used in Matthews et al. (in 1982) thus appear to be interim figures taken from
Feinstein’s working papers that were then superseded by the data series subsequently
published in 1988.19
However uncertain and transitory the data, it is the capital stock figures that enable
the two further identities to be determined:20
(a) gross profit rate [gross profits to capital stock]
1856 168/1273 = 13.2%
1873 337/2407 = 14.0%
1913 625/5297 = 11.8% and
(b) the gross capital-output ratio rate [capital stock to gross trading income (i.e. non-
farm, non-rent GDP)]:
1856 1273/471 = 2.7 times
1873 2407/887 = 2.71 times
1913 5297/1851= 2.86 times.
These in turn produce:
16 Feinstein, Capital formation, pp. viii, 78. The earlier estimates were of UK capital stock at 1860 at replacementprices (Feinstein, National income, Table 46) and the later of GB capital stock at 1851-60 replacement costs.17 In Feinstein and Pollard, Studies in capital and Mitchell, British historical, particularly Table NA4, pp. 828-30.18 Mitchell’s sources for 1855-1948 are given as Feinstein, National income, ‘and extensions and revisions theretokindly provided by Professor Feinstein’; Mitchell, British historical, p. 828.19 Matthews et al., British economic growth, p. v. Matthews et al. identified a gross reproducible capital stock (atcurrent prices) in 1856 of £1273m, far lower than the data in Feinstein, National income, Table 46, as withdrawn,and of the same approximate order of magnitude as Feinstein, Capital formation, where only the 1860 data wasgiven (on an illustrative basis). The data used in the Matthews et al. series was based upon 'a preliminary version'of the Feinstein and Pollard data (Studies in capital, p. 260, fn 2) that would be further refined before publicationin 1988.20 Matthews et al., British economic growth, Fig. 6.3, p. 184.
9
(c) the defined proportion of gross profits to gross trading income (i.e. non-farm, non-
rent GDP) as their arithmetic product. Thus in:
1856 GP 168/NF NR GDP 471 = 35.6%; (13.2%* 2.7 =35.6%)
1873 GP 337/NF NR GDP 887 = 38.0%; (14.0%*2.71 =37.9%)
1913 GP 625/NF NR GDP 1851 = 33.8%; (11.8%*2.86 =33.7%).21
Estimation errors in the gross profit rate series (a) will necessarily be offset by
compensating (i.e. opposed) errors in the gross capital-output ratios (b) although only
the former appears to be of any real importance in its own right.
Precisely matching definitions can then be used to compute the net (i.e. post-
depreciation) equivalents of the above.22 Thus, NNP differs from GNP by capital
consumption, GDP differs from GNP by property income from abroad and NDP is
GDP less capital consumption and net trading income is NDP less rent less farm
income.23
More directly, the net share of (net) trading income differs from the gross share of
(gross) trading income, as computed above, in that capital consumption for the year is
deducted from both the numerator and denominator. For the key years:
1856 (168-21)/(471-21) = 147/450 = 32.7%
1873 (337-40)/(887-40) = 297/847 = 35.1%
1913 (625-105)/(1851-105) = 520/1746 = 29.8%.24
This primary ratio can in turn be partitioned into:
(a) net profit rate [net profits to net capital stock, where the net capital stock is (gross)
capital stock less accumulated capital consumption]. Again, in the key years:
1856 147/(1273-331) = 147/942 = 15.6%
1873 297/(2407-690) = 297/1717 = 17.3%
1913 520/(5297-1807) = 520/3490 = 14.9% and
(b) the net capital-output ratio rate [net capital stock to net trading income]:
21 Thus in 1856, for example, the obverse of the 35.6% is the labour share of NF NR GDP (336 - agriculturalwages 60 + non-farm self employment 27) = 303/471 = 64.4%.22 See Matthews et al., British economic growth, Fig 6.4, p. 185.23 The components of farm income are set out in Feinstein, National income, T60, Table 23.24 As in Matthews et al., British economic growth, Table 6.8, p. 178.
10
1856 942/450 = 2.1 times
1873 1717/847 = 2.03 times
1913 3490/1746 = 1.99 times.25
In principle, given the lower numerators and denominators used in the net profit rate
calculation, the net profit rate could be greater or smaller than the gross profit rate,
depending on 'the length of life of the capital stock, its age distribution, the rate of
growth and so forth' 26 although, based on a reading of the two graphs, until 1964 the
net rate was always the higher of the two. In the three key years for which Matthews
et al. provide unequivocal figures this was clearly the case, as follows:
1856 net 15.6% gross 13.2%
1873 17.3% 14.0%
1913 14.9% 11.8%.
II
The gross and net profit series that appear in Matthews et al. provide potentially the
most important macro-based indicator of the returns to risk capital invested in the
industrial sectors of the British economy, particularly during the period from 1855 to
1914.
The series have, however, been seen to suffer from three main deficiencies. Firstly,
they are specified only in graph form, which cannot be readily converted to
unambiguous numerical values. Second, some of the underlying data is contained in
private working papers, rather than in publicly available sources and, finally, the
figures for capital stock used in Matthews et al. have undergone substantial further
revision since that date. In this section, the Matthews et al. methodology, in a
modified, extended and updated form, is used to identify the rates of return on UK
risk capital, as invested in the main sectors of the economy in the period 1855 to
1914, insofar as they can be derived from the most recent, publicly available macro-
economic data sources.
25 As in Matthews et al., British economic growth, Table 6.13, p. 186.26 Matthews et al., British economic growth, Figure 6.3, p. 186. The estimation of depreciation by Feinstein andother authors (for 1855-1920) is discussed in Feinstein, National income, pp. 24, 26-8, 41, 152-3, 156-7.
11
This again requires the definition and identification of relevant data concerning the
three key identities, national income, profit and capital, in each case at the gross and
net levels. As far as gross measures are concerned, the most logical starting point for
national income would appear to be the broadest measure available, GNP. Starting
from this base, the overseas, property rental, agricultural and industrial sectors can
then be viewed as the major alternative investment areas.
There are series in Mitchell for GNP and for the income generated by each of these
sectors (as set out in Table 1, cols. 1 to 6).27 Similarly, Feinstein and Pollard identify
the latest estimates of the gross capital stock employed in each sector (see Table 1,
cols. 7 to 11). Over the period 1855-1910, overseas assets comprised 20.8 per cent,
agricultural assets 23.2 per cent, property assets 30.7 per cent and trading capital 25.3
per cent of the gross capital stock.28 The bringing together of the two sets of data
enables the sectoral gross rates of return (and a composite measure) to be calculated
(see Table 1, cols. 12 to 16 and Figure 1). It is clear that the capital invested in
overseas, agricultural and property rental assets generated similar annual returns,
averaging 4.3, 4.8 and 4.7 per cent respectively. The annual returns on overseas and
property assets were extremely consistent (ranging from 3.7 to 4.9 per cent and from
4.1 to 5.2 per cent respectively; see Table 1, cols. 13 and 15). There was slightly more
variation in the returns on agricultural assets, with an observable low point in 1879 of
3.0 per cent and a distinct upturn in 1902-10 to average returns of 6.3 per cent (see
Table 1, col. 14).
These returns were, however, considerable below those on the (higher-risk)
investment in trading capital, which averaged 18.5 per cent (Table 1, col. 16) or 13
per cent under the earlier Matthews et al. series (see also Figure 1). The greater
variability in the trading capital return series is also apparent; a slight but persistent
downward trend is observable from the peak of 1871, particularly from 1899 until the
start of the First World War. Nonetheless, if returns did indeed range from 15 to 25
27 See Mitchell, British historical, Tables Agr 15 and NA4, pp. 220, 828-30.28 In some cases (overseas assets prior to 1870, and both agricultural and non- agricultural land, the annual figuresare linear interpolations of the ten-yearly figures given in Feinstein and Pollard, Studies in capital.
12
per cent across the period, they would have provided fairly substantial risk premia, at
least until the effects of capital consumption are taken into account.
The series in Mitchell for national income (GNP less capital consumption) as set out
in Table 2 (col. 1) and its subdivision similarly provide sectoral net income figures (as
set out in Table 2, cols. 2 to 6).29 Feinstein and Pollard again identify the latest
estimates of gross capital stock and of capital consumption which, taken together,
provide the net capital stock figures (as set out in Table 2, cols. 7 to 11).30
Bringing the income and capital figures together enables the sectoral net rates of
return (and a composite measure) to be calculated (see Table 2, cols. 12 to 16). Figure
2 shows the various rates of return, including the earlier Matthews et al. series
(although the rate of return on overseas capital has been omitted as being identical
with its gross equivalent).
Capital invested in agricultural and property rental assets generated net returns
averaging 5.0 and 5.2 per cent respectively. Net returns in these two sectors were
slightly above their gross equivalents but otherwise followed very similar paths. The
net returns on agricultural assets fell to a low point of 2.9 per cent in 1879. Until 1899
they remained below 6 per cent in every year, although they were then to average 6.5
per cent through to 1910, with a peak of over 7 per cent in 1906-08 (see Table 2, col.
14).31 Annual net returns on property assets across the period1855-1910 ranged from
4.5 to 5.9 per cent, increasing very slightly from 1895 (see Table 2, col. 15). The rate
of return on overseas assets and the gross and net rates on agricultural assets and
property are also shown in graph form in Figure 3.
29 See Mitchell, British historical (Tables NA4 and Agr 15, pp. 220, 828-30).30 The figures for depreciation of UK capital stock in Feinstein and Pollard for 1855-1914 (1988, Table I, pp. 427-8) are identical with, and presumably the source for the figures for capital consumption in Mitchell 1988 (TableNA4, pp. 828-30). The former work does not provide sectoral figures for depreciation, and these have beencalculated using the information therein, as explained in the notes to Table 2 below, for two sectors, agricultureand property: depreciation on trading capital is a residual.31 Feinstein and Pollard (1988) provide only decennial estimates for land values, and annual figures were derivedby linear interpolation. It was not considered reasonable to do this for the period 1910-1920, and accordingly theseries for those sectors in which land is a constituent (agricultural and property rental assets) end in 1910.
13
III
The main focus of this paper has, however, been the returns on UK trading capital and
the gross and net series, both according to Matthews et al. and in a revised and
updated form. These are shown in graphical form in Figure 4 (see also Table 1, cols.
16 and 17; Table 2, cols. 16 and 17). It is apparent that both series provide rates of
return that are higher in every year in their net (of depreciation) form than in their
gross.32 Equally, the new net and gross series provide rates of return that are
substantially above the equivalent earlier series in Matthews et al. Although there are
differences in the definitions used in the two sets of series, the effects of these
differences on the computed results is very slight; in large measure the higher rates in
the new series reflect the substantial reductions in capital stock figures between the
interim calculations used in Matthews et al. (in 1982) and their final form in Feinstein
and Pollard (in 1988).
The revised net series shows returns on UK trading capital averaging 25.3 per cent
across the period 1855 to 1910. The overall range was from 19.8 to 34.4 per cent with
some evidence of decline over time, although one masked to some extent by cyclical
variations. Nonetheless, the returns of 30 per cent or more experienced in 1862-64
and 1870-72 (with a peak of 34.4 per cent in 1871) were never to be repeated and
average returns from 1900-10 of 21.6 per cent were somewhat below the levels of
earlier parts of the period.
Because of a number of difficulties concerning the specification of, and the
accessibility and reliability of some of the data used in, the earlier series, this paper
has put forward a modified, extended and updated version of the series showing the
rates of return on UK risk-bearing capital in a number of sectors in the period 1855 to
1914, insofar as they can be derived from published macro-economic data sources.
The primary result, the series for net rates of return on UK trading capital, averaging
25.3 per cent across the period 1855 to 1910 (and 26.1 per cent from 1855 to 1899)
32 See previous note.
14
has been derived from the most reputable and authoritative published macro-
economic series.
If the results are to be believed, investment in British industry across the second half
of the nineteenth century provided returns on risk-bearing capital that were a
substantial multiple of the 'riskless' rate available on government consols and would
then seem to have represented a 'golden era' in the history of British industrial
investment. The level of returns indicated is, however, sufficiently high as to invite
some scepticism and to suggest the need for further research into macro-economic
data series before any firm conclusions can be drawn about the returns to risk-bearing
capital across the important period from the middle of the nineteenth century until the
start of the First World War.
15
Sources
Capie F and A Webber, Profits and profitability in British banking, 1870-1939,Discussion Paper of the Centre for Banking and International Finance, City UniversityLondon (1985).
Davis L E and R A Huttenback, Mammon and the pursuit of Empire: the politicaleconomy of British imperialism 1860-1912, (Cambridge, 1986).
Edelstein M, ‘Realized rates of return on UK home and overseas portfolio investmentin the age of High Imperialism’, Explorations in Economic History, 13 (1976), pp.283-329.
Edelstein M, Overseas investment in the age of High Imperialism: the UnitedKingdom, 1850-1914, (New York, 1982).
The Editor of the Statist, ‘Wholesale Prices of Commodities in 1916’, 80 (1917),Journal of the Royal Statistical Society, pp. 289-309.
Feinstein C H, ‘Income and investment in the United Kingdom, 1856-1914’,Economic Journal, 71 (1961) pp. 367-85
Feinstein C H, National income, expenditure and output of the UK 1855-1965(Cambridge, 1972).
Feinstein C H, Statistical tables of national income, expenditure and output of the UK,1855-1965 (Cambridge, 1976).
Feinstein C H, ‘Capital formation in Great Britain’ in Cambridge Economic Historyof Europe, VII, The industrial economies: capital, labour and enterprise (Cambridge,1978), pp. 28-96.
Feinstein C H and S Pollard (eds), Studies in capital formation in the UK, 1750-1920(Oxford, 1988).
Jones, G T, Increasing Return (Cambridge, 1933).
Maiwald, K, ‘An Index of Building Costs in the United Kingdom, 1845-1938’,Economic History Review, ns. 7 (1954), pp. 187-203.
Matthews R C O, C H Feinstein and J C Odling-Smee, British economic growth1856-1973 (Oxford, 1982).
Mitchell B R, British historical statistics (Cambridge, 1988)
Mitchell B R and P Deane, Abstract of British historical statistics (Cambridge, 1962).
Sauerbeck, A, ‘Prices of Commodities and the Precious Metals’, Journal of theStatistical Society of London, 49 (1886), pp. 581-648.
Sauerbeck, A, ‘Prices of Commodities During the Last Seven Years’, Journal of theRoyal Statistical Society, 56, pp. 215-254.
Sauerbeck, A, ‘Prices of Commodities in 1908’, Journal of the Royal StatisticalSociety, 72, pp. 68-80.
16
Table 1: Gross Returns on Capital in the UK 1855-1914
(1) = GNP at current prices(2) = Total property income(3) = Net overseas income(4) = Total agricultural property income(5) = Property (non-agricultural) rental income(6) = Gross profits of industrial (and commercial & public sector) enterprises(7) = Gross capital stock including overseas assets(8) = Overseas assets(9) = Agricultural assets(10) = Property (non-agricultural) assets(11) = Trading capital (and public sector assets)(12) = Return on overall capital including overseas assets(13) = Return on overseas assets(14) = Return on agricultural assets(15) = Return on property (non-agricultural) assets(16) = Return on trading capital (and public sector assets)(17) = Return on trading capital (gross profit rate) per Matthews et al (1982)
Sources
Col 1 Mitchell (1988, NA 4, pp. 828-30).Col 2 Col 3 + Col 4 + Col 5 + Col 6.Col 3 Mitchell (1988, NA 4, pp. 828-30).Col 4 Mitchell (1988, Agr 15, p. 220).Cols 5-6 Mitchell (1988, Agr 15, p. 220; NA 4, pp. 828-30).Col 7 Col 8 + Col 9 + Col 10 + Col 11.Col 8 Feinstein and Pollard (1988, Table XVII, pp. 462-3; Table XVIII, p. 464;
Table XX, pp. 466-7) give figures for overseas assets at current prices only forselected years (usually decennial), but also give figures for annual netinvestment from 1870. In the above table, stock in intervening years has beencalculated by cumulating annual net investment figures at constant prices andmultiplying the resultant by the implied index used to convert constant tocurrent prices. Before 1870 linear interpolation has been used.
Col 9 Feinstein and Pollard (1988, Table 18.5, p. 400; Table IV, pp. 433-34;for land values, the annual figures are linear interpolations of ten-yearlyfigures).
Col 10 Feinstein and Pollard (1988, Table 18.5, p. 400; Tables IV-V, pp. 433-36; forland values, the annual figures are linear interpolations of ten-yearly figures).
Col 11 Feinstein and Pollard (1988, Tables IV-V, pp. 433-6).Col 12 Col 2/Col 7.Col 13 Col 3/Col 8.Col 14 Col 4/Col 9.Col 15 Col 5/Col 10.Col 16 Col 6/Col 11.Col 17 Matthews et al. (1982, Fig. 6.3, p. 184).
20
Figure 1: Gross returns on capital in the UK, 1855-1914
0
5
10
15
20
25
30
1855
1857
1859
1861
1863
1865
1867
1869
1871
1873
1875
1877
1879
1881
1883
1885
1887
1889
1891
1893
1895
1897
1899
1901
1903
1905
1907
1909
1911
1913
Return on trading capital, Matthews et al. Return on trading capital Return on overseas assets
Return on agricultural capital Return on property Return on overall capital
21
Table 2: Net Returns on Capital in the UK 1855-1914
(1) = National Income(2) = Net property income = National Income less Employment Income(3) = Net overseas income(4) = Total net agricultural property income(5) = Net property (non-agricultural) rental income(6) = Net profits of industrial (and commercial & public sector) enterprises(7) = Net capital stock inc overseas assets(8) = Overseas assets(9) = Net agricultural assets(10) = Net property (non-agricultural) assets(11) = Net trading capital (and public sector assets)(12) = Net return on overall capital including overseas assets(13) = Return on overseas assets(14) = Net return on agricultural assets(15) = Net return on property (non-agricultural) assets(16) = Net return on trading capital (and public sector assets)(17) = Net return on trading capital (net profit rate) per Matthews et al. (1982)
Sources
Cols 1 to 3 Mitchell (1988, NA 4, pp. 828-30).Col 4 This is gross agricultural property income (Table 1 Col 4) less depreciation on
agricultural capital other than land. Depreciation has been computed byconstructing annual figures at constant prices for the components of non-landcapital, i.e. buildings and equipment, from the decennial figures given byFeinstein and Pollard in Table 10.2 (p. 272) and Table 10.4 (p. 278)respectively, and converting these into current prices by the indices used inthat work. For buildings the index is given in Table XXIII (pp. 470-1); aspecial index was used for equipment, which is not in Feinstein and Pollardand apparently has not been published. This index combines the index for‘General plant and machinery’ (Table XXIII, pp. 470-1) with an index forhorse-drawn vehicles, with weightings of 3:1. This latter is apparentlyunpublished, but is ‘constructed by combining wages of carpenters and joiners(weight 45) with price indices for iron (35) and timber (20)’ (p. 327). The firstis taken from Jones (1933, pp. 261-3). The index for iron is the series ‘Iron A’used by Maiwald (1954), except that sources for figures for 1856-8 and 1901-14 could not be traced, and so the price of iron bars taken from Sauerbeck(1886; 1913) and ‘The Editor of the Statist’ (1917) has been used to completethe index for this paper. The index for timber is constructed from figures for‘Sawn or Split’ timber in Sauerbeck (1886; 1893;1909) and ‘The Editor of theStatist’ (1917). See Feinstein and Pollard (1988, pp. 279, 327 and 310, fn. 4).
Col 5 This is Gross Property (Non-agricultural) Rental Income (Table 1 Col 5) lessdepreciation on Dwellings (excluding farmhouses) and Industrial and
25
Commercial buildings. Depreciation has been calculated in constant pricesfrom the figures for capital formation and capital stock in Table X (pp. 446-7),Table XII (pp. 450-1) and Table XIV (pp. 454-5), and then converted tocurrent prices using the index of building costs computed by Maiwald (1954)which is reproduced in Feinstein and Pollard (rebased to 1900=100) as theprice index of 'Buildings including dwellings' in Table XXIII (pp. 470-1). SeeFeinstein and Pollard (1988, pp. 292, 309, 390).
Col 6 Col 2 - [Col 3 + Col 4 + Col 5].Col 7 Col 8 + Col 9 + Col 10 + Col 11.Col 8 See Table 1 Col 8.Col 9 Feinstein and Pollard (1988, Table 18.5, p. 400; Table VI, pp. 437-8; in the
former table (of farmland) ten-yearly figures only are given: annual figures arelinearly interpolated until 1910.)
Col 10 Feinstein and Pollard (1988, Table 18.5, p. 400; Table VII, pp. 439-40; in the former table (of non-farm land) ten-yearly figures only are given: annual figures are linearly interpolated until 1910.)
Col 11 Feinstein and Pollard (1988, Tables VI-VII, pp. 437-40): Total Net Stock ofDomestic Reproducible Fixed Assets, less Agricultural assets, Dwellings andIndustrial and Commercial buildings.
Col 12 Col 2/Col 7.Col 13 Col 3/Col 8.Col 14 Col 4/Col 9.Col 15 Col 5/Col 10.Col 16 Col 6/Col 11.Col 17 Matthews et al. (1982, Fig. 6.4, p. 185).
26
Figure 2: Net returns on capital in the UK, 1855-1914
0
5
10
15
20
25
30
35
40
1855
1857
1859
1861
1863
1865
1867
1869
1871
1873
1875
1877
1879
1881
1883
1885
1887
1889
1891
1893
1895
1897
1899
1901
1903
1905
1907
1909
1911
1913
Return on trading capital, Mathews et al. Return on trading capital Return on agricultural capital
Return on property Return on overall capital
27
Figure 3: Sectoral returns in the UK, 1855-1914
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
1855
1857
1859
1861
1863
1865
1867
1869
1871
1873
1875
1877
1879
1881
1883
1885
1887
1889
1891
1893
1895
1897
1899
1901
1903
1905
1907
1909
1911
1913
Return on overseas assets Gross return on agric assets Net return on agric assets
Gross return on property Net return on property
28
Figure 4: Gross and net returns on UK trading capital, 1855-1914
0
5
10
15
20
25
30
35
40
1855
1857
1859
1861
1863
1865
1867
1869
1871
1873
1875
1877
1879
1881
1883
1885
1887
1889
1891
1893
1895
1897
1899
1901
1903
1905
1907
1909
1911
1913
Gross Net Gross, Matthews et al Net, Matthews et al