¹ British Industry 1700–1950. By Hoffmann, W.. Translated by Chaloner, W. H. and Henderson, W. O.. Oxford: Basil Blackwell, 1955. Pp. xxiii, 330. 35s.Google Scholar

² Hoffmann, W., “Ein Index der industriellen Produktion für Grossbritannien seit dem 18. Jahrhundert,” Weltwirtschaftliches Archiv, XL (1934, ii), 383–98Google Scholar; Wachstum und Wachstumsformen der englischen Industriewirtschaft von 1700 bis zur Gegenwart (Jena, 1940).Google Scholar

³ Professor Ashton, T. S., An Economic History of England: the Eighteenth Century (London, 1955), P. 124Google Scholar, uses the exports of the Northern field as a guide. He uses them to demonstrate that there was a substantial increase in production. But that is all; and he states that it is his opinion that the inland coal fields were growing more rapidly.

See also, Gayer, A. D., Rostow, W. W., and Schwartz, A. J., The Growth and Fluctuation of the British Economy, I790–1850 (Oxford, 1953), p. 693.Google Scholar

⁴ The index of engineering production is founded on much worse material and constructed with very much less care than the one recently disowned by Professor Cairncross, A. K., Home and Foreign Investment, 1870–1914 (Cambridge, 1953.), p. 158.Google Scholar

⁵ See the statistics given in “The Third Fiscal Blue Book,” British and Foreign Trade and Industry 1834–1908, in Parliamentary Papers, 1909, CII (Cd. 4954), pp. 806–7, Table 44.Google Scholar

⁶ See Heaton, H., The Yorkshire Woollen and Worsted Industries (Oxford, 1920), p. 258Google Scholar, and Clapham, J. H., “The Transference of the Worsted Industry from Norfolk to the West Riding,” Economic journal, XX (1910), 195–210Google Scholar, for the rise of the Yorkshire industry at the expense of the rest of the country.

R. P. Beckinsale, “West of England Textile Industry” (unpublished D.Phil, thesis, Oxford Univ.), pp. 86–87, states that the West of England industry stagnated completely in the closing years of the eighteenth century because of competition from Yorkshire.

Miss Phyllis Deane, of the Department of Applied Economics, Cambridge University, tells me that after studying contemporary estimates she has no doubt at all that Yorkshire's share did increase. Miss Deane also makes the further important point that the series used by Hoffmann, broad and narrow cloths, may cause him to overestimate the growth of output in Yorkshire if, as seems probable, the proportionate importance of these cloths increased there.

⁷ The published histories of the Yorkshire trade are not adequate for the period after 1850, but in what is available there is general agreement on the story of Bradford's boom and subsequent depression. See, for example, the articles that appeared in the Bradford Observer in November and December 1883. For Bradford in the period before 1860, see the chapter by E. Sigsworth in Fay, C. R., Round About Industrial Britain 1830–1860 (Toronto, 1952), pp. 114–31Google Scholar. Another good source on the boom phase is Baines, E., “The Woollen Manufactures of England,” a paper read to the British Association in 1858 and revised and reprinted in Baines, Yorkshire Past and Present (London, 1871).Google Scholar

⁸ Shannon, H. A., “Bricks—A Trade Index, 1785–1849,” Economica, n.s., I (1934), 300–18.CrossRefGoogle Scholar

⁹ A good deal more information on the period after 1870, both about housing and the investment of local authorities, is to be found in A. K. Cairncross, Home and Foreign Investment, pp. 142–58. The most recently published paper on housebuilding is Weber, B., “A New Index of Residential Construction, 1838–1950,” Scottish Journal of Political Economy, II (1955), 104–32.Google Scholar But inclusion of these sources, though it would improve the Hoffmann estimates, would still leave them a very far from perfect guide to total building activity.

¹⁰ This problem of quality changes is all-pervasive. Perhaps even the quality changes of so apparently homogeneous a commodity as coal ought not to be neglected. It may be that the decline in the rate of growth of tonnage mined was offset by increasing proportions of washed coal and of steam coal—both more valuable types.

¹¹ Fairly reliable weighting is possible for 1907 because the first census of production was held in that year. Final Report of the First Census of Production of the United Kingdom, in Parliamentary Papers, 1912–1913, CIX (Cd. 6320), 1.

¹² Perhaps a simple, if rather lengthy, numerical example may make this point clear.

Let us suppose that an economy has only two sectors, A and B. Annual production data are available for sector A, whose output grows from 100 units in year O, to 200 in year XV, to 400 in year XXX—a percentage rate of growth of 5 per cent per year. Annual data are not available for B, but it is known that in years O, XV, and XXX its output was 100 units; and this knowledge is used in the reweighting of the constituents of the aggregate index in those years.

For the first period, industry A will be weighted 50 in the total index, and if the index of aggregate production in year O is 100, it will be 200 in year XV. For the second period, XV to XXX, industry A will be weighted 66⅔, and the index of production in this period will increase from 150 in XV to 300 in XXX; when the weights are revised again in XXX, A will be given a new weighting of 4/5 and the index for XXX will be 250. Within each period, therefore, the procedure adopted would produce an overestimate of the rate of growth of the whole economy (roughly 5 per cent per year instead of the true rate of under 3 per cent in the first period and under 3½ per cent in the second). There is still the effect of the revision of the weights to be considered; but before it can be considered we have also to apply the “smoothing” procedure to our results. If we decide to retain year O as our base year, the smoothing procedure will require the multiplication of all indices for the period XV to XXX by that multiple which is needed to make the second estimate of year XV coincide with thefirst, i.e., by 4/3. We finish with a series in which the index rises from 100 in O to 200 in XV to 400 in XXX. The overestimates of the short series have thus been perpetuated in the final, longer, series because the “smoothing” procedure has an effect that is equal and opposite to the effect of the periodic changing of the total weights of the included series.

¹³ In pointing out this serious error, I do not wish to imply that there is an easy alternative method of constructing an index. Any attempt to make the total of weights have operational significance raises the further problem, “How ought the total of weights to be calculated?” Hoffmann's weights, which are based on employment, take into account changes in productivity as well as changes in production and would still lead to disputable estimates of the over-all rate of growth. Thus, for example, if all the industries included in the index increased both in production and in productivity by 5 per cent per year while all other industries completely stagnated, the relative weightings of the two groups would remain the same with the consequence that the estimate of the over-all rate of growth would also be 5 per cent.

The answer to this problem is, of course, that if we know only the relative employments of the two sectors and the change of production in one sector, we cannot know the over-all change. The best we can do is to assume that the sector we do know about is representative of the whole economy—an assumption that in many cases will be highly disputable.

¹⁴ Hoffmann, W., “The Growth of Industrial Production in Great Britain: A Quantitative Survey,” Economic History Review, n.s., II (1949), 162–80.CrossRefGoogle Scholar It cannot be claimed that the topic receives adequate treatment in the article; but it, together with the next topic, are at least mentioned there.

¹⁵ See n. 9 above.