^{Note (1) page 45} A. D. Smith, D. M. W. N. Hitchens and S. W. Davies, Inter national Industrial Productivity: a comparison of Britain, America and Germany, National Institute of Economic and Social Research,Occasional Paper XXXIV, Cambridge University Press, 1982. A short summary article also appeared in the National In stitute Economic Review, August 1982.

^{Note (2) page 45} To integrate the results of this study with previous sectoral international productivity comparisons undertaken in this series. retailing has been defined on the basis of the 1968 British Standard Industrial Classification, comprising Minimum List Headings 820 and 821. This procedure excludes car dealers, filling stations hotels and restaurants etc, which are classified as retailing in the US, but as ‘miscellaneous services’ (Order XXVI) in Britain.

^{Note (3) page 45} M. Hall, J. Knapp and C. Winsten, Distribution in Greai Britain and North America, Oxford University Press, 1961.

^{Note (4) page 45} For statistical reasons (see footnote(2)) 1975 was the year for which more reliable comparisons could be made of comparative sectoral sizes.

^{Note (1) page 46} Basic British and American data were derived, respectively, from: Business Statistics Office, Department of Industry, Report on the Census of Distribution and Other Services 1971, Business Monitor, HMSO, and Bureau of the Census, Department of Com merce, 1972 Census of Retail Trade. The 1971 British distribution census related only to Great Britain. For convenience it has been assumed that the picture of the British retail trade which it portrays is applicable to the whole of the UK. The international price indices were based on those presented in I. B. Kravis, Zoltan Kenessy, Alan Heston and Robert Summers, A System of Inter national Comparisons of Gross Product and Purchasing Power, John Hopkins University Press, Baltimore and London, 1975.

^{Note (2) page 46} See Appendix 1.

^{Note (3) page 46} See Appendix 1.

^{Note (4) page 46} International output and productivity comparisons are usually more favourable to the domestic economy when measured by ‘other country’ rather than by ‘own country’ weights.

^{Note (1) page 47} K. D. George, Productivity in Distribution, University of Cambridge, Department of Applied Economics, Occasional Papers 8, Cambridge University Press, 1966, p.17.

^{Note (2) page 47} Aggregation requires, first, the derivation of an output (in this case, sales) per head figure for the whole of retailing in each country as a weighted average of sales per head in each component retail trade, with each trade's share of the retail labour force as weights. At a second stage, these two retailing per capita sales figures are compared between Britain and America.

^{Note (3) page 47} US, Monthly Retail Trade Report, January 1974.

^{Note (1) page 48} See Smith, Hitchens and Davies, op.cit.

^{Note (2) page 48} A degree of interaction exists between some of the factors considered. For instance there is an overlap between the growth of multiples and the development of self service techniques.

^{Note (3) page 48} Including difficulties which arise when matching individual British and American trades. For example the door-to-door distribution of milk in Britain is classified with grocers, but in the US, to the extent that it exists, with direct selling. Consequently it would be wrong to attach undue significance to relatively small differences in the percentages presented in the table.

^{Note (1) page 49} The wider US definition includes, it should be recalled, garages etc. and catering. Whilst American data are not available for the narrower retail definition on the same basis as the British data presented in Appendix table A4, it is possible to check the extent to which the distribution of American retailing between States differs using, respectively, the narrower and broader definition of the sector. The test reveals that the two geographical patterns are almost identical.

^{Note (1) page 50} Including cooperatives among the multiples. This distinction between independents and multiples differs from one sometimes used in British censuses where the line is drawn between organisations operating fewer than 10 shops and those with ten shops or more.

^{Note (2) page 50} Five of the original 34 trades have been excluded. The concept of a shop has little relevance to mail order, vending machines and some of the residual categories. Additionally, variety shops are omitted because they vary from very large Woolworth type stores to small general village shops, in proportions which differ signifi cantly between Britain and America. Department stores have been excluded because in both countries their average size is so many magnitudes greater than any of the other retail trades that their inclusion could distort the results.

^{Note (3) page 50} This finding is by no means novel. Such an association was detected at an early stage by Hall et al. (Distribution in Great Britain and North America, op. cit) and also between towns in Great Britain, by George (Productivity in Distribution, op. cit).

^{Note (1) page 51} The median shop size measured by persons engaged was about the same in the two countries.

^{Note (2) page 51} There are sound reasons when comparing comparative US/UK shop size and productivity levels for leaving out of account the ‘mixed retailing’ category denoted in table 5. In the US small general village stores feature quite prominently in this ‘mixed’ category, whilst in the UK large department stores are relatively more important. As a result of this imperfection in the ‘matching’ of individual shop categories, the average size of American ‘mixed’ shops—in sharp contrast to any other category—is only half the British size, and in the case of independent mixed shops, less than a tenth of the British size.

^{Note (3) page 51} Comparable data by size of shop is not available for Britain in 1977.

^{Note (1) page 52} Data relating to firms have been compiled on the basis of the UK and US census definitions. In both censuses efforts were made to classify different types of shop (grocery, footwear, etc.) oper ated by a single firm to the appropriate kind of trade.

^{Note (2) page 52} i.e. on the assumptions, respectively, that British firms were the same size as American firms and American firms the same size as British firms. That the two measures can yield rather different answers reflects discrepancies in the slopes of the British and American productivity curves over the crucial ranges. Another simplification in the analysis is that both British and American firms are assumed to be concentrated at their respective average sizes.

^{Note (3) page 52} In the case of some individual trades—food, apparel and household goods—the corresponding size effect ranges from 10 to 28 per cent, 12 to 46 per cent and 13 to 30 per cent.

^{Note (1) page 53} This is by no means a perfect measure of per capita national income since, in particular, it fails to take account of savings and of expenditures that are not channelled through the retail trade.

^{Note (2) page 53} The British retail censuses contained quite detailed area data for 1971, none for 1977. The data presented in the 1972 US census related to selected trades only.

^{Note (3) page 53} These differences in coverage are due to the preponderance of ‘general’ type shops in the table—groceries, department stores, chemists (drug stores)—the sales of which bulk large in the US. Comparative sales area data are also available for ‘variety’ and general stores, but this category has been excluded due to the size distortions which arise where shops such as Woolworths are lumped together with general village stores.

^{Note (4) page 53} It should be noted that this productivity differential differs little from the comparable ratio of 2.36 obtained for the whole retail sector.

^{Note (1) page 54} K. D. George, Productivity in Distribution, op. cit. Based on a multiple correlation of differences in sales per head in British towns. George found productivity to be positively associated with tightness of the local labour market, shop size, proportions of multiples and independents and regional differences in per capita income.

^{Note (1) page 55} Hall et. al., Distribution in Great Britain and North America, op. cit.; and Kravis et. al., A System of International Comparisons of Gross Product and Purchasing Power, op. cit.

^{Note (2) page 55} UK, Census of Distribution and Other Services, 1971, op. cit.; and US, 1972, Census of Retail Trade op. cit. The 1971 British distribution census was the last comprehensive one. More recent British censuses (though not American) would be less suited to the kind of analysis undertaken in this study—even if contemporary US/UK price ratios were available.

^{Note (3) page 55} Certain non-shop activities are covered: mail order, credit retailing and vending machines. Market stalls and mobile shops, accounting for 0.9 per cent of British retail sales, have been ex cluded, as have electricity and gas board showrooms which in the British 1968 SIC are classified with utilities. The complete specification and matching of these retail trades—which in the case of the US is based on the US 1972 Standard Industrial Classification—is available from the authors at the National Institute.

^{Note (1) page 56} For both Britain and the US, table 1 is based on establishment (i.e. shop) tables in the respective censuses, so that in the case of both countries ‘central organisations’ are excluded.

^{Note (2) page 56} On the conventional assumption of two working proprietors per partnership.

^{Note (3) page 56} Distribution in Great Britain and North America, op. cit., p.58.

^{Note (4) page 56} Census of Distribution 1971, Part 1, table 2.

^{Note (5) page 56} Distribution in Great Britain and North America, op. cit., p.58.

^{Note (6) page 56} John L. Carey and Phyllis Flohr Otto, ‘Output per unit of labour input in the retail food store industry’, Monthly Labor Review, January 1977.

^{Note (7) page 56} US Department of Commerce, Survey of Current Business, July 1973.

^{Note (8) page 56} To check this result the percentage of part-time retail workers in 1972 was estimated by an alternative means using a 1981 retail part-time percentage (supplied directly by the US Bureau of Labor Statistics, Washington) in conjunction with a 1972 part-time per centage for the whole of the American distributive sector. The outcome was an estimated 34 per cent of part-time workers in retailing.

^{Note (9) page 56} A System of International Comparisons, op. cit., p. 213.

^{Note (10) page 56} This procedure yields two measures of comparative real sales, for which, effectively, the British pattern of relative prices of the items in question, and the American pattern of relative prices, are used in turn as weights for the quantities involved.

^{Note (11) page 56} US Monthly Labor Review, and UK Department of Employ ment Gazette.

^{Note (1) page 57} Since the complete census of 1971 only annual sample censuses of the retail trade have been taken.

^{Note (2) page 57} The American labour force indices relate to changes in num bers of employees and of estimated working proprietors.

^{Note (3) page 57} Excise taxes on gasoline, liquor, tobacco, etc., paid by the manufacturer or wholesaler and passed on to the retailer and consumer were included in both censuses.

^{Note (4) page 57} US Department of Commerce, Bureau of the Census, Retail Trade 1978.