EH.net is owned and operated by the Economic History Association with the support of other sponsoring organizations.
Market Services and the Productivity Race, 1850-2000: British Performance in International Perspective
Published by EH.NET (June 2007)
Stephen Broadberry, Market Services and the Productivity Race, 1850-2000: British Performance in International Perspective. Cambridge: Cambridge University Press, 2006. xvii + 409 pp. $95 (cloth), ISBN: 0-521-86718-5.
Reviewed for EH.NET by Alex Field, Department of Economics, Santa Clara University.
As I was preparing this review of Stephen Broadberry's latest book, my daughter leaned over to see what I was currently reading. She glanced quickly at the cover, shook her head and said, in a voice tinged with pity, "Dad, that just looks incredibly boring." Then, catching sight of the word 'productivity' in the title, and knowing something of my own recent work, she added, apologetically, "Oh, did you write this?"
Such are the occupational hazards facing those of us who study productivity history, hardy but sensitive souls whose pulse can quicken in the presence of minute differences in rates of TFP increase. For a general audience, this book will indeed be slow going. But for those interested in the motors of economic growth and development, considered particularly from a comparative perspective, this work is a notable achievement. In it, Broadberry brings together research on the comparative productivity performance since the 1870s of Britain, the United States, and Germany. The emphasis is on Britain, and Broadberry's main theme is the necessity of focusing on the service sector, not just or even principally manufacturing, in understanding Britain's relative decline. Differential trends in service sector productivity, not industry, he argues, explain most of the movement in the comparative productivity performance of Britain's aggregate economy.
How can this be? The basic story is that U.S. productivity in manufacturing in 1870 was already about fifty percent higher than it was in Britain, and the same was true 120 years later. Britain was ahead of the U.S. and Germany in 1870 not because of its lead in industry but because of its lead in services. By 1890, however, the U.S. had caught up and subsequently forged ahead in services, widening its lead in the sector up through the Second World War. Thus it was differential productivity trends in services, not any fundamental alterations in trends in manufacturing, that account for most of the overtaking and surpassing of Britain by the United States in aggregate labor productivity.
Critical to the story is the Chandlerian theme of the transition from low volume, high margin to high volume low margin operations, particularly in transportation, communication, and distribution (Chandler, 1977). The models for such transition originated in the United States, and it was the more rapid diffusion of such organizational methods in the U.S. that enabled it to overtake Britain in services and in the overall economy by 1890. But not all service sectors were equally amenable to this transition, explaining why Britain's relative decline was not more severe. With respect to Germany, Britain's continuing lead in services relative to Germany accounted for its ability to maintain an overall productivity advantage with respect to that country through about 1960.
Modern business enterprise originated in the U.S. in the railroad and telegraph sectors, spreading to wholesale and retail distribution, before being adopted in some sectors of manufacturing. Britain's relative productivity decline was most notable in transport and communication and distribution, whereas the country retained a strong position in finance, a sector which offered unfavorable terrain (at least until recently) for the adoption of high volume low margin operations. Broadberry attributes the slower service sector productivity growth in Britain to labor resistance to work intensification combined with the fact that many services, unlike manufactures, weren't tradeable, and therefore didn't face the same competitive pressures. For industrial products, poor productivity performance eventually meant they were replaced by imports and the sector shrank. For nontradeable services, and those protected by non-tariff regulatory barriers, poor productivity performance could persist, dragging down comparative productivity performance for the entire economy. Still, Broadberry is at pains to present a variegated view of the British service sector, emphasizing areas where its comparative advantages in organizing through networks rather than internal firm hierarchies continued to serve it well. Germany's success in overtaking Britain after 1960 is explicable in terms of a different structure of labor relations and patterns of human capital formation after the Second World War.
The book has three sections. The first is concerned principally with laying out the argument and presenting the data. The second focuses on broad explanations for the comparative productivity trends (I did not find that the theoretical model in chapter 5, based on the work of Broadberry and Ghosal, added much). The third section presents a series of detailed historical/empirical chapters dissecting the performance of different parts of the British service sector in different time periods, carrying the story forward through the 1990s. Taken together this material provides a compelling case that it is both possible and necessary to focus on the service sector in understanding productivity history at the national level and in comparative perspective. The data for services may not always be quite as good as they are for industry, but there is much more here than one might imagine, and it is unworkable and unjustifiable to restrict our attention to manufacturing simply because "the light is better there."
One of the difficulties with the book and the way the data are presented, however, is the melding of a narrative structure focusing on the one hand on the history of British productivity growth and on the other hand on its comparative productivity performance with respect to the U.S. and Germany in different time periods. From an historical perspective, one is principally interested in what drove aggregate productivity forward, both in individual countries and in the world as a whole. There are of course also reasons for measuring comparative productivity performance at different times, but excessive focus on these numbers risks obscuring important aspects of the process of economic growth. For example, compare two time periods in which there is apparently no change in comparative performance. It makes a big historical difference whether this is because there was no change in levels in either country or because levels in both countries increased at the same rate. To his credit, Broadberry provides detailed time series data on the evolution of levels for the individual countries. Still, because the narrative is principally couched in terms of the explanation of relative decline, Broadberry finds himself repeatedly compelled to emphasize that even in periods in which the country may have performed poorly in comparison with other countries, the absolute growth rate of labor productivity in Britain was high by world historical standards.
The point I am making is separate from but in a sense related to the important methodological issues which have motivated the exchanges between Broadberry on the one hand and Ward and Devereux on the other. To calculate comparative productivity ratios one chooses a benchmark year, calculates value added per worker or per hour in the domestic currencies of each country, and then converts the value added in the two countries to a common metric using a PPP adjusted exchange rate. Using domestic rates of economic growth, one then projects forward and backwards. Because of index number and other problems, however, growth rates from one PPP based comparative productivity benchmark will not always or necessarily produce a later or earlier comparative productivity measure consistent with a later or earlier PPP based benchmark. Broadberry has tried to address this problem by choosing 1937 as a benchmark (unlike Maddison, who used 1990), and by using other benchmarks as a check on the projections forward and backwards using domestic growth rates. He is persuaded the methodology is defensible, and the remaining discrepancies tolerable, but he will probably not satisfy all of his critics. Those seeking more detailed discussion of these problems, which also bedevil work on contemporary economies, should consult, for example, Ward and Devereux (2003).
There are some aspects of Broadberry's treatment of U.S. productivity history with which I differ. For example, he endorses Abramovitz and David on the importance of capital accumulation rather than TFP in explaining trends in U.S. labor productivity growth in the last third of the nineteenth century (p. 109). But TFP growth in the U.S. private domestic economy was over 1 percent per year from the early 1870s through 1906, lower of course than rates experienced in the second quarter of the twentieth century but certainly respectable and indeed substantially higher than during a comparable period of the twentieth century (Field 2008, forthcoming).
Secondly, Broadberry correctly identifies very strong labor productivity and TFP growth in transportation and communication in the 1930s as part of what underlay the continuing advance of U.S. service sector productivity and the consequent widening of the US/UK productivity gap during the Depression years. But this was not necessarily a period of increasing "industrialization" of transport services as Broadberry suggests (p. 12). The introduction of modern business enterprise in railroads occurred in the last third of the nineteenth century. The growth sector within transportation in the 1930s was trucking, a sector comprised of small, non-hierarchically organized companies. What is striking about the 1930s in the U.S. is not only the extraordinarily high TFP growth in trucking (a rapidly growing sector), but the relatively strong advance in the railroad sector, which, in contrast with the 1920s, was experiencing capital shallowing. Since railroads were still such a large part of the economy, this mattered (see Field, 2006). As Broadberry's narrative indicates (pp. 230-235), nothing comparable happened in railroads in Britain. Greater emphasis on the importance for the United States of the Depression build out of the surface road network, and the role of street and highway spending in the U.S. in generating spillovers in transportation would have made the narrative historically richer (Field, 2003).
There is also some tendency to be too dismissive of trends in industry in explaining the huge gap opened up between the U.S. and Britain by the 1950s. The driver of aggregate trends in comparative productivity, Broadberry asserts, are trends within the respective service sectors. But his own data show that the widening of the US/UK productivity differential between 1870 and 1960 was more marked in industry (250.4/153.6) than it was in services (137.7/85.9) (Table 2.1). Table 2.1 includes data for industry — construction and mining as well as manufacturing — and it is true that the increase in manufacturing alone, although substantial, was not as dramatic as in the service sector as a whole. My point is that whereas the surge in U.S. service sector productivity certainly played a very important part in widening the productivity lead opened up by the U.S. at mid-century, trends in industry and in manufacturing also contributed substantially. Broadberry's larger thesis is sustained, however, because whereas by 1990 faster productivity growth in British industry and manufacturing had restored the US/UK advantage to roughly where it had been in 1870, the same cannot be said for services.
Finally somewhat more emphasis, particularly in the US/UK comparisons, on changing sectoral shares in explaining aggregate productivity growth would have been useful. For example (p. 28) "over the long run, Britain was overtaken at the aggregate level because of a loss of labour productivity leadership in services." It would be fairer to say that Britain was overtaken because of a loss in productivity leadership in services combined with the very sharp growth in the share of services, particularly in the United States. (The UK-German analysis does place considerable emphasis on the retention of a large agricultural sector in German and its implications in terms of a smaller service sector.)
These are small bones to pick. Broadberry has written an impressive book that will interest economic historians and students of economic growth, particularly those focusing on Britain.
Chandler, Alfred D. 1977. The Visible Hand: The Managerial Revolution in American Business. Cambridge: Harvard University Press.
Field, Alexander J. 1987. "Modern Business Enterprise as a Capital-Saving Innovation," Journal of Economic History 47 (June 1987): 473-85.
Field, Alexander J. 2003. "The Most Technologically Progressive Decade of the Century," American Economic Review 93 (September): 1399-1414.
Field, Alexander J. 2006. "Technological Change and U.S. Economic Growth in the Interwar Years," Journal of Economic History¬¬ 66 (March 2006): 203-36.
Field, Alexander J. 2008. "U.S. Economic Growth in the Gilded Age," Journal of Macroeconomics 29 (prepared for submission to a special issue).
Ward, Marianne and John Devereux. 2003. "Measuring British Decline: Direct vs. Long Span Income Measures," Journal of Economic History 63: 826-51.
Alex Field is the Michel and Mary Orradre Professor of Economics at Santa Clara University. His current research interests include U.S. productivity history and the implications of evolutionary theory for the behavioral sciences.