Convergence of income levels has preoccupied economic research at least since Adam Smith. The starting point should be to gather reliable and comparable data on income and factor price differences between countries. However, as Svante Prado—a promising young Swedish economic historian—shows in his published Ph.D. thesis, previous international comparisons have not paid enough attention to whether various series are congruent, and could even rest on deficient data. Prado's examines the development of factor costs in 1860–1950, by comparing Sweden, a poor country in 1860, with the United States and Britain. He focuses on real wages, the wage-rental ratio, labor productivity, and total factor productivity (TFP). And although Sweden is a small country, it provides an interesting case, because of its rapid growth during the studied period and because of the high international standard of its historical data. As Prado points out (p. 140), while Swedish Industrial Statistics has been published annually since 1858, the British Census of Production was only published beginning in 1907, and then only every second year.

In Chapter 2, Prado criticizes Jeffrey Williamson's (“The Evolution of Global Labor Markets Since 1850: Background Evidence and Hypothesis.” Explorations in Economic History 32, no. 2 [1995]: 141–96) view on the convergence of real wages after 1850. Scandinavian countries accounted for most of the convergence found in Williamson's sample. According to Prado, Williamson uses a flawed series for Swedish real wages that only represents a small part of the employed. Prado presents a more comprehensive series, implying that some of Sweden's catch up with the United States and the United Kingdom slips away. Furthermore, the wage rate for the United States is exaggerated by Williamson, and there is a clear difference in the development of wages of skilled and unskilled labor. While according to Williamson the U.S./Swedish wage ratio dropped from 410 to 157 percent between 1870 and 1914, another series displays a smaller decline, while a third reveals no trend at all. These are, clearly, huge disparities. Unfortunately, the reader is somewhat left in the dark as to which series are the best to employ. While Prado writes that a final affirmation “has to await future research efforts” (p. 53), it would have been nice if he had argued more forcefully for the series that he thinks best suits international comparisons.

Chapter 3 is written together with Jan Bohlin, and deliberates convergence in the wage-rental ratio between the Old and New World in the decades around 1900. As is well known, declining transport costs allowed labor to move to the New World, while grain imports in Europe put downward pressure on land prices. By presenting a new series of land values, the chapter reveals that the wage-rental ratio in Sweden before World War I did not increase as much as previously thought, at least by comparison with other countries. Sweden's development was in fact similar to other protectionist countries in Europe, although the chapter makes it clear that land prices are difficult to compare, since they are significantly affected by variations in property rights, taxes, and locational rents.

Chapter 4 compares the development of labor productivity in manufacturing. Prado contends that Stephen Broadberry's view that convergence in GDP per capita mainly took place through the reallocation of employment towards activities with high value added per labor unit needs to be modified as far as Sweden is concerned. Prado finds find that there was considerable Swedish convergence in the labor productivity of manufacturing around 1900 and in the 1930s and 1940s. He also makes a significant contribution to Swedish historical national accounts by presenting new estimates of output and employment in manufacturing, which is based on more realistic methods to fill in gaps in official statistics than has been the case with previous research. Unfortunately, the book only makes available the data for manufacturing as whole. Prado should remedy this deficiency by making his new series at branch level available to international researchers, for example, through an online publication.

Chapter 5 ponders the development of total productivity in Swedish manufacturing. Prado concludes that while real cost reduction (or increased TFP) was uneven around 1900 (so-called mushroom growth), a more uniform pattern was embarked on afterwards (so-called yeast growth). This shift in the growth process points towards the establishment of a General Purpose Technology, which Prado suggests could be electricity. Since the concept of TFP is difficult to interpret—despite its wide acceptance among economic historians—the outcome of this chapter is less clear cut, although it has other merits. Technological change (as Prado points out) is not the same as TFP growth, since technical change is usually embedded in capital. Furthermore, it is well known that high capital growth rates could entail low or even negative growth rates in TFP, but this would not necessarily imply technological stagnation—for instance, if increased used of capital is accompanied by declining real costs (the price deflated by some more comprehensive price index, for example, CPI) of physical capital. How to best measure aggregate technological change is still largely an unresolved problem.

The book's greatest strength is the empirical analysis, but it also makes important contributions to recent theoretical discussions on convergence, which Prado carefully relates to in all the chapters. In this way, the book places itself into the forefront of the economic-historical debate. However, I miss a summary chapter at the end that could bring the different investigations together—for example, how the development of wages, labor productivity, and TFP could be related to each other—even if the first chapter partly fulfills that role. Overall, the book exemplifies how more in-depth analysis of the data of various countries, such as the one Prado provides for Sweden, is needed to settle the debate on convergence. Good economic analysis can only rest on good historical data collection.