EXPLORATIONS IN ECONOMIC HISTORY 32, 141-196 (1995) The Evolution of Global Labor Markets since 1830: Background Evidence and Hypotheses* Jeffrey G. Williamson Harvard University Debate over the economic convergence of currently industrialized nations has suffered a number of shortcomings. First, the underlying data base has typically been limited to Angus Maddison's GNP and GNP per worker-hour estimates. This paper offers a new data base, purchasing-power-parity-adjusted real wage rates for unskilled labor. Second, the debate has typically focused on end points from the 19th century to the present, paying little attention to differential behavior in four distinct regimes: 1830 to midcentury, midcentury to World War 1, the interwar decades, and the post-World War II experience. Third, with some recent exceptions, the search for explanations has focused primarily on technological advance, while ignoring the potential role of global factor and commodity market integration (and disintegration). The new real wage data base confirms some old stylized facts and offers some new ones. It also points out how these four regimes differed. They differed enough to suggest that different explanations will be nec- * This paper is part of a larger project on "World Labor Market Integration and Disintegration over the Past Century" supported by the National Science Foundation, Grants SES-9021951 and SBR-92-23002. I am grateful for the excellent research assistance of Toni Estevadeordal, George Fan, Carter Hood, and, especially, Bill Collins, Boris Simkovich, and Alan Taylor. In addition, generous help has been supplied by many colleagues in constructing the national real wage data base underlying the project. They are Bob Allen, George Boyer, Luis Catao, Sebastian Coll, Roberto Cortes-Conde, Fergus D'Arcy, Jan DeVries, Stefano Fenoaltea, Ian McLean, Cormac O'Grada, Kevin O'Rourke, Stefan Op-pers, Gerry delta Paolera, Leandro Prados de la Escosura, Peter Scholliers, Pierre Sicsic, Gabriel Tortella-Casares, Annelies Vermaas, Glenn Withers, and Vera Zamagni. The comments of Moe Abramovitz, Brad DeLong, Claudia Goldin, Tim Hatton, Peter Lindert, Boris Simkovich, Alan Taylor, and participants at the NBER/DAE Conference (March 14, 1992) are acknowledged with pleasure. This paper is a revision of an earlier version entitled "The Evolution of Global Labor Markets in the First and Second World Since 1830: Background Evidence and Hypotheses" (October 1991). The data base is revised, and thus the present version takes precedence over the earlier version. A number of appendices are omitted from this version in the interests of cost containment. They are available upon request from the author at 216 Littauer, Harvard University, Cambridge, MA 02138 (617) 495-2438 or e-mail jwilliam@husc4.harvard.edu). 141 0014-4983/95 $6.00 Copyright © 1995 by Academic Press. Inc. All rights of reproduction in any form reserved. 142 JEFFREY G. WILLIAMSON essary to account for the convergence over the past century and a half. © 1995 Academic Press, Inc. I. WHERE DOES THE RECENT CONVERGENCE DEBATE LEAVE US? Four strands of literature initiated in the 1950s and 1960s seem to be converging on better explanations of long run growth: empirically based country studies led by the cliometricians of the late 1950s and early 1960s; debates in comparative economic history about latecomers (Gerschenk-ron, 1952), about the demise of British leadership, and about the rise and fall of America's industrial supremacy (Abramovitz, 1986; Baumol, 1986); the empirical sources of growth tradition launched by Abramovitz (1956), Solow (1957), Kendrick (1961), and Denison (1962); and the formal models of the 1960s which have recently blossomed into the new "endogenous theories of growth." This paper reports a new data base for 15 European Old World and overseas New World countries covering the past century and a half. My hope is that this new data base will contribute to the new agenda designed to achieve the lofty goals set by those pioneers in the 1950s and 1960s. While I view this paper as only a provisional start of a longer project, it offers a novel data base which confirms some old facts and uncovers some new ones which I believe should be central in guiding our search for new and better theories of growth which are well grounded in history. These four strands of literature almost always use GNP per capita or per worker hour as the dependent variable. This paper uses instead purchasing-power-parity-adjusted real wage rates per worker (typically urban unskilled) as the dependent variable. While I do not argue that the real wage data base used in this paper is necessarily superior to the GNP estimates, it can be argued that factor prices generally, and real wages specifically, are the better yardstick for assessing sources of long run convergence. First, labor's marginal and average products differ, and all the more so as other factors of production, like natural resources and human capital, are added to conventional capital in the analysis. GNP per worker is, after all, nothing more than a sum of per unit factor returns weighted by specific factor endowments per worker. Any model of convergence must offer an explanation for the forces driving factor return (or price) equalization. Furthermore, factor price equalization rarely implies that all factor prices in the rich country will fall relative to the poor. For example, suppose the initial rich countries are New World land-abundant and labor-scarce, while the initial poor countries are Old World land-scarce and labor-abundant. While convergence implies that wages in the Old World catch up to those in the New, it also implies that land rents in the New EVOLUTION OF GLOBAL LABOR MARKETS 143 World catch up to those in the Old.1 All of this suggests that wage convergence is likely to be far more dramatic than GNP per worker convergence. In short, GNP per worker is much too coarse an index to understand adequately the forces of convergence. It seems to me highly desirable to study the convergence of wage rates separately so as to isolate convergent behavior in national labor markets. Of course, what I have just said about wages could also be said about land rents, skill premia, and the rates of return to capital, a data base currently under construction. Second, the GNP and the wage deflators differ. In a world of very incomplete commodity price equalization, the difference may matter a great deal, especially since laborers heavily consume commodities which are resource-intensive (e.g., food and dwelling space), items which are most expensive to move internationally, a statement that holds true with greater strength the farther back in history we look. These influences are likely to diminish over time for three reasons: transport costs decline, thus contributing to commodity price equalization (O'Rourke and Williamson, 1992); the ratio of value added to crude material inputs increases even for resource-intensive products; and Engel effects assure that resource-intensive products diminish as a share of workers' budgets (the share spent on food declines). Third, the aggregate labor participation rate is likely to differ greatly between countries and over time in an environment of migration and differential rates of population growth, driving a wedge between per capita and per worker indices. This relationship is likely to matter especially in any comparison involving historical growth patterns in high-wage New World countries and low-wage Old World countries. After all, native labor supplies are responsive to conditions of labor scarcity and surplus, and these tend to influence child dependency rates and labor participation rates. Thus, high-wage New World countries have higher fertility and lower child mortality rates, both serving to raise the child dependency rate in contrast with those of low-wage Old World countries.2 These distinctions may have mattered more in the 19th than in the 20th century. Thus, to the extent that per capita rather than per worker-hour figures typically characterize 19th century estimates, real wage rate data are likely 1 The ratios of wage rates to farm land rents or land values converge dramatically between 1870 and 1913. In the New World, where land was abundant and labor scarce, the wage/rental ratio plunges downward from high levels across the period. In the Old World where free trade prevailed (e.g., Britain and Scandinavia), and where land was scarce and labor abundant, the wage/rental ratio surges upward across the period. In the Old World where the protectionist response was strong (e.g., Germany, France, and Spain), the wage/rental ratio was more stable but still rose. See O'Rourke et at. (1993). 2 A survey of this evidence and its impact on domestic savings behavior, and thus on capital flows, can be found in Taylor (1993) and Taylor and Williamson (1994). 144 JEFFREY G. WILLIAMSON to be an improvement in tracking productivity for those important decades. Fourth, even when expressed in per worker-hour terms, GNP can be viewed as a coarse index of long run performance, although it is used almost exclusively in what has come to be known as the "convergence" literature. Indeed, as Wolff (1991, pp. 568-569) has recently reminded us, almost everyone engaged in the long run convergence debate uses Maddison's (1982) output and hours worked data for the pre-World War II years. By his own admission, Maddison's pioneering 19th century GNP and man-hour estimates are inevitably based on partial data, often constructed by backward projection from assumed average growth rates. Perhaps in recognition of that fact, scholars using Maddison's data have selected only benchmark years separated by a decade or even longer. While it may suffer other flaws, my annual real wage data base does not, at least, suffer this disadvantage. While it has little to do with the use of GNP measures, previous empirical studies have tended to sleight important historical episodes and important country observations which are likely to enrich our understanding of the underlying forces producing economic convergence and divergence between nations. For example, when Baumol (1986) and his critics (DeLong, 1988) use the century 1870-1979 to explore the convergence hypothesis, they tend to ignore the intervening observations and focus on the end points. Even in Baumol's more recent work with collaborators (Baumol et al., 1989, Chap. 5), the analysis tends to focus on end points with little attention to pre-World War II epochs. This seems a pity, since there may be quite different growth regimes within the century which are likely to offer additional insight into the growth process.3 There is another reason why I have made the effort to develop this real wage data base, and it should be apparent from the title. I am especially interested in the evolution of global labor markets since 1830, and rarely is there any mention of the role of international migrations and the development of integrated global labor markets in the convergence accounts.4 Nor is there any mention of the role of international commodity markets and the venerable factor price equalization theorem.5 Certainly there is no mention of either influence in Baumol and in the stream of literature that his work has sparked. Nor is there any significant mention of such forces in the empirical applications of the "new endogenous growth 3 A point with which Abramovitz (1986) agrees. 4 A recent survey by Hatton and Williamson (1992b) points out this imbalance, and two empirical papers have tried to fill the gap, one on Anglo-America (O'Rourke et al. 1993) and one on Ireland (Boyer et al. 1993). 5 However, O'Rourke and Williamson (1994) have shown recently that commodity price equalization between Britain and the United States had a profound impact on real wage convergence between the two countries 1870-1913. EVOLUTION OF GLOBAL LABOR MARKETS 145 theories." While this may be a useful simplification for the post-World War II decades, it is unlikely to be very helpful in understanding the late 19th century decades of free migration, or in understanding their contrast with the interwar decades of restricted migrations. This paper deals next in Section II with a brief description of how the data base was constructed. The details are offered in three copious appendices; thus we are free to turn to the critical issue in Section III: what does the real wage history since 1830 tell us about international convergence? I conclude with an agenda in Section IV. II. CONSTRUCTING THE REAL WAGE DATA BASE I have been able to construct time series on real wage rates over the past century and a half for 15 countries, 4 in the overseas New World and 11 in the Old World. While I am still looking to augment the sample,6 I believe these 15 countries include most of the ones that matter. As I pointed out in the introduction, most participants in the recent convergence debate have relied on Maddison's (1982) GDP data. Thus, for example, Baumol and his collaborators (1989) use Maddison's sample of 16 countries. The Maddison sample includes 4 countries which are excluded here: Austria, Finland, Japan, and Switzerland. Japan could have been included in my sample, but since it was not a significant participant in international commodity and factor markets until the turn of the century, I have chosen to omit it. I have not been able to find adequate real wage evidence for Austria, Finland, and Switzerland, but at least two of these are very small countries and, hopefully, they are adequately represented in our sample by others like them. There are three very important additions to our sample that are missing from Maddison's: Argentina, Ireland, and Spain. (Brazil and Portugal are also in the Appendices, but not in the sample. See footnote 6.) Ireland offers an important observation given its enormous post-Famine emigration rates and the remarkable speed with which it became integrated into the global English-speaking labor market.7 And the addition of Argentina and Spain 6 Actually, the Appendices include information on Brazil and Portugal, countries excluded from the sample of 15 used throughout the paper. They were constructed too late to be included in the analysis. 7 The best statement on the integration of Ireland into a global English-speaking labor market was by O'Rourke (1989, 1990), although he and two collaborators have recently given even more empirical content to that process (Boyer et al,, 1994). New World Old World Australia Argentina Canada United States Belgium Denmark France Germany Great Britain Ireland Italy Netherlands Norway Spain Sweden 146 JEFFREY G. WILLIAMSON (joining Italy) makes it possible to say something about New World and Old World connections along Latin lines, as well as to say something about the integration or segmentation of global labor markets between the North and South within both the New and Old World. The data base is built along the following lines. First, I construct nominal wage time series from sources that country experts tend to favor (Appendix 1). These refer to wage rates, and, prior to 1914, they refer to unskilled labor with very few exceptions. Wherever possible, they measure hourly, daily, or weekly wages, although for a few countries early in their wage histories I had no other option but to use monthly wages. Only for Belgium do I use annual earnings. The goal is to focus on the cost of labor per unit of time, and to control as well as possible for the work/leisure choice. For the pre-WWI period, I try to restrict the focus wherever possible to unskilled wage rates; until the interwar period, the use of average wages in, say, manufacturing would be a mistake since the skill premium varied widely over time and across countries,8 as did the skill mix. Since these problems seem to have become far less serious by the 1950s, and since the data are readily available in standard ILO and OECD publications, I use manufacturing hourly wages for the post-World War II regime. Furthermore, and with few exceptions, the unskilled wage rates refer to city labor rather than farm labor. Like the skill premium, wage gaps between farm and city vary widely over time9 thus I have made every effort to stick with urban unskilled wage rates in what follows (a goal violated between 1860 and 1913 only for Belgium, France, Italy, and the Netherlands). Sources and methods underlying the nominal wage time series for each of the 15 countries are described in Appendix 1. Second, I construct and apply cost of living deflators to the nominal wage time series. The cost of living figures refer to urban areas and, where possible, are derived from budget weights of the low-wage unskilled. Typically, the cost of living indices are comprehensive, including detail on foods, dwelling rents, fuel, light, and clothing. The cost of living series for each of the 15 countries are described in Appendix 1; like nominal wages, they come from sources that country experts tend to favor. The third step is to convert these national real wages into internationally comparable units of measurement by establishing benchmarks, years where wages are quoted for comparable jobs and comparable workers, typically for the unskilled in the building trades, and deflated by purchasing-power-parity (PPP) price indices for comparable market baskets 8 The literature is large, but see most recently Williamson and Lindert (1980), Williamson (1985), and Allen (1990). 9 Some comparative evidence on this point can be found in three papers by Hatton and myself (1991a, 1991b, 1992a). EVOLUTION OF GLOBAL LABOR MARKETS 147 (Appendix 3). Such benchmarks are constructed at three points in time: the decade or so prior to World War I, based on the British Board of Trade, augmented by information reported in U.S. Senate documents published shortly before the turn of the century; the late 1930s, based on those rich ILO Yearbook sources which appear with great detail in either the 1927 or 1938 issue; and the post-World War II period, based on the purchasing-power-parity deflators reported by Kravis and his associates (1978, 1982). These sources make it possible to construct comparable real wages across countries at each of these three benchmarks, to which the national real wage time series are then linked (Appendix 2). The PPP real wage benchmarks serve to define three distinct periods and I have made little effort to extend any of the national real wage time series across the breaks separating them, namely, the two World Wars. This point deserves stress. Since country-specific relative prices changed markedly over this century and a half, real wage cardinal rankings projected forward (based on the national time series) from the pre-World War I benchmark to 1938 would almost certainly yield a real wage cardinal ranking different from that implied by the 1938 ILO data themselves. Similarly, real wage cardinal rankings projected backward from a 1938 benchmark to the pre-World War I years would also differ from that implied by the Board of Trade benchmark. The problems are somewhat less severe when a 1927 benchmark is used for the interwar period, as I do here, but they do not disappear. These index number problems are certainly not unique to these real wage data: indeed, they even beset the famous PENN data set generated by Kravis and his associates for the post-World War II period. Thus, I make little effort to leap over these benchmarks in the time series analysis which follows.10 Of course, there have been others who have offered comparative real wage estimates even prior to World War I, and even for samples larger than mine. Perhaps the best known are Mulhall's (1896) estimates of "annual earnings per inhabitant" for 1894. His earnings figures are not PPP-adjusted, but rather converted into common currency units by use of quoted exchange rates. Furthermore, they are average annual earnings, not wage rates for comparable work. When Mulhall's estimates are regressed on those reported here, we find (t statistics in parentheses) Mulhall, 1894 = 19.50 + 0.64 Williamson, 1894 R2 = 0.71. (1.76) (5.37) 10 In contrast, Maddison uses 1970 relative price benchmarks to project his GNP statistics backward into the 19th century. While my real wage data set may have other flaws, at least it offers three benchmarks, not just one. 148 JEFFREY G. WILLIAMSON Thus, his figures tend to overstate the real wages of poor countries and understate those of rich, and the differences are often large. The new national real wage time series covering the past century and a half are presented in the Appendix 2. For the pre-World War I period, the national real wage observations are standardized by setting the British wage equal to 100 in 1905; for the interwar period, the British wage is set equal to 100 in 1927; and for the post-World War II period, the British wage is set equal to 100 in 1975." To illustrate how the data in Appendix Table A2.1 can be used comparatively, U.S. real wages were more than four times those of Italy in 1895 (151/37 = 4.08) while Irish real wages were 71% of British in 1870 (49/69 = 0.71). The real wage data base in Appendix 2 is, of course, only as good as the underlying national time series. If Canadian wages for the late 19th century are really of such doubtful quality as some suggest, then the fact that Table A2.1 has Canada's real wage almost 30% above that of the United States in 1913 will also be doubtful. Indeed, some of these series are undergoing additional repairs, while others, like Portugal and Brazil, can now be added. This then should be viewed as a provisional data base. If nothing else, my explicit use of them here for comparative purposes may provoke others to improve them. III. THE REAL WAGE EVIDENCE: 1830-1988 Four Regimes since 1830 The evidence presented in this paper suggests that there have been four distinct global labor market regimes since 1830. The first was associated with early industrialization in the Old World, settlement in the New World, modest international migrations, high transport costs on commodity trade, and, for the most part, barriers to trade. The regime covers the four decades from 1830 to 1869. Although the sample is relatively small (7 countries in the first half of the regime, rising to 11 at the end), it suggests nonetheless that disequilibrium characterized the first two decades, when real wages diverged sharply, reaching peak dispersion in the mid 1840s. The second covers the period 1870 to 1913, the classic dating for what Argentineans call the "belle epoque," what North Americans call the post-Civil War age of industrialization and free international migration, what the English call the great Victorian boom amidst an age of high imperialism, and what most of us are taught was a classic world boom under free trade and the gold standard.12 If the decade and a half prior " Throughout this paper, I invoke poetic license and refer interchangeably to England and Britain. In fact, the figures refer to England up to 1913 and Great Britain thereafter. Ireland, of course, is always treated separately. 12 And thus what Maddison (1982, p. 92) calls the "liberal" phase. EVOLUTION OF GLOBAL LABOR MARKETS 149 Dispersion 0.4 i- 0.1 0.2 0.3 0 1830 1835 1840 1845 1850 Year 1855 1860 1865 1870 Fig. 1. International real wage dispersion, 1830-1869. to 1870 is included, it was by far the most dramatic period of real wage convergence since 1830, including the better-known convergence of the post-World War II era. Most of the convergence, however, was completed by the turn of the century. The third covers the two World Wars and the interwar period when world commodity and factor markets break down.13 Between 1914 and 1938, measures of real wage dispersion do not fall at all, implying that these two decades served to halt a spectacular real wage convergence which had been at work for six decades. World War II served to increase real wage disparities so much that our measure of global labor market (dis)integration retreats back to the levels of the late 1870s. The fourth is the most studied regime, 1946-1988. The levels of global labor market integration which characterized the 1920s were not regained until the mid 1960s, after which the great pre-World War I convergence is resumed following a half-century pause. As we shall see, not only do each of these regimes exhibit different convergence behavior, but the components of that behavior also differ. They may be sufficiently different, in fact, to warrant different explanations. Disequilibrium and the Industrial Revolution: Pre-1870 Figure 1 documents real wage dispersion between 1830 and 1869. The summary statistic C(N) plotted there, the unweighted coefficient of vari- 11 And thus what Maddison (1982, p. 92) calls the "beggar-my-neighbour" phase. 150 JEFFREY G. WILLIAMSON ation (where N is the sample size), has been used extensively in the convergence debate (e.g., Wolff, 1991, p. 565). Based on a sample of seven countries for which data are available (France, Great Britain, Ireland, the Netherlands, Spain, Sweden, and the United States), C(7) rises from about 0.14 in 1830 to about 0.28 in 1846. (Each of the C(N) series is extended as a dashed line in Fig. 1 when a new C(N + X) series, augmented by X new observations, is added.) That is, the index of real wage dispersion doubles over the two decades. As Appendix Table A2.1 shows, this global labor market disequilibrium was driven primarily by a wage boom in the United States and a wage slump in Spain and the Netherlands. Relative to the European leader of the pack (Britain), Ireland and France held their own while Sweden lost some ground. While events in Europe were contributing to this disequilibrium phase up to the mid 1840s, the sharp rise in C(7) is driven primarily by New World success: the United States increased her real wage advantage over England from 40% in 1830 to 83% in 1846.14 This is a surprising finding given that the British standard of living debate points to these decades as unambiguous evidence supporting the optimists' position. It is also surprising that American "exceptionalists" do not make more of the relative wage performance on this side of the Atlantic. While I do not have similar real wage data for, say, Canada or Australia, the American evidence certainly suggests that the global labor market disequilibrium was being driven primarily by wage gaps between Old World and New, not by some more general process associated with Ger-schenkron-like industrial revolutionary leader versus latecomer dynamics centered in Europe. Having said as much, we must remember an inherent selectivity bias determining this small sample of seven countries: since the availability of real wage data is correlated with the beginnings of modern economic growth, the sample is biased (since it excludes so many poor, latecomers) and the dispersion within Europe may have risen far more than these figures show. By 1854, the sample is augmented to include Australia, Belgium, Germany, and Norway, and C(U) tells the same story as C(7): the dispersion statistic falls from the series' start in 1854. When the sample is augmented still further in 1864 to include another important New World country, Argentina, the summary statistic behaves the same way (not shown in Fig. 1). Figure 1 suggests a secular turning point somewhere in the mid 1840s. It appears to initiate a long run convergence in international real wages that extends into the 1870-1913 regime. While a good share of the real wage convergence from the mid 1840s to 1865 can be explained by the " While per capita incomes were, apparently, still quite a bit higher in England, real wages were a different story. EVOLUTION OF GLOBAL LABOR MARKETS 151 well-known collapse in American wages during the Civil War (DeCanio and Mokyr, 1977; Williamson, 1974), the story is more general than that (see the Appendices for the American evidence and for the rest of what follows in this paper). First, the American post-Civil War "catch-up" in real wages (Goldin and Lewis, 1975) never regained the high wages relative to England achieved at the peak in 1854-1856,97%. Second, relative wages fell everywhere in the New World, at least based on Australian and U.S. experience after 1854. Third, and once again, the results were mixed in Europe. While Sweden gained a lot of ground on England between the mid 1850s and 1869, none of the other European countries in our sample did (with the possible exception of Belgium), and two suffered losses, France and Norway. As with the sharp rise in real wage dispersion in the two decades or so following 1830, the fall in C up to 1869 was being driven primarily by the erosion of the wage gap between the Old World and the New. There are three morals which emerge from this look at global labor markets between 1830 and 1869. First, there was a very sharp divergence in real wages up to the mid 1840s. Second, what turns out to be a long run convergence in real wages started after the mid 1840s. In that sense, the second regime of convergence might be dated 1845-1913. I have resisted this temptation since the New World sample is so small prior to the mid 1850s. The total sample rises to 15 by 1870 (including 4 New World countries); thus we should be more certain about these convergence trends starting then, a date which is commonly used by economic historians in describing other events anyway. Third, there is no comprehensive European support for the Gerschenkron hypothesis that some leader and laggard dynamic was contributing to divergence and labor market disintegration. Fourth, and perhaps most important, the dispersion statistic was driven primarily by the behavior of wage gaps between the Old World and the New, rather than by divergence or convergence patterns within the New World or within the Old World. A Century of Convergence: 1870-1988 Typically, the convergence hypothesis is usually tested by using end points. Long run convergence documented by others using GNP per capita or per worker evidence is confirmed with this new data on real wages. As Fig. 2 illustrates with the late 19th century decades, countries with high real wages in 1870 (like those important three in the New World-Australia, Canada, and the United States), underwent relatively slow real wage growth up to 1913; countries with low real wages in 1870 (like the poorest in the Old World—Denmark, Italy, Norway, and Sweden) underwent relatively fast real wage growth up to 1913; and those clustered in the middle (Argentina and the early European industrializers or their contiguous neighbors—Belgium, France, Germany, Great Britain, Ire- 152 JEFFREY G. WILLIAMSON land, the Netherlands, and, surprisingly, Spain) started with average wages and underwent average wage growth. How fast was real wage convergence over the full century following 1870 compared with what other researchers have found when using Mad-dison's GNP estimates? The following regression results (constant omitted), compared with revisions of DeLong's (1988, p. 1139), implies that real wage convergence was far stronger than GNP per worker hour, even when the analysis is limited to those 12 countries which appear in both samples. Independent variable Dependent variable Slope coefficient (r statistic) Sample description Log 1870 labor productivity Log 1870 labor productivity Log 1870 income per capita Log 1870 income per capita Log 1870 wage Log 1870 wage Annual percentage growth labor productivity Annual percentage growth labor productivity Log difference of 1979 and 1870 incomes Log difference of 1979 and 1870 incomes Log difference of 1988 and 1870 wages Log difference of 1988 and 1870 wages -0.748 Maddison's 16: (10.542) DeLong revision -0.887 Overlapping (9.069) sample of 12 -0.959 Maddison's 16: (10.241) DeLong revision -1.045 Overlapping (6.683) sample of 12 -1.215 Williamson's 15 (4.930) -1.159 Overlapping (9.032) sample of 12 Real wage convergence among the lesser skilled over the past century has been more dramatic than GNP per worker hour, a result consistent with the prediction offered earlier in this paper. Namely, it suggests that EVOLUTION OF GLOBAL LABOR MARKETS 153 0.4 0.3 0.1 0.2 0 1870 1875 1880 1885 1890 1895 1900 1905 1910 Fig. 3. International real wage dispersion, 1870-1913. as New World relative resource abundance diminished over time and as land rents in the New World tended to converge on the Old {O'Rourke et al., 1993), the contribution of real wage convergence to total income or output convergence was partially offset, and thus average labor productivity converged less dramatically than did real wages. Four Decades of Convergence: 1870-1913 As Fig. 3 shows, the striking convergence which started after the mid 1840s continued up to 1900, after which it ceased. In fact, the coefficient of variation is almost cut in half over the three decades 1870-1900 (falling from 0.24 to 0.14), and perhaps by two-thirds over the 46 years 1854— 1900. The unweighted coefficient of variation, C, can be decomposed into three additive parts: D„, dispersion within the New World, a variable weight times the coefficient of variation there; D0, dispersion within the Old World, a variable weight times the coefficient of variation there; and Dno, dispersion between the Old World and the New, a variable weight times the square of the average wage gap between the two. Along with C, Fig. 3 plots each of these three components. The results are striking and repeat those we found for the first regime. First, throughout the period 1870-1913, the average wage gap between the New World and the Old accounts for about 60% of the real wage variance across these 15 countries. The remainder, 40%, is explained by the variance within the Old and New Worlds. Furthermore, real wage variance within the New World accounts for more of the total variance than does real wage 154 JEFFREY G. WILLIAMSON Dispersion 0.16 i- 0.14 - 0.12 - 0.1 - 1B70 1875 1S80 1885 1890 1895 1900 1905 1910 Fig. 4. International real wage dispersion, Europe, 1870-1913. variance within the Old World, 24 and 16%, respectively. All of this implies that real wage variance among our 11 late 19th century European countries was a very modest part of "global" real wage variance. Second, about 60% of the convergence between 1870 and 1900 is explained by the collapse in the wage gap between New World and Old. It appears that this regime of dramatic convergence is primarily a story about the Old World catching up with the New, and of Argentina and Canada catching up with Australia and the United States. It is much less a story about latecomers catching up to leaders in the Old World. Nonetheless, and as Fig. 4 documents, convergence did take place within Europe. Furthermore, the European pattern (Fig. 4) closely resembles that of the full sample (Fig. 3), with one exception: while European convergence continued throughout the late 19th century, it ceased in the full sample around the turn of the century, the latter reflecting an American surge to industrial dominance. The convergence in Europe following 1870 deserves a more detailed look. Given the great debate about Britain's loss of industrial leadership to her close competitors, there is a tendency to look for evidence of, say, German catch-up on the leader. What matters far more, however, is the behavior of the poorer European countries relative to the richer, and the latter included Belgium, France, and Germany, not just Britain. From 1870 to 1900, five of the poorest countries improved their real wages relative to the average: Denmark, from 52 to 85% of Britain; Ireland, from 71 to 88; Italy, from 38 to 40; Sweden, from 41 to 82; and Norway, Year EVOLUTION OF GLOBAL LABOR MARKETS 155 from 41 to 65; only one, Spain, lost ground (a lot of ground, from 74 to 48). Over the same period, and with the exception of Great Britain, all but one of the richer countries underwent a real wage deterioration relative to the average: Germany, from 84 to 83% of Britain; France, from 72 to 68; the Netherlands, from 75 to 71; the exception being Belgium, from 87 to 91. There was indeed convergence within Europe between 1870 and 1900, but, ironically, Britain was not a major part of it. Throughout the late 19th century, convergence in Europe was slower than it was globally in our total sample (between 1870 and 1900, the dispersion statistic declines by about 0.1 in Fig. 3, or by about 40%; it declines by about 0.03 in Fig. 4, or by about 33%). Figure 4 makes it apparent why: it was not due to some weakness of convergence trends in the non-Latin North since that convergence continues and is very robust (at least to 1905: DNEUR, Fig. 4). Convergence is relatively slow in Europe as a whole because of a rise in the wage gap between the Latin South and the non-Latin North (DNSEUR, Fig. 4), especially up to the mid 1890s. This rise in the historically persistent wage gap between the Latin South and the non-Latin North accounted for much of the slow European convergence in the late 19th century, and this in spite of so much attention to an alleged late Victorian and Edwardian failure in England.15 Late Victorian and Edwardian failure helps explain continued convergence in the North of Europe, but what dominated European experience was not so much Britain's failure (which contributed to convergence), but rather the failure of the Latin economies (which retarded convergence). Let us now return to the average wage gap between New World and Old, the variable which drove such a large share of the convergence over the three decades or so after 1870 and the five decades or so after the mid 1840s. Three countries illustrate the process best, Ireland and Sweden (with heavy emigrations from the late 1840s onwards), and the United States (with heavy immigrations from the late 1840s onwards). In 1856, real wages in Sweden were only 47% of Britain, while in 1913 they were 89%, an impressive doubling in Sweden's wage relative over the 57 years. In 1852, and shortly after the famine, real wages in Ireland were only 59% of Britain, a figure that had changed hardly at all over the previous three decades. Real wages in Ireland started a dramatic convergence on Britain during the 1850s (and, notably, in the absence of any Irish industrialization16) so that they were 71% of Britain by 1870. By 1905 they were 92% of Britain, before slipping a bit to 82% by 1913. Ireland was transformed over this period of convergence from a poverty-stricken, peasant economy which had served as a source of elastic labor supply for ls See, for example, the seminal survey by McCloskey (1970). 16 Abramovitz (1986, p. 398), O'Rourke (1989, 1990), and, more recently, Boyer et al. (1994) all make the same point. 156 JEFFREY G. WILLIAMSON Britain's booming cities, to an economy at the start of the 20th century which boasted wages close to those prevailing across the Irish Sea, and which came to exceed British wages in the 1920s (O'Rourke, 1990). The Irish convergence toward real wages in the United States must have been even more dramatic since relative real wages were falling in America during most of this period. In 1855, real wages in the United States were 98% above Britain, while in 1913 they were 54% higher, almost a halving in the American wage advantage over Britain, a spectacular decline that has gone almost unnoted by American economic historians.17 These patterns were comprehensive enough to have contributed to real wage convergence over the half century, and, as we have seen, it was the decline in the wage gap between the New and Old World which was doing most of the work. But there are some deviant countries and periods well worth our attention. First, the Latin experience was very different. Through dramatic booms and busts, Argentina increased her real wage advantage over Spain and Italy, the source of the vast majority of her immigrants from the Old World. Indeed, Argentina improved her real wage position relative to Britain, from 66% in 1864 to 84% in 1913, and her real wages actually exceeded Britain in 1893, 1899, 1900, and 1904, an achievement that Argentineans view with nostalgia (Cortes-Conde, 1979). Second, the experience in the English-speaking New World varied over the regime. While Australia experienced a steady and much-studied (McLean and Pincus, 1983) erosion in her real wage position over the whole period of convergence—from 138% above the British real wage in 1854, to 84% above in 1870, to 35% above in 1890 and to just 16% above in 1913, the other New World countries enjoyed a partial resurrection in their real wage advantage late in the regime. This was especially true of North America. Relative to Britain, real wages in the United States, were 98% higher in 1855, 67% higher in 1870, 32% higher in 1880, but 54% higher in 1913. Real wages in Canada were 43% higher than in Britain in 1870, 42% higher in 1880, but 99% higher in 1913 (after the great wheat boom and railroad expansion of which so much is made by Canadian economic historians18). In short, both Canada and the United States bucked the convergence tide after the mid 1890s. This result is consistent with the North America's emerging industrial dominance about that time (Wright, 1990), and it makes America's successful defense of her economic leadership for so long thereafter all the more impressive. Thus, even during this period of dramatic convergence, when trends in 17 With the exception, perhaps, of Shergold (1982). 18 For the classic paper, see Chambers and Gordon (1966). EVOLUTION OF GLOBAL LABOR MARKETS 157 the wage gap between the Old World and the New were doing most of the work, there was a variety of experience that remains to be explained. It seems worth noting that the two most prominent contributors to the historical convergence literature, Abramovitz and Baumol, make very little of the convergence forces 1870-1913 which seem to be so pronounced in the real wage data used here. In Abramovitz's words, "the rate of convergence . . . showed marked strength only during the first quarter-century following World War II," and "in the years of relative peace before 1913 . . . the process [of convergence] left a weak mark on the record" (1986, pp. 385 and 395). These are puzzling statements since Abramovitz's own Table 1 reports the coefficient of variation falling by more than a third, from 0.51 in 1870 to 0.33 in 1913 (Abramovitz, 1986, p. 391). True, our real wage data document an even greater convergence, and it is also true that in percentage change per year, Abramovitz's variance statistic drops faster 1950-1973 than with 1870-1913, but his use of Maddison's GNP data reveals a very strong convergence prior to 1913 nonetheless. Based on his own evidence, it is not clear why Abramovitz thinks convergence left only a "weak mark" on the record. Since Baumol and his associates also use Maddison's data, it will come as no surprise that Productivity and American Leadership replicates Abramovitz's findings. In their Fig. 5.2, they show the coefficient of variation falling between 1870 and World War I and further state that "the downward trend in this dispersion measure is strong and steady in each of the two periods separated by World War II" (Baumol et al., 1989, p. 92). Having confirmed Abramovitz's finding, Baumol and his associates move on to the post-World War II period where concern about America's loss of leadership pulls them, ignoring this earlier and spectacular period of convergence for the remainder of their book. It seems to me that the pre-1913 convergence deserves far more attention than the literature has given it thus far. After all, no other period since the mid-19th century shares so much in common with the amazing post-World War II epoch. Convergence Ceases: 1914-1945 The World Wars and the interwar decades offer nothing but contrasts to the long run convergence experience initiated in the mid-19th century. As Fig. 5 confirms, the convergence ceased from 1914 to 1934 since C is roughly the same in both years. The cessation of real wage convergence documented here offers a very different characterization than that found in Productivity and American Leadership. When Baumol and his associates plot the coefficient of variation (based on Maddison's GNP data) beyond 1913 and up to the mid-1950s, their C continues its long run decline 158 JEFFREY G. WILLIAMSON 1939 1919 1924 1929 1934 Year Fig. 5. International real wage dispersion, 1914-1945. initiated in 1870.19 Indeed, they state that convergence "has proceeded steadily, with the exception of a brief but sharp fallback during and after World War II" (Baumol et al, 1989, p. 92). The real wage data suggest the contrary: long run convergence ceased between 1914 and 1934. Fur- Dispwskxi 0.16 0.14 012 0.08 0 06 1914 1919 1924 1929 1934 1939 1944 Year Fig. 6. International real wage dispersion, Europe, 1914-1945. Abramovitz (1986, Table 1, p. 391) found the same. EVOLUTION OF GLOBAL LABOR MARKETS 159 thermore, after World War I it was the variance in real wages within the Old World (Dc) that dominated, not the average real wage gap between New World and Old that was so true of the classic period of convergence prior to 1913 (see Fig. 6). In addition, divergence took place after 1934 and up through World War II and it took place everywhere—within the Old World, within the New World, and between the two. A large share of that divergence was driven by the spectacular surge in real wages in the United States.20 The result of all this was that the level of real wage dispersion of the early 1880s had been regained by 1945: the Great Depression and World War II lost almost everything that had been gained over the four decades or so prior to 1913. Indeed, since our real wage rates do not take account of unemployment, and since unemployment rates in the United States in 1934 were higher than those elsewhere, the surge in American unemployment-adjusted real wages would be even greater and the measured divergence greater as well. As we shall see, a good part of the post-World War II convergence served simply to regain what had been lost between 1934 and 1945. The most interesting question suggested by the war and interwar experience, it seems to me, is how much of the cessation in long run real wage convergence can be explained by the breakdown of international commodity and factor markets. This research avenue seems at least equal in promise when compared to that well-traveled street which appeals to the cessation of international technological transfer. Convergence Resumes: 1946-1988 The post-World War II convergence has, of course, been well studied. But there are two aspects of this experience that may not have been fully appreciated. First, and in contrast with the position taken by Abramovitz and others, it was not the period of most dramatic convergence. I believe the pre-1913 epoch deserves that prize. Second, there were three distinct episodes within the period, not just one. Figure 7 displays C over the full postwar period. Postwar recovery in the Old World generated a short, dramatic convergence so that by 1950 much, but not all, of the global labor market integration that had been lost after 1913 was regained. Across the 1950s, there was no convergence, and this was not simply due to continued American success in retaining her leadership. Figure 8 shows that within Europe there was no convergence at all until the mid 1960s. Indeed, there was some divergence at work, and it was driven entirely by that North-South wage gap along the Latin divide, a repeat performance of the 1890-1913 experience. Our real wage evidence therefore suggests that the pre-1913 long run 20 See also Abramovitz (1986), p. 395, and Wolff (1991), p. 569. 160 JEFFREY G. WILLIAMSON Dispersion 0.4 i- 0.3 - 1946 1951 1956 1961 1966 1971 1976 1981 1986 Year Fig. 7. International real wage dispersion, 1946-1988. convergence did not resume until the mid-1960s. Post-World II real wage convergence, therefore, is a relatively recent story that started unfolding only 25 years ago. The story has two parts: the first is well known—the European Old World's final catch up to the New World, especially the leader, the United States—and the second is less well known—the spec-Dispersion 0.16 i-. 0.14 -0.12 -0.1 - 1950 1955 1960 1965 1970 1975 1980 1985 Year Fig. 8. International real wage dispersion, Europe, 1946-1988. EVOLUTION OF GLOBAL LABOR MARKETS 161 tacular rise in Old World Latin real wages in Italy, Spain, and even France relative to everyone else. This latter event is especially notable since these three countries as a group hadn't made much progress at all in joining the convergence club from the mid-19th century onward. In 1870, Italy, Spain, and France had real wages that were, respectively, 38, 74, and 72% of Britain, for an average of 61; in 1913, they were 50, 46, and 60, for an average of 52; in 1938, they were 42, 47, and 67, for an average of 52; in 1946, they were 38, 77, and 51, for an average of 55; and in 1980, they were 115, 112, and 91, for an average of 106. Over the seven decades between 1870 and 1946, these three countries were able to erase none of the 39% wage gap between themselves and Britain, while over the four decades between 1938 and 1980, they were able to erase all of it and more. IV. A RESEARCH AGENDA This real wage data base should add new fuel to the fires burning on the determinants of economic growth generally, and the forces of convergence and divergence specifically. The evidence confirms much of what has been said about convergence based on Maddison's GNP data. By itself, the addition of a completely new data base which reinforces some of the conventional wisdom about long run growth should be welcome. But the new data base also rejects some conventional wisdom, adds new facts, and sharpens the agenda. It might be helpful to summarize the new facts and confirmation of the old. Recall, however, that these facts are based on the experience of a sample of 15 countries, all of whom have achieved economic success since 1830. The sample excludes all of Eastern Europe and includes only Italy and Spain from Southern Europe. The New World group is limited only to 4—Argentina, Australia, Canada, and the United States. No Third World countries are included, or even Japan for that matter. Subject to that sample limitation, here is what we find: * The convergence that others have documented for the period between 1870 and 1988 is confirmed. However, wage convergence started in the mid-1840s, not in 1870. * Real wage convergence since 1870 has been considerably more dramatic than GNP per capita or GNP per worker-hour convergence. * Wage convergence between the mid-1840s and 1913 was at least as dramatic as it has been since 1950, and probably more so. * The long run wage convergence between the mid-1840s and 1913 has two phases, a very steep descent to about 1900, and stability thereafter. *Between the mid-1840s and 1913, the average wage gap between the New World and the Old was doing most of the work in driving overall convergence (and divergence). The cessation of wage convergence after 1900 was driven by a cessation in the erosion of the average wage gap 162 JEFFREY G. WILLIAMSON between the Old World and the New, in particular by a relative wage boom in North America as it rose to world industrial leadership. The cessation of convergence within Europe was driven by a rising wage gap between the Latin South and the non-Latin North. * Long run wage convergence ceases between 1913 and the mid-1930s, and sharp divergence takes place thereafter until 1945. This war and interwar episode was sufficiently dramatic that much of the long run convergence achieved after 1870 was lost by 1945. * The post-World War II wage convergence has been very recent— since the mid-1960s—and it has been driven primarily by an erosion of two wage gaps—between the Old World and the New, and between the Latin South and the non-Latin North within the Old World. In short, there has been significant variance in the rate of convergence since the mid-19th century, so much so that it suggests that the world economic environment mattered a great deal, and that different explanations may be more relevant for some epochs than for others. I do not mean by this that a "general theory" of convergence is out of reach, but only that the forces driving convergence (or divergence) are likely to have had very different quantitative significance within different epochs. What remains is to uncover the sources of convergence within these epochs. How much can be attributed to labor supply responses, and international migrations in particular? We have only started to attack this important question for the late 19th century decades of mass migrations.21 How much can be attributed to capital accumulation responses, and international capital flows in particular? After all, capital typically chased after labor migrating to the New World thus muting the convergence impact of international migrations on capital-labor ratios in the labor-scarce New World and the labor-abundant Old World (Hatton and Williamson, 1992b). How much can be attributed to world commodity market integration and disintegration, and thus to the real wage gap implications of Heckscher-Ohlin thinking embedded in trade theory? In the late 19th century, it appears that commodity price convergence—generated by transport improvements—made a profound contribution to real wage convergence.22 And exactly how did experience in each of these three markets—labor, capital, and commodities—interact? It seems to me that economic historians should attack these issues first before elevating international technological transfer to the status of prime mover, a thesis so ably argued by Gerschenkron that it has dominated the convergence debate ever since. 21 The argument is offered in Hatton and Williamson (1992b). The argument is supported with Anglo-American evidence in O'Rourke etal. (1994), and with Irish evidence in Boyer et a!. (1994). 22 At least between Britain and the United States; see O'Rourke and Williamson (1994). EVOLUTION OF GLOBAL LABOR MARKETS 163 APPENDIX 1 Nominal Wage, Cost of Living, and Real Wage Series: Sources and Methods Wages are specified as being hourly, daily, weekly, etc., unless this information is omitted in the sources. Resulting real wage indices are reported in Table Al.l. In most cases, the indices in Table A 1.1 have a base year of 1900 for the entire 1830-1988 period. Breaks in the real wage indices could not be avoided, however, for Belgium, Brazil, Denmark, Germany, and Portugal. Such breaks are indicated in Table Al.l by a horizontal line. Observations in Table Al.l subsequent to a break have a base year of either 1929 or 1955. Argentina (ARG) Nominal wage. 1864-1870: Simple average, average monthly wages of porteros and peones (pesos fuertes); Republica Argentina, Ley de Pre-supuesto General, various years, Buenos Aires. Linear interpolation used for peones in 1866. 1870-1883: Simple average, average monthly wages of porteros and peones (pesos fuertes); Republica Argentina, Memoria de Hacienda, various years, Buenos Aires. Linear interpolation used in 1871, 1875, and 1882. 1883-1903: Average monthly wage, peones de policia (pesos moneda nacional); R. Cort6s Conde, El Progreso Argentino, 1880-1914, Buenos Aires, Editorial Sudamericana, 1979, p. 226. 1903-1914: Daily wage of peones albaniles (pesos mn); Republica Argentina, Boletin del Departamento Nacional del Trabajo, Buenos Aires, no. 21, Nov. 30, 1912, p. 460, and no. 33, Jan. 30, 1916, p. 132. Linear interpolation for 1913. 1914-1940: Average nominal wage in Buenos Aires, 1929 = 100; Republica Argentina, Investigaciones Sociales 1943-1945, Buenos Aires, Direction de Estadistica Social, 1946, p. 258. Cost of living. 1864-1890: Cost of living index, 1882 = 100; R. Cortes Conde, unpublished worksheets. Based on wholesale prices of 16 items with fixed weights. 1890-1910: Cost of living index, 1903 = 100; R. Cortes Conde, El Progreso Argentino, 1880-1914, p. 226. Straight line interpolation based on food price index (ibid.) in missing years. 1910-1914: Cost of living index, 1910 = 100; A. E. Bunge, Los Prob-lemas Econdmicos del Presente, vol. 1, Buenos Aires, n.p., 1920, p. 269. 1914-1940: Cost of living index for Buenos Aires, 1929 = 100; Republica Argentina, Investigaciones Sociales 1943-1945, p. 258. Real wage. 1940-1980: Average real wage in manufacturing, 1970 = 1; J. J. Llach and C. E. Sanchez, "Los Determinantes del Salario en la TABLE A 1.1 National Real Wage Indices, 1830-1988 (1900 = 100) Year ARG AUS CAN USA BEL DEN FRA GER GB IRL ITA NET NOR SPA SWE BRZ POR 1830 47 52 50 36 50 132 30 55 1831 48 53 51 35 45 110 30 54 1832 49 56 54 37 47 95 30 72 1833 46 66 59 39 50 98 29 32 1834 52 67 60 41 54 119 29 32 1835 47 64 58 44 60 108 29 48 1836 54 63 52 38 58 97 29 49 1837 61 60 51 37 60 84 28 45 1838 62 58 49 35 51 83 28 39 1839 52 55 47 34 48 94 28 42 1840 50 55 49 34 48 97 28 35 1841 54 61 50 36 46 110 28 37 1842 62 60 54 38 46 103 28 55 1843 69 48 61 60 43 48 114 28 53 1844 74 48 61 58 41 52 118 28 54 1845 71 46 61 59 42 47 132 28 45 1846 69 46 57 56 40 43 105 28 42 1847 54 42 56 49 34 35 89 28 48 1848 64 41 59 60 42 52 99 28 45 1849 70 45 62 66 46 53 128 28 53 1850 68 46 65 79 65 53 62 34 132 28 60 79 1851 69 49 66 72 68 48 61 31 126 28 55 94 1852 70 51 64 61 66 46 54 32 132 28 59 95 1853 68 53 63 57 55 49 56 32 119 28 68 75 1854 89 65 46 61 53 49 43 53 36 122 29 54 67 1855 74 64 46 60 50 49 49 53 39 118 30 53 76 1856 84 66 47 61 53 51 53 56 39 99 30 56 60 1857 80 64 50 63 67 55 47 59 42 103 31 56 76 1858 76 68 52 69 68 65 56 65 40 105 31 62 86 1859 74 67 56 73 67 64 54 65 41 112 32 64 76 1860 85 68 56 66 65 55 48 63 41 109 33 72 80 1861 89 66 53 64 55 55 46 60 40 103 40 72 77 1862 75 59 49 65 64 55 46 61 41 108 48 71 88 1863 66 53 53 65 69 54 45 55 43 100 46 66 89 1864 40 76 50 60 66 75 55 45 57 41 99 42 63 85 1865 43 58 53 61 67 71 56 46 63 42 103 44 65 89 1866 50 67 56 60 67 70 61 51 65 40 101 35 64 85 1867 49 79 60 60 67 60 63 53 67 40 98 36 73 76 EVOLUTION OF GLOBAL LABOR MARKETS 165 0--*0—'--iO—■ ^^OOOOOv<>OvO^OvOOOO^OOOO\OOa>OOCO ooooo»-*oo--o 't- W( X' U'i IT, ITi kA ir-j "Ai in \C r- r^j r- ^ r ^ 00 er ^ ^ ^ ^ ^ ^ o r> ^> ^< rsi —^!^J2?2 — SfSf 3w-, r4 r J c4 c~j ^ rj ^ o ^ cc ^ ^ ^2o^^co^^^^M^^^^«MM»KMoco^ao^ooocooooooo'0^o^ac^o^o\ ^voovcr^r-r^oor^r^r^r^r^r^r^wccxocoo^ooo^xxo^^o^o^^^oc ■ ^a>©oo©o©© vcr^r^r^r^r^^^ox^r-t^r^r^r^ooccocoocc^ocoo^Ose^ f"-J Ov VI Vi \C O1, f^J (""J CN (N *—| VI CN t*~- I CT1 00 00 © JT* 1—1 £1 £-1 10 t" 2* 2 2t 1— Ex 5 I^ t ^ 3D sC^r^t^r^r^r^oooür^t^r^r^l^QOCCccaoococoo x m cc c go \£ w"/ vi ^ oi T+ ia r^; ^ ^ oo ^ w o ^ ^ o ^ w q 2 3" ic' i 2: 15 3! S 5 ^ iA rf in «a ^ Tt ^ vC »r, v£ iOO00QO«Q0«M<»00»XO0O00O»i § ^ §^ § ^ ^ ^ ^ TABLE Al.1 -Continued Year ARG AUS CAN USA BEL DEN FRA GER GB IRL ITA NET NOR SPA SWE BRZ POR 1911 89 98 134 106 92 121 94 105 93 101 126 96 111 113 125 134 95 1912 92 94 130 105 96 120 96 104 91 100 134 98 110 108 120 120 94 1913 87 92 140 109 102 118 % 109 97 100 136 100 125 107 119 129 1914 85 89 136 108 107 62 94 105 85 103 152 100 136 106 119 149 1915 76 78 132 109 82 93 74 94 154 89 119 104 110 129 1916 71 81 131 124 72 83 69 82 134 87 113 100 108 120 1917 61 81 120 132 49 69 74 70 88 129 88 121 99 106 94 1918 52 79 126 141 56 68 76 80 101 123 100 110 102 110 76 1919 71 75 131 143 85 73 91 162 123 139 107 120 75 1920 73 79 125 141 92 93 82 97 178 123 142 136 138 69 1921 91 101 129 131 108 106 95 106 184 152 150 143 162 102 1922 104 104 131 129 120 97 96 95 175 158 147 149 144 100 1923 107 101 139 139 115 89 97 93 164 156 142 147 147 84 1924 106 105 139 144 109 89 96 85 95 135 159 150 133 145 148 86 1925 111 105 131 139 112 95 96 98 96 121 158 150 136 145 149 83 1926 112 106 129 140 96 100 92 105 97 121 159 156 142 152 157 107 _ 1927 118 109 134 145 106 99 85 105 100 129 172 159 142 147 158 112 70 1928 126 110 137 148 114 99 91 111 99 124 171 160 147 152 160 108 1929 124 108 138 152 125 100 97 114 100 124 127 167 156 138 170 112 1930 113 112 143 153 130 107 97 121 103 125 168 173 159 155 179 130 1931 122 112 146 161 125 113 98 124 109 132 174 181 169 152 183 82 1932 129 111 144 157 110 115 100 120 109 133 168 178 162 157 183 88 1933 119 111 143 166 112 112 105 118 112 138 173 176 162 174 180 96 1934 123 109 144 191 117 109 111 113 110 136 175 169 167 170 178 1935 126 108 162 193 111 105 113 109 110 133 167 168 166 168 181 94 1936 118 109 163 193 112 106 106 107 110 131 155 172 167 166 181 1937 119 110 151 212 128 104 108 106 108 124 146 170 172 163 183 114 1938 119 113 149 222 135 108 101 107 110 129 143 171 176 140 189 1939 121 113 150 229 133 109 98 107 109 127 165 173 178 123 196 85 1940 122 111 153 233 97 98 104 108 111 160 159 159 109 185 89 1941 120 112 149 248 93 90 102 107 104 147 148 91 174 92 1942 120 112 173 254 94 76 100 115 94 144 137 100 178 73 1943 124 115 186 264 100 65 100 120 87 105 134 101 181 79 1944 141 116 192 272 104 92 125 87 40 130 137 97 183 _ 92 1945 134 117 194 273 110 104 128 89 39 133 154 95 192 74 85 1946 129 118 198 279 _ 122 99 139 93 88 152 174 80 204 80 90 1947 158 124 198 279 86 125 103 143 114 145 148 189 71 223 68 94 1948 194 127 195 275 83 138 98 139 121 185 149 200 66 241 70 0\ CK m 3 to a r r > 2 O Z EVOLUTION OF GLOBAL LABOR MARKETS 167 _j ,_ ,_, ^ _^ _M H^HHHrtNHrtHHr-HHHp-fNIrNNMNHtNNMMMrNNNn 168 JEFFREY G. WILLIAMSON Argentina. Un Diagnostico de Largo Plaza y Propuestas Depolfticas," Estudios, aiio VII, no. 29, Enero/Marzo 1984, p. 5. 1980-1984: Real wage in industry, 1982 = 100; Estudio M.A.M. Brody y Asoc, Carta Econdmica, ano 2, no. 23, April 1985. From the FIEL (Fundacion de Investigaciones Economicas Latinoamericanas) database. 1984-1988: Real wage in industry, usual calculation, January 1984 = 100; Estudio M.A.M. Brody y Asoc, Carta Econdmica, ano 6, no. 73, June 1989. From the FIEL database. Australia (AUS) Nominal wage. 1854-1860: Unweighted average of indices of wage rates in New South Wales and Victoria; indices derived from annual percentage changes reported in R. Maddock and I. McLean, "Supply-Side Shocks: The Case of Australian Gold," Journal of Economic History 44(4), 1984, Table 2, p. 1055. The New South Wales and Victoria wage rates are themselves employment-weighted averages of wage rates (mostly daily and weekly) of seven occupations: farm laborers, shephards, workers in building trades, blacksmiths, female servants, general laborers, and gold-miners (implicit wage rate). 1860- 1861: Money wages of urban general laborers in Victoria; G. Withers, unpublished database, n.d. 1861- 1913: See Real wage below. 1913-1975: Money wages in industry; B. R. Mitchell, International Historical Statistics: The Americas and Australasia (hereafter IHS), Gale Research, Detroit, 1983, Table C4, p. 177. 1976- 1977: Money wages in industry; International Labour Office (hereafter ILO), Yearbook of Labour Statistics, Geneva, 1980. A continuation of the above. 1977- 1988: Hourly wage in all activities; OECD, Main Economic Indicators 1969-1988, (hereafter MEl), Paris, 1989, p. 159. Cost of living. 1854-1861: Consumer price index; I. W. McLean and S. J. Woodland, "Consumer Prices in Australia, 1850-1914," mimeo, Department of Economics, University of Adelaide, revised April 1992, Series W6, Appendix Table 3. 1861-1913: See Real wage below. 1913-1975: Consumer price index; B. R. Mitchell, IHS, Table 12, p. 841. 1975-1988: Consumer price index; OECD, MEl, p. 160. Real wage. 1861-1900: Geometric mean of real wages of urban general laborers in Queensland, South Australia, and Victoria, weighted by each colony's workforce share; based on colony-level nominal wage, price level, and workforce estimates from G. Withers, unpublished database, n.d. (The colony-level price indices in the Withers database are derived from N. G. Butlin, Australian Domestic Product, Investment and Foreign Bor- EVOLUTION OF GLOBAL LABOR MARKETS 169 rowing, 1861-1938/39, Cambridge Univ. Press, Cambridge, 1962, Table 78, p. 158.) 1900-1913: Real wage of laborers in Sydney; R. C. Allen, "Real Incomes in the English Speaking World," mimeo, University of British Columbia, Vancouver, November 1990, p. 45. Belgium (BEL) Nominal wage. 1843-1939: Composite index of annual wages in several occupations (varying from 6 to 25); P. Scholliers, "A Century of Real Wages: A Methodological and Empirical Contribution on Belgium, 1840-1940," paper presented at the conference on International Migration and Labor Market Integration in the 19th and 20th Centuries, Bellagio, Italy, June 14-18, 1993, Annex 2, "Scholl-3, 4, 5" series are used, pp. 21-22. 1940-1946: Data not available. 1947-1969: Daily wages of males in manufacturing; B. R. Mitchell, European Historical Statistics 1750-1975 (hereafter EHS), Columbia Univ. Press, New York, 1978, p. 75. 1969-1988: Hourly wage in manufacturing; OECD, MEl, p. 240. Cost of living. 1843-1880: F. Michotte, "Ľévolution des prix de detail en Belgique de 1830 ä 1913," Bulletin de l'Institut des Sciences Econo-miques de Louvain, May 1934, pp. 354-357; reprinted in J. Singer-Kerel, Le coüt de la vie ä Paris de 1840 á 1954, A. Colin, Paris, 1961, pp. 108-109. 1880-1939: Scholliers, "A Century of Real Wages," Annex 1, "Scholl-1, 2, 3" series are used, pp. 19-20. 1939-1969: Consumer price index; Mitchell, EHS, p. 781. 1969-1988: Consumer price index; OECD, MEl, p. 244. Brazil (BRZ) Nominal wage. 1830-1930: Unweighted average of monthly wages for laborers, carpenters, bricklayers, and porters in Rio de Janeiro; E.M.L. Lobo, História do Rio de Janeiro (do capital commercial ao capital industrial e financeiro), Instituto Brasileiro do Mercado de Capitais (IB-MEC), Rio de Janeiro, 1978, pp. 803-813. 1930-1937: Hourly wage of unskilled building laborers; ILO, Yearbook of Labour Statistics, Geneva, 1938. 1938-1944: No data available. 1945-1988: Industrial wage index; J. Wilkie (Ed.), Statistical Abstract of Latin America, UCLA Latin American Center Publications, Los Angeles, Vol. 29, Pt. 1, p. 534. The years 1960-1962 are missing. Cost of living. 1830-1850: Price index using 1919 consumption weights; Lobo, História do Rio de Janeiro, pp. 748-751. 1850-1870: Price index; K. W. Goldsmith, Brasil, 1850-1984, Harper & Row, Sao Paulo, 1986, pp. 30-31. 170 JEFFREY G. WILLIAMSON 1870-1913: Wholesale price index; L. A. V. Catao, "A new wholesale price index for Brazil during the period 1870-1913," Revista Brasileira de Economica 46(4) (October/December, 1992), Appendix 1, Table 1, p. 530. 1913-1947: Global price deflator; Estatisticas Histöricas do Brasil, IBGE, Rio de Janeiro, 1987, Vol. 3, p. 159. 1947-1988: Consumer price index; Wilkie (Ed.), Statistical Abstract of Latin America, Vol. 29, Pt. 1, p. 534. Canada (CAN) Nominal wage. 1870-1889: Wages of construction workers; T. O. Dick, "Output, Prices and Real Wages: The Canadian Experience 1870-1915," mimeo, Harvard University, 1982, p. 25. 1889-1901: Average daily wage, laborers in the building trades (Ottawa and Toronto); F. H. Leacy (Ed.), Historical Statistics of Canada, Ottawa, 1983, Tables D-472 and D-480. 1901-1974: Weighted average by provincial population of hourly wages of laborers in the building trades (Halifax, Montreal, Toronto, Winnipeg, Vancouver); ibid., Tables E-248 and E-267. 1974-1988: Hourly earnings in manufacturing; OECD, MEI, p. 30. Cost of living. 1870-1913: Wholesale price index; Leacy, Historical Statistics of Canada, Table Jl. 1913-1969: Consumer price index; ibid., Table K8. 1969-1988: Cost of living index; OECD, MEI, p. 34. Denmark (DEN) Nominal wage. 1870-1980: Hourly wage of unskilled males in crafts and industry; H. Chr. Johansen, Dansk 0konomisk Statistik, 1814-1980, Copenhagen, 1985, pp. 294-296. 1980-1988: Hourly earnings in manufacturing; OECD, MEI, p. 271. Cost of living. 1870-1980: Consumer price index; Johansen, Dansk 0konomisk Statistik, 1814-1980, pp. 298-302. 1980-1988: Consumer price index; OECD, MEI, p. 273. France (FRA) Nominal wage. 1830-1840: All France wage rate index; M. Levy-Le-boyer and F. Bourguinon, The French Economy in the Nineteenth Century: Essays in Econometric Analysis, Cambridge Univ. Press, Cambridge, 1990, Table A-IV. 1840-1939: All France, nonfarm daily wage rate index; J. Singer-Kerel, Le coüt de la vie, pp. 536-537. 1939-1946: Hourly wage rates, males, Paris; ibid., pp. 538-539. 1946-1969: Hourly industrial wage; Mitchell, EHS, p. 75. 1969-1988: Hourly industrial wage; OECD, MEI, p. 335. EVOLUTION OF GLOBAL LABOR MARKETS 171 Cost of living. 1830-1840: M. Levy-Leboyer and F. Bourguinon, The French Economy in the Nineteenth Century, Table A-IV. 1840-1948: J. Singer-Kerel, Le coat de la vie, pp. 141, 452-453. 1948-1969: Wholesale price index; Mitchell, EHS, pp. 388-390, 392. 1969-1988: Consumer price index; OECD, MEI, p. 338. Germany (GER) Real wage. 1850-1871: Real wage; J. Kuczynski, Darstellung der Lage der Arbeiter in Deutschland von 1789 bis 1949, Berlin, 1961, p. 246. 1871-1943: Real wage of unskilled in the building trades; G. Bry, Wages in Germany 1871-1945, Princeton Univ. Press, Princeton, 1960, Table A-4, pp. 335-336 (no data available 1919-1923). Nominal wage. 1950-1969: Weekly industrial wage, West Germany; Mitchell, EHS, p. 197. 1969-1988: Hourly wage in manufacturing; OECD, MEI, p. 384. Cost of living. 1950-1969: Cost of living index; Mitchell, EHS, p. 781-783. 1969-1988: Consumer price index; OECD, MEI, p. 385. Great Britain (GB) Nominal wage. 1830-1834: Weekly farm wages; Mitchell, EHS, p. 78. 1834-1860: Weekly wage of a common laborer in the building trades computed as an unweighted average of wages in Manchester, Southern England and nine towns. Sources: Up to 1838, A. L. Bowley, Wages in the United Kingdom in the Nineteenth Century, Cambridge Univ. Press, Cambridge, 1900, p. 60; 1839-1860, A. L. Bowley, "The Statistics of Wages in the UK during the last 100 years," Journal of the Royal Statistics Society 43, June 1900, pp. 300-311. 1860-1880: Weighted average of weekly wages of bricklayers' laborers in Birmingham, Leeds, London, and Manchester. Source: G. Boyer, T. Hatton, and K. O'Rourke, "Emigration and Economic Growth in Ireland, 1850-1914," paper presented to the conference on International Migration and Labor Market Integration in the 19th and 20th Centuries, Bellagio, Italy, June 14-18,1993; based on Board of Trade's unpublished document, Rates of Wages and Hours of Labour in Various Industries in the United Kingdom, 1908. 1880-1913: Weighted average of weekly wages for bricklayers' laborers in 10 cities. Source: G. Boyer, T. Hatton, and K, O'Rourke, "Emigration and Economic Growth in Ireland." 1913-1969: Weekly wages, adult males, in manufacturing; Mitchell, EHS, p. 71-76. 1969-1988: Industrial wages; OECD, MEI, p. 698. Cost of living. 1830-1851: P. H. Lindert and J. G. Williamson, "English 172 JEFFREY G. WILLIAMSON Workers' Real Wages: Reply to Crafts," Journal of Economic History 45, March 1985, Table 1, pp. 148-149. 1851-1860: J. G. Williamson, Did British Capitalism Breed Inequality?, Allen & Unwin, London, 1985, Table A.8, p. 220. 1860-1913: C. H. Feinstein. National Income, Expenditure and Output of the UK, 1855-1965, Cambridge Univ. Press, Cambridge, 1972, Table 140. 1913- 1914: C. H. Feinstein, "A New Look at the Cost of Living, 1870-1914," in J. Foreman-Peck (Ed.), Reinterpreting the Victorian Economy: Essays in Quantitative Economic History, Cambridge Univ. Press, Cambridge, 1990, Table 4, pp. 32-33. 1914- 1969: Cost of living index; Mitchell, EHS, pp. 780-782. 1969-1988: Consumer price index; OECD, MEI, p. 702. Ireland (IRL) Nominal wage. 1830-1860: Daily wages of general laborers in the Dublin building trades; F. A. D'Arcy, "Wages of Labourers in the Dublin Building Industry, 1667-1918," Saothar 14, 1989, Tables 6 and 7, pp. 23 and 24. 1860-1913: Weighted average of laborers' weekly wages in Dublin and Cork building trades. Source: G. Boyer, T. Hatton, and K. O'Rourke, "Emigration and Economic Growth in Ireland;" based on F. A. D'Arcy, "Wages of Labourers" and the Board of Trade's Rates of Wages and Hours of Labour in Various Industries in the United Kingdom, 1908. 1913- 1918: Daily wages of general laborers in the Dublin building trades; F. A. D'Arcy, "Wages of Labourers," Table 8, p. 26. 1918-1931: Weekly wages in agriculture, males; International Labor Office, Annual Review, Geneva, 1931, Table 1, p. 323; and ILO, Yearbook, Geneva, 1931, Table 1, p. 297. 1931-1967: Industrial wages, in [Irish] Statistical Abstract, several issues. 1967-1969: Weekly earnings in manufacturing; Mitchell, EHS, p. 76. 1969-1988: Hourly wage in manufacturing; OECD, MEI, p. 441. Cost of living. 1830-1860: K. H. O'Rourke, Agricultural Change and Rural Depopulation in Ireland 1845-1876, Ph.D. Thesis, Harvard University, 1989, pp. 129, 212-226. 1860-1914: Cost of living in United Kingdom (see GB sources). 1914- 1969: ILO, Yearbook, several issues. 1969-1988: Consumer price index; OECD, MEI, p. 443. Italy (ITA) Nominal wage. 1871-1890: Hourly industrial wage, males; Mitchell, EHS, p. 72. 1890-1913: Daily wage in industry; V. Zamagni, "An International Comparison of Real Industrial Wages 1890-1913: Methodological Issues EVOLUTION OF GLOBAL LABOR MARKETS 173 and Results," in P. Scholliers (Ed.), Real Wages in 19th and 20th Century Europe: Historical and Comparative Perspectives, Berg, New York, 1989, p. 134. 1913- 1922: Daily wage for industrial operatives; V. Zamagni, "La al-terazione nella distribuzione del reddito in Italia neH'immediato de po-guerra 1918-1922," in La transizione dell'economia di guerra all'economia di pace in Italia e in Germania dopo la Prima Guerra Mondiale, Societa editrice il Mulino, Bologna, 1983, p. 531. 1922-1969: Daily wage in industry; Mitchell, EHS, pp. 74-76. 1969-1988: Hourly industrial wage; OECD, MEI, p. 468. Cost of living. 1871-1890: Mitchell, EHS, p. 778. 1890-1914: ISTAT cost of living series, in V. Zamagni, "An International Comparison of Real Industrial Wages 1890-1913: Methodological Issues and Results," in P. Scholliers (Ed.), Real Wages in 19th and 20th Century Europe, p. 134. 1914- 1969: Cost of living index; Mitchell, EHS, pp. 780-783. 1969-1988: Consumer price index; OECD, MEI, p. 474. The Netherlands (NET) Real wage. 1830-1850: Real wage of craftsmen in Amsterdam; H. Nus-teling, Wetvaart en Werkgelegenheid in Amsterdam 1540-1860, De Ba-taafsche Leeuw, Amsterdam, 1985, Table 6.3, p. 265. Nominal wage. 1850-1913: Wages of unskilled laborers in building trades; based on A. Vermaas, "Real Industrial Wages in the Netherlands, 1850-1913," paper presented to the conference on International Migration and Labor Market Integration in the 19th and 20th Centuries, Bellagio, Italy, June 14-18, 1993, unskilled component by personal communication. 1913-1939: Daily wages in industry; D. J. van der Veen and J. L. van Zanden, "Real Wage Trends and Consumption Patterns in the Netherlands, 1870-1940," in P. Scholliers (Ed.), Real Wages in 19th and 20th Century Europe, pp. 205-228. 1939-1969: Monthly/weekly industrial wages; Mitchell, EHS, p. 76. 1969-1988: Hourly wages in manufacturing; OECD, MEI, p. 514. Cost of living. 1850-1913: Cost of living index, 1910 weights; A Vermaas, "Real Industrial Wages in the Netherlands," Appendix 1. 1913-1939: van der Veen and van Zanden, "Real Wage Trends," pp. 227-228. 1939-1969: Cost of living index; Mitchell, EHS, p. 282. 1969-1988: Consumer price index; OECD, MEI, p. 517. Norway (NOR) Nominal wage. 1850-1940: Average daily wage, workers at public road and railway construction; F. Hodne, O. Grytten, and J. Alme, "Norwegian Real Wages: Trends in Prices and Wages, 1850-1940," paper 174 JEFFREY G. WILLIAMSON presented to the conference on International Migration and Labor Market Integration in the 19th and 20th Centuries, Bellagio, Italy, June 14-18, 1993, Appendix 1, Table 4.3; based on data from Norges Officielle Statistik and the Wedervang archive. 1940-1969: Hourly wages in industry, adult males; Mitchell, EHS, p. 198. 1969-1988: Hourly earnings in manufacturing, males; OECD, M EI, p. 548. Cost of living. 1850-1940: Ramstad-CBS index as constructed in F. Hodne, O. Grytten, and J. Alme, "Norwegian Real Wages," Appendix I, Table 3.1. 1940-1969: Cost of living index; Mitchell, EHS, pp. 780-782. 1969-1988: Consumer price index; OECD, ME1, p. 550. Portugal (POR) Nominal wage. 1850-1912: Unskilled urban nominal wage index con-stucted using series I, J, M, and N from D. Justino, "Prenos e Salarios em Portugal (1850-1912)," História Económica, Banco de Portugal, Lisbon, no date, p. 22. 1913-1930: Data not available. 1931-1955: Wages of laborers in building trades; International Labour Review and Yearbook of Labour Statistics, ILO, Geneva, various issues. 1955-1988: Industrial wage index; B. R. Mitchell, International Historical Statistics: Europe 1750-1988 (hereafter IHSE), Stockton Press, New York, 1992, pp. 186-187; supplemented by agricultural wage index when industrial wage data are missing (1969-1971), pp. 194-195. Cost of living. 1850-1912: Surrogate cost of living index; A. B. Nunes, E. Mata, and N. Valério, "Portuguese Economic Growth, 1833-1985," Journal of European Economic History 18, 1989, pp. 292-295. 1913-1930: Data not available. 1931-1988: Cost of living index; Mitchell, IHSE, pp. 849-851. Real wage. 1927-1930: Extrapolation from the 1931-1988 real wage series based on data described above. Spain (SPA) Nominal wage. 1830-1899: Unweighted average of wage indexes, workers in Madrid building trades ("peon" and "official" series); D. Reher and E. Ballesteros, "Precios y salarios en Castilla la Nueva: la construction de un indice de salarios reales, 1501-1991," Revista de História Económica II, 1(1993), Appendix 1, Table 1, in press. 1899-1925: Weekly wage of unskilled laborers in textiles (wool); E. Déu, "Els salaris de la industria textil llanera a Sabadell 1896-1925," UB Conference paper on Wages and Labor Markets in Spain, Barcelona, March 1987. EVOLUTION OF GLOBAL LABOR MARKETS 175 1925- 1933: Average wage in the textile industry; J. Maluquer de Motes, "Precios, Salarios y Beneficios. La Distribuci6n Funcional de la Renta," in A. Carreras (Ed.), Estadisticas Historicas de Espana, S.X1X-XX, Fun-dacion Banco Exterior, Madrid, 1989, Table 12.14, p. 520. 1933-1941: Unweighted average of wage indexes, workers in Madrid building trades ("peon" and "oficial" series); D. Reher and E. Ballesteros, "Precios y salarios en Castilla la Nueva," Appendix 2, Table 6. 1941-1963: Hourly wage of unskilled laborers in railways; S. Garcia, "Los Salarios de la 'Maquinista' 1940-1985," UB Conference paper on Wages and Labor Markets in Spain, Barcelona, March 1987. 1963-1983: Average hourly wage of unskilled laborers in the building trades; Maluquer de Motes, "Precios, Salarios y Beneficios," Table 12.12, p. 523. 1983-1988: Hourly earnings, all activities; OECD, MEI, p. 592. Cost of living. 1830-1899: General price index; D. Reher and E. Ballesteros, "Precios y salarios en Castilla la Nueva," Appendix 1, Table 1. 1899-1913: Price index reported in F. Bustelo and G. Tortella-Casares, "Monetary Inflation in Spain, 1800-1970," The Journal of European Economic History 5(1), 1976, Table I, p. 142, and Table 2, p. 145. 1913-1933: Consumer price index (Barcelona); Maluquer de Motes, "Precios, Salarios y Beneficios," Table 12.12, pp. 518-519. 1933-1941: General price index; D. Reher and E. Ballesteros, "Precios y salarios en Castilla la Nueva," Appendix 1, Table 1. 1941-1983: Consumer price index; Maluquer de Motes, "Precios, Salarios y Beneficios," Table 12.16, pp. 521-522. 1983-1988: Consumer price index; OECD, MEI, p. 595. Sweden (SWE) Nominal wage. 1830-1860: Daily wage, males, in agriculture (average of counties, mid-point average for 5-year periods, interpolating between mid-points); L. Jorberg, A History of Prices in Sweden, 1732-1914, CWK Gleerup, Lund, Sweden, 1972, Vol. II, p. 229. 1860-1913: See Real wage below. 1913-1926: Daily earnings in industry; Statistik Arsbok, Helsingfors, 1924, Table 168, p. 194, and subsequent issues. 1926- 1938: Daily wage in industry, commerce and communications; Mitchell, EHS, p. 196. 1938-1950: Hourly wage, adult males, in industry, commerce and communications; ibid., p. 198. 1950-1969: Hourly wage, adult males, in industry; ibid., p. 198. 1969-1988: Hourly wage, manufacturing and mining; OECD, MEI, p. 624. Cost of living. 1830-1860: Cost of living index (mid-point average for 176 JEFFREY G. WILLIAMSON 5-year periods, interpolating between mid-points); L. Jorberg, A History of Prices, p. 350. 1860-1913: See Real wage below. 1913-1926: Cost of living index; Staff of the Institute for Social Sciences, University of Stockholm, Wages, Cost of Living and National Income in Sweden, 1860-1930, P. S. King, London, 1933, Vol. I, col. 7, Table 28, p. 189. 1926-1969: Cost of living index; Mitchell, EHS, pp. 782-783. 1969-1988: Consumer price index; OECD, MEI, p. 627. Real wage. 1860-1913: Real unskilled industrial wage Index constructed from Staff of the Institute for Social Sciences, University of Stockholm, Wages, Cost of Living and National Income in Sweden, 1860-1930, Vol. II. United States of America (USA) Nominal wage. 1830-1856: Unweighted average of: (a) wages of civilian common laborers and teamsters at army forts in the northeastern United States, from C. Goldin and R. A. Margo, "Wages, Prices, and Labor Markets before the Civil War," in C. Goldin and H. Rockoff (Eds.), Strategic Factors in Nineteenth Century American Economic History: A Volume to Honor Robert W. Fogel, Univ. of Chicago Press: Chicago, 1992, Table 2A.2, p. 95; and (b) wages of nonfarm common or unskilled labor, from P. A. David and P. Solar, "A Bicentenary Contribution to the History of the Cost of Living in America," Research in Economic History 2, 1977, Table B.l, pp. 59-60. 1856-1889: Wages of urban unskilled workers; J. G. Williamson and P. H. Lindert, American Inequality: A Macroeconomic History, Academic Press, New York, 1980, Appendix G, p. 319. 1889-1945: Wages of nonfarm common or unskilled labor, from P. A. David and P. Solar, "A Bicentenary Contribution," Table B.l, pp. 59-60. 1945-1985: Average hourly earnings, all manufacturing, in U.S. Department of Commerce, Historical Statistics of the United States, Part I, Bureau of the Census, 1975, Table D-802, pp. 169-170, and ILO, Yearbook, various issues. 1985-1988: Hourly wages in manufacturing; OECD, MEI, p. 76. Cost of living. 1830-1856: Unweighted average of: (a) price index for the northeastern United States, from C. Goldin and R. A. Margo, "Wages, Prices, and Labor Markets," Table 2A.4, p. 97; and (b) index of consumer prices, from P. A. David and P. Solar, "A Bicentenary Contribution," Table 1, pp. 16-17. 1856-1974: Index of consumer prices; P. A. David and P. Solar, "A Bicentenary Contribution," Table 1, pp. 16-17. EVOLUTION OF GLOBAL LABOR MARKETS 177 1974-1985: ILO, Yearbook, various issues. 1985-1988: Consumer price index; OECD, MEI, p. 79. APPENDIX 2 International Real Wage Series The national real wage indices in Appendix 1 have been used in conjunction with benchmark real wage comparisons in Appendix 3 to calculate international real wage indices for which the real wage in GB for some given year (Table A2.1, 1905; Table A2.2, 1927; Table A2.3, 1975) is set equal to 100. The derivation of these indices utilizes the identity where w\ is the real wage in country i in year t. The ratios w'Jw'r&nd w?B/w?B are calculated from the individual national real wage indices. The term w'r/w?B is the benchmark comparison—derived using nominal wage quotations and purchasing power parity measures—of real wages in country i in year r and real wages in GB in year s. Via the additional identity the values of vvJ/vv,GB have been used along with the GB national real wage index to derive a set of international real wage indices in which each country's real wage is expressed relative to the real wage in GB in a particular year q. These indices are summarized in Tables A2.1-A2.3. Please note that each country's real wage in each year is expressed relative to GB = 100 in year q. For those using these data for country time series, this basing will not matter (and the results should be consistent with Table A 1.1). For those interested in wage gaps between countries, divide one series by the other. A few of the values in Tables A2.1-A2.3 have been calculated based on interpolation or extrapolation of the underlying w'Jw^jB series. Results derived by linear interpolation are accompanied by a t; those derived by extrapolation of a five-year trend are followed by the letter e. An asterisk, *, accompanies those values in Tables A2.1-A2.3 which, for the sake of sample size consistency, have been excluded from the dispersion calculations summarized in Figs. 1 and 3-8 of the text. APPENDIX 3 Purchasing Power Parities and Real Wage Benchmarks Purchasing Power Parities and Real Wages, 1905-1914 Notes and definitions. The relevant data are presented in Tables A3.1 and A3.2. (W',/Wr) Wr TABLE A2.1 International Real Wages, 1830-1913 (100 = GB Real Wage in 1905) Year ARG AUS CAN USA BEL DEN FRA GER GB IRL ITA NET NOR SPA SWE BRZ POR 1830 74 36 53 32 36 63 24 37 1831 75 37 54 31 32 53 24 36 1832 76 39 57 33 34 45 24 48 1833 72 45 61 35 36 47 24 21 1834 82 46 63 36 39 57 24 21 1835 73 44 61 39 43 51 24 32 1836 85 43 54 34 41 46 24 33 1837 96 42 54 33 43 40 23 30 1838 97 40 51 31 36 39 23 26 1839 81 38 49 30 35 45 24 29 1840 78 38 51 31 35 46 24 23 1841 85 42 53 32 33 52 24 25 1842 96 41 57 34 33 49 24 37 1843 107 44* 42 63 38 35 54 23 36 1844 116 44* 42 60 36 37 56 23 36 1845 110 42* 42 62 37 34 63 23 30 1846 108 42* 40 59 36 31 50 23 28 1847 85 39* 38 51 30 25 43 23 32 1848 101 38* 41 63 37 37 47 23 30 1849 109 41* 43 69 41 38 61 23 35 1850 106 42* 45 67* 69 47 45 22* 63 23 40 34 1851 107 45* 45 60* 71 43 44 21* 60 23 37 40 1852 110 47* 44 51* 70 41 39 21* 63 23 39 40 1853 106 48* 43 48* 58 44 40 21* 56 23 46 32 1854 124 102 42 42 44 52 39 38 24 58 24 36 28 1855 104 101 42 41 42 51 43 38 25 56 24 35 32 1856 118 104 43 42 44 53 47 40 26 47 25 38 25 1857 111 100 46 43 56 57 42 42 28 49 25 37 32 1858 106 106 48 48 57 68 50 47 27 50 26 42 37 1859 104 105 52 50 56 67 48 46 27 54 26 43 32 1860 119 107 51 46 54 58 43 46 27 52 27 48 34 1861 124 103 49 44 46 56 41 43 26 49 33 48 33 1862 105 92 45 45 54 59 41 44 27 51 40 48 37 1863 92 83 49 45 58 63 40 40 29 48 38 44 38 1864 42* 106 77 55 45 63 64 40 41 27 47 34 42 36 EVOLUTION OF GLOBAL LABOR MARKETS 179 r<-, c^i o j r* i cN r» cn i tT e*i cn i r*"} f*"! f*i r*i r*i ^ ^ ""t vi v> *0 *o '•Ü *o t I ■ (NDNNNnMiNnMD^fnr, rr, m f*") fl f«"» f*^ ci r«"i n c n n n n 'T ^ t ^ w> ^ *c r—• r» r— ]00000OM<»00O0 00 0O00C»0OQ00ri00 00Ovr>i TABLE A2.1- Continued Year ARG AUS CAN USA BEL DEN FRA GER GB IRL ITA NET NOR SPA SWE BRZ POR 1903 100 129 174 164 85 89 71 83 100 92 42 73 68 48 87 88 40 1904 107 134 183 163 86 92 73 85 100 92 43 68 74 48 90 88 45 1905 94 124 174 167 81 90 75 84 100 92 44 73 78 48 88 94 48 1906 92 12? 192 172 90 92 76 87 110 91 45 70 80 52 88 95 47 1907 89 -.4 184 171 84 94 76 89 108 89 45 72 81 50 90 85 43 1908 89 122 193 166 84 94 75 86 107 93 48 73 76 52 95 88 43 1909 88 123 197 172 85 100 76 86 107 93 50 71 75 52 96 86 44 1910 95 135 205 170 87 99 71 87 105 91 50 70 70 52 100 85 42 1911 94 137 210 166 84 105 64 88 104 90 51 69 73 54 103 90 40 1912 97 132 204 165 88 104 66 87 101 89 54 71 73 51 99 80 40 1913 92 128 219 169 94 102 66 92 110 90 55 72 82 51 98 87 TABLE A2.2 International Real Wages, 1914-1945 (100 = GB Real Wage in 1927) ARG AUS CAN USA BEL DEN FRA GER GB 1RL ITA NET NOR SPA SWE BRZ FOR 1914 71 111 120 124 41 92 69 86 85 88 50 78 101 39 79 48 1915 64 98 116 124 35t 80+ 60 77 74 80 50 69 88 38 73 41 1916 60 101 115 142 31 + 74+ 53 68 69 70 44 67 84 37 72 38 1917 51 101 105 150 29+ 74 50 61 70 75 42 69 90 37 70 30 1918 44 98 110 161 32+ 83 50 63 80 86 40 78 82 37 73 24 1919 60 94 115 163 35+ 127 53 71+ 91 101+ 53 95 103 39 79 24 1920 62 98 110 161 36 139 60 75+ 97 109+ 58 96 105 50 91 22 1921 76 127 113 150 42 158 70 80+ 106 121+ 60 118 112 53 107 33 1922 88 130 115 147 47 144 70 72+ 96 111+ 57 122 109 55 95 32 1923 90 126 122 159 45 132 71 69+ 93 111+ 54 121 105 54 97 27 1924 89 131 122 164 42 132 70 70 95 115 52 117 99 53 98 28 1925 93 131 115 159 43 142 71 80 96 103 52 116 101 53 99 27 1926 94 133 113 160 37 149 67 86 97 103 52 121 105 56 104 34 1927 99 137 117 166 41 147 62 87 100 110 56 123 105 54 105 36 22 1928 106 137 120 169 44 147 67 91 99 105 56 125 109 56 106 35 22+ 1929 105 134 122 174 48 149 71 94 100 106 41 130 115 51 112 36 23+ 1930 95 140 126 175 50 159 71 99 103 107 55 135 118 57 118 42 24+ 1931 102 141 128 185 48 168 71 102 109 113 57 141 125 56 121 43+ 26 1932 109 139 126 180 42 171 73 99 109 114 55 139 120 58 121 42+ 28 1933 100 139 125 190 43 166 77 97 112 117 57 137 120 64 119 42+ 30 1934 103 137 126 219 45 163 81 93 110 116 57 132 124 62 117 40+ 30+ 1935 106 136 142 221 43 157 83 90 110 114 55 131 123 62 120 39+ 30 1936 99 136 143 221 43 157 78 88 110 111 51 134 124 61 120 38+ 29+ 1937 100 137 132 243 50 154 79 87 109 105 48 132 128 60 121 37 29t 1938 100 142 131 254 52 160 74 88 111 109 47 133 130 52 125 29+ 1939 102 141 132 262 51 163 72 88 109 108 54 134 132 45 130 28 1940 103 139 135 266 144 72 86* 108 94 53 123 118 40 122 27 1941 101 140 131 283 138 66 84* 107 88 48 115 109+ 33 115 29 1942 101 139 152 290 139 56 82* 115 80 47 107 109+ 37 118 23 1943 105 144 163 302 148 48 82* 120 74 34 104 106+ 37 120 25 1944 119 145 169 311 155 68 125 74 101 101 36 121 29 1945 112 146 170 312 164 76 128 76 103 115 35 127 29 TABLE A2.3 International Real Wages, 1946-1988 (100 = GB Real Wage in 1975) ARG AUS CAN USA BEL DEN FRA GER GB IRL ITA NET NOR SPA SWE BRZ POR 1946 20 37 50 98 50 23 47 26 18 50 48 36 48 17 19 1947 24 39 50 98 41* 51 24 49 32 29 49 52 32 52 14 20 1948 29 40 49 97 40* 57 23 47 35 37 49 55 30 56 14 21t 1949 30 40 50 101 42* 58 23 48 37 39 47 57 30 57 18 23 1950 29 41 51 105 44 56 24 34 49 37 41 47 58 28 58 19 23 1951 27 42 54 105 45 56 23 35 49 34 42 45 57 26 60 18 22 1952 25 50 56 109 47 60 26 37 48 41 42 46 58 31 66 18 22 1953 26 51 61 114 46 62 28 40 50 42 42 47 60 29 69 15 22 1954 29 51 62 116 46 63 31 41 53 42 43 50 61 28 71 19 22 1955 29 51 65 122 48 61 33 43 55 41 44 52 63 30 75 21 21 1956 30 50 67 125 49 63 34 46 57 43 44 53 66 30 78 21 21 1957 30 51 70 128 51 66 35 47 58 42 45 56 68 40 78 22 21 1958 32 51 74 128 51 67 35 48 58 42 45 57 67 36 80 22 22 1959 24 52 78 132 52 71 36 50 61 43 46 58 72 35 83 22 23 1960 25 53 84 134 54 73 37 54 62 46 47 61 74 36 84 22t 24 1961 28 53 83 136 55 79 39 59 66 46 47 63 78 37 90 24t 26 1962 27 54 88 138 58 82 42 62 65 50 48 65 81 35 93 24t 27 1963 27 54 90 141 61 83 44 65 67 50 50 66 83 33 96 25 27 1964 30 55 93 143 65 87 46 68 70 53 54 72 83 34 102 25 28 1965 33 55 96 145 68 91 49 73 73 51 56 76 87 35 106 24 30 1966 33 55 109 147 71 96 50 74 72 51 56 79 90 37 108 29 31 1967 33 57 114 148 73 98 53 74 73 67 58 81 93 42 113 29 31 1968 31 59 117 151 76 102 59 77 75 70 59 83 97 44 118 31 32 1969 32 61 128 152 79 109 60 83 78 74 62 87 103 47 125 32 32 1970 33 63 131 151 85 114 63 90 84 77 72 92 101 51 132 32 34 1971 34 67 144 154 88 124 67 97 86 86 77 97 107 54 133 34 34 1972 32 68 152 160 96 130 70 100 91 83 81 102 109 59 143 43 36 1973 34 73 157 161 105 137 75 103 96 88 91 106 111 64 146 47 39 1974 38 82 155 157 112 142 78 107 98 91 93 116 120 75 147 47 48 1975 37 78 162 157 118 154 82 110 100 95 100 119 130 82 154 53 59 1976 24 78 171 161 121 158 85 112 101 98 105 118 139 92 165 55 61 EVOLUTION OF GLOBAL LABOR MARKETS * TT in ifl irt in m o \C r-IT, in ^ 't O fS r*l in m m 00 \C —• \fi <ß in O P 2 3 p \o ^ \o •£> « X \0 vC >n m W) o 00 ir-, ^ r*i r^, O0 CT- o *n fN N fl m r- vc m r-N N (N rj — oj rN c m r- CT M N N O n-i Q rr, m H Ifl O (N rs (N t- r^j ^ CTv o © o o OJ (N CN (N g g n >—< r— r*j rs ri m NO» r**, r^i rs h oo a oc" 00 <> ct (N m t ^ CO CO OO 00 CT* 0*> CT> CTv vO h « oo oo oc TABLE A3.1 Purchasing Power Parities and Real Wages, 1905-1914 USA CAN AUS ARG DEN IRL GB SWE GER FRA BEL ITA Year 1909 1905 1905 1914 1905 1905 1905 1914 1905 1905 1905 1905 d $mn d d ore d d d lira 00 Tea and coffee (lb) 15.415 — — 0.950 — 22.750 18.000 71.202 11.000 19.250 7.625 2.023 Sugar (lb) 2.750 — - 0.177 - 1.979 2.000 29.025 2.375 2.875 3.000 0.676 Bacon and sausage (lb) 9.250 - — 0.727 - 8.667 8.000 61.224 9.525 9.611 7.875 0.726 Beef and veal (lb) 7.375 — — 0.179 - 8.750 8.000 47.779 8.250 8.438 7.938 0.862 Pork (lb) 6.500 — — 0.262 — 8.200 8.000 58.050 10.000 9.250 8.500 0.762 Lamb and mutton (lb) 7.375 — - 0.109 - 8.542 8.250 61.678 9.750 8.375 7.000 0.621 Cheese (lb) 10.000 — — 0.405 - 8.083 7.0O0 53.968 7.500 8.831 8.500 0.590 Butter and margarine (lb) 16.750 — - 0.650 - 13.750 13.000 88.504 13.875 12.500 13.000 1.270 Potatoes (7 lb) 7.000 — — 0.350 - 2.583 3.000 20.317 2.625 3.000 2.750 0.381 Flour and meal (7 lb) 12.500 — — 0.509 — 9.417 9.000 96.466 12.625 13.750 9.625 1.206 Bread (4 lb) 11.125 — — 0.364 — 5.417 5.000 77.040 5.625 5.750 4.750 0.653 Milk (qt) 4.500 — — 0.160 - 2.750 3.500 10.727 2.625 2.500 2.250 0.227 Eggs (doz) 14.400 - - 0.810 - 16.248 12.000 102.600 9.842 12.000 11.077 1.261 Rent d $mn d d ore d d d lira 3 rooms/week 32.667 - - - - 13.167 23.250 224.359 24.500 19.083 10.417 - Exchange rates (per £) $ £ £ $mn Kr £ £ Kr mark Fr Fr lira £ 4.86 1.00 1.00 11.45 18.16 1.00 1.00 18.16 20.43 25.23 25.23 25.23 PPP 6.48 - - 14.84 18.71 0.94 1.00 19.27 21.05 25.78 20.35 25.88 Price levels PF 132 — — 105 — 105 100 102 103 107 92 103 PR 141 — — 344 — 57 100 128 105 82 45 99 P 133 - - 130 103 94 100 106 103 102 81 103 Nominal wages s/wk $mn/day ore/hr s/wk s/wk Kr/wk s/wk s/wk s/wk s/wk Skilled building 98.14 — — — — 33.94 37.69 31.52 28.54 26.15 22.35 14.00 Unskilled building 59.38 - - 2.83 - 17.00 25.64 27.13 21.75 17.83 16.50 9.00 Skilled engineering 75.88 - - - 45.00 36.25 34.50 - 29.90 27.28 22.21 17.00 Unskilled engineering 40.63 - - - 34.00 17.00 20.00 - 20.00 17.38 14.58 12.00 Real wage index 171.58 173.80 123.52 89.24 89.93 92.33 100.00 98.29 84.26 74.58 80.86 44.23 a «: o r r C/i O Z EVOLUTION OF GLOBAL LABOR MARKETS 185 TABLE A3.2 Purchasing Power Parity Budget Shares, 1905-1914 USA GB SWE GER FRA BEL ITA Average Year 1909 1912 1914 1905 1905 1905 1905 Tea and coffee 0.067 0.068 0.040 0.047 0.051 0.051 0.051 0.053 Sugar 0.051 0.059 0.084 0.028 0.031 0.020 0.038 0.044 Bacon and sausage 0.053 0.058 0.040 0.143 0.038 0.072 0.048 0.064 Beef and veal 0.199 0.180 0.060 0.121 0.151 0.155 0.106 0.139 Pork 0.056 0.020 0.041 0.088 0.034 0.054 0.072 0.052 Lamb and mutton 0.041 0.062 0.003 0.014 0.230 0.013 0.018 0.054 Cheese 0.017 0.033 0.025 0.024 0.031 0.021 0.039 0.027 Butter and margarine 0.113 0.128 0.190 0.125 0.071 0.178 0.018 0.117 Potatoes 0.069 0.055 0.041 0.062 0.047 0.079 0.044 0.057 Flour and meal 0.076 0.064 0.091 0.020 0.000 0.020 0.148 0.060 Bread 0.081 0.138 0.128 0.196 0.223 0.243 0.298 0.187 Milk 0.091 0.076 0.202 0.089 0.050 0.051 0.078 0.091 Eggs 0.087 0.060 0.056 0.043 0.044 0.045 0.044 0.054 Total 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 Food 0.733 0.800 0.745 0.836 0,857 0.888 0.880 0.820 Rent 0.267 0.200 0.255 0.164 0.143 0.113 0.120 0.180 Total 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 E Official exchange rate. PPP Purchasing power parity based on food and rent relative prices. PF Relative price of food converting via E. PR Relative rent converting via E. P Relative price of food and rent based on weighted average of PF and PR. The principal sources used are abbreviated as follows: BOTBEL Board of Trade, "Report of an Enquiry by the Board of Trade into Working Class Rents, Housing and Retail Prices Together with the Rates of Wages in Certain Occupations in the Principal Industrial Towns of Belgium," Parliamentary Papers, 1910, Accounts and Papers (37), cd. 5065, vol. XCV. BOTFRA Board of Trade, "Report of an Enquiry by the Board of Trade into Working Class Rents, Housing and Retail Prices with the Rates of Wages in Certain Occupations in the Principal Industrial Towns of France," Parliamentary Papers, 1909, Accounts and Papers (42), cd. 4512, vol. XCI. 186 JEFFREY G. WILLIAMSON BOTGER Board of Trade, "Report of an Enquiry by the Board of Trade into Working Class Rents, Housing and Retail Prices Together with Rates of Wages in Certain Occupations in the Principal Industrial Towns of the German Empire," Parliamentary Papers, 1908, Accounts and Papers (47), cd. 4032, vol. CVIII. BOTUK1 Board of Trade, "Report of an Enquiry by the Board of Trade into Working Class Rents, Housing and Retail Prices Together with the Standard Rates of Wages Prevailing in Certain Occupations in the Principal Industrial Towns of the United Kingdom," Parliamentary Papers, 1908, Accounts and Papers (46), cd. 3864, vol. CVII. BOTUK2 Board of Trade, "Report of an Enquiry by the Board of Trade into Working Class Rents and Retail Prices with the Rates of Wages in Certain Occupations in Industrial Towns of the United Kingdom in 1912," Parliamentary Papers, 1913, Accounts and Papers (26), cd. 6955, vol. LXVI. BOTUSA Board of Trade, "Report of an Enquiry by the Board of Trade into Working Class Rents, Housing and Retail Prices Together with the Rates of Wages in Certain Occupations in the Principal Industrial Towns of the United States of America," Parliamentary Papers, 1911, Accounts and Papers (44), cd. 5609, vol. LXXXVIII. ISS The Institute for Social Sciences, University of Stock- holm, Wages, Cost of Living and National Income in Sweden 1860-1930, London, P. S. King, 1933. The following abbreviations are used for countries and their currencies: USA United States $ CAN Canada £ AUS Australia f ARG Argentina $mn DEN Denmark Kr IRL Ireland f GB Great Britain f SWE Sweden Kr GER Germany mark FRA France Fr BEL Belgium Fr ITA Italy lire U.S. dollar. Pound sterling (see GB). Pound sterling (see GB). Peso moneda nacional (peso papel). Kronor; 1 kronor = 100 ore. Pound sterling (see GB). Pound sterling; s, shilling; d, penny; £1 = 20s = 240d. Kronor; 1 kronor = 100 ore. German mark. French franc. Belgian franc. Italian lire. Data are not available to calculate PPPs for BRZ, POR, NET, NOR, EVOLUTION OF GLOBAL LABOR MARKETS 187 and SPA. Interwar benchmarks are carried back through the pre-World War I period for NET, NOR, and SPA. The BRZ real wage is set equal to ARG in 1905, while the POR real wage is set equal to SPA in 1905. Budget shares. These are detailed in Table A3.2. A simple average of budget weights in the seven-country Board of Trade sample is used. These are simply expenditure shares of the given items in the food consumption basket computed from price and quantity data, and the weights scaled to sum to unity. The average represents an international budget weight and is used for all countries in the sample. The food and rent shares are drawn from budget data in the same sources, again with weights scaled to sum to unity. Food prices. Averages used as appropriate. GB BOTUK1, pp. xiv, xxxiii, xvi-xxviii. GER BOTGER, pp. xx-xxi, xxv, Appendix III. FRA BOTFRA, pp. xvii, xxii. BEL BOTBEL, pp. xiii, xvii. ITA V. Zamagni, "The Daily Wages of Italian Industrial Workers in the Giolittian Period (1898-1913)," Rivista di Storia Economica 1 (International Issue), 1984, p. 90, except prices and budget shares of tea and coffee and eggs taken from FRA data. Other weights re-normalized as appropriate. SWE ISS, vol. I, pp. 161-162. USA BOTUSA, pp. xxix, xlvi, 1. IRL BOTUK2, pp. 286-297. ARG Tornquist, Ernesto & Co., Ltd., The Economic De- velopment of the Argentine Republic in the Last Fifty Years, Buenos Aires, Tornquist, 1919, pp. 267-269. Ham proxies for bacon. Rent. Averages used as appropriate. Based on rates for three rooms. GB BOTUK1, pp. xiv. London and towns, minimum and maximum. GER BOTGER, pp. xiii. Berlin and towns, minimum and maximum. FRA BOTFRA, pp. xii. Paris and towns, minimum and maximum. BEL BOTBEL, pp. xiii, xvii. Towns, minimum and max- imum. SWE ISS, vol. I, p. 178. Note that 1 kronor = 100 ore. Kitchen counts as one room. USA BOTUSA, pp. xxii. Towns, minimum and maximum. 188 JEFFREY G. WILLIAMSON IRL BOTUK2, pp. 286-297. Belfast, Cork, Dublin, Lim- erick, Londonderry, and Waterford. Exchange rates. E is the pre-World War I (gold standard) par value. Quotes are taken from The Economist, London, January 3, 1920, p. 37. Purchasing power parities and price levels. Cobb-Douglas indices throughout, weighting by budget weights. GB, GER, FRA, BEL, SWE, USA, ITA, ARG: PPPF and PF from price data, E, and international average budget shares (Table A3.2). ITA: PR from Zamagni, "The Daily Wages of Italian Industrial Workers in the Giolittian Period (1898-1913)," p. 91. Implicit PPPR derived via E. ARG: PR from A. E. Bunge, Riqueza y renta de la Argentina, su distribuciön y su capacidad contributiva, Buenos Aires, Agencia General de Libreria y Publicaciones, 1917, p. 267. Based on three-room rate. GB, GER, FRA, BEL, SWE, USA: PPPR and PR from rent data. P and PPP derived from PF and PR using international average food and rent budget weights (Table A3.2) and E. DEN: Use P from GER. Nominal wages. In Board of Trade sources, skilled building is an average wage for bricklayers, masons, carpenters, stonemasons, joiners, cabinet makers, plumbers, plasterers, iron workers, stucco workers, and painters (as available); unskilled building is an average wage for bricklayers' laborers, masons' laborers, plasterers' laborers, and other laborers; skilled engineering is an average wage for moulders, fitters, turners, machinists, smiths, and pattern-makers; unskilled engineering is an average wage for laborers. GB GER FRA BEL ITA SWE USA IRL DEN BOTUK1, p. xxxiii. England and Wales. BOTGER, p. xxx. Municipal workers excluded. BOTFRA, p. xxvi. Municipal workers excluded. BOTBEL, p. xxi. Zamagni, "An International Comparison of Real Industrial Wages 1890-1913: Methodological Issues and Results," in P. Scholliers (Ed.), Real Wages in 19th and 20th Century Europe, Table 5.1. ISS, Vol. II, p. 27. Municipal workers only, as a proxy for building trades. Based on a 57-hour week (see p. 15). BOTUSA, p. xvi. BOTUK1, p. xxxiii. Hourly wages of skilled and unskilled males in crafts and industry; H. Chr. Johansen, Dansk 0konomisk Statistik, 1814-1980, Copenhagen, 1985, pp. 294-296. EVOLUTION OF GLOBAL LABOR MARKETS 189 ARG Daily wage of peones albaniles, representing unskilled work in the building trades. Republica Argentina, Bol-etin del Departamento Nacional del Trabajo, Buenos Aires, no. 33, Jan. 30, 1916, p. 132. Real wages. AUS and CAN: For these countries analagous calculations comparing real wages in terms of purchasing power have already been performed by R. Allen, "Real Incomes in the English Speaking world," mimeograph, Vancouver, University of British Columbia, November 1990, pp. 45-46. Average of bricklayers and laborers relative real wage index (GB = 100). For AUS use Sydney, for GB use Manchester, and for CAN average Toronto and Vancouver. All other countries: Simple average of relative real wage index (GB = 100) in all categories available, with conversion using PPP. DEN, assume a 56-hour week for comparison; ARG, assume a 6-day week for comparison; BRZ, real wage set equal to ARG in 1905; POR, real wage set equal to SPA in 1905. Purchasing Power Parities and Real Wages, 1927 Notes and definitions. The relevant data are presented in Tables A3.3 and A3.4. E Official exchange rate. PPP Purchasing power parity based on food and rent relative prices. PF Relative price of food converting via E. PR Relative rent converting via E. P Relative price of food and rent based on weighted average of PF and PR. The principal sources used are International Labour Office, Yearbook of Labour Statistics, Geneva (hereafter denoted YLS or YLS x, where x denotes a particular year), and International Labour Office, International Labour Review, Geneva (hereafter denoted ILR). The following abbreviations are used for countries and their currencies: USA United States $ CAN Canada $ AUS Australia £ ARG Argentina peso DEN Denmark D Kr IRL Ireland £ GB Great Britain £ SWE Sweden S Kr GER Germany RM U.S. dollar. Canadian dollar. Pound sterling (see GB). Peso moneda nacional (peso papel). Danish kronor. Pound sterling (see GB). Pound sterling; s, shilling; d, penny; £1 = 20s = 240d. Swedish kronor. German Reichmark. TABLE A3.3 Purchasing Power Parities and Real Wages, Interwar Period Year Food prices ARG 1938 peso AUS 1927 d CAN 1927 cent USA 1927 cent BEL 1927 B Fr DEN 1927 D Kr FRA 1927 F Fr GER 1927 RM GB 1927 d IRL 1927 d ITA 1927 lira NET 1927 G NOR 1926 N Kr SPA 1927 pta SWE 1927 S Kr BRZ 1937 mil POR 1927 esc © Bread (kg) 0.32 6.9 17.0 20.7 2.56 0.83 2.10 0.78 5.0 5.8 2.00 0.25 0.88 0.65 0.74 1.60 2.40 Rye Bread (kg) (0.32) (6.9) (17.0) (20.7) 2.50 0.24 (2.10) 0.46 (5.0) (5.8) (2.00) 0.18 0.46 (0.65) 0.64 1.50 2.40 Flour (kg) 0.19 5.1 12.8 11.5 3.51 0.40 3.05 0.58 6.0 5.7 2.20 0.29 0.52 0.80 0.47 1.80 3.80 Rice (kg) 0.46 8.5 24.9 24.9 5.29 0.84 4.80 0.68 6.4 6.6 2.33 0.25 1.01 1.00 0.52 1.60 2.40 Beef (kg) 0.40 17.9 52.0 84.5 25.34 2.73 19.00 2.30 41.5 37.5 9.50 1.40 3.21 5.40 2.09 2.50 8.60 Bacon (kg) 0.75 52.9 78.5 100.8 17.08 2.35 13.80 2.80 36.4 33.1 8.00 1.00 2.44 3.20 2.37 4.50 10.00 Milk (liter) 0.15 9.6 8.8 11.5 1.90 0.33 1.60 0.33 6.1 5.3 1.50 0.17 0.35 0.80 0.26 1.05 2.20 Butter (kg) 1.60 59.o 93.7 119.5 29.88 3.85 23.55 4.30 49.0 46.3 17.50 2.53 4.43 5.00 3.85 9.50 26.00 W Cheese (kg) 0.89 36.5 76.0 88.2 (24.16) 1.73 17.15 1.50 29.1 35.3 13.50 1.40 0.84 5.00 1.81 5.95 20.00 Eggs (each) 0.05 1.8 4.9 4.3 1.31 0.16 1.01 0.14 2.3 2.5 0.70 0.11 0.16 0.20 0.15 0.22 0.55 Potatoes (kg) 0.14 4.8 4.6 7.7 0.94 0.21 0.85 0.10 2.4 1.9 0.85 0.10 0.12 0.27 0.18 0.95 0.80 >< Sugar (kg) 0.42 10.2 ]-"'.8 14.8 3.63 0.60 3.90 0.64 7.9 7.7 7.00 0.52 0.83 1.60 0.64 1.25 3.80 p Coffee (kg) 1.14 57.9 , v'.4 86.4 (21,20) 3.35 24.00 6,60 (61.7) 61.7 29.00 1.39 4.42 9.00 3.60 3.70 10.00 Rent peso t $ D Kr F Fr RM £ £ G S Kr r 3 rooms/month — — _ — _ — — _ 2.50 _ — _ _ — 85.50 — — > 4 rooms/month — 4.36 20.16 43.58 — — - — 2.57 1.74 — 24.42 — — — — — 5 rooms/month - - - - - 54.70 227.4 - 3.24 - - - - - - - o Exchange rates (per £) peso f $ $ B Fr D Kr F Fr RM £ £ lira G N Kr pta S Kr mil esc E 20.55 1.00 4.86 4.86 174.7 18.19 123.9 20.46 1.00 1.00 94.41 12.12 18.66 28.49 18.13 57.30 96.66 PPP 9.27 1.27 5.67 7.57 117.6 16.86 89.1 18.08 1.00 0.90 73.21 8.99 22.08 25.58 20.45 42.29 74.97 Price levels PF 35 116 105 121 64 93 77 92 100 98 85 73 102 104 96 67 86 PR 100 170 161 349 78 93 57 78 100 68 57 78 189 57 189 100 57 P 45 127 117 156 67 93 72 88 100 90 78 74 118 90 113 74 78 Nominal wages/hr peso f $ $ B Fr D Kr F Fr RM £ £ lira G N Kr pta S Kr mil esc Unskilled building 0.630 0.118 0.450 0.850 3.25 1.68 3.75 1.06 0.07 0.07 2.80 0.75 1.57 0.94 1.45 1.05 1.13 Real wage index 100.4 136.8 117.2 165.9 40.8 147.1 62.2 86.6 100.0 109.6 56.5 123.3 105.0 54.1 104.7 36.7 22.2 EVOLUTION OF GLOBAL LABOR MARKETS 191 TABLE A3.4 Purchasing Power Parity Budget Shares, Interwar Period Average Bread 0.114 Rye bread 0.103 Flour 0.048 Rice 0.012 Beef 0.091 Bacon 0.134 Milk 0.139 Butter 0.067 Cheese 0.026 Eggs 0.059 Potatoes 0.050 Sugar 0.064 Coffee 0.091 Total 1.000 Food 0.763 Rent 0.237 Total 1.000 FRA France F Fr French franc. BEL Belgium B Fr Belgian franc. NET Netherlands G Gulden. NOR Norway N Kr Norwegian kronor. ITA Italy lira Italian lira. SPA Spain pta Spanish peseta. BRZ Brazil mil Milreis. POR Portugal esc Escudo. Wage and food price data in particular cities are generally used for the construction of the interwar benchmarks. In particular, wage and food price data for AUS are from Sydney, for CAN are from Ottawa, for USA are from Philadelphia, for BEL are from Brussels, for DEN are from Copenhagen, for FRA are from Paris, for GER are from Berlin, for GB are from London, for IRL are from Dublin, for ITA are from Rome, for NET are from Amsterdam, for NOR are from Oslo, for SPA are from Madrid, for SWE are from Stockholm, for POR are from Lisbon, and for BRZ wages are from Rio de Janeiro while food prices are from Rio de Janeiro and Sao Paulo. Budget shares. Budget shares are based on price and quantity data from a large sample of countries. An international set of budget weights was 192 JEFFREY G. WILLIAMSON computed for selected items in the food basket. Likewise for food and rent shares. Items chosen were common to most countries. Sources are YLS 1933, pp. 497-503, Tables IX, XI, X. YLS 1934, 1935, pp. 144-148, Tables XIV, XV, XVI. YLS 1935, 1936, pp. 177-193, Tables XV, XVI, XVII, XVIII. YLS 1937, pp. 189-192, 195, Tables XVIII, XIX, XX. YLS 1939, pp. 176-183, 187, 192-193, Tables XVII, XVIII, XX, XXII. Food item budget shares were computed for the following countries: Austria, Belgium, Brazil, Bulgaria, Colombia, Czechoslovakia, Denmark, Estonia, Finland, Great Britain, Germany, Hungary, the Netherlands, Norway, Poland, South Africa, Sweden, Switzerland, and the United States. The shares were then averaged across countries. The resulting average budget shares were then rescaled to sum to unity. Note that expenditures on bacon, pork, and lard are combined. Food and rent shares are averages across the following countries: Germany, Austria, Argentina, Belgium, Brazil, Bulgaria, Canada, China, Colombia, Denmark, Estonia, the United States, Finland, Hungary, India, Japan, Latvia, Mexico, Norway, New Zealand, the Netherlands, Poland, Sweden, Switzerland, Czechoslovakia, and South Africa. Based on YLS 1939, Table XX, p. 187. Food prices and rents. Food prices are taken from 1LR, January, 1928. Since the YLS does not report 1927 rents in national currencies, the rent figures for 1927 (with the exception of AUS) are calculated using 1938 rents (in currency units from YLS) which are then carried back to 1927 using time series for national rent indexes. Australian food and rent prices are from the Quarterly Summary of Australian Statistics, December, 1927, p. 79, and have been converted to metric units. There are no rent data reported in YLS for ARG, GER, BEL, NOR, ITA, SPA, BRZ, and POR for 1938 or earlier years. Bread is occasionally used to proxy rye bread prices, the home produce ribs price represents beef prices, Ireland's coffee price proxies Britain's, France's cheese price proxies Belgium's (converted via E), and Portugal's cheese price is for 1926. Proxies are signified by parentheses. Exchange rates. E is the average official exchange rate over the course of the given year; from League of Nations, Statistical Year-Book of the League of Nations 1927, Geneva, 1928, Table 101, pp. 202-207. Purchasing power parities and price levels. Cobb-Douglas indices are constructed using budget weights. PF is based on food prices and food item budget weights. PR is based on rent data, except for ARG (uses 100), GER (uses NET), BEL (uses NET), NOR (uses SWE), ITA (uses FRA), SPA (uses FRA), POR (uses FRA), and BRZ (uses 100). P is based on PF and PR using food and rent shares. PPP is based on P and E. Nominal wages. Nominal wages are average hourly for unskilled building occupations in the given year. The Australian wage is reported in TABLE A3.5 Purchasing Power Parities and Real Wages, 1970s ARG AUS CAN USA BEL DEN FRA GER GB IRL ITA NET NOR SPA SWE BRZ POR Year 1975 1975 1975 1975 1975 1975 1975 1975 1975 1975 1975 1975 1975 1975 1975 1975 1975 Exchange rates (per $) peso A$ C$ $ B Fr D Kr F Fr RM £ E£ lira G N Kr pta S Kr cruz esc E 36.60 1.31 1.02 1.00 36.78 5.75 4.29 2.46 0.45 0.45 652.85 2.53 5.23 57.41 4.15 8.13 25.55 PPP 19.03 1.51 1.02 1.00 39.72 7.07 4.76 2.85 0.41 0.38 581.04 2.73 6.32 41.33 5.02 5.77 17.89 Price level PC 52 115 100 100 108 123 111 116 91 84 89 108 121 72 121 71 70 Nominal wage 21.50 3.59 5.06 4.83 144.32 33.50 U.99 9.69 1.26 1.10 1794 9.95 25.30 104.73 23.79 9.32 32.60 Real wage index 36.75 77.53 161.87 157.14 118.21 154.21 81.99 110.46 100.00 94.57 100.45 118.52 130.15 82.44 154.05 52.58 59.30 TABLE A3.6 International Real Wage Benchmark Summary ARG AUS CAN USA BEL DEN FRA GER GB IRL ITA NET NOR SPA SWE BRZ POR 1905 1909 1914 1926 1927 1937 1938 1975 - 124 174 172 81 90 75 84 100 92 44 - - - - - - 89 - - - - - - - - - 105 - 98 - - 100 37 137 117 166 41 147 62 87 100 110 56 123 54 105 37 22 78 162 157 118 154 82 110 100 95 100 119 130 82 154 53 59 194 JEFFREY G. WILLIAMSON ILR, April, 1927; the Norwegian wage is reported in 1LR, January, 1927; the Brazilian wage is from YLS 1938, and the Argentinian wage is from YLS 1939, Table XIV. All other countries reported in ILR, January, 1928. Real wages. The relative real wage in unskilled building occupations is based on nominal wage rates, with conversion using PPP. Purchasing Power Parities and Real Wages, 1970s Notes and definitions. The relevant data are presented in Table A3.5. E Annual average market exchange rates. PC Price level of consumption. PPP Purchasing power parity based on E and PC. Exchange rates. E from International Monetary Fund, International Financial Statistics Yearbook 1982, Washington, DC. Purchasing power parities and price levels. PC from R. Summers and A. Heston, "Improved International Comparisons of Real Product and its Composition, 1950-1980," Review of Income and Wealth, 1984, series 30, no. 2, pp. 207-262. Nominal wages. Average hourly wage in 1975 of male and female employees in all manufacturing industries. Most observations are from International Labour Office, Yearbook of Labour Statistics 1980 (hereafter denoted YLS 1980, with issues for other years referred to as YLS 19xx) Geneva, 1982, Table 18A. Exceptions are AUS Weighted average of male and female manufacturing wage rates in 1975 from YLS 1980, Table 18A; weights are the number of male and female wage earners (respectively) in all manufacturing industries in 1975, from YLS 1982, Table 5A. ARG Unweighted average of wages in 1975 of unskilled and skilled workers in all manufacturing industries, from YLS 1982, Table 17A. GB Weighted average of male and female manufacturing wage rates in 1975 from YLS 1980, Table 18A; weights are the number of male and female wage earners (respectively) in all manufacturing industries in 1975, from YLS 1980, Table 6A. NOR Average hourly wage in 1975 of male and female employees in all manufacturing industries, from YLS 1982, Table 17A. BRZ Average monthly earnings in manufacturing industries in 1975, from YLS 1980, Table 18A. Real wages. Relative real wage in manufacturing based on nominal wage rates, with conversion using PPP. International real wage benchmark summary presented in Table A3.6. EVOLUTION OF GLOBAL LABOR MARKETS 195 REFERENCES Abramovitz, M. (1956), "Resource and Output Trends in the United States Since 1870." American Economic Review 46, 5-23. Abramovitz, M. (1986), "Catching Up, Forging Ahead, and Falling Behind." Journal of Economic History 46, 385-406. Allen, R. C. (1990), "Real Incomes in the English Speaking World, 1879-1913." Discussion Paper 90-32, Department of Economics, University of British Columbia (November). Baumol, W. J. (1986), "Productivity Growth, Convergence, and Welfare: What the Long- Run Data Show," American Economic Review 76, 1072-1085. Baumol, W. J., Blackman, S. A. B., and Wolff, E. N. (1989), Productivity and American Leadership: The Long View. Cambridge, MA: MIT Press. Boyer, G. R., Hatton, T. J., and O'Rourke, K. (1994), "The Impact of Emigration on Real Wages in Ireland, 1850-1914." In T. J. Hatton and J. G. Williamson (Eds.), Migration and the International Labor Market, 1850-1939. London: Routledge. Chambers, E. J., and Gordon, D. F. (1966), "Primary Products and Economic Growth: An Empirical Measurement." Journal of Political Economy 74, 315-322. Cortes-Conde, R. (1979), El Progreso Argentina, 1880-1914. Buenos Aires: Editorial Su- damericana. DeCanio, S. J., and Mokyr, J. (1977), "Inflation and Wage Lag during the American Civil War," Explorations in Economic History 14, 311-336. DeLong, J. B. (1988), "Productivity Growth, Convergence, and Welfare: A Comment," American Economic Review 78, 1138-1154. Denison, E. F. (1962), The Sources of Economic Growth in the United States and the Alternatives Before Us. Washington, DC: Committee for Economic Development. Gerschenkron, A. (1952), "Economic Backwardness in Historical Perspective." In B. F. Hoselitz (Ed.), The Progress of Underdeveloped Areas. Chicago: Univ. of Chicago Press. Goldin, C. G., and Lewis, F. D. (1975), "The Economic Costs of the American Civil War: Estimates and Implications," Journal of Economic History 35, 299-326. Hatton, T. J., and Williamson, J. G. (1991a), "Integrated and Segmented Labor Markets: Thinking in Two Sectors," Journal of Economic History 51, 413-425. Hatton, T. J., and Williamson, J. G. (1991b), "Labour Market Integration and Rural-Urban Wage Gaps in History." Paper presented at the conference on The Role of Long Run Analysis in Economics, Australian National University, Canberra, July ll~ 12. Hatton, T. J., and Williamson, J. G. (1992a), "What Explains Wage Gaps Between Farm and City: Exploring the Todaro Model with American Evidence 1890-1941." Economic Development and Cultural Change 40, 267-294. Hatton, T. J., and Williamson, J. G. (1992b), "International Migration and World Development: A Historical Perspective." Paper presented to the conference on Economic Aspects of International Migration, Vancouver, Canada, September 5-6. Kendrick, J. W. (1961), Productivity Trends in the United States. New York: National Bureau of Economic Research. Kravis, I., Heston, A., and Summers, R. (1978), International Comparisons of Real Product and Purchasing Power. Baltimore: Johns Hopkins Univ. Press. Kravis, I., Heston, A., and Summers, R. (1982), World Product and Income: International Comparisons of Real Gross Product. Baltimore: Johns Hopkins Univ. Press. Maddison, A. (1982), Phases of Capitalist Development. Oxford: Oxford Univ. Press. McCloskey, D. N. (1970), "Did Victorian Britain Fail?", Economic History Review 23, 446-459. 196 JEFFREY G. WILLIAMSON McLean, I. W., and Pincus, J. J. (1983), "Did Australian Living Standards Stagnate Between 1890 and 1940?," Journal of Economic History 43, 193-202. Mulhall, M. G. (1896), Industries and Weallh of Nations. London: Longmans, Green and Company. O'Rourke, K. (1989), "Agricultural Change and Rural Depopulation: Ireland, 1845-1876." Ph.D. thesis, Department of Economics, Harvard University. O'Rourke, K. (1990), "International Migration and Wage Rates in Twentieth Century Ireland." First Boston Working Paper Series, Graduate School of Business, Columbia University (December). O'Rourke, K., Williamson, J. G., and Hatton, T. J. (1994), "Mass Migration, Commodity Market Integration and Real Wage Convergence: The Late Nineteenth Century Atlantic Economy." In T. J. Hatton and J. G. Williamson (Eds.), Migration and the International Labor Market, 1850-1939. London: Routledge. O'Rourke, K., Taylor, A. M., and Williamson, J. G. (1993), "Land, Labor and the Wage-Rental Ratio: Factor Price Convergence in the Late Nineteenth Century." Paper presented at the European Historical Economics Society Conference on Market Integration in Europe, Lerici, Italy, April 1-4. O'Rourke, K. and Williamson, J. G. (1994), "Late 19th Century Anglo-American Factor Price Convergence: Were Heckscher and Ohlin Right?" Journal of Economic History 54, 1-25. Shergold, P. R. (1982), Working-Class Life: The "American Standard" in Comparative Perspective 1899-1913. Pittsburgh: Univ. of Pittsburgh Press. Solow, R. M. (1957), "Technical Change and the Aggregate Production Function." Review of Economics and Statistics 39, 312-320. Taylor, A. M. (1993), "External Dependence, Demographic Burdens, and Argentine Economic Decline After the Belle Epoque," Journal of Economic History 53, 907-936. Taylor, A. M., and Williamson, J. G. (1994), "Capital Flows to the New World as an Intergenerational Transfer," Journal of Political Economy 102, 348-371. Williamson, J. G. (1974), "Watersheds and Turning Points: Conjectures on the Long Term Impact of Civil War Financing." Journal of Economic History 34, 636-661. Williamson, J. G. (1985), Did British Capitalism Breed Inequality? London: Allen and Unwin. Williamson, J. G., and Lindert, P. H. (1980), American Inequality: A Macroeconomic History. New York: Academic Press. Wolff, E. N. (1991), "Capital Formation and Productivity Convergence Over the Long Term," American Economic Review 81, 565-579. Wright, G. (1990), "The Origins of American Industrial Success, 1879-1940." American Economic Review 80, 651-668.