Introduction
The early modern European paradox—a divergence between the slow but unmistakably rising trend in per capita gross domestic product (GDP) and the long decline in real wages—has invited scholarly speculations on how to best ‘model’ an economy exhibiting such seemingly conflicting tendencies.
No consensus has emerged so far. What does seem certain at this stage of research, however, is that these contrasting trends were the results of market-led, ‘Smithian’ growth, associated with the rise of metropolitan cities, overseas and long-distance trade, proto-industrialization, the increasing pace of proletarianization and, in the case of England and the Low Countries, agricultural investments made by enterprising landlords. All this, as Jan Luiten van Zanden (Chapter 7) and Philip Hoffman et al. (Chapter 6) hint in their contributions to this volume, seems to suggest that the growth of the European economy was accompanied by widening income differentials during the early modern period, as has been observed for many countries in later periods and as the Kuznetz curve indicates for modern growth. According to this interpretation, therefore, pre-industrial growth also bred inequality.Should we expect that East Asian peasant economies experienced similar changes in the level of real wages and wage differentials during the period before the onset of industrialization? In his recent article on Asia, Jeffrey Williamson (2000) found that real wages lagged behind per capita GDP in all Asian countries before the First World War, which suggests that inequality increased generally during the first globalization boom. Japan's wage—GDP ratio too, according to his estimates, underwent a sharp decline from the 1870s onwards before flattening out in the interwar period. But, apart from the estimation problem concerning the Japanese case, what about trends before the impact of world trade was felt? Did growth in pre-globalization Asia also breed inequality?
The present chapter will address this question by looking at wage data in Japan for 1727—1894, linking the latter half of the Tokugawa to the early Meiji era.
InJapan, as in many other countries, this period saw the growth of commerce andindustry. During the late Tokugawa years, it was primarily in the countryside where trade and industry proliferated and grew steadily, if at a modest pace. After the opening of the country to world trade in 1859,’ the sudden emergence of export markets enabled the rural economy, especially of silk-producing eastern Japan, to grow further. In short, growth in eighteenth- and nineteenth-century Japan was market-led and rural-centred (Smith 1973; Saito 1983; Shimbo and Saito 2004).
However, since there were long swings in the series of wages and prices over the 170-year period, and since there is no single time-series covering the entire period, I should like to examine the evidence I have marshalled so far (in Saito 1998), phase by phase. The focus is not simply on the trend in the general level of real wages, but more on its relationship with the trend in wage differentials within the working population, in the context of the changing phases of the eighteenth- and nineteenth-century economy. Wage labour in that period was supplied in most cases from peasant households. In such a society, it is true that wage earnings were not synonymous with the family income of the poor peasant household. Most data are of wage rates, not of total earnings that workers actually received, since wage work was often undertaken to supplement the peasant household's meagre earnings from farming; thus it is not a straightforward matter to translate the evidence on wage differentials into income inequality. However, given the paucity of pre-modern data on income differentials, the first step is to explore wage data thoroughly, and then to examine their relationship to macro measures of output growth.
The first phase, 1727—1820, was a period in which the Kinai economy grew. The Kinai was Tokugawa Japan's most advanced core region. But its growth slowed towards the end of the eighteenth century.
Productivity growth in agriculture tapered off, and some of the region's manufacturing activities died back as a result of competition from remote regions that started ‘exporting’ labour-intensive products of cottage industries to the core region. In the Kinai during this period, a contraction in occupational wage differentials took place with the real wage level for unskilled farm workers rising. This was the conclusion from my own observations based on nominal and real wage series for the Kinai area, more specifically for a village near Osaka and for Kyoto (Saito 1978). Section 2 of the present chapter will confirm this conclusion with a slightly different set of real wage series.Regarding the trends in the 1820—94 period, there is no agreement among the specialists. The general price level changed course around 1820 as the shogunate debased the currencies and the nation entered an inflationary phase. Scholars working on contemporary sources such as the Mitsui House's wage books find that the real wage level for builders declined sharply towards the end of the last decade of the Tokugawa regime, that is, the late 1860s (Umemura 1961; Kusano 1996), while those using data for carpenters and other building craftsmen in Edo (later renamed Tokyo) compiled retrospectively by Meiji trade associations tend to emphasize that real wages did rise between 1830 and 1894 despite erratic movements due to short-term monetary shocks and external disturbances such as famines and earthquakes. In particular, the Sano series of carpenters' wages has often been
referred to as evidence of a growing economy in the late Tokugawa period (Sano 1962; Hanley and Yamamura 1977; Hanley 1997). It was also chosen by Williamson (1998) for his database of Asian wages and relative factor prices, which in turn has been used by Robert Allen (Chapter 5, this volume) to bridge a gap between the earlier series and the modern, government statistics-based series. Undoubtedly there are data problems with all of these materials.
Data from Tokugawa merchant houses tend to show nominal wage rates as fixed for long periods, suggesting that adjustments to price changes were made, at least in part, with allowances and other extra payments. On the other hand, a close look at the retrospective surveys reveals irregular fluctuations in the individual nominal wage-rate series, which casts doubt upon their consistency as time-series data. In Section 2, I present a revised series of builders' real wages in Edo/Tokyo plus another comparable series of soy-sauce makers' real wages in Choshi, a town in the same Kanto area. The two series, together with some additional sets of evidence for eastern Japan, suggest that the period should be divided into two, that is, the phase between 1820—70 in which the general level of real wages was declining while wage differentials seem to have widened, and the phase after 1870 when the trends were clearly reversed.Then in Section 3, attention will be turned to the relationship between trends in real wages and output growth in the longer-run. I shall argue that traditional Japan did not exhibit the drastic divergence between wage change and output growth that was experienced in early modern western Europe. In the final section I shall briefly explore the factors that could account for the differences between the two pre-modern growth processes.
2.