Discussion
Robert Allen (Chapter 5, this volume) suggests that Japan's history of real wages was characterized by ‘both close integration of the [labour] markets and the absence of the dynamic urban economy' of the northwest European type, the latter of which echoes the point made by Thomas Smith in his classic essay on ‘pre-modern economic growth' (Smith 1973).
He contrasted Tokugawa Japan's rural-centred development with western Europe's urban-centred growth, noting that the lack of overseas trade and virtually static population in late Tokugawa Japan induced the decline of towns, especially of large towns, which meant that losses were imposed on the merchant class. With the Tokugawa shogunate's imposition of tight restrictions on foreign intercourse in the mid-1730s, the days of merchant-adventurers had gone. Instead, large-scale mercantile enterprises that concentrated on domestic, inter-regional trade emerged and merchant houses based in Osaka came to dominate the nation's economy. Organizationally they were big businesses with a multi-unit, complex managerial structure. Their profit rates were no longer very high but they traded extensively and efficiently. This type of merchant became a casualty of the growth of rural commerce and industry in the eighteenth and nineteenth centuries. The age of rural development saw the decline of large cities such as Edo and Osaka, on the one hand, and the rise of small country towns, on the other. After the opening up of trade with foreign countries in 1859 the large cities re-emerged as dynamic centres of growth. Yet, as the available evidence indicates with respect to wage differentials between agriculture and industry, those between large and small firms within industry and trends in income distribution (Minami 1998; Saito 1998), and contrary to the remark made by Williamson (2000: 34—5), it was not until after the 1910s that the growth in earnings of the urban rich and middleclasses started to outstrip that of the labouring public.
Pre-First World WarJapan had not yet broken away from the traditional growth regime.That said, however, another set of factors should also be commented on. Elsewhere, based on the observation that in 1840s Chδshu, a province in western Japan, the average wage rate for unskilled workers in salt farms was equal to the marginal labour productivity in farming derived from production function estimates (Nishikawa 1978), I suggested that it was agricultural productivity that determined the supply price of unskilled labour from the peasant farm household and, hence, that wage differentials contracted in the phase up to 1820 due to the rising productivity level in the farm sector (Saito 1978). As long as a peasant farm household was an independent decision-making unit of production, this is the mechanism which enabled the market wage rate for the non-farm unskilled labour force to keep pace with output growth. Given the evidence we now have for the second and third phases, and given the regional pattern of change in real wages and rice yields that we have established for the entire period, this mechanism must have been working throughout the period in question. In other words, rising productivity in farming acted as one factor that prevented the widening of the real wage—GDP gap in eighteenth- and nineteenth-century Japan.
This consideration has an important implication for the discussion of pre-modern growth. Agricultural progress of this type acted as a brake on the process of proletarianization, which is believed to have taken place across early modern European countries. The proposition that peasants' asking price of labour was determined by the marginal productivity of labour on their own farm implies that if the market wage rate was low in comparison with their farms' productive capacities, then labour would not be supplied. They would stay on the land. Hence, there would be no proletarianization. Here, by ‘productivity’ I do not mean just rice and other food crops.
Any item the farm household produced should be included in our consideration. Indeed, as I demonstrated elsewhere by using early Meiji household returns in the sericultural province of Kai and district statistics in the cotton-growing prefecture of Osaka, the introduction by peasant cultivators of cash crops such as cotton and mulberry trees had an effective countervailing effect opposing the oft-assumed tendency for rural commercialization to create a landless labour force (Saito 1986). Throughout the Tokugawa-Meiji years, and even in the period between the two World Wars, contemporaries were concerned with tenancy problems but never mentioned the emergence of a rural landless class as a social problem. To quote Ronald Dore writing on the post-war land reform:Tenancy increased while the number of hired labourers was never, before the last war, higher than some 300,000 (out of a total of 14 million agricultural workers) and most of these were either young men and girls, half labourers and half family ser vants, living in and supplementing the labour of the family on a bigger-than-average peasant holding, or else casual labourers with no permanent employer. (Dore 1959: 17)
As noted earlier, the emergence of a well-integrated labour market between the peasant farm household and urban sectors was one feature of the economic history
of late Tokugawa Japan. The situation prior to the onset of industrialization, on the face of it, resembled what the Lewisian model of unlimited supply of labour postulated. In such a setting, the peasant sector would certainly appear as a pool of cheap labour if marginal productivity in farming were low and static. With a slow but steady rise in productivity in a period of static population, however, the farm sector could determine the direction and pattern of ‘pre-modern economic growth'. This too was a variant of ‘Smithian' growth. It was this pattern that characterized eighteenth- and nineteenth-century Japan and distinguished it from early modern western Europe.