Urban Development & Agrarian Change in Early Modern Europe

Urban Development and Agrarian Change in Early Modern Europe The paper assesses neo-liberal and neo-Marxist theories dealing with agricultural institutions and productivity growth. Both argue that English institutions (enclosures and capitalist agriculture) were particularly conducive to development. Labour and total factor productivity indices are presented for the main European countries, and they show that three countries had high levels of efficiency--England, the Netherlands, and Belgium. Since they did not share common tenurial institutions that distinguished them from the rest of Europe, such institutions do not explain agricultural revolutions. What the three countries had in common and what set them apart from others was higher rates of urbanization and higher levels of wages. Tight labour markets related to a large urban sector help explain high levels of agricultural productivity and institutional changes like enclosure and large farms. Class structure and conflict, the issues emphasized by neo-Marxists, are important in explaining the differences between Belgium, the Netherlands, and England as well as differences between eastern and western Europe. What caused agricultural productivity growth in early modern Europe? The modernization of agrarian institutions is the most common explanation. Neoliberals contend that the replacement of communal ownership or management with private property was the key to progressive, while Marxists put the emphasisis on the transition from feudalism to capitalism. This paper will examine these theories as they have been formulated by Douglas North and Robert Thomas (1973), on the one hand, and by Robert Brenner (1976, 1997), on the other. A European-wide data base on the history of wages and prices provides a new way of evaluating and interpreting these ideas. North and Thomas have focused on institutional change as the central dynamic of economic development. They argue that "the development of an efficient economic organization in Western Europe accounts for the rise of the Western World." Efficiency "entails the establishment of institutional arrangements and property rights that create an incentive to channel individual economic effort into activities that bring the private rate of return close to the social rate of return." (North and Thomas 1973, p. 1). In the North and Thomas view, individuals act to raise their incomes. What they do depends on the balance between their costs of acting and the gains to be had. If externalities are internalized so that the private gain from change equal the social return and if the cost of making that change is low, then technology and institutions will evolve in the way that maximizes economic growth. 5 This framework has a particular application to agriculture. Private property was the key to bringing private and social returns into line. Open field farming and common pastures were inefficient because they introduced gaps between social and private returns. For instance, when a farmer considered how many cows to graze on a common pasture, his reckoning included only his own costs and not the higher costs that other farmers would bear as a result of the higher stocking density. Over grazing was the result. There was some recognition of these problems as evidenced by voluntary agreements to mitigate the inefficiencies--"successive approximations toward exclusive property rights developed in cultivated areas" (North and Thomas 1973, p. 151)--and finally by enclosure when state policy switched from opposing to facilitating it. "The great agricultural revolution in England...came from eliminating many of the common property aspects of land ownership and increasing the return to the cultivator from using more efficient techniques." (North and Thomas 1973, p. 151.) The history of the Netherlands confirms the progressive role of enclosure and private property, for "the owner-opertors of the small [Dutch] farms, free from the restrictions of common fields, could adjust quickly to changes in market demand and, since the proprietors alone captured the inherent rewards, had every incentive to do so." (North and Thomas 1973, p. 143.) North and Thomas's view that enclosure was the basis of agriculture progress is, of 6 course, consistent with much English historiography, although his confidence in the progressiveness of the small scale, owner-occupying Dutch cultivator is unusual in the historical literature. While North and Thomas emphasize the importance of enclosure, Brenner's highly influential paper "Agrarian Class Structure and Economic Development in Pre-Industrial Europe" (1976) stresses the social relations of production. The fundamental question addressed by Brenner is why the Industrial Revolution happened first in England. His answer is in two parts. First, England developed a unique system of agrarian capitalism. In his 1976 paper, Brenner related this argument, in the classic manner, to the rise of the great estate. Most English land was owned by the gentry and aristocracy (on the order of 20,000 members in the eighteenth century) and leased to farmers who cultivated it in large units--75 or more hectares was not unusual. The large farms required a large workforce of servants and day labourers. No other country in Europe had a similar rural social structure, and that structure led to England's early industrialization. According to Brenner (1976, p. 49), "it was the emergence of the 'classic' landlord/capitalist tenant/wage-labourer structure which made possible the transformation of agricultural production in England, and this, in turn, was the key to England's uniquely successful overall economic development." 7 Recently, however, Brenner (1997) has amended the argument to place less stress on the use of wage labour. He regards the large family farmers operated by yeomen in the sixteenth and seventeenth centuries as conducive to productivity growth. In his view, these were still capitalist enterprises since arbitary feudal exactions had ended--so the peasants had the security to accumulate--and since most of the yeomen held their land on copyholds or leases for lives. To renew these agreements, they had to accumulate large cash balances to pay high entry fines, so they had to be efficient producers. They could not, in other words, withdraw from the market and satisfy their needs with meagre production as peasant proprietors could do. Brenner sees the English yeomen as quite distinct from the peasant proprietors common on most of the continent and from the pre-plague English peasants. England did not always have a unique rural social structure. In the middle ages, the land was organzied in manors containing demesne, free, and servile lands, as it was in most of Europe. In much of western Europe, this system gave way to small scale peasant proprietorship--a mode of production which Brenner (unlike North) believed was ill adapted to economic development. Eastern Europe possessed a lax feudal system in the high middle ages. This was replaced by a draconian system of serfdom in the sixteenth and seventeenth centuries. Only in England, was agrarian capitalism the outcome in the eighteenth century. The 8 big question is why. Brenner answers the question by effectively challenging two established explanations. In one, the growth of commerce and towns (or the lack of such growth) is the prime mover that explains the divergent patterns of rural development; in the other, population change is the prime mover. Brenner attacks both explanations by showing that the same development, be it of commerce or population, could produce different results in different regions. Population collapse after the Black Death, for instance, was related to the end of serfdom in western Europe and its intensification in eastern Europe. Some mediating variable must, therefore, have been operating to account for the different outcomes, and for Brenner it is "class structure" or, more exactly, the prior history of class struggle. Thus, peasants in western Europe had centuries of experience in resisting feudal exactions and this experience had created both village organizations and a mentalité that they used to good effect in the fifteenth century to defeat the nobility. In contrast, the agricultural development of Poland and East Prussia had involved immigration, the establishment of new villages under the patronage of lords, and very low feudal exactions. Without the need to resist the nobility, peasant agriculture was far more individualistic than in the west and lacked the village-level institutions to resist aristocratic encroachments. Hence, east European peasants were not able to effectively defend their 9 rights when they were challenged by the nobility in the sixteenth century and so lost their freedom. The history of rural class struggle is ultimately the explanation for England's unique social structure and is the second part of Brenner's answer to the question of why the industrial revolution happened in England. In this essay I will test and refine the theories of North and Brenner by confronting them with a cross-European sample of wages and prices (Allen 1998a). For a century, historians have been recording time series of prices and wages in leading European cities from the middle ages to the nineteenth century. I have collated much of this material and converted the wages and prices to grams of silver and the weights and measures to metric equivalents. In this way, consistent comparisons of wages and prices can been made across space and over time. The data base currently includes Antwerp, Amsterdam, London, Paris, Strasbourg, Florence, Milan, Naples, Valencia, Madrid, Augsburg, Leipzig, Munich, Vienna, Gdansk, Krakow, Lwow, and Warsaw. Supplementing these data with additional information on the population histories and land area of the countries concerned allows new estimates of agricultural productivity and opens up the possibility of modelling the process of economic development. In this paper, I examine as many aspects of the theories of North and Brenner as possible with these data. The form of the argument draws on the styles of both Brenner 10 and North. Like Brenner in his critiques of the demographic and commercial models, I use comparative history to test theories. In this case, I identify the parts of Europe that achieved high agricultural productivity by the eighteenth century. They were the Netherlands, Belgium, and England. The question then is what these countries had in common and what distinguished them from the rest of Europe. The answer is not land tenure or rural social structure. This finding calls into question the progressive role assigned to enclosures by North and the tripartite class structure emphasized in Brenner's original formulation. What the three countries did have in common was unusually high rates of urbanization and high levels of real wages. While agricultural advance, urbanization, and high wages were all interrelated, I analyze here the impact that urbanization and wages had on the agricultural sector. I argue that many of the distinctive features of agriculture in northwestern Europe make sense when considered in the context of labour scarcity and high wages caused by large or rapidly growing cities. Reasoning in the manner of North, these conditions created the incentive or demand for techological and institutional change. The history of class struggle emphasized by Brenner determined the response of the system or the supply of institutional change and thus the differences between the three high productivity countries. Finally, I take up the question of serfdom. Why did it disappear in western Europe after the Black Death--a development which 11 Brenner and North both rightly recognize as a key to western development--while it was greatly intensified in eastern Europe? The answer lies in both the structure of the economies, which implied differential returns from serfdom in different regions, and the prior history of class struggle, which again played a role in explaining the different historical trajectories. I. Productivity levels in Europe To approach the question of agricultural modernization from a pan-European perspective, we need to be able to compare productivity across Europe. For the years before 1800, most of the evidence currently at hand relates to crop yields (e.g. Bairoch 1997, p. 48). While this evidence is not without its problems, it suggests that productivity c. 1800 was highest around the North Sea--in England and the present day countries of Belgium and the Netherlands. Productivity may also have been high in adjoining regions of France and Germany. Some regions in these countries also showed high productivity in the middle ages, although it may have fallen after the Black Death and rebounded in the early modern period. In other regions, there was probably advance in the early modern period from low medieval levels. Labour productivity is arguably more important than crop yields in understanding economic development, and there have been several attempts to measure it (Wrigley 1985, Persson 1988, 1991, 1993). Recently, I have extended these procedures and developed 12 estimates of output per worker in agriculture for most parts of Europe from the middle ages to 1800 (Allen 1998b). Figure 1 plots output per worker for England, Belgium, Italy, Spain, France, and the Netherlands over this period. The graph show several patterns of development. Italy is one paradigm. Output per worker rose from 1300 to 1400 due to the population losses of the Black Death, and then slowly declined until 1800 as the population grew. The data for other continental countries start only in 1400, but they show similar levels of output per worker and the same long run decline as Italy up to 1800. There is some suggestion that French productivity started to rise before 1800. This pattern of productivity change marks the unsuccessful economies of early modern Europe. The Low Countries show a second pattern of productivity change. Belgian productivity was impressively high from the middle ages to 1800, but did not grow. Dutch productivity in 1500--when the data begin and its population was light--was at Italian levels but rose to Belgian levels during the Golden Age. England is the third paradigm. Its data begin in 1300. From 1300 to 1600, output per worker in English agriculture followed a pattern like Italy's but the post-plague fall in productivity was delayed a century since the English population did not recover before the sixteenth century: Productivity rose from 1300 to 1400, remained high until 1500 in view of the 13 stagnation of population, and then fell as population began to grow again after 1500. The seventeenth century, however, marked a decisive break with the continent. Even though population grew rapidly, labour productivity rose sharply. By 1750, output per worker in English agriculture had surpassed Belgian and Dutch levels. There was no growth, however, between 1750 and 1800 despite historiographical enthusiasm for an agricultural revolution in that period. England, Belgium, and the Netherlands still stand out as the leaders when total factor productivity is used as the measure, but, in this case, the Low Countries are identified as the most productive. Total factor productivity was measured in a two stage procedure. Since the trajectories of output per worker were similar in France, Spain, Italy, Germany, and greater Austria and since there was no indication of productivity growth, these countries were analyzed separately. A constant returns to scale Cobb-Douglas production function was estimated by regressing the logarithm of output per worker on the logarithm of the land-labour ratio. This equation was used to predict output per worker for all countries including England, Belgium, and the Netherlands. The ratio of actual to predicted labour productivity equals total factor productivity. Figure 2 plots the total factor productivity index. A value of one indicates performance at the average level of the nondevelopers. The productivity indices of Italy and Spain fluctu14 ate around that line, as do those of Germany, Austria, and Poland, which are not shown. Evidently, these countries did all operate at the same level of efficiency. Before 1700, French productivity was also at that level, but there is a strong suggestion of rising productivity in eighteenth century France. Total factor productivity was much higher in northwestern Europe. Belgium was the productivity leader with a level of efficiency consistently twice that of other continental countries. Farmers in the Netherlands in 1500 and in England before 1600 performed at the same low level of efficiency as their continental counterparts. Then the Dutch and English agricultural revolutions began. The total factor productivity of Dutch agriculture rose strongly after 1500 as did the productivity of English agriculture after 1600. By 1800 farmers in these countries were approaching Belgian levels of performance. The relative performance of Belgium, the Netherlands, and England depends on the productivity index used. The Low Countries score best in terms of total factor productivity since farmers in Belgium and the Netherlands achieved their high levels of labour productivity despite using less land per worker than did English farmers. Whichever measure is chosen, these three countries perform at decidedly higher levels of efficiency than the rest of Europe. Any exercise of this sort can be challenged on theoretical or empirical grounds. An important item in favour of these 15 productivity estimates is that they do not overthrow what we might otherwise have said about European agricultural productivity. It is not surprising to find that England and the Low Countries had high productivity in the eighteenth century, whatever indicator is chosen. It is surprising, perhaps, that the data of other countries follow a single pattern. Who would have imagined that a simple production function like that estimated here would fit the data so well? What Figures 1 and 2 provide is a more systematic account of European agricultural productivity than was available previously, but not one overturns all previous thought. II. What distinguishes the high productivity regions Figures 1 and 2 do raise questions about productivity and institutional change in a European wide perspective. What the figures show clearly is that England and the Low Countries were more productive that other parts of Europe. The important question is what Belgium, the Netherlands, and England had in common and what distinguished them from the rest of Europe. If Brenner's and North's views were correct, we should find that Belgium, the Netherlands, and England shared unique institutions that distinguished them from France, Germany, Italy, etc. And yet they did not. Brenner's (1976) formulation is right that England's tripartite rural class structure and large scale farms were unusual. The problem is that they were too unusual and not 16 shared by the Netherlands and Belgium. English exceptionalism takes a hard knock since English performance was no better--perhaps a bit worse--than that of the Low Countries. Explanations of agricultural development that emphasize the importance of exclusively English institutions are hard to credit. The same problem afflicts North's view that the enclosure of open fields and commons was the key to growth. He portrays the Dutch Republic as entirely enclosed in the Golden Age, which is far from true. Large parts of modern-day Belgium were also in open fields, as, indeed, were most of the English midlands. While enclosure may account for the slightly higher level of labour productivity in England than in the Low Countries, the high levels of efficiency achieved there militate against the argument that enclosure was necessary for high agricultural productivity. Similar arguments could also be made for grain yields--and grain was the major crop of open field farmers--for yields were high in Belgium and the Netherlands as well as in England (Bairoch 1997, p. 48). Comparisons with the Low Countries undermine the view that enclosure was the key to progress just as they undermine the claims for agrarian capitalism. Perhaps the common feature lies elsewhere. Revisionist views of English history highlight family farmers operating in the open fields as the heroes of the agricultural revolution (Allen 1992). Permanent or long term occupation at minimal and fixed rents supposedly provide the financial incentive for rising 17 productivity: With fixed rents, reductions in farm cost or increases in revenue accrued to the farmer rather than the lord, so the farmer had the incentive to raise yields or save inputs. Such institutions were common in the Netherlands, but they were also found elsewhere in Europe, so they do not distinguish the North Sea countries from the rest of the continent. Nor were family farms held for money rents found only in England as Brenner (1997) suggests. Moreover, it is not obvious why tenurial arrangements like share cropping, which split output gains between landlord and tenant, should not also have given the farmers an incentive to innovate. There does not seem to be an important difference in tenurial arrangements between the countries that modernized and those that did not. Hence, tenure and property rights did not matter. Was there anything that England, the Netherlands, and Belgium had in common and that distinguished them from the rest of Europe? There were two important characteristics--urbanization and wages. Table 1 shows urbanization rates for the major countries in Europe from the middle ages to the nineteenth century. These figures derive from Bairoch's (1988) investigations and show the fraction of the population living in settlements of 5,000 people or more. Throughout the early modern period, the Dutch and Belgian urbanization rates were usually about one third and the highest in Europe. English rates were low in the middle ages and 18 began to rise dramatically in the seventeenth century and reached Low Country levels in the eighteenth century. No other country had a comparable growth in the fraction of its population living in cities. Growth was as important as level for the argument I will develop, which emphasizes rural-urban migration. Urbanization, then, was one important factor that distinguished the Netherlands, Belgium, and England from the rest of Europe. The level of wages was a second distingiushing characteristic of Northwestern Europe. After 1600, wages were low in most of Europe, whereas they were high in the countries with high agricultural productivity. Figure 3 plots the real wage--i.e. the wage adjusted for differences in the cost of living--of building labourers in London, Amsterdam, Northern Italy, Valencia, Krakow, and Strasbourg from 1350/99 to 1750/99. In the late middle ages, wages were high everywhere, but they diverged thereafter. Real wages were maintained across the early modern period in London and Amsterdam but fell by 50% or more on the continent. Real wages in Antwerp lagged a bit c. 1800 but were similar to those in London and Amsterdam and remained much above those elsewhere in Europe. A large gap between wages in the North Sea countries and those elsewhere in Europe emerged in the early modern period. High wages, rising agricultural productivity, and urbanization were all interrelated in the long run. The population grew everywhere in Europe between 1400 and 1800. On most of the 19 continent, there was no long run process of capital accumulation or technological progress, so the population rise led to falling wages due to diminishing returns of labour in agriculture. In England, Belgium, and the Netherlands, however, economic development offset population growth maintaining or increasing the wage despite rates of population growth that exceeded those elsewhere in Europe. Successful early modern development involved exogenous factors like successful imperialism and positively reinforcing devepments in the urban and agricultural economies. The growth of the urban economy affected the development of agriculture. Product markets provided one link between these these sectors: Von Thunen's ring model is a classic formulation in which urban demand and transportation cost imply a systematic spatial pattern of agricultural production and intensity. While not meaning to deny the role of product demand, the argument developed here emphasizes labour demand. In the present analysis, the labour market provided the key link between the city and the country. Growing cities implied a strong demand for labour that drew people from the country. Even a large urban sector that was not growing generated labour demand since city mortality was so high the urban population could be maintained only by attracting migrants from the countryside (van Zanden 1993, Wrigley 1967). Wages rose as a result. The question is how the countryside responded to the job opportunities in the city. 20 III. Responding to the High Wage Challenge The early modern agricultural revolutions make a lot more sense if we analyze them as a response to urbanization and the high wage environment of Northwestern Europe than if we search for a--nonexistent--common tenurial pattern. The growth of high wage job opportunities in the city posed a challenged to rural society. Agricultural revolutions were the response to that challenge. We can specify the challenge numerically by comparing the ratio of the wage rate to the price of food. Figure 4 plots this ratio for the same six countries whose real wage was examined earlier. The ratio is considerably higher for England, Belgium, and Poland than for other countries. (Poland does much better by this measure of the real wage than it did in Figure 3, a point to which we will return in discussing the imposition of serfdom in that country.) The ratio of the wage rate to the price of food is important since profit maximizing farmers should equate the value of the marginal product to this ratio. When the wage was high, they did this by cutting back on farm employment. Many of the distinctive features of agriculture and institutions in Northwestern Europe are understandable in this context. Three obvious strategies were available to farmers around the North Sea to reduce farm employment. 1) increase the size of the farm Aside from demesnes, most farms in medieval Europe were 21 small. A farm less than 25 hectares could be largely, if not entirely, cultivated with the labour of a family. However, a farm of that size used more labour per hectare than a farm of 75 or 100 hectares. The larger farm employed more people--but fewer per hectare--and the greater number of people allowed a more efficient division of labour (Allen 1992, pp. 211-27). One could offset high labour costs, therefore, by amalgamating little farms into big ones. England was the classic case of large scale farming, and it can be most easily understood as a response to high wages. There were moves to large scale farming in much of the Netherlands as well, and De Vries and van der Woude (1997, pp. 219, 551-7) explain it in exactly the terms suggested here: "If rural dwellers were not to lag far behind the growth of urban prosperity in the century after 1580, farm revenues would have to be raised." This effort required a reorganization of the farm. "Success in this redeployment of farm labour required capital investment in agriculture...and the achievement of an increased scale of operation." 2) enclosure Enclosure can also be seen as a response to high wages. Until the late eighteenth century, most enclosures were for the conversion of arable to pasture. In particularly fertile regions, revenue per hectacre might rise, but generally it fell in these enclosures. Employment costs fell more, however, so the 22 rental value of land rose. Enclosure can be thought of as a way to reduce labour costs, so it is not surprising that enclosures took place in England in the high wage part of Europe. Enclosure was not universal throughout the Netherlands or Belgium--nor, indeed, were large farms--despite the high level of their wages. Why? Because increasing farm size and converting arable to pasture reduced employment even as it raised the value of land. As a result the attractiveness of these reorganizations depended on the structure of property ownership. In a society like England's where much land was owned by one set of people and farmed by another, the landowners could always raise their income by enclosing and amalgamating farms, so there was a powerful social group always pushing for these changes. In contrast, owner-occupying farmers would have been concerned with the net income change--the decline in labour income as well as the rise in land value. If enclosure led to the conversion of arable to pasture and farm revenues per hectare fell--and this was often the case--then owner-occupiers would lose income and oppose it even though it would have been favoured by a class of landowners. In a less extreme case like farm amalgamation where revenue per hectare would probably have been unchanged, owner-occupiers might still oppose the change since it would reduce the number of farms. Certainly, the peasant proprietor mode of production would self-destruct if too much farm amalgamation took place. 23 These considerations show that the response of the farm sector to the challenge of a large, high wage urban sector depended on the structure of property ownership in the countryside. A society of peasant proprietors, and certainly one where the members felt strongly about preserving their way of life, would respond differently from a society of landlords and tenants. The differences in the development of agrarian institutions around the North Sea are explained by the considerations that Brenner has highlighted. Other responses to the challenge of high wage, urban jobs would have been far more attractive to peasant proprietors. 3) protoindustry The peasants could try to reduce employment within the context of the family farm by contracting highly labour intensive activities, so that labour could be allocated to nonfarm employment at high wages. The shift of activity from grain to livestock with a concommittant redirection of labour towards protoindustrial employment would have been a rational response to a high off farm wage rate. Thirsk (1961) attributed this pattern to geography, but it may also have been a response to labour market developments. Protoindustry had causes besides that indicated here. In backward parts of Europe, it was a response to underdevelopment, but in the advanced economies, it was a response to a high urban demand for labour. Urbanization in early modern Europe meant a 24 substantial migration of people to the cities. But why should the people move to the jobs; why shouldn't the jobs move to the people? The spread of industry to the countryside is usually seen either as an escape from the restrictions of guilds, or as a response to a growing excess supply of labour in the countryside, or as a response to "deurbanization." While these factors may all play a role (with an importance that varied with time and place), rural industrialization should also be expected as a result of a boom in the urban economy itself. 4) increase farm output The three strategies just reviewed were adjustments within a defined range of techniques to a high wage. In the metaphors of economists, they involved reducing employment along an unchanging labour demand curve. Another kind of response would have been to shift the labour demand curve so that employment could have been maintained at the higher wage. Such a strategy would have involved technical change and raising farm revenue per hectare. One way to do that would have been to develop new techniques that raised crop yields, so that a farm of a given size would have generated a greater cash flow. Alternatively, new crops or income generating activities might have been explored. Output expansion was a viable way for the small farm sector to try to match high urban incomes without self-destructing. That was the incentive for the change. Technical change involved experimentation. There were, of 25 course, no experimental stations that would discover and perfect new ways of farming. The farmers--or some of them--had to do that themselves. Any experimenting involved the risk of failure. If farm incomes were low and close to subsistence, then the costs of failure would have been catastrophic and not much experimenting would have taken place. As incomes rose, experimentation was encouraged since the dangers of failed experiments became less consequential. There was thus an ascending spiral of progress (to borrow a phrase of Havinden (1961)). High urban labour demand and wage rates led to off-farm employment and attempts to raise output. To the degree these were successful, farm incomes rose. As farmers became more prosperous, they could experiment more and technical progress in agriculture increased. North and Thomas's emphasis on incentives has been helpful in explaining why the North Sea differed from the rest of Europe, and why developments diverged within that region. Agricultural productivity growth and institutional change were faster in that region because the incentive to make those changes was greater there. The large size and growth of the urban economy generated tight labour markets and high wages in the face of expanding population. Output increased and labour was shed in response. England, Belgium, and the Netherlands differed in their response to the urban challenged, however, and the class arrangements that Brenner has emphasized explain those differences. In England, with its tripartite rural social structure, there was far more 26 enclosure and more extensive farm amalgamation than in the Netherlands or Belgium. IV. Poland and the Second Serfdom The explanation just offered for the agrarian revolutions in Northwestern Europe hinges on the free mobility of labour from farm to city. That mobility presupposes the end of serfdom. The end of serfdom also plays a central role in Brenner's account of early modern history. Serdom disappeared in western Europe late in the middle ages, while it was intensified in eastern Europe. Historians have explained this divergence with appeals to both towns and population, but Brenner rejects both arguments in favour of the prior history of class struggle. How do these claims look in terms of the pan-European evidence considered here? First, consider the "demographic" explanation for serfdom in Poland and East Prussia. The argument starts with the region's low population density after the Black Death. Vacant land and empty farms meant that peasants could leave their lords and cultivate on their own. Had they been allowed to do so, the aristocracy would have lost its income. To prevent that, a harsh serfdom was imposed, which tied the serfs to the land. Denied the chance to move, their lords then increased the share of farm output they appropriated. Free land led to serfdom. Brenner (1976) has criticized this argument by showing that 27 similar demographic conditions in western Europe led to the abolition of serfdom as the lords bid against each other for tenants, lowering rents in a effort to attract cultivators for vacant farms. A first question to ask in exploring this issue is how different were eastern and western Europe in terms of population densities? Some insight is provided by Figure 5, which plots the land-labour ratio in six countries. The ratio was highest in Poland in 1400, but England recorded the second greatest value. Evidently, the population losses following the Black Death created a labour shortage in England almost as great as that in Poland. Indeed, England and Poland vied for the highest landlabour ratios throughout the early modern period. Since England and Poland followed such different courses of development while sharing the same land-labour ratio, that ratio cannot explain the differences in their histories. Brenner was right on that matter. Labour scarcity can be measured with the wage as well as with the land-labour ratio. Scarce labour means high wages. Were real wages high in Poland? Poland turns out to be an unusual case, for it is the only country for which the answer to this question depends on how the real wage is measured. Figure 3 shows the wage rate deflated by a consumer price index. By that measure, Polish wages in the fifteenth century were not unusually high, and they fell to low levels thereafter. What was the point