The Great Leveler, page 14
Once the plague abated in the late fifteenth century, European population began to recover. Economic development reached new heights, as did inequality. The formation of fiscal-military states in Europe, the creation of overseas colonial empire, and an unprecedented expansion of global trade fostered institutional change and new networks of exchange. Although commercial and tributary exchanges had always existed side by side, the former gradually became dominant as commercialization transformed tributary states and increased their dependence on commercial revenue. The growth of a more integrated world system supported by bullion extraction in the New World and intercontinental trade mobilized wealth and widened the gap between rich and poor on a global scale. As Europe became the hub of a worldwide exchange network, development gave more power to mercantile elites and drew rural majorities into market activities that put pressure on their attachment to the land. Tribute-taking elites morphed into commercial and entrepreneurial landowners, and merchants established closer ties to governments. Peasants were gradually separated from the land via enclosure, tax, debt, and the commodification of landholding. Traditional means of enrichment rooted in the predatory exercise of political power persisted alongside these modernizing market-based processes: stronger states offered attractive pathways to riches. All this put upward pressure on wealth inequality.8
Late medieval and especially early modern Europe occupies a special place in the historical study of material inequality. For the first time, quantitative evidence for the distribution of wealth (though not yet income) becomes available that allows us solidly to track change over time and compare developments in different areas. These data are primarily derived from local registers of taxable property and complemented by information about land rents and the incomes of workers. In the following, I use information about the distribution of wealth and income side by side. Systematic disaggregation of these two indices is not normally possible for this period: students of premodern inequality generally need to be more eclectic in their choices than modern economists might prefer they be. This is not a major problem: in preindustrial societies, trends in wealth and income inequality could hardly have moved in different directions.9
Even though these data sets do not add up to genuine national statistics of inequality, they put our understanding of the structure and evolution of wealth concentration on a much more solid footing than for earlier periods. Thanks to their internal cohesiveness and their consistency over time, some of these late medieval and early modern data series may be a more reliable guide to the overall contours of change than modern attempts to reconstruct countrywide trends from disparate sources, even for the nineteenth century. Taken together, this evidence from several western and southern European societies shows that resources were more unequally distributed in large cities than in smaller towns or in the countryside, that inequality was generally going up after the end of the Black Death, and that this rise occurred under a wide variety of economic conditions.
Greater division of labor, differentiation in skills and incomes, spatial concentration of elite households and mercantile capital, and the inflow of poorer migrants had always driven up urban inequality. According to the Florentine census (catasto) of 1427, wealth inequality was positively correlated with the scale of urbanism. The capital city, Florence, boasted a Gini coefficient of wealth distribution of 0.79—probably closer to 0.85 if the unrecorded propertyless poor are included. Gini values were lower in smaller cities (0.71–0.75), lower still in the agricultural plains (0.63), and lowest in the poorest areas, the hills and mountains (0.52–0.53). Top income shares likewise varied accordingly, from 67 percent for the top 5 percent in Florence to 36 percent for their peers up in the mountains. Much the same picture emerges from other Italian tax registers. From the fifteenth to the eighteenth centuries, reported wealth concentration in the Tuscan cities of Arezzo, Prato, and San Gimignano was consistently higher than in the surrounding rural areas. The same pattern can be observed, albeit to a lesser extent, in Piedmont.10
High wealth inequality of at least 0.75 was a standard feature in the major cities of late medieval and early modern western Europe. Augsburg, one of Germany’s leading economic centers in this period, provides a particularly extreme example: recovery from plague-related leveling witnessed an increase in the urban wealth Gini from 0.66 in 1498 to a stratospheric 0.89 by 1604. It is hard to imagine a more polarized community: a few percent of residents owned almost all assets, whereas between a third and two-thirds had nothing worth reporting at all. I return to this case in detail at the end of chapter 11. In the Netherlands, large cities experienced similarly high levels of wealth concentration (with Ginis of 0.8 to 0.9), whereas smaller towns lagged far behind (0.5–0.65). Urban income inequality was also very high in Amsterdam, where the relevant Gini coefficient reached 0.69 in 1742. English tax records from 1524 to 1525 reveal urban wealth Ginis that were generally above 0.6 and that could be as high as 0.82 to 0.85, well in excess of rural values of 0.54 to 0.62. The distribution of assets in probate inventories of personal estates was likewise correlated with settlement size. Urbanization rates held steady between 1500 and 1800 in some of these regions, most notably in Italy as well as in the Iberian peninsula, but grew considerably in England and the Netherlands, thereby raising overall levels of inequality.11
Beginning in the fifteenth century, the low point of the leveling brought about by the Black Death, inequality increased in virtually all parts of Europe for which we have data. The Netherlands provide some of the most detailed information. A precociously advanced economy with what was almost certainly the highest per capita GDP in the world at the time, it documents the disequalizing effects of commercial and urban development. By the late seventeenth century, the urban share of the population reached 40 percent, and only a third of the population worked in agriculture. Large cities manufactured and processed for export markets. A weak nobility had been eclipsed by a commercial elite that enjoyed freedom from despotic predation. Cities were highly unequal thanks to the urban concentration of capital and the urban residence of many landowners. In Amsterdam in 1742, almost two-thirds of all income was derived from capital investment and entrepreneurship. In response to a shift from labor-intensive to capital-intensive production techniques and the steady inflow of foreign labor that depressed real wages, the share of capital income in Holland rose from 44 percent in 1500 to 59 percent in 1650.12
Economic development and urban growth raised inequality over time as a small fraction of the Dutch citizenry captured a disproportionately large part of the newly created wealth even as the ranks of the urban poor kept expanding. In the longest available time series for reported wealth, from the city of Leiden, the wealth share of the top 1 percent grew from 21 percent in 1498 to 33 percent in 1623, 42 percent in 1675, and 59 percent in 1722. During the same period, the proportion of households whose assets did not reach the minimum taxation threshold increased from 76 percent to 92 percent. Most relevant information is derived from tax registers that record the annual rental value of houses in different parts of Holland, a more indirect and imperfect proxy of overall asset inequality that is likely to underestimate the latter considering that the wealthy spend a progressively smaller share of their income on housing the richer they become. A weighted value for much of Holland shows a sustained increase, from 0.5 in 1514 to 0.56 in 1561, 0.61 or 0.63 in the 1740s, and 0.63 in 1801. Between 1561 and 1732, the Gini coefficients of rental values went up everywhere, from 0.52 to 0.59 in the cities and from 0.35 to 0.38 in the countryside. The most recent standardized survey of material from fifteen Dutch towns finds a general upward trend from the sixteenth all the way into the late nineteenth century.13
Economic progress provides only a partial explanation for this phenomenon. Sometimes wealth concentration kept rising even once economic growth had stalled. Only in the northern Low Countries did the upward trend in inequality coincide with economic growth, whereas in the southern Low Countries there was no systematic relationship between these two variables at all. Divergent paths of economic development did not affect a shared tendency for inequality to increase. Nor did different tax regimes: whereas a strong emphasis on regressive consumption taxation in the south would have had disequalizing consequences, taxes in the Dutch Republic to the north were in fact uniquely progressive, focusing as they did on luxuries and real estate. And yet inequality tended to widen across the entire region.
This is not surprising: in the more dynamic north, the disequalizing forces of global trade and urbanization were complemented by growing wage dispersion that was at least in part rooted in sociopolitical power relations. In Amsterdam between 1580 and 1789, the wages of senior administrative officials, clerks, schoolmasters, and barber-surgeons rose more quickly—by a factor of five to ten—than did those of carpenters, which merely doubled. For some professions, such as surgeons, this may reflect the attachment of greater importance to their skills, even though skill premiums for workers did not generally increase during this period. However, generous raises for government officials and cognate knowledge workers such as schoolmasters may well have been driven in the first instance by a desire to keep up with those in the same bourgeois stratum who benefited from growing capital income. Thus commercial capital income seems to have had a striking knock-on effect on wages for certain socially privileged groups. Rent-seeking by elites had a polarizing effect on the income distribution.14
In the territory (contado) of Florence, wealth inequality documented by property registers grew from a low of 0.5 in the mid-fifteenth century to 0.74 around 1700. In the city of Arezzo, it rose from 0.48 in 1390 to 0.83 in 1792, and it grew from 0.58 to 0.83 in Prato between 1546 and 1763. This concentration was very much driven by the growth of top wealth shares: between the late fifteenth or early sixteenth century and the early eighteenth century, the share of reported assets owned by the richest 1 percent of households rose from 6.8 percent to 17.5 percent in the contado of Florence, from 8.9 percent to 26.4 percent in Arezzo, and from 8.1 percent to 23.3 percent in Prato. Comparable trends are visible in registers from Piedmont, where wealth Ginis increased by up to 27 points in a number of cities and on a similar scale in some rural communities. In Apulia in the Kingdom of Naples, the wealth share of the richest 5 percent rose from 48 percent around 1600 to 61 percent around 1750. In Piemont and the Florentine state, the proportion of households whose wealth reached at least ten times the local median value grew from 3–5 percent in the late fifteenth century to 10–14 percent three centuries later: polarization intensified as more households pulled away from the median.15
Unlike in the Netherlands, much of this change took place in the context of seventeenth-century economic stagnation and an even more prolonged lack of net advances in urbanization. Three major disequalizing forces have been held responsible: demographic recovery from the attrition of the Black Death, the gradual expropriation and proletarianization of rural producers, and the formation of the fiscal-military state. As elsewhere in Europe, a growing supply of labor depressed its value relative to that of land and other capital. More and more land was acquired by the elite, a process we witness in the Netherlands and France as well. In addition, city-states endowed with autonomous communal traditions and powerful notions of citizenship and republicanism were subsumed within larger and more coercive states that levied heavier taxes. In Piedmont as well as in the southern Low Countries, public debt channeled resources from workers to wealthy creditors.16
These case studies highlight the long-term continuity of mechanisms of disequalization. Reaching back at least to the ancient Babylonians, intensive economic growth, commercialization, and urbanization had boosted inequality. The same was true of the Roman period and the High Middle Ages. As we have seen, appropriation of land by affluent capital owners and elite enrichment sustained by fiscal extraction and other state activities have an even longer pedigree, going back to the Sumerians. Early modern income and wealth concentration merely differed in terms of style and scale: alongside more conventional strategies of rent-seeking, elites might now benefit from purchasing public debt instead of stealing or extorting resources outright, global trade networks opened up unprecedented investment opportunities, and urbanization began to exceed any earlier levels. Yet deep down, the principal means of disequalization had remained fundamentally unchanged and once again powerfully reasserted themselves after a temporary hiatus that had been induced by a violent shock.
The effective complementarity of these well-established disequalizers goes a long way in accounting for similar outcomes across a wide range of economic and institutional conditions (Fig. 3.2). In the Dutch Republic, inequality rose thanks to global trade, economic growth, and urbanization, whereas fiscal pressures appear to have been the most crucial factor in Piedmont and rural proletarianization in Tuscany, and both of these mechanisms operated in the southern Low Countries. In England, the most dynamic economy of this period after the northern Low Countries, commercialization and urban expansion boosted material disparities: wealth Ginis in Nottingham rose from 0.64 in 1473 to 0.78 in 1524, and in one survey of probate inventories of personal estates increased from 0.48–0.52 in the first half of the sixteenth century to 0.53–0.66 over the following eighty years. Across nine samples of such records, the richest 5 percent had held 13 percent to 25 percent of all assets at the beginning of this period and 24 percent to 35 percent later on.17
Figure 3.2Gini coefficients of wealth distribution in Italy and the Low Countries, 1500–1800
Economic conditions were strikingly different in Spain, which experienced ruralization, a shift from husbandry to farming, and low wages. In this context of stagnation or even retrenchment, the ratio of nominal per capita GDP to nominal wages climbed fairly steadily from the 1420s to the end of the eighteenth century, reflecting an ongoing disequalizing devaluation of labor as real wages fell, a phenomenon we also observe in many other European countries. The ratio of land rents to wages, another indicator of inequality, fluctuated more across this period but was likewise much higher in 1800 than it had been 400 years earlier (Fig. 3.3). These findings comport well with the observation that in the province of Madrid, wealth inequality as reconstructed from tax records increased between 1500 and 1840, albeit in a discontinuous fashion.18
Figure 3.3Ratio of mean per capita GDP to wages and real wages in Spain, 1277–1850
In the French countryside, beginning in the sixteenth century, the twin pressures of demographic recovery and growing estates hollowed out the middle and polarized local communities between larger landowners and smallholders whose farms were too small to feed them, forcing them into tenancy and wage labor. For the time being, Portugal remains the only documented outlier. According to tax rolls, overall income inequality declined somewhat between 1565 and 1700 in an environment defined by stagnating economic development and urbanization and a weakening of overseas empire. Skills premiums largely held steady in this period, whereas the ratio of land rents to wages fell throughout the seventeenth century before only partly recovering by the 1770s. Yet, looked at more closely, the moderate reduction in income inequality was mostly a phenomenon of small towns and rural communities, whereas urban inequality changed little, if at all, in the long term.19
In the absence of violent compressions, inequality might rise for a variety of different reasons that were determined by local economic and institutional conditions, but rise it (almost) always did. For what they are worth, the results of modern attempts to devise national Gini income coefficients for this period are largely consistent with the trends revealed by more localized empirical data sets. Overall income inequality in the Netherlands is thought to have increased from 0.56 in 1561 to 0.61 in 1732 before falling back to 0.57 by 1808, the period of the Napoleonic Wars. Considering the shaky foundations of the underlying schematic computations, these numbers are probably best seen as an indication of fairly high and stable inequality. The corresponding Gini values for England and Wales increased from 0.45 in 1688—well above the putative medieval peak of 0.37—to 0.46 in 1739 and 0.52 in 1801. At around 0.56, it was also high in France in 1788. All of these values are higher than those for the Roman and Byzantine empires, as was per capita output: roughly four to six times minimum subsistence in the Netherlands, five to seven times in England and Wales, and four times in France, compared to something like twice the base minimum in Rome, Byzantium, and medieval England. However, as already noted, economic development as such was not the only pathway to higher inequality: at two and a half times minimum subsistence, Old Castile in 1752 did not boast per capita surpluses much larger than ancient Rome’s but experienced high income inequality (0.53), which reflects the effects of strong social and political forces of disequalization.20
In all cases in which they can be roughly estimated, effective extraction rates—the actualized proportion of maximum feasible inequality at a given level of per capita GDP—either remained flat or rose between the sixteenth and the beginning of the nineteenth century. Three centuries after the abatement of the Black Death, income inequality in the better-documented parts of western and southern Europe had reached levels that in nominal terms—expressed in gross Gini coefficients—for the first time surpassed those of the Roman period. When adjusted for effective subsistence requirements that were sensitive to per capita GDP, they roughly approximated those experienced in classical antiquity and the High Middle Ages. Without exception, by 1800 real wages of urban workers were lower than they had been in the late fifteenth century, and although “real” inequality adjusted for divergent cost-of-living indices for high and low income groups was somewhat more volatile than nominal measures, the overall trend was likewise upward.21
BEYOND EUROPE
What about the rest of the world? Ottoman probate inventories from four cities in Asia Minor that record complete estates, including all real and personal property as well as cash, credits, and debts, shed some light on the evolution of wealth inequality between 1500 and 1840. As in Europe, mean wealth and levels of inequality were positively associated with city size. In three cities with extensive data series, Gini coefficients of asset concentration in 1820 and 1840 were higher than they had been when these series began, varying from the early sixteenth to the early eighteenth century. The same broadly applies to the top decile of wealth shares. Aggregate Ginis for rural probates rose from 0.54 in the 1500s and 1510s to 0.66 in the 1820s and 1830s, an increase that may be linked to the commercialization of agriculture and changing property relations characterized by diminishing state control over land and expanding privatization. The observed rise in wealth inequality is also consistent with evidence for falling real wages elsewhere in the Ottoman empire. Thus trends in inequality east of the Aegean quite closely resembled those in western and southern Europe.22
