Patterns and Determinants of Wealth Inequality in Late-Nineteenth-Century Ontario: Evidence from Census-Linked Probate Data

This article shows that late-nineteenth-century wealth inequality was associated with rising wealth levels supporting the existence of a Kuznets-type curve, but this curve is not unconditional. The tendency of wealth inequality to vary with age means that wealth inequality was also a function of the changing age composition of the population and may have been the result of portfolio allocation decisions across the life cycle. Canada’s population “aged” during the late nineteenth century, with the proportion of population under age 20 dropping from 53% in 1871 to 43% by 1911. The general aging of the population could have increased inequality in both wealth and income. These results follow recent work by Jeffrey Williamson (1998), who argues that Kuznets curves are not unconditional. In other words, as the results of this article confirm, wealth inequality is the outcome of a complex economic process, not a single determinant cause.

A literature dealing with wealth accumulation and inequality in nineteenth-century North America has emerged that quantitatively assesses the extent of wealth holding and inequality. Studies for Canada and the United States have found high degrees of wealth inequality even in frontier areas of recent settlement. ¹ These studies have made use of census, assessment roll, and probate data but rarely in combination, with an exception being Di Matteo and George 1992, 1998, which links probated decedents in Wentworth County, Ontario, to census and assessment rolls for the period 1872 to 1902. Linking individuals across data sources creates data sets with more socioeconomic variables, which enhances explanatory power.

This article uses 3,515 census-linked probated decedents from Ontario in 1892 to examine some of the patterns of wealth inequality using basic inequality measures, such as Gini and Theil coefficients, as well as coefficients of variation. ² The size of the data set and the distribution of the probated decedents both regionally and across age, gender, and religious lines allow for the examination of wealth inequality across subgroups of the data. The results show that while wealth inequality in late-nineteenth-century Ontario was high, there are variations across regions and age groups, as well as across wealth levels and socioeconomic status groups. Moreover, regression results found significant statistical relationships between inequality and wealth levels, as well as age, farm employment, portfolio composition, region of residence, and literacy.

Issues in Wealth Inequality and Growth

Wealth inequality can be defined as the differences that individuals exhibit in their control over or access to the stock of economic resources. The determinants of inequality are complex but ultimately depend on the ability to take advantage of economic opportunities and the capacity to absorb economic shocks. The study of wealth inequality using historical micro-data must be placed into the context of Simon Kuznets’s (1955, 1966) pioneering work on the inverted U-hypothesis relationship between economic growth and inequality. Most of the literature in this area has dealt with income and inequality, though the concepts and arguments are readily extended to wealth, which is highly correlated with income. Also, most studies of the Kuznets hypothesis have relied on cross-sectional country studies whereby intercountry differences are used to make inferences about intracountry differences. Such cross-sections have been criticized, first, because there are obvious deficiencies in using static cross-sections to address issues of dynamic change over time, ³ and second, because there are deficiencies in the quality of the data. ⁴

Several studies of inequality in North America that have explored the Kuznets hypothesis are relevant here. Williamson (1965) found that during the economic growth of the United States, the period 1840 to 1880 was one in which regional inequality increased, 1880–1920 was a period of stabilization, and the post-1920 period witnessed a decline in regional inequality. ⁵ Lindert and Williamson 1985, Williamson and...