In order to measure economic development and living standards in the past, economic historians have constructed real wages for many countries. They have generally found that Britain, the Low Countries, and British North America had substantially higher real wages than the rest of the world. One of the assumptions of this methodology, however, is that family size has been constant across time and between countries. This blog post tests that assertion by building a demographic simulation to understand how family size has changed over time in England. I find that, contrary to the claims of some historians, family size was quite low in England and did not change substantially over time. Thus, existing estimates of real wages accurately capture average family welfare at the average point in the family life cycle.
In the past ten years economic historians have become increasingly interested in comparing living standards across countries and over time. One of the favourite methods for measuring living standards is to calculate real wages for unskilled labourers using comparable, subsistence-level price baskets (Allen et al., 2011; Allen et al., 2012). These studies have generally found that Britain, the Low Countries, and the British North American Colonies had wages that were two to three times higher than real wages in the rest of the world in the seventeenth and eighteenth centuries (Figure 1). This little divergence is important because it helped spark the Industrial Revolution in Britain, which in turn, led to the ‘Great Divergence’ between the wealthy developed world and the poor developing world that we see today.
Figure 1: Welfare ratios (real wages) around the world in the early modern period.
This research has provided many interesting conclusions about relative economic performance in the early modern world, but in order to make the real wages internationally comparable, Allen and his co-authors (including myself) had to make some simplifying assumptions. One of the assumptions was that the average family’s required consumption was equal to the consumption of three adult males. This assumption was held constant over time and across countries.
This may seem like an incredibly bold assertion, especially to demographers who are aware of the tremendous demographic heterogeneity in societies around the world. However, there is unfortunately little data that would help economic historians adjust real wages to family size over time. Even for Britain, whose demographic history has been well documented, the only measure of family size that existed for the early modern period was completed family size, or the number of children born to married mothers who survived to the age of 50.
Therefore, I set out to estimate a more realistic metric for family size from which family consumption requirements could be calculated: the average number of children being supported in a household at any given time. This allowed me to adjust the real wage estimation for changes in family size over time (Schneider, 2013). To do this, I built a demographic simulation based on the family reconstitutions of 26 English parishes (1580-1837) compiled and analysed by Wrigley et al. (1997). The simulation has two components. First, the model predicts individual families, showing how their required consumption and real wages change over time as additional children are born, die, and leave the household. Parents are also given a certain probability of dying in any given year, and if the father died, the family’s income was cut in half. Second, I predicted 20,000 families recording the median and maximum number of children being supported at any given time and the median and minimum family real wage across the family life cycle.
After running the simulations, I found that family size was much smaller than some historians had thought: in the period 1750-1800 the average family was supporting a median of 2.29 children across their family life cycle. Family size was low because wide birth spacing limited the number of children who could cohabit at any given time, many children died before reaching maturity, and some mothers died before reaching menopause. In addition, Allen’s estimate that the average required family consumption was equal to the equivalent consumption three adult men was approximately correct. The family size adjusted real wages were within 10 per cent of Allen’s own calculation. However, these family size adjusted real wages only refer to the average family at the average point in their family life cycle. The average family was much worse off at the minimum real wage in their family life cycle, suggesting that Allen’s estimate of real wages hides life cycle poverty.
Although the simulation I described only adjusted real wages for England, it can be extended to the global story raised at the beginning of this post. Family sizes in Asia, Latin America and the rest of Europe would have to be much smaller than Britain’s family size to overcome Britain’s real wage advantage. This seems unlikely given Britain’s late age at marriage and nuclear household norm. Thus, the early modern British wage advantage appears to have withstood demographic scrutiny!
Allen, Robert C., Jean-Pascal Bassino, Debin Ma, Christine Moll-Murata and Jan Luiten van Zanden, ‘Wages, Prices, and Living Standards in China, 1738-1925: in Comparison with Europe, Japan, and India’, The Economic History Review, 64, no. S1 (2011), pp. 8-38.
Allen, Robert C., Tommy E. Murphy and Eric B. Schneider, ‘The Colonial Origins of Divergence in the Americas: A Labour Market Approach’, The Journal of Economic History, 72, no. 4 (2012), pp. 863–894.
Schneider, Eric B., ‘Real wages and the family: Adjusting real wages to changing demography in pre-modern England’, Explorations in Economic History, 50, no. 1 (2013), pp. 99–115.
Wrigley, E. A., R. S. Davies, J. E. Oeppen and R. S. Schofield, English Population History from Family Reconstitution, 1580-1837 (Cambridge, 1997).