Ye Olde Inæqualitee Shoppe

A quick note : Income inequality in pre-industrial societies was, in general, lower than in modern industrial societies, but traditional agrarian economies tended to be closer to their “maximum feasible inequality” than modern ones.

In “Russian Inequality on the Eve of Revolution” (I’m obliged by protocol to perform some kind of headgear ritual with Ben Southwood) there is a table which shows various measures of income inequality in pre-revolutionary Russia, along with some comparator societies :


It’s not surprising that income inequality should be low in agrarian, pre-industrial societies. You actually need some income to achieve substantial inequality ! But on second thought, that doesn’t properly put things into perspective, because you should compare actual inequality with the “maximum potential inequality” possible given the society’s level of per capita income. This point is vividly illustrated in Milanovic, Lindert & Williamson, “Pre-Industrial Income Inequality”, which supplies the interesting “inequality possibility frontier” curve :



The idea is, there is an absolute level of subsistence income for the mass of peasants who are creating value in the traditional society. Above that level is the surplus, or the amount of income which can be expropriated by elites ; and below that amount… starvation for the producers in the economy.

Thus, in a very poor society whose average income is close to the absolute survival level, the surplus extracted is pretty small and therefore inequality cannot be very high. But as average income rises, there is potentially more to extract. So the interesting question becomes, did pre-industrial societies at different levels of income have different “extraction levels” ? Put another way: did peasant incomes also rise when the average income rose or did the increase simply lead to more elite extraction ?


(You can click the above image to enlarge.) Subsistence income is assumed to be $300 in 1990 international dollars.

Incomes may have been more equally distributed in China in 1880 than in England in 1688, but that’s only because the average income was quite low in China. Perhaps more importantly, China in 1880 was closer to its maximum potential inequality than England in 1688. The chart below compares preindustrial societies with modern ones :

inequality possibility frontier

Some countries considered very unequal, like modern Brazil, are definitely more unequal than most pre-industrial societies, but it’s much farther from its “potential maximum inequality” precisely because it produces much more income than Peru in 1876. There’s just more to go around.

So for preindustrial or developing countries with very low incomes, it’s best to consider what Milanovic et al. call “inequality extration ratios” — the ratio of inequality to “maximum feasible inequality”, both measured in Gini terms. As average income rises with development, inequality also usually rises, but the extraction ratio should fall. Yet for some very poor countries today the “inequality extraction ratio” remains almost mediaeval :

IER table

Here’s also a full page of comparisons with modern and preindustrial societies.

Kenneth Pomeranz, in The Great Divergence, observed :

There are signs that Indian income distribution was significantly more unequal (and so popular consumption more limited) than it was in China, Japan, or western Europe. A study of Mogul land taxes for 1647 finds that 445 families received 61.5 percent of all revenues, which were about 50 percent of gross agricultural output, and that roughly one-quarter of the revenue flow to those families represented actual personal income. (The rest was consumed in various expenses of office.) [526] If this is accurate, these 445 families— presumably less than .002 percent of the population—would have received an income from their offices alone equal to 7.5 percent of total agricultural output, or perhaps 6 percent of the society’s total income! [527] An estimate based on Shireen Moosvi’s reconstructions for 1595 [528] is similar: it suggests that 1,671 Mughal nobles would have had a net personal income from their claims on government revenue alone equal to about 7 percent of total empire-wide output.

Mughal India does appear (in the table in the link above) to have been particularly extractive, but the other two historical values for India — especially the one for Bihar circa 1800 — not nearly as much. And you’d expect Bihar — one of the most benighted places on earth — to be a particularly exploitive place ! So Pomeranz may have been overgeneralising from the Mughal intensification of the preexisting zamindari system in north India.

About pseudoerasmus
This entry was posted in Branko Milanovic, Income distribution, Inequality and tagged , , , , . Bookmark the permalink.

3 Responses to Ye Olde Inæqualitee Shoppe

  1. Pingback: On the relevance of the income inequality to expected health expenditures. – Random Critical Analysis

  2. Johannes Parvus says:

    This is very interesting, but the graph and commentary seem to disagree with the full page table regarding 1880 China. In the table, China 1880 is at 55.2% extraction, lower than 1668 England and Wales’s 57.2% extraction.


  3. Johannes Parvus says:

    Looking at the graph more closely, it agrees with the table that China 1880 was far from its potential maximum inequality, even further than 1688 E&W. This surprises me, and makes dying Qing China an outlier among non-European pre -industrial societies, but that what these data say.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.