Founding editor of The Independent Review Robert Higgs reminds us that
[i]ncome inequality is a statistical artifact. . . . The aggregate of the measure
is arbitrary: why, for example, should inequality be measured for the entire
U.S. population rather than for the population of the city or state in which one
lives, the entire North American population (including Mexico), the entire Western
Hemisphere population, or indeed the entire world population? (2014).
He might have added or indeed the entire world population throughout all of
history. In fact, measuring inequality this way may be very illuminating. Therefore,
in the calculations below I take a preliminary stab at putting the incomes of todays
poor Americans into the broadest historical perspective.
Branko Milanovic (2010) provides data showing where people in different
income ventiles (twentieths of the distribution) in several countries fall within the
worlds income distribution. He finds that the income of the poorest 5 percent of
Americans puts them at the 68th percentile of the worlds income distribution in 2010.
Poor Americans are rather rich by the worlds standards today. However, they are far
richer in comparison to all the people who have ever lived.
Virtually everyone in the premodern era lived at a small multiple of the subsistence
level. Although subsistence cannot be precisely measured, Milanovic, Lindert,
and Williamson (2007) estimate (following Angus Maddison 2001) that subsistence
equals $400 per year in 1990 dollars, which equals about $855 per year today. Jeffrey
Williamson (2009, 2930) provides estimates of the ratio of the mean income to the
subsistence income for a wide range of countries over a period of almost two thousand
years. In the years before modern economic growth began, these ratios averaged
about 3.1falling below 2 in places like medieval Byzantium (AD 1000), Mughal
India (1750), Qing China (1880), and colonial Kenya (1927), while slightly exceeding
2 in ancient Rome and exceeding 5 in a handful of places, including England
(1759) and Holland (1732). Although a few places still have very low mean-income to
subsistence-income ratiosboth Tanzania and the Democratic Republic of Congo
fell below 2 at the turn of the millenniumthe ratio is now over 60 in places like
Britain, France, Germany, Italy, Japan, and the United States (where it is over 80).
The U.S. Censuss Annual Social and Economic Supplement reports incomes of
$0 and $800 for the bottom two percentiles in 2020 (DQYDJ 2021), which is obviously
incorrect. Income estimates are well below consumption levels for the American
poor, as shown by data from the Bureau of Labor Statistics (2020). The BLSs
Consumer Expenditure Survey calculates that households with incomes in the bottom
10.4 percent (those earning less than $15,000 in 2020) earned 0.93% of total income
in the United States while their consumption was 4.7% of the total. The gap is due to
transfers from government and nongovernment sources (such as family members), unreported
income in the formal and informal economy, and other factors. Consumption
by the American poor is roughly five times (5.065) their income levels, although this
ratio is probably an even higher multiple for the poorest of the poor. Accordingly, in
comparing the bottom ventile (5 percent) of modern Americans to all the people who
have ever lived, I use the consumption estimate, although one could make alternative
adjustments. In 2020, average household incomes in the bottom ventile were $4,300 per
year, so annual consumption was closer to $21,800. There is an average of 1.7 people in
these households (called consumer units in the Consumer Expenditure Survey report),
which means that the consumption per person averaged about $12,820.
Max Roser (n.d., using data from Van Zanden et al. 2014) estimates that among
the worlds 1.09 billion people living in 1820 about 30.5 million had incomes above
$12,800 per year (in todays dollars), putting todays Americans from the bottom
ventile a bit above the 97th percentile (97.2) of the 1820 population. Angus Maddison
(2001, 28) estimates that average income levels of AD 1 and AD 1000 were about
one-third below those in 1800. If all the income estimates given by Roser are reduced
by one-third, this implies that todays Americans from the bottom twentieth would be
a bit above the 98th percentile (98.1) among people living in the premodern period.
Kaneda and Haub (2020) estimate that about 94.4 billion of the 116.8 billion
human beings who have ever lived were born by 1650, and 101.6 billion were born
Using these numbers (see table 1), about 1.8 billion of the people born by 1650
were richer than those in the bottom five percent of Americans living today, and
about 200 million of those born between 1650 and 1850 were richer than todays
There are currently about 7.9 billion people living, and Milanovic estimates
that about 32 percent of todays population have incomes above the bottom ventile
of the United States, so about 2.53 billion people in the world have higher incomes
than the poorest 5 percent of Americans. Similar calculations using Rosers estimates
put modern Americans of the bottom ventile at the 80th percentile worldwide in the
middle of the twentieth century. There are about 7.3 billion people who were born
after 1850 and who are no longer living. If modern Americans in the bottom ventile
have incomes at the 80th percentile among them, this makes 1.46 billion of this final
group who had incomes higher than the poorest 5 percent of todays Americans.
There are surely some mismeasurements here, but Ive tried to be conservative,
for example, by ignoring bias in the consumer price index, which understates the
value of todays dollars versus the 1990 dollars used by Roser. I have assumed that
consumption equaled income before the advent of the welfare state in recent times.
Table 1 summarizes these results. Adding together the four groups (2.53
billion living + 1.46 billion born after 1850 but no longer living + 200 million from
between 1650 and 1850 + 1.8 billion born before 1650), approximately 5.99 billion
people have had incomes higher than the bottom 5 percent of Americans today.
In other words, the poorest Americans have higher incomes than roughly 95 percent of
all the humans who have ever lived.
This is a profound and little appreciated point: the poorest among us today are
very well off economically by the standards of all of world history. They live lives that
exceed even the aspirations of most people who have ever lived. Virtually everyone
living in the United States today is an economic winner. Will anyone among us decry
 Leeson, Hardy, and Suarez (2022) report that panhandlers in Washington, D.C. metro stations take
in about $6.10 per hour.
 Rosers figures are in 1990 dollars, which Ive converted into 2020 dollars using the consumer price
index, which estimates prices to be 2.14 times higher in 2020. This ignores the upward bias in the
CPI due to substitution bias, quality change bias, new goods bias, and other biases. See, for example,
Hausman 2003. Stripping out these biases would reduce the 2.14 ratio to about 1.67 by my back-of-theenvelope