Fall 2024 Featured Research Profile

Climbing the Economic Ladder, and Counting the Rungs: Economists Debraj Ray & Garance Genicot Explore a New Method to Measure Upward Mobility

Traditional economics has long divorced questions of economic growth from those of
economic equity. The (in)famous Kuznets Curve posits that, as countries’ economies grow,
inequality will rise but ultimately fall again. The prominence of this theory, and others like it, has
led mainstream economists of yesteryear to focus almost exclusively on growth, handwaving
away concerns about inequality as an unfortunate side effect that can be cured by more growth.

As appealing as that conclusion was, in the decades since the 1960s, developing countries
have seen significant increases in inequality despite sustained economic growth. This has led
many economists to foster a more nuanced understanding of growth, as potentially both inclusive
and exclusive, depending on how wide a subset of a country’s population shares in that country’s
overall growth. Though, understanding growth equity requires measuring it. One such area
economists are interested in measuring is upward mobility: the ability and frequency with which
individuals move to higher socioeconomic positions.

When it comes to measuring upward mobility, there’s good news and bad news. The good
news is that economists have developed multiple clever and statistically strong ways to quantify
this important area of interest. The bad news is that many of these measures require panel data. Panel data is a type of dataset that collects information over time from the same set of subjects,
whether it be people, companies, countries, or some other category. This type of data allows
economists to leverage powerful statistical estimation techniques, but is expensive and
time-consuming to collect, as it requires observations from multiple time periods on the same
entities. As a result, the reliable panel data on income and wealth that you’d need to estimate
upward mobility can be hard to come by, especially so if you’re interested in studying
developing countries.

This is where economists Debraj Ray & Garance Genicot come in. In their paper
“Measuring Upward Mobility,” the two researchers craft a mathematically robust and statistically
rigorous measure of upward mobility that is panel-independent – requiring only income or wealth
information across time, not necessarily from the same individuals. They start with a central, and
rather intuitive, assumption they refer to as Growth Progressivity, which assumes that an increase
in a poorer individual’s income growth rate relative to a decrease in a richer individual’s income
growth rate constitutes an increase in upward mobility. With this baseline, and a few other key
assumptions, Ray & Genicot are able to construct their new measure of upward mobility, which
turns out to be robust enough to accommodate dynamic population sizes and stratification by
social groups with only some minor modifications.

Not only do Ray & Genicot construct this new method, but they conclude their paper by
taking it to the data for three empirical test runs. The first concerns the work of famous public
economist Raj Chetty, who’s work on upward mobility relies on measuring the share of families
whose children earn more than their parents. While an appealingly intuitive measurement, this
method is panel dependent, and Ray & Genicot demonstrate that essentially identical conclusions
can be drawn using their measurement of upward mobility without using panel data. Ray & Genicot’s new method is justified formally via their mathematical proofs, but the fact that this
measure mimics the finding of a well-known panel-dependent study offers some informal
support.

Emboldened by this support, the paper then covers new ground, by using its measurement
to investigate the upward mobility of a selection of countries where panel data isn’t widely
available. The paper looks at Brazil, India and France from 1980 to 2010, and concisely plots the
countries’ growth rate, upward mobility (μ), and relative mobility (ρ) all as annualized
percentages over ten-year periods, seen below:

Finally, the paper concludes by examining the “Great Gatsby Curve,” a plot made popular by
economist Alan Krueger that argued for a negative relationship between intergenerational
income mobility and income inequality. The curve is essentially a visualization of the argument
that, as a country becomes more unequal, it becomes increasingly hard for the poor if said country to advance their economic position generation on generation – the inequality acting as a
poverty trap that prevents upward mobility. Similarly to the first empirical exercise, Ray &
Genicot’s new method mimics the negative correlation finding of Krueger when looking at the
same countries. What’s interesting here is that, without the panel data restriction, when
expanding the set of countries to 71 the correlation actually becomes very slightly positive!

These three empirical examples are brief, but they offer a jumping off point for further
study. Not only does the new method mimic past reliable work on upward mobility, but it
expands the number of countries that work can be done in. And as the final example
demonstrates, this expansion into developing countries has the possibility.