Minnesota Big Data and Global Wage Inequality

Professor Fatih Guvenen, set against a blurred backdrop and wearing a blue button down shirt

Professor Fatih Guvenen of the Department of Economics knows income inequality is one of the more misunderstood economic trends discussed in popular media today. Although his research conducted through the Minnesota Big Data Institute (MEBDI) attests to the oft-cited “shrinking middle class,” it pushes back on the assumption that income inequality is rising, writ-large, in every corner of the globe.

“If you look at our data,” says Dr. Guvenen, “you’ll see that while income inequality rose significantly in the U.S. since the 1980s, income distribution became more equitable in places like Brazil and Argentina, and stayed relatively stable in France.”

An International Team and Scope

Although MEBDI’s research is international in scope, the organization’s database, the Global Repository of Income Dynamics (GRID), is hosted and operated here at the University of Minnesota. Boasting income data from 13 different countries: Argentina, Brazil, Canada, Denmark, France, Germany, Italy, Mexico, Norway, Spain, Sweden, the UK, and the US, phase one of the GRID project includes 51 economists from all over the world to study the state of income inequality. 

To Guvenen, it’s useful to have a single database that explores wage distribution globally. As he puts it, “income inequality can vary remarkably around the world, and the world is not just the United States.” Conversely, when income distribution does shift simultaneously across the globe, like during the COVID-19 pandemic, then researchers can have a better understanding of the scope of the changes at play.

“If you see inequality among working-class people rising in the United States but nowhere else, it would indicate that the issue is one of U.S. policy,” Guvenen says. “But, if income changes are reflected in the data as a global phenomenon, then we know that what’s occurring is much more fundamental.”

A Global Income Trend

Income volatility, or the likelihood of major fluctuations in wage, is one measure in which the GRID project found global trends. In his research paper “Global trends in income inequality and income dynamics,” Guvenen and his colleagues discuss that, across countries, income volatility is highest among workers at the very top and very bottom of the wage distribution.

While low-income workers are likely to see sharp increases in their wages—if you make $10,000 a year it is easier to double your income than if you make $60,000 annually—they also face a higher probability of sharp decreases. The same goes for the very top earners as well, as these individuals’ earnings are more closely tied to the fortunes of their employers that are subject to change depending on market conditions. This is not to say that all high-income earners experience the same amount of income volatility, just the individuals at the very top of the distribution.

Working With a Unique Dataset

GRID is not the only data project that investigates income inequality and volatility, however their methodology does bring unique contributions to the table. While most other datasets rely on income distribution among randomly-selected population samples, GRID actually follows the income shifts of the same people, year after year. By partnering with the statistical agencies of each country studied, the project gains access to what researchers call “income dynamics.”

Unlike random samples that change each year, the study of longitudinal income dynamics allows researchers to see the way a person’s wage changes throughout their life, shedding light on the big and small income shifts we have all experienced, from time to time. Studying income distribution this way allows MEBDI researchers to take a closer look at how social mobility may depend on individual characteristics, and better reflects the eb and flow of real life. 

For example, a man who graduates college, begins his career, and then takes a two-year hiatus from the workforce to be a stay-at-home parent, would have all of these changes reflected in GRID. Changes that may be missed were the dataset to rely on random samples alone.

With phase one of the GRID project now complete, Guvenen and his colleagues at MEBDI are on track to expand the project in phase two. By more than doubling the number of countries involved in the study, the hope is to create the most comprehensive open-source investigative tool for the study of global income dynamics available to researchers today.

Research papers from the 13 countries in phase one are available now on the GRID project website.

Share on: