Income inequality is higher in America than in most European countries. As a typical example, here is a comparison of the U.S. to Germany — Europe’s largest economy with an income distribution that is pretty typical for the continent:

% Shares of National Income by Income Quintile: Germany vs. America

The poorest fifth of the German population has about 9 percent of national income versus about 5 percent for the poorest fifth of Americans, and the richest fifth of Germans have about 37 percent of national income compared to about 46 percent in America.

In order to think about this, imagine a stylized example of a very homogeneous, middle-class town of 10,000 people next to a very homogeneous, poor town of 10,000 people. Call the middle-class town Laketown, and the poor town Hillville. Both Laketown and Hillville are homogeneous, so each has very low inequality by all standard statistical metrics. 

Now, imagine that the state government establishes a statistical reporting district defined by the sum of Laketown and Hillville. Call it Lakeville. Lakeville has dramatically higher inequality metrics than either town. Everybody’s life is exactly the same, but now each of them lives in a “place” with much greater inequality. 

The key point is a simple one, but is often overlooked in these debates: inequality statistics are determined by the definition of the reference population.

What is the “true” inequality experienced by a given person in Lakeville? This is in part a subjective question — to whom do I choose to compare myself: my cousins, other people in my town, other people in Lakeville, or America, or the world, or my parents, or other people at my company, or some other group? And it is in part an objective question — with whom am I in a community required to create and sustain my standard of living? 

Suppose I told you that very few people move residences between the towns, but that all of the gardeners, garbagemen and store clerks in Laketown live in Hillville, and all of the doctors, lawyers and engineers that serve the people of Hillville live in Laketown. They share a common police department, but maintain separate fire departments. In Laketown, the fire department is all-volunteer. Hillville pays a fee to the county for fire protection. They have separate elementary schools, but both feed into a common regional high school. They share little league and soccer leagues. The local Elks Club covers both towns, but each town has a separate chamber of commerce. 

Residents of Laketown often claim to live in a very egalitarian community. This is not an empty boast, but it’s not so obvious that the whole picture is captured by some statistical metric that covers only the income distribution for legal residents of Laketown.

Think about this in real-world terms for Europe (where I live) and America. The EU has about 500 million people compared to about 300 million in the U.S. Distinct national cultures exist across Europe, but there is substantial shared sovereignty between countries, extensive law and regulation is established centrally, and most of the countries share a currency and a central bank. Though they are substantially less integrated than the states of the U.S., the countries of Europe exist in an economic union that has massive internal trade, but external trade between the EU as a whole and the rest of the world that looks a lot like America’s level of external trade. 

Suppose we compared inequality for America to inequality for Europe as a whole — how would it look?

I took the data on income by quintile for each country in the EU, and ordered them from poorest to richest for Europe as a whole. If you start with the 1.5 million people who are the lowest quintile in any country of the EU (it turns out to be the lowest quintile in Bulgaria), then add the 0.5 million who are the lowest quintile in Latvia, and keep going until you have about 100 million people, you have an estimate for the lowest quintile for Europe as a whole.  This bottom quintile includes, for example, a majority of the population of Bulgaria, Latvia and Romania, and the bottom quintiles of the UK, Ireland, Italy and France. You can then build each of the quintiles this way. (This will tend to make Europe look more equal than if I had individual-level data, and further, the data I had is on a purchasing power parity (PPP) basis, which will also tend to make this look more equal than if I had market-rate exchange data for the non-Euro members.)

Here’s the result:

% Shares of National Income by Income Quintile: Europe vs. America

It turns out that America and Europe as a whole have extremely similar levels of economic inequality — Europe’s is just chunked by country, 

The only academic analyses that I have seen that tried to do the same kind of analysis have come to the same basic conclusion that Europe as a whole has a roughly comparable level of inequality as does America (a little less on a PPP basis, and little more on an exchange-rate basis). There is a detailed methodology debate here.

Of course, this is not a complete comparison. America and the EU as a whole both have large trading partners (even though at a much smaller scale relative to GDP than intra-EU trade for a typical European country); why wouldn’t we include those somehow? How aware are people in Helsinki of incomes in Bulgaria, and how much does this affect their perceptions of inequality? Further, relative income inequality between two jurisdictions is not simply an issue of size — e.g., most American states have greater income inequality than most European countries — but is influenced by practical opportunities for internal migration, intermarriage, human capital stocks, rates and sources of immigration into the national society, and many other factors. In general, I believe that most people have a shifting, nuanced, and layered view of inequality that sometimes crosses national boundaries and sometimes doesn’t cross the street. And so on. 

I have written previously that I believe inequality in America to be a symptom of deep problems, but the real story is lot more complicated than the cartoon of “Europe equal, America unequal.”