January 29, 2014

New academic study links rising income inequality to ‘assortative mating’

Bride and groom holding hands
© Dejan Ristovski / istockphoto

Here’s another reason the rich are getting richer and the poor are falling farther behind:  A new working paper by an international team of economists finds that better educated people are increasingly more likely to marry other better-educated people while those with less formal schooling are more likely to choose a less well-educated partner.

As a consequence, income inequality has increased because education is strongly correlated with income—the more schooling you have, the more money you typically earn, according to a team of economists headed by Jeremy Greenwood of the University of Pennsylvania.

Economists call the tendency of people with similar characteristics to marry “assortative mating.” For their study, Greenwood and his team tracked patterns in marriages grouped by education level from 1960 through 2005 using U.S. Census data.

Their analysis identified three distinct trends.  Consistent with previous research, they found that “the degree of associative mating [by education level] had increased” over that time period, according to the working paper published by the National Bureau of Economic Research. (For a detailed look at marriage patterns of couples, see this Pew Research report.) 

But the big surprises came in household income trends among couples with relatively more and relatively less education. Virtually across the board, the income gap between couples with relatively high and those with relatively low levels of education had widened substantially since 1960 relative to the average household income.

For example, in 1960, a husband and wife, each with a high school education, would earn about 103% of the average household income.  But in 2005, that same couple would earn only about 83% of the average. At the other end of the education spectrum, a couple in which both partners had done post-graduate work earned about 176% of the mean household income in 1960 but a whopping 219% in 2005.

Expressed another way, the relative earnings of couples with high school degrees had fallen by 20 percentage points relative to the average while the household incomes of highly educated husbands and wives had increased by 43 points.

To assess the overall impact of these trends on income inequality, they conducted a novel test. They first computed the overall level of income inequality in 1960 and 2005.  Then they estimated what income inequality would have been if couples were randomly matched by education level. In effect, they asked what income inequality would have been if education didn’t matter in selecting a spouse, and if men and women with lots of schooling were as likely to marry people with relatively little education as they were to choose better educated partners.  The difference in those two numbers would mark the impact of associative mating by education on income inequality.

The statistic they used to gauge income inequality was the Gini coefficient, which measures income inequality on a scale from zero to 1.  Zero represents no inequality—as if everyone earns exactly the same amount—and 1 represents perfect inequality, which would occur if one person earns everything and everybody else makes nothing.

Greenwood and his colleagues estimated that the Gini coefficient was .34 in 1960, or about a third of the way to complete inequality. When they randomly matched people by education level and recalculated the coefficient, the answer was basically the same:  The Gini coefficient still stood at .34, suggesting that assortative mating by education played little, if any, role in income inequality.

Then they applied the same method to 2005 data. Now the overall Gini coefficient was .43, an increase of about .09 since 1960 and consistent with other research. But when they randomly matched people by education and re-ran their analysis, the Gini index plummeted to .34, showing that today, “assortative mating is important for income inequality.”

One reason for these changes is because more married women than ever are joining the labor force (and marrying similarly educated men), which reinforces the income gains for better educated couples.  Their evidence: When they randomly matched men and women by education level, income inequality in 2005 declined. (Other studies have also shown that the increase in married women’s labor force participation has not been the same across education groups.  College-educated married women have increased their work hours, so it has become even more valuable for college-educated guys to have college-educated wives, at least in monetary terms.)

The study is the latest entry in a contested area of research to examine inequality and income of married couples.  On the one hand, economist Gary Burtless of the Brookings Institution has found that between 10 percent and 16 percent of income inequality in the United States “was caused by the growing correlation of earned incomes received by husbands and wives.”

Researchers Deborah Reed and Maria Cancian reported  in 2001 that the increasing correlation of husbands and wives’ earnings in the late 1960s through the mid-1990s worked to worsen inequality. They also found, contrary to some researchers, that changes in men’s earnings was the largest source of rising income inequality while changes in women’s earning actually reduced the disparity.

About the authors: Jeremy Greenwood is a professor of economics at the University of Pennsylvania and a research associate at the university’s Population Studies Center. Nezih Guner is an adjunct professor at the Universitat Autònoma de Barcelona and Research Professor of the Barcelona Graduate School of Economics. Georgi Kocharkov is an assistant professor of economics at the University of Konstanz in Germany. Cezar Santos is an assistant professor of economics at the University of Mannheim in Germany.

Category: Social Studies

Topics: Educational Attainment, Income, Income Inequality, Marriage and Divorce

  1. Photo of Rich Morin

    is Senior Editor at the Pew Research Center’s Social & Demographic Trends Project.

Leave a Comment

Or

All comments must follow the Pew Research comment policy and will be moderated before posting.

15 Comments

  1. Bob B3 months ago

    A 15-page paper about assortive mating and its consequences for society, and the words intelligence and IQ are not mentioned once.

    Do the authors of this study – or PEW for that matter – not understand the education is generally – generally, I said – a proxy for IQ. Therefore, the real issue with assortive mating on such a massive scale isn’t that highly-educated people are separating themselves; it’s that high-IQ people are finding each other and mating in a way never seen before in history.

    If IQ is ~50% inherited, and you systematically put together high-IQ people – who also tend to have the best environments for learning – you’re going to create a permanent upper class. More often than not – again, in general – their kids will inherit high-IQ genes, which coupled with good environments will lead the vast majority of those kids to do quite well in school and life. (Of course, a few of those kids will fall down the ladder and a few working-class kids will rise.) Rinse, wash, repeat. Over time, you will have near permanent class of high-IQ people – say 10-20% of the population – that will have no connection to the rest of the population and a nearly insurmountable advantage over them.

    Now, obviously, such a future would never play out perfectly like that. We are humans after all. There will be upper class kids who have emotional issues or who simply prefer to work with their hands or who fall in love with the daughter of their mechanic. However, even if high-IQ people mate with each other only 75% of the time, it will slowly create a new upper class that will little in common with and, therefore, likely little sympathy for the rest of the population.

    Not sure how you stop this from happening, but it doesn’t bode well for the country.

    Reply
  2. Steve3 months ago

    I like the term “associative ” mating (also used in this article), better than “assortative” mating. Although the article really is talking about marriage, rather than just mating. I Don’t have raw data but it seems more people, since 1960, are living together before marriage or in lieu of marriage which may serve as a trial period that may illuminate differences in education levels and personal goals that ultimately (notice the word mate in there) may become a deal breaker.

    Reply
  3. kristina b3 months ago

    to everyone complaining….its not particularly about the conclusion, but the percentages. its about the AMOUNT of inequality and mate-matching NOT the fact that it exists. ugh

    Reply
  4. Dan3 months ago

    Hopefully this study was not done with some grant funded by taxpayers.
    Any educated person knows and has known this from childhood

    Reply
  5. Mike3 months ago

    Duh……isn’t this something you have been told your entire life. Get the education and you can increase your life time earnings. How much money as spent on this research?

    Reply
  6. JMK3 months ago

    You had to do a study to come to that conclusion? Are you going to do a study next to see what color people perceive the sky to be?

    Reply
  7. john3 months ago

    oh, so “assortative mating” is a concept from economics????

    Reply
  8. Sharon3 months ago

    At last, an article which states that levels of education help to elevate people to better lives.
    The national school dropout rate is very high. Poverty is also fostered by unmarried mothers with babies, drug and alcohol problems, living in the moment without a life plan.
    Self responsibility, staying in school is a worthwhile start to a better life.

    Reply
  9. slk3 months ago

    give a million people, a million each, and after some time, how many will be worse off then before!!! 46% of all jackpot winners go broke!!! it’s human nature that’ll keep many people with almost nothing!!!

    Reply
  10. edw. zimmermann3 months ago

    Are there statistics available as to the percentage of younger students/workers who are becoming aware of the given statistics? Has this percentage advanced or declined over the recent past?
    Has it affected the percentage of unmarried couples vs. married couples; positive or negative?
    why or why not?

    Reply
  11. Yezdyar Kaoosji3 months ago

    Assortative Mating! Even spell-check does not know that word!
    Do you need research to draw this commonsense conclusion?
    People will most naturally meet and marry those in the circles in which they live and work.

    Reply
  12. SZachry3 months ago

    This is the same conclusion that Charles Murray reached two years ago in his book Coming Apart. There are many reasons for the current state of the wealth divide and they do not fall neatly along partisan lines. Whatever the reasons, Americans had better wake up and do something about this; nothing less than the stability of our democracy is at stake, in my opinion.

    Reply
  13. Silence DoGood3 months ago

    People must understand. If you want to not be broke. You must save money. Its the only way. If you dont save money no matter what your income is you will be broke. When you get the wage increase save it. Or when the next time this discussion comes back around. You will still be broke. Possibly even more broke.

    Watch this and learn how to save money. Join the dark side of the force.

    youtube.com/watch?v=nVat9VBsSCg

    Reply
  14. Jay B+Bern3 months ago

    Poorer people are more likely not to marry at all, but they do drive themselves deeper into poverty by having children they can’t afford (education, proper nutrition and secure living situations).

    Reply
    1. slk3 months ago

      unfortunately thats a business for some!!!

      Reply