Are Incomes in the United States Becoming More Unequal?

Today we’re going to delve into Income Equality, a topic which has been fresh in the American collective consciousness ever since the Great Recession ended.

Gini Index

If you click the above image, you’ll see a graph from the St. Louis Fed (Census Bureau Data) showing a measure known as the ‘Gini Index’ headed up and to the right.  A Gini Coefficient measures inequality in a distribution, but is most often used in the press as a shortcut to demonstrate rising wage inequality in a country.  In America, this tendency has lead to a fair amount of Sturm und Drang amongst the Literati about how {insert policy or political scapegoat here} is destroying the middle class, starving the poor, enriching the wealthy, or something else generic and scary.

You’re looking at a graph of the Gini coefficient spanning back to 1967 (.397) and ranging to 2011 (.477).  In essence?  The graph is showing that income inequality is widening in the population sampled – the population being households of the United States.  Here are the numbers on some other countries.

An Important Distinction

Now, nobody is questioning the validity of the data shown in the graph.  Undoubtedly, household incomes are diverging and becoming more unequal.  Note, however, that I keep mentioning the word ‘households’.  Are we really measuring what we want to measure?  If our concern is with incomes in the United States… shouldn’t we actually be looking at individual incomes?

At this point, I should note that this line of questioning isn’t home grown.  Ironman at Political Calculations has already used published Census tables to calculate individual income Ginis (this led to some nonsensical arguments from people politically opposed to the idea that household inequality is overrated).  Mathematician and Physicist Ivan Kitov has done some work with Census data back to 1947.  The main arguments against their methods?  Garbage in, garbage out.

You see, Census income data is ‘topcoded’.  That means that incomes are bracketed by the whims of the Census, with ranges arbitrarily picked back in the beginning of census-taking and the ‘top income’ set at some value like $100,000 or $250,000.  The argument is there are uncaptured differences because of the low resolution of the lower brackets and the top coding (since >$250,000 ‘only’ puts you around or at the top 1% of individuals).


Okay, so the issue here is people opposed to the individual income argument don’t like the data being used to create the models.  Luckily for us, we can go a step further than the Census data, and pull microdata from the University of Minnesota, who release anonymous Current Population Survey data to anyone who can fog a mirror (well, and sign up for an email address).  A great grant supported service if I’ve ever seen one, IPUMS-CPS was particularly helpful in getting me microdata going all the way back to 1962 – and as an added bonus, since March is the goto heavy interview month of the CPS, we even have 2012 in there.  Is the data topcoded?  Sure is… but at $999,999, the number of people falling out of that range is so miniscule you’d really be clutching at straws to keep hating the data.  Let’s see what we found:

Data before 1968 is a bit weird, but contact me and I can provide you with the full set.

Explaining the Graph

Why 2 lines?  Well, we wanted to preempt complaints about rich people receiving money from non-wage income so we did it for both wages and all sources.  You’re welcome.

As you can see, there is an easy way to equalize incomes – just have a recession!  The tech bubble was a tremendous income equalizer, as was the end of Jimmy Carter’s presidency/the beginning of Ronald Reagan.


Miriam King, Steven Ruggles, J. Trent Alexander, Sarah Flood, Katie Genadek, Matthew B. Schroeder, Brandon Trampe, and Rebecca Vick. Integrated Public Use Microdata Series, Current Population Survey: Version 3.0. [Machine-readable database]. Minneapolis: University of Minnesota, 2010.

(To reproduce my data, select all March samples from 1962-2012, weight by WTSUPP, and use the weighted Gini calculator in the R package laeken).

Wrap it Up…

So, what exactly is going on here?  Individual incomes are less unequal than they were previously, yet household income inequality is reaching new highs.  Well, actually, that explains it – something is up with households.  That makes logical sense – if incomes are depressed, perhaps some of the following things are happening?

  • The lower earners are less likely to marry other lower earners, meaning that poorer households are actually single person households.
  • Students earning less money than expected move back home after college, increasing middle class incomes (turning 1 and 2 earner households into 2 or 3)
  • Folks of all stripes are holding off on divorce, meaning that already formed households have 2 earners instead of one.
  • (I kid, mostly…) The well off are pooling resources in communal living setups.

So perhaps we’ve been going about this ‘issue’ all wrong.  Instead of trying to dictate economic and tax policy (and protests!) on this flawed number, perhaps we should try to fix it with social policy?  Here’s one proposal: make the rich marry the poor!

PK for President, 2016!


  1. says

    You have far more data and a more concise answer to this. Thank you!

    My post today about moving back from the US to permanently live in Canada because it’s better for me financially, was mostly about what I personally saw, read, and felt.

    • says

      You’re the first person to call one of my articles concise! Thanks, haha.

      If there’s one thing we do horrendously wrong it’s skilled immigration. Living and working as I do in the Bay Area (and attending school at USC, which prides itself on representing the most countries of any University), I think it’s ridiculous the barriers we throw up to educated immigrants. I know in your article you call it economic slavery, but the term I like better is “indentured servitude”. The fact that some of my foreign friends are beholden to a firm until they get a green card is just an absurd notion to me.

      In essence, we draw the smartest people in the world to our shores because of the best education in the world, then either tell them to suck it up for 7-10 years for a green card (if you can get work with a mid-large company) or go home. What’s the point?

      • says

        Concise enough for me! Data-heavy = Good. I’m very lazy in that respect when I write posts so I’m very impressed by those who can put all that info into one post without making me read for half an hour.

        What I don’t understand why they WOULDN’T want me to work as a freelancer. I could be helping their companies that don’t want to pay for a large corporation (or those who won’t look at them as clients), and I’d be pushing money into their economy.

        I get that maybe they don’t like foreigners in general, but Canada isn’t really that foreign..

        We’re skilled, they need us, but only under their terms, which I am not willing to put up with.

        In contrast, Canada just has a rule where you need to have a (real) degree. Then you come here, and better have enough money to support yourself while you look for a job.

        If you can’t hack it, they know you’ll leave to go back. Simple as that.
        Plus the cold weather keeps people away.

        • says

          On the flip side, it did take me a few hours to write it, haha. Some of that was waiting for my computer to spit back results, of course – but any trip to IPUMS is going to be a long one.

          I’ve heard it kicked around the political circles a bit, but here’s a short description of my favorite skilled immigration ideal: we staple a green card to every single advanced degree we hand out. PhD in Physics? Why… here’s your Green Card! Makes sense to me.

          Even if you argue that not all degrees are the same, and one degree means different things at different schools, degrees are still a useful sorting mechanism:

          1) You need to be smart enough to get into the school
          2) You need to be motivated enough to actually finish the degree
          3) People who fit 1 & 2 are rarely a net drain on society (even if they come out with a ton of loans, they generally pay more back into the system over a career)

          Did you try working for a big company? My last 2 jobs have been at big companies (read: big enough that most people have heard of them). They are big and they have tons of resources – if you work for a large or even a medium firm the legal department will generally do what it takes to get you a green card. The gears turn slowly, but I think big companies get a bad rap on PF blogs. It takes a certain size of firm to be able to work through all the ridiculous requirements (like the absolutely stupid H1B quota).

          • says

            I like it. :)

            I’d have gotten the Green Card without a doubt then. I also like expanding the idea to stapling that green card to proven occupations that are on the US “Desperately Need These People” list.

            I looked into bigger companies but ultimately decided against joining them.
            1) I don’t want to wait 10+ years to get a green card before I can quit and freelance.

            I was willing to wait 5 years, max, for a Green Card. Otherwise, it’s just limiting my income.

            2) I can already freelance in Canada and make at least 2-3X what they’d pay me for those 10+ years, working half the time.

            3) I am a Canadian, so we don’t have H1B’s where they make you switch it to a GC after 6 years. TN’s are hypothetically renewable forever in 3-year increments.

            This makes it very easy for a company to say: *Nope, not interested in making you a GC-holder. You stay on the TN.*

            At the end of it all, staying in Canada is not so bad. There’s more than just the GC problem for skilled immigrants for me. As a freelancer, the crazy healthcare cost was a big problem.

            To put things into perspective, to get the equivalent healthcare in the US to what I have in Canada, I’d have to pay $1000 a month, net.

            In Canada, I took private insurance for a month, and it cost $85.

          • says

            If you come back, come to the SF Bay and get into one of your big companies? Our timeline is 4-6 years on the green cards, but from what I hear from people who have done it we treat them right. I also have a few finance friends who did what you described – got green cards, waited a year, then went off on their own.

            I don’t know enough about incomes in Canada vs. the United States to tell you what to expect, though! I know that as an Engineer, I’m better off in Silicon Valley, but as in all things… your mileage may vary.

  2. freeby50 says

    Interesting stuff.

    So the Fed data for households shows GINI going from 0.395 to 0.475 roughly. Looking at the individual income numbers it goes up from 0.51 to 0.54-0.55.

    So if you look at incomes it is worse actually as 0.55 is worse than 0.47. But its not growing as fast as the GINI for households. However in either case its still getting worse.

    I would guess that having more women in the workforce might have impacted the difference. That would worsen the difference between single people and families as well as between single parents and married couples. I would also bet that an aging population would make income disparity look worse. As more people age and retire you get more households among the retired population who look poor by comparison.

    • says

      Yes- but I have no further detail on how the Fed calculates their Gini numbers. I can tell you that taking the weighted median and mean from the March CPS data gives higher resulats than the Census Bureau. I would roll through all of the years to match it up, but for a large number of samples only the March data is available.

      Well, what is a “good Gini”? Do we care about income Ginis, household income Ginis, or something else, like a wealth Gini? (This disparity is most clear for Sweden, where the household income Gini is in the .30 range, while the wealth Gini is somewhere between .79 and .9).

      • freeby50 says

        Yeah I would think that wealth would matter more than income. Someone sitting on $1M in the bank with $50k income is doing considerably better than someone with $0 bank account and $60k income.
        I found a list on Wikipedia for the ‘wealth gini’ per nation and the US is among the highest at 0.8 That list has Sweden at 0.74.

        • says

          Are there any countries with wealth Ginis less than incomes? I’m interested in the QOL of a society like that, heh.

        • sullivanb3 says

          Great point. Romney’s wife tried to claim that they were both jobless at one point and really struggling to survive but yet Mitt had millions in investments. I’m sure they were living in the projects at the time.

  3. JT says

    I think its a function of number 1 and 2. Lower income earners are probably less likely to marry other low income earners. Students and graduates are probably having some slight effect on total household income. Since this data doesn’t take into effect the cost of debt service for an individual, it makes perfect sense. Quite possible for someone who earns $40k to have just about nothing after loan payments, other expenses, yet when added to a middle class income, $40k would have a huge effect.

    • says

      Well, they are more likely to marry other low income earners than high income earners (same deal for the rich. So 1x + 1x << 10x + 10x << 100x + 100x. Social policy can fix this!

  4. Funancials says

    Kings will marry servants and all will be well. I like it.

    I like what Sam had to say about income inequality…does income inequality really matter if social equality is there? The people that are “worse off” are living like ballers.

    • says

      I agree completely – rising tides lift all ships!

      Would you rather be poor in America or poor in Europe 1,000 years ago?

  5. AverageJoe says

    I’d have to ask my wife first if it’s okay if I marry someone else…. Great stuff. I find it hard to believe that someone would try to “dictate economic and tax policy on this flawed number….” Politicians wouldn’t do that….would they?

    • says

      Reminds me of this image!

      I really don’t like it. But, you know me, I’m a Milton Friedman negative income tax man. Dictator for a day I’d spend a ton of time revamping the EITC and eliminating wasteful welfare programs instead of worrying that rich people were marrying other rich people and their combined income is more than some other household.