Wednesday, April 21, 2021

Inequality mirage?

David Splinter and Gerald Auten gave last week's Hoover Economic Policy Working Group seminar, summarizing their past and some work in progress on the distribution of income.  Link in case the above embed does not work. A recent paper. Splinter's web page

Splinter and Auten are very even handed, just-the-facts, economists. I'll pass on their facts. Grumpy interpretations are my own. 

It is a fact generally accepted that income inequality has grown a lot recently, and this is a "problem" to be "solved." So what if the great inequality crisis simply isn't true? Let's leave aside whether income is a good measure (it isn't), let's just look at the fact, has income inequality substantially increased? 

No. Here is the headline result. In their careful redoing of the numbers, the top 1% share of income has barely budged since the 1970s. (And, by the way, if you think the mid 1970s economy was the great happy prosperity we should try to reestablish, you're too young to remember the 1970s.)

Now we get in to a deep under the hood exercise about costing up income, and where did Piketty and Saez go wrong. The video has some of that. The papers have more, and a long list of back and forth, including comparisons with many other studies. I'll name just a few.

Omitted income.   Piketty Saez leave out many kinds of income. Auten Splinter attribute all national income to somebody.  Before 1986 many wealthy people were incorporated.  Leaving out corporate income biases the early shares down. Auten Splinter fix that. Pre-tax and transfer income! Who cares about pre-tax income! Auten Splinter calculate income after taxes at the top -- lower -- and including transfers at the bottom -- higher. Demographics. Marriage rates have fallen, so Auten Splinter calculate income by individuals. Benefits! They include benefits like employer-provided health insurance.

One can quibble, but one can quibble. At least this cornerstone "fact" of political debate is a lot less sure than it looks. 

If the rich aren't getting richer, the poor aren't getting poorer:

More interestingly, let's get past shares and look at the second great article of contemporary faith: Have people of lower incomes really not done any better in the last decades?  What is their actual income, not just their share?

Actual income, including all the things Piketty and Saez left out, has risen. Actual income including transfers has risen a lot.

Even this is misleading, because the actual people are different. Those with low income today are different people than those with low income in 1970. Even if the the PS graph were correct, it would be false to say that "the poor got poorer" because they are different people. Suppose, for example that everyone stays the same over a lifetime, but that wages grow faster over time. The poor (young) would be poorer and the rich (old) richer, but yesterday's poor are today's rich, and the distribution of lifetime income unchanged. That's not what happened, just a story to illustrate the confusion 

The "cross section" approach looks at income growth of each percentile, i.e. how much is the income of today's top 1% different from the 1980s' top 1%. It does look like "the wealthy got all the gains." 

But the "panel approach" looks at the experience of actual people. If you look at people who were in the first decile in 1980, how much has their, personal income grown over time? We see the opposite pattern. People who had very low incomes saw much more growth. 

A third article of current faith, "the rich don't pay their fair share, and are paying less taxes" In fact, the US tax and transfer system is one of the most progressive in the world. 

This is a lovely graph as it properly includes all taxes not just federal income taxes. 

These are just taxes, including tax credits to generate negative numbers at the bottom. They do not include transfers, i.e. negative taxes. 

Ah the good old 1950s when we really stuck it to rich people? Also false. 

Even in events like Hoover donor reunions I am asked "what is the conservative answer to inequality?" My stock answer is that inequality is not a problem. If you think inequality is a problem, then you think the world is better off if Bill Gates is $1000 poorer and you are $10 poorer. A good interpretation of the feeling "inequality is a problem" is not the intolerable envy of the super-rich but the barriers to opportunity that people with low incomes face in the US. Now we have something to talk about and a fixable problem. Let's start with atrocious public schools in thrall of teacher's unions. 

But even if you don't buy every one of Auten and Splinter's numbers, it is not at all obvious that the "problem" of increasing inequality is real in the first place.

The problem of lack of opportunity, barriers to advancement, remains. And even if it isn't getting a lot worse, it was bad enough in the 1970s. Let's get to work on the real problems.  


  1. Although you touched upon this issue in another post recently, I am still surprised by the additional details presented here. Rising income inequality has been splattered across quite a solid number of pages over the last decade -- and not just in popular culture, but also in top journals in economics.

    It's somewhat of a shock that we could get a basic picture so deeply and severely wrong. I can't imagine that in current political climate we will get anything like a sane public conversation about it, especially not where people acknowledge almost everyone left and right were wrong and shift their focus toward providing opportunities to help people build their way out of poverty.

    1. It is not a shock at all. Read Caplan's The Myth of the Rational Voter and would be easy to understand why the voters vote what they vote.

      Politicians perfectly understand that too

  2. Prof., hope all is well. WODR Inequality is huge problem...Sincerely, your Student

  3. Paul Graham has an interesting recent blog post on this issue - "How People Get Rich Now". I'd be interested to know what you think of it.

  4. I am trying to square in my mind the following:

    pre tax and pre tax plus transfers have definitely gone up over time while the after tax numbers have not

    The top 1 percent tax numbers have stayed flat.

    The first statement would suggest that taxes have gone up in lockstep with the pre tax income growth while the latter would seem to contradict that. Perhaps another reading suggests most taxes have increased on high income earners but not on the top 1 percent?

    Or perhaps John is right and income is a poor measure to begin with.

  5. You are measuring the wrong thing! Income now is a MIRAGE....where the tax credit buffet and scams galore(offshore patent holding companies, non-profits paying people millions, etC) erase, hide income.

    You should only measure Wealth increase which Asset growth+income per year. The richest...say Warren Buffet, Bezos, Gates, don't take income and heck now they just push it into a FAKE lifestyle charity where they hold complete sway for decades! If I own piece of art(why the rich invest) increase value from $10 to a billion a year...did I get richer?

    Now redo this by WEALTH GROWTH(Income+ASSET GROWTH) and see what you get!
    This headline should help "US wealth inequality - top 0.1% worth as much as the bottom 90%"

    1. There's merit to your point but Wealth is hard to measure and very sensitive to assumptions that John laid out in his prior posts.

  6. Re: "Even this is misleading, because the actual people are different."

    This is probably most true of the top 1%. According to IRS data published in the NY Times circa 2000 and as stated by Thomas Sowell, most of the top one-percenters will be there only once in a ten-year period. The portion in the top 1% twice in a ten-year period is small.

    I suspect the top 1% is primarily composed of individuals that have experienced a once-in-a-lifetime economic event.

    R. S. Cox, CPA
    Austin, Texas

  7. It might be of some interest to compute the GINI value for the chart titled "Main Results: Top 1% Income Shares" for the upper and lower traces shown in the chart. The upper trace is labelled "Piketty-Saez with capital gains". The lower trace is labelled "After-tax income".

    For the period 2010-2020, the upper trace lodges around 20% of income while for the lower trace a figure of roughly 10% of income will serve the purpose for computing the GINI value of the income distribution.

    Suppose that the income distribution follows a Pareto distribution (cf., This is said to be a power-law distribution suitable for the characterization of population vs income distributions in many countries. The computation of the GINI value is described on the wikipedia webpage cited.

    For the upper trace, the Pareto Index (╬▒) is 1.967; for the lower trace, the Pareto Index is 2.53. The corresponding GINI values are 0.341 for the upper trace and 0.246 for the lower trace.

    The GINI value of 0.341 is consistent with GINI values given for the United States as well as for a number of other western OECD countries. This validates the method, within the bounds of accuracy of the data presented in the blog article.

    The lower trace, labelled "After-tax income", by this method gives a GINI value of 0.246 which indicates that the degree of inequality is reduced by the force of taxation consistent with the redistributive characteristic of the U.S. income tax code.

    Redistribution is the means that economist T. Piketty favors to improve equality (reduce inequality).

    The presentation of economists David Splinter and Gerald Auten confirms (1) that taxation for purposes of distribution reduces inequality, and (2) that the distribution of gross income has become more concentrated over the course of time (i.e., since circa 1980).

    As the American-trained French economist Avram Florin remarks in his on-line published notes on the topic of stochastic finance, wealth is a multiplicative process. For investor Warren Buffet, it is 'the miracle of compound interest' at work.

    We would expect to see increasing inequality in the distribution of total wealth over time, ceteris paribus. Senators E. Warren and B. Sanders would tackle that phenomenon by a tax on accumulated net wealth over and above a certain threshold or baseline level of wealth. Others might consider applying a lump-sum tax on wealth (i.e., a one-time levy applied in one tax period).

    Bottom-line: While the analyses performed by economists David Splinter and Gerald Auten is useful, it fails to address the issue of accumulated wealth and the perception of increasing inequality in the distribution of that wealth amongst the American population. That perception is the tougher nut to crack.

  8. Inequality goes down even further when you consider the present value of all government promises made: US Wealth is $100Tr[1], while we have ~40Tr in unfunded social security[2] and ~$45Tr in unfunded Medicare[3]. Even optimistically assuming wealth grows at twice the rate of obligations, the US has already voted to spend a 35% wealth tax, we just haven't got around to voting to pay for it.



Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.