Income Info

June 25th, 2011 at 12:05 am

In discussions of tax policy, folks like me tend to throw around references to “income fifths,” “the top 1%,” and such without enough context.  Below is a graph which shows the actual income levels in 1979 and 2007.  Unfortunately, 2007 is most recent year for this very comprehensive income data set published by the Congressional Budget Office.  But it’s still elucidating re the actually dollar values behind these income classifications and how they’ve changed over the past few decades.

The chart shows average household income levels for each fifth of the income scale, along with the top 10, 5, and 1 percent of households, in real 2007 dollars.

The fact that incomes at the top of the range have grown so much over these years kind of messes up the scale (yet another problem associated with the growth of income inequality—it’s getting hard to plot!) so I include a table as well.

Back in 1979, the average income of the top 1% was about 33 times that of the bottom fifth.  In 2007, that ratio had grown to 100.  The real income of middle-class households grew 19% over these years, less than 1% per year, while that of the top 5% grew by about 150% and the top 1%, by 240%.

Now you’ve got some context.  Sorry you asked??


Source: CBO



Print Friendly, PDF & Email

17 comments in reply to "Income Info"

  1. Saad says:

    A question, just conceptually why us it a problem if almost all income levels are up, if some only very modestly? I’m trying to understand why I should care that some who used to make 1 million a year went to 10 million a year, as long as my income in inflation adjusted dollars also went up. Or am I missing something?

    • Christopher says:

      For decades as the fortunes of the United States increased (GDP, wealth, productivity, etc.), the fruits of that prosperity were shared across the income spectrum much more fairly than the years since 1979. Now it isn’t. Something broke.

  2. tom says:

    Saad, I will give four reasons:

    1. Income inequality is bad per se. See the book “The Spirit Level” for details.
    2. Considered in terms of real dollars, income levels increasing but more slowly may not seem bad, (but note that the bottom quintile income increased less than 0.5% per year.) However, as a share of total income, their income decreased.
    3. People are working harder, and with more two income households, just to barely stay ahead of inflation.
    4. Typical household consumption has increased more than inflation. Maybe you want to call that ‘living beyond ones means’, but today, living without things like cell phones, cable tv, internet, etc. puts you well outside the mainstream of society.

  3. tom says:


    Is there a version of this data that accounts for the different composition of households: married, single, two income?

  4. Ken says:

    What about linking corporate tax rates to unemployment? As the unemployment rate decreases the corporate tax rate would be lowered.

  5. tom says:

    Alright, I went and tried to do the homework that I assigned. I did not find the quintiles adjusted for household composition, but I did find some info from the Census family income tables. From Table 16:

    Husband working full-time, year-round, one or more related children under 18 years old
    Work experience of wife
    Year Total Worked Worked full-time, year-round Did not work
    2009 19507 13911 8383 5596
    1987 19464 13807 6264 5657

    Median Income
    2009 87,091 95,646 101,968 60,789
    1987 72,856 76,644 84,355 62,077

    What I see from this is that family incomes when the wife doesn’t work have stagnated. All the increase comes from women being paid better, as well as more wives working full time.

  6. readerOfTeaLeaves says:

    This kind of contextual information is marvelous. Thank you!

    With respect to the chart, it is helpful in some respects: you can easily contrast 1979 against 2007 and that’s helpful.

    What I find misleading is the width of the bars.
    For instance, that bottom quintile — to be more visually accurate — should be 25x as wide as that top 1% bar, should it not?

    Because although it is wonderful to see the 1979-2007 differences in the blue-red bars, the width of the bars makes that lowest quintile seem to represent ‘about the same number of people’ as that 1% bar on the far right.
    They seem ‘about equal’ because the bars are the same width!

    This is actually visually inaccurate in terms of the actual relationships within the data, and the visual distortion that it inadvertantly creates results in misinformation IMVHO. (Or to be more *verbally* accurate IMVSHAPOO: In My Very Strongly Held And Possibly Obnoxious Opinion.)

    Maybe someone can correct me, but it seems extremely important to ‘correct the background information about population distributions’ by making it visually distinctive and accurate: aren’t we are actually talking about a POPULATION DISTRIBUTION that looks more like this?

    xxxxxxxxxxxxxxxxxxxxxxxxx bottom quintile
    xxxxxxxxxxxxxxxxxxxxxxxxx second quintile
    xxxxxxxxxxxxxxxxxxxxxxxxx third quintile
    xxxxxxxxxxxxxxx fourth quintile up to the top 10%
    xxxxxxxxxx = top 10%
    xxxxx = top 5%
    x = top 1%

    I am trying my best, with words and x’s, to point out that the bottom quintile

    xxxxxxxxxxxxxxxxxxxxxxxxx is not the same thing as


    And now, if we try to show the difference between 99% and 1%, one way is to do it with bars or x’s




    The differences between 1979 and 2007 are hugely important, but by collapsing the data about the actual population distributions, I think we are missing some key information.

    Nevertheless, the added context and the post are really interesting.
    But I think the graph could still be improved upon: the scope and magnitude of the shifts since 1979 need to be crystal clear to policy makers, and the rest of us.

  7. pjr says:

    For context it’s also helpful to know approximately what income puts someone into the top 1 percent, 5 percent, 50 percent and the like. Here’s some data–rounded figures for 2009 from Table 7 located at

    Remember that this is based on federal individual income tax returns:
    Top 1 percent: over $380,000 AGI
    Top 5 percent: over $160,000 AGI
    Top 10 percent: over $114,000 AGI
    Top 25 percent: over $67,000 AGI
    Top 50 percent: over $33,000 AGI

  8. tom says:

    ROTL, you might be looking for is the cumulative income distribution, or Lorenz curve. We can compute the shares of aggregate income for each quantile using the data table Jared posted. I get

    Shares of aggregate income
    Lowest Second Middle Fourth Highest
    Quintile Quintile Quintile Quintile Quintile Top 10% Top 5% Top 1%
    1979 5.18% 11.27% 16.89% 22.85% 43.80% 28.54% 19.42% 8.59%
    2007 3.80% 8.78% 13.32% 19.43% 54.67% 40.74% 31.56% 19.34%

    From this, you see that the aggregate income of the top 1% is roughly halfway between that of the bottom 20% and the second quintile in 1979, while by 2007 the share of income of the top 1% is equal to the share of the fourth quintile.

    Some other interesting observations: the share of income of all but the top quintile fell, and the share of income of the top quintile excluding the top 1% was essentially unchanged.

    • readerOfTeaLeaves says:

      Thanks for your thoughtful reply — I saw the data as you describe it.

      My point is that when you look at the bar graph, the bar graph makes it ‘look like’ we are breaking down 8 equally-sized groups of people.

      I’m saying, we’re breaking down 4 groups – by quartile.
      Within that fourth quartile, we break out subgroups (for 10, 5, 1).

      But if you look at the way the bar chart displays, it kind of ‘looks like’ we’re talking about 8 groups of 12.5 people in each group — because all of those bars are the same visual width. Therefore, the ‘top 1%’ looks as ‘wide’ as the lowest quartile.
      It’s a visual distortion that affects the way we think about the data.

      Yes, it’s super important to show **how much more** the top 1% had in 2007 as contrasted with 1979. That’s key information.

      But because all of the bars ‘look the same’, it ‘looks like’ we are breaking out 8 equal groups within the population and that does not accurately reflect the population distributions this data contains.

      But I follow your point and agree.
      It’s a good point.

      Just not the point that I was attempting to make.

  9. Kevin Rica says:

    But it is NATIONAL POLICY — signed by the Dems and Ted Kennedy to anchor wages at the bottom with immigration policy.

    What causes wages to rise at the aggregate level is an “excess demand for labor:” More jobs than workers. That is the mechanism in every macroeconomic textbook. Jam that mechanism and wages won’t rise.

    But when an excess demand for labor begins to emerge (employers can’t find enough workers), the Chamber of Commerce yells “labor shortage” (a synonym) and the National Democratic leadership, Nancy Pelosi, Harry Reid, and Teddy Kennedy in his day, start wetting their hankies and Poise Pads for men and crying compassionately (they are sensitive) we need more immigrants — Americans are too lazy!

    That may be the worst thing that happens in the immigration debate, a “shortage of labor,” the mechanism that drives wages up is misrepresented as some sort of economic emergency.

    And when the Chamber of Commerce refuses to allow the “guest workers” to be covered by the Fair Labor Standards Act, Teddy Kennedy takes their side:

    That’s the devil’s bargain that the national Democratic Party leadership has struck. Lower wages for more immigration. The rank and file doesn’t necessarily agree. That is why Hilary Clinton lost a commanding lead among the Democratic primary electorate when she muffed the drivers’ licenses for illegal immigrants question in the debate.

    So it’s national policy: lower wages for more immigrants. Ted Kennedy and his allies at the Chamber of Commerce signed off on the bargain.

    So those who accept this “Grand Bargain” with the Devil and the Chamber of Commerce should not complain about low incomes at the bottom of the income distribution. That was the bargain signed in others’ blood! A deal is a deal!

  10. steve says:

    “So those who accept this “Grand Bargain” with the Devil and the Chamber of Commerce should not complain about low incomes at the bottom of the income distribution. That was the bargain signed in others’ blood! A deal is a deal!”

    Literature suggests some immigration effects at the bottom, but we still have stagnation for everyone else but the top 1%-2%. Also, I had no idea that the Chamber of Commerce was such a left leaning group.


    • Kevin Rica says:

      “I had no idea that the Chamber of Commerce was such a left leaning group.”


      The CoC doesn’t have to lean, the left does the leaning.

      The modern left’s (Not “left” as Harry Truman would understand it) position has been to lean over backwards (what anthropologists would call “presenting”) to accomodate the CoC on immigration.

  11. Geoff Freedman says:

    I believe wealth redistribution to the top is one of the reasons for our current economic problems.

    During the 1920’s income tax rates were sharply regressed and the top rate went from 76% to 25% in 1925 via a series of tax cuts in 1922, 1924, and 1925. I believe there were other tax changes as well regarding investments and capital gains. In 1929, we have the stock market crash, and the Great Depression, bank failures and other financial issues follow and the Depression lasts until 1941.

    In 1981 Reagan began a series of tax initiative which again were essentially regressive in nature and resulted in declines in top rates. On October 19, 1987 we had Black Monday when the stock market crashed, and the recession of the early 1990’s followed along with the problems of the Savings and Loan industry, problems in Real Estate and other financial industry issues.

    Clinton introduced tax legislation that effectively made the tax code more progressive and the top income tax level was raised to 39.6 (effective rate with surcharge).

    In 2001 and 2003, Bush pushed passage of tax cuts (EGTRRA and JGTRRA) which were regressive tax cuts.

    The point of all this is that regressive tax cuts tend to place a greater percentage of net wealth into the hands of the top decile of wage earners, and I believe there is a tipping point when this increase in net wealth at the top results in investment of this excess capital into more risky and highly leveraged financial instruments (stock bought on low margin requirements in the 1920’s, real estate loans with lower requirements with savings and loans, stock and commodity options in the late 1980s and other forms of derivatives like collateralized debt obligations, and credit default swaps and real estate purchases in the 2000’s).

    The increase in assets (bubbles) caused by the this speculative investments masks the stagnating income that occurs to the rest of the population as one result of this wealth imbalace.

    The result is over time even though the consumptive economy seems to be humming along, bad things tend to happen in an imperfect and under regulated free market economy influenced by many other variables other than those just discussed.

    We need more a more progressive tax rate as one tool to get more wealth into the lower 80% of population to help prime our consumptive free market economy.

  12. Dick C says:

    Do these income figures include include capital gains?

    • Jared Bernstein says:

      Yep–realized cap gains included.

      • readerOfTeaLeaves says:

        Is there any way to break out capital gains in these graphs…? In other words, is there a quick and simple way to separate out the capital gains to see what percentage they are of the totals…?