email print
Blog Post

Animated chart of the day: America’s middle-class is disappearing…. but it’s because they’re moving up

AEIdeas

The new animated “bar chart race” visualization above is a dynamic version of the second static chart, and both show the percent shares of US households by total money income for three income categories annually from 1967 to 2017: a) low-income households earning $35,000 or less, b) middle-income households earning between $35,000 and $100,000 and c) high-income households earning $100,000 or more (all in constant 2017 dollars). The income data are from the Census Bureau report “Income and Poverty in the United States: 2017.”

Here’s how I explained the income share trends displayed above in a post on CD last fall:

Yes, the “middle-class is disappearing” as we hear all the time, but it’s because middle-income households in the US are gradually moving up to higher income groups, and not down into lower-income groups. In 1967, only 9% of US households (only 1 in 11) earned $100,000 or more (in 2017 dollars). In 2017, more than 1 in 4 US households (29.2%) were in that high-income category, a new record high. In other words, over the last half-century, the share of US households earning incomes of $100,000 or more (in 2017 dollars) has more than tripled! At the same time, the share of middle-income households earning $35,000 to $100,000 (in 2017 dollars) has decreased over time, from more than half of US households in 1967 (53.8%) to less than half (only 41.3%) in 2017. Likewise, the share of low-income households earning $35,000 or less (in 2017 dollars) has decreased from more than one-third of households in 1967 (37.2%) to below one-third of US households last year (29.5%), a near-record low.

Bottom Line: As can be seen in the visualization above, America’s middle class did start largely disappearing in the 1970s, but it was because they were moving up to higher-income groups, not down into a lower-income category. And that movement was so significant that between 1967 and 2017, the share of American households earning incomes above $100,000 more than tripled, from 9% to 29.2%. Many prominent people like Paul Krugman and progressive politicians like Sen. Bernie Sanders and Sen. Elizabeth Warren claim that American’s middle class has been declining, disappearing, collapsing, losing ground, vanished, stagnated, etc. But the Census Bureau data on household income over time displayed above demonstrate conclusively that those assertions are incredibly and verifiably wrong.

Think about it for a moment and let it sink in — in 2017 nearly one out of three (almost 37 million) US households had annual incomes of $100,000 or more. And the share of American households with that level of income has increased by more than three times since 1967! Then compare that picture of a prosperous America with millions of middle-class households moving up into higher income groups to the narratives we hear all the time that the American middle class is: losing ground, falling behind, collapsing, stagnating, disappearing, fill in the blank ___________.

 

Discussion (41 comments)

  1. tkc says:

    How government can solve this ‘problem’!

    Soak the Almost Rich.

    https://newrepublic.com/article/110859/soak-the-almost-rich

    I’ll just show myself out.

  2. Seattle Sam says:

    Last night we heard over and over that none of the trillions of dollars of new government programs would be paid for by the “middle class”. I would predict there will be a lot of households with two earners and $100K in income who will be very surprised when they get the bill.

    1. Citizen Buddy says:

      Although I do not agree with Senator Sanders on socialism, he is the only Democrat presidential candidate that states taxes on the middle class will go up to pay for one choice medical coverage. All the others smile and lie.

      1. Ron H. says:

        “You’re gonna have single payer medical coverage whether you want it or not, and you’re going to pay through the ass for it. Vote for me.”

        Sure thing.

      2. Tom Terwilliger says:

        Yes, at least Sanders (unlike most of the others) is honest. And I guess its better to be honest and wrong than dishonest and also wrong.

  3. Vic Volpe says:

    Unanimated chart of the day — Income inequality by quintile —
    https://www.facebook.com/photo.php?fbid=2639422086102706&set=a.165771190134487&type=3&theater

    Source: U.S. Census Bureau, Current Population Survey, Annual Social and Economic Supplements. For information on confidentiality protection, sampling error, nonsampling error, and definitions, see //www2.census.gov/programs-surveys/cps/techdocs/cpsmar18.pdf.

    1. Not Sure says:

      Here’s a small sampling of my findings done a few years ago (which you can explore in more detail here). I have since updated these findings to 2019; the trend of falling work-time costs continues.

      – In 1975 a ten-cup drip coffee maker cost the ordinary American worker almost 8 hours of work time; today it costs that worker only 45 minutes.

      – In 1975 a pair of all-cotton jeans cost 1.5 hours of work time; today it costs 20 minutes.

      – In 1975 a car battery cost 9 hours of work time; today it costs 3.6 hours.

      – In 1975 a microwave oven cost 93 hours of work time; today a microwave oven of similar size and power costs 6 hours.

      – In 1975 an automatic dishwasher cost nearly 50 hours of work time; today a dishwasher costs 12 hours.

      – In 1975 an exercise bike cost 16 hours; today it costs 5 hours.

      – In 1975 a low-priced standard-size electric washer-dryer combo cost 70 hours of work time; today this combo costs about half that, at 36 hours.

      https://www.aier.org/article/myth-american-middle-class-stagnation

        1. Not Sure says:

          Things that are heavily regulated or nearly completely run by the government cost a lot? Who could have guessed?

          Why are you so fixated on encouraging people to be envious of what other people have?

        2. Ron H. says:

          Vic

          Yes, as Not Sure pointed out, those things with heavy government involvement seem to have increased in price much faster than overall inflation as opposed to those things left to the market.

          Notice also that average hourly wages have increased at a higher rate than overall inflation.

      1. Vic Volpe says:

        Half of items you listed have had some type of government support — coffee, cotton, car batteries, microwaves — not exactly market forces at work, at least not without some form of government involvement.

        1. Ron H. says:

          Vic

          It’s important to differentiate between intended and unintended consequences.

          For those items you mentioned the *intent* of government action was to make them more expensive and that was known at the time. Lucky for us, markets are often so powerful they actually overcome those government efforts, and prices eventually drop anyway ( e.g. Not Sure’s list).

          But when the stated goal of government meddling is to make things more affordable, as in medical care, hospital costs, education, and housing – watch out! Predictably those prices will go through the roof as we can see on Mark’s “Chart of the Century”.

        2. Jon Murphy says:

          Half of items you listed have had some type of government support — coffee, cotton, car batteries, microwaves — not exactly market forces at work, at least not without some form of government involvement.

          That’s true but also irrelevant. We’re talking degrees here. Housing, medical, education, range from extremely controlled to outright owned and operated by government. The other things you list are considerably less so. It’s not a “one single drop” argument.

          Also, for more on rising costs of education etc, see this recent paper by Alex Tabarrok and Eric Helland. It’s a tad technical, but not too bad.

      2. James Steele says:

        See my response below to the Census bureau data – stagnation is real. What you are talking about is efficiency of production. How about medical costs? Do you think the same comparisons between 1967 and 2019 would hold true? What about dental? What about housing? The things that are outside of manufacturing ARE more costly. Are you are citing is an expectation – manufacturing productivity DOES increase every year so we would expect to see the things you are citing.

        1. Tom Terwilliger says:

          Medical and dental costs have risen much more rapidly because of government involvement. And because there are so many medical options that did not exist 50 years ago. Whenever some politician claims that someone has a “right” to all these medical advances I always note that my grandparent’s rights must have been continually violated, (so maybe they should have sued someone??) because they had none of these medical advances. Comparing medical care of today with that of 50 years ago is truly an apples to oranges comparison.

          Housing can’t really be compared either because houses today are so much larger than houses from 50 or 75 years ago.

    2. Jon Murphy says:

      Ah yes, Vic. More evidence of Mark’s point.

    3. Ron H. says:

      Nice cat. Nice dog Vic.

      BTW Vic, a better explanation of that picture at your facebook link titled “”Life On Hollywood Boulevard” might be the following:

      “A Courteous pedestrian respects a woman’s privacy by walking quietly past her as she sleeps on the sidewalk in her accustomed spot.

      The pedestrian’s expression of compassion and concern the previous day had been met by screaming and cursing from the woman who apparently resented being awakened by a meddling stranger.”

  4. Ted says:

    Couldn’t the rise of $100K+ households in the animated graph be because women spouses have been working more in the past few decades instead of staying at home while at the same time middle class income has gone down for each person working?

    1. Jon Murphy says:

      Possible, but not probable. The number of earners per household has been generally flat/falling.

      But, even if it were the case, that’s still progress and an indication that people are living better.

      1. Ron H. says:

        Yes. More former stay-at-home spouses entering the workplace indicates the cost of maintaining a household has decreased compared to the value of their labor in the workplace.

        Thank you technology. Thank you inexpensive immigrant labor.

    2. Citizen Buddy says:

      Ted, umm, if incomes had not gone up, then the trend for households would not have risen. Thus, middle and lower income classes would not have moved up over the last fifty years.

  5. MortMain says:

    The government pays people not to work.

  6. Matt says:

    This graph seems misleading. $100,000 is worth a lot less today than it did in 1967. Shouldn’t this need to count for inflation?

    1. Not Sure says:

      Both of them say they are measuring in 2017 dollars.

      1. Mark Perry says:

        Yes, the text of the post and charts actually reports “2017 dollars” seven different times!

  7. cc says:

    Part of the problem I see here is rising expectations. In 1967 the average house size was much smaller and on a per person basis even smaller. Furniture was not as nice. Parents had just gotten our first dishwasher. Cars were crap. There were 4 channels on black and white TV. No internet or iPhone or GPS. People had much smaller wardrobes. Eating out was rare.
    The other irony here is innumeracy: when you define the “middle class” as is done as between fixed $ values ($35k to $100k), the median income almost by definition stays constant. Anyone whose income rises moves into the higher class, as in the graph. Thus “stagnant wages” of an imaginary group of people.

  8. Ted says:

    If the number of earners per household has been generally flat/falling during the 50 year period in Mark Perry’s animated graph, then Mark should add that annotation to his graph. As the graph is currently presented, it could be seen as a clever way to obscure decreased earnings per worker, i.e., more spouses are having to work in order to make ends meet (especially to cover medical insurance costs which in earlier decades were more commonly provided by employers).

    1. Mark Perry says:

      The average household size has declined from 3.28 in 1967 to 2.54 in 2017, meaning that median household income per household member has gone up over time more than household income has gone up.

      For two-earner households their real median household income went up by more than 30% between 1987 and 2017 to an all-time high of more than $98,000 (in 2017 dollars). For all US households, real median income went by less than 14%.

      1. Ted says:

        Thank’s for the additional information Mark. Still, it seems to me that household income obscures things. Why not use income *per earner* in your animated graph to show the changes over the last fifty years in low/middle/high income brackets? My guess is that income per earner data (not sure if this is the correct technical term) is not available going back to 1967?

        1. Mr. Econotarian says:

          How about Real Compensation Per House going back to 1947: https://fred.stlouisfed.org/series/COMPRNFB

          Here are “number of earners per household” going back to 1980:
          https://caseforcapitalism.files.wordpress.com/2018/11/stagnant-wages-average-earners-per-household-1980-2017.jpg

          “Since 1980, the average number of earners per household has actually dropped, from 1.39 earners to 1.26”

  9. James Steele says:

    Your article doesn’t quite jive with the census data report. According to the census report median household income went from about $46,000 in 1967 to $61,372 (in 2017 dollars) which represents an annualized increase of 0.5% per year (in 2017 dollars). See the chart on page 5 of the report https://www.census.gov/content/dam/Census/library/publications/2018/demo/p60-263.pdf

    The census data supports the wage stagnation which is what feels correct. According to table A-2 on page 35 of the report even the 95th percentile limit (the upper 5% income bracket), which is the largest growth sector over the 50-year period, only changed by an annualized 1.33% rate in terms of 2017 dollars.

    What makes the numbers visually appealing is that they are over a 50-year period so that even a relivatively flat 1.33% change per year ends up looking huge over that long period. This is hardly an incredible shrinking middle class. A 5-10% change per year, yes that would be the incredible shrinking middle class, not a puny 1.33%/year.

    Now what we really need to think about is that this is household income – how many more couples work today than in 1967? Quite a few more. Considering that, I think that we see this doesn’t represent income growth at all.

    1. Mark Perry says:

      The real median family income for married couples with both spouses working has more than tripled, from $34,800 to $111,000 in 2017 (in 2017 dollars).

      1. James Steele says:

        How do you explain Figure 1 (page 5) then where the median family income for 2017 is listed as $61,372 (in 2017 dollars). Can you give a reference in the report to the data you are citing?

        1. Mark Perry says:

          The source for: “The real median family income for married couples with both spouses working has more than tripled, from $34,800 to $111,000 in 2017 (in 2017 dollars)” is below, “Table F-7. Type of Family (All Races) by Median and Mean Income”

          https://www.census.gov/data/tables/time-series/demo/income-poverty/historical-income-families.html

          1. James Steele says:

            Yes, I see the table, but…

            1) Married Couple family data from Table F-7. 2017 median income – $90,148, 1967 median income – $54,317. This gives us a 1.42% increase/yr (2017 dollars) hardly blockbuster growth.

            2) Wife in paid labor force data from Table F-7. 2017 median income – $110,893, 1967 median income – $64,066. This gives us a 1.10% increase/yr (2017 dollars) again hardly blockbuster growth.

            3) Percentage of wife in paid labor force of the married couples. Using raw numbers what we see is that in 2017 the ratio is 58.1 % and in 1967 the ratio is 36.6%. This represents a 59% increase in the number of wives in paid labor force over 1967. So using table F-7 all families: 2017 median income – $75,938, 1967 median income – $51,048 – this represents a growth rate of 0.797%/yr (in 2017 dollars) but we know that in this 59% more working wives are in the labor force. All together this spells STAGNATION.

          2. Mark Perry says:

            The main point of the post is this, not a story of stagnation, but a story of significant upward mobility:

            In 1967, only 9% of US households (only 1 in 11) earned $100,000 or more (in 2017 dollars). In 2017, more than 1 in 4 US households (29.2%) were in that high-income category, a new record high. In other words, over the last half-century, the share of US households earning incomes of $100,000 or more (in 2017 dollars) has more than tripled! At the same time, the share of middle-income households earning $35,000 to $100,000 (in 2017 dollars) has decreased over time, from more than half of US households in 1967 (53.8%) to less than half (only 41.3%) in 2017. Likewise, the share of low-income households earning $35,000 or less (in 2017 dollars) has decreased from more than one-third of households in 1967 (37.2%) to below one-third of US households last year (29.5%), a near-record low.

    2. Mark Perry says:

      Your article doesn’t quite jive with the census data report.

      Note: The article and graph are based on data from the Census Bureau report.

      1. James Steele says:

        The 50 years period causes a large number gain – that’s why I looked at growth rates. But there are no triple values as you suggest. All I saw from the same sets of data are the numbers that I gave above.

        Just look at growth rates alone. Overall growth rate from 1967 to 2017 for median family income sees a annualized 0.797% growth rate.

        More women work outside the home. Look at table F-23 from the data set that you cited for the percent of wives earning more than their husbands. 2017 is almost twice that of 1981 and the data doesn’t go back any further than that. It’s almost a linear relationship between 2017 and 1981 – it might be linear all the way to 1967 for all we know.

        When we have a 59% increase in the number of working women and their salaries have continued to go up over time. That is the only factor that leads to your “growth”. You can’t look at absolute numbers and conclude anything – you need to look at growth rates. That eliminates time from making the numbers look so big over 50 years.

        1. Mark Perry says:

          Think about it for a moment and let it sink in — in 2017 nearly one out of three (almost 37 million) US households had annual incomes of $100,000 or more. And the share of American households with that level of income has increased by more than three times since 1967! Then compare that picture of a prosperous America with millions of middle-class households moving up into higher income groups to the narratives we hear all the time that the American middle class is: losing ground, falling behind, collapsing, stagnating, disappearing, fill in the blank ___________.

        2. Mark Perry says:

          Families and household have changed over time, so a comparison of 1967 to 2017 has to account for that. For example, real median household income increased by 33.8% from 1967 to 2017. But median household income per household member increased more than 2X as much, by 72.8%.

Comments are closed.