January 27, 2014

Despite recovery, fewer Americans identify as middle class

Despite a slowly recovering economy, the proportion of Americans who identify themselves as middle class has dropped sharply in recent years. Today, about as many Americans identify themselves as lower or lower-middle class (40%) as say they are in the middle class (44%), according to a recent Pew Research Center/USA TODAY survey.

FT_14.01.24_middleClass_line_420The nationally representative survey of 1,504 adults conducted Jan. 15-19 found that the share of Americans who identify with the middle class has never been lower, dropping  to 44% in the latest survey  from 53% in 2008 during the first months of the Great Recession.

At the same time, the share of the public who says they are in the lower or lower-middle classes rose by 15 percentage points, from 25% in 2008 to 40% today.

As a consequence of these shifts, the difference in the proportion of Americans who say they are middle class and those who say they are in the lower classes has fallen from 28 percentage points in 2008 (53% vs. 25%) to a statistically insignificant 4 points today (44% vs. 40%).

The survey also found that the upper classes have not been immune from the changes reshaping the middle and lower classes. In 2008, about one-in-five Americans (21%) said they were in the upper or upper-middle class. Today, only 15% place themselves at or near the top of the social and economic order.

The findings from the latest survey suggest a significant part of the change in the size of the middle class and lower classes has happened in just the past two years. In a July 2012 Pew Research survey, about half (49%) of the public identified themselves as middle class, five percentage points more than do so today.

In a closer look at the data, we found that virtually every key demographic group saw its social standing slip as the proportion of Americans who identified themselves as lower or lower middle class increased over this time period. But the decline was not experienced equally by all groups.

For example, in 2008 about a quarter (24%) of those with some college experience but no bachelor’s degree placed themselves in the lower or lower-middle classes. Today, that proportion has about doubled to 47%, a 23 percentage point increase compared with the overall 15-point rise. In contrast, the percentage of college graduates who identify themselves as being in the lower or lower-middle class rose only 8 percentage points to 20%.

Young adults ages 18 to 29 also were disproportionately more likely to report a decline in social standing. In 2008, a quarter of all young adults identified as being in the lower or lower-middle class; today about twice as many do (49%), a 24-point increase. In contrast, the share of adults 65 and older who say they are in the lower classes increased by 10 points to 31%.

The increased reluctance of the American public to self-identify as middle class is consistent with recent economic trends. According to the Census Bureau, the median household income in the U.S. decreased by 8% since the Great Recession started in December 2007—from $55,627 in 2007 to $51,017 in 2012 (all figures in 2012 dollars). The median income in 2012 was at the same level it was in 1995, a setback of 17 years.

Moreover, losses have been greater for households in the middle of the income distribution than for households at the top. For households in the third quintile of the income distribution, average income fell by 8% from 2007 to 2012. Meanwhile, for households in the highest quintile, average income fell by only 2%. The net result is growing inequality in the income distribution.

A separate study we conducted in 2012 also found a hollowing of the middle of the income distribution. The study used Census Bureau data to classify adults into lower, middle or upper income tiers depending on the income of their household relative to the overall median. It finds that the share of adults in the middle income tiers has fallen in each of the last five decades, from 61% in 1971 to 51% in 2011.

Economists also report a lack of jobs growth in middle-skill, middle-income jobs. An analysis by the New York Federal Reserve Bank found that employment in middle-skill jobs increased by 46% from 1980 to 2009. Meanwhile, employment in low-skill jobs increased 110% and employment in high-skill jobs increased 100%. This phenomenon of “jobs polarization” is perhaps most assiduously studied by David Autor, an MIT economist. His research demonstrates that employment growth over the past three decades has steadily gravitated toward low-skill jobs.

Topics: Income, Wealth, Middle Class

  1. Photo of Rakesh Kochhar

    is an associate director of research at Pew Research Center.

  2. Photo of Rich Morin

    is a senior editor focusing on social and demographic trends at Pew Research Center.


  1. Charles Hickman3 years ago

    Please state in dollars what is low, middle, and upper income classes/levers! Thanks and have a bless day/night. Charles H

    1. Rakesh Kochhar3 years ago

      In this analysis, the classification of people into lower, middle or upper class is based on their self-identification in the survey. In other words, we did not use income to determine the distribution. For an analysis of how people are distributed into lower, middle and upper income groups based on their household income see pewresearch.org/fact-tank/2015/0…

  2. Travis3 years ago

    Why is it significant to know what people “say” they “think” their class is? A study with concrete numbers and specific brackets for each class would be much more helpful as people search for financial class information online.

    1. Jane3 years ago

      they have charts like this too, What people think is important, if you struggle day by day and week by week with no savings and no money to have extra you know your in the lower class. and anyone who thinks this is not a problem when single moms can’t afford to buy their kids shoes is not thinking abt the good of americas future for the working class, and indeed for the majority.

  3. Allauoi3 years ago

    Most memoirs I read about the coirevsnon of the religious to the non-religious describe studying the bible or becoming aware of other religions as being the core cause of disbelief. That’s why a lot of churches have Bible Study courses (because it’s only been written in English for a few centuries, so let’s have a class on how to read a book), so that they can control how their flock interpret what is written. If you sit with that book by yourself and read it critically, it becomes ridiculous, and even offensive to a modern sense of morality. But when studied in a group led by a preacher, you get that group-think factor and the the ringleader can better control your reaction to otherwise horrifying stories. Preacher, mentalist, seminarian: same mechanism, same bull. I would even extend this to teachers, not to be exclusive of positive examples of this practice. However, I can say that learning evolution in school was much more zombie-ish than re-learning and re-discovering it for myself in my thirties. This is why I’m more supportive of one-on-one learning and self-learning (both parentally inspired of course) than the education-mills that our classrooms have turned into. Then again, when I show my kids the gorgeous images of galaxies that the Hubble has brought us and explain the immense numbers of stars with as much awe and excitement that this fact brings me, the only response I get is yeah, real cool, dad, we get it, lots of stars, whoopee! So maybe such realizations aren’t possible without at first a sense of interest.

  4. VonVictor Roesnchild4 years ago

    Very interesting article, and I thank all those that contributed to it, and those that presented it. I remember a few years back, right when the Global Recession was about to kick in ‘full force’, that there were more than a few executives working at major financial institutions at that time, were saying; “there is no such thing as Middle Class in America. They went on to say, that there exist now the following social levels: Poor ->Working Poor -> The Rich/The Wealthy. Those that use to fit the Middle Class, have now been forced to come out of retirement, take more loans against their homes, and some forced in to the lowest states of poverty, and ending up in Homeless Shelters, which in and of themselves have seen a mark rise across the United States. There are more people living in Homeless Shelters than living in homes.

    1. Amy Livingston2 years ago

      That’s an alarming statistic. May I ask where it comes from, and how it’s defined? Do you mean that there are more people in homeless shelters than in homes that they own, or more in shelters than in homes of any kind – owned, rented, or shared?

  5. Joseph Malgeri4 years ago

    This post from PEW points to the significant erosion of the middle class. Pay attention.

  6. Daniel Sullivan4 years ago

    This study is one of the most important wake up calls america could receive. I sincerely hope the Obama administration has read it has thoughtfully considered its implications and will address what to do about it tonight with the Presidents address. there are two other Professors at the Sloan School who may have the explanation for this see Race Against Machines technology is and always has destroyed jobs the issue never seriously questioned before is does it create as many new jobs as it destroys and more concerning who will fill those jobs? not the unemployed even if they have a college degree.. I am currently writing a book on technology and inequality and your study confirms my worse fears.

  7. Dan Ryan4 years ago

    Unless you look jointly at change in my income and change in my self-identified class it’s really hard to see how you tease apart the effects of “rhetorical culture” (is it OK to call yourself upper or lower or middle?) changing definitions, and subjective assessment of economic status. Perhaps you’ve done this but the blog post doesn’t communicate it.

  8. Chris Knight4 years ago

    Begs the question. Who’s economic recovery are you talking about in the first sentence?