The Morass That Swallowed the Middle Class

By Matthew Shaw

While much of the inequality debate focuses on the gains of “the 1%,” less attention has been paid to the economic well-being of what is broadly termed the middle class, which is all too often just lumped into the other “99%.”  However, focusing the debate on only the 1% obscures important trends within each of these groups, including that there is ample evidence that the gains of the 1% are largely driven by the wealthiest among this already-elite group along with diminishing prospects for the rest of us.  Today, we look at one of these groups with diminishing prospects and a concerning trend recently highlighted by IMF staff: the “hollowing out” of the U.S. middle class. 

Even though many Americans across the income spectrum consider themselves to be middle class – an icon of American self-image – they’re not.  This is a shift with profound policy and political consequence.

First, some basics.  The IMF paper defines the middle class as households with incomes greater than 50% but less than 150% of the national median in a given year.  It then extrapolates what the middle class’ cumulative share of income would be in a benchmark, “zero inequality” society – i.e., one where the bottom 10% of households earn a 10% income share, the bottom 50% earn 50%, the bottom 90% earn 90%, etc.  The benchmark middle class thus represents the 25% of the population both above and below the median household income and includes exactly half of all households.  This conception of the middle class is a different equality benchmark than some use, but we prefer it to models that define equality as everyone earning about the same – i.e., a more Marxist construct.

The IMF staff’s benchmark case is then juxtaposed with the actual income data from the U.S. Population Survey through 2016 by the Census Bureau and the authors’ own estimates to show the actual position of this 50%-to-150%-of-median-income group over time.  Much of recent inequality literature has focused on the separation between the benchmark, zero-inequality distribution (demonstrated in the charts below from the IMF staff paper by the 45 degree line emanating from the origin) and the actual distribution (demonstrated below by the curve below the 45 degree line).  This separation is referred to as the “Gini coefficient,” and as shown below it has been growing for nearly half a century. 

While using both census data and the Gini coefficient are sometimes contested from a methodological perspective, these data nonetheless show clearly that the middle class is taking home a smaller slice of the income pie:

Even more interesting is that, in 1970, the middle class accounted for 57.6% of households; by 2016 it fell almost ten percentage points to 48.3%.  Furthermore, a greater percentage of households had incomes less than 50% of the median income in 2016 compared to 1970, meaning the trend is not purely attributable to middle-class household movement into the upper-class.  Some of the shrinkage is almost certainly due to this as households with greater than 150% of the median income also comprise a larger percentage of the population in 2016 compared to 1970.  However, the change has not been even: the paper reports that the sub-50%-of-median-income group has grown by about 30% since 1970 while the above-150%-of-median-income group has grown by only about 15%. 

What does it all mean? Fewer Americans today share in the common experience of a middle-class lifestyle that at least since the end of World War II has been an essential part of the “American Dream.”  While in 1970, 83.2% of American households earned at least 50% of the country’s median income, as of 2016 this fell to 77.6%.  Put another way, an additional seven million U.S. households earn less than 50% of the country’s median income today than would be the case if income distribution in 2016 mirrored income distribution in 1970.  In 2016 dollars, that’s an additional seven million households – accounting for approximately 18 million Americans – living on less than $29,000 per year before accounting for tax inflows or government transfer payments that boost income.  It’s also worth noting that the middle-class share fell even as the average number of wage-earners per household has increased as the percentage of women in the workforce grew.

New research from the Federal Reserve Bank of Minneapolis presents one possible explanation as to why the middle class’ income-share, and the middle class itself, is diminishing.  Using data provided by both the Social Security Administration (SSA) and the IRS to examine differences in labor- and non-labor related income, it concludes that income-share gains by the very well-off have been almost entirely driven by increases in non-labor related income, such as increased earnings for pass-thru entities resulting from decreased tax rates.  When looking solely at labor-related income – which is used in the IMF paper and accounts for the vast majority of income earned by middle-class households – the discrepancy between the income-share growth of the well-off and that of the middle class is significantly muted even though wealth inequality grows worse.  Further, new data from the WID.world database that includes all sources of income finds that, since 1980, the richest 10% of the U.S. population has seen its income share continually rise while the share owned by the bottom 50% of the population dropped. 

Data from the three papers noted above all add up to one clear conclusion: income gains for the top of the distribution have, for almost half a century, largely come at the cost of a diminished middle class and a poorer lower class.  Increased post-tax earnings for top-income earners have been largely driven by non-labor related income and have failed to trickle-down to middle-class wages.  Those residing in the middle and lower classes, for which labor constitutes a substantial and sometimes the only source of income, have seen their incomes stagnate and when adjusted for inflation decline while the very wealthiest households have increased their personal incomes and their income-share as a group.  Coupled with challenges to upward mobility – as shown by the limited increase in the number of households above the 150%-of-median-income threshold – it’s no wonder average workers and not just the least well off are viewing capitalism and free-market economics less favorably than just a generation ago.