Capital Matters

The U.S. Middle Class’s Exceptionally Long Decline — and Recent Recovery

Median household income has been on the rise since Trump's election. Will it outlast his presidency?

A single day often appears, in the fullness of time, to represent a moment of transition between two eras. One such day appears to be October 10, 2000, when President Clinton signed the legislation that permanently normalized U.S. trade relations with China. Another appears to be November 8, 2016, when the American people elected a president who promised to shred this once-bipartisan consensus on economic engagement with China on their behalf.

In the time between these two inflection points, trade with China would join events such as 9/11 in reshaping how ordinary Americans experience globalization, and the growing headwinds facing America’s middle class would both blow in from abroad and swirl at home. No one can yet know what dates, if any, will come to mark passages between two eras under the Biden administration. But for now, all of us, including the incoming administration, would do well to take stock of a single fact that summarizes in no small part the story of the once-brave new world that began in 2000 and ended in 2016.

That fact: The ordinary (or market-generated) income of the median American household failed to regain its 2000 peak until 2016. In every single annual observation from 2001 through 2015, adjusted for inflation, as the chart above shows, the household at the center of America’s income distribution earned less money than the household at the center did in 2000. Economists and pundits continue to debate whether America’s middle class started to fail itself, or if something started failing the middle class. Regardless of its ultimate interpretation, however, a fact remains a fact.

Specifically, the chart shows the median household income’s deviation from what it was in 2000, in percentage terms. For instance, in 2012, the median U.S. household income of $56,900 was 9.0 percent below the $62,500 peak it had reached in 2000 (both in 2019 dollars). Only in 2016 did Americans’ median household income surpass its 2000 level, reaching $62,900 and putting the measure in the positive territory from which it has never since registered a fall from. By way of comparison, however, over the same period, Canadian median household income managed to rise steadily. If global factors such as automation (which also affected Canada and other peer economies) had been the reason behind the median household’s plunge in the U.S., this divergence ought not to have occurred.

Some who assess trends in the median household’s income prefer to include government transfers (things such as cash welfare or Social Security payments as well as food stamps). When you do that, the decline in median household incomes between 2000 and 2016 does indeed disappear. But does it make more sense to look at ordinary (market) incomes or at incomes net of government transfers?

It depends. If you care about the purchasing power of the household as an individual and discrete unit of society, at a single point in time — as many economists do — any distinction between earned income and transfer income is necessarily irrelevant. If you care about the median household income for its use as a broad barometer of the well-being of American society over time and as a whole, however, the case for looking at market incomes becomes compelling.

The need for the government to borrow money to fund the transfer payments that plug the hole in household market incomes leaves America today unambiguously poorer, and more debt-saddled, than it would be if market-income growth had produced an equivalent increase. Issuing government debt to fund transfer payments redistributes purchasing power from tomorrow’s America to today’s. Transfer payments also have indirect economic and social costs, such as disincentives to work. A society where the median household’s income relies on government transfers to grow is, in ways large and small, unlikely to resemble one where the growth is coming from earned income. However significant these concerns ought to be to society as a whole, they are obscured by a focus on trends in the net-of-transfer figure rather than on the evolution of market incomes of households. Given this article’s focus on trends in society over time rather than household finances at a given moment, the chart’s focus on market rather than net-of-transfer income seems appropriate.

No matter what, the era of middle-class prosperity that feels like it could be end-dated to October 10, 2000, is never coming back. No policy can fully replicate the economic environment that prevailed at home and abroad before then. The question now is whether the incipient recovery of the median household income that lines up with the era that began on November 8, 2016, will continue. On this, the recovery of America’s median household incomes, regardless of political persuasion, we should all hope for continuity between the outgoing and incoming administrations.

Some former colleagues from my time at the White House Council of Economic Advisers (which spanned 2017 through 2019) estimate that a Biden agenda would cost the median household as much as $6,500. I hope they’re wrong. If not, however, America may end up rich in, if nothing else, irony: The era that began on November 8, 2016, may end with a return to the middle-class stagnation that created it. Who says history’s turning points can’t form a circle?

Joseph W. Sullivan served at the White House Council of Economic Advisers as the special adviser to the chairman, as well as a staff economist, from 2017 to 2019.


The Latest