Warren’s wage claim about the bottom 90 percent is based on comprehensive, yet limited, data
“From 1935 to 1980, we’re coming out of the Great Depression. What do we do? We do two things. We put a cop on the beat on Wall Street, and we started investing in our future together. . . . And we build America’s great middle class: the 90 percent, everybody outside the top 10 percent. The 90 percent get 70 percent of all wage growth in this country. . . . Trickledown economics hits in the 1980s, and you just watch this reverse. So from 1980 to 2012, that’s the latest year for which we have data, the 90 percent, everybody outside the 10 percent, do you know how much they got? They got zero percent of income growth.”
—Sen. Elizabeth Warren (DMass.), remarks in various venues.
Warren has repeated a version of these remarks in a variety of high-profile settings, including on the “Late Show with Stephen Colbert.” Readers have asked whether her frame of reference is correct. Did the bottom 90 percent of income earners in the United States get 70 percent of the wage growth from 1935 to 1980 — and then zero percent of the growth after 1980?
This is a very interesting but highly technical issue. Warren pitches it in a fairly partisan manner — everything turned south after “trickle-down economics” (i.e., President Ronald Reagan) was introduced. She also offers somewhat simplistic reasons (such as the creation of the Securities and Exchange Commission) for the shift in income growth, though her staff says she was not intending to exclude other factors.
Where do these numbers come from?
Warren derived her numbers from data collected by two wellknown economists, Emmanuel Saez of the University of California at Berkeley and Thomas Piketty of the Paris School of Economics. Most economists agree that the two men have assembled a comprehensive data set that helps examine the distribution of market income over time.
Here are the factors captured in the data: wages, net selfemployment income reported on income tax returns, taxable interest and dividends, rental income, and so forth.
But there are limitations. Social Security and unemployment benefits are not counted. Neither are government transfer payments such as food stamps and veterans benefits. The data also does not include noncash items such as employer-provided health insurance. Last, the impact of income and payroll taxes are not part of the calculations, even though the rich pay most of the taxes — and some lower-income workers receive payments such as the earned-income tax credit that offset all of their federal payroll and income taxes.
In other words, the numbers show part of the income picture— essentially the wages. They do not necessarily show everything, be it food stamps or company-provided health insurance, that Americans receive to help them pay for what they consume.
“The income items Piketty and Saez omit tend to equalize incomes of rich and poor,” said Gary Burtless, an economist at the Brookings Institution who specializes in income distribution. “More importantly, an increasing percentage of the gross incomes received by Americans is excluded from the income measure tabulated by Piketty and Saez.”
In essence, Warren is just focusing on the income that is reported on a tax return — not all of the items that help families pay for what they consume.
There is a bit of an irony here. By focusing on wage income, Warren is not counting the impact of programs often celebrated by Democrats as success stories. Indeed, the official poverty rate also does not count the impact of those programs, which is why the Census Bureau in recent years has developed an alternative version that does include it.
But a fuller set of data goes back only to the early 1960s, making the lengthy comparison highlighted by Warren all but impossible. But the income divide highlighted by Warren shrinks when the missing data is part of the equation.
The Congressional Budget Office in 2014 produced a report that focused on the after-tax resource flow available to a family to pay for its consumption. Families in the top 1 percent saw their after-tax income triple from 1970 to 2011, but other groups also saw a sizable improvement in household incomes.
Another method, used by Cornell University professor Richard V. Burkhauser with colleagues, looks at “post-tax, post-transfer, size-adjusted household income including the ex-ante value of in-kind health insurance benefits.” In particular, his approach includes capital gains, which has been criticized by some economists because the volatility of the stock and housing markets means the results can change significantly depending on the base year chosen.
In any case, his research reaches the controversial conclusion that once government taxes and transfers are included and you account for changes in household size, the rich did get richer, but so did everyone else.
“The great advantage of tax record data is that it goes back to 1913,” Burkhauser said. “That is what is so special about the Saez numbers. But they can easily be misused by those who confuse the market income to tax units with the full after-tax income of households. Elizabeth Warren is one of those people.”
Burkhauser noted that the introduction of Social Security in the 1930s meant that many older people stopped working — and thus negatively affected the “market income” of older people.
Saez acknowledged the limitations of his data and said that he and Piketty are trying to improve it. “Both pre-tax and post-tax are interesting precisely to assess how government programs help,” he said. “We are going to produce both in the (hopefully) near future.”
But Saez also said he agreed with the broad thrust of Warren’s statement that the top 10 percent is suddenly doing better.
Lacey Rose, a spokeswoman for Warren, responded that market income “paints a better picture of the plight of middleclass families than looking at an income measure that includes transfer payments.” Rose noted that transfers are illiquid and not fungible —housing vouchers cannot be used to fix a car—and that refundable tax credits such as the earned-income tax credit are erratically timed.
“The low income is a fact; transfer programs are a policy response to that fact,” she said. “They don’t negate the existence of the fact.”
The Pinocchio Test
We dislike leaving readers with an economic muddle. But this is a classic case in which the outcome depends on the data selected. Should you count government transfers or benefits such as health insurance? Do you count families or households? Do you measure before or after taxes?
Each of those choices can change the outcome in significant ways — and reasonable people can argue about the importance of those choices. If one includes health insurance, others might argue that it’s silly to say a person’s wealth includes payments to medical professionals.
Broadly speaking, most economists would agree that the income growth has increasingly benefited the top 10 percent. Warren’s reasons for that — and her specific numbers — are certainly debatable. She would do better to make clear to listeners that she is citing a particular set of data that has certain limitations. After all, there is no certitude in economics.
We can’t award her statement a Geppetto Checkmark. But neither does it quite venture into Pinocchioland. So this statement will receive no rating.