Innovation Files has moved! For ITIF's quick takes, quips, and commentary on the latest in tech policy, go to

A Response to the Critics

I expected that there would be a lot of critical responses to my argument that Emmanuel Saez had “cooked the numbers,” in his study of income inequality,to show that virtually all of income growth during the “recovery” after the Great Recession went to the wealthiest 1 percent. I had a strong feeling that most people would miss my narrowly framed argument and think that I was belittling the negative effects of inequality on our population. Despite attempts to inoculate myself from this criticism by showing the relatively low share of income held by the top one percent in 1979, various commentators have criticized me on several grounds: not discussing wealth inequality; not seeing the long rise in inequality; picking selective years to make my points; helping the right wing; overemphasizing the effect of transfers because of the rise of Social Security and Medicare and failing to appreciate the difficulties of middle class people and exaggerating the effects on the rich.

I have an odd intellectual history in that I was one of the first researchers to report on rising inequality in the late 1970s  and 1980s, yet have for the last 10 years argued that many researchers, in particular Saez and Thomas Piketty and, have overplayed the inequality card by arguing that the middle class received very few gains in their real standard of living after 1980.

I think that it matters to have some sense of how bad inequality is. In multiple venues, I have argued that one cannot understand the politics of the last 20 years if you think that the middle class has been stuck in a rut and only gotten a sliver of economic growth (just 9 percent for the bottom 90 percent between 1979 and 2007). Further, according to Piketty and Saez, all of these gains were wiped out in the Great Recession as the real incomes of the bottom 90 percent in 2011, 2012, and 2013 were 9 percent lower that they were in 1979. By contrast, the Congressional Budget Office finds that middle class families (the third income quintile) had real incomes that were 35 percent higher in 2011 versus 1979 (less than one percent gain per year).

If you believe the Piketty and Saez figures then you are likely to agree with Thomas Frank when he says In What’s the Matter with Kansas:

“The countryseems more like a panorama of madness and delusion…of sturdy blue-collar patriots reciting the Pledge while they strangle their own life chances; of small farmers proudly voting themselves off the land; of devoted family men carefully seeing to it that their children will never be able to afford college or proper health care; of working class guys in Midwestern cities cheering as they deliver up a landslide for a candidate whose policies will end their way of life.”

In other words, you’ll rub your eyes in disbelief at the results of the 2014 elections and continue to lament that electorate votes in ways that are against their own interests.

With this said, let me turn to the specific criticisms of my argument.

Wealth Matters

Ben Walsh in the Huffington Post posits that the headline of the New York Times story (“Inequality Has Actually Not Risen Since the Financial Crisis”) is “flat-out wrong” because wealth inequality rose according to a study of Saez and Gabriel Zucman. He criticizes David Leonhardt and me for not including a discussion of wealth in a short column and report on income inequality.

There are three problems with his argument. First, it misses the point of the study which was to critique Saez’s widely-cited finding on exploding income inequality.  He does not mention Saez’s work at all or the fact that Saez wrote about income inequality without mentioning wealth inequality. Over the last 20 years, the number of scholarly papers on income inequality is at least 20 times greater than the numbers of papers on wealth inequality, and virtually none of these papers on income inequality felt the need to simultaneously present data on wealth inequality.

Second, there are very few sources on wealth inequality and Walsh cites a single source and not the disagreement that exists across sources on what is happening to wealth inequality (see comments on the problems with the wealth data here and here).  Further, the value of assets is uncertain and subject to wide swings. For example, consider the volatility of wealth over the last 20 years—first the explosive growth of tech stocks, then the collapse of the bubble with the NASDAQ composite index falling by nearly 80 percent, followed by the rapid rise in housing prices and the eventual bursting of that bubble. Finally, while most people’s main source of wealth is the equity in their home, rich people’s wealth mainly comes from owning companies, stocks, and bonds. Since the period after 2007 witnessed a devastating decline in home prices, it is certainly unclear whether the current rise in wealth inequality found in several data sets is a temporary blip or part of a longer trend

Third, which would you rather have $50,000 a year for 20 years or a lump sum of $700,000? This is close to the choice that many lottery winners have and reflects the fact that a stream of payments can be “capitalized” into a single value today. By the same logic, the stream of cash payments and medical services that retirees get through Social Security and Medicare can be translated into a single amount that you would need today to buy an annuity adjusted for inflation and the guarantee of medical insurance for the rest of your life. For people who earned average amounts during their career, this value would be many hundreds of thousands of dollars and would approximately triple the net worth of people at retirement.

The notion that this sum should be considered personal wealth is quite reasonable. Consider the following alternative of what life would be like without these government programs. In this alternative universe, there is a government mandate that workers had to put aside 16 percent of their earnings in a separate account to prepare for their retirement years (Chile has a version of this). Since economists believe that it is total compensation that determines employer wage decisions, cash earnings now would be higher by the amount that companies paid to government in payroll taxes. Consequently, in this scenario, people would approach retirement with much higher levels of private wealth and would use this money to buy annuities and long-term health insurance.

Today, wealth inequality is much higher than income inequality. Including the capitalized value of Social Security would reduce wealth inequality substantially and would certainly affect the trends in wealth inequality especially since the cost of Medicare has been rising so much. So for this reason and others discussed above, the vast majority of socio-economic research in the future will continue to focus on the income distribution alone because this shows in an immediate way how different people live.


Ben Walsh and others are also upset with me for using 2007 as the starting point and not seeing the long upward movement of income inequality. Notice that they don’t critique the use of 2009 to 2012 by Saez. Without understanding what I am trying to do, it is of course odd that I would only look at 2007 on. But my purpose in starting with 2007 is to show that Saez is only focusing on the “bounce-back” effect from the huge decline in the incomes of the top one percent from 2007 to 2009. In every piece that I have written on the topic of income inequality, I have talked about the huge rise in inequality from 1980 through the present. I fully understand that a pattern has not yet emerged from the up and down movements in inequality following the onset of the Great Recession. In my piece criticizing Saez, I was careful not say that this was the beginning of a long-term decline in inequality. Consequently, I find this criticism is another out of context attempt to undermine the findings without having to deal with the specifics of my critique of Saez’s claim.

Helping the right wing and not seeing how little the super- rich were affected.

Yes Fox News did have fun with this Times story. But Mitch McConnell and John Boehner started their 60 Minutes interview with an attack on Obama as presiding over an economy that was only helping the rich. The bottom line is that each political party will “spin” data to show that their policies are being vindicated.

The recession obviously hurt lower and middle income people hardest in real life terms. The rich may have lost more percentage-wise, but going from $1.5 million per family to $1.1 million does not really cause pain. The Piketty and Saez insistence that inequality is driven by the share of the top 1 percent is iffy. By and large it follows other inequality indicators, but it doesn’t have to and it is much, much more volatile than other measures.

My purpose was not to say that inequality was decreasing but to say that the statement that the rich got virtually all of the income growth in the past few years was deceptive on two grounds. First, the methodology of using shares of growth focused on changes in income levels in two periods is inappropriate. If you use short time periods, this change can be very small; and if the change is small, a small gain at the top can look like inequality is exploding over these years. The second problem is that the years were carefully selected to make the point. This became very clear when the 2013 data became available: rather than including these data as part of the recovery period, Saez kept his focus solely on 2009 to 2012 because the 2013 data showed a steep decline in the incomes of the top one percent (due to the raising of the marginal tax rates for rich people, which led them to take capital gains and some business and bonus income in 2012 rather than 2013).

Are Increased Transfers All Due to Social Security and Medicare?

There is no question that the value of Social Security and Medicare has grown substantially over time. For elderly households in the third income quintile in 2011, their market incomes averaged $20,800 while their transfers accounted for $35,300 (which was up for $16,100 in 1979). Because of this rise in transfer incomes the share of elderly households in the third quintile rose from 16 to 24 percent over these years. Therefore, it is tempting to think that elderly transfers are the main driver of overall transfer income.

But, the CBO has data on two other types of households: the non-elderly without children and households with children (which are almost entirely composed of non-elderly adults). For both types of households, added transfer income offset market income losses substantially for the bottom sixty percent of the income ladder: for the non-elderly childless, income losses for the bottom 60 percent were trimmed by about 5.5 percent, turning an 8.5 percent income loss into a 3 percent decline. Transfers to households with children were more concentrated among households with the lowest incomes: for those in the lowest income quintile, an 8 percent income loss was changed into a 2 percent gain once transfers were included; in the second income quintile, the effect was also quite large as a 12 percent income loss became a post-transfer loss of 3 percent.  The targeting of added transfer benefits to families with children was evident in that there were only tiny declines in income losses for the wealthiest 40 percent of households while the third income quintile had a modest offsetting effect with market incomes declining by 8.6 percent and post-transfer incomes declining by 4.2 percent.


If I had thought harder, I might have been able to anticipate the criticisms discussed above. But in the end, no one really found fault with my analyses. They just changed the topic and implied that I didn’t appreciate the negative effect that the high level of inequality had on our country.  This is too bad, since all policy debates benefit from more, not fewer facts and analysis.


Print Friendly, PDF & Email