05/22/2014 05:00 pm ET Updated Jul 22, 2014

Arguing Taxes, Transfers, and Market Outcomes

LEON NEAL via Getty Images

Inequality analyst Scott Winship feels "systematically ignored" by "writers and researchers" including myself. Since he's a smart and interesting critic in this space, not to mention an old fellow traveler, allow me to correct that.

Scott is one of a number of analysts who criticize Thomas Piketty's work for focusing largely on market incomes, thus leaving out the impact of taxes and transfer payments. Include those sources, these critics assert, and both real income and inequality trends appear considerably more favorable than Piketty's data show.

My analysis, with which Scott takes issue, is that this "transfer defense" as I call it, has merit, though less in terms of inequality (more on that below) and more in terms of real income growth for the middle class. But my key point is that while analysts should look at all income sources, the market story Piketty tells remains true. To the extent that middle-class families got ahead, it was through transfers and tax cuts more so than earnings. As I read Scott's own data, they further corroborate this insight.

Scott complains that my analysis failed to separate out elderly and non-elderly households. In fact, I fully agree with the differences he notes, and noted them myself, stressing the role of Social Security and Medicare in particular on "aging boomer households." I devoted a precious paragraph (Scott gets more space than I do!) to the fact that "for the average elderly household that receives retirement benefits from the program, Social Security provides just under two-thirds of income."

No question, the figure I posted looks different for different family types, with transfers obviously playing a lesser role -- though still a larger one than many recognize, as I show below -- for non-elderly households. And no question, as I acknowledge, the changing composition (more of those "aging boomers") means more transfers in the middle class.

That's exactly my point. The broad "middle class" -- elderly, non-elderly, with kids, without -- faces all the economic and fiscal pressures we're currently debating, particularly in the age of Piketty, including stagnant earnings, attacks on social insurance, inequality channeling growth away from the middle, and the increasing dependence, particularly but not exclusively among the elderly, on tax cuts and increased transfers to lift wage-starved incomes (in fact, boomers are not just aging, they're also working longer).

Scott's own data neither assuage these concerns nor contradict these points. He suggests that if we look at non-elderly households we'll see the "strong gains in earnings" that I missed. Except that his table shows real wages and salaries of this family type grew all of 5 percent, from 1979-2010. That's 0.2 percent per year and it is not "strong" growth. Add in the value of employer-provided health care (which I argue these CBO data may overestimate) and other employer benefits and you're up to 0.4 percent per year. Drill down further to working-age families with kids and you get 0.4 percent per year in real wages/salaries; 0.7 percent including health care, etc. (14 percent and 23 percent over 31 years).

In what world is that strong earnings growth? The economy's productivity was up 89 percent over those years, or 2.1 percent per year, ten times the Scott's allegedly strong gains in wages for middle-class, working-age households. Moreover, as EPI's Josh Bivens shows in the income chapter of State of Working America, the earnings gains these families achieved derived not from higher hourly wages or benefits, but from more family members working more hours per week and more weeks per year.

Sticking with non-elderly households, Scott shows that wage gains explained 13 percent of the after-tax income growth they achieved over these years, while transfers and taxes accounted for 54 percent. Of course that's less than the 91 percent that I get by looking across the broad middle class, but the fact remains that more than half of the income gains from non-elderly households derived from higher transfers and lower taxes. Only 13 percent came from earnings.

The CBO data show that these working-age families relied on wages for 70-80 percent of their income in these years, yet wages provided just 13 percent of the overall gains, even accounting for the extra hours and weeks worked. In fact, for every family type he examines, taxes and transfers explain a larger share of their gains over these years than do earnings (e.g., for families with children, real earnings growth explains 30 percent of the gains; taxes and transfers explained 35 percent). Again, how does this support Scott's contention that market-based outcomes are working out fine for the middle class?

I fact, I suspect such numbers will lead objective readers towards the same conclusion of my Upshot piece: "... the extent of wage stagnation and its corollary, the increased role of transfer income and tax cuts in raising middle-income living standards, [is alarming]. Instead of hacking away at the safety net, the data reveal the need to preserve it while increasing the quantity and quality of employment opportunities and the real growth rate of earnings for the majority of the work force."

Final point. I'm not sure if Scott holds this position, but many who ply "the transfer defense" seem to believe that if Piketty had included taxes and transfers his conclusions regarding increased inequality would be altered. As I wrote, "...while the progressive taxes and transfers that don't show up in Mr. Piketty's data reduce the level of inequality at any point in time, they don't have that much impact on its growth. The share of comprehensive income going to the top 1 percent grew 6 percentage points before taxes and transfers from 1979 to 2010, and 5.4 points after taxes and transfers. (If one stops at 2007, before the recession, the same comparison yields an increase of 9.8 points before tax and 9.3 after)."

I'd invite Scott to weigh in on that point. I definitely agree with him that it's important to include these income sources in inequality and living standards analyses, though for reasons I stress in my piece, I place a lot of weight on Piketty-style market data too--as market inequality rises, you simply cannot ignore market outcomes unless you want to ratchet up the redistribution function every year to offset it. I also find it awfully dissonant to hear some conservatives (not Scott, I don't think) proffer the transfer defense on Monday, only to argue that we can't afford social insurance on Tuesday.

This post originally appeared at Jared Bernstein's On The Economy blog.