“Our Broken Economy, in One Simple Chart” (nytimes) shows that the highest-income Americans are also the ones with the “largest income growth.” I wonder if Thomas Piketty and friends followed individuals over time or just looked at brackets of income that contained different people from year to year. If they followed individuals then these data suggest that the most energetic and motivated Americans are the richest. Instead of slacking off and enjoying their yachts, Gulfstreams, and 7 luxury homes worldwide, they are doing something that gives them a 6% pay raise every year.
If they didn’t follow individuals, though, aren’t the stated conclusions wrong? Suppose that the income of rich people is highly variable, e.g., inflated one year due to selling a company and comparatively depressed the next year. In that case, from simple volatility and economic growth you might see that the high end of the income scale was doing well (how else did those people get to the high end of the scale?) but it wouldn’t be the same people year after year. Similarly, for those with low income, someone who goes onto a diet of SSDI and OxyContin might have the same income as last year’s consumer of SSDI/OxyContin.
[Separately, the chart note suggests that it includes “transfers and non-cash benefits” for the poor, but I wonder how that is possible. Ever since the Clinton-era “welfare reform,” simple cash transfers have been a small component of modern-day welfare in the U.S. The non-cash stuff is tough to track and value. When a new apartment building is constructed, for example, the developer may be required to hand over 10 percent of the units to a government housing ministry for distribution to the poor. The value of these units are not on the government’s books. And if real estate prices go up, does the person who lives in a free apartment in Manhattan experience a boost in income according to Piketty and friends? Collecting child support and alimony is a big part of the U.S. economy and there are no convenient authoritative sources for the total cashflow (generally from higher-income defendants to plaintiffs with lower wage income).]
The article came to me from a hedge fund manager friend:
First, there has always been a distribution of income and we know it is skewed right. It pretty much has to be, if income is bounded by 0 but not limited on the upside. Allowing for negative income tax rates (the earned income credit) creates some weird growth rates. Overall, growth rates are compressed for people receiving income-based transfers when those transfers are progressive.
There is a tautological aspect to the results. People whose income rose the fastest (e.g. Steve Jobs) ended up in the 1% or the 0.1% because of that growth. So it shouldn’t be surprising that people who now have the highest incomes also had faster rates of growth in that income. That is how they got into the right tail. Few people complain when Steve Jobs or Mark Zuckerberg leapfrogs into the high income brackets but it really bothers economists like Thomas Piketty that the averages behave different from the averages of other income brackets. Remember, we are measuring how fast the average moved, not how fast the income of a third-generation trust baby’s income rose over that period. Their income may have determined the average in 1980 but they are not part of the tiny super-high income brackets any more. Those brackets are reserved for people who grabbed the brass ring and held on for a meteoric rise in pay.
Related:
Few people complain when Steve Jobs or Mark Zuckerberg leapfrogs into the high income brackets but it really bothers economists like Thomas Piketty that the averages behave different from the averages of other income brackets. Remember, we are measuring how fast the average moved, not how fast the income of a third-generation trust baby’s income rose over that period. Their income may have determined the average in 1980 but they are not part of the tiny super-high income brackets any more. Those brackets are reserved for people who grabbed the brass ring and held on for a meteoric rise in pay.
This doesn’t appear to explain how the US economy has changed over the past few decades.
Vince: Nobody said that hedge fund managers were always right! In fact, about half of them underperform the S&P 500.
Aren’t incomes at the top mostly driven by highly paid employees – CEOs and the financial sector – rather than founders? CEOs have a lot of influence on their own pay; norms regarding self-restraint seem to have weakened. (I think in another context you described this as “looting of public companies.”)
Economic Policy Institute:
Discussions of inequality at the top always make me think of the Breaking Bad scene where Walter and Skyler are contemplating the giant pile of cash that they can’t spend. How much money can one person spend?
Agreed, Russil. But why has the looting at the top purportedly become dramatically more effective each year in recent years? Profit growth for big companies has been strong and the people at the top are scooping up most of it?
There’s a couple of other interesting points that can be made. This discussion of Steve Jobs and Mark Zuckerberg on one hand and SSDI, EITC and subsidized housing on the other hand ignores a large majority of the population in the middle. The graphs in the NYT article show that people exactly in the middle enjoyed a 2% increase in their income in 1980, but their raise was only half as much in 2014. It’s unclear why both Philip and the hedge fund manager would choose not to comment on that fact. There may be something interesting going on there.
While we’re on the topic of welfare benefits, Silicon Valley billionaires and hedge fund managers, we should remember that people like Jobs and Zuckerberg accumulated their fortunes working in an industry that was built on corporate welfare from its earliest days. Hedge fund managers work in an industry that requires federal bailouts every 7 – 10 years. That may be one big difference between 1980 and 2014. Before the bailout of Continental Illinois of 1984, bank rescues were fairly uncommon. I don’t recall every reading the term “too big to fail” before the 1990s. It’s been around 9 years since the last Wall Street bailout, so we should expect that Wall Street will looking for another one in the near future.
Vince: https://fred.stlouisfed.org/graph/?g=eKCl shows GDP growth from 1957 to the present. It looks to me as though average GDP growth rates were higher in the old days so you might expect the average person’s income to grow at a faster rate, no?
Also, the education level of the average American was going up circa 1980. Today I don’t think you can argue that a young American is better educated than a middle-aged American, unless you’re going to look at knowledge about transgender issues, microaggressions, speaking truth to power, etc. Why would there be a lot of income growth for workers that aren’t better-educated than last year’s workers?
There is a lot more immigration today than there was in 1980. See http://www.migrationpolicy.org/programs/data-hub/charts/immigrant-population-over-time ; the availability of immigrants to employers should limit wage increases available to native-born Americans.
Finally, elite Americans aren’t too exposed to foreign competition whereas average Americans are.
“But why has the looting at the top purportedly become dramatically more effective each year in recent years?”
Two words: stock-based compensation. Stock prices are based on expectations, and CEOs have a great deal of influence over investor expectations. Roger Martin discusses this at length in his book Fixing the Game (2011).
philg 4, you do not describe GE
Anonymous: http://philip.greenspun.com/blog/2017/05/14/stocks-for-the-long-run-ge-since-2001/ certainly shows that GE is not an example of profit growth. But of course they are a great example of management looting. I’m not sure that the looters at the top of GE today are doing that much better than the looters at the top were doing 10 years ago, though.
Comment number 6 has quite a few interesting mistakes. The comparison under discussion is between 1980 and 2014. The year 1957 is not under discussion. That Fred graph indicates that growth was higher in 2014 than 1980.
The workforce is actually a bit more educated in 2014. Around a third have bachelor’s degrees today, as opposed to a quarter in 1980. Of course, that means that college education was irrelevant to the incomes of a large majority of the people in the middle of income distribution in 1980. So it’s unclear what you’re getting at with remarks about education. People with more education tend to have higher incomes, but the graph of 1980 shows that those with lower incomes, thus generally less education, were getting bigger raises in 1980 than those with higher incomes. It’s possible that you referring to the phenomenon of people going to college at night and getting degrees in their 30s or 40s. Maybe you think that that was more common a few decades ago. It would be interesting to evidence of that if it exists.
Young workers make up a pretty small portion of the total workforce, so whether or not they’re more educated than they’re parents, even though they are, would have to play a pretty small role in the statistics regarding one year, 2014. I’m not sure why you mention microaggressions and so forth. Those issues are completely irrelevant. I’m sure that there were other issues back in 1980 that young people were interested in and old fogies were complaining about.
There’s some research that immigration does bring down wages for some workers. There’s plenty of research that shows that most workers aren’t affected.
However, the sentence in comment #6 does make sense. Corporate America, in its neverending quest for higher profits, has pitted many American workers against those in the Third World, driving down wages for many. The rich and powerful have not been exposed to much of that same competition.
Philip: “I’m not sure that the looters at the top of GE today are doing that much better than the looters at the top were doing 10 years ago, though.”
The graph compares two 34-year periods, 1946-1980 and 1980-2014. The top end did far better in 1980-2014 relative to everyone else – they took much more of the economic pie – than in 1946-1980.
Roger Martin traces the widespread use of stock-based compensation to a 1976 paper by Jensen and Meckling. Their proposal was intended to align the incentives of CEOs with those of shareholders. Unfortunately it turns out to have mostly been a way for CEOs to take home ever-larger paychecks.
Russil: Understood that the chart is for 34-year periods (not just two, right? There is an animated one that shows everything ending from 1980-2014, right?). I was trying to say that for the 34-year growth rate to accelerate over time the looters at a company such as GE should be doing better today compared to 10 years ago even. But now that you mention it… if there was kind of a step function change starting 30-40 years ago that could explain it the data even with flat looting from 2007-2017.
Vince: I recognize that more people have a bachelor’s today than in 1980. But they’re not better educated! That’s why the bachelor’s is the new high school diploma. On the other hand in the decades leading up to 1980 there was a big improvement in American education. That’s why I would expect to see typical workers getting a lot of income growth.
http://philip.greenspun.com/book-reviews/academically-adrift
provides some data on how college students in the 1960s worked more and learned more compared to college students today.
Philip: a step function change is a good way of looking at it.
The data presented is really only for two 34-year periods. If you look at any particular point on the x-axis, you’ve got one set of 34 values (1946-1980), multiplied together and then averaged; and a second set of 34 values (1980-2014). As the animation proceeds, it’s just removing one value from the first set and adding a value from the second set.
US workers may not be better educated than in 1980, but they have much more capital to work with. So average US labor productivity is considerably higher than it was in 1980: the increase from 1973 to 2013 was 74%. It’s just that most of the gains have gone to the top.
Discussions of inequality often mention the decline of unions as a factor. The bargaining power of workers in the middle has weakened, at the same time that the bargaining power of the highest-paid employees (particularly CEOs and the finance sector, but also other knowledge workers like you and me) has increased. Capital versus talent: the battle that’s reshaping business.