Thursday, March 28, 2013

Where Do the Rich Get Their Money, Again?

This was an early topic at the Slack Wire, but worth revisiting.

There's this widespread idea that the rich today are no different from us. We no longer have the pseudo-aristocratic rentiers of Fitzgerald or Henry James, but hard-working (if perhaps overcompensated) superstars of the labor market. When a highbrow webzine does an "interview with a rich person," it turns out to be a successful graphic designer earning $140,000 a year.

Sorry, that is not a rich person.

The 1 percent cutoff for household income is around $350,000. The 0.1 percent, around $2 million. The 0.01 percent, around $10 million. Those are rich people, and they're not graphic designers, or even lawyers or bankers. They're owners.

From the IRS Statistics of Income for 2010:



Wages and Salaries Pensions, Social Security, UI Interests, Dividends, Inheritance Business Income Capital Gains Total Capital Income

Total 64.5% 18.5% 6.1% 7.2% 3.8% 17.1%

Median Household 72.7% 21.5% 2.4% 2.4% 0.0% 4.8%

The 0.01% 14.6% 0.7% 23.1% 19.0% 40.6% 82.7%











As we can see, for households at the very top of the distribution, income overwhelmingly comes from property ownership. Total property income at the far right, the sum of preceding three columns. (The numbers don't add to quite 100% because I've left out a few small, hard-to-classify categories like alimony and gambling winnings.) The top 0.01 percent's 15 percent of labor income is not much more than the same stratum got from wages and salaries in 1929. No doubt many of these people spend time at an office of some kind, but the idea of "the working rich" is a myth.

Here's the same breakdown across the income distribution. The X-axis is adjusted gross income.


So across a broad part of the income distribution, wages make up a stable 70-75 percent of income, with public and private social insurance providing most of the rest. Capital income catches up with labor income around $500,000, making the one percent line a good qualitative as well as quantitative cutoff. It's interesting to see how business income peaks in the $1 to $2 million range, the signature of the old middle class or petite bourgeoisie. And at the top, again, capital income is absolutely dominant.

It's an interesting question why this isn't more widely recognized. Mainstream discussions of rising inequality take it for granted that “those at the top were more likely to earn than inherit their riches," with the clear understanding that "earn" means a paycheck. Even very smart Marxists like Gerard Dumenil and Dominique Levy concede that "a large fraction of the income of the wealthiest segments of the population is made of wages," giving a figure of 48.8 percent for the wage share of the top 0.1 percent. Yet the IRS figures show that the wage share for this stratum is not nearly half, but less than a third. What gives?

I think at least some of the confusion is the fault of Piketty and Saez. Their income distribution work is state of the art, they've done as much as anyone to bring the concentration of income at the top into public discussion; I'd be a fool to criticize their work on the substance. They do, however, make a somewhat peculiar choice about presentation. In the headline numbers in much of their work, they give not the top 0.01, 0.1, 1, etc. percent by income, but rather the top percentiles by income excluding capital gains. [*] This is clearly stated in their papers but it is almost never noted, as far as I can tell, by people who cite them.

There are various good reasons, in principle, for distinguishing capital gains from other income. But in an era when capital gains are the largest single source of income at the top, defining top income fractiles  excluding capital gains seriously distorts your picture of the very top. For instance, you may miss people like this guy: In both 2010 and 2011, the majority of Mitt Romney's income took the form of capital gains.

"They have taken untold millions that they never toiled to earn," or if you prefer, "Save your money -- same like yesterday."


[*] The fractiles are defined this way even when capital gains income is reported. You have to dig around a bit in their data to find the composition of income by raw income fractiles, equivalent to my table above.

16 comments:

  1. This is a bit off topics, but I think that also a big part of what is counted as "wages" for big name CEOs should, in theory, be counted as "profits" and not really wages.
    The logic is this:
    - profits is what remains to the owners after the costs of production (wages, materials and fixed capital) are deducted;
    - wages are usually what people are payd for for producing stuff;
    - but a big part of the job of top CEOs is to squeeze wages for other workers, for example relocating jobs in cheaper nations or negotiating advantageous financial deals - not producing stuff in itself;
    - in other words, to a certain degree (that increases as we go up the bureaucratic piramyd) manager are not paid to increase the pie, but to cut a bigger share of the pie and sen it to the owners;
    - therefore I think that an (unspecified) portion of the wages of CEOs and top managers should be considered a share of the profits that owners redirect to managers so that they perform the "work squeezing" job.
    Also, for what I can see in Italy, usually people who get top jobs get those jobs because of connections, so that it really is a sort of "class" thing more than a "career" thing (though many people say that Italy is worse than the rest of the world in this respect).

    ReplyDelete
    Replies
    1. This is all true, and people like Dumenil and Levy (who I think you would like, if you haven't read) spend quite a lot of time making exactly this point. My point though is that it's less necessary to ask how much apparent labor income is really capital income, when you realize how much of the income of the top stratum shows u as capital income to begin with.

      I don't know what the numbers look like for Italy, though. It may well be that there (or elsewhere) the income of capital-owners does mainly take the form of salaries.

      Delete
  2. Ah! But the question then becomes how are "capital gains" generated? What is it about the prevailing system the distributes alleged increases in productive surpluses to rising prices of "financial assets"? Isn't this basically a matter of economic rent extraction, and, if so, what are the specific "mechanisms" that give rise to this particular form of their crystallization?

    ReplyDelete
    Replies
    1. Right, that's the question. Or rather, two questions. First, how the share of profit increases, and second, why at the household level profits take the form of capital gains rather than (as in the past) interests, dividends and business income. With respect to the second of the questions, a big part of the answer is that corporations distribute more of their profits via share buybacks. That in turn may have to do with taxes, but I'm pretty sure there are deeper reasons as well.

      Delete
    2. I think the answer to the second question is quite simple:
      - as the share of profits (SOP) rises, the rate of profits (ROP) also increases initially, because total profits increase;
      - but then, "capitalists" (actually businesses) reinvest those profits in capital goods. More capital goods with fixed wages means capacity underutilisation, hence in the medium term a rise of the SOP causes a fall of the ROP;
      - the price of unelastic capital goods then bubbles up to match the ROP of elastic capital goods;
      - but a low ROP with bubbly asset prices create an enviroment where speculation pays more than long term investiment ("casino' capitalism"), therefore owners try to buy capital assets that they think will bubble up, and the financial industry provides them by share buybacks etc. .

      Delete
  3. Plus, in the top x%, a lot of the nominal income is in the form of bonuses, which are dependent on profits, so that's really business income as well.

    Cheers!
    JzB

    ReplyDelete
  4. Does capital gains as a category include the case of CEO compensation with stock options (which at some point seemed like the major source of CEO income)?
    There is also the second order effect (i.e. moral hazard) that the CEO maximize the value of the stock often in ways that indifferent or inimical to the firms longer term profitability.

    ReplyDelete
    Replies
    1. Exercised stock options should fall under capital gains, yes.

      I think there are a whole range of important consequences when wealth owners are more concerned with maximizing the market value of their assets than in the flow of income from them.

      Delete
  5. Notice that capital bump around $10k? the Minneapolis Fed had a nice paper on it. Although I'm sure without googling you already know or can guess what it is.

    ReplyDelete
    Replies
    1. Part time working retirees and rich risk takers who didn't make a lot this year but havea lot of wealth stored up. They were pointing out that the poverty line misclassifies wealthy with low earned income.

      Delete
    2. oops. I should have said "part-time working rich" — but the focus is on business owners with losses in a particular year but longitudinally generally high incomes // high wealth.

      http://www.minneapolisfed.org/research/qr/qr2121.pdf page 10 "Income poor"

      Two updates sincehte 1997 paper: both before the mainstream press incorporated inequality as a regular topic.

      Quadrini's page has nice stuff as well: http://www.econ.umn.edu/~vr0j/ec8503-12/vicincinq2.html

      Delete
  6. Yaj Reizarb:

    1. Whenever you are talking about income inequality, it's worth considering the persistence of filers at the top, and looking at lifetime earnings trajectories and their composition rather.

    Quite a few people that fall into the top 1% or 0.1% in any given year have been using wage-income to accumulate a position in an asset that they sell off at the end of their working life, and the large gain that they realize gives them a one-year income that's vastly higher than the average income that they realized over the course or their working life.

    I wonder about the extent to which one-off gains are a driver of measured income inequality. To the extent that they are driving increases in measured income inequality - if they are - it's worth asking the extent to which metrics that capture one-off gains but don't recognize them as such are really affecting lived/perceived inequalities in the real world.

    Put another way, the accountant who buys a series of duplexes with his wage income over the course of his working life and sells them when he's 67 years old had to schelp himself to the office every day, fret over his bills, retirement savings, braces, etc, etc, etc, just like his neighbors. He'll get a nice payout after 30 years of tending to them and sinking his savings into them, but he had the same sort of house, same sort of car, sent his kids to the same schools as they did, etc. It's not clear how many filers in the top 1% fall into this category, but they shouldn't be omitted from the analysis entirely.

    The paper below isn't perfect, but it might be a useful starting point for exploring this question.

    Nicholas Turner. "Income Inequality, Mobility, and Turnover at the Top in the US, 1987–2010" American Economic Review (Papers and Proceedings) 103.3 (2013): 1-7.



    ReplyDelete
  7. I wonder about the extent to which one-off gains are a driver of measured income inequality.

    Right, this is a very good question. To the extent this is true, we should give capital gains a lower weight in exercises like this -- but I'm far from convinced that giving it a zero weight, as in Piketty and Saez's headline numbers, is an improvement. And remember, other forms of income, including wages, are lumpy too.

    But yes, one really would like to know how stable the distribution of capital gains income across households is. I was discussing this with Seth Ackerman of the Jacobin the other day, and he took the same position you do, that figures like these exaggerate the importance of capital gains at the top by including people whose average incomes are much lower and are enjoying a onetime windfall from the sale of their main asset. Personally, I'm not convinced. I think a large and increasing share of "normal" capital income takes the form of capital gains -- for one thing, as I mention in the post, stock buybacks are now as important as dividend as a channel for distributing corporate profits, and all of that shows up as capital gains. But we should settle this by data, not intuitions. If I can find some good empirical work on this -- I'll take a look at the article you mention -- I'll post an update.

    ReplyDelete
  8. Money now is predominantly numbers on computer screens. Many people advocate the "gold standard", but gold doesn't actually represent capacity to produce anything. Up till the last 100 years or so, we were not degrading the environment in ways that threatened its production capacity (fisheries, farmland, energy, forest materials, minerals, etc). Money should represent the ability of produce food, materials, energy, clean water, etc. What will happen as the world population rises and more and more try to attain a middleclass lifestyle? Likely greater pressure on those systems. Simultaneously, those with the most decimal places in their bank accounts are going to want to add even more digits.

    Just look at China and India. They are increasing their "wealth" at the expense of natural resources and the environment. Are they actually more wealthy? Can their land produce the same quantity of goods, provide the same amount of clean water? No, not sustainably so.

    I always laugh at those who push "Emerging Markets". they seem to think there will always be an emerging market somewhere. Well, one day there won't be, unless we discover a 3rd world country on Mars perhaps

    ReplyDelete