“Rich Get Richer” Theories
Which of Piketty’s “Rich Get Richer” Theories Matters More?
– by Steve Roth
For to every one who has will more be given, and he will have abundance; but from him who has not, even what he has will be taken away. —Matthew 25:29, Revised Standard Version
Steve Randy Waldman reminded me recently of this great 2014 Matt Bruenig post from Demos. (The post is missing on the Demos site; the link here is from archive.org.) It unpacks some fairly abstruse reasoning in Piketty’s Capital in short, clear, simple form.
Piketty actually has two theories for The Matthew Effect: why the rich get richer. Since Piketty uses “capital” and “wealth” as synonyms,1 I use wealth here (or more precisely, assets) as the key term. The two theories:
- Wealth grows faster than GPD/GDI. Since wealth holdings deliver “property” income to the wealth holders, over time the share of share of national income captured by asset-holders in return for owning stuff increases, vs. the income captured by workers in earned labor compensation, for working.
- Wealth (assets) concentrate into fewer hands. So, the unearned property income received by asset holders is more concentrated as well, in a self-perpetuating cycle where more-concentrated wealth delivers more-concentrated (property) income and vice versa.
One or both of these is very clearly at play in the U.S., and over a very long period:
The red line is the standard-issue measure; the denominator (household income) is the NIPAs’ radically balance-sheet-incomplete measure of household income.² The blue line divides labor income by households’ total income including total return on assets, which necessarily includes the “capital” or “holding” gains accrued and accumulated by people, households, families, and dynasties over years, decades, generations, and…dynasties.
But which of the two theories most powerfully explains the big labor-share decline, and hence the runup in the property-income share? For the moment I’ll just share two quick graphs. (This topic merits deeper drill-down; if readers are interested I’ll explore more.)
First, the wealth:income ratio: Piketty’s β(eta).³
Next, a significant representative measure of wealth concentration.4
Eyeball-analyzing, wealth concentration (theory #2) appears to be a much stronger driver of The Matthew Effect than the growth of national wealth relative to national income.
As always, comments, suggestions, and especially critiques are very welcome from my gentle readers.
1 “I use the words ‘capital’ and ‘wealth’ interchangeably.” Capital, p. 47 Full kudos to him for stating this assumption explicitly; it’s baked into many economic theories and their implementations — notably Solow-style growth models — but it’s ~universally left unstated, silent, and hidden.
2 For comparison, The BLS “labor share of nonfarm business output” measure averages 59% over the period. The Penn World Tables’ U.S. labor share of GDP averages 60%.
3 This graph and the labor share graph at top are made possible by the Total Household Income Accounts, or THIAs.
4 This graph is adapted from Saez and Zucman’s gabriel-zucman.eu/files/SaezZucman2020JEPData.xlsx (thanks to Gabriel for his help). It’s updated through 2023 using data from their realtimeinequality.org site.
One way to explain the power law distribution is to look at it from Ben Thompson’s point of view. He sees it as being about aggregation. If there is anything to be gained by placing oneself in a gateway position, such as being an employer or middleman, you will accumulate more surplus than the parties you are aggregating. As one might expect, aggregating aggregators produces even higher returns. This gibes nicely with capitalist theory in that the rewards flow to whomever can most effectively organize information.
However, simulations show that aggregation isn’t necessary. All you need is people interacting and a random number generator to determine who gains the most from each transaction. Power law distributions flow mathematically from interaction and accumulation. In the real world, transactions are not evenly biased. Everyone has to spend a certain amount just to stay alive, but as one accumulates this is a smaller and smaller portion of what one has accumulated. With even a slight bias like this, it is no surprise that the power law distribution gets increasingly steeper.
In other words. You don’t need a lot of economic theory to explain this. There’s no need for moralizing about hard work, saving, intelligence or the like. It’s just that there are advantages to simple possession. There’s a reason casinos have a house limit. A gambler with the resources to keep rolling double or nothing will eventually win everything. If there are two gamblers and one has greater resources than the other, that gambler will eventually win even with a fair coin.
In the economic realm, the only way to counter this increasing steepness is to limit possessions. This can be done with estate taxes that seize possessions on death of their owner. This can be done by progressively taking an increasing portion of gains in possession. This can be done, as it was in the Bronze Age, by seizing possessions at regular intervals or at the accession of a new ruler. We know that this works. It’s just politically impossible nowadays.
@Kaleberg,
Kudos!
>there are advantages to simple possession.
Exactly right. It doesn’t even require interaction/trade. Even in the simplest Epstein/Axtel Sugarscape simulation (agents wander a grid harvesting sugar that emerges, and die if they don’t harvest enough), it takes very few generations for wealth gini to skyrockets. https://www.asymptosis.com/how-perfect-markets-concentrate-wealth-and-strangle-growth-and-prosperity.html
Born out by 5000 years of recorded human history. Known instances of wealth dispersal are rare. Examples include post-Black Plague, US 1930s–1970s, other 20th C examples.
JW2A: Lest we not forget the Golden Rule – “The guys with the most gold rule.”
The key here is leverage and tax breaks. The rich have assets to leverage. They can borrow money at low interest rates. Though today’s interest rates may not seem that low, but after tax rates may well be below inflation. So as long as the return on the overall investment exceeds inflation, you come out well ahead. The more leverage, the better the return on equity.
But you have to have assets to leverage to start with. Most of us don’t start out with any assets and accumulate them only over time. A large portion of most Americans’ retirement savings consists of the house they live in. It is a perfect example of the combination of high leverage (sometime approaching 100% mortgage on appraised values), and tax deductibility of interest payments.
Mathew is right. “For to every one who has [significant assets to start with,] will more be given, and he will have abundance [as long as he takes advantage of leverage and tax breaks;] but from him who has not, even what he has will be taken away.
This is why there is constant pressure on the Fed to lower interest rates by investors who want to leverage more at lower rates. After the Great Recession when the Fed introduced interest rates near zero, bank profits recovered within 3 -4 years as did stock prices. But unemployment rates stayed relatively high for most of a decade and real wages stagnant for almost as long, despite the Fed, the media and politicians incessant hectoring us about how they had lowered interest rates to benefit workers!
>The key here is leverage and tax breaks.
I’ll disagree that leverage is “they key” even though it has a very powerful effect. An asset-holder gets unearned income for simply owning assets (Kaleberg: “posession”) even it the liability side of their balance sheet is zero.
I will 100% agree that (extremely) progressive taxation especially of unearned property income is always and everywhere a necessary corrective to reduce wealth concentration, coupled with downward wealth redistribution. The redistributive mechanisms take many different forms, from public goods provision (eg. healthcare) to explicit xfers (universal income FTW!) to…
Kind of an aside but I should add that IMO universal income should be quite progessively “clawed back” from higher-income/wealth households on their tax returns. So as of April 15th, sure, it’s not really “universal” anymore. But it’s means-testing the rich (which we already do, on their tax returns), not the poor.
Steve:
I have set money aside to cover such occurrences where my position was eliminated, transfer or the company moved. State could sponsor such funding in addition to their own. Small interest rate paid on it . . .
Leverage is, it seems, a special case of possession, allowing accelerated accumulation of assets.
There is also a special case of leverage which obviates the need for possession in the first round: work in leveraged finance. One can benefit from leverage using other peoples’ assets in the initial round. Many of the world’s rich got started this way, more or less.