"Inequality" is a popular political issue, but it's a paradoxical one. There are many kinds that are debatable—most countries have substantial athletic inequality, beauty inequality, music appreciation inequality, etc., and we mostly don't consider that a problem. To the extent that we do, we care about raising the floor rather than lowering the ceiling: public health campaigns focus on things like vaccination and hand-washing, not snipping athletes' ACLs so they hobble as fast as we walk. Power inequality is hard to measure, but it certainly exists, and one form it takes is that politicians can direct people to do things they wouldn't otherwise do (this is pretty much their job, and it's an important one). So when a politician wants to expand the scope of the government as part of a plan to combat inequality, they're also saying that we should aggravate power inequality in order to reduce wealth inequality.

(There is a reason I'm not a professional speechwriter.)

But even that kind of inequality starts to get sort of slippery. If you compare the top to bottom quintiles, the rich side has about 18.3x the wealth of the poorest. They have about 14.4x the income. But they only have 4.3x the consumption. In other words, you can eliminate about three quarters of inequality if you switch from measuring how much people could theoretically consume in the future to measuring how much they consume right now. 4.3x is still a big deal; a household spending $50k a year lives very differently than one spending $215k. But that gap illustrates several reasons that wealth inequality is not quite what it looks like.

For one thing, wealth is the cumulative total of savings (plus transfers) and capital appreciation. A natural financial lifecycle for a household is to start out with roughly zero, then go negative to spend money for education, claw their way back a bit, then go into debt to buy a house.¹ So measured income inequality is partly a measure of growing lifespans and a partially-privatized retirement system. And, if you're tracking wealth in accounting terms, you're also missing intangibles. Some of these are hard to live on—you're not going to retire on your pleasant memories of Ibiza—but it's fair to say that an annuity is a financial asset with a readily-determined value, and one of those, Social Security, represents about half of the wealth of the bottom 90% of the wealth distribution. And it's gone up a lot over time, partly because there are more earners, and partly because its value is a function of interest rates and lifespans. When rates decline, the present value of a fixed-income asset goes up. This is very easy to see when market multiples expand as rates decline, making the paper wealth of the richest much higher. But someone with a government-guaranteed annuity probably has more duration exposure than the typical equity-owning rich person. So, any assessment of changes in wealth inequality that excludes Social Security and starts any time before 2007 is systematically excluding the biggest source of gains for the majority of Americans.

And this highlights yet another point. The rich have more volatile net worths. Social Security has cost-of-living adjustments, so in 2022 the nominal value had to be adjusted up in response to inflation. But equities do not have a direct means-testing component; when inflation spikes and rates go up, valuations tend to come down. And individual equities are much more volatile than indices; plenty of people have gone from multibillionaire status to mere centimillionairehood and at least one person has had great luck recently in the other direction: going from mere multibillionaire status to centibillionairehood in just a handful of years

Taxes and transfer payments complicate things further. They chip away a bit at inequality (the top 1% get about 16% of pretax income and 13% of after-tax income), but not in a way that materially changes the picture. What really matters is that people vary in how motivated they are by money compared to other things. They vary in their baseline consumption needs, too; some 22-year-olds really want to be investment bankers, others really want to be poets, and many of the would-be poets are willing to forgo a banker lifestyle to pursue their passion. For other people, it's fuzzier, but it's not unheard of for someone to turn down a promotion that means working longer hours, or to stay in a city they like even if they could move somewhere else and get a raise. If that's true, and the government also uses economic redistribution to flatten consumption inequality, then higher income inequality is partly a measure of how well it does this. The easiest thought experiment here is to ask whether or not you'd quit your job (either to retire or pursue a passion project) if you were given an annuity that provided you with the same real income. Certainly, many people would. But not everyone! And that not-everyone cohort would now be providing relatively scarce labor in an economy whose consumption demand was undiminished. So, every time public benefits hit a new threshold, some people at that level see their income drop (it's quitting time! Or at least finally-write-that-novel time) and others see it grow ("Bob has decided to retire in order to focus on oil painting; you up for some overtime?").

If you ran the world's most successful social democracy, where you combined generous transfer payments with a flexible labor market and financial system, you'd expect extreme income inequality: you'd have a smaller labor force, because people wouldn't work so much; you'd also have lower wages for jobs that pay mostly in prestige—many middle-class jobs would be easier to get, and pay pretty well, but art, sports, and academia would all be much more brutally competitive, since so many people opt out of those fields for financial reasons.² Meanwhile, the people who are really money-motivated would have plenty of end consumers for their output, and would thus get a lot richer.

Since consumption inequality is much lower than wealth inequality, one of the important questions to ask is: in what ways does wealth matter? If we gave everyone in the US a bank account with $1m, but with the limitation that they could never withdraw money from it, borrow against it, or sell it to anybody else, we'd theoretically reduce wealth inequality while having zero real-world effect. And if a rich person has a sufficiently low marginal propensity to consume, much of their wealth is, from a consumption standpoint, equivalent to that. Someone else's wealth makes you poorer if they compete with you for access to scarce goods and services—the existence of rich people makes you poorer if one rich person with three full-time nannies is the reason you can't find a babysitter one night a month. But if that rich person gets a kick out of driving a used car and shopping at Target, it doesn't affect you if they're a millionaire or a billionaire.

Or, at least, it affects you in an indirect way. Having wealth in the form of equity tends to correlate with having control, and it's nonlinear; someone who owns 25% of a company has a lot more than 25 times as much influence as someone who owns 1%. If the company is public, and doesn't have other big shareholders, 25% is de facto control unless they really mismanage things. So high wealth inequality probably means that more of the marginal income in a given country goes to investment rather than consumption. That investment eventually yields consumption—assets only produce a return if there's someone to buy their output—but the cost of getting more of it is getting it later. One could imagine a world where the assets of the wealthy compound indefinitely and they end up owning everything, but they don't compound indefinitely: one symptom of actually-excessive wealth inequality is that investments' returns decline, because economic output is being invested and not consumed, so the pool of returns-producing consumption shrinks while there's more money chasing investments. (This also means that the wealth is much more volatile when inequality is high: fixed-income assets are more rates-sensitive, and high-multiple stocks tend to react more strongly to incremental news flow.)

If your wealth hasn't reached escape velocity by the time this happens, it's not exactly fun sitting around in a low-growth, low-return environment where it takes forever to save for a down payment, waiting for things to mean-revert. But low rates also mean that investments that pay out further in the future are more worthwhile today, and encourages R&D to find new useful products. But a maximum-redistribution economy enjoys its own kind of stasis. The problem there isn't that rich people will simply go on strike, it's that higher taxes slow the pace at which money flows from bad companies to good ones, and that it also blunts the difference between bad managers and good ones. The case against inequality is pessimistic at both ends: high wealth inequality is a problem if disruptive technologies never arrive, and if the endpoint of the economy is a handful of people cashing in the treasury bonds they bought for a 1% return and rolling over the proceeds into bonds that yield 0.5%. In the maximum equality scenario, you have to hope that the mix of assets and companies we have is the right one, because it isn't going to change any time soon. In lower-growth economies, this leads to a different kind of volatility: they tend to revert to more egalitarian economies through some mix of total war, revolution, state collapse, or pandemics.

Meanwhile, back in the real world, we're pretty far from either pole. If there's a country that's achieved maximum inequality and zero-growth stasis, it's North Korea. But the rich world does, from time to time, have debates over policies to mitigate inequality. As noted above, these are really policies to shift inequality, moving power from rich people to legislators. That can be good or bad, depending on who goes into business versus politics and who succeeds at either. But one thing to note is that if you're worried about entrenched power, you should probably skew things towards wanting higher wealth inequality: last year's Forbes 400 had 14 new entrants and 22 returning members, for an annual turnover of 9%. Meanwhile, in the last election cycle, 97% of incumbents who ran were reelected, and even if you include people who declined to run, the annualized turnover rate was just 7%. The half-life of political power is longer than the half-life of financial power, and even that understates things, since seniority affects how much authority individual legislators have.

There's probably a level of wealth turnover and industry shifts that the average person would find unappealing even if it made them better-off in the long run. The market-clearing wage is lower for jobs with great job security, and a lot higher for the ones where you could be fired at any time. The tradeoffs are real, and different electorates can reasonably prefer different arrangements. Whatever arrangements they end up making.

Inequality and its cycles drive some interesting results, which we’ve covered in The Diff. For example, see:

• A solution in search of a problem is a low-rates phenomenon.

• The case for independent fiscal policy ($). 

• Why buy-borrow-die ($) increases intergenerational mobility but also whoever turns out to be richest much richer.

• We’ve considered the wealth-creation speed limit.

• Technological shifts and well-exploited luck can create a temporary cohort of very rich people ($).

Join the discussion!

1  One could imagine other models, like giving people loans to start small businesses or to spend a few years in apprenticeships. There will be adverse selection problems here, whereas housing tends to be homogeneous enough that you can use pricing of comparable units to figure out roughly what it's worth. Schools do face that kind of adverse selection; if someone has a high discount rate, they can borrow money to treat college as a sort of temporary retirement where they mostly have fun before un-retiring. Of course, it can also be incredibly valuable, but if you have a system that treats everything in a category as equivalent, it'll tend to relatively subsidize whatever member of that category is easier to manufacture. And it's a lot easier to expand Party State U. than to scale Harvey Mudd.

2  In a US context, sports probably wouldn't be affected that strongly, because athletic accomplishment is weighted in college's acceptance decisions and colleges still affect incomes. So America has a system where if you're reasonably athletic, it probably makes sense to invest relatively more time in sports, even at the expense of schoolwork, specifically to maximize your academic options. And if you take a normal high school path, and then go to college, you've used up many prime years of athleticism already! Athletic performance tends to peak in the mid to late 20s, but for team sports one likely factor is that playing on the same team for a while means getting better at working with those specific teammates, so the underlying athletic skill presumably peaks earlier.