What drives the rise in top wealth inequality?

This is the first Forbes 400. At No. 1 is the shipping magnate Daniel Ludwig, worth $2 billion. It took $75 million to make the list.

Ten years later, only of these 400 people were still on the list.

By 2002, .

In 2026, remain: .

Everyone else died, split their fortune among heirs, or was overtaken by someone richer.

These are the 40 richest Americans in September 2020, on the left, and a year later, on the right. 33 people were in the top 40 both years. The gray lines connect their ranks. Jeff Bezos was first both times.

One of the 2020 top 40 died before the next list: the casino owner Sheldon Adelson, in January 2021. His fortune leaves the top 40.

Six fell below the top 40, including the investor Carl Icahn, the hedge fund founders Ken Griffin and Steve Cohen, and the real estate developer Donald Bren. Their fortunes did not collapse. Others simply grew faster.

Six others took their seats. Sam Bankman-Fried, the founder of the crypto exchange FTX, was not on the list at all in 2020. A year later he was No. 32, with $22.5 billion. Nvidia's Jensen Huang rose from No. 51 to No. 34.

Adelson's fortune passed to his widow, Miriam Adelson, who enters the top 40 by inheritance. Deaths and heirs are what economists call demography.

Pick any pair of years:

Lists

Each dot is a fortune. The richer it is, the higher it sits. The 40 dots above the dashed line are the top.

Now let one year pass, one step at a time.

First, some people die: here two people at the top die. Their fortunes leave the top, and the average is now computed over those still alive. This is the exit term, the first part of demography. Here it raises the average slightly, because their average wealth was below the top's average.

Follow only the people who started the year in the top and are still alive, shown in blue.

Over the year, each of their fortunes grows at its own rate. Some rise a lot, some fall, some slip below the line. Each bar shows how far one fortune moved. The change in their average wealth is the drift: the average of these moves, with the biggest fortunes counting the most.

Drift is what most people have in mind: the rich own stocks, and stocks do well.

Now turn to everyone who was not in the top. They are highlighted, and the people we just followed are faded.

Their fortunes moved during the year too. Many of them also rose above the initial cutoff.

Now rank everyone still alive and keep the richest 38, the number of survivors from the original top 40. Some who started below the top are now above the new cutoff, and some who started in it are now below. The first replace the second.

This is the displacement term, which contributes to the growth of top wealth shares. It is positive whenever people swap places: as the bars show, those who broke in sit above the new cutoff and those who fell out sit below it, so the newcomers are always richer. This means the top share can rise even if the fortunes at the top grow no faster than the economy. It is larger when fortunes diverge more.

Last, new people join. The fortunes of those who died pass to their heirs, who are new to the top: here a widow inherits one fortune and enters the top. Three children split the other, and none is rich enough to enter, so the second seat goes to the richest person below the cutoff.

Opening the top to these newcomers is the entry term. Heirs usually hold less than the person they replace, so entry tends to pull the average down. Exit and entry together make up demography.

Each dot is one year. It shows how fast the Forbes top wealth share grew between two lists. On average it grew a year.

Drift is the growth of the surviving fortunes already at the top, relative to total U.S. household wealth. It accounts for most of the ups and downs.

Displacement is never negative. It adds a year on average.

Demography is a small, steady drag: a year on average. It barely changes over four decades.

Now gather each line's years into a column. Each cloud shows how much that term varies from year to year: drift swings widely, while displacement and demography barely move.

Average each cloud over all the years from 1984 onward. The Forbes top wealth share grew a year. Drift added , displacement , and demography .

Displacement accounts for about half of the rise.

Let's now zoom in on the year-to-year variations of each term, starting with drift.

Drift follows the stock market, because the fortunes at the top are mostly stock: in the dot-com boom, in the bust, in the financial crisis, in the rally after COVID.

Booms and busts mostly cancel out. On average, drift adds a year: the surviving fortunes at the top grow faster than total U.S. household wealth.

Displacement comes in waves. It is high when a new technology or a new kind of finance lets fortunes grow very fast, and low when the existing fortunes consolidate.

Below the chart are the five biggest newcomers of each wave, ranked by their wealth relative to the average American household.

The 1980s: corporate raiders like Saul Steinberg, and the first personal-computer fortunes, such as Microsoft's Paul Allen. Displacement averaged a year.

The dot-com boom: the fiber-optics founders David Huber and Gururaj Deshpande, eBay's Pierre Omidyar, and Ty Warner, the maker of Beanie Babies. Displacement averaged a year.

After a very calm period, displacement came back. Crypto, SPACs and pandemic stocks brought in Sam Bankman-Fried, the trader Jeff Yass and Coinbase's Brian Armstrong. Displacement reached in 2021, its highest level since 2001.

The AI boom brought in OpenAI's Greg Brockman, the robotics founder Brett Adcock, Surge AI's Edwin Chen and AppLovin's Adam Foroughi. Six co-founders of Anthropic also entered in 2026. Displacement rose from in 2023 to in 2026.

Across these waves, displacement has declined. Before 2000, it added a year on average. Since 2000, it has added only , even with the crypto and AI booms.

Why did it fall? The contribution of displacement to the growth of the Forbes top wealth share depends on two forces: the volatility of fortunes and inequality within the top.

To see each at work, take a top of 40 fortunes, as in the illustration above, and let a year pass. The people who swap places are in pink, and each bar shows how far they end up above or below the new threshold. Here displacement adds .

The first force is the volatility of fortunes: how much they move in a year. Now let every fortune move twice as much.

Twice as many people swap places, and each swap is twice as large. Both effects add up: displacement rises from to , four times as much.

The second is inequality within the top. Go back to the original volatility, and make the top twice as unequal instead: the fortunes above the threshold spread out.

This matters in two ways. Fewer people sit close to the threshold, so there are fewer swaps. And the threshold sits further below the average fortune in the top, so each swap moves the average less. Displacement falls from to , about half.

The following graph tries to quantify the influence of this force. The new green line shows the model's prediction if we start from average displacement in 1984–88, hold volatility fixed, and let only inequality within the top change.

It falls from to about a year since 2020. In this calculation, rising inequality within the top cuts the displacement term by about half.

The difference between the two lines, shaded in orange, reflects changes in the volatility of fortunes and departures from the model's approximation. In the dot-com and crypto booms, displacement rose above the line. In the calm 2010s, it fell below.

Since 2020, the actual term, a year, has averaged close to the green line. So rising inequality within the top accounts for most of the decline since the 1980s. When the top is very unequal, each new fortune counts for less.