That Nearly Secret IRS Personal Wealth Study

Last spring, some fascinating, nearly secret IRS personal wealth information slipped out into public view.

And, just so there’s not confusion, I am not talking about the ProPublica folks and their semi-self-righteous reporting about the hacked tax returns of billionaires.

No, what I’m referring to? The IRS Personal Wealth Study 2016. That study, based on estate tax returns filed for people who died in 2016, appeared in May of 2021. And it provides a fascinating window into the wealth and finances of the top third of the top one percent. So the richest 734,000 people in the country.

Why this Matters

Folks who want to understand mechanically how someone gets really wealthy benefit by digging into the data. So, I’m thinking policy makers, sure. But also voters and citizens. Also maybe parents talking to their kids about jobs and careers?

Small business owners who want to understand how entrepreneurship impacts wealth benefit by looking at the data. Interesting intersections exist, it appears, between wealth building and entrepreneurship.

Finally, those professionals among us who serve these folks benefit from carefully looking through the data and assessing what it means. The accountants, attorneys, investment advisors, bankers and so on.

You’re probably going to want to read the summary reports the IRS published if you find this discussion at all interesting. (I provide links to those at the end of this blog post.) But let me give you an overview of the study and the data it summarizes. And then let me also point to some things that jump out to me.

IRS Personal Wealth Study  2016: An Overview

First, though, a quick overview of how the study works.

The IRS requires estate tax returns be filed when a taxpayer’s death may trigger federal estate taxes. The threshold for when estate taxes kick in changes semi-regularly. But in 2016, any taxpayer dying with an estate in excess of $5.45 million potentially owed federal estate taxes.

Accordingly, in 2016, tax law required the estate of any taxpayer who died with gross assets greater than $5.45 million to file an estate tax tax return. Roughly 14,000 estates filed such a tax return.

The IRS then coupled that estate tax data with mortality information to estimate how many other living taxpayers’ gross assets exceeded $5.45 million. Again, this would be in 2016.

A simplified example to illustrate: If the estates of one percent of 40-year-old males who died had gross estates exceeding $5.45 million? The IRS thinks that one percent of gross estates of 40-year-old males who lived had gross estates exceeding $5.45 million.

Some really interesting statistics drop out of their study. A few morsels of which I’ll point out in the paragraphs that follow.

The Birds-eye View of the Top Third of the Top One Percent

A good place to start? With a birds-eye view of the folks the IRS calls top wealth holders.

The table below gives this snapshot, providing the net worth data and then the number of individuals in a handful of wealth brackets:

Gross Assets Brackets	Count	Mean Net Worth
Under $5.45 million	151,529	3,686,752
$5.45 million under $10 million	373,730	7,133,350
$10 million under $20 million	137,773	13,356,950
$20 million under $50 million	52,325	29,607,930
$50 million or more	18,995	147,877,427
Total	734,352	12,831,708

Explaining the Table’s Information

For example, that “Under $5.45 million” bracket counts up and calculates the mean net worth for the taxpayers whose gross estates exceeded $5.45 million but whose net worths in the end fell under that threshold (probably due to debts.) As the table shows, 151,529 individuals fall into this bracket. Their mean net worth? Almost $3.7 million.

The next bigger bracket up reports on individuals with net worths of $5.45 million to $10 million. That bracket actually includes more than half of the top wealthholders. Their mean net worth equals slightly over $7.1 million.

The even bigger dollar brackets? Ever bigger net worth numbers. Ever smaller groups.

Two other notes: First, those mean averages run way higher than median averages according to the IRS personal wealth study. While the mean net worth of the top third of the top one percent equals roughly $12.8 million—see that last row in the table—the median net worth of these folks is around $7 million. (Figure C in the IRS Personal Wealth Study 2016 document gives this median information.)

Second, the IRS says the threshold wealth required for an individual to join the top third of the top one percent equals roughly $1.5 million. (Figure E in the IRS Personal Wealth Study 2016 supplies this insight.)

If you and I assume that every individual top wealth holder is married to another individual top wealth holder—which isn’t as farfetched as you might think based on the data—that means in 2016 a household or family needed about $3 million of net worth to join the top third of the top one percent.

And now, let’s dig into the data…

Big Surprise in the Personal Wealth Study

The big surprise the data provides? The IRS says far fewer rich people exist that many think and than the media commonly reports.

For example, you just read the IRS study and data suggest a family or household needs about $3 million to join the top third of the top one percent.

Many economists like Eric Zwick, Emmanuel Saez, Gabriel Zucman as well as the Survey of Consumer Finances sponsored by the Federal Reserve estimate that a family or household needs about $3 to $4 million to join the top one percent. In other words, you commonly hear or read that to join the ranks of the richest 1,250,000 families in the country, you need $3 million to $4 million in net worth.

But the IRS says their data show you can’t possibly have more than about 365,000 households with $3 million or more of net worth. That would be the absolute maximum. And you could have far fewer.

Why the Disagreement?

Why this disparity occurs? It’s complicated. But in the case of the economists mentioned above it appears to boil down partially to the way the economists look at the income the wealthy report on their tax returns. The economists don’t really know how much wealth people possess. So they look at people’s tax return income and then convert (or “capitalize”) that income into wealth.

But then other additinal factors explain why the IRS’s numbers come in so low, too. For one thing, some economists think the wealthy do a good job at moving their wealth out of their estates before they die. (I doubt this—at least for the great majority of wealthy folk—based on how I personally see people plan their estates.) For another thing, the IRS’s approach to extrapolating from estates to the living shows extreme sensitivity to small changes in input–which could mean their formula doesn’t work well.

Where this conflict between economists and the IRS leaves those of us who look at and like the IRS’s statistics? Not sure. But maybe we remind ourselves the IRS data skew much lower than conventional wisdom. And as a result, we maybe need to approach that data with skepticism. But then at the same time, this companion idea: Maybe we also keep in mind the possibility that IRS may be right.

A final further thought about the low IRS numbers: If you do work in a role where you deliver services to the top wealth holders? The IRS personal wealth study probably provides more practical details about how the wealthy work, live and invest.

Which nicely leads to my next point…

The Archetype Multimillionaire According to the IRS

The IRS personal wealth study allows one to describe the archetype multimillionaire by looking at the wealth bracket that includes half of the individuals in the top third of the top one percent.

The table below summarizes some of the key financial bits of information about this bracket.

	Mean	Percentage
Net worth	7,133,350	N.A.
Personal residence	806,165	72%
Other real estate	1,308,722	55%
Closely held stock	2,335,341	31%
Publicly traded stock	1,869,889	81%
Noncorporate business assets	1,720,369	38%
Farms	3,371,305	13%
Cash	620,841	98%
Retirement accounts	1,229,309	71%

Again this caution: Those dollar values shown in the table represent arithmetic means. Not medians. The median averages probably run about half or maybe two-thirds of the mean. (The IRS provides some of these values in its very useful summary.)

But ignoring that mean-versus-median wrinkle, the average multimillionaire—so the average person in the top third of the top one percent—possessed about $7.1 million of net worth in 2016. (I guess the average person in the top third of the top one percent probably had a median net worth of a little over $5 million based on the tables shown in the actual IRS personal wealth study.)

About 72 percent of these individual owned their own homes. A very nice home obviously if the mean equals $800,000. (Remember, we are talking 2016 dollars. Also for what it’s worth, I’m guessing that the median home value is probably half the mean average.)

On average, the person held a huge pile of cash. That makes sense given the thing I talk about in the next section.

Seventy percent held a retirement account and these folks enjoyed  a very decent sized retirement account balance: More than $1 million on average.

These people also often held a big, two-million-ish chunk of publicly held stock.

Obviously, these top-third-of-the-top-one-percent-ers have made a way better-than-average living. They also bought a nice house. They probably followed the standard advice to save for retirement.

But in many respects the data suggest they look pretty unremarkable. Except for one thing I’ll discuss next…

Business Ownership Dominates

The big striking feature that appears, at least to this accountant’s eyes, are those business-y assets.

About a third of these folks hold a two-million-dollar chunk of stock in a closely held corporation.

Nearly 40 percent own an investment in a noncorporate entity like a professional services partnership. Or an interest in an S corporation.

A little over 10 percent own an interest in a farm or ranch. Or all of a farm or ranch.

Slightly more than half, 55 percent, possibly own investment property.

You can get an Excel spreadsheet with all the gritty details (see link to IRS statistics page below). But back of the envelope calculations suggest and the Federal Reserve in its survey confirms that a big percentage, around 70 percent, own a business or a part of a business.

The point  here? These people have also often combined good luck with the right skills to own all or part of a successful small business.

Risk and Illiquidity Premiums?

In fact, can I share a hunch? I am pretty sure—no, wait, sorry, let me change that. I am certain that these people have borne significant business and financial risk to join the top third of the top one percent.

People who become business owners do not automatically join the top third of the top one percent. At least ten million small businesses exist in America. Some counts go as high as thirty million small businesses. Only a fraction of these folks end up in the top third of the top one percent, according to the IRS. Like maybe a quarter a million? Or maybe at absolute most half a million?

And so I think what jumps out here are two interesting features of all these business owners in the top third of the top one percent.

First, they’ve gotten lucky in a sense by getting a great long-run return on their high-risk business investment.

And then a second something connected to this, I think. They’ve unfortunately or fortunately been forced to hold and probably add to their business investment over years and years. Maybe over decades and decades.

That combination of a risk premium and a illiquidity premium—in this very best case scenario—delivers a great financial outcome.

Growth in the Ranks of the Wealthy

One final observation, and one I make for the benefit of the accountants, investment advisors, bankers and attorneys who read this blog.

That big bracket that represents the lion’s share of the top third of the top one percent? You know, the bracket showing people with a net worth of $5 to $10 million? Not only do most multimillionaires fall into that category. But that group appears to have grown over time relative to the population.

The IRS talks about that in their study. And I think that means if you provide services to this group—most tax accountants do—your business or professional practice probably should grow too.

By the way, the even higher brackets in the IRS personal wealth study report on people with $10 million to $20 million, $20 to $50 million and over $50 million. But the numbers of individuals in those categories are not growing in relative terms. Those market niches appear more stagnant in terms of real growth.

Which maybe makes sense. Because trees don’t grow to the sky. And because very possibly people move out of alternative assets (like a small business) and into traditional assets (like the publicly traded stocks and then bonds) as their wealth grows. Obviously, those moves may natually produce lower returns. And make it easier to spend.

Other Sources

The IRS’s short summary appears here: Publication 5536: Personal Wealth 2016.

A full analysis appears here: IRS Personal Wealth Study 2016 by Aaron Barnes.

The IRS provides Excel spreadsheets with all the agregated wealth data here.

The many components of the 2019 Survey of Consumer Finances appear here (that survey includes the 2016 numbers mentioned above). The best component to grab and read first is probably this one: Changes in Family Finances 2016 to 2019.

Finally, I want to provide a link to the defense by Professors Saez and Zucman of their wealth estimation methodology. I disagree with their tax-the-wealthy conclusions. And I suspect their methodology inflates their wealth estimates. But they’ve made wonderful contributions to the public debate on this subject. So here are two papers of theirs that address the issues discussed in this blog post. The Rise of Income and Wealth Inequality in America: Evidence from Distributional Macroeconomic Accounts and Wealth Inequality in the United States Since 1913: Evidence from Capitalized Tax Data