A fashionable theory of how the rich avoid taxes captures something real—but it misses what’s mostly going on. Consider two of the wealthiest billionaires in the US: Jeff Bezos and Elon Musk. The theory says they never sell their stock and never draw a real salary. Instead, they borrow against their appreciated shares to finance yachts, jets, and everything in between. When they die they’ll pass their unsold stock to heirs on a “stepped-up” tax basis, resetting the stock’s value to its price at their death and wiping out any capital gains during their ownership.
Buy, borrow, die? A catchy, scandalous story.
This story is everywhere, from Ezra Klein’s interview with Professor Ray Madoff about her acclaimed book “The Second Estate,” to newspaper stories and cable segments across the country.
The catch? It isn't the full story. Buy-borrow-die is a real loophole, and in absolute terms the rich borrow a lot—but the data say the very wealthy are mostly saving, not borrowing.
Using two decades of household data, we measured the annual borrowing of the top 1 percent of American wealth-holders. That borrowing comes out to roughly 1 to 2 percent of their economic income (which includes unrealized capital gains). Meanwhile, their unrealized gains over the same period were 20 to 40 times larger.
If buy-borrow-die were the dominant playbook, the rich would be leveraging aggressively against their ballooning portfolios. They aren’t. Borrowing against unrealized gains is, if anything, more of a middle-class habit than a billionaire one. (Figure 1)
How are the ultra-rich funding their lives, if they’re not selling or borrowing much?
The answer is boring: They earn plenty of taxable income—a lot more than they spend—in the form of salaries, dividends, interest, business profits, and realized capital gains. They reinvest the substantial difference between their taxable income and their consumption. So the dominant tax strategy of the super-rich isn’t some exotic loan-against-stock scheme. Instead, they save a large share of their liquid, taxable income while unrealized gains on their assets compound, untaxed.
Call it “buy, save, die.”
This strategy matters because it changes the arithmetic of tax reform. Including unrealized gains, the current income tax base captures most economic income: about 60 percent for the top 1 percent, and roughly 71 percent after adjusting for inflation. That suggests raising rates on income already in the tax base would raise meaningful revenue, though it would not solve the problem of untaxed gains. (Figure 2)
The headline cases that fuel the buy-borrow-die narrative—like ProPublica's reporting that Elon Musk paid a “true tax rate” of 3.7 percent—are real, but they reflect extremes. They are not the typical experience of affluent households, or even of most taxpayers usually described as “the rich.” (Even the richest like Musk and Bezos also in some years have much higher “true tax” rates when they sell tens of billions of their stock).
A less catchy but more powerful reform: Raise rates on income already in the tax base
Critics are right that the tax code is riddled with problems, that the ultra-wealthy enjoy advantages unavailable to wage earners, and that the step-up in basis at death is indefensible.
Buy, borrow, die is a problem, but one that sits fairly far down the list. As a result, we would focus a reform agenda in a different place: raising rates.
The IRS already taxes much of the income earned by the very rich, including salaries, dividends, interest, business income, and realized capital gains. But pushing up ordinary and capital-gains rates on those top earners would collect serious money without requiring Congress to design a constitutionally contested wealth tax or a mark-to-market regime for illiquid assets.
And higher rates would hit the “save” part of buy-save-die where it lives—on the taxable returns that help finance the accumulation in the first place. Of course, if rates went up, behavior might respond, but reliable estimates take that into account—and still say that increasing rates even a few percentage points would raise hundreds of billions of dollars.
As to buy, borrow, die, one possible response is a borrowing tax: a policy that would tax or limit the advantage of borrowing against appreciated assets instead of selling them and realizing capital gains. We outlined one approach ourselves, and Sen. Ruben Gallego (D-AZ) and Rep. Dan Goldman (D-NY) recently released a similar proposal informed by our analysis.
A recent Washington Post editorial also cited our analysis, arguing “buy, borrow, die” is a “myth.” But taxing the super-rich’s borrowing has merits. The top 1 percent borrow over $1 trillion, potentially unlocking hundreds of billions of dollars of unrealized gains without paying tax. Nevertheless, taxing the super-rich’s borrowing, despite its merits, is aimed at only a small share of the untaxed gains accumulating at the top.
Another way to shut down buy, borrow, die would be to reform the step-up in basis at death. The step-up allows a large fraction of capital gains accruing to the rich to go entirely untaxed. Reforming it would raise serious revenue and reduce the incentive for the rich to avoid realizing gains in the first place.
If we want reform that works, it must be based on two truths. The rich do pay taxes, more than the conventional wisdom suggests. But they also shelter enormous gains from taxation, less by borrowing cleverly than by holding assets as their value grows. Reformers who want to raise meaningful revenue from the top should spend less time chasing loans against Tesla or Amazon shares and more time focusing on the rates we already have on the books.
Zachary Liscow is an affiliated scholar with TPC and a professor of law at Yale Law School. Edward Fox is a professor of law at the University of Michigan.
