When Donating to Charity Makes the Rich Richer: A Loophole in Raising Capital Gains Taxes

Andrew Granato
6 min readApr 28, 2021

The following is a slightly more detailed write-up of an observation economist Nic Duqette and I made on Twitter: that under President Biden’s reported forthcoming tax proposal, the combination of (1) the proposed federal top marginal income and long-term capital gains tax rates, (2) the charitable deduction, (3) California’s top marginal income and capital gains tax rates, and (4) that donations of assets are not taxable means that there will be an accidental loophole where Californians in the top tax bracket will be able to *gain money* by donating highly appreciated assets to charities rather than selling them to get cash. Here’s how it works.

It is reported that President Biden is going to propose raising the top marginal federal income tax rate to 39.6% and raising the top marginal federal capital gains tax rate to 39.6% (43.4% when you include a 3.8% existing Medicare-related tax on investment gains and income called the Net Investment Income Tax, or NIIT). The top marginal income and capital gains state tax rates (applying to people who earn over $1m a year) in California are 13.3%.

Let’s look at what happens when we look at an example where the California taxpayer in the top income bracket (let’s call this person T) has $100 of income taxed at the top income rate. T also has an asset worth $100 that has a basis (the original price of something when T acquired it) of $0. (For example, following from Duquette’s thread, let’s say T is a tech entrepreneur who started a company from scratch and now has a share of stock of that company that trades at $100 per share).

Scenario 1: T sells the stock

If T chooses to sell the stock, T will pay capital gains tax on the gain in price that the stock experienced. The stock went from $0 to $100, so the capital gain will be $100. The federal capital gains tax that T will pay will be 43.4% of that $100, so that will be $43.40, and the California capital gains tax that T will pay will be 13.3% of that $100, so that will be $13.30. The total capital gains tax that T will pay will be $43.40 + $13.30 = $56.70.

T will also pay $39.60 in federal income tax and $13.30 in state income tax on the $100 of income, so T will pay $52.90 in income taxes.

In total, T will pay $56.70 + $52.90 = $109.60 on the $100 of income and the realization of the $100 asset, leaving T with $100 + $100 — $109.60 = $90.40.

Scenario 2: T donates the stock to a recognized charity

If T chooses to donate the stock, because the donation of an asset means you don’t pay taxes on it and because there are deductions on charitable contributions in both the federal and California tax codes, T will not pay any capital gains tax on the donated stock and T’s income will be lowered by the amount of the stock that T donates. T’s income is $100 and the value of the stock that T donates is $100, so for tax purposes T’s income = $100 — $100 = $0. This means that T will not pay any income tax either, leaving T with $100.

In the above two scenarios, T comes out ahead by donating the stock to a charity rather than realizing the gain on the asset by selling it for cash. This is an incredible loophole, one that, in my view, potentially risks devastating capital gains tax revenue, particularly in California.


First caveat: this is only a pure loophole (where the taxpayer comes out truly ahead by donating the asset rather than selling it) when the basis on the asset is zero or near-zero.

The math here can get a bit more complicated, but think about it like this: let’s say that another top-bracket California taxpayer, A, bought a stock for $100 in the prior year and it has now appreciated in value and is worth $200. The taxable capital gain, if sold, is $100 (same as above). Scenario 1’s description of the tax consequences is the same, and additionally it’s as if A put in $100 into a vault for a year and then got it back as A paid $100 to get the stock and, when selling the stock for $200, only paid taxes on $100 of that $200. (For the tax heads out there, there is a time value of money component to this, but for simplicity let’s ignore that as it will be very minor here).

But in scenario 2, though the tax consequences to A are the same, A still had to pay $100 in the prior year to acquire the asset. So (once again, we’re bracketing the time value of money here), A isn’t as well off as T because A paid $100 last year for something that’s just gone now. The more that A had to pay to acquire the asset (the higher the basis of the asset), the less of a pure gain A is able to get by donating the asset relative to selling it.

Second caveat: there are limitations on the amount of appreciated assets that T can donate while still getting the federal and state charitable deductions on those donations. At the federal level, a taxpayer is permitted to donate up to 30% of their adjusted gross income to charity while still getting the charitable deduction. I believe that California has the same rule [though honestly I am struggling a bit with the wording of the California statutes, and this line may be edited later if I find out I am reading it incorrectly]. However, for a person in the top tax bracket, earning over a million dollars a year, this is still an enormous amount of money.

Third caveat: there are other tax provisions that will modify the tax consequences faced by real people, but not in a way that materially changes the core calculation above of the immediate difference in gain to the taxpayer from donating the assets relative to selling the assets. Some relevant provisions: (1) Up to $10,000 of state and local taxes, including non-income taxes like property taxes, is deductible for federal income tax purposes. (2) The amount collected by the NIIT cannot be reduced by charitable deduction. (3) There is a 6.2% payroll tax on wages that falls on employees and a 6.2% tax that falls on employers for Social Security, and a corresponding 1.45% tax falling on each for Medicare, that cannot be reduced by charitable deduction. (4) The United States has an estate tax that in 2020–2021 kicked in at estates valued at over $11.7 million, which further incentivizes charitable donation by the very affluent.


In scenario 1 (selling the stock), T pays $39.60 + $43.40 = $83 in taxes to the federal government and $13.30 + $13.30 = $26.60 in taxes to the California state government. In scenario 2 (donating the stock), T pays $0 (zip, nada, zilch) in taxes to both the federal and California governments. Because of a confluence of provisions, under President Biden’s proposal, extremely wealthy taxpayers in California will have a direct financial incentive to donate mass amounts of heavily appreciated assets to charities to make themselves richer than they would be if they cashed out their shares. I will note here that California is the home base of Silicon Valley, where there are billions upon billions of dollars of highly appreciated, unrealized securities that started from a very low basis.

I believe there is a risk that substantial amounts of tax revenue could be lost to this loophole, one that is created due to a confluence of different tax provisions that enable a small set of extremely wealthy people to get a highly disproportionate gain. Donating to charities is often virtuous, but the charitable deduction is already highly distortionary in favor of the rich, and in this scenario the cost of the charitable deduction is likely to be extremely disproportionate to the benefits. President Biden’s proposals generally seek to increase the progressivity of the US tax structure, but this loophole would significantly hamper that goal. While full policy prescriptions are a bit outside of the scope of this post, a much tougher limitation on the amount of appreciated securities that can be deducted from income when donated to charity on both the state and federal levels would probably be a bare minimum.

Update, April 28, 10:30 pm EST: the third caveat has been modified to more comprehensively account for additional tax provisions. A thank you to Daniel Hemel and Zachary Liscow for the guidance.



Andrew Granato

JD-PhD Candidate (Financial Economics) at Yale University