Thread by @dylanmatt

I rarely do Twitter threads but I have a lot of feelings on the unrealized gains proposal, and I think it's badly misunderstood, so here goes.

The core problem with the capital gains tax right now is that the realization rule (you only are taxed when you sell) creates massive economic distortions.

budgetmodel.wharton.upenn.edu/issues/2022/3/…

It effectively offers capital income an interest-free loan: if I get $1 billion in unrealized gains at a 2010 IPO and owe $250 million in taxes, I can wait a decade and still only pay $250 million, with no interest, if I just don't sell in the interim.

This creates lock-in, preventing investors from directing capital to new projects if they've already gotten big gains in an existing one. It's actively bad for startups.

It also is part of why you've seen a move away from dividends and toward stock buybacks, which let investors enjoy this benefit from deferring realization.

Economists and tax lawyers have been writing about this for decades. William Vickrey has a paper on it from 1939. Everyone agrees this is a problem. This is not a sudden Biden-era thing.

journals.uchicago.edu/doi/10.1086/25…

If you want to do an income tax rather than a consumption tax you need to structure it as taxing what's called Haig-Simons income, which treats unrealized gains as income. This is basic optimal tax stuff.

jct.gov/CMSPages/GetFi…

We already use mark-to-market taxation (the technical term for taxing unrealized gains) on some income. This is what the US v. Moore case was all about.

taxpolicycenter.org/taxvox/examini…

The alternative to mark-to-market that also fixes the deferral bias is to add a "deferral charge": you tax assets when they're sold but charge interest for each year they weren't sold. Auerbach had the classic proposal for this in 1988.

nber.org/papers/w2792

For illiquid assets like real estate and art, you have to use a deferral charge, because there's not a reliable way to value the assets before they're sold. (Tho I've seen proposals to use local property tax assessments for real estate.)

The Biden proposal uses a deferral charge not just for real estate and art but for stock in privately held companies.

If the stock is not valued on a public exchange, and more than 80+% of your assets are like this or otherwise illiquid, you don't have to prepay.

This is what drives me crazy about the Silicon Valley complaints about shares in non-public startups. They haven't read the plan and don't realize they're exempt.

The problem with deferral charges, and the reason why Biden didn't just go with them to cover absolutely everything, is political uncertainty.

Let's say Harris signs a law adding 3 percent to everyone's capital gains bill for each year they don't sell. The best move is to sell nothing until 2029 or 2033 when there's a Republican president and Congress and the charge is repealed.

Policymakers, in the process of trying to reduce lock-in, wind up increasing it dramatically. The unrealized gains tax is an attempt to avoid this problem by making it clear that you can't just wait indefinitely.

The normie alternative is to just end step-up in basis.

That rule says if I buy a Basquiat painting for $1 in 1983 and it's worth $10 million when I die, and my kid then sells it for $11 million, they only pay tax on the $1 million diff btw the sale price and the price when they inherited.

You could do "carry-over" basis (they pay when they sell, but on the full $11 million) or "realization at death" (they pay when they inherit on the $10 million).

Former is more administratively complex and doesn't raise money as fast, but latter raises all the "do I have to sell the farm?" issues the estate tax does.

Carry-over/realization at death are great, they're no-brainers, but they don't solve the core deferral problem.

Zuckerberg can still get an interest-free loan from the government on the tax from his initial Facebook windfall for decades and decades before he and Priscilla die. If it's carry-over, his kids get the windfall for decades more too!

Biden's initial BBB plans included realization at death and an anonymous group of rich families hired a motley crew of ex-Dem Senators (Baucus, Heitkamp, Breaux) to kill it.

nymag.com/intelligencer/…

Jon Tester ultimately killed it.

wsj.com/articles/sen-j…

A fair question is "if you couldn't put together 50 votes for realization at death, why even try for something more aggressive like taxing unrealized gains."

I think the answer is a mix of (a) there's public outrage at Musk and Bezos's low tax rates and DfP polling indicated that taxing unrealized gains was popular with the general public

dataforprogress.org/blog/2022/4/18…

(b) it's good policy from an optimal tax perspective, and why not put it out there (c) Biden needed to reduce deficits in his budgets and this raises a lot of money.

US v. Moore convinced me and a lot of other people that SCOTUS would strike down any unrealized gains tax.

vox.com/scotus/355969/…

Kavanaugh's majority opinion threaded the needle carefully to say he wasn't ruling against a wealth tax or unrealized gains taxes, but was also full of indications that he thinks there are constitutional arguments against either.

Barrett used her concurrence to argue that they're unconstitutional, and I suspect she has four people on the court who agree with her.

In light of that, there are a few options to address the problems with the cap gains tax. There's realization at death, an oldie but a goodie. There's deferral charges, though the political uncertainty problem where investors would just wait for a Republican admin remains immense

.@ZLiscow and Ed Fox have a clever plan to tax ultrarich people who borrow cash against their wealth, so you can't do the "buy, borrow, die" strategy that enables them to have cash to live on without selling their investments.

papers.ssrn.com/sol3/papers.cf…

@ZLiscow The constitutional argument here is rock-solid: it's an excise tax on borrowing, and the feds are absolutely allowed to levy excise taxes. I think this is what I'd prioritize.

@mcuban @ZLiscow In the long-run, interest rates are probably going to be lowish, because growth is sluggish and the population is aging, so the loans might make a comeback

newyorkfed.org/newsevents/spe…

@mcuban @ZLiscow I think you're right that taxing borrowing would push people into selling their assets. From a revenue perspective, that's fine: they'd pay capital gains upon realization. I think it's probably better efficiency-wise too by reducing lockin. But could be wrong on that.