OP here. I sort of agree with you, but it's a bit complicated and it wasn't obvious what the right way to word that was.

To get granular:

Wealth taxes are a bit distinct from taxing unrealized gains directly (even though they often have that effect, and there is some obvious overlap). If we look at, say, the Netherlands, they're assessing a ~1.7% wealth tax on assets over €1m independent of the performance of the assets over the year in question, which then exempts the payer from capital gains taxes upon sale. So we can call this taxing unrealized gains, but it's a bit imprecise in that they aren't taxing the gains themselves (which are unknown, and could be losses), but rather wealth at a prior point in time based on a fixed formula. If the asset in question went up 20%, the Dutch gov isn't going to tax the excess or force realization on any specific timeline. They'll just keep taking their 1.7% every year on whatever is there on Jan 1st.

There also aren't a lot of countries doing anything like this as touching upon non-real estate investments, and most European countries that have experimented with them are in the process of reversion (e.g., Norway, France) as it basically hasn't proved workable in most instances.

So maybe I should have left something like this as a footnote to clarify.

What? This is extreme hair splitting. Switzerland, Norway, etc, do tax unrealized gains. They net tax wealth of saleable assets along with cash at a specific point in the year, as you note.

That includes unrealized gains.

It’s totally unclear to me why you think the distinction here somehow invalidates the ProPublica piece.

If what he wrote about Dutch taxation, this tax is a replacement for income / capital gains tax, not an addition (and so far none of the "tax the rich" schemes that I saw included removing income or capital gains tax).

At 1.7% a year it gives you quite a few years before it would even reach the same level of taxation as income / capital gains tax.

To make it concrete: last year I made $200k in short term capital gains in stock market. It was taxed at my high marginal tax rate of 35%+, that I had to pay this year.

In Dutch system I would be much better off: I would pay 1.7% on that increase this year and continue paying 1.7% for 20 years before that would equal the amount I paid this year.

And if I managed to, say, 4-8x that $200k in 20 years (at 15% you double the money in 5 years, at 7.5% you double in 10 years, which is a return on S&P index fund), I would be even better off in Dutch system.

So US system, at least regarding stock (which is how all those billionaires in ProPublica article created their wealth) is better only if you hold your stock for 10-20 years, which is generally regarded as a good thing. It both provides stability for stock price and shows the confidence of a person in the business.

Few points:

- Yes, I don't know exhaustively, but a number of countries with wealth taxes do not have gains taxes.

- Your comparison is a little bit mistaken, I think; you're looking at the tax only on the gain, whereas (obviously) the tax would be on the entire principle. I'm too lazy to do the math, but a wealth tax is easy to factor in, since it's a fixed reduction in total yield on wealth. The long-term historical return on capital in the west is about 4%/year; a 1% wealth tax is thus averaging about 25% of that. Gains taxes are typically lower than that amount. (Note that in the US the long-term gains tax is 20% at the upper bracket.)

ProPublica said that these billionaires were avoiding taxes because their unrealized gains weren't being taxed, and implied that they should be in a direct sense for fairness.

(Note that ProPublica mentions wealth taxes at the end, and then immediately dismisses them. Their object here really was/is gains taxation in a direct sense.)

My point was that this isn't tax avoidance, that there are reasons we don't tax unrealized gains directly, and that virtually no countries do it that way.

It's true that a few (and very much declining) number of countries have some form of wealth tax that includes some amount of unrealized gains. The Netherlands, for example, marks-to-market on Jan 1st of the tax year then doesn't actually track gains/losses over the next 364 days. So that obviously isn't a direct tax on gains. But it's certainly adjacent, and I said in the parent comment here.

What? How else would a tax on unrealized gains work if not mark-to-market at a fixed day of the year?

This is exactly how Wyden’s proposal works as well, so I fail to understand the point you think you are making.

Personal wealth taxes are literally a tax on personal wealth including unrealized gains.

“I was wrong” is much easier to type than your confusing reply. :)

There are taxes A B. B includes A.

ProPublica says "we'd love to see more of A, which is separate from B, as B doesn't seem especially workable".

A and B are not the same, even if one encompasses the other in some meta sense.

A is actionable in a direct sense. And if you wanted to do it in a direct sense, the M2M date you'd pick would not be the first day of the tax year. And you wouldn't use a fixed formula that ignored actual results of the asset.

> A and B are not the same, even if one encompasses the other in some meta sense.

What does "meta" mean in your usage? I may be missing some point you are trying to make. Personal wealth taxes encompass unrealized gains taxes, not in a "meta" sense, but in actuality.

Do you disagree?

> ProPublica says "we'd love to see more of A, which is separate from B, as B doesn't seem especially workable".

I took that to be ProPublica saying "one way to increase the total effective tax rate on the ultra-rich would be to tax unrealized gains. Wealth taxes do that; here are some downsides of wealth taxes as implemented elsewhere."

This strikes me as...fair and balanced reporting?

> A is actionable in a direct sense.

I don't know what this means. Hypothetically the US could have either a personal wealth tax or (only) a tax on unrealized gains.

> And if you wanted to do it in a direct sense, the M2M date you'd pick would not be the first day of the tax year.

No, I agree with this. The Swiss system uses mark-to-market at the end of the tax year. So...point you?

If ProPublica suggested the beginning of the year instead of the end of the year, I overlooked that. Good job catching them on this error, I guess? Or something?

> And you wouldn't use a fixed formula that ignored actual results of the asset.

I don't know what you mean by this. Can you rephrase?

It's the same as property taxes. Do they factor capital gains in some obvious sense? Sure! They're based on assessed value, which should be closer to market value than original sale price. But no one looks at them as the same as forcing gains realizations. Just two different things conceptually. Just so for wealth taxes, with are just property taxes for non-RE property.

Backing up a bit, you wrote,

"It's true that a few (and very much declining) number of countries have some form of wealth tax that includes some amount of unrealized gains. The Netherlands, for example, marks-to-market on Jan 1st of the tax year then doesn't actually track gains/losses over the next 364 days. So that obviously isn't a direct tax on gains."

I wrote,

"What? How else would a tax on unrealized gains work if not mark-to-market at a fixed day of the year?"

How do you believe an unrealized gains tax would work if not mark-to-market on a fixed day of the year?

Backing up even further, as I said, this is extreme hair splitting. Major economies do have wealth taxes, which include a tax on unrealized gains via mark-to-market. Is your point just "lol ProPublica r idiots nobody taxes just unrealized gains"?

> That includes unrealized gains.

And also includes everything else.

You pay the same tax if you have $1m in shares that you bought long ago for $10k or $1m in shares that you bought last month for $2m.

Indeed. An unrealized capital gains tax by itself (which does not tax the basis) probably requires you adjust the taxable basis to ensure there's no double taxation at sale. A wealth tax irrespective of basis is administratively easier, I think.

The part about Norway is total bullshit. The "reversion" is solely because the Conservative party, the party for rich people, is in power.

There's no evidence it hasn't proved workable. The left coalition has promised to increase the wealth tax back again to recent levels if they gain power in the September election.

Edit: Just got banned from his blog for making this comment... https://i.imgur.com/ZkaLuRT.png

It's certainly true that political winds are driving the reversions. But what's driving the politics?

Wealth taxes face three classic problems: (1) they're very hard to set in any fair sense given complexities of markets and assets, (2) they can induce liquidation in ways that are net bad for the economy, and (3) capital flight.

3 seems to have been a problem in Norway (https://archive.is/L6kyZ), as it was in France and elsewhere in Europe prior. If your tax system causes a net decrease in receipts because you're golden geese have fled, that's a workability problem!

That article consist of just a Conservative politician doing politics and a survey by various Chamber of Commerce's, not exactly good evidence.

Of course, the taxed are going to say that they don't like paying taxes.

Real evidence of capital flight in larger amounts than the taxes gained are harder to come by.

If you'd been respectful, there'd be no ban!

Propublica and OP talk about taxing gains (as though the gains were income), while this wiki link refers to taxes on wealth. Those aren’t the same thing. The wealth tax would be a low rate that applies to your entire wealth each year, while the tax on gains would be higher but only apply when the gains occur.

I’ll grant that these would be two different ways to achieve very similar results, but I don’t read where propublica proposed a direct tax on capital gains rather than a more generalized wealth tax, which very much is a real thing.

Propublica calculates a “true tax rate” which uses the unrealized capital gains as the denominator. They don’t make any public policy recommendation in the article, but I think computing the tax rate in this way implies that the unrealized gains should be taxable income. They also mention that Senator Ron Wyden has proposed doing this.

Why? Calculating the tax rate as “taxes paid” / “net wealth” is just what the effective wealth tax would be if one existed. This is consistent with advocating for a wealth tax, which, as I noted elsewhere, does in fact exist in other OECD countries than the US.

Wealth taxes use wealth as the denominator. Propublica uses change in wealth as the denominator.

True. I find the denominator chosen to be weird.