Tax Bill 3: Don’t Mess With Taxes

I.

Thanks to everyone who commented on my last two posts, especially the many people who disagreed with me. Two things I will admit I got mostly wrong:

1. I was wrong to say there was “no case” for the tax bill. Aside from all of the minor provisions which can be good or bad, the case for slashing corporate rates is that they’re more distortionary and less efficient than other forms of taxation. Thanks to everyone who pointed this out to me.

2. Several people brought up problems with the article saying CEOs say they will just give the money back to shareholders, most notably that giving money back to shareholders may stimulate the economy in other ways.

But two things I still think are true:

1. Seriously, guys, I admit I don’t know as much about economics as some of you, but I am working off of a poll of the country’s best economists who came down pretty heavily on the side of this not significantly increasing growth. If you want to tell me that it would, your job isn’t to explain Economics 101 theories to me even louder, it’s to explain how the country’s best economists are getting it wrong. You may find this book review relevant.

2. I stand by my claim that I care less about economic growth than about where the money goes. That includes caring less about distortionary taxation, deadweight loss, and all those other concepts.

Suppose Alice is an effective altruist who supports whatever charity you think is most important and does a really good job of it. Every dollar she spends saves multiple lives. She lives in a town of 1000 people where nobody else is an effective altruist and everyone else just lives a pretty decent life and spends their extra money on, I don’t know, breeding virtual cats or something.

A demon places a curse on Alice’s neighbor Bob. Every time Bob pays a dollar in taxes, it destroys a random two dollars’ worth of wealth somewhere in the town.

The town elders meet and decide that for some reason they have to lower taxes either on Alice or Bob. The economic case for Bob is overwhelming – taxes on him are especially inefficient because of the extra wealth they destroy.

Still, I would want a tax cut for Alice. It seems like the only important thing that happens at all in this town is Alice’s charitable donations. The amount I care about this town’s utility focuses pretty much entirely on that. We could give the break to Bob, and have a nominally better economy, but it would just lead to more people buying virtual cats. It could be that the extra two dollars’ of wealth destroyed by Bob’s taxes was some sort of useful machinery, and so taxing Bob harms economic growth. Again, it is hard to care, except insofar as that hurts Alice, the only person in town whose wealth matters much for anyone’s utility.

I can imagine a world in which Bob’s curse was stronger, and every dollar Bob was taxed destroyed a million dollars in value, and soon any tax on Bob meant the citizens of the town were starving to death and all of them including Alice went bankrupt. But right now the tax on Bob isn’t big enough to be worse for Alice than a tax on Alice, and since Alice is the only important person in this situation, I don’t care.

I can also imagine a world where a wise economist comes to town. She says “Alice’s work is the most important thing in this town, but taxing Bob destroys wealth for no reason. Some of the town elders support tax breaks for Bob, and others support tax breaks for Alice. But we can give the tax break to Bob, and then all the people who saved $2 each from the curse not being activated can give $1.50 to Alice. That way Bob is better off, Alice is better off, and potential curse victims are better off.”

This is the best argument in favor of wealth creation instead of redistribution. But right now we’re not doing that. We just create the wealth and then don’t redistribute it, except through charity, which is a rounding error, and taxes, which everyone agrees this bill causes there to be less of. If we actually had Pareto-optimal wealth redistribution, then of course, create as much wealth as possible and redistribute it Pareto-optimally. Since we don’t, we’re kind of stuck.

My takeaway from this story is that in societies with a lot of marginal-value-of-money inequality, economic growth is potentially less useful than working to keep the money with people who can spend it on higher-marginal-value things. Consider three variables:

1. How low is the marginal utility of money for the person holding the average dollar, if no efforts are made to redistribute it?

2. How much economic growth are we sacrificing by choosing redistribution?

3. How high a marginal utility of money do we get by redistributing it?

Point 1 is why I stress the research showing increasing inequality eg most money going to people rich enough not to really have much use for it.

Point 2 is why I stress the economists saying that the gains from cutting corporate taxes really won’t have that much effect on growth.

Point 3 is the one I’m least sure about. If the government were a perfect effective altruist, it would be no contest – them having the money would be thousands of times more effective than random corporations (or even random middle-class people) having it. Even if the government were to give the money as a tax break to the working classes, it still seems really obvious to me that the increased utility swamps any effect from higher economic growth. In reality, the tax cut is being funded by increasing the deficit. I don’t know whether that means we need to compare it to whatever is bad about having a higher deficit, or else take as a given that the deficit has a certain amount of slack, and then compare it to other things we could do with the same money.

Imagine the government went $100 billion into debt to build a giant bronze statue of George Washington. Should we be debating whether running up the deficit is really that bad? Should we be debating the artistic merit of giant bronze statues of Washington, and whether it’s actually a pretty good statue that boosts tourism in the area? Or should we be comparing it to the best possible use for that money?

(added: I would be 100% happy with a bill that cut corporate taxes exactly this much, then raised taxes somewhere else in an equally progressive way, causing there to be the same amount of taxes with less distortion)

II.

The fairest thing I can think of is to compare this use of $100 billion to just spreading $100 billion evenly among all the government’s existing priorities.

Suppose that this tax cut was vastly better at stimulating economic growth than any reasonable person expects, and it increased growth by 1% per year. Then it would create $200 billion in value. With extreme good luck, 3% of that might go to the poorest quintile, giving them an extra $6 billion.

Or suppose the government keeps the $100 billion and distributes it evenly according to its existing priorities. Half of the budget is entitlement programs, and 32% of those go to the poorest quintile, so they would get an extra $16 billion.

I’m sure these numbers are wildly off. But it’s hard to come up with remotely plausible numbers in which the poor and working-class are better off with the tax bill than without it. I think the assumptions I plugged in were overly generous: the bill won’t really increase growth 1%, and although poor people have 3% of income they get much less than 3% of economic gains. Still, even under these generous assumptions, this bill gives poor people less money than the default case of not doing it.

One could argue that poor people are better off with $6 billion in actual money than $16 billion in government programs purporting to help them. But although I agree there’s a multiplier, I don’t know if it’s this big. And government programs would also disproportionately help the poorest of the poor, compared to economic gains which would disproportionately help the richest of the poor.

I think the marginal utility from an extra dollar to the poor (and the working class, etc) is orders of magnitude higher than the same dollar going to something else. So if you want to get me to support the tax bill, don’t tell me yet another reason why you think it would make the economy more efficient. Tell me why I’m wrong about this.

[EDIT: Commenters point out I was mistaken about the speed at which this would compound. See here. If the real growth from the bill was as high as 1%, it would probably be better for the poor than the lost government spending; if it were lower, it would take several decades to break even. So the best way to convince me to support this bill would be to find a plausible estimate of what level of growth is expected. My best guess from the economist poll is still “approximately zero”. ]