Nordics Taxing the Rich

Jonathan Chait said:

The most generous tax-and-transfer systems in the world — such as in Denmark, a country Sanders has cited as a model — do not tax the rich all that much more heavily than the United States does.

I said:

This is simply untrue. The Nordic countries do tax the rich much more heavily than the US does. They tax everyone much more heavily than the US does.

Since this type of discussion is typically focused on income tax rates in the US, I provided both maginal and effective labor tax wedges to make my point:
 
 
 
The marginal labor tax wedges are quite a bit higher, especially in Sweden and Finland. Denmark's and Norway's are also higher than the US, though to a lesser degee. It's clear at least on this labor income metric that the countries with the most generous tax-and-transfer systems do tend to have much higher taxes on the rich than the US does. It's also worth emphasizing again that the usual retort about how they tax the middle more never seems to appreciate that given the way tax rates work, this necessarily means they tax the rich more too. This is why across-the-board rate cuts like Bush did do not benefit the rich only in proportion to the reduction in the top marginal rate but also as a result of reductions in the bottom and middle rates.
 
Chait II
 
In response to my post, Chait quotes a Jonathan Cohn piece from 2007 (a favorite of his), which states without any links or citation that:
 
Denmark’s “high rates on personal income,” as Jonathan Cohn wrote in 2007, “mask relatively low rates on investment capital and corporate earnings.”
 
Just as Cohn gave Chait the wrong impression about Danish labor market institutions (see here), he's also given the wrong impression on the nature of Danish income taxation. Comparing corporate income tax rates across countries is fraught because statutory rates have little to do with effective rates, given that a popular form of business support is targeted corporate tax relief. As Chait explained in 2011, the US "corporate income tax has high statutory rates but, with its vast loopholes and exceptions, low effective rates."
 
Taxes on the investment income of shareholders are somewhat easier to compare reliably by looking at institutional details because they feature far fewer carveouts and the carveouts they do have are generally applied across the board in ways that make it easy to work out effective rates from the institutional details alone.
 
 
 
Denmark dominates the field by quite a significant margin. The others are similar to the US. It's worth noting in the Norwegian case that they also have a Piketty-style annual tax on net worth set at 0.85%. If you were trying to figure out some kind of all-in tax on shareholders, you'd need to factor that in as well. Some basic back-of-the-envelope math that assumes the rate of return is steady at 5% means that a 0.85% wealth tax is equal to around 17% of the annual rate of return, this on top of the 27% above.
 
After quoting that old Cohn piece, Chait shares an email message from Kenworthy that confirms my basic point:
 
Your piece made the point, assuming I understand you, that Denmark's tax system isn't more progressive than ours. Bruenig is making the point (though he doesn't have the data) that Denmark's effective tax rate on the 1% is higher than ours. Both are true. As best I can tell, Denmark taxes everyone at a higher rate than we do, and their tax system is no more progressive than ours — probably a bit less progressive (the tax wedges Bruenig looks at don't include consumption/VAT taxes).
 
I wasn't focused just on the 1%, but Kenworthy grabs on to precisely the confusion on this point that I have been trying to fight for years. The confusion is that people take the point "Nordic tax systems are less progressive than US tax systems" and then erroneously conclude that it means "Nordic tax systems tax the rich less than or the same as the US." But these are not the same thing and the second point is, as Kenworthy confirms, not true. The former point is often quite muddled for reasons I've explained before (see here), but is true under conventional definitions.
 
It's hard to compute someone's Actual Tax Rate in some grand universalist sense, not just because of the data but also because the terms are inherently disputable. For instance, Greg Mankiw argues (in so many words) that any tax system with inheritance taxes (or logically net worth taxes like in Norway) has an effective tax rate that approaches 100%. This is because any bit of saved labor income would, relative to a no-tax counterfactual, eventualy be gobbled up almost 100% by an infinite number of inheritance (or net worth) tax iterations.
 
Likewise, consumption taxes are often held to be "regressive" but a truly compehensive rate might observe that all income is eventually consumed and so the rates are actually flat. Indeed even if you don't play around with the time horizon on consumption taxes, it's not hard to see that the fact that the rich pay around 25% VAT in Nordic countries causes them to be taxed even more extremely than their higher labor taxes suggest because in the US that same rich person would pay maybe 8% in sales tax depending on where they live.
 
There is also the matter of what the numerator and denominator should be, which seems straightforward but is anything but. In the numerator, do you include literally all taxes paid, including consumption tax? If so, then dividing current tax by current income would have some people paying infinite tax rates, as they may be using prior income to fund current consumption (and pay the consumption tax) while having no current income of their own. In the denominator, do you include only factor income or also transfer income? Whose numerator does a gift or inheritance tax go into? The donor or the donee? I could go on.
 
Despite these complications in quantifying what the Actual Tax Rate of anyone is, it seems fairly clear from the institutional details that, based on most people's notion of what higher taxes are, the Nordic rich pay them, as Kenworthy confirms. The marginal and effective labor tax rates are much higher. The consumption taxes they pay when they spend their net labor income is much higher. The investment income taxes are similar or, in the case of Denmark, much higher.
 
Lastly Chait turns to a practical point:
 

Again, given the political constraints posed by public opinion in the U.S., it makes sense to focus on heavier taxes on the rich. It also makes sense to find opportunities for other sources of revenue to fund worthy investments, like prekindergarten education in Philadelphia. The point of my post was that it is ideologically self-defeating for Bernie Sanders to oppose such an effort. Do Bruenig and other Sanders enthusiasts disagree?

At the very end of my post, I defended the point that eventually you'll obviously want to bring in the VAT. And I've long advocated for a VAT. As far as particular excises like soda taxes go, I don't really know enough to say. The proximate reason you go for an excise rather than an increase in the general sales tax (or VAT) is not the revenue (which you can get equally from upping the sales tax or other taxes really) but the targeted impact on consumption. I have no idea at what level a soda tax impacts consumption, how valuable that impact is, and how it is distributively different from a general sales tax. I know Finland (my fav) had a sugar tax that didn't really work and that they eventually nixed because of its failure and for EU-related reasons. But I can't say I know much beyond that.
 
As a general matter though, I am not sure what is meant by saying it is ideologically self-defeating. In his original post, Chait seemed to identify this as the self-defeating character of it:
 
The largest single impediment to the kind of welfare state most liberals favor is finding ways to pay for it. There’s more money to be raised by taxing only the rich, but not enough to finance the kind of social benefits American liberals would prefer.
 
The confusion I have with this is that people who advocate for larger welfare states, such as Sanders, do ask for large levies on the non-rich via labor taxes. Indeed they face attacks from centrist liberals for doing so. Bernie's remarks on the soda tax seem to be rooted in the reflexive liberal opposition to consumption taxes more generally as "regressive," not the centrist liberal opposition to all additional taxes on anyone with income below $250,000.
 
So the question is how self-defeating it is for liberals to maintain their disdain for consumption taxes because of their "inherently regressive" character. I would say it's somewhat self-defeating in the very long run, but it's not much of an impediment at all right now for "finding ways to pay" for the "kind of welfare state most liberals favor."
 
 
If the Nordics are any guide, the US has a lot of room available in its income tax level. Without ever touching consumption taxes, the US could add something like 10 to 15 points of GDP to its tax level, which in dollar terms is $1.8 trillion to $2.7 trillion per year. By comparison, estimates of the cost of universal pre-K put it somewhere around $15 billion per year. Finding ways to fund something like that is laughably trivial, which you could probably guess by the fact that a small excise tax on a single product is apparently enough to fund it in Pennsylvania.

 

Comments