Scott Sumner on Taxes
I had urged him to write up an “explainer” type piece, summarizing key lessons on taxes. He came through. So far, his is the single best piece I’ve seen on the issue of border tax adjustment. I am working on some blog posts / podcast episodes to walk through these things as well.
I’m not going to bother excerpting from Scott’s piece, because for its full punch you need to just read it. He starts out with the tax incidence stuff–does it matter whether the government levies a 25-cent tax on the gas station or on the consumer? Then he ends up showing how a tax on imports coupled with a subsidy to exports won’t significantly change the volume of trade, so long as the currency adjusts. *That* is the key claim that guys like Krugman and others are citing, but nobody has slowly spelled out why it works.
To reiterate, Scott’s piece is the single best one I’ve seen so far, if you want someone to explain that initially counterintuitive result.
P.S. Scott is still wrong about the Fed.
Sumner made a pretty shocking concession in this post.
Bravo! Dr Murphy’s disclaimer at the end wins.
There’s a lot that disturbs me about Sumner’s approach here. Lack of humility mostly, but also all the things that economists are really very sure of but haven’t been verified by observation. For example, Sumner boldly declares that if the US government changed the 50/50 split of payroll tax between employers and employees then wages would neatly adjust… it’s good theory, I see where he is coming from, but no government has ever tried the experiment and it seems highly unlikely that any government ever would. For what it’s worth, in Australia the effective equivalent tax (they call it “Superannuation Guarantee Levee” should anyone be interested) falls 100% on employers and yet wages in Australia for many jobs are higher than in the USA (the opposite of what Sumner predicts) but then cost of living is also higher in Australia, and there are different taxes at work (e.g. GST in Australia which is same as 10% VAT). Mind you, some employers will quote wages with the Superannuation component included, which is a bit confusing, then you need to spend the time arguing it out.
I guess my point is that even on the things we are very very sure about, there’s still a lack of empirical confirmation. You understand, this makes me skeptical. I’m pretty sure Sumner would be the guy to tell me we are very very sure that humans are causing Global Warming, and if I was in any way doubtful then I should look at all that scientific evidence (i.e. measurement) because only measurement can settle such a question. BTW: I do look at the measurement, I just question the adjustment and homogenization process but that’s another story, at least from the standpoint of methodology Sumner is basing his entire approach on theory, while not making that clear to his audience.
Sure, in theory floating currency can always adjust to settle out any balance of trade in the long run (regardless of tax, subsidy or whatever). Problem is that we have a practical example in the USA where over quite a significant chunk of all our lifetimes this theory has not worked (including today it still isn’t working). I cannot prove that your “long run” conclusion is wrong (because I know I won’t live forever) but I am tempted to ask whether the theory is useful if by “long run” you mean an indeterminate length of time, well beyond any horizon we can meaningfully conceive of or operate within.
Sumner talks about a “dark matter” debate, and maybe he’s right that USA is getting money direct from foreign investments… but that does not explain why every year the government needs to sell more Treasury Bonds to foreigners. Subject for a different blog post, but there it is.
Finally from the end of Sumner’s post:
OK, if you believe in the concept of freedom and therefore believe in free markets as a structural device to help deliver freedom, then you kind of need to accept that the agents within said free market are acting in their own interest. If you don’t accept this, and you believe these agents are “irrational” then there’s no way to build an operational free market. You might possibly argue that some small percentage of agents are defective in some way (perhaps drug addiction, inadequate education, unfortunate accident causing brain damage, what have you) but when Sumner says “shows just how irrational we are when it comes to trade” this does seem to completely contradict any possible market structure that Sumner wants to propose (other than central planning which demonstrably does not work).
Sumner’s example following this is highly disingenuous (I hate that word, but it’s the only one that applies, open to better suggestions). Typical million dollar house in Sydney is approx 80% land value and 20% house value. Put the exact same house at Cooper Pedy and you will find you can’t sell it at all, because very few people want to live there. But wait, what is it that makes land in Sydney so valuable? Hmmm, all sorts of stuff: infrastructure, availability of jobs, amenities, stable political system with good rule of law, people who almost never ping you with sniper fire from the rooftops (don’t laugh, that’s a handy thing when raising a family), pleasant social environment, nice weather, stylish harbour, good reputation… none of which was built in the same way that blue collar workers are building jets in factories.
The world is not made of convenient, easily measurable economic units.
Tel, the concept of tax incidence is know purely from the logic of supply and demand and not subject to empirical verification.
This is a great comment.
WRT that last part, I would make 2 observations —
1) It’s probably useful to consider that imported goods arriving in developed economies from less-developed economies (or, generally, all imported goods) as *unfinished goods* — because they haven’t yet been delivered to the consumer. So — just how much value is added in that last stage of production, especially crossing that border? If we look at global markets, holy cow, it’s a lot, at least for goods moving from less developed economies to more developed ones. Like your house in Cooper Pedy getting loaded on a trailer and dragged to Sydney.
Where does that value come from? Who exactly is creating it, and who benefits? If other ‘value creation’ activities are taxed, and this one isn’t, what’s to come of things? Hmmmm… Obviously, this isn’t as much of an issue between economies which are relatively similar, but between very different ones, it seems at least something to think twice about. One of the better arguments for capitalism is that by attaching incomes and ownership to sources of value, you get rationing/economization, which overall makes everyone better-off. But clearly we know there are all sorts of income generating mechanisms which don’t fit well in this regime, which would seem to be the case here. And this is clearly a fairly large and important one. How is that to work out? Tragedy of the commons, anyone?
2) One of the questions not addressed by Sumner in this particular post (or most people, for that matter) which maybe it might be a good idea to ask is what one ‘gets’ — in a rather direct sense — for paying one’s taxes (i.e., why pay them, as opposed to not paying them?) I’m not talking about ‘government services’ here — I mean really, what do you get in the most immediate sense? The quick (and wrong) answer is ‘not getting fined or thrown in jail.’ But I think it’s pretty obvious that you can do a lot of things and not get taxed or thrown in jail, you just more-or-less can’t earn any cash income.
But you can certainly grow oranges in your back yard, and consume them yourself. You can do lots of things economically. But the main thing you can’t do, generally speaking, is have access to the division of labor. For the most part, as soon as you do that, you start paying taxes. A lot of things start making sense when you think of it that way — graduated income taxes immediately come to mind. People in high income brackets sure seem to tolerate some really high taxes without being discouraged from producing, but that totally makes sense if their benefit from additional access/engagement is very high…
So, if taxes are effectively ‘access taxes’ for the division of labor, at least for the most part, what is the effective policy of heavily taxing access to the most value-creating divisions of labor in the world for those who are largely responsible for having created and sustaining them, but not those remote and external to them?
Hmmmm….. I’m totally down with the Austrian comment ‘well, this is just an argument against income taxation in general.’ But — if we *are* to have taxation…well… then what?
Anyway… my 2 cents.
Scott Sumner says:
“So far I’ve been ignoring trade deficits. When we run a trade deficit, it means we import stuff now in exchange for exporting stuff later. If it seems like we’ll never have to pay for the imports (as trade deficits seem to run on year after year) it’s because the accounting is flawed. To some extent we are paying for imports by earning big profits on overseas investments. That’s the whole “dark matter” debate, which is worth a blog post on its own. Or we might give them some of our property. But the key point is that we must give other countries something of real value (not just paper) for all the cars they send us, unless the other countries are essentially giving us Lexuses and BMWs as gifts.”
Bob, you’re the king of analogies so here’s one you may appreciate:
“Tony Soprano runs a deficit with every one in his neighborhood, he receives clothes, paintings, meals in restaurants, merchandise, cars, etc. etc. If it seems like he’s never have to pay for the imports…I mean, the products…it’s because the accounting is flawed. Tony Soprano must give everyone else something of real value (not just paper) for all the cars they sent him, unless other people are essentially giving him Lexuses and BMWs as gifts”
There was a reason Economics was once called Political Economy. To even begin treating U.S (Russia, China) trade balance without adjusting for very difficult relations between countries is beyond naive.