ARNOLD HARBERGER ON TAXATION by Ihering Guedes Alcoforado

“ Arnold Harberger: “That is the way I think. If you read the different articles that I have written about taxation: We put on one tax and we get a triangle in its market; now we put on a second tax and we get a triangle in this second market, but we get a new rectangle in the first market, or a parallelogram or a trapezoid. Then we put on a third tax and we get a triangle there, plus two rectangles on the goods with the first two taxes, and so on. This is general equilibrium. As in a reduced-form situation, the instrument that we’re playing
with is the tax rate. We can think of the tax rate as being imposed, like a 10% tax or, more elegantly, squeezed in as in a line integral. You squeeze up a tax rate to generate a triangle in its market, but you can get rectangles or trapezoids in related markets if they have preexisting distortions. As we change the price of beer, we have to look at what happens to the demand for wine. If the beer price is going up, then the demand for wine will go up. At every point in the raising of that tax on beer, the person who jumps over to
wine is just barely indifferent at that point. It’s an elegant, sweet concept, and I think of this as something that people absolutely have to understand in order to understand applied welfare economics well.” [HARBERGER, p.10

ARNOLD HARBERGER: The difference between the partial and the general equilibrium levels — “[…] the difference between the partial and the general equilibrium levels is not in the size of the triangle for a given taxed commodity. Rather, it is that in the general equilibrium case you take account of preexisting distortions that are affected by your move, and if you’re just doing partial equilibrium, you don’t. It’s very tricky. If you look at the tax on telephone calls, you put the tax on telephone calls and
look at the telephone call market, and you’ve got this triangle. When I’m teaching a class, and I have that triangle, I’m telling my students, “If the only tax we have is the tax on telephone calls, this is what it looks like; it stands alone.” Another way to say this is,
“If this is a general equilibrium situation, and we have a tax that has ramifications in other distorted markets, for us to be left with just a single triangle, these other ramifications must cancel out.” That’s a perfectly reasonable thing to say to the class at that level.
But that same class should be taken through an analysis that says, “When we put this tax on telephone calls and we have an already existing tax here and another already existing subsidy there, this is how we are supposed to take those ramifications into account.” Then somebody says, “These things can be infinite.” I say, “If you’re a damned perfectionist, you’re probably right because the dXi =dPj are probably not all going to
be precisely zero.” But to be practical, you would take into account those places to which people’s demand shifts when they reduce their demand for telephone calls. What are the three or four most important places? Which of these three or four most important places has a significant distortion? Maybe only one, maybe only two; so we take into account the intersection between the items whose quantities are significantly affected by the tax you’re imposing (or the other disturbance you’re creating) and the set of items
with significant preexisting distortions. You don’t need to catch all of them; you only need to deal with those that have significant distortions.” [HARBERGER, 2012:11]

ARNOLD HARBERGER: The monopoly markup as a private tax — “The monopoly markup is really just a private tax […] Life is tough. At one level, there are many, many, many entities that have some degree of control over the prices of their products. That is to say, they face somewhat-downward-sloping demand curves. We know better than to think that they look only at their existing competitors when they estimate the long-run demand curve facing themselves. They consider potential entry, as well as potential expansion of production by existing competitors, in determining the elasticity of that demand curve. If you want to think about monopoly power in the company world, I would say, “Let’s look at profits that are measured. Look at the return to capital. Let’s measure profits — everything, and say how much of that can plausibly be thought of as monopoly profits.” You’re not going to have a technical way to measure that. But I think you can say a lot of things.
You can identify some industries as pretty competitive and figure that the median one of them is earning a competitive rate of return. If somebody’s earning more than that rate of return, the difference is very likely the maximum amount that you would assign as monopoly profits. Then what you want to do is take those monopoly profits and spread them over the whole value of the goods and services produced by the entity, because monopoly profits are not a proper part of the return to capital at all”[HARBERGER, 2012:16]

“ They’re effectively the return of an excise tax, so they ought to be treated as we would an excise tax by spreading them over the whole value of the product. When you do that you get a lot of insights about the economy, and this is somewhat important in deriving our economic opportunity cost of capital for cost-benefit purposes. If we think there’s a lot of monopoly in the economy, we ought to make that spread. In making that spread, you can spread the monopoly markup over the full value of the product, or
you can pretend that materials come in fixed proportions to the product. If you do that, you can spread monopoly profits just over the return to labor and capital, which makes it simpler. In that case, you’re just dealing with two big factors of production, and you can handle the problem more nicely” [HARBERGER, 2012:16/17}

“ Definitely, the way I would go about looking for monopoly elements in an economy is: I would look at the business sector of the economy. I would look at their profits and see if they are substantially and repeatedly in excess of what that economy might reveal as a competitive return. If you have an invention or a new product, it’s a monopoly all right, but it isn’t something that you want to fight particularly, like new drugs.”[HARBERGER, 2012:17]

For example, in the drug area, we seem to want monopoly profits to prevail for awhile, for we think of them as a good incentive for innovation in the drug and other industries. That doesn’t stop them from being monopoly profits, but it may stop you from doing what you might otherwise want to do about the monopoly profits [HARBERGER, 2012:17]

HARBERGER, MONOPÓLIO DA INDUSTRIA FARMACEUTICA E PATENTES “Absolutely. There you’ve got a real problem (ots of people who could be helped by a particular drug that they can’t get access to because of a 90% profit margin on some patented pharmaceutical). If we wanted to nationalize or internationalize the pharmaceutical industry, would we be getting better results than what we get with what we’ve got? Again, the length of time of patents is another story. It seems to be a very good thing that in the drug industry, these patents do in fact run out, giving us generics all over the place. It seems to be a question of time before
they get to be cheap, and the issue to which you refer is something within this time frame.”

ARNOLD HARBERGER: The Benefit-cost analysis can be applied to taxes and subsidies — “The Benefit-cost analysis can be applied to taxes and subsidies and what not, as well as to projects. From that point of view, I think of our story as centering on a profile of benefits and costs over time. This profile, itself, is the difference between two moving pictures: One is how the economy would have evolved if we didn’t pursue this investment project (or economic policy change), and the other is how the economy is
likely to evolve if we do make such a move. The profile is the difference between those two moving pictures. It doesn’t have to be a dam or a park (project). It can be a change in the law (taxes and subsidies). You can still make those two moving pictures, take their difference through time, and then take the present value. [HARBERGER, 2012:19]

BIBLIOGRAFIA

HARBERGER, Arnold., A Conversation with Arnold Harberger and Richard Just IN Annu. Rev. Resour. Econ. 2012: 4: 1–26