What you need to understand about taxation and externalities

Economists like to talk about externalities. An externality is just a benefit or a negative which the originator either can’t be paid for (if it’s a positive externality) or whose cost they don’t bear (if it’s a negative externality). Pollution is a negative externality – the cost of air pollution in terms of asthma and deaths is not born by big air polluters like coal plants, for example. Public roads provide positive externalities – they enable commerce and industry to ship goods and gain profits and those who build the road don’t directly receive a cut off the new business they make possible. Another example of a positive externality is good health care – the value of people not being sick does not go to health care providers, but to the employers of the healthy people.

When you’re designing tax policy a general principle is that you want to tax people the cost of their negative externalities. So, if an industry pollutes, they should pay the cost of cleaning up the pollution and the healthcare costs of those who are made sick by their pollution. Not only is this fair, but it makes markets function better because it means the true costs of an economic activity are actually accounted for where they belong. In a pollution scenario the costs of polluting are born by individuals, hospitals, other business in terms of absenteeism, and by governments who pay for health care – not by the polluting company. Since they aren’t paid for by the polluter, that polluter’s goods are underpriced – they don’t include all the costs of the activity.

A good which is underpriced will be overproduced.

This is also an argument for differential taxation. Let’s take railways versus roads. Roads – or rather cars, trucks and so on – produce a lot more pollution than railways. In addition they are, in effect, subsidized by public expenditures building and maintaining the road network.. On the other hand railways move people with a lot fewer negative externalities – fewer accidents, less pollution, less noise. However railways bear much more of their own infrastructure cost because they have to build and maintain their own railways or rent the use of them from other rail lines that do. When you add up all the numbers, rail travel costs less than road travel, but that cost isn’t born equally with people who travel by road. (Trucks, for example, are effectively massively subsidized). The same is true of airline travel, which is subsidized far, far more than rail travel. The cost to society of moving a thousand people or a thousand tons by train is a lot less than the cost to society of a thousand people in car or plane. But because railways are subsidized much less it winds up costing much more. However, since the real costs of rail are less than the real costs of air or road travel, it should cost less. The current market and government regime of subsidies doesn’t do that. Instead it underprices road and air travel and overprices rail travel.

A good which is underpriced will be overproduced. And a good, like rail travel, which is overpriced, will be underproduced.

This is also the argument for subsidizing the arts, universities and so on. There would have been no Massachusetts miracle without universities, no Silicon Valley without universities. Museums, theaters and so on bring economic activity to the communities that host them. But most of the money they bring to a community they don’t capture. Restaurants, hotels, cab drivers, airlines and a host of other businesses receive the vast majority of the money that a museum or a theater district, say, is responsible for. How many people visit New York or Toronto for theater? How much money do they spend? And how much of that do the theaters themselves get? Since positive externalities are produced by those institutions – externalities they cannot capture, it is not unreasonable to subsidize them some portion of how much extra economic activity they bring to their communities.

It is the job of government to produce or subsidize that which has strong positive externalities. The reason for this is that the government doesn’t have to know who is going to benefit, only that someone does. It’s not necessary for government to know which specific businesses will benefit from a good university in town, it is only necessary to know that some will, and that you’re taxing them. On the other hand, in a pure for-profit university, it would be necessary to charge enough people who would benefit. That would mean the students themselves (rather than the businesses who also benefit from having educated workers) and those businesses who are given access to research. The result of this would be fewer students getting a unversity education (because they are paying full price, even though much of the benefit will be received by the businesses who hire them) and research will not be widely disseminated but will be used only by those businesses who can afford to pay. Unfortunately the greatest economic benefit is that research be widely disseminated so that anyone can use it, because you never really know who might be able to profit. All those businesses that could never profit, or that never existed, because they never knew of the new research, are part of the cost of private research.

An entirely private university system in which research is also private, then, produces much less benefit for society as a whole. If everyone had to pay Harvard rates for an education, there’d be a lot fewer educated people. And as we move towards a research system in which private interests pay, and lock up the research, there is less and less benefit to society.

Subsidies are a problem when they go out of whack with positive externalities. If something produces a dollar of positive externalities and you’re subsidizing it for more than a dollar, that’s a problem. More practically when the government can’t capture enough benefit through taxes to pay for the subsidy, it’s time to stop. So if the government has a recapture taxation rate of, say 20%, then you need $5 in positive externalities to pay for $1 of subsidies.

The higher the taxation rate, then, the more subsidies are possible, not just because the government has more money, but because they become economically viable for the government to pursue.

This doesn’t apply to negative externalities, mind you. There is no reason not to tax a company for the full value of its negative externalities, not just a percentage of them, unless it is also producing positive externalities you want to capture. Still, as a matter of simplicity, it is probably better to tax for the full value, then subsidize, so you know what the value of each is.

This probably strikes a lot of people as messing with the free market. And it can be – but done properly it’s actually the opposite. It makes markets more efficient by making them price things properly. A market which doesn’t price in the cost of pollution isn’t operating with all the facts and is thus underpricing whatever it produces. Likewise people who produce a positive externality the value of which they can’t capture deserve to be compensated for making society, literally, richer. If they aren’t compensated for it, they’ll stop doing it.

Originally posted Nov 10, 2007.