Controlling Externalities

 

Externalities (“neighborhood effects”, spillovers)  = the actions of one agent affect another agent’s welfare by means other than through trade.

 

Economists who paid central roles in developing this topic:  Sidgwick 1887, Alfred Marshall 1890, A.C. Pigou 1920 (Supply & Demand graphs with marginal damage curves), Ronald Coase 1960 (control via property rights), Kenneth Arrow 1972 (markets for pollution)

 

Negative externalities = Pollution, noise, ugliness (Prairie Moon?), emotional distress

 

Positive externalities = Beauty, (good) music, clearing swamps to prevent mosquitoes.

 

Private agents will produce and buy up to the point were Private Marginal Benefit = Private Marginal Cost.  (At this point there is zero dead weight loss.)

 

Ideally, society chooses a quantity to produce where: Marginal Social Cost = Marginal Social Benefit.

 

Note:  We can find the socially optimal quantity be either adding MD to MC, or subtracting MD from MB.  Either gives the same result for the optimal social quantity produced and sold. 

 

How do we control externalities to minimize DWL?  We have made great progress on this in the last thirty years. 

 

Six Methods for Controlling Externalities

1) Command and control, Regulation - Set pollution to zero or some other low level by regulating how much firms in an industry can produce, or how they produce it. 

If regulators can only observe output, how would they mandate cutbacks in production to reduce pollution? Optimally, firms that can reduce pollution cheaply in terms of lost output (dirty plants) should do so, and firms where the loss in output would be high for a given pollution reduction (clean plants) should reduce less.  Clearly in such a case it would be sub-optimal for both firms to cut back by the same amount or percentage.

What difficulties might arise in running such a system?  Would getting firms in concentrated industries to jointly reduce production lead to more oligopoly profits?  Would each firm that tried to enter the market have to go to the regulators to find out how much it would be allowed to produce? If entry would reduce prices and profits for firms in the industry, incumbents in the business might even press for more stringent standards to keep firms out of the business. This has actually happened.

2) Pigouvian taxes - Unit tax on firm output equal to the marginal damage created by the firm's output. This will increase marginal private costs to the point where they are equal to marginal social costs. (Pigouvian taxes were worked out by A.C. Pigou.  Arthur Cecil Pigou, who you just know got beat up in gym class every day.  At Cambridge, he was a rival of John Maynard Keynes.)

Such a tax would increase the price of the "under-priced" social resource, and simulate a market where the market fails to exist due to a lack of property rights. In effect, the tax attempts to serve as a price for the un-priced input (such as clean air), as judged by the quantity of output.  It would probably be better to tax the pollution created directly but that can be very hard (expensive) to do.  As such, since we can’t tax the externality directly, this is a “second best” solution.

It is always difficult to calculate marginal damage, particularly when the marginal damage may not be a linear function of output as it is in this example. The optimal Pigouvian tax would be equal to the marginal damage that would be created if the firm were producing at its optimal level of output.

One very nice feature of Pigouvian taxes is that they bring in tax revenue (although these taxes may be expensive to enforce). These taxes can even be used to encourage beneficial externalities as well. Evanston could put a tax on people who do not plant flowers in their yards.

3) Pigouvian subsidies - Pay unit subsidy to firm for each unit it reduces its output. The subsidy would be equal to the marginal damage created by firm output to encourage firm to reduce its production of the pollution causing good. 

This would increase the opportunity cost of producing goods.  A firm would know that for every unit it produces, it would lose (say) $5 in subsidy.

This is a pretty stupid idea really. It is expensive to come up with the revenue for these subsidies and this increases taxation DWL in other markets. It may cause firms to enter market just to get the subsidy. This effect could be reduced in the short run by "grand-fathering" in and limiting the subsidy to the original outputs of incumbent firms in the market. Certainly, firms in the industry prefer Pigouvian subsidies to Pigouvian taxes.

There are government sponsored "buy-back" programs for guns that are examples of a Pigouvian subsidy.  (LaToya Jackson held some concerts where a person could be admitted by turning in a gun.  I wonder if many of the people who turned in guns for those tickets later regretted not having a gun when LaToya began to sing and there were long lines at the exits.  I wonder if any of them where unsatisfied with the concert and demanded their guns back.  My guess is that the real criminals probably held on to their weapons anyway. Oh well, at least there are fewer stray guns around for kids to use as playthings. Illinois considered a buy-back program to get high-polluting old cars off the streets. However, they realized that any price they would set would bring lots of currently dead cars out of junk yards just to get the subsidy. The supply of cheap junk cars is highly elastic. It is not even clear that this is a good policy. It pollutes to make a car and many of the old junkers are driven very little. So if people turned in their old cars and bought new ones, pollution might even go up.)

4) Auction Pollution Permits - Calculate how much pollution is acceptable and sell rights to produce this pollution. 

The Clean Air Act of 1990 created a market in pollution rights. Some people see buying a right to pollute as being akin to buying a right to murder someone. However, it has its advantages. It is a flexible system where quantities are easily altered. Market based approaches are nice in this way, environmental groups could just buy up these permits and give us zero pollution (except for Volvo's). As compared to strict regulation, this program can shelter jobs from changes in environmental regulations and allow time for adjustments to be made.

However, it is very hard to monitor compliance and so far works only on the pollution from firms. To date, this approach does nothing about the biggest polluters of all, automobiles. If policing can be worked out, this system may make for a practical method of organizing international coordination on pollution. It is very tough to set regulations and taxes between nations but buying and selling is quite common. Rather than having the rich countries tell the poor ones not to pollute, the rich countries could pay them not to.

If set to the optimal amount and adequately policed, this system will drive the price of this "under-priced" good to its "true" value, which should be equal to the Pigouvian tax.  Making the firms pay for the pollution they have to "buy" would shift their supply curves upward. If the shift is too little, DWL will remain. If the shift is too far, DWL will be created.

One potential problem with this sort of policy is that it is very sensitive to local concentrations.  Reducing some pollution in fairly clean Kansas with a tradeoff of increased pollution in dangerously smoggy LA is a bad outcome. 

Examples

In 1992 Mobil Oil paid $3 million for the right to emit a certain amount of pollutants in Torrance, California. General Motors sold the rights to Mobil after it closed down its plant in the town. GM's saleable rights were less than the pollution it previously emitted and the profit from the sale was an incentive to close the old, polluting plant. Commodities exchanges like Chicago's Board of Trade have begun selling rights and futures for these markets. By 1995, the Clean Air Act will force all power plants in the US to pay for the right to emit sulfur dioxide. The credits are denominated in pounds of pollution per day and typically sell for $1,000 to $4,000 per pound. This has become a major expense of opening a new plant.

In 1992, the Tennessee Valley Authority paid nearly $3,000,000 to a Wisconsin utility for the right to emit 10,000 tons of sulfur dioxide. This is the cheapest way for the two utilities to cut the cost of their pollution. This system also makes pollution reducing devices like stack scrubbers financially attractive to firms.

 

5) Internalized the Externality  Economically connect the producers of the externality with those who feel its effect.  For instance, get the up—stream polluting paper mill to merge with the down stream fishery.  Then losses to the fishery will hurt the company and so it will have incentives to try to reduce the pollution from its paper mill.

 

6) Property rights, Coase Theorem

If transactions costs don't overwhelm gains, saleable property rights will bring social resources to their socially efficient uses, even when there are externalities. Social welfare will be maximized. This is a simple variant on the pollution permit auction.

 

Questions, comments, typos?  mwitte@northwestern.edu