change. Yet the cost of achieving it will fall. Firm As costs will rise by $10, the marginal cost of the 51 % level of abatement. Firm Bs costs will fall by $15, the marginal cost of the 50% level of abatement. Therefore, the total cost of abatement will fall by $5 ($15 - $10). We are not finished saving money yet. Let fu-m A abate another percent more to 52% and let firm abate another percent less to 49%. As before, the total amount of abatement is unchanged. Yet, the total cost of achieving it has fallen, this dme by $3 ($14 - $11). Repeadng the procedure by moving Firm A up to 53% abatement and by moving Firm dovvn to 47% vdll save another $1 ($13 - $12). At this point, the marginal cost of the last unit abated is equal at $12, and our total saving comes to $9 ($5 + $3 + $1). No alternative

combinadon of abatement levels can achieve die inidal 50% level of abatement at a lesser total cost For example, if Firm A were to abate by another percent more and Firm were to abate by another percent less, we would save $12 from Firm B, but at a cost of $13 to Firm A, causing our total cost to rise rather dian fall. We can summarize all this by saying that the total cost of achieving a given amount of abatement is minimized when the marginal cost of abatement is the same for all firms.

Under the standard, society spends more than need be to achieve the mandated level of abatement. In other words, we coidd have the same level of cleanup at a lower total cost. Making the same point a slightly different way, we could have a greater amount of cleanup at the same total cost by a reallocadon of the abatement burden between firms. Either way we and up with inefficiency because weve left unexploited opportunides for making others better off without harming anyone else.

All of this makes good intuidve sense. Under the standard, nobody specializes in abatement; everyone produces the same equitable 50% cleanup, even though the opporturuty costs of doing so are far from equal. On the other hand, if the abatement burden is allocated efficiendy, the lower cost company, Firm A, specializes in it As we saw in Secdon I, specializadon increases the output of all good things, polludon control included. After all, where would we be if we shared the burden of growing food "equally" as we do vddi the burden of supplying polludon control?

In any case, a tax of the right size on emissions can achieve efficiency automatically. Suppose the tax is set at $12 per unit of emissions. Then we can con-

Fig. V An Emission Tax

Tax=$i2

44 45 46 47 48 49 SO 51 52 53

Abatement

sider the behavior of the two firms of Table I in Figure V. To make the analysis simple, lets assume that in an uncontrolled enx-ironment both Firms A and would emit 100 units of polludon. Naturally, as profu maxi-mizers, both would seek to avoid the tax by controlling their emissions.

Consider first the behavior of Firm B, and lets jump into the middle of the story by starting with the forty-fourth, rather than the first, unit of abatement. If the firm fails to abate the 44th unit of emissions, it will pay a tax of $12 for not doing so. Therefore, the firm would be willing to abate that unit provided it can do so for any cost less than $12. Looking at the marginal cost of abatement curve, we see that it costs only $9 to abate that unit. Hence, the firm can save itself $3 ($12 - $9) by abadng it, and so will do so. Now consider the 45th unit. If abates this unit, it can save the difference between the $12 tax and the $10 marginal cost of abatement, so it will surely choose to abate. By stepping up abatement by sdll another unit to 46 units, yet another dollar is saved ($12 - $11). Only when abatement reaches 47 units, where the marginal cost of abatement equals the polludon tax, will the firm stop abadng. For example, the firm would not wish to abate by 48 units; doing so would add $13 to its cost at the saving of only $12 in taxes. To summarize everything, the firm will condnue to abate undl the marginal cost of abatement just equals the tax on polludon.

Firm A would likewise abate undl its marginal cost of abatement equals the $12 polludon tax, in this case at 53 units. As a result, the low cost company. Firm A specializes in abatemenL Moreover, in equilibrium the marginal cost of abatement is equal for both firms at $12. But as we just saw, equating the marginal cost of abatement among firms, minimizes the total cost of producing a given level of abatement. The polludon tax, unlike the standard, is in this respect efficient.

THE THEORY & ROLE OF GOVERNMENT

While in principle government intervention can correct inefficiency in markets characterized by positive and negadve externalities or public goods, it does not necessarily follow that in fact the government either can or will do so. As we have stressed throughout, the government (Congressmen, BureaucraLs, the lobbies, etc.) has interests of its own. Economic efficiency is, however, rarely one of them.

While a complete theory of government would doubtlessly explain the governments behavior through the interests of those involved, work on this subject by economists is in its infancy. Nonetheless,

some striking differences between political decision making and market processes have emerged from current research.

Decision Making. A consumer in the market for a car can pick the car and options of choice down to a considerable degree. If a consumer doesnt want a particular feature or option, he can usually reject it without rejecting the whole can When the same consumer votes, however, he selects the single politician who promises the best overall package of promises, even though many features of the selected package may be quite disagreeable. Nonetheless, he accepts such undesired features as part of the cost of attaining the desired ones. Political decision making thus allows considerably less discretion than prirate decision making.

Protection from Fraud. A consumer who buys a car based upon false representations made by a salesman may be able to sue for damages. Less recourse is offered for fraudulent political promises. Politicians, after all, have exempted themselves from suits for fraud.

Information Costs. A consumer buying a car has strong incentives to learn the characteristics of competing models. It pays to bear these search costs because the consumer knows hell be buying the car On the other hand, it may not pay to acquire much information about politicaLchoices, since.th e. probabil-ity that a consumers vote will swing the election is nearly zero. Moreover, because information is cosdy, voters have litde incentive to look beyond the obvious and superficial effects of programs toward the more important but hidden and deeper effects. As Aveve seen, rent controls, input price controls, die windfall profits tax, the two-tier crude pricing system, and minimum Avage laws are examples of programs with surface appeal but litde substantive merit. Unfortunately, a complete list of similar programs would prove almost endless.

Oftentimes the government disguises the real nature of its acdities under some misleading tide or another For example, "Conservation of Agricultural Resources" is the current euphemism for Agriculture programs to keep farmers from growing crops. And if you want to find how the federal government subsidized the 1984 Olympics, you will have to look in - of all places - the Department of Defense Budget. Of course, the higher these information costs, the less people are willing to acquire information.

Performance Evaluation and Discipline. In a competitive industry inefficient firms go bankrupt and disappear. Government bureaucracies, on the other hand, are shielded from competition. Therefore,

there is no ready yardsdclc for measuring whether a government enuty does a good or poor job. Moreover, the compeddve forces which exert disciphne over private firms are absent for the government. UnHlce private firms, which are threatened with bankruptcy if they perform poorly, the government will always be there whether it does a good job or not.

Perverse Incentives. Stockholders attempt to give management financial incentives to manage the company well. Mimagers of government bureaucracies, on the other hand, may have incentives to maximize the size, budget, and influence of their departments. Such personal incentives militate against efficiency.

Concentrated Interests. Politicians solicit campaign contributions in exchange for promising to support the interests of the donors. For example, UCSB students, the oil industry, the teachers union, or the

farmers, etc., will invest time and money heavily in candidates who support the special interests of these groups. Those harmed by the resulting special interest programs do not form opposing organizations because their personal stake is too small and because it is often too cosdy to study the potential results.

Of course, none of the foregoing explains the behavior of the government in the same thorough way that the theory of the firm can be explained. Nonetheless, at the very least, it does point up the error, common in the past and still often seen today, of treating the government as if it had the potential for being a benign dictator dispassionately acting outside the economy to correct this or that blemish in the market system. In brief flaws in the market system are not necessarily "problems for the government" any more than flaws in the government are necessarily problems for the market. In the 20th century the government is

1. Coase, Ronald H., "The Problem of Social .-Journal of Law tc Economics (October, 1980)

2. Johnson, Op. cit.

3 Nortli, Douglas. "Private Property & The American Way," NaUanaLEfildcw 0" 8, 1983)

4. Friedman, Milton & Friedman Rose, The Tyranny of the Sutus Quo (San Diego: Harcourt. 1984) pp 20-21