Are Pigovian Taxes the “Answer” to Environmental Problems?
All environmental problems arise from the existence of “externalities.” Many of the common worries about water quality, air quality, etc. result from negative externalities – value producing activities generating costs on unwilling third parties. Some worries are also the result of positive externalities – because free-riding is difficult to prevent for certain classes of goods. But any action that mitigates a negative externality can be considered a positive externality in its own right. As such, we do not need separate theories to deal with each of these cases. Considering this there are several reasons why the canonical “solution” to dealing with environmental problems (e.g. Pigovian taxes and subsidies) does not exactly follow from the theory. A Pigovian tax (roughly) is one where policymakers tax the activity that produces the externality in an amount equivalent to the amount of damages being caused by that activity. It is an alluring and intuitive idea, but per usual in economics, the devil is in the details. Here are some reasons to be cautious in the application of Pigovian taxes. Each of these is worthy of several posts, but for now I would like to merely list them.
- The Coase Theorem: So long as transactions costs are low enough, a simple assignment of property rights to the open-access resource will be sufficient to get us to an efficient pollution level (the distributional consequences will vary however). The Pigovian solution is, in fact, just a special case of the Coase Theorem. To talk of them as being different beasts is incorrect.
- Identifying high transactions costs is not enough to get a tax/subsidy solution to work. Pigouvian taxes are only efficient solutions in the special case when the transactions costs of negotiation between affected parties are prohibitively high AND when the taxing authorities levy fines on those parties that are the least cost mitigators of pollution. For example, if a steel factory’s emissions are causing a problem for some people downwind, it is not clear that asking the steel factory to reduce emissions is the best or even a desirable way to deal with the problem. Note to people who think I am some market zealot who ignores the problem by saying this – my comment does NOT imply that the steel firm should not be financially responsible for the cleanup (nor does it imply that it definitely ought to be), it is a comment about what mechanisms are best for everyone irrespective of distributional issues.
- It is “efficient” to have some pollution and degradation. Just pointing to a cost does not imply that the cost should be eliminated. The point of Pigovian policy is to get us to the point where the marginal social benefits of the last unit of an activity equal the marginal social costs generated from the production of that last unit. In a world where knowledge is tacit and dispersed knowing exactly what the social costs of an action are or the social benefits is a heuristic process. In other words, it is not clear that policymakers would be able to find the right tax (or output level). This does not mean that they could not get close, or do a better job as they learn more.
- Just pointing to a positive or negative externality does not imply that policy should do something about it. Information is costly to obtain and cost-benefit analysis requires that we direct our attention to the policies that generate the largest net benefits. Moral and ethical considerations matter too. For example, if Wintercow and his wife go on vacation for a week, they are more charitable to their neighbors when they return. It does not follow that society ought to subsidize the Wintercow family’s camping trips just so they produce an efficient level of this activity.
- Not all external impacts work in the same direction. Why? Because people and ecosystems are heterogeneous. Some goods that generate positive externalities will also generate negative ones and vice versa. For example, my planting of flowers is generally a benefit to my neighbors. But some neighbors undoubtedly prefer shrubs and trees only, or get allergies from the pollen in my flowers – so the case that my planting of flowers is an action that generates positive external benefits is muted – what we care about is the NET external effect of my actions. Similarly for actions that produce negative costs, the canonical size of the damage is often too large if we focus only on the costs of activities. For example, CO2 certainly causes problems, but it does improve the ability of certain agricultural products to grow – a benefit of me emitting CO2. So simply pointing to the negative impacts of CO2 will overstate the amount of necessary mitigation.
- Not all externalities are relevant. In particular, for a certain class of goods that are nonrival in consumption (i.e. my consumption of it does not inhibit your ability to also consume it – such as looking at a sunset), the conditions for optimization are far less difficult to achieve than what would be suggested from a theory designed to discuss a different class of goods. For example, for goods that are nonrival in consumption, the number of parties that are required for negotiation to reach optimal solutions are substantially smaller that in the classical case.
- Even if we are in the condition where a tax / subsidy is the appropriate policy – remember that when we present a theoretical model of externalities that we are assuming a level of taxes and subsidies or zero. But in reality, we already tax economic activity extensively in the US – by somewhere in the $5-$7 trillion dollar range in direct taxes and dead weight losses due to taxes. In a $14 trillion economy, that is up to 50% just from taxes and deadweight losses alone. On top of that, throw in the hundreds of billions (trillions?) of added costs due to environmental and other regulation, and it is not clear that an efficient solution would call for a reduction in many taxes.
- The canonical theory presumes the existence of perfect political solutions. The lesson of public choice economics is that political actors are as self-interested as are economic actors. As such, regulatory failure is quite common. Much of environmental policy in fact turns out to NOT be about the environment at all.
- Even if there are no public choice problems, political solutions will “fail” for precisely the same reasons that market solutions might “fail.” Quis Custodiet Ipsos Custdodes. Who watches the watchdogs? In other words, to correct a market failure, government action is often called for. But this action also requires monitoring and the creation of an incentive compatible mechanism to ensure good outcomes. But this monitoring activity is also a public good – and the canonical solution is for the government to step in and provide this service (because it would otherwise be underprovided). But therein lies the problem – we have the government overseeing itself who should be overseeing itself … and it is turtles all the way down.
- Just pointing to the existence of an externality does not mean that these things will not be overcome/solved absent government action. Many public goods have, and continue to be, provided by private non-coercive alternatives. In fact , the 2009 Nobel Prize in economics was awarded to two social scientists that have studied and recognize the vast array of voluntary cooperative actions that citizens around the world have come up with to overcome the free-rider problem. Markets are but one small option among the myriad private non-coercive actions that people have chosen. In fact, among the goods most commonly cited as being intractable public goods problems, many have in fact been solved privately. Several illustrations are possible, and here they are without analysis:
- Contracting and property rights solutions such as those described in Cheung’s the “Fable of the Bees.
- Packaging solutions such as how bundling lighthouse fees with port use fees solved the lighthouse problem (and many ultimately were built and run privately
- Charity: many public goods survive and are produced because of voluntary contributions from high demanders. E.G. environmental organizations, art galleries, hiking clubs, college, etc
- Evolution: people have evolved to be more cooperative than theory might suggest we are
- Price discrimination: allowing producers to charge each customer his maximum willingness to pay can overcome the high-fixed cost, low marginal cost problem associated with many public goods.
- Contracting innovations: makes it credible for producers to withhold production of a good unless contributions are made
- Strategic anti-free riding: many people do not free ride simply because they feel that they would be subject to free riding behavior in the future … and many more possibilities.
OK, so back to our list:
11. Technology can quickly change the “publicness” of a good. In some cases, it can work against the interests of the environment but in many it can work in its favor. This technology will not have the incentives to be developed if government policies distort prices and rewards for developing technologies that improve our ability to exclude non-payers from accessing open-access resources.
12. Transactions costs are just an economic cost like any other cost. Why are transactions costs relevant? How are they different than any other cost that might prevent beneficial transactions from emerging? By pointing to a cost that is not permitting otherwise beneficial actions from occurring, does that mean government tax and subsidy policy should intervene in all those cases? For example, until recently, high transportation costs made it virtually impossible for me to get goods and services from certain places around the world. Would that be justification for subsidy and tax policy to overcome this “failure”? Or until recently, the costs of producing MRI machines exceeded the benefits – so in 1900 would you point to the non-existence of MRI machines as a market failure? In fact, would you then point to the entire history of mankind (average income for thousands of years was roughly $800-$1000) as a gigantic market failure? It would seem an odd argument when in fact it was market institutions that emerged to reduce the costs of securing the goods and services that we all wanted.
13. While the government has indeed had some success in solving some environmental problems (e.g. eradicating smallpox, municipal sewer systems (at least until recently), SOx emissions reductions), its record in general has not been good. In other words, public provision of public goods is no guarantee that the good will be provided adequately. A minor glimpse into this record is to recall the stories of Ravenna Park in Seattle, the Kesterton Wildlife preserve in California, the poor record of parts of the Endangered Species Act, the relatively poor management of national forests as compared to state and private forests, the Clean Air Act, etc.