In this essay I will explain what externalities are, why they can be problematic, how they can be addressed, the role of government and the potential effects of how governments choose to intervene, concluding that transaction costs are a major determinant of the best policy response to the issue of externalities.
WHAT ARE EXTERNALITIES?
Connolly & Munro (1999) describe an externality as “an action by one agent which affects directly the well-being or production possibilities of other agents, but is chosen without regard to those consequences”. Externalities can be positive or negative. I’m living in oil-rich Bahrain, with irritants such as pollution from oil production, smoke from sheesha pipes, traffic jams from excessive use of motor cars, construction noise in the new apartment block where I reside and nuisance from jet skiing activities at the beach.
These negative externalities resulting from others’ activities are a nuisance, but if those who partake in these activities were to stop, they would lose out instead of me. Conversely, I don’t complain about the pleasant oudh scent in shopping malls which I can enjoy without purchasing anything, night lighting provided by our neighbouring hotel, or the agreeable views of my neighbour’s gardens, beach and marina, enjoyed from my balcony. These are all positive externalities which my neighbours would not produce unless there was some benefit to them, but they cannot prevent me from enjoying either.
WHAT IS THE PROBLEM WITH EXTERNALITIES?
In private transactions, the wider social effects are generally not taken into account. In a Pareto-efficient market, the private and social costs would be equal, but in reality, the market often fails to produce a balance. Private costs ≠ social costs in competitive markets. Private producers will continue production while it is profitable, but, if there is a social cost, such as pollution, the more they produce, the higher the social cost – the more they pollute the environment. Yet, many market activities that create a social cost are allowed to continue; and manufacturers carry on producing goods that give free benefits to others…why?
The goods they produce are still valued in some way and, therefore, to eliminate them completely would not necessarily be an optimal outcome. Instead, there must be a trade-off between production and externalities – a balance must be found between the two. Private enterprises will only continue to produce goods that create a social benefit (positive externalities) to the point where they remain profitable. However, they will not voluntarily take measures to reduce social costs (negative externalities), so some form of intervention is required.
HOW CAN THE PROBLEM BE ADDRESSED?
There are a number of ways of addressing social costs. A lot will depend on how ‘property rights’ are distributed in the individual situation. Do oil production companies have a greater right to pump as much oil as possible out of the ground than the local residents have to a clean and healthy environment? Morally and ethically, you’d think the answer would be “no”, but it’s not as simple as it seems. Without oil production, the Middle East would never have achieved its economic strength and resultant power, yet the region would become uninhabitable if all its resources were given over to oil production. If oil companies owned the legal right to a clean environment, the local inhabitants would have to try to induce companies to reduce production; whereas, if the inhabitants owned the legal right, the oil companies would have to pay them off in order to continue production. The costs in reaching a solution to the externality problem are called ‘transaction costs’.
The Coase Economic Theorem states that “bargaining over externalities will lead to Pareto efficiency, provided property rights are well-defined” (Connolly & Munro, 1999). For example, if a fisherman has a clear right to fish where an oil production company wishes to drill oil, he may receive a payoff from the company – a small price to pay for the right to gain massive profits, while destroying the local ecosystem! However, transaction costs can be so high that they exceed the benefits to be gained by negotiating. The fisherman may decide he cannot afford a lawyer to pursue the oil company (perhaps the legal rights are not clearly defined) and he may simply decide to fish elsewhere instead. Conversely, if the site was near a popular tourist area and several large hotels were determined to protect their livelihood, it may not be feasible or profitable for the oil company to pay them all off. None of the above outcomes would produce an optimum result and that is usually why governments must step in.
POLICY RESPONSES AVAILABLE TO GOVERNMENTS
A visual aid helps demonstrate the effects of taxation levied by governments on production of a good/service which creates negative externalities, in order to better balance social and private marginal costs, termed ‘Pigouvian tax’, after the economist who first examined their use.
Figure 1 – Optimal taxes on externalities. Source: Connolly & Munro, 1999, p.79 Figure 5.5
As output increases, marginal costs also increase, whereas marginal benefits (MB) decrease. The oil company will produce oil up to the point at which their private marginal cost (PMC) equals their private marginal benefit (Q°). Oil production also creates a social marginal cost (SMC) for local residents, which will be above the private cost. If the oil company ignores the social cost, the private optimum (Q°) is above the social optimum (Q*). If a tax (t) is levied on the oil company, this raises private marginal cost (PMC¹).
As a result, the oil company cuts back production to Q*, which is the social optimum. In this situation, how the government uses the money raised by taxes will determine who gains from imposing the tax. If the money goes into general taxation or is given to those suffering from the environmental effects, there may be some resistance from the oil company because they have lost out. If a cutback in oil production was subsidised by the government, this would reduce benefits to oil producers who expand output; optimal output would be the same as if a tax had been imposed on production, but with different distributional consequences.
The problem could be internalised by forcing the polluter to merge with the pollutee. Now they will have to find a balance between the two enterprises/activities, in order to take advantage of positive externalities, or address negative externalities. In an ideal world, private enterprises would internalise the pollution problem, given that they are also inhabitants of the environment, and take measures to reduce these negative externalities.
* Direct controls
Sometimes governments are forced to take direct controls to address a negative externality. For example, in the Middle East, taxation on vehicle use will not be a high enough incentive to reduce the number of cars on the road, yet pollution and traffic congestion is a major problem. Governments could address this problem directly by closing the city centre to traffic, rather than taxing vehicle use.
* Quantity controls
Problems also arise in cases of differential costs. Getting back to my local example, there is more than one oil company in Bahrain. “Direct controls [don’t work so well] when costs differ between firms and it is not feasible to set controls which differentiate between firms.” (Ibid.) Assuming the relationship between oil production and pollution is the same for each company and the higher marginal cost differs between companies, only the total quantity of oil production matters to the environment – society doesn’t care who is producing the pollution.
Figure 2 – Taxes superior to direct controls with differential costs. Source: Connolly & Munro, 1999, p.81 Figure 5.6
Let’s say BAPCO and AGOCO are two oil companies in Bahrain, but AGOCO has a higher marginal cost (MC) than BAPCO, but they both have the same relationship between output and pollution. The level of pollution is related to the total quantity of output, regardless of who produces the oil. The government decides first to impose a tax to restrict output. BAPCO and AGOCO both face the same price in equilibrium P*. AGOCO produces at QH and BAPCO produces at QL. Total output is Q, therefore, Q = QH + QL in this scenario. Government policy changes and quantity restrictions are introduced, so that each firm is allowed to produce Q/2 (half the total output). AGOCO now produces more at Q/2 than it did under QH, whereas BAPCO now produces less at Q/2 than it did under QL. AGOCO’s costs rise by the shaded area ABC, but BAPCO’s costs lower by DEF. The net change is ABC – DEF. Total industry costs rise, however, because ABC > DEF. The total costs for the industry rise when government policy changes from taxation to quantity controls.
UNCERTAINTY AND THE RISKS INVOLVED
The government does not know the marginal cost curves of BAPCO and AGOCO, yet it has to choose between setting a price level or an output level. Figure 3 shows the social marginal benefit curve and potential marginal cost curves.
Source: Connolly & Munro (1999, p.83, Figure 5.7)
Frank (2005) states that “the best laws regarding harmful effects cannot be identified unless we know something about how much it costs different parties to avoid harmful effects.” The actual marginal costs could be high or low, averaging out midway between the two. The optimum level of output (MC=MB) could be at B or G, even though it is estimated at E (average). Let’s say the government chooses to set a price level = p*. Output will be at the point where p* = MC on the cost curve. The estimated point is E, but it could be at A or J – very different extremes! If the cost curve is low, output will be at J, but the optimum output level on this curve is G. This creates a welfare loss of GJK. If the cost curve is high, output will be at A, but the optimum output is at B, creating a welfare loss of ABC. Both scenarios are clearly inefficient. Setting an output level is also risky. Output will be set at Q* but if the MC curve is low, this results in a welfare loss of EGD; and if the MC curve is high, there is a welfare loss of EFB. In this example, the size of the welfare loss is greater when using price controls, compared to using output controls.
Without knowing the actual position of the marginal cost curve, the government will either decide on direct controls or a price level, yet the option chosen may or may not be appropriate for the actual marginal cost curve. However, if the MC curve is steeper than the MB curve, price controls are the better option; whereas output controls are better where the MB curve is the steeper.
Although Connolly and Munro (1999) argue that “uncertainty about the position of the MB curve does not have consequences for the prices versus quantities decision”, a clear trade-off between social marginal benefits and costs is demonstrated by Gruber (2010) where the social marginal benefit curve differs. He uses the example of global warming (where the benefits of pollution reduction are initially high, but additional reductions are not so beneficial) versus nuclear leakage (where each reduction in leakage is highly beneficial and, therefore, a priority).
Source: Gruber, J. (2010, p. 143, Figure 5-10). DWL = Deadweight Loss / Market inefficiency
As global warming is a global problem accumulated over many years, the social marginal benefit of reductions in one country will have little impact, or global social marginal benefit, so the SMB curve is very flat – modest reductions in carbon dioxide emissions yield little benefit to society. Leakage from a nuclear power plant can have catastrophic effects. Any reduction in leakage is highly beneficial, so the SMB curve is very steep. MC1 represents the government’s estimated costs of pollution reduction, but they may well have under-estimated the true marginal cost – it could be much higher, as in MC2.
As with the above example, when the government intervenes on the output side, using its estimated curve MC1, output is set at the “optimum” Q1, whereas, if the actual marginal costs are at MC2, the optimal output should be set at Q2, where SMB = MC2. Here, the social inefficiency is represented by ABC. We can see that setting output levels for global warming has far greater inefficiency risks than it does for nuclear leakage, given the difference in size of ABC in each scenario. The introduction of a tax (price level) is represented by t. Government sets the tax that would result in a reduction in output to the “optimum” level C, whereas the actual outcome could be E as indicated on MC2, causing an inefficient outcome EDB. The inefficiency for global warming is much less significant than for nuclear leakage.
CONCLUSION: WHY TRANSACTION COSTS ARE A MAJOR DETERMINANT OF THE BEST POLICY RESPONSE “When one party’s actions affect another party and the first party doesn’t fully compensate (or get compensated by) the other for this effect, then the market has failed and government intervention is potentially justified” (Gruber, 2010). Governments can intervene by choosing price or output controls. These controls will distort the market with varying effects on the level of social costs and benefits, depending on the effect the control has on transaction costs for those creating the externality and how they respond to those transaction costs.
Governments must be careful to choose the right measure in the given circumstances. Where the marginal cost curve is steep, price controls result in a more efficient outcome, whereas when the marginal benefit curve is steep, price controls result in a much less efficient outcome. A lot will depend on the core aim of the intervention – is it government policy to prioritise the reduction of output, e.g. to protect the environment from very serious risks that could have a catastrophic effect on society, or to ensure that private enterprises are protected from very high transaction costs, e.g. to encourage entrepreneurialism? In any case, governments must choose the correct measure, depending on transaction costs, to achieve the desired outcome.