A typical example is pollution of air, water or soil, were the pollution is the unintended by-product of production which affects the value of the resource to other users or enjoyers, but which is not accounted for in the markets for either the production inputs and factors, or the markets for the products of the production process. Other relevant examples are pretty countrysides and wildlife, which are largely a consequence of farming practices in the UK, either producing these "by-products" or not, depending on the practices adopted.

These external effects are important because they mean that market transactions do not account for all the benefits and costs of particular activities, so the outcome of markets cannot ensure a social optimum. They arise because of two major factors:

• first, transactions costs - the necessary transactions to account for the benefits and costs (to internalise the effect within a market mechanism) are difficult and expensive in time and effort;
• second, because the property rights to pretty countrysides or wildlife are unclear or contested - who owns the landscape or view, or the wildlife? Without clear ownership rights, there is no legitimate basis for the negotiations about rewarding benefits or compensating losses.

We can picture the situation as in the following diagram.

The point of private profit maximisation is Qm. At this point (and corrpsonding production activity), the damage suffered by the river users and enjoyers is very substantial. On the other hand, there is a level of production and associated production practice which would eliminate the pollution entirely (shown by the intersection of the MD curve with the horizontal axis - implying in this illustration that some level and form of dairy farming is actually beneficial to the river environment).

Suppose that the farmer owns the river, and is alive to the business opportunities of the river. It might be worth his while to find out how much people were prepared to pay for enjoying or using the river. The answer would be that they would only be prepared to pay for a cleaner and less polluted river. The marginal damage (MD) curve shows how much they would be prepared to pay for a cleaner river, reading this curve from right to left, since it shows the monetary equivalent of the damage they suffer from the polluted river. They would be prepared to pay up to this damage suffered to avoid it (prefering to pay less of course). The sensible farmer would then negotiate with the potential users, agreeing to take steps to reduce pollution so long as the price paid for a cleaner river were greater than the reduction of net revenue earned from the farming operation as a result of reducing pollution. This reduction is shown by the MC curve, again reading from right to left. The final agreement between farmer and user would be at Q* pollution, with a payment of T* per uint of cleaner river by the users to the farmer.

What if the farmer does not own the river? Suppose the river is owned by someone else, who is also concerned about its value to them and other users. In this case, these owners can require the farmer to take steps not to damage their resource, and to pay compensation to the owners if such damage occured. The compensation required by the owners is shown, again, by the MD curve (reading from left to right) - a little compensation for a small amount of pollution and a lot for a large amount. What would now be the farmer's best option? Think, before you read on.

The farmer increases his private net revenues the more pollution he causes, but at a reducing rate, as shown by the MC curve, left to right. But he is now required to pay for this pollution at an increasing rate, according to the MD curve, left to right. The profit maximising optimum for the farmer is now where the reducing marginal benefit of polluting (the MC curve) crosses the increasing marginal cost to him of polluting - the MD curve. Once again, the market equilibrium solution (the optimum) is at pollution level Q* and the farmer paying T* per unit of pollution to the owners for the priviledge of polluting the river.

Conclusions on externalities

Thus, according to these principles, there does exist a logic which would allow for negotiations between owners and beneficiaries of a resource and polluters or damagers of (in effect using up) this resource to agree on an appropriate transaction and contract which balances the interests of the producers and users - allowing some pollution, but preventing levels of pollution where the gain from additional pollution is less than the cost of this additional pollution. In effect, our marginal damage curve above is the marginal social cost of pollution, while the marginal cost curve is the marginal social benefit of pollution.

This last remark might raise some eyebrows. The marginal net benefit of private production, however, is also a social benefit (reflecting the market demands for the product in question, milk and beef in this case). Of course, we can object that these markets are distorted by government policies and intervention, and thus do not properly reflect the social benefit of dairy farming. Quite right. The analysis above assumes that the markets for the products produced with pollution are perfectly competitive and not distorted.

The analysis also points up the often very serious difficulties with externalities. Three of the most important are as follows.

• First, it is often very difficult to estimate the monetary equivalent of the damage caused by pollution (or equivalently, the value of cleaner environments - more diverse wildlife and and more valuable landscapes etc.). Voluntary donations to organisations such as the RSPB strongly suggest that there is a large demand amongst at least some of the general public for cleaner and more natural environments. It might be that, in particular circumstances, the MD curve in the above diagram lies everywhere above the MC curve, in which case the market (economic) optimum level of pollution would be zero (that level of production which does not damage the environment at all). In other cases, however, it could be that the actual willingness to pay exhibited by the general public for cleaner environments is simply insufficient to outweigh the indirect benefits of pollution (the value of the production and associated consumption). A case in point is traffic pollution - generally, the public is concerned to reduce this, but is not prepared to do without all the indirect benefits of traffic pollution. So we try to insist that research and development is devoted to making our "necessary" traffic cleaner, and penalising only those who cause excessive pollution, rather than everyone.

• Second, there are often very considerable transaction and negotiation costs associated with getting everyone concerned with the problem together and reconciling their differences. Frequently this is because the externality has at least some of the properties of being also a public good (see below) - like better environments, cleaner air and rivers etc. But, even where this is not the case, the fact that the side-effect has not already been accounted for in market transactions and remains an externality itself implies that the costs and effort required to undertake the necessary transactions are too great to be worthwhile. Externalities are, therefore, an indication that the transactions necessary to internalise them are more costly than the benefits which might be obtained from carrying out the transactions. The existence of organisations like the RSPB illustrate the market response - try and develop institutions and organisations to reduce the transactions costs and negotiate collective agreements to reduce negative (bad) externalities and encourage positive (good) externalities.

• Third, the ownership of the property rights on the natural resource is often contested. Who should pay for pollution depends on who owns the affected resource or property, as shown above. The notion that farmers should be subsidised (paid) to avoid polluting the natural environment implies that farmers own the natural environment. If they do not, then the implication is that farmers should be taxed (required to pay compensation) for any pollution they cause. The general presumption (outside agriculture) is that air, water, and even soil are common property - belonging to everyone - so the polluter pays principle applies. Should farmers be different, and if so, why? Thoughtfull and well argued Answers to me, please!

It is partly because of these difficulties, and also because the effects of pollution (or, more generally, environmental degradation) tend to be specific to particular locations and different between different locations and practices (that is, the effects are highly heterogeneous), that governments are asked (by their electorates, assisted by pressure groups) to intervene and find solutions to the externality problems. This analysis suggests that the appropriate solution is a system of pollution taxes or subsidies to avoid degradation. It can be shown that, under restrictive conditions, the tax/subsidy system is more efficient than regulation - blanket restrictions on levels of pollution and fines/penalties for contravention of the regulations (see here for a demonstration of this conclusion, and reasons why regulation might, nevertheless, be a preferable option in many cases, as is frequently observed in public opinion preferring regulation to taxes and subsidies).