Its all very simple in a first best economy. To get the maximum total of consumer and producer surplus a natural monopoly must produce where price = marginal cost. As marginal cost is easy to measure, this rule would be easy to enforce. However, natural monopoly is likely going to face very high fixed costs, and very low marginal costs, so the price would be very low. Therefore, the company would likely make a loss. On the other extreme, when allowed to operate freely, the price would be set where MC=MR and too much consumer surplus is probably lost.
One can finance the losses by using Ramsey taxes (or Clarke tax or whatever) on other industries and subsidise this. Another way would be to aim for a profit target, and restricting either prices or quantities. The price vector restriction has been especially popular, with the RPI-X formulas adopted for most privatised industries. This is because it leaves the industry with quite a bit of discretion. As there is likely going to be large asymmetric information problems with regulating industries (regulators do not know enough), some discretion is optimal.
There are also social aims that the industry might have to fulfil. For example, some argue that it is easier for the nation to live in a country where water costs the same everywhere. Or that is politically more feasible to provide welfare benefits in kind for equitable purposes. These rules are very hard to formalise, thus no clear rules have emerged.
Lastly, the industry must allow competition. However, there are problems with stranded contracts and cream-skimming when competition is introduced. When the monopoly cross-subsidises between consumers, incumbent could just take over the customers that are paying too much. This might create unequitable outcome, however, it is more efficient. Strangled contracts occur because of the need to cover fixed costs by the monopoly. Thus, they have long contracts with the consumers. However, once the resources have been committed, consumers have an incentive to bargain for better deals, as the investment is irreversible.
On the other hand, lack of competition can damage consumers. Normally the monopolies have vertically merged, so that the generation and distribution are together. The distribution bit has an incentive to discriminate against other providers, which allows generation to make higher profit. This reduces social welfare. Furthermore, the regulatory board can also sometimes fail, with the result being, that more monopoly pricing is introduced.
Although most people tend to dislike the possibility of price discrimination, I think it is crucial for the success of natural monopolies. Although it is a transfer of consumer surplus to producers surplus, it often allows more consumers to be served, thus actually increasing consumer surplus as well. And once the discrimination is in place, state can tax (or price cap) away the gains. So I think that complex pricing mechanisms should be allowed to formulate, and no rule to rigidly fix prices should be adopted.
Having an optimal amount of investment from the natural monopoly is very hard. Firstly, monopolies use investment and research as barriers to entry. Secondly, if the price caps are not universal, obviously the investment becomes distorted. Thus, the regulation should in first instance be set as to prevent these distortions.
Secondly, facing no real competition the natural monopolies would probably not modernise their equipment fast enough. That is why it is important to separate the generation, where technical advances occur, from transmission, which has a stable technology. This will mean that the costs for undercutting a monopoly with better quality machinery are not that large, and more investment occurs.
Thirdly, the public company should be floated in the stockmarket. Then participants can invest based on the q theory. If extra investment would benefit their shareholders more, they would invest indeed, otherwise not, although the projects might still be profitable. If these conditions are satisfied there should be no distortions to investment. This has happened in the electricity industry where small new plants have been set up to provide for peak time demand.
As the control and ownership become separated, it is harder to make managers pursue the aims of shareholders. One simple option would be to give them share options – if company does well, so do they. Furthermore, there are likely going to be more checks in the private system than in the government one anyway. Therefore, I think one should not worry to much about the motivation of management. Seeing the huge salaries they get publicised lately, and added to that the moral satisfaction one gets from leading a big utility, there should be no significant problems with management.
In a natural monopoly consumers will have nowhere to go if they are not satisfied with the quality they get. Thus setting standards is important. However, in practice I think that the crucial fact is to allow firms to differentiate and price discriminate with quality. Then the government could observe the demands and essentially overcome its informational deficiencies as to what is a feasible quality. Also, the threat of regulation if the quality is not adequate should suffice to make the monopolist set its own quality standards.
There is very much different experience from the UK. I am briefly going to describe three: the overall feasibility of privatisation, the investment and the time one should spend in formulating the regulation.
Firstly, the privatisation has widely been regarded as a success. This was shown in the high initial gains in the share price of floated utilities. Also, the regulation has not disappeared and the abuse of monopoly power is not widespread. There have been considerable gains in efficiency.
Secondly, the evidence of investment and quality has been mixed. Not enough investment was achieved in the water industry, despite the higher prices they were allowed to charge. It can be said that the clean drinking water has positive externalities, so instead of hoping the firm to provide sufficient investment, maybe it is better to subsidise its investing.
Lastly, many of the privatisation programmes have seemed to be operated in a hurry, when government needs cash fast. The costs of delaying the privatisation should not be that large, so maybe it is sometimes more beneficial to allow for longer transition periods.