Privatisation has been one of the weakest links when it comes to market-oriented policies. Although some privatisation schemes have been immense successes, others have resulted in similar, or even worse, levels of service to the end consumer, while only benefiting the private corporations which have taken over the previously state-owned companies.
The problem, I think, stems from forgetfulness. Specifically, the forgetfulness of policymakers. Privatisation is not an ends in itself. Privatisation is a means to an end — and this is something which policymakers have forgotten.
Rather, they seem to think having privatised a formely state-owned industry is enough. If all that results is the shifting of monopoly power from the state to a corporation, nothing has changed except who reaps the profits.
The point of privatisation is to stimulate the usage of the market to allocate goods and services in an economy, rather than relying on the state. The purpose is also to increase efficiency in the provision of goods and services, in the hope that the resulting cost cuts will be passed on to the consumer.
These ends are often forgotten in the mad rush to privatise. If a privatised industry remains a monopoly, then government must either regulate it in order to ensure that any falls in costs are passed on to the consumer in reduced prices, or stimulate competition in the industry.
Privatisation is often attacked by people who believe that state control of an industry has benefits. These normally include providing employment that would not otherwise be available (at the expense of efficiency, due to redundant workers) and lower prices than might otherwise be expected (basically government subsidisation of the industry).
These are not good reasons to oppose privatisation, however. For certain non-critical things, it may be wiser for the public sector to allow the private sector in.
The employment problem can be resolved by providing adequate unemployment benefits (mainly through the negative income tax, ideally) and also by providing retraining and education, as well as capital for unemployed people with an entrepreneurial streak.
If subsidisation is absolutely warranted, there's no reason it can't continue under a privatised industry; furthermore, if the privatisation is correctly effected, the increased efficiency will lead to a fall in costs (and thus prices), which may partially or even completely obviate the need for subsidies.
Privatisation is an excellent way to encourage efficiency in the provision of goods and services. Let us take the hypothetical example of the post. If the postal service were not a guaranteed monopoly of the state, but rather was a job granted to a particular firm for a certain period of time based on a contract (of course open to bidding), then there would be a strong incentive for the operator of the postal service to be as efficient as possible, so as not to lose the contract once it expired and bidding commenced for the next one.
Privatisation is a great tool. But like all tools, in the hands of careless or foolish men, its positive impact can be severely stunted. Privatisation must be utilised and applied correctly and carefully if it is to yield the fruits that its advocates so lavishly promise.