That the government isn't the best provider of many goods and services is a common enough argument, with plenty of empirical evidence to back it up. The (theoretical) question is, Why? I'm increasingly thinking that the answer is simply the incentives that government ownership or provision entail, or in short, the politicisation of provision.
It could be argued that it is an issue of corporate governance and that the problem with government-owned firms is poor selection and motivation of managers and workers. That within the bureaucracy government managers and workers just need the correct incentives and all will be well. But what underlies bad corporate governance and incentives?
I noted in the posting on Last Rites for the Cullen Fund? the comment by Roger Kerr on the politicisation of decisions to do with the fund,
Moreover, there was never any chance that the Guardians of the Fund would remain free from political interference in their job of maximising its returns (for a given level of risk). Predictably, in a few short years we have seen political parties variously proposing to bias the Fund’s investments towards infrastructure, green projects, venture capital and investment in New Zealand.I also noted in the posting on Public-private partnerships one of the reasons that PPPs have not worked as expected was
[...] the operation and outputs of PPP schemes have often been subject to substantial political and bureaucratic intervention. As seen with some of the public transport PPPs, a hostile relationship may develop between the counterparties. There can even be politically-motivated attempts to subvert the viability of projects. This makes it more difficult both to raise private finance and transfer risk. Investors are more likely demand a premium and contractual guarantees if they perceive political risks as high.When applying their grabbing hand theory of government to privatisation Shleifer and Vishny write,
Unlike the helping hand perspective, it suggests that the key problem of state firms is government interference in their activities to direct them to pursue political rather than economic goals, such as excess employment. [...] Indeed, assigning to the government the role of actively finding better managers or of restructuring monopolies contradicts the essential premise of the grabbing hand approach: that government control is itself the fundamental problem.In their discussion of New Zealand's SOEs, Spicer, Emanuel and Powell (1996) warned that there are two pressures on SOE's: the first being towards privatisation since the productivity and efficiency gains achieved by SOE are in danger of being eroded over time due to political pressure and the second being towards a return to being part of the public sector where social and political objectives can be more readily be meet. Most interventions in SOE's operations seemed to be more politically motivated.
If this is right and the fundamental problem with state ownership is the politicisation of provision, then it suggests that any reform of the public provision of goods and services has to be aimed at reducing as much as possible the ability of politicians and the bureaucracy to intervene. This is far from easy. Privatisation, for example, is often seen as one way to depoliticise a firm. But governments often keep "golden shares" in the privatised firm just so they can intervene when it is politically advantageous to do so. Shleifer and Vishny argue that
[...] the design of privatization must focus on restricting the possible future influence of the state on privatized firms, through subsidies, regulations, or even minority ownership.But the compete avoidance of government influence on firms is impossible. However the above does suggest that for the example of privatisation, the sale of the total government interest in a firm is needed along with no use of residual control via things like "golden shares".
Other approaches to the depoliticisation problem have been tried in other areas of the economy. New Zealand tried to make its SOEs work like private business by writing this requirement into the SOE Act. But as Spicer, Emanuel and Powell (1996) make clear this hasn't been totally successful, political interference can still take place. Depoliticisation of the operation of monetary policy is attempted in New Zealand via the Reserve Bank Act which gives the Reserve Bank instrument independence, but not goal independence. But there is also debate as to how successful this has been. The Kerr quote above suggests that the independent Cullen Fund may not be so independent and the problems of the PPP schemes noted above also suggests that these approaches to political free investment may not be as successful as they could be.
So the politicisation of provision may be an obvious problem but it doesn't seem to have an obvious solution.