Australians returning from holidays to work this week were met with headlines reading: “ACCC calls for big asset sell-off”. Australian Competition and Consumer Commission chief Rod Sims had reportedly suggested the federal government sell off both Medibank Private and Australia Post.
Later that day the ACCC issued a “clarification”, stating that while Sims believed that the private sector could run commercial enterprises more efficiently than government, there had been “ no reference made to privatise any specific Commonwealth owned entity”.
That didn’t squash speculation, however. Former ACCC Commissioner Stephen King specifically called for Australia Post to be sold, and Radio National’s Breakfast Program on Tuesday ran a debate between Julie Novak, a Senior Fellow at the Institute of Public Affairs and ACTU president Ged Kearney, expanding the discussion into selling electricity assets and even the ABC and the SBS.
The debate around these issues has been partisan and emotive – advocates are either “for” or “against” privatisation, almost regardless of the circumstances.
On one side is the argument that businesses owned by government are generally less efficient than those owned by the private sector. With such a generalisation it’s easy to find examples to support one’s partisan view. Anecdotes on both sides were flowing freely in the Breakfast interview.
It is useful to study efficiency and to make comparisons, not only between the private and public sectors, but also between government entities performing comparable services. That task is well-handled by the Productivity Commission in its regular reviews of government service provision and in its specific inquiries such as into private and public hospitals (which found no discernible difference between the sectors).
These inquiries and other studies usually confirm economic theory about the complexity of making comparisons. “Efficiency” is not a single, easily-measured property. It has many dimensions, and what is efficient by one criterion often comes at the cost of efficiency by other criteria. For example an electricity utility with high supply reliability may have achieved this through over-investment in peak load infrastructure.
Governments are often left with the hard cases, such as services for which the private sector cannot make a profit (country mail deliveries), or which present too much uncertainty for risk-averse investors (disability insurance). And that’s before we get into the complex world of allocative efficiency. No matter how technically efficient a government or private business is in delivering a service, if those resources can be put to better use there is allocative waste. Police and customs may be very good at intercepting narcotics, but to reduce drug dependence it may be far better to allocate funding to public health measures.
Studies which reveal the complexity of measuring efficiency don’t support Sims’ categorical statement: “If all you’re after is maximum efficiency then there’s no question that you’d have those assets owned by the private sector”. But many studies do reveal occasions where government enterprises can improve their performance. Common problems are poor labour productivity, either because of restrictive work practices and featherbedding, or because of under-capitalisation.
In many state-owned enterprises, particularly power and water utilities, under-capitalisation results from governments taking too much as dividends at the expense of retained earnings. And there is often poor governance and management when board and management appointments are used as forms of political patronage. (When people resist privatisation for employment reasons, it’s a pretty good sign of poor labour productivity.)
Such findings, however, do not establish a case for privatisation. Rather, they establish a case for reform. That’s what a large company does when it finds one of its divisions is operating inefficiently. Privatisation of government enterprises is generally a lazy and costly substitute for reform.
Decisions on what should be in public hands versus privately owned should be based on economic criteria rather than sweeping and untested generalisations about efficiency. Where private markets work well, bringing benefits associated with competition, there is a case for privatisation. Where private ownership does not bring such benefits – in other words where there is significant market failure – there is a case for public ownership or control. The clearest examples are those of natural monopoly, where the market can support only one supplier, but there are many other situations of market failure requiring strong intervention.
That’s conventional economic theory, but it is getting little airing. Rather than articulating that theory, Sims suggested that the only sound reason for government ownership is to achieve particular social objectives. Of government enterprises, he said: “If you’re continuing to own them by government, then that’s because you’ve got some social objective to achieve”. In her debate with Novak, Kearney endorsed Sims’ “social benefit” justification.
That’s a vague and weak basis for public ownership, implying that we must bear the supposed intrinsic inefficiency of public ownership in order to achieve “social objectives”, whatever they may be.
The division between public and private ownership should based on well-established economic principles rather than the outcome of a partisan conflict or vague claims about “efficiency” and “social benefits”. We are already bearing the economic cost of inappropriate privatisations – toll roads, energy and water utilities, health insurance – where commercial incentives conflict with efficient resource allocation.
Are we letting a a partisan stoush displace good economics? Or is it that our politicians and commentators aren’t even aware of the economic principles which should guide the division between the private and public sectors?