OPINION: In the wake of the Queensland state election, and in the run-up to the election in NSW privatisation is, once again, a hot topic.

Predictably, Labor is running a scare campaign that makes Henny Penny look like an optimist. If we privatise electricity distribution in NSW then not only will the sky fall in but the ground on which we walk will be in trouble, too. We will rue the day we deigned to discuss it. The Liberal party is no slouch in the scare stakes either: privatisation is so obviously right that we'll tie building roads and basic infrastructure to it. No sell the poles and wires, no good stuff.

I'm not so sure.

There are sensible arguments on both sides of the privatisation debate. We just haven't heard them yet. So here goes.

Most people would agree with the following statement: there are some things that government should pay for. National defence, primary and secondary education, basic healthcare – these are things that almost all of us acknowledge that government should pay for.

But what should government own? After all, we could give students (or their parents) a voucher for education primary and secondary education. The same goes for healthcare. Just because government should pay for some stuff doesn't mean they should own it. After all, the government can outsource all sorts of functions to private-sector companies who face competitive pressure to do a good job at a low cost.

Yet sometimes government should not just pay for things, it should own them, too. Not often, but sometimes. And it's useful to understand when and why. The most insightful example is prisons – as Harvard economists Oliver Hart and Andrei Shleifer pointed out in 1997, along with Robert Vishny of the University of Chicago, in a now classic academic paper in the world's oldest and most highly regarded economics journal: The Quarterly Journal of Economics.

Imagine privatising a prison. The downside of public ownership is that the warden who runs the prison has terribly weak incentives to cut costs and find efficiencies. He or she is just an employee, after all. In contrast, private ownership will provide strong incentives for cost reduction because the private provider gets all the benefit of cost reductions. But if quality – such as how prisoners are treated and what rehabilitation services are provided – is hard to describe in a contract then these incentives might be too strong. Private prisons could cut costs by providing substandard treatment and insufficient rehabilitation services.

This is hard to dismiss as ivory-tower academic waffle. There's plenty of evidence about how privatised prisons treat prisoners. And it ain't pretty.

In a world of perfect foresight and perfect contracts none of this would matter. But in the real world there is a trade-off. Private ownership provides big incentives for cost cutting, but it typically comes at the expense of quality. Public ownership provides weak incentives across the board. The key question is how hard it is to contract on what the government wants from a private provider.

The public debate that I listen to right now involves pretty silly arguments on both sides. On the left: "These are essential services" – subtext: "I wish the government could run everything, but we've lost that fight, and this is the last asset standing." On the right: "Business is more efficient than government" – subtext: "I don't trust public-sector-types to run a bath, let alone a $20 billion business."

The key in all of this is that privatised entities will respond to incentives – they will cut costs. But experience teaches us that they will cut costs across the board, and there are some things we, as a society, want more rather than less of. "Quality" and "innovation" are chief among those things.

So what about poles and wires in New South Wales? To make a serious case against privatisation requires pointing to big issues that the government cannot contract on in the process. It's hard to see what those might be. This is a simple business with easily measurable outputs. We all know what quality electricity provision means and the technology is well understood. Privatisation would provide incentives for cost reduction and free up much-needed capital for the state government.

Government should own things when, and only when, it is hard to describe what "quality provision" means and hence residual rights of control are important. Poles and wires in New South Wales is not such a case. Both sides of politics would do well to drop the ideological pronouncements and get pragmatic.

Privatising these assets is neither "right" nor "wrong". It's just efficient and sensible. Like most things in life, it involves a tradeoff. But this one is clear. The New South Wales government has no business being in the electricity business.

Richard Holden is Professor of Economics at the UNSW Business School.  

This opinion piece was first published in the Australian Financial Review.