Published in The Guardian, Tuesday 30 September
The pet shop galahs are talking about infrastructure. We even have a Working Dog satire, Utopia, on the wonders of nation-building. If only Australian governments would spend more on infrastructure, the story goes, our economic problems would be solved.
Alas, life is not so simple. Infrastructure can certainly provide economic, social and cultural benefits, but only if it is the right infrastructure, in the right place, built at the right time for the right price.
Spending well on infrastructure is first-best policy. A 21st century economy is increasingly about services, particularly the high-value services that cluster in and around the centres of cities. Without the transport infrastructure to get people there easily, we all lose.
Socially, infrastructure matters as well. Long commutes are bad for you – they go with lower well-being, poorer health, lower productivity at work, and just about everything else that people care about.
However, not all infrastructure spending justifies the cost. We have a national monument, usually known as the Alice-to-Darwin railroad, to prove it. This is but one in a long list. As the Productivity Commission observes in its recent report on infrastructure, Australian governments could do a lot better. They tend to pick the projects that instinctively appeal rather than those that would make the most difference. Governments tend to pick large projects – they sound so much more impressive – but smaller projects usually have a better payoff.
Indeed, Australian governments have spent a lot on transport infrastructure over the last six years – about 1.1% of GDP, substantially higher than the average between 1987 and 2006 of around 0.8% Yet there has been no obvious economic lift as a result.
It is understandable that it is getting harder to find good projects. If we are doing a moderately good job of project selection, then by definition we tend to do the high pay-off projects first (think Hume Highway between Sydney and Melbourne), leaving lower pay-off projects till later. Local communities are always happy to have new infrastructure – provided that someone else is paying for it.
It doesn’t help that Australian governments often try to keep infrastructure details under wraps. The Victorian government has now signed contracts for the first stage of the East West Link road that will cost about $6bn, but has so far refused to release the detailed analysis justifying the project. It has claimed that exposure would reduce the government’s negotiating power. But surely it is better to pay contractors a bit more for the right project, than a lot for the wrong project. In this area of government administration, like nearly all others, sunlight is the best disinfectant.
Even when the original case looks good on paper, the outcomes tend to be worse than planned. Road projects both in Australia and around the world tend to carry substantially less traffic than projected. Costs tend to be 20% to 45% higher than forecasts.
As a result, we tend to get third best policy – governments spending a lot of money on unproductive infrastructure. There is a real danger that advocates of infrastructure as a cure to sluggish economic growth will ultimately get more of the same.
Requiring state governments to live within their means – and in particular to fund infrastructure out of recurrent surpluses – may not be the best policy in an ideal world. But it creates a lot more pressure to pick the best project on offer.
For example, in the lead-up to the Victorian election, the electorate seems to have realised that within the next decade, government will not be able to build both the East West Link and the Melbourne Rail Link (the latest name for the Metro Rail project). As a result, there is genuine political contest over what would be the best transport infrastructure choice.
Yet changing infrastructure priorities at every election can waste a lot of time and energy. To improve decision making, all states need an independent statutory infrastructure body that investigates projects, analyses them carefully, and reports publicly on the relative merits of various proposals. This can both improve choices, and help to build genuine bipartisan commitment.
Infrastructure NSW was a step in this direction. Yet the model could be improved, by clarifying whether its job is to critique government proposals, or to devise its own.
Infrastructure has become a political football. Large commitments are being made, with little information available to the public. There are too many bad decisions.
First-best policy might be to spend even more on infrastructure, with much better decisions. A more realistic second-best would be to improve the quality of decision-making first. It won’t be easy to get the pet shop galahs talking about the quality of infrastructure decision-making. But that’s what we need.