Published by Australian Financial Review, Monday 13 March
Reserve Bank of Australia governor Philip Lowe is talking up public investment in infrastructure. We need more of it, he says, because it “increases demand, takes the pressure off ultra-low interest rates, and increases the productive capacity of the economy”.
Not so fast, governor. Until Australia gets a lot better at assessing and selecting which projects to back, there is high chance that much of the new spending will be misdirected or wasted.
It’s not just the Reserve Bank. The International Monetary Fund and the OECD are also championing infrastructure spending to boost growth. But to see public investment solely in these “macro” terms is to miss the critical point that each individual project needs careful evaluation to ensure taxpayers get the best value for their money.
Harvard economist Ed Glaeser advocates a project-by-project approach to infrastructure policy. He rightly worries that a focus on jobs and macroeconomic impact leads to the wrong infrastructure in the wrong places – such as responding to a halving of Detroit’s population by building a new monorail to glide over the empty streets. “If you have a focus on jobs and macroeconomic effects, it leads to infrastructure in the wrong place,” Glaeser said recently in debate with former US Treasury secretary Larry Summers.
Dr Lowe’s comments would fit better in a world where we could be confident that project evaluation and selection led to the right infrastructure in the right places. But Australia lacks a strong and transparent governance framework for assessing individual projects.
Our politicians sometimes commit to projects before Infrastructure Australia has the chance to independently assess the business case – or even before there is any business case at all. Grattan Institute analysis shows that over the past 15 years, 32 per cent of transport infrastructure projects valued at $20 million or more were announced before a funding commitment had actually been made. These prematurely announced projects accounted for 74 per cent of the total value of cost overruns. And it’s an expensive folly: over the past 15 years, Australian governments spent $28 billion more on transport infrastructure than they told taxpayers they would
If the evaluation is poor, the selection of projects is sometimes worse. Australian politicians have a long history of backing economically wasteful projects for base political purposes. Melbourne’s East-West Link road, if it had been built, would have provided only 45¢ of value to the community for every dollar spent on it. And if some of the marginal projects currently on the drawing board incur the average cost overrun, they will prove not to have been worth building. For example, Brisbane’s Cross River Rail, and the Inland Rail project between Brisbane and Melbourne, teeter on the brink of being unviable. Both are expected to provide $1.10 of benefits to the community for every dollar spent – but only if they do not run over their budgets.
Some argue government should take advantage of historically low interest rates to borrow more money for infrastructure. The implied assertion is that the discount rate – the device used to transform future costs and benefits into today’s terms, to compare what else could have been done with the money – should have fallen in recent years. This argument seems reasonable. Typical discount rates used in individual project evaluations are probably too high, and therefore the number of projects passing muster may be too low. But until we can solve the problem of “optimism bias”, a high discount rate is probably a useful counterweight against wasteful spending sprees.
Of course, politicians naturally hope that their own pet projects will be different, free of optimism bias, and won’t end up as expensive white elephants. Dream on.
Even now, there is every sign of optimism bias on some of the large projects being developed or built. Melbourne’s Western Distributor and Sydney’s WestConnex each only allow 6 per cent more for the “worst case” cost estimate than the amount governments say the projects will cost. Canberra’s Light Rail allows only 7 per cent. But history tells us that the “worst case” cost estimate should be more like 26 per cent above estimates of their probable cost, to account for the inevitable nasty surprises complex projects can produce.
There is a macroeconomic case, as Lowe argues, for more investment in public infrastructure. But not before Australia has greater discipline in project selection and appraisal. Politicians should not be able to commit public money to infrastructure until a rigorous independent evaluation has been tabled in the Parliament, along with the project’s business case. And the Commonwealth should publish reviews of completed projects on the data.gov.au website, so engineers – and politicians and the public – can learn from mistakes.
Yes, Australia’s macro-economy might benefit from extra public infrastructure investment. But it’s dangerous to imagine that our prosperity is determined by the number of hard hats and cranes, rather than how well public investment supports the community.