Taxpayers are footing higher infrastructure bills than is strictly necessary due to poor data sharing between various levels of government, the Grattan Institute has found.

In its report Orange Book 2022: Policy Priorities for Federal Government, the Melbourne-based think-tank found poor access to historic cost data has left cost estimators unable to make accurate forecasts for the pricing on new projects.

“Large public infrastructure projects are funded wholly or mostly by taxpayers,” the report said. 

“Therefore the community has a stake in knowing how projects turned out, whether costs were well managed, and whether the initial promises were delivered. But at present, information on project delivery is not presented in a clear way, if at all.”

One “clear manifestation” of this lack of data accessibility, the report said, is that cost estimates for new projects “continue to make insufficient provision for ‘worst case’ cost outcomes”.

“Business cases typically include an estimate of the median cost, or ‘P50’, and the worst case, or ‘P90’. In business cases produced in recent years, the difference between P50 and P90 cost estimates has generally been about 7%. 

“But the experience of the past two decades has shown that the cost over and above the median cost estimate is actually 49%, on average.”

The Grattan Institute used the Inland Rail Project to highlight this discrepancy, noting that this project’s initial P90 cost estimate was $10.657 billion.

If this estimate was correct, the report said, it would indicate the probability of the Inland Rail Project exceeding this cost would be only 10%; current estimates place that figure at $15.5 billion already, and this could still increase.

“Relying on undercooked cost estimates to make an investment decision distorts investment planning,” the report said.

This skews project selection in three ways: underestimating the full cost means the cost/benefit ratio is also overstated; decisions to invest in a project are therefore made on “an incorrect basis”, the report said; and importantly, unrealistic estimates are misleading to the public.

In order to fix this issue, the Grattan Institute proposed creating a ‘cross-jurisdictional database’ capturing information on all projects valued at $20 million or more.

And although state governments predominantly hold this data, the collective database should be co-ordinated at a federal level, the report said, to give estimators a better understanding of how their forecasts align with actual experience.

Sources

• Grattan Institute, Orange Book 2022