Large projects currently under development or construction (source: Grattan Institute)
Large projects currently under development or construction, showing the assumed average difference is median vs worst case cost estimates is much lower (9.%) than the historical difference across all completed projects (26%). Ssource: Grattan Institute

The report released last week by The Grattan Institute, Cost overruns in transport infrastructure, informs us that over the last 15 years Australian taxpayers outlaid 24% more for transport projects than they were told they would pay. The “promise premium” totals $28 billion.

The Institute examined 836 projects valued at $20 million or more and planned or built since 2001. It concludes the main cause of overruns is political opportunism – politicians announce projects before they’ve been evaluated:

Projects that are announced prematurely have larger and more frequent cost overruns than those announced at a more mature stage of development. This is true not just in the run-up to a formal cost assessment but throughout the project lifecycle. Limiting premature announcements could substantially reduce cost overruns.

My go-to example is Melbourne’s South Morang rail extension; it was initially promised at an estimated cost of $8 million by the Bracks Government but ultimately cost around $250 million (or $562 million depending on whose counting).

Unfortunately, the Institute concludes there’s no grounds for believing the problem has been fixed:

On the contrary, projects currently in the planning and delivery stages have cost estimates that do not take account of the experience of the past 15 years, and many may well be at risk of significant cost overruns.

The exhibit indicates governments persist with taking an unjustifiably optimistic view of large transport projects, even those that are well advanced. It shows the difference between best case (P50) and worst case (P90) cost estimates is generally much smaller than the 26% average revealed by the real-life experience of the last 15 years:

Judging by recent history, this indicates that either the median cost estimate is too high, or – more likely – the “worst case” cost estimate is too low.

The Institute recommends several key actions, including:

  • A rigorous, independent evaluation of proposals is undertaken before government’s commit funding, with the business case tabled in parliament and published.
  • Post-completion reports are published and used to monitor the cost of projects via a database of completed projects.
  • Stand-alone legislation is enacted for each project that costs $1 billion or more.
  • Project contingency funds are kept in a portfolio pool at arm’s length from the project manager.

Politicians have strong incentives to understate the cost of projects. It makes their (politically) preferred project look more affordable than it really is. The downside is it might be a dog and there might be other better ways the money could be spent.

I think the Grattan Institute is on the right track with its emphasis on transparency. Trying to take the politics out of politicians – via setting up bodies like Infrastructure Australia – is a hit-and-miss solution with often disappointing results (e.g. see Can the politics be taken out of infrastructure planning?). The most plausible way to better outcomes, I think, is putting information in the public realm.

Improved transparency, though, relies on the mainstream media to do its job of holding politicians to account. Unfortunately, even a quality publisher like Fairfax is nowadays too often more interested in shaping the news in ways attractive to its readers prejudices than in objectively reporting it or questioning it (e.g. see Is Fairfax off-track (again) on airport rail?).

Kudos to the Grattan Institute for doing the sort of policy-relevant work that many Australian universities choose to ignore.