Public sector engineering construction work - %GDP (source: Grattan Institute)

Whatever misgivings I’ve had about some other reports by the Grattan Institute, the latest one demonstrates the value of the Melbourne-based think tank.

Game changers: economic reform priorities for Australia seeks to identify those workable economic reforms that would produce the greatest benefits for the nation by 2022. It proposes three:

  • Broaden the GST to cover education, health and food, but simultaneously lower income and corporate taxes
  • Change the benefits for women with young children so as to reduce disincentives to paid work
  • Increase the age at which people can access their superannuation and the aged pension.

I don’t agree with everything in this report (summary here), but I think the approach is impressive. It asks the salient questions and seeks to follow the evidence to its logical conclusion rather than fawning at sacred cows. This is the sort of independent, hard-nosed and even unfashionable thinking a think tank should be doing!

Along the way it evaluates – and rejects – a number of common suggestions for lifting economic growth (it’s upfront that the focus is on economic outcomes). Most of them it finds wanting, either because they don’t deliver big enough returns and/or because there’s insufficient evidence they’ll deliver the change sought (note though that political feasibility is deliberately not considered).

I’m not going to discuss the merits of the three recommended “game changers”. My primary interest is in urban issues, so the bit I want to focus on is the discussion (from page 26) of the potential of expanded investment in transport infrastructure to drive economic growth.

Contrary to the conventional wisdom, the Institute finds it would not make a huge economic difference. In fact it says the potential is “surprisingly small.” It’s not just transport infrastructure though – it reaches the same conclusion about innovation policy, industry policy, oligopoly regulation and industrial relations reform.

In essence, it argues the transport projects that could be delivered within the time frame don’t have very large economic benefits (there might be non-economic benefits though) relative to their costs. It’s not that transport infrastructure is a poor investment in-principle – it’s the underwhelming economic pay-off from the specific projects that governments around Australia are considering.

That’s interesting stuff! The discussion of transport infrastructure in the Institute’s report is only short, so the best thing is to read it yourself – here’s the full text (I’ve deleted the footnotes and references – refer to page 26 of the report for these):

Some issues might be expected to feature as game-changing reforms, but don’t appear among this report’s proposals. The prime example is transport infrastructure.

It is often claimed that Australia’s infrastructure is poor and declining, and under-investment is holding back economic growth. However, investing faster in Australian infrastructure is unlikely to substantially increase the size of the economy over the next decade. Australian infrastructure spending is already at historic highs. Theoretical work on infrastructure over the last few years casts doubt on claims that infrastructure spending has a major impact on growth of a developed economy. Project-by-project analysis by Infrastructure Australia reveals relatively few projects ready to proceed, and most have modest net benefits.

Of course, much infrastructure has non-economic benefits, including public amenity, social cohesion, and environmental impacts. On these grounds alone, individual projects may well be worth pursuing. But in strictly economic terms, infrastructure does not have economic impacts large enough to change the game of Australia’s economic growth.

Many have decried the state of Australia’s infrastructure. Engineers Australia has estimated a $700 billion shortfall. It rated much of the nation’s infrastructure as needing major changes to be fit for current and future purposes. While its assessment identified changes that would make the infrastructure “fit for purpose”, costs would exceed the economic benefit of many of these improvements. The World Economic Forum rated Australia as 24th of 142 countries for its infrastructure, although this ranking was largely driven by the self-assessment of surveyed Australian executives. Similarly, the OECD concluded that “Australia has an important infrastructure deficit.”

Despite – or perhaps because of – these claims, Australian government spending on infrastructure has increased rapidly over the last decade (see exhibit). And of course given the mining boom, private sector spending on infrastructure has increased even further.

Obviously, infrastructure can significantly increase economic growth, particularly if it facilitates trade. Although work from 1989 suggested very high economic returns, more recent analysis suggests that the benefits are smaller, so that a 10 per cent increase in the stock of infrastructure increases GDP by 1 per cent. Such multipliers must take into account the opportunity cost of the government funds involved – including the value generated by taxes that are lower because infrastructure spending is less.

Even if infrastructure is productive on average, it is only economically productive if it is “the right infrastructure, in the right place at the time and accessible at sensible prices”. Simply spending money on infrastructure is not enough to get a return if the cost benefit case is not there.

Australia has improved the rigour of infrastructure project assessment through Infrastructure Australia, whose assessments have become an important part of Commonwealth government spending allocations. Their assessments between 2009 and 2012 suggest that it is unlikely that more than $10 billion of new positive cost benefit projects will be sufficiently prepared to proceed each year. Between 2009 and 2011, Infrastructure Australia identified $42.2 billion of ‘new’ projects that were at least on the threshold of being ready to proceed; $25.5 billion has been allocated to fund most of these projects over the last three years. Only two projects, costing $7.4 billion, remain in the pipeline as “ready to proceed” but unfunded. It is doubtful that these projects are in fact “ready to proceed”: the $5.9 billion Managed Motorways project depends on a pilot program now underway, and the $1.4 billion Melbourne Metro 1 project depends on the outcome of design and preconstruction work.

Based on the published cost benefit analysis of 1.5 for the 2010-2011 projects (excluding the technologically unproven Managed Motorway project), the net forecast economic benefit of completing $10 billion projects would be around $5 billion. This includes consideration of the “wider economic benefits” such as increased agglomeration economies and greater labour supply, which are inherently difficult to quantify.

Spending of $10 billion per year on new major transport infrastructure projects is already built into government spending plans, with the ABS reporting public construction work in the pipeline as close to record highs. Multiples of this investment would be needed to result in an increase in net economic growth – assuming such projects could be found.

The realised benefits are likely to be substantially lower. A large survey of infrastructure projects across the world found that project costs are typically at least 20 per cent higher than forecasts, and benefits (particularly for rail) substantially lower. Even if the project costs and benefits are realised, the net benefit of $5 billion per year is much smaller than the game-changing opportunities to increase economic growth identified elsewhere in this report.

Anyone who doubts that transport projects have modest BCRs should look at this evaluation prepared for Melbourne’s East West Needs Assessment Study. It estimates a ratio of 1.2 for a public transport option and 1.0 for a combined road and public transport option. These authors reckon even these estimates are optimistic.