More dodging of cost-benefit tests

LABOR has changed the rules so there will be no costing analysis of any election promises, says Henry Ergas.

EXACTLY 45 years ago, the Vernon inquiry into Australia's economic performance lamented the poor quality of investment assessment in the public sector and the resulting waste of scarce capital. A year later, Treasury emphasised that the waste was so great that "the problem of efficient allocation of public capital might well be thought to take precedence over any question of increases in its supply".

The only solution, Treasury argued, lay in ensuring public investment decisions were carefully assessed "against proper economic criteria, [including] the economic return that would have obtained had the resources been put to their most productive alternative use". From that insistence sprang a tradition of rigorous cost-benefit analysis that has added many billions of dollars to the wellbeing of Australians.

This is not to suggest that only projects passing a cost-benefit test saw the light of day. But it is telling that the two that most spectacularly failed proper investment appraisal, and went ahead, were the Adelaide to Darwin rail link and the Sydney Cross City Tunnel. Their outcomes speak for themselves.

Yet at least in those cases there was a proper appraisal. And the public knew what the appraisal showed. As a result, governments could be held to account. Now that accountability is under greater threat than at any time since the 1960s.

Earlier this year, we had the allocation of substantial infrastructure funding to projects that had not met Infrastructure Australia's evaluation criteria, with little disclosure of the reasons for Infrastructure Australia's decisions and even less for those of the government. Then, slightly before the election, and again with scant publicity, the government amended the guidelines on regulation impact statements so as to effectively exempt election promises ("advance auction sales of stolen goods", H. L. Mencken called them) from cost-benefit evaluation. A timely move, for now we have the deal with the independents, which commits vast new funding for projects without any mechanism for measuring and disclosing those projects' opportunity cost.

Hovering above all this is the shadow of the National Broadband Network, the largest single public investment in Australian history.

More than any other, the failure to subject that project to cost-benefit appraisal has compromised the credibility, standing and future of public investment evaluation in this country.

The trouble, we are told, is that doing a cost-benefit appraisal of the NBN would be difficult because of the uncertainties involved. But in the Garnaut report, Treasury modelled the economic consequences of climate change policy a century out. As anyone who has built telecommunications cost and demand models knows, the challenges they involve are far more tractable.

Indeed, sensible cost-benefit evaluations of NBN-like projects have been done in a number of countries, providing a good foundation on which to build. An evaluation would be all the easier given that KPMG and McKinsey, in exchange for $25 million of taxpayers' money, have developed detailed, geographically disaggregated forecasts of costs and demand for the NBN.

Those forecasts, which include estimates of the use of the NBN by hospitals, schools, public entities, businesses and households, provide much of the information a proper cost-benefit appraisal would require.

True, there are some effects such an appraisal might not capture. But unless one knows how big the shortfall is, after counting only those benefits that can be estimated, how can one evaluate whether what has been left out could plausibly make up the difference?

Moreover, without knowing that gap, how could one analyse whether it is best to forge ahead with high-speed connections to all premises, or whether a more targeted approach might not allow the bulk of the benefits to be secured while being more prudent and more efficient?

And how, without that modelling, can the public know the consequences of the government's intention to force the closure of the NBN's chief competitor, or what new burdens are created by its concessions to the independents?

There is consequently an overwhelming case for making public the implementation study's detailed estimates, as well as those by NBN Co. How then can the government, which in its first term vaunted a commitment to "the highest level of transparency", refuse to release those estimates? The reality is the government, not wanting to know the answer, would prefer it if no one could tackle the question.

But experience shows that poor public decision-making and its cousin, lack of transparency, are a gangrene. It is an illusion to believe the decision-making process can be kept half-rotten.

Rather, to require cost-benefit analysis for projects that will readily pass the test, while exempting those one fears will fail, merely ensures an ever worse project mix.

Little wonder, then, that hot on the heels of its NBN decision, the government committed to the Epping to Parramatta rail link and the inland rail project.

For the first, there is not even a business case, much less a proper cost-benefit assessment. For the second, the cost-benefit assessment suggests waste that on a dollar-for-dollar basis could mimic the NBN.

Ultimately, the quality of project evaluation depends on the importance governments place on it. Governments that view rational economic assessment as merely a nuisance that stands in the way of the decisions they want to take, and that believe they can get away with no evaluation or poor quality evaluation, will invariably succeed in devaluing the evaluation process.

The scope to transfer the costs of wasteful projects to future generations, which have little voice in the political process, as well as to future governments, which will bear the political consequences of pulling the plug on failed ventures, then magnifies the risk of enduring inefficiencies.

The resulting costs invariably are made all the greater when those future governments, so as to manage the fallout from the bad decisions they inherited, make bad decisions of their own. It is easy to criticise the telecommunications reforms of the early 90s for the market structure they put in place. But those decisions were shaped by the pressing need to bury mounting losses from the AUSSAT venture which, like the NBN, had proceeded without any careful cost-benefit appraisal.

No easy bulwarks can contain these risks. But it would be helpful if Treasury showed the courage and conviction on these issues it displayed all those years ago. The alternative is a return to the waste Vernon so rightly denounced and that Treasury once fought so long and hard to cure.