How a cost-benefit analysis could help the NBN case

OPINION: It may be entirely appropriate that NBN Co deliver the Federal Government ongoing cost-benefit analysis around the national broadband network rollout: so long as stakeholders understand how to interpret the results, writes David Havyatt.

David Havyatt

The liveley public debate over the merits of a cost-benefit analysis for the NBN reminds me of a scene from a telco senior executive meeting. 

The Chief Financial Officer (CFO) and the Chief Marketing Officer (CMO) are in furious disagreement about the plan for a new product. The company has made an investment in a Voice over IP platform. The marketer is presenting a business case for the launch of the first products on that platform.

The CFO argues that the business case didn't justify the product, as a sufficient return can't be earned from the original investment. The CMO retorts that since the investment was now a sunk cost (i.e. the network was built), the business case didn't need to justify it. And regardless, there were also other products that would be built using the platform. The CMO doesn't think a business case is really relevant for merely trying to get a marginal benefit from the investment.

This is an argument familiar to boardrooms, where participants display great knowledge but little wisdom.

In defiance of the precepts of "logic", both were both right. The CMO was right that it was not the function of a business case for one product to justify the sunk investment in a platform. The CFO was right in understanding that the discounted cash flow impacts of additional investment and costs versus future revenue was the important measure for making budget decisions. 

The error was in deciding that the analysis was either a hurdle to be cleared or equally that it was irrelevant as the bulk of the investment had already been made.

The utility of a cost-benefit analysis

The methodology of a cost-benefit analysis is to look at total social costs and benefits, not just the private firm costs and revenues, as occurs in a business case. 

But the methodologies are similar. Both require estimation of future events and a discounting methodology to compare events in different time periods. Both are conducted in the context of great uncertainty, common costs and multiple separate benefits. 

More importantly, they are similar as they should not be seen as hurdles but merely as tools to analyse components of the decision (fibre versus wireless, distribution through direct sales versus through resellers) and to identify the significance of any uncertainties and to decide whether additional investment is required to reduce the uncertainty.

This is a practical reason for engaging in both business case planning and cost-benefit analysis. 

Minister Conroy and ACCC Chair Graeme Samuel have both dismissed the utility of cost-benefit analysis due to the great uncertainty of the benefits. However, as Doug Hubbard argues in How to Measure Anything, in the process of attempting to measure the unknown, both the extent of the uncertainty can be reduced and the significance of the uncertainty can be revealed.

That is the practical value of using cost-benefit analysis. It can be applied to parts of the project, not just a one-shot total project analysis. 

As opposed to asking the question "does the cost-benefit analysis justify the NBN", it is more useful to ask questions such as "does the marginal benefit of increasing access speeds from 12 to 100 Mbps justify the cost of extending the fibre reach from 90 percent to 93 percent". 

Social and business benefits

The financial case for the NBN is better on the inside-out model, wherein people who already get fast broadband are the first to get it faster. But the priority might be different if the social benefits (as well as the costs) were assessed in prioritising the provision of broadband to those who can't access seriously fast speeds (anything above 2 Mbps) already.

There is also a political reason to embrace a cost-benefit analysis. In understanding the political benefit, it is also necessary to understand that cost-benefit analysis is not a perfect policy tool. 

A cost-benefit analysis only measures the efficiency of a project, not its equity; in fact a cost-benefit analysis is inherently anti-egalitarian. The methodology bases efficiency on the concept of "Pareto improvement"; an outcome is said to be a Pareto improvement if it makes at least one person better off without making anyone else worse off.  In practical application though, cost-benefit analysis only measures "potential Pareto improvement" based on the "compensation principle". In this case a project is to be preferred if it makes at least one person better off and if the gainer could compensate the losers.

Hence, were the project to make Australia's richest man $100 better off, and Australia's poorest person $90 worse off, a cost-benefit analysis would suggest the project should proceed. However, we really know that the richest person's "utility" in that gain is not really greater than the poorest person's utility loss.  

Just because something fails a cost-benefit analysis, it does not make it the wrong thing to do.  The social equity grounds are simply not included in the calculation. Let's assume that a decision to go to 93 percent fibre comes out as having a cost/benefit outcome of $1 billion less than only 90 percent fibre. I would argue that it can still be good public policy to go to 93 percent on equity grounds.

It is perhaps one weakness of the existing equity provisions in telecommunications that the extent of cross-subsidy is hidden from the public. There may be a good reason why I'm not an elected politician, but to me it really seems to make sense to show people the size of the benefit you are delivering to them, not attempt to hide it from scrutiny. 

So who should do the study?

If both business cases and cost-benefit analysis are required, the question is who should do them.  Clearly NBN Co is best placed to do the actual business plan, a key reason why the Government had to establish NBN Co first and get the plan second.

The real question is whether the whole plan should be made public or only parts of it. Here we already hear the cry "commercial confidentiality". This is usually a spurious claim to justify hiding details from the public. It can be valid if release of the plan were to change the behaviour of others.  For example, potential vendors to the NBN might adjust their prices to reflect the NBN Co business case rather than their genuine best price.

To commission a cost-benefit analysis is harder. The resources that used to be available to Government, first as the Bureau of Transport and Communications Economics and then as the Communications Research Unit are no more. These suffered under the Howard Government which cut the economic research capabilities of all departments and funnelled them to the Productivity Commission, and otherwise "outsourced" public policy analysis. 

The last days of the PMG provide some guidance. The PMG commissioned internal work that ultimately was published by Telecom Australia as Telecom 2000, the start of a commitment to genuine public planning. This was a commitment by the then government-owned telco to a public planning process. It is desirable that NBN Co should be required to also co-ordinate all the work of undertaking ongoing cost-benefit analysis to inform the Board and to support their submissions to the Minister about their plans.

But if anyone were serious about seeing this occur, they would not be introducing a private member's bill which smacks of a "stunt". 

The exposure draft of the NBN Co legislation created the requirement for NBN Co to provide business plans for the Minister. 

It would seem to be more appropriate for the Coalition to wait for that legislation and attempt to amend it.