Model shows $66bn in benefits, but it’s not the last word

Universities are feeling the pinch. This year and next, course subsidies from the federal government are being frozen.

Universities are feeling the pinch. This year and next, course subsidies from the federal government are being frozen in nominal terms, meaning there is no increase, not even for inflation.

So it’s no wonder that the response of the Group of Eight universities was to commission a new study, launched at the National Press Club yesterday, which enumerates the benefits of the $12 billion or so the nation spends on these eight institutions each year.

The answer from consultants London Economics, which did a similar study for Britain’s Russell Group universities, is that the annual economic benefit to Australia is $66bn, or more than five to one.

On anybody’s scale, this is a massive level of return.

If the motorways and railways being built in Australia’s current infrastructure boom had benefit-cost ratios anything like this then governments would be pouring even more money into these projects. What’s more, they would have no problem in justifying large amounts of borrowing to do so.

Clearly, the consultants’ work needs to be carefully examined. So here we go.

The bulk of the benefits identified by London Economics is in the spillover impact of research, which is worth nearly $24bn.

What is this? It’s where research knowledge “spills over” into other areas of the economy, improving productivity, so that the overall economy grows as a result.

How is this measured? Obviously not directly.

Tracing the paths through which each piece of research has an impact on other sectors of the economy is not possible. So it is estimated through econometric modelling.

London Economics has relied on work done by Amani Elnasri and Kevin Fox, two economists at University of NSW, who analysed nearly 20 years of numbers to estimate the relationship between spending on university research and productivity in the economy.

In its report, London Economics explains how it used Elnasri and Fox’s work to derive the ­numbers.

I’d like to know more about the sensitivities in the relationship. What is the level of confidence in the result and what is the impact if the relationship is different?

Another key part of the paper estimates the benefit to the economy of the teaching and learning carried out by universities. In other words, it is the impact of graduate skills as measured by the greater incomes they earn as a ­result of their studies.

Clearly, this is an effect that lasts over a lifetime, so an estimate needs to be made of the impact many decades into the future.

Once again, data is available to use an econometric model to estimate the answer.

In this case, it is the University of Melbourne’s long-running Household, Income and Labor Dynamics in Australia survey.

Because it is necessary to calculate earnings way into the future, a crucial part of the analysis is the discount rate being used to turn ­future value into present value.

London Economics says that because it has conservatively chosen a high discount rate (7 per cent) it is likely to be erring on the low side in estimating a $4.9bn benefit from teaching and learning. (Taking the rate down to 3.5 per cent yields a higher benefit of $17.5bn from teaching and learning.)

To me, another problem here is the lack of a counterfactual.

Certainly we can compare the actual earnings of people who, in past years, have gone to university with those who have not. But do we reliably know what would have happened to those who did go to university if they hadn’t gone?

None of this is to cast doubt on the value of the report.

This is work that needs to be done and governments need to make more, not fewer, of their decisions based on thorough economic analysis.

But I reckon this report should be a starting point for informed debate, not the last word.

[e-mail]