Is the Chief Scientist’s recent science & economy report really pseudo-science?

Fairfax journo Gareth Hutchens has an article in today’s Herald with the headline:

Australia’s scientists forced to rely on pseudo-science to be taken seriously in Canberra

Well, it is April the first.

Hutchens has a go at the recent Chief Scientist and Australian Academy report: The importance of advanced physical and mathematical sciences to the Australian economy (pdf). They engaged the Centre of International Economics to conduct an economic analysis that used the MMRF-NRA computable general equilibrium model to estimate the impacts of a number of input assumptions on the contribution of the physical and mathematical sciences to the economy. The report estimates that the direct contribution of the advanced physical and mathematical sciences is
equal to 11% of the Australian economy (about $145 billion per year). Along with the direct contribution, the report estimates additional and flow-on benefits of another 11%, bringing total benefits to just over 22% (around $292 billion per year).

I felt I had to defend the report, which is not perfect but necessary (I think I also agree with the headline). In doing so I find myself in defence of CGE economic models (which I can’t quite believe I’m doing). Basically, Hutchens reckons that by engaging with an economic consultancy and an economic model, the Chief Scientist and Australian Academy of Science have prostituted themselves (my words) to the same economic lobbying that everyone else in Canberra uses to argue for government support. Here’s my response posted to the comments of the article (slightly edited – doesn’t seem to have made it through, either):

This is a really disappointing article because it completely fails to grapple with the conceptual framework used to inform this work and takes it all out on the model (which was developed the Centre of Policy Studies formerly at Monash, now VU and adapted by the PC). The project was not “get a hired gun”, it was clearly a collaboration.

The report discusses the difficulty with estimating scientific contribution to the economy – to dismiss it as garbage in is to actually dismiss a very simple concept. Each contribution was assessed by someone who knew that industry, but as educated guesswork, most will be wrong. The issue is to whether collectively they are biased too high or low. So they are normative. All applied science is normative because it’s done for a purpose (the stuff about curiosity driving science in the report is cute, but only correct some of the time).

At this level, this concept could be applied to a simple input-output table. The question is whether the basic approach is valid. The model was used to estimate added economic contributions and is almost certainly wrong, but is it useful?

The overly complicated model argument is tosh – it’s exactly the same one aimed at climate models. Granted, a CGE model has lots more assumptions than a scientific model, but the basic issue is whether the effect it simulates is roughly in the right direction and of the right size.

The sad thing is not that the Chief Scientist and the AAAS had to argue the economics of science – the sad thing is that the connection between science and economy has become so tenuous in the current political environment that they had to make the point. In a knowledge-driven economy, the link between knowledge and economics needs to be known.