We are failing to think though the interdependencies in an economy.
We miss economist Bryan Philpott (1921-2000). I miss him personally – we used to have such great discussions – but this column is about how the nation misses his economic experience and wisdom.
We tend to propose economic policies without thinking of their wider repercussions. It is just so easy to say ‘we should do this’ and ignore the consequences. This was nicely, and sadly, illustrated by a recent controversy between Keith Woodford, retired professor of agribusiness at Lincoln University, and the NZIER, which wrote a report on the contribution of agriculture to the economy. It argued that the sector’s contribution was small (4 to 5 percent of GDP); Woodford argued that the estimate was misleading about the significance of the farm contribution.
Bryan would have been irate, because the analytic framework the NZIER report was so limited. It seemed to be saying that the farm sector was so small we could almost neglect it. That would encourage those who want to diminish or even close down the agricultural sector.
How important is farming to the New Zealand economy? To answer that sort of question, we need to consider the following (‘counterfactual’) scenario. Suppose the entire farm sector was to close down; what would be the impact on the economy?
The loss would not be just the amount it contributes to GDP, but the activity of those who supply inputs to the farm sector would also be reduced. In addition, those who process farm outputs – dairy factories, freezing works and transport to the wharves, for instance – would also close down.
Implicit in the NZIER approach would be that all these other activities would switch to doing other things. But to which? (A freezing works becoming an aluminium smelter?) And would productivity be as high in the new activity as it is currently?
Moreover, some of the new activities would have to export or import substitute to replace the lost foreign-exchange earnings currently provided by the farm and allied industry sectors. I take it that the real exchange rate would have to fall to make the new activities profitable.
The result would be a reduction in real incomes. By how much? Almost certainly more than that a misinterpretation of the NZIER report might imply.
Bryan knew how to give quantitative answers to such questions using a suite of ‘computable general equilibrium models’ (CGE) that he and his students developed. I am not going to argue that they gave precise answers, but they gave directions and orders of magnitude, enabling the economy to be seen as a whole, rather than a lot of tiddly little corners.
Bryan’s research program was recognised as world leading, or so those overseas in CGE research told me in the early 1980s. However, the New Zealand models suffered from one major defect: they did not give the answers that the neoliberals of the 1980s wanted, for they were much more pessimistic about the impact of the Rogernomic policies. (And they proved to be right, famously predicting the slow growth that actually happened.)
Did the neoliberals engage with Bryan, drawing attention to the model assumptions with which they disagreed? Certainly not! They closed down the research program’s funding and used their superior political position to undermine the standing of those involved. (This is not a peculiarity of the extreme right; Stalin did the same to mathematical modelling in the Soviet Union; later the principal economist, Leonid Kantorovich, was awarded a Nobel prize in economics.) Even Bryan’s university gave the project little support.
And so the research program ground to a halt; it hardly exists today although two of Bryan’s students – Ganesh Nana of BERL and Adolph Stroombergen of Infometrics – are respected members of the New Zealand economics profession. There are still CGE models around but none have the richness, complexity and insight of Bryan’s program.
(An infamous case was when a couple of overseas economists were commissioned by the neoliberals to demonstrate how much better the Rogernomics tax regime was than its predecessor. To everyone’s surprise, including mine, their ad hoc CGE model concluded the new tax regime was less efficient than the one it replaced. It was a nonsensical conclusion. As Bryan pointed out, it assumed the higher unemployment under Rogernomics was caused by the tax system rather than by macroeconomic mismanagement. The neoliberals dropped this modelling pretty quickly, while FIFO consultants (Fly-ins, Fly-outs) do not have to live with their mistakes.)
It takes a lot of time, resources and experience to keep a CGE model in good shape. More fundamentally, we have forgotten the lessons it teaches; economies involve interdependencies; mucking around with one sector can impact on many others. It is a very narrow economist who ignores this lesson.
In good hands a CGE model is a potent way of exploring economic issues. For instance, there is a view that a part of the farm sector’s success comes from exploiting the environment. What happens if we adopt policies which reduce all this exploitation? Presumably it will reduce the size of the sector and average incomes. How much?
Or we talk about the farm sector filling in the gap from reduced international tourism and from reduced educational services for overseas students. Can it?
Is the closure of the Tiwai smelter as huge as some say? What are the energy implications?
So thirty years later we still need to remember the lesson of interdependencies. Sadly we cannot quantitatively estimates them nowadays. To do so, would certainly help us better understand what the economic impact of the policy options. Instead, one is flying blind like an anti-empirical neoliberal. Bryan, we miss you.