Wonky economics and the energy transition

This is a guest post by Tim Worstall

The Climate Change Committee uses a specific economic model to look at the effects of the planned energy transition. This is sensible; no one can even dream of what the economic effects of something are going to be without having some model of human behaviour to aid in logical thinking. The interesting question is whether the correct model is being used and whether, therefore, we gain useful results.

The answer here, in the grand and overall sense, is "no". This is simply because, in the model being used, it is assumed, at the start, that government spending and government intervention will grow the economy. Therefore the result of the model calculations -- showing that government intervention and spending will grow the economy -- is not an interesting one. We have here petitio principii, or the logical fallacy of "begging the question". We have assumed at the beginning the thing which we now claim to be our result.

If we assume that roses are red and then find that roses are red we have not exactly advanced human understanding by that finding.

The model, which comes from Cambridge Econometrics, includes in its underlying assumptions amusements such as this:

A particular feature of the UK economy is the high proportion of imports in the supply of oil and gas. As the economy transitions from ongoing spending on imported oil and gas in favour of low-carbon domestic investments, leakage from the UK economy is reduced and the implied economic multiplier increases, leading to increases in GDP and employment"

That treatment of imports as “leakage” is arguable, and adherents of most standard theories would argue with it. However, in terms of pure GDP accounting it is correct – imports are a deduction from GDP, while exports are an addition to it. We'd not support this in a general or useful sense as a truth, but in the accounting it works.

Well, OK. That means that if we issue more North Sea licences, or go fracking, then we will gain the same effect upon the economy. Even if all of that fracked gas is exported (it won't be) so that domestic gas prices don't fall (they would), we still gain that effect on GDP: imports are reduced, exports are increased, GDP rises.

This is not, therefore, a useful justification of a move to renewables, for an intensification of fossil fuel production leads to the same result.

We can also point out that this is a minor league version of what is wrong with the modelling as a whole. We wish to find out what are the economic effects of more renewables? Things that can be achieved by not just more renewables, but also by fewer, cannot be ascribed as "effects of renewables". We can – and should – therefore reject the trade effects - fewer imports of oil and gas - argued here concerning renewables. Not because they won't happen, but because they're not caused by renewables. They're an effect of more domestic energy production of any kind – nuclear would achieve the same result just as well and also in a carbon neutral manner.

Here's another of Cambridge Econometrics' views:

Of dynamic innovation, as low-carbon technologies are invested in at scale the costs fall significantly. The cost of renewable energy technologies has already fallen significantly beyond most expectations”

That's also a lovely assumption. If we have more of something, with more investment, then the costs will fall. But that then also applies to – say – fracking or nuclear. France certainly built a nuclear fleet on a cost per reactor basis very much cheaper than other places did. They did this by building a fleet, not one power station at a time. But again, this is something which is true of many technologies; it is not something specific to renewables.

Then there's this:

Moreover, given a likely continued slowdown in the UK economy because of Covid-19 and Brexit, there is more scope than ever for a green economic stimulus. During these times of higher unemployment and low interest rates, there is a strong economic rationale for the government to invest heavily to bring about the transition directly. This would lead to an even faster transition and stronger economic performance."

Well, yes. Many an economic forecast has been brought low by MacMillan's “Events, dear boy, events”. Such as the fact that, since the release of this analysis, the major British macroeconomic problem has been inflation due to excessive stimulation of the economy by government. So we might also want to park that justification to one side.

We don't need to get down and dirty with their socioeconomic assumptions, with their predictions for which technologies will mature at what price. For these are not the things that drive their finding. They're not even the things that drive the model. That trade effect, which we now know we shouldn't use, is a heavy driver of the outcome.

We do, however, need to consider the base problem with the very nature of the analysis. No model, ever, encapsulates all that happens in reality. Every model is an abstraction of a few principles that we wish to examine. This must be true, for if it weren't then we'd not be using a model, we'd be examining reality. What matters, therefore, is in the selection of the model. We need one that explores the effect of the specific action we wish to explore.

But that's exactly what this research does not do. As above, we've a begging of the question instead.

What we want to know is whether government spending upon the renewables transition makes us richer or not. Therefore we need to use a model that is equivocal – or possibly even neutral – about the effect of government spending. Only if we start from the point that government spending is neutral can we then go on to ask whether government spending on windmills (or, to be less perjorative, the green transition) is beneficial. This is a piece of logic that should be obvious but sadly does need to be spelt out.

That's because the model that is being used here assumes that government spending grows the economy. The results gained from the model do not, in fact, distinguish between spending upon green or upon anything else. The effect is purely driven by that original assumption: that more is better.

The details of the economic model being used are at E3ME (or here for the full gory detail). In this model – whether they're right or wrong as assumptions – we find these two:

• Markets are subject to frictions in both the short- and long-runs; prices do not automatically balance supply and demand.
• There is usually spare capacity in the economy, including, for example, unemployed workers.

Both lead inexorably to the conclusion that government spending increases GDP. It's simply what they mean. The conclusion is therefore baked in with the initial assumptions. It doesn't matter what the spending is upon either; the stimulus still happens. So, government spending upon fracking has the same effect as government spending upon renewables. But the Cambridge Econometrics result is presented as if government spending upon renewables will – uniquely – expand GDP.

Which leaves us with just the one final observation. E3ME is a widely used model among EU types who would like to see more government control of investment activity – as the model creators themselves boast. Meanwhile, the model assumes that more government control of investment is a good idea, and thus gives results that suggest that...more government control of investment is a good idea. Funny the people who think that more government control of investment is a good idea like to use it, isn't it?