Stern is wrong again

Lord Stern's latest paper on climate change suffers from the usual problems, but there's a nasty failure of logic at the heart of it, which is rather larger than his standard errors.

Stern notes that renewables and other green technologies have been getting cheaper over the years. They are also, as he estimates, likely to continue doing so. And he's probably right in this assumption. The illogic comes in insisting this means that we should be making swift and large efforts to roll those technologies out now. The correct reaction to something becoming cheaper by the day is to delay getting hold of it.

The reporting on Lord Stern's paper is, as in The Guardian, about how the discount rates used undervalue young lives. This is an assumption that doesn't quite work; by using a very low discount rate – as recommended by the Stern Review of course – we are already close to being indifferent as to when a year of life takes place. In other words, this is already a problem solved by the earlier calculations.

In the paper itself we gain the usual...well, casuistry isn't too harsh a word:

“Our current emissions pathway implies that we are headed for temperature increases of more than 3°C”

No, it doesn't. We've already done enough to avoid that, even using the standard IPCC assumptions. The developments of renewables and their likely rollout in the coming decades have diverted us from that path. The problem is revealed only a couple of sentences later:

Under a business-as-usual scenario, one of the most densely populated regions in the world, the North China Plain, would likely experience deadly heatwaves later this century with “wet-bulb” temperature exceeding the threshold defining what people can tolerate while working outdoors (Kang and Eltahir, 2018).

As we've noted in an earlier paper, we need to be extremely careful when climate activists start throwing around the term “business as usual”. For checking that referenced paper we get this:

We perform simulations for historical period (1975–2005), as well as future climate (2070–2100) assuming two scenarios of GHG emissions28 (BAU scenario (RCP8.5) and moderate mitigation scenario (RCP4.5))

The definition of business as usual being used is RCP 8.5, a greenhouse gas emissions pathway we already know isn't going to happen. So we're being offered a possibly mild overestimate of how bad it's going to get, then scared silly with examples of what absolutely isn't going to happen, in order to, what, scare us? That this is a common technique doesn't make it an acceptable one.

As we explained in that earlier paper, RCP 8.5 is based upon the earlier A1FI scenario. This was originally designed as the very far edge of the envelope, a little beyond possible bounds, to act as something between a warning and an outer limit to estimations if absolutely everything was done the wrong way. It assumes that as a society we don't develop renewables, we don't frack for unconventional oil and gas, we run out of conventional such and so go back to being a coal powered society. So much so that we end up gaining more of our energy supply from coal than any previous global society has done, more even than in the steam age. This never really was going to happen and now definitely isn't. We have developed fracking, we have developed renewables and the use of coal is falling precipitately as compared to the model's assumption of ever more in both volume and intensity per unit of GDP.

RCP 8.5 simply is not a valid assumption to make about any future.

But Stern's really big logical error is in this from the precis:

Any reasonable estimate of the costs of inaction would be still higher now, and the costs of action lower, than in 2006.

This is then used as the basis of the argument that we must be doing much more right now. Which is entirely the opposite of the correct deduction from that fall in the costs of the new technologies. That something is getting cheaper day by day – as above, this is something Stern assumes will continue to be true, and we agree – is an argument for delay in installation, not for an acceleration. Because things that are getting cheaper day by day are cheaper tomorrow than they are today.

Take this from the paper itself:

Installing solar today costs us only 18% (by their numbers note, not anyone else's) of what installing the same power supply would have done as recently as 11 years ago. The paper, probably correctly, assumes that something like this cost reduction is going to continue. This is an argument for delay in deployment, not acceleration.

As to when is the right time to deploy, clearly that's when the new is cheaper than the old. Stern's claim is that this is already true. His claimed concern is also that whatever is built now will be extant for some decades, so it's right that we should be installing emissions-free. But if that's already cheaper, then no effort is required to encourage it. If it's not then those price reductions prophesied mean that the deployment date should be delayed until it is. This is true whether we use raw costs or those loaded with claimed climate costs; the logic is the same in either case.

One possible exception would be if it were deployment itself which was reducing costs. At least a nod is made to the idea that this is so - that it is entirely economies of scale of deployment which are driving down costs. Which is a most odd thing for an economist to be saying. Certainly, there's a correlation between deployment volume and cost, but this can and does work either way. A new technology becoming cheaper can and does drive greater adoption – that's just standard supply and demand curves. To try to insist that it is fully and only economies of scale driving cost reductions is a necessary precondition for the argument that we must “go big” now and much too strong an assumption to bear the weight of the argument.

This before the economists point that, while there are indeed economies of scale, so also are there diseconomies. We know, for example, that grids become very much more expensive once the load of wholly variable renewables rise above a certain portion of total energy supply; an issue we saw just a few weeks ago as the wind stopped blowing.

The macroeconomic part of the paper suffers from having been prepared in the summer. The contention is that the economy suffers from too much unemployment and therefore investment in green schemes can boost matters. In the current economy suffering from inflation and labour shortages this is a less sensible idea.

It's tempting simply to say that, given the imminent start of COP 26, arguments for doing much more, and immediately, must be made. Continued falls in the costs of renewables are arguments in favour of delay in deployment, not acceleration. But only a cynic would assume that arguments are being made because they're fashionable, or politic, rather than correct.