Net Zero Substance

You can find a potted version of this article on the Institute of Economic Affairs' blog.

Introduction

The Global Warming Policy Foundation’s new board member, Steve Baker MP, published an article in The Critic, cautioning about the cost of Net Zero, and the lack of critical scrutiny of the policy in Westminster and the media.

One of many critical responses came from Tim Lord, Senior Fellow, Net Zero at the Tony Blair Institute for Global Change, and former civil servant at the Department for Business, Enterprise and Industrial Strategy, where he was a Director responsible for the UK government’s decarbonisation strategy.

Many of Tim’s and Steve’s points were their estimates of the likely costs and impacts of Net Zero. As someone who has been in the business of decarbonising energy for over 30 years, I too have an opinion on the likely impact of Net Zero. I lean mostly towards Steve’s view.

BUT, I recognise two limitations to my assessment:

1. I might be wrong. No one has a perfect crystal ball. Innovations may surprise us all. That is a good thing, to be encouraged.
2. These debates are stagnant, because no one can prove they are right or wrong about the future until we get there. If policy is determined by the predictions of government advisers, no amount of evidence and arguing that they are wrong will change it. It comes down to power, not reason. We get what those in power think we ought to get. Anyone outside government’s circle of influence who has ever responded to a consultation knows the truth of this.

Net Zero is a symptom of a bigger problem – the way that government policy is set. It cannot be tackled by offering different assessments of the future, however well-founded.

I was reminded about modern governments’ repudiation of the knowledge problem and about outsiders’ impotence by one of Tim’s tweets. He described the International Energy Agency (IEA) assessment that “rapid transition this decade can create more jobs than it costs” as “evidence”. Are predictions about the future evidence? It made me realise that we need to step back and look at the epistemics:

1. How do we deal with compound uncertainty?

2. How do we draw the system boundaries?

3. Where does humanity fit in?

Caveats

1. This is not about the probability of climate change. Some of the points are relevant, but this is about the uncertainties of how to proceed assuming climate change is real and man-made.
2. The risk of ecological disaster in no way converts fallacies into truths.

1. How do we deal with compound uncertainty?

Nothing about the future is known. I would prefer that “evidence” refers only to data that support or falsify hypotheses, which inherently do not exist yet for predictions. But I am not resting on semantics. We can consider predictions as evidence if you want to abuse language that way.

What sort of evidence are predictions? At best, they are probabilistic. In most cases, they are not even that, at least not explicitly. The crystal-ball-gazers combine a bunch of assumptions without stopping to consider the degree of confidence for each, let alone how their probabilities combine.

It is a common error to think that combining uncertainties reduces rather than increases uncertainty. Clever minds did that with sub-prime derivatives. Theirs were sophisticated methods applied to limited uncertainty, compared to intellectuals prognosticating about the state of the economy, environment and technology decades in the future.

Even if each of the assumptions under-pinning the supposedly low-cost routes to Net Zero 2050 were almost certain, there are huge numbers of them. The product of hundreds of near certainties is almost complete uncertainty.

Many of the assumptions are far from certain. It is said (e.g. by John Kerry, Bill Gates and the IEA) that the technologies needed for half of Net Zero have not been invented yet. Some dispute that, either semantically (they have been invented, they’re just too immature to be viable), or conspiratorially (there are cheap solutions available now, that just haven’t been widely deployed for some strange reason). It is immaterial as far as uncertainty goes. They highlight, not diminish, the uncertainty.

If our knowledge were greater, Net Zero advocates might claim to use the probability distribution around each assumption to mitigate this compounding of uncertainty. But we cannot have such knowledge with regard to speculation about the future. In practice, they don’t even try.

Typically, their assumptions are optimistic (or for sceptics, pessimistic), i.e. they would be on the tail, if normal distribution were relevant or considered, which it’s not. It is not even treated as a bare probability, let alone a distribution. Mostly, the assumptions are chosen for convenience, with only a narrative justification of their likelihood. It’s just “we’re assuming this because we think it’s possible and our model needs it”. This is a quicksand foundation for any policy structure.

Even the pathways to 2030 are immensely (but tacitly) uncertain, let alone to 2050. They try to fudge the issue by saying “they’re pathways not predictions”, but what value are they in that case? If I am allowed to discount uncertainty on that basis, I can find a pathway to almost anything.

If the pathways are “evidence”, it is because we are confident they can be followed, i.e. pretty certain. If we recognise the uncertainty, they are evidence of nothing. You can have 100 pathways. If they are all highly uncertain, they are cumulatively evidence of nothing.

“This is nihilism: do nothing because we don’t know the best way for sure.”

No. Shape your institutions to the nature of the challenge. Dirigisme is not the only option. Pervasive uncertainty does not prevent action, but a different approach is better than assuming the uncertainty away.

Hayek differentiated between taxis (human-imposed order) and cosmos (emergent order). A taxis is applicable where its components can be known and structured by a central planner. Where the complexity and uncertainty are too great for a taxis, order is better achieved through a cosmos. Net Zero is an attempt to impose a taxis where the uncertainty mandates a cosmos.

[Hayek sketched these ideas in “The confusion of language in political thought” and then fleshed them out in Law, Legislation and Liberty.]

As an emergent order without a central planner, a cosmos cannot be harnessed to deliver a particular outcome. But its rules (nomos) can be augmented to embed considerations within the choices of the actors. The outcome will reflect those considerations, but not necessarily what a central planner might think ought to be the outcome of those considerations. In a cosmos, outcomes are discovered, not planned. “Internalising the externality” (i.e. polluter pays) embeds climate change as a consideration in the choices of all actors within the cosmos that is our society and economy.

With the externality internalised, our cosmos might end up at Net Zero, but (a) it does not aim for it, and (b) the chances are tiny that the welfare-maximising outcome happens to be Net Zero. That is a feature, not a bug.

If the outcome is other than Net Zero, there would have been a net loss of welfare had we aimed for Net Zero.

If the outcome is Net Zero, then there is no disadvantage to having taken the decentralised option (cosmos). In fact, it was probably an advantage, unless the discovered route to Net Zero was exactly the same as the planned one would have been.

In other words, there is no world in which a centrally-planned approach would produce a more welfare-optimising outcome than a decentralised approach, for problems as uncertain and complex as addressing climate change.

One of these is a fallacy:

(a) The problem is so severe that dirigisme is essential. Besides, alternative approaches are unpalatable. Dirigisme needs confidence that the plan will have the expected outcomes. We can’t have that, but the need for this approach to be viable means we can behave as though we can.

(b) We can’t be confident in any plan, so dirigisme is not sensible. We should use means that are more effective under uncertainty, even if they are unpalatable. Better to promote the solution that is rational but unpalatable than the solution that is irrational but palatable.

Irrational policies don’t work. The comfort is illusory. The right option is rarely easy. If it were, we wouldn’t need governments. A politician’s job is not to pander to comfortable fallacies. It is to explain that the uncomfortable reality is in people’s interests in the long-run.

2. How do we draw the system boundaries?

Everyone plays with system boundaries. The ubiquitous example in this field is failing to distinguish between energy and electricity. But I am thinking of a more subtle example from the economic modelling. The IEA’s Net Zero roadmap will add 25 million jobs net by 2030, apparently (p.17):

Let’s put aside how they know that. That is covered in point 1 above. Even if true, is it the whole picture? Increasing the labour required for a key economic input is not usually a recipe for prosperity.

So it is not only the extraordinary capital investment whose costs have to be covered. It’s the increased labour costs too. It is a reasonable inference that this will have an impact on energy costs (especially considering experience in the UK over the past decade).

If so, what is the impact on the rest of the economy? Some other consumption must give, and jobs in those areas will go. Are they factored into the impact of Net Zero on jobs? Of course they aren’t. Factoring that in can yield a very different answer.

On the other hand, the extra direct jobs in the Net Zero pathway could mean more spending, and more jobs outside the system boundary. Indirect factors can flow both ways.

If government-directed investment generated wealth and jobs more effectively than undirected investment we would see a lot more of it and stronger growth. The net effect is highly uncertain. The system boundary is drawn tightly around the direct elements not because the indirect elements are insignificant, but because incorporating them would exacerbate the endemic uncertainty.

But the effects are not limited in practice to the direct impacts, even if we have a reasonable inkling what they are. We should choose a system that takes account of all important factors, not exclude important factors because they make the chosen system unworkable.

Jobs are just one example of this problem. If you think you can counter this point by pretending to understand enough to quantify the jobs impact across the economy, you have failed to understand the point in two crucial ways.

3. Where does humanity fit in?

By “humanity”, I don’t mean mankind as a species. I mean human nature.

The human species is unique for the extent that it can abstract, hypothesize, experiment, refine, communicate and learn.

For most of mankind’s existence, the benefit of these cognitive capabilities was limited by top-down social orders.

Humanity’s true flourishing occurred when feudal structures decayed and social structures emerged that facilitated cooperation outside the “tribe”. These allowed for progress not only when the elite happened upon an innovation that was useful to them, but also for innovations outside the elite and/or that benefited others than the elite. The coincidence of aptitude and experience being significantly random, progress is bound to accelerate where there are opportunities for contributions outside the ruling group to prosper.

The trend back to dirigisme is a reversal of this development. The only innovation that is welcome in a dirigiste system is one that furthers the central plan. There is no space for heresy. This is de-motivating for the people whose imagination a system ought to harness to maximise innovation.

Businessmen in a dirigiste system will do better as “yes-men” than as radicals. They must maximise compliance as much as utility. At best, innovators will waste resources trying to make their ideas compliant. Many will be deterred by the barriers to their non-compliant ideas.

It is not a minor consideration that a cosmos is not only more rational than a taxis for these conditions, but also more conducive to making the most of human nature. It is a better world to live in. Why would anyone want to downplay the abundant uncertainty in order to justify an approach that is as unattractive as it is irrational?

The 2003 Energy White Paper

Governments have been increasingly engaged in imposing taxis where cosmos would be more suitable. Hayek noted that it was already happening in 1968.

The 2003 Energy White Paper (introduced by the then Prime Minister, Tony Blair) paid only modest lip service to uncertainty and discovery. Mostly, it viewed innovation and investment as things for government to direct.

We can judge how well-founded was its confidence in dirigisme, because it predicted (pp.18-19) the state of the energy industry in 2020, resulting from the proposed measures. A large number of them fell wide of the mark. Many reflected over-optimism about cost reductions in technologies that are once again being talked about as major contributors in 2030 or 2050 once they have been “driven down the learning curve” (successfully, this time, they hope).

• Nuclear would have been shut down by now,
• “Fuel cells will be playing a greater part in the economy”,
• Hydrogen for the fuel cells “will be generated primarily by non-carbon electricity”,
• Hybrids “will be commonplace” (PHEV market share in 2020: 4.4%)
• “Substantial and increasing use of low carbon biofuels” (stagnated at <5% until boost to c.8% in the last couple of years)
• “Hydrogen will be increasingly fuelling the public service fleet and utility vehicles. It could also be breaking into the car market”
• “Nuclear fusion will be at an advanced stage of R&D”
• Reductions in domestic gas use will be “offset by demand for gas from CHP”
• “Many buildings will have the capacity to reduce their demand on the grid, for example by using solar heating systems to provide some of their water heating needs”
• “There will be much more micro-generation, for example from CHP plant, fuel cells in buildings, or photovoltaics”
• “Some of those large power stations will be offshore marine plants, including wave, tidal and windfarms” (1/3 ain’t good)
• “There will be much more local generation, in part from medium to small local/community power plant, fuelled by locally grown biomass” etc.

The paper had already banked (p.12) the projected delivery of the Renewables Obligation by 2010: 10% renewable electricity, costing £1bn p.a. It aimed to double that by 2020. This was exceeded in practice. But they would have been surprised to learn that by 2020 costs had risen not fallen. We spend more than £10bn p.a. supporting renewable electricity. That does not pay for 10 times what £1bn was expected to pay for in 2010, let alone the higher ratio that falling costs would imply. In reality, the cherry-picking effect (people do the easy things first) and diminishing marginal returns often outweigh the learning curve.

Nevertheless, we are told that renewables costs are falling. Very clever government advisers, and highly reliable metrics like LCOE say so. Rising consumer costs are purely coincidental. Oligarchy is fool-proof and in everyone’s best interests. “Evidence” of the future (i.e. expert projection) disproves the failings of the past. This time is different.

An analysis of any other government energy policy of the past 30 years would lead to a similar conclusion. Reality usually varies significantly from the government’s expectations. We have a long enough track record that we should no longer fall for the pretence that government can predict how the energy industry ought to look and direct it efficiently to that outcome. But if it can’t, what substance is there in the claims about the costs of Net Zero?

Conclusion

Net Zero is a depressing, inhuman, irrational, impoverishing concept.

It cannot be de-coupled from central planning. If uncertainty were acknowledged and discovery incorporated, it would have to allow that one discovery could be that Net Zero is not the optimal level.

The Net Zero debate will continue to focus on the credibility of different pathways. That is a red herring, and an unwinnable war.

There is no real evidence until we get there, by which time it is too late. Advocates can’t prove they are right, but sceptics can’t prove they are wrong.

It comes down to power. Sceptics won’t win that, and neither will society. Besides the direct costs of imposing detrimental policy, there is also huge damage to democracy when governments claim to have capabilities that they don’t, and are exposed by unintended outcomes.

Trust and social cohesion are lost. They are key to economic cooperation. We lose directly and even more indirectly.

We do best when people understand the knowledge problem, and when governments limit themselves (outside narrow areas where taxis is applicable) to supporting the nomos.

Debates about the costs and benefits of an immensely complex and uncertain government plan are beside the point. We will all continue to engage in it, because the questionable assumptions make it hard to resist a challenge. But it is largely performative. It will make little difference.

If we are in a decentralised system, we will discover what works and what is worth it. If we are in a centrally-planned system, we will never know if better options were available.

If you think Net Zero is the right target and certain technologies will get us there cheaply, if you are right, discovery will validate your analysis. Ironically, that is not so under central planning. If you are wrong, discovery will allow alternatives to emerge, whereas dirigisme would have locked us in to the wrong course.

Under dirigiste climate policy, we will not discover if the dominant analysis is wrong until serious damage has been done. If it is sub-optimal but not disastrous, we may be stuck with it forever, unable to prove that we could have done better.

It is a miserable vision with a tiny chance of doing as well as a more rational approach, and an overwhelming likelihood of doing worse. That’s not because the experts aren’t clever and doing their best, but because it is based on a misunderstanding of knowledge and humanity. There is no virtue to it.