Has wind power saved UK households vast sums since 2010?

Usually I try to avoid commenting in this blog on matters of current controversy However, I am going to break that guideline to discuss a paper by three geographers at University College London that has garnered a lot of publicity in the Guardian and among environmental lobby groups. My reason for doing this is that the paper is a classic example of dressing-up a standard piece of economic analysis in environmental clothes and then making grand claims without understanding the potential weaknesses in the argument.

Let me start with a different example. For more than three decades there have been enthusiastic proponents for adopting some form of carbon taxation as a way of penalising the externality caused by emissions of greenhouse gases. This is a standard application of what economists call Pigouvian taxation of externalities, because it was first developed by the brilliant Cambridge economist A C Pigou. The idea of taxing externalities has been applied in many areas from smoking to road use.

However, advocates of carbon taxes go further. They claim what is called a “double dividend”. The idea is that the revenue from carbon taxes can be used to reduce other distorting taxes, such as ones on employment or labour supply. Thus, it is argued that there are two benefits from taxing carbon emissions – a reduction in the harm due to the externalities of carbon dioxide emissions and a reduction in the economic losses caused by taxes on labour. What a wonderful wheeze – two benefits for the price of one!

Of course, it doesn’t occur to such advocates – or they don’t want to acknowledge it – that the same logic could apply to all kinds of sin taxes. There is nothing special about environmental externalities in the logic, yet strangely the advocates of taxes on smoking or other sins rarely make the same argument. On reflection the reason should be obvious. The whole point of taxing externalities is to discourage them. The goal of Pigouvian taxation is not to raise revenue by taxing smoking but to reduce smoking to the lowest possible level. Thus, success means that there will be no revenue available to reduce taxes on labour supply. The double dividend is temporary at best.

Advocates of environmental taxes should have learned the lessons from practical experience. There was great enthusiasm for pollution charges in Europe in the 1980s and 1990s. The idea was the revenues would be paid into Environmental Funds, which would finance investments in pollution abatement. The approach was widely adopted in Eastern Europe during and after the transition from socialist economies. Initially it worked but within a few years the revenues from pollution taxes fell dramatically because either (i) the polluting industries contracted or closed, or (ii) the taxes and investment funds persuaded firms to reduce their emissions below the permitted levels. So, the approach was a success in achieving its primary goal, but it was self-liquidating as a way of financing investment in pollution abatement.

What any fiscal economist knows is that sin taxes and any other Pigouvian taxes are only useful as sources of tax revenue if they cease to be taxes on externalities. Instead, they become like any other tax on goods or services for which demand is inelastic – i.e. relatively unresponsive to price. It is a standard piece of tax analysis that the welfare loss of using excise or commodity taxes to raise revenue – i.e. taxes on specific items of consumption – are minimised by imposing the highest taxes on items whose demand is least affected by price. That is why taxes on cigarettes and alcohol are favoured because, at least up to a point, their consumption is not greatly affected by price. In contrast, taxing cinema tickets is not an efficient or effective way of raising substantial revenue because there are many alternative forms of entertainment or (today) ways of watching films.

All of this is routine fiscal economics. There is a huge literature on what is called optimal taxation – the kind of stuff that I used to teach at Cambridge 50 years ago. The environmental dressing-up became popular in the 1980s, but within 15 years it was obvious to even the most enthusiastic that pollution charges were not quite the silver bullet that their advocates had assumed a decade earlier. The double dividend arguments for carbon taxes lasted longer but it is little more than a variation on the standard theme of taxing petrol (gasoline) – i.e. demand for fuel does not respond to price, so the welfare loss from taxing it is small. The trouble is that every energy economist knows the underlying assumption is wrong. It is correct that demand for petrol does not respond immediately to higher prices by much, but it turns out to be more responsive to prices in the medium and longer term.

I have explained this example at some length because it illustrates how environmental enthusiasm often leads to a misleading presentation of what are mundane economic arguments. This is exactly what has happened in the case of the paper by O’Shea et al, though it is obvious that they have no idea that they have just dressed up what is a standard but very limited argument in economics.

At the heart of the claims made by O’Shea et al is a simple piece of economic analysis, which runs as follows:

Step 1: Investment in wind generation results in a reduction in demand for gas for electricity generation.

Step 2: The supply of gas is (relatively) unresponsive to the market price, while demand for gas other than for electricity generation is also unresponsive to the market price, Hence, the reduction in demand for electricity generation leads to a fall in the market price of gas.

Step 3: The lower price of gas benefits all users of gas, but primarily those using it for heating, industrial processes, and other purposes other than electricity generation. The benefits of lower prices outweigh the costs of promoting investment in wind generation.

This is an old argument in economics that may be expressed in more general terms. If the supply of a commodity is inelastic – i.e. does not respond significantly to market prices – then any policy that reduces demand will lead to lower prices which benefit consumers of the commodity. The argument is the standard basis for buyer cartels that can drive prices by restricting demand.

The presentation of the argument is a pure case of environmental dressing-up combined with modelling fluff. The logic would be no different if we had substituted all references to wind power in the paper by nuclear power or, indeed, coal generation. There is nothing special about wind power in the economic logic – it is simply a variant of implementing a buyer cartel to drive down the market price of an inelastically supplied commodity. But what correspondent for the Guardian would either understand the point or explain it if they did understand it? So, we get the usual outpouring of self-interested nonsense about renewable energy.

It is not my intention to delve into the technical details of the paper, since it relies on an elaborate and questionable superstructure to dress up a very simple argument. Instead, we should examine the simple but crucial assumption: Is it true that promoting any alternative to gas generation in either the UK or Europe more generally will drive down the European price of gas permanently – not just temporarily? That runs counter to what we have learned about commodity and energy markets over the last eight decades. There have been repeated attempts to exercise monopoly power over either supply or demand in commodity markets. While they may enjoy temporary success, invariably they have failed in the longer term.

Of course, many may wish to point to the influence of OPEC over oil supply and oil prices. The general view of energy economists is that this influence is, as for other commodity markets, much greater in the short term than over periods of 5 or 10 years. More recently, the goals of OPEC have been (a) to stabilise market prices as far as possible, and (b) to preserve its members’ share of world oil supply. The kind of long run influence on gas prices postulated by O’Shea et al’s argument is certainly not consistent with the evidence from other commodity markets.

There is a slightly more complicated logic for European gas that the authors may have in mind. Gas is not like oil because transport costs account for a far higher share of the delivered price of gas in Europe and the UK. Transport involves either large pipelines or heavy investment in LNG gasification and liquefaction facilities. So, the logic runs: the commitment to huge investments in pipelines or LNG infrastructure means that major suppliers like Algeria, Norway and Russia are reluctant to reduce supplies of gas to the European market when prices fall as they need the revenue to meet their financial commitments to build these transport facilities.

However, major gas suppliers are not stupid. They can easily see that they might be vulnerable to the exercise of market power by a buyer cartel. That is why they have usually signed supply contracts with gas customers to match financial commitments for new investments. These contracts are normally take-or-pay arrangements with prices linked to the world price of oil or a similar indicator. Market pricing has become more important as the financial commitments linked to large investments in transport infrastructure have tailed off.

What matters to major gas suppliers is the net revenue – or wellhead price – that they earn after deducting transport costs from the delivered market price. They have a choice between either (a) producing gas now and earning this wellhead price, or (b) leaving it in the ground and selling more gas in the future. While there may be various reasons to keep the supply of gas flowing even if wellhead prices fall, most suppliers will reduce supplies sooner or later if wellhead prices fall significantly. Thus, the assumption that supplies of gas to the European market are unresponsive to market prices is at most only true in the short term and is certainly not true over the full period analysed by O’Shea et al.

Equally important, electricity generation accounted for under 30% of the total demand for natural gas in the UK from 2020-24. Households accounted for 36% of total demand over the same period, with industry and other sectors accounting for the remaining 34%. There is certainly no reason to believe that household and other demand was unresponsive to market prices.

Overall, the logic that a buyer cartel can transfer substantial sums from producers to consumers of gas in Europe is implausible in the longer term. Adding the assumption that such a buyer cartel can, in effect, be organised by promoting investment in wind power extends the argument from one that is thin to one that is little more than posturing. The dressing-up in environmental and renewable energy clothes may encourage an audience with no understanding of the basic economics but it doesn’t alter the logic.

None of this will prevent this and related articles being published in a well-known and peer reviewed journal such Nature Climate Change. After all, the dressing-up is exactly what the journal’s editors want to put out, as it will generate lots of publicity and citations. Such is the state of academic publishing today. One might at least hope that someone will ask the authors to point out that the net benefits have nothing to do with wind power. Indeed, they would have been even larger, because of savings in system costs, had there been a deliberate policy to promote nuclear or coal power.

However, any reference to alternative ways of exercising buying power in gas markets is likely to be very sotto voce and ignored by lobbyists. The dressing-up is the whole point of the exercise for them.

The same was true of the double dividend. Any competent fiscal economist realised that revenue from any tax can be used to reduce distortionary taxes on labour or other factors of production. The fact that a carbon tax is one option to reduce distortionary taxes is uninteresting in the larger context. More important is the question: why countries persist in relying heavily on what are known to be inefficient tax instruments?

In this case any reader should ask why, if buyers can collectively influence the market price of gas, they don’t do so? Perhaps the answer is that the assumption is just wrong, or the effect is too small to justify the costs involved.