There is a consensus among energy experts, environmental economists and some of the world’s most venerable institutions that one of the most promising solutions to the CO2 problem lies in encouraging us to use energy in a more efficient manner.
It’s not hard to see the attraction behind the idea: benevolent governments, faced with a need to reduce CO2 emissions while at the same time not threatening the living standards of their electorates, should order that machines and appliances use less energy to do the same things. The rest of us (in the developed world at least) can then carry on driving, heating our homes, using aeroplanes and seeing things at night while emitting fewer gigatonnes of CO2. By virtue of enlightened commandment, the amount of energy used declines and with it, our CO2 emissions.
Oh, and did I mention that at the same time, by using less energy, we all save money? Yes, you read it correctly the first time round. Not only can improvements in energy efficiency cut mankind’s emissions of CO2, they also make us better off at the same time. By this logic therefore, the very first thing that sensible governments should do to cut CO2 emissions is to make us all use energy as efficiently as possible, as quickly as possible.
This all sounds good, doesn’t it? And, in fact, almost everyone seems to agree on this. It’s highly likely that most Departments of the Environment, consultancies and industrial think-tanks will provide online links to policy plans that would allow their government to help consumers choose energy efficient appliances. They may go still further and set out standards to which manufacturers of machines and appliances should conform (I’ll be looking at the claims of the most famous energy efficiency advocates in a later article), so that consumers don’t even need to think about efficiency when out shopping.
However, I believe that the logic underlying this happy story is unfortunately wrong. I do not mean that it is completely wrong – as aspects of improving energy efficiency can, and do, have merit. The energy efficiency story is partially wrong, but sufficiently flawed that the advocates of increasing energy efficiency will not, as they hope, help deliver us from the CO2 problem. In fact, they and their simplistic way of thinking about the CO2 problem, pose a real danger to the climate.
Objections to the efficacy of increased energy efficiency (ie, in reducing energy consumption) have a long and sometimes overlooked history. Back in the 1860s, William Jevons argued that technological improvements that increased the efficiency with which a resource was used would, instead of reducing the rate of consumption of that resource, lead to further demand for it, and thereby greater consumption. When Jevons came up with this idea, the thermal efficiency of coal-burning steam engines was increasing rapidly (from single to double figures). Expectations that this gathering technological revolution would conserve resources foundered on the reality of rocketing demand for coal.
The rebound effect was born: an increase in efficiency leads to less demand locally in the short term, but at the same time, makes the unit cost of a particular resource lower, which actually encourages further consumption.
In my first article, I introduced the admittedly clunky term ‘man-and-machine-kind’, and described how the beginnings of the industrial revolution allow mankind to harness the earth’s resources to its own ends, but how, at the same time, this kicked-off a CO2 problem (already in 1861 Tyndall wrote about the heat absorbing properties of CO2 and in 1896 the distinguished Swedish scientist Arrhenius worked out the link between higher atmospheric CO2 concentration and the retention of greater amounts of heat energy in the atmosphere). The duality in the existence of man-and-machine kind hints at the rebound effect being all too plausible. If our more efficient machines require us to dig up fewer resources every time we flick the on-switch, we’ll want to use them just that little bit more, thereby offsetting some of the original promise in energy efficiency.
This argument about the nature of improvements in energy efficiency has gone on almost continually since then. In the 1970s, following on from the 1973 oil shock, the economists Daniel Khazzoom and Leonard Brookes looked at the effects of changes in energy efficiency from a macroeconomic perspective and confirmed Jevons’ earlier hypothesis. Micro-level energy efficiency improvements may work locally (ie, where they are applied), but at a macro level energy consumption ends up rising.
Have these objections had much of an impact on how people think about energy efficiency? Have they given pause for thought to those who are inclined to commend increased energy efficiency as a solution to the CO2 problem, as set out above? On both counts, the answer can only be ‘no’. Absent of any definitive proof that the rebound effect could be measured, energy efficiency advocates argue quite fairly that at least most of the improvements from energy efficiency remain.
Attention among academics and econometricians (whose tools have been getting sharper faster than reliable data can be collected for them) has been concentrated only on this “rebound effect” discourse. Obediently, study after study has been undertaken to try to pin-down the rebound effect. Is the rebound effect nearly one, meaning that all increases in energy efficiency are counteracted by an equal (and offsetting) increase in demand? Or might it be closer to zero, in which case, we can sweep up the ideas of Jevons and put them into history’s bin?
As it turns out, those studying the rebound effect econometrically ended up stumbling upon the world’s first truly computer-aided random number generator – albeit a very slow one. Given ever-greater volumes of data, and econometric models of varying complexity, the distribution of estimates of the size of the rebound effect is embarrassingly random. ‘Somewhere between zero and one’ is still about as much as we can say.
Well, it’s almost as much as we can say. We can actually add that from a CO2 point of view, the rebound effect is a mere distraction: even if you could measure it correctly, you’d still be missing the point. What no one has thought about is linking energy efficiency to the amount of resources in the ground that can be economically extracted, regardless of the rate of extraction, or when that extraction takes place.
To use a crude medical analogy, if a patient is suffering a fever, do you look only at alleviating the symptoms (in which case, placing the feverish patient in a bath of icy water would be effective), or do you consider the underlying cause? If we have a CO2 problem, do we try to tackle it by tackling the rate at which we use fossil energy (in which case, increasing energy efficiency will help, rebound effects notwithstanding), or do we take a step back and ask why we are digging up and burning fossil energy in the first place?
Perhaps the easiest way to think about the problem with energy efficiency is to set ourselves, say, 10,000 years into the future. Let’s also imagine that we can see and account for how many resources remain in the earth’s crust, just as we can already measure the composition of the atmosphere.
So we look 10,000 years into the future, and we’re be able to know just how much coal we dug up and burned, how much oil and gas we drilled and burned, and how much remains. Importantly, from the point of view of the planet at such a distant point in the future, the rate at which we extract fossil fuels and burn them to emit CO2 today (and even in the coming century) will cease to matter when compared with how much we ultimately extract. If in the end, all our attempts to solve the climate problem fail and we dig up sufficient amounts of fossil fuels to cause a climatic disaster, then who will really care if that climate disaster occurs in 100 years, 200 years or 5,000 years? If the best of our knowledge around climate science turns out to be correct, then man-and-machine kind will have caused the climate to change disastrously: the moral responsibility for that fact stands first and foremost – arguments around when and how it might have happened will be trivial by comparison.
By this means of looking at the climate problem, we need to concentrate first and foremost on how much coal, oil and gas can be economically dug-up and burned. The rate at which we do this today, tomorrow or in 10 years’ time is simply not that important from a climate point-of-view (while emissions now clearly don’t help, they are actually not the issue – the issue is around the timing of the last lump of carbon that we feel the need to dig up).
Where does energy efficiency come into this, then? Well, just as Jevons pointed out, increasing energy efficiency makes the per-unit cost of using that energy cheaper. So if you buy a new car which, thanks to a decisive government’s efforts to tackle the CO2 problem, is twice as efficient as your old car, then with fuel prices constant (and in the short-term they might even drop….) you spend half as much per mile as with your old car.
Now, instead of engaging in arguments about short and long-term rebound effects, let’s think about whether an improvement in the efficiency of a car will have any impact on how much oil can be economically extracted. We can set up a little thought experiment….
So, first of all, you’ve bought a new car and found to your delight that it contains new technology, as encouraged by your government, and only costs you £30 per week in petrol instead of the £60 per week you paid with your old car (that is, petrol costs £6/gallon; you drive 300 miles per week; your new car does 60 miles per gallon where you old one did only 30 miles per gallon). If Jevons is wrong, and you don’t change your driving habits, (continuing to drive 300 miles per week, even if it’s much cheaper to drive now) then whereas before you were consuming 10 gallons of petrol per week, you only consume 5 gallons now.
At this stage, advocates of energy efficiency will have us believe that you’ve cut your CO2 emissions by 50% (well done, you!). However, just as the icy bath soothes the feverish patient, the measure designed to help tackle the CO2 problem might not be doing quite as much good as they’d have you believe.
Looking a few more years into the future, more people have these nice, efficient new cars. They carry on driving, happy in the belief that their new cars have cut their road transport CO2 emissions by 50% (compared to the old dirty 30mpg cars that they used to drive). Oil is being extracted still, but even with Jevons being ‘proven’ wrong, the rate at which it is being extracted has dropped – a sure sign of success, perhaps.
Extraction of oil continues apace, and about 20 years later the price has to rise, as the most productive fields have been exhausted, and new oil fields cost a lot more to develop per barrel of oil than the ones they replace. Eventually oil gets so expensive that petrol ends up costing £12 per gallon. Doubling the price of oil! That must have an impact, surely? Well, no. At £12 per gallon, at 60 miles per gallon, and still driving 300 miles per week, our drivers will still be able to afford to drive using oil-based fuels. Why might this be?
Well, years before, when their old cars only did 30 miles per gallon and petrol cost £6/gallon, their 300 miles per week cost them £60. Now that petrol costs them £12/gallon, in their new cars it still costs £60 to drive their 300 miles per week. This sort of calculation is highly significant, as doubling the energy efficiency of their cars meant that drivers would still be able to afford to use their cars, even if the price of petrol doubled (in this case, caused by an increase in the cost/price of oil, rather than by an increase in fuel taxes).
Let’s take this one step further. If the efficiency of cars were to double again, then drivers would still continue to use oil-based fuels if the price of oil rose so high that petrol ended up costing £24/gallon. Those 300 miles per week would still cost the same, even if the rate of consumption of petrol had dropped also.
This illustrates the potential perversity of government policies to increase energy efficiency: by making machines even more useful to mankind (as each unit of service they provide to man ‘costs’ less), increases in energy efficiency increase the ‘usefulness’ of fossil fuels. That is, for a given unit of service provided by a machine, we are willing and able to pay more for the fossil fuels that run the machine. And the higher the price we can afford to pay, the more of those fossil fuels can be economically dug-up and burned. Another way of putting it is that we dig up fossil fuels and burn them because it is useful – if we can convert them to useful energy service more efficiently, then fossil fuels become even more useful, irrespective of the rate at which we extract them.
So, if pursuing increased energy efficiency is flawed idea, why isn’t it completely wrong?
In my view, there are two forms of energy efficiency improvement, just as in macroeconomics literature Hicks-neutral, Harrod-neutral and Solow-neutral technological advances model the impact of technological improvements on demand for labour and capital.
By this analogy, the first type of efficiency improvement can be labelled “carbon-augmenting” technology changes. Just as set out above, carbon-augmenting technology changes increase the efficiency with which specifically fossil fuel is converted into end use (be it, miles driven, degrees-days heated, etc). These types of efficiency improvement, by themselves, increase the amount of fossil fuels that we can ultimately afford to dig up and burn. Hence by themselves, carbon-augmenting technology changes make the CO2 problem ultimately worse.
If, like me, you reject the idea that biofuels can ever replace oil in transport (or gas for that matter) without a disastrous impact on the world’s rainforest cover then the following efficiency improvements can be looked at as carbon-augmenting: high efficiency central heating boilers, variable valve timing technology, vehicle hybridisation, common-rail diesel injection, automotive turbochargers, use of carbon-fibre in aeroplane construction, and turbines which enable higher temperature and pressure combustion in jet-engines. This is by no means an exhaustive list!
The second type of efficiency improvement can be labelled as – you guessed it – “carbon neutral” technology changes. These mean things such as: home and office thermal insulation, low-loss power transmission lines, high efficiency light bulbs, high efficiency household appliances, and electric-only vehicle propulsion. While this type of efficiency improvement still enables users to be able to afford to use more expensive fossil fuels, it also allows users to afford similarly more expensive non-carbon energy as an input, be that renewable energy or nuclear energy. From both a short and a long-term viewpoint, this can be quite useful, as renewable energy is limited in most places, so it is actually imperative to be able to extract as much end use as possible from whatever renewable energy can be harvested from the environment. However, carbon neutral energy efficiency improvements are of limited ultimate use unless consumers themselves are incentivised to switch to non-fossil technologies (or otherwise fit CCS – Carbon Capture and Sequestration, although this is still more of an idea than a proven technology).
To re-cap, carbon-augmenting improvements in energy efficiency are actually counter-productive to reducing CO2 emissions. By contrast, while carbon-neutral improvements in energy efficiency make no real difference to the relative end-user competitiveness of fossil and non-fossil fuels, they at least enable the same level of increase in the affordability of non-fossil sources of energy as of fossil fuels. Furthermore, they make it more likely that the demand for energy service can be satisfied within the constraints of local availability of non-fossil energy.
Finally, there is one last major aspect of energy efficiency that needs to be understood properly: tax. Taxes on fossil fuels are, in my view, our only effective tool to prevent emissions of CO2 (by this, I mean either a direct tax on emissions, or a traded emissions certificate-type scheme). If we understand the interaction between fuel taxation and energy efficiency choices properly and that there are no alternatives to taxes on carbon-based fuels, then the political resistance to fuel taxation will be more easily overcome. Moreover, both resources and valuable political attention will not be wasted on the designing muddle-headed interventions to encourage the wrong types of energy efficiency improvements.
Absent a tax or price on CO2 emissions, pushing for ever-greater increases in energy efficiency is like attempting to play snooker with a length of rope. Instead, governments should use heavy taxation of CO2-emitting fuels to drive permanent and irreversible reductions in CO2 emissions. By making fossil fuels more expensive to end-users through tax, policy can act to prize apart the carbon linkage in the man-and-machine-kind relationship. Energy efficiency would then be seen (in my view, rightly) as having a subordinate role – namely, to enable consumers to counteract some, or all, of the taxation on fossil fuels.
From the ‘10,000-year perspective’, taxes on fossil fuels bring forward the point at which the extraction of fossil fuels is deemed too expensive to bother with, and will enable us to leave those fuels in the ground. This stands in stark contrast to policies which would encourage improvements in carbon-augmenting energy efficiency – by postponing emissions rather than stopping them outright, and by increasing the affordability of extraction, carbon-augmenting efficiency improvements lead us down the path of disaster, to an atmosphere of ever greater CO2 concentration with ever more serious consequences for the planet.