Emissions trading systems are being operated, constructed or expanded in Europe, China, India, Brazil, parts of the USA, South Korea, New Zealand, and other places. What should governments expect them to achieve?
‘Decarbonisation at least cost’, is what policymakers often say they expect carbon pricing to achieve. That they should express this belief is hardly surprising, since it is a claim that has often been put forward by respected economists and authoritative international organisations.
The first surprising thing about this claim is how confidently it is put forward despite there being no evidence to support it. The academic debate on carbon pricing is complex and nuanced (as in any field) and there is evidence that it can be effective when used in the right way, in the right context, in combination with other policies. But there is no evidence at all to support the claim that it is the most efficient or cost-effective way to achieve decarbonisation (and in fact, considerable evidence to the contrary). This claim is entirely theoretical.
The second surprising thing is how wrong this claim is. Not just a bit wrong, but a lot wrong.
The idea is that carbon pricing allows businesses to find the least cost way to reduce emissions at any moment in time. The mistake is to assume that this equates to reducing emissions at least cost over the course of time. These things are not the same because the economy is path dependent: actions taken at one point in time determine the options available in future. There is no reason to believe that the easiest next step leads along the easiest path to your destination.
Early in the transition, in any sector, zero emission technologies and solutions are likely to be high cost. This was certainly the case with solar power, wind power, and electric vehicles before their widespread deployment, and is generally still the case across the energy intensive industries. Under those conditions, there are often lower cost ways to reduce a tonne of emissions: increasing the efficiency of fossil fuelled processes and systems, switching from one fossil fuel to another, or blending fossil fuels with biofuels. These are the measures that carbon pricing tends to encourage.
The problem is that while these are the easiest steps, they do not lead to the destination at all. The only way to reach the end-goal of the transition – the elimination of emissions – is through the development, deployment and diffusion of zero emission solutions. When governments rely on carbon pricing early in a transition, and businesses respond with incremental improvements to fossil fuelled systems, it only delays that process. During the delay, more money will be spent on burning fossils and invested in fossil-burning assets. This is like making improvements to a building that is about to be demolished.
With this understanding, it becomes clear that when carbon pricing is used in the wrong way, at the wrong time, with the wrong expectations, it can waste both time and money.
The two main forms of carbon pricing, a fixed tax and a cap-and-trade system, are not equal in this respect. An emissions trading system creates a balancing feedback: if you reduce your emissions, your demand for emissions permits decreases; that causes the price of emissions permits to fall, and the incentive for me to reduce my emissions is less. That makes cap-and-trade even less efficient than a tax (all else being equal), and likely to lead to even more wasted investment. Colleagues and I explored these dynamics with a theoretical argument, systems mapping study, and conceptual agent-based model in a paper published yesterday. Of course, how a policy works in practice depends on many factors – including its design, its context, and the presence of other policies. We do not wish to over-generalise. But understanding the core dynamics of a policy – knowing whether it is inherently self-amplifying or self-limiting – is important.
Does cap-and-trade get you decarbonisation at maximum cost? Well, not in an absolute sense. It would cost more to build a new power plant every Tuesday and blow it up every Wednesday. (Trump’s policies seem to be getting close to this). But by incentivising the least cost marginal emissions reductions at each moment in time, it risks delaying the necessary investment in zero emission solutions for as long as possible, allowing a maximum of wasted investment – within the boundaries of attempting to achieve a given transition in a given period of time.
None of this means that carbon pricing, even in the form of a cap-and-trade system, cannot be useful when done in the right way, together with other policies. So, what do you do if you have already institutionally committed to a cap-and-trade system and want to make it useful?
First, implement some other policies. Targeted subsidies, public procurement, and clean technology mandates (for either production or use) are likely to be needed to introduce zero emission solutions to the market. Later, infrastructure investment, market reforms, and many other policies will become important.
Second, create a carbon price floor. There are many ways of doing this: a complementary tax, a permit auction reserve price, or a ‘market stability reserve’, among others. Set the level of this so as to make the most carbon intensive technology more expensive than its nearest alternative.
Third, put a hard cap on emissions and give it a trajectory that goes all the way to zero. This can create helpful clarity over the direction, pace, and destination of the transition.
Finally, adjust your expectations. The cap-and-trade system will not give you decarbonisation at least cost. It is highly unlikely to introduce new zero emission solutions to the market. But it can contribute to reducing emissions along the way. It can increase incentives for material efficiency and circularity. And when other policies have established the clean technologies and spread them through markets, it can gradually remove the fossil fuel assets from the system, kicking out the dirtiest ones first.
A cap-and-trade system is not an ice-breaker ship that carves out a path at the forefront of the transition. But it can be a road-sweeper, following along at the back of the transition and clearing up the mess.