The inefficiency of New Zealand's emissions trading scheme

A couple of days ago I wrote a post about the game theory of climate change negotiations. One of the conclusions of that post was that there was a dominant strategy for countries not to reduce their greenhouse gas emissions. Another problem might be that countries reduce emissions, but not by as much as they should (in order to achieve the Paris Agreement goal of no more than two degrees of temperature increase over pre-industrial levels).

Potentially, even worse might be that countries find inefficient ways of meeting their emissions reduction goals, and I believe there is a strong case that New Zealand is in the inefficient camp. New Zealand introduced its emissions trading scheme (ETS) in 2008, and it was later amended in 2009 (and has been reviewed twice since). Under the scheme (described here), "certain sectors are required to acquire and surrender emission units to account for their direct greenhouse gas emissions or the emissions associated with their products".

As Megan Woods notes, one of the main problems with the ETS is that agriculture is not included in the scheme, and farmers have been told that there are no plans to change that in the near future. Agriculture is responsible for about half of New Zealand's greenhouse gas emissions (see page 4 of this fact sheet from NZAGRC).

This creates a problem because, in order to meet the overall goal of emissions reduction, other sectors must reduce emissions by more to compensate. To see why this is inefficient, consider the diagrams below. Say there are just two markets: (1) agriculture (on the left); and (2) all other sectors (on the right). Both markets produce a negative externality, represented by the difference between the supply curve (the marginal private cost or MPC curve, since it includes only the private costs that producers face) and the marginal social cost (MSC) curve (made up of MPC plus the marginal external cost (MEC), which is the cost of the externality to society). In both cases the market, left to its own devices, will produce at the quantity where supply is equal to demand - at Q0 in the agriculture market, and at Qa in the other market. Society prefers each market to operate where economic welfare is maximised. This occurs where MSB is equal to MSC - at Q1 in the agriculture market, and at Qb in the other market.

In the agriculture market, total economic welfare is equal to the area ABD-BFE, and there is a deadweight loss of BFE [*]. In the other market, total economic welfare is GHL-HMJ, and the deadweight loss is HMJ [**]. The value of the externality is represented by the area DFEC in the agriculture market, and by the area LMJK in the other market.

Now consider the implementation of two different emissions trading schemes, as shown in the diagrams below. In the first scheme, both markets are included. Firms must either reduce emissions directly, or buy credits to cover their emissions.  Either of these is costly, and forces the producers to internalise the externality. The markets both move to operating at the point where MSB is equal to MSC, maximising economic welfare at ABD in the agriculture market and GHL in the other market (there is no longer a deadweight loss in either market).

In the second scheme, agriculture is excluded but the same total emissions reduction is desired. This means that the other market must reduce emissions by more to compensate. The other market reduces quantity to Qc (note that the reduction of the value of the externality in this case is double what it was in the first scheme). Total economic welfare in this market reduces to GNSL, with a deadweight loss of NHS. This market over-corrects and produces too little relative to the welfare maximising quantity (Qb). Meanwhile, the agriculture market continues to produce a deadweight loss of BFE.

Notice that the size of the combined deadweight losses across the two markets is pretty much the same under the second scheme (BFE + NHS) than it was without any emissions trading scheme at all (BFE + HMJ). So compared with the first scheme, the second scheme leads to a loss of economic welfare - it is inefficient.

Emissions trading schemes create a property right - the right to pollute (if you have purchased ETS units, you are allowed to emit greenhouse gases). In order for property right to be efficient, they need to be universal, exclusive, transferable, and enforceable. In this case, universality means that all emissions need to be covered under the scheme, and all emitters must have enough rights to cover their emissions. Exclusivity means that only those who have rights can emit greenhouse gases, and that all the costs and benefits of obtaining those rights should accrue to them. Transferability means that the right to emit must be able to be transferred (sold, or leased) in a voluntary exchange. Enforceability means that emissions should be able to be enforced by the government, with high penalties for those who emit more than they are permitted to.

Clearly, the current New Zealand ETS fails under universality as agriculture is not included. And the previous analysis above shows why this leads to inefficiency (loss of total economic welfare). The government is simply passing the buck by avoiding the inclusion of agriculture in the ETS (e.g. see Paula Bennett here). If we want to efficiently reduce our greenhouse gas emissions, agriculture must be included in the scheme.

*****

[*] The total economic welfare in the agriculture market is made up of consumer surplus of AEP0, producer surplus of P0EC, and the subtraction of the value of the negative externality DFEC.

[**] The total economic welfare in the other market is made up of consumer surplus of GJPa, producer surplus of PaJK, and the subtraction of the value of the negative externality LMJK.