Supply side effects of direct action

This is a guest post by Evan Calford, a PhD Student at UBC

This post seeks to describe one particular concern with the implementation of the Australian Government’s Direct Action policy, as laid out in the recent Green Paper. There are many concerns with the Green Paper that have already been pointed out which, in the interests of space, I shall ignore. I wish to focus solely on the potential change in producer behaviour induced by Direct Action. I shall call this change in behaviour the supply-side response to the policy.

Direct action works by providing incentives to producers to reduce their carbon dioxide equivalent (CO2-e) emissions; the Government will, through a reverse-auction tender process, pay companies for reducing their emissions beneath their historical emissions levels. This causes firms to face an effective marginal cost for emissions, when they are operating beneath their baseline emissions levels (the direct action plan will not include penalties for firms that exceed their historical emissions levels).

Some media coverage has been worried that the lack of upside penalties means that firms won’t make any emissions reductions. This seems backwards, as it implies that firms will leave cash on the table. The effect of using a reverse-auction tender process will be to pay firms more than the actual cost of their emissions reductions. If the tender process uses a second price approach (so that all firms will be paid at the rate of the most expensive reductions contracted for at that auction) then this conclusion is straightforward. Even in a first price auction (each firm is paid what they bid) the conclusion is still valid – in equilibrium, firms will bump up their asking prices above actual emissions reduction costs.

This implies that firms (and industries) who make emissions reductions are being subsidized relative to the rest of the economy. It is unclear, given the contents of the green paper, what the relationship between quantity of production and the size of subsidy will be. The exact accounting mechanism used for changes in production and new entrants will be critical in determining this relationship. This point should be obvious to anyone who has thought about how to design carbon pollution abatement schemes, and seems to be understood in the green paper. A more subtle point is that, in a dynamic world, even if changes in production and new entrants are accounted for in a manner which does not distort incentives the mere existence of a direct action plan, and associated subsidies, will increase production levels in high emissions intensity industries.

To understand why this is true, consider a simplified scenario in which a firm that reduces emissions is paid a for the costs of these reductions, plus an annual lump sum payment (to a first approximation, this is what a well-designed direct action scheme will achieve). Under a CPRS or carbon tax, some high emissions intensity firms would shut down due to the higher costs. These shutdowns may not occur under the direct action scheme because the annual lump sum payment will decrease the firm’s average variable costs and affect the shut down decision. The direct action plan will see less reallocation of resources from high emissions intensity sectors to low emissions intensity sectors.

Furthermore, the annual lump sum payment can influence future investment decisions by changing relative rates of returns between industries. In a competitive environment we would expect, in the long run, rates of returns to be equalized across industries; externally induced changes in rates of return can lead to changes in resource allocations between industries.

It is natural to ask “how much effect will the supply response have on emissions levels?” This is a very difficult question to answer. A paper published by ABARE (now ABARES) in 2010 tried to estimate the supply response effect for the agriculture industry [disclosure – I am the lead author of the ABARE paper]. The paper made use of a computable general equilibrium model, developed by ABARE and Treasury, called the Global Trade and Environment Model (GTEM) which embeds many assumptions (see the paper for details).

supplyeffect

The key graph in the paper is figure d on page 24. This graph demonstrates that, by 2030, applying an emissions offsets scheme to agriculture (the emissions offsets scheme modelled is roughly comparable to the Government’s Direct Action policy) would have reduced Australian agricultural emissions by almost 8 million tonnes of CO2-e, relative to 2030 emissions without any specific emissions reduction policies for agriculture. However, this 8 Mt conceals the fact that agriculture would have made efficiency improvements equivalent to a reduction of over 10 Mt of CO2-e. The difference between these two numbers (2.6 Mt of CO2-e, or roughly 25% of the efficiency improvements) is the effect of the supply response caused by subsidies to the agricultural industry. The subsidy causes overall agricultural production to increase and causes a substitution away from low emissions intensity cropping to high emissions intensity beef cattle and dairy production.

To put it another way, you need to pay for 10 million tonnes of `direct action’ to get 8 million tonnes of emissions reductions.

Note that these figures are specific to agriculture, and depend on a specific set of modelling assumptions. Even when viewed in context, these numbers are best thought of as an educated best guess rather than a pinpoint prediction. The supply response effect of the Direct Action plan could be bigger, or smaller, than the 25% estimated by ABARE for the agricultural sectors. I don’t know the answer to the question “how much effect will the supply response have on emissions levels?” But I do know that the Government (in particular, ABARES and Treasury) have the expertise to try and answer this question, and that knowing the answer is crucial for designing an effective Direct Action policy.