Prior to the last national election in Australia, our system of carbon credits was strongly criticised in the media by researchers from the Australian National University (see PD368). The then-opposition Labor party promised to hold an independent inquiry into the scheme if they won the election. They duly won, and the inquiry is currently underway, chaired by former Chief Scientist, Ian Chubb. 

This seems to be the ideal vehicle for airing my criticisms of the soil carbon method within the scheme. My recent Pannell Discussions on this have covered both its limited attractiveness to farmers and its weaknesses as a policy. My submission to the inquiry focuses on the latter. (I’ve given several presentations to groups of farmers on the former.) Here is what I submitted.

This submission addresses the following parts of the terms of reference:

1. The integrity of ACCUs issued under the Carbon Credits (Carbon Farming Initiative) Act 2011, with specific reference to:

b. Whether the methods by which ACCUs are generated meet the offsets integrity standards

The submission focuses on one component of the ERF: the 2021 method dealing with soil carbon sequestration.

Although the number of projects registered under this method is not large, it is important for a couple of reasons.

Firstly, the previous government explicitly stated that soil carbon sequestration would provide a sizeable proportion of the mitigation necessary to achieve its overall targets. If the expectations about carbon sequestration using this method are not realistic, then it will be more difficult to achieve the national targets.

Secondly, the problems with the soil carbon method are relevant to other ERF methods as well. Understanding the nature and severity of the problems is important not just for soil carbon.

Under the 2021 soil carbon method, farmers can choose to implement one of 13 types of farm management changes. Soil carbon is measured prior to adopting the new management changes, and then re-measured over time to determine the amount of carbon sequestered. Based on the amount of measured sequestration, ACCUs are issued to farmers.

Various adjustments are made to the measured amount of carbon sequestration when determining how many ACCUs to issue. Adjustments include: a deduction of 5% for “risk of reversal” (release of sequestered carbon due to events such as drought), a deduction of 20% for impermanence (if farmers choose to commit to sequestering soil carbon for 25 years instead of 100 years), deductions for increases in other emission as a result of adopting the new practice (such as methane from extra livestock or N2O from soils), and a small deduction to allow for the potential for unrepresentative sampling when measuring soil carbon.

Problem 1: The deduction for impermanence is too small

Most of the adjustments built into the scheme are reasonable, but there is one that should be revisited and reconsidered: the deduction of 20% for impermanence.

When calculating the appropriate adjustment to make when converting a sequestration value for a 100-year commitment to what would be appropriate for a 25-year commitment, one of the key factors that affects the result is the discount rate. Economists use a discount rate to convert future economic values to an equivalent present value. Essentially, discounting allows for the opportunity cost of capital that is spent to generate a benefit: the return on capital raised from industry, and/or the cost of borrowing funds to spend on consumption. The higher the discount rate, the lower the deduction for a 25-year commitment period, because values in the more distant future count for less in the present.

At the time when the ERF was established, the real discount rate specified by the Australian Government for economic calculations was 7%. This implies an opportunity cost of capital that is 7% higher than the inflation rate. From my calculations, the 20% deduction for impermanence (for a 25-year commitment period) is broadly consistent with a real discount rate of 7%.

Since that time, Australia has experienced a period of very low interest rates, which suggest that a real discount rate lower than 7% would now be appropriate. Indeed, for some of the recent past, real interest rates for government have been negative (i.e., nominal interest rates have been less than the rate of inflation). Although interest rates are now rising, so are inflation rates, so real interest rates remain low.

Australian economists are now generally using real discount rates below 7%. For example, in recent analyses I have seen economists using 2% to 5%. This makes a large difference to the deduction that should be made for impermanence (Table 1).

Table 1. The calculated deduction for impermanence for a 25-year commitment period for a range of real discount rates.

In my judgement, the deduction for impermanence currently being applied should be at least double the 20% that is being used. Unfortunately, the 20% deduction is hard-wired into the ERF. There is no opportunity for it to be adjusted in response to changes in economic conditions. This means that farmers are receiving substantially more ACCUs for their soil carbon than is justified considering the impermanence of that sequestration. The same is true for other ERF sequestration methods that use this 20% deduction for 25 years’ commitment. When purchasers pay for an ACCU generated by soil carbon sequestration over 25 years, the ACCU is associated with less than one tonne of CO2-e. In other words, the true cost to society of securing sequestered soil carbon is higher than it appears to be based on the current carbon priced.

Recommendation 1. The quantitative level of the deduction to be made for impermanence should not be set at a fixed level in ERF. It needs to be possible to adjust the deduction in response to changes in the discount rate resulting from changes in economic conditions.

Recommendation 2. The deduction for impermanence should be increased above 20% to correspond to the real discount rate that is currently judged to be appropriate for economic analysis of public investments in Australia.

Problem 2: The newness criterion is a weak measure of additionality

A key factor influencing the integrity of ACCUs awarded in the ERF is the accuracy of efforts to detect additionality. Detecting additionality accurately is a great challenge. In principle, it requires a comparison of what the farmer does as a participant in the ERF with what he or she would have done in the absence of the ERF (the counterfactual). The two key challenges in doing this comparison are as follows.

(a)    The counterfactual is dynamic. Even if the farmer currently has not yet adopted a carbon-sequestering practice at the time when they register a project in the ERF, it may be that they would have done so at some subsequent time even in the absence of the ERF.

(b)   The counterfactual for an individual farmer cannot be observed once the farmer joins the ERF. It is impossible to see inside farmers’ heads to know their true intentions regarding adoption of a new practice in the counterfactual scenario. Indeed, even if they do not currently intend to adopt, that may change over time.

The method used to assess additionality in the ERF is the newness criterion. This assumes that any practice that has not been used in the 5-year baseline period (prior to registering a project) would not have been adopted subsequently if not for the ERF. In other words, it assumes that the counterfactual scenario is static, which may be true for some farmers for some practices but will be untrue in many cases.

The degree of inaccuracy introduced by using the newness method for assessing additionality will vary for the different ERF methods in which it is used. In my judgement, it is likely to be especially problematic for soil carbon. The reason is that, for most farmers, the financial value of ACCUs potentially available under the soil carbon method is small or very small relative to their other revenues and costs. This means that the availability of ACCUs will make a decisive difference to farmers’ adoption decisions for only a small minority of farmers.

At the same time, some of the key eligible actions under the 2021 soil carbon method are likely to provide private benefits for farmers, and in some cases those private benefits would be sufficient to eventually motivate adoption without the ERF. This will not be obvious at the time that farmers register a project in the ERF. A wealth of research on farmers’ adoption of new practices shows that it is often slow – it takes a decade or more for a new practice to spread through the population of farmers. Combining those observations, it means that out of the set of farmers who have not yet adopted the practice and who wish to register a soil carbon project in the ERF, those who are additional could easily be swamped by those who are non-additional. This would be true even if the number of non-additional participants is modest because the number of additional participants will always be small (given that financial value of ACCUs awarded will be small relative to farmers’ other revenues and costs).

To illustrate the problem, consider zero tillage – one of the 13 practices that can qualify as an eligible practice in the 2021 soil carbon method. If the ERF had existed in 1990 and used the newness criterion to assess additionality, around 90% of crop producers would have been eligible to be awarded ACCUs if they adopted zero tillage. However, almost none of that new adoption would have been additional, because the great majority of crop farmers were about to start rapidly adopting zero tillage due to the large private benefits it generates. The few additional adopters would have been swamped by the many non-additional adopters.

Some of the other actions that are potentially eligible for ACCUs in the 2021 soil carbon method are likely to be similar to this, to some degree. Listening to advocates for some soil-carbon-sequestering practices that are currently not widely adopted (e.g., regenerative agriculture), they often emphasise that the practices generate substantial private benefits for farmers. Currently, farmers who are not yet using these practices would qualify for ACCUs if they did adopt them, but a proportion of them would not be additional.

Even a modest proportion of non-additionality can make a large difference to the cost of sequestered carbon to the Australian community.

For example, Table 2 shows some results from modelling I have done on the purchase cost per tonne of soil carbon sequestered under a livestock/pasture system, assuming a real discount rate of 4% and correcting the deduction for impermanence to match 4% (i.e. the excessive costs purely reflects lack of additionality, not problems with impermanence).

Table 2. Cost of purchasing sequestration of soil carbon ($/tonne CO2-e) from adoption of a new carbon-sequestering practice in a livestock/pasture farming system.

Note that “75% additionality” refers to the proportion of the whole population who would not adopt a soil-carbon-sequestering practice in the counterfactual scenario. Only a minority of those farmers would be prompted to adopt the practice due to the ERF, whereas all of the 25% non-additional farmers would like to register a project if they are allowed to.

The results in Table 2 include the payments to both additional farmers and non-additional farmers for whom participation in the soil carbon method would be beneficial, but only count soil carbon for the additional farmers, since non-additional farmers would have adopted the practice without the scheme.

Preferably, the costs of soil carbon sequestration shown in Table 2 would be no higher than the price of carbon, but the results show that, even for the relatively favourable case of a practice that is additional for 75% of farmers in the population, the purchase cost per tonne of CO2-e generated by soil carbon is very much greater than the price of carbon. Unless an accurate method for detecting additionality can be devised, it seems that there would be many very much cheaper methods available to the Australian community for mitigating climate change. For example, there are plenty of options for reducing emissions for which the cost per tonne of CO2-e would be a small fraction of the costs indicated in Table 2, and therefore reduction of those emissions would be a preferable avenue for the Australian society to pursue. The costs in Table 2 are well above what I consider to be plausible estimates of the social cost of carbon, especially for the scenario with lower additionality.

Recommendation 3. Unless a reasonably accurate method for determining additionality can be devised, the purchase cost per tonne of additional soil carbon sequestered is likely to be so high that Australia would be better off cancelling the 2021 soil carbon method and ceasing to register any new projects under that method.

Recommendation 4. Other methods that rely on the newness criterion to assess additionality should be modelled to evaluate whether they are likely to provide mitigation at a reasonable cost. If not, they should be withdrawn as well.

Unfortunately, given the challenges of accurately determining additionality outlined earlier, I am pessimistic about the prospects of an accurate method being devised. Nevertheless, it would be worth trying.

Recommendation 5. New efforts should be undertaken to develop a more accurate and reliable method for assessing additionality to replace the newness criterion. The method needs to account for a dynamic counterfactual scenario.

I noted earlier that advocates for practices that sequester soil carbon often emphasise the private benefits to farmers from adopting those practices. This is, perhaps, not surprising, but I was surprised to read an ABC article based on an interview with a senior staff member in the Clean Energy Regulator that did the same thing. The article quoted Shayleen Thompson (Executive General Manager, Scheme Operations Division) as saying that farmers will be “getting a new revenue stream, improving the productivity of their operations, and they’ll be making their land more resilient to things like drought. … It really is a win-win-win opportunity.”

I was struck that this senior manager has apparently failed to recognise that the ERF should not be supporting win-win-win opportunities because they will not be additional. There is a risk in such rhetoric of feeding an expectation amongst some farmers that they should be rewarded for adopting carbon-sequestering practices even if they are not additional. It potentially flags that there is insufficient understanding of and attention to the additionality principle within the Clean Energy Regulator.

Recommendation 6. The Clean Energy Regulator should ensure that its public-facing staff understand the principle of additionality and the importance of only supporting practices whose adoption would be additional.

Further reading

Pannell, D.J. (2022). Serious weaknesses in the Emissions Reduction Fund, Pannell Discussions 368.

Pannell, D.J. (2022). Challenges in making soil-carbon sequestration a worthwhile policy. Pannell Discussions 371.

Pannell, D.J. (2022). Soil carbon webinar and written Q&A, Pannell Discussions 374.