Much scientific attention is directed at developing and improving the reliability of techniques to estimate the stock of soil carbon at a particular time. This is important and essential work, and more progress is needed. There will, however, always be uncertainty and inaccuracy in these measurements. Measurement error need not be fatal to including carbon sequestration in a cap-and-trade system. A trading system can operate as long as the measurement process is accepted as defining an authoritative measurement - it need not be accurate with certainty. The process might include not only a technical approach to measurement but also the ability to challenge a measurement and a process for resolving questions or challenges, and final certification. While error in measurement can be tolerated, it would be hard to create legal authority if errors were so large and random as to appear to lack any scientific foundation.
A more subtle problem, however, is a compromise of the effectiveness of the system if there is a bias in the measurement process. If on average the measurement process systematically underestimates carbon stored, the system will provide too little incentive to sequester carbon, whereas if it systematically overestimates carbon, the cap will be met legally, given that the measurement system is legally accepted, but the effect on the atmosphere will be less than expected. This can be remedied by further tightening the cap to meet the atmospheric target, but carbon sequestration will be overused compared to emissions reductions because pricing does not reflect the actual carbon sequestered.
An even subtler problem of bias arises when a measurement process has been constructed to be unbiased based on experimental measurement, but the incentives to participate lead to bias in the actual application. Consider the following situation in which a practice is extensively evaluated through experiment and an average sequestration level is attached to that practice. The average sequestration achieved by the practice becomes a part of the measurement method - the part of the model by which carbon inventory is estimated. Under real conditions, the vigour with which the practice is implemented may be subject to variation. If it costs more to vigorously implement the practice, actual landowners may have an incentive to minimally implement the practice. This, in turn, results in less carbon stored on average under real conditions than the average experimental result. Another way this may happen is if the cost of implementing the option is correlated with an environmental condition that also affects the amount of carbon stored. If the correlation is such that lower costs are associated with lower carbon uptake, again the average uptake under real conditions will be less than the experimental average, because the activity will be implemented at the low-cost sites but it may not be economic to implement at the higher-cost sites. Reliance on practice rather than actual measurements tends to increase the chance of these incentive effects to create bias in the estimates.
Finally, enforcement is a necessary element of a successful system and must be part of the design. One of the surprising aspects of a cap-and-trade system as described by Ellerman et al. (2000) is that enforcement has been much more successful. There is no direct reason in economics for this, but rather it appears that regulators are more willing to enforce a cap when they can point to allowances that can be purchased to meet it. Since all entities have opportunities to purchase in the same market, the claim of some entities facing special hardships that prevent them from complying is less compelling. Hardship is more compelling in systems where entities must comply with an individual limit, and experience shows that exclusions are often granted, and so the environmental target is rarely met. Consider the case of wildfire that resulted in carbon emissions. If the landowner were required to meet some level of seques tration, and keep the carbon sequestered for some minimum period of time, enforcement in the face of fire becomes problematic. The landowner may have little ability to actually comply. The enforcement agency can levy a fine, but this can appear unreasonable given that the landowner could not prevent the fire. This would likely give rise to hardship exclusions. In any case, the carbon is in the atmosphere, and levying a fine would not remove it. With cap-and-trade, where the landowner can purchase or sell allowances, a fire is a hardship but the landowner can still comply by purchasing allowances.
Again, homeowners and businesses that choose to locate in areas prone to disasters mostly face the economic consequences of these disasters, and therefore presumably try to limit their exposure to, and the effects of, these disasters. The same principles should be applied to carbon sequestration. To the extent emergency aid or disaster assistance would apply to carbon losses, care is required to structure the aid so as not to undermine the incentives to reduce the chance of losing the carbon. Completely exempting the landholder from any responsibility to cover these emissions with credits would certainly undermine these incentives. One alternative worth further consideration that could provide some relief would be that if the landowner demonstrated effort to re-establish the forest and restore the carbon, he or she could borrow against that planned future replacement of carbon to cover the catastrophic loss. Such borrowing automatically occurs within the inventory period, and so a long period such as 20 years automatically gives the landowner a chance to restore carbon catastrophi-cally released in, for example, year 3 into the inventory period. However, the fixed period still creates the possibility that catastrophe in year 19 or 20 would leave the landowner short. Additional borrowing provisions could ease this problem.
Was this article helpful?