OPINION: Carbon credits as credence goods: Why does it matter?
Quantum Commodity Intelligence - Donna Lee is co-founder of carbon credit ratings agency Calyx Global
I recently read a manuscript1 about why the voluntary carbon market (VCM) has failed to deliver on quality. One of the issues raised was that carbon credits are 'credence goods'.
Because of this, there is a higher opportunity for conflicts of interest to create problems in such a market. This article is my own spin on the issue — and the first part of a blog series I am writing on this topic.
What the heck is a 'credence good'?
Economists, as well as the US Securities and Exchange Commission, define three different types of goods:
Search goods: those with attributes that can be evaluated prior to purchase or consumption. For example, you can go to a store and sit in a chair before buying it.
Experience goods: those that can be accurately evaluated only after the product has been purchased and experienced. An example is a haircut — you may not know how it will turn out, but you'll know afterwards if you're happy with the service.
Credence goods: those that are difficult or impossible to evaluate even after consumption has occurred. Carbon credits are a great example of credence goods.
There are several reasons why carbon credits fall squarely into the category of credence goods.
First, most consumers lack the expertise needed to understand if the emission reductions (or removals) were calculated well.
Was the oxidation factor used for the landfill gas project accurate? Was the biomass change over time measured correctly? Is it fine that the project did not use a dynamic grid emission factor? Most buyers of credits do not have such technical knowledge.
Secondly, there is a large information asymmetry between the seller of the credits (e.g. the project developer) and the buyer of the credits.
Another example of credence goods are services provided by car repair companies. Most people (like myself, who are not experts in car mechanics) are at the mercy of the repair shop. They may suggest that certain items need to be fixed, but really, I have no idea if that is true or not. I must just trust them. I drive away and it is still not clear to me if everything they said happened really did, and/or if such repairs were actually needed.
Finally, carbon credits are a bit strange in that one may never know whether the claims are true because all carbon credits (including removals) involve a counterfactual baseline.
In some cases, one can intuit that the counterfactual has a very high likelihood of being true… but in other cases, it is very difficult to know and most carbon methodologies do not require considering the uncertainty of the assumptions or models that go into the baseline scenario.
Why does this matter?
This matters because credence goods are particularly prone to bias and conflicts of interest - or even worse, fraud. For those (like me) who have been watching carbon markets for decades, we have seen all of the above occur. Therefore, for carbon markets to operate well, there need to be safeguards in place. Here I propose a few that are critically important:
#1 - Preventing fraud: Fraud can occur more easily with credence goods. Think back to the auto repair example. I may never know if fraud occurred with regard to the service rendered because I know nothing about automobile mechanics.
How can fraud be mitigated in the VCM? There are several ways to do this.
One is through the practices of credit issuers, who should have in place requirements that create checks on project developers.
Many of the more established carbon crediting programs have such rules in place, but buyers should beware with newer entrants who have not developed such practices. Another is through regulations — such as those under the Commodity Futures Trading Commission — that impose fines or penalties on those who commit fraud.
#2 - Reducing conflicts of interest: The impact from conflicts of interest will also be stronger with credence goods. Why is that the case? Because the asymmetries of information are greater, and the feedback loops to call out issues are weaker or non-existent. Battocletti et al provide a useful study on the conflicts of interest within the VCM, including those of standard-setters and validation/verification bodies — both of whom have incentives to be more generous to project developers (who are their customers).
This is a well known issue in the VCM. I wrote about it earlier in another blog, i.e. Why the VCM is facing headwinds. My hypothesis is that there have been too many conflicts of interest in the market to date. The VCM needs greater checks and balances (more on this in the next section) in order for the market to build trust.
#3 - Tackling bias: Bias can easily creep into credence goods because of the information asymmetry that sellers have vis-a-vis buyers. Sellers can "over-deliver" a service, or over-charge for a product, because they know more than consumers. And furthermore, consumers may never return the product or complain about the service because they never know that this occurred.
One of the best examples of bias creeping into the VCM is REDD baselines. By now, it's quite clear that REDD baselines using older methodologies were significantly overestimated. Many academics have written about this, and the analysis at Calyx Global also demonstrates that bias occurred in REDD baseline setting. Project developers were acting rationally, however — stretching the flexibilities in the methodologies to maximise credit generation.
One of the key improvements in Verra's new REDD methodology (VM0048) is that the baselines are no longer set by project developers, but by an independent third party.
Checks and balances in the VCM
In order for the VCM to build trust — and crawl out of the crisis of confidence we are facing — there needs to be checks and balances. In other words, the market will be stronger if, built into the ecosystem, are a range of organisations with different motivations and financial interests.
To date, as I explained in my previous blog, most entities were incentivised to "make the market" — in other words, by increasing the volume of credits in the market. However, we now see some organisations emerging with different incentives, which is positive for the maturity of the market.
For example, insurance providers are not incentivised to increase marketable volumes, but are incentivised to simply understand risks well - whether delivery risk, non-permanence risk, or whatever risks they are insuring against.
Ratings agencies are another new entry into the ecosystem. Ideally, such organisations position themselves to be incentivised to 'get it right' — i.e. they help to provide the right check, and balance, in the market. But what is the best way they can do this? I will cover this in Part III of this series
In the meantime, it's useful to think about the implications of carbon credits as 'credence goods' and how we can evolve the market responsibly around such goods. While the consumer or buyer of credits may never know the true impact… that is not true of our planet, who bears the brunt if we don't get this right.
Note: [1] Vittoria Battocletti, Luca Enriques and Alessandro Romano. The Voluntary Carbon Market: Market Failures and Policy Implications. European Corporate Governance Institute (EGCI) Working Paper Series in Law N° 688/2023, July 2023.
This article was first published on Donna Lee's blog page, which can be found here. Parts two and three of the this series will appear at this site.