in Uncategorized

Larry Lohmann: “The problem is not ‘bad baselines’ but the concept of counterfactual baselines itself”

In June 2016, I wrote a post based on a paper published in the International Forestry Review titled, “The ‘virtual economy’ of REDD+ projects: does private certification of REDD+ projects ensure their environmental integrity?”.

The authors of the paper, Coline Seyller, Sébastien Desbureaux, Symphorien Ongolo, Alain Karsenty, Gabriela Simonet, Jean-François Faure, and Laura Brimont, looked at the baseline scenario in two REDD projects: Mai Ndombe in the Democratic Republic of Congo and the Corridor Ankeniheny-Zahamena project in Madagascar.

In both cases, the authors found that the baselines were unreliable in determining what might happen in the absence of the REDD project. They wrote that, “the baseline scenarios in REDD+ projects amount to untestable guesses”.

I concluded my post about the paper with the following two paragraphs:

The authors of the paper are careful to talk about project developers “optimizing the parameters” or using a “convenient baseline scenario”.

Fraud would be a better way of describing what REDD project developers are doing when they set bogus baselines. The voluntary certification systems, such as VCS, are complicit in this fraud.

Larry Lohmann, writer and activist with The Corner House, left a comment in response:

Larry LohmannThanks, Chris, for the very useful piece and pointer to this important article.

I’m wondering, however, whether the charge of “fraud” at the end of your piece doesn’t actually let REDD proponents off the hook.

If baselines are indeed unverifiable, then there can be no distinction between “fraud” and “nonfraud”, and it doesn’t make sense to say that any particular baseline, or baselines in general, are fraudulent.

To say that REDD accounting is fraudulent implies that with some reform, and some honest work on the part of VCS and others, it might be made nonfraudulent. This is simply not the case with offsets. The situation is unfortunately far worse than that.

Similarly for related words like “bogus” and “nonadditional”. If baselines are unverifiable, there can be no effective criteria for distinguishing between “bogus” and “nonbogus” baselines. And to say that a project is “nonadditional” implies that with some effort, it might be made “additional”, but again this is not the case. If there can be no criteria for distinguishing between “additional” and “nonadditional” projects, we might as well just come out and say frankly that there is no such thing as additionality and we ought to jettison the whole concept. After more than a decade, I am still waiting for most academic critics of REDD to take what seems to me to be this obvious step.

You might dig in your heels and say that REDD project developers ALWAYS set bogus baselines because there isn’t any other kind of baseline, and that therefore ALL REDD projects are fraudulent. But to my mind that would be to drain the concept of fraudulence of its content. If everything is fraudulent, then nothing is fraudulent. To attribute the contradictions in REDD accounting to “fraudulence” (or “non-additionality”) is to underestimate wildly the absurdity of the enterprise.

I wrote to Larry, who I’ve known for many years through our work with the World Rainforest Movement, asking him to clarify this sentence: “If everything is fraudulent, then nothing is fraudulent.” It seemed to me that if baselines cannot be verified, then REDD is a fraud. That does not imply that VCS can work harder to verify baselines, because baselines cannot be verified. It just means that REDD is a fraud.

Here’s his response:

There’s a distinction between saying that REDD is a fraud and saying that any particular baseline is a fraud. I agree with you that REDD is a fraud, or totally bogus across the board. But I don’t agree with the claim made by vast numbers of academic and policy authors that the fraudulence or bogusness of particular REDD projects is due to their non-additionality, i.e. to fraudulent or badly-calcuated baselines. This is not only wrong, but tends to encourage the further proliferation of REDD.

One way of saying this is to say that REDD is bogus, or a fraud, because it is premised on the existence of something that does not exist, namely additionality. Since no intelligible distinction could ever be made out between additionality and non-additionality, that means there is no such thing as non-additionality, either.

So when academic and policy authors say that this or that REDD project is bogus or a fraud because it is non-additional, they are talking nonsense. No REDD project could ever be either additional OR non-additional. To put it yet another way, the problem is not “bad baselines” but the concept of counterfactual baselines itself. That reality does more than invalidate any particular REDD project. It invalidates REDD (and all other offsets) as a whole.

Whatever their intentions might be, such authors are in practice encouraging the further development of REDD because they are implicitly endorsing the basic premise of REDD (and other offsets) that there is such a thing as additionality/nonadditionality. They are, in other words, implicitly denying what you and I agree on, that REDD as a whole is bogus or a fraud.

The Seyller-Karsenty article you did the post about is a good example. The authors do a good job of showing that, intentionally or unintentionally, you can set pretty much whatever baseline you like — that there are a lot of pressures to set profitable baselines but no standards to constrain you that would enable you to distinguish between additionality and nonadditionality. They suggest not only that VCS has failed to “remove doubts” about the two projects’ “additionality”, but that VCS will never be able to do so. The authors even admit that “there is a kind of irreducible uncertainty regarding what the ‘right reference scenario’ should be” (i.e., no one can give a criterion for distinguishing additionality from non-additionality, i.e., that “additionality/nonadditionality” is not a useful concept for distinguishing fraud from non-fraud).

And yet they fail to follow through. They claim that setting the “right” baseline is after all nothing more than a “challenge” that someone someday will be able to meet (p. 234). They insist that there is such a thing as “non-additionality” (“… in-depth analysis of large-scale renewable and non-renewable CDM energy projects in India and China revealed a lack of additionality for these projects (Michaelowa and Purohit 2007, Shishlov and Bellassen 2012)” (p. 234)). They say that baselines can be “inflated” (p. 243) — as if some sharp distinction could be made out between “inflated” and “noninflated” baselines. For them, additionality is merely an “issue” (p. 244) to be resolved at some unspecified time in the future. And so they end up maintaining their position as part of the colonialist offset establishment, like so many other academics and NGOs.

I’ve never been able to figure out whether this phenomenon is due to native timidity (“Oh dear! If I tell the truth, my technocrat friends will stop talking to me, my articles will not be published in the journals, and I will lose my job and respectability. And anyway, doesn’t the UN tell us that there is such a thing as additionality? I guess I’d better go on lying.”) or to intellectual confusion, or something else entirely. But whatever the reason, it’s been going on way too long.


Leave a Reply

  1. Absurdity upon absurdity. REDD is a PES, but nobody is willing to pay for the opportunity costs of not deforesting. So it is asking the people who forego the (possibly short-term) benefits of deforestation to subsidise the climate change mitigation effects that the rest of us are asking for. In the absence of a genuine political and financial commitment to REDD, researchers should stop studying it. They should instead turn back to the public and the policy makers and say that it is not moral to study REDD and contribute to the illusion that it is serious when nobody is willing to pay for it.

    Of course, even if REDD became realistic – ie if there was enough money to properly compensate people for not deforesting. And if one could magically create trustworthy institutions and monitoring mechanisms to move that money from the people who want climate change mitigation to the people who live off deforestation. It would still be the shortest of short-term measures. Strangling the supply of forest products would just drive up the price until the money for REDD was insufficient again.

    REDD is surely just a distraction from the challenge of decarbonising the world economy. As yet there is still not genuine political support for that. Scrap REDD! Make a bigger, better, more ambitious and more expensive plan for climate change mitigation and go out and get broad-based political and economic support for it.

  2. Dear Chris, dear Larry,

    Thank you for sharing your interesting discussion on the website and to allow me to step in this chat. I am globally in agreement with Larry’s perspective – and I admit that some of the language used in our article may seem ambiguous. But I would not say that additionality/non-additionality concept is simply to dismiss. Let’s recall we are talking about scenarios, and given the fact that one reference scenario should be designed to measure the “performance”, it is not a scenario but a prediction. This is why, as we agree, the scenario will unverifiable (even ex-post, as it is a without the project scenario and the project will take place). Therefore, criteria for assessing the additionality would require being able to design the “right” scenario, which is impossible (to verify, then to set). However, if demonstrate additionality proves impossible for a scenario, one can suspect the likely non-additionality of some chosen scenarios. If, in the range of potential futures, the reference scenario selected is extremely different from the historical trend (Madagascar case) or if the “reference area” cannot compare with the project area (DRC case), a suspicion of deliberate choice of a likely non-additional scenario might be a reasonable assumption. Even though a given scenario cannot be coined as “impossible”, only more or less unlikely.

    But the criteria of additionality and of counterfactual are essential to keep for the sake of evaluation. Ex-post evaluation of projects/policies aimed essentially at measuring the net effect (with and without the policy), that is to assess the degree of additionality. Methodologies are well known (inspirited by drug trials) even though seldom possible to apply in our fields of interest. It might be interesting for policy design to know that more than 90% of the recipients PES in Costa Rica would not have deforested even without payment, and econometric studies seem to converge on similar results for this country.

    Business-as-usual scenarios are also useful within democratic debates about the future we want to build (or to avoid) collectively (for outlook/prospective exercises). The trouble with “REDD+ baseline reference scenarios” is that they are used for generating offsets. Remunerating countries with money rather than carbon credits can be inefficient, as the same issue of “reference level” will arise, but it will be less disastrous than allowing the circulation of “fake climate money”, i.e. REDD+ hot air.
    I am often questioned (by my “technocrat friends”) on the alternative I propose instead of REDD+, notably on the “performance-based” reward principle. I generally put forward two points:

    1. If donors cannot live without the performance-based principle (as it seems to be the case currently), then “performance” should be understood in a broad sense, with a mix of indicators based on the effective and sustained implementation of forest-related policies (all national policies that impact forests) measures and some elements of performance that can be considered as correct “proxies” for reduced emissions and on which governments can act (like fragmentation or the area of intact/natural forests).
    2. Before being able to deliver on performance, there is a need for investment to design and implement the concrete instruments that are needed to curb deforestation (e.g. land tenure reforms, etc.). The first requirement for this is a credible and functioning state that can deliver on the implementation of policies and enforcement of law. It entails strengthening institutions and civil society that are needed to shape more democratically expressed collective choices.

    Best wishes,