Why trade-off analysis leaves me un-enthused

By Joern Fischer

Last night, I had an interesting chat over dinner about ecosystem services and trade-offs. For a long time, I’ve felt that somehow trade-off analysis just doesn’t really excite me. I think I’ve finally figured out why.

My reading of the ecosystem services literature is that it developed a bit like this. Once upon a time, people recognised that the idea of conserving biodiversity for its own sake did not appeal sufficiently to the general public. So the metaphor of ecosystem services was created as a way to emphasise to people that nature actually matters not only for its own sake, but also for human well-being.

Armed with a lot of enthusiasm about this new framing for conservation, people set out to identify the possibilities of this new metaphor. I’d say people at this early stage of ecosystem services research set out to find new synergies — for example, was it possible to conserve species simply by appealing to the conservation of ecosystem services?

Then, as the ecosystem services concept began to mature in theory and application, an increasing amount of economics came into it. Services were being quantified in biophysical, and increasingly also in monetary terms. Optimisation started becoming possible: where should we have which kind of service? And what will be the cost of having one kind of service to another kind of service? Thinking in decision-relevant trade-offs had arrived.

Right now, my sense is that thinking about trade-offs is increasingly fashionable in the same research community that once upon a time emphasised the need to find synergies. Selling win-win solutions — or framing papers around win-win solutions — has got more difficult. People will be skeptical and will say that’s just seeing the world through rose-coloured glasses and in fact there are important trade-offs that must be considered.

True, I guess, but to my mind also rather uninspiring.

Let me draw a parallel to forecasting versus backcasting. Forecasting is the dominant scientific mode; we generate numbers about the future based on our understanding of the present. And then we work out things will continue to get worse. (Hooray.) Backcasting — the alternative — starts by asking where do we want to be, and what do we need to do to get there? What do we need to do to break out of current trends, rather than assume they must continue into the future? I like backcasting because it lifts us out of the present set of perceived constraints, and opens our minds to possibilities and new, creative solutions.

To me, this is also the reason why I’m un-enthused about trade-off analysis. We can understand in ever more detail the various constraints imposed on current management decisions. But this psychological space of perceiving the world in constraints — to my mind — is likely to keep us locked into the same general structures and thought processes that quite possibly, we need to break through. Seeking win-win solutions — just like backcasting — may be less “realistic” right now, but as a mindset, it paves creative avenues to finding a better future.

I won’t deny that we should be understanding constraints and trade-offs more carefully in some situations, for example to avoid unintended side-effects on marginalised groups. But the overwhelming emphasis on trade-offs in the literature at present, instead of on synergies and win-wins strikes me as also, inadvertently, constraining our thinking to what is — rather than opening it to what can be.

10 thoughts on “Why trade-off analysis leaves me un-enthused

  1. Dear Joern,

    enthusiastic or not, but, in my humble opinion, your writing of history (I paraphrase…) “people invented ecosystem services in order to get more conservation” is only one strain of history. The other, quite independent strain is the one that says “climate only earnestly became an objective of policy once we could, a) tell people why it mattered to them, and b) that, in a lot of ways, it might matter more than other aspects of environmental degradation”. One can like that way of thinking or not, but at some point, if you want to argue about values, then you need to meet the person who might be affected by some loss, speak her language, and you have to try and compare one loss of something valuable with another.

    I see the problem elsewhere: the simplistic and shortsighted land cover based monetarisation tools we see pop up everywhere now have created a market for often inadequate tradeoff analyses. Inadequate, first, because a lot of non-marketable values usually don’t enter the tables, and second, because synergies are rarely accounted for properly.

    I therefore think that one needs to more, rather than less, conceptual and quantitative tradeoff analyses. But the benchmark must be that a comprehensive view on values (including, but not limited to, the famous “option values”) must be taken, and synergies must of course be part of the analysis.

    At that point, biodiversity and ecosystem services will finally meet climate and other crucial environmental agenda items. I am perhaps naively optimistic, but I believe that doing so will improve environmental policy.



    • Thanks Wolfgang! It’s always nice to get comments, and especially so if they are thoughtful such as yours above! From my perspective, what you’re saying is not necessarily opposed to my own thoughts: I am not at all against highlighting where some choices affect somebody negatively, and reality is, that humanity probably has to make some choices where some people will be worse off. (In order to make sure future generations can still be well off.) What I wanted to highlight is that there is another level to this. The more we focus on the tangible choices, with current constellations of actors and institutions, the more we risk getting locked into a “solution space” that is potentially narrower than it needs to be. In short: yes, in those situations where decisions need to be made, the trade-offs among different options should be understood. But at the same time, keeping an open mind towards the potential to broaden the solution space — e.g. by looking for win-wins or synergies — could be just as important. One way or the other, you got me thinking … perhaps hoping for more win-wins is just overly naive! I’d be interested in other opinions on this issue… Thanks again!

  2. I have no problem with trade-off analyses for their particular contributions. Perhaps the sentiment is there may currently be more trade-off centered effort than one is comfortable with? Within a finite space where you have finite resources their allocation seems to necessarily lead to prioritization and a zero-sum game.

    But I do like the more optimistic approach of testing and probing to discern relationships and mechanisms where synergy can be better understood and increased with minimal cost. I agree that if we lock into current understandings as ‘final answer’ and calculate all relationships going forward as trades within a fixed space then we miss looking for better appreciations of what is happening around us. We don’t have all the answers, indeed as things change (climate, population density, etc) the answers we think we have become suspect. Context matters. New synergies may arise or well known synergies may be lost as change occurs.

    For me then both trade-off analyses and continued search for win-win relationships have their place. Do we need a trade-off analysis to determine optimum levels research on trade-off research (… or is that TOO meta? 🙂

  3. One of my continued problems with trade-off analyses is similar to yours, Joern, but I don’t know that I see it as the result of a trade-off approach but rather a result of a certain lack of “backcasting” and an appeal to conservative assumptions that limits choices. (Conservative in both the political and epistemological senses, here.) To whit:

    It is “conservative” in both senses to presume pareto optimality is a goal and/or outcome restraint: someone is better off, while no-one is worse off. (This is NOT conservative in the biophysical sense, in that pareto optimality typically does allow, at least implicitly, decreases in/damages to natural systems, processes, biodiversity, etc.) But this assumption of the desirability of economic efficiency has little to do with, say, thermodynamic efficiency, and it very avowedly declares that it has nothing to do with justice, appropriate distribution, etc. Distribution is particularly interesting/important here. As Herman Daly has said “the Pareto
    principle has its own extreme—one person could have all of the
    surplus and everyone else live at subsistence (or die for that
    matter!), and there would still be no case for arguing that
    redistribution would increase net social benefit. Within limits,
    therefore it is reasonable to say that redistribution can give us
    economic growth in sense (2), but not in sense (1)—another reason
    why ecological economists pay more attention to distribution than
    do neoclassicals” (http://www.greeneconomics.net/2003HDaly.pdf). If we are not permitted to presume anyone “loses”, even those who are consuming much–to say nothing of often OVERconsuming much–then trade-off analyses in ecosystems services will always be unsatisfying, because at some point, there must be a limit to how much humans without sufficient rights and consumption can increase their *appropriate* consumption, while neither cutting into the consumption of high-consumers nor unsustainably increasing net conversion/consumption of resources.

    This, then, is the crux of the flaw for “win-win” to me: let’s say that, even under a strictly neoclassical analysis, there is a net benefit to some action (saving an ecosystem service, say), but this is a benefit to many in a community but a cost (or foregone benefit) to a small number of the better-off. Let’s say, however, there is also a positive benefit to conservation/biophysical sustainability. Under certain circumstances, a “net” win for society and a clear win for the environment IS a win-win. But under traditional economics, it is not if some of the people in the community lose. That they lose something that only marginally affects their high quality of life is immaterial (as Cosma Shalizi says, the principle of Pareto optimality shrugs and says “who are we to judge?”) It is a win-lose because SOME people lose.

    I could go on and on, and will some time, but we now come to this: I think trade-off analyses often miss several related, crucial questions that often look too “political” or “radical”:
    1) Is our analysis limited by what we view as “politically feasible” rather than biophsyically possible?
    2) Has our “political feasibility” assessment come from a deep, prolonged engagement and collaboration with social scientists, historians, behavioral economists, etc., producing a political possibility frontier based on reasoned scholarship, at least informally? Or
    3) Has our “political feasibility” assessment come from our informal judgment of what that means; reference to some piece of popular debate or political will (e.g. citing David Brooks or the Economist or some such to “show” that increase regulations or taxes are currently impossible); or from presuming political feasibility is the same thing as, say, pareto optimality?
    4) Have we presumed pareto optimality without examining its historical validity as a political (not just economic) rubric?
    5) **Have we considered that if our goals are (a) human well-being; and (b) environmental sustainability, that (c) there may not be *pareto optimal* solutions available, but that there may indeed be redistributive options available? And IF redistributive options allow increased well-being for ‘under-consumers’ but a marginal (if real) decrease in wealth for ‘over-consumers’, is that not a significant finding that we should not pre-judge as politically impossible, but rather, at the very least, present as a biophysically plausible and empirical result that should be reported and examined in detail?

    One you get to #5 of my adamant, ranty list of questions, then much REALLY interesting work can be invoked, as it is completely true that is politically infeasible to yank resources from one group and distribute them to another ad hoc. (Or at least, my informal assessment based on on-going engagement with political science and history tells me so 🙂 But there are actually many and varied mechanisms and pathways to redistribution, few of which are EVER studied in trade-off analyses, which usually precludes them right away. Society has sanctioned and developed countless redistributive schemes, from “land reform” to reparations to social security/social insurance to single-payer/universal health case, manumission and freeing of slaves, increased rights and income parity for women, and on and on. By making the POLITICAL presumption of pareto optimality, our “win-win” trade-off analyses fail to engage in backcasting and considering real ways that redistribution already happens, or could happen. By “conservatively” assuming that recent government inaction is intractable, rather than temporal (e.g. government and society views change, but usually on a decadal scale) we foreclose proper academic evaluation and many possible solutions.

    And finally: many say in response to this that, for example, climate change is too urgent to “wait” for social change of the kind I’m thinking about. (Though of course, that again is an interesting presumption because the rhetorical “wait” implies a lack of action, instead of organizing action towards speeding up social change–e.g., it has presumed, (often) without academic study that necessary societal change is too slow *no matter what actions we take*, and so “waiting” and “acting on” societal change are indistinguishable. Obviously, at least two other choices are (a) that social change is possible, and possible fast enough if we put ourselves to it (though this may mean putting ourselves to it in ways quite outside our bounds as scientists but in the bounds of our responsibilities as citizens!), and (b) the understudied possibility that social change *may be necessary*. That is, the urgency of a situation DOES NOT CHANGE its “objective” dynamics. Rather than assuming social change is too slow, why not study social change? Or, if you’re in a gloomy cast of mind, it is still an open question: is there a “politically feasible” road to the change we need that doesn’t require the social change I envision? To be scientific, we must consider that the answer to this MIGHT BE “no”.

    But given that social change ALWAYS happens (though at various time horizons), I much prefer to stake my interest and passion at these long-term, irregular but not totally unpredictable social changes. The thing is, no one has shown, to my knowledge, the efficacy of papers in social change is greater than creating and maintaining widespread, healthy community links. (we needn’t all be organizers, but perhaps we do all need to be good local, regional, national, or even global citizens–e.g. MEMBERS of grassroots organizations). What of the (slightly tongue in cheek) hypothesis that the changes we need are better served by more time with our fellow (less well-off) citizens and less with our fellow (academic, scientific) colleagues writing papers and going to conferences?

    • Hi Jahi,

      while I agree with much of what you wrote (when was it ever not thus?) I would note that there is noting remotely inherent in trade-off analysis (or even utilitarianism) that requires a Pareto efficiency framework. It entirely possible to have a utilitarianism which is equity maximizing.

      Moreover,if you have a fair initial allocation then Pareto efficiency is entirely unproblematic, which does at least suggest that perhaps we have to focus on justice before we start thinking about efficient allocation and trade-offs.

      • I agree, Dave. Indeed, in my rant, my initial point got lost: “I don’t know that I see it as the result of a trade-off approach but rather a result of a certain lack of “backcasting” and an appeal to conservative assumptions that limits choices.”

        And of course you’re right about fair initial allocation, but that is rarely the actual nor proposed starting condition, no? Hence, the importance conversations about values and “ends/means” that we often bang on about 🙂

  4. Or another Shalizian koan: “Restoring stolen goods to their rightful owners is not pareto-improving.”

    All the more striking to consider the implications of this for today if you accept that any number of past actions/current institutions may be based on something like stolen, ill-begotten, or questionably monopolized resources.

  5. Thanks Jahi — and wow! — congrats on the longest comment yet 🙂

    I find it very engaging, so keep the comments coming!

    I think this part of your response is particularly pertinent:

    >>> Rather than assuming social change is too slow, why not study social change? Or, if you’re in a gloomy cast of mind, it is still an open question: is there a “politically feasible” road to the change we need that doesn’t require the social change I envision? To be scientific, we must consider that the answer to this MIGHT BE “no”. <<<

    Cheers! — J.

  6. Very interesting blog post and even more so are the comments!

    Besides the already discussed ambiguities, I think trade-off analysis leaves many of us un-enthused because the way it is use in ecosystem services research exemplifies the inability of research fields to learn from each other. In fact, research on trade-offs has been around for several years in other (related) research fields (i.e. resource management, sustainability assessment, etc.). Although ecosystem services research on trade-offs mirrors many of the same challenges and concerns that were dealt with 10-15 years ago little has been learned.

    For example, trade-offs in sustainability assessment are well researched because they are addressed explicitly, alongside policies and development targets. Extensive research has resulted in well-defined procedures when and how to balance between two desirable but incompatible features. General rules have been formulated to avoid narrow solution space by delineating when trade-offs are acceptable and under what condition (see Gibson, 2006, p. 176, doi: 10.3152/147154606781765147). However, as Wolfgang pointed out, ecosystem services trade-off analyses are often based on weak foundations and are sloppily done. Instead of using trade-off decisions as tools that can effectively move us towards sustainability (addressing them explicitly alongside synergies), its application reinforces lock-in (detached analysis).

    I can think of two reasons why research on ecosystem services is particularly excited about trade-offs. Firstly, novelty is an important criterion in the editorial decision process of peer-review publishing. This serves as a strong incentive to present old wine in new bottles. Secondly, we have a limited capacity (interest, time, access) to screen other fields for potentially relevant knowledge (as authors and as reviewers). Thus, new research fields tend to reinvent the wheel (another example is transdisciplinary modes in climate change research).

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s