The Post-Cautionary Principle

(Reprinted from July 13, 2009.)

You-all regular reader types know, by now, that I really really don’t like lazy articles that drum up an impression of a false symmetry where there is none to be had in reality. But I’ve stumbled across a very striking and real symmetry, and one that I think is at the core of our lazy polarization.

In my talk last week, I attempted to take the edge of the partisan global warming debate by emphasizing that neither “yes” nor “no” are meaningful answers to the question “how much CO2 should we tolerate?” So many people, though, aren’t thinking in those terms, but rather in absolutes.

On the one hand, we have the “precautionary principle”:

No action of any kind should be tolerated unless it can be proven beyond doubt that it has no net negative impact on the earth as a whole system.

On the other, we have a libertarian philosophy which can be summarized as almost the diametrical opposite:

No restrictions of any kind on individual behavior should be tolerated unless it can be proven beyond doubt that they infringe on the rights of others.

It doesn’t really have a name that I know of, but it certainly has some proud adherents.

Now, see, most people think science is about “proof”, which shows what a terrible job we are doing of conveying the stuff of science to the world, but that’s another topic. The point here is that in both cases the person making the claim is asking for a sort of proof that is totally impractical even in the best of circumstances. To take matters further, of course, people who choose not to be convinced by evidence of something they find intuitively distasteful just won’t be convinced. Your evidence, no matter how well corralled and paraded before them, will surely be insufficient, and they will have a vast array of well-formatted pseudoscience, replete with charts and graphs for every legitimate piece of work you show them.

Yes, as I look at it the symmetry to me seems almost perfect.

What’s more, both principles, that of maximum self-determination, and that of maximum respect for future generations, have a core dignity and appeal to them. It’s a hard heart that isn’t tempted to sign up for both sides!

Unfortunately, both principles, however noble in intent, are unworkable. The real world is one of guesswork, tradeoffs, and tightropes. Neither principle offers useful guidance. The implicit contention of the two is doing real, fundamental damage to our ability to compromise and reason about our circumstances.

Life is uncertain. We might just screw up the economy AND get no real benefit on the climate change front. That’s an easier and likelier outcome than the opposite one, really. As long as we are pulling in opposite directions, the odds of us coming through this mess in a semblance of dignity keep going down.

The times don’t call for small measures, but they don’t call for digging our heels in and sticking to absolute principles either.


  1. Wikipedia says:

    "People’s freedom to innovate technologically is highly valuable, even critical, to humanity. This implies several imperatives when restrictive measures are proposed: Assess risks and opportunities according to available science, not popular perception. Account for both the costs of the restrictions themselves, and those of opportunities foregone. Favor measures that are proportionate to the probability and magnitude of impacts, and that have a high expectation value. Protect people’s freedom to experiment, innovate, and progress."

    I think this is a more reasonable proposition than my admittedly straw man construction.

    I don't entirely agree with it, though it makes for a more philosophically challenging contrast with the more reasonable formulations of the precautionary principle.

  2. I've always taught that both positions use the precautionary principle and the only difference is whether we're precautionary in favor of protecting the environment from harmful interference or protecting the market from harmful interference. This is not an original insight but one I've cribbed from other scholars, such as Kip Viscusi and Cass Sunstein. Others have written that the only difference between the two is who bears the burden of proof, or what we take to be the null hypothesis.

    Coase's theorem is an attempt to put this equivalence on a rigorous mathematical basis.

    There's a nice new-ish book that attempts to revitalize the precautionary principle, and answer the symmetry objection: Alan Randall's "Risk and Precaution" (Cambridge, 2011). Randall offers spirited and sensible criticisms of the kind of equivalence arguments Viscusi, Sunstein, and others make, and is causing me to rethink the way I construe precautionary principles.

  3. In my previous comment, I wrote that Coase's theorem is an attempt to put this symmetry on a rigorous mathematical basis. That's not quite right. Coase wasn't thinking about precaution an uncertain harms, but about definite quantifiable harms.

    The application of his theorem to the symmetry between both of Michael's precautionary principles is a corollary to Coase's main thrust.

  4. In the fisheries my agency helps to manage, we (biologists) have sometimes been asked to provide a risk management framework and assessment. Interestingly, being biologists, we're not qualified to do economic risk management assessments. So we have provided a biological risk management assessment to managers (including fishing industry reps), who seemed to use it in a seat-of-the-pants manner -- there is no formal, quantitative, transparent model applied to incorporate their economic data into negotiations for fishing decisions. Yet somehow our process seems to work.

    Why does it work, and what relevant differences are there for contrast with climate? Perhaps it's the heirarchy for achievement of objectives -- achieving adequate spawning escapement comes first. Or perhaps it's the binding agreement on these rules among parties to the Treaty. Or maybe it's because the impacts are relatively short term (next generation of catchable fish) compared to climate disruption. Or it could be that costs/benefits of making mistakes are borne somewhat equally and in similar currencies by parties to the Treaty.

    Whatever the differences, our success (so far) seems to occur despite uncertainty remaining in both the biological and economic models. We acknowledge uncertainty, but our knowledge of it isn't (yet) applied in a rigorous manner. We have a Treaty framework consitent with negotiating around trade-offs. We still have a long way to go in optimizing our fisheries management, but we have an agreement that puts us on a path to doing that optimization. Uncertainty should not be allowed as a legitimate barrier to making progress in climate negotiations.

  5. The characterization of the precautionary principle in this post is a pure version that does not really correspond to real-world proposals.

    The proposed rule in the EU, for example, would only have applied the regulations to substances used in great enough quantity, effectively exempting R&D efforts and focusing the regulation compounds produced in commercial quantities.

    Another good practical example is BPA. Proving hormonal disruption is not easy but we can see the existence of intersex species in the wild. There is a real risk and there are solid indications from lab testing that BPA can cause these kinds of impacts. So it is a simple case of adding two and two. Maybe all canned food should not be manufactured with BPA. That's how I see the precautionary principle.

  6. I agree with Dean about the statement of theprecautionary principle that MT offers; it's purely impossible, so it's easy to dismiss. (I believe it's MT's own phrasing, I haven't found it in use elsewhere.)

    “If you have the choice between a hypothetical situation and a real one, choose the real one.” – Joan Baez (to Michael Krasny, KQED radio, Feb. 4, 2003)

    Looking at real statements may be helpful; they're not that simple:

    The Precautionary Principle; 2005

  7. I do recognize the original statements are pure to make the point -- that "both principles, however noble in intent, are unworkable." I don't know if the 'ibertarian one is codified anywhere outside of science fiction worlds, where it seems to work excellently well.

  8. In re-reading this after the fact, it strikes me that it is very much about the concept of "proof", even in more careful statements about the principles.

    In real life we mostly care about balance of evidence, not about proof. It's clear that when the law becomes involved it is necessary to draw sharp distinctions, but this legalistic idea of proof (which in fact is highly problematic) doesn't much affect our mundane decision making and it shouldn't much affect our collective decision making either.

    The more we can cast our decisions on a behavior continuum, the better.

  9. " when the law becomes involved it is necessary to draw sharp distinctions, but this legalistic idea of proof..."

    Not quite. May be different for common law, and simpler for criminal law, but when you're drafting legislation and regulation what you're usually doing is setting boundaries. There's a really nice description of the issue in Sir Ernest Gowers' "Plain Words" neatly delineating why and how public servants had (and still have) great difficulty in explaining laws to citizens.

    Basically, you can't rely on positive terms in legislation. Positive statements are reserved mainly for preambles, not for the enactments themselves. If you want to grant something to a certain class of people, the only way to ensure that the benefit flows to those people - and only those people - (not to others) is to describe the limits. And that is always done in the negative, by proscribing granting the benefit to those who don't meet the definition of the target group.

    Some limits are easy - age for instance. But most are much more difficult - say, people adversely affected by a flood or a bushfire in a particular place.

    There have been advances since Gowers' time. But some of them, like the drive to express legislation in "plain English" can be a bit counter-productive.

    As for 'proof'. The requirements change depending on whether we're talking criminal or civil or administrative proceedings and what may or may not count as suitable or definitive in that environment.

    By and large, balance or weight of evidence is our usual way of doing things in society at large. It is constrained a bit in civil law, defined for various administrative matters, and tightly controlled (and contested) in criminal proceedings. Legalism requires different things in different circumstances.

    My feeling about people who demand 'proof' of various things are relying on a fuzzily recalled year 9 approach to maths or science - aligned with a distaste for committing themselves to a statement of their own values or reasons for (dis)agreeing with a particular proposition.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.