Tuesday, June 17, 2014

Uncertainty and Risk in the Context of Environmental Policy and National Defense

I have been thinking of a course on environmental policy. Given that I had been teaching a new course on defense policy, perhaps it was not too surprising that I realized that lots of what we know relevant to the environment comes from defense analysis. Defense analysis needs some other ways of thinking to free it from its usual polarities. As I reread this draft, I realize I need to make clear how the defense examples apply to environmental policy.

The basic idea is to consider matters of uncertainty and risk, in the contexts of national defense and environmental policy. 

Risk and Uncertainty in Environment and National Defense

0.       Background
1.       The Unbounded Challenge.
2.       Bounding Uncertainty
3.       Risk and Probability
4.       Decisions
5.       Insurance
6.       Firefighting
7.       A Standing Army
8.       An Urbanized Bureaucratic World
9.       History and Intra-generational Transformations
10.    Values and Goods

0.Background: There will always be surprises, and risks you choose not to defend yourself against (cost, likelihood is low). Your security umbrella, even if an empire or an imperium, will allow you to get wet, and sometimes even drenched. So you will choose. Although you may present yourself as total, you are in fact selective.

                Perhaps the most crucial factor, rarely addressed, is the hope and spirit of the people and institutions. During WWII, Japan believed they were superior in these, sacrifice being the norm.

1.The Unbounded Challenge. I had been teaching defense policy and proposed a course in environmental policy, and realized that there was lots I wanted to teach to each group, and that each field could learn from the other field. In each case, the other/opponent gets a vote, where that other may be an enemy, Nature, or other people. And others are sufficiently inventive in what they do, that it is best to have a Red Team that spends its time on figuring out your vulnerabilities. You can gather data and intelligence, but the uncertainties and the vagaries of bureaucracies, yours and others’, that you have to make judgments and suppositions.  And you are living in societies where legitimacy of actions and governments are subject to question.  Sustainability never addresses this.

2.Bounding Uncertainty: Your opponents are adaptive and they will alter their actions to suit yours. They are likely to shave the rules and game the system, finding loopholes you had not imagined—for they can then get rich or accumulate power.  You are not in the realm of problem-solving, but of strategy and tactics. To improve your insight, you have Red Teams devoted to anticipating what you’ve not thought, and your assessment of your strengths and vulnerabilities is in terms of your opponents’ capacities and what they understand about you. The theme here is Constant Vigilance Without Paranoia. CVWP could lead to infinite regress—what would they do if we do…, but action is crucial.

                Also, you want to manage your risks, purchase insurance, but uncertainty may be more significant than risks you can assess. I suspect that what you want to do here is to improve your awareness, especially of the limits of your imaginations.

                Example: Global warming: The wide range of estimates does not mean that global warming is uncertain. It does mean that you need to vet the reliability of estimates, be able to have a sense of what is at risk, which risks you might well sacrifice, the costs of avoidance including societal reorganization. There are surprises we might imagine, and those beyond our ken (or imagined by someone but not someone taken as reliable). Constant vigilance involves developing bureaucratic means of filtering, highlighting, and comparing accounts.

                Intelligence sets the context, and is perhaps most unreliable about particular contingencies. Noise is prevalent, and in retrospect looks more informative than it was in actuality. Ideally, you have an uncertainty bearing organization, one that sends risky prospects to the risk-analysts. And that uncertainty is hierarchized, and given background.

3. Probability and Risk: You have information, you have predilections, and you estimate probabilities. Those are always accompanied by ranges or errors. Very small probabilities are unlikely to be reliable since external effects will overwhelm their causes. You can delineate a tree of decisions or events, assign probabilities, and likely you will miss some steps—ok. Again, you try to keep track of costs of avoidance and consequence. You might look at events that are several sigma away from the estimated probability, and if you don’t have an estimate of the error in your probability, you won’t know how reliable your claims are.  You might estimate Value at Risk, but beyond that you are beyond probabilities of the costs/consequences. Also, many probability distributions have heavy tails, so the reserves you need may be much larger than you first estimated.

                It may be rational to go for broke, if what you need is X and anything less is useless to you. Lots of small bets will get you sure defeat by the law of large numbers. One possibility is to find a more realistic path to low probability or go for broke situations, but this may not be possible. Still, maybe you can get reinsurance or joint bettors in a market with speculators.

4.Decisions: You often make decisions by inadvertence, by unintended consequences, by others’ moves and your response or not, etc. You only sometimes make a decision.

Estimates of expectation value, range, etc, have problems if discount rates have a big effect. Moreover, there are distortions of conventional decision-making theory, ala Kahneman-Tversky.

                Doom scenarios are apocalyptic and religious, and they divert attention from useful action, and encourage extreme responses. In general, the survivors do not envy the dead, so horrible events are horrible not apocalyptic. Moreover, doom does not encourage coping, nor thinking about surviving to fight another day—the latter being what’s crucial. Losses can be awful, but what’s crucial is whether you can go on and do better. Fear and scare are paralyzing.

                Infinite sequences of anticipated moves—if I do, what would they do, and then what would I do,… are too smart by a half. You may want to act given a finite or even short horizon, perhaps you opponent will still be thinking through that infinite game. (Of course, the infinite game may have a convergent answer, but not likely.) Your goal should be to degrade the opponent, deter their action, find out more of real consequence (by acting and finding out the answer to your if question). Hence you may not want to pay too much attention to the longer term, because a short-range action may change future contingencies.

What’s crucial is hope, to have a society that is not demoralized. Resilience is surely material and organizational, but it is as well ideational or moral. Recovery is often remarkable, although sometimes it takes too long and suffering becomes unbearable.

5. Insurance: Insurance allows you to prevent bad consequences (risk management), ameliorate their consequences, and compensate for losses. The problem is to find someone who will sell you insurance, or if you self-insure, where will you get reinsurance. Is pooling possible? Or perhaps you can find a Lloyds of London. Insurance cuts current spending for future security, and that may not be popular.

                Ideally, you could modularize your risks, erect fire breaks, and find smaller modules that would allow for more conventional insurance. You might be able to purchase a real option, allowing you to make decisions in the future, so you are not irreversibly committed and may respond to new information and contingencies. Scandalously, you might define acceptable losses and deductibles; the problem of explicitly acknowledging such is enormous. Likely you have such losses implicitly defined, in any case.


6. Firefighting: You might respond to risks and the uncertain by having a fire department, first responders who are sent out when called. Mobility and backup—N-alarm fires—and ways of dealing with coincidences that may demand weakening defenses in some places. What’s crucial here is to understand precursors so you are more aware of risk and can cut that risk, in effect fire-insurance maps (See above on Insurance, and here the problem become reserves and reinsurance, and an insurer of last resort—in effect the society.) You also want to have after-action reports, or lessons-learned, or “post-cursors.” But you might imagine how such a system would fail if fires were set all over, by conspiracy, by chance, or by a massive external source (fire-bombing).

7. A Standing Army: A standing army is a massive resource, located at all sites of potential conflict, with a high level of readiness. This is the complement of the fire department (real armies are a balance of fire department and a ready standing army).  This puts your resources in harm’s way, and your opponents can in effect determine your actions by choosing to attack you at a particular place and time. There is much to be said for selective engagement (vs. domino effect or an imperial force), choosing your battles. Do you then have to allow for sacrifice of some of your deployed troops, having decided not to go to war. The problem is how to diplomatically “choose your battles.” You might have “special forces,” forces that are not salient but go out and do the work in an impromptu way.

                Forward presence is expensive, may be a deterrent (or a temptation), but it does express intent and power and resources.

The World:

8. An Urbanized Bureaucratic World: If you are in an urbanized bureaucratic world, interconnection, density, concentration, and bureaucratic autonomy are the facts. Collateral damage is likely, and it may be hard to have what people would call a surgical strike (although surgery itself is not so “surgical.”) Moreover, it may be difficult to identify non-combatants and civilians, for they are too intertwined with the fabric and the problems.  There is much to be said for standard operating procedures in government and firms, for they provide for responsiveness in crisis and systematic organization. The instruments for dealing with problems—police, tax, regulation, law, markets, informal sectors—provides for many modes of intervention.

9. History and Intra-generational Transformations: If we examine US, European, and East Asian history, post WWII, we have a sense of intractable problems that sometimes went away, were altered, or persisted. Social conventions actually may change dramatically even if folk beliefs remain. What seemed to be settled, can then return anew with greater perversity (Russia, 2014?). Moreover, what we call ideologies—capitalism, communism—would seem to have national variations that turn out to make those ideologies rather constumed traditional modes (Russian vs. Chinese communism). Resources respond to technologies, and vice versa, and so technologies can alter politics. In any case, resources are denominated by price and quantity, and by alternative technologies—and one should note that invention may not come on when needed so that resource conflicts and economic disruption may be serious. For most societies, stability and legitimacy are crucial, and policies that uproot stability or legitimacy, under rubrics of such as efficiency or fairness, are likely to meet resistance. Put differently, rent-seeking is everyone’s game.

10.Values and Goods: Survival is rarely if ever the problem. Rather, what kind of society do we wish to have, what is earned and deserved, what is hereditary, what is schemed and gamed and legitimate, and what is not? We are not likely to defeat many enemies, since new ones would seem to arise. (Why?)  But we can prevail, keep things from getting too much out of control, and provide for stability and change. Institutions alter and transform, wither away, die and are born.

In all these observations, what is needed are examples and models. For to think about the environment or defense, one has to avoid abstractions that trap one in crises or dilemmas, and to have at hand enough examples where movement would appear to be possible. (Ideologies may encourage and empower people, especially if we do not let them get in the way of practical action.)

No comments: