Tuesday, July 29, 2014

The immaculate conception theory of decision-making

An essay from a decade ago about foreign policy that provides great counsel and insight on decision-making in general, Foreign Policy Immaculately Conceived by Adam Garfinkle.
When a talented but untutored journalistic mind focuses on a foreign policy issue, particularly one that editors will pay to have written about, an amazing thing sometimes happens: All of a sudden, crystalline truth rises from the clear flame of an obvious logic that, for some unexplained reason, all of the experts and practitioners thinking and working on the problem for years never saw. This is the immaculate conception theory of U.S. foreign policy at work.

The immaculate conception theory of U.S. foreign policy operates from three central premises. The first is that foreign policy decisions always involve one and only one major interest or principle at a time. The second is that it is always possible to know the direct and peripheral impact of crisis-driven decisions several months or years into the future. The third is that U.S. foreign policy decisions are always taken with all principals in agreement and are implemented down the line as those principals intend — in short, they are logically coherent.

Put this way, of course, no sentient adult would defend such a theory. Even those who have never read Isaiah Berlin intuit from their own experiences that tradeoffs among incommensurable interests or principles are inevitable. They recognize that the urgent and the imminent generally push out the important and the eventual in high-level decision making. They know that disagreement and dissension often affect how public policy is made and applied. More than that, any sober soul is capable of applying this elemental understanding to particular cases if he really puts his mind to it.
I would recast this. The immaculate conception theory of decision-making holds that:
There is one and only one goal to be attained and that there are no trade-offs between goals.

There is perfect knowledge of cause and effect and that consequences can be accurately forecast two or three removes from the action (both in terms of time and proximity).

All affected parties agree not only on the primary goal but also on the ordinal ranking and the relative priorities of subsidiary goals.

That there is no fresh knowledge likely to arise between conceptualization and implementation and there are no feedback mechanisms that change priorities over time.
Obviously all four of these assumptions are wrong and most people would readily acknowledge that they are wrong. But in our Monday morning quarterbacking, we behave as if these four maxims were, in fact, true.

Garfinkle provides a foreign policy example.
How many times have we heard the clarion claim that the covert U.S. effort to aid the Afghan mujahedeen through the Pakistani regime during the 1980s was, in the end, a terrible mistake because it led first to a cruel Afghan civil war and then to the rise of the Taliban? I have lost count.

This argument is about as cogent as saying to a 79-year old man — Ralph, let’s call him — that he should never have gotten married because one of his grandsons has turned out to be a schmuck. But a person does not consider marriage with the character of one of several theoretical grandchildren foremost in mind. It was not possible at the time of the nuptials for Ralph to have foreseen the personality quirks of a ne’er-do-well son-in-law not yet born; so, lo and behold, the fine upbringing that he bequeathed to his children somehow got mangled in translation to the next generation. These things happen.

Similarly, in 1980, when the initial decision was made (in the Carter administration, by the way), to establish links with the mujahedeen, the preeminent concern of American decision makers was not the future of Afghanistan, but the future of the Soviet Union and its position in Southwest Asia. Whatever the Politburo intended at the time, the consolidation of Soviet control in Afghanistan would have given future Soviet leaders options they would not otherwise have had. In light of the strategic realities of the day, the American concern was entirely reasonable: Any group of U.S. decision makers would have thought and done more or less the same thing, even if they could have foreseen the risks to which they might expose the country on other scores.

But, of course, such foresight was impossible. Who in 1980 or 1982 or 1985 could have foreseen the confluence of events that would bring al Qaeda into being, with a haven in Afghanistan? The Saudi policies that led to bin Laden’s exile and the Kuwait crisis that led to the placement of U.S. forces on Saudi soil had not yet happened — and neither could have been reasonably anticipated. The civil strife that followed the exit of the Red Army from Afghanistan, and which established the preconditions for the rise of the Taliban government, had not yet happened either. Of course, despite the policy’s overall success in undermining the Soviet position in Afghanistan, entrusting Pakistan’s Inter-Services Intelligence Directorate to manage aid to the mujahedeen turned out to be problematic, but who of the immaculate conception set knows whether there were better alternatives available at the time? There weren’t; a tradeoff was involved, and it was a tradeoff known to carry certain risks.
Summing up the difficulties of foreign policy decision-making:
American presidents, who have to make the truly big decisions of U.S. foreign policy, must come to a judgment with incomplete information, often under stress and merciless time constraints, and frequently with their closest advisors painting one another in shades of disagreement. The choices are never between obviously good and obviously bad, but between greater and lesser sets of risks, greater and lesser prospects of danger. Banal as it sounds, we do well to remind ourselves from time to time that things really are not so simple, even when one’s basic principles are clear and correct.
All true. So why do we both claim to see the past so clearly and hold the past accountable to the knowledge of the present? Lots of reasons can be adduced ranging from simple ignorance to political expediency.

Obviously hindsight bias plays a role as exemplified by the current Obamacare contretemps related to the Halbig v. Burwell decision (in which the court held that the plain language of the Patient Protection and Affordable Care Act law had to be read as such, i.e. subsidies are only available to those in State exchanges). Megan McArdle focuses on the comments of Jonathan Gruber, one of the chief architects of PPACA, in 2012 in two different forums that comport with the court's interpretation but which Gruber now disavows.
I believe that Gruber sincerely does not remember making these remarks. Memory is fallible; at some point, Gruber probably changed his mind and forgot that he had ever believed otherwise. People show a strong tendency to edit their recollections of prior beliefs to reflect the "correct" answer, and even brilliant economists are not immune to this common cognitive bias.

But though I do not fault his honesty, I also think that in January 2012, Gruber did believe that premium tax credits would only be available on state-created exchanges, and that this would give states a strong incentive to create exchanges.
Hindsight bias is definitely part of the phenomenon but I think there is more going on than that.

I think the inclination towards hindsight bias is facilitated by two general weaknesses in human decision-making. 1) We are often exceptionally poor at creating, articulating, and measuring a coherent, logical, rational, and empirically robust argument towards some goal. If we are honest with ourselves, there is a lot we don't know and there are many uncertainties. As the physicists Niels Bohr is reputed to have said, "Prediction is very difficult, especially about the future." To chase down all the unknowns, imponderables, uncertainties, contingencies, etc. is mentally taxing and time consuming. We are cognitively conservative and take all sorts of shortcuts. "Everyone knows that . . . "; "Obviously . . . "; "That's an exception . . ."; etc. By taking these cognitive shortcuts, the lack of rigor and coherence is hidden. We sacrifice strategic effectiveness (likelihood of achieving the desire outcomes) in order to achieve tactical efficiency (faster, easier decision).

By not formalizing our thinking in advance, we give ourselves lots of wiggle room to reinterpret that thinking in a favorable fashion as circumstances change.

2) We plan strategically but implement tactically. We continually adjust our goals to accommodate emerging information and issues without ever revalidating the underlying premises. We start with a broad vision but then implement on a circumscribed basis. While implementing, we usually see and focus on only what is immediately in front of us and rarely lift our eyes to the horizon to make sure we are still broadly headed in the right direction. Consequently, and too often, we end up at unexpected destinations without understanding how we got there.

There are all sorts of grounds for criticizing decisions made in the past, and many lessons to be learned about how to be more disciplined in decision-making in the future. What we cannot do with integrity and honesty is to criticize past decisions based on the immaculate conception theory of decision-making.




No comments:

Post a Comment