Wednesday, July 12, 2017

Decision-Making: Some Thoughts

     The political news is all froth and blather, and there are certainly enough willing froth-and-blather miners out there to deal with it, so today I’ll spend a few minutes on a subject that’s been much on my mind lately: how to know whether you have command of the important aspects of a pending decision.

     If you’re familiar with the enormous unwillingness of persons in public offices to make clear, consequential decisions, you’ve probably wondered what could be done about it. As sad a subject as it is, the problem isn’t confined to public officialdom. It exists in virtually any hierarchical structure that grows beyond a certain threshold size. Atop the unwillingness problem, the decisions issued by such persons tend to be of the nebulous sort that can be word-minced a posteriori, such that the issuer can qualify his statements afterward to avert any responsibility for what followed. The unwillingness problem and the ambiguity problem proceed from the same set of incentives, disincentives, and constraints.

     However, solving the problems of sloth and evil is beyond the scope of this tirade. Instead let’s look at the bare bones of decision-making: generically speaking, what a responsible decision-maker should assure himself of before committing to a course of action.

1. Who has the authority?

     “Both for practical reasons and for mathematically verifiable moral reasons, authority and responsibility must be equal - else a balancing takes place as surely as current flows between points of unequal potential. To permit irresponsible authority is to sow disaster; to hold a man responsible for anything he does not control is to behave with blind idiocy.” -- Robert A. Heinlein

     A common observation about hierarchies is pertinent:

In any hierarchical structure,
Authority is pulled upward,
While responsibility is pushed downward.

     The dynamic of power guarantees this. Persons avid for power want it for its positive features, most notably its perquisites, the admiration of the crowd, and the propitiations of the importunate. They’ll do what they can to shed its negative features, which can be summarized as responsibility for the consequences of decisions made. As a man rises in such a hierarchy, both his incentive to avert responsibility and his array of tools and methods for doing so will increase.

     But Heinlein’s observation comes into play. If Smith has the authority but Jones has the responsibility, Jones will lack the power required to effectuate a decision once made. Smith will lack a personal commitment to the consequences of the decision. Should disaster and mutual finger-pointing follow, Jones is the more easily and cheaply sacrificed scapegoat. Any chastisement Smith will experience will be far less.

     A typical case arises in matters of estimation. Imagine that Smith has asked Jones to prepare an estimate for the duration of an effort. Jones does his research and returns with an estimate of one year. Smith immediately demurs and says “It must be complete in six months.”

     Who owns the estimate? More to the point, should the effort take more than six months, who will be treated as owning the estimate?

     Similar cases can be found in all kinds of decision-making, and in all kinds of hierarchies.

2. Does the decision-maker know what he needs to know?

     There are known knowns. These are things we know that we know. There are known unknowns. That is to say, there are things that we know we don't know. But there are also unknown unknowns. There are things we don't know we don't know. – Former Secretary of Defense Donald Rumsfeld

     Knowledge is never absolute nor complete. What we hope for, in facing a decision with significant consequences, is to know enough, and with enough confidence. Yet these, too, are difficult thresholds to part because no one really knows where they are.

     Former Secretary of Defense Donald Rumsfeld was often ridiculed for his “known unknowns” versus “unknown unknowns” formulation. The adversary press treated his statement as gibberish. Yet this is among the wisest things to come from a public official in my lifetime.

     Among the most important bits of knowledge one seeks before making a significant decision are the breadth and depth of the risk factors:

  • Do we know all the risk factors?
  • Do we have a sense for the downside, should those risks materialize?
  • Have we thought about what could be done to mitigate or eliminate those risks?

     Sometimes it’s possible to know all the risk factors that pertain to some undertaking. Sometimes it’s possible to be braced for all of them. But if the project is large, neither condition can be satisfied without serious study. Indeed, sometimes neither condition can be satisfied, period.

     Consider a contractor who must subcontract parts of a project. If he’s fortunate, he’ll be well acquainted with the abilities and histories of all his subcontractors. He’ll have a good sense for the quality and reliability of their work. Yet there are aspects he’ll never be able to control: for example, whether particular artisans will remain available for the duration of the job. Unless he’s far better informed than most persons in his position, he won’t know whether all his subcontractors are financially sound. And of course, many a subcontractor will hedge against future idle periods by overloading itself with current commitments, often to the detriment of promised delivery dates.

     The acknowledgement of known risk factors, and the admission – to oneself, at least – that there could be “unknown unknowns” as well, is a spur toward risk management planning: i.e., what to do if the more plausible unfavorable possibilities should come true. While no amount of risk management planning can be genuinely comprehensive – try to concoct a risk-mitigation tactic for the emergence of a volcano in the middle of your conference room – the risks most likely to reify should be addressed frankly. “Unknown unknowns” can be met, to a degree, with insurance.

3. What is the objective?

     I left this for last because it’s more pertinent to the mid-level decision maker than to the CEO or the private individual acting alone. In addition, it synergizes with topic #1: the problem of authority decoupled from responsibility. It’s far too often the case that no matter how clear the decision-maker’s vision of his objective, persons hierarchically above him will be seeking far different things. When this is the case, the support from above that any mid-level manager needs will be uncertain, perhaps even at cross-purposes to the overt objective of the effort.

     Governmental hierarchies are beset with this problem all the time. Consider the Veterans’ Administration scandals of recent years. I have no doubt that some of the persons involved in the corrective efforts were sincere in wanting to clean up the VA’s act. Yet I also have no doubt that there were persons involved who merely wanted to “display good intentions,” but leave the status quo essentially untouched. We can also see this effect in the shuffling of incompetent or morally deficient public school teachers and administrators to “turkey farms” out of the public eye, thus satisfying the political needs of the higher-ups while mollifying the private parties who forced the matter into the light.

     A decision-maker who suspects that those to whom he reports have a divergent agenda will be powerfully tempted to “sit on his hands,” delaying a decision in the hope that what is hidden might, given time, be revealed. Though such hopes are not often gratified, even less often is there anything else he can do.

     I’ve been fascinated by decision-making approaches for decades: approximately since I went into software engineering. While nothing so diverse can be reduced to a formula, there are generalizable aspects to the thing. Were it otherwise, no one would concern himself with procedure analysis, the therblig, or operations research.

     He who takes a deep interest in this subject is likely to want much more than the surface treatment of its most general aspects. That gets mathematical pretty quickly. However, the windy generalities addressed here might be enough to stimulate a few Gentle Readers into looking deeper.


  1. Never heard the term "turkey farm" in that context.

    In education, the shifting of personnel has been called "the dance of the lemons" - i.e., those that cannot be fired, but no longer can be left in that position, at that location.

  2. A good assessment, but I would make one change. Responsibility isn't pushed downward, it's _spread_ downward. And the larger the hierarchy, the more diffuse the responsibility becomes. When the organization gets as large as the Federal bureaucracy, you'll be hard-pressed to find anyone to be responsible for anything (until a scapegoat is needed).


Comments are moderated. I am entirely arbitrary about what I allow to appear here. Toss me a bomb and I might just toss it back with interest. You have been warned.