Decision Theory for Planners #113 To decide or not to decide? – the Delay Dilemma
When it is not necessary to make a decision, it is necessary not to make a decision.” ~ Lord Falkland
“A good plan, executed now, is better than a perfect plan next week.”
- General George S. Patton, Jr.
The above two quotes show two sides of the delay dilemma. For Falkland the risks of deciding now outweighs the benefits. Better information, or changed circumstances make deciding now risky. For Patton the risks of waiting outweighs the benefits, we might have good understanding of the situation now, but not in the future, by when circumstances might not be in our favour. Patton of course was one for putting the other side on the back foot, with his nostrom of ‘attack is the best means of defence’.
We then seem to have two factors that weigh on the issue of the meta-decision, that is how, when and why we make a formal decision. How fast circumstances are changing and the quality of our information and the perception of how this may change in the future.
You might think then that a simple approach of rational certainty. I.E a two factor optimisation model would suffice of the form of: If the benefits of making one of two or more options now outweigh the uncertainties and dis-benefits of making the decision in the future then make the decision.
Such a theory though does not stand up to serious thought.
When we make a meta-decision to decide we are making two decisions, to decide and the choice we decide. when we make a meta-decision not to decide we make only one.
The rational certainty approach does not stack up because the decision to decide and the choice we make are always and indivisibly one decision. We decide because our choice and its risks have advantages to the staus-quo ante when projected forward, even where projected forward one infinitesimally small division in time.
Where we cannot decide it is either because the uncertainties of delay outweigh the advantages of delay or we are struck by geniune indecision between two or more choices and are seeking resolution of those choices.
Meta-decisions therefore have a fundamental antisymmetry in time compared to other decisions. Indeed the uncertainties associated with meta-decisions seem to dominate the public policy process.
What is more because the future is fundamentally unknowable rational ignorance is essential for making a decision at all. As you can always seek out more information it might seem more rational to delay, but that delay has an opportunity cost in terms of future information seeking yourself or from others.
A decision to decide is effectively saying to yourself I cant predict the future but im going to make an assumption that the future will not be significantly divergent from the future I have mentally projected. That is the future will not be surprising.
If though that future is surprising to someone else we may have an advantage over them, particularly in terms of economic advantage. Seen in this way it supports GLS Shackles view that economic decisions are made in terms of assessment of competing degrees of potential surprise, and that decision making is a creative act in terms of how we shape the future.
As well as not being able to disentangle a meta-decision of yes from a decision we cannot disentangle such decisions from other actors.
“In some negotiations, the tactic of one side might be to delay negotiations indefinitely. For example, environmentalists can often discourage a developer through protracted litigations.” – Howard Raiffa, in The Art and Science of Negotiation, 1982
A decision space is not just defined by the actor making the decision but those who supply information and can react to a decision seeking to overturn it or by supplying information in the hope that the decision is delayed. Such examples are legion in planning.
Strategic delays are often used by disputants who do not have the power to win directly in court or through procedure. Rather than waste their resources on a fight they cannot win, some use delaying tactics to frustrate their opponent hoping that circumstances turn in their favour, such as a change in national policy. If a group is part of the decision-making process but does not want change, the slower the process, the better. If they do not have a say in the design of the process, then they can deliberately delay it by stalling on their involvement.
Most recently the crackdown on corruption in India has become so great that civil servants are trying to prevaricate on all decisions by giving excessive briefing notes to ministers because they know that all decisions will subject them to scrutiny whether innocent or not.
Also complex decisions are interconnected, often part of a social mess, so we are often left in a position of a possibly infinite regress of not being able to decide one thing without deciding everything. This might seem rational looking at individual widenings of the decision space but overall the dis-benefits, of a decision being delayed exponentially, force a cut of in terms of the scope of the expected interconnected impacts of our decisions.
The final problem with rational certainty is that the vast majority of decision made are not made on the basis of assessment of more than one option.
Gary Kein’s research, set out in his book ‘Intuition at Work: Why Developing Your Gut Instincts Will Make You Better at What You Do (2002) derived from his work in the US miltary on how commanders make decisions.
He found that 95% of decsions are made without options, without concious consideration of evidence, they were made on instinct. This led him to wonder, with only one option waht was the decision compared to?
He believes that a given situation, the decision maker will pick up cues and indicators that let them recognise patterns. Based on these patterns and the decision they have to make, the person chooses a single course of action, an ’ action script‘, that they consider will achieve the outcome – they ran a mental simulation.
The mental simulation was based on mental models that the decision maker had developed through experience. In other words, the decision maker has an idea how things work based on the knowledge that has been gained from experience. S/he compares the option against what is known to work.
This is intuitively correct. Inexperienced planners will often be paralysed by the need to seek more evidence and the choice of options, experienced planners will often leap to a conclusion and then either seek evidence to support it or disprove it. As people become more expert in their chosen field, as they have more and more experience, their ability to recognise patterns is enhanced. This gives them more options to choose from. Which means that, more often than not, the first option they choose will work.
If the decision maker considers the action script will achieve the outcome, they go ahead.
If they consider that it might not work because of a potential problem, they may try and alter the action script in some way. If mentally they don’t think it will work, they discard it completely, and choose a second action script.
This is then mentally rehearsed and so on until they find an action script that they think will work. This is then utilized. Note that in this recognition primed decision making model there is no comparison of alternatives. If this pattern is used often then the decsion can be routinized. Routinized decisions can also be instituationalised as policy, indeed this is the source of policy that works, based on expereince of achieving successful outcomes.
This theory is very similar to Gerd Gigerenzers ideas of gut instincts we met in the first part of this series.