Today I was interviewed by a Japanese TV programme for about an hour. I was asked what I want to do in the future, and I thought to myself that I am making a decision that I can't lose. It is called Robust Decision Making.
I'm reposting this again because I just found it the other day while reading my Oxford blog. It's something I put together 13 years ago, but I think it's still relevant. Uncertainty is on the rise, so one strategy is to adapt to an uncertain future at critical points. A vision of forging your own path through an uncertain future is just the opposite of "not to lose" decision-making.
Areas where "not to lose" decision-making theory works
For decades, there have been solid alternatives to utility-maximising decision theory, such as cost-benefit analysis in operations research. Utility maximisation is "to win" decision making. At that time, researchers developed the "robust" decision theory as an alternative to the "to win" decision theory because they thought that there should be a more defensive decision theory.
When I read the article by Rosenhead*1 in 1972, which is still referred to in many places, I found that the methodology is still applicable, but the part of the introduction, "Robust" decision theory is good for CEOs of huge companies", may no longer applicable.
The areas where "robust" decision theory is good are
Uncertainty is a big part of decision making.
Long term decision making.
Areas where competition is not fierce.
Areas where there is absolutely no room for error.
I think this is the area where I work with the poor in Africa and elsewhere. Uncertainties, such as climate change and political factors, are reflected in decision-making in agriculture. It is a world where competition is not fierce, but there are no second chances, and if you fail, there are no seeds for the following year.
Conversely, there are areas where 'robust' decision-making theory is of little use: for example
Short-term decision-making, where uncertainty is not so much a factor.
Highly competitive fields.
Areas of risk-sharing.
Recent decisions made for short-term profit and exposure to unprotected international competition should not be based on "not to lose" decision-making. The joint-stock company system itself is risk-sharing, so decisions should be made "to win", and in a world where there are only a few winners' chairs, a company that makes decisions "not to lose" will not stand a chance. The big companies that used to have a hand in every industry are slimming down.
In the developed world, the only areas where "robust" decision-making is useful are those where the long-term vision is important, such as information and transport infrastructure, and where the state supports it, I thought. I think the emphasis on "never-fail" may be a bit over-emphasised and over-optimistic.
*1:ROSENHEAD, J., ELTON, M. and GUPTA, S.K., 1972. Robustness and Optimality as Criteria for Strategic Decisions. 1970-1977), 23(4), pp. 413-431.