Strategic Thinking

Foundations and Methods

It is easy to formulate a strategy or vision for R&D if you do not care what theoutcomes are. Just convene a meeting of powerful stakeholders, smile a lot, get them to free associate out loud,and then put together a collage of the most harmless, watered-down wordsthey emitted, making sure they are all sure that the check is in the mail tothem personally. This happens often enough in the world, and it is a majorsource of programs that are very effective in enriching intermediaries inthe short term and of producing absolutely nothing in the end.

But if we want results, the strategy formulation for R&D is nodifferent in principle from the strategy formulation in other areas. First,we must have some idea of what kinds of results we actually want. Inprinciple, we need a kind of global metric to define what are better resultsand what are worse results. For many years, economists inspired byutilitarian philosophers like John Stuart Mill, argued that we need todefine a utility function U(X), which gives an overall rating or score to any possible outcome X. The vector X is simply acollection of all the n variables, X1 through Xn, which we view as important inevaluating the outcomes. At this stage, they do not have to be measured variables; the challenge, to begin with, isto define what we really want and whether we can measure it.

If we have one idea of what we want on Mondays and a different idea onTuesdays, we usually end up fighting ourselves. We end up in a situationthat is worse according to both systems of preferencethan some other situation we could have gotten to if we were moreconsistent. Thus, if we want results, we will try to be consistent aboutwhat we want, at least at any one time, as we are formulating a strategy. Weneed a utility function. This is a key principle. Of course, we also need tobe very adaptive as well in S&T policy, as new information comes in,changing our options and what we expect to get from them.

...