Friday, June 8, 2007

Decision making

Some time ago, I had a discussion with friends on the theory of decision making. Decision making theory is one of my favourite subjects in college. In fact it is an examination question I chose to answer which helped me get good marks. My early study was based on a paper by Amitai Etzioni. The paper introduced three models of decision making: the rational model, the incremental model and the mixed scanning model. The last one is also called the hybrid model where one scans the ground with loose criteria based on incremental consideration and then zooms in for a rational analysis. I wish to use a simple example to illustrate their decision making models.

Suppose we need to organize a get-together party. The decision required is a good restaurant in convenient location, with good food and service at a reasonable price. With the rational model, we need to collect detailed information and compared them to find the best. There is an element of choice and the tyranny of choice comes in. Are we sure that we have not missed a good restaurant? What if we pick a bad one? Should we first sample the food and service of each of them, or read all gourmet magazines for commentaries. What location is convenient to all? What price is reasonable to all. Even with such a simple decision, the resource required to make a sufficiently rational choice is huge. For an extreme maximizer, he will surely get clinical depression.

Clever persons normally use the incremental model. Just find out where did we eat last time. We could just book the same restaurant and same food. But to show that we have done our job, we could consider another restaurant nearby with a slightly better reputation. Waa Laa, a job is done.

A more conscientious person but equally lazy will use the hybrid model. We could first find out comments from the last get-together. Did anyone complaint about the place, food or price? Did anyone have suggestion on improvement? We then look for a few restaurants that could avoid the same pitfall, and then make a detailed rational comparison, perhaps with some site visits, before making the decision.

Our public policy decision making is rarely based on the rational model. Although some issues are researched by academics or by our Central Policy Unit, political decisions are seldom based on the rational approach. It is too difficult, too resource-demanding, and never complete. Charles Lindblom's book of the Science of Muddling Through paints a good picture of the psychological process of decision making of an average person. Civil servants are just average at best. Etzioni does have a point. You just cannot make decision so lazily. Thus upon pressure from the public, academic and politicians, the use of the hybrid model can show that at least some work is done. If you read some of the consultation papers, white papers or LegCo papers, they all offer a few (insufficient and incomplete) choices and then do a detailed analysis on them before recommending one for final decision. That's how we run our business.

No comments:

Post a Comment