Paul Krugman discusses the value of models.
But he misses an important point: the value of models is also their greatest flaw. Models simplify things so everybody can follow the train of thought, but that simplification means they’re always wrong. In other words, they are excellent tools for communicating ideas and highly flawed when it comes to understanding complexity.
So the problem with models is that they served to hide the fact that this assessment of the Depression is also entirely true. “Faced with the Depression, institutional economics turned out to have very little to offer, except to say that it was a complex phenomenon with deep historical roots, and surely there was no easy answer.”
As far as I’m concerned the only solution to this problem is to develop competing models and demand that each individual researcher decide for him or herself what weight should be put on each model in explaining a given circumstance. That is, to acknowledge that there is no easy answer and enjoy the consequent debate.
Thus the task of economics is to encourage and facilitate the development of competing models. This will require that far more respect be paid to verbal models, as qualitative models are needed to lay the foundations for the development of new formal models.