Was the Financial and Economic Crisis Predictable?

The economic science came under fire during the last financial and economic crisis for many reasons. One thing that came frequently under attention was the supposedly inability of economists to predict the crisis. Well, it seems that the crisis was predicted to a certain extent by some economists, at least that’s what the paper by Bezemer shows.

Noah Smith has an interesting post on this. I would make several points about his blog post. First of all, his detailed remarks about what it means to predict crises are a bit misplaced. By chance, I am a familiar with papers published in economic forecasting and I don’t think that they are usually required to pass such rigorous tests for forecast accuracy. To be more precise, no one would require predicting “the exact timing” of the beginning of the crisis. Nor would anybody expect predicting the exact duration of the crisis as well as the size of the recession in a certain year. On the contrary, forecast accuracies are usually judged on a sample of reasonable size based on different statistics and tests. So, in the end, I think that Noah Smith’s criteria for prediction do more harm than good.

Then the paper proceeds by picking up one of the candidates who supposedly predicted the crises, namely Steve Keen. I am not sure whether this is a random choice, or just a way to prove that Steve Keen was only slightly better in predicting the crisis than mainstream economists.

However, the main point of my comment is that the paper should have focused on the paper by Bezemer (and I agree with one of the commentators on Smith’s blog that made a similar remark) instead of just mentioning it. Indeed, Noah Smith suggests that his analysis of Keen’s predictions can be applied to any other pundit that pretended to have the crisis predicted. But I think that a more constructive approach would have in discussing why those pundits or economists made these predictions and why mainstream economists did not do the same predictions. Of course, a second issue would have been discussing Bezemer’s model for predictions (using accounting macroeconomic models).

Leave a Reply

Your email address will not be published. Required fields are marked *