Well today the National Bureau of Economic Research (NBER) declared the ending date of the Great Recession. They put the end of the current downturn as June of 2009. In my charter blog post from March 6th 2010 I placed the end of the downturn as July of 2009, confirming that my approach was sound.
I think this helps prove my point with respect to how economic data mining can be effectively used: (1) intuitive but simple approaches that work, or (2) excessively complex models or methods that take months or years to "prove" correct.
My point with the initial post was to illustrate the practical benefits of the former over the latter. When one can adequately predict the key points with as few inputs as possible, then it should be deployed as a first-response diagnostic. The benefits are that policy makers, business leaders, and voters(!) can therefore more quickly react to a new economic reality and mitigate the tendency of a cycle to swing in excessive directions.
Of course in-depth research should be done to better understand how we collectively create messes that are costly and time-consuming to clean up, but these lags in time should not be used (as they are today) as an excuse to befuddle the masses with time-wasting debates.
As it stands today we have quite enough economic illiterates and short-term thinkers sitting in positions of power across government, the media and business sectors, who make decisions that affect a great many people, so the less time we data mining analysts give them to speculate on whether the earth is round the better off we will all be, on average, over the long term.
No comments:
Post a Comment