He writes,

First, let’s do more research to help reduce the uncertainty regarding the fiscal situation we face and the new, modern and more complex economy in which we live. This step will involve de-emphasizing a number of metrics underlying macroeconomics built around national totals, such as national income, GDP and the aggregate unemployment rate. Instead, we are wise to take a more geographically granular view of our economy that measures regional, state and local economic activity and adapts policies specific to these areas. Focusing upon the national aggregate or the national average masks the extraordinary variation among markets in this country and, indeed, can even make it harder to identify seriously stressful events until it’s too late. This is difficult to do, but c’est la vie.

Pointer from Richard Green (who liked that quote also), in turn from Mark Thoma. The usefulness, or lack thereof, of macroeconomic aggregates is a debate that is worth having. With PSST, of course, I am very much on Follain’s side.

He continues,

Second, and at least as difficult to implement, we need to challenge those who base their policy prescriptions on overly simplistic paradigms and do more “stress testing” of the plans being put forward using multiple potential scenarios

He cites CBO forecasts and bank capital adequacy estimates as examples where such stress testing is most needed.

Finally, he writes,

Third, let’s include more “contingent” budgeting, incorporating explicit connections between future spending and how the economy evolves. One of my favorite examples of this point is a 1999 agreement between the State of California and one of its major pension plans for state employees.[2] An assumption underlying the agreement was that “the Dow Jones Industrial Average would reach 25,000 in 2009 …” The Dow is currently about half of that. The type of contingency budgeting I have in mind would call for adjustments to the pension plan in light of such an erroneous assumption underlying the agreement.

These are all excellent points about the shortcomings of current methods of central planning. However, Follain is trying to come up with ways to make central planning work. In principle, central planning could be improved by using more information more intelligently. In practice? … I would instead look for ways to reduce our reliance on central planning.