Truck and Barter Where Sympathy and Hedonism Collide
Thursday, October 31, 2002
A comment on Lean Left has made me realize the very high regard that many people have of cost-benefit analysis. It appears, for many, to be a requirement for objective analysis. But what happens to the objectivity if our analysis if the LSE and Austrian traditions are right about cost, and cost is a subjective ex ante view of immediate action? Clearly, we must make confidence intervals along with our point estimates; these confidence intervals reveal uncertainty about how people value changes in objective data, and uncertainty on how to make interpersonal comparisons of value, in addition to uncertainty about the objective data.
It seems that the usefulness of cost-benefit analysis will depend on each application.
The PPI in 9/02 is up 0.1%. And? This doesn't tell me much at all. Foods are down, energy is up. The whole is still down about 2% from a year ago.
Crude Goods are up 0.6%, and intermediate goods 0.5%.
The releases go into deep detail about all producer goods, although I don't know the value of looking at each and every number. Which numbers are relevant? How do I identify price changes (or price stabilities) that are indicative of important movements in the economy?
At some level, all this relative price data--both between categories and over time--indicate important movements in the macroeconomy. But I want to note two important caveats outside of the usual technical objections:
Many, if not most, economic transactions are missed by the sampling process--and the sampling process is not necessarily representative.
We really don't know how to utilize the data we do have--e.g. how am I to relate these relative price movements to vertical (dis)integration or new firms.
Aren't some of these prices predetermined in contracts, and others the outcome of search procedures and bargaining? Then many economic actors know some prices ahead to time, and for the ones they don't, they do their best to lower their factor prices and increase their output prices? What is the aggregate price information in this report telling us about the economy?