Philip A. Klein
When approaching the analysis of business fluctuations, conventional statistics courses are probably still organized so as to give the student the impression that the field consists of a stock of techniques. Acquire them and one is equipped for a lifetime of work in the statistical analysis of—in this case—business cycles. This chapter is devoted to demonstrating how wrong this conclusion would be. The approach taken in the early days of the National Bureau of Economic Research (NBER) by Mitchell, Arthur Burns, Geoffrey H. Moore and their associates is a monument to the cogency of the view that there is, in fact, a perennial need for statistical innovation in a dynamic research area such as cyclical instability. Here we shall focus on the recent past.
It is a cliché to note that the economy is dynamic and that interrelationships critical in one cyclical crisis may vary both in their impact and in their importance in other crises. What is seldom realized is that statistical techniques optimally effective in exposing these changing critical relationships may themselves need to be varied or changed altogether over time. The most obvious example, perhaps, would be seasonal adjustments, the use of which was recognized but which were made only sporadically a generation ago. They were arguably less critical in earlier times, when inflation rates were less significant in their day in, day out impact on the functioning of the economy. Today, there are relatively few time series used in cyclical analysis which are not initially and routinely adjusted for seasonal variation. But there are many other examples and this chapter will concern itself with some selected from Geoffrey Moore’s work.
The bulk of his work is a testimony to the assertion that the assessment of cyclical developments over any substantial period of time with an unchanging arsenal of statistical techniques is likely to prove increasingly inadequate as the economy moves further away from the time in which the techniques were developed.
One of the strengths of the National Bureau of Economic Research in its