Some New Evidence on the Timing of Consumption Decisions and on Their Generating Process.

While quarterly consumption data are known to be well fitted by an integrated first-order moving average process--IMA(1, 1)--with a positive coefficient, monthly consumption data are found to be well fitted by an IMA(1, 1) process with a negative coefficient. Without measurement errors, one implication is that, if R. Hall's (1978) random walk model of consumption behavior is true, then the agents' decision interval must be greater than a month. (In particular, this evidence rejects the possibility of continuously taken decisions.) Another implication is that, if consumption decisions are generated by an IMA(1, 1) process at intervals shorter than a month, the coefficient must be negative. The paper also discusses the case of monthly data corrupted by measurement errors. Copyright 1989 by MIT Press.