OPEN-ECONOMY MARKET: Existing Arguments for NIT 2

The target inflation rate would be more accurately achieved than with monetary aggregate targets, the argument goes, because advance knowledge of average velocity growth would not be required. In addition, real output and employment fluctuations might be smaller on average than with pure inflation targeting because of the implied response of the rule to unusually high or low output growth rates.

The latter result can not be assured, because of the profession’s ignorance concerning the mechanism by which nominal income growth is split between inflation and real output growth components, but the conjecture was that it would be achieved in a wide variety of models — this is the robustness idea — and in actuality. Also, in recent writings (including McCallum [1997b] and McCallum and Nelson [1998]) it has been stressed that NIT — especially in its growth-rate version — avoids dependence on unreliable measures of the “capacity” or “natural” output level that are required in several prominent rules, including the one proposed in Taylor’s influential “Discretion versus policy rules in practice” (1993a).

Regarding the relative desirability of nominal income or inflation targeting rules, McCallum (1997a, p. 232) has suggested that “because the prices of goods and services… react more slowly than output in response to monetary actions, cycling and dynamic instability are more likely to occur with a price level or inflation target. In other words, the problem of ‘instrument instability,’ which would render the targeting attempt entirely unsuccessful, is intensified.”

Also, the same passage suggests that it may be “more difficult to devise a policy rule for hitting inflation targets than nominal GDP targets, because the former requires an understanding of the forces that determine the split of nominal GDP growth into its inflation and real growth components.”4 These conjectures are not buttressed by any theoretical results, however, so it is at best an open question whether they are consistent or inconsistent with quantitative results in (a variety of) carefully-specified models.

Tags: , ,