OPEN-ECONOMY MARKET: Existing Arguments for NIT

To a considerable extent, the case for NIT as developed in the early 1980s grew naturally from a perception that the then-prominent strategy of monetary aggregate targeting could be improved upon. Large and unpredictable changes in payments industry technology and regulatory practices had led to well-publicized instability in the short-run relationships between monetary aggregates, such as Ml, and nominal GDP (as well as other measures of aggregate nominal spending).

In addition, several economists held the belief that a policy that smoothed out fluctuations in nominal GDP would be more effective in stabilizing real output and employment than a policy that smoothed the path of a monetary aggregate. These ideas were put forth in papers by Meade (1978), Tobin (1980), Hall (1984), Gordon (1985), Taylor (1985), and McCallum (1985) read only.

At about the same time, papers by Bean (1983) and Aizenman and Frenkel (1986) developed analytical results suggesting that NIT would be superior to other targeting schemes in terms of the implied automatic policy responses to shocks of particular types. Root-mean-square (RMS) deviations of macroeconomic aggregates relative to desired values would be smaller, that is, according to small analytical models in which such magnitudes could be calculated precisely. More recently, this line of work has been extended by Frankel and Chinn (1995) and Ratti (1997).

It has been shown by various authors, including West (1986) and Henderson and McKibbin (1993), however, that claims for the theoretical superiority of NIT have been overstated in some of these studies as a consequence of the non-robustness of results to details of model construction, including the failure to take account of some types of shocks. Other notable recent contributions include Feldstein and Stock (1994), Hall and Mankiw (1994), and Cecchetti (1995).

The criterion of robustness to model specification has been a prominent element in the NIT discussions of McCallum (1988, 1993, 1997b). The argument in this work has been that keeping nominal GDP, or some other measure of nominal spending, close to a target path that grows smoothly, at a rate equal to the long-run average rate of growth of real output plus a desired inflation rate, would result in an average inflation rate close to the desired value and perhaps in reduced fluctuations of real output and employment

Tags: , ,