Knowledge that Transforms

Monetary economists concerned with the indicators of monetary policy issue are like dermatologists in that their patient seems incapable of either dying or getting well. The indicators issue cannot be expected to die so long as the two sets of prime indicator candidates-interest rates and monetary stocks-frequently yield conflicting information about the stance of monetary policy. If it is unlikely that the patient will die a quiet death, some consider it even more unlikely, in the near future of course, that he will be cured to everyone's satisfaction through the concentrated application of evidence. Rather than presenting additional hypotheses and evidence, this paper is concerned with revealing the implications of some existing hypotheses for the indicators issue. These hypotheses include four large-scale econometric models referred to here as FRS-MIT [5], FRS-CHI-MIT [13], Brookings [8], and Ando and Goldfeld [1]. Being hypotheses of unknown quality, these models are not likely to either kill or cure the patient, but they do contribute to the debate in several interesting ways. Detailed discussions of the indicators issue are available in the literature [2] [3] [16], so only a brief summary is offered here: (1) An indicator of monetary policy is a variable which yields information about the thrust of current monetary policy actions on a goal variable in the present or some future time period. Commonly mentioned indicator candidates are prices and quantities observed in money markets. (2) In comparing the quality of two indicator candidates, one is said to be superior if it is more likely to yield true and less likely to yield false information than its competitor. Stated differently, an indicator of monetary policy should be relatively insulated from shocks other than those generated by monetary policy actions. These conditions are set out rigorously in the next section. (3) An ideal indicator, which reveals the separate, quantitative effect of policy actions on the goal variable in the present and in all future time periods, is differentiated from a less than ideal indicator, which may reveal only the direction of policy effects with some probability of being correct. According to the models mentioned above, the open market policy instrument (bank reserves or monetary base) approaches being an ideal indicator if no other policy instruments are concurrently used. This assertion is also elaborated in the following section. It is expedient and I believe correct to divide the participants in the indicators debate into price or interest rate watchers and quantity watchers. Although price watchers tend to be those stressing Keynesian income-expenditure theories and quantity watchers tend to be those stressing quantity theories of national income determination, there is no logical relationship between the choice of a theory and the choice of an indicator. For instance, even a perfectly constant ratio of money to income by itself implies nothing for the indicators issue (the money stock may be independent of monetary policy actions). Similarly, a theory with an explicit interest rate transmission mechanism from monetary to expenditure sectors does not imply that interest rates are the best indicator of the separate effect of monetary policy actions on expenditure variables. Noting that these are logically separate issues aids in explaining some of the implications of the models. In deriving the implications of the models for the indicators issue, the field of indicator candidates is reduced to one price, the Treasury Bill rate r, and one quantity, total privately owned bank deposits M.' Two further simplifications are (1) of all the exogenous shocks which could make r or M yield false indicator information, only past changes in the open market instrument and current changes in fiscal policy represented by government expenditures are considered2 and (2) price watchers are assumed to interpret increases in r as an indication of tight policy and decreases in r as indicating loose policy, while quantity

The allocation of investment to particular industries is a major problem in economic projection and planning. The purpose of this paper is to examine the stability of the sectoral incremental capital-output ratios over time and their differences among countries. It is clear that the usefulness of these ratios in projecting capital requirements will depend on how stable they are, at least over the periods used in development programs. Section I examines the problem of measuring the incremental capital-output ratio and lists the factors affecting it. Section II offers the empirical results of the behavior (trend and variability) of the ratio in a number of sectors of several countries. Section III deals with the possibility of misallocation of investment even if the ratio is stable, and offers an interpretation of the findings.