RDP 7704: Money and Money–Income: An Essay on the “Transmission Mechanism” VI. Summary and Conclusions

This essay has covered a great deal of ground but its principal theme is easily summarized. The IS-LM model provided a common theoretical framework in terms of which a spectrum of viewpoints about the nature of the “transmission mechanism” of the effects of monetary policy could be reduced to questions about the empirical magnitudes of the parameters of particular behavior relationships. The model treated the money supply as an exogenous variable, independent of fiscal policy, dealt with a closed economy, and offered no satisfactory way to analyze the division of changes in money income between real income and prices. Each of these shortcomings seemed open to a remedy that involved extending the structure without at the same time altering its basic properties, but, in the event, the attempt to solve these problems has led economists to a basic reassessment of the nature of macroeconomic analysis.

At the heart of this reassessment lie the results of the search for the “missing equation” that divides up changes in money income between real income and prices. To begin with, it has become apparent that the processes underlying this division must be treated as forming part of the transmission mechanism that links money and money income, rather than as involving a subsequent and analytically separable series of events. The expectations augmented Phillips curve, which is the most popular candidate to fill the role of “missing equation,” rests on theoretical foundations which are as yet not fully developed. There seems to be a good deal of empirical evidence in its favor, but what kind of micro postulates it rests upon, particularly as regards the labor market, is still an open and controversial question. It is hardly surprising that much work on the transmission mechanism has in recent years centered on labor market behavior.

Even so it is not doubts about the way in which the forces of supply and demand operate in the labor market that have undermined our convention way of thinking about macroeconomic problems. Rather it is the role that price expectations play in determining all kinds of behavior (including labor market behavior), that forces us to rethink so much of our analysis. If such expectations were related to the behavior of particular economic variables in a stable and discoverable way, then even if such relationships differed between agents, the way in which they were determined could be treated as part of the structure of the economy. We could argue that, to the extent that we had not yet discovered such relationships, there existed a gap in our understanding of the structure of the economy, of the transmission mechanism through which policy, including monetary policy, operates to influence prices and output. That would be a problem, but not a fundamental one.

However the Issue raised for macroeconomics by the emphasis we now lay upon expectations is more basic. If economic agents are capable of recognizing that economic policy, and the institutional framework against whose background it is carried out, influence the environment in which they themselves operate, then it is plausible to postulate that information about the conduct of economic policy, and about that institutional background, will be used by at least some of them in forming their expectations. Hence, the way in which fiscal policy, or the exchange rate regime adopted by an open economy, interacts with monetary policy will influence not just the values of those expectations but the very manner in which they are formed.

This would imply that there is no such thing as a unique transmission mechanism for monetary or any other kind of policy, knowledge of which will enable us first to discover how, in particular historical episodes, alternative policies to those actually implemented might have worked, and second to choose in any current situation, the best policy from a menu of alternatives. Rather, there is potentially a different transmission mechanism for every policy regime. If this is so, the interior of the famous “black box” has no unique structure to its contents, which certainly explains the difficulties we have encountered over the last two decades in trying to discover what that structure is. Whether it will prove possible to generate empirically useful hypotheses about the way in which the nature of the transmission mechanism varies with that of the policy regime, so that it becomes feasible systematically to allow for the endogeneity of the transmission mechanism in building macro theories and designing macro policies must remain to be seen. This question, though, must surely be the crucial one for future work in macroeconomics.