RDP 8301: Financial Innovations and Monetary Policy, A Preliminary Survey V Alternative Approaches

What options other than pure monetary targeting have been suggested for those who formulate monetary policy? Innovations shift and turn the LM schedule so that the conditions which have justified targeting the money supply tend to break down;[11] redefinition as an ex post correction neither precludes further disturbances to monetary aggregates nor ensures policy accuracy by restoring the former historical relation between GDP and money.

One approach is to push ever harder on the segments still under direct monetary control so that, despite innovation and slippage, the desired end result is achieved. The second option involves widening the base for monetary policy, thus spreading the burden of regulation and control. Unfortunately, both these options involve the possibility of the financial system innovating further to evade control and restriction as long as the costs of being regulated remain high. The impact of policy measures will be diluted again. In the extreme, the pressure of control might render traditional monetary policy completely ineffective if liabilities of non-controlled institutions become acceptable as transactions balances, bypassing the bank payments system.

A third course of action would be to abandon monetary targeting.[12] For example, returning to a regime of interest rate targets has been proposed. In a world of high rates of inflation, however, the interest rate most relevant to activity levels is one of the expected real (after-tax) rates of interest, which would be unobservable. A current nominal rate – even a current real rate – is a poor substitute target.

Benjamin Friedman (1982), Jacob Cohen (1982) and Henry Kaufmann (cit. Morris, 1982, p. 8 n. 2) present the case for employing outstanding debt of the non-finance sector as the operating target of monetary policy. Basically their argument rests on an historical relation between non-financial economic activity and the aggregate outstanding indebtedness of all non-financial borrowers, a relation which is at least as stable and reliable as that between activity and any monetary aggregate. In addition, it is argued that this stability is manifest not only in the US economy, but also in Britain, Canada, Germany and Japan. The observed correlation, of course, proves nothing. Even if there were a causal influence from credit to GDP, it might be argued that Goodhart's law will again operate if a credit aggregate were to become the operational target of monetary policy.

Finally, explicit nominal GDP targets have also been proposed, for example by James Tobin and James Meade (cit. Morris, 1982, p. 8 n. 2). Apart from the difficulties in securing timely and reliable data and in directly controlling what is an ultimate aim of policy, it has been pointed out (by Albert Wojnilower, for example) that the announcement of a target for GDP comes perilously close to announcing an unemployment rate and this step could often be politically unacceptable.[13] The links from policy instruments to GDP are also far less clear than when money is the target variable.

Since each of these proposed alternatives carries significant negative aspects, perhaps the most sensible option is to follow a programme of internally consistent multiple targets. Within such a programme, credit, interest rates, money stock and other direct indicators of economic activity might play a role. This amounts to maintaining a broad bank of economic information, minimising economic instability and the overreaction to an unforeseen disturbance in any one target, and tracking and reconciling divergent developments in the indicators.[14]


For a wider discussion of these conditions see Friedman (1982). [11]

See “Discussion” following Lindsey (1982), Friedman (1982), Morris (1982) and Federal Reserve Bank of Boston (1980). [12]

“Discussion” following Lindsey (1982), p.269. In this context, it was proposed that a two or three quarters moving average of nominal final sales could become the intermediate target. [13]

Advocates of this approach include Modigliani in “Discussion” following Lindsey (1982), Porter, Simpson and Mauskopf (1979), Friedman (1980, 1982). [14]