^{page 40 note (1)} Jevons, Marshall, Irving Fisher and Friedman are prominent examples. There are probably others.

^{page 40 note (2)} There may of course be other (more broadly defined) social costs of inflation—for example, its divisiveness.

^{page 41 note (1)} These problems are more fully discussed by J. S. Flemming and I. M. D. Little in ‘Why We Need a Wealth Tax’, Methuen 1974, see especially pages 6-7.

^{page 41 note (2)} Of course the tax is not payable until the ‘gains' are realised.

^{page 41 note (3)} These gains are real only to the extent that the prices of companies' stocks rise more quickly than the general level of prices in the rest of the economy.

^{page 42 note (1)} This argument is intended to be analogous to that used by Professor Arrow in his discussion of how the non-existence of some markets can help to explain the non-existence of others. See American Economic Review, March 1974, ‘Limited knowledge and economic analysis’, especially page 9.

^{page 42 note (2)} The typical mortgage contract in Britain (in which the nominal rate of interest can vary during the life of the mort gage) is certainly not the same as a fully indexed mortgage. John Whitley's article in this issue explains the differences. But variable rate mortgages come nearer to neutralising the effects of inflation on mortgage repayments than the traditional US system of fixing the nominal interest rate for the whole term of the mortgage.

^{page 43 note (1)} The type of control which I am considering is one which sets ceilings on the allowable rates of increase of wages or prices; these ceilings may be given by formulae in which some variables are not known when the controls are imposed (the allowance for thresholds and the productivity deduction are recent examples). I specifically wish to exclude both anti- trust laws and legal constraints on the institutional framework in which wage negotiations are conducted from ‘controls' as defined here.

^{page 43 note (2)} This point has been emphasised by S. Brittan, ‘Govern ment and the Market Economy’, Hobart, Institute of Economic Affairs, 1971. See especially pages 52 and 55-56.

^{page 43 note (3)} Monetary Policy in 1974 and Beyond’, Brookings Papers on Economic Activity, 1974 (1). See especially Fig. 3, page 230.

^{page 43 note (4)} If applied in an identical situation the US method of measuring unemployment would give a higher estimate than the British method. See National Institute Economic Review, no. 56, May 1971, pages 66-69.

^{page 44 note (1)} Both the index-linked and the non-index-linked systems will give rise to similar anomalies: those who are due for wage increases are liable to do worse than those who have just re-negotiated contracts.

^{page 44 note (2)} Fortune, July 1974.

^{page 44 note (3)} Though preferably less severe than the restraint operating during the first three quarters of this year.

^{page 44 note (4)} It would not be appropriate in a survey on index-linking to discuss the grounds for rejecting other possible arguments for permanent controls, such as their alleged ability to reduce the harmful effects on allocative efficiency (and especially unemployment) of the monopoly powers of trades unions.

^{page 44 note (5)} To provide the authorities with detailed, operational policy rules is difficult and liable to large errors (like most other things in economics): the money supply, on whichever definition, is not directly controlled by the authorities and the velocity of circulation can and does change. One can move towards an operational rule for policy in the following way: the total demand debt of the government held by the private sector should not be allowed to grow (either by deficit spending or through the operation of the Exchange Equalisation Account) faster than the tolerated rate of inflation plus the rate of growth of real output, unless there is some good reason-such as an anticipated rise in the real rate of return on government securities or perhaps a call for increased special deposits-for expecting the private sector to plan to increase the real value of its holdings of currency and govern ment securities relative to its real income by the required amount. Even this rule is not really operational without a detailed forecasting model.

^{page 45 note (1)} ‘Inflation Policy and Unemployment Theory’, Macmillan 1972.