^{page 78 note (1)} The paper was written by Andrew Britton with the help of Simon Wren-Lewis, drawing on discussions in which Simon Brooks, Stephen Hall, Peter Hart, Brian Henry, Chris Johns, David Savage and Chris Trinder all took part. The main author should be held responsible for all errors, omissions or idio syncrasies of judgment.

^{page 78 note (2)} In the Economic Review for May 1984 we forecast a level of unemployment of 3 million at the end of this year and also at the end of 1985. Our medium-term projections, published in the Economic Review for November last year, unemployment of about 3½ million in 1988.

^{page 78 note (3)} The tracking record of the Institute model is discussed by Simon Brooks and Brian Henry in Chapter 11 of Britton. A. (ed), Employment, Output and Inflation, Heinemann Educational Books, 1983.

^{page 78 note (4)} Reference should also be made to the Centre for Labour Economics, Discussion Paper 165, ‘On Vacancies’, by R. Jackman, R. Layard and C. Pissarides. This paper makes use of male unemployment as a better indicator of the pressure of demand for labour than the ‘bi-sexual’ rate. The implications are similar to those we draw from Census and Survey data about the change in the propensity of unemployed women to register.

^{page 79 note (1)} See S. Wren-Lewis, ‘The roles of output expectations and liquidiiy in explaining recent productivity movenient’. NIER, May 1984.

^{page 79 note (2)} This remark is made despite the existence of much interesting work designed to test the hypothesis of continuous market dearing, for example ‘Neo-classical demand for labour functions for six major ecofiomies’, by J. Symons and R. Layard, Centre for Labour Economics, Discussion Paper 166 and ‘The aggregate labour market, an empirical investigation into market dearing’, by M. Andrews, Centre for Labour Economics. Discussion Paper 154.

^{page 79 note (3)} Lucas, in ‘Tobin and Monetarism: a review article’, Journal of Economic Literature, 1981, criticises Tobin for apparently giving up the long run to the monetarists whilst claiming the short run for the Keynesians. He argues that such a view is untenable.

^{page 79 note (4)} The difficulty of identifying factor substitution may reflect the difficulty of measuring the cost of capital. Neo-Ricardisus on the other hand deny the theoretical importance of factor substitution.

^{page 80 note (1)} Several alternative wage equations are compared by Brian Henry in his paper on ‘Econometric Wage Equations’ in the series of ‘Systematic Econometric Comparisons’ produced by the National Institute last year. Other recent empirical studies include D. Grubb, R. Layard and J. Symons. ‘Wages, unemployment and incomes policy’, Centre for Labour Economics, Discussion Paper 168; N. Foster, B. Henry and C. Trinder, ‘Public and private sector pay’, National Institute Economic Review, February 1984; and S. Wren-Lewis, ‘A model of private sector earnings behaviour’, Government Economic Service Working Paper No 57. The equation now used in the Institute Model is described by Stephen Hall, Brian Henry and Chris Trinder in Chapter 5 of Employment, Output and Inflation (edited by Andrew Britton), Heinemann 1983.

^{page 80 note (2)} For a recent example of an empirical wage equation explicitly based on ideas of this kind see S. Nickell and M. Andrews, ‘Unions, real wages and unemployment in Britain 1951-79’, Oxford Economic Papers, 1982.

^{page 80 note (3)} O. Hart, ‘A model of imperfect competition with Keynesian features’, Quarterly Journal of Economics, 1982, provides an ex ample of a model based on imperfectly competitive goods and labour markets where this is true.

^{page 80 note (4)} Recent cross-section studies, showing the effect of unemploy ment experience on subsequent employment include ‘Repeat unemployment spells’, by Jon Stern and ‘State dependency in youth unemployment’, by Lisa Lynch. Centre for Labour Economics, Discussion Papers 192 and 184 respectively.

^{page 80 note (5)} Registered unemployment, covering 12½ per cent of employees, is on some grounds an underestimate of the resources wasted but on the other grounds an overestimate. The shortfall of the current level of output below its pre-1978 trend also suggests a waste of resources on broadly the same scale.

^{page 81 note (1)} The social costs of unemployment, identified by Adrian Sin- field in What Unemployment Means, Martin Robertson, 1981, include the insecurity of those threatened by redundancy, the frustration of those unable to move away from unsatisfactory or inappropriate jobs, and the hostility expressed by the unemployed towards minority groups. The relationship between unemploy ment and crime is discussed by Carr-Hill and Stern in ‘Crime, Unemployment and the Police’ (1983). Work in progress on the relationship of unemployment and health includes studies of unemployment and the use of health care (at York), of unemploy ment and heart disease (Royal Free Hospital), of unemployment and attempted suicide (Royal Edinburgh Hospital), of the psychological impact of unemployment (Warwick University), and of unemployment and mortality (OPCS).

^{page 81 note (2)} ‘The British Experiment’, being the Sixth Mais Lecture delivered at the City University on July 18.

^{page 81 note (3)} There is an interesting discussion of the merits and demerits of ‘cross-linking’ in rather different context in Chapter 2 of ‘Demand Management’ by D. Vines, J. Maciejowski and J. Meade. The argument there against cross-linking rests primarily on the need to decentralise ‘wage-fixing’. Otherwise cross-linking would generally be more efficient. It is possible that similar considerations arise in drawing a distinction between macroeconomic and microeconomic policy, but it is not obvious that this is the case.

^{page 82 note (1)} An interesting review of the evidence will be found in A. J. Brown, ‘Accelerating Inflation and the Growth of Productivity’, Chapter 4 of ‘Slower Growth in the Western World’, edited by R. C. O. Miatthews, Heinemann 1982.

^{page 82 note (2)} In addition to the article by Simon Wren-Lewis already cited, see S. Hall, S. G. B. Henry, and S. Wren-Lewis. ‘Manufacturing stocks and forward-looking expectations in the UK’, NIESR Discussion Paper no. 64, 1984.

^{page 82 note (3)} This is the term sometimes used to describe the view that monetary targets are needed not so much because ‘money matters’ as because money is thought to matter.

^{page 83 note (1)} See Willem Buiter, ‘Measuring aspects of fiscal and financial policy’, Centre for Labour Economics. Discussion Paper No 193, and references cited therein. International comparisons are regularly made in the OECD Econotnic Outlook. Some related issues were discussed in a note on public sector borrowing by the present author in the National Institute Economic Review for February 1983.

^{page 83 note (2)} This is true of the simulations described by Simon Brooks and Brian Henry in Chapter 10 of Ernployment. Output and Inflation. It is also true of the simulations reported in the Treasury's Macroeconomic Model Technical Manual, November, 1982.

^{page 83 note (3)} The idea has been mooted of internalising the externality of unemployment benefit in the private sector by making trade unions responsible for some part at least of its cost. As yet this interesting concept has not been used to produce a workable policy proposal.

^{page 83 note (4)} The case was argued with almost excessive zeal by the present author in an article offering a ‘free lunch’ in the Observer, August 1983.

^{page 83 note (5)} A simulation of the effects of reducing NIS is included amongst those in the Institute reference cited above.

^{page 84 note (1)} The chapter included a set of simulations, using optimal con trol techniques, in which policy was assumed to adapt to alternative circumstances. The resulting level of unemployment at the end of the period (1988) was lowest, at 2.6 million, in the case of high world growth, as compared with a central projection of 3.5 million on the basis of existing policies.