Policy options under a labour government.

Anderton, Bob ; Gregg, Paul ; Soteri, Soterios 等

Our forecasts for the domestic economy in Chapter 1 assume the continuation of present economic policies, and hence implicitly that the Conservatives are returned to power at the next General Election. Here we look at some of the changes in policy which might follow a Labour victory. As yet, relatively few firm commitments have been made by the Labour Party, so we cannot make any overall assessment of their plans. Instead we consider a range of options based on the general statements of objectives which have been made.(1) The aims and priorities of a Labour Government would differ somewhat from those of the present Government; and a Labour Government would consider a wider range of options for policy change, including some that have been ruled out explicitly by government in recent years.

A new Government does not set up business on a green field site and judging by its recent policy statements, the Labour Party, recognises this. But the case for a fresh look at economic policy is strengthened by the disappointing performance of the economy in the last few years. There could be radical change in some areas, but important elements of continuity as well. There is no intention of conducting policy in the way which has characterised previous Labour Governments. This is reflected in the options we have chosen to analyse in this chapter: for example we do not include indicative planning, trade protection or incomes policy.

We begin by discussing relations with the rest of the European Community. The extent to which economic policy is harmonised within Europe narrows the range of options open to a Labour Government, in monetary policy, in industrial policy and in virtually every branch of economic policy. We then discuss the macroeconomic issues, starting from the firm commitments already made for extra spending and tax changes. We assess the trade-off between inflation and unemployment, drawing on an analysis using optimal control techniques in Annex A. There is a separate section on the possible role of credit controls as an additional instrument of policy. In Part Two we consider some of the microeconomic or structural policy options already under debate, especially the effects of higher investment and training. The setting of a minimum wage is discussed in Annex B. In a chapter of this length we cannot hope to be comprehensive-we do not, for example, discuss industrial relations, housing or local government finance-but conclusions do emerge about some aspects of Labour Party policies. Where possible we quantify these conclusions using the Institute's model of the UK economy.(2)

Policy towards Europe

In recent years the Labour Party has been more enthusiastic about the moves towards European economic integration than has the Government. This is most evident in the areas covered by the Social Charter, where the approach in Brussels is more in line with that of the Labour Party than the Conservatives in Britain. The attitude of a future Labour Government to European integration would be influenced by political developments in other major member countries, especially France, where the strongly pro-European policies of the Socialist Government provide an important precedent for a future Labour Government in this country.

Britain is now a full member of the EMS. This means that monetary policy cannot be conducted in this country independently of policy elsewhere in Europe. If the system evolves, as is the declared intention of the other member states, towards a full economic and monetary union, then the remaining possibility of realigning the exchange rate will be progressively eliminated. The Labour Party was ahead of the Government in advocating full membership of the EMS. They have indicated a willingness to contemplate a new European system of central banks' provided that it is politically accountable'. They would also hope to negotiate additional financial transfers, favouring the less prosperous regions of the EC, from which the UK might gain.

The pace of progress towards EMU is currently in question. A Labour Government elected in 1991 or 1992 would have to decide whether it wished to retain any exchange-rate flexibility, and if so how it wished to use it. We discuss the costs and benefits of a realignment after the election . We assume however that a Labour Government would wish to be less obstructive to plans for EMU than the present Government, rather than more.

Economic integration in Europe is not just a matter of macroeconomic policy, although that has attracted most attention in the last year or two. The single market programme rules out any policy measures which are openly, or even tacitly, protectionist. Public sector procurement for example, where extra spending could well follow a Labour election victory, has to be opened up to competition from other member states. Any policies which seem to promise a subsidy to British industry will be subject to careful scrutiny in Brussels. Even in the case of rates of taxation there has to be some presumption in favour of harmonisation. Special employment measures have to be designed in such a way as to avoid the suspicion that they are a disguised form of trade protection. This does not mean that there is no scope for a distinctive employment or industrial policy under a Labour Government, but it is a consideration which would have to be kept in mind every time a proposal is put forward.

Thus the process of economic integration in Europe, and the closer policy coordination it entails, necessarily narrows the range of options open to any Government in Britain. This has advantages for a Labour Government, even if they would find it frustrating in some respects. The fear that a Labour Government would introduce controls on trade or capital flows for example could arise and be damaging to confidence, even if the new Government has no such intention in fact. The existence of the European 'constraint' on policy could in these circumstances reassure industry and the markets in a way that would help to underpin the strategy actually adopted.

PART ONE. MACROECONOMIC POLICY

When the medium term financial strategy was first introduced its title had to be chosen with some care to avoid the use of the word plan', with its connotations of socialism. Whatever changes it made to the name or the format of the MTFS, an incoming Labour Government would need some medium term objectives for macroeconomic policy and a framework within which their achievement could be monitored year by year. Unless the economic situation changes rapidly between now and the election, the objectives of the plan would have to include the reduction of the rate of inflation and of the deficit on the balance of payments.

No Government is in fact indifferent to the cost of reducing inflation in terms of lost output or higher unemployment. Traditionally the Labour Party has been thought to be more concerned than the Conservatives to keep unemployment low, if only because its own supporters are more at risk of losing their jobs. This emphasis has been less evident, however, in recent policy statements. We would not want to suggest that policy under a Labour Government would necessarily be more expansionary in current circumstances than policies under a reelected Conservative Government. A Labour Government which accepted an early date for EMU would need to follow less expansionary policies than a Conservative Government which did not. Moreover, a change of Government would add to exchange-rate uncertainty, perhaps for several years, until the markets came to accept the credibility of new policy commitments. For that reason, other things being equal a rather larger risk premium would be attached to relative interest rates in the UK.

Firm commitments

The firm Labour party commitments for tax and benefit changes are summarised in table 1.(3) Our estimates are based on the assumption of an increase in pensions of [Br pound]5 per week for a single person and [Br pound]8 per week for a married couple (in the first year after the election). Thereafter, pensions would be linked to average earnings, instead of price inflation, so that the cost of the commitment rises year by year. The assumption for child benefit is that it is uprated to the 1987 value in real terms. No other spending commitments are included in the table although the Labour Party has made many policy statements which imply substantial extra spending at some stage.

On the revenue side we show the estimated yield by the end of the year of the increases in tax and contribution rates to which firm commitments have been made. (No effect is shown for the cost or yield of changes in income tax bands as no commitments have been made as to their width.)

Initially these firm commitments actually raise more revenue than they spend, but this gives a misleading picture of the implication of Labour's policy statements as a whole for the balance of revenue and spending. There are many 'nearly-firm' commitments to spend in a whole range of areas, and few if any suggestions of additional means of raising revenue. Moreover, the cost of the commitment to raise pensions would, on its own, use up the yield of the tax increases within a few years. The declared intention is 'to spend only what the country can afford', but to give a higher priority to public spending increases than to cuts in rates of taxation. This would make the fulfillment of their spending aims dependent on the performance of the economy.

The exchange rate and fiscal policy

We assume here, as we do also in Chapter I describing 'existing policies', that EMU is the eventual destination, whatever party is in power. Nevertheless, the possibility of realignment remains under the present EMS arrangements (in Stage One), and a Labour Government would have to decide whether to make use of it.

There is no indication of Labour Party thinking on this issue at present. From our analysis in previous issues of the Review a case could be made out that the UK has entered the ERM at too high a rate : the UK has a current account deficit which must be corrected in the medium term by a fall in the real exchange rate; to bring that about by lower prices rather than by a lower nominal exchange rate requires our rate of inflation to be below the average of the rest of Europe; starting as we do with inflation above the average of the rest of Europe the period of transition must be very protracted; in the meantime the commitment to the exchange rate may not be believed. (Interest-rate differentials suggest that the market also expects sterling to depreciate at some time in the future.)

It does not follow, however, that devaluation is inevitable. Eventually the real economy will adjust to any level of the nominal exchange rate. Moreover, the fact of keeping the exchange rate fixed would, by reinforcing expectations of stability, make the adjustment to DM 2.95 quicker and less costly. For an incoming Labour Government to devalue on taking office would be to risk undermining the whole strategy based on EMS membership which is now supported, with more or less enthusiasm, by most informed opinion in this country.

With monetary policy dependent on EMS membership, any substantial policy measures to change the pressure of demand at home must involve changing the balance of taxation and public spending. But even in fiscal policy decisions the range of choices is not unlimited. The 'inter-temporal Budget constraint' requires that spending cannot be raised for an indefinite period of time without an increase in taxation to finance it. A temporary increase or decrease is permissible, however, leading to a higher or lower stock of debt at the end. Some economists would view this constraint as applying to the position of the public sector; our own practice is to concentrate rather on the position of the economy as a whole, reflected in the balance of payments and the stock of net foreign assets. The implications for fiscal policy are much the same either way.

For this reason the policy analysis shown in Annex A considers the timing of public spending increases. It takes it for granted that the end position must be one of external balance. In the short run however, a choice can be made between more or less inflation, less or more output. The analysis shows a trade-off or menu of choices, using the technique of policy optimisation' applied to the Institute's model.

If the exchange rate and interest rates are fixed, that is to say unchanged when fiscal policy is tightened or loosened, then the scope for varying inflation or the pressure of demand is not very great. The average growth rate of the economy over a five-year period and the average rate of inflation have to be much the same irrespective of the policy maker's priorities.

It is interesting therefore to see how the results differ if the policy maker is also allowed to change the exchange rate at the start of the period. A policy maker who cares only about reducing inflation will want the exchange rate to be realigned up, and the effect of that on the balance of payments will be offset by a tightening of fiscal policy. From this point of view the extra degree of freedom provided by realignment is clearly advantageous. A policy maker who is more concerned about the level of output and less concerned about inflation might want to take the opportunity to realign downwards. The examples in the Annex suggest however that the advantage to him of doing so would be relatively small.

There is an asymmetry here. The kind of policy maker who would want to appreciate the exchange rate in the circumstances predicted for the mid-1 990s has to be a policy maker who is so keen to achieve zero inflation quickly as to be indifferent to the output cost of doing so. Hence the opportunity to raise the exchange rate is an unmixed blessing for him. On the other hand the kind of policy maker who wants to depreciate the exchange rate by a similar amount would be one who cares about limiting inflation as well as sustaining the growth of output. Hence he will feel the results of his actions as a mixture of losses and gains.

If one adds that the danger of undermining the commitment to the EMS is greater for a depreciation than for an appreciation, then a case might be made that an incoming Labour Government, even if it wanted to raise output and employment, would not change the exchange-rate path substantially from that shown in Chapter 1. The implication is that demand management could play only a small part in its overall economic strategy.

Credit controls

In August 1988 we published a survey by Jon Shields of the range of measures which might be introduced to slow down the growth of credit to the personal sector. None of the measures was painless, but some stood a good chance of being effective. One possible approach would involve regulation of the terms on which mortgages could be granted, another was a form of tax on new credit extended. The Labour Party has expressed interest in a variety of different forms of credit control, some restricting borrowing, others restricting lending. The openness of financial markets limits the use that can be made of quantitative restrictions on the banking system of the kind familiar before 1980.

In 1988 the Institute took the view that the most appropriate way of moderating the growth of consumer spending was to raise income tax or national insurance contributions rather than to restrict the availability of credit. In the event taxes were not raised, and neither was credit restricted, with the result that demand grew much too rapidly for several years. Subsequently, the growth of credit has slowed down whilst the housing market has passed from boom to slump. Clearly the introduction of credit controls, particularly controls linked to the housing market, would not be appropriate at the present time. The situation could have changed however by 1991 or 1992. One of the benefits of full EMS membership will be a decline in UK interest rates. The housing market should recover (sooner or later) and the growth of consumer credit should re-accelerate.

Table 2 shows the effects of restricting the growth of personal sector credit, as compared with those of an increase in income tax to achieve a similar reduction in consumer spending in the medium term. As might be expected credit control acts more quickly and its effect is concentrated on consumer durables. Raising income tax probably adds to pressure for wage increases, whilst credit control probably does not. Hence the model simulations show a better medium-term outcome when consumer spending is cut by restricting credit, with a lower price level and less of a reduction in output. Concentrating the reduction on durables means that the benefit to the balance of payments is nevertheless greater.

The precise measures required to hold back the growth of credit, by about 10 per cent over a five-year period in the way shown in the simulation, have not been specified, but if they were similar to those used in the past they might be more acceptable as a temporary expedient than as a permanent restriction on consumer choice. Puffing the restrictions on would produce a reduction in spending as shown in the simulation; taking the restrictions off again might produce a corresponding surge which is not shown.

At present the regulation of consumer credit is not harmonised across EC countries, and regulation could be introduced in this country which could restrict borrowing irrespective of the nationality of the lender. Other EC countries however are in the process of liberalising their credit arrangements, so a tightening of controls here might be seen as a retrograde step. in the longer term, if the EC is moving towards a monetary union , there will be a need to bring the methods of monetary control into line. Relative to the present regime this could involve a tightening of control over the banking system in Britain. Increasingly, if EMU goes ahead, the control of credit will pass from the national monetary authorities to the new European central banking system.

PART TWO. MICROECONOMIC POLICY

Dissatisfaction with the performance of the UK economy is not just related to the excessive pressure of demand in the late 1980s and the subsequent slowdown in the growth rate. As well as the problems of demand management, requiring a reassessment of macroeconomic policy, there are underlying weaknesses of the supply side. These show up in a relatively slow growth in output over the medium term and in a level of unemployment which has remained high even at the peak of the cycle. The influences on the sustainable growth rate and on the NAIRU, or sustainable level of unemployment, can be examined using the framework of the Institute's model.

The sustainable level of output depends on technology, skills, business efficiency and the allocation of resources, as well as the quantity of labour and capital employed. The determination of the level of unemployment in the long run is less certain. In the Institute's model it is explained by the need to reconcile competing bids' for factor incomes. The NAIRU is the rate of unemployment just high enough to induce wage bargainers to set a real wage level that matches the income available for distribution. As such the NAIRU rises when tax rates are increased, or when the price of imported goods rises relative to home production.

The supply side of the Institute's model has been extensively developed in the last few years, making it much more useful for the analysis of microeconomic measures of the kind discussed in Labour Party policy statements. Even so it cannot take account of the detailed differences between sectors of the economy. Much of what we can say about policy options in this area still depends on 'off-model' calculations, or simply on considerations of economic theory.

This part of the chapter is in four sections: first we assess the long-run supply-side effects of increasing public spending instead of cutting taxation, assuming that the policy switch has no net effect on demand. We treat this as a microeconomic measure because it has implications for the supply side of the economy and the structure of output. Then we consider in turn policies to raise fixed investment and labour productivity. The model cannot tell us how to organise vocational training! But it can provide a framework for assessing the broad effects of raising output per head, with or without a corresponding increase in wages. Lastly we look at the effects of 'special employment measures', which could include vocational training, and which might serve to reduce the mismatch' between the kinds of labour in demand and the kinds of labour actually available.

Extra public spending and taxation

There are many examples of aspirations in Labour party statements to increase a wide range of public spending programmes. These fall into a variety of economic classifications: more grants and benefits which support the incomes and expenditure of (for example) pensioners; more aid to developing countries; more investment in the public and the private sector; higher wages for public servants; more spending on education and the health service; but perhaps less spending on defence.

These increases must be financed either by borrowing or by tax increases. It is possible that a distinction could be made between capital expenditure, which is financed by borrowing, and current expenditure, which is financed by taxation (or tax cuts foregone) - but since revenue is not 'hypothecated', this distinction is a difficult one to draw in practice.

Table 3 below shows the effects of an increase in current expenditure on goods and services financed by foregoing cuts in income tax (that is by a higher level of taxation than there would have been if expenditure had not been raised). The nominal exchange rate and the nominal interest rate are both fixed this being the most convenient assumption to make if the UK is moving towards membership of an EMU.

This combination of policy measures raises GDP in the short term, because the tax increases have only a delayed effect on consumer spending. For the same reason the initial effect is to make the balance of payments position worse. The point to emphasise however, is that in the long run the sustainable level of unemployment is not significantly reduced. The supply side of the institute's model now takes account of the incentive effects of income tax changes. This offsets the beneficial effect on unemployment of the changes in the composition of demand which follow from switching demand out of private consumption into public spending.

Fixed investment

Labour party policy statements emphasise the case for more investment in the private sector (as well as in the public sector). This will come about as a result of lower interest rates, which we would expect in the medium term as a result of ERM entry whichever party is in power. One of the distinctive Labour policies is the establishment of a new national Investment Bank, which would make additional funds available for businesses which now have difficulty in borrowing for fixed investment. Restrictions on takeover activity are also proposed within European rules', with a view to encouraging longer-run planning and discouraging distribution of dividends. More generally there seems to be a greater willingness than the present Government has shown to use regional policy, support for science and technology and support for small firms as a means of offering an implicit or explicit subsidy from public funds to firms who are trying to expand production.

The economic case for providing extra incentives of this kind is that firms (or their bankers and shareholders) are excessively short-sighted, or because the benefits of investment do not accrue only to the firms which pay the costs. it is very likely in practice that some of the benefit of extra investment goes to the employees of the firm in higher wages as well as to the shareholders in higher profits.

In the Institute's model higher investment has several beneficial effects for the economy as a whole: it raises the sustainable level of output, by progressively raising the capital stock; it eventually reduces the share of imports in domestic expenditure, although investment goods themselves are largely imported in the short run; it raises productivity and hence real wages as well as real profits. But to get these benefits in proportion it must be remembered that each year's investment is a relatively small proportion the capital stock. Thus a sustained 5 per cent increase in investment would raise the capital stock in manufacturing by about 1 1/2per cent after five years and 2 1/2per cent after ten years (see table 4).

The new machines which embody new technology will be more productive than the existing capital stock. The Institute's model seeks to capture this effect in a vintage production function'. Even so, the rise in productivity following a period of extra investment is, according to the model, no more than proportionate to the rise in capacity.

The model also predicts, based on past experience, that much of the demand for investment goods would be met by imports. With a fixed exchange rate, the consequence would be an increased deficit on the current account of the balance of payments. If domestic demand had to be cut back to preserve external balance the level of output and employment would be lower. In the longer term it is possible that higher investment would improve the balance of payments by enhancing the 'non-price' competitiveness of British industry.

The scope for giving direct assistance to UK industry to improve its international competitiveness is limited by the EC rules prohibiting protection against imports, or implicit export subsidy. The implementation of the single market programme will make this constraint on industrial policy increasingly effective. It will remain possible nevertheless to offer assistance to firms in certain regions, and inducements to encourage new investment. These regions include the whole of Scotland, Wales and Northern Ireland, as well as much of Northern England. Thus the emphasis on regional policy in recent Labour party policy statements may be an appropriate one in view of developments in Brussels.

It is impossible to make precise estimates of the difference that an active policy of stimulating investment in the private sector could make to the performance of the economy. As a broad order of magnitude the figures shown in the table above may be a reasonable guess. A 5 per cent increase in private investment is conceivable as a result of a policy change, although its public expenditure cost might be considerable. An increase of about 1 to 1 1/2 per cent in the sustainable level of output after five years of such a policy seems if anything a generous estimate of the benefits to be expected.

Training

In terms of standard welfare economics the case for incentives to increase expenditure on vocational training is more conclusive than that for encouraging fixed investment. The benefits to private sector firms of extra spending on training their employees are limited by the transferability of skills between firms. To the extent that individual workers are reluctant or unable to meet the cost of training themselves, there is a case for subsidy from public funds. As with fixed investment subsidy, however, there is no way of avoiding some deadweight loss' in the process of subsidisation.

The emphasis on the need for better education and training in the Labour party policy statements has received wide support. It is appropriate given the shortcomings of training in the UK identified in a number of studies undertaken at the Institute and elsewhere. It is not at all easy however to quantify the scale of improvement which could result from new policies or from extra expenditure.

The operative constraint on increasing formal education and vocational training for young people may be the shortage of teachers and lecturers. A recent estimate (4) is that raising the rate of participation of 16-18 year olds from 35 per cent to 85 per cent would cost about an extra [Br pound]1 billion a year in current expenditure, and require about an extra 50 thousand teachers and lecturers. It would clearly take a number of years to build up to such a high level of participation. The highest economic returns probably came from improving training for this age group, but additional resources could also be devoted to re-training workers already in the labour force. 'Job-specific' training or retraining of adult workers is relatively cheap in terms of teachers' pay, but relatively expensive in terms of hours of work lost to production.

As an order of magnitude, suppose that an extra [Br pound]3 billion a year in total were to be devoted to education and training over a five-year period (including direct expenditure and income foregone). The stock of investment in 'human capital' would then amount to [Br pound]15 billion. If the rate of return were as high as 15 per cent '51 the corresponding flow of benefit would be [Br pound]2 1/4 billion a year or about 1/2 per cent of GDP. This calculation may seem to take a generous view both of the likely scale of extra spending on training that could be induced over a five year period, and of its rate of return. Nevertheless the overall potential for raising economic performance by improved education and training could be under estimated. International comparisons suggest that a better deployment of existing resources could on its own improve the quality of the labour force substantially in the medium or long term.

The effects of better training would be to raise labour productivity and hence the sustainable level of output. But improvements in productivity do not necessarily reduce the NAIRU. if the sustainable level of unemployment reflects the process of wage bargaining, then the outcome depends crucially on the way in which productivity gains enter that process. If they are simply reflected one-for-one in wage settlements, then the process of reconciling claims to real income is in no way assisted, and the level of unemployment will not fall.(6)

In the Institute's model productivity improvements are normally reflected one-for-one in wages, so a shift to a higher level of productivity would not reduce unemployment even in the long run. This seems too pessimistic a view to take of the results of more or better vocational training. To the extent that employers are persuaded to meet the cost of training, the rise in productivity may be reflected in higher profit margins or better competitive performance.

A cautionary note is also needed about the effects of higher productivity on the external balance. Higher output, even if it is the result of supply-side improvements, requires a higher level of imports of materials, which have to be paid for by more exports. In the absence of improvements in non-price competitiveness' this can be achieved only by reducing export prices (compared with what they otherwise would have been). The consequence will be a further reluctance (or inability) of firms to raise wages, thus intensifying the conflict of aspirations in the labour market, tending to raise the sustainable level of unemployment. But, if as seems likely, non-price competitiveness is also improved by better training, then the macroeconomic effects will be more favourable.

A Labour Government would be heavily committed to encouraging extra effort both on training and on fixed investment. To the extent that it has to choose between the two, one advantage of training is that the immediate expenditure is almost all on domestic resources. Teachers can be produced at home, more readily than machine tools, it seems.

Special employment measures

A Labour Government might well take off ice at a time when unemployment is rising. The discussion so far suggests that neither their approach to macroeconomic policy, nor the structural changes they propose, will quickly reverse that trend. They will need therefore to consider which of the special employment measures now in place they wish to continue.

Increasing the numbers assisted in this way would reduce the claimant count; its effect on the labour market is more difficult to judge. in the institute's model the pressure of demand for labour is measured by the proportion of the labour force not working, irrespective of whether they are in receipt of benefit or not. Thus SEMs have no effect at all on wage pressure unless the participants are competing for jobs with other workers. The other economic effects of SEMs depend on the level of pay to participants. if this is no more than enough to replace the benefit they would have been entitled to, then the effect on household income is zero. In other words SEMs shift the measured level of unemployment without having any other significant effect on the economy in the short term. The longer-term benefits depend on the effectiveness of the training being given, which has been discussed in general terms above.

One aim of retraining the unemployed is to reduce the mismatch' in the supply and demand for particular skills. Existing training programmes however may be too short to have a major impact of this kind. (An effective craft' training programme would normally take at least a year to complete.) In estimating our model equation for earnings (7) we found that a mismatch in this sense tended to raise average real earnings at any given level of unemployment and productivity. The model can give some account therefore of the overall macroeconomic effects of improved matching. These effects, as one would expect, are small but generally beneficial (see table 5).

The simulation shows the consequences of reducing real earnings in this way by just 0. 1 per cent. The effect is to reduce both nominal wages and prices by a more substantial amount, about 1/2 per cent after five years. This improves competitiveness if the exchange rate is fixed. Output is gradually raised and unemployment is eventually reduced. This very gradual effect on unemployment might be accelerated if steps were also taken to reduce friction' in the placing of newly trained workers in appropriate jobs. Statutory minimum wage

Looking to the future' contains the following commitment. 'Labour will introduce a national legal minimum hourly wage, starting at a level of 50 per cent of the mid-point of men's earnings (the median). Currently, this is [Br pound]3.10 an hour.' This is explained as a move into line with European practice, but it is one which is unlikely to take place under a Conservative Government in the foreseeable future. The motivation is not mainly economic, although it is claimed that raising minimum wages will help to 'create a high skill, high value, economy'. The effects on inflation and unemployment are discussed in Annex B to this chapter.

The scale of the impact depends first on the enforcement of the legislation and then on the effect on the wage distribution. One of the groups mainly affected is part-time workers, for whom enforcement may be a problem, and whose pay may have relatively little influence on that of most full-time workers. If the scale of the impact was to raise average real earnings by 1 to 1 1/2 per cent, this would be associated with wages 3 to 5 per cent higher and prices 2 to 3 1/2 per cent higher after about two years (for a fixed exchange rate). Thus the inflationary cost of the proposal is not negligible. Output and employment in the medium term would be reduced by perhaps per cent. The effect on claimant unemployment would be very small, because the majority of those losing their jobs would not be eligible for benefit.

Narrowing the wage distribution by imposing a statutory minimum wage might encourage employers to offer more job-specific training, if they were confident that the staff would not leave when their training was complete. On the other hand the incentive to the individual worker to enhance his earning power by training would be reduced. Many of those whose wages would be increased most by the proposal would be young unskilled workers.

CONCLUSIONS

In the longer term supply-side measures of the kind proposed by the Labour Party, if they were efficiently implemented, could raise the sustainable level of output in the UK and might also bring about a lasting reduction in unemployment. As with most supply-side measures the time scale of their effect is necessarily long term. Raising the capital stock and improving the skills of the labour force both take time, because only a small fraction of the total is renewed each year.

The process is also an expensive one, not least because it is impossible in practice to ensure that the implicit or explicit subsidy is directed only to marginal decision makers. If a Labour Government wishes to divert resources to investment and training on a large scale, then it will not also be able to afford for some time all it would like to spend in other areas of public expenditure.

Spending more on investment and training could eventually tackle some of the underlying weaknesses of the UK economy. In the meantime a Labour Government (or any other Government) would face the unpleasant choice of more or less inflation, less or more unemployment. That choice can be exercised by delaying or bringing forward the tightening of fiscal policy to correct the balance of payments deficit. So long as the exchange rate is not 'irrevocably' fixed the option of devaluation cannot be ruled out, but to exercise it would be to risk undermining the credibility of the Government's commitment to financial rectitude'.

The assumption of this chapter is that a Labour Government would opt for a strategy of maximising the advantages of association with the rest of Europe; we have stressed the advantage of that approach, whilst recognising the way it narrows the range of policy choice.

The upshot is that the economic policy differences between the two major parties are narrower now than they have been for about twenty years. That in itself is not unwelcome. The extent of the political differences over economic policy in Britain in the 1970s and 1980s has been a source of weakness rather than strength.

NOTES

(1) A policy document was issued by the Labour Party earlier this year with the title 'Looking to the future'.

(2) The model simulations were carried out by Bob Anderton, Paul Gregg, and Soterios Soteri.

(3) This table is based on that prepared by Adrian James of NatWest (18th May 1990) but only firm commitments to spend are included: We have uprated the tax estimates given in that paper to a 1991 tax base, but kept the levels of benefit unchanged (since the commitment is to a fixed sum in [Br pound] per week). The proposal to make unearned income subject to National Insurance Contributions is included in the table as one of Labour's firm commitments'. It should be recognised however that the taxation of investment income may require harmonisation as capital markets become more closely integrated.

(4) Finegold et aL A British 'Bacca/aureat' IPPR Education and Training Paper no.1.

(5) A relevant study is G. Psacharopoulos and R. Layard (1979), Human capital and earnings: British evidence and a critique', Review of Economic Studies, vol. 46 pp. 485-503. This was based on a cross-sectional analysis of a sample of 6,783 men aged 16-64 from the 1972 General Household Survey. From an estimated earnings function, they computed private rates of return to years of schooling, to (post-school) training' (including simple on-the-job learning) and to specific qualifications (0-levels and A-levels). From the way in which rates of return have been computed, the social rate of return (as customarily defined in the human capital literature) would necessarily be lower, but not by much, since the difference is the cost of tuition, which is usually found to be a small proportion of the total.

Their main findings were as follows:

(a) The rate of return to 1 or more A levels (over 5 or more 0 levels) was between 9 and 12 per cent. (Unfortunately, no calculations were made or were possible for the more vocational qualifications.)

(b) The rate of return to post-school training rises with the number of years of schooling: someone with 1 1 years of schooling had a 23 per cent rate of return to training, while someone with 13 years of schooling had a return of 33 per cent.

These results refer of course to a period nearly 20 years ago. But they do give some support to a high estimate of the social rate of return, especially if one takes account of the second finding-high levels of formal qualification make it worthwhile offering on-the-job training.

(6) The point can be demonstrated algebraically. Suppose that the determination of nominal earnings is as follows:

w = p + [alpha]g - BU

where w, p and g are the growth rate of earnings, prices and productivity respectively and U is the level of unemployment.

Suppose also that prices are determined as follows:

p = a (w-g) + (1-a) m

where m is the growth rate of import prices. Substituting the second equation with the first, and rearranging terms, will give an equation for the sustainable level of unemployment.

BU = ([alpha]- 1)g + (m-p)

If productivity improvements are reflected one-for-one in earnings, [alpha] = 1, they do not also reduce the sustainable level of unemployment.

(7) See Moghadam and Wren-Lewis, (1990), Are wages forward looking?', National Institute Discussion Paper no. 163.

ANNEX A: EVALUATING MACROECONOMIC POLICY OPTIONS by Peter Westaway

This annex examines a menu of alternative macroeconomic policy options that might be available to an incoming Government elected in Autumn 1991. The starting point for the analysis is an updated version of the forecast outcome as described in the August Review which assumes the continuation of currently announced policies by the Conservative administration. Any policy stance will reflect the relative priorities of policy makers with regard to different objectives of policy. There will be an implicit trade-off between say inflation and output growth even for a Government as avowedly anti-inflationary as the present one. The techniques of policy optimisation allow us to make rough calculations as to the nature of this trade-off. But more interestingly in the context of an incoming Government, it is possible to illustrate how fiscal and monetary policy might be changed in order to achieve a different balance between inflation and output. This annex investigates that question under a number of different assumptions.

In fact, the analysis to be adopted here can equally well apply to the choices facing a Conservative administration as to a new Labour or indeed any other complexion of Government. However, as noted in the main text, it is particularly relevant to the analysis of the room for manoeuvre of an incoming Government since they are likely to consider items on the menu previously ignored. Of course realistically, the macroeconomic policy environment facing different administrations may not be the same even if the Labour Party's intended fiscal and monetary stance was identical to that of the outgoing Conservative administration. First, it might take time for the new Government's policy intentions to be fully believed by the private sector. For example, if their anti-inflationary resolve were believed to be weaker than that of the previous administration, there would be a temporary period of weakness of the exchange rate until the credibility of the new regime were established. Alternatively, this uncertainty might manifest itself in a higher risk premium on sterling assets which would similarly cause an initial fall in the exchange rate on unchanged policies. In the analysis described below we do not explicitly consider the credibility of different Governments although in principle it is possible to do so.

Three broad scenarios are considered. First, we consider the options available if the new Government wishes to manipulate fiscal policy in such a way that the impact on the exchange rate is neutral. This would seem a plausible assumption if the incoming Government decides that the prevailing level of the exchange rate should be held because to do otherwise would be too costly on credibility grounds. However, as a second scenario, it is possible that the foreign exchange markets might allow a new Government a single realignment without necessarily assuming that the commitment to the EMS has collapsed. In this case, the Government would effectively gain an extra policy instrument in the form of the 'entry rate'. Of course, if this realignment is anticipated then it will already have affected the level of exchange rate. We ignore this issue here by assuming that any realignment is unanticipated. Finally, we consider the freedom of scope that an incoming Government might have in manipulating interest rates. In an eventual monetary union, of course, no such freedom would be available, but over the short term while UK inflation is still converging towards that of other EMS members, there is still some degree of flexibility. Therefore, in the third scenario, we examine the implications of shifting the mix between fiscal and monetary policy.

To illustrate how changes in Governments' priorities can affect the optimal policy stance we must quantify the policymaker's objectives. We assume that the Government sets policy with regard to two final objectives, price inflation (on a CPI basis) and output. Policy makers are assumed to attach a cost to any inflation outcome away from zero and any level of GDP which falls below the long-run sustainable level, assumed to grow at 2 1/2percent per annum.(1) There are also constraints on the extent to which fiscal policy can be altered. In all the examples shown, we assume that Government expenditure rather than taxes is varied.(2) A cost is attached to large changes in Government expenditure relative to base levels. There is also an important constraint that the long-run growth of Government spending must be consistent with GDP growth in order to preserve a constant equilibrium share of Government spending in output. Thus, in our analysis, Government spending is only varied over the first 8 years before returning to its equilibrium growth path. Of course, this does not mean that the long-run ratio of Government spending to GDP cannot be altered. For given monetary policy any short-term increase in Government spending must eventually be offset by a long-run fall to finance the extra debt interest payments.

Formally, we can characterise these Government policy priorities in terms of an objective function or cost function whose value policy makers will attempt to minimise. i.e.

[FORMULA OMITTED]

where p is inflation, GDP and GDP* are the actual and desired levels of GDP and AG is the change in government spending, a, b and c are the weights which reflect Governments' relative priorities.

The cost function used here assumes that policymakers are far-sighted and compute the implications of their policies through to the next century, discounting future costs at a rate of 1 per cent per quarter (so the government values 1 per cent on the inflation rate today the same as 2 per cent in 2005). Of course, this assumption may be bestowing too great a degree of forward-lookingness on the part of politicians whose true policy horizons may be the date of the next election; this more myopic type of behaviour could be captured in this framework by evaluating costs over the near future only, or by adopting a much higher discount rate. This route is not followed here. (3)

It is important to notice the absence in the cost function of certain policy variables which might be considered as policy targets, for example the exchange rate, the current account of the balance of payments or the PSBR. We argue that these ought to be excluded on the grounds that they are intermediate targets which only matter to policy makers because of their implications for the final targets, ie inflation and output. This is not to say, of course, that the government can afford to be indifferent about these outcomes since the long run inter-temporal budget constraint of the government and the country as a whole acts as a crucial constraint on the policy process. (4)

We start by reviewing the base forecast. Interest rates are assumed to fall gradually from their current level of 14 per cent to 12 per cent in 91 Q4 at the time of the election and to 7 per cent by 1996. By that time a long run differential of 1 per cent is assumed to be held with European rates which, assuming a 1 per cent risk premium on sterling ensures that the exchange rate stays at a constant level relative to the DM. Tax rates are set at their current levels while government expenditure is assumed to grow broadly in line with GDP. Given this policy stance, price inflation (as measured by the consumer price index) falls from current levels above 6 1/2 to around 1 per cent a year by 1996; this fall in inflation is slightly larger than that in our current forecast. GDP growth gradually recovers from around 1 per cent in 1991 to reach 2 1/2 per cent a year by 1995. in terms of the underlying cost function, the forecast policy stance is broadly optimal if the weighting parameters on inflation and GDP are set to 1 and 1/2 respectively. in all the charts, the base run will be referred to by the mnemonic 'B'.

Now consider the implications of varying fiscal policy, first under the assumption that policymakers place no weight on GDP performance (weights of 1 and 0), second under the assumption that there is a higher weight on GDP (weights of 1 and 2). These alternatives will be referred to by the mnemonics 'P' and PG' respectively. Charts 1 and 2 show the different outcomes for inflation and GDP growth while chart 3 shows the required profiles for government spending required to achieve these. With a larger penalty attached to inflation, it is possible to achieve moderate reductions in the rate of inflation, by up to 1 per cent in 1992-3. However, this necessitates a severe cutback in government spending of nearly [Br pound] 10 billion over the first five years which results in GDP levels falling by around 0.5 per cent during the first two years. In the 'PG' variant, with a much higher weight on output gains, these relatively unspectacular effects on policy objectives are reversed. Slightly higher growth in the first two years is achieved by a [Br pound]2 billion per annum increase in government expenditure which is sustained until 1998 but this results in a moderate deterioration in inflation which remains higher until 1996.

So far we have seen that if any new government is bound by the exchange rate commitments of its predecessor then there is relatively little scope for radically changing the balance between inflation and GDP growth. in fact, this conclusion may be seen as an attractive one by an incoming Labour Government since it would give them some short term scope for expenditure increases without precipitating too great a surge in inflation. Nevertheless, it is still of interest to examine scenarios where more leverage on policy outcomes is possible. How much scope is available to policy makers improving the balance between inflation and output if they also decide to re-align the exchange rate upon election? We assume here that if this were done it would be unexpected and that the exchange rate would subsequently move in the same manner as before, that is, consistent with the risk adjusted interest differential. First consider the actions of a Government placing a high cost on inflation ('EP'). Because an appreciation in the exchange rate has short term beneficial effects on inflation mainly because of the fall in import prices, the optimal response is to engineer a higher exchange rate. The example given shows an 8 per cent appreciation (see EEP in chart 4) which is achieved by a gradual but eventually severe tightening of fiscal policy (see QEP in chart 5) with Government spending some 3 per cent lower in the long run. This duly achieves a sharp fall in the inflation rate which falls to zero by the end of 1992 (chart 6). The cost in terms of lower short-term output growth is significant (see chart 7), with GDP growth on average 0.5 per cent below base for the first two years although by 1996 GDP levels have returned to base levels.

As before, we now consider the alternative policy preference where the weight on GDP is raised (variant 'EPG'). This time, the exchange rate depreciates by some 6 per cent and government spending is increased by an average of [Br pound]1 billion a year over the first five years and importantly is higher as a proportion of GDP even in the longer run (see QEPG in chart 5). The short term effect on inflation is a sharp short term deterioration to nearly 7 per cent by the end of 1992, staying higher until 1998. GDP growth temporarily reaches 3 per cent by mid 1992 and remains slightly higher until 1999.

Overall it is clear that the ability of policymakers to manipulate the effective entry level of the nominal exchange rate within a fixed exchange rate system like the EMS represents a powerful degree of freedom. The examples shown illustrate clearly that, in contrast to the fixed exchange rate cases, policymakers are allowed a fairly large amount of discretion with regard to the short term output inflation trade-off. Of course, this 'freedom' to alter the exchange rate crucially depends on the willingness of the authorites to validate their intentions by the appropriate use of fiscal policy. It also depends on the willingness of the private sector, in particular the foreign exchange markets, to believe these policy announcements.

Tables 1 and 2 emphasise this point by quantifying the 'welfare gain' available to policymakers under the two regimes. Table 1 shows the inflation costs incurred by the purely anti-inflationary policymaker while table 2 shows the outcomes when GDP costs are also taken into account. The total inflation cost represents the discounted cumulated deviation of inflation from zero while the GDP cost is the discounted cumulated deviation of GDP levels from a 2/2 per cent growth path.(5) Table 1 shows that there is a 100 point improvement in cost when the exchange rate is free to vary compared to a much more modest 24 in the fixed exchange rate case. When GDP costs are also scored, table 2 shows an improvement of 134 when a re-alignment is allowed compared to only 81 points when the exchange rate is completely constrained. Of course some care should be taken in comparing the units of tables 1 and 2 but interestingly the welfare improvements available to the policymaker who also values GDP are much less clear cut; any gain in GDP can only be bought at the expense of higher inflation. It is the relative weightings on inflation and GDP which determine how much of this cost is worth paying.

So far, it has been assumed that monetary policy is taken as given by the incoming Government. However, a new regime might wish to alter the fiscal-monetary mix by having a lower level of interest rates on the grounds that the use of high interest rates as a weapon of demand management may have undesirable long term supply-side implications. In principle, the level of interest rates can also be chosen optimally as part of the overall policy design problem. However, we consider a simpler alternative where interest rates are brought into line with German levels more quickly so that they converge to their long run differential of 1 per cent by 1994. For unchanged fiscal policy, this causes the exchange rate to fall immediately by 8.5 per cent, but subsequently it depreciates less because of the lower differential (see chart 8). This fall can only be mitigated by tightening fiscal policy. The optimal response of the Government who attaches a high weight to inflation 'REP') is to tighten fiscal policy sufficiently so that the weakening in the exchange rate is not only avoided, but there is a small appreciation of 3 per cent. This in combination with the subsequent higher level of the exchange rate compared to the base case causes a substantial improvement in inflation along similar lines to the 'EP' case. However, as chart 9 shows, the required fiscal tightening is implausibly large. Indeed, the magnitude of this required fiscal tightening to offset looser monetary policy clearly explains why the current Government can not bring interest rates down to German levels immediately and hence why they must accept a transient period of depreciation against the D-Mark (see Wren-Lewis et al, (1990)). The contrast with the response of a Government who places a higher weight on GDP 'REPG') is predictable. Most of the ex ante depreciation is allowed to occur at the expense of much higher short run inflation and lower output.

Has either type of policy maker benefitted from lower interest rates? With an inflation target only, the inflation profile is very similar to the 'EP' case while the cost is only slightly worse. But this has only been achieved by a very severe cutback in Government expenditure, even compared to the large cutbacks required in the high interest-rate case. A policy of high short-term interest rates would therefore seem to be more desirable for this policy maker. Turning to the Government who also worries about GDP growth the situation is less clear. Now the overall outcome for GDP is less favourable than in the 'EPG' case, this despite a larger surge in growth in the first year. Inflation falls more quickly in this case however but overall the total welfare gain is more akin to the rather modest improvement achieved when the exchange rate was fixed. Again we might conclude therefore that there are relatively few advantages to cutting interest rates prematurely.

NOTES

(1) For a discussion of the importance of the difference between a GDP growth target and GDP level target used here, see Westaway and Wren-Lewis, (1990).

(2) This reliance on expenditure rather than tax changes is done for illustrative purposes only.

(3) The calculation of policy outcomes over a long horizon is also necessary because of the forward-looking nature of the National Institute model which requires solving to a terminal date which is sufficiently far in the future.

(4) We also ignore the possibility that there may be political reasons for attaching a cost to the current account deficit, or indeed that there may be costs or benefits not captured by the model which are associated with Government spending.

(5) GDP is calibrated in units of GDP growth per quarter. The desired level of GDP is set at 2 percent above the 2 1/2 percent growth path to avoid penalising growth above the long run rate.

REFERENCES

Westaway, P.F. and Wren-Lewis, S., (1990), 'Is there a case for the MTFS?', CEPR Discussion paper No. 41 1.

Wren-Lewis, S., Westaway, P. F., Barrell, R. and Soteri, S. (1990), Choosing the rate; an analysis of the optimum level of entry for sterling into the ERM', National Institute Discussion Paper No. 171.

ANNEX B: A NATIONAL MINIMUM WAGE by Paul Gregg

One of the most precise policy commitment of the Labour Party is to the introduction of a National Minimum Wage (NMW). The twin justifications given for a NMW are a desire to combat the perceived inequitable distribution of pay in Britain and a desire to reduce low skill employment.

That it will have short-term economic costs, higher inflation and higher unemployment is widely accepted. These costs will be minimised where the workers effected are isolated and facing a monopsonistic employer for then there will be little knock-on effects to other workers and few jobs lost. The costs will be greatest where a NMW effects workers in large organisations with rigid pay structures who are in an industry with a high elasticity of demand for labour.

Table B1 shows those affected by a NMW at [Br pound]3-10 per hour in 1990 and by how much their earnings would rise on average. We have excluded workers not on adult rates from this analysis, as young workers and those in training are likely to be exempt from coverage of a NMW. The table shows that over 15 per cent of employees would be affected, or a little over 3.5 million employees. Their pay would rise by nearly 15 per cent on average, giving a wage gap between current pay levels and those consistent with NMW of 1.0 per cent of the total wage bill. This is smaller than the gap calculated by Bazen (1990) primarily because we have compared the level of [Br pound]3.10 per hour with 1990 NES data, rather than 1989, in line with Labour's commitment. Table B1 also shows how the NMW largely effects women and particularly part-time women.

Table B2 shows a breakdown of the proportions of male and female full timers on adult rates across industry, region and age group. Similar calculations, for part timers are made difficult by the small samples for many categories within NES. What Table B2 shows is the concentration of wage earners with pay below 23-1 0 per hour in certain industries, (agriculture etc, distribution and other manufacturing), in all regions except Greater London and the South East and amongst younger workers.

Economic effects of a NMW

Having calculated the number of people affected by a NMW on the lines of the Labour Party's proposals and by how much pay will rise among this group to give a wage bill cost of the proposals, we can attempt to estimate the economic effects of such a policy shift. There are number of additional issues that need to be considered.

To what extent do workers just above the minimum rate try to restore their differentials with those receiving earnings increases? in the 1980s the dispersion of earnings has widened dramatically, having been more or less constant for the whole of this century (see Table B3), with the lowest decile of Manual males having suffered over a 5 per cent relative earnings decline with respect to the Median wage since 1980. This widening of relative wages shows that flexibility exists in the system and that a NMW could restore some of this relative decline. Neuburger (1984) estimates that the earnings distribution can be crushed with a 'knock-on' effect of 60 per cent of the original wage bill costs of a minimum wage.

It seems plausible that the restoration of differentials will occur most where workers are engaged in similar functions and work in the same environment. If so, manual male differentials will be little affected by the wage changes of female part-timers but heavily so by those of other male manuals. This would imply that the knock-on effect would be positively related to the level of the NMW, as more workers would be covered across more groupings. If this is the case then Labour's proposals to raise the NMW progressively over the years to % of male median earnings will have greater knock-on effects than the initial introductory rate.

If Neuburger's estimates are accepted it would imply that the total wage shock from the introduction of NMW would be a 1.6 per cent rise in the wage bill. Brosnan & Wilkinson argue however that because of concentration of low pay in highly segmented labour markets the knock-on effects will be minimal. We therefore show below the effects of two wage bill shocks. Table B4A has just the 1.0 per cent in the first year, whilst table B4B has an additional 0.6 per cent knock-on in the second year.

The effect on the public sector wage bill is likely to be somewhat less than proportionate, as the wage distribution in the public services is less dispersed than the private sector (the lowest paid decile of full-time adult public sector workers earned less than [Br pound] 139.40 in 1990 relative to [Br pound] 123.60 in the private sector), although large numbers of workers would be effected in certain public services especially the NHS. This implies that the public sector wage bill would only undergo 80 per cent of the shock of the whole economy or 1.3 per cent of the wage bill. A NMW will also create benefit savings on payments to those on low wages. This will largely be from Family Credit which cost [Br pound]0.5 billion in 1988/9, and from Housing Benefit which cost 23.9 billion in 1988/9.

The NMW will mainly effect females and particularly part-time females. Hence jobs lost will disproportionately be from these groups and there may also be some substitution of male/full-time jobs for female/part-timers. This will reduce the effect of the NMW on the claimant unemployed count, as women part-timers are less likely to be able to sign on (see Gregg 1990).

The results presented below assume that enforcement is 100 per cent successful for those not in the black economy. This is optimistic for a number of reasons. Firstly the evidence of wages councils rate enforcement is not very encouraging though a NMW may prove easier to implement. Secondly employers may force workers to adopt self-employment status, which would make enforcement difficult. Thirdly the requirement to pay higher wages in the legitimate economy might encourage increased activity in the black economy.

Table B4A shows the economic impact of a 1.0 per cent shock to the wage bill. The shock was introduced into the cointegrating vector of the average earnings equation in the National Institute Model 11.4. The policy move is inflationary in the first three years adding 1 per cent to inflation at its peak in year 2. Part of the attempted rise in real wages is offset by the loses of employment and the general downward pressure on wages this produces. Thus the policy initially stimulates activity because of a general rise in real spending power, but the resulting inflation produces a loss of competitiveness. The fall in real profitability causes firms to destock and reduce investment as their financial situation deteriorates.

These results differ from those of the Bazen article for three reasons. Firstly he uses a higher wage bill cost because of definitional differences discussed earlier. Secondly he used an older version of the Institute's model. Thirdly and most importantly he applies only a nominal wage shock, so that the real wage gain is only 0.63 per cent, the rest being eroded by higher prices. This implies a real pay cut for those not effected by the NMW and implies also that the NMW is not uprated in line with that higher inflation in subsequent years.

Table B4B allows for secondary knock-on effects equivalent to 60 per cent of the initial wage bill shock (i.e. a 1.6 per cent total wage bill shock). The results are broadly in proportion with those in table B4A except for the output consequences which are somewhat less than proportional and the inflation effects which are slightly more than proportional.

So even with a 60 per cent knock-on effect the output costs of a minimum wage after five years are only of the order of 1/2 per cent and the effects on prices around 3 1/2 per cent over three years.

REFERENCES

Bazen, S. (1990), 'On the employment effects of introducing a national minimum wage', British Journal of Industrial Relations, no. 28:2, July.

Brosnan, P. and Wilkinson, F. (1988), A national, minimum wage and economic efficiency', Contributions to Political Economy, 7:1-48.

Gregg, P. (1990), 'Out for the count-a new approach to modelling claimant unemployment', National institute Discussion Paperno.167.

Neuberger, H. (1984), From the Dole Queue to the Sweatshop, London, Low Pay Unit.