A refreshing feature of the regular Medium-Term Reviews undertaken by the Economic and Social Research Institute has always been a willingness to review previous projections and to analyse why from time to time events turned out differently from what had been forecast.

All six previous reviews since the inception of this series in 1986 were criticised at the time of publication for being over-optimistic. But instead, all but one of them proved in due course to have been to be too pessimistic.

The exception was the 1989 review: understandably, it failed to foresee the collapse of Berlin Wall and German re-unification, which set back Europe's growth for three years. Even in this instance, although our early 1990s growth rate dropped below that projected by the ESRI, the institute's employment forecast was once again exceeded.

Is it possible that the drop of one-quarter in our growth rate that is now projected for the period 2000-2005, and the further drop of over one-fifth projected for the following five years, may again prove over-pessimistic? This is, of course, possible but because of the labour supply constraint that now faces us, it is less likely than in the past. Previous growth projections proved too low because the demand for Irish labour in Ireland was under-estimated. But, throughout all the review periods from 1986 to 1999 this higher-than-expected demand was capable of being met by flexibility in our labour supply.

By contrast, even if in the years ahead the demand for Irish labour in Ireland were to continue at the recent very high level - and rising labour costs make this unlikely - from now on we will find it very difficult to supply from our own domestic resources the additional labour that would be needed to raise growth.

The institute estimates that, in order to achieve a 1 per cent higher growth rate than it is forecasting, net immigration during the years ahead would probably have to rise from a projected annual inflow of 18,000 to 33,000 a year - and most of these additional immigrants would have no previous connection with Ireland.

That would involve building another 80,000 dwellings in addition to the 225,000 dwellings required by the ESRI's central forecast for the 2000-2005 period. It is hard to see why we would want to undertake such an additional burden, generating further inflationary pressures for the sake of a higher growth rate.

Could these projections prove, on the other hand, to be over-optimistic - as was the case in 1989? Yes, in certain circumstances they could - if, for example, we experienced an external shock similar to German re-unification or the oil crises of 1973 and 1979. Or alternatively, if we were to go completely overboard on pay increases and tax cuts.

Possible external shocks, the implications of which are considered in some detail in the report, include a sudden two-point rise in euro-zone interest rates, or a US stock market collapse knocking 25 per cent off share values. The report suggests that either of these events could reduce our GNP by three percentage points below its projected level, and that a recovery thereafter to or above the longer-term growth rate would take three to four years. An important factor in such a GNP drop would be the potential vulnerability of our housing market to shocks of this kind. Because housing demand is running ahead of housing supply, pushing prices up to a level above that in major cities of our principal trading partners, some house-buyers are trying to beat inflation by buying earlier than they might otherwise have done.

The ESRI believes that even a modest dip in prices because of an unexpected external shock could reverse this buyer psychology, leading some people instead to postpone house purchase in the hope of future price falls. This is the so-called "bubble effect".

However, encouragingly, the ESRI points out that the "bubble effect" here would be much less severe than it was in the US MidWest in the mid-1970s, Britain in the late 1980s and Scandinavia in the early 1990s. Because of EMU membership, there would be no knock-on effect on Irish interest rates; our banking system is not unduly exposed to the housing market; the large part of our economy serving export markets would not be affected; and our economy is currently well-placed to cope with such a crisis because of its competitiveness.

Overall, the institute believes that the most likely outcome for the housing market is a "soft landing", with house prices stabilising, and eventually falling slowly in real terms.

However, the ESRI does express concern lest pay increases in the forthcoming round, even if sustainable in the short-run, could eventually make the economy uncompetitive. And it agrees with Professor Paddy Geary of NUI Maynooth that it might be better if some private sector pay increases were taken as part of a profit-sharing or gain-sharing arrangement.

Nevertheless the ESRI accepts that "some rise in wage inflation compared to the last agreement is warranted", and the Central Forecast upon which its growth projections are based provides for very much higher future wage rate increases than those of the past six years: increases at an annual rate of 6 per cent, as against the recent rate of 3.5 per cent a year.

This 6 per cent figure allows for the impact of the educational upgrading of the labour force as each year a large new generation, of whom three-fifths have had experience of higher education, replace a much smaller number of retiring workers, two-thirds of whom never got beyond primary school. This educational upgrading factor adds about 1 per cent to the total annual wage bill - and, of course, even more to labour productivity - so the 6 per cent projection for wage rates implies an increase of about 5 per cent a year for those currently in the labour force.

The increase in the after-tax income of workers would, however, be very much greater than this. Between 1993 and 1999 the average industrial wage rose by 22 per cent but, because of tax cuts, after-tax income increased by 37 per cent. And this review provides for a cut of 15 per cent in personal taxation over the next dozen years - a reduction of a magnitude we have never previously experienced.

The ESRI also projects that even with an economic growth rate almost one-third lower than that which we have recently been experiencing, it will be possible to raise average living standards during the next six years by an amount very close to that experienced since 1993 - viz by 26 per cent as against 29 per cent.

This is possible because, with our economy now in good shape, we can afford to raise the labour share of GNP, which fell sharply during the 1990s as the profit share rose from an inadequate to a more-than-comfortable level. In suggesting how the additional resources accruing during the next six years might be used, the ESRI seems to have set itself a number of targets.

First, it proposes that the Budget surplus should be raised over the next couple of years to £3 billion, then held at that figure until 2006, and over the following decade gradually reduced to nil. In this way, our national debt would be repaid within the next 10 years and a nest egg would be built up thereafter in order to give us the leeway to meet any future crises. This is an ambitious target - some might feel over-ambitious - which has few parallels elsewhere.

Next, the review proposes that, because of our huge infrastructural deficit, the share of GNP devoted to investment be held at its present very high level of over 25 per cent for some years to come. Few will quarrel with that.

In the review, government expenditure is projected to rise by just over 6 per cent a year in current money terms - a similar figure to that of the past six years. However, when debt interest is excluded, the growth in the residue of current spending in the past six years has averaged 7.4 per cent a year - and a similar increase is projected for the six year ahead.

However public service pay, which absorbs over two-fifths of current spending other than debt interest, is slated to rise by an annual 9.5 per cent - comprising a 2.5 per cent increase in public service numbers, and pay increases of just under 7 per cent. Consequently, the provision made for social welfare is much lower - around 5.9 per cent a year. This would enable social welfare payments to rise in line with wages, which has not been the case here in recent years.

Taking all these factors into account, overall Government day-to-day spending will still continue to fall as a percentage of GNP, since GNP growth remains strong. Since 1993, public sector spending has fallen from 44.5 per cent of GNP to less than 36 per cent, and the review proposals would reduce it further to 31 per cent by the year 2005. Tax revenue would absorb a similarly reduced proportion of GNP - down from 41 per cent to 36 per cent. There will be some who will argue for lower tax cuts and more spending on social services.

At a time when wages are about to receive a large and overdue boost, adding to the economic stimulus of the recent sharp fall in interest rates, the review is insistent on the need to postpone significant tax cuts for two years to avoid further overheating of the economy and to consolidate the budget surplus.

With two years to go before the next election, this could make political as well as economic sense. In practice, however, even with generous pay provisions, it may be difficult to avoid agreeing tax concessions as part of the forthcoming pay deal. And there lies a real danger for an economy which is already facing growing inflationary pressures. The review raises once again the question of broadening the tax base in ways that would make economic sense as well as facilitating the transfer of many over-taxed workers from the top 46 per cent marginal tax rate to the 24 per cent band. These proposals include carbon taxes, which our Government has been unwisely opposing at EU level, and road pricing to reduce traffic congestion at peak times. A change of heart on these issues is long overdue.

What this Review provides is a context for policy-formulation. Its suggested allocation of resources between tax cuts, current and capital spending and debt reduction simply provides a starting point for political decision-making by both Government and Opposition. Each party will have its own views as to how best to deploy additional resources, but in planning ahead, the parties would be unwise to try to move outside the parameters here provided, which represent a fair assessment of the scale of resources likely to be available during the years ahead.

This review was produced under certain constraints so far as statistical data is concerned. At the time the work on the review model was completed, the 1998 National Accounts, which are compiled on a new and more comprehensive basis (ESA95) now employed throughout the European Union, were not available to its authors, and in any event are not available far enough back to provide a valid model of the economy. The Review projections were thus based on the older ESA79 basis that was previously employed. And because the new basis is more comprehensive, both the level of GNP today and its growth since 1991 are now known to be significantly higher than in the data available to the ESRI in preparing the Review.

Moreover, a similar problem of lack of earlier data on the ILO method of assessing employment and unemployment, and discontinuity arising from the recent substitution of quarterly for annual Labour Surveys, have also made it necessary for the ESRI to use the earlier Principal Economic Status method, which shows unemployment at a higher level than the ILO method.