Unemployment continues to fall; manufacturing output has accelerated; the latest CBI survey reports renewed optimism: clearly the economy is going through a period of reasonably rapid growth. The reasons for this better performance are no doubt manifold, but amongst them we would emphasise the fall in the sterling exchange rate which took place last year, especially against the Deutschemark.

In the ‘overall outlook’ section we focus on the influence of the Group of Five (G5) ‘accord’ on exchange-rate prospects, and compare the expected impact of an appreciating currency on growth prospects in Japan and Germany. The short-term forecast is then discussed in more detail in sections covering the six major economies, commodity prices and trade. In the second half of the chapter we examine economic prospects in the world economy over the next five years, focusing particularly on the outlook for current balances. In our central medium-term forecast exchange rates move sufficiently to bring about a large, if not total, reduction in the present level of trade imbalance. Our forecast for a substantial fall in the US deficit is unusual, and we examine the key issues involved as well as looking at associated changes in the Federal budget deficit. Finally, we present two simulations which represent alternative routes to removing current account disequilibrium. The first involves substantial, and as yet unannounced, fiscal reflation in Japan and Germany, which is sufficient to broadly maintain current exchange-rate parities. By contrast, a second simulation illustrates a collapse in the dollar, and the consequences of exchange-rate overshooting.

Reliable measures of the gross replacement cost of fixed assets are needed for a variety of reasons. It is generally accepted that this is the valuation concept which is most appropriate for assessing the output capacity of capital and for analysing developments over time in output, labour force and capital stock relationships. The gross replacement cost of fixed assets is also the benchmark for deriving measures of net capital stock which, taken in conjunction with other capital elements and profit data, yield rates of return to capital. Furthermore it is the concept which constitutes the fundamental ingredient in the quantification of the depreciation allowances that are used in national accounts.

Better pre-planning of work is a familiar requirement for improved productive efficiency; this has manifested itself for some decades in a tendency for an increasing proportion of the workforce to be transferred from physical production on the factory floor to ‘paper work’ in offices—that is, to administrative and clerical activities. That tendency has been offset, especially in the past decade, by advances in data processing machinery allowing clerical activities to be performed very much more quickly and by fewer staff. The total number of employees required for administrative and clerical tasks has fluctuated according to the relative strength of these tendencies. The required quality, or technical capabilities, of such staff has also fluctuated: for 100 years or more, clerical and office staff were usually required to follow formal procedures on direct instructions from a superior; much work was of a purely routine and repetitive nature. The first phase of automation using mainframe computers and central processing facilities did little to alter this situation, and may even have led to further fragmentation and simplification of tasks.

Comparative industrial labour costs have been analysed in three previous articles contained in earlier issues of this Review. Each time the analyses were based on the same Swedish publication, which contains hourly wages, social charges and total labour costs for most of the OECD countries, and converts them into a common currency using official exchange rates. This provides an acceptable starting point for international comparison since the products of manufacturing industries compete on international markets on the basis of these exchange rates.

In November 1985 the National Institute introduced a new version of its quarterly forecasting model which included a large number (approximately 45) of explicit expectations terms. The model was solved on the basis of model consistent expectations.