Is there a point to forecasting?

The BCIS has produced a two-year forecast of tender prices for some time, but the request of subscribers to come up with a five-year forecast presents a whole range of different challenges.

Nothing in the future is certain but over a two-year period it seems reasonable to rely on the lead indicators:

• New orders ought to predict the pattern of output

• Housing completions will surely follow the pattern of starts

• Changes in fixed investments must include some building work

• Long-term wage agreements for one trade will indicate future settlements for others.

Of course, few of these things turn out to be true.

There is some comfort offered, however, from many other economic forecasters predicting over the same period. For example, the Treasury's Forecasts for the UK Economy quotes 41 forecasts from city firms, brokers, banks and research organisations which all provide predictions for inflation, GDP and so on, for the next two years. Beyond that it gets a bit lonely, with only 12 forecasters prepared to offer advice up to 2010.

The known unknowns

The current economic climate, and decisions and actions being taken now, will undoubtedly affect the construction market in five years time - but which actions and how they will influence the future are unknowable. The linkages and chain of events are not always obvious, even in hindsight. The further you look into the future the more of Donald Rumsfeld's "the known unknowns" there are likely to be.

So why do longer term forecasting? As the economist David Cadman says: "All forecasts are wrong. The question is, are they useful?" Feedback from BCIS subscribers shows that the answer to that question is undoubtedly "yes". BCIS collects together long-term statistical indicators, examines the linkages between them, discusses possible outcomes and explains the conclusions it draws from them. Subscribers then have a base from which they can make their own forecasts and offer advice to their clients.

So what runes do BCIS consult when forecasting construction tender prices? Construction is a free market in economic terms and prices react more to demand pull than to cost push. The key direct drivers are:

• Input costs - labour, material and plant

What will be the effect of joining the Euro or diverting resources to deal with the bird flu crisis?

• Construction demand - orders and output

• Capacity - availability of labour, material, plant and management.

The indirect drivers - the things in the general economy which influence the above include availability of money, interest rates, GDP, capital investment, earnings and government policy on just about anything.

Government policy affects construction both directly and indirectly. The effect of an increased hospital building programme may be obvious but what, for example, will be the effect of joining the Euro or diverting resources to deal with the bird flu crisis?

In preparing its current five-year forecast BCIS looks at many long-term relationships, including inflation and tender prices, GDP and construction output, and construction output and tender prices.

Figure 1 shows that inflation in late 1979 and in 1980, as measured by the Retail Prices Index, rose to very high levels, peaking at 22.4%. It fell away to 3.5% in 1983 and peaked again at 10.6% in 1990, since when rates have generally remained between 1% and 4%. Since the Bank of England was handed responsibility for controlling inflation, it has managed to keep it relatively low and stable and in June 2001 the figure was 1.9%. Tender prices have been much more volatile with higher peaks, much lower troughs, precipitous falls and negative inflation. Significantly, changes in the direction of tender price inflation since 1989 have tended to precede similar movements in retail prices. As for the few forecasters on the Treasury's panel who predict for 2010, all expect inflation to be between 2% and 3%. History would suggest that tender price inflation will be above that.

Figure 2 shows that since the 1950s there has been a fairly close relationship between annual GDP and annual construction output. The latter tends to be more volatile and for the past five years both have been relatively stable. Again, all long-term economic forecasters predict steady GDP growth of between 2% and 3%, which would suggest continued growth in construction.

The BCIS also uses econometric models to aid its deliberations plus we all have our own pet economic theories.

I am a believer in the output gap theory, which basically says that when output is rising faster than its long-term trend, inflation tends to rise. When output is rising slower than the trend, inflation falls.

Having examined all these factors, what does BCIS think will happen? Our next five-year forecast, due out in August, will have full details. Following a year when new work output fell, new work output is expected to return to growth in 2006, albeit growing a little behind trend, thereafter rising ahead of trend over the next three or four years. The industry is expected to be kept busy by investment in education and health, either traditionally funded or via PFI, and through work in the infrastructure sector. The 2012 Olympics will boost construction output, particularly in London, to an increasing degree as the event approaches. On the downside, the Channel Tunnel Rail Link is due for completion in 2007. All this adds up to pressure on prices, together with upward pressure from input costs, and the BCIS believes that prices will rise by more than double the rate of general inflation over the next five years.

How good have the BCIS forecasts been? Compared with our 2001 prediction, actual growth in new work output over the five-year forecast period was a little stronger and the annual average inflation in tender prices over this period has been 6.9% compared with our forecast's average of 5.4%.