Probing the figures for signs of trade crisis

One of the questions most asked in the financial markets today is whether a balance of payments crisis is brewing in the UK. The answer is: of course a balance of payments crisis is brewing, it always is. No one who has observed the history of our economy in the past few decades could sensibly conclude otherwise. But the real question is whether such a crisis is imminent, and I think the answer to that is no.

The trade area is bedevilled by more data problems than most. At present, there are two important minefields to negotiate - the so- called balancing item problem, and the Intrastat problem. Recent reports from economists at UBS and elsewhere have argued that these two factors indicate that the deficit is much larger than the official data show, and that the next crisis may be just around the corner.

By contrast, throughout the 1980s there were some congenital optimists - including Lord Lawson at one point - who argued that Britain's trade position was much better than the official data showed, primarily because of the very same balancing item, both here and in other countries. The term 'balancing item' is a euphemism for data errors that we know must be there because some figures in the accounts do not balance as they should. But while we know the mistakes must be there, we do not know why, and that is the problem.

Take the global trade numbers. In 1993, recorded world imports exceeded world exports by pounds 70bn, clearly an impossibility. The most likely cause of this discrepancy is that some service exports are under-recorded, and since the UK is a large service exporter, it is reasonable to assume that part of the error occurs here. A rough guess would be that the UK share of this discrepancy might be about pounds 4bn a year.

Furthermore, this happens to be a similar figure to the average balancing item in the UK's own trade statistics, which indicates either that the trade and current account deficits must be smaller than officially recorded, or that capital inflows must be larger than recorded. So far then, it would appear that the official statistics might be tending to exaggerate the extent of the UK's trade problem, if only slightly.

There are, however, two main reasons why this might not be the case. First, there is a possibility that the balancing item in the UK accounts might represent under-recording of capital inflows - which would then indicate either that foreigners have more claims on UK assets or that UK residents have borrowed more than we think overseas. Either way, we would have to pay interest on these unrecorded capital account items, in which case the current account deficit may be under-recording interest outflows.

This is a rather convoluted argument, since it starts with a factor - the balancing item - which should be good news for the current account, and contorts the case until it re-emerges as potential bad news. Nevertheless, it is a possibility, and it could worsen the current account by up to pounds 4bn a year.

The second and much more concrete way in which the figures might be worse than they appear concerns the errors that are now emerging in the European Union's new Intrastat system. In the old days before the single European market started in January 1993, we used to calculate the trade statistics by adding up the goods that passed through the ports - a highly reliable method. The statisticians are no longer allowed to do this in the single market, so they have to conduct sample surveys asking firms how much they have imported and exported in a given month.

Not surprisingly, many firms are unclear about this, especially when it comes to reporting imports. Consequently, there has been a tendency throughout the EU to under- record trade flows in both directions, but to understate imports by more than exports, therefore giving a flattering picture of the overall trade balance.

We know this for sure because we can add up the trade figures of all the EU countries, focusing only on trade with each other. Obviously, this should produce a net position of zero, as indeed it did in 1992. However, in the first half of 1993, the sum total of EU internal trade produced an overall surplus of ecu20bn (pounds 15bn). This is the extent to which the Intrastat system has flattered the net trade position of the 12 members of the EU by systematically underestimating imports relative to exports. If the application of the new system in the UK has been about average, then the UK trade deficit with the rest of the EU has been overstated by about pounds 2bn in the first half of 1993.

Assuming that the figures became a little more accurate as firms became used to the new system in the second half of 1993, this implies that the maximum effect in 1993 as a whole was to underestimate the UK trade deficit by pounds 3-4bn. (Furthermore, this estimate may well be too high, since the Intrastat system may for the first time correctly attribute imports coming in through Rotterdam to the rest of the world category and not to the Benelux countries. This may previously have overstated imports from the EU countries.)

Where does all this meandering lead us? A cursory examination of balancing items here and elsewhere suggests that the UK trade deficit is overstated by some pounds 4bn a year. However, there are reasons for believing that the UK's outflow of interest payments might be underestimated by up to pounds 4bn, and that the trade deficit with the rest of the EU in 1993 might be flattered by (say) pounds 3bn. Overall, this indicates that the official figures might be a little optimistic for 1993, but probably by less than pounds 3-4bn, or 0.5 per cent of GDP - a trivial amount in the great scheme of things.

What really matters is not whether the trade figures are fractionally optimistic or pessimistic in any given year, but whether the trend is so adverse that inevitable trouble is just around the corner. Certainly, over long periods trends look grim, but in the very recent past our trade performance has not been too bad.

The graph shows the UK deficit of goods and services, both as published by the Central Statistical Office and adjusted for variations in the economic cycle here and abroad. (This latter measure is probably the more relevant since it is not distorted by different positions in the economic cycle.)

It is clear that the underlying position of UK trade improved gradually from 1989 to 1992, and then improved quite sharply in 1993 as the impact of the sterling devaluation worked into the system. In fact, the cyclically adjusted trade deficit narrowed from pounds 15.5bn in 1992 to pounds 6.7bn in 1993, a gain equivalent to about 1.6 per cent of GDP. This is a measure of how successful the 1992 devaluation has proven to be - unusually so, in fact. It has certainly bought us a good deal of time, during which it would be sensible to start addressing the deeper- seated problems of the economy.

One last related point. Among other problems, the trade statistics have certainly been overestimating export prices in 1993, and correspondingly understating export volume. Adjusting for this, the economy might well be growing at a healthy 3 per cent rate, not the more anaemic 2 per cent shown in the official data.

(Graph omitted)