Measuring international price and cost competitiveness
BIS Economic Papers No 39
Few economic indicators attract as much controversy as those of international competitiveness. One reason for this is the imprecision of the concept: in common parlance "competitiveness" can be used to cover almost any aspect of market performance. Product quality, the ability to innovate, the capacity to adjust rapidly to customers' needs and the absence of restrictive practices in the labour market are frequently evoked in discussions of competitiveness. This paper, however, will focus on a much narrower meaning, that based on relative prices or costs. It might be stressed at the outset that the link between this narrow concept and economic performance more generally is not unambiguous. The ambiguity arises from the fact that its international relative price or cost position can be both cause and result of a country's economic performance. On the one hand, it is clear that if relative costs are too high, the ability to compete internationally can be compromised. On the other hand, successful economic performance can lead to an exchange rate appreciation, and thus to higher relative costs or prices. For instance, if enterprises in a country become more successful in the non-price dimensions of performance - if they are innovative, flexible, produce high-quality goods and so on - then the real exchange rate would be expected to strengthen. Price and wage competitiveness - the narrow concept - would thus appear to "worsen". But such "deterioration" would of course be a symptom of success, not of failure. A second reason for controversy is that even the narrow concept of competitiveness can be given many distinct statistical forms, using prices, wages and other costs. There is no one ideal measure, and the large number of different measures that are in common use often diverge appreciably. One purpose of the present study is to survey these measures and to examine key trends of the major currencies in the light of the various indicators, attempting where possible to account for the divergences observed.
Two general issues are raised in the construction of indices of real effective exchange rates. The first is the choice of currencies to be included in the calculation of relative indices. This is reviewed in the first part of Section I of this paper. One important aspect is the increased number of countries that "count" in international trade. In particular, the rise of the Asian NlEs and other dynamic economies in South-East Asia has increased the number of currencies that may need to be included in effective exchange rate calculations. The inclusion of such currencies implies significant modifications to the movement in real effective exchange rates. This is considered in more detail below.
The second general issue concerns the choice of price or cost measure used. As for industrial countries, there are basically three sorts of measures in common use: those based on unit labour costs in manufacturing industry; those based on consumer prices (or some other broadly-based price measure); and those based on export unit values. These classic measures are reviewed in the second part of Section I; recent revisions to the indices calculated by the BIS are also discussed. The final part of Section I uses these measures to review actual developments in some major currencies over the past twenty years, paying particular attention to how and why different indicators can "tell a different story". The broad developments in competitiveness between Europe, the United States. Japan and the dynamic Asian economies provide a central focus. The successive exchange rate crises in Europe from September 1992 have of course given measures of intra-European competitiveness added interest, and what the different measures tell about this is also considered. The issues raised in measuring the competitiveness of commodity exporters (mainly in the developing world) are rather special and these are the subject of Section Ill.
However, few observers any longer rely solely on the classic real effective exchange rate indices. An apparently relatively simple extension is to take the ratio of one measure to another to paint a wider picture of a country's competitive position. Most common among these are ratios of price to cost indices as a proxy for profitability; but other ratios have also been used. These are considered in Section II of this paper.
A more radical departure from the classic real effective exchange rate is the greater emphasis placed on levels of competitiveness. The standard measures of relative costs and prices are limited by dependence on a quite arbitrary choice of base year. They do not allow statements such as "unit labour costs were X% higher in country X than in country Y in 1990" to be made; only statements about relative changes are possible. Yet popular assertions about actual differences in labour costs are legion. Translating this perception into operational measures has, however, always faced formidable difficulties - notably as regards the valuation of output at a consistent set of prices. But the considerable research effort made in recent years to develop carefully constructed measures of relative productivity and to compute detailed estimates of purchasing power parities has begun to tip the scales in favour of developing level-based measures, at least for a number of industrial countries. Section IV reviews some recent work in this area.
A final element of recent efforts to broaden the scope of competitiveness measures is the greater attention paid to non-manufacturing: as the relative importance of manufacturing industry declines, the need for such a re-emphasis is likely to grow. Not only are non-manufacturing outputs increasingly traded, but service inputs are frequently key components of traded goods even if these inputs are themselves not directly traded. The advent of detailed sectoral national accounts in most industrial countries has greatly widened the range of measures that can be construcred. The penultimate section of this paper uses national account statistics to take an empirical look at the distinction between tradable and non-tradable goods production, examining in particular relative productivity and profitability. This analysis is, of course. highly preliminary; nevertheless, it does serve to warn against placing undue reliance on any one measure, and also uncovers other important aspects or symptoms of competitiveness.
Looking carefully at all the different measures of competitiveness cannot but instil a good deal of reticence about basing strong conclusions on any one measure. Indeed, this is the single most important point underlined in the concluding section. Nevertheless, there are a number of conclusions that would seem to be borne out by several measures. The first concerns the competitiveness of the main areas. South-East Asia remains highly competitive. The United States has become more competitive in recent years. The Japanese situation was always more ambiguous because of marked differences between sectors, with the country appearing extremely competitive in certain goods - notably in the electronics area - and over-priced in others. At any event, the recent sharp appreciation of the yen brought about a marked loss in the country's earlier competitiveness. Europe was much worse placed than the others in the early 1990s, with many indicators pointing to a serious competitiveness problem. Exchange rate changes in recent months have gone some way towards alleviating this. The second conclusion is that a number of European countries had become relatively uncompetitive even within Europe: here too the exchange rate adjustments since September 1992 have done much to correct divergences in intra-European competitiveness.