Exchange Rate Fluctuations
Permanent Damage?

There have been dramatic swings in exchange rates during recent years: sterling appreciated sharply in 1979-82 and the value of the dollar soared between 1981 and 1986. Such temporary exchange rate overvaluations may cause long-lasting or even permanent damage to industrial competitiveness, according to Research Fellow Charles Bean at a lunchtime meeting on 26 June. Bean highlighted several reasons why a transitory appreciation of the exchange rate could lead to a fall in both the demand for and supply of domestically produced goods, which would not be completely reversed even if the exchange rate returned to its previous level. His research suggested that a period of undervaluation was necessary to restore the status quo ante. For this reason the dollar might have to fall a further 20-30% to restore US current account balance, and the decline in North Sea oil revenue in the next decade will require a significant real depreciation of sterling to attract resources back into exporting and import-competing industries.

Charles Bean is Reader in Economics at the London School of Economics and a member of the Macroeconomic Policy Group of the European Commission. His talk drew on research published in Discussion Paper No. 177. The lunchtime meeting at which Bean spoke was sponsored by the German Marshall Fund of the United States.

Between 1978 and 1987, the nominal exchange rate of sterling rose by 17%, with an associated loss in UK export price competitiveness (that is, an appreciation of the real exchange rate) of 23%. Bean reported the results of simulations using a small macroeconomic model of the UK economy which suggested that around 50% of this real appreciation of sterling could be attributed to North Sea oil. The remainder resulted from the combination of restrictive monetary policies and adverse supply- side developments. Economic theory suggests that as the revenue from oil declines, UK price competitiveness must improve in order to maintain current account balance. Bean's own research suggests, however, that small exchange rate movements will not be sufficient to achieve this. The sterling appreciation and loss of competitiveness between 1979-82 has had permanent effects on the ability of British producers to compete with foreign rivals , and a corresponding period of sterling undervaluation may be necessary.

This is an example of a hysteresis effect, in which the equilibrium value of a variable is affected by the recent history of the variable itself. Hysteresis effects have figured prominently in recent analyses of labour markets and unemployment. Bean's research applied this concept to markets for traded goods. Hysteresis in the demand for such goods might arise because uncertainty about product quality leads consumers to stick with products that they already know, or because of significant costs attached to switching suppliers. These arguments suggest that the demand for a firm's product is likely to depend on the previous level of its sales.

Hysteresis effects could also arise on the supply side of a goods market. If there are significant fixed costs of entry into a market, such as establishing a distribution network, then a large exchange rate appreciation, which leads producers to abandon foreign markets, may reduce permanently the equilibrium value of exports: markets once lost are not easily regained. A large temporary undervaluation may therefore be necessary to encourage domestic firms to re-enter these markets. Technical progress through learning-by-doing may have similar effects; if the level of productivity depends on past levels of output then a loss in competitiveness, which leads to a fall in production today, will lower productivity and the ability to compete in the future.

In order to assess the importance of these factors, Bean examined data on British export performance since 1900. Confining attention to just a few years of data, he argued, would severely limit the precision with which one can estimate the possible long-run effects of misalignments lasting one or two years. The long historical data set used by Bean had other advantages as well; while the 1979-80 sterling appreciation is perhaps the most pronounced fluctuation, there are other periods since 1900 when UK competitiveness changed very sharply. Bean estimated a model in which the demand for UK exports depended on the relative prices of UK and foreign-produced goods and on the level of world economic activity. The estimates suggested that transitory fluctuations in price competitiveness have had a permanent (or very long-term) effect on export shares. A temporary deterioration in competitiveness of 20%, sustained for two years, will permanently lower the level of exports by around 3%. The fall in exports will not be reversed by a return to the initial level of the exchange rate; a corresponding period of undervaluation is necessary.

Bean also estimated a model of UK exporters' behaviour in which producers incur adjustment costs when they change the level of exports. These costs may differ according to whether exports are increasing or decreasing, capturing the idea that producers may face higher costs in entering new markets or expanding in existing ones. Bean's estimates confirm that the adjustment costs of exporters are indeed asymmetric and differ according to whether exports are rising or falling. Holding other variables constant, UK export prices have tended to be higher when exports are increasing than when they are decreasing.

Bean's analysis also has implications for the behaviour of the dollar in the immediate future. Many commentators believe that the recent fall in the dollar may have been enough to eliminate the US current account deficit. Bean disagreed: if US producers have been squeezed out of foreign markets because of the earlier appreciation, it may be necessary for the dollar to fall by as much as a further 20-30%.

The apparently permanent effects of exchange rate fluctuations suggested by this analysis also strengthened the case for stronger coordination of exchange rate policies, Bean concluded, such as recent proposals for a 'target zones' regime. After the talk Bean was asked how policy should respond to real shocks to the exchange rate, within the EMS for example. Bean responded that if there was a significant change in the real exchange rate then the 'target zones' would have to be adjusted. If this appreciation was likely to be temporary, its effects could be alleviated by providing the export sector with a temporary tax subsidy to prevent a loss of market share. Bean accepted, however, that such 'temporary' support measures were easier to introduce than to remove: there was hysteresis in the political process as well.