Conflict between national economic policies and the possible benefits of internationally coordinated policies have become increasingly important questions in policy debates. Greater policy coordination, it is argued, might avoid the imbalances, instability and insecurity which have affected the world economy in the last decade. A workshop in International Economics focussing primarily on these issues was held in Stockholm on September 10-13. It was jointly organised by the Institute for International Economic Studies and CEPR; participation by CEPR Research Fellows was financed by grants from the ESRC and the John D and Catherine T. MacArthur Foundation.

The first paper was presented by Tony Venables (Sussex University and CEPR) on the subject of 'Industrial Policy Coordination under Imperfect Competition'. His paper studies the effects of industrial policy in one country on the welfare of its trading partners. It compares the level of industrial subsidies chosen independently by a single country with the level that would be desirable from the point of view of all countries together. Venables considers an 'industrial policy' which takes the form of a subsidy reducing firms' marginal cost. The effect of such an 'industrial policy' depends crucially on assumptions about entry of new firms into the industry and whether the subsidy changes the number of firms in the industry. If the number of firms in each country is fixed, then an independent 'nationalistic' choice of subsidy will result in subsidies smaller than those which would maximise world welfare. If, however, the number of firms can be changed - either directly through anti-trust policy, or indirectly through entry and exit in response to changes in profits - then independently chosen national policies will lead to higher subsidies and more firms than will internationally coordinated policy.

Recent literature on trade and industrial policy under imperfect competition by Brander and Spencer has identified cases in which it is optimal to use policy interventions such as tariffs and subsidies that would be undesirable in a competitive environment. This literature is based upon simple and rather special models. In their paper on 'Industrial Policy Under Monopolistic Competition' Harry Flam (IIES) and Elhanan Helpman (Tel-Aviv University) analyzed the effects of such policies in a more general model of imperfect competition. They conclude that the response of the economy to policy interventions can be predicted only if one has detailed knowledge of the economy's structure. Although there is potentially a case for industrial policy, the detailed specification of such a policy depends on information about the structure of industries, the nature of factor market interactions, the degree of competition, and the structure of preferences. In addition, any practical application would have to take into account the effect of similar policies by trading partners and the possibility of retaliation. Flam and Helpman conclude that one should be very cautious in using simple models to advocate the desirability of activist industrial policies.
Is it desirable for a country to subsidise research and development in order to give its firms an advantage in international competition? Avinash Dixit (Princeton University and CEPR) addressed this question in his paper 'The Cutting Edge of International Technological Competition'. Dixit modelled technological competition as a race from which one firm emerges as the winner and obtains increased profits from the patents it gains. Losers in the race may also make some smaller gains, perhaps through opportunities to license the new technology. An important assumption is the free entry of firms into the subsidised industry. The relationship between the amount of R & D undertaken and its cost is assumed to be the same for all firms in a country. This relationship gives international competition a 'knife-edge' property. If the average cost of R & D is even slightly lower in one country than in another, then all firms in the higher-cost country will choose not to enter the R & D race! This analysis seems to imply that subsidies to R & D will be an effective means of promoting a country's interests. Dixit argued, however, that this is largely an illusion: free entry of firms implies that, taking the unsuccessful participants in the race together with the winner, aggregate industry profits are zero. Consumers in the model face the same prices whether they are supplied by domestic or foreign firms; they are no better or worse off. Domestic firms as a whole are no better off either; even if the subsidy drives out foreign firms, free entry of other domestic firms drives industry profits down. This makes the economic case for a policy of promoting R & D a rather weak one.

International trade in textiles is subject to a wide variety of quantitative restrictions. Carl Hamilton (IIES) considered the effect of such restrictions on Hong Kong's exports in his paper 'An Assessment of Voluntary Export Restraints on Hong Kong Exports to Europe and the USA'. Such quantitative restrictions result in higher prices for consumers, as do taxes on imports (tariffs). Unlike tariffs, however, the revenue from the higher prices due to quantitative restrictions accrues to the foreign producers and not the home government. Import licences are therefore valuable, since they permit exporters to earn higher profits, and such licences are freely traded in Hong Kong. Hamilton used data on the traded value of such licences in order to construct estimates of the 'tax equivalents' of the quantitative import restrictions - i.e. that tariff rate which would have yielded the same increase in price as the quantitative import restriction itself. These 'tax equivalents' are useful measures of the effect of import restrictions on the consumer, and Hamilton was able to construct the measure for a large number of countries over a long time span and for a large number of commodity groups. Hamilton found that the market prices of textiles were very similar in different countries in the European Free Trade Association (including the European Community), so that the effects of particular countries' individual trade restrictions seemed to be nullified by trade between them. The value to Hong Kong of the increased profits or 'rents' associated with European and American restrictions on its textile exports was estimated to be 1.7 per cent of GDP and in excess of 10 per cent of value added in the apparel industry in 1983.

Countries experiencing balance of payments or debt service difficulties are often advised to devalue their currency. Thorvaldur Gylfason (University of Iceland and IIES) and Marian Radetzki (IIES) asked 'Does Devaluation Make Sense in the Least Developed Countries?' Their model is one of an economy in which a large part of the population receives only a subsistence wage, and in which macroeconomic policy is constrained not to reduce this real wage. With real wages held constant, devaluation has a greater contractionary effect on GNP than if nomimal wages were fixed. Using admittedly incomplete data on 12 countries, Gylfason and Radetzki estimate the inflows of foreign aid necessary to maintain GNP in the face of a devaluation and fixed real wages. They then argue that if real wages were allowed to fall, relatively modest inflows of aid would be sufficient to maintain GNP or total wage earnings in a successful devaluation.

'Does policy coordination necessarily reduce inflation?' was the question posed by Marcus Miller (Warwick University and CEPR). In a model in which inflation depends not only on past inflation but also on future monetary policy through the exchange rate, he compared 'time consistent' policies with policies which involve 'pre-commitment'. In this context 'pre-commitment' means that the government, having announced a policy, will stick to that policy in later periods, even if they are tempted to deviate from it. In the absence of such 'pre-commitment' the only credible policy for the government is a 'time-consistent' one, in which the government has no incentive in later periods to deviate from its previously announced policy. In general, 'time-consistent' policies, although credible to the private sector, are inferior to policies to which the government can pre-commit itself.

In Miller's model, for example, a high exchange rate helps to reduce inflation, and high real interest rates push up the exchange rate. The optimal inflation-reducing policy is to choose an interest rate now that is relatively low, but announce that the interest rate in the future will be higher. This policy, which couples leniency towards inflation now with a threat of toughness in the future, will, if it is believed, actually reduce inflation now through its effect on the current exchange rate. The policy is effective only if it is believed, however, and this credibility may be in doubt, for the government will have an incentive in the future to renege on its threat of toughness, and the private sector knows this. Miller considered alternative policies that avoid this difficulty. A simple policy rule in which the government ignores the effect of its monetary policy on the exchange rate turns out to be preferable to a rule satisfying time-consistency. Miller discussed other results related to the issue of international policy coordination. International coordination of anti-inflation policy has no effect if governments adopt rules which ignore the effects of interest rates on exchange rates. Miller also argued that coordination may actually lead to outcomes inferior to uncoordinated policy, if governments have to follow time-consistent policies.

The paper by Torsten Persson and Lars Svensson (IIES) on 'International Borrowing and Time-consistent Fiscal Policy', was also concerned with the problem of time-consistency. Lucas and Stokey have established that a sufficiently varied structure of government debt can allow government to design a time-consistent fiscal policy that is as good as the optimal policy under pre- commitment. Persson and Svensson extended this analysis to open economies in which both the government and the private sector may have foreign debt. They find that in a large open economy, the Lucas-Stokey result does carry through, but it does not hold in an economy that is too small for its policies to affect its world interest rates. They find that the use of debt restructuring to make optimal policy time-consistent requires the government to have as many effective debt instruments there are choice variables facing governments in later time periods; by choosing its debt instruments now, the present government can effectively induce future governments to choose the optimal 'pre-commitment' policy. The feasibility of such a policy depends on its changing world interest rates; this is possible in a 'large' open economy but not in an open economy that is too small to influence interest rates.

The efficacy of trade embargoes has been the subject of much recent discussion. In 'Embargoes and Multinational Corporations', Alasdair Smith (Sussex University and CEPR) addressed the question of how the investment behaviour of a multinational corporation could be affected if trade in goods might be blocked by an export embargo imposed by the government of its home country, while production in its foreign subsidiaries could avoid such an embargo. Unsurprisingly, this gives the multinational an incentive to invest abroad rather than supply foreign markets through export sales. Smith noted that even without the prospect of an embargo foreign firms already face strong incentives to invest abroad rather than to export, because of the strategic advantages that investment in a particular country gives in competing with firms in that country or in deterring them from entry. Further, firms located in the country subject to the embargo may respond to an embargo by entering the market even if they had previously chosen not to enter. Smith concluded that an embargo is likely to be effective in only a very narrow range of circumstances.

Lars Svensson (IIES) presented a preliminary version of a joint paper with Sweder van Wijnbergen (World Bank and CEPR) on 'International Transmission of Domestic Policies'. Their objective was to analyze how exogenous shocks to the economic system in one country are transmitted to other countries. Their model is one in which the demand for money is determined by a 'cash-in-advance' constraint - purchases can be made only if the buyer has cash available. Prices are 'sticky', in the sense that they are set by firms in the absence of full information about the current state of the economy. In general, both monetary and real disturbances in one economy affect the consumption and trade balance of other economies.

In the final paper of the workshop Joseph Stiglitz (Princeton University) surveyed 'The Causes and Consequences of the Dependence of Quality on Price'. He argued that markets in which there are asymmetries in information and in which prices convey information about quality behave very differently from the textbook equilibrium model. In the labour market, one implication is that wages may not fall in response to unemployment, if firms believe that a lower wage will attract lower quality employees. Other phenomena which Stiglitz argued can be explained in this way are the use of lay-offs rather than work-sharing, differential unemployment rates for different groups of workers and similar firms paying different wage rates (and having different labour turnover rates) for the same work. In such models, the market equilibrium is generally not socially optimal, and there is a case for government intervention.

The research discussed at the workshop ranged over a wide variety of topics. The issue of policy conflict and strategic behaviour did recur, however, in the papers by Venables on industrial policy, Dixit on subsidies to high technology industries and Smith on embargoes. This is likely to lead to further meetings and research on policy conflict and strategic behaviour within the international economy, in the context of CEPR's programme on international security, funded by the MacArthur Foundation.