Recent events in both Europe and the US including the conclusion of the North America Free Trade Agreement, the completion of the European single market and recent developments in the Exchange Rate Mechanism of the EMS have highlighted the need for greater understanding and cooperation between economies on both sides of the Atlantic. Specialists in international macroeconomics and trade from Europe and North America gathered to discuss the significance of these events at a conference on `The New Transatlantic Economy', held at Georgetown University, Washington DC, on 5/6 May. The conference formed part of CEPR's research programme on `Market Integration, Regionalism and the Global Economy', supported by a grant from the Ford Foundation. It also represented part of an ongoing collaboration between the Center for German and European Studies (CGES) and the Program for International Economic Studies (PIES) of Georgetown University, supported in part by a grant to CGES from the German government and by funds from the Walsh School of Foreign Service in honour of the School's 75th anniversary. The conference was organized by Matthew Canzoneri, Professor of Economics at Georgetown University and Director of PIES, and Vittorio Grilli, member of the Italian Treasury Ministry's Council of Experts, both Research Fellows in CEPR's International Macroeconomics programme, and by Wilfred Ethier, Professor of Economics at the University of Pennsylvania.

In `Transatlantic Policy Cooperation with Sticky Labour Market Regimes: The Reality of the Real Side', written with Yue Ma, Andrew Hughes Hallett (Princeton and Strathclyde Universities and CEPR) explored the conditions under which the operations of a European Central Bank dominated by Germany may be compatible with its members' incentives. Assuming that Germany is more inflation-averse than other members, they compared three possible ERM regimes: in the `hard EMS', Germany targets its domestic variables and the other members' `reaction functions' are determined by their commitment to remain within an exchange rate band; in the `preference transfer solution', exchange rate bands are much tighter and realignments ruled out, so the non-German members adopt Germany's preferences as well as its policies; and in the `domain solution', Germany applies its own policies but targets average European variables under an agreed set of preferences. The domain solution yields the highest overall gains for Europe, but Germany only gains if its own preferences change to reflect overall European welfare or the dynamic gains from greater European stability feed back to Germany. European labour market rigidities imply that the gains from transatlantic policy coordination are greater for the US than for Europe under all three regimes.
Nouriel Roubini (Yale University and CEPR) pointed out that all three `cooperative' solutions were based on exchange rate bands with intramarginal intervention, while the model's instruments were also target variables, so most of the gains would be realized in both Nash and cooperative settings; he therefore wondered to what extent these results rested on cooperative behaviour. The potential gains from cooperation would also be exaggerated if the inflationunemployment trade-off were worse in reality than in the model. Alberto Giovannini (Columbia University and CEPR) agreed and suggested recalibrating the model so that actual policy gains corresponded to the Nash outcome.

In `International Policy Coordination and the Future of EMU: A German Perspective', Axel Weber (Universität Bonn, Universität Gesamthochschule Siegen and CEPR) presented statistical evidence to show that the G3 economies have relied on policy coordination to maintain exchange rate stability and only used market interventions in periods of crisis. He used US and German data to show that the two countries' interventions had never conflicted, while sales and purchases consistent with the goal of exchange rate stabilization within the EMS suggested their shared purpose and probable coordination of policies. All market interventions since 1985 had also been sterilized in the long run. Despite this policy coordination, such interventions appeared to have had little effect; imposing long-run sterilization of market interventions revealed that these had proved ineffective in stabilizing exchange rates in the long run. Weber then applied this framework to analyse the ERM crises of September 1992 and August 1993 to show that a greater German defence of the EMS target zone through intramarginal interventions would have been ineffective and also imposed heavy costs on the German economy by driving up interest rates, inflation or both. He concluded that the ERM's failure had resulted not from a lack of Bundesbank commitment to maintain the parities but rather from a lack of coordination of monetary and interest rate policies across Europe.

Joseph Gagnon (US Department of the Treasury) commented that assuming that market interventions have no short-run impact and then examining the model's long-run implications for long-run sterilization and policy effectiveness might be a more natural restriction for Weber's VAR analysis. Giorgio Basevi (Università degli Studi di Bologna) pointed out that his model did not allow for policy coordination among other members of the EMS excluding Germany. Matthew Canzoneri stressed the difficulties of interpreting policy responses to speculative attacks with monthly rather than daily intervention data.

Eric Bond (Pennsylvania State University) presented `Trading Blocs and the Sustainability of Inter-Regional Cooperation', written jointly with Constantinos Syropoulos, which focused on whether trading blocs increase their market power relative to the rest of the world or enhance inter-bloc trade. In their model, agreements within trading blocs are commitments while those between blocs achieve tacit cooperation in a repeated game. GATT therefore plays a coordinating role among the multiplicity of trade agreements, with the Nash bargaining solution chosen from the various possible Pareto-optimal agreements. For a symmetric case with blocs of equal size, inter-bloc free trade agreements become more difficult to support as the size of each bloc increases. If blocs differ in size, the welfare of a representative country is greater in the larger bloc than in the smaller.

Peter Moser (Universität St Gallen) commented that interpreting repeated game models is difficult when enforcement requires strong assumptions about information. Konstantine Gatsios (Athens University of Economics and Business and CEPR) noted that little is known about how differences in technologies and endowments among the potential members of a customs union affect its formation. Giorgio Basevi suggested examining whether trading blocs' stability varies with their size and the number of them in the world economy.

In `Effects of Trade Liberalization on the Members of a Common Market: A Lumpy Country Analysis', Alan Deardorff (University of Michigan) considered the effects of world trade liberalization on members of the European Union. Modelling a two-country common market as a `lumpy country', in which differences in immobile factor endowments, consumer amenities or agricultural subsidies between its regions prevent factor prices from equalizing, Deardorff showed that the effects of reducing agricultural subsidies and liberalizing trade on regional welfare differentials depend critically on the patterns of regional specialization. Any empirical analysis of the welfare implications of the Uruguay Round should therefore take these differential effects on regional welfare within trade blocs into account.

L Alan Winters (World Bank, University of Birmingham and CEPR) noted that revenue effects should also be incorporated in models of welfare analysis of this type. He also pointed out that the agricultural subsidies applied by Common Agricultural Policy apply throughout the European Union, unlike those in the model: differences in the level of subsidies across EU regions can only arise from regional specialization.

In `The Increased Importance of Direct Investment in North Atlantic Economic Relationships: A Convergence Hypothesis', written jointly with Anthony Venables, James Markusen (University of Colorado) compared patterns of horizontal direct investment and trade, motivated by the observation that levels of multinationals' direct investment in the North Atlantic area has risen faster than trade volumes, despite reductions in both tariffs and transport costs. Markusen captured such effects using a model of an imperfectly competitive homogeneous goods sector with firm- and plant-specific fixed costs. Firms can locate either in their home countries or in both countries with firm-specific costs split between each plant. In this framework, direct investment increases relative to trade as either country size or factor productivity increases, or as firm-specific fixed costs become more important relative to plant-specific fixed costs. A reduction in transport costs has ambiguous effects: investment rises if this leads to a fall in marginal production costs, but the decline in the relative price of foreign goods raises the volume of trade. As countries become more similar in their factor endowments, size and production costs, the ratio of investment to trade rises.

Alasdair Smith (University of Sussex and CEPR) questioned Markusen's assumptions that multinationals split their fixed costs equally among plants and that countries' differences in factor endowments or technology are qualitatively significant. Jacques Mélitz (Institut National de la Statistique et des Etudes Economiques, Paris, and CEPR) noted that the harmonization of institutional structures across countries may also give rise to increased horizontal direct investment.

In `Speculative Attacks on Pegged Exchange Rates: An Empirical Exploration with Special Reference to the European Monetary System', Barry Eichengreen and Andrew Rose (University of California, Berkeley and CEPR) and Charles Wyplosz (INSEAD and CEPR) examined the distributions of policy variables before and during speculative attacks to determine whether exchange rate adjustments result from policy changes before the attacks from self-fulfilling speculative behaviour. They considered attacks that were both `successful' and `unsuccessful' in effecting realignments. They considered government budget deficits, real exchange rates, the trade balance, interest rates, foreign reserves, money supply and credit, to determine whether their observations were drawn from a common underlying distribution. While they rejected this hypothesis for all of the variables except money and credit growth, for the OECD countries from 1967 to date, they could reject it only for reserves and credit growth for a sub-sample of ERM countries. They concluded that theories of speculative attacks based on changes in fundamentals are less appropriate to ERM than to non-ERM countries, which may indicate that the greater capital mobility of the ERM period has reduced the costs of launching speculative attacks.

Robert Cumby (Georgetown University and Council of Economic Advisors) commented that only a small probability of an exchange rate realignment, perhaps resulting from a small change in policy, is needed to induce a speculative attack, and that these small changes in policy fundamentals might not be observable under this approach. Also, even if changes in policy variables cause an attack, there is no reason to expect policy to reverse following an attack. Alberto Giovannini noted that the fundamentals driving speculative attacks may not lie in financial markets, and Susan Collins (Georgetown University and CEPR) added that the chosen set of current fundamentals may not capture expectations about future policy changes. Robert Flood (IMF) pointed out that the volatility of future shadow exchange rates is itself a fundamental that is not determined in the case of non-ERM flexible exchange rates.

In `Central Bank Independence and the Role of Reputation', Marcus Miller (University of Warwick and CEPR), Ben Lockwood (University of Exeter and CEPR) and Lei Zhang (University of Warwick) developed a model of the optimal choice of central banker to focus on reputation effects when the natural rate of unemployment is inefficiently high, based on a trigger strategy equilibrium in which the government delegates the conduct of monetary policy to a conservative central banker as a substitute for its own conservative reputation. The Fed's ability to provide liquidity during the recent recession without inducing inflation demonstrates that a central bank can conduct an active stabilization policy and achieve target rates of unemployment once it has established a record of anti-inflationary commitment. This may explain the greater persistence of unemployment in many European countries. If their labour markets cannot be made more flexible, as has happened recently in the UK, the monetary policy of a newly established European Central Bank with no established reputation may prove less accommodating as unemployment persistence increases, which risks giving Europe less stabilization than the US, when it logically requires more.

Stanley Black (University of North Carolina) wondered whether a trigger strategy solution in which a far-sighted banker can be less conservative was a plausible model of reputation formation. A signalling game in which credibility has to be earned might yield more credible solutions than one in which central bankers and governments start with reputations, and he noted that a myopic central banker could achieve more stabilization in this trigger strategy model than a conservative one with no reputation.

In `Trade Liberalization and Trade Adjustment Assistance', Robert Staiger (University of Wisconsin) and K C Fung (University of California, Santa Cruz) argued that import-competing resources adversely affected by increased trade may deserve special treatment relative to resources displaced for other reasons. They used a two-country model to show that trade adjustment assistance introduced by one country as part of a cooperative trade agreement reduces the distortionary costs of its partner's tariffs, raises the value to the trading partner of maintaining the agreement, and tempers the enforcement problems associated with such agreements. If each country implements adjustment assistance, enforcement problems are symmetrically relaxed, tariffs may be symmetrically reduced, and both can achieve higher welfare.

Constantinos Syropoulos (Pennsylvania State University) questioned the robustness of their conclusions and suggested generalizing the model to consider cases in which agents use adjustment assistance in a non-cooperative manner, receive large rather than marginal subsidies, or face asymmetric technology shocks. Alan Deardorff welcomed the model's useful contribution to considering the effects of domestic adjustment assistance on foreign countries. Staiger added that foreign countries had asked the US to provide trade adjustment assistance to its domestic markets during the Kennedy Round negotiations, which supported the findings of his model.

In `Trade Liberalization as Politically Optimal Exchange of Market Access', Arye Hillman (Bar-Ilan University) and Peter Moser presented a model of trade policy decision-making based on income distribution and producer interests to show why liberalization may take place when governments are more concerned about political support than social welfare. In their two-country framework, governments' political support depends on the income shares of the factors of production. Cooperative tariff reductions may enable both governments to benefit each others' exporters and thus effect income redistributions which they could not have achieved unilaterally. This may enable them to enhance their political support without necessarily enhancing overall domestic welfare. If one government liberalizes unilaterally because it prefers a more liberal trade regime, however, the other will respond by increasing protectionism in order to secure its own domestic political support.

Martin Richardson (Georgetown University) argued that many of their results depended on the two-country nature of the model and suggested extending it to consider a three-country framework, and also to incorporate other policies such as domestic producer subsidies.

In the panel discussion that concluded the conference, Alberto Giovannini noted that few of the papers had addressed the changes in the international economic environment resulting from developments in NAFTA and the EU. While Hughes Hallett and Ma had showed that macroeconomic spillovers between the US and the EU were small, US relations with Mexico or Japan and EU relations with Eastern Europe were of equal interest. He also noted the discussion at the conference mirrored that at the recent G7 summit in focusing on structural aspects rather than the coordination of responses to economic fluctuations.

Dale Henderson (Federal Reserve Board, Washington DC) noted that the international coordination of trade policies had made greater progress than that of macroeconomic or exchange rate policies. Trade liberalization has been a gradual process, and macroeconomic policy coordination might best follow a similar route. Weber had showed that US and German foreign exchange market interventions were not at cross-purposes, but that did not indicate the kind of coordination witnessed in trade policy in the post-war period.
Paul Wonnacott (Institute for International Economics, Washington DC) replied that this was because the gains from trade are easily demonstrated, while many issues concerning the optimality of currency areas or the effectiveness of exchange rate intervention remain unresolved.

Richard Portes (CEPR and Birkbeck College, London) stressed the contrast in the perspectives of US and European academics concerning integration policy: Europe takes a gradualist approach whereas the US advocates policies of shock therapy. For example, US advocacy of voucher-systems in privatization programmes that would instantly create stock markets in Eastern Europe contrast with West European proposals for direct sales to a developing middle class. There is also widespread academic support in the US for adjustment loans to be made conditional on far-reaching macroeconomic stabilization programmes rather than success in micro-level institution-building. Portes noted that this difference in focus emphasized the benefits for academics from both sides of the Atlantic of meeting in this conference to discuss these issues and compare perspectives.