Collaborative research between economists in the United Kingdom and France has assumed growing importance in recent years.
The Sixth Anglo-French Colloquium in Political Economy was held at Birkbeck College, University of London, on 26 and 27 April 1984. It was jointly financed by the ESRC and the Maison des Sciences de l'Homme and organised by CEPR and the Centre d'Economie Quantitative et Comparative.

The theme was international monetary economics. Most of the seventeen papers presented dealt with economic interdependence and the coordination of national policies, issues which appeared especially relevant in the light of the forthcoming London economic summit.

The Colloquium began with two explicitly comparative cross- country studies. Michael Artis spoke on 'Wage Inflation in Western Europe'. He had obtained empirical estimates of wage inflation equations on a comparable basis for several European countries. The underlying model was eclectic, being broadly consistent with both augmented Phillips curve and target real wage interpretations. The treatment of inflation expectations led to two sets of equations based on Monetarist and Keynesian views of the inflationary process. But the results were not much affected by which predictor was used to generate inflation expectations. They confirmed that wage inflation could be explained by expectations, inertia, real wage resistance and pressure of demand for a variety of Western European countries.

Claude Bismut presented a paper (co-authored by Patrick Artus), 'Exchange Rate and Wage Price Dynamics.' This was one of several papers extending Rudiger Dornbusch's model of exchange rate 'overshooting,' in which the initial adjustment to a 'shock' or new information causes the exchange rate of 'overreact' and move beyond its new long-run equilibrium value. Bismut stressed the interaction between wage and price determination with feedbacks from the exchange rate and interest rates. The structural model was estimated for five major economies and the results used for a simulation exercise. The Japanese case gave the only anomalous simulation results: there a positive monetary shock led to explosive exchange rate undershooting. This is consistent, however, with backward-looking expectations formation.

Patrick Artus's paper, 'La boucle prix-salaires en economie ouverte avec imparfaite mobilite des capitaux', modelled wage- price inflation in an open economy, again building on the Dornbusch framework, but allowing less than complete capital mobility. Government behaviour is derived from a loss function including the exchange rate and level of reserves as arguments. The empirical results suggested exchange rate overshooting in the UK, Germany and Japan.

One participant questioned whether the wage equations in these papers rested on properly specified optimising or bargaining processes. All the authors responded that since they were concerned with fairly broad macroeconomic issues, it seemed sensible to take functional forms prevalent in the literature or employed in large macro models.The Bismut paper provoked further comments. The Japanese results were attributed to the particular form of quasi-rational expectations employed. Since most of the potential dynamics are suppressed, the results must be treated with caution.

Matthieu Feroldi discussed 'La dynamique du taux de change: du long au court terme'. This too extended the overshooting model by allowing imperfect capital mobility and modelling long-run trade and capital account constraints. Stabilisation policy is hindered by the difficulty of distinguishing the long-run determinants of the real exchange rate, or by limited foreign reserves, or by the policy makers' ignorance of wherethe shocks affecting the system originate. Another interesting feature of the model was a 'double-shooting' phenomenon in response to negative monetary shocks: short-run overshooting followed by relatively long-lived undershooting.

In 'An Empirical Investigation into the Causes of the Failure of the Monetary Model of the Exchange Rate', Michael Wickens started from the poor track record of empirical monetary exchange rate models. Using a novel methodology discussed more fully in the summary of CEPR Discussion Paper no.7 (Bulletin No 2) he 'models the misspecification' of the monetary approach. This allows the 'blame' for misspecification to be distributed (in a statistical sense) between the various building blocks of the monetary approach. Interestingly, the uncovered interest parity condition did not figure importantly - the major culprits appeared to be the purchasing power parity assumption and the assumed functional form of the money demand equation.

How does a monetary union such as the EMS affect economic interdependence? This question featured in the final paper of the first day by Jean-Pierre Laffargue: 'Politique budgetaire et union monetaire'. He examined the effects of fiscal policy in countries of different sizes belonging to the same monetary union in a model with a fixed exchange rate between two (European) countries (France and Germany) and floating rates between them and a third country (the USA). If two countries were of a similar economic structure but of different sizes, monetary union would not change the impact on them of the third country's fiscal policy. An expansionary fiscal policy on the part of one member of the union would have a deflationary impact on the other. One participant remarked that the model had long-run neoclassical properties.

Most model builders must now take seriously the idea of 'rational expectations' in their work. David Begg's paper, 'Rational Expectations and Risk Premia', examined the consequences of extending the conventional rational expectations framework of linear models and certainty equivalence to the non-linear case for financial markets. Financial panics can occur in such models, and this suggests a rationale for day-to-day intervention by the monetary authorities. Conventional rational expectations models are often criticized for assuming that economic agents are 'neutral' in their attitudes toward risk. Begg took a case in which investors were risk averse. The prevailing monetary policy will then feed back onto the nature of the risk premia. The discussion centred on the problems and advantages of relaxing the naive features of current rational expectations models. One discussant noted that the utility function in the paper was myopic and that future research might examine the intertemporal problem more fully.

Bernard Dumas's 'Introduction de l'incertitude en economie monetaire' set forth a model whose main feature was the non- neutrality of monetary policy. The model assumed rational expectations and slowly adjusting prices. Its features were fairly Keynesian. A positive external shock to the system led to an immediate and sustained increase in consumption, an improvement in the external account and an increase in inflation in the short but not the long term. A positive internal monetary shock gave a slight increase in consumption, a deterioration of the external account and a short-run reduction in inflation. The discussion located the source of the monetary non-neutrality in the use of the rate of expansion of the money supply (rather than its level) as a policy instrument.

Marcus Miller's paper (co-authored with Mark Salmon), 'Policy Coordination and Dynamic Games', described a game-theoretic approach to policy design in a two-country model with policy interdependence. In particular, it examined the impact of monetary policy on real exchange rates via real interest rates. Relative to the optimal coordinated policy design, non- cooperative behaviour substantially alters the form of policy interdependence between the two countries, given a certain set of 'plausible' parameter values for the model. In particular, non- cooperation implies that domestic policy becomes much less sensitive to overseas inflation, while policy impact on the real exchange rate becomes much more pronounced.

Anton Brender spoke on 'Le bloc monetaire international du modele Simulo: une interpretation des evolutions de change de 1981 et 1982'. The 'world' which the model depicts is divided into twenty zones, and the paper features the economic interdependences between the zones, not only in the flow of goods and services but also in financial and international monetary relationships. In particular, the model sought to explain autonomous capital inflows; for example, the high level of the US dollar appeared as the counterpart to a reluctance of US banks to invest abroad.

Andrew Britton addressed European trade policy, drawing on material assembled over the past year by a RIIA/NIESR study group. Pressures for European protectionism arise first from the high level of European unemployment and doubts about classical arguments for free trade when labour markets fail to clear, as well as the belief that certain European industrial sectors (notably technically advanced product industries) are lagging behind US and Japanese dominance. Britton also briefly touched upon the idea of coordinated European expansion. With such a policy, however, there would remain a problem of unbalanced trade within Europe due to differences in import propensities. Coordinated expansionary policy might therefore require regulated intra-European trade.

Patrick Minford's paper, 'A New Classical Strategy for Multi- Country Modelling', described the linkage of nine small macro models of identical structure, with investigations into the effects of fiscal and monetary shocks in the model. An interesting simulation result was that a bond-financed US budget deficit does not affect either US or world output but significantly raises world real interest rates - i.e., there is 100% crowding out. The discussion pointed out that a high US budget deficit and high external borrowing will also have crowding out effects in Europe. Another participant queried the role of wealth effects in the model - the generally 'new classical' character of the model would seem to imply a Ricardian view of the national debt, so that wealth effects should not be present.

Georges de Menil's paper, 'La surveillance multilaterale: un nouveau dialogue macroeconomique pour les Cinq?' took as its starting point the statement of Ministers at Versailles and the Williamsburg Communique. It considered the prospects for coordinated European expansion - i.e. a macroeconomic 'locomotive'. This approach to recovery was discussed at Bonn several years before, but the Bonn program was swiftly followed by the second oil crisis and a shift towards firm anti- inflationary policy. At Versailles and Williamsburg there was again a discussion of convergent policy, in particular, of coordinated monetary policy. Failure to carry out these initiatives was probably due to the ensuing debt crisis, political inertia and a belief in the necessity of reducing inflation. There is no reason, however, to preclude fresh efforts in this direction. Such a coordinated policy should encompass fiscal as well as monetary policy. In devising such a strategy one would have to perform various technical exercises in comparing alternative scenarios, and this would require a permanent professional secretariat. One discussant pointed out the importance of possible structural differences in the coordinating countries (e.g., differences in real wage rigidity or import propensities).

Richard Brown's 'Thoughts on the Longer-Term Evolution of the International Financial Institutions' covered four areas of the international monetary system: macroeconomic performance and policies, exchange rate arrangements, the debt overhang and external flows. He argued that the influence of the IMF should be maintained more through its 'seal of approval' than through its resources. In the discussion, one participant suggested that the various programmes applied by the Fund when lending to LDCs have typically been severely deflationary and perhaps counterproductive. The reply was that the IMF's aim was to reduce the external deficit, and that Fund programmes frequently include supply-side or microeconomic policy as well as overall deflationary measures. In connection with debt rescheduling, the problem of moral hazard and the difficulty of discerning credible precommitment were mentioned.

In 'The Welfare Case for the European Monetary System', Jacques Melitz looked at the EMS as a coalition in a game-theoretic setting. He saw the gains from membership of the EMS in the avoidance of bargaining costs. Absent such costs, some other co- operative solution would generally be superior. With them, however, membership of an exchange rate union like EMS outperforms both non-cooperative solutions and even co-operative solutions (if bargaining costs are sufficiently high). Melitz argued that economic differences among the partners require occasional realignments of the EMS. Such a realignment may change the rules for distributing the costs of exchange intervention as well as the parities.

David Currie presented a paper, 'Simple Macropolicy Rules for the Open Economy' (co-authored with Paul Levine), which develops techniques for the design of simple optimal rules in linear stochastic rational expectations models. Using an open economy model with a developed wage-price sector, a government budget constraint and a floating exchange rate, they construct simple 'decoupled' monetary and fiscal policy rules. The best simple rule for a single open economy is a price rule (interest rates proportional to the price level), but this performance does not carry over to the world level. One conclusion which echoed points made by Britton, de Menil and Melitz, is the inherent difficulty of designing simple coordinated policy rules. There may be incentives for countries to adopt policy rules that give better individual performance when used singly but perhaps not when adopted generally. An important issue raised in the discussion was the implicit assumption of policy-invariant parameters within the model. Currie replied that the Lucas critique only seriously affects policy simulation results with econometric models, an exercise quite different from his own.

An excellent conclusion was provided by Charles Wyplosz's paper, 'Sterling and the Fiscal Aspects of North Sea Oil' (jointly authored with Francesco Giavazzi and Jeffrey Sheen). This examined the role played by fiscal policy in appropriating and disposing of the revenue stream accruing from a resource discovery. If the revenue is used to retire external debt then there will be an imbalance in debt service; this must be reflected in the long run in the trade account, leading to a permanent loss of competitiveness through exchange rate appreciation. Fiscal policy then has both short and long run effects: the real exchange rate appreciation remains even when the resource has been exhausted. Wyplosz was asked whether the model could admit real assets and capital accumulation, so as to generate more traditional 'Dutch disease' results. The model was designed, however, to illustrate a specific aspect of resource discovery; introducing real assets would almost certainly complicate it to the point of requiring a simulation exercise in order to derive clear-cut results.

The participants agreed that the Colloquium had been exceptionally productive in assembling so much closely related work from the two countries. This opened up a wide range of possibilities for collaborative research, which will be especially fruitful in an area so closely tied to current policy issues. Some cross-country cooperation among the participants was already under way, and other projects could be expected to result from the work presented.