The Nordic countries have traditionally been very similar and closely integrated, both economically and politically. The last ten years have seen the comprehensive liberalization of their financial markets and restructuring of their monetary and exchange rate policies. These issues formed the main focus of a CEPR joint conference with the Bank of Finland on `Exchange Rate Policies in the Nordic Countries', held in Helsinki on 21/22 September, at a time of considerable turbulence in international capital markets. The conference was organized by Alberto Giovannini, Jerome A Chazen Professor of International Business at Columbia University, Member of the Italian Treasury Ministry's Council of Experts, and Co-Director of CEPR's International Macroeconomics programme, and by Johnny Kerholm, Head of the Central Bank Department of the Bank of Finland.

In his opening presentation, `Equilibrium Real Exchange Rates', Alberto Giovannini noted that such equilibrium is an elusive concept for policy-makers. Intuitively, the real exchange rate should respond only to real shocks that affect the demand for and supply of goods whose prices it measures, but it also reacts to monetary phenomena. He sought to identify sources of real exchange rate fluctuations, using a two-sector model of a small open economy with a managed exchange rate. With imperfect competition in both non-traded and traded sectors, the traded sector faces a given world price and thus behaves competitively, but firms in the non-traded sector set prices at a constant mark- up over wages, and increased competition reduces their equilibrium prices and hence the real exchange rate. Giovannini considered both monetary determinants of real exchange rates, such as expectations of nominal exchange rate changes which have real effects when money is non-neutral, and real determinants, such as exogenous and endogenous productivity movements and shocks to real wages and public spending.

Eduard Bomhoff (Erasmus Universiteit Rotterdam) noted that a shift in government spending towards tradables could lead to a devaluation and stressed the need for caution in defining domestic real interest rates in such two-sector models. Joseph Zeira (Hebrew University of Jerusalem) suggested extending the analysis to consider inflation, excess volatility, imperfect credit markets and income (or wealth) distribution.

In response to the debate within economics during and since the Bretton Woods era over the relative merits of fixed versus floating exchange rates, Feldstein has proposed that countries base their choice more on political pressures than on any economic rationale. In their paper, `Country Characteristics and the Choice of the Exchange Regime: Are Mini-Skirts Followed by Maxi?', Seppo Honkapohja (University of Helsinki and CEPR) and Pentti Pikkarainen (Bank of Finland) examined 140 countries' choice of regime in 1982 and 1992 and found that flexible exchange rates had gained in popularity: the share of countries pegging to a single currency fell from 40% to 26% and that of `floaters' rose from 8% to 22%, while that of countries pegging to a currency basket remained around 24%.

Honkapohja and Pikkarainen then estimated logit and probit models linking countries' choice of regime to their total size, degree of economic development, international goods and financial market integration, diversification of foreign trade, and terms-of-trade fluctuations. Small countries with low export diversification were the most likely to peg their currencies, while other characteristics had hardly any effect. Developing countries had moved towards more flexible exchange rates, while those with well-diversified exports had moved towards more rigid arrangements. Increased monetary cooperation among EC member countries appears inconsistent with the `conventional wisdom', since they are rich and open and have diversified export structures, which supports Feldstein's `political' explanation or at least suggests that the exchange rate regime has no long-run real effect. Alternatively, the status of the central bank and the conduct of monetary policy may be more important than the choice of exchange rate regime per se.

Bernard Delbecque (Université Catholique de Louvain) argued that the lack of a formal theoretical structure to support the authors' hypothesis weakened their empirical results. Alan Sutherland (University of York and CEPR) was not surprised that so few statistically significant factors can account for countries' choice of exchange rate regime. There is no unique theoretical approach to the choice of optimal regime, which may also depend heavily on policy objectives and credibility considerations. William Branson (Princeton University and CEPR) and Charles Wyplosz (INSEAD, DELTA and CEPR) argued that the IMF classification scheme used by the authors is quite arbitrary, and many of the countries studied do not have well-organized capital and exchange markets.

In 1982 the Danish government announced a policy programme of `no more' devaluation and disinflationary stabilization, in response to a period of high inflation, growing unemployment, economic growth fuelled by budget deficits, a severe balance-of-payments deficit, and the erosion of international competitiveness through wage and price inflation. Despite membership of the ERM since 1979, the krone had been repeatedly devalued, and interest rates were among the highest in the OECD. In his paper, `Monetary Policy in Denmark in the Last Ten Years', Peter Erling Nielsen (Kobenhavn Universitet) focused on the role of long-term bond markets and the interactions between the budget deficit and deficit financing, interest rates, capital flows and monetary policy. He found that monetary policy in particular the management of interest rates had stabilized exchange rates and capital flows and thus made a small but effective contribution to the improvement of Denmark's macroeconomic performance, although part of the interest rate differential with Germany remained. Svensson's simple test of target zone credibility indicated that long-run interest rates are still too high to be compatible with the EMS bounds, although short-run interest rates appeared low enough to vindicate the authorities' credibility.

Susan Collins (Georgetown University and CEPR) argued that the key role of short-run interest rates and capital flows suggested the need for close attention to the estimated equation for net capital flows, since the method of its construction and possible simultaneity problems complicate the interpretation of its results. John Driffill (Queen Mary and Westfield College, London, and CEPR) questioned whether the Danish experience should be considered a success, since unemployment is still high and indeed growing.

In his paper, `Disinflationary Stabilization Policy Denmark in the 1980s', Torben Andersen (Aarhus Universitet and CEPR) focused on the credibility of the shift to a disinflationary regime that may risk a recession, formalizing the `production strategy' of shifting resources to tradables in a simple, two- sector model of a small open economy with endogenous, centralized wage bargaining. A reduction in government demand for nontradables does not reduce activity but instead induces a price reduction that moderates nominal wages, crowds in private demand for nontradables and hence also the supply of tradables, which improves the trade balance.

Andersen then used various measures to evaluate the credibility of this disinflationary policy. The interest rate differential between Danish and German long-term bonds had fallen to approximately 1% by 1991, but there had been a non-trivial spread for most of the 1980s, and Svensson's test indicated that credibility had been gained only recently. Wage inflation has also fallen, although it remains above its target levels. Unemployment has remained quite high, but private sector expansion in response to the 1982 policy shift more than compensated the fall in public sector employment. This further fuelled a powerful upturn in investment, and financial deregulation and the wealth increases induced by the sharp fall in nominal interest rates provided a further boost to private demand. Andersen concluded that the erosion of policy objectives in the mid-1980s had mainly derived from poor demand management and stressed the role of fiscal instruments in supporting real adjustment under fixed exchange rates.

Kari Puumanen (Bank of Finland) questioned Andersen's evaluation of the Danish experience as a success, since structural problems remain, notably high unemployment. Antti Suvanto (Bank of Finland) noted that Andersen's paper had not mentioned the behaviour of the real exchange rate, which was critical to the success of Denmark's `production strategy'.

In 1982, the new Swedish government also announced a disinflationary stabilization policy programme based on a fixed exchange rate, but starting with a `once-and-for-all' devaluation to ensure some immediate change. This policy incurred credibility problems in the medium term by fuelling inflation and thus creating incentives to devalue again, whereas the Danish government had faced major strains in the short run as it established the credibility of its low-inflation policy. In their paper, `Has the Swedish Krona Turned into a Hard Currency?', Lars Hörngren (Sveriges Riksbank) and Hans Lindberg (Institute for International Economic Studies, Stockholm) considered `relative' and `absolute' definitions of a hard currency, based on fixity of the exchange rate and a stable value vis-à-vis a consumption basket respectively. In the 1980s, Sweden underwent substantial deregulation and institutional reform, involving the abolition of domestic credit market, interest rate and foreign exchange controls; the introduction of a ban on government borrowing in foreign currency; and the Riksbank's announcement of an `official' target zone of 1.5% around a bench-mark value to replace its previous `unofficial' 2.25% band. Excess demand in the mid-1980s, driven by wealth effects and credit expansion, demonstrated that the development of fiscal instruments for demand management lagged behind these reforms.

Hörngren and Lindberg then considered the development of Sweden's hard currency policy as a sequence of tests. The Riksbank was forced to intervene substantially and raise its overnight rate by up to 6 percentage points to stem capital outflows in February and October 1990, and again in December 1991, which indicated that devaluation expectations and capital flows had become more sensitive to the exchange rate's position within the band. This suggested responding earlier with the overnight interest rate, as the Riksbank did succesfully in April 1992. Hörngren and Lindberg concluded that recent developments are consistent with the krona's status as a relative and absolute hard currency: its present real depreciation against the ecu reflects the fall of Swedish inflation below the EC average. Swedish experience also indicates the key role of fiscal policy in the formation of expectations and hence the need to safeguard the long-run sustainability of budgetary policies as an integral part of a long-run hard currency policy.

Zhaohui Chen (Columbia University) suggested spelling out more carefully the relationship between the long-run hard currency strategy and the choice of short-run policy instruments. Niels Thygesen (Kobenhavn Universitet) noted the similarities between Swedish and French exchange rate policies prior to 1982 and attributed the krona's exchange rate movements vis-à-vis the ecu in the 1980s to the large weight of the dollar in the currency basket.

In the 1980s, not only the Nordic countries but most other industrialized economies adopted disinflationary policies. In `Disinflation Experience in Finland Compared with Other OECD Countries', Palle Andersen (Bank for International Settlements) discussed credibility, initial conditions, the choice of nominal anchors and the role of other policies in complementing anti- inflationary monetary policy. He found that labour and product market flexibility are essential to successful shock treatment and that countries using monetary aggregate or exchange rate targets performed better than those with no medium-term nominal inflation anchors. He also found some evidence that tight programmes undertaken under severe economic stress are more credible than those introduced under more favourable conditions.

Andersen also studied Finland separately, since the sacrifice ratios used to measure the (transitional) costs of disinflation and its implied ranking took no account of specific unfavourable factors that underlie its current problems. He conjectured that policy did not yet assign instruments efficiently and transparently to the principal targets. Given Finland's relatively low wage rigidities, a fixed exchange rate could serve as an effective nominal anchor for anti-inflationary policy with low output costs. Tax-based incomes policies could also reduce inflation. Whether disinflation will resume once the current crisis is over will depend critically on the size of the cumulative output and employment costs.

William Branson noted that observed sacrifice ratios based on employment or output policies are usually influenced by factors unrelated to the disinflationary policies they are intended to measure. He suggested using time-series properties of unemployment instead, coupled with estimated wage and inflation equations to infer a given country's sacrifice ratio (and any hysteresis effects).

In the final paper, `Currency Band and Credibility: Finnish Experience', Olli-Pekka Lehmussaari and Antti Suvanto (Bank of Finland) and Laura Vajanne (Union Bank of Finland) assessed the credibility of the markka's target zone. Pressures on the peg had gathered momentum whenever the economy entered a slow-down, even though the authorities had been placing greater emphasis on exchange rate and price stability and less on competitiveness since the early 1980s. The greatest speculative attack to date occurred in 1986 and was followed by a sharp increase in the central bank's call money rate. This turbulence fell only after a strong and unexpected cyclical upswing, which led to the overheating of the economy and hence to an appreciation to the strong edge of the band. Daily and monthly interest rate differentials have displayed greater volatility since 1988 especially in short maturities than before, which may result from the widening of the band in November 1988 but more probably reflects increased uncertainty about future economic developments. Discretionary measures such as the surprise (downward) realignment of the band, supplementary reserve requirements on banks and changes in the mechanisms of liquidity control could all have contributed to the increased interest rate volatility.

Applying target zone models to Finnish data for 1987-90, the authors found that the authorities appeared to have defended a narrow, `unofficial' band within the official band. Svensson's credibility test revealed that the computed forward exchange rate remained inside the band for most of the period for maturities of up to one year, but credibility problems accumulated from autumn 1990 onwards. The March 1989 revaluation had little effect on interest rate differentials in short maturities, but it may have created expectations of a future reversal and undermined the band's credibility. There seems to be a lack of long-run credibility: the five-year forward exchange rate has consistently remained above the weak edge of the band.

Since the revaluation, the expected future exchange rate has generally exceeded the current spot rate while the devaluation risk has been highly volatile and increased towards the end of the period. Finally, causality appears to run from interest rate differentials to the exchange rate, suggesting the presence of intramarginal intervention, especially in the post-revaluation period. Taken together, these results indicate that the markka's band has become less credible since the revaluation, with an increasing but still highly volatile devaluation risk.

Lars Svensson (Institute for International Economic Studies, Stockholm, and CEPR) supported the authors' approach, arguing that the money market equilibrium condition can be used to interpret a central bank's policy activism towards the domestic economy, once it is decomposed into the expected rates of depreciation within the band and devaluation.

In the closing panel discussion, on `Problems of Economic and Monetary Policy in a Large Monetary Union', Johnny <143>kerholm pointed out that Finland's problem of countercyclical interest rate movements reflecting similar movements in devaluation expectations will not apply in a monetary union. Its large terms-of-trade fluctuations will be exacerbated, however, by asymmetric shocks: fiscal policy cannot control the effects of such fluctuations, and demand management in the nontradable sector will only make matters worse.

William Branson argued that the EMS faced a `prisoner's dilemma', especially after German unification: non-German members realized that a Deutschmark appreciation would be an equilibrium outcome but would not admit it because they had invested their credibility in pegging to the Deutschmark and therefore could not agree the necessary realignments. Now that individual countries have devalued, the outcome is less good in credibility terms; this `prisoner's dilemma' makes Stage II of EMU inherently unstable.

Hans Denkov (Danmarks Nationalbank) emphasized the role of short- term interest rates in securing the union's long-run credibility. He stressed the key role of member countries' short-run management of their interest rate differentials vis-à-vis Germany as an indicator of commitment to the fixed parities. Unlike the UK, Denmark has accepted that fixed rates cannot admit an independent monetary policy.

Lars Hörngren pointed out that we know neither the optimal combination of monetary policies within a target zone area nor whether central banks' high interest rate strategy is optimal or `only' something they have always done. This strategy and the adjustable peg will no longer be needed in a monetary union, but we do not know the costs of giving up short-term discretion over national interest rates, which may also lead to political tensions if national fiscal policies can no longer cope with asymmetric shocks.

Jan Qvistad (Norges Bank) noted that the choice of fixed rather than floating rates has never been challenged in Norway, where debate has focused on the choice of peg. In 1990, Norway unilaterally switched from a trade-weighted basket to the ecu, mainly because of inflation targeting and not because this was optimal for Norway. The last few months have shown that this unilateral peg must be strengthened, for example by multilateral agreement with other central banks including easier short-run credit facilities and effective multilateral surveillance systems.