August 1998 Preliminary version The Political Economy of European Exchange Rates: An Empirical Assessment Jeffry A. Frieden Department of Government Harvard University Cambridge, MA 02138
2 For thirty years, the member states of the European Union have pursued monetary stability in the region. In practice, this has meant fixing national currency values against the Deutsche mark (DM), the continent s anchor. This macroeconomic goal has been reached with varying degrees of success. The snake and early European Monetary System (EMS), from 1973 until about 1983, were of limited scope, while the later EMS went through a cycle of optimism (1984-1991), crisis (1992-1994), and renewed optimism (1995-) in the runup to Economic and Monetary Union. And while some European countries were able to persist in pegging their exchange rates to the DM, others were quite unsuccessful. This suggests two major unanswered questions: why did the fortunes of monetary integration (MI) vary over time, and why did its attainment vary so much among European countries? This paper looks at the statistical record of exchange rate movements in Europe since 1973 to see how they might be explained. Specifically, I examine currency movements against the DM and explore the impact of a variety of economic, political, and political economy variables on exchange rate policies. I evaluate a series of factors that might help explain European monetary integration, and currency policy more generally. These can usefully be divided into four categories. First are variables of most relevance for the aggregate welfare effects of exchange rate policy, from which come implications about what policy would be were it chosen purely on the grounds of economic efficiency. Second are European regional political and institutional factors, especially links between MI and other aspects of European integation, and the institutional character of exchange rate arrangements. Third are political economy factors, which emphasize the impact of electoral and other political institutions on macroeconomic policy. Finally, interest-group variables stress the distributional impact of currency policy and how this affects policy. These are not mutually exclusive sources of trends and variation in European monetary integration; indeed, they almost certainly were all at work. I am especially interested in interestgroup considerations because they have been little discussed in this context. Such economicinterest explanations differ from most interpretations of currency union in their focus on the
3 disaggregated distributional effects of fixing exchange rates. Particularly, I emphasize support for MI from cross-border investors and exporters of specialized manufactures who are especially concerned about currency volatility; and opposition to fixed rates from those, especially import and export competers, who are primarily concerned about the potential loss of the ability to engage in currency depreciations to gain (or regain) international competitiveness. The paper presents a simple statistical evaluation of potential causes. These include proxies for economic-efficiency, regional institutional, political economy, and distributional factors; I also include controls for broader macroeconomic trends which are expected to affect exchange rates. These are used to try to explain two dimensions of currency movements. The first is the general propensity of the exchange rate to depreciate against the Deutsche Mark (the region s implicit or explicit anchor); the second is the volatility of the exchange rate against the DM. Several factors are uniformly important. Positive macroeconomic trends movements in the terms of trade, overall economic growth, the payments balance reduce the propensity to devalue and also reduce currency volatility. Also important are two measures used as indicators of private-sector interests with regard to currency policy: the higher the level of manufactured exports to Germany and the more positive are changes in the trade balance (controlling for the state of the current account), the less likely are depreciations and the less volatile the exchange rate. Capital controls are associated with both more depreciation and greater volatility, although the causal implications are unclear. Membership in regional monetary agreements reduced the depreciation rate, albeit somewhat weakly; it had a much clearer impact on reducing volatility. More independent central banks had no impact on the propensity to depreciate, but did reduce currency volatility. Similarly, the partisan composition of government did not affect depreciation, however left-wing governments were associated with more stable currencies. Somewhat simpler (and less reliable) data tend to show that governments in highinflation countries tended to delay depreciations until after an election. This is consistent with an exchange rate-based political business cycle, in which governments facing elections used a real
4 appreciation to artificially increase incomes. It is also consistent with the use of the exchange rate to enhance credibility, in the sense that politicians and voters might regard the currency peg as an indicator of policy credibility. Contrary to much discussion, indicators of the net benefits of inclusion in an optimal currency area the most important measure of the efficiency of MI have no measurable impact on exchange rate policy. Nor does unemployment, or conditions associated with government stability (the size of the ruling majority and the numbers of parties in government). As mentioned above, the partisan character of government and central bank independence appear only to affect volatility, and not the overall stance of currency policy. These results indicate that the state of the macroeconomy, particularistic private interests, regional institutional membership, monetary policy institutions, election timing, and (less clearly) the desire for credibility are important determinants of European currency policy. Favorable macroeconomic conditions and high levels of manufactured exports to Germany are associated with lower levels of depreciation and less volatility against the DM, while increased net import competition (controlling for the state of the current account) tends to lead to depreciation and volatility. Membership in the snake or ERM tended to reduce both depreciation rates and volatility, while capital controls had the opposite effect. Central bank independence and left-wing governments are associated with reduced currency volatility. In countries with relatively high inflation, governments tended to delay depreciations until after elections. The analysis gives little support to the relevance of optimal currency area criteria, government stability, or the unemployment rate for government decisions on the exchange rate, or to the relevance of central bank independence or partisan politics for broad trends in currency policy (although these last two do affect volatility). I begin with a summary of the course of the political economy of European monetary integration. Then I discuss the most compelling factors that might explain the currency policies of the past 30 years; I emphasize my view that the distributional effects of exchange rate
5 policies are crucially important. I go on to present statistical analyses of European currency policies since 1973, which give rise to some conclusions. European monetary integration: dimensions of variation As early as 1969, member countries of the European Union expressed a desire to stabilize their exchange rates. 1 This task became urgent with the collapse of the Bretton Woods system between 1971 and 1973. The first formal attempt to create the desired zone of monetary stability was the formation of the snake in the tunnel, soon only the snake, in 1973. Within a few months only Germany and Benelux were full and reliable participants, with Denmark sometimes cooperating, and this remained the case until 1978. In that year, plans for a new European Monetary System and its exchange rate mechanism (ERM) were unveiled, and the system came into operation early in 1979. The EMS appeared to have added little to the snake for its first five years: only Germany and the Benelux countries, and now more reliably Denmark, were able to keep their currencies more or less aligned. But between 1983 and 1985 three high-inflation laggards, France, Italy, and Ireland, began to stabilize and converge monetarily with the Deutsche mark zone that was the core of the EMS. From 1985 until 1992 a zone of monetary stability indeed became more and more of a reality among EU members, eventually attracting adherence from such improbable candidates as the United Kingdom (long unwilling) and Spain and Portugal (long unable). The Nordic countries and Austria, not EU members but considering joining, also tied their currencies to the EMS. In this setting, member states developed and began to implement plans for a common European currency within a broader Economic and Monetary Union (EMU). This idyll was shattered in the aftermath of German unification and the macroeconomic mix with which the German authorities financed it. High German interest rates forced upon national governments a choice between sticking to a fixed DM parity and enduring a steep recession, or letting the currency float away from its rate in the ERM and hoping that this would stave off a downturn. Most post-1985 adherents to the zone of monetary stability eventually let
6 exchange rates move with at least a widening of the acceptable target zone, at most a substantial depreciation. Momentum for EMU was rebuilt after the currency crises faded. Eleven EU will start the final steps toward a single currency in 1999, and at least that many are likely to finalize a full currency union by 2002. The 1973-1994 experience raises important analytical questions. Attempts to hold to fixed exchange rates 2 were much more successful at some times than at others within Europe as a whole. In addition, EU member states had highly varied experiences within the snake and EMS. This means that there is meaningful variation both over time and among countries. The policy choice most in need of explanation can be expressed simply: the degree of fixity (flexibility or volatility) of the nominal exchange rate against the Deutsche mark. This definition of the thing to be explained, which might be questionable in other historical and regional contexts, is justifiable in the post-1973 European setting. First, exchange rate stability was a publicly stated goal of all European countries. Second, it was clear early on that such stability implied fixing against the Deutsche mark. Third, despite reservations about a focus on nominal as opposed to real variables, the attention of all relevant actors policymakers, observers, economic agents was on nominal exchange rates. 3 In what follows, then, the dependent variable is the stability of nominal currency values against the Deutsche mark. As we shall see, this changed substantially between 1973 and 1994, and among countries. There have been times when European currencies have fluctuated quite wildly against each other in general, and times when they have been quite stable; some national currencies have been much more volatile than others; and individual currencies have gone from being very unpredictable to being very stable. All three dimensions of variation are worthy of explanation. Potential determinants of European currency policy: Efficiency, regional institutions, political economy, special interests
7 Exchange rate policy requires policymakers to weigh different values and make tradeoffs among them. Currency flexibility is valuable because it allows policymakers to vary the exchange rate for economic or political reasons. These reasons might include making domestic products cheaper relative to foreign goods, or using monetary policy to respond to such unexpected shocks as a world recession or a deterioration in the terms of trade, or manipulating the currency for electoral purposes. On the other hand, exchange rate stability is valuable because it reduces economic and policy uncertainty about a price that is of great importance to the economy. Policymakers might care about stability in order to lower the cost of currency volatility to those involved in cross-border economic activity, or to signal the reliability of their own macroeconomic policy. By the same token, a currency s value, the exchange rate, can also involve potentially contradictory goals. 4 Policymakers might value a weaker (more depreciated) currency in order to help national producers of tradable goods. Depreciation raises the (domestic-currency) price of foreign products, which helps domestic producers who compete with imports; it also lowers the (foreign-currency) price of domestic products, which helps domestic producers for export. This is often referred to as the competitiveness motive for depreciation. 5 It makes national tradables more competitive on world and home markets, reflecting the substitution effect of currency movements, in which a depreciation leads to substitution out of foreign and into domestic goods. On the other hand, policymakers might well prefer a stronger currency for the other effect of a currency s value, the income effect: a stronger (more appreciated) currency raises the purchasing power (thus real income) of national residents. An appreciated exchange rate increases national purchasing power and might well be broadly popular, despite its negative impact on tradables producers. Many things can affect the weights policymakers give to currency flexibility and stability, and to the currency s value, both in general and in the context of European monetary integration (MI). These factors, which can be taken as potential determinants of the policy choices made, can be divided into four categories. The first is associated with aggregate economic effects --
8 social efficiency -- and especially the potential that regions might be natural parts of an optimal currency area. The second emphasizes institutional factors, such as membership in organizations and the links between MI and other policies. Third are political economy factors, such as partisan and electoral politics, government stability, and central bank independence. Finally, there are distributional factors that focus on the role of special interests that can put pressure on governments to make policies in their favor. I now turn to a brief presentation of potential explanations of European currency policy, organized into these four broad categories. Efficiency. Many accounts of European monetary politics emphasize the role of monetary integration as an instrument to achieve national macroeconomic goals, to enhance aggregate social welfare. There are in fact two well-developed arguments in which stabilizing exchange rates at the limit, merging them might be an efficient national policy: optimal currency area and credibility approaches. The first such argument is that of the optimal currency area (Mundell 1961, McKinnon 1963, and Kenen 1969 are early classics; Masson and Taylor 1993 and Tavlas 1994 are recent surveys). The approach specifies the circumstances under which it is welfare improving for a nation to give up its exchange rate autonomy. 6 This is the case where exchange rate policy would otherwise be superfluous, either because it would be ineffective or because it can better be carried out by a bloc of national monetary authorities than alone. High levels of factor mobility among countries make individual national currency policies by any one of them ineffective, while a production structure that leads to correlated exogenous shocks makes such policies unnecessary. In other words, the more mobile factors are across countries and the more similar their susceptibility to external shocks, the more desirable is a monetary union. Convincing and well-developed as this may be for normative purposes, it can only be a partial explanation of European policy choice. Existing studies have found too little labor mobility among European countries, and too little intra-european similarity in production structures and sensitivity to exogenous shocks, to explain the level of interest in currency unification. The general conclusion is that Europe is not an optimal currency area, and even the
9 hard core of the EMS may not have been one at the time it was established. 7 Of course, on both dimensions there is variation among EU member states, so that some might be more appropriate members of a currency union than others. While the course of monetary integration in Europe cannot be explained on the basis of optimal currency area criteria alone, they may have had differential effects on different countries and are worth considering. A second argument often made is that European countries pegged to the Deutsche mark in order to import German anti-inflationary credibility (Giavazzi and Pagano 1989, Weber 1991). There are various explanations proposed as to why a currency peg might itself have been more credible than simply committing to lower inflation. 8 This is hard to assess empirically, as it is difficult to know how to determine what portion of the observed inflation reduction was due to the DM peg. 9 The general point is that the exchange rate can be used as a nominal anchor for credibility-enhancing purposes, and that this might explain some of the European experience. 10 There is, however, reason for concern about explanations of national policy that rely on the efficiency of the chosen policy. We know better, from generations of political economy research, than to assume that national policymakers will pursue policies solely because they are welfare improving. This is especially true in macroeconomic policy, where Pareto improvements appear possible in most countries monetary and fiscal policies most of the time. So it would in fact be surprising were European currency policies powerfully motivated by concern for national efficiency gains. In addition, neither approach would be very useful in explaining the dramatic shifts in the policies of national governments. These include, for example, France, Ireland, and Italy between 1982 and 1985, during which their commitment to the EMS became generally credible; or the rapid turnarounds in British and Swedish policy toward the EMS in the early 1990s, which led them to effectively abandon the system. It is hard to imagine that either optimal currency area criteria or national desire for inflation fighting credibility changed as quickly as did national policy in these and many other instances. Optimal currency area and credibility concerns are
10 structural and cannot do much explanatory work with regard to short-term changes in policy outcomes. Nonetheless, both optimal currency area and credibility considerations have probably played a role in European monetary integration. Their importance has almost certainly varied among countries: the fact that the core around Germany (Benelux and Austria, especially) is closest to qualifying as an optimal currency area probably helps explain its considerable success in stabilizing exchange rates, while the DM peg probably played a significant role in helping reduce inflation in Italy, Spain, Portugal, and Finland. Regional institutions. The weakness of aggregate economic explanations has led many analysts to look for European political factors that might help explain the course of European monetary integration (Eichengreen and Frieden 1994 present a survey). The most common is that monetary union involved an inter-state bargain with linkage across issue areas, and was made possible by its connection to the broader institutions of European integration. Many supporters of monetary union are such primarily because they think it will speed and deepen European integration more generally. In this view, adopting a single currency is itself a way of committing to further European unification, as it will almost certainly bring a host of follow-on integration initiatives to the fore. It is often remarked that German support for EMU is largely of this type, to signal Germany s commitment to its European partners rather than for any inherent desire to give up the Deutsche mark. This is closely related to the point that national support for monetary integration has been bargained away in return for other policies that are in fact more highly valued. It is common, for example, for analysts to assert that Germany supported EMU in the early 1990s as part of deal with France which gave Germany needed support in its foreign policy initiatives in Central Europe (Garrett 1994). This sort of linkage politics is seen as essential to European integration broadly and monetary integration more narrowly (Martin 1994). A variant of this is the widely held view that many countries want to participate in European monetary integration primarily to
11 ensure that they have a seat at the table, as first-class citizens, for the EU s continuing negotiations over its future. While there is much to this argument, it obviously cannot apply for all countries in the EU for linkage politics to work, some had to value MI highly on its own, else there would have been no scope for gains from political exchange! So while it may be the case that some countries participated in efforts for monetary integration more consistently than they would otherwise have because it was tied to other European policy issues, this can only be a valid explanation for the actions of a subset of EU member states. 11 In any event, EU institutions undoudtedly influenced the character and course of MI. Macroeconomic political economy factors. There has been an explosion of systematic political economy analysis of monetary policy in the past fifteen years (excellent surveys or examples include Alesina 1989; Alesina and Summers 1993; Alesina and Roubini with Cohen 1997; Blackburn and Christensen 1989; Grilli, Masciandaro, and Tabellini 1991; and Price 1997). Virtually all of it has focused on inflation, and despite differences there is general consensus on the sorts of variables that are conceivably important in explaining different rates of inflation among nations and over time. Electoral politics typically matters either in a partisan manner, with Left governments more likely to be inflation-prone, or in an opportunistic election-cycle manner, with inflation more likely in the runup to an election. Given the putative tradeoff between the short-term positive impact of inflation and its long-term costs, governments with shorter time horizons are more likely to be inflation-prone; these are especially weak governments, such as those with slim majorities (or minorities) or shaky multi-party coalitions, or those facing imminent elections. In other versions, a divided or coalition government makes it difficult for voters to assign credit or responsibility for macroeconomic policies to particular parties (analogously, parties can impose externalities on each other when there are more in government) so that there is a greater incentive to undertake policies with clear benefits but uncertain costs. 12
12 While this literature has focused almost exclusively on monetary policy, similar political economy factors might operate in the making of exchange rate policy. Devaluation often involves similar political tradeoffs as inflation between the short run and the long run: a devaluation s near-term stimulative impact comes at the cost of longer-term credibility (and, perhaps, higher inflation as input costs rise). However, there may also be reasons to associate shorter time horizons with a desire for a real appreciation (or postponing a depreciation). Postponing a depreciation in order to sustain an appreciated real exchange rate artificially increases incomes, which might be politically useful. 13 Because the application of recent macroeconomic political economy to exchange rates is almost non-existent, it is difficult to know what exactly to expect. Shorter political time horizons could conceivably be associated with a propensity to depreciate, so that less stable governments and those facing imminent elections would be more likely to devalue. But they might also be associated with real appreciations and the delay of devaluation, as weak governments try to prop up voters purchasing power with an appreciated currency. The problem is that both the income and the substitution effects of currency movements are potentially politically relevant: voters might like the increase in income brought on by delaying a depreciation, but they might also like the increased international competitiveness brought on by a depreciation. Which effect dominates politically presumably varies among countries and over time. The partisan connection is no clearer. If Left parties are associated with higher inflation, they would also probably be associated with general exchange rate weakness. By the same token, labor unions may be prone to devaluations for their impact on the competitiveness of basic industrial products, in which unions have long found a traditional base. It is striking that the vast literature on the politics of monetary policy has not been brought together more with the equally vast literature on European currency problems, despite the fact that the most prominent monetary issues of the past decades have had to do precisely with European currencies. 14 This obscures the implications of the political economy approaches
13 for exchange rate policy. Nonetheless, these sorts of political economy factors are worth evaluating in the European context. Distributional (particularistic) interests. This is an especially underdeveloped area. In fact, much of the received wisdom in macroeconomics and political economy rejects the view that there are constituencies for and against fixed currencies. Macroeconomists tend to believe that the distributional effects of currency regimes (fixed or floating) are either unclear, or vanishingly small, or both (see Giovannini 1993 for an example). Whether or not this is true in the steady state, it ignores the problem faced by countries that fix when inflation rates differ substantially, which is usually the case. Almost every attempt to fix exchange rates in the last 25 years has involved substantial real appreciations, with equally substantial distributional implications. Even in the steady state, it is not obvious that volatility is distributionally neutral, both in general and with regard to exchange rates; at the very least, this is a hypothesis for which clear evidence has not yet been presented. 15 Many political scientists, for their part, believe that substantial collective action problems preclude serious politicking over currency values (Gowa 1988 is a classic statement). However, the extraordinary political prominence of exchange rates in history and today seems to call the assertion into question. From the 1860s until the 1930s, the gold standard was a major, and mass, political issue in most countries; and since 1980 exchange rates have been domestic high politics in many developed and developing countries as well. 16 Both the volatility and the level of nominal exchange rates can have substantial short- and medium-term distributional effects. And these effects, and their anticipation, have played an important role in the politics of European currency policy. 17 The tradeoff between exchange rate stability and the freedom to vary the currency s value tends to pit two broad groups against one another, based on how highly they value the two conflicting goals. Both import-competing and exporting firms are helped by depreciation. This was especially relevant in Europe, as most attempts to fix DM exchange rates involved countries with inflation higher than Germany s so that some inertial real appreciation was very common.
14 In this context, reducing the flexibility to change the currency s value put competitive pressure on import-competing and exporting firms. For this reason, I expect opposition to fixing exchange rates to come especially from import-competing and exporting sectors. On the other hand, exchange rate volatility principally affects those with substantial cross-border contractual interests. Foreign investors, lenders and borrowers dislike the unpredictability associated with substantial fluctuations in currency values, which are often not amenable to hedging at longer time horizons. In addition, exporters of goods with limited passthrough that is, goods whose prices to consumers do not fully reflect exchange rate movements, typically due to substantial product differentiation or long-term contracts are also typically harmed by volatility. 18 I expect those with cross-border economic interests to be more oriented toward fixing the value of the national currency. 19 There is one category of firms that can be torn in confusing ways by this tradeoff, manufactured exporters. In general, exporters favor maintaining the exchange rate as an active policy instrument. The exporters and import competers most sensitive to nominal exchange rate levels are those whose product prices are more or less fully passed through, typically standardized products commodities, clothing, footwear, steel. But the impact of the level of the exchange rate is mitigated in the case of industries with little pass-through; an appreciation does not cause an analogous rise in the (foreign-currency) price of exports, nor does a depreciation significantly increase (domestic-currency) export prices. In these instances, the exchange risk is carried by the export-producer, so that currency volatility can be quite costly. A common example is that of automobiles, which are priced to local market conditions. If the yen appreciates against the DM, studies find, Japanese car exporters do not raise German prices out of fear they will lose market share. For this reason, exporters of specialized, productdifferentiated manufactured goods which are typically the most dynamic European exporters are less likely to want a weak exchange rate and more likely to value currency stability. To summarize, then, I expect a distributionally driven division between economic actors who support and oppose fixed rates. In favor will be cross-border investors and financial actors,
15 as well as export-competing producers of specialized manufactured goods. Against fixed rates in favor of maintaining the national ability to depreciate the currency will be producers of standardized import-competing and export goods. This reflects the tradeoff mentioned before, between stability, a predictable currency value, and flexibility for competitiveness, a currency value at which standardized (price-taking) tradable producers find it easy to compete with foreigners. This masks much nuance and complexity, of course. There are firms for which the tradeoff between reduced currency volatility and the loss of exchange rate autonomy is not clear, either because both are important or because neither is important. And I have (largely for the sake of brevity) ignored the interests of nontradable producers, such as public sector employees and small businesses, which typically favor maintaining monetary policy autonomy rather than sacrificing it to stabilize currency values which have little direct impact on them. The principal conclusion is that the choice of the desired flexibility of the exchange rate, and the associated choice of its real level, has prominent enough distributional effects to matter politically. Specifically, an important feature of European currency politics has been that its principal supporters are firms and industries with major cross-border investments, markets, or other business interests; while its principal opponents are producers of standardized importcompeting and export products. In national political debates, this sometimes takes the form of allegations that MI is a tool of big business, or that opposition to MI comes from more backward and uncompetitive sectors. I expect the support of the former for fixing exchange rates to be relatively constant, while the opposition of the latter should increase at times of a real appreciation and associated competitive difficulties for national producers. 20 This distributional aspect of European currency politics has been absent in most analyses of European monetary integration. 21 These, and other, views about sources of European monetary unification will rise or fall on empirical grounds. There is little question that all make logical sense, and are based on
16 explanatory traditions of long standing. It is to such an empirical evaluation of the causes of European currency policies that I now turn. Analyzing European currency politics: Variables of interest for a statistical assessment This section and the next presents statistical assessments of causes of the currency policies of European countries. They examine the movement of national exchange rates against the Deutsche mark from 1972 to 1994, and the impact on these movements of a series of economic and political factors. The results allow for some conclusions. Several purely macroeconomic factors, used largely as controls but also of interest in and of themselves, affect exchange rate policies in expected ways. Specifically, slower growth and a weaker current account are associated with depreciation. Capital controls were associated with more depreciation and more volatility, probably because they are more likely to be adopted in countries with more unstable currencies. Two measures used to proxy distributional considerations are strongly associated with currency policy. One is the importance of a country s manufactured exports to Germany (and its longtime currency partners in Benelux), or alternately to the EU as a whole: the more integrated, the more likely to fix the currency. Another is changes in the trade balance (controlling for the current account): net import surges tend to lead to depreciations. I interpret the former result as consistent with pressures for currency stability from such cross-border economic interests as manufactured exporters, and the latter as consistent with pressures for a competitive exchange rate from import- and export-competers. Membership in the two regional monetary institutions of the period, the snake and the ERM, had only a weak impact on the stance of currency policy. However, both it and central bank independence tended to substantially reduce shorter-term exchange rate volatility. This implies that regional institutions and domestic monetary institutions can help limit currency fluctuations among their members but do not alter substantially the broad character of their macroeconomic policies. While the partisan composition of government had no impact on the
17 longer-run propensity to depreciate against the DM, it did reduce volatility: somewhat surprisingly, Left governments were associated with more stability. Indicators of fit with typical optimal currency area criteria are not associated with a tighter exchange rate link with the DM. Specifically, the similarity of national production or trade structures with those of Germany a good measure of suitability for an optimal currency area seems to have no effect on the propensity to fix with the DM. Most of the traditional political economy factors had limited effects. Weaker and less stable governments those with more parties in the ruling coalition, and with weak majorities or pluralities showed no appreciable effect on policy. Direct electoral effects were assessed with a different (and very simple) method of empirical evaluation. From this it appears that the prospect of elections considerably reduces the propensity of the government to devalue in a macroeconomic context of high inflation, but not otherwise. This is in line with an electoral motive driven by the fact that delaying depreciation in a high-inflation setting has a positive impact on income; it is also conceivable that the political resistance to depreciation in such circumstances is a result of the connection between the exchange rate and credibility. In summary, macroeconomic and distributional forces dominated the making of exchange rate policy. Membership in the snake and ERM, central bank independence, and Left wing governments, were associated with greater short-term currency stability but not with the general orientation of currency policy. The dependent variables. The statistical analyses use two intuitive and computationally simple measures of the trends of national currency values against the Deutsche mark. The first is the annual rate of nominal depreciation, the second the annual coefficient of variation of monthly exchange rates. The former directly measures the general trend of the currency against the DM anchor (all European currencies decline over the period, so there are no appreciating currencies). The latter looks more at volatility itself within each year than at the trend of the currency s value.
18 Table 1 shows these two measures of the stability of European currencies against the Deutsche mark. The table includes the thirteen (pre-emu) EU currencies other than the DM (Luxembourg shares a currency with Belgium), plus that of Norway. 22 The table is divided among four groups: hard-currency countries are those that were always members of both the snake and the ERM, soft-currency countries are those which were not been continual members of either, and intermediate countries are those which were members of the ERM but not the snake. The four non-eu members are shown separately. The simplest way to measure the relationship between exchange rates is the rate of change in their nominal values, in this case the average annual rate of depreciation against the DM, as presented in panel A of Table 1 and graphically in Figure 1. This has the advantage of transparency of interpretation; however, it does not indicate potential currency volatility. For this purpose, the coefficient of variation of national currencies against the DM is presented in panel B of Table 1. 23 The two measures produce very similar classifications of countries and country-years, and when they are used in statistical analysis they give rise to virtually identical results. However, the differences are also interesting, as they pick up (inasmuch as they differ) differences between determinants of broad currency policy and of shorter-term policy toward volatility. The explanatory variables used are of course limited by data availability. They can be categorised along the lines of how I interpret their fit with the factors discussed above, which is by no means uncontroversial. In all instances, recall that the dependent variables are the depreciation rate and volatility against the DM so that a negative (positive) sign means higher values of the explanatory variables are associated with less (more) depreciation and volatility. I start with generally accepted macroeconomic control variables, and move on to the four categories above. Macroeconomic factors. It is important to take into account developments in national macroeconomic performance that can reliably be expected to affect the propensity to allow the currency to depreciate. While the arguments for depreciation in each of these instances are not
19 unproblematic, they are common enough that it seems worth trying to assess or control for such macroeconomic trends. This is especially the case when looking at annual data: particularly difficult years can be expected to be associated with a weaker currency. 1. Growth rates. Recessions may increase the propensity of monetary authorities to use a depreciation to stimulate the economy. This of course depends on the tradeoff between the income and substitution effects of a depreciation, but the consensus is that depreciations can be stimulative in the short run. Variable name: lagged growth rate of GDP (-). Expected sign: negative (i.e. the stronger GDP growth, the less depreciation). (All variables and their construction, along with other details of the data, are described in detail in the Appendix.) 2. Unemployment. This can be expected to be of significance for the same reason as the overall rate of economic growth. Variable name: lagged unemployment (+). Expected sign: positive. 3. The current account. The weaker a country s current account, the more difficulty it is likely to face in defending its currency and the likelier a depreciation. Variable name: lagged current account balance as percent of GDP (-). Expected sign: negative. 4. The terms of trade. In this case, it is the difference between movements in the country s terms of trade and those of Germany that are expected to affect the currency. The more the country s terms of trade deteriorate relative to Germany, the harder it should be to sustain a fixed exchange rate. As measured, a positive number here means that the terms of trade improved in the year relative to Germany s, while a negative number means they deteriorated. This implies that increases in the measure should make it easier to sustain the currency peg, and vice versa. Variable name: difference in terms of trade relative to Germany (-). Expected sign: negative.
20 As can be seen from the variable names, all these are lagged one year except for the terms of trade figure. This is because policy can be expected to respond to such macroeconomic trends only with something of a delay, except for the terms of trade which is a price-based measure and thus should have nearly immediate effect. In any case, using simultaneous (lagged, in the case of the terms of trade) data makes no difference to the results. The current account is expressed as a percentage of GDP, unemployment is share of the labor force, GDP growth is of course a rate of (real) change, and the terms of trade are also a rate of change; all are expressed in percentage points. Efficiency considerations. These are factors associated with optimal currency area and related criteria. It is notoriously difficult to measure factor mobility across countries, and although indicators of financial integration do exist (Frankel and Wei 1995) they are available only after about 1980. An easier approach is to look at structural characteristics of national economies that imply, in the OCA framework, a reduced need for or possibility of independent monetary policies. Here I use the following measure: 1. The correlation of a nation s industrial structure with that of Germany. This should indicate how different the exogenous shocks affecting the two countries are likely to be. Variable name: industrial correlation with Germany (-). Expected sign: negative. Other related measures might be used. The correlation of a nation s trade structure with that of Germany has attractions (as it is more directly related to pressures on the exchange rate), but risks endogeneity, as trade structure is much more likely to be affected by exchange rate policy than overall industrial structure. In any case, the two measures are highly correlated and give nearly identical results. Other measures of optimal currency area criteria (see, for example, Gros 1996) tend to give rise to very similar categorizations of countries. In the case of the measure of industrial structure, the greater the correlation with Germany the more likely the country is, by optimal currency area criteria, to maintain a fixed exchange rate with the Deutsche mark.
21 Although credibility has been adduced as an important potential source of national desire to link to the DM, it is hard to imagine any clean measure of the demand for credibility. Of course, poor macroeconomic performance implies a greater need for credibility; but it also implies a higher cost of achieving it. I return to credibility factors below, as they arise in the interpretation of the results. Regional institutions. Ideally, we would like some measures of such things as overall commitment to European integration, or the presence of linkage politics. A more basic example of such regional factors is simply national membership in the EU institutions that have been devoted to monetary integration. Presumably, if EU institutional factors have been important, they would show up here in one way or another. 1. Membership in the snake or EMS. Of course this is endogenous, but many believe that the snake and EMS as international (regional) institutions may have had a substantial independent impact on government behavior. This is a dummy variable that takes the value 1 if the country was a member of one of the two exchange rate mechanisms, 0 otherwise. Variable name: member of snake or ERM (-). Expected sign: negative. Macroeconomic political economy. Electoral, partisan, legislative, and administrative characteristics of national political systems are commonly incorporated into political economy explanations of macroeconomic policy. Their impact on the exchange rate is less well developed, as discussed above, but some reasonable expectations might be as follows. 1. Partisan effects. To the extent that the Left is more inflation prone than the Right, we expect the Left to be associated with a depreciating the currency; and some scholars in fact regard this to be the case independent of inflationary propensities. The variable used here measures the partisan (Left-Right) nature of the cabinet in power; parties are coded on a widely accepted scale and weighted according to their importance in the cabinet. In this scale, lower numbers are more to the Left. (Alternate measures of the legislative center of gravity, or the government s ideology,
22 which use similar scales, yield nearly identical results.) Variable name: cabinet center of gravity (-). Expected sign: negative. 2. Election timing. In the spirit of the political business cycle, governments may be expected to manipulate the currency in the runup to an election. What in fact they do depends on the relative desirability of the stimulative effect of depreciation, and the income effect of an appreciation. However, the traditional view of inflation and depreciation as similar in source and effect would lead us to expect elections to be associated with depreciations. The measure here is simply whether an election occurred in the year in question, which has its problems (again, I use different methods to assess the impact of elections below). Variable name: Election (+). Expected sign: positive. 3. Government stability. It is a commonplace of macroeconomic political economy that less stable and/or more fragmented governments have a propensity to inflate, which implies to depreciate the currency. As mentioned above, for most this is due to the shorter time horizon of policymakers; for others, to the unclear responsibilities for imposing costs and benefits. I use two measures, which are not closely related in institutional terms. The first is the share of all legislative seats held by the governing coalition, which indicates roughly the security of the government in office. (A measure that uses share of all votes gives the same results.) The bigger this seat share, the more stable the government and the less likely a depreciation. The second measure is the number of parties in government, which gives a rough sense both of stability and of the assignment problem mentioned; more parties in government should increase the propensity to depreciate. 24 Variable names: Percent of seats held by government parties, number of government parties (-, +). Expected signs: negative, positive. 4. Central bank independence. Inasmuch as the independence of the central bank is associated with lower inflation, and lower inflation facilitates the maintenance of a
peg with low-inflation Germany, this should also be tied to less depreciation against 23 the DM. 25 The measure used is that created by Cukierman, Webb, and Neyapti (1992) in their influential study. Variable name: central bank independence (-). Expected sign: negative. 5. Capital controls. Controls on capital movements should facilitate the maintenance of a fixed exchange rate. Of course, countries whose exchange rates face market skepticism for other reasons such as macroeconomic fundamentals or political instability are more likely to impose capital controls in the first place, so it may not be clear what to expect. However, in general it seems consistent with the literature to expect countries with capital controls to be less likely to depreciate, all else equal. The measure used is a composite created by Dennis Quinn and drawn from the IMF s categorization of restrictions on capital movements. Variable name: capital controls (-). Expected sign: negative. These variables are all roughly comparable to the sorts of factors commonly adduced to explain inflation in work on macroeconomc political economy (Alesina 1989 and Grilli, Masciandaro, and Tabellini 1991 are classic examples). Inflation-based precursors lead us to expect Left, unstable or fragmented governments, impending elections, dependent central banks, and the absence of capital controls to be associated with more depreciation. Evaluating the validity for exchange rates of these expectations, developed almost exclusively in reference to inflation, has two implications. First, it checks the applicability of these typically closedeconomy models to an open-economy setting. Second, independently of the relationship between inflation and the exchange rate, it can help tell us something about the impact of political institutions on macroeconomic policy. Private interests. The distributional effects of exchange rate politics, and thus the interests of private agents, have been underemphasized as sources of variation in European monetary integration. Attempts to evaluate this claim are hampered by the general unavailability of data on special interests. This is especially the case inasmuch as arguments to this effect rely