De facto monetary policy objectives in developing and emerging countries. David Cobham

Transcription

1 De facto monetary policy objectives in developing and emerging countries David Cobham Heriot-Watt University, Edinburgh preliminary version, May 2015 Abstract A statistical methodology is used to determine the de facto monetary policy objectives of developing and emerging countries since It shows that these countries (particularly developing) continue to use single exchange rate objectives to a large extent, but have also shifted (particularly emerging) towards a single objective of inflation and multiple objectives subject to a price stability constraint. The latter type of objective turns out to be associated with the lowest inflation rates, but the different objective types do not seem to affect economic growth. While specific attention to inflation produces lower inflation at no cost in terms of growth, single exchange rate objectives (where appropriate anchor currencies exist) continue to offer a decent alternative for countries which have problems with data availability. JEL: E52, E58, F33 Keywords: monetary policy, de facto, exchange rate, inflation targeting, multiple objectives, developing countries, emerging countries

2 1 Introduction While there is a considerable literature on de facto exchange rate regimes which identify what governments (or central banks) do rather than what they say they do with respect to their exchange rates little attention has been paid to the issue of de facto monetary regimes. Analyses of the effects of inflation targeting on inflation, for example, typically use de jure identifications of the inflation targeters, occasionally adding the US and/or the Eurozone as informal inflation targeters. This paper sets out to identify not full monetary regimes which would include the way policy is formulated and implemented together with its effects on expectations - but what the monetary authorities have actually succeeded in stabilising in different countries and at different times. Possible policy objectives can be formulated in terms of intentions to stabilise the growth rates (which could be zero or non-zero) of variables such as money, prices and exchange rates. It is assumed that a policymaker aiming to stabilise only or above all the rate of inflation, for example, will generate a growth rate of prices which has been stabilised both within some specific margins and more narrowly than that of any other relevant variable. In that case the variable successfully stabilised may reasonably be regarded as having been the single objective of policy. On the other hand, a policymaker may take the view that he or she should not focus solely on one objective, in which case they may be thought of as trading off between objectives or pursuing multiple objectives in the way envisaged by Theil, as opposed to the n-objectives/n-instruments policymaker portrayed by Tinbergen (Shaw, 1977). The methodology used here is essentially that of Cobham (2015), which analysed single and multiple de facto objectives in a sample of around 30 advanced economies. In that paper the emphasis was on the choice between single and multiple objectives, and the aim was to shed light on the issue of whether a monetary policymaker which pursued more than one objective 1

3 notably if he or she adopted a leaning against the wind strategy which reacted to sharp and sustained movements in asset prices was likely to produce much worse performance in terms of inflation and economic growth (the evidence suggested otherwise). In this paper the concern is more with the evolution of monetary policy possibilities and objectives over time in both emerging and developing countries. Exchange rate fixity turns out to be much more important (and price stabilisation less important) in this sample of countries, and the global financial crisis had much smaller effects on policy choices, but some of the comparisons between the different types of objective are comparable to those in advanced economies. Section 2 presents the methodology to be used. Section 3 notes some of the data limitations before implementing that methodology for the 122 emerging and developing countries considered, 1 where it distinguishes between five different subperiods. Section 4 splits the sample into 28 emerging and 94 developing countries, but focuses on just two larger subperiods: and Section 5 presents first conditional and then unconditional analyses of the economic performance associated with each of the twelve types of monetary policy objective considered. Section 6 discusses the findings and concludes. 2 Methodology The distinction between de jure and de facto classifications of exchange rate regimes dates largely from the findings of Calvo and Reinhart (2002) to the effect that a number of developing countries that claimed to having floating rates were in fact intervening much more in foreign exchange markets (and manipulating their interest rates for exchange rate reasons) than high-income countries such as the US, Japan and Australia which were widely recognised as free-floaters. Of the two main de facto classifications which have been developed, that of Reinhart and Rogoff (2004) uses pre-announced targets as well as 2

4 statistical data over rolling windows of five years: the classification is more clearly one of regimes and is partly de jure as well as de facto. That of Levy Yeyati and Sturzenegger (2005), on the other hand, uses purely statistical data for 12-month calendar years to identify fixed, crawling, dirty-floating and fully flexible exchange rates. 2 The focus in this paper, as already noted, is not on more widely-defined regimes but on short term stabilisation achievements, and the methodology is therefore closer to that of Levy Yeyati and Sturzenegger. However, the concern is with monetary policy, not just exchange rates, so the possible objective variables are taken to include broad money, real income, nominal income and prices, as well as exchange rates against a range of possible anchor currencies. The key metric is the standard deviation of the four quarter growth rates of the possible policy variables. In each case the successful pursuit of a single objective will be identified by a relative criterion its growth rate must have been stabilised more narrowly (around a zero or non-zero trend) than that of any other possible objective variable and an absolute criterion: the standard deviation of its growth rate must have been below some threshold level. These thresholds, which will be presented below, will be drawn sufficiently tightly that it will be highly unlikely that an objective is justified just by chance. 3 Because there are a range of alternative anchor currencies to which countries can peg, it is necessary to identify first which anchor, if any, a country may be pursuing before bringing that objective into comparison with the domestic objectives. For that reason the analysis starts by comparing the different possible domestic objectives formulated as inflation, real income growth, nominal income growth and monetary growth and (separately) identifying the relevant possible external anchors dollar, Deutsche mark, French franc, ecu and euro before combining the domestic and external results. 3

5 Finally, it is assumed that the monetary authorities could always, if they really wanted to, control a single objective variable (though the costs might be high), so that if they do not do this they are implicitly trading off between more than one objective, that is, pursuing multiple objectives. However, the very large category of country/year cases where no single objective can be identified clearly covers a wide range of monetary policy strategies and experiences, from the well-structured and disciplined to the poorly organised and undisciplined. It therefore makes sense to distinguish between multiple objectives subject to a price stability constraint and multiple objectives pure and simple. This is roughly comparable to Reinhart and Rogoff s (2004) distinction between freely floating and freely falling exchange rates, in which they use the criterion of inflation > 40% to identify the latter category and keep countries in that situation separate from genuine floaters with low inflation. The initial data are in nearly all cases quarterly data taken from the IMF s International Financial Statistics (IFS). Monthly data are typically available for inflation and exchange rates, but only quarterly data are typically collected for real or nominal GDP. For each calendar year, data on the four quarters of that year, together with the last quarter of the previous year and the first quarter of the following year, are used to infer the type of objective pursued in that year. The use of the two quarters preceding and succeeding the calendar year ensures an overlap which avoids sharp end-year changes going unnoticed, while standard deviations measured over six rather than over four quarters are less affected by individual outliers. On the domestic side, data on the percentage growth rates since four quarters before of consumer prices, real GDP, nominal GDP and broad money are used to calculate standard 4

6 deviations of growth rates over the six quarters. These standard deviations are then used to identify single de facto objectives as follows: type of objective relative criteria threshold criterion single objective of inflation SP single objective of real income growth SY single objective of nominal income growth SPY single objective of monetary growth SM σ π < each of σ gy, σ gpy, σ gm σ π < 0.5 σ gy < each of σ π, σ gpy, σ gm σ gy < 0.5 σ gpy < each of σ π, σ gy, σ gm σ gpy < 0.5 σ gm < each of σ π, σ gy, σ gpy σ gm < 1 where σ indicates standard deviation, π is inflation, gy is real income growth, gpy is nominal income growth, and gm is monetary growth. The idea behind the relative criteria is that if the policymaker is seriously pursuing a single objective, he or she must be stabilising its growth more than that of any of the other variables. The effect of the threshold criteria for SP, SY and SPY is that (under a normal distribution) the growth rate of the variable must in 68% of cases be no more than 0.5 away from the mean (e.g. in 68% of cases inflation must be no more than 0.5% away from its mean), and it must in 95% of cases be no more than 1 away from the mean; this is comparable to the numerical targets typically adopted by formal inflation targeters. 4 The corresponding threshold criterion for monetary growth is set at 1, on the grounds that monetary growth is inherently less stable (or inertial) in the short run than the growth of prices or income; even so, as will be shown below, relatively few cases of SM are identified. On these criteria it turns out that in many country/year cases policymakers have not been pursuing single objectives, from which it is inferred that de facto they have been following a 5

7 number of different objectives, but these cases vary widely in terms of policy operation and performance. Given the primacy typically accorded in theoretical and policy analysis to price stability, the following simple distinction would seem to be appropriate: type of objective multiple objectives subject to a price stability constraint MC multiple objectives (unconstrained) MU criterion no single objective threshold criteria fulfilled but average inflation < 5% no single objective threshold criteria fulfilled, and average inflation 5% On the external side, data on exchange rates of currencies against the US dollar, the Deutsche mark (DM) and the French franc (FF) up to the end of 1998, the ecu (data are available from 1979 Q3 to 1998 Q4) and the euro (from 1999) are used to generate four quarter percentage growth rates, from which the standard deviations over the six quarters are calculated. 5 These standard deviations are then used to identify single de facto exchange rate objectives as follows: type of objective relative criteria threshold criteria single objective of exchange σ gusd < each of σ gdm, σ gff, σ gecu, σ gusd < 1 rate against dollar SUSD σ geuro, σ grand, σ grup as appropriate in different subperiods single objective of exchange σ gdm < each of σ gusd, σ gff, σ gecu, σ gdm < 1 rate against Deutsche mark SDM σ grand, σ grup as appropriate in different subperiods single objective of exchange rate against French franc SFF σ gff < each of σ gusd, σ gdm, σ gecu, σ grand, σ grup as appropriate in different subperiods σ gdm < 1 6

8 single objective of exchange rate against ecu SECU single objective of exchange rate against euro SEURO single objective of exchange rate against South African rand SRAND single objective of exchange rate against Indian rupee SRUP σ gecu < each of σ gusd, σ gdm, σ gff, σ gecu < 1 σ grand, σ grup σ geuro < σ gusd, σ grand, σ grup σ geuro < 1 σ grand < each of σ gusd, σ gdm, σ gff, σ gecu, σ geuro, σ grup as appropriate in different subperiods σ grup < each of σ gusd, σ gdm, σ gff, σ gecu, σ geuro, σ grand as appropriate in different subperiods σ ggbp < 1 σ ggbp < 1 where gusd, gdm, gff, gecu, geuro, grand and grup represent respectively the growth rates of exchange rates against the dollar, the mark, the French franc, the ecu, the euro and the Indian rupee. The penultimate row relates only to Lesotho, Namibia and Swaziland, which had hard pegs to the South African rand throughout the period within the Common Monetary Area, so that their σ grand was equal to 0; and the final row only to Bhutan, whose currency was interchangeable with the Indian rupee throughout the period, so that its σ grup was equal to 0. The relative criteria here ensure that a government or central bank has stabilised its exchange rate on one anchor currency more closely than on any other. The threshold criteria require that 68% of the time the exchange rate growth must be no more than 1 percentage point away from its mean, and 95% of the time it must be no more than 2 away from the mean. This is less strict than the threshold criteria used for the domestic objectives other than money, but exchange rates are inherently more volatile than most of the domestic variables, and these criteria are comparable to the exchange rate bands used in the Bretton Woods system from late 1971 and in (the narrow band of) the European Exchange Rate Mechanism from 1979 to 1993 (2.25% in either direction). The criteria here do not include policy interest rates (which are typically more stable than exchange rates), and they do not include foreign exchange reserves, as discussed above. However, the thresholds are sufficiently strict that they identify, as will be shown below, rather few unexpected cases of de facto exchange rate objectives in the sample, which suggests that the method is unlikely to be picking up spurious examples of 7

9 chance exchange rate stability. Cases where no de facto exchange rate objective is being pursued are labelled as follows: type of objective no exchange rate objective NER criterion no single exchange rate objective threshold criterion fulfilled The final step is to choose between domestic and external objectives where both have been identified for a particular country/year. This is done by comparing the standard deviation of the domestic and external objectives identified for each country/year to find which is lower (and for this purpose the standard deviation of inflation is used for MC as well as for SP). The algorithm used also ensures that a country/year case where there is a domestic objective of some kind but no external objective is classified in terms of the domestic objective, while a case where there is MU in terms of the domestic objective but a clear exchange rate objective is classified in terms of the latter. The overall default category is MU, no clear domestic or external objective and average inflation 5%. 3 Implementation The availability of data is a serious issue here: for many of the countries in the sample quarterly GDP data, and in a few cases quarterly monetary data, are available only from well after the starting date (1973), as is shown in Appendix Table A1 which lists the first calendar year in each case for which consistent quarterly data can be found in IFS. For country/year cases where quarterly data for GDP (or money) are not available it is assumed that the policymakers could not have been trying to stabilise real or nominal income (or monetary growth) because of the lack of data. The lack of quarterly GDP data must also make inflation targeting and even monetary targeting more difficult. But this implies that a large number of 8

10 countries have quite limited monetary policy choices: Figure 1 shows the number of countries, out of the 122 in the sample, which had consistent quarterly data of each kind exchange rate, inflation, broad money, real GDP, nominal GDP in 1971 and in the closing years of each of the subperiods used below. By 1998, only 33 countries had quarterly real GDP data and 34 nominal GDP data. Even by 2013 these numbers had risen to only 52 and 50, while only 47 countries had quarterly data for all five indicators (that 47 included 24 out of the 28 countries designated below as emerging countries). 6 Table 1 shows the domestic objectives identified initially (before consolidation with the external objectives), in five subperiods: , , , and The first three subperiods are divided more or less equally; 1999 sees the disappearance of the DM, the FF and the ecu as anchors and their replacement by the euro; and 2008 marks the start of the global financial crisis. There are a number of cases where no objective can be defined, because the data or the country do not exist (e.g. Belarus pre-1996); these cases are denoted by X. It is clear that the dominant objective type overall is MU, multiple objectives unconstrained (inflation 5%), but there are also large numbers of MC, multiple objectives subject to a price stability constraint; and smaller numbers of SP, single objective of inflation, SM, single objective of monetary growth, and SY, single objective of real income growth (but hardly any of SPY, single objective of nominal income growth). Over the first three subperiods there are some rises in the frequency of SP and MC (and to a lesser extent SM), but it is clear that the major change is between the third and fourth subperiods, with more SP, much more MC, fewer MU and some SY cases post

11 Table 2 shows the external objectives identified, before consolidation with the domestic objectives. Here there are a large number in every subperiod of NER, no exchange rate objective, together with large frequencies of SUSD, single objective of exchange rate against dollar, SFF, single objective of exchange rate against the French franc, and from 1999 SEURO, single objective of exchange rate against the euro. The last two subperiods see the replacement of the European legacy currencies as anchors by the euro, together with a small decline in the frequency of NER. Table 3 shows the objectives as identified after the consolidation of domestic and external analyses. The main effects of the consolidation are that the frequencies of MU and MC are now lower, because many of these countries are now identified instead as having an exchange rate objective. There is also some reduction in the number of SP cases. On the other hand, there are few reductions in the exchange rate objectives identified (that is, few of these are displaced by domestic objectives). Overall, it is clear that for this sample of countries the dominant objective types are MU and SUSD, together with SFF pre-1999 and SEURO and MC from Of the domestic single objectives, SP is the most important, particularly in the last two subperiods. It is also worth noting that the average duration of the objective types that is, the number of years for which the same objective persists in a given country varies widely: for SFF it is over 10 years, for SEURO over 9 years, for SUSD over 5 years, for MU around 5 years and for all the other objective types 2 years or less. 4 Emerging versus developing countries Tables 1-3 indicate strongly that the big change is between the first three and the last two subperiods, both in terms of the anchor currencies and in terms of the frequency of domestic and multiple objective types. On the other hand it is of interest to distinguish between 10

12 emerging and developing countries. Table 4 presents the objective data, therefore, for two larger subperiods and for emerging and developing countries, where emerging countries are broadly those identified as emerging in Laurens, Arnone and Segalotto (2009), and developing are the remainder. 7 In the first large subperiod, 1973 to 1998, the emerging economies have more (in percentage terms) of each of SP, SM and MU than the developing countries, but less of SUSD and much less of SFF. In the second large subperiod, , the balance shifts for both groups from MU to MC, but more strongly so for the emerging economies. SFF is slightly more than replaced by SEURO for the developing countries, but for the emerging countries the frequency of SEURO is much higher, mainly for Bulgaria, Estonia, Latvia and Lithuania which had mostly been MU or SUSD (or, earlier, X) before. On the other hand, the incidence of SUSD falls a little for the emerging but rises a little for the developing countries. There is also a much larger rise in the frequency of SP for the emerging than for the developing countries, and an important element of SY for the former but not the latter. These trends are illustrated by Figures 1 and 2, which show the changing proportions of each objective type over time for the emerging and developing country groups respectively. The advanced economies covered in Cobham (2015), on the other hand, had as shown in Table 5 a much higher incidence of each of SP and MC, a much lower incidence of MU, and a markedly lower incidence of single exchange rate objectives (10.3% in total for , and 2.5% in , as against 19.6% in and 25.9% in for the emerging country group, and 39.9% in and 42.9% in for the developing country group). 8 In addition, Table 5 shows a marked decline in SP in response to the crisis, which could be regarded as a sensible response by advanced economy policymakers to the depth of the crisis and the ensuing recession; in the developing and emerging countries there is no 11

13 such decline (Table 3 shows a small rise), which can be attributed in part to the fact that the crisis and recession hit those countries much less hard, at least in the immediate period. It is also useful here to consider a few examples, referring to the full data set out for every country in Appendix Table A2. Among countries known to fix their exchange rates, Bahrain shows up as SUSD since 1975 while Jordan has been SUSD since 1997, and both of these are in line with conventional views, for example those of Ilzetski, Reinhart and Rogoff s (2011) updated country chronologies. The CFA countries show up as SFF in every year to 1998 except when they were mainly MU (they devalued against the franc in January 1994 and that change remains in the four quarter growth rates until 1994 Q4, which is included in the calculation for 1995), but SEURO from 1999 onwards. China (mainland) is mainly MU in the 1970s and 1980s but SUSD from 1996 to 2004 and more mixed since then (it is also SUSD in 2006 and 2011): this is broadly consistent with Ilzetzki et al. (2011). India is mainly MU to the mid-1990s and since then a mix of SM, SP, MC, MU and SY. Argentina was SUSD from 1993 to 2000 (the period of its currency board on the dollar), but since then mostly MU. Mexico, Peru and Chile have been SP or MC most of the time since they formally adopted inflation targeting (in 2001, 2002 and 1999, respectively): in these countries inflation has typically fluctuated around 4-5% and has been considerably more variable than that for advanced (formal or informal) inflation targeters such as the UK, the US and the Eurozone (Cobham, 2015). These examples clearly suggest that the methodology is making broadly plausible identifications of de facto monetary policy objectives. 5 Economic performance Table 6 shows the inflation and growth performance associated with the different objective types. 9 For inflation it is striking that the lowest numbers by some distance are for MC, 12

14 multiple objectives subject to a price stability constraint, for both subperiods and both groups of countries, while the highest are for MU, unconstrained multiple objectives. In the second large subperiod the inflation associated with SUSD and SEURO is not bad, but that for the exchange rate objectives in the first large subperiod is less good. The single domestic policy objectives (SP, SM, SY and SPY) have much better performance than MU but less good than MC. For growth the picture is more mixed, partly because there is less variation in the data. Growth under MC is mostly around under the average, while growth under SP is generally a little above, and growth under some but not all exchange rate objectives is at or above the average. 10 The countries concerned differ widely and in many respects, notably in their underlying growth rates. It is therefore essential to control for relevant factors (along with country fixed effects). Tables 7 and 8 present the results of regressions of first inflation and then the growth of real GDP per capita on standard explanatory variables: openness and monetary growth for inflation; openness, population, population growth, average schooling (of over-25s) and the investment/gdp ratio for growth. 11 Results are reported for emerging and developing countries, and for each of the large subperiods. The aim of the exercise is to identify the effects of the different objectives, rather than to provide full explanations of the determination of inflation or growth. The lack of data means that, particularly for , the numbers of observations used in the regressions are well below the totals in Table 4. In addition, to avoid the results being dominated by outliers, the relatively small number of observations where inflation exceeds 40% (the cut-off for freely falling as opposed to freely floating in Reinhart and Rogoff, 2004) are excluded. 13

15 In Table 7 the control variables are mostly significant, with coefficients not far from those in Cobham (2015). Of the objective types, the high-frequency single exchange rate objectives SUSD, SFF (for developing countries) and SEURO are associated with significantly lower inflation than the default objective, MU, which is consistent with findings elsewhere (Tavlas et al., 2008). MC is associated with lower inflation than the other objectives (except for SFF in developing countries in the first subperiod and SPY in emerging countries in the second subperiod, when there were only two cases of SPY), but SP does consistently better than SM. The growth regressions of Table 8 are less satisfactory. For the developing country group none of the control variables and none of the objective types are significant for the first subperiod, while only the investment ratio and MC are significant in the later subperiod. For the emerging countries only the investment ratio and SFF are significant in the earlier subperiod (and SFF is significantly negative, but there were only two such cases). For the later subperiod population and population growth are also significant, but none of the objective types are. These simple regressions are therefore poor in explaining the variations in growth in these countries (where output is consistently more volatile than in advanced economies, as discussed, for example, by Koren and Tenreyro, 2007). However, it is reasonable to conclude that the different objective types are not clearly or strongly associated with different rates of growth Discussion and conclusions The basic features and trends in the de facto monetary policy objectives of developing and emerging countries revealed by the analysis are these: a high frequency of single exchange rate objectives (pegs to the US dollar, French franc and then euro); a high frequency of 14

16 multiple objectives, particularly in the years up to 1998, but some decline from 1999 together with a shift from unconstrained multiple objectives MU to price stability constrained multiple objectives MC, a shift that is stronger for the emerging countries; and a much lower incidence of domestic single objectives (for prices, money, real GDP or nominal GDP) before, but a little higher after, The big differences from the advanced countries are that the developing and emerging countries have far more exchange rate objectives and follow a slower and weaker trend towards the single objective of inflation SP and/or MC. However, the emerging countries, in particular, have a lot of MC in the years from 1999, so the associations of MC and other objective types with levels of economic performance remain important. Here the basic findings are close to those for the advanced countries: MC is associated with the lowest inflation (in both subperiods and for both country groups), and there seem to be no significant differences between objectives in terms of economic growth. The implication is that policymakers should not be obliged to confine their attention to single objectives (such as inflation) but can reasonably consider other objectives as well. Finally, while a specific focus on inflation (in MC or SP) seems to ensure lower inflation, it does so at no cost in terms of economic growth. However, since many developing countries are impeded by the lack of good statistical data of the kind that would be required for inflation forecasting, fixing the exchange rate continues to offer a decent alternative, although at that point more country-specific information is needed about international trade and investment patterns and other things before sound choices can be made. 15

17 Notes 1 These are the full sample of emerging and developing countries in the IMF s International Financial Statistics, plus Estonia and Hong Kong, minus small countries (population < 250,000) and countries for which the data were completely inadequate. 2 See Tavlas et al. (2008) for a survey of the issues involved in exchange rate classification and the findings on economic performance under different methods. 3 Calvo and Reinhart (2000) and Levy Yeyati and Sturzenegger (2005) addressed this issue by examining data on the use of the policy instruments (foreign exchange reserves, also interest rates). Complications could arise if this was done here from the stability of policy interest rates, particularly given the focus on individual calendar years and the inclusion in the sample of many countries with relatively underdeveloped financial sectors and policy rates that are stable over long periods. 4 Of the sixteen emerging and developing countries whose inflation-targeting details are given in Hammond (2012), eight (Chile, Guatemala, Indonesia, Mexico, Peru, Philippines, Poland and Romania) had targets specified as x ± 1%; two (Serbia and Thailand) had targets of x ± 1.5%, and three (Brazil, Ghana and Turkey) had targets of x ± 2%, while two others specified ranges (Colombia 2-4%, South Africa 3-6%). The threshold used here is therefore slightly tighter than some emerging/developing inflation targeters targets, but it is convenient to retain this for comparability with the advanced country results in Cobham (2015). 5 At start- or end-points for different anchor currencies, e.g. the replacement of the DM by the euro at the start of 1999, the standard deviations are calculated over five rather than six quarters. 16

18 6 The issue of data availability for countries moving to adopt inflation targeting or otherwise modernising their monetary policies is a serious one. See Al-Mashat (2010, p. 249), for example, on the role of data improvement in Egypt s journey towards IT. 7 The list of emerging countries used here, which is set out in Appendix Table A1, excludes from Laurens et al. s list the Czech Republic, Israel, Slovakia and Slovenia, which are all counted by IFS as advanced countries and treated as such in Cobham (2015), but includes Estonia and Hong Kong. 8 It should be noted that in that study the Eurozone is treated as a single currency area, and the countries adopting the euro drop out of the analysis from the date of adoption. 9 Where quarterly data on inflation or real GDP growth are not available but annual data are, the latter are used. However, for some countries in the earlier years not even annual data are available, in which case the averages exclude those cases. For Zimbabwe the inflation data for 2008 are rising exponentially to a rate that completely dominates the average data, but only two quarters of data are available for that year (and none thereafter), so the number for 2008 has also been excluded. 10 There are a few odd results for objectives which have very low frequency e.g. growth under SFF, emerging countries, 2 cases only and these should probably not be given much importance. 11 See Ghosh, Gulde and Wolf (2002) and Husain, Mody and Rogoff (2005) for comparable analyses of the association of exchange rate regimes (under various different classifications) with inflation and growth, and Cobham (2015) for comparable regressions to those here for advanced economies. 12 See also Ghosh, Gulde and Wolf (2002) for the result that the link between the exchange rate regime and growth was much less robust than the link between regime and inflation (p. 206). 17

19 References Al-Mashat, R. (2011), The evolution of monetary policy in Egypt and steps towards inflation targeting, in M. Boughzala and D. Cobham (eds), Inflation Targeting in MENA Countries: An Unfinished Journey, Basingstoke: Palgrave Macmillan Calvo, G., and Reinhart, C. (2002), Fear of floating, Quarterly Journal of Economics, 107, Cobham, D. (2015), Multiple objectives in monetary policy: a de facto analysis for advanced countries, Manchester School, forthcoming Ghosh, A., Gulde, A-M., and Wolf, H. (2002), Exchange Rate Regimes: Choices and Consequences, Cambridge, Mass.: MIT Press Hammond, G. (2012), State of the art in inflation targeting, Handbook no. 29, Centre for Banking Studies, Bank of England Husain, A., Mody, A., and Rogoff, K. (2005), Exchange rate regime durability and performance in developing versus advanced economies, Journal of Monetary Economics, 52, Ilzetzki, E., Reinhart, C., and Rogoff, K. (2011), The Country Chronologies and Background Material to Exchange Rate Arrangements into the 21 st Century: Will the Anchor Currency Hold?, available at Country_Chronologies_2011.pdf Koren, M., and Tenreyro, S. (2007), Volatility and development, Quarterly Journal of Economics, 122(1), Laurens, B., Arnone, M., and Segalotto, J-F. (2009), Central Bank Independence, Accountability and Transparency: A Global Perspective, Basingstoke and New York: Palgrave Macmillan 18

20 Levy-Yeyati E. and Sturzenegger, F. (2005), Classifying Exchange rate Regimes: Deeds vs. Words, European Economic Review, 49, Reinhart, C., and Rogoff, K. (2004) The modern history of exchange rate arrangements: a reinterpretation, Quarterly Journal of Economics, 119, 1-48 Shaw, G.K. (1997), Introduction to the Theory of Macroeconomic Policy, 3 rd edition, London: Martin Robertson Tavlas, G., Dellas, H., and Stockman, A. (2008), The classification and performance of alternative exchange-rate systems, European Economic Review, 52,

21 Appendix Table A1: Countries, emerging or developing, availability of quarterly data Emerging? exchange rate broad money inflation Bangladesh Bhutan real GDP nom GDP Brunei Cambodia China,P.R.: Mainland X China,P.R.:Macao China,P.R.:Hong Kong X Fiji India X Indonesia X Lao People's Dem.Rep Malaysia X Maldives Mongolia Myanmar Nepal Papua New Guinea Philippines X Solomon Islands Sri Lanka Thailand X Timor-Leste Vietnam Albania Bulgaria X Croatia X Estonia X Hungary X Latvia X Lithuania X Macedonia, FYR Poland X Romania X Serbia, Republic of Turkey X Belarus Georgia Kazakhstan Kyrgyz Republic Russian Federation X Ukraine Algeria

22 Bahrain, Kingdom of Egypt X Iran, I.R. of Iraq Jordan X Kuwait Libya Mauritania Morocco X Oman Pakistan X Qatar Saudi Arabia Syrian Arab Republic Tunisia Yemen, Republic of Cameroon Central African Rep Chad Congo, Republic of Equatorial Guinea Benin Burkina Faso Côte d'ivoire Guinea-Bissau Mali Niger Senegal Togo Angola Botswana Burundi Cabo Verde Ethiopia Gambia, The Ghana Kenya Lesotho Liberia Madagascar Malawi Mauritius Mozambique Namibia Nigeria Rwanda

23 Sierra Leone South Africa X South Sudan, Rep. of Swaziland Tanzania Uganda Zambia Zimbabwe Argentina X Bahamas, The Barbados Belize Bolivia Brazil X Chile X Colombia Costa Rica Dominican Republic Ecuador El Salvador Guatemala Guyana Haiti Honduras Jamaica Mexico X Nicaragua Panama Paraguay Peru X Suriname Trinidad and Tobago Uruguay Venezuela, Rep. Bol. X Notes: this table shows whether countries are classified as emerging or developing, and the first calendar year after 1970 for which consistent quarterly data are available for the indicators specified. 22

24 APPENDIX TABLE A Bangladesh MU MU MU MU MU MU MU MU MU MU MU MU MU MU SUSD SUSD MU MU MU MU Bhutan SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP Brunei Darussalam MU MU MU MU MU SECU MU MU MU MU SECU MU MC MC MU MU SUSD MU MU MU Cambodia MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU China Mainland MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU SUSD MU MC MC MU China Macao MU MU MU MU MU MU MU MU MU MU MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD China Hong Kong MU MU MU MU MU MU MU MU MU MU MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Fiji MU MU MU MU MU MU MU MU MU MU SP MU MC MC MU MU MU MU MU MC India MU MU MU MC SDM MC MU MU MU MU SM SM SM SM MU MU MC MU MU MU Indonesia SUSD SUSD SUSD SUSD SUSD MU MU MU MU MU MU MU MC MU SP MU SP SUSD SP SUSD Lao People's Dem.Rep MU MU MU MU MU MU MU MU MU MU MU SUSD MU MU MU MU MU MU MU SUSD Malaysia MU MU MC MC MC MC MC SM MU MU SECU SECU MC SP SP MC SP MC MC SP Maldives MU MU MU MU MU MU MU MU SUSD MU MU SUSD SUSD MU MU MU MU MU MU MU Mongolia X X X X X X X X X X X X X X X X X MU MU MU Myanmar MC MU MU MU MC MC SECU MC MC MU MU MC MU MU MU MU MU MU MU MU Nepal MU MU MU MC SUSD MU MC SUSD SM MU SM MC MU MU MU SM SP MU MU MU Papua New Guinea MU MU MU MU MC SP MU MU MU MU MU MU MC MU MC MU MC MU MU MC Philippines MU MU MU MU SUSD SUSD SUSD MU MU MU MU SM MU MC MC SUSD MU MU MU MU Solomon Islands MC MU MU MC MU MU MU MU MU MU MU MU MU MU MU MU MU MU SUSD MU Sri Lanka MU MU MU MC MC MU MU MU MU MU MU MU MC SUSD MU SUSD MU MU MU MU Thailand MU MU SUSD SUSD SUSD SUSD SUSD SUSD MU MU SUSD MC MC MC SUSD SP MU MU MU MC Timor-Leste X X X X X X X X X X X X X X X X X X X X Vietnam X X X X X X X X X X X X X MU MU MU MU MU MU MU Albania X X X X X X X X X X X X X X X X X X X X Bulgaria MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU Croatia X X X X X X X X X X X X X X X X X X X X Estonia X X X X X X X X X X X X X X X X X X X X Hungary MU MU MU MU MC MC MU MU MC MU MU MU MU MU MU MU MU MU MU MU Latvia X X X X X X X X X X X X X X X X X X X X Lithuania X X X X X X X X X X X X X X X X X X X X Macedonia, FYR X X X X X X X X X X X X X X X X X X X X Poland MU MU SUSD SUSD SUSD MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU 23

25 Romania MU MU SUSD SUSD MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU Serbia, Republic of X X X X X X X X X X X X X X X X X X X X Turkey SM MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU Belarus X X X X X X X X X X X X X X X X X X X X Georgia X X X X X X X X X X X X X X X X X X X X Kazakhstan X X X X X X X X X X X X X X X X X X X X Kyrgyz Republic X X X X X X X X X X X X X X X X X X X X Russian Federation X X X X X X X X X X X X X X X X X X X X Ukraine X X X X X X X X X X X X X X X X X X X X Algeria MU MU MU MU MU MU MU MU MU MU MU SUSD MU MU SECU MU MU MU MU MU Bahrain, Kingdom of MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Egypt MU MU SP SUSD SUSD MU MU SM SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU MU MU Iran, I.R. of MU MU MU MU SUSD SUSD SUSD MU MU MU MU MU MC MU SM SM MU MU MU SM Iraq MC MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Jordan MU MU MU MU MU SECU MU MU MU MU SECU MC MC MC MC MU MU MU MU MC Kuwait MU MU MU MU SUSD MU MU MU MU SUSD MC MC SP SP SP MC SP MC MU SP Libya MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU MU MU MU MU MU MU Mauritania MU MU MU MU MU MU SUSD SUSD MU MU MU MU MU MU MU MC MU MU MU MU Morocco MC MU MU MU MU SFF SECU SDM MU SECU MU MU MU MU MC MC MC MU MU MU Oman MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU MU SUSD SUSD SUSD SUSD SUSD Pakistan MU MU SUSD SUSD SUSD SUSD SUSD SUSD MU MU MU MU MU MC SM MU MU MU MU SP Qatar MU MU MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Saudi Arabia MU MU SUSD SUSD SUSD MC MC SUSD MC SP SUSD SUSD SUSD SUSD MC SUSD SUSD SUSD SUSD SUSD Syrian Arab Republic MU MU SUSD MU MC SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU MU SUSD SUSD SUSD Tunisia MU MU MU MU MU MU SUSD MU MU MU MU SECU MU MU MU MU MU SP MU MU Yemen, Republic of X X X X X X X X X X X X X X X X X X X X Cameroon SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF Central African Rep. SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF Chad SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF Congo, Republic of SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF Equatorial Guinea MU SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF 24

26 Benin SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF Burkina Faso SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF Côte d'ivoire SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF Guinea-Bissau MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU Mali SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF Niger SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF Senegal SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF Togo SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF SFF Angola MU MU MU MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU MU Botswana MU MU MU MU MU MU MU SUSD MU MU MU MU MU MU MU MU MU SP MU MU Burundi MU MU SUSD MU MU SUSD SUSD SUSD SUSD SUSD MU MU MC MC MU MC MU MU MU MC Cabo Verde MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MC MU MU MC Ethiopia MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU Gambia, The MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU Ghana MU MU SUSD SUSD SUSD MU MU SUSD SUSD SUSD MU MU MU MU MU MU MU MU MU MU Kenya MU MU MU MU MU SECU SECU MU MU MU MU MU MU MC MU MU MU MU MU MU Lesotho SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD Liberia SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Madagascar SFF SFF SFF SFF SFF SFF SFF SFF SFF MU MU MU MU MU MU MU MU MU MU MU Malawi MU MU MU MU MU SECU SECU MU MU MU MU MU MU MU MU MU MU MU MU MU Mauritius MU MU MU MU MU SECU MU MU MU MU MU MU MU MC MC MU MU MU MU MC Mozambique MU MU MU MU MU MU MU MU MU MU SUSD SUSD MU MU MU MU MU MU MU MU Namibia SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD Nigeria MU MU MU SUSD MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU Rwanda MU MU MU MU MU SECU SECU MU MU MU MU MU MC MC MC MC MC MC MU MU Sierra Leone MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU South Africa MU MU MU SP SUSD SUSD MU MU MU MU MU MU MU MU MU MU MU SY SY MU South Sudan X X X X X X X X X X X X X X X X X X X Swaziland SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD Tanzania MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU Uganda MU MU MU MU SDM SECU SECU MU MU MU MU MU MU MU MU MU MU MU MU MU Zambia MU MU SUSD MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU 25

27 Zimbabwe MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU Argentina SUSD MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU Bahamas, The SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Barbados MU MU MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Belize MU MU MU MU MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Bolivia MU MU SUSD SUSD SUSD SUSD MU MU MU MU MU MU MU MU MU MU MU MU MU SUSD Brazil MU SUSD MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU Chile MU MU MU MU MU MU MU MU SUSD MU MU MU MU MU MU MU MU MU MU MU Colombia SUSD MU MU MU MU MU SUSD MU SUSD MU MU MU MU MU MU MU MU MU MU MU Costa Rica SUSD MU MU SUSD SUSD SUSD SUSD MU MU MU MU MU MU SUSD MU MU MU MU MU MU Dominican Republic SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU MU MU MU MU MU MU MC Ecuador MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU El Salvador SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU MU SUSD SUSD MU MU MU Guatemala SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU MU MU MU MU MU Guyana MU MU MU MU SUSD SUSD SUSD SUSD MU MU MU SM MU MU MU MU MU MU MU MU Haiti SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU Honduras SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU MU MU Jamaica MU MU SUSD SUSD MU MU MU MU SUSD SUSD MU MU MU MU MU SUSD MU MU MU MU Mexico SUSD SUSD SUSD MU MU MU SUSD SUSD MU MU MU MU MU MU MU MU MU MU MU MU Nicaragua SUSD SUSD SUSD SUSD SUSD MC MU MU SUSD SUSD SUSD MU MU MU MU MU MU MU MU MU Panama SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Paraguay SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU MU MU MU MU MU MU MU MU Peru SUSD SUSD MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU Suriname MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Trinidad and Tobago MU MU MU MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU MU MU MU SUSD SUSD SUSD Uruguay MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU Venezuela, Rep. Bol. SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU MU MU MU MU MU MU MU MU 26

28 Bangladesh SUSD SUSD SUSD MC SM SUSD MU MC SP SM SM SM MU SP MU SUSD SUSD SM MU SM SP Bhutan SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP SRUP Brunei Darussal MU MC MU MC MC MC MU MU SUSD MC MC MC SP SP MC SP MC SP MC MC SP Cambodia MU MU MU MU MU MU MC MC MC SUSD SUSD SUSD SUSD MU SUSD MU SUSD MC MU SUSD SUSD China Mainland MU MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MC SUSD MC MU MC MC SUSD SY SY China Macao SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD China Hong Kong SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SP SUSD SUSD SUSD SUSD SUSD SUSD SUSD Fiji MU MC MC MC MC MU MC MC MC MC MC MC MC MC MC MU MC MU MU MC MC India MU MU MU MU SM SM SY MC MC SP SP MC SP MU SP MU MU MU MU SY SY Indonesia SUSD SUSD SUSD SUSD MU MU MU MC MU MU SY MU MU SY SP MU MC SY SY SY SY Lao People's DR SUSD SUSD MU MU MU MU MU MU MU MU MU MU MU MU MC MU MC MU MU SUSD MU Malaysia MC MC SP SP MC MU MC SUSD SUSD SUSD SUSD SUSD SP SY MC MU MC SM SY MC MC Maldives MC MC MU SUSD SUSD SUSD SUSD SUSD MC MC SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU SUSD Mongolia MU MU MC MU MU MU MU MU MU SUSD SUSD SUSD MU MU MU MU MU MU MU MU MU Myanmar MU MU SFF MU MU MU MU MC MU MU MU MC MU MU MU MU MC MU MU MC MU Nepal MU MU MU MU MC MU MU MC SP MC MU MC MU MU MU MU MU MU MU MU MU Papua New Gui MC MC MU MU MC MU MU MU MU MU MU MC MC SUSD SEUR MU MU MU MC MC MC Philippines MU MU MU MU MU MU MU MC SY MC SP MC SP SY MC MU MC SP MC MC SPY Solomon Islands MU MU SUSD SUSD MU MU MU MU MU MU MU SUSD SUSD SUSD SUSD MU MU SUSD MU MU SUSD Sri Lanka MU MU MU MU SUSD MU SM SM MU MU MU MU SM MU MU MU MC MU SY MU SM Thailand SUSD SP MU SUSD MU MU MC MC MC MC SP MC MC MC MC MU MC MC MC SP MC Timor-Leste X X X X X X X SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Vietnam MU MU SUSD MU MC MU MC MC SUSD MC SUSD SUSD SUSD SUSD SUSD MU MU MU MU MU SUSD Albania MU MU MU MU MU MU MC MC MC MU MC MC SP MC MC MC MC SP MC MC SP Bulgaria MU MU MU MU MU MU SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Croatia MU MU MC MC MC MU MC MC SEUR MC SP SP SP MC SEUR MU MC MC SEUR SEUR SEUR Estonia X SDM SDM SDM SDM SDM SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR MC SEUR SEUR SEUR Hungary MU MU MU MU MU MU MU MU SY MU MC MU MC MC MU MU MC MC SPY MU MC Latvia MU MU MU MU MU MC MC MC MC MC MC MU SP SEUR SEUR SEUR SEUR SEUR MC MC MC Lithuania MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD MC SP SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Macedonia, FYR X X MU SDM MC MC SEUR SEUR SEUR MC MC SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Poland MU MU SM MU SECU MU MU MU MU MC MC MC MC SP SY MC SP MC SY MC SUSD 27

29 Romania MU MU MU MU MU MU MU MU MU MU MU MU MU MU MC MU MU MU MU MC MC Serbia X X X X X X X X X X MU MU MU MU MU MU MU MU MU MU MU Turkey MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU Belarus X X X MU MU MU MU MU MU MU MU MU SUSD SUSD SUSD MU MU MU MU MU MU Georgia X X X X SUSD MC MU MC MC MU MC MU MU MU MU MU MC MU MU MC MC Kazakhstan X X X MU MU MU MU MU SUSD MU SP SP MU SP MU MU MU MU SUSD SUSD MU Kyrgyz Rep X X MU MU MU MU MU MU MU SUSD MC MC MC SP MU MU MU MU MU MC MU Russian Fed X MU MU MU MU MU MU MU MU MU MU MU MU SY SY MU MU MU MU MU SP Ukraine X X MU MU MU MU MU MU MU MC SUSD SUSD MU MU SUSD MU MU MU SUSD SUSD MC Algeria MU MU MU MU MU MC MC MC MC MC MC MC SUSD MC MC MU MU MC MC MU MC Bahrain SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Egypt SUSD SUSD SUSD SUSD SUSD SUSD SUSD SP SP SY SY SY MC MU MU MU MU MU MU MU MU Iran, I.R. of MU MU MU SUSD SUSD SUSD SUSD SUSD MU MU MU SUSD SUSD SUSD SUSD MU MU MU MU MU MU Iraq SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MU MU MU MU MU MU MU MU SUSD SUSD SUSD SUSD Jordan MC MC MC MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Kuwait MC SP SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SEUR MU MC MC MC MC SP Libya MU MU MU MU MU MU MU MU MU MC MC MC MC MC MU MU MC MC MU MU MC Mauritania MU MC MU MC MC MU MC MC MC MC MU MU MU SUSD MU MU MC MU SP MC MC Morocco MU MU SFF SM SECU MC MC MC MC MC SEUR MC MC MC SEUR MC MC MC MC MC SEUR Oman SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Pakistan MU MU MU MU MU MU MC MC MC MC SM MU MU SUSD SUSD MU MU MU MU MU MU Qatar SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Saudi Arabia SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Syria SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Tunisia MC SP MU MC SP SP SP MC MC MC MC MC SM MC SEUR MC MC MC MC SP SP Yemen, Rep. X SM MU MU MU MU MU MU SUSD SUSD MU MU MU MU SUSD SUSD MU MU MU SUSD SUSD Cameroon MC MU MU SFF SFF SFF SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Central Afr Rep MC MU MU SFF SFF SFF SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Chad MC MU MU SFF SFF SFF SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Congo, Rep MC MU MU SFF SFF SFF SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Equatorial Gui MU MU MU SFF SFF SFF SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR 28

30 Benin MC MU MU SFF SFF SFF SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Burkina Faso MC MU MU SFF SFF SFF SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Côte d'ivoire MC MU MU SFF SFF SFF SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Guinea-Bissau MU MU MU MU MU SDM SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Mali MC MU MU SFF SFF SFF SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Niger MC MU MU SFF SFF SFF SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Senegal MC MU MU SFF SFF SFF SP SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Togo MC MU MU SFF SFF SFF SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Angola MU MU MU MU MU MU MU MU MU MU MU MU MU MU SP MU SP MU MU MU SUSD Botswana MU MU SP MU MU MU MU SP MU MU MU SP MU MU MU MU MU MU MU MU MU Burundi MU MU MU MU MU MU MC MU MU MC MU MU MU MC MU MU MU SUSD MU MU MU Cabo Verde MU MC SFF MU MU MC MC SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR SEUR Ethiopia MC MU MU MC MC MC MU MC SUSD MC SUSD SUSD SUSD SUSD MU MU MU MU MU SUSD MU Gambia, The MU MC MU SUSD MC SUSD MC MC MC MU MU MU MC MC MU MC MC MU MC SP SP Ghana MU MU MU MU MU MU MU MU MU MU MU MU SUSD SUSD MU MU MU MU SP MU MU Kenya MU MU MC MU MU MU MU MU MU SUSD MU MU MU MU MU MU MU MC MU SY SY Lesotho SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD Liberia SUSD SUSD SUSD SUSD MU MU MU MU MU MU MU MU MU MU MU SUSD MU MU SUSD SUSD MU Madagascar MU MU MU MU MC MU MU MU MU MU MC MU MU MU MU MU MU MU MU MU SP Malawi MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU SUSD MU MU MU MU MU Mauritius MU MU SP MU MU MU MU MC MU SM MC MC MC MU MU MU MC MC SM SY SY Mozambique MU MU MU MU MU MC MC MU MU MU MU MU MU MU MU MU MC MU MU MC MC Namibia SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD Nigeria MU MU SUSD SUSD SUSD MU MU MU MU MU MU MU SUSD SUSD SEUR MU MU MU MU MU SUSD Rwanda MU MU MU MU MU MU MC MC MC MC MU MU MU SUSD SUSD MU MU MC MU MU SY Sierra Leone MU MU MU MU MU MU MU MC MC MC MU MU MU SUSD SUSD MC MC MU MU MU SUSD South Africa MU MU MU SPY MU SY MU SY MU MU MU MC SY MC MU MU MU MC SM MC MU South Sudan X X X X X X X X X X X X X X X X X X X X SUSD Swaziland SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD SRD Tanzania MU MU MU MU MU MU MU MU MU MU SUSD MC MU MU MU MU MU MU MU MU MU Uganda MC MU MU MU MU MC MU MC MC MC MU MC MU MU MU MU MU MC MU MU MU Zambia MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU SP SP SP Zimbabwe MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU 29

31 Argentina SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD MC MU MU MC MU MU SUSD MU MU MU MU SP SP Bahamas, The SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Barbados SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Belize SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Bolivia SUSD MU MU SUSD MC SUSD SUSD SUSD SUSD SUSD MC SP MU SUSD MU MU MC SUSD SUSD SUSD SUSD Brazil MU MU MU MU SUSD MC MC MU MU MU MU MU MU MC MC MU MC MU MU MU SP Chile SP MU SP MU MU MU MC MC MC MC SY MC MC SY MC MU MC MC MC SY MC Colombia MU MU MU MU MU MU MU MU MU MU MU SP SP SP MU MU MC SP SY MC SP Costa Rica MU SUSD MU MU SUSD SUSD SUSD MU SUSD MU SUSD SUSD SP MU MU MU MU MU SP MC MC Dominican Rep MU MU MU MU MU MC MU MU SUSD MU MU MU MC MU MU MU MC SUSD SUSD SUSD MC Ecuador MU MU MU MU MU MU MU MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD El Salvador MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Guatemala MU MU MU MU MU MU MU MU MU MU SP MU MU SUSD SUSD MU MC MC MU SY SP Guyana MU MU MU MU SUSD MC MU MU SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Haiti MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU MC MU MU SUSD SUSD Honduras MU MU MU MU MU MU SUSD SUSD SUSD SUSD SUSD SUSD MU SUSD SUSD SUSD SUSD SUSD SUSD SP SP Jamaica MU MU MU MU MU SUSD MU MU MU MU MU MU MU MU MU MU MU MU MU MU MU Mexico SUSD MU MU MU MU MU MU MU MU SP MC SP MC SP SP SM MU SP SP SP SP Nicaragua MU MU SUSD SUSD SUSD SUSD SUSD MU SUSD MC SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Panama SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Paraguay MU MU MU MU MU MU MU MU MU MU MU MC MU MU MU MU MC MC MU MC MC Peru MU MU MU MU MU MU MC SP MC MC MC MC MC MC SEURO MU MC MC SUSD SUSD SP Suriname MU MU MU MC MU MU MU MU MU MU MU MU MU SUSD SUSD SUSD SUSD MU MU MU SUSD Trinidad & Tobago MU MU MU MC SP SUSD SUSD SUSD SUSD SP MC SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD SUSD Uruguay MU MU MU MU MU MU SUSD MC MC MU MU MU MC MU MU SP MU SP SP SP SP Venezuela MU MU MU MU MU MU MU MU MU MU MU MU MU MU SUSD SUSD MU MU MU MU MU 30

32 Table 1: Domestic objectives by subperiod, pre-consolidation SP SM SY SPY MU MC X Total Table 2: External objectives by sub-period, pre-consolidation SUSD SDM SFF SECU SEURO SRD SRUP NER X Total Table 3: Objectives as finally identified, by sub-period no. % no. % no. % no. % no. % SP SM SY SPY MU MC SUSD SDM SFF SECU SEURO SRD SRUP X Total

33 Table 4: Emerging versus developing country objectives EM no. EM % Dev no. Dev % EM no. EM % Dev no. Dev % SP SM SY SPY MU MC SUSD SDM SFF SECU SEURO SRUP SRD X Total Note: the percentages are of the non-x totals. Table 5: Objectives in advanced economies number % number % number % number % SP SY SPY SM MC SUSD SDM SECU SEURO SGBP MU X Total Note: percentages are relative to the total of non-x cases, where X includes the cases of anchor currencies as well as cases where the country or data are not available. Source: Cobham (2013, Table 3). 32

34 Table 6: Inflation and growth, emerging and developing Inflation (%) Growth (%) EM Dev EM Dev EM Dev EM Dev SP SM SY SPY MU MC SUSD SDM SFF SECU SEURO SRD SRUP All Note: All is a weighted average (excluding X cases) where the weights correspond to the relative incidence of each type of objective. 33

35 Table 7: Inflation regressions Emerging countries Developing countries constant (2.72)*** 2.83 (2.37) 8.37 (1.72)*** 5.37 (1.44)*** growth (0.09)* (0.10) (0.01) (0.02)** openness 0.04 (0.03) 0.04 (0.02) 0.05 (0.01)*** 0.04 (0.01)*** monetary growth 0.16 (0.04)*** 0.15 (0.05)** 0.06 (0.02)** 0.07 (0.02)*** MC (1.26)*** (0.58)*** (0.99)*** (0.82)*** SDM (1.45)*** (6.51) - SECU (1.59)* (2.24) - SEURO (0.92)*** (1.36)*** SFF (1.27) (1.81)*** - SM (2.60) (2.79) 2.90 (4.01) (0.74)*** SP (1.31)*** (0.61)*** (1.02)*** (0.58)*** SPY (1.84) (0.49)*** - - SUSD (1.75)** (1.12)*** (1.07)*** (0.65)*** SY 1.23 (1.59) (0.73)*** (0.75)*** observations R-squared Notes: robust standard errors (clustered on country) in brackets; ***, ** and * indicate significance at 1%, 5% and 10% respectively; R-squared is within. 34

36 Table 8: Growth regressions Emerging countries Developing countries constant (83.39) (133.00)*** (55.05) (71.67) openness (0.03) 0.07 (0.03)** 0.03 (0.02) (0.02) investment ratio 0.28 (0.06)*** 0.22 (0.12)* 0.05 (0.05) 0.10 (0.04)** population (4.95) (7.58)*** (3.57) 4.02 (4.33) population growth (0.43) (0.46)** 0.07 (0.49) (0.38) average schooling (0.68) 0.08 (0.53) 0.04 (0.68) 0.38 (0.81) MC 0.74 (0.74) 0.20 (0.38) (0.70) 0.80 (0.37)** SDM 0.14 (2.74) SECU 1.15 (0.94) (1.58) - SEURO (1.18) (1.27) SFF (0.70)*** (0.98) - SM 1.61 (1.13) 0.23 (0.69) (1.75) 0.23 (0.29) SP 0.86 (0.73) 0.60 (0.58) 1.58 (2.75) 0.77 (0.64) SPY 3.84 (1.06) 2.12 (1.50) - - SUSD (0.63) (0.83) 0.22 (0.65) 0.53 (0.42) SY 0.46 (0.69) (0.56) (0.81) observations R-squared

37 Figure 1: Countries with consistent quarterly data at different dates 36

38 Figure 2: De facto objectives of emerging countries 37

39 Figure 3: De facto objectives of developing countries 38