BRUEGEL WORKING PAPER 2012/04

THE G20: CHARACTERS IN SEARCH OF AN AUTHOR

IGNAZIO ANGELONI* AND JEAN PISANI-FERRY**

Highlights
The G20 has produced mixed results. After initial high hopes and some success, negotiations within the G20 forum have slowed, progress is less visible and disagreement rather than agreement has come to the fore. Against this background, this paper revisits the basic economic and geopolitical motivations for the G20, in order to review its performance and attempt to draw lessons for the path ahead. We conclude that: (1) in todays global economy (with its trade and financial market integration and its institutional architecture) a G20-type institution is necessary if it didnt exist, it should be created; (2) the G20 had its high noon moment in 2008-09 and some recalibration of expectations was inevitable, but its achievements in 2010-11 have nevertheless been disappointing; (3) to be fair there is, in detailed and technical work, more progress than there seems to be at first sight; (4) from a governance standpoint, the G20 is not an efficient forum; improvements in working methods are urgently needed; (5) more fundamentally, for the G20 to retain its role, its members need to develop a common vision of global economic problems and the way to approach them. * Bruegel and European Central Bank; [email protected]. ** Director of Bruegel; [email protected]. An earlier version of this paper was presented at the Bank of Korea Annual Conference held in Seoul, 26-27 May 2011. We are grateful to our discussants, Yiping Huang and Sungmin Kim, and to Barry Eichengreen, Avinash Persaud, Andre Sapir, Tom Sargent, Larry Schembri, to name just a few, for stimulating comments. Christophe Gouardo and Silvia Merler provided excellent research assistance. The views expressed in this paper are personal and do not reflect those of either the European Central Bank or Bruegel.

MARCH 2012

Siamo qua in cerca dun autore! Dun autore? Che autore? Duno qualunque, signore.1 L. PIRANDELLO

1.

Introduction: new actors on a new stage

It is commonplace to assert that the crisis that started in 2007 has changed the world financial and economic stage in a fundamental way. Less obvious is to explain how and why. A key novelty is the increased interconnection among phenomena previously deemed unrelated. This interconnection predates the crisis, but before 2007 few were aware of its true extent. It has several dimensions. Spatial interconnection, first of all: the ability of asset markets to transmit shocks across space was known, but few anticipated, only three years ago, that financial integration was so advanced that a crisis in the US would automatically lead to a crisis in Europe in a matter of days. Market interconnection, second: once the financial turmoil hit, it spared nobody: banks, other intermediaries, corporates, households, real estate, stocks, eventually governments and their debt. And in the end, unavoidably, policy interconnection: in a crisis so extensive and devastating all instruments of economic policy (monetary, banking and financial, fiscal in all its ramifications) are called into action, often together, to rescue and to repair. The faith in a pre-arranged and rigid assignment of policy instruments to specific goals is an intellectual casualty of this crisis. Another key aspect of the picture is the birth, or move to centre-stage, of new actors on the world economic scene. National actors, most evidently: the main emerging countries, previously invited when lucky to sit in a dinner or speak at a seminar at the margin of a G7, since 2008 have become equal partners in the global economic decision making process. More than that: they have played, as we shall discuss, a fundamental role in providing much needed support to the global economy during the 2009 recession. The simple fact that a contribution of the emerging countries to solving the crisis in the euro area has been discussed, as indeed happened 2011, is indicative of the magnitude of the change. But, in addition to national actors, we are now witnessing the rise of a whole new class of institutional actors. The financial architecture reform underway has produced a new family of financial market regulators or empowered existing ones. The Financial Stability Forum has changed name and grown in size and scope. Systemic risk councils have been created on both sides of the

'We are here in search of an author! An author? Which author? Whoever, sir'. Quote from Luigi Pirandellos play Sei personaggi in cerca dautore, 1921.

Atlantic. In Europe, supervisory and regulatory coordination, recently a vision of few farsighted2, is now largely achieved through the creation of three new supranational agencies, respectively supervising banking, markets and institutional investors. And the list could continue. Among the institutional actors, a most prominent one, with the ambition of playing the protagonist role, is the new G20. 'New' because a G20 already existed since 1999, but at a ministerial level and with much lower visibility and responsibility. In the days following the fall of Lehman Brothers (September 2008), when global finance seemed about to melt down, the political leaders reacted with a new imaginative idea, to convene a meeting of G20 leaders to find a common response to the present danger and to send a reassuring message to the financial markets and the public opinions. It was not obvious that a financial crisis originated and confined in the advanced countries should be approached by strengthening a grouping like the G20. yet the idea worked, and contributed positively to the restoration of calm and initiating, through successive meetings in Washington, London, Pittsburgh, Toronto, Seoul and Cannes, a phase of effective policy coordination in several key areas, including notably financial market reform, macroeconomic policies, and IMF governance. The overall performance of the G20 over time is the object of dispute. After considerable initial excitement, hope and success, leaders and international bodies got tangled in complex details. Negotiations slowed, progress became less visible, disagreement resurfaced, and outside criticism increased. Some ask if the G20 is already pass. Against this background, the purpose of this paper is to revisit the basic economic and geo-political motivations for the G20, to review its performance and attempt to draw, from the two, lessons for the way ahead. We start in section 2 with a discussion on the rationale for the G20 in its current format. We then review in section 3 its short history, outlining what have been the ambitions and the achievements of the successive presidencies. In section 4 we move to providing assessment, partly judgemental and partly by using more systematic indicators available in the literature. Section 5 is devoted to governance and the reforms that could help make the G20 more effective and efficient. Section 6 concludes. In short, our main conclusions are the following: (1) in todays global economy (with its trade and financial market integration and its institutional architecture) a 'G20-type' institution is necessary if it didnt exist, it should be created; (2) the G20 had its high noon moment in 2008-09 and some recalibration of expectations was inevitable, but its achievements in 2010-11 have been
2 We pay tribute here to the late Tommaso Padoa-Schioppa, who in 2007 proposed to the Ecofin, receiving almost no support, changes in Europes financial regulation that were very similar to those eventually enacted.

nevertheless disappointing; (3) to be fair there is, in detailed and technical work, more progress than appears at first sight; (4) from a governance standpoint, the G20 is not an efficient forum; improvements to its working methods are urgently needed; (5) more fundamentally, for the G20 to retain its role, its members need to develop a common vision of global economic problems and the way to approach them. 2. Do we need a G20 at all?

Is there a need for economic coordination in todays global economy? Can policy decisions be improved by a collective consultation process, similar to that taking place in the G20, or should such decisions be taken by national authorities acting in isolation and based primarily on domestic conditions? Even if coordination is needed, should it involve emerging as well as advanced economies? These are the question we are looking at in this section. The focus is not on whether coordination has worked in practice (we will examine the performance issue next), but on whether it can serve a useful purpose as a matter of principle. If the answer is in the affirmative, then the debate can move on to examine how this function can be performed, and by whom. The question is neither rhetorical nor new. For long time, for example under the classical gold standard (nineteenth and early twentieth centuries), national monetary policies followed by and large stable rules, shaped by incentives and constraints dictated by the working of the international monetary system (Keynes famous 'rules of the game')3. The inherent constraints on the conduct of monetary policy implicitly provided policy discipline in other domains as well. Nevertheless, no explicit, discretionary coordination mechanism existed at the time. Rules made national policy decisions relatively simple and predictable. As discussed below in more detail, this lesson from history namely, that not all successful international economic policy arrangements necessarily rely on coordination received considerable backing from economic analysis. Theory and evidence in the recent decades have tended to support the view that, under plausible circumstances concerning the working of the international economy, the most efficient and effective arrangement for policymaking corresponds to each country acting in isolation, pursuing national objectives. While acknowledging that these views carry weight, and that the issue is far from settled, we take a different view here. The point we make in this section is that the world economy has evolved, in recent years, in a way that makes the benefits from policy coordination at G20 level more likely and more substantial. The same developments have made the analytical results in the economic literature, generally denying the existence of significant benefits from coordination, less relevant.

See McKinnon (1993). The relevance of the concept of rules of the game was challenged in the well known essay by Triffin (1985).

We will proceed in three steps, first examining links between coordination and developments in the international monetary system in the post-war period; second by revisiting and putting in context the arguments and evidence produced by the economic literature on the gains from coordination; third by evaluating the degree of integration between advanced and emerging economies. 2.1 Monetary arrangements and international coordination since WW2 The world monetary order established after WW2 was centred on the hegemonic position of one country, the US, and it currency, the dollar. The US acted as the main source of global demand and of global liquidity, with all other industrialised countries (mainly, Europe) linked to the dominant country and to it currency by a set of rules, amounting to what was called the 'gold exchange standard' (Angeloni et al, 2011). Europes post-war reconstruction and development needs required an expansion of international demand, which was initially provided by the move of the US into current account deficit, and a constant expansion of international reserves, which took place through a persistence of the US financial account deficits, the accumulation of dollar balances in the books of the other countries central banks, and ultimately a gradual transfer of US gold reserves to the rest of the world. As well known, the system worked acceptably for many years but in the end was undermined by its own success (the 'Triffin dilemma'). Eventually, the continuous and growing US deficit brought the system to reach the limit of the US availability of gold reserves. What matters more for our purpose, however, is that the system designed at Bretton Woods was built on rules that produced an automatic self-rebalancing mechanism for all countries except the US. For this reason it did not require explicit coordination mechanisms. With the important exception of the US, the emergence of an excessive external imbalances in any country was automatically cured by the constraint of the availability of dollar reserves (and indirectly, gold) triggering macroeconomic adjustment, either by market forces (deflation), or domestic policies or, in occasional circumstances, via exchange rate adjustments. The breakdown of Bretton Woods in 1971 and the subsequent move to floating exchange rates in 1973 changed this situation radically. First and foremost, the possibility of exchange rate floating removed the constraint on domestic policymaking arising from the need for foreign exchange reserves. Some countries adopted other forms of domestic monetary anchors most notably, Germany introduced monetary targets in the mid-1970s. Other European countries continued to peg their exchange rate, adhering first to the 'snake' and later, in 1979 and following years, to the European Monetary System (a fixed exchange-rate system among European countries that mimicked the Bretton Woods system on a regional scale). Most developing economies initially

remained on some sort of peg or managed exchange rate system (Ilzetzki, Reinhart and Rogoff, 2008). The demise of the Bretton Woods regime and the first oil shock in 1973-74 triggered a new wave of cooperation arrangements. The first important decision was taken in 1975 when the G5 (which would soon become the G7) was created by initiative of the French President Giscard dEstaing. As recovery from this shock disappointed, the G7 soon embarked on a coordinated stimulus exercise on the occasion of the 1978 Bonn summit. The attempt ended in a failure after the second oil shock in 1979 called for a reversal of priorities in favour of fighting inflation and it remained in the policymakers memory as an illustration of the risks of coordination. In the eyes of many policymakers, joint action had led to embark on a wrong course and to fight a demand shock while the world economy had been suffering from a supply shock. Especially, German policymakers resented to have bowed to US pressure, which led the country to be caught in a difficult position when the second oil shock hit. There would be no further meaningful attempt at coordinated stimulus until the 2008 crisis. In the mid-1980s, following the rise and fall of the US dollar exchange rate, the G7 embarked on an attempt at monitoring exchange rates. The Plaza accord of 1985 and the Louvre accord of 1987 resulted in the setting of soft target zones for exchange rates between the dollar, the deutsche mark and the yen. These guideposts would remain more or less in place until the first half of the 1990s, after which they were de facto abandoned. But again, it was felt afterwards in the policy community that coordination on exchange rates had gone too far (from the mere avoidance of blatant misalignments to the near-fixing of real exchange rates) and again, it left one country unhappy. In that case it was Japan whose monetary policy had been geared towards external objectives instead of addressing the build-up of the housing and stock-market bubble4. Apart from these two significant attempts, the G7 in the reminder of its 35-years existence refrained from attempting at effective coordination and limited its ambition to setting broad objectives. At finance minister level, it was also instrumental in setting directions for multilateral responses to developing-countries crises, serving as a de facto steering committee for the IMF. The emergence of the G20 was initially a response to the recurrence of crises in emerging countries. Its creation at ministerial level was triggered by the Asian crisis of 1997-98, and in its first ten years of existence it served as a consultation forum and did not attempt at effective policy coordination. The global financial crisis of 2007-08 led to its upgrading to head of state and government level and,

Shirakawa (2011).

in 2009, to its elevation to the status of main forum of global economic coordination, overtaking the G7/G8. 2.2 Gains from coordination: the academic literature revisited Policy developments in the 1970s and the 1980s triggered research on the benefits and costs of macroeconomic policy coordination. Research in this area, both theoretical and empirical, was very active in the 1980s and then again in the last 10 years. The two strands of literature were very different in method and emphasis, but converged, for different reasons, to a common conclusion: that the benefits of policy coordination are likely to be too small to balance the costs of negotiating policies among independent governments. In many respects, the benchmark view was put forth by Milton Friedman (1953). According to Friedman, flexible exchange rates have the property of insulating the national economy from domestic and foreign shocks, deriving from technological or market factors as well as from policies. By 'flexible', Friedman and his followers meant floating, that is, as clearly explained by Harry Johnson (1969) foreign exchange rates that are determined daily in the market for foreign exchanges, by the forces of demand and supply without restrictions imposed by governmental policy on the extent to which they can move. If the insulation property is in fact fulfilled, cross-border effects stemming from policies or other economic shocks disappear and national policymakers can concentrate on conducting optimal policies based on domestic objectives alone. International coordination becomes, by definition, pointless. Does the exchange regime influence real macroeconomic performance? Baxter and Stockman (1989) examined the statistical properties of the main macro variables of a large sample of industrialised countries, before and after the end of the Bretton Woods regime. They concluded that, though the macro performance (measured by volatility and correlation of the main macro indicators) changed in a number of respects across the two periods, this was not related to the exchange regime. This evidence has traditionally been considered as supportive of Friedmans view, based on the reasoning that, if no gain in terms of macroeconomic performance can be achieved by fixing or pegging exchange rates, it may be better to opt for flexible rates. a simpler solution that also increases the degrees of freedom for policymakers and allows to choose and pursue country specific objectives. In fact, however, the same evidence could be read differently: if macroeconomic volatility is not altered by the exchange regime, the insulating property of floating rates is unlikely to hold, at least in the strict sense advocated by Friedman.

Floating exchange rates furthermore do not make coordination irrelevant. Models with strategic interactions among policymakers la Hamada (1974, 1979), Canzoneri and Gray (1985) and others showed that monetary policy could be used strategically, with the result that the uncoordinated outcome was Pareto-inferior. This finding provided an argument for coordination, but the question was how significant it was empirically. In the 1970s and 1980s, many approached the issue of the gains from coordination using macroeconometric models, with blocks of countries linked by international trade. Capital flows were absent or passive, reflecting the reality of the times. As a consequence, the international spillovers incorporated in these models depended on trade integration typically limited at the time, but rising. Following Oudiz and Sachs (1984), many attempted to provide empirical estimates of the gains from coordination. The central message of this literature, reviewed by Canzoneri and Henderson (1990), was mixed: on the one hand, there was clearly a role for coordination, generated by the fact that domestic policy choices did have cross-border effects in particular, beggar-thyneighbour policies were possible by manipulating the exchange rate. On the other, these effects were quantitatively small, and so were the potential gains from coordination. Frankel and Rockett (1988) added another argument based on imperfect information: even though there would be gains in a fullinformation context, uncertainty over the true structure of the economy made them elusive at best. In short, the literature of the 1970s and the 1980s did not deny the possibility of coordination gains, but it claimed that there were too small and uncertain to be pursued by policymakers. The key issue was an empirical one: the problem is correctly measuring the importance and implications of crossborder interconnections stemming from trade and potentially from capital flows, using realistic models. The subsequent literature, however, abandoned this line and took a markedly different route. A new generation of open-economy macroeconomic models appeared after the turn of the millennium (Obstfeld and Rogoff, 2001; Corsetti and Pesenti, 2001; Clarida, Gali and Gertler, 2002; Canzoneri, Cumby and Diba, 2002), which differed from the earlier ones in two respects. First, they possessed explicit (but highly stylised) microeconomic foundations, which allowed, with assumptions about individual preferences, to directly measure the effects of policy coordination on the welfare of economic agents. Second, they featured price rigidity and monopolistic competition on the supply side. These market imperfections made competitive equilibria suboptimal and opened the way to a role for macroeconomic policy in removing or alleviating the distortions. In the logic of these models, gains from coordination arise if policymakers can correct the market distortions better by coordinating than by acting in isolation focusing on domestic objectives alone. Moreover, further gains can be achieved if international financial markets are imperfect for example, if securities

cannot be freely traded across borders so that by coordinating policymakers manage to achieve a higher degree of risk sharing (Bergin, 2008). Two additional assumptions of these models are however worth emphasising, because they almost automatically exclude sizeable gains from international coordination. First, individual preferences are (at least nearly) logarithmic and the elasticity of substitution between domestic and foreign consumption is (at least approximately) unitary. This implies that changes in the terms of trade do not significantly affect the external balance. The current account remains close to equilibrium regardless of any shocks, the output gap essentially coincides with that prevailing in a close economy, and hence, not surprisingly, inward oriented monetary policy turns out to be optimal. Second, these models typically do not contain sources of persistence, such as capital accumulation or debt build-up. This rules out persistent disequilibria in the balance of payments as well. Note that these properties of the 'new generation' open economy models do not depend on the degree of international openness; even if trade or capital market integration increases, as it has in recent years, under the restrictive assumptions just described international spillovers and gains from coordination would be automatically excluded. The new open-economy macroeconomic literature of the 1990s and the 2000s reinforced the conclusions from the earlier research that coordination could in principle help improve upon the Nash equilibrium but that gains were unlikely to be significant enough to justify incurring the cost of coordinating national policies. They reinforced them because they made them independent of the external openness of the countries concerned. As Obstfeld and Rogoff (2002) put it, lack of coordination may not always be a big problem, even in a world with significant economic integration, adding that continued improvements of monetary policy institutions at the domestic level, coupled with the further broadening of world capital markets, may render [partial coordination] schemes superfluous or even counterproductive. The insights from this literature are important, as they indicated that the standard policyoptimisation argument for coordination is likely to carry less empirical weight than initially thought. However, even without subscribing in full to Goodharts dismissive judgement on the usefulness of these models for policy purposes5, one must admit that they offer little guidance to address the problems faced recent years. Neither the open-economy models la Mundell-Fleming of the 1980s nor those, la Obstfeld-Rogoff, that were developed in the 1990s, offer much insight into the type and extent of interdependence observed in recent years and documented in the IMF Spillover

There is simply nothing in these models that is of the slightest interest to a central banker. This quote is reported by White (2010).

Reports (2011a, 2011b)6. Empirical research on the spillover effects of national developments and policies, as undertaken by the Fund, highlights that interdependence through traditional channels can be dwarfed by that arising from gross holdings of financial assets and the bellwether role of US capital markets. Except for countries like Canada, Mexico, China and Saudi Arabia, for which the US is primarily an export market, asset price links are significantly more important than traditional links and taking them into account typically multiplies the spillover effects of US shocks by a factor comprised between two and five, or even more. Furthermore these linkages are asymmetric as US developments affect the rest of the world much more than vice-versa. These phenomena, which constitute the bread and butter of policy discussions at global level, are often assumed away in standard models. 1.3 Coordination with whom? Whereas the creation of the G20 occurred on a particularly dramatic occasion, it had been in the making for several years. From about 2000 onwards, it became clearer that the G7 was not the appropriate venue anymore to address the major issues for the world economy. The leaders from Brazil, China, India, Mexico and South Africa were invited to have coffee with the G7 leaders on the occasion of the G8 summit in Heiligendamm in 2007. Absent the global crisis, the process of enlargement would probably have taken place anyway, although in a more gradual way. The reason for this evolution can easily be illustrated with a few graphs. To start with, growth in the emerging and developing countries accelerated markedly in the 1990s, implying that their share in the world economy (at PPP exchange rates) increased from 31 percent in 1990 to 48 percent in 2009 (Figure 1). Figure 1: GDP per capita in the advanced and emerging economies, 1970-2009
Advanced countries
35000 30000 25000 20000 15000 10000 5000 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 0 2000 1800 1600 1400 1200 1000 800 600 400 200 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 0 3000 2000 1000 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 0

Emerging and developing countries

7000 6000 5000 4000

World

Source: World Development Indicators (World Bank), authors calculations.

There are exceptions, for example by Faia (2007).

Second, the participation of these countries in world trade also increased dramatically (Figure 2). Whereas until the 1990s the bulk of world trade consisted in exchanges of goods and services between advanced countries, its pattern underwent major changes in the last two decades. As a consequence the de facto structure for trade negotiation was significantly transformed: with the launch of the Doha round the traditional Quad consisting of the US, the EU, Japan and Canada all G7 members was replaced by for agricultural negotiations by a new informal grouping, the FIPs (Five Interested Parties) of the US, the EU, Australia, Brazil, and India. This change is widely regarded as irreversible as it is in line with the change in the structure of world trade. Figure 2: Trade to GDP ratios, 1970-2009
Advanced countries
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.7 0.6 0.5 0.4 0.3 0.2 0.1 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 0

Emerging and developing countries

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

World

1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008

Source: World Development Indicators (World Bank), authors calculations. Trade is equal to sum of exports and imports.

The situation is less clear-cut for finance. While financial account liberalisation was completed in the early 1990s in advanced countries, it is far from being complete in most emerging countries. As a consequence financial opening measured by the ratio of assets and liabilities to GDP exceeds 400 per cent in the advanced countries against 100 per cent in the emerging and developing countries (Figure 3). Figure 3: Assets and Liabilities as a percentage of GDP, 1970-2009
Advanced countries
3 2 1 0 -1 -2
1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

Emerging and developing countries

5 4 3 2 1 0 3 2 1 0 -1 -2
1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

5 4 3 2 1 0

3 2 1 0 -1 -2
1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008

World
5 4 3 2 1 0

-3

-3

-3

Assets + Liabili es

Liabili es

Assets

Assets + Liabili es

Liabili es

Assets

Assets + Liabili es

Liabili es

Assets

Source: Lane and Milesi-Ferretti (2009), IMF, authors calculations.

10

As a consequence, the emerging countries are much less integrated in the networks of global finance. Recent research by Kubelec and S (2010) provides data on bilateral financial claims that make possible to draw both a map of global trade integration and a map of global financial integration (Figures 4a and 4b). The graph for trade highlights the key role of China alongside the traditional trade powers and its strong interconnection to both all major markets and significant prediction networks in East Asia. It also shows that key emerging countries such as Korea and Mexico (and to a lesser extent Brazil and India) have become significant parts of the world trade web. The graph for finance highlights the pivotal role of five world hubs, the US, the UK, Germany, France and Japan all G7 members which are both tightly interconnected and connected with all or most other important centres. Interestingly, however, the same data also indicate the emergence of financial centres in emerging countries: mainland China, Hong Kong, Korea and Singapore in East Asia; India in South Asia; and Argentina, Brazil and Mexico in Latin America. Interestingly, 14 of the 18 nodes represented on the graph are G20 members, one (Spain) is a permanent guest, one (Hong Kong) is represented through China and one (Singapore) was a guest to the Cannes summit. Figure 4a: International trade networks, 2005

Source: Kubelec and Sa (2010). Links are given by the sum of bilateral exports and imports divided by the sum of the GDPs of the source and host countries. The size of the nodes is proportional to the countrys trade openness, measured by the sum of its total exports and imports.

11

Figure 4b: International financial networks, 2005

Source: Kubelec and Sa (2010). Links are given by the sum of bilateral assets and liabilities divided by the sum of the GDPs of the source and host countries. The size of the nodes is proportional to the countrys financial openness, measured by the sum of its total external assets and liabilities.

Clearly, there is a gap between the degree of economic integration between advanced and emerging economies and their degree of financial integration. Data on the size of domestic financial markets however indicate the fast-rising financial potential of the emerging world (Figure 5). Whereas the rise of emerging finance will ultimately depend on liberalisation decisions, it is fair to expect it to vindicate the inclusion of major emerging countries in the G20 group. Figure 5: Size of domestic financial markets, 1988-2010
Advanced
90000 80000 70000 60000 50000 40000 30000 20000 10000 0 18000 16000 14000 12000 10000 8000 6000 4000 2000 0

Emerging and developing

120000 100000 80000 60000 40000 20000
1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 1988 1990 1992

World

1990

1992

1994

1996

2000

2002

2004

2006

2010

1988

1998

2008

1994

1996

1998

2000

2002

2004

2006

2008

Market capitalization

Market capitalization Stocks traded

Market capitalization Stocks traded

Source: World Development Indicators (World Bank), authors calculations.

12

2010

3.

The G20 saga

How successful was the G20 so far? The limited time (3 years) elapsed since its creation or rather its upgrading, in November 2008, from an informal ministerial group to a forum of global economic coordination makes it difficult to answer this question. Its agenda has evolved in line with economic developments and policy priorities. Its scope is still shaping up and implementation is very much in progress, not only because of delay. During its first three years it has mostly been chaired by G7 countries (US, UK, France), with the exception of Korea. The successive Mexican, Russian, Australian and Turkish presidencies of 2012-2015 will be a true test of the changes it can bring. Moreover, there is also another reason why a performance judgement is premature. Success ought to be measured against explicitly stated goals. The 'new' G20, convened in a rush in the dramatic postLehman weeks, has so far had neither the time nor the desire to embark in a soul-searching discussion about ultimate objectives. The G20 exists, but its mission and role in the world economy are not well defined. For the moment, its focus evolves over time, driven by successive presidencies and urgencies (Table 1). Hence, the only systematic way to provide elements for a judgement is to study the G20 compliance with its own specific commitments, as contained in official communications mainly communiqus released at the end of each meeting. This admittedly narrow approach facilitates the task, because the G20 has been in most cases rather explicit and detailed in spelling out its future work agenda. We will do this in two ways. Our first approach is to revisit the sequence of meetings mainly summits, and where necessary ministerial and discuss their outcomes. This narrative unavoidably reflects our own biases, concerning for example what the G20 should do and what expectations could justifiably be entertained ex-ante. We move to a more objective assessment in the next section. Table 1: From Washington to Cannes: An evolving agenda
Summit Washington London Date November 2008 April 2009 Headline priorities Pittsburgh September 2009 Financial reform Global stimulus Financial reform International financial institutions G20 governance Rebalancing of world economy Financial reform

13

Toronto

June 2010

Rebalancing of world economy Financial reform Rebalancing of world economy International financial institutions International monetary system Commodity prices Euro crisis

Seoul

November 2010

Cannes

November 2011

3.1 Washington The Washington summit (15 November 2008) was convened, as we mentioned, at the peak of the global financial market tensions that followed the failure of Lehman Brothers. Several later claimed paternity of the idea, including the then UK Prime Minister Gordon Brown, the outgoing Bush administration and president Sarkozy of France, which held the rotating EU presidency at the time. As the US was holding presidential elections, leadership was more than usually exercised by the Europeans, among whom coordination had strengthened after an emergency meeting of euro area heads of state and government (and the UK), convened at the initiative of the French on 12 October to define common responses to the banking crisis.7 Regardless of who can legitimately claim the initiative, two things about the Washington meeting are clear. The first is that it was unexpected, the second that it was well received by public opinions and financial markets. It amounted to positive news, which contributed to stabilise financial markets (table/chart with some market data). In a nutshell, it conveyed to market participants the sense that authorities were capable of agreeing on a forceful coordinated response to what appeared to be, at a time, a likely meltdown of the global financial system. Lets examine the outcome of the Washington meeting in some detail. The concluding statement is rather short by usual standards (5 pages) and fully concentrated on the situation in the financial markets and on the actions to be taken to stabilise them. The diagnosis of the crisis only mentions macroeconomic factors (global imbalances, macro-policies) in a oblique way, mentioning unsustainable global macroeconomic outcomes. This was reportedly because the Chinese did not
7 That was still a time in which many, in Europe and elsewhere, thought that the transatlantic repercussions of the financial crisis would remain limited. The euro, so it was thought, was providing and would continue to provide an effective shelter against global financial instability.

14

want global imbalances to be explicitly mentioned among the root causes of the crisis, as this would have suggested that their surplus was somehow to blame for it. Macroeconomic issues would come later. Here is a quote from the relevant section: During a period of strong global growth, growing capital flows, and prolonged stability earlier this decade, market participants sought higher yields without an adequate appreciation of the risks and failed to exercise proper due diligence. At the same time, weak underwriting standards, unsound risk management practices, increasingly complex and opaque financial products, and consequent excessive leverage combined to create vulnerabilities in the system. Policy-makers, regulators and supervisors, in some advanced countries, did not adequately appreciate and address the risks building up in financial markets, keep pace with financial innovation, or take into account the systemic ramifications of domestic regulatory actions. Major underlying factors to the current situation were, among others, inconsistent and insufficiently coordinated macroeconomic policies, inadequate structural reforms, which led to unsustainable global macroeconomic outcomes. These developments, together, contributed to excesses and ultimately resulted in severe market disruption. The action plan, in addition to committing to all the necessary macroeconomic stimulus (monetary and fiscal) with due regard to individual country needs and conditions a notion dear to the Europeans is divided in 5 sections, all focused on financial markets and financial institutions: strengthening transparency and accountability; enhancing financial regulation, promoting integrity; reinforcing international cooperation and reforming the international financial institutions (IFIs). The agenda is further broken down into detailed tasks, in two parts: immediate actions to be implemented by 31 March 2009; and medium term actions. The G20 ministers, supported by an enlarged Financial Stability Forum (FSF), standard setting bodies and national supervisors, as well as the International Monetary Fund (IMF), were entrusted with the task of monitoring and ensuring progress on the action list. In an informal way, the triangular governance structure (G20 on top, supported by FSF and IMF) that would be set up later, was taking shape. The Washington communiqu conveys a sense of urgency, focus and concreteness that could not be found in the traditional G7/G8 declarations. Instead of broad, often nebulous, open-ended political declarations encompassing a wide range of topics, it reads like what it is an extremely focused action plan. The traditional set of commitments that can suit every participant, because it merely restates what they are already committed to nationally, cannot be found anywhere in the text. Instead, the language is precise, even technical - specialised institutions in charge of carrying out

15

work the IMF, FSF, the Basel Committee, national regulators, etc. are named and they are given strict deadlines for implementation.8 3.2 London The London summit on 1-2 April 2009 is likely to remain in history as the moment when the international community united to ward off the risks of recession and protectionism. Indeed, it took place at a time when world output and trade were only beginning to stabilise and this wasnt known in real time. Observers were wondering whether the world was heading towards another Great Depression (Figure 6). In the same city where the countries participating in the World Economic Conference of 1933 had failed to find common ground, it provided a major impetus toward cooperation. Figure 6: Output and trade in the Great Depression and the Great Recession World Industrial Production World Trade

Source: Eichengreen and ORourke (2010).

The leaders convened in London just after the deadline of completion of the 'immediate actions' decided in Washington and aimed at restoring the functioning of financial markets. A Progress Report on the Washington Action Plan, published in London on 2 April, suggests with hindsight that most of the major areas of financial reform, such as bank capital strengthening, the definition of a new capital framework to avoid pro-cyclicality, new liquidity and risk management standards, IMF surveillance (including the decision to conduct a FSA for the United States), internal incentives and
8 For example, the first item of the financial regulation action plan, for implementation by 31 March 2009, reads The IMF, expanded FSF, and other regulators and bodies should develop recommendations to mitigate pro-cyclicality, including the review of how valuation and leverage, bank capital, executive compensation, and provisioning practices may exacerbate cyclical trends.

16

compensation practices, were underway, though in many cases they would be better defined later. Less clear was the progress achieved on transparency and accountability. As in earlier cases, progress towards harmonisation of accounting standards remained elusive. On financial regulation the London summit maintained the momentum launched in Washington, and also brought several new results. The final statement was again relatively short and to the point, explicitly indicating concrete actions to the undertaken. The leaders decided to reshape the FSF, transforming it into a Financial Stability Board (FSB) with broader representation (mirroring that of the G20) and an enhanced mandate. The FSB would, from then on, act as coordinator of all actions undertaken, in the area of financial regulation and supervision, by national and international standard setters. The leaders also started establishing the broad principles that would characterise, after a transition, the post-crisis bank capital standards: in the short term, until the macroeconomic recovery would strengthen, minimum capital requirements would remain unchanged or even decline, to facilitate lending; subsequently, prudential standard would be strengthened, building capital buffers above regulatory minima, increasing the quality of capital, mitigating the pro-cyclicality of capital ratios, (including a requirement that banks build capital buffers in good times), supplemented capital requirements with non-risk based measures, improving risk management incentives and enhancing liquidity buffers. In addition to capital standard, the London conclusions include provisions on hedge funds (registration, information gathering), credit derivatives (establishment of central counterparties), managerial compensation and bank board risk control responsibilities, credit rating agencies (registration and supervision, according to IOSCO rules). Still tentative were, on the contrary, the initiatives agreed in London concerning systemically important banks (SIFIs), shadow banks and accounting standards. The London summit also broadened the scope for action to include macroeconomic topics. The leaders stated that their countries were implementing fiscal expansion, monetary expansion and banking sector repair, indicating that these actions constituted the largest fiscal and monetary stimulus and the most comprehensive support programme for the financial sector in modern times and they pledged to conducting all [their] economic policies cooperatively and responsibly with regard to the impact on other countries not an insignificant commitment for a number of non-G7 countries which were used to regarding sovereignty over macroeconomic policy as nearly absolute. It is to be noted, however, that the declaration included no specific commitment in terms of either effort or date. The true innovation was the strengthening of the resources available to emerging and developing economies through the IFIs. The summit decided a $750bn overall increase of resources available through the IMF (of which $250bn in immediate, temporary bilateral loans, to be substituted at a

17

later stage by $500bn in form of an expansion of the Funds New Agreements to Borrow, plus SDR allocations for $250bn), as well as an increase in the capital endowments of the multilateral development banks, part of which earmarked for low income countries, and finally an increase in trade credit at the global level, to be channelled through the IFC. All this amounted to a historically unprecedented increase in the resources available to multilateral institutions. This support more than offset the restrictions that had been enacted in the preceding years, in a benign economic and financial environment. The reason for this massive increase in the resources available to support emerging and developing countries was the fear that at a time when advanced economies were struggling with the domestic fallout from the financial crisis, advanced and emerging countries would suffer from a major reversal of capital flows. Indeed, these countries, which had mostly been immune from the financial crisis until the Lehman shock, experienced a sudden stop in the third quarter of 2008. This was especially the case in emerging Europe and in Asia (Figure 7). Figure 7: Private capital Flows to Emerging Countries, 2001-2010
Percentage of recipient regions GDP

Source: Bruegel calculations with IMF data.

18

3.3 Pittsburgh Five months later (September 2009), the summit in Pittsburgh marked what can be regarded as a sort of watershed. In a number of ways, in a climate of low expectations Pittsburgh achieved important results, but also coincided with a marked slowdown in the productivity of the G20. By the time the leaders gathered it was clear that a recovery was under way and optimism was on the rise: the October IMF forecast envisioned 3.1 per cent world growth in 2010, against 1.9 for the April forecast. The sense of urgency that had characterised the two previous meetings had abated and the policy focus was shifting on making preparations for more normal times. Nevertheless, the communiqu unambiguously emphasised that it was yet too early to remove the stimulus: We pledge today to sustain our strong policy response until a durable recovery is secured. We will act to ensure that when growth returns, jobs do too. We will avoid any premature withdrawal of stimulus. At the same time, we will prepare our exit strategies and, when the time is right, withdraw our extraordinary policy support in a cooperative and coordinated way, maintaining our commitment to fiscal responsibility. A first important result attained in Pittsburgh concerned institution-building. The leaders decided that the G20 summit would become a regular event, replacing the G8 at the top of the international financial architecture. In this framework, Finance Ministers would act as deputies, preparing agendas and implementing decisions, supported by the FSB and the IMF the first responsible for financial markets, the second for macroeconomic surveillance and last-resort lending. The leaders also committed to a strengthening the voice of emerging and developing countries in the IMF through shifting by January 2011 at least five per cent of the quotas from overrepresented to underrepresented countries. However, because of US reluctance, they could not find agreement to announce that the heads of the Fund and the Bank would be appointed without condition of nationality. Regarding financial reform, a key challenge in Pittsburgh was to continue to exert guidance and preserve the reform momentum, while at the same time avoiding micro-management and excessive top-down command of what was bound to be increasingly technical discussions. In this area, the leaders struck a reasonable balance between direct guidance and delegation to the FSB and to the national authorities. On risk control the message was clear strengthen capitalisation; extend the focus to leverage and liquidity; avoid pro-cyclical regulation but implementation was left in the hands of the Basel Committee. A number of broad principles on compensation practices were established, but the task of working out the implications and, most importantly, to implement, was
19

left to the FSB and especially to national supervisors (which should have the responsibility to review firms compensation policies and structures with institutional and systemic risk in mind...). On moral hazard, the leaders asked for the establishment of bank resolution procedures and standing crisis management groups in all systemically important institutions, especially those operating across borders. The area where the G20 made the most important headway in Pittsburgh was that of global imbalances. The issue had been left aside in the two previous meetings, first to avoid turning the summits into US-Chinese confrontations and second, because the priority was to address financial regulation failures and the common risk of a depression. But it was becoming too important an issue for it to continue being avoided. First, in discussion on the causes of the crisis, the idea was gaining ground that large and persistent payments disequilibria among currency areas were among the contributors to systemic risk-building in previous years9. Second, the IMF was projecting for the medium term a rebound of imbalances (the October 2009 WEO envisaged that they would stabilise at about 2 per cent of world GDP a projection that has not changed much since against 2.5 to 3 per cent in 2006-2007). Third, it was feared that in the years ahead demand in the advanced countries would remain subdued because of the extent of deleveraging and the coming fiscal retrenchment, and that to sustain global growth there was a need to foster demand in the emerging countries. On the eve of Pittsburgh it seemed unlikely that leaders would enter this contentious territory, but the outcome surpassed expectations. At the initiative of the USA, a 'framework' for macroeconomic policies was agreed upon, with the aim of making national policies ('fiscal, monetary, trade, structural') consistent with balanced growth, including regular consultations on commonly agreed policies and objectives. Agreement on the Framework represented an ambitious international coordination endeavour, which set differentiated goals to deficit countries and surplus countries10. Furthermore, the leaders instructed their Finance Ministers to put in place a surveillance process, the 'Mutual Assessment Process' (MAP) to evaluate the collective implications of national policies for the world economy. The IMF was asked to give technical support to this exercise in multilateral surveillance, working with G20 ministers and central banks, and to report regularly to the G20 leaders. Unimportantly for the short term, but noticeably, Pittsburgh was also characterised by a significant broadening of the agenda: for the first time the communiqu mentioned at length energy security,

This link was made explicit in two influential European reports by Jacques de Larosire (2009) and Adair Turner (2009) on the reform of financial regulation. However, the idea remained disputed among academics and policymakers. 10 As argued by Vines (2011).

20

climate change, poverty, jobs quality, and trade and investment. All important issues for sure, but on which commitments were more verbose than on core G20 business. 3.4 Toronto In the year following Pittsburgh, the G20 calendar included two summit meetings under a joint Canadian-Korean chair: Toronto (June) and Seoul (November). In our view the Toronto meeting on 27 June 2010 marked the lowest point in the short G20 history. The macroeconomic coordination framework virtually stalled against the test of delivering a joint assessment. The IMF, which was steering the process although its responsibility was in principle one of technical assistance only, was reluctant to move too fast to policy conclusions as it wanted first to educate governments to the process of information-sharing. At the same time specific circumstances contributed to rendering agreement difficult. In spring 2010 it was undisputable that the recovery was underway, but the fears of double-dip recession were still widespread. In all countries monetary and fiscal authorities, while beginning to think about 'exit strategies' and refer to them in public communication, maintained de-facto an accommodative stance. But the awareness was increasing that the revenue shortfall provoked by the recession (and to a lesser extent by the stimulus) would give rise, in time, to a historically unprecedented debt explosion. In the midst of the trade-off (expand now, consolidate later), positions were divided, Europeans being especially concerned about fiscal risks after the outbreak of the Greek crisis (which had led, weeks earlier, to what many regarded as a first controversial step towards a fiscal union, the creation of the European Financial Stability Facility) and the US, where unemployment had risen much more, being still decisively positioned in favour of fiscal expansion. The majority of participants in the summit emphasised the fiscal risks and the need for consolidation, as evident in this passages from the final statement: Recent events highlight the importance of sustainable public finances and the need for our countries to put in place credible, properly phased and growth-friendly plans to deliver fiscal sustainability, differentiated for and tailored to national circumstances. Those countries with serious fiscal challenges need to accelerate the pace of consolidation. and Sound fiscal finances are essential to sustain recovery, provide flexibility to respond to new shocks, ensure the capacity to meet the challenges of aging populations, and avoid leaving future generations with a legacy of deficits and debt. [] Those with serious fiscal challenges need to accelerate the pace of consolidation. Fiscal

21

consolidation plans will be credible, clearly communicated, differentiated to national circumstances, and focused on measures to foster economic growth. Not least because of a convergence of views between the new British Prime Minister, David Cameron, and German Chancellor Angela Merkel, the tone of the final statement concerning macroeconomic policies was surprisingly emphatic about the need for fiscal consolidation, an orientation that left the US uncomfortably isolated. The advanced G20 countries agreed to halve fiscal deficits by 2013 and to reduce public debt ratios by 2016 if not before. The whole macroeconomic discussion was very much reminiscent of traditional US-European disputes, with the emerging countries playing a secondary role. In this respect, as well as in the fairly general tone of the discussion, the Toronto G20 summit was closer to a traditional G7 summit than to the meetings in London and Pittsburgh. At the same time the financial reform agenda had visibly slipped off the hands of the heads of state and government. It was left entirely in the hands of the FSB, without further political stimulus or guidance. This was probably an inevitable development in view of the technicality of the topics, but contributed to give the impression that the Toronto summit had very little to deliver. 3.5 Seoul The Seoul G20 summit was the first to be chaired by a non-G7 country and Korea was especially keen on making it a success. It was not an easy task: on the macroeconomic front, a swift return to normality was reviving old problems. The currency dispute between the US and China was again making headlines and further controversies had erupted as many emerging countries were overwhelmed by capital inflows. 'Currency wars', in the words of Brazilian Finance Minister Guido Mantega, was the theme of the day. Simultaneously, the Mutual Assessment Process was proving more cumbersome and controversial than expected. In this climate, the Korean presidency was not able to achieve breakthrough on the controversial issues. An attempt was made to open a new road towards compromise between China and the US by stating that current account balances should remain below 4 per cent of GDP, but no agreement could be found in time for the summit.11 Heads of state and government could only agree to name the problem (a change from the initial reluctance to mention imbalances) and call for further work by their Finance Ministers. The corresponding sentences in the communiqu were particularly convoluted: Persistently large imbalances, assessed against indicative guidelines to be agreed by our Finance Ministers and Central Bank Governors, warrant an assessment of their
11 According to former UK Prime Minister Gordon Brown (2011), this was because of premature US public declarations.

22

nature and the root causes of impediments to adjustment as part of the MAP, recognizing the need to take into account national or regional circumstances, including large commodity producers. These indicative guidelines composed of a range of indicators would serve as a mechanism to facilitate timely identification of large imbalances that require preventive and corrective actions to be taken. An outcome from the MAP was a set of 'policy commitments' by the G20 members, in which each country indicated its policy goals in several fields (fiscal, financial, monetary and exchange rate, structural, development, other policies) and how they contributed to the Framework goals. The very fact that all participants agreed to commit to policy actions vis--vis partners was politically significant, but the commitments themselves did not represent departure from pre-existing policy course. Seoul was more successful on two other fronts: first, it delivered ahead of time on the Pittsburgh commitment to reform IMF governance, with a reform that shifted 6 per cent of quota shares toward under-represented countries.12 Second, it could take stock of agreement reach in the Basel Committee to revise bank capital adequacy ratios. Korea had been keen to open two new chapters in the international discussion. It had first proposed reflection and action on what it called 'financial safety nets', in other words mechanisms for giving countries access to liquidity when facing capital outflows. What was intended through providing better insurance was to remove the motive for self-insurance through reserve accumulation, which made discussion on the issue another, more systemic contribution to the discussion on global imbalances. The validity of the theme, that could encompass IMF facilities, regional agreements and swap agreements with the major central banks, was widely recognised, but achievements remained limited. The IMF announced in August 2010 a new low-conditionality facility, the Precautionary Credit Line (PCL) to complement the pre-existing Flexible Credit Line (FCL), but no agreement could be found on the more ambitious Global Stabilisation Mechanism (GSM). The theme however served as a bridge to discussions on reforming the International Monetary System under French presidency. The other chapter was development. It was indicative of the broadening of the G20 agenda from an initial focus on financial regulation and crisis management to a much broader, potentially very large set of issues. Clearly, Korea had substantive and political motives to open a new chapter and so would future presidencies. But the downside was, inevitably, a lack of focus and significant deliverables.

12 Agreement was reached in a Finance Ministers meeting in Gyeongju on 22-23 October 2010.

23

3.6 Cannes The 2011 French presidency started off with an ambitious agenda. President Sarkozy announced from the outset that sticking to the agenda [of completing work in progress] would condemn the G20 to failure and the world to new crises13 and he proposed to add new items to the agenda. In the event the summit itself was dominated by developments of the Greek crisis and a dispute among Europeans about the resources of the euro-area financial facility and proposals to increase of the resources at the disposal of the IMF for European operations.14 Nevertheless preparatory work covered a broader scope, very much in line with the three new French priorities and the leftovers from previous presidencies. The first new topic was the reform of the International Monetary System, which had not been addressed explicitly under the previous presidencies, though it was implicit in several topics under discussion such as the 'international financial safety nets'. The theme initially ('a monetary system to succeed Bretton Woods') was broad and grand enough to leave room for many interpretations, but discussions among Finance Ministers led to more modest outcomes: a broad consensus on the management of capital flows (but not the code of conduct initially contemplated), principles for cooperation between the IMF and regional financial arrangements, an action plan to develop local bond markets, and (instead of an actual inclusion of the renminbi) a commitment to review the composition of the Special Drawing Rights (SDR) basket in 2015, with a view to enlarging it. France had hoped to be supported by China in its reform efforts, but the Chinese government remained throughout the discussions more cautious than suggested by a widely-quoted 2009 statement by the governor of its central bank (Zhou, 2009). The second element added was an entirely new chapter with on the volatility of commodity prices. The initial aim was to make progress towards limiting volatility through the regulation of derivative markets and improvements in market transparency, while at the same time finding ways to limit the adverse consequences of volatility. Here again, the outcome was more modest: the most notable achievement was the creation of an Agricultural Markets Information System. The third new element marked for the G20 the beginning of a more reflexive approach to its own role. Indeed it was not entirely clear whether the G20 was intended to substitute existing structure (as suggested by the statement that it had become 'the premier forum' for coordination), or to complement them. At the request of the French presidency, British PM David Cameron presented a report on G20 governance where he emphasised the need for the G20 to remain an informal body,

13 Speech at the French Ambassadors Annual Conference, 25 August 2010. 14 An outline agreement on a special SDR allocation was rejected by Germany at the request of the Bundesbank.

24

reach out to non-members, and prioritise areas where existing global governance structures were in need for improvements. These conclusions were endorsed by the summit, which made clear that the G20 is part of the overall framework of international governance. Following Camerons recommendations, the decision was also taken to formalise the Troika of past, present and future presidencies. France also followed Korea and the US with a renewed stimulus towards the set-up of a scoreboard to monitor macroeconomic developments and policies, with a view to reducing global imbalances. The 2011 ministerial meetings led to an agreement on a two step approach: in the first stage, countries whose size and/or conditions may imply systemic risks for the global economy would be singled out, based on a limited set of indicators and a mix of analytical and statistical approaches. In the second stage, an in-depth analysis would be conducted on these countries15. This outcome is valuable because it breaks the stalemate prevailing in earlier meetings, in which leaders and ministers alike had been unable to give shape to their commitment to operationalise the 'Framework for sustainable and balanced growth' agreed in Pittsburgh. Furthermore, in the context of renewed economic concerns the Action Plan for Growth and Jobs agreed upon at the summit was fairly specific on the policy measures expected from the participants.16 3.6 Summing up From November 2008 to the end of 2011 the G20 seems to have gone through a cycle. The initial, 'emergency' period (Washington and London summits) was marked by swift action on financial reform and crisis mode. The Pittsburgh summit, while very effective in terms of institution building (establishment of a permanent G20, plus the macroeconomic 'Framework') marked the transition to the second stage, in which, in the context of fragile economic normalisation, renewed asymmetry between advanced and emerging countries, and reduced financial market tension, progress became slower.

15 The negotiation was particularly long and complex, largely around semantics. Reportedly, in the Paris meeting the drafting came to a gridlock over the acceptable use of the term current account, and the meeting lasted more than 14 hours. In the end, as reported by the press, the list of countries singled out for second-stage examination, kept confidential, includes the US, Japan, Germany, France, the UK, China and India. 16 For example, Italy committed to reaching a rapidly declining debt-to-GDP ratio starting in 2012 and close to a balanced budget by 2013 while Australia, Brazil, Canada, China, Germany, Korea and Indonesia, where public finances remain relatively strong, taking into account national circumstances, [agreed] to let automatic fiscal stabilisers work and, should global economic conditions materially worsen, [agreed] to take discretionary measures to support domestic demand as appropriate, while maintaining their medium-term fiscal objectives.

25

4.

Measures of success

We now move on to assess achievements. Our first approach is to employ a set of performance scores available in the literature, that compiled by the University of Toronto. We then dig deeper with specific assessments of achievements in the two fields where the G20 has focused during its first three years, financial regulation and macroeconomic policies. 4.1 Overall compliance We start by examining the compliance of the G20 with its own stated objectives, using the indicators elaborated at the University of Toronto (UoT). Researchers at this university have watched, for several years using a consistent methodology, the G8 and G20 processes, and compiled detailed scoreboards. There is a farly long time series of these scoreboard for the G8, and a much shorter one for the G20. In a nutshell, the methodology is the following: UoT research teams catalogue, for every G20 summit, the commitments expressed in the final statement, and then monitor compliance with these commitments in the period up to the next summit. For each commitment and each meeting, each country is judgmentally assigned a value of 1 if the commitment was fulfilled fully or nearly, of 0 if the commitment could not be fulfilled or was fulfilled only to a limited extent, and -1 if the country did not act17. The next step is to calculate average measures of the degree o compliance, separately for each meeting or each topic. A positive value means that there was at least partial compliance on average with the commitment made, while a value close to 0 or negative signals limited or no compliance. Table 2 reports simple average scores, divided by topic (upper part of the table) and meetings (lower part). Among the topics, for simplicity we have restricted attention to four: macroeconomic policy, financial reform, IFI reform and others (a heterogeneous mix including trade, development, climate change, terrorist financing and money laundering, etc). The last summit for which scores are available is Toronto. We calculated scores for the G20 as a whole and for the following sub-groups of members: the advanced countries, the advanced countries with a current account deficit, the emerging countries, the emerging countries with a current account deficit, the G7 and G7 Europe. We have also calculated the standard deviation across the G20, as a measure of the cross-country dispersion of the degree of compliance.

17 The basic methodology is due to von Fuerstenberg and Daniels (1992). See detailed explanations in the technical notes available from the University of Toronto G8 Information Center website (http://www.g20.utoronto.ca/analysis/index.html)

26

Table 2: Unweighted compliance scores 2a Scores by topic

Macroeconomic policy 0.55 0.66 0.67 0.46 0.32 0.65 0.73 0.31 Financial Reform 0.27 0.57 0.60 -0.06 -0.04 0.63 0.74 0.41 IFI 0.48 0.89 0.83 0.10 0.20 0.93 0.88 0.49 Other 0.31 0.58 0.62 0.03 0.13 0.60 0.66 0.35 Average 0.40 0.67 0.68 0.13 0.15 0.70 0.75 0.39

G20 Advanced Advanced in deficit Emerging Emerging in deficit G7 G7 Europe Memo: St-Dev G-20 total

2b. Scores by summit

Washington 0.51 0.72 0.80 0.19 0.28 0.72 1.00 0.59 London 0.23 0.49 0.57 -0.04 0.08 0.54 0.65 0.43 Pittsburgh 0.24 0.58 0.56 -0.08 -0.08 0.57 0.59 0.45 Toronto 0.28 0.58 0.59 -0.02 -0.09 0.59 0.61 0.34 Seoul 0.50 0.66 0.71 0.32 0.36 0.65 0.71 0.24 Average 0.35 0.61 0.65 0.08 0.11 0.61 0.71 0.41

G20 Advanced Advanced in deficit Emerging Emerging in deficit G7 G7 Europe Memo: St-Dev G-20 total

Source: University of Toronto G20 Information Center, authors calculations

Two pieces of evidence are immediately apparent. First, almost all numbers are positive. This means that, based on the criteria used, there was at least some compliance in most cases. For the G20 as a whole, and as an average across all meetings, the score is 0.40, with a standard deviation of 0.39, which means that a large part of the distribution, including that comprised between sigma, lies in the positive range. The only negative numbers regard emerging countries in the area of financial reform. Surprisingly, the overall G20 compliance in the area of financial reform is rather low (0.27). This, however, depends entirely on the negative score of emerging countries (where financial reform may be perceived as less urgent that other areas), and on the fact that the averages reported in the table are unweighted (which implies for example that the US, with a GDP share in total G20 close to 25 percent and an even larger share in terms of financial market size, is weighted equally to Indonesia, that has a share of 1 percent in total GDP). This suggests that performance in this domain tended to be stronger for large developed countries. Weighted scores will be presented later.

27

The second finding is that compliance drops after the first meeting (Washington), remained low for London and Pittsburgh, and recovered somewhat for Seoul. Interestingly, both the decline after Washington and the recovery for Seoul are visible in both the advanced and the emerging country groups, which suggests that there was a genuine meeting effect. However, the two groups of countries are characterised by sharply different levels of compliance, the advanced ones having a much higher score than the emerging ones. To the extent that this concerns financial reform, the signal is not necessarily disappointing, since reforming the financial sector was (and remains) a priority principally in the advanced countries. The G20 financial reform agenda was meant to trigger action in the advanced countries and it should be no surprise to observe that it is in these countries that it was most effective. Note that, after London, the average compliance increased steadily, with a somewhat stronger increase in the Seoul meeting. This suggests that the commitments, while perhaps no longer constituting breakthroughs relative to expectations, gradually became more realistic and achievable. More generally, the indicators in the table measure the effectiveness or success in translating G20 decisions into national legislation, whereas our qualitative judgements in the previous sub-section were based on the communiqus after G20 meetings. The two approaches are different, and indeed complementary because the earlier one focused on the quality of the commitments expressed at the meetings, while the indicators measure the extent to which those commitments wee actually put into practice. One striking implication of this difference regards the judgement on the London meeting. This meeting's outcome was judged by us and others very favourably, based on the commitments expressed in the final statement. However, compliance with those commitments appears to have been less positive. Implementation was still good in 'resisting protectionism and promoting global trade and investment', but it was low in 'ensuring a fair and sustainable recovery for all' (official development assistance, particularly for the poorest countries) and 'the scope of regulation' (financial regulation). It should be mentioned that the increase of IMF resources, a major achievement of London, is not among the 'priority commitments' chosen by the UoT to measure expost compliance with that meeting. The reason is unclear, hence this may introduce a negative bias in the ex-post judgement of the London meeting. It is interesting to note that the UK was the only country that complied fully with all commitments on that occasion, according to the UoT indicators; a possible 'chair effect'.

28

Among the topics, the highest scores are obtained by macroeconomic policies and IFI reform. Again, advanced countries show higher compliance18. On macro policies, the positive score, despite slow progress in setting up the 'Framework for Sustainable and Balanced Growth', is attributable to the timely enactment of the stimulus policies agreed mainly in the Washington meeting. Table 3 reports similar scores, but weighted by GDP19. The main difference with the earlier table is in the much higher score observed in the area of financial reform, that in many cases now surpass that of macro policies. This shows that the degree of compliance in the commitments relating to finance and banking was highly skewed towards the large countries: among the advanced countries the US, and among the emerging ones China and India. All in all, the GDP-weighted scores are uniformly higher across the table. Table 3: Average compliance scores (GDP-weighted) 3a Scores by topic
Macroeconomic policy 0.47 0.44 0.36 0.49 0.25 0.42 0.74 0.31 Financial Reform 0.52 0.63 0.67 -0.05 -0.12 0.65 0.75 0.41

G20 Advanced Advanced in deficit Emerging Emerging in deficit G7 G7 Europe Memo: St-Dev G-20 total

IFI 0.64 0.95 0.93 0.06 0.06 0.97 0.90 0.49

Other 0.44 0.49 0.47 0.08 0.13 0.49 0.66 0.35

Average 0.52 0.63 0.61 0.14 0.08 0.63 0.76 0.39

3b Scores by summit
Washington 0.77 0.76 0.89 0.23 0.11 0.76 1.00 0.59 London 0.36 0.46 0.49 -0.14 0.02 0.47 0.68 0.43 Pittsburgh 0.37 0.55 0.54 -0.06 -0.27 0.54 0.61 0.45 Toronto 0.43 0.49 0.46 0.14 -0.03 0.48 0.60 0.34 Seoul 0.58 0.55 0.54 0.40 0.39 0.54 0.69 0.24 Average 0.50 0.56 0.58 0.11 0.04 0.56 0.72 0.41

G20 Advanced Advanced in deficit Emerging Emerging in deficit G7 G7 Europe Memo: St-Dev G-20 total

Source: University of Toronto G20 Information Center, authors calculations

18 The compliance with the IFI reform commitments is calculated with reference to the national ratification of the international accords, and/or with the countries taking an active stance in addressing the reforms. This explains why compliance scores differ across countries, even if these reforms are collective in nature. 19 Details on calculations and weighting. Are in the appendix.

29

Table 4 shows a comparison of historical records between the G20 and the G7/G8. The message here is that the compliance of the G20 is not much worse than that of its much older and more experienced 'brother'. This is a non-negligible achievement in view of the ambitious agenda of the recent G20s and of the much more precise character of the commitments. Again, the difference that exists is attributable to the lower level of compliance in the emerging components of the G20, considering that the scores of the US and Europe are not much different in the two 'Gs'. Table 4: Comparative performance of G20 and G8
G20 (established 2008) London Group average US EU 0.23 0.4 0.6 Pittsburgh 0.24 1 0.38 Toronto 0.28 0.33 0.57 Avg 0.25 0.58 0.52 G7/G8 (established 1975) 1990s 0.37 0.47 0.37 2000s 0.49 0,44 0.54 Avg 0.44 0,45 0.47

Calculations based on data from the Compliance Reports of the G20 and G8 Information Centers of the University of Toronto. The entries are simple averages of elementary compliance scores, ranging from -1 (no compliance) to +1 (full compliance). Elementary scores are judgemental estimates and are reported separately for each country and for groups of main agenda items (macro policies, financial regulation, development aid, trade, etc.)

The staff of the IMF also prepared for the Cannes summit a 'G20 Accountability Report' (IMF, 2011d) that reviews the implementation of the Toronto and Seoul summits in four areas (monetary and exchange rates, fiscal, financial and structural policies). However, this report provides a broad assessment of the appropriateness of the policy course in G20 countries, taking as benchmarks the summit communiqus but also policies that Fund staff considers appropriate in view of the evolution of economic conditions. Furthermore, it does not provide country-by-country assessments. For this reason it cannot be used as a basis for a systematic assessment. Overall, the report does not contradict the above findings. It assesses that there has been more progress toward satisfying the letter of the Summit declarations than addressing the difficult reforms needed for long-run sustainability and balanced growth and finds that progress on the financial and structural reforms has been slow. 4.2 Financial sector reform Rottier and Vron (2010) have conducted a more in-depth, detailed examination of the implementation of the leaders commitments at the Washington summit, with a focus on financial regulation. They therefore provide a useful complement to the UoT assessment. Rottier and Vron list 47 action items in the financial communiqu, 39 of which relate to financial regulation. They consider three criteria:

30

Effectiveness of implementation, grading it from 0 (nothing done) to 5 (fully implemented); Cross-border consistency of implementation measures, ranking from 0 (no coordination) to 3 (fully consistent or harmonised); Follow-up initiatives (scoring from 0 to 2).

Rottier and Vron do not grade countries separately but they distinguish between channels for implementation, depending on the degree to which implementation relied on voluntary action by national authorities or, on the contrary, fell into the remit of international institutions. This approach leads them to distinguish four groups (Table 5). Table 5: Rottier and Vrons assessment of the implementation of the Washington commitments Effectiveness score (betw. 0 and 5) Group 1 (national authorities) Group 2 (FSB) Group 3 (BCBS, IASB, IOSCO, IAIS, FATF) Group 4 (IMF, OECD, World Bank)
Source: Rottier and Vron (2010).

Consistency score (betw. 0 and 3)

Follow-up score (betw. 0 and 2) 0.6 1 1.28 1.45

Total score (betw. 0 and 10) 3.2 4.63 6.28 7.73

2 2.63 3.28 3.82

0.6 1 1.72 2.45

What is interesting in these results is that they highlight the degree to which the effectiveness of the G20 depends on that of the implementing agent national authorities or international agencies. Rottier and Vron show that the latter have on average been more effective much more when they could claim full authority as opposed to a mere coordination role. The difference in the quality of implementation is unsurprisingly especially striking as regards cross-border consistency. This analysis portends more general conclusions, to which we will return in the discussion on governance. The G20 is a political body and it does not have sub-structures to rely on, not even a secretariat. Depending on the topic, it has therefore relied on voluntary, non-binding commitment by national authorities, on stricter forms of cooperation, or on supranational regulation. Unsurprisingly, Rottier and Vron find that the first one, which is the more politically expedient form of coordination, is also the less effective.

31

4.3 Macroeconomic coordination: Phase 1 As indicated in section 3, G20 macroeconomic coordination went through two successive phases. In the first one, from Washington to Pittsburgh included, the focus was on stimulating the economy across the board. All countries were requested to contribute, to the extent permitted by the domestic fiscal situation. In the second one, from Toronto onwards, it shifted towards a more complex set of objectives, with the aim of combining continued support for growth, budgetary consolidation, and the prevention of imbalances. There was also much less consensus in the second phase than in the first one. Achievements of the first phase are relatively easy to assess. On the fiscal front, data confirm that a stimulus was engineered not only in the advanced G20 group but also, and to a broadly similar extent, in the emerging group (Figure 8). Russia, India and China were among the countries where the 2008-2009 stimulus effort was the largest.20 Whereas it was natural for emerging countries to rely more on discretionary action because government expenditures represent a lower fraction of GDP and, as a consequence, automatic stabilisers are smaller, the full participation of the emerging group to the concerted stimulus was nevertheless a remarkable event. Emerging countries were traditionally viewed as passive players in a global macroeconomic coordination game dominated by the members of the G7. The fact that they fully took part in the stimulus was indicative of their new global role and was an ex-post vindication of the very creation of the G20.21

20 A problem with measuring the fiscal impulse by changes in the structural primary balance, as done here, is

the latter is affected by changes in the evaluation of potential output or the elasticity of taxes. These technical revisions are normally of limited impact, but they were substantial in the aftermath of the financial crisis. For this reason our measurement is likely to overstate the size of the stimulus. A better approach is what the IMF (2010) called the action-based approach. However available data do not allow implementing it for the G20 countries.
21 There was significant heterogeneity within the emerging G20 group, but not markedly more than within the advanced countries group.

32

Figure 8: Fiscal Impulse in the G20, 2008-2010

Note: Fiscal impulse is measured by the change in the cyclically-adjusted primary balance. Data are from the IMFs Spring 2011 Fiscal Monitor.

There was more heterogeneity on the monetary front because situations differed markedly. In Europe and the US, central banks had had to resort to enhanced credit or liquidity support policies as early as in 2007, more than a year before the Washington summit, but no such action was in order in Japan or the emerging world. Even after the Lehman shock, access to domestic-currency liquidity remained much less problematic in the emerging world and in Japan that in the US and Europe. What is harder to assess is to what degree action undertaken at national level was triggered by G20 coordination. In a situation of global demand shortfall, high risk aversion and partial paralysis of financial markets, the policy prescription was about the same everywhere, irrespective of the degree of coordination. It is likely, however, that the G20 meetings helped focus the policymakers attention on a well-defined policy package, facilitating also domestic consensus, and that they made each and every government more secure that it would not act in isolation. Implementation of the multilateral part of the G20 crisis agenda is by contrast easy to assess. As indicated in section 3, the London summit agreed on a $500bn increase in IMF resources (through temporary bilateral loans for $250bn, to be later substituted by an increase of the NABs), and on a special allocation of SDRs. The increase in IMF resources was enacted swiftly and made possible a large increase in lending through standard programmes, as well as the granting of credit lines to selected countries through two new facilities, the FCL and the PCL. Figure 9 plots the evolution of IMF resources and commitments from 2007 to 2010. It shows that without the replenishment of resources at the time of the London summit the commitment capacity
33

of the Fund would have been severely constrained already in 2009. It also indicates that by end2010, half of these new resources had been almost entirely used up, as the Funds uncommitted usable resources and one-year commitment capacity were both at, or slightly above their 2007 level.22 With hindsight, the increase in Fund resources seems to have been of the right magnitude, at least taking into account the size of the subsequent assistance programmes.23 Figure 9: IMF resources and commitments, 2007-2010

Source: IMF, Authors calculations.

Other initiatives with the Fund have not yet delivered major results. By end-2011 only three countries, Colombia, Mexico and Poland, had had access to the Flexible Credit Line (FCL), and only one, the FYROM (Macedonia) had had access to the Precautionary Credit Line (PCL). None had drawn on these facilities. As to the exceptional $250 SDR allocation, subsequent data suggest that effective usage of SDR by IMF members was limited and restricted mainly to small countries several of which under IMF programme. It seems unlikely on this basis that the allocation contributed significantly to revive global demand and growth.

22 As the expanded NABs were only made effective in March 2011, resources recorded in Figure 3 only include $250bn in temporary bilateral loans to the Fund. 23 In 2008-2010 the IMF concluded 19 stand-by agreements, including large-scale ones with Greece, Romania and the Ukraine and four extended arrangements, including a large-scale one with Ireland.

34

4.4 Macroeconomic coordination: Phase 2 Whereas Phase 1 was conceptually simple, Phase 2 was more complicated because it was meant to involve a differential treatment of participating countries. The intellectual background to it was the Obama administrations and the IMFs fear that the recovery would leave international imbalances largely untouched. Writing at the end of 2009, Blanchard and Milesi-Ferretti (2009) were warning that One of the three central adjustments emphasized in the earlier multilateral consultations has taken place, namely the increase in U.S. private saving. Two remain to be implemented, lower fiscal deficits in the U.S., and lower current account surpluses in China and a number of other emerging market countries. If these do not take place, there is a high risk that the recovery will be weak and unbalanced. Staying in midstream is dangerous. Against this background the goal, to quote from the Pittsburgh declaration, was to develop a forward-looking analysis of whether policies pursued by individual G20 countries are collectively consistent with more sustainable and balanced trajectories for the global economy that would feed into the leaders discussions, which could result in agreeing on actions to meet our common objectives. This was the purpose of the 'Mutual Assessment Process' (MAP). The name of the game was to make all participating governments more conscious of the international spillover effects of their actions and, through peer pressure, to lead them to amend policy course in case of global inconsistency. The initial strategy for making coordination work was to ask each country to submit medium-term policy frameworks and plans. The staff of the IMF was entrusted with the task of checking consistency of national assumptions and policy directions, providing feedback to G20 members and evaluating policy alternatives. This was intended to be a multi-stage iterative process involving: (1) initial submissions by G20 governments; (2) aggregation and multilateral consistency check by the IMF; (3) evaluation of alternative policy paths by the IMF; and (4) discussions on policy adjustments among G20 members (Box 1).24

24 See the communiqu of the St Andrews G20 Ministerial Meeting in November 2009, IMF (2009) and IMF (2010b, 2010c and 2011c).

35

Box 1: Early steps of the Mutual Assessment Process The MAP started in early 2010 with the submission by G20 members of medium-term 'policy frameworks' (read: projections) by individual G20 members according to a common template. These 'raw' projections included a number of cross-country inconsistencies because they had been prepared at different times, did not cover the same period, were based on different assumptions as regards, for example, the price of oil, or evaluated differently the consequences of the crisis on potential output. Furthermore, each included (explicitly or implicitly) assumptions as regards policy and outcomes in the partner countries that were not mutually consistent. These 'raw' projections served as inputs for preparation by Fund services of a 'refined' scenario exempt of major inconsistencies. This sometimes resulted in significant adjustments to individual country projections (for example, US GDP was revised downwards by 0.8 per cent over the 2010-2014 period). In a second step, the Fund evaluated through model simulations whether concerted policy actions by the major players could improve the global outcome. Three 'layers' of action were identified and evaluated: (a) reforms to boost internal demand in emerging surplus countries (primarily China); (b) further fiscal consolidation in advanced economies (primarily the US); (c) across-the-board structural reforms to increase the growth potential. This set of policies was evaluated to result in higher growth, faster fiscal consolidation and a reduction in global imbalances. The MAP exercise was carried out in Spring 2010 for the preparation of the Toronto summit and in Autumn 2010 for the preparation of the Seoul summit. Growth projections in the second vintage of submissions was more cautious than in the first one, that Fund staff had assessed too optimistic. This suggests that there has been some learning in the process.

As conducted for the Toronto and Seoul meetings, the MAP was a cumbersome exercise technically and it resulted in projections of uncertain accuracy. Discrepancies between MAP and World Economic Outlook projections are supposed to signal biases in the evaluation by G20 countries of the likely global outlook in its report for the Cannes summit, for example, the IMF staff (2011c) assessed national projections underlying the MAP outlook as 'too sanguine' but they could also indicate forecast errors by Fund staff. The coexistence of two sets of projections, both of which emanate from the Fund, was also confusing for observers and policymakers. Furthermore, the MAP were not a necessary input to policy simulations: those could equally be carried out on the basis of WEO projections. The value of the MAP was probably more in the process leading to them: whereas WEO projections are entirely of the responsibility of Fund staff, MAP projections were the result of a

36

bottom-up process. More than a top-down exercise, this may have facilitated genuine discussions on the challenges facing the world economy and national economic policies. As indicated above, at the Seoul meeting it was agreed to 'enhance' the MAP by outlining 'concrete policy commitments' for each of the members and by assessing 'the nature and root causes of impediments to adjustment' behind 'persistently large external imbalances'. This agreement by itself a progress in comparison to the initial silence on the issue of global imbalances opened the way to a more ambitious attempt at multilateral surveillance.25 Open discussion may or may not trigger policy action. As already indicated, a set of indicators and guidelines intended to help tackle global imbalances through policy adjustment in the key countries was adopted in April 2011 at the G20 ministerial in Washington, DC (Box 2). These indicators were in turn used by the IMF staff to identify seven key countries experiencing imbalances, to provide a broad-brush assessment of their underlying causes, and to make corresponding recommendations (IMF, 2011c). In effect, the Fund essentially indicated that imbalances had been driven by saving behaviour and it recommended fiscal consolidation for some (France, Japan, the UK, the US and India), the removal of distortions that keep Chinese savings artificially high, and measures to lower corporate savings in Japan and Germany. These recommendations were in part taken on board in the Cannes G20 Action Plan: leaders expressed agreement to differentiated budgetary consolidation strategies, including through letting automatic stabilisers work in Australia, Brazil, Canada, China, Germany, Korea and Indonesia (without excluding further discretionary stimulus if needed). This was a non-negligible achievement but, obviously, does not guarantee implementation.

25 Building on the agreement, IMF staff prepared for Cannes a whole set of reports: an Accountability Report (vis--vis the Seoul commitments), a MAP Report, a Sustainability Report (with appendix reports for seven countries identified as experiencing imbalances: the US, China, Japan, India, France, Germany, and the UK) and an Umbrella Report. See IMF (2011c).
37

Box 2: Indicators and guidelines for identification of required policy action The G20 Finance Ministers in February and April 2011 agreed on: A process leading to the identification of countries whose policies deserve closer examination and discussion. A set of indicators to monitor. These are (i) internal imbalance indicators (public debt and fiscal deficits; private savings rate and private debt) (ii) external imbalance indicators (current account balances, though they are not named because of Chinese reluctance to have them explicitly included in the list). External imbalance assessment is to take due consideration of exchange rate, fiscal, monetary and other policies. Indicative guidelines against which each of these indicators are to be assessed. It is stated that while not policy targets, these guidelines establish reference values for each available indicator allowing for identification of countries for the second step in-depth assessment. Four approaches to assess individual country positions. These are (i) a 'structural approach' presumably inspired by the IMFs GGER methodology for the assessment of equilibrium exchange rates (IMF, 2006); (ii) a statistical approach which benchmarks G20 countries on the basis of their national historical trends; (iii) a statistical approach which benchmarks G20 country's historical indicators against groups of countries at similar stages in their development; (iv) a statistical approach which draws on data, benchmarking G20 country's indicators against the full G20. The three statistical approaches are primarily based on data for the 1990-2004 period and they are expected to be based on simple methodologies. In all cases, forecasts for 2013-2015 are to be assessed against the four guidelines. A categorization of countries into two groups: seven systemic countries, and the rest of the G20. Selection criteria will be stricter for the second group, so that they will only be selected for review if they depart significantly from benchmarks. The goal is to help the process focus on the most important countries presumably again the US and China.

Whether the MAP will have lasting traction and help change the policy conversation in the main participating countries also remains to be seen. One potential difficulty is that it de facto focuses on a narrow subset of countries those with large external surpluses and deficits but it is not clear which of these countries is ready to trade a change in its own policy for a change in its partners policy. Would, for example, a Chinese exchange-rate adjustment facilitate a US budget agreement? The political economy of international horse trading is highly uncertain.

38

Finally, a potentially problematic issue is that the whole exercise is predicated on the assumption that global imbalances remain a serious concern for the world economy going forward. Indicators, guidelines and processes may serve coordination well if this assumption proves correct. Should other problems say, public debt risks in the advanced countries, or global inflation because a major cause for concern, they may rather prove to be a distraction. There is a difficult trade-off here: to keep focusing on the same issue helps narrowing down differences through the development of common concepts, indicators and guideposts. As indicated by the European experience, this process takes time, and for the outcome of this process to influence national policies even more time is needed. The same requirement applies even more to coordination within a large group whose participants are not used to speaking openly to the others about their policy choices. But keeping the focus on a particular set of issues involves the risk of focusing the policymakers attention on a certain set of problems as the expense of others. Again, Europe provides a clear case of attention distraction: its focus on making its fiscal pact operational has distracted the policymakers attention from the build up of large imbalances in the private sector.
4.5 Conclusions Four conclusions stand out. The first is that the G20 process is an ambitious one and that it has delivered more than the G7 in its last decades. The second is that degree of compliance of the G20 is, on average, far from disappointing. This conclusion emerges both from a systematic and rather blind analysis of the follow-up to all commitments and from a more focused analysis of some key issues. The third is that there were important signs of fatigue after the first meetings, at the time the first signs of international recovery became visible, but that there was some recovery lately. The last is that the degree of compliance of emerging countries is on average lower than that of advanced countries.

39

5.

Looking ahead: small fixes and big hurdles

Is the 'new' G20 working as it was initially hoped, or at least sufficiently well? Is it evolving, and in what direction? What lies ahead? For an institution that assigned itself the role the 'premier forum of economic cooperation' (Pittsburgh summit, final statement) presumably for an indefinite future, three years of life if a remarkably short time to base a judgement on. Even more so it is to venture into predictions on its eventual impact or survival, on which the verdict will belong to historians. Yet this debate has already started, involving economists and political scientists. Not surprisingly there are two camps; those who support the new formation, or are at least willing give credit to it by adopting a wait-and-see attitude, and those inclined towards early disappointment and dismissal. Absent a solid and sustained performance record, both camps have apparently convincing arguments. 5.1 Crisis management vs. long-term governance Towards the beginning of 2011 observers on either side agreed that a transition had occurred, between a crisis management mode, ie a phase in which leaders reacted under pressure to dramatic events in the financial markets with the single clear aim of preventing a global financial meltdown, to a sort of longer-term governance mode. In the latter and more recent mode of operation, the immediate risks of financial collapse or of a repeat of the Great Depression had receded, and the focus had shifted to the objective of managing interdependence and designing and implementing macroeconomic and financial reforms for the longer term. While this transition is natural and in fact desirable, it created three problems: First, the objectives were no longer so clear. Putting the world economy on a balanced and sustainable growth footing, or avoiding another crisis, are very general statements, of little help when it comes to restricting the range of concrete policy choices; Second, not unrelated, responses involved more asymmetry across countries and there was more scope for disagreement. This applies by definition to the management of imbalances but to many other policy fields also, because in the aftermath of the crisis advanced and emerging countries found themselves in very different situations as regards, for example, growth, inflation and public finances. As regards financial regulation, there was asymmetry too between the advanced countries that had experienced the crisis and where there was (admittedly somewhat uneven) appetite for reform, and the emerging countries where concern were inevitably less widespread;

40

Third, as a consequence, public communication had become more of a challenge. What is not well understood is difficult to summarise, the old motto goes; certainly it is to choose effective words for a final communiqu when the underlying substance is elusive or controversial.

When did the transition occur? As discussed in section 3, the first 'mode' clearly characterised the Washington and London summits (November 2008 and April 2009), the first taking place in the midst of the financial turmoil and at the start of the recession, the second when the recession was ending but evidence of this was not yet available. The transition to the second phase is located somewhere in the vicinity of the Pittsburgh meeting (September 2009). Woods (2010) considers London as the highpoint of cooperation, whereas in her view Pittsburgh marked the return to politics as normal, and Toronto achieved little. We agree on Toronto but take a partly different view on Pittsburgh. This summit in our view carried substantial weight, in itself and relative to expectations, because of the announcements of a novel policy architecture, orbiting around the G20 and with the FSB, the IMF and other standard setters as operational arms (see Angeloni 2009a and 2009b). The other substantial contribution from that summit was the launch of the 'Framework for Growth' embodying the IMF-driven 'Mutual Assessment Process' (MAP). We do concur, however, that subsequent work on implementing the Pittsburgh agreements fell short of hopes, for over a year. In subsequent articles we were also rather critical of the second phase see Pisani Ferry (2010) and Angeloni (2011). At the time of writing (end-2011) a relevant question is whether the G-20 should still play by the rules of this longer-term mode or revert to the crisis management mode. Certainly the combination of a sovereign crisis in Europe, a slowdown in the US and a (hopefully soft) landing in China calls for stepping up coordination and cooperation. It is hard to assess the Cannes summit in this respect because leaders were taken aback by unforeseen development in Greece, but at the very least it did not go far in the definition of a joint response to the worsening of the global situation. 5.2 What should the G20 do? If views diverge on timing, even more they do on interpretation. Some regard the passage to from the 'crisis' to the 'governance' phase as a sign of maturity, the G20 being able to demonstrate (or at least not to immediately contradict) its ability to last beyond the emergency and make a positive contribution also in normal times. Some see it following the path that led the G7 to transform from focused and effective to grandiose, all-encompassing and largely ineffectual. In the crisis management mode the heads of state took direct responsibility for both macroeconomic and financial regulation decisions. A simple comparison between the broad and vague G7

41

communiqus and those issued at the first G20 summits provides strong evidence of this extraordinarily hands-on behaviour. This was evidently not sustainable. It is not the role of heads and state and government to dictate the details of financial regulatory reform and a normalisation was both inevitable and desirable. In normal times it is more appropriate for the G20 to function as a steering committee that provides political energy and direction to international standard setters and also assesses progress of implementation, to quote from Dan Price (2011), the former Sherpa of President George W. Bush. One of us (Angeloni, 2009), in commenting post-Pittsburgh developments, has taken a similar view. Observers also emphasise the change in style and the deterioration in public communication between the two phases. Critics argue that the G20 will not be able to display, in a climate of economic recovery and stabilised financial markets, the same alertness and effectiveness it showed as a crisis manager. History can be suggestive; Woods (2010) notes that a similar regress characterised, years earlier, the G20 Finance Ministers formation. Created in 1999 to respond to the global risks generated by the Asian crisis, hence in a crisis management spirit, the G20 ministerial formation gradually transformed itself, inadvertently, in a non-committal discussion forum, and its influence declined. G20 summits could follow a similar fate. To a certain extent, therefore, the move from a 'crisis management' to a 'governance' mode results more from the G20s current capabilities than from the nature of the situation it is confronted to. Against the background of a serious deterioration of the global economic situation, this evolution involves risks: a G20 that would fail to address the priority problems of global adjustment and growth would soon lose the legitimacy earned in the management of the 2008-2009 global crisis. 5.3 Large but not universal With 19 countries plus the two-headed EU, the G20 is a large group, even more so when taking into account the participation of Spain (permanent guest), the countries invited on behalf of regional groupings (Equatorial Guinea, Ethiopia, Singapore and the UAE), and international organisations (FSB, ILO, IMF, OECD, UN, World Bank, WTO). This is clearly too large a group for dealing with macroeconomic and monetary issues, for which only a few countries or groupings are relevant. There were five participants in the Multilateral Consultations (China, the euro area, Japan, Saudi Arabia and the US), the G20 now has singled out seven systemic countries (China, India, France, Germany, Japan, the UK and the US), and there are talks of enlarging the SDR so that it would consist of five currencies (dollar, euro, pound sterling, renminbi and yen). Clearly, the international community is trying to find ways to narrow down the discussion to those who really matter.

42

At the same time the G20 is still not universal: it represents 85 per cent of world GDP, but only twothirds of the total population and one-tenth of the total number of countries. This is problematic for two reasons: first, countries not represented in it naturally question its legitimacy and request a representation in it (Vestergaard, 2010). This is a significant difference with the G7 which clearly had the character of a leadership group and was for three decades able to resist enlargement. Second, as noted by Woods (2010), non-universality formally limits the G20 to the role of an agenda-setter, not a rules-setter because it cannot make rules for the countries not represented in it. This distinction is important for regulatory matters where the G20 has de facto acted as a rule-setter (even though formal decision-making was delegated to the competent bodies). It has been advocated, notably by former UK Prime Minister Gordon Brown and Pascal Lamy, the head of the WTO, to resolve this tension by moving to a constituency system analogous to that in use at the IMF.26 This would require several member countries to take responsibility for representing current non-members. An evolution of this sort is probably inevitable in the long run if the G20 retains a major role in world governance. A formal constituency system would however involve the weighting of votes and joint decision-making, neither of which squares well with representation at head of state and government level. 5.4 Making the best of diversity A major feature of the G20 is its diversity. In this respect it is not just a G7 with extra chairs (Bnassy-Qur et al, 2009). This dimension was overlooked in the urgency context of the first meetings but it is nevertheless of major relevance. It is graphically capture by Figure 10 where we plot the distribution of gross financial assets held cross-border and net savings international savings flows. The former are indicative of the degree of financial integration while the latter are indicative of the contributions to current account imbalances. Clearly, financial integration remains very much a G7 business (if anything, enlargement to Switzerland and the Netherlands should have been considered) whereas the G7 is largely inadequate to discuss global imbalances. In setting it agenda and goals, the G20 should make full use of the asset constituted by its diversity. The presence around the same table of advanced and developing economies, deficit and surplus countries, high and low savers, debtors and creditors, naturally leads to a possibility of exploiting complementarities, assuming cooperation works. Large creditors with vast pools of liquid reserves, like emerging Asians, can, for example, contribute to stabilise distressed debt markets in advanced countries. This would be in their own interest; in this way, they would contribute to mitigate risks of adverse feedbacks against their own export-oriented economies.
26 In a debate at the Brussels Forum on 26 March 2011.

43

Figure 10: Whose role? G7 vs. G20

World distribu on of current account balances, 2012
World distribution of external assets and liabilities, 2007

G7 Other advanced non-G7 G20 Others

G7 Other advanced non-G7 G20 Others

Source: Authors calculations with Lane and Milesi-Ferretti data

Source: Lane and Milesi-Ferreti (2007), IMF WEO (2011), authors calculations.

In its first three years of existence the G20 has attempted to strike a balance between the priorities of its advanced and emerging members but, in part because of the crisis, the former have played a more prominent role in setting the agenda. In 2008-2009 its agenda was dominated by the management of the global turmoil, the provision of financial resources to countries in crisis, and the rebuilding of financial regulation a rather G7-like agenda. Furthermore, the framing of issues was strongly influenced by the nature of the problems in the advanced countries. As put by Andrew Sheng, rules for global finance were designed for Ferraris, not for Tatas27. The Korean presidency was in this respect a noteworthy moment as agenda shifted towards issues of importance for emerging and developing countries, such as global safety nets and development. The renewed focus on global imbalances also brings to the table an issue on which participation of major emerging economies is indispensable. But inflation, a major concern in that part of the world, remains of secondary importance in G20 discussions. More needs to be done to ensure that all G20 members feel ownership in the organisation and it can be expected that the forthcoming Mexican and Russian presidencies will indeed reshape the agenda. 5.5 Going forward In front of these challenges and given the considerable uncertainties, anyone should be reluctant in taking the risk of making explicit recommendations on how the G20 process should develop going forward, on how best to fulfil the promises of a 'global economic steering committee'. We would, however, miss an opportunity in this paper if we did not attempt, and perhaps shy away from our duty as observers. Our thoughts at this moment focus on two separate and only loosely connected areas. We feel there are, at the same time, a few simple and technical adjustments that could be made ('small fixes'), mainly of organisational nature. Alongside with them, there are much deeper

27 In remarks at the Bruegel-CEPII-ICRIER 2009 G20 conference, Delhi, September 2009.

44

areas of reflection that the G20 needs to undertake ('big hurdles') in a longer term horizon. The first could be decided and implemented, in principle, in a short time, even in a single meeting. The second are more fundamental. The challenge here is to develop the G20 mission, sharpening its common understanding and ownership of it, by developing an agreed set of share values (White, 2010). Let us comment briefly on both. The G20 is a quintessential intergovernmental organisation: participants (heads of state or government in this case) take part in it entrusted with their own powers, derived from the national political systems. The body itself is not delegated any authority: only the individual participants are. The chair, rotating annually (sometimes shared, as in 2010), bears sole responsibility for the agenda and the calendar of work, with the support of the two other members of the 'troika', composed by the rotating chairs of the preceding and the following year. Inevitably, the annual agendas have strong idiosyncratic components, reflecting national interests or philosophical biases. Work areas and deliverables each year are decided in essence by one country, while others are asked to patiently wait in line for their chance to make a mark. We believe the intergovernmental nature of the G20 is an asset to be preserved; attempts to enter the more technical sphere, invading eg the area that belongs to the IMF, should be resisted. Nonetheless, the exclusive reliance on the rotating chair for setting agendas is already proving to be a limitation to the effectiveness of the group. It is increasingly evident that this is a source of fragmentation and short-termism when the aim is to agree on and implement policy reform programs of a longer term nature. Strategic objectives cannot be agreed upon and dealt with within narrow yearly timeframes. The troika system does not fix this problem because 3 members out of 20 do not ensure a sufficient degree of representation. Limited improvements in working arrangements can help strengthen the continuity to the G20 agenda. First of all, strategic directions should be discussed and agreed upon in specific sessions, freed from the constraints of the annual programs dictated by the rotating chair. Second, input from independent experts could help avoid agenda inertia and break path dependency, as suggested by Indian Sherpa Montek Singh Alhuwalia.28 Third and more ambitiously, a steering group (SG, similar to that set up in the FSB), with a mandate extending beyond the annual chair, could be foreseen. The SG would have responsibility to coordinate and bring together the output of working groups on specific areas, ensuring the unity of action of the G20 and its consistency over time. A SG would also contribute in ensuring a systematic liaison with the 'operational arms' (IMF, FSB, other standard setting bodies); the latter in principle should report to the G20, and are delegated tasks by it, but it is unrealistic to expect that the rotating chair alone can ensure the effective functioning of the the
28 In remarks at the Bruegel-CEPII-ICRIER 2010 G20 conference, Delhi, September 2010.

45

delegation structure. The SG should itself rotate in its composition, but in any given point in time should possess sufficient representation and members should remain in charge for a sufficiently long time to provide continuity (three years seems to be a minimum). Yet higher in terms of ambition would be the establishment of a permanent secretariat. Though the simple name often evokes ghosts of bureaucracy and red-tape, a small secretarial structure located at, and administratively dependent on, an existing institution (like the IMF) could act as an important catalyst and a source of stimulus. If the secretariat is kept small and with a narrow and well-focused mandate, such a reform could result in dramatically improving the efficiency of the process. It would make possible for countries in the group with less administrative capabilities and international experience to hold the chair and would prove to be a complement, not a counterbalance, to the prerogatives of the chair. Again, the precedent of the FSB, whose secretariat is hosted by the BIS in Basel, is useful references. This said, it is clear that organisational changes can never be a substitute for substance and vision. As it stands the G20 cannot provide such vision fundamentally because it lacks a shared philosophy, a common understanding of the economic priorities of our time and the way to approach them. A serious debate, though warranted and mature, has not started yet on a number of questions of central importance for the global economy. How will the global community cope with the limits to global resources in this century and beyond? How will the inescapable aspirations of economic newcomers (the emerging world) be made compatible with the requirements of the veterans, in a balance of interests that may benefit all? How can within- and between-countries inequalities generated by the globalisation process be tackled? How, in the political economy of this balance, should enter the interests of future generations? What economic system, specifically what position in the multifaceted spectrum between free and regulated markets, offers the best chance of making the reconciliation of this multidimensional set of interests easier? And, not least nor final, what are the specific orientations to be taken in the key areas facing the global economy going ahead energy, environment, financial regulation, international monetary relations, crisis and natural disaster management, just to name a few? The need for a shared vision, or at least to start a dialogue, on these themes did not arise because of the G20; rather, it was the fruit of the same historical forces (globalisation and new equilibria in international relations) that gave rise to the G20. These are admittedly politicised and controversial issues, on which the risk of stalemate is significant. The G20 can and should respond to this challenge by helping develop this dialogue in the most harmonious possible way.

46

6.

Conclusions: searching a script for the characters

Our assessment of the first three years of the G20 is largely positive. Whereas other schemes for enlarging the G7 could have been considered and could perhaps have been more appropriate discussions on this point should not conceal the essential outcome, which is that this long-overdue enlargement was a major positive result of the crisis. The group earned political capital and built team spirit in the successful fight against the global recession in 2008-2009. Whereas later achievements were certainly less spectacular, they still compare favourably to the results of the G7/G8 over the previous decade. Size, which is admittedly a difficulty, and diversity have not prevented the G20 from being more effective than its more cohesive predecessor. Is this enough to claim success? Hardly. As we have argued, the G20 is facing three important challenges. The first is the challenge of legitimacy. The more the G20 proves effective, the more all countries in the world want to be represented in it. Why is Argentina a member, and not Chile? Why Turkey, and not Egypt? Why is most of Africa excluded? The more the group is hands-on, the more it raises questions about decision without representation. This is a hard challenge to address, because the G20 was born too large to evade it by pretending to confine itself to providing leadership. We have sympathy for the idea of forming constituencies, but we doubt this is a route that can be formally followed until its end. In the realm of immediate feasibility, we think that the G20 should as far as possible work through existing universal international organisations, so that decisions go through a formal process of deliberation and approval. The second is the challenge of effectiveness. We are afraid of diminishing returns. To address the problem, the G20 needs to equip itself with efficient working methods. We have made suggestions a rotating steering group, a secretariat embedded in an existing international organisation. They can in principle be enacted quickly, but obstacles are in the details. We also believe that the G20 would gain effectiveness in limiting itself to the roles agenda-setter and compromise-builder, while leaving technicalities to specialised international organisations. This would help it staying ahead of the curve, instead of falling into the trap of path-dependency. The third is the challenge of vision. Until now the G20 has worked on the urgencies of the moment and those inherited from the past. This was fully justified at the time of the global crisis but is hardly sufficient in normal times. It now needs to make the best of its diversity to set itself a forwardlooking agenda.

47

As in Pirandellos play but hopefully, with less drama and despair the right characters are there, but the plot is not written. Economic globalisation and the birth of global politics require the development of global political institutions. The G20 is the seed, but fulfilling its promises will require time. .

48

REFERENCES Angeloni, Ignazio (2009a) 'The Pittsburgh G20 Checklist', Bruegel Policy Contribution, September Angeloni, Ignazio (2009b) 'The Pittsburgh G20 Summit', Intereconomics Angeloni, Ignazio (2011) 'Lets bring our acts together', The Bruegel G20 Monitor http://www.bruegel.org/65, February Angeloni, Ignazio, Agns Bnassy-Qur, Benjamin Carton, Zsolt Darvas, Christophe Destais, Ludovic Gauvin, Jean Pisani-Ferry, Andr Sapir, and Shahin Valle (2011) Global currencies for tomorrow: a European perspective, Bruegel Blueprint No 13, August Baxter, Marianne and Alan Stockman (1989) 'Business cycles and the exchange regine, some international evidence', Journal of Monetary Economics Bnassy-Qur, Agns, Rajiv Kumar and Jean Pisani-Ferry (2009) 'The G20 is not Just a G7 with Extra Chairs', Bruegel Policy Contribution No 2009/10, September Bergin, Paul (2008) 'International Policy Coordination', New Palgrave Dictionary of Economics Blanchard, Olivier, and Gian Maria Milesi-Ferretti (2009) 'Global Imbalances: In Midstream?', IMF Staff Position Note 09/29, December Brown, Gordon (2011) Keynote speech at the INET Bretton Woods Conference, 9 April Canzoneri, Matt, Robert Cumby and Behzad Diba (2002) 'The Need for International Policy Coordination: Whats old, Wants new, Whats yet to come?', Journal of International Economics 66 Canzoneri, Matt and Joanna Gray (1985) 'Monetary Policy Games and the Consequences of Noncooperative Behavior', International Economic Review Canzoneri, Matt and Dale Henderson (1990) Noncooperative Monetary Policies in Interdependent Economies, MIT Press, Cambridge Clarida, Richard, Jordi Gali and Mark Gertler (2002) 'A Simple Framework for International Policy Analysis', Journal of Monetary Economics 49

49

Corsetti, Giancarlo and Paolo Pesenti (2001) 'International Dinensions of Optimal Monetary Policy', NBER Working paper n. 8230 Eichengreen, Barry, and Kevin ORourke (2010) 'A Tale of Two Depressions; What do the New data Tell Us?', VOX, 8 March

Faia, Ester (2007) 'Finance and International Business Cycles', Journal of Monetary Economics 54
Frankel, Jeffrey, and Katherine Rockett (1988) 'International Macroeconomic Policy Coordination When Policymakers Do Not Agree on the True Model', American Economic Review 78(3), pp. 318-340 Friedman, Milton (1953) 'The Case for Flexible Exchange Rates', in Essays on Positive Economics Hamada, Koichi (1974) 'Alternative exchange rate systems and the interdependence of monetary policies', in Aliber, R.Z. (ed.), National Monetary Policies and the International Financial System, Chicago University Press, Chicago Hamada, Koichi (1979) 'Macroeconomic strategy coordination under alternative exchange rates. In: Dornbusch, Rudiger, Frenkel, Jacob (Eds.), International Economic Policy, Johns Hopkins University Press, Baltimore, pp. 292324 Ilzetski, E., C. Reinhart, and K. Rogoff (2008) 'Exchange Rate Arrangements Entering the 21st Century: Which Anchor Will Hold?', available on Carmen Reinharts website, http://terpconnect.umd.edu/~creinhar/ International Monetary Fund (2009) 'The G-20 Mutual Assessment Process and the Role of the Fund', mimeo, 2 December International Monetary Fund (2010a) 'Will it Hurt? Macroeconomic Effects of Fiscal Consolidations', Chapter 3 of the October World Economic Outlook International Monetary Fund (2010b) 'G-20 Mutual Assessment Process Alternative Policy Scenarios', Report to the G20 for the Toronto summit, June International Monetary Fund (2010c) 'G-20 Mutual Assessment Process IMF Staff Assessment of G-20 Policies', Report to the G20 for the Seoul summit, November

50

International Monetary Fund (2011a) Consolidated Spillover Report: Implications from the analysis of the Systemic-5, IMF, July International Monetary Fund (2011b) United States: 2011 Spillover report Selected Issues, July International Monetary Fund (2011c) IMF Staff Reports for the G-20 Mutual Assessment Process, prepared by the Staff of the Fund for the Cannes summit International Monetary Fund (2011d) Review of the Funds Involvement in the G-20 Mutual Assessment Process, 13 May Johnson, Harry (1969) 'The Case for Flexible Exchange Rates', St Louis Fed Review Kubelec, Chris, and Filippa S (2010) 'The geographical composition of national external balance sheets: 1980-2005', Bank of England Working Paper No 384, March Lane, Philippe, and Gian-Maria Milesi-Ferretti (2006) 'The External Wealth of Nations Mark II: Revised and Extended Estimates of Foreign Assets and Liabilities', IMF Working Paper WP06/09 Larosire, Jacques de (2009) Report of the High-Level Group on Financial Supervision in the EU, Brussels, February McKinnon, Ronald (1993) 'The rules of the game: International Money in Historical Perspective.' Journal of Economic Literature 31 (March) pp.1-44. Obstfeld, Maurice, and Kenneth Rogoff (2002), 'Global Implications of Self-Oriented Monetary Rules', The Quarterly Journal of Economics 117(2), pp. 503-535 Oudiz and Sachs (1984) 'International Policy Coordination in Dynamic Macroeconomic Models', NBER Working Paper n. 1417 Pisani-Ferry, Jean (2010) 'Back to the G7?', The Bruegel G20 Monitor http://www.bruegel.org/65, June Pisani Ferry, Jean and Andre Sapir (2011) 'Half empty, half full'; The Bruegel G20 Monitor http://www.bruegel.org/65, February Price, Daniel (2011) 'G20 version 2.0 will appease the skeptics', Financial Times, 31 March Rottier, Stphane, and Nicolas Vron (2010) 'An Assessment of the G20s Initial Action Items', Bruegel Policy Contribution 2010/08, September
51

Shirakawa, Masaaki (2011) 'Global Imbalances and Current Account Imbalances', in Global Imbalances and Financial Stability, Financial Stability Review, Banque de France Triffin, Robert (1985) 'The Myths and realities of the so-called gold standard', in Eichengreen, B. (ed.) The Gold Standard in Theory and History, Methuen, London Turner, Adair (2009) The Turner Review: A Regulatory Response to the Global Banking Crisis, London, Financial Services Authority Vestergaard, Jakob (2010) 'The G20 and Beyond: Towards effective Economic Governance', Danish Institute for International Studies Report 2011:04 Vines, David (2011) 'The G20MAP, Global Rebalancing, and Sustaining Economic Growth', mimeo, April Von Fuerstenberg, George and Joseph Daniels (1992) 'Economic Summit Declarations 1875-1989: Examining the Written Record of International Cooperation', Princeton Studies in International Finance 72, February White, William (2010) 'The importance of collective action and sound policy frameworks', address given at the G20 workshop on The Framework for Strong, Sustainable and Balanced Growth, Toronto, May Woods, Ngaire (2010) 'The G20 leaders and global governance', GEG Working paper 59, Oxford University Zhou, Xiaochuan (2009) 'Reform the International Monetary System', PBOC, 23 March

52

Appendix: scoring calculations using the University of Toronto G20 compliance reports29 The average compliance scores, by summit and by topic, were constructed using the results of the compliance reports produced by the G-20 Research Group at the University of Toronto (see http://www.g20.utoronto.ca/analysis/index.html). It should be noted that the UoT results are based on samples: at each summit a limited number of commitments from the leaders declaration are chosen to be monitored, and the overall compliance is based only on them. On the selection of commitments, the latest report (Toronto Compliance Report) states that: The selection should be balanced and represent the summit priorities, be important for the G20 and the world, be comprehensive and cover most of the summit domains and decision areas, and represent both individual and collective pledges. The selected commitments should also meet secondary criteria relating to performance measurability, ability to comply within the monitoring period and the feasibility of establishing interpretative guidelines. Most importantly, the reports keep up with the evolving list of issues raised during the successive summits and the number of commitments recorded in each iteration of the report is different. The Washington report features 230, the London report features 5, and 8 items are monitored in the reports for Pittsburgh and Toronto. Summit scores Following von Frstenberg and Daniels (1992), the UoT analysts assign to each commitment a score of -1 (non compliance), 0 (partial compliance) or +1 (full compliance). Our summit scores were computed as simple averages of the scores assigned to each of the commitments monitored at each summit, regardless of their belonging to one of the four groups outlined below. For the Toronto report, where there are several macroeconomic commitments that do not apply to all countries, only one overall score if used for macroeconomic policies. Because of missing data for the Trade commitment, the score for the Washington summit is only available for 14 countries.

29 Prepared by Christophe Gouardo. 30 In fact, there is no comprehensive compliance report available for Washington. We use the scores from two

individual compliance reports focusing specifically on trade and financial regulation, and compute the summit score as the weighted average of these two issues.
53

Issue scores The scores for each of the four areas Macroeconomic Policy, Financial Regulation, IFI Reform, and 'Others' are first computed for each of the areas, and then the average is calculated. The Macroeconomic Policy, Financial Regulation, IFI Reform and Others areas each appear at 3, 4, 2 and 4 summits respectively. In practice each area is represented by a single item, except 'Others', which changes in perimeter depending on the issues discussed at each summit (broadly, the 'Others' category aggregates Trade, Development, Energy, Corruption, Climate Change, and Food and Agriculture). Because of the way the average scores are constructed (and especially the fact that not all commitments feature at each summit and that the total number of commitments, both at each summit and within each issue group, is different), the 'Average' columns at the end of the tables are not identical. Group averages The group averages are simply the averages of the scores obtained by individual countries at each summit/on each issue area. When weighted scores are calculated, the weights are on the basis of the countrys share of GDP at market prices in the year 2010 within each group. The weights for each the groups are presented below. The groups are as follows: G-20 (20 countries): all 19 G-20 national countries (Argentina, Australia, Brazil, Canada, China, France, Germany, India, Indonesia, Italy, Japan, Korea, Mexico, Russia, Saudi Arabia, South Africa, Turkey, United Kingdom, United States) plus the European Union. Advanced (9 countries): Australia, Canada, France, Germany, Italy, Japan, Korea, UK, USA Advanced in deficit (6 countries): Australia, Canada, France, Italy, UK, USA. Emerging (11 countries): Argentina, Brazil, China, India, Indonesia, Mexico, Russia, Saudi Arabia, South Africa, Turkey. Emerging in deficit (6 countries): Brazil, India, Mexico, South Africa, Turkey. G7 (7 countries): Canada, France, Germany, Italy, Japan, UK, USA. G7 Europe (4 countries): France, Germany, Italy, UK.

54

Table A1: GDP weights used in calculations, at market prices in 2010

Advanced Emerging Emerging in deficit in deficit 2.53% 5.07% 14.29% 36.25% 6.46% 40.18% 10.60% 10.51% 26.67% 4.83% 8.44% 7.10% 18.02% 10.01% 3.03% 2.44% 6.20% 5.07% 12.86% 9.23% 60.19% -

G20 Argentina Australia Brazil Canada China France Germany India Indonesia Italy Japan Korea Mexico Russia Saudi Arabia South Africa Turkey UK USA EU 0.57% 1.90% 3.21% 2.42% 9.04% 3.97% 5.10% 2.36% 1.09% 3.16% 8.39% 1.55% 1.60% 2.25% 0.68% 0.55% 1.14% 3.46% 22.53% 25.03%

Advanced 3.62% 4.61% 7.57% 9.71% 6.02% 15.99% 2.95% 6.58% 42.94% -

G7 4.94% 8.10% 10.40% 6.44% 17.12% 7.05% 45.96% -

G7 Europe 25.32% 32.50% 20.15% 22.03% -

Source: IMF WEO (2011).

55