Chapter 3 Use and Delivery

International Monetary Fund. Independent Evaluation Office
Published Date:
September 2006
  • ShareShare
Show Summary Details

We now consider issues related to the use and delivery of the IMF’s multilateral surveillance products. We look first at how multilateral surveillance informs bilateral surveillance. We then discuss whether the multilateral surveillance message is reaching the intended external audience, including how well the delivery of the message is exploiting the potential for peer pressure. We consider whether the presentation of surveillance products is sufficiently focused on areas where the IMF can offer the greatest value added and is done in a way that maximizes its impact. We then offer a few observations on the ultimate impact of IMF multilateral surveillance on policies adopted in major countries.

Informing Bilateral Surveillance

To assess the extent to which multilateral surveillance provides global perspectives to bilateral surveillance, we paid particular attention to the two most recent staff reports (issued through December 2005) for Article IV consultations with a sample of 36 countries,1 as well as evidence collected from surveys and interviews. We review below (1) how the staff reports cover the linkages between domestic macroeconomic issues and global economic conditions; (2) how area department staff incorporates global capital market perspectives in its country work; and (3) how the IMF is trying to promote the integration of multilateral perspectives in bilateral surveillance.2

Macroeconomic surveillance

While the staff reports we reviewed all make some reference to global linkages, in many cases the analysis was not central to the overall discussion, with hardly a paragraph devoted to them. In some reports, the references were limited to recent world oil price developments, but this was an event difficult to ignore for any country (Box 3.1). Additional references to global and regional spillovers were found in a few accompanying selected issues papers, but the IEO’s overall assessment does not change.3 The global or regional ramifications of domestic policies were discussed only for China in 2004, the United States in 2004 and 2005, and Japan in 2005. Even for Germany or Russia, there was virtually no discussion of the regional (let alone global) spillover effects of their policies (for the euro area as well, discussion of global and regional issues was found to be more inward-looking).

Why such limited coverage of global and regional linkages? At least three interpretations are possible. First, linkage issues may be discussed with the authorities but not written up in staff reports. Interviews with area department staff have suggested this possibility. Second, the absence of discussion of global or regional linkages in staff reports may only suggest that these issues are not pertinent to policy discussions in a particular context. A third interpretation, of course, is that multilateral surveillance fails to affect the conduct of bilateral surveillance.

These interpretations are not mutually exclusive, but the last possibility is implied by the IEO surveys of senior area department staff and national authorities (Figure 3.1). Focusing specifically on the WEO, only 7 percent of the IMF’s senior staff and 10 percent of the authorities surveyed “strongly agreed” with the statement that “WEO topics are discussed during Article IV discussions,”4 although there may be a considerable variation across the area departments.5 The split views of the staff and the authorities in the next two categories (“agree” and “neutral”) may mean that limited time is spent discussing global issues during Article IV consultations (thus leading to divergent assessments). Although progress may have been made in terms of references to economic-linkage-related issues in Article IV staff reports, there is obvious scope for bringing global perspectives further into bilateral surveillance.

Figure 3.1.Senior Area Department Staff and National Authorities on the WEO

(In percent)

Box 3.1.Global Linkages and Spillovers in Individual Country Reports, 2004-05

References to global linkages and spillovers vary across Article IV staff reports. Some reports—specifically those on Algeria, Israel, Namibia, the Netherlands, Nigeria, Pakistan, and Spain—made few or no references to global linkages beyond the implications of oil price developments. By contrast, the country reports for Chile, China, Mexico, Singapore, Turkey, and the United States offer numerous references and detailed analysis of global linkages.

The staff report for Chile, for example, mentions that robust growth in the global economy has helped boost the price of copper. It also points out that Chile’s real GDP growth was aided, among other things, by low world interest rates. Chile’s trade links with China are also discussed in detail. The report refers to regional developments—such as Argentina’s decision to reduce natural gas exports—forcing Chilean electricity generation to switch to more costly coal- and diesel-generated power. It also discusses the possible effects of a sharp rise in world oil prices.

The Article IV report for China discusses the possible regional impact of a slowdown in that country. It also looks at greater exchange rate flexibility and its impact on the domestic economy. Analysis of the prospects for higher U.S. interest rates is featured in some sections of the report.

Economic links with the United States are discussed in great detail in the Article IV report for Mexico, which also discusses the increased Chinese exports to the United States and its effect on Mexico’s export share.

The U.S. Article IV report analyzes the effect on domestic activity of an increase in oil prices and discusses the implications of the large current account deficit and associated risks of market disruption. The report points out that the low U.S. national saving rate could be a significant drain on global saving as the world recovery matures, potentially dampening global investment and growth. As to the resolution of global current account imbalances, the report discusses the importance of stronger growth abroad and the need for a stronger U.S. fiscal position; it also analyzes the spillover effects of U.S. economic policies on global investment.

Capital market surveillance

Most IMF area department economists appear to be paying attention to international capital market issues. According to the survey of area department staff, more than 70 percent of respondents said they were either “very familiar” or “familiar” with the Global Markets Monitor, a daily summary of market developments produced by the International Capital Markets Department. Indeed, more than half of them said that they used it “frequently.” The review of the staff reports for our sample of 36 countries also indicates that the reports for 25 of them included some discussion of the implications of global capital market developments for the economies concerned. The countries for which no mention of global capital market developments was made mostly included countries with relatively limited access to international capital markets, such as Cameroon, Kenya, and Namibia.

In almost all cases where references to international capital market issues were made, however, they were limited to a few lines; rather surprisingly, no mention of global market developments was made in the staff reports for the Netherlands, Singapore, and Spain. The references were mostly related to capital flows or interest rates, issues that are readily translatable into macroeconomic terms. The frequent absence of references to global capital market developments in staff reports, and the terseness of the references that are made, may partly reflect the relative paucity of global capital market issues that have immediate macroeconomic consequences.

Turning to the use of the GFSR, most members of the IMF staff and the Executive Board interviewed by the IEO had a generally negative view of the GFSR, mainly because of what they consider its excessive length and overuse of market jargon. The views expressed by national authorities were more positive. More than 80 percent of respondents considered the analytical rigor of the GFSR as either “completely adequate” or “adequate.” Somewhat fewer respondents gave such high marks to the GFSR in such other categories as timeliness and identification of vulnerabilities, but the outside perception of the GFSR was considerably better than that of those within the IMF.

How do we reconcile this generally positive assessment of the GFSR by national authorities with the generally negative view held by those within the IMF? We believe it can be explained by the relatively small readership of the GFSR, as revealed by interviews with IMF staff, market participants, and member country officials. According to our survey, only 4 percent of area department staff uses the GFSR in its country work “regularly” (and hardly anybody uses it “frequently”); 76 percent said that they seldom read the document. But those in member countries who responded to the surveys actually read the GFSR. We suspect that when the surveys reached the finance ministries and central banks concerned, the task of responding to the survey was assigned to those who actually read it.

The limited audience for the GFSR within the IMF has much to do with the nature of the issues that the publication has dealt with in recent years (Table 3.1). Discussion of recent market developments in the review chapter of the GFSR offers little new information to most IMF staff members, beyond what is already available from the financial press and other routine market intelligence. Likewise, the analytical chapters have almost exclusively covered issues related to the transfer of risks across sectors and medium-term institutional development issues. These issues have little immediate consequence for the IMF’s routine Article IV country work. Outside the IMF, some of these chapters have been highly appreciated by those who read them. A senior major country official responsible for financial stability explained to the IEO team that risk-related issues covered in the GFSR were critically important, but only to a relatively small group of risk managers in the public and private sectors.

These considerations suggest that if the GFSR is to offer greater value added to those within the IMF and also to a wider audience outside, its authors must better exploit their comparative advantage. Indeed, IMF staff has a unique position that corresponds to: (1) its ability to analyze market developments from a perspective that is detached from the marketplace; (2) its neutrality characterized by the absence of a profit motive; and (3) its access to the IMF’s extensive global macroeconomic information. This underscores the need for a more collaborative approach to producing the GFSR, seeking greater inputs from various IMF departments at different stages.

At the same time, the GFSR should aim to provide IMF staff with an analytical view of market developments that identifies the implications of these developments for the IMF’s country work, including short-term risks and vulnerabilities in the global financial system. The more specialized studies on medium-term risk transfer and institution-building issues could be redesigned to target the relatively small group of interested specialists and could be featured in a separate publication.

Bringing more multilateral perspectives into bilateral surveillance

The IMF has several mechanisms in place to bring multilateral perspectives into bilateral surveillance. In recent years, it has taken steps to strengthen these mechanisms and develop new instruments for this purpose. We highlight below the recent modifications of the vulnerability exercise (VE); a prospective enhancement of the work of the Coordinating Group on Exchange Rate Issues (CGER); and the increasing use of global economic models in bilateral surveillance. We also review the participation of RES and ICM staff in area department missions in recent years.

First, the IMF substantially modified the vulnerability exercise in mid-2005, to strengthen its topdown element. The primary objective was to enhance the analytical underpinning of vulnerability assessments. The modification also included a new mechanism to select countries on the basis of objective criteria, so as to ensure that all systemically important countries are covered regardless of the judgments of area departments. The final assessment of vulnerability still rests with area departments. Although most area department staff does not seem to find much value added in the exercise, the in-dicatorbased rating scheme at least imposes discipline on the process, with any major differences of view communicated to management.

Second, work is under way to expand the coverage of the CGER to include emerging market economies, in order to increase the multilateral aspect of the exercise and its usefulness in bilateral surveillance. Considerable data collection work is required before the methodological approaches can become operational. When the work is completed, CGER assessments will provide additional multilateral inputs into exchange rate surveillance by country teams for a much larger set of countries.

Third, multilateral perspectives can be brought into bilateral surveillance by using global economic models. Two models, GEM and MULTIMOD, have been used by IMF staff in some bilateral surveillancesituations. 6 By the very nature of these models, their use, at least in their multicountry versions, implies that economic linkages across countries are explicitly incorporated in the analysis. The models have often been applied to relevant policy issues and have helped draw some sharp policy implications. An IEO review of recent staff reports indicates that, especially with the development of GEM, model-based simulations have become an increasingly frequent feature of bilateral surveillance. Whereas staff used only a handful of global model-based simulations each year during 2000-04, they were used in at least 10 cases in 2005.7 We may see greater use of global models as GEM develops further and as more area department staff members become familiar with using its framework.

Finally, although not a formal process, the participation of RES and ICM staff in area department missions could potentially contribute to bringing more multilateral perspectives into the IMF’s bilateral surveillance work. Over the six-year period FY2000-05, 88 RES and ICM economists (52 for RES and 36 for ICM) participated in Article IV consultation missions.8 ICM participation appears to have picked up in 2005, largely because of its increased participation in missions to emerging markets, particularly WHD countries.9 RES participation in industrial country missions has been largely limited to a few countries, notably, Belgium, Australia, Germany, and Japan. On the whole, participation by RES and ICM staff in the IMF’s country mission work is limited, and considerable scope may exist to promote the integration of multilateral and bilateral surveillance through this channel.

Reaching the Intended Audience

Identifying the intended audience

The external products of IMF multilateral surveillance—particularly the WEO and GFSR—have multiple audiences in the public and private sectors. The WEO receives a particularly wide press coverage, which has grown over the years (Figure 3.2).10 Notable growth appears to have taken place in non-English language outlets, where citations have more than quadrupled, even as English language citations have grown by 50 percent.11 The coverage of the WEO is particularly extensive around the time of its release: over a 12-day period following the release, the WEO typically receives close to 600 references in the press throughout the world.12 There is clearly media interest in IMF forecasts, implying that effectively communicated multilateral surveillance messages can potentially have substantial impact on public debate.

Figure 3.2.The Press Coverage of the WEO, 2000-05

(Total references to the WEO)

Although the private sector is an increasingly important audience, interviews with staff members suggest that many seem to consider the policymakers of member countries as the primary audience of multilateral surveillance outputs. National authorities also seem to see themselves as the main audience. More than 90 percent of respondents said that the audience for the WEO and GFSR was “policymakers and public sector economists,” while about 50 percent also considered academics and research institutes to be an audience. In contrast, the over-whelming majority of respondents did not consider the media or the private sector (including the markets) as the primary audience of these publications. Given the acknowledged importance within the IMF of informing the public, this is a rather surprising result.

Interviews with senior officials of IMF member countries indicate that the “message” of these publications does reach the relevant officials in finance ministries and central banks. Given the time constraints of senior officials and the length (and sometimes complexity) of these documents, however, those who actually read them are more likely to be “working level” officials assigned to prepare short summaries for their superiors. According to the survey, about 25 percent of respondents replied that “senior policymakers” read much of the WEO and the GFSR (Figure 3.3). The meaning of “senior policymakers” was not clearly specified in the survey. On the basis of interviews with officials of major countries, the IEO evaluation team believes that much less than 25 percent of ministers and of governors and their deputies actually read the WEO or the GFSR.13

Figure 3.3.How Multilateral Surveillance Is Used by National Authorities

(In percent)

Is the Executive Board effective as a peer pressure group?

The Executive Board’s role in multilateral surveillance has three dimensions. First, members of the Board represent their national authorities and, as such, the Board itself can serve as a vehicle of peer pressure. Second, the Board processes the information it receives on global economic and market developments (e.g., through WEMD sessions) for its oversight of bilateral surveillance work.14 The Board can thus serve as an additional mechanism for integrating multilateral and bilateral surveillance. Finally, the Board reviews the multilateral surveillance documents it receives (the WEO and GFSR) for factual accuracy and, possibly, also for its political sensitivities prior to the reports’ publication.15 We focus here on the role of the Board as a peer pressure group.

As a forum for peer pressure, the Executive Board’s contribution to multilateral surveillance may be limited by the current setup for its involvement. When Executive Directors meet to discuss the WEO twice a year, in March and August, virtually all of their initial interventions are prepared statements that are circulated before the meeting and entered into the official minutes.16 These written statements—called “grays” in IMF jargon—are useful because they often incorporate the carefully articulated and drafted views of national authorities. But these statements have not facilitated an active, free exchange of views on the substance of policy issues, and a number of senior officials in capitals maintain that whatever discussion does take place at the Board has not fed into their policy discussions.

The limited role played by the Executive Board under the current setup of multilateral surveillance is illustrated by examining the minutes of Board discussions on the WEO. We found that, during 2000-05, about three-quarters of the additional interventions by Executive Directors concerned factual clarifications, drafting suggestions, and other procedural comments (such as content and future topics). Indeed, Board members had only a limited exchange of views on an average of four topics, resulting in 10-12 additional interventions by Executive Directors (Table 3.2). The Summing Up of the Board discussion—appended to the WEO report—does not receive much press coverage, because the press conference that follows the Board discussion focuses on the WEO as a staff document. The table indicates a similar story for the Executive Board meetings on the GFSR.

Table 3.2.Executive Board Meetings on the WEO and the GFSR, 2000-051
Number of meetings4n.a.4644
Length3 hours and3 hours and2 hours and2 hours and1 hour and
46 minutes11 minutes40 minutes50 minutes38 minutes
Number of “grays” issued18n.a.18182424
Number of interactions on policy issues33 (12)n.a.4 (12)1 (4)4 (10)0 (0)
Factual clarification or drafting
Other procedural comments312n.a.15161118
Source: Official minutes of relevant Board meetings.Note: n.a. = not applicable; publication of the GFSR began in 2002.

Averages for each meeting.

Includes only regular spring and fall sessions.

The average number of topics on which some exchange of views took place among Executive Directors (number of such interventions in parentheses).

Percent of total Board discussion.

Source: Official minutes of relevant Board meetings.Note: n.a. = not applicable; publication of the GFSR began in 2002.

Averages for each meeting.

Includes only regular spring and fall sessions.

The average number of topics on which some exchange of views took place among Executive Directors (number of such interventions in parentheses).

Percent of total Board discussion.

Contact with intergovernmental groups

Given their share of the global economy, the G-7 (accounting for about 65 percent of world GDP) and the G-20 (about 80 percent of world GDP) represent the most important policy forums to which the IMF has access.17 These are the forums of systemically important countries where policymakers (including ministers and central bank governors and their deputies) meet to discuss policy issues of mutual interest. The limited number of attendees permitted at these meetings can facilitate frank exchanges of views among the most senior policymakers, and peer pressure can thus be exerted. The effectiveness of IMF multilateral surveillance, therefore, depends critically on the effectiveness with which the institution interacts with these major intergovernmental groups. Within the IMF, however, there is ambivalence toward these groups, which are seen more as competitors than as allies.

This ambivalence is partly reflected in a lack of continuity in institutional representation and in a poor infrastructure to support the IMF’s inputs into these and other similar groups. While the Managing Director attends the meetings of these intergovernmental groups at the ministerial level, the First Deputy Managing Director and other senior officials have represented the IMF at the various meetings of the deputies. Surveillance notes for the G-7 and the G-20 have typically been prepared by a single B-level staff member in the Research Department.18

As a result, the IMF has not proactively engaged with these intergovernmental groups; only highly standardized surveillance notes containing an updated summary of the latest WEO are routinely prepared at the request of a country holding the chairmanship. Occasionally staff has found itself overstretched when a last-minute request for a special thematic piece is made. If more resources were allocated to these activities, not only could such requests be welcomed as a possible opportunity to help shape the policy debate, but the IMF could also initiate new analysis as an input into these forums. The lack of leadership, focus, and resources has precluded pursuing creative ways of engaging with these groups of senior policymakers.

When sufficient attention and resources are committed to intergovernmental group issues, a more proactive engagement can indeed be expected. This is illustrated by the preparation of a special note on “Adjustment of Global Imbalances” by an interdepartmental group of staff from the Asia and Pacific Department (APD), European Department (EUR), Western Hemisphere Department (WHD), ICM, PDR, and RES in January 2004, ahead of the G-7 meetings in February. Responding to a request from management, this team quickly prepared a highly technical note that analyzed: (1) how large external imbalances were creating divergent changes in net foreign asset positions across countries; (2) how much the U.S. dollar would need to adjust over time to accommodate these changes; and (3) the impact of such an exchange rate adjustment. This was in the end packaged as a five-page supplement to the surveillance note.

Regional outreach

Regional outreach is another channel through which the IMF can communicate the products of its multilateral and regional surveillance to targeted audiences. In mid-2005, area departments initiated efforts to deliver the key messages of the WEO to various regions. These regional outreach initiatives have generally been well received. Area departments can also take advantage of the access they have to regional forums of senior policymakers, including meetings of the Asia Pacific Economic Cooperation Council (APEC) and Western Hemisphere Finance Ministers. While it is not possible to assess these efforts in detail, the experience of the European Department’s engagement with EU institutions in recent years suggests the potential to increase the impact of IMF multilateral surveillance through these channels.

Within the EU, the IMF has enjoyed extraordinary access to policy discussions between senior national policymakers in the framework of the Economic and Financial Committee (EFC) of the Council for Economic and Financial Affairs (Ecofin Council).19 The Director of the IMF’s European Department has regularly been invited to attend the Eurogroup and Eurogroup Working Party meetings through an informal arrangement.20 Some senior European officials indicated to the IEO that the IMF was the only outside entity to be invited to the Eurogroup Working Party, and the IMF’s interventions were appreciated and sometimes sparked subsequent policy debate.

The reasons for the success of regional outreach in Europe may or may not apply to other regions. First, European governance structures and institutions have been undergoing rapid change. This has helped increase policymakers’ interest in the views of the IMF, as a global organization with a politically dispassionate perspective. Second, the IMF has been able to focus selectively on key issues and to deliver views constructively. This has helped secure its access to ministerial discussions. Finally, IMF analysis is considered by many officials to be both analytically stronger and more forward-looking than similar analyses done by others.

Presenting the Message

In a world in which the informational advantage of the IMF has declined and there are competing sources of analysis, the IMF cannot expect to affect the policy debate unless it has a clearly articulated message that is skillfully delivered through an effective communications strategy. An effective message must provide value added but this is not sufficient in itself. The message must also be packaged in such a way that it is quickly understood and absorbed. Interviews with senior policymakers and journalists in the elite financial press strongly suggest that the IMF’s multilateral surveillance outputs would benefit from greater focus and pointedness. In this section we discuss these issues for the WEO and the G-7 and G-20 surveillance notes.21


The WEO is highly regarded and virtually all those interviewed by the IEO held positive views about its quality. Yet, only a small number actually read a major portion of the report. As noted earlier, the key messages do reach senior policymakers but only through a summary prepared by working-level officials. Outside the public sector, only a small minority in the financial press and academia seems to be reading much of the document. A leading journalist in the elite financial press told the evaluation team that he considered his role to be “deciphering” what the WEO said and to communicate that message to the public. Many interviewed by the evaluation team considered the WEO to be too “dense” and too long to excite their interest.

Undoubtedly, each of the individual topics discussed in the WEO is useful to a particular audience. But the fact that some topic is useful to somebody does not necessarily mean that it should receive heavy coverage, even if the IMF is the most efficient producer of information on that topic. A publication that tries to satisfy the needs of every segment of the broader audience tends to expand in volume and length over time. It thus risks losing focus and its appeal to potential readers.

For example, the WEO’s review chapter (Chapter I) expanded from 40 pages in 2000 to 60 pages (or 90 pages with appendices) in 2005, owing partly to the greater attention paid to economic developments in emerging and other developing countries (Figure 3.4). While this may have increased the WEO’s appeal to a wider geographical readership, it may also have contributed to the widely voiced reservations about its length. Although the overall length of the WEO has remained roughly the same, the key multilateral surveillance chapter (Chapter I) gets the most attention. There is thus considerable scope to streamline this chapter by narrowing its focus.

Figure 3.4.Evolution of WEOChapter 1

(Number of pages)

Model-based scenario analysis is a resource-intensive exercise that cannot easily be replicated by other institutions. It may thus deserve a more prominent place in the WEO, which many feel should focus more on prospects and risks. To be sure, global models were used for policy simulation purposes in every issue of the WEO during 2000-05 (Table 3.3), but the results were rarely cited in the main body of Chapter I; they appeared instead in an appendix, a text box, or in a section of the analytical chapters. Most of these simulations were designed to calculate the impact of various shocks on growth, inflation, and exchange rates in industrial countries. Global imbalances, the sustainability of U.S. external deficits, and the impact of changes in U.S. fiscal deficits have figured prominently as topics, and pointed policy recommendations were occasionally spelled out. More prominent treatment of these simulation results would have increased the WEO’s value added.

The staff’s reluctance to give more prominent treatment to simulation results is understandable, as such results are specific not only to the particular model but also to the particular set of parameter values. In order to increase the general appeal of such model-based scenario analysis, IMF staff should include more parameter values and consider explicitly the implications of model uncertainty for policymakers. Decision-making tools to deal with these aspects of uncertainty are currently being developed and applied for various long-term planning purposes. If scenario analysis is to become a more central part of the WEO exercise, consideration could be given to combining the IMF’s existing models with such emerging methodologies.22

At the same time, in order to reach a wide range of audiences with diverse needs, a variety of products may be needed. For busy senior policymakers, for example, a concise summary of the WEO might be helpful. Similar types of summaries are currently provided to journalists when the document is released. But most important, the IMF’s multilateral surveillance products should have a “customer” focus. Each product must be designed for a specific purpose and with a specific audience in mind. Although the press and working-level officials play useful roles in disseminating information, the IMF should not count entirely on them to “decipher” the WEO’s messages.

G-7 and G-20 surveillance notes

In assessing how G-7 and G-20 notes are presented to their audiences, we relied on interviews with senior officials of several G-7 and G-20 governments, as well as on our review of a number of surveillance notes prepared during 2003-05. At present, G-7 surveillance notes are prepared by IMF staff four times a year, and G-20 notes three times a year. Each note is transmitted to the relevant authorities ahead of the meetings of ministers and governors and their deputies. G-7 and G-20 surveillance notes are very similar. When G-7 and G-20 notes are prepared at about the same time, the two are almost identical. The only difference is that the G-20 note includes a summary of economic developments in a larger group of countries and may also include an annex prepared at the request of the G-20 secretariat. We focus below on the G-7 notes, but almost identical points can be made for the G-20 notes.

A G-7 surveillance note is an 8-15 page, single-spaced summary of global prospects, financial and commodities market developments, and developments in individual G-7 member and major nonmember countries. Often, but not always, one or two short pieces on current policy issues are attached to the note. These annexes can be as short as two pages or as long as seven. Sometimes, they are prepared at the request of the country holding the chairmanship. Recent annexes have included topics related to tax policy, oil market developments, and global imbalances. Text boxes in the main surveillance note have occasionally discussed policy issues.

The summary of prospects and developments constitutes the surveillance note per se. It presents an overview of key prospects and developments succinctly but with sufficient detail. A number of people interviewed by the IEO who had read these notes and participated in subsequent ministerial or deputies’ meetings viewed them as technically competent and useful for providing background for subsequent discussions. The summary is descriptive, however, and largely devoid of policy discussion or implications. IMF staff seems to believe that its primary role is to provide such a summary and is reluctant to step outside this self-imposed boundary. This is not the role which senior policymakers indicated to the IEO that they foresaw for the IMF.

Three aspects of the current mode of presentation limit the value of surveillance notes. First, although each note is well written and short, yet rich in detail, its value added becomes almost nil when seven such notes are prepared (for both the G-7 and the G-20) every year in almost identical format. Each note meticulously updates the previous one, but rarely does a major development change the overall picture, so that any difference from the previous note is usually marginal. Second and more important, the IMF’s policy stance is never stated forcefully, and whatever position the IMF may have is given secondary treatment, by being relegated either to a text box or to an annex. Finally, as short as these surveillance notes may be relative to the WEO, most senior policymakers still do not read them. The information provided in the notes is typically incorporated in the pre-meeting briefing for ministers and governors and their deputies. For such busy people, even a 10-page note is often too long to read unless it is timely and offers clear value added.

These observations suggest that the IMF must radically change the way its surveillance notes are presented. For them to have a real impact, the IMF should more explicitly spell out economic interactions and the consequences of policy spillovers, and where appropriate, provide sharp policy prescriptions. The staff should also attach a brief, double-spaced, one-page summary of its key message to the front of the note, to be read directly by ministers and by governors and their deputies.

Identifying the Impact of IMF Multilateral Surveillance

If IMF multilateral surveillance is to have impact on the policies pursued by systemically important countries, its messages must:

  • reflect the comparative advantage of the IMF and offer value added over information obtainable elsewhere;
  • relate to issues of current interest;
  • reach the intended audience, particularly senior policymakers, and exploit the potential for peer pressure; and
  • be presented effectively.

This evaluation has so far addressed each of these ingredients of effectiveness, noting that while IMF multilateral surveillance has largely succeeded in identifying the right issues for analysis, there was considerable scope for improving the way its messages are delivered.

The ultimate test of effectiveness is the final impact of multilateral surveillance on policies adopted by systemically important countries. Such an assessment is difficult to make for a variety of reasons. Countries make policy decisions as a result of multiple influences, of which IMF advice may be only one and is likely not the most important one. Even if there is evidence that policy adjustments are made as a result of international discussions, it is still difficult to know how much of that is attributable to IMF analysis and advice, as opposed to the persuasiveness of partner countries. Governments in any case rarely wish to acknowledge that policy changes are made for international rather than domestic reasons, still less that they are made as a result of advice from an international institution. The IMF’s influence may be indirect, for example, through its influence on the public debate. Finally, the IMF may be most effective when it has persuaded those being advised to take full ownership of policy decisions.

With this caveat, the IEO has identified, on the basis of interviews with senior member country officials, the following instances in which IMF multilateral surveillance can be said to have influenced policy debate or policymaking in systemically important countries:

  • Risk transfer to household balance sheets. The IMF introduced this issue at the Financial Stability Forum (FSF), and subsequently contributed significant supporting material to the discussion. Several officials told the IEO that the IMF material influenced their own preparation for discussion at the FSF. Similar remarks were made about the IMF’s work on hedge funds.
  • Structural reform in Europe. WEO analysis of this issue had influenced public debate, and the IMF analysis had been frequently quoted, including by the President of the European Central Bank.
  • Global imbalances and oil prices. Officials in several countries cited recent IMF analysis on these topical issues as having usefully informed the internal debate. While in many cases these issues were already on the agenda, policymakers considered the IMF work to be for the most part timely and of value in developing their thinking.
  • House prices. The analysis of house prices in the WEO spurred debate in at least one large European country about why prices had risen sharply in some European countries but not in others.
  • Sustainability of public debt. IMF work on debt sustainability was cited as having prompted further work by officials of a G-7 country.
  • Foreign direct investment (FDI) in financial services. The Basel-based Committee on the Global Financial System (CGFS) took up the issue of FDI in financial services, and one G-7 country official noted that the IMF representative at the meeting was able to contribute a great deal of detailed knowledge to the discussion. Another (from a different G-7 country) called the IMF’s contributions useful and timely.
  • Some G-7 officials told the IEO that IMF research mentioned in the First Deputy Managing Director’s speech on outsourcing services had affected the domestic policy debate, leading some government officials to alter their previously negative view of outsourcing (see Krueger, 2004).

More generally, we found a widely shared view among the policymakers interviewed that IMF multilateral surveillance, if and when effectively conducted and communicated, was an important global public good, with the potential to provide valued inputs into national and (to the extent it exists) global economic policymaking.


    Artis, Michael J., 1997, “How Accurate Are the WEO’s Short-Term Forecasts? An Examination of the World Economic Outlook,” Staff Studies for the World Economic Outlook, World Economic and Financial Surveys (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Boughton, James M., 2001, Silent Revolution: The International Monetary Fund 1979-1989 (Washington: International Monetary Fund).

    Ghosh, Atish, CharisChristofides, JunKim, LauraPapi, UmaRamakrishnan, AlunThomas, andJuanZalduendo, 2005, The Design of IMF-Supported Programs, IMF Occasional Paper No. 241 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Guitián, Manuel, 1992, “The Unique Nature of the Responsibilities of the International Monetary Fund,” IMF Pamphlet Series No. 46 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    International Monetary Fund (IMF), 1999, External Evaluation of IMF Surveillance: Report by a Group of Independent Experts (Washington).

    • Search Google Scholar
    • Export Citation

    Krueger, Anne, 2004, “Trade, Jobs and Growth: Why You Can’t Have One Without the Others,” Address at the Reuters Trade, Globalization and Outsourcing Conference, New York City, June15.

    • Search Google Scholar
    • Export Citation

    Mussa, Michael, 1997, “IMF Surveillance,”American Economic Review, Papers and Proceedings, Vol. 87 (May), pp. 28–31.

    Popper, Steven W., Robert J.Lempert, andSteven C.Bankes, 2005, “Shaping the Future,”Scientific American, Vol. 292, No. 4, pp. 66–71.

    • Search Google Scholar
    • Export Citation

    Timmermann, Alan, 2006, “An Evaluation of the World Economic Outlook Forecasts,”IMF Working Paper No. 06/59 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    U.S. General Accounting Office (GAO), 2003, “International Financial Crises: Challenges Remain in IMF’s Ability to Anticipate, Prevent, and Resolve Financial Crises: Report to the Chairman, Committee on Financial Services, and to the Vice-Chairman, Joint Economic Committee, House of Representatives,” GAO-03-734 (Washington).

    • Search Google Scholar
    • Export Citation

In addition to the 19 G-20 countries (Argentina, Australia, Brazil, Canada, China, France, Germany, India, Indonesia, Italy, Japan, Korea, Mexico, Russia, Saudi Arabia, South Africa, Turkey, the United Kingdom, and the United States), the sample was expanded to include a broad cross section of the IMF’s membership: Algeria, Cameroon, Chile, Colombia, Egypt, Israel, Kenya, Malaysia, Namibia, the Netherlands, Nigeria, Pakistan, the Philippines, Poland, Singapore, Spain, and Ukraine. As a reference, we also looked at the staff reports (and accompanying selected issues papers) for euro area policies.


Integration of multilateral and bilateral surveillance is a high-priority item in the latest Surveillance Guidance Note (issued in May 2005), which suggests that Article IV reports should highlight the domestic consequences of global developments and that surveillance in large countries should pay close attention to the systemic impact of their policies. Guidance Notes are issued by PDR to provide operational guidance to staff on various topics.


Selected issues papers frequently included some analyses of cross-country issues.


Although not shown in the figure, only 15 percent of senior area department staff surveyed “strongly agreed” that “the WEO provides effective support to Article IV consultations.” Another 15 percent “agreed” with the statement.


In commenting on the draft report, the European Department indicated that the WEO more regularly informed bilateral surveillance in that department. Note that the statistical margin of error (at the 95 percent level) for the senior staff survey was about 16 percent.


MULTIMOD is a dynamic multicountry macroeconomic model of the world economy designed to study the transmission of shocks across countries as well as the short-run and medium-run consequences of alternative monetary and fiscal policies. GEM is a new multicountry model with firmer microeconomic underpinnings.


See Background Documents, pp. 43—44; available via the Internet at


See Background Documents, p. 45; available via the Internet at


Barbados, Brazil, Chile, the Dominican Republic, Mexico, and Venezuela.


By comparison, the coverage of the GFSR is much more modest. For example, there were just over 400 press references to the GFSR during 2005, though this may be an underestimation of the actual press coverage. The term GFSR (or its full title) is not as firmly established as the WEO and consequently the Factiva search may not have caught all the relevant references to the publication, especially in non-English languages.


It should be noted that the Factiva database, from which the EXR data are taken, has grown from about 7,000 sources in 2001 to over 9,000 in 2005. The database covers sources in 21 languages.


For example, in the fall of 2005, there were 554 press references to the WEO (compared with 382 references over a similar period for the OECD’s Economic Outlook released in December of the same year).


In our survey, none of the G-7 countries reported that senior policymakers read the WEO and only one country reported that senior policymakers read the GFSR. The interview results are more in line with these survey results for the G-7 countries.


Such information is also communicated to capitals. However, WEMD sessions are informal, and no Board decision or Summing Up is produced.


Whereas in the past the Executive Board may have exercised an unwritten prerogative to suggest that certain politically sensitive information be deleted before publication, the IMF’s more recent transparency policy has transformed the role of the Board in the review process. RES staff indicated to the evaluation team that it now considers Board comments—which are more focused on facts and interpretations, rather than on excising politically sensitive material—to be helpful in ensuring the production of an accurate and well-balanced document.


This is a widespread practice aimed at limiting the length of meetings and is not limited to Board discussions of the WEO.


The figures come from the WEO database. G-20 GDP does not include those EU member countries that are represented at the forum through the European Commission.


Only in 2005 was one-half staff year of dedicated support allocated to this senior staff member.


The EFC has reportedly been increasingly successful in generating common EU positions on economic issues over the past 10 years.


The Eurogroup is an informal group of Ecofin Finance Ministers, established in 1997. The Eurogroup Working Party is the Eurogroup’s counterpart to the Ecofin Council’s EFC. The ministers are accompanied by one EFC member each. Currently, the Eurogroup has 12 member countries represented. One senior official in Brussels noted that the Ecofin Council (with 25 member countries) was large and that more frank discussion took place in the Eurogroup.


We do not consider the GFSR here for two reasons. First, its review chapters, while useful to some, do not offer much value added. Second, its analytical chapters, in the current format, have been of interest mostly to a small group of specialists.


See Popper and others (2005) for one such possible approach.

    Other Resources Citing This Publication