2011 Triennial Surveillance Review - External Study - IMF Surveillance
Coverage, Consistency and Coherence

External study prepared by Stephen Pickford, former IMF/World Bank Executive Director for the United Kingdom and former Managing Director of International and Europe at H.M. Treasury and G-20 Finance Deputy, United Kingdom: Surveillance (both bilateral and multilateral) is a key instrument for the Fund’s crisis prevention role, analyzing economic developments and policies at national, regional and global levels. It also identifies risks and vulnerabilities, and forms the main basis for the Fund‘s discussions with policy-makers.


External study prepared by Stephen Pickford, former IMF/World Bank Executive Director for the United Kingdom and former Managing Director of International and Europe at H.M. Treasury and G-20 Finance Deputy, United Kingdom: Surveillance (both bilateral and multilateral) is a key instrument for the Fund’s crisis prevention role, analyzing economic developments and policies at national, regional and global levels. It also identifies risks and vulnerabilities, and forms the main basis for the Fund‘s discussions with policy-makers.

II. Introduction

1 Surveillance2 (both bilateral and multilateral) is a key instrument for the Fund’s crisis prevention role, analyzing economic developments and policies at national, regional and global levels. It also identifies risks and vulnerabilities, and forms the main basis for the Fund’s discussions with policy-makers.

2. Bilateral surveillance is a universal activity of the Fund, conducted for each of its 187 members. It not only provides advice to governments, but also contributes to the policy debate in wider society, and is a source of information to the wider international community. Multilateral surveillance goes beyond the national level, to the regional and global level: analyzing the linkages between countries, and assessing the potential for policy cooperation and coordination. The Fund has also in recent years analyzed more comprehensively national and global risks and vulnerabilities, and the spillovers from national policies onto other countries; this forms a bridge between bilateral and multilateral surveillance. There are also strong interactions between bilateral and multilateral surveillance Good multilateral surveillance depends on a deep and thorough knowledge acquired from bilateral surveillance of individual countries and their policies; and good bilateral surveillance takes into account the economic conjuncture and spillovers from policies in the rest of the world, and draws on the experience of what has (and has not) worked in other countries. Ultimately the success of surveillance rests on whether it helps policy-makers make better decisions for the benefit of the countries concerned and/or the wider international community. To do so surveillance needs to be knowledgeable and persuasive, and the recent economic and financial crisis has added to the pressure on the Fund to ‘get surveillance right.’

3. Having a consistent and coherent set of views across countries (while taking due account of differences in circumstances) is important to provide good analysis and advice, and to ensure that the Fund’s advice is persuasive. Increased globalization also makes interlinkages and spillovers between countries quantitatively and qualitatively more important. And in practice the distinction between bilateral and multilateral surveillance is becoming more blurred. Also the growth of regional and other supra-national economic groups brings a new dimension and complexity to surveillance, both in terms of substance of surveillance and its process. This report, which is an input to the 2011 TSR, looks at three main issues:

  • Coverage : are there gaps or overlaps in Fund surveillance?

  • Consistency and coherence : are the various surveillance products consistent in their analysis and policy advice, and well-integrated?

  • Other fora: what are the implications for the Fund of other bodies carrying out surveillance-like assessments?

III. Coverage

4. Article IV reports—the ‘workhorse’ of bilateral surveillance—are prepared for each of its 187 members, usually once a year.3 Bilateral surveillance has been primarily concerned with the internal coherence of a country’s policies; it has not looked consistently at whether a country’s policies have adverse effects on other countries. The recent crisis highlighted the potential for these negative externalities, and the Fund is currently experimenting with ways to analyze these impacts, in particular through ‘spillover reports’ which look at the international impacts of a country’s policies.

5. The Fund’s multilateral surveillance and related work operates at many levels:

  • the highest profile multilateral surveillance is carried out at the global level: the main products are the World Economic Outlook (WEO), the Global Financial Stability Report (GFSR), and the Fiscal Monitor (FM);

  • at the sub-global level, reports are prepared on common currency areas4 (the Euro area, WAEMU, ECCU and CEMAC) as part of the Article IV reports on individual members of the currency area. Also, each area department now produces a Regional Economic Outlook (REO), drawing in part on the global products;

  • Fund staff also prepare reports for specific groupings of countries: assessments are prepared for each Ministerial meeting of the G-7 and G-20; and in 2009 a consolidated report was produced bringing together the main conclusions of Article IV reports for 5 systemic countries; and

  • multilateral products also include thematic reports which cover a group of countries, focusing on a specific theme or issue. Examples here include: the Multilateral Consultation in 2006/7 which looked at the issue of global imbalances; more recently, cross-cutting thematic reports (which have covered oversized banking sectors, employment, and capital flows); and Vulnerability Exercises and the Early Warning Exercise which regularly assess global risks.

A. Gaps

6 Significant gaps in the coverage of surveillance were identified following the onset of the recent crisis. In particular the 2008 TSR concluded that the Fund needed to strengthen its efforts in: risk assessment; macro/financial linkages; multilateral perspective (including spillovers and cross-country analysis); and exchange rate assessments5. The Fund responded by introducing a number of new products and processes aimed directly at filling these gaps

7. On risks and vulnerabilities, the WEO explicitly considers 'baseline risks’ to the central forecast. The GFSR in addition looks at 'tail risks’ which could pose potential risks to financial stability. And Vulnerability Exercises (VEs) for advanced economies, emerging markets and LICs and the Early Warning Exercise (EWE) presented at the IMFC look at a broader range of tail risks6.

8. Nevertheless, surveys carried out for the TSR7 indicate that this is still seen as an area where more could be done. There is general recognition both outside and inside the Fund that there has been a significant improvement in risk assessment in bilateral surveillance. But countries, EDs and market participants all feel that Article IV reports should pay still more attention to risks, especially for advanced economies. And in interviews countries said they wanted more discussion of risks to their economies from international developments, and possible policy responses.

9. On macro-financial linkages,8 efforts have been made to strengthen the Fund’s analytical base. For example, integration of the WEO and GFSR has been improved (including in one instance a joint foreword), and a number of departments have stepped up analytical work (including setting up a macro-financial unit in the Research Department).

10. New products and processes have also been introduced in recent years to improve analysis of interlinkages and spillovers:

  • new products include: the 2009 report on 5 systemic countries, inputs to the G-20 (both the surveillance note and MAP inputs), and spillover reports (currently in preparation); REOs also cover regional interlinkages to some extent; and

  • new processes aim to strengthen bilateral-multilateral links: cross-departmental surveillance meetings; reforms to the review process for country reports; and regional issues units have been set up in all area departments.

11. Nevertheless, the Fund is still perceived as needing to do more. The TSR surveys report that country policy-makers think Fund surveillance contributes relatively little to their thinking on spillovers (both inward and outward) compared to other issues. The Fund will review again its coverage of spillovers in the light of experience with spillover reports.

12. And on exchange rate issues,9 the 2007 surveillance decision adopted by the Board puts external stability (including exchange rate issues) at the heart of surveillance. The TSR surveys show that countries and financial markets regard exchange rate issues as an area where surveillance is making relatively less contribution to their thinking.

13. Some of these new products and processes have been introduced only very recently, so it is too early to tell how successful they have been in addressing fully the issues raised in 2008. The general conclusion of this review is that in most areas the Fund has made significant moves in the right direction but more is expected from them. The survey results suggest that most users of Fund surveillance do not see big gaps remaining in its coverage, but there are areas where more attention should be paid.

14. Risks and vulnerabilities are one particular area where there appears to be a gap in the Fund’s public output. While the Fund has done a great deal of work in developing methodologies for identifying risks and vulnerabilities, little of that work is published. There can be good reasons why the VEs and EWE remain confidential to the Board and the IMFC, not least because of the risk that Fund statements in public could be destabilizing and could increase the probability of risks materializing. The possibility of publication could also lead to staff being more reluctant to identify sensitive risks, which in turn could have reputational costs for the Fund.

15. Managing more explicit public presentation of the Fund’s views on risks would require careful handling, in a way which does not lead to sharp market reactions. It would be necessary to reach a judgment on each occasion about how much of the Fund’s analysis of risks and vulnerabilities could be published (as opposed to being incorporated in confidential briefings for Ministers). And that judgment would depend in part on the circumstances at the time. In general, the Fund should be cautious about publishing conclusions that could generate destabilizing near-term movements in asset prices; and where risks are identified that could trigger sustained longer-term market reactions, these should be accompanied by advice on policy actions that would mitigate the risks.

16. But the Fund has already moved towards being more explicit about risks, for example in its public warnings on the balance sheet risks of banking systems and the need for greater write-downs of bad assets and bank recapitalizations. And many of the risks identified to date in its confidential reports had also been highlighted publicly by market and other commentators. The conclusion of this report is that publication of more of the Fund’s thinking on risks and vulnerabilities would be in the public interest; and that provided due regard is paid to market sensitivity there seems no overriding argument for maintaining the current level of confidentiality surrounding these issues.

Recommendation 1:The Fund should incorporate the main conclusions of its work on vulnerabilities and risks in its main multilateral surveillance reports.

B. Overlaps

17. While gaps have largely been addressed, the introduction of new products has increased the level of overlaps between different products. There are now a wide range of surveillance products, both bilateral and multilateral (and some bridging that divide), covering similar issues to a greater or lesser degree (see Table 1).

article image

Surveillance products are italicized in blue bold. FSSAs are regarded as surveillance products for 25 systemically-important financial sectors Notations

As in Recommendation 3 in this paper. Yes

Indicates average lengths of each of the products.

Further updates have also been published on occasion between the regular publication dates.

18. At the latest count, the Fund publishes every year around 2,300 pages of multilateral surveillance products and some 3,200 pages of Article IV reports. While each product is differentiated to some degree (in country coverage, depth of analysis, audience or communication channel) it is valid to question whether so many different products are required, or whether some could be dropped or combined, or produced less frequently. There is also a general presumption in this report that the Fund should reduce the number, and size, of its products, rather than increase them. At the same time the Fund is expected to be comprehensive in its surveillance.

19. The three flagship reports overlap to some degree (Box 1). All three cover both conjunctural and analytical issues and risks, though from somewhat different perspectives. There are also clear interlinkages between the issues covered in each of the three reports, and some issues cut across all three reports (for example, problems in the euro area periphery, and international capital flows). But in general the overlaps appear to be kept to a manageable level.

The IMF’s Flagship’ Reports

The IMF produces three 'flagship’ products: the WEO, the GFSR and the FM. Each is published twice a year, with up to two additional updates through the year. And each covers both conjunctural (recent developments, prospects and policies) and analytical issues.

The distinction between conjunctural and analytical issues is made more clearly in the WEO (chapters 1 and 2 cover global and regional prospects and policies respectively) and the GFSR (chapter 1 covers current risks and challenges). Although some analysis is included in boxes in these conjunctural chapters, the more theoretical analytical content appears in subsequent chapters. The FM is organized rather differently, with four chapters covering: deficit and debt; financing and debt markets; shocks to the outlook; and risks and policy responses. Analytical issues are covered both in the four chapters and in appendices.

There are clear interlinkages between the issues covered in each of the three reports; and there is a question whether they should remain as three separate products, or whether they should be combined in some way. It is difficult to measure the overlap precisely, but as a rough indication the April 2011 editions show that in the WEO, out of a total 28 pages in chapter 1 (excluding appendices) approximately 4 pages covered financial developments and issues, and 2 pages covered fiscal issues. Both the GFSR and the FM spend relatively little space discussing the wider macroeconomic situation and policies, except where it affects financial and fiscal issues. But both reports cover the relationship between public deficits and debt and financial conditions in bond markets.

There are also policy issues which necessarily cut across all three reports. For example, the April 2011 editions of the three reports all discuss three topical cross-cutting issues: problems in the euro area periphery; support mechanisms for financial sectors; and international capital flows.

All three reports include a discussion of risks. The WEO is focused more on 'baseline’ risks (probability distributions around the central baseline projection), whereas the GFSR (and FM to a lesser extent) consider also 'tail risks’ - less probable events, but with significant impacts. The GFSR and FM also have adopted classifications of types of risk to sustainability which are assessed regularly - the GFSR looks at macroeconomic, emerging market, credit, and market and liquidity risks; the FM judges them against fiscal indicators, long-term fiscal challenges, liability structures, macroeconomic, financial sector, and policy implementation risk.

All three also assess vulnerability to shocks and policies to mitigate or build resilience to them. Some of the risks and vulnerabilities identified are specific to one of the three reports. But others (for example, fiscal exposure to financial sector stress) are common to all three.

If the reports remain separate, there is a strong argument for ensuring that overlaps are minimized, and that each document cross-refers to relevant data and analysis in the other documents. For policy issues, risks and vulnerabilities which cut across the different reports, it is important that the policy messages are consistent.

There is also the question of whether they should be published so frequently, given the resource cost involved. When events and policy challenges are changing rapidly, there is a stronger case for more frequent publication. In general the underlying data on financial and general economic developments is of higher frequency than fiscal data, as is also the case with policy 'events,’ with annual fiscal budgets still the norm in most countries.

20. All area departments now publish Regional Economic Outlooks (Box 2), covering the current economic situation, prospects and policy challenges for the region; and all include cross-country analytical work of regional relevance. The REOs are able to provide more detail on conjunctural issues in the region than the WEO and the GFSR, though they are based on the WEO projections. Some REOs include a short summary of the WEO projections, and these ensure consistency. All REOs also incorporate analysis of crosscountry and regional themes and issues, which countries want the Fund to do more of (although in many cases the implications go wider than the region itself).

Regional Economic Outlooks

All area departments now publish twice-yearly reports for their region. These reports are somewhat different in scope and coverage, but all include a substantial section on the current economic situation, prospects and policy challenges for their region; and all include cross-country analytical work, both within the conjunctural sections and in separate chapters (as in the WEO and GFSR).

One issue with these reports is whether the ‘region’ (as defined by the country coverage of the Fund’s area departments) is an economically coherent grouping. In all cases there are important differences between the countries making up the region, and this is reflected in the REOs. The conjunctural sections of each report are divided into more coherent sub-groupings:

  • Sub-Saharan Africa: resource exporters; middle-income diversified exporters; and low-income country diversified exporters;

  • Europe: advanced Europe (EU and euro area as sub-components); and emerging Europe;

  • Middle East and Central Asia: Middle East, North Africa, Afghanistan and Pakistan; and the Caucasus and Central Asia;

  • Asia Pacific: Asia; and low-income countries and Pacific Island economies; and

  • Western Hemisphere : the US and Canada; and Latin America and the Caribbean.

The REOs are seen by many Fund staff as having an important role as outreach and communications tools, with a number of area departments reporting strong interest from policy-makers and other audiences in their region. This is confirmed by survey results from countries, though there is much stronger interest from EMEs and LICs than from advanced economies.

One role for the REOs is to provide more detail of conjunctural issues in the region than the WEO (and to a lesser extent the GFSR) is able to. For example, in the April 2011 WEO chapter 2 (Country and regional perspectives) ran to 30 pages, while the REOs in total contained nearly 200 pages on conjunctural issues. The difference was even more marked for the smaller and less systemic countries and regions: the WEO chapter 2 spent about 3 pages each on Latin America and the Caribbean, sub-Saharan Africa, and the Middle East and North Africa; the respective REOs covered these regions in more depth (devoting around 25 pages in each case).

In addition, the REOs incorporate additional analysis of cross-country and regional issues.10Examples of regional analysis in the April 2011 issues of the REOs include: the impact of the East African Community on growth; imbalances in the euro zone; youth unemployment and social protection in the MENA region; the impact of capital inflows in Asia; and the effectiveness of foreign exchange market intervention, in particular in Latin America.

In many cases the implications of these analyses go wider than the region itself. It is important to ensure both that these regional analyses are consistent with the experience of other regions, and also that (where relevant) policy conclusions are taken account of in other regions.

21. The survey results give some pointers to the value placed on the Fund’s multilateral products by different groups. The WEO is rated the most highly by all groups, outside and inside the Fund, for its usefulness. But for the other products there are some important differences between groups’ ratings:

2011 TSR Survey Results

(Usefulness of multilateral surveillance and related products to: 1 = most useful; 7/8 = least useful)

article image

22. Media coverage is also concentrated on the WEO. At the Spring Meetings this year, media interest in the Fund’s surveillance products gauged by media articles in the first 48 hours after publication were by far the highest for the WEO. Media Interest in Fund’s Multilateral Surveillance Products (2011 Spring Meetings)

Media Interest in Fund’s Multilateral Surveillance Products

(2011 Spring Meetings)

article image

23. Reviewing the contents of the various multilateral products points to the following conclusions:

  • Although the main surveillance products appear to address somewhat different audiences, there may be scope for some rationalization One possibility would be to merge the WEO and the GFSR (which might also help fuller integration of macroeconomic and financial issues). But given their rather different focus and audiences, and the high ratings these publications receive, this step should not be taken lightly. There is a stronger case for merging the WEO and FM because there is greater overlap of the issues covered, but again they are likely to have rather different audiences among national policy-makers;

  • However, it is worth giving serious consideration to publishing a separate report which brings together the conjunctural material contained in all three flagship publications. This would be additional to the full versions of the WEO, GFSR and FM; but it would replace the ‘summary version’ of the GFSR (which currently comprises the conjunctural chapter of the GFSR, with web links to the analytical chapters);

  • An alternative is that the GFSR, FM and REOs could include a very short summary of the global situation and prospects taken from the WEO, as background for the subsequent discussion of financial and fiscal issues respectively. This already happens in some of the REOs (and should be the practice for all REOs); and

  • In some cases a lower periodicity of publication would be possible For example, the main fiscal policy decisions in many countries are made annually. This would suggest that the frequency of the FM could be reduced to once a year. It is also not clear that the benefits in terms of communication and outreach based on frequently updated REOs outweigh their costs of production twice a year, given that they draw to some extent on Article IV reports which are on an annual cycle. Publishing the FM and REOs once a year would also allow more resources to be directed to analytical work.

Recommendation 2:A rationalization of the flagship multilateral surveillance products should be considered along the following lines11:

a. theWEO and GFSR produced twice a year (at the Spring and Annual Meetings). In normal circumstances additional updates should not be required;

b. the FM produced once a year at the Annual Meetings, and focused more on longer term analysis (and less on conjunctural issues, which could be covered more fully in the WEO);

c. the GFSR ’summary version’ could be replaced by a separate report, bringing together the conjunctural material from the WEO chapter 1, GFSR chapter 1, and FM ; or alternatively the GFSR and FM could include a short summary of the global situation and prospects from the WEO; and

d. REOs produced once a year, providing an expanded regional dimension to the conjunctural analysis in the WEO and GFSR, but with greater focus on analytical work of particular relevance to the region. This regional analytical work should be dovetailed with wider analytical work (for example as part of the WEO, GFSR and FM), to draw together regional and global analysis, ensure consistency, and avoid unnecessary duplication.

IV. Consistency And Coherence

24 Consistency and coherence across all products is very important for the Fund to produce high-quality analysis and gain acceptance for its policy advice At a minimum it is important that the various surveillance products are not inconsistent. But to achieve full coherence requires that connections (economic linkages, policy linkages, or intellectual linkages) are made across the full range of surveillance; and that similar conclusions are drawn in similar circumstances (across countries or time).

25. Looking across the range of the Fund’s surveillance products, consistency and coherence needs to be achieved:

  • across countries: ensuring that bilateral surveillance delivers consistent analysis and advice to countries in similar circumstances;

  • between bilateral and multilateral: making sure that advice in an Article IV context is consistent with, and is not contradicted at, the multilateral level (and vice versa); and

  • across multilateral products: ensuring that the different products, which are all to some degree interrelated, give consistent messages.

26. There are a number of ways that consistency across the range of surveillance products can be measured and achieved. Some are relatively straightforward: using common and consistent data sets; using consistent projections, including between the WEO and the Fund’s country forecasts; and ensuring that headline policy messages tie up across the various multilateral products and country-specific reports.

27. These elements are important, and it is essential to have processes in place to guarantee this level of consistency. Some countries and EDs, in particular emerging markets and LICs, felt that the Fund needed to improve the consistency of its policy advice across countries.12 But in other areas, although some inconsistencies do appear on occasion (including on policy messages13), interviews with Fund staff and EDs suggested that cases of inconsistencies were relatively limited. This suggests that the review processes in place to join up different parts of the Fund across multilateral products, and between bilateral and multilateral surveillance, do seem to be working fairly effectively to prevent serious consistency problems.

28. However, there are a significant number of products which provide ’top line’ messages on global policies and priorities. In addition to the ‘flagship’ products, the MD’s statement to the IMFC, G-20 surveillance notes, and REOs all carry some ‘top line’ messages. And the Vulnerability Exercises and Early Warning Exercise, while not published, contain further analysis of risks. But at present no one document draws them all together.14 There is a strong case (in the interests of consistency and clarity of message) for a single concise15 overarching statement of the Fund’s view across the board on prospects, risks and policy priorities:

Recommendation 3: At the centre of the Fund’s multilateral surveillance output should be a single concise overarching document, covering global prospects, policies, spillovers and risks, which candidly and clearly brings out the key policy messages from the Fund. This report would be presented by the MD to Ministers (and published) at the Spring and Annual Meetings, with two mid-term updates around January and July (which could also be the basis for discussions at the G-20 meetings).

A. Integration

29. The previous section has been primarily concerned with avoiding inconsistencies between products. But consistency should go further than that minimum level, ensuring that consistent approaches are applied across the Fund’s surveillance activities, and that connections are made across countries and across issues. This ’super-consistency’, or integration, of surveillance is harder to achieve, and requires more than simple crosschecking processes.

30. Consistent approaches to similar issues would encourage the spread of best practice across the full range of surveillance work. Current examples include: using consistent methodologies for measuring how much exchange rates deviate from equilibrium; the Fund’s work to develop a common methodology for assessing the adequacy of foreign exchange reserves; and the models developed to identify risks in the VEs.

31. Drawing out the inter-connections between countries and between issues is equally important. One of the Fund’s comparative advantages lies in its ability to look across its entire membership, and to use that experience to provide appropriate policy advice: on national policies, based on the experience of other countries in dealing with similar issues; on policies that would minimize the potential for adverse spillovers onto other countries, and to mitigate risks; and on policies to deal with issues that need to be addressed through policy cooperation across countries (e.g. global imbalances).

B. Connecting the Dots

32 One conclusion that has featured in many reviews of Fund surveillance over the years is that it needed to get better at ’connecting the dots’ 16 This means that it should do better at:

  • Connecting bilateral and multilateral surveillance;

  • Connecting macro and financial surveillance;

  • Making connections across countries; and

  • Thematic surveillance which looks at an issue across countries and regions.

33. Again, a number of changes have been introduced in recent years to address these issues. In order to improve connections between bilateral and multilateral surveillance:

  • Weekly surveillance meetings are used to disseminate information and encourage cross-departmental discussion; and

  • Policy coordination meetings and policy issues notes are prepared before each Article IV mission, and cross-country analysis is commissioned in advance of the mission to inform discussions with the country authorities.

34. But in interviews with countries, EDs and staff some felt there was a need to improve further the links between bilateral and multilateral surveillance, in particular to incorporate multilateral conclusions and financial sector work into Article IVs. The following further process steps would help to more fully integrate bilateral and multilateral surveillance:

Recommendation 4: To increase the consistency of bilateral surveillance with multilateral products:

a. where Financial Stability Assessments are carried out, they should be fully integrated with the Article IV mission, initially at least in the 25 countries for which stability assessments under the FSAPs will be mandatory. Also, for these countries the Fund should conduct financial stability assessments more frequently than once every 5 years;

b. all Article IV teams should be required to explicitly consider whether the main policy issues raised in the single overarching document17 are relevant for that country. This should be discussed at each policy consultation meeting prior to the Article IV mission;

c. the weekly surveillance meeting should be developed further as a systematic channel for linking bilateral and multilateral developments;

d. composition of Article IV teams should fully reflect the policy issues relevant for the country (including those identified in the single overarching document), and incorporate staff from all relevant functional departments; and

e. presentation of multilateral surveillance outputs to national policy-makers (and more widely) should routinely feature as part of Article IV missions.

35. Cross-country issues are also an area where there is continuing demand for the Fund to do more work. For example, the TSR surveys showed that greater use of crosscountry comparisons was thought by countries to be one of the most important areas for the Fund to improve in its surveillance. Countries also reported that inward and outward spillovers were the areas where surveillance had made least contribution:

Recommendation 5: To strengthen cross-country work:

a. outward spillovers should be a) systematically identified and b) routinely incorporated in Article IV consultations, at least for the most systemically important countries; and

b. reports drawing out the themes from Article IV reports for systemic countries should be produced on a regular basis.

C. Thematic Surveillance

36. The Fund has always carried out thematic work, but recent innovations include: the VEs and the EWE which look systematically at risks and vulnerabilities18; reports on cross-cutting themes (including on countries with over-sized banking sectors, and employment); and the cross-departmental working group recently set up to work on international capital flows. Also there has been more thematic work on issues affecting specific regions.

37. However, the Fund seems to find it easier to draw these connections using special projects and inter-departmental teams. Cross-departmental teams and working groups on specific issues should be encouraged. Ideally this should become a more generalized feature of the Fund’s work, and the challenge is to find management structures that encourage these types of analysis.

Recommendation 6: Cross-cutting themes should play an increasingly important part of the Fund’s analytical surveillance work, and the Fund’s working methods need to allow and encourage this. In the first instance there should be greater use of cross-departmental teams and working groups on specific issues.

38. Finding a way to encourage staff to make these connections spontaneously, without setting up new processes and structures, is more of a challenge. The IEO’s view was that this shortcoming was endemic and stemmed from a silo mentality.19 In the longer term

management will need to send clear messages about the value placed by the organization on collaborative and cross-departmental styles of working, and set out incentives to encourage them, using the internal staff assessment process to recognize and reward them:

Recommendation 7: Cross-cutting working should increasingly be recognized as a valued part of staff members’ output. In particular the internal staff assessment process should explicitly recognize and reward membership of cross-departmental units/projects, virtual teams, and input into the work of other departments.

V. Implications of Regional and Other Groups

39. Although other institutions (such as the OECD, the European Commission, rating agencies and market analysts) have for many years made country assessments covering similar issues to Fund surveillance, a number of other governmental-based fora are now carrying out self-assessments of their members’ economies and policies. Some of these fora have their basis in the obligations placed on countries as members of a currency area or of a regional financing arrangement (including the Eurozone, and AMRO, which is being set up to support the Chiang Mai Initiative). And in both these cases they have a potential role in providing finance through mutual support mechanisms to members in difficulty (sometimes jointly with the Fund). Other bodies reflect the common and mutual interests of their members, for example because of economic links or financial inter-connectedness (in particular the G-20 and FSB).

40. Since its creation as a Leaders’ level body, the G-20 has increasingly concerned itself with assessments of its members’ economic policies, and their interrelationships. The G-20 has relied heavily on the IMF for analysis and advice, through the surveillance note provided for each G-20 meeting of Finance Ministers and Governors, and of Leaders. And the IMF is playing a key role in the G-20 Mutual Assessment Process (launched at the Pittsburgh Summit in 2009), providing technical support and advice, and commenting on the consistency of country forecasts and policies.

41. The existence of these bodies, which are carrying out self-assessments similar in scope to the IMF’s surveillance, raises some important questions for the Fund: how far the Fund should be involved in the activities of these other fora; and whether there are lessons for ways to improve the Fund’s own surveillance.

42. The emergence of these fora raises the possibility of overlap and duplication with the Fund’s own surveillance role. Their outputs do frequently cover very similar ground. But there are distinct differences between the way they operate: the groups carry out self-assessments through a dialogue between member countries’ senior policy-makers, with the Fund providing analytical and technical support; whereas Fund surveillance involves dialogue between staff and country authorities, and peer review carried out by members’ representatives in the Executive Board and through discussion at the IMFC.

43. As far as the Fund’s potential links with these fora are concerned, they provide a challenge for the Fund, but also an opportunity. Because these groups only cover a portion of the Fund’s membership, Fund involvement with these groups can be seen as challenging the universal nature of the Fund’s remit. There is also a risk that the Fund’s independent role could be compromised if it is seen to be too closely involved in the work of these groups.

44. The Fund therefore needs to be careful to maintain a clear distinction. In providing analytical and technical support to these fora it is responsible and accountable for its own input to the process; but the ultimate judgment on countries’ economies and policies (the output from the process) lie with the members of the fora. On the other hand in the Fund’s own surveillance work, both bilateral and multilateral, it is responsible not only for the inputs but also for the judgments embodied in the outputs.20

Recommendation 8: Collaboration with supra-national groupings requires clear protocols and structures for the Fund’s involvement. In particular the Fund will need to be clear on its role: when it is just providing advice and assistance, when it is a full participant in the groups’ discussions, and when it is jointly responsible for the output of the discussions.

45. Although some observers are concerned by the Fund’s involvement with these groups, this report takes the view that on balance there is a strong case for the Fund to provide technical support to these groups. The Fund has much to offer in the way of expertise; and it can help reduce unnecessary duplication of effort for these groups to leverage off the Fund’s own existing surveillance work (at relatively low cost to the Fund itself). More importantly, involvement with these groups can also help increase the traction of the Fund’s surveillance work, both bilateral and multilateral. By allowing a direct dialogue with groupings of policy-makers who can influence each other, this provides the Fund with another avenue to inform and influence national policies.

46. There are also implications for the way the Fund organizes itself to interact with these groups. It will need internal structures that allow the Fund to organize its input to these groups and to maximize the synergies with its own surveillance activities. In some cases this happens already: for example, interactions with the European Commission on euro zone discussions are led by the Euro area team in EUR. And while staff are already building up relations with the Chiang Mai Initiative, there is also a case for setting up a specific team in APD to manage interactions with AMRO.

47. It will be more difficult for the Fund to internally manage these interactions when the groupings cross internal departmental boundaries. But if the Fund is have an ongoing interaction with and input to other groups, there is a strong case for setting up similar arrangements.21

Recommendation 9: The Fund should align its organizational structures to strengthen links with regional and other groups.

48. There may also be lessons for the Fund in the way it carries out its own surveillance, both on substance and process. In particular these groups are undertaking self-assessments which focus on cross-country connections and spillovers, which could inform how the Fund approaches these issues. For example, the indicators being developed for the G-20 MAP may suggest ways in which Fund surveillance can be improved. The direct in-depth dialogue between policy-makers in these groups may also suggest ways to improve the Fund’s own consultations with its members.


This paper represents the views of the author and does not necessarily represent IMF views or IMF policy. The views expressed herein should be attributed to the authors and not to the IMF, its Executive Board, or its management.


Stephen Pickford is a former IMF/World Bank Executive Director for the United Kingdom and former Managing Director of International and Europe at H.M. Treasury and G-20 Finance Deputy, United Kingdom. Research support was provided by a team consisting of Gillian Nkhata, Kingsley Obiora, Tetsuya Konuki, and Alison Stuart.


Report recommendations are summarized in bold; they are contained in full in the body of the report.


Article IV of the Fund’s Articles of Agreement forms the basis for surveillance, requiring the Fund to ―oversee the international monetary system ... and the compliance of each member with its obligations.‖ Surveillance therefore has a specific legal definition, including concerning the role of the Executive Board. This report is not primarily focused on the legal issues; it covers both Fund surveillance and related activities.


There are important questions about whether bilateral surveillance is carried out at the appropriate level for country circumstances and whether it treats all countries equally - the issue of ‘evenhandedness’ - but these are outside the scope of this report, and are covered elsewhere in the TSR.


Technically these reports are part of bilateral surveillance.


Surveys carried out for this TSR (see TSR Health Check and Statistical Information) show that financial market participants rate the Fund’s contribution in Article IV reports (compared to the input of other commentators) least highly in the areas of exchange rates, cross-border risk transmission, and transmission of risks between the financial sector and real economy - though in all cases the Fund is still seen as adding value.


However, there does not appear to be a clear distinction between 'baseline’ and 'tail’ risks. Greater clarity on the coverage of risks in different products would be helpful.


For example, 40% of countries (and 22% of EDs) thought that Fund advice for other countries signaled risks too infrequently (though only 9% of countries, and no EDs, felt the same was true for their own countries).


See also Chapter II of TSR Staff Background Studies.


Covered in more detail in Chapter I of TSR Staff Background Studies.


The IEO evaluation report on Research at the IMF found that most country authorities (except in Africa) did not consider the REOs and their analytical chapters very useful or insightful; and that the technical quality of analysis in REOs was lower than for other publications (although rising over time). However, in the TSR surveys, country authorities in emerging markets and LICs reported that REOs were the second most helpful multilateral surveillance product (after the WEO).


See also the following section of this report.


The issue of even-handedness across bilateral surveillance is not addressed in this report, but is covered elsewhere in the TSR.


For example, some EDs took the view that the latest WEO and GFSR reached rather different conclusions about the speed of bank recapitalization required.


The April 2011 IMFC Communiqué also called for ’ consolidated multilateral surveillance report.’


If a separate surveillance report was published, bringing together the conjunctural material in the WEO, GFSR and FM – as suggested in Recommendation 2(c) – this would in April 2011 have been around 80 pages long (or well over 150 pages including boxes, annexes and references). The document envisaged in Recommendation 3 would have a wider coverage, but be considerably more concise than this.


For example, the recent IEO report on the IMF’s surveillance in the run-up to the recent crisis looked exhaustively into its quality, and in particular into how well it identified in advance the risks and vulnerabilities that led to the crisis of 2007 onwards. In particular it concluded that:

  • IMF surveillance "did not convey a clear message to the membership about the urgent need to address financial sector risks", and "paid insufficient attention to risks of contagion or spillovers from a crisis in advanced economies";

  • "Some of the risks ... identified at different times in the GFSR ... were not reflected in the WEO or in the IMF’s public declarations"; and "discussion of the risks and vulnerabilities ... never found its way into the bilateral surveillance of the largest systemic financial centers";

  • There was a "lack of a coherent macro-financial storyline";

  • "The internal review process failed to 'connect the dots'" : it "did not connect bilateral and multilateral surveillance, or coordinate the analysis of the WEO and the GFSR"

The Fund staff’s own analysis, "Initial lessons of the crisis" (2009), also concluded that 'a fragmented surveillance system compounded the inability to see growing vulnerabilities/risks. ’


See Recommendation 3.


The VE for emerging markets has been conducted regularly since 2001, but the VE for advanced economies was only introduced in 2009. The VE for LICs will commence this year.


The IEO report into the recent crisis concluded that “turf battles and silo behavior ... made it difficult to integrate multilateral with bilateral surveillance, to link macroeconomic and financial developments, and to draw lessons from cross-country experience.” And the TSR survey of Mission Chiefs indicated that lack of knowledge sharing within and across departments made surveillance work more difficult.


Clarity on the Fund’s role is even more important where joint financing arrangements are involved.


For example, cooperation with the FSB will need to involve in particular both MCM and the country teams covering the individual FSB members. Interaction with the G-20 involves an even greater number of departments.

2011 Triennial Surveillance Review - External Study - IMF Surveillance: Coverage, Consistency and Coherence
Author: International Monetary Fund