## _073014h - EXECUTIVE SUMMARY ## Source details **Canonical URL:** [_073014h - EXECUTIVE SUMMARY](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2014/_073014h.pdf) ## Other formats - [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2014/_073014h.pdf.md) - [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2014/_073014h.pdf.json) --- ### Importance of evenhandedness - Evenhandedness is a “cornerstone” for a cooperative institution like the Fund and directly impacts legitimacy and the effectiveness of surveillance. - IMFC communiqué (October 2, 2004): “Effective and even-handed IMF surveillance across the whole membership is central to promoting high and sustainable growth in member countries and to crisis prevention.” - Managing Director (December 2011): credibility and effectiveness of the IMF depends “on our honesty, on the quality of our work and on evenhandedness.” - Perceptions that the Fund is not even-handed can influence behavior and reduce receptiveness to Fund advice. ### Conceptual framework: inputs vs outcomes - Evenhandedness should be assessed on the “inputs” to surveillance (the way surveillance is conducted), not only on “outcomes” (policy assessments and advice in reports). - Inputs include decisions over resources deployed, issues to be covered, depth of analysis, policy advice provided, and presentation of advice. - Tailoring surveillance to country circumstances can produce different outcomes without implying lack of evenhandedness provided inputs are unbiased. ### Findings on perceptions and evidence - Significant, often long-standing, perceptions exist that the Fund is not even-handed; many perceptions do not relate directly to surveillance but can affect effectiveness of advice. - Differences in surveillance were identified but differences alone are not definitive evidence of lack of evenhandedness. - Differences fall into three categories: - Appropriate tailoring of inputs to country circumstances. - Occasional differences not well-justified; inconsistencies problematic and should be addressed. - Pockets of systematic differences across country groups representing lack of evenhandedness, for example: - less rigorous surveillance for program countries, - minimal reporting on the response to past Fund advice for advanced economies, - a steep reduction in resources from functional departments for surveillance for emerging markets and low-income countries, - main multilateral surveillance products paying little attention to developing countries. - Note: the latter two aspects could be explained by a risk-based approach to surveillance. - Review did not find evidence of a pervasive lack of evenhandedness in surveillance, but small instances matter given entrenched perceptions. ### Perceptions by stakeholder groups - Advanced countries more inclined to consider Fund surveillance even-handed, though many can cite instances where the Fund was not sufficiently critical. - Mission chiefs generally felt unconstrained in expressing views, with some acknowledging greater pressure on major advanced or emerging market cases. - Authorities from emerging markets and developing countries expressed more concern over lack of evenhandedness, often linking it to governance and the need to reform Fund governance arrangements. ### Recommended measures to strengthen evenhandedness (overview) - The review recommends extra measures to strengthen and demonstrate strengthened evenhandedness: - Clarify what is meant by being even-handed, including more explanation in the staff guidance note on surveillance. - Provide more explanation in Article IV reports on tailoring and on why advice differs from other countries; encourage greater use of cross-country comparisons and thematic papers. - Highlight the role of all reviewing mechanisms—senior staff in area departments, reviewing departments, Management and the Executive Board—in ensuring evenhandedness. - Raise and address concerns over lack of evenhandedness in a transparent and well-substantiated manner; staff should prepare a periodic report on evenhandedness for the Executive Board, with the first report outlining respective roles of Management, staff and the Executive Board. ### Positive recent developments - Increased focus on spillovers. - Adoption of the integrated surveillance decision (ISD). - Requiring systemic countries to undergo FSAPs. - More flexible IMF stance on capital controls and countercyclical fiscal policies. ### Context and drivers of perceptions - Many initiatives responded to the global financial and economic crisis; perceptions arose that the IMF had not been even-handed, particularly regarding advanced economies. - IEO assessment of surveillance in the run-up to the crisis identified shortcomings including insufficient attention to spillovers and contagion risks and a lack of candor and evenhandedness. - IEO noted a relatively biased belief in the ability of advanced economies and markets to withstand or mitigate a severe shock. - Interviews with mission chiefs suggested improvement in criticism of advanced economies; staff noted an improved image of the Fund in Africa. ### Examples of perceived lack of evenhandedness (reported in interviews) - Not sufficiently recognizing improvements in some economies and overly highlighting risks in others. - Not pressing fiscal consolidation in some large advanced economies. - Inconsistency in treatment of pension reform across developing countries. - Perception that countries involved in the Arab spring received “lighter” treatment. - Perceived bias in support for small economies relative to global importance. - FSAPs perceived as not equally tough and “extra critical” for some advanced economies. - Insufficient critical examination of spillovers from unorthodox monetary policy by advanced economies (e.g., quantitative easing seen as financial repression). - Some emerging markets felt the IMF was not sufficiently critical of some other emerging markets. - Multilateral surveillance products largely ignored developing countries; Fund research program did not sufficiently address concerns confronting developing countries (data constraints acknowledged). - Fund advice on similar issues differed across countries (examples: directed lending, use of subsidies, tax base widening). - Concerns over frequent change in mission chiefs for developing countries (example reported: three mission chiefs within 12 months). - Perception that advanced economies could make more changes and deletions in staff reports than other members; capacity constraints in developing countries to review reports noted. - Some Asian economies noted a lack of evenhandedness over exchange rate surveillance and an “Asian bias”; linked to governance concerns. ### Evidence review: scope and approach - Review of Article IV reports for 25 countries (subjective assessment); focus on outputs while emphasizing assessment should be on inputs. - Analysis included staff forecasting record (1990–2010) and numbers of staff working on surveillance across countries to detect bias. ### Coverage and consistency of surveillance - Differences in depth of analysis on issues (e.g., spillovers); background analysis for the TSR of 50 recent Article IV reports found coverage of inward spillovers increased since 2011, but only about half of reports assessed contained “extensive” analysis of inward spillovers. - No skew in coverage favoring one group of countries over another found. - Differences among Article IV reports for advanced economies and for emerging markets/developing economies exist and may reflect tailoring. - Some instances of insufficient rigor in surveillance in a given year suggest deficiencies in internal review processes; Board discussion and external commentary often lead to corrective action. - Multilateral surveillance products have considerably less coverage of developing country issues compared with major advanced and emerging markets. ### Reference to cross-country experiences - Use of cross-country experiences in policy advice is generally limited; background analysis found only about a third of recent Article IV reports included an in-depth analysis of policy experiences in other countries. - No apparent bias in referring to experiences of other countries. ### Forecasting performance (1990–2010) - Attachment 2 analyzes Fund forecasting of real GDP growth for 1990–2010. - No marked difference in forecasting performance across income groups overall, with complexities: - Forecasting optimism for program countries (Chart 2(i)). - Tendency to under-estimate growth in hydrocarbon energy-exporting countries (Chart 2(iii)). - For non-program countries by region: forecasts for AFR, EUR and WHD countries seem to be over-estimated while APD and MCD countries are underestimated (Chart 2(v)). - Overall, no evidence of a systematic bias in forecasting performance; consistent with IEO draft findings. - Methodological note: ‘forecast error’ = difference between Fund forecasts and actual growth performance over a three year forecast horizon; covers 184 countries using all vintages of the WEO between 1990 and 2010. ### Focus on income inequality - Concern over rising inequality is raised in some emerging market and developing country reports but receives little, if any, coverage in Article IV reports for advanced economies. - An Executive Director from a large advanced economy noted inequality was covered in a selected issues paper in a recent Article IV consultation but was not considered a priority given other challenges. - Given the Fund’s recent emphasis on income inequality and its impact on sustainability of growth, surveillance coverage in this area needs general improvement. - On current treatment of income inequality, there appears to be some lack of evenhandedness. ### Position of smaller economies - Depth and coverage of reports for smaller economies are generally less than for major advanced and emerging markets, though not uniformly. - Some smaller economies’ Article IV reports were quite extensive. - Explanations: data availability, risk-based allocation of resources (greater attention to larger economies because of larger spillovers), and higher word count limits for systemic countries. - Little coverage of smaller or low-income countries in multilateral surveillance products such as the World Economic Outlook. ### Program versus non-program countries - Combined Article IV and program reviews show limited coverage of Article IV issues, with most of the report focusing on the program. - Objectivity of Article IV may be impacted in program countries due to concerns that criticizing a program country’s policies could be seen as criticizing the Fund-supported program. - Example: considerations about how critical an Article IV could be for a country with a Precautionary and Liquidity Line program “without jeopardizing the insurance.” ### Number and experience of staff on country teams - Decisions on number and quality of staff working on surveillance are an “input” and should be free of bias. - More experienced staff work on more systemically important countries; strong staff demand to work on certain major emerging markets. - 2014 TSR survey of mission chiefs: - 58 percent of mission chiefs for advanced economies felt they had a less experienced team than in 2011. - 55 percent for emerging market mission chiefs felt they had a less experienced team than in 2011. - 34 percent for those leading teams to low income countries felt they had a less experienced team than in 2011. - Only eight percent of advanced economy mission chiefs recorded a decline in support from functional departments between 2011 and 2014. - Emerging and low income country mission chiefs recorded falls of 48 and 46 percent, respectively, in support from functional departments. - Results suggest a broad-based decline in staff skill levels over the past three years, possibly a consequence of significant downsizing of the Fund prior to the crisis. - Strong demand from staff to work on larger economies may be influenced by nature of work and perceived career impact. ### Evolution of staff time on surveillance (2010–2013) - Attachment 3 findings: - No significant change in staff hours spent on bilateral surveillance in recent years (Chart 3(i)). - Staff spent more time on surveillance of emerging markets compared with advanced and low-income countries, with some increase in hours on advanced countries since the crisis. - For multilateral surveillance, staff increased time spent on all countries, especially advanced economies (Chart 3(ii)). - Functional departments spent more time on surveillance of advanced economies since the crisis, with implications for time spent on other countries, particularly low-income countries (Chart 3(iii)). ### Structure and presentation of Article IV reports - References to authorities’ response to previous Fund advice vary significantly; background analysis found past policy advice was discussed much less frequently in reports for advanced countries. - Reporting of authorities’ views varies: some reports place authorities’ views in a separate section; in other cases, particularly major advanced economies, the report is drafted as an interchange between staff and authorities. - Historical calls to differentiate authorities’ and staff views; staff responses vary. - 2014 TSR survey of mission chiefs on presentation of authorities’ views: - 54 percent of mission chiefs to advanced economies said the IMF was either “not at all,” “to a limited extent,” or “to some extent” even-handed in presentation of the authorities’ views. - Corresponding figures for mission chiefs from emerging and developing economies and low income countries were 77 and 68 percent, respectively. - Finding suggests considerable room for improvement in reporting authorities’ views, particularly as perceived by mission chiefs for emerging and developing economies. ### Recording and reporting of authorities’ views (buff statement) - The buff statement issued by Executive Directors is an opportunity for authorities to indicate if their views were not appropriately reflected in the Article IV staff report. - One case identified where authorities’ views were recorded differently in the report compared with the Executive Directors’ buff statement reporting significant disagreement. - Mission chiefs’ observations on incorporating authorities’ views: - Broad agreement makes integration easier. - Significant differences require clear distinction and outlining. - Some staff place authorities’ views under a separate heading, sometimes saying little more than “the authorities broadly agree.” - Separating authorities’ views can make it easier to check accuracy with authorities. - Concern: Variation in approach can give impression some countries receive more favorable treatment. - Recommendation: More guidance and elaboration of good practices for recording authorities’ views to ensure decisions are robust and free from bias. - Observation: Integrating authorities’ and staff views in the body of reports gives impression of a dialogue rather than targeted IMF policy messages followed by an authorities’ response; if widely adopted for advanced economies, this may reinforce perceptions that reports for advanced economies are less critical. ### Tone and candor of reports - No evident persistent bias favoring advanced economies being less critical or using softer language across a cross-section of Article IV reports. - Variation in language across reports: authorities may be told they “must,” be “urged,” or that they “should” pursue policies. - Stronger language generally used where countries face deeper economic difficulties. - Survey contrast on tone: - 36 percent of advanced economy mission chiefs indicated they thought the tone of Article IV reports was even-handed to a great extent. - Emerging and low income mission chiefs reported 8 and 12 percent, respectively. ### Pressures affecting candor - IEO evaluation: “almost 60 percent” of mission chiefs surveyed who worked on advanced countries felt pressure to dilute candor to avoid upsetting country authorities. - This review’s discussions did not reflect widespread belief in self-censorship for fear of upsetting countries. - Main identified sources of pressure: - Direct involvement of senior management. - Extra attention and scrutiny for systemically important countries. - Multiple Fund staff making public comments on major economies complicates message consistency. - Large shareholders pressuring the Fund to approve programs where they have national interests. ### Bias in corrections and deletions - Perception that advanced economies could “get away with more” regarding changes or deletions to staff reports. - 2013 Transparency Policy review findings: - Rate of deletions in staff reports increased since 2009. - Deletion rates highest for advanced economies; advanced economies requested more deletions. - Difference in deletion rates between advanced and emerging markets diminishes when splitting data into program and non-program countries. - Detailed sample examination for deletions and corrections for “evident ambiguity” revealed only a few cases inconsistent with policy and little evidence of bias. - Recommendation: present an annual table to the Executive Board detailing all rejected modification requests. ### Forecast bias and projections (further detail) - Analysis of Fund forecasts of real GDP growth for 1990–2010 reveals general bias toward optimism overall. - Does not suggest a marked or sizeable difference in optimism—or “forecast error”—based on countries’ income levels. - Nuances: - Program countries tend to have more optimistic growth forecasts. - Under-estimation tendency for hydrocarbon energy-exporting countries. - For non-program countries by income levels: forecasts more optimistic for advanced countries on the whole. - Regional variation for non-program countries: optimism for AFR, EUR and WHD; underestimation for APD and MCD. - Comparison across regions for non-hydrocarbon energy-exporting countries shows APD and MCD countries having more pessimistic forecasts on average; growth forecasts for MCD appear biased downward in general. - Conclusion: many factors influence forecasting performance; overall there does not appear to be a systematic bias in the Fund’s forecasts. ### Staff time on surveillance (charts and indices) - Chart references and sources: - Chart 3 (i) Staff Hours on Bilateral Surveillance by Country Income Level, 2010-2013 (Index, Year 2010=100) — source: IMF ACES and TRACES. - Chart 3 (ii) Staff Hours on Multilateral Surveillance by Country Income Level, 2010-2013 (Index, Year 2010=100) — source: IMF ACES and TRACES. - Chart 3 (iii) Functional Departments' Total Contribution to Surveillance Work by Country Income Level, 2010-2013 (in staff years) — source: IMF ACES and TRACES. ### Overall conclusions - Little evidence that surveillance is systematically biased in favor of one group of Fund members, especially advanced economies; differences in approach reflect country circumstances. - Significant perceptions remain that the Fund is not even-handed; such perceptions are held by some staff and can impact behavior. - Evenhandedness is essential to IMF legitimacy and effectiveness; governance reform is frequently cited by developing countries and emerging markets as needed to ensure evenhandedness. - If governance reforms are delayed, the Fund must demonstrate that this does not impact evenhandedness and must take steps to address perceptions. ### Recommendations (A–D and follow-on actions) - Recommendation 1 (A) - Clarify what is meant by being even-handed in surveillance. - Provide more elaboration in the staff guidance note on surveillance. - Focus on ensuring the “inputs” to the surveillance process are even-handed: every decision taken during surveillance should be sound, analytically robust and free from any bias. - Evenhandedness does not mean a “one size fits all” approach; every country should receive the same sound, objective analytical inputs for surveillance decisions and policy advice. - Recommendation 2 (B) - Staff reports should set out the rationale for the approach and policy advice being offered. - Where appropriate, explain how surveillance is adapted to a country’s specific circumstances and why the approach may differ from that for other countries. - Differences in approach, where valid, should be recorded in the staff report; the review process should detect such differences. - Recommendation 3 (C) - All reviewing mechanisms—senior staff in area departments, reviewing departments, Management and the Executive Board—must be vigilant in pursuing evenhandedness and have their roles highlighted. - Parties should periodically review their contribution to ensuring evenhandedness. - The internal review process is critical for ensuring consistency and identifying relevant cross-country experiences. - More thematic reviews should be undertaken to promote evenhandedness (e.g., fiscal policy, pension reform) and examine treatment across multiple Article IV consultations. - The role and activities of the review process, including contributions by Management and the Executive Board, should be publicly emphasized. - Recommendation 4 (D) - Concerns over a lack of evenhandedness should be raised and addressed in a transparent and well-substantiated manner. - Specific concerns should be fully discussed and appropriate action taken, which may include adjustments to subsequent Article IVs. - Staff should raise concerns with senior offices; Executive Directors and authorities should raise specific issues with departmental directors or Management. - Staff should prepare a periodic report to the Board outlining steps being taken to ensure surveillance is even-handed, including specific concerns raised and actions taken. - The first periodic report should outline the respective roles of Management, staff and the Executive Board in ensuring evenhandedness in surveillance. *Source: 2014 TSR—REVIEW OF EVENHANDEDNESS (excerpts from the supplied content unit).* ### EXECUTIVE SUMMARY ### _073014h - EXECUTIVE SUMMARY ### Importance of evenhandedness - Evenhandedness is described as a “cornerstone” for a cooperative institution like the Fund and directly impacts legitimacy and the effectiveness of surveillance. - The IMFC’s October 2, 2004 communiqué stated that, “Effective and even-handed IMF surveillance across the whole membership is central to promoting high and sustainable growth in member countries and to crisis prevention.” - The Managing Director (December 2011) is cited: credibility and effectiveness of the IMF depends “on our honesty, on the quality of our work and on evenhandedness.” - Perceptions that the Fund is not even-handed can influence behavior and reduce receptiveness to Fund advice. ### Conceptual framework: inputs vs outcomes - Actual evenhandedness should be assessed on the “inputs” to surveillance (the way surveillance is conducted), not only on “outcomes” (policy assessments and advice in reports). - Inputs include decisions over resources deployed, issues to be covered, depth of analysis, policy advice provided, and presentation of advice. These decisions must be free from bias and based on sound, robust and objective considerations. - Outcomes will differ when surveillance is appropriately tailored to country circumstances; tailoring does not imply lack of evenhandedness if inputs are unbiased. ### Findings on perceptions and evidence - There are significant, often long-standing, perceptions that the Fund is not even-handed. Many perceptions do not relate directly to surveillance but can affect the effectiveness of Fund advice. - The review identifies instances of differences in surveillance across countries, but differences alone are not definitive evidence of lack of evenhandedness. - Differences identified in this review can be grouped into three categories: - Differences in the approach—or “inputs”—to surveillance that reflect appropriate tailoring to country circumstances. - Occasional differences that are not well-justified. While there is not a systematic difference in treatment across countries or country groups, these inconsistencies are problematic, can contribute to perceptions of a lack of evenhandedness and should be addressed. - Pockets of systematic differences in treatment across country groups that represent a lack of evenhandedness, for example: - less rigorous surveillance for program countries, - minimal reporting on the response to past Fund advice for advanced economies, - a steep reduction in resources from functional departments for surveillance for emerging markets and low-income countries, - the main multilateral surveillance products paying little attention to developing countries. - Note: the latter two aspects could be explained by a risk-based approach to surveillance. - The review did not find evidence of a pervasive lack of evenhandedness in surveillance. However, even small instances matter given entrenched perceptions. ### Perceptions by stakeholder groups - Advanced countries are more inclined to consider Fund surveillance even-handed, though most can cite instances where the Fund was not sufficiently critical. - Mission chiefs generally felt unconstrained in expressing their views, with some acknowledging greater pressure on major advanced or emerging market cases. - Authorities from emerging markets and developing countries expressed more concern over a lack of evenhandedness, often linking it to governance and to the need to reform the Fund’s governance arrangements. ### Recommended measures to strengthen evenhandedness - The review recommends that the Fund take extra measures to strengthen and to demonstrate strengthened evenhandedness. - Recommendation 1: - Clarify what is meant by being even-handed, including providing more explanation in the staff guidance note on surveillance. - Recommendation 2: - Provide more explanation in Article IV reports on how surveillance has been tailored to country circumstances and why the approach and advice differs from that offered to other countries that appear to be facing similar circumstances. - Encourage greater use of cross-country comparisons (currently limited). Facilitate this by periodically selecting policy topics—such as fiscal policy or pension reform—and examining the handling of the issue in the Article IV consultations for a number of countries in thematic papers. - Recommendation 3: - All reviewing mechanisms within the Fund—senior staff in area departments, reviewing departments, Management and the Executive Board—have a significant role in ensuring evenhandedness. They must be vigilant in this regard and their role should be highlighted. - Recommendation 4: - Concerns over a lack of evenhandedness—be it by country authorities, Executive Directors, staff or other stakeholders—should be raised and addressed in a transparent and well-substantiated manner. - Staff should prepare a periodic report on evenhandedness for the Executive Board. Further to Recommendation 3, the first periodic report should outline the respective roles of Management, staff and the Executive Board in helping to ensure evenhandedness in surveillance. *EXECUTIVE SUMMARY, 2014 TSR—REVIEW OF EVENHANDEDNESS (July 30, 2014).* ### 19. Some positive recent developments in terms of promoting greater evenhandedness include: ### _073014h - 19. Some positive recent developments in terms of promoting greater evenhandedness include: ### Positive recent developments - Increased focus on spillovers. - Adopting the integrated surveillance decision (ISD). - Requiring systemic countries to undergo FSAPs. - A more flexible attitude of the IMF on issues such as capital controls and counter cyclical fiscal policies. ### Context and drivers of perceptions - Many initiatives responded to the global financial and economic crisis, which may have influenced perceptions that the IMF had not been even handed with surveillance, particularly with respect to advanced economies. - The IEO’s assessment of the Fund’s surveillance in the run-up to the crisis identified shortcomings including insufficient attention to spillovers and contagion risks and a lack of candor and evenhandedness.5 - The IEO noted a relatively biased belief in the ability of advanced economies and markets to withstand or mitigate a severe shock. - Interviews with mission chiefs suggested the Fund was improving in being more critical of advanced economies, citing the Managing Director’s forceful remarks on slow progress in addressing banking weakness in Europe and the political fight on raising the US debt ceiling.6 - Staff noted an improved image of the Fund in Africa, reflecting more focus on the region and improved regional performance. ### Examples of perceived lack of evenhandedness (reported in interviews) - The Fund had not sufficiently recognized improvements in some economies’ performance and excessively highlighted developments in some countries as risks to others. - The need for fiscal consolidation in some large advanced economies was not sufficiently pressed by the Fund. - Inconsistency in treatment of pension reform across a number of developing countries; similar reforms were assessed differently across countries. - Perception that countries involved in the Arab spring received “lighter” treatment compared with other developing countries. - Bias in support for small economies, with excessive focus on initiatives for small states relative to their limited importance to the global economy. - FSAPs were not equally tough across the membership and were “extra critical” with respect to some advanced economies. - The Fund was insufficiently critical in examining spillovers from unorthodox monetary policy by advanced economies; some viewed quantitative easing pursued by some advanced economies as a form of financial repression that would be heavily criticized if undertaken by an emerging market or developing country. - Some emerging markets felt the IMF was not sufficiently critical of some other emerging markets. - Multilateral surveillance products largely ignored developing countries; Fund research program did not sufficiently address concerns confronting developing countries (data constraints acknowledged). - Fund advice on similar issues differed across countries (examples: directed lending, use of subsidies, tax base widening). - Concerns over frequent change in mission chiefs for developing countries (example: three mission chiefs within 12 months); perception that advanced economies did not confront the same problem. - Perception that advanced economies could “get away” with more changes and deletions in staff reports than other Fund members; capacity constraints in developing countries to review reports noted. - Some Asian economies noted a lack of evenhandedness over exchange rate surveillance and an “Asian bias”; lingering lack of trust in Asia linked to Fund governance concerns. ### Evidence review: scope and approach - Article IV reports of 25 countries were reviewed (subjective assessment). The review focused on outputs but emphasized that evenhandedness should be assessed on inputs to surveillance (decisions made during the surveillance process). - Staff forecasting record (1990–2010) and numbers of staff working on surveillance across countries were analyzed to detect bias. ### Coverage and consistency of surveillance - Differences in depth of analysis on issues (e.g., spillovers); staff background analysis for the TSR of 50 recent Article IV reports found coverage of inward spillovers increased since 2011, but only about half of reports assessed contained “extensive” analysis of inward spillovers. - Review found no skew in coverage favoring one group of countries over another. - Differences exist among Article IV reports for advanced economies and for emerging markets/developing economies; differences expected when surveillance is targeted to country circumstances. - Some instances where surveillance was not rigorous enough in one year, suggesting deficiencies in internal review processes; Board discussion and external commentary often lead to corrective action. - Multilateral surveillance products have considerably less coverage of developing country issues compared with major advanced and emerging markets, consistent with focus on countries with largest impact on global developments. ### Reference to cross-country experiences - Use of cross-country experiences in policy advice varies and is generally limited; no apparent bias in referring to experiences of other countries. - Background analysis for the TSR found only about a third of recent Article IV reports included an in-depth analysis of policy experiences in other countries. ### Forecasting performance - Attachment 2 (analysis) covers Fund’s forecasting of real GDP growth for 1990–2010. - No marked difference in forecasting performance across income groups overall, though complexities exist: - Forecasting optimism for program countries (Chart 2(i)). - Tendency to under-estimate growth in hydrocarbon energy-exporting countries (Chart 2(iii)). - For non-program countries by region: forecasts for AFR, EUR and WHD countries seem to be over-estimated while APD and MCD countries are underestimated (Chart 2(v)). - Factors influencing results include rapid growth in some regions, volatility, and data limitations. - Overall, no evidence of a systematic bias in forecasting performance; consistent with IEO draft findings. ### Focus on income inequality - Reference to income inequality and inclusive growth differs between advanced economies and emerging markets: - Concern over rising inequality is raised in some emerging market and developing country reports but receives little, if any, coverage in Article IV reports for advanced economies. - An Executive Director from a large advanced economy noted inequality was covered in a selected issues paper in a recent Article IV consultation but was not considered a priority given other challenges. - Given the Fund’s recent emphasis on income inequality and its impact on sustainability of growth, this area needs general improvement in surveillance. - On current treatment of income inequality, there appears to be some lack of evenhandedness.7 ### Position of smaller economies - Depth and coverage of reports for smaller economies are generally less than for major advanced and emerging markets, though not uniformly: - Some smaller economies’ Article IV reports were quite extensive. - As a rule, staff reports for large advanced economies and major emerging markets are normally longer than for smaller economies, particularly low-income countries. - Explanations include data availability, risk-based allocation of resources (greater attention to larger economies because of larger spillovers), and higher word count limits for systemic countries. - Little coverage of smaller or low-income countries in multilateral surveillance products such as the World Economic Outlook. ### Program versus non-program countries - Combined Article IV and program reviews show limited coverage of Article IV issues, with most of report focusing on the program. - Objectivity of Article IV may be impacted in program countries, reflecting concern that criticizing a program country’s policies could be seen as criticizing the Fund-supported program. - One mission chief noted considerations when undertaking an Article IV for a country with a Precautionary and Liquidity Line program about how critical the Article IV could be “without jeopardizing the insurance.” ### Number and experience of staff on country teams - Decisions on number and quality of staff working on surveillance are an “input” and should be free of bias. - More experienced staff work on more systemically important countries; strong staff demand to work on certain major emerging markets. - 2014 TSR survey of mission chiefs: - 58 percent of mission chiefs for advanced economies felt they had a less experienced team than in 2011. - 55 percent for emerging market mission chiefs felt they had a less experienced team than in 2011. - 34 percent for those leading teams to low income countries felt they had a less experienced team than in 2011. - Only eight percent of advanced economy mission chiefs recorded a decline in support from functional departments between 2011 and 2014. - Emerging and low income country mission chiefs recorded falls of 48 and 46 percent, respectively, in support from functional departments. - Results suggest a broad-based decline in staff skill levels over the past three years, possibly a consequence of significant downsizing of the Fund prior to the crisis. - Strong demand from staff to work on larger economies may be influenced by nature of work and perceived career impact. ### Evolution of staff time on surveillance - Attachment 3 shows staff time on bilateral and multilateral surveillance: - No significant change in staff hours spent on bilateral surveillance in recent years (Chart 3(i)). - Staff spent more time on surveillance of emerging markets compared with advanced and low-income countries, with some increase in hours on advanced countries since the crisis. - For multilateral surveillance, staff increased time spent on all countries, especially advanced economies (Chart 3(ii)). - Functional departments spent more time on surveillance of advanced economies since the crisis, with implications for time spent on other countries, particularly low-income countries (Chart 3(iii)). ### Structure and presentation of Article IV reports - References to authorities’ response to previous Fund advice vary significantly; when included it is usually for developing countries and less often for advanced economies. Background analysis found past policy advice was discussed much less frequently in reports for advanced countries. - Reporting of authorities’ views varies: some reports place authorities’ views in a separate section; in other cases, particularly major advanced economies, the report is drafted as an interchange between staff and authorities. - Historically, mixing of authorities’ and staff views prompted calls to differentiate them clearly; staff responses vary. - 2014 TSR survey of mission chiefs on presentation of authorities’ views: - 54 percent of mission chiefs to advanced economies said the IMF was either “not at all,” “to a limited extent,” or “to some extent” even-handed in presentation of the authorities’ views. - Corresponding figures for mission chiefs from emerging and developing economies and low income countries were 77 and 68 percent, respectively. - Result suggests considerable room for improvement in reporting the views of authorities, particularly as perceived by mission chiefs for emerging and developing economies. *2014 TSR—REVIEW OF EVENHANDEDNESS, INTERNATIONAL MONETARY FUND* ### 38. The buff statement issued by Executive Directors is an opportunity for country authorities to ### _073014h - 38. The buff statement issued by Executive Directors is an opportunity for country authorities to ### Recording and reporting of authorities’ views - The buff statement issued by Executive Directors is an opportunity for country authorities to indicate if they consider that their views have not been appropriately reflected in the Article IV staff report. - One case was identified where the authorities’ views had been recorded differently in the report compared with the Executive Directors’ buff statement, which reported that the authorities significantly disagreed with staff on some policy issues. - Mission chiefs’ observations on incorporating authorities’ views: - If there is broad agreement, it is easier to integrate the authorities’ views into the discussion of staff’s views. - If there is a significant difference of views, it is important to clearly distinguish and outline the differences. - Some staff place the authorities’ views under a separate heading, sometimes saying little more than “the authorities broadly agree.” - Separating out authorities’ views can make it easier to check with the authorities whether their views had been accurately recorded. - Concern: Variation in approach can give the impression that some countries receive more favorable treatment than others. - Guidance need: More guidance and elaboration of good practices are recommended for recording authorities’ views to ensure decisions are robust and free from bias. - Observation: Integrating authorities’ and staff views into the body of the report gives the impression of a dialogue rather than targeted IMF policy messages followed by an authorities’ response; if widely adopted for advanced economies, this may inadvertently reinforce perceptions that reports for advanced economies are less critical. ### Tone and candor of reports - Review finding: No evident persistent bias in favor of advanced economies being less critical or using softer language across a cross-section of Article IV reports. - Variation in language across reports: - Some reports state authorities “must” take certain action. - In other cases authorities are “urged” to pursue a policy path. - In other reports staff say authorities “should” pursue certain reforms. - Determinants of language strength: - More forceful language generally used where countries face deeper economic difficulties. - Larger economies with systemic implications may justify greater surveillance focus and forceful recommendations, although this was not always observed. - General tone: Reports generally provided positive comments acknowledging positive policy measures by the authorities. - Survey contrast: - 36 percent of advanced economy mission chiefs indicated they thought the tone of Article IV reports was even-handed to a great extent. - Emerging and low income mission chiefs reported 8 and 12 percent, respectively. ### Pressures affecting candor - IEO evaluation reported: almost 60 percent of mission chiefs surveyed who worked on advanced countries felt pressure to dilute the candor of staff reports to avoid upsetting country authorities. - This review’s discussions with mission chiefs did not reflect widespread belief in self-censorship for fear of upsetting countries. - Main sources of pressure identified: - Direct involvement of senior management. - Extra attention and scrutiny associated with surveillance of systemically important countries. - Maintaining consistency of message challenged by multiple Fund staff making public comments on major economies. - Cases where large shareholders put pressure on the Fund to approve a program for another country where they have national interests. - One mission chief reported pressure from within the Fund and from some other country authorities to take a more critical approach, but resisted when inconsistent with his analysis. - Interpretation: The IEO finding of 60 percent may reflect staff difficulty balancing frank advice that receives a hearing versus advice perceived as excessively critical and immediately rejected. ### Bias in corrections and deletions - Perception: Advanced economies could “get away with more” regarding changes or deletions to staff reports. - 2013 Transparency Policy review findings: - Rate of deletions in staff reports had increased since 2009. - Deletion rates were highest for advanced economies; however advanced economies requested more deletions. - Difference in deletion rates between advanced and emerging markets diminishes when splitting data into program and non-program countries. - Detailed examination of a sample for deletions and corrections for “evident ambiguity” revealed only a few cases inconsistent with the policy and little evidence of bias. - Recommendation from 2013 review: present an annual table to the Executive Board detailing all rejected modification requests. ### Forecast bias and projections - Analysis of Fund forecasts of real GDP growth for 1990–2010: - Reveals a general bias toward optimism overall. - Does not suggest a marked or sizeable difference in the degree of optimism—or the “forecast error”—based on countries’ income levels. - Nuances (“unpacked” data): - Program countries tend to have more optimistic growth forecasts. - Tendency to underestimate real GDP growth in hydrocarbon energy exporting countries. - For non-program countries by income levels, forecasts on the whole are more optimistic for advanced countries. - Regional variation for non-program countries: optimism for AFR, EUR and WHD and underestimated growth for APD and MCD; interpretation is difficult. - Comparison across regions for non-hydrocarbon energy-exporting countries shows APD and MCD countries having more pessimistic forecasts on average; growth forecasts for MCD appear biased downward in general. - Conclusion: Many factors influence forecasting performance; overall there does not appear to be a systematic bias in the Fund’s forecasts. Regional or program-status biases may exist, but differences may reflect forecasting difficulties in certain circumstances (e.g., pre- and post-crisis volatility, commodity-related volatility, less predictable growth patterns in early stages of development, or poor data quality). - Methodological note: The ‘forecast error’ was calculated as the difference between Fund forecasts and actual growth performance over a three year forecast horizon. It covers 184 countries, drawing on all vintages of the World Economic Outlook between 1990 and 2010. ### Staff time on surveillance (2010–2013) - Chart indices and sources reported: - Chart 3 (i) Staff Hours on Bilateral Surveillance by Country Income Level, 2010-2013 (Index, Year 2010=100) — source: IMF ACES and TRACES. - Chart 3 (ii) Staff Hours on Multilateral Surveillance by Country Income Level, 2010-2013 (Index, Year 2010=100) — source: IMF ACES and TRACES. - Chart 3 (iii) Functional Departments' Total Contribution to Surveillance Work by Country Income Level, 2010-2013 (in staff years) — source: IMF ACES and TRACES. ### Overall conclusions - Little evidence that surveillance is systematically biased in favor of one group of Fund members, especially advanced economies; differences in approach reflect country circumstances. - Significant perceptions remain that the Fund is not even-handed; such perceptions are held by some staff and can impact behavior. - Evenhandedness is essential to the legitimacy and effectiveness of the IMF; governance reform is frequently cited by developing countries and emerging markets as needed to ensure evenhandedness. - If governance reforms are delayed, the Fund must demonstrate that this does not impact evenhandedness and must take steps to address perceptions. ### Recommendations (A–D and follow-on actions) - Recommendation 1 (A) - Clarify what is meant by being even-handed in surveillance. - Provide more elaboration in the staff guidance note on surveillance. - Focus on ensuring the “inputs” to the surveillance process are even-handed: every decision taken during surveillance should be sound, analytically robust and free from any bias. - Evenhandedness does not mean a “one size fits all” approach; the core principle should be that every country receives the same sound, objective analytical inputs for surveillance decisions and policy advice. - Recommendation 2 (B) - Staff reports should set out the rationale for the approach and policy advice being offered. - Where appropriate, explain how surveillance is adapted to a country’s specific circumstances and why the approach may differ from that for other countries facing similar circumstances. - Differences in approach, where valid, should be recorded in the staff report; the review process should detect such differences. - Recommendation 3 (C) - All reviewing mechanisms—senior staff in area departments, reviewing departments, Management and the Executive Board—must be vigilant in pursuing evenhandedness and have their roles highlighted. - Parties should periodically review their contribution to ensuring evenhandedness. - The internal review process is a critical element for ensuring consistency and for identifying relevant cross-country experiences. - More thematic reviews should be undertaken to promote evenhandedness (e.g., fiscal policy, pension reform) and examine treatment across multiple Article IV consultations. - The role and activities of the review process, including contributions by Management and the Executive Board, should be publicly emphasized. - Recommendation 4 (D) - Concerns over a lack of evenhandedness should be raised and addressed in a transparent and well-substantiated manner. - Specific concerns should be fully discussed and appropriate action taken, which may include adjustments to subsequent Article IVs. - Staff should raise concerns with senior offices; Executive Directors and authorities should raise specific issues with departmental directors or Management. - Staff should prepare a periodic report to the Board outlining steps being taken to ensure surveillance is even-handed, including specific concerns raised and actions taken. - The first periodic report should outline the respective roles of Management, staff and the Executive Board in ensuring evenhandedness in surveillance. *Source: 2014 TSR—REVIEW OF EVENHANDEDNESS (excerpts from the supplied content unit).* --- _Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2014/_073014h.pdf_