## EXECUTIVE SUMMARY ## Source details **Canonical URL:** [EXECUTIVE SUMMARY](https://www.imf.org/-/media/files/publications/pp/2024/english/ppea2024058.pdf) ## Other formats - [Markdown version](/-/media/files/publications/pp/2024/english/ppea2024058.pdf.md) - [Structured JSON version](/-/media/files/publications/pp/2024/english/ppea2024058.pdf.json) --- ### Publication Trends - Since the 2013 Transparency Policy review, the publication rates of Board documents, including both country and policy documents, have improved and have reached about 96–98 percent in recent years. - Nearly all Board documents are published. - Country papers not published are generally concentrated in a few countries and their non-publication tends to reflect country-specific factors. - The timeliness of publication of country papers has improved, although an important share of these papers is published with long delays. - Since the last review, country papers were on average published 16 days after Board consideration. - About a third of these papers were published more than 14 days after Board consideration. - A considerable share is still published with delays exceeding four weeks. - The average publication lags for policy papers have lately increased, reflecting delays in publishing specific papers with long processing timelines. - Press releases for program cases are published immediately after Board consideration, but press releases for Article IV consultations are published with considerable delay. - Press releases for Article IV consultations are on average issued 10 days after Board consideration. ### Modifications of Board Documents - Most country papers are published with some corrections (about 75 percent on average). - The share of country papers published with deletions has declined since the last policy review, reflecting reduced concerns about financial sector issues after the 2008 global financial crisis. - Most correction requests and a large share of deletion requests are approved. - There is no evident evenhandedness issue in the application of the correction and deletions rules of the Transparency Policy. - Statistical evidence of acceptance and rejection rates of modification requests and a detailed assessment of a large sample of modification requests show no systematic bias in the approval of deletion and correction requests across countries, income level groups, regions, and economic size. ### Publication Rates: Trends and Analysis - Overall publication performance - Since the 2013 review, publication rates for Board documents have edged higher, and nearly all Board documents are now published. - During 2014–22, the share of Board documents covered by the policy that are published increased from an average of 89 percent during 2009–12 to 97 percent in 2022 with an average of 95 percent during 2014–22. - Improvement reflects increases in publication rates of both country and policy papers. - Country papers - During 2014–22, about 94–98 percent of country papers were published compared to 91–94 percent in 2009–12. - Article IV (AIV) reports publication improved from an average of 89 percent in 2009–12 to 94 percent in 2014–22. - Publication of country papers for Emerging Markets (EM) and Low-Income Countries (LIC) increased to 93–98 percent from 89–93 percent, approaching the 100 percent publication rate of country papers of Advanced Economies (AE). - Use-of-Fund-Resources (UFR) and related reports - Average publication rate was 96 percent over 2009–12 and edged up to 99 percent during 2014–22. - Other country documents - Financial System Stability Assessment (FSSA) reports improved to an average of 86 percent during 2014–22 (from 73 percent over 2009–12) but remains volatile. - Selected Issues Papers (SIPs) publication rate remained broadly constant around 88–91 percent, on average below the publication rate of associated Article IV reports. - Policy papers and multi-country documents - Over 2014–22, the average publication rate of policy papers was 97 percent (up from 90 percent during 2009–12). - 23 percent of policy papers presented to the Executive Board during 2014–22 were not covered by the presumption of publication regime. - Since 2013, 35 multi-country documents were issued to the Executive Board, 30 were presumed to be published, and among those presumed to be published, only one was not published. ### Can Higher Publication Rates Be Achieved? - Unpublished country papers, 2014–22 - Since 2014, 65 country papers (out of 1,628 papers) were not published—about 7 country papers per year. - Half of these unpublished papers were for EMs, and the rest were for LICs. - Countries in LAC had the largest share (29 percent) of unpublished country papers. - Role of modification requests in non-publication - Since 2014, 78 percent of unpublished country papers had either no modification requests (52 percent) or had all modification requests approved (26 percent). - The remaining 22 percent (14 country papers) that were not published had some rejected modifications, and only 5 percent (3 papers) had rejected deletions. - This suggests neither market sensitivities nor early policy disclosures have been major driving factors preventing publication. - Concentration of non-publication cases - Over 2014–22, 35 countries had at least one unpublished country report. - 11 countries had two or more unpublished papers, accounting for 63 percent of all unpublished reports. - Five countries (Antigua and Barbuda, Bahrain, Brunei Darussalam, Oman, and Turkmenistan) accounted for about 42 percent of all cases of unpublished country papers. - Increasing the publication rate further would require addressing idiosyncratic factors preventing publication in a few specific country cases. ### Translation and Languages Other Than English - Translation rates - About 31 percent of press releases (PRs) of country papers were translated into languages other than English during 2019–22 (PR data available only for this period). - Only about 14 percent of all country-related documents were published in languages other than English during 2019–22, with no major improvement compared to 2010–12 (12 percent). - For policy documents, the translation rate over 2014–22 was 1.4 percent. - Note: Translation is outside the scope of the Transparency Policy, but there is scope to increase translations given English is the “de-jure” common/official language in only about one third of IMF member countries. ### Publication Lags — General Timeliness - The publication timeliness of country papers has improved since the last policy review. - A majority of country and policy papers are now published promptly (i.e., within 14 days). - However, a number of country papers are still published with long delays (i.e., 28 days or more), mainly due to delay in obtaining publication consent. - Board documents - The average publication lag of country papers declined during 2014–22 to 16 days, from 33 days in 2009–12. - Since 2013, average publication lag of FSSAs and SIPs declined to 25 and 20 days, respectively (compared to 52 and 25 days during 2009–12). - The 2013 Review redefined “prompt publication” as publication within 14 days from Board consideration or Lapse-of-Time (LOT) decision date. - During 2014–22, 67 percent of country papers were published promptly (within 14 days), compared to 45 percent at the time of the last review. - 45 percent of country papers were published within 7 days, compared to 23 percent in 2009–12. - Improvements were driven by a substantial increase in prompt publication for EMs and LICs (62 percent during 2014–22 versus 35 percent in 2009–12). - Prompt publication for AE increased from 88 percent to 93 percent. ### Publication lags of country papers — overall patterns - Publication lags for EMs and LICs remain more than double than that for AEs. - For EMs, the average publication time for country papers ranged from 8 days for countries in Central and Eastern Europe to nearly a month for countries in Middle East and North Africa. - For LICs, the average lag ranged from 15 days for countries in Asia to 30 days for LAC. - During 2014–22, on average about 33 percent of country papers had publication lags that exceeded 14 days. - During 2014–22, about 12 percent of country papers were published more than 28 days after Board consideration (referred to as “long-lag papers”). - About 57 percent of these long-delayed country papers were for EM economies, followed by LICs (41 percent) and AEs (2 percent). - Some papers were published more than 90 days after Board consideration and followed a lower-profile publication procedure. ### Distribution, multi-country papers, and differences over time - Average publication lag for multi-country papers during 2014–22 was 21 days, compared to 16 days for single country papers. - About 60 percent of multi-country papers were published promptly (i.e., within 14 days), compared to about 67 percent for single country papers. - The longer publication time for multi-country documents likely reflects a more complex publication consent process, potentially requiring publication consent from all relevant countries. ### Policy papers publication timing - Between 2014–22, the median publication time for policy papers was 11 days, compared to 8 days during the previous review period (2009–12). - The share of policy papers published promptly (i.e., within 14 days) declined to about 61 percent (from 68 percent in 2009–12). - The average lag of the policy papers published in more than 28 days after Board consideration reached 88 days during 2014–22, compared to 80 days in 2009–12. ### Press releases (PRs) — timing and patterns - During 2019–22, PRs for country papers were, on average, published 6 days after the Board meeting. - About 22 percent of PRs took more than 7 days to be issued, while 43 percent of PRs for country papers were issued on the same day of the Board meeting. - Excluding 2020 data, the share of country-paper PRs issued on the same date of the corresponding Board meeting would be 35 percent. - For policy papers, the average publication timing in 2019–22 was 26 days as PRs for these papers are usually issued together with the corresponding policy paper. - During 2019–22, 91 percent of PRs for UFR cases were issued on the same day of the relevant Board discussion. - For Article IV consultations, PRs were on average published 10 days after Board consideration; 20 percent of these PRs were published more than two weeks after Board consideration. - Only 2 percent of Article IV consultation PRs were issued on the same day of the relevant Board consideration; about 29 percent were issued within 3 days after Board consideration. - During 2019–22, about 71 percent of AIV PRs were issued together with the corresponding staff report, and about 47 percent of them were issued more than seven days after Board consideration. - For policy documents, during 2019–22, 92 percent of PRs were issued at the time of the publication of the corresponding policy document; on average PRs for policy documents were issued about 26 days after the relevant Board meeting with 62 percent of PRs issued within 14 days. ### What drives publication lags of country documents and PRs? - CSO surveys noted long delays in the publication of some country reports and requested more information about the Executive Board’s activities. - Long publication delays for country papers are associated with idiosyncratic factors prevailing in some countries and delays in obtaining publication consent from members. - During 2014–22, country papers published more than 28 days after Board consideration covered 91 countries; these were mostly EMs and LICs. - Seven countries alone accounted for about 1/4 of all long-lag country papers. - The practice of seeking explicit publication consent (rather than relying on non-objection) and the lack of a specific timeline to obtain consent can result in long publication delays. - During 2014–22, 16 countries that have opted out of non-objection regime accounted for 28 percent of all country papers published more than 28 days after Board consideration. ### Role of modification requests in delays - During 2018–22, about 17 percent of modifications to country papers are estimated to have been made after Board consideration. - Among the long-lag papers, only 11 percent had the modification requests settled more than 28 days after Board consideration. - For most long-lag papers, final modification requests were submitted after the 28-day mark; thus the post-Board modification process under the TP was not the driver of the publication timing for most cases. - About 90 percent of country papers not published promptly (after 14 days but less than 28 days) had no modification requests or no pending requests at the 14-day mark, suggesting delay in obtaining publication consent was the likely cause of non-prompt publication. - For most papers with modification requests settled after the 14-day mark, the final requests themselves were submitted after the 14-day mark, again indicating the modification process was not a direct cause of non-prompt publication. - Delayed publication involved a limited number of small developing countries and other LICs and were generally one-off events associated with specific reports during 2018–22. ### Press releases drivers — Article IV PRs and staff report linkage - Long publication time of PRs for Article IV consultations reflects current publication rules and the expectation that PRs are published with the staff report. - For UFR cases PRs are expected to be issued immediately after Board consideration. - For Article IV consultations, regional surveillance discussions, and stand-alone FSSA reports, PRs are expected to be issued along with the corresponding staff report if the latter is expected to be published within seven days after Board consideration; otherwise, a PR is expected to be published shortly after the Board consideration. - Evidence suggests PRs for Article IV consultations tend to follow the publication timing of staff reports even when staff report publication is delayed. - During 2019–22, about 71 percent of AIV PRs were issued together with the corresponding staff report, and about 47 percent of them were issued more than seven days after Board consideration. - The expectation of issuing PRs with the staff report and the tendency to do so even when staff report publication is delayed are main reasons for significant PR lags for Article IV consultations. ### Modifications — deletions and corrections trends - The share of country papers published with some deletions has sharply declined since the last review. - In 2014–22, on average about 12 percent of published country papers had some deletions compared to around 21 percent before the 2013 review. - The decline reflects a return to deletion levels prevailing before the Global Financial Crisis. - The decline in deletions was largely driven by a drop in the share of papers with deletions for financial sector issues, which declined to 5 percent during 2014–22, from 12 percent during 2009–12. - Deletions for exchange rate issues remained broadly constant. - The decline in deletions was associated with a reduction in approval rate of the deletion requests since 2014. - A large share of country papers continues being subject to corrections and most deletion and correction requests are approved. - An analysis of approval/rejection rates of deletion and correction requests and an ex-post assessment of a large sample of such requests indicate no evident bias in the application of the modification rules to country documents. ### Deletion trends by country group - During 2014–22, the share of EM country papers with deletions was 17 percent, compared to 12 percent for AE papers. - This represents a shift from the period before the previous review when AEs recorded higher shares (35 percent for AEs, compared to 24 percent for EMs). - During 2014–22, deletion requests averaged: - 0.6 deletions per paper for EMs - 0.2 deletions per paper for AEs - 0.3 deletions per paper for LICs - Overall, in 2014–22, about 51 percent of all deletion requests were for country papers for EMs. - For policy papers over the same period, deletion requests were low: on average only one in eight policy papers had some deletion requests. ### Approval rates of deletion requests - During 2014–22, on average about 64 percent of deletion requests to country papers were approved. - The median approval rate across papers for 2014-19 and 22 was 83 percent. - Approval rates for country papers (and all country documents) have remained stable over time, ranging from 59 to 76 percent once outlier cases are accounted for. - Approved deletions by stated reason (percent, excl. 2020 and 21): - Highly Market Sensitive Material, 76 - Premature Disclosure of Policy Intentions, 23 - Removal of References to Unpublished Documents, 1 - About three-quarters of the approved deletions were for market sensitivity concerns and the remaining quarter for premature disclosure of authorities’ policy intentions. ### Corrections: incidence and types - During 2014–22, about 75 percent of published country papers had corrections, compared to about 65 percent in 2009–12. - Excluding typographical errors, correction requests for 2018–22 were: - 52 percent to rectify factual errors - 41 percent for evident ambiguities - 7 percent amending the characterization of authorities’ views - Correction requests were concentrated: about 7 percent of countries that requested corrections during 2014–22 had more than 150 requests, which accounted for 28 percent of total correction requests. - For 2014–22, average number of correction requests per country paper by income group: - AE country papers: 18 - EM country papers: 9 - LICs: 3 - Article IV reports accounted for 77 percent of correction requests for country papers during 2014–22; on average Article IV reports had about 9 correction requests per report, compared to about 5 for Combined Article IV and UFR reports and about 3 for UFR reports. - About 17 percent of correction requests (about 18 percent of all modification requests including deletions) were received after Board consideration during 2018–22; the share was slightly higher for EM and LIC than for AE. ### Approval rates of correction requests - During 2014–22, the average approval rate of correction requests for country papers was 87 percent (ranging 77–93 percent across years). - Approval rates of correction requests varied little across types of requests and document types. - Country-level distribution of approval rates for correction requests, 2014–22: - 74 percent of countries had approval rate of correction requests of 80 percent or above - 91 percent had approval rates of 70 percent and above - Only eight countries had approval rates lower than 65 percent (two standard deviations below the mean); these cases were driven by small sample bias or single report weight rather than systemic bias ### Evenhandedness and cross-group analysis - Stakeholder surveys indicate perceived uneven application of the policy has abated: about 37 percent of responding EDs perceive uneven application, down from 50 percent in 2013. - Statistical evidence does not signal systemic issues with evenhanded application of modification rules: - No evidence of systematic biases in approval of correction requests across countries. - No evidence of systemic bias in approval rates of modification requests across country income levels: - Average approval rate of deletion requests, 2014–22, by income group: - AE: 65.1 - EM: 64.6 - LIC: 61.9 - Average approval rates for correction requests, 2014–22, by income group: - AE: 88.5 - EM: 86.2 - LIC: 87.8 - No evidence that largest economies (G-7 or G-20) benefit in application of modification rules; approval rates for non-G-7 and non-G-20 countries were broadly similar to G-7 and G-20 groups over 2014–22. - Approval rates for correction requests varied little across regions (range 84 to 93 percent over 2014–22). - Approval rates of deletion requests showed more geographical variation (range 22 to 85 percent) with particularly low rates in EMs and LICs in Asia and Central and Eastern Europe; however, these low rates were attributable to small sample biases or a few country reports in specific years. ### Sample design and compliance findings (ex-post assessment) - Sample design and scope - Staff examined 20 percent of the modification requests to country documents received during 2019, 2021, and 2022 — covering AEs, EMs, and LICs — to assess whether the outcome conformed with the TP’s rules. - Documents considered included Article IV staff reports, UFR staff reports, combined Article IV-UFR staff reports, and FSSAs. - 2020 was excluded from the sample because postponement of surveillance-related documents due to the COVID-19 pandemic resulted in a smaller sample size that year. - Staff focused on corrections for evident ambiguity and deletions. - Regional deletion request counts (2014–19 and 22) - Central and Eastern Europe: 22 deletion requests. - Asia: 69 deletion requests, of which 54 requests were rejected. - Among the rejected requests in Asia, 43 were for four specific papers. - Compliance findings from ex-post assessment - Corrections for evident ambiguity: about 1 percent of the corrections for evident ambiguity (four cases) did not comply with the policy. - Deletions: about 5 percent of the sample (2 cases) does not appear to have fully complied with the policy. - In the identified corrections and deletions that did not comply, staff concluded the modification requests should have been rejected. - Given the small sample numbers in each group, staff could not identify any signs of bias. ### Borderline and difficult cases brought to management - Frequency and share - Since the last policy review, 25 modification requests (out of 11,502 modification requests) were considered controversial and brought to management attention. - These borderline cases covered about 0.2 percent of all requests and less than 1.8 percent of country papers presented to the Executive Board over the period 2014–2022. - Composition of borderline cases - 19 Article IV reports, 5 UFR reports, 1 SIP, and 1 FSSA. - These cases largely covered EMs and LICs (with the exception of two cases) and tended to cover Article IV reports. - Management outcomes - In 20 of these borderline cases (80 percent) the original rejection was confirmed or alternative modifications, acceptable under the policy, were agreed. - In the remaining 5 cases the original rejection was overruled, and the modification requests approved. - Effect on publication decisions - The rejection of modification requests in these borderline cases had limited effect on the decision by country authorities to authorize the publication of their country reports. - Publication outcomes: - About 80 percent of reports with disputed modifications were published. - Specifically, 70 percent of country reports with confirmed rejections (9 reports) were published and 4 reports were not published. - All reports for which alternative modifications were agreed or the original modification requests were accepted (11 reports) were published. ### Appendix I. Data — key methodological and dataset notes - Publication timing and scope - Documents counted refer to those considered by the IMF Executive Board during the calendar year, and published within six months of the end of the calendar year (e.g., data for 2021 refer to documents discussed by the Board in 2021 and published by June 30, 2022). - April 2020: Art. IV consultations and FSAs were temporarily suspended, with a few exceptions. - Data include documents issued for the Board's consideration in a meeting or on a lapse-of-time basis. SMPs, which are for information only, are included. - Publication presumption and exclusions - For the application of the stronger presumption of publication, published UFR staff reports include requests and reviews but exclude extensions and modifications, in accordance with the Transparency Policy (TP). - The presumption of publication does not apply to certain papers dealing with administrative matters. - Policy papers that are presumed to be published are tracked separately from the broader class of memos/all policy papers. - Country papers: definitions and classifications - “Country Papers” include staff reports for Article IV surveillance, Use of Fund Resources (UFR), and Article IV combined with UFR; they exclude reports for staff monitoring program, policy support/coordination instruments, post-program monitoring, and ex-post program assessments/evaluations. - Classification of economies and regions is based on World Economic Outlook (WEO) definitions; from 2020 the income and regional classifications refer to the 2019 WEO. Countries not included in the WEO are classified based on their GDP. - Publication lags and timing metrics - Publication lags are measured as calendar days between the Board date and the publication date; the dataset reports average lags and the percent of published reports with lag > 1 month & <= 2 months, and percent with lag > 2 months. - The reporting window for publication within six months is a consistent metric used across tables and years. - Deletions and corrections (data recording notes) - The dataset records deletion and correction requests received by country desks and reviewed by SPR; it does not include requests received and rejected by country teams that were not forwarded to SPR for review. - Each deletion or correction is counted separately; partially rejected requests were not counted to avoid double-counting. - In June 2022, the Transparency Policy was modified to allow for targeted deletions related to the new Debt Sustainability Framework for Market Access Countries (MAC SRDSF) based on market sensitivity; these accounted for 20 percent of deletions in 2022. - With the launch of the Transparency Portal on January 1, 2017, modification requests reflect only approved requests for which authoring departments sought SPR sign-off; for consistency from 2017 the number of papers with approved corrections includes an estimate of corrections for typographical errors and inconsequential factual errors. - Members and publication consent - Members can elect to “opt out” of the new publication guidelines, which presume consent to publication on the external website, unless the Fund is otherwise notified. Members that have chosen to opt out require their explicit consent prior to publication of their country and related policy intention documents. - Members requesting explicit consent for publication of their country documents (listed in the dataset) include Bahrain, Brazil, Bulgaria, Cabo Verde, Dominican Republic, Ecuador, Egypt, Guyana, Haiti, Iraq, Jordan, Kuwait, Lebanon, Maldives, Nicaragua, Oman, Panama, Qatar, Saudi Arabia, Suriname, Syrian Arab Republic, Timor-Leste, Trinidad and Tobago, Turkmenistan, United Arab Emirates, Uzbekistan, Yemen. *Source: ppea2024058 - Appendix I. Data* ### EXECUTIVE SUMMARY ### EXECUTIVE SUMMARY ### Publication Trends - Since the 2013 Transparency Policy review, the publication rates of Board documents, including both country and policy documents, have improved and have reached about 96–98 percent in recent years. - Nearly all Board documents are published. - Country papers not published are generally concentrated in a few countries and their non-publication tends to reflect country-specific factors. - The timeliness of publication of country papers has improved, although an important share of these papers is published with long delays. - Since the last review, country papers were on average published 16 days after Board consideration. - About a third of these papers were published more than 14 days after Board consideration. - A considerable share is still published with delays exceeding four weeks. - The average publication lags for policy papers have lately increased, reflecting delays in publishing specific papers with long processing timelines. - Press releases for program cases are published immediately after Board consideration, but press releases for Article IV consultations are published with considerable delay. - Press releases for Article IV consultations are on average issued 10 days after Board consideration. ### Modifications of Board Documents - Most country papers are published with some corrections (about 75 percent on average). - The share of country papers published with deletions has declined since the last policy review, reflecting reduced concerns about financial sector issues after the 2008 global financial crisis. - Most correction requests and a large share of deletion requests are approved. - There is no evident evenhandedness issue in the application of the correction and deletions rules of the Transparency Policy. - Statistical evidence of acceptance and rejection rates of modification requests and a detailed assessment of a large sample of modification requests show no systematic bias in the approval of deletion and correction requests across countries, income level groups, regions, and economic size. ### Publication Rates: Trends and Analysis - Overall publication performance - Since the 2013 review, publication rates for Board documents have edged higher, and nearly all Board documents are now published. - During 2014–22, the share of Board documents covered by the policy that are published increased from an average of 89 percent during 2009–12 to 97 percent in 2022 with an average of 95 percent during 2014–22. - Improvement reflects increases in publication rates of both country and policy papers. - Country papers - During 2014–22, about 94–98 percent of country papers were published compared to 91–94 percent in 2009–12. - Article IV (AIV) reports publication improved from an average of 89 percent in 2009–12 to 94 percent in 2014–22. - Publication of country papers for Emerging Markets (EM) and Low-Income Countries (LIC) increased to 93–98 percent from 89–93 percent, approaching the 100 percent publication rate of country papers of Advanced Economies (AE). - Use-of-Fund-Resources (UFR) and related reports - The stronger publication presumption for reports on the UFR introduced in 2013 had limited impact because rates were already high: average publication rate was 96 percent over 2009–12 and edged up to 99 percent during 2014–22. - Other country documents - Publication rate of Financial System Stability Assessment (FSSA) reports improved to an average of 86 percent during 2014–22 (from 73 percent over 2009–12) but remains volatile. - Publication rates for SMP, EPA, EPE, PPM, PSI, and PCI have improved. - Selected Issues Papers (SIPs) publication rate remained broadly constant around 88–91 percent, on average below the publication rate of associated Article IV reports. - Policy papers and multi-country documents - Over 2014–22, the average publication rate of policy papers was 97 percent (up from 90 percent during 2009–12). - 23 percent of policy papers presented to the Executive Board during 2014–22 were not covered by the presumption of publication regime. - Since 2013, 35 multi-country documents were issued to the Executive Board, 30 were presumed to be published, and among those presumed to be published, only one was not published. ### Can Higher Publication Rates Be Achieved? - Unpublished country papers, 2014–22 - Since 2014, 65 country papers (out of 1,628 papers) were not published—about 7 country papers per year. - Half of these unpublished papers were for EMs, and the rest were for LICs. - Countries in LAC had the largest share (29 percent) of unpublished country papers. - Role of modification requests in non-publication - Since 2014, 78 percent of unpublished country papers had either no modification requests (52 percent) or had all modification requests approved (26 percent). - The remaining 22 percent (14 country papers) that were not published had some rejected modifications, and only 5 percent (3 papers) had rejected deletions. - This suggests neither market sensitivities nor early policy disclosures have been major driving factors preventing publication. - Concentration of non-publication cases - Over 2014–22, 35 countries had at least one unpublished country report. - 11 countries had two or more unpublished papers, accounting for 63 percent of all unpublished reports. - Five countries (Antigua and Barbuda, Bahrain, Brunei Darussalam, Oman, and Turkmenistan) accounted for about 42 percent of all cases of unpublished country papers. - Increasing the publication rate further would require addressing idiosyncratic factors preventing publication in a few specific country cases. ### Translation and Languages Other Than English - Translation rates - About 31 percent of press releases (PRs) of country papers were translated into languages other than English during 2019–22 (PR data available only for this period). - Only about 14 percent of all country-related documents were published in languages other than English during 2019–22, with no major improvement compared to 2010–12 (12 percent). - For policy documents, the translation rate over 2014–22 was 1.4 percent. - Note: Translation is outside the scope of the Transparency Policy, but there is scope to increase translations given English is the “de-jure” common/official language in only about one third of IMF member countries. ### Publication Lags - General timeliness - The publication timeliness of country papers has improved since the last policy review. - A majority of country and policy papers are now published promptly (i.e., within 14 days). - However, a number of country papers are still published with long delays (i.e., 28 days or more), mainly due to delay in obtaining publication consent. - Board documents - The average publication lag of country papers declined during 2014–22 to 16 days, from 33 days in 2009–12. - Publication lags improved for both Article IV and UFR reports. - Since 2013, average publication lag of FSSAs and SIPs declined to 25 and 20 days, respectively (compared to 52 and 25 days during 2009–12). - The 2013 Review redefined “prompt publication” as publication within 14 days from Board consideration or Lapse-of-Time (LOT) decision date. - During 2014–22, 67 percent of country papers were published promptly (within 14 days), compared to 45 percent at the time of the last review. - 45 percent of country papers were published within 7 days, compared to 23 percent in 2009–12. - Improvements were driven by a substantial increase in prompt publication for EMs and LICs (62 percent during 2014–22 versus 35 percent in 2009–12). - Prompt publication for AE increased from 88 percent to 93 percent. _Prepared by the Strategy, Policy, and Review Department; approved by Mark Flanagan (SPR), Gilles Bauche (SEC), Magi Clavé Badia (ITD), Bernhard Steinki (LEG). October 1, 2024._ ### 13. However, the publication time of ### ppea2024058 - 13. However, the publication time of ### Publication lags of country papers — overall patterns - Publication lags for EMs and LICs remain more than double than that for AEs. - For EMs, the average publication time for country papers ranged from 8 days for countries in Central and Eastern Europe to nearly a month for countries in Middle East and North Africa. - For LICs, the average lag ranged from 15 days for countries in Asia to 30 days for LAC. - During 2014–22, on average about 33 percent of country papers had publication lags that exceeded 14 days. - During 2014–22, about 12 percent of country papers were published more than 28 days after Board consideration (referred to as “long-lag papers”). - About 57 percent of these long-delayed country papers were for EM economies, followed by LICs (41 percent) and AEs (2 percent). - Some papers were published more than 90 days after Board consideration and followed a lower-profile publication procedure. ### Distribution, multi-country papers, and differences over time - Average publication lag for multi-country papers during 2014–22 was 21 days, compared to 16 days for single country papers. - About 60 percent of multi-country papers were published promptly (i.e., within 14 days), compared to about 67 percent for single country papers. - The longer publication time for multi-country documents likely reflects a more complex publication consent process, potentially requiring publication consent from all relevant countries. ### Policy papers publication timing - Between 2014–22, the median publication time for policy papers was 11 days, compared to 8 days during the previous review period (2009–12). - The share of policy papers published promptly (i.e., within 14 days) declined to about 61 percent (from 68 percent in 2009–12). - The average lag of the policy papers published in more than 28 days after Board consideration reached 88 days during 2014–22, compared to 80 days in 2009–12. ### Press releases (PRs) — timing and patterns - During 2019–22, PRs for country papers were, on average, published 6 days after the Board meeting. - About 22 percent of PRs took more than 7 days to be issued, while 43 percent of PRs for country papers were issued on the same day of the Board meeting. - Excluding 2020 data, the share of country-paper PRs issued on the same date of the corresponding Board meeting would be 35 percent. - For policy papers, the average publication timing in 2019–22 was 26 days as PRs for these papers are usually issued together with the corresponding policy paper. - During 2019–22, 91 percent of PRs for UFR cases were issued on the same day of the relevant Board discussion. - For Article IV consultations, PRs were on average published 10 days after Board consideration; 20 percent of these PRs were published more than two weeks after Board consideration. - Only 2 percent of Article IV consultation PRs were issued on the same day of the relevant Board consideration; about 29 percent were issued within 3 days after Board consideration. - During 2019–22, about 71 percent of AIV PRs were issued together with the corresponding staff report, and about 47 percent of them were issued more than seven days after Board consideration. - For policy documents, during 2019–22, 92 percent of PRs were issued at the time of the publication of the corresponding policy document; on average PRs for policy documents were issued about 26 days after the relevant Board meeting with 62 percent of PRs issued within 14 days. ### What drives publication lags of country documents and PRs? - CSO surveys noted long delays in the publication of some country reports and requested more information about the Executive Board’s activities. - Long publication delays for country papers are associated with idiosyncratic factors prevailing in some countries and delays in obtaining publication consent from members. - During 2014–22, country papers published more than 28 days after Board consideration covered 91 countries; these were mostly EMs and LICs. - Seven countries alone accounted for about 1/4 of all long-lag country papers. - The practice of seeking explicit publication consent (rather than relying on non-objection) and the lack of a specific timeline to obtain consent can result in long publication delays. - During 2014–22, 16 countries that have opted out of non-objection regime accounted for 28 percent of all country papers published more than 28 days after Board consideration. ### Role of modification requests in delays - During 2018–22, about 17 percent of modifications to country papers are estimated to have been made after Board consideration. - Among the long-lag papers, only 11 percent had the modification requests settled more than 28 days after Board consideration. - For most long-lag papers, final modification requests were submitted after the 28-day mark; thus the post-Board modification process under the TP was not the driver of the publication timing for most cases. - About 90 percent of country papers not published promptly (after 14 days but less than 28 days) had no modification requests or no pending requests at the 14-day mark, suggesting delay in obtaining publication consent was the likely cause of non-prompt publication. - For most papers with modification requests settled after the 14-day mark, the final requests themselves were submitted after the 14-day mark, again indicating the modification process was not a direct cause of non-prompt publication. - Delayed publication involved a limited number of small developing countries and other LICs and were generally one-off events associated with specific reports during 2018–22. ### Press releases drivers — Article IV PRs and staff report linkage - Long publication time of PRs for Article IV consultations reflects current publication rules and the expectation that PRs are published with the staff report. - For UFR cases PRs are expected to be issued immediately after Board consideration. - For Article IV consultations, regional surveillance discussions, and stand-alone FSSA reports, PRs are expected to be issued along with the corresponding staff report if the latter is expected to be published within seven days after Board consideration; otherwise, a PR is expected to be published shortly after the Board consideration. - Evidence suggests PRs for Article IV consultations tend to follow the publication timing of staff reports even when staff report publication is delayed. - During 2019–22, about 71 percent of AIV PRs were issued together with the corresponding staff report, and about 47 percent of them were issued more than seven days after Board consideration. - The expectation of issuing PRs with the staff report and the tendency to do so even when staff report publication is delayed are main reasons for significant PR lags for Article IV consultations. ### Modifications — deletions and corrections trends - The share of country papers published with some deletions has sharply declined since the last review. - In 2014–22, on average about 12 percent of published country papers had some deletions compared to around 21 percent before the 2013 review. - The decline reflects a return to deletion levels prevailing before the Global Financial Crisis. - The decline in deletions was largely driven by a drop in the share of papers with deletions for financial sector issues, which declined to 5 percent during 2014–22, from 12 percent during 2009–12. - Deletions for exchange rate issues remained broadly constant. - The decline in deletions was associated with a reduction in approval rate of the deletion requests since 2014. - A large share of country papers continues being subject to corrections and most deletion and correction requests are approved. - An analysis of approval/rejection rates of deletion and correction requests and an ex-post assessment of a large sample of such requests indicate no evident bias in the application of the modification rules to country documents. *Source: IMF staff calculations.* ### 26. With declining concerns about financial sector issues, deletion rates have tilted ### 26. With declining concerns about financial sector issues, deletion rates have tilted ### Deletion trends by country group - During 2014–22, the share of EM country papers with deletions was 17 percent, compared to 12 percent for AE papers. - This represents a shift from the period before the previous review when AEs recorded higher shares (35 percent for AEs, compared to 24 percent for EMs). - During 2014–22, deletion requests averaged: - 0.6 deletions per paper for EMs - 0.2 deletions per paper for AEs - 0.3 deletions per paper for LICs - Overall, in 2014–22, about 51 percent of all deletion requests were for country papers for EMs. - For policy papers over the same period, deletion requests were low: on average only one in eight policy papers had some deletion requests. - Note: Analyses of 2022 deletion requests exclude targeted deletions introduced for MAC SRDSF (unless otherwise noted). ### Approval rates of deletion requests - During 2014–22, on average about 64 percent of deletion requests to country papers were approved. - The median approval rate across papers for 2014-19 and 22 was 83 percent. - Approval rates for country papers (and all country documents) have remained stable over time, ranging from 59 to 76 percent once outlier cases are accounted for. - Approved deletions by stated reason (percent, excl. 2020 and 21): - Highly Market Sensitive Material, 76 - Premature Disclosure of Policy Intentions, 23 - Removal of References to Unpublished Documents, 1 - About three-quarters of the approved deletions were for market sensitivity concerns and the remaining quarter for premature disclosure of authorities’ policy intentions. ### Corrections: incidence and types - During 2014–22, about 75 percent of published country papers had corrections, compared to about 65 percent in 2009–12. - Excluding typographical errors, correction requests for 2018–22 were: - 52 percent to rectify factual errors - 41 percent for evident ambiguities - 7 percent amending the characterization of authorities’ views - Correction requests were concentrated: about 7 percent of countries that requested corrections during 2014–22 had more than 150 requests, which accounted for 28 percent of total correction requests. - For 2014–22, average number of correction requests per country paper by income group: - AE country papers: 18 - EM country papers: 9 - LICs: 3 - Article IV reports accounted for 77 percent of correction requests for country papers during 2014–22; on average Article IV reports had about 9 correction requests per report, compared to about 5 for Combined Article IV and UFR reports and about 3 for UFR reports. - About 17 percent of correction requests (about 18 percent of all modification requests including deletions) were received after Board consideration during 2018–22; the share was slightly higher for EM and LIC than for AE. ### Approval rates of correction requests - During 2014–22, the average approval rate of correction requests for country papers was 87 percent (ranging 77–93 percent across years). - Approval rates of correction requests varied little across types of requests and document types. - Approval rates of correction requests by type, 2018–22 (percent): - Factual Errors: approval rate shown in charts (range indicated in source) - Evident Ambiguity: approval rate shown in charts - Mischaracterization of the Views of the Authorities: approval rate lower but driven by three specific papers; excluding these, approval rate would be 74 percent (almost same as for evident ambiguity). - Country-level distribution of approval rates for correction requests, 2014–22: - 74 percent of countries had approval rate of correction requests of 80 percent or above - 91 percent had approval rates of 70 percent and above - Only eight countries had approval rates lower than 65 percent (two standard deviations below the mean); these cases were driven by small sample bias or single report weight rather than systemic bias ### Evenhandedness and cross-group analysis - Stakeholder surveys indicate perceived uneven application of the policy has abated: about 37 percent of responding EDs perceive uneven application, down from 50 percent in 2013. - Statistical evidence does not signal systemic issues with evenhanded application of modification rules: - No evidence of systematic biases in approval of correction requests across countries. - No evidence of systemic bias in approval rates of modification requests across country income levels: - Average approval rate of deletion requests, 2014–22, by income group: - AE: 65.1 - EM: 64.6 - LIC: 61.9 - Average approval rates for correction requests, 2014–22, by income group: - AE: 88.5 - EM: 86.2 - LIC: 87.8 - No evidence that largest economies (G-7 or G-20) benefit in application of modification rules; approval rates for non-G-7 and non-G-20 countries were broadly similar to G-7 and G-20 groups over 2014–22. - Approval rates for correction requests varied little across regions (range 84 to 93 percent over 2014–22). - Approval rates of deletion requests showed more geographical variation (range 22 to 85 percent) with particularly low rates in EMs and LICs in Asia and Central and Eastern Europe; however, these low rates were attributable to small sample biases or a few country reports in specific years. *Source: IMF staff calculations (excerpts from the provided content).* ### 35. The statistical analysis above was complemented by an ex-post assessment of a large ### ppea2024058 - 35. The statistical analysis above was complemented by an ex-post assessment of a large ### Sample design and scope - Staff examined 20 percent of the modification requests to country documents received during 2019, 2021, and 2022 — covering AEs, EMs, and LICs — to assess whether the outcome conformed with the TP’s rules. - Documents considered for the sample included Article IV staff reports, UFR staff reports, combined Article IV-UFR staff reports, and FSSAs. - 2020 was excluded from the sample because the postponement of surveillance-related documents due to the COVID-19 pandemic resulted in a smaller sample size that year. - Staff focused on: - corrections for evident ambiguity, and - deletions, two types of modifications that are usually more likely to be controversial. ### Regional deletion request counts (2014–19 and 22) - Central and Eastern Europe: 22 deletion requests. - Asia: 69 deletion requests, of which 54 requests were rejected. - Among the rejected requests in Asia, 43 were for four specific papers. ### Compliance findings from ex-post assessment - Corrections for evident ambiguity: - About 1 percent of the corrections for evident ambiguity (four cases) did not comply with the policy. - This compares to 10 percent of cases when staff conducted a similar assessment as part of the 2013 review. - Deletions: - About 5 percent of the sample (2 cases) does not appear to have fully complied with the policy (broadly similar results to the assessment in the 2013 review). - In the identified corrections and deletions that did not comply, staff concluded the modification requests should have been rejected. - Given the small sample numbers in each group, staff could not identify any signs of bias. ### Borderline and difficult cases brought to management - Definition and identification: - Borderline and difficult cases are those brought to management attention because of serious disagreements with a member regarding that member’s request for modifications. - These cases represent the most egregious disagreements where country members felt an acceptance under the policy was warranted and not granted. - Frequency and share: - Since the last policy review, 25 modification requests (out of 11,502 modification requests) were considered controversial and brought to management attention. - These borderline cases covered about 0.2 percent of all requests and less than 1.8 percent of country papers presented to the Executive Board over the period 2014–2022. - Composition of borderline cases: - 19 Article IV reports, 5 UFR reports, 1 SIP, and 1 FSSA. - These cases largely covered EMs and LICs (with the exception of two cases) and tended to cover Article IV reports (19 Article IV reports, 5 UFR reports, one SIP, and one FSSA). - Management outcomes: - In 20 of these borderline cases (80 percent) the original rejection was confirmed or alternative modifications, acceptable under the policy, were agreed. - In the remaining 5 cases the original rejection was overruled, and the modification requests approved. ### Aggregate table highlights (2014–22) - Total modification requests since last review: 11,502. - Total number of requests going to Management: 26 (less than 1%). - Note: Of these, one item was withdrawn before management could make a decision. - Composition of management requests: 19 Article IV, 5 UFR, 1 SIP, 1 FSSA. - Total number of requests overruled by Management: 5 (20% of cases). ### Effect on publication decisions - The rejection of modification requests in these borderline cases had limited effect on the decision by country authorities to authorize the publication of their country reports. - Publication outcomes: - About 80 percent of reports with disputed modifications were published. - Specifically, 70 percent of country reports with confirmed rejections (9 reports) were published and 4 reports were not published. - All reports for which alternative modifications were agreed or the original modification requests were accepted (11 reports) were published. *2024 TRANSPARENCY POLICY REVIEW AND OPEN ARCHIVES POLICY⸺KEY TRENDS, INTERNATIONAL MONETARY FUND* ### Appendix I. Data ### ppea2024058 - Appendix I. Data ### Publication timing and scope - Documents counted refer to those considered by the IMF Executive Board during the calendar year, and published within six months of the end of the calendar year (e.g., data for 2021 refer to documents discussed by the Board in 2021 and published by June 30, 2022). - To deal with the exceptional circumstances of the COVID-19 pandemic, in April 2020, Art. IV consultations and FSAs were temporarily suspended, with a few exceptions for practical and operational reasons. - Data include documents issued for the Board's consideration in a meeting or on a lapse-of-time basis. SMPs, which are for information only, are included. ### Publication presumption and exclusions - For the application of the stronger presumption of publication, published UFR staff reports include requests and reviews but exclude extensions and modifications, in accordance with the Transparency Policy (TP). - The presumption of publication does not apply to certain papers dealing with administrative matters, for instance internal audits and papers on personnel and human resource issues. - Policy papers that are presumed to be published are tracked separately from the broader class of memos/all policy papers. ### Country papers: definitions and classifications - “Country Papers” include staff reports for Article IV surveillance, Use of Fund Resources (UFR), and Article IV combined with UFR; they exclude reports for staff monitoring program, policy support/coordination instruments, post-program monitoring, and ex-post program assessments/evaluations. - Classification of economies and regions is based on World Economic Outlook (WEO) definitions; from 2020 the income and regional classifications refer to the 2019 WEO. Countries not included in the WEO are classified based on their GDP. - The dataset includes discussions held with respect to members' territories and currency unions in the context of Article IV consultations with members. ### Publication lags and timing metrics - Publication lags are measured as calendar days between the Board date and the publication date; the dataset reports average lags and the percent of published reports with lag > 1 month & <= 2 months, and percent with lag > 2 months. - Publication lags and rates are reported for multiple document categories including Article IV, UFR, Combined, Advanced Economies, Emerging markets, Developing countries, Emerging Market and Developing Economies, Africa, Asia, Central and Eastern Europe, CIS and Mongolia, Middle East, and Western Hemisphere. - The reporting window for publication within six months is a consistent metric used across tables and years. ### Deletions (Country Papers) - The dataset records deletion requests received by country desks and reviewed by SPR; it does not include requests received and rejected by country teams that were not forwarded to SPR for review. Each deletion is counted separately, implying there may be multiple deletion requests for each document. - The number of deletion requests is reported as the sum of requests approved and rejected. Partially rejected requests were not counted to avoid double-counting. - In June 2022, the Transparency Policy was modified to allow for targeted deletions related to the new Debt Sustainability Framework for Market Access Countries (MAC SRDSF) based on market sensitivity; these accounted for 20 percent of deletions in 2022. - Deletion statistics are presented by economic and regional groupings, and include breakdowns for deletions related to exchange rate issues and financial sector issues. ### Corrections (Country Papers) - Correction requests received by country desks and reviewed by SPR are recorded; this aggregate does not include requests received and rejected by country teams that were not forwarded to SPR for review. Each correction is counted separately, implying multiple correction requests may exist for a single document. - The number of correction requests is the sum of requests approved and rejected. Partially rejected requests were not counted to avoid double-counting; many partially rejected requests were resubmitted to finalize modification. - With the launch of the Transparency Portal on January 1, 2017, modification requests reflect only approved requests for which authoring departments (ADs) sought SPR sign-off; this tends to exclude corrections of typographic errors and inconsequential factual errors. For consistency in the time series, from 2017 the number of papers with approved corrections includes an estimate of corrections for typographic errors and inconsequential factual errors for which ADs were no longer requesting sign-off and were not reported. - Correction approvals and correction approval rates are reported by economic and regional groupings. Corrections are classified by type, including “Evident Ambiguity” and “Mischaracterization of Authorities’ views.” ### Members and publication consent - Under the Fund’s Transparency Policy members can elect to “opt out” of the new publication guidelines, which presume consent to publication on the external website, unless the Fund is otherwise notified. Members that have chosen to opt out require their explicit consent prior to publication of their country and related policy intention documents. - Members requesting explicit consent for publication of their country documents: - Bahrain - Brazil - Bulgaria - Cabo Verde - Dominican Republic - Ecuador - Egypt - Guyana - Haiti - Iraq - Jordan - Kuwait - Lebanon - Maldives - Nicaragua - Oman - Panama - Qatar - Saudi Arabia - Suriname - Syrian Arab Republic - Timor-Leste - Trinidad and Tobago - Turkmenistan - United Arab Emirates - Uzbekistan - Yemen *Source: ppea2024058 - Appendix I. Data* --- _Source: https://www.imf.org/-/media/files/publications/pp/2024/english/ppea2024058.pdf_