40. Various factors played a role in the IMF’s failure to identify risks and give clear warnings. Many of these factors represent long-standing problems that had been highlighted for over a decade.²⁴ In this section, these factors are grouped into the following broad categories: analytical weaknesses, organizational impediments, internal governance problems, and political constraints.²⁵ There are considerable interconnections among these categories, and their relative importance is based on subjective judgments. The IMF’s ability to correctly identify the mounting risks was hindered by a high degree of groupthink, intellectual capture, a general mindset that a major financial crisis in large advanced economies was unlikely, and incomplete analytical approaches. Weak internal governance, including unclear lines of responsibility and accountability, lack of incentives to work across units and raise contrarian views, a review process that did not “connect the dots” or ensure follow-up, and an insular culture also played a big role, while political constraints may have also had some impact. Interviews with country authorities (Annex 7) and survey evidence from staff (Annex 8) echo many of the same factors.

A. Analytical Weaknesses

41. Analytical weaknesses were at the core of some of the IMF’s most evident shortcomings in surveillance, particularly for the largest advanced economies. These weaknesses were of two broad types: groupthink and other cognitive biases, and analytical approaches/ knowledge gaps. Neither implies that the IMF staff lacked skills or expertise; the first type is about thought processes and decision-making, the second is about the approaches and tools that staff used.

42. Several cognitive biases seem to have played an important role. Groupthink refers to the tendency among homogeneous, cohesive groups to consider issues only within a certain paradigm and not challenge its basic premises (Janis, 1982). The prevailing view among IMF staff—a cohesive group of macroeconomists—was that market discipline and self-regulation would be sufficient to stave off serious problems in financial institutions. They also believed that crises were unlikely to happen in advanced economies, where “sophisticated” financial markets could thrive safely with minimal regulation of a large and growing portion of the financial system.

43. IMF staff was essentially in agreement with the views of the U.S., U.K., and other advanced country authorities that their financial systems were essentially sound and resilient. Staff also concurred with the paradigm that the system could not only allocate resources efficiently, but also redistribute risks among those better prepared to bear them. Moreover, IMF staff felt uncomfortable challenging the views of authorities in advanced economies on monetary and regulatory issues, given the authorities’ greater access to banking data and knowledge of their financial markets, and the large numbers of highly qualified economists working in their central banks. The IMF was overly influenced by (and sometimes in awe of) the authorities’ reputation and expertise; this is perhaps a case of intellectual capture.

44. Confirmation bias is a well-documented cognitive bias that refers to the tendency of people to only notice information consistent with their own expectations and to ignore information that is inconsistent with them (Bazerman and Moore, 2009). This may explain staff’s focus on the IMF’s primary concern—global imbalances and a disorderly dollar decline—as the key risk to global stability, largely ignoring evidence pointing to other risks.

45. The choice of analytical approaches and important knowledge gaps, some of which were shared by the whole profession, also played a role in the failure to identify risks and vulnerabilities. The linking of macroeconomic and financial sector analysis remained inadequate, even though a series of evaluations since the Asian crisis had called for enhanced attention to macro-financial linkages in the IMF’s surveillance (Caprio, 2011, and Annex 6). This reflected the lack of a suitable conceptual framework for analyzing such linkages within the economics profession at large, as well as the view common among IMF economists that financial issues were not central.²⁶

46. IMF economists tended to hold in highest regard macro models that proved inadequate for analyzing macro-financial linkages. The dynamic stochastic general equilibrium (DSGE) model that was the work horse for policy discussions introduced money and asset markets in only the most rudimentary manner. Work is now ongoing to develop models that can incorporate financial frictions. Perhaps more worrisome was the overreliance by many economists on models as the only valid tool to analyze economic circumstances that are too complex for modeling.²⁷

47. Balance sheet analysis was used insufficiently and occasionally incorrectly, despite the fact that sometimes this approach captures risks and vulnerabilities better than would a typical open-economy macro model. As one senior staff member put it, “balance-sheet analysis was the missing link in macro analysis.” Unfortunately, sometimes when this approach was used, it yielded misleading results as it did not account for the ongoing bubble in asset prices.

48. FSAPs used stress testing to help determine the soundness of banking systems. While stress tests are useful for a first-round examination of risks, they typically do not capture second-round effects or liquidity shocks. As a result, a number of authorities and staff believe that stress tests could have led to complacency, because their limitations were not explicitly discussed.

49. Lack of data and information, while a problem, was not a core reason behind the IMF’s performance. First, much available data were ignored or misinterpreted (e.g., credit growth, leverage, the growth of high-risk instruments, and household balance sheets).²⁸Indeed, the April 2008 GFSR estimate of financial sector losses was produced without any additional access to data. Second, the lack of data did not prevent the IMF from praising the state of some financial systems nor the risk-diversification features of securitization. Moreover, the relative paucity of data in some emerging markets did not prevent the IMF from raising the alarm in these countries. Finally, advanced country surveillance teams typically received the information that they requested, and in any case it is unclear how they would have used additional data on individual financial institutions given their prevailing conceptual framework on macro-financial linkages.

B. Organizational Impediments

50. An important organizational impediment that hindered IMF performance was its operating in silos, that is, staff tend not to share information nor to seek advice outside of their units. This has been blamed for the IMF failure to “connect the dots” in the run-up to the crisis. The silo behavior is a long-standing problem; and it occurs between departments, within departments, within divisions, and even within Management, adversely affecting the IMF staff’s ability to learn lessons from each other’s experiences and knowledge.²⁹

51. The silo behavior made it difficult to integrate multilateral with bilateral surveillance, to link macroeconomic and financial developments, and to draw lessons from cross-country experience. Discussion of the risks and vulnerabilities that led to the crisis never found its way into the bilateral surveillance of the largest systemic financial centers, even though some of these risks were laid out in GFSRs. A survey done for the IEO’s research evaluation (IEO, 2011) suggests that almost half of respondents in area departments admitted to seldom or never using the GFSR; the most mentioned reason was that its analysis did not lead to country-specific insights.

52. The internal review process failed to “connect the dots” and to ensure follow-up of concerns raised by the Board, Management, and internal reviewers. It did not connect bilateral and multilateral surveillance, or coordinate the analysis of the WEO and the GFSR. Formal interdepartmental review typically took place at a late stage in the production process of flagship documents, country briefs, and staff reports. In part, this explains its failure to ensure coordination and cross-fertilization. By the time comments were received from other departments, views had already crystallized, and it was often too late to make significant changes. Hence, comments were only minimally addressed.³⁰

53. IMF reports rarely referred to work by external analysts pointing at the mounting risks in financial markets. Rather than lack of awareness, it is likely that this was an example of the IMF’s insular culture, as this was also common in much of the surveillance-related analytical work, which made little reference to research from outside the IMF (IEO, 2011).

54. IMF macroeconomists, particularly in area departments, did not sufficiently appreciate the skills and experience of financial sector experts. At times there was a “culture clash” between macroeconomists and financial sector specialists and their analytical approaches. In addition, bilateral surveillance missions to systemic financial centers were not always staffed with the most experienced financial sector specialists.

C. Internal Governance Problems

55. Internal governance refers to the incentives and management processes that apply to IMF staff and the organization as a whole. The evaluation found that incentives were not well aligned to foster the candid exchange of ideas that is needed for good surveillance—many staff reported concerns about the consequences of expressing views contrary to those of supervisors, Management, and country authorities. It also found lapses in oversight and weak accountability.³¹

56. Staff reported that incentives were geared toward conforming with prevailing IMF views. Several senior staff members felt that expressing strong contrarian views could “ruin one’s career.” Thus, views tended to “gravitate toward the middle” and “our advice becomes procyclical.”³² Staff saw that conforming assessments were not penalized, even if proven faulty. A lack of accountability was frequently highlighted as a serious obstacle to getting the incentives right.

57. Many area department economists felt that there were strong disincentives to “speak truth to power,” particularly in large countries, as there was a perception that staff might not be supported by Management if they disagreed with these authorities. One senior staff member asserted that area departments were “unduly captured by countries” that they worked on. Analytical work was geared to “justify” the authorities’ policy proposals. All this was “driven by the agenda of getting on well with” country authorities.

58. High staff turnover was a frequent complaint of country authorities and an issue that has been raised by several previous IEO evaluations. High turnover left mission teams in constant need of getting up to speed on country-specific issues, which made it difficult to come up with alternative views and policy options. Staff working on countries with complex, systemic financial systems reported that they felt uncomfortable raising difficult issues during their first mission.

59. Turnover of Management and senior staff also weakened IMF effectiveness during the run-up to the crisis. In this period, the IMF had three Managing Directors, and an Acting Managing Director who served for three months. This high turnover led to shifting priorities, gaps in attention to the challenges facing the global economy and the IMF, and weak oversight over senior staff. IMF effectiveness also suffered as a consequence of changes in the First Deputy Managing Director, and Economic and Financial Counsellors during this period.

60. Turf battles, closely related to the issue of silos and incentives, were reportedly a major impediment to cooperation and collaboration. These were further evidence of a lack of sufficient oversight and follow through by senior staff and Management. The IMF was often described as a tightly-run, hierarchical organization, with clearly defined boundaries. According to one senior staff, “the Fund operates as little fiefdoms.” Staff attributed the failure to integrate bilateral and multilateral surveillance and macro-financial issues in part to such turf battles.

D. Political Constraints

61. What role might political constraints have played in the run-up to the crisis? The answer is multifaceted because political constraints have many dimensions, including requests to alter messages in staff reports, demands by authorities to replace specific mission members, perceptions of pressure from authorities leading to self-censorship, and requests to pursue certain policy initiatives. To varying extents, each of these factors influenced IMF surveillance during the evaluation period. But with the possible exception of self-censorship, they were not a major factor in the IMF performance in the run-up to the crisis.

62. On the messages from surveillance, the perceived degree of explicit or implicit political pressure from authorities varied significantly by country. On the United States, for example, staff and Management indicated that there was no overt pressure to change the mission’s messages. In some other large advanced economies, however, staff noted that the authorities took a heavy-handed approach, exerting explicit pressure to tone down critical messages. As one staff member who worked on a large country explained, “it was hard to give difficult messages to the authorities even if the team had the analysis … the concluding meetings were really just negotiation sessions on language.” In contrast, teams seemed more comfortable in presenting hard-hitting analysis to smaller advanced and emerging markets, confirming some authorities’ belief that there was a lack of evenhandedness in surveillance.

63. Self-censorship appeared to be a significant factor even in the absence of overt political pressure. Many staff members believed that there were limits as to how critical they could be regarding the policies of the largest shareholders—that “you cannot speak truth to authorities” since “… you’re owned by these governments.” Moreover, staff perceived that in case of disagreement, Management would end up endorsing country authorities’ views instead of those of staff. Sometimes country authorities would ask that a mission chief or other mission members be replaced. While at times there might be valid reasons for such requests, for example, a mismatch of skills or even personalities, such changes should be explained clearly and openly to staff or they could have a chilling effect on staff willingness to disagree with country authorities. While there have been few such cases, it is clear that staff across the IMF was aware of them and that this may have led to self-censorship.

64. Pressure to adopt certain initiatives distracted Management from more urgent concerns in the world economy, and their implementation diverted staff’s attention. In a multilateral organization like the IMF, it is natural for country authorities to influence the launching and design of policy initiatives. Indeed, there was a perception that the largest shareholders were the driving force behind certain initiatives that are seen as having distracted the institution while the crisis was emerging. The two main examples were the discussions leading to the adoption of the 2007 Decision on Bilateral Surveillance, which directed staff attention to exchange rate analysis and reinforced the focus on global imbalances; and the IMF’s 2008 downsizing, which absorbed the attention of Management and senior staff at a particularly important time.

65. Many authorities from member countries and other stakeholders pointed at problems in overall IMF governance as critical to understanding the institution’s performance in the run-up to the crisis. They indicated that to enhance its effectiveness, the IMF needed to clarify the roles of the Board, the Managing Director, and his Deputies, and to establish a clear accountability framework. In a survey conducted in 2007, Board members pointed to the lack of evenhandedness and weak accountability as hindering the capacity of the IMF to react effectively to emerging risks (IEO, 2008).