61. To incorporate social protection considerations into IMF operational work, the Board has consistently called on Fund staff to rely on the expertise of other relevant institutions (Abrams, 2017). According to economist staff surveyed by this evaluation, the World Bank was by far the IMF’s major partner on social protection issues: 80 percent of survey respondents reported interactions with Bank staff ranging from periodic or occasional meetings and information-sharing to joint missions. On the other hand, almost 75 percent and 90 percent of survey respondents respectively reported minimal to no interaction with UN agencies (including the ILO) and the OECD (Figure 12).79
Staff Survey Results on Extent of Interaction with Other Institutions
Source: Wojnilower and Monasterski (2017).Staff Survey Results on Extent of Interaction with Other Institutions
Source: Wojnilower and Monasterski (2017).Staff Survey Results on Extent of Interaction with Other Institutions
Source: Wojnilower and Monasterski (2017).A. The World Bank
62. This evaluation generally found an effective division of labor and good cooperation between the IMF and the World Bank on social protection issues. In areas such as social safety nets and pension reform (as well as energy subsidy reform), it was well understood that the World Bank would take the lead, and the IMF by and large agreed with the Bank’s approach. In the LIC and EME case studies examined for this evaluation, what sometimes appeared to be a low degree of IMF involvement in social protection issues was often explained by the fact that the World Bank was already actively engaged in that area.80 IMF staff almost always deferred to the Bank when it came to the design and implementation of social protection schemes.81 In interviews for this evaluation, IMF staff almost without exception expressed appreciation for the input and assistance of their World Bank colleagues; and World Bank staff likewise expressed appreciation for the IMF’s efforts to mobilize support for social protection issues and maintain the momentum of reforms. In the staff survey, 32 percent of respondents rated their cooperation with World Bank staff as “highly effective” and 55 percent rated it as “moderately effective” (Wojnilower and Monasterski, 2017).
63. While past IEO evaluations expressed concern at the lack of collaboration between the IMF and the World Bank on PSIA, in practice the IMF was able to undertake the analysis needed over the past decade largely using its own resources.82 Since 2008, PSIA has been “mainstreamed” into FAD’s expenditure policy work, particularly in the context of food and fuel subsidy reforms, and staff has not had to rely on the World Bank for such analysis. According to staff interviewed for this evaluation, FAD continues to field over 10 PSIA-related TA missions per year, and the number of staff familiar with this methodology has grown. This evaluation heard no complaints within the IMF about any difficulty in getting PSIA done on a timely basis by FAD during the evaluation period.
64. However, the World Bank’s approach to social protection may be evolving, which could complicate future collaboration. In 2015 the World Bank joined the ILO in support of universal social protection. A joint statement by the heads of the two institutions announced that they shared the objective to increase the number of countries that can provide universal social protection by supporting countries in designing and implementing universal and sustainable social protection systems (World Bank Group and ILO, 2015).83 Given the ILO’s grounding in the rights-based approach to social protection, it is not clear what the implications are for the World Bank’s long-standing approach of targeting social protection and for future Bank-Fund collaboration in this area. World Bank staff informed this evaluation that they are currently working with the ILO on a number of practical issues involved in achieving the goal of universal social protection that would have direct implications for targeting, fiscal costs, and choosing the right combination of instruments for a given context.84 IMF staff contacted by this evaluation indicated that they have not seen a departure from the Bank’s standard advice on designing and implementing social safety nets so far.
B. Other Institutions
65. The IMF’s collaboration with the ILO on the Social Protection Floor Initiative in 2011–13 was not very successful. The initiative aimed to ensure a basic level of social protection (access to essential services and social transfers for the poor and vulnerable) in every country.85 The idea for IMF-ILO collaboration originated at a 2010 conference in Oslo jointly hosted by the IMF Managing Director and the ILO Director-General (IMF, 2010b). Tree countries—El Salvador, Mozambique, and Vietnam—were chosen as pilot cases. Of the three pilots, this evaluation found effective collaboration only in Mozambique, where the IMF country team strongly believed in the need for greater social protection and worked closely with the ILO social protection specialist to successfully persuade the government to adopt a social protection floor.86 In the other two pilots, there was very little interaction, let alone collaboration, between IMF and ILO staff. According to IMF and ILO staff involved, there was no buy-in from Fund staff or the country authorities for the ILO’s social protection floor proposals.
66. Yet it was decided to scale up the Social Protection Floor pilots in 2014. Within the IMF, response to the successful Mozambique pilot was lukewarm. While the Communications Department (COM) highlighted the pilot in a factsheet, neither SPR, FAD, nor AFR considered using the case as a good-practice example of how to implement social protection measures within a sustainable macroeconomic policy framework.87 According to IMF staff interviewed for this evaluation, there was lingering skepticism among staff of the relevance of the initiative to the Fund’s mandate or department work programs and a broad sense that collaboration was difficult because IMF and ILO staffs “did not speak the same language.” Nonetheless, an unpublished joint IMF-ILO stocktaking in April 2014 concluded that the pilots had been very successful and proposed expanding the number of countries covered by the exercise. The IMF proposed some additional countries to the ILO in 2014, and the Managing Director’s Global Policy Agenda in the fall of that year included a commitment to “[continue] to analyze—jointly with the International Labor Organization—social protection floors, particularly for Africa, ACTs [Arab Countries in Transition], and Latin America” (IMF, 2014e). However, the momentum for IMF-ILO collaboration stalled shortly after that and no new exercises were initiated. Senior IMF and ILO staff interviewed for this evaluation could not explain why the decision was made to scale up the pilots and why the effort was subsequently abandoned.
67. The IMF has not been an active participant in the ILO- and World Bank-led Social Protection Interagency Cooperation Board (SPIAC–B). The SPIAC-B was established in 2012 in response to the G-20 request for international organizations to improve coordination of their efforts to provide social protection financing and technical advisory services to developing countries. The SPIAC-B met nine times between July 2012 to April 2017 to discuss, among other things, ongoing social protection initiatives; inter-agency joint work at the country level; collaboration on social protection assessment tools; and cooperation in the field of social protection statistics. However, the IMF was represented in only three meetings, in only one of them by a senior official. SPIAC-B member representatives told the IEO that the IMF’s absence was felt as there were areas where its input would have been constructive.
68. The IMF’s collaboration with UNICEF during the evaluation period was also not very successful. The two institutions had had a thorny relationship since the 1980s, when UNICEF called on the IMF (and the World Bank) to achieve “adjustment with a human face” (Jolly, 1991).88 In 2008, following the food and fuel price shocks and the financial crisis, IMF staff agreed to collaborate with UNICEF on achieving “recovery with a human face.” The collaboration took place at two levels: (i) the headquarters (or institutional) level—involving an exchange of views in joint meetings of senior staff from both institutions; and (ii) the country level—involving enhanced staff contacts in the field with the aim of protecting core social spending in 11 pilot countries.89 According to interviewees, collaboration at the headquarters level was interrupted by a clash of viewpoints (and personalities) after UNICEF staff published research on the extent fiscal consolidation in low- and middle-income countries resulted in cuts in social protection for children and poor households, which IMF staff saw as a “betrayal of trust” (in the words of one interviewee).90 Collaboration at the country level fared somewhat better, with about half of the 11 pilots able to point to some tangible results, even if it was just a mention in a staff report or in the authorities’ Memorandum of Economic and Financial Policies or PRSP.
69. The IMF-UNICEF collaboration was not sustained after 2011, even though both institutions rated the experiences of the pilot countries as “overwhelmingly positive” overall.91 According to IMF and UNICEF staff involved, contributing reasons included: staff turnover in both institutions; change in country circumstances; change in external circumstances (food and fuel prices fell and the financial crisis abated, hence some of the initial concerns became less urgent); and change in focus (the IMF turned its attention to its collaboration with the ILO, and UNICEF turned to other partners as well).
70. The IMF did not have any formal collaboration with the OECD on social protection during the evaluation period. As noted in the staff survey results (Wojnilower and Monasterski, 2017), IMF staff did draw on OECD data and research in their work, mostly in the context of surveillance. However, there were few cases in which staff consulted with the OECD on social protection issues in the handful of OECD member countries that had an IMF arrangement during the evaluation period.92 OECD staff interviewed for this evaluation confirmed that their views on social protection issues were rarely sought by the IMF during the euro area crisis, and suggested that this was a missed opportunity for closer collaboration between the two institutions.
71. Cooperation with regional institutions on social protection issues took place mainly at the country level. Among the regional development banks, IMF missions worked with the ADB and the IDB in several countries in a program and/or TA context during the evaluation period. This evaluation did not encounter any problems with the cooperation. In European Union (EU) member countries, IMF missions collaborated with the EC on social protection issues in the context of programs (particularly on issues of pension reform) but the experience was not always smooth. As noted in Heller (2017), two principal challenges that arose in the program context were: (i) differences in the decision-making time frame, where the EC negotiating team confronted different institutional clearance requirements that complicated IMF staff’s discussions with the authorities; and (ii) differences in policy priorities, where the EC had to consider, for example, that some policy reforms might set a precedent that would apply to other EU members.
