The IMF’s Independent Evaluation Office rarely lives up to its pretention to operate in real independence from the Fund, as its name implies, writes Peter Bavkis. One is hard-pressed to cite IEO reports that embrace thorough critiques of IMF policies, despite their frequency in academic circles and among civil society organisations, or that call for substantial revamping of policies and practices. The IEO report The IMF and Social Protection fits into this pattern.
However, by operating as a sounding board for IMF officials (ex-staffers make up most of its personnel), the 16-year-old IEO produces reports that often reveal useful information about Fund operations and staff attitudes. From that point of view, the report on social protection does not disappoint. Among other things, it provides a detailed account of the IMF’s hostility to concepts such as universal social protection.
Perhaps most surprising, for those who have not followed the IMF’s evolution in recent years, is the invitation to the Fund to question whether its endorsement in 2015 of the Sustainable Development Goals is compatible with its reaffirmed support for narrowly targeted safety nets.
Those of us in the trade union movement who have encouraged the international financial institutions (IFIs) to support UN-inspired initiatives such as the Social Protection Floor have found the IMF to be a reluctant participant, at best. In 2010, the Fund announced joint work with the ILO to identify “fiscal space” for financing national floors but, after making a serious effort in only one country (Mozambique), it abandoned the collaboration three years later. In 2012, the IMF was invited to join a Social Protection Inter-Agency Coordination Board, created at the behest of the G20 for expanding social protection coverage, but it declined to participate in all but a few of the meetings, even when they took place in New York where the IMF has permanent representation.
It is not a lack of involvement in social protection issues that explains this reluctance but, rather, the fact that the IMF has increasingly been the odd man out in its attitude to social protection. On the basis of interviews with (unidentified) Fund staff, the report reveals strong antipathy at the IMF towards cooperating on social protection issues with the ILO and other UN agencies. It states that the IMF’s targeting approach does not mesh well with the “rights-based approach” to social protection espoused by the other agencies.
The IMF’s suspicion of other agencies’ motives extends even to its sister IFI, the World Bank. In 2015, the Bank joined with the ILO to endorse the objective of universal social protection, which, the IEO asserts, “could complicate future collaboration” with the IMF. However reassuringly (for the Fund), the report states that so far it has “not seen a departure from the Bank’s standard advice” on the design of social safety nets, which presumably includes the type of targeting favoured by the Fund. It also admits that, because the IMF has no in-depth expertise in social protection, it has no choice but to continue to cooperate “at the least … with the World Bank”.
The IMF’s perspective, as the report explains, is driven by a focus on fiscal cost. Thus, as regards old-age pensions, the IMF has not been interested in “social issues” such as coverage or replacement rates but rather “macro-critical issues … such as fiscal sustainability and the short-term expenditure burden”. The need to reduce fiscal costs above all else also explains the emphasis on targeting social safety nets to the poorest.
Although the IEO’s analysis ends in 2015, an IMF policy paper on social safeguards in low-income countries released in June makes clear that strict targeting of social assistance so as to “narrowly cover the needs of the most vulnerable” remains the central focus of the Fund’s approach. Whether its emphasis on narrow targeting means that assistance actually goes to those who need it most is something that both the IEO report and the social safeguards policy paper decline to examine. Ample evidence exists to show that the proxy means tests promoted by the IFIs for targeting in developing countries typically exclude 50 per cent or more of those who should be eligible by virtue of their income level.
The negative effects of such policies have been felt in IMF borrowing countries. In Tunisia, which began the first of two loans in 2013, social expenditures at the end of the first programme were 14 per cent below the symbolic social spending floor in the lending agreement. An IMF staff report blames the missed target on failures of the delivery mechanism for the “well-targeted social safety net” that was supposed to mitigate the impact of increased energy prices on low-income people.
In a loan to Mongolia approved in May 2017, the IMF insisted on better targeting of the national universal child cash benefit to replace a previous universal benefit. According to an analysis by two UN agencies, the programme was particularly effective in reducing poverty among Mongolia’s rural families, many of whom had not received benefits from an earlier selective scheme due to “exclusion errors associated with proxy means testing”.
While making clear that the IMF disassociates itself from agencies that support so-called rights-based approaches to social protection, neither the IEO report nor the social safeguards paper attempt to determine whether support for narrow targeting has been effective in reducing poverty and inequality, despite the attention the Fund has given to these issues in pronouncements. The IEO declines, because of “conceptual and measurement problems”, to assess whether IMF programmes led to increased social protection, as Fund officials have frequently claimed.
A preferred IMF tool in developing countries has been to include “indicative targets” such as social spending floors in lending programmes. Although the targets appear in the conditionality tables of IMF loan reports, they are not binding and thus have no impact on loan disbursements when not attained (as was the case in Tunisia in 2015), contrary to hard conditions on items such as total government spending ceilings or deficit floors. Interestingly, the social safeguards paper mentions that 20 per cent of IMF mission chiefs supported converting social spending targets into binding loan conditionality, but neither the policy paper nor the IEO report retain this suggestion.
Finally, without making an explicit recommendation, the IEO report seems to question the wisdom of the IMF having supported the Sustainable Development Goals: “The IMF’s endorsement of the SDGs has raised questions about consistency with its continued support for targeted (means-tested) social protection schemes.” It may be recalled that SDG target 1.3 includes the goal of social protection for all.