The IMF tried last year to counter long-running accusations that its lending and advice programmes damage health outcomes in developing countries, but the independent evidence points in the opposite direction. The question is whether the IMF will use this year’s reviews of its conditionality and of lending facilities for low-income countries (LICs) to switch approach and start helping Sustainable Development Goal (SDG) three to “ensure healthy lives and promote well-being for all at all ages.”
An IMF blog from March last year claimed that: “A number of studies have found that IMF support for countries’ reforms, on average, either preserve or increase public health spending.” However, the evidence provided was weak. Of the six studies referenced, one, by Oxford and Cambridge university researchers, which we discuss below, flatly contradicts this claim. Two were not related to health expenditure: one looked at revenue, not expenditure, and the second had a broader remit and contained no new evidence on the IMF and health. One was over a decade old and did not directly support the claim, while another was a link to an IMF page on the Ebola crisis. Therefore the only actual study that was referenced that supported the claim was written by the staff who authored the blog.
In another blog, the IMF Managing Director, Christine Lagarde stated that “our latest research shows that health and education spending have typically been protected in low-income country programs” noting “the IMF’s strong commitment to protect health and education spending and the most vulnerable during challenging economic reforms.”
Again, it’s worth looking deeper into the evidence cited to back up this claim, which is provided by an internal review of the IMF’s programmes in (LICs). This examines IMF programmes since a change in the Fund’s policy in 2009 which mandated the use of conditionality to protect social spending in its dealings with LICs.
Essentially, the IMF is arguing that this policy change has had two impacts. First, the IMF can no longer be caricatured as a champion of austerity, as the internal review claims that “LIC programmes are broadly divided between those entailing fiscal expansion [more spending] and those seeking fiscal consolidation [cuts].” Second, the IMF claims that it has used conditionality to ring-fence “social spending” and also to insist on policy reforms - “if feasible and appropriate” - to protect “the most vulnerable” from “any adverse effects of programme measures”. Unfortunately, neither of these claims hold up well under scrutiny.
Firstly the internal review that states that “LIC programs are broadly divided between those entailing fiscal expansion and those seeking fiscal consolidation” is not supported by the evidence in that paper. This graph, taken from the paper, gives totals for the number of IMF programmes in LICs according to whether they mean cuts in expenditure or increases, with a small number that are ‘fiscally neutral’ having neither. This shows that 34 of the 68 programmes studied mandated cuts in government expenditure while fewer (29) supported fiscal expansion.
However, including all types of IMF programmes is misleading. It makes sense to compare the two main types of IMF programme alone: the Extended Credit Facility (for medium to long term lending) and the Standby Credit Facility (SCF) for shorter term lending. Here the expenditure the gap is quite large, with 28 programmes associated with budget cuts versus 17 with expansion. The other two programmes, excluded in this comparison, are the Policy Support Instrument– which is advisory and entails no IMF lending, and the Rapid Credit Facility – an emergency programme for countries in trouble, which unsurprisingly has a majority of programmes that increase expenditure.
The second problem is that the statistical analysis in the review of IMF programmes between 1988 and 2014, showing “no evidence that fiscal adjustment policies in LIC programmes come at the expense of health and education spending,” is disputed by independent experts. Researchers at Oxford and Cambridge Universities published a thorough review of the IMF’s claims, finding that “the methodological strategy employed in the IMF analysis is unsound.” The researchers ran their own analysis, covering the same years as the IMF study and found that “an additional year of IMF programme participation decreases health spending, on average, by 1.7 percentage points as a share of GDP.”
This chimed with an earlier study by the same researchers of IMF programmes between 1995 and 2015 in 16 West African countries. This study found that “IMF policy reforms reduce fiscal space for investment in health, limit staff expansion of doctors and nurses, and lead to budget execution challenges in health systems.” If we are to meet the SDGs we need to have exactly the opposite outcome: significantly increased funding on healthcare, estimated to be in the order of an additional $274 billion per year by 2030.
The IMF has also faced criticism for its policies and conditionality aimed at protecting the “most vulnerable,” in particular over its approach to ‘social protection’ policies, which aim to reduce risks and vulnerability, and include the provision of welfare services to vulnerable groups, pensions and unemployment insurance. Such programmes also affect gender equality where the Fund has promised to deepen its commitment, even if critics argue it has a long way to go. Social protection programmes are important for gender equality not just because women and girls are likely to benefit from those programmes, but also because without them, the burden of caring for vulnerable groups such as the young or old is largely borne by women.
A coalition of civil society organisations (CSOs) wrote to the IMF in October 2017 saying that “the IMF’s approach towards social protection has been principally oriented around the desire to reduce social protection coverage and contain expenditure, rather than ensuring adequate levels of protection for all.” The fundamental issue is the IMF’s preference for targeting social protection spending towards sub-groups of the population.
A recent review by the International Labour Organisation, The South Centre and Columbia University “found that the targeting approach was considered in policy dialogue between the IMF and 68 developing countries”. The review goes on to list many reasons why targeting is not the best approach, particularly in developing countries where a large proportion of the population are living in poverty. These include adding to costs and administrative burden and creating two tier systems, in addition to leading to under-coverage, meaning many of the most vulnerable may be excluded. For example, the proxy means-testing (PMT) approach, championed by the World Bank, is beset by ‘exclusion errors’ which can mean that from 50% to 93% of the desired target group do not receive the proper benefits, research shows. A recent World Bank paper, analysing the PMT approach in social programmes in nine Sub-Saharan African countries, found average exclusion errors of 81 per cent.
Despite this major failing of targeted approaches, the IMF keeps pushing targeted social programmes through its loan conditionality and policy advice. For instance, in Kyrgyzstan when the government adopted a law introducing universal child benefits, the IMF continued to advocate for targeting, despite the poor performance of PMT-methods in the country. In Mongolia, the government was satisfied with its universal child benefit programme but, critics argue, confronted with the potential of the interruption of loan disbursements, it eventually accepted a loan condition on targeting child benefits. More recently, people took the street in Iran to protest IMF-backed reforms that would end the universal cash transfer scheme.
The CSO coalition also notes that the IMF’s targeting approach undermines global agreements and international law that social protection should be universal –and rights based – available to all as a right, and therefore should not be threatened when IMF programmes mandate reductions in expenditure.
The CSO letter came in response to a report by the IMF’s arms-length evaluation unit, Independent Evaluation Office (IEO), which found that the IMF’s increased attention to social protection over the past decade had met with “mixed success in implementation” and “at times it seemed that attention to social protection in surveillance devolved into a box-ticking exercise.” In response to this report the IMF published an implementation plan, which promises a new policy or ‘institutional view’ on social protection by early 2019. However, this plan already signals that the IMF’s support for targeting is not up for discussion, by “noting that targeting of scarce resources may be required to fill existing education, health and social protection gaps among disadvantaged groups.”
The IMF’s increased attention to social protection issues and its concern not to be seen impacting health expenditure in the poorest countries could be viewed as an improvement in the IMF’s approach. However, it seems clear that IMF conditionality does constrain expenditure on health and other related services, and is at odds with global efforts to achieve universal, rights-based social protection provision.
The next scheduled review of how IMF funding to low-income countries operates is planned this year. Unfortunately, judging by the questions posed in a public consultation last year, the IMF review may be missing the point. The impacts on health and other social expenditure arise not primarily because of the access to IMF financing – which the review focuses on – but on the conditionality attached to that financing. More promisingly, the IMF 2018 Executive Board work programme also promises a review of conditionality, but , as yet, there is no public information on the scope of this review.
It is time for a much broader reform of IMF conditionality. We hope that the Fund’s upcoming review has a broad scope, with a full and fair consultation process. Eurodad’s detailed study, in 2014, found that IMF conditions are often highly controversial and intrusive on key economic policy issues that should be the crux of democratic debate in country, not mandated from Washington. Crucially, we also found that “almost all the countries [that had IMF lending programmes during the period studied] were repeat borrowers from the IMF, suggesting that the IMF is propping up governments with unsustainable debt levels.” Yet the IMF is supposed to provide temporary support for countries facing balance of payments problems, not to become deeply engaged in policy making, affecting all areas, including health.
The IMF would do well to consider extending the approach of its Flexible Credit Line to all IMF facilities – requiring no conditionality other than the repayment of the loans on the terms agreed. This would remove the conditionality that is at the heart of so much of the damage that IMF lending can do in developing countries. This would allow more space to focus on the real solutions to increase fiscal space for health and social protection, such as the creation of an independent debt work-out mechanism to assess and cancel the unsustainable and illegitimate debt that holds many countries back.