A/HRC/38/33 46. Third, proposing to replace one social protection system (subsidies) with more “efficient” systems starts from the assumption that highly complex and politically sensitive systems can be successfully created on a drawing board. However, experience shows that successful rights-respecting systems are the outcome of “contestation, be it from the electorate, from competition between political parties, or from public, civil society pressure”.107 47. Finally, fiscal savings are one of the key attractions of subsidy reform, for IMF and for Governments. Given the strong preference of IMF for macrostability, which favours minimizing public expenditures, and most Governments’ preference to use savings for more politically popular programmes, it is unsurprising that many subsidy reforms lead to significant net reductions in social protection spending.108 48. None of this is to suggest that energy subsidy reform is undesirable, but that serious efforts to establish alternative social protection systems require a more concerted and committed approach than has generally been used to date. IMF is well aware of the obstacles to far-reaching reforms, 109 but instead of embracing a politically and socially sustainable social protection policy, it emphasizes communications strategies, sequencing and “depoliticization” as solutions. In other words, rather than tackling the more complex substantive issues, the challenge is seen largely as a question of marketing. Social spending floors 49. In theory, IMF has given strong recognition to social protection through the inclusion of “indicative targets” for social spending floors, especially in loan agreements for low-income countries. The goal is to maintain or increase public spending on education and health care and sometimes on social safety nets. Since 2009, about 90 per cent of programmes for low-income countries has included such targets. But IMF itself acknowledges severe shortcomings in that approach. Some targets are too broad to be meaningful (covering total spending on health and education, for example) or include spending on infrastructure and other budget lines not directly related to social protection. And memoranda of understanding between Governments and IMF “often lack details on coverage and monitoring, and on reporting arrangements”.110 50. External critiques have been even stronger. One study showed that social spending floors for sub-Saharan African countries since 2000 had been implemented only around 50 per cent of the time, with even lower success rates in West Africa. And even in contexts in which IMF balanced budget conditions are met, the accompanying social spending floors are generally not. 111 This may be because spending floors are non-binding indicative targets, while key macroeconomic targets in IMF loans are binding. While IMF claims that public spending on education increases in low-income countries that implement IMFsupported programmes, and that there is no statistically significant effect on health spending,112 those claims have been strongly contested.113 Critics have argued that the 2017 IMF Policy Paper is methodologically unsound,114 and that IMF programme participation 107 108 109 110 111 112 113 114 Gabriele Koehler, “The Politics of Rights-Based Transformative Social Policy in South and Southeast Asia”, in International Social Security Review (Vol. 70, Issue 4, 2017). p. 107. “For example, in Ghana, the eliminated fuel subsidy would have cost over US$1 billion in 2013, whereas the targeted LEAP programme costs only about US$20 million per year (where did the rest of savings go?).” Ortiz, Cummins, and Karunanethy, p. 7. In Mauritania, the elimination of food subsidies was expected to provide a windfall for the general budget, rather than for other social protection activities. See A/HRC/35/26/Add.1, para. 59. See Clements et al, Energy Subsidy Reform, ch. 4. IMF, “Social Safeguards”, pp. 13–14. Stubbs and Kentikelenis, “Targeted Social Safeguards in the Age of Universal Social Protection: the IMF and Health Systems of Low-Income Countries”, in Critical Public Health (Vol. 28, Issue 2, 2018), pp. 136–37. IMF, “Social Safeguards”, p. 11. Stubbs et al, “The IMF and Government Health Expenditure: A Response to Sanjeev Gupta”, in Social Science & Medicine (vol. 181, 2017), pp. 202–204. Stubbs and Kentikelenis, “Targeted Social Safeguards”, pp. 133–134. 15