Exchanges on subsidy reform between policymakers and economic analysts, particularly analysts from international financial institutions (IFIs), have become somewhat stylized. The standard dance begins with a situation where the subsidy leads to a large fiscal deficit and an imperative to manage this deficit. Economic analysts from IFIs and elsewhere then do an incidence analysis of a subsidy, for example a fuel subsidy, which shows that it is very badly targeted toward the poor. The strong implication is that removing the subsidy will not impact poverty levels “too much.” But of course it will have some impact. It is then suggested by analysts that better targeted transfers that focus on the poor could mitigate the poverty impact.
Policymakers and politicians considering subsidy reform, and indeed any policy reform, are of course interested in the impact on poverty. But, equally, they are aware of and interested in the broader political economy of subsidy reform. More generally, they are interested in the full range of winners and losers from subsidy reform. To the extent that compensation is on the agenda, they are as interested in finding possible mechanisms for mitigating impacts on politically salient groups as they are in mitigating impacts on the poor. But here the analysis that is usually available to them—the typical incidence analysis of the subsidy, done with greater or lesser sophistication depending on data availability—does not help very much.
How can analysts respond to this key concern of policymakers? One answer is to still do incidence analysis, but of a different sort. Such incidence analysis would differ from the conventional type in two ways. First, it would look at incidence on socio-politically salient groups that go beyond just those in income poverty. Such groups include the middle classes, labor groups, regional groupings, and others. Second, the analysis would look at the incidence of not just removal of subsidies, but of alternative policy packages that include instruments beyond subsidies. The exercise would thus show the winners and losers from a range of fiscally equivalent packages.
Identifying the alternative instruments that go beyond targeted cash transfers to the poorest groups is most important: is it wage policy, or any other instruments? This raises the issue of the feasibility or not of reaching some groups who may be strongly affected by the subsidy reforms but are difficult to identify and target, and its political implications. It also raises the issue of the macroeconomic impact of using these alternative instruments such as wage policy.
Clearly, such an exercise would need close collaboration between policymakers and analysts to identify packages that are feasible from the implementation standpoint, and to identify socio-politically salient groups. But once these are identified, the analysts can use their expertise to conduct incidence analysis that can be modified as the discussion evolves. Such a process would surely be more productive than the stalemate we see in many countries now.
Other questions where analysts can be helpful involve the design of the subsidy reform itself. For instance, should policymakers go for a “big bang” or a more gradual approach and what are the macroeconomic implications when the size of the subsidies is quite large. The reform in Iran is a case where the macroeconomic implications led eventually to the failure of a major policy experiment. Another question is whether the price reform should aim to keep relative prices consistent with the structure of international prices or accept a distortionary structure by adjusting less aggressively the prices of commodities consumed more by the poor. What are the equity-efficiency trade-offs?
One additional problem often ignored is that of the cross-border implications of subsidy reforms. When neighboring countries have large subsidy programs for the same commodities, reforms in one country may have significant implications in terms of smuggling and cross-border trade. These are of concern to policymakers, but are usually not part of standard analysis of subsidy reform, which is done on a country-by-country basis.
To be sure, there exist some nascent analyses along the lines suggested here. What is needed is a more systematic integration of policymakers’ political-economy concerns into the typical subsidy reform analysis.
Homi Kharas delivered the keynote address at IFPRI’s annual staff retreat on September 12, 2018. He explored the evolving development agenda and its implications for policy research.