- Charalambos Christofides, Atish Ghosh, Uma Ramakrishnan, Alun Thomas, Laura Papi, Juan Zalduendo, and Jun Kim
- Published Date:
- September 2005
IMF-supported programs require an analytical basis to ensure that policy advice is coherent, that conditionality comprises measures that are critical for program success, and that the intended objectives will indeed be achieved. While the analytical challenges of setting coherent policies and goals are not unique to IMF-supported programs—national authorities face such challenges on a daily basis, and the IMF always strives to ensure that its policy advice is coherent and apt—they take an added importance in the context of IMF-supported programs for a number of reasons. First, members typically seek IMF support at a times of crisis or when significant external adjustment is required—that is, when a policy response (usually different from the country’s previous policies) is required. Second, at such times the economy may be experiencing shifts in existing economic relationships, thus complicating the formulation of economic policies and making it more likely that policymakers will want to draw upon the IMF’s advice and expertise. Third, the IMF needs to assess whether the authorities’ policy program is likely to achieve its intended objectives, both to ensure that the member country addresses its economic problems and to help safeguard IMF resources.
Therefore, while this paper pertains to economic policy setting in general, it focuses on those economic programs in which a country’s authorities request the use of IMF resources.1 The first part of this paper lays out the process of program design and briefly describes some of the analytical tools—including the financial programming framework, the balance sheet approach, and the debt sustainability template—employed by IMF country teams in advising national authorities on policy formulation. The second part of this paper seeks to assess how well this process works in practice. Since this is difficult to do directly, the approach taken here is to examine whether there are systematic errors in program projections of key near-term macroeconomic variables—output, inflation, and the current account balance—and in the relationships between policies and targets implicit in the design of IMF-supported programs. Turning to longer horizons, the paper examines the record on program projections of real GDP growth and external debt dynamics. It bears emphasizing that the record on the quality of program projections sheds light on how well the analytical tools and approaches to policy formulation work as a modeling process—not on whether program objectives themselves were appropriate and achieved.2
The main conclusions are as follows. First, no single model or framework is universally applicable—policy formulation relies on a variety of models, techniques, and economic judgment. A key feature of this eclectic approach is its adaptability to evolving economic conditions, with program reviews providing an opportunity to reassess policies subject to conditionality (see Mussa and Savastano, 1999). In this regard, the role of financial programming is to inform and tie together projections of individual sectors (external, monetary, fiscal) into a coherent macroeconomic framework, rather than to pin down precisely the parameters of the financial program. Second, capital account crises pose challenging analytical problems and the balance sheet approach can help to assess the potential magnitude of capital flows and their implications for the efficacy of policy instruments. Third, analytical tools for understanding the factors driving sustained output growth are limited, and such tools as do exist are not always fully utilized in program design. For example, greater use of cross-country growth models could be helpful in informing and disciplining medium-term growth projections. Fourth, the IMF’s debt sustainability template complements the macroeconomic projections underlying IMF-supported programs by articulating their implications for debt dynamics and subjecting these dynamics to systematic stress testing exercises.
These conclusions are reflected in the record of program projections. Projections over the short term are relatively accurate and do not exhibit systematic biases with respect to inflation or output growth (except in capital account crises). This is important in that policies in IMF-supported programs are seldom formulated for more than a few months without an opportunity for revision at the time of quarterly or semi-annual reviews. Moreover, the relationships between macroeconomic instruments and targets assumed in programs are generally consistent with the actual relationships. At longer horizons, however, the quality of projections deteriorates markedly, with systematic biases in long-run growth projections that can, in turn, undermine assessments of debt sustainability.
The plan of this paper is as follows. Section II discusses the process of program design and the analytical tools used to help set macroeconomic and structural policies to achieve program objectives. Section III turns to the record on projections of key macro-economic variables and on relationships between policies and targets assumed in program design. Section IV concludes.