This paper has made a comprehensive appraisal of the arguments surrounding, and the empirical evidence on, the global effects of Fund-supported adjustment programs. The paper’s main findings can perhaps best be summarized as follows.
First, in assessing not only the size but even the direction of program effects, it is important to recognize that alternative definitions of program effects can yield markedly different results. The review of five alter-native definitions found that measured program effects can vary substantially depending on seven different considerations: (1) whether changes in non-program factors (such as economic activity in the industrial countries, world interest rates and world oil prices, or weather conditions in program countries) between the pre-program and program period are taken into ac-count; (2) whether program targets incorporate accurate forecasts of the global economic environment during the program period; (3) whether program countries are systematically different from non-program countries prior to the program period in ways that matter for subsequent performance; (4) whether non-program countries are themselves indirectly affected by Fund programs; (5) whether the medium- and longrun (beyond-one-year) effects of programs are considered in addition to the impact of short-run effects; (6) whether, because of confidence and credibility, the imposition of a given policy within the context of a Fund program has different effects than without it; and (7) whether the most relevant comparison for actual results under a Fund program is what would have happened in its absence or instead, what could have happened under some (hypothetical) optimal set of policies. All these suggest that one reason the evaluation of Fund-supported programs has produced such widely varying verdicts is that different judges have often applied different “yardsticks” to the same data.
The Fund’s own interpretation of the measurement of program effects can perhaps best be summarized as follows: (1) actual outturns in program countries should be compared with what would have happened in these countries in the absence of Fund programs; (2) in forming a judgment about what would have happened in the absence of programs, the Fund’s direct and indirect catalytic role in providing additional finance to program countries must be considered; this means that any compression of expenditure and imports during the program period needs to be weighed against the (larger) expenditure and import changes that are likely to have occurred in the absence of this program-induced financing; (3) Fund-supported programs should be characterized by the full range of policy measures included in typical past programs and not just by government budget targets, domestic credit ceilings, and exchange rate changes alone; similarly, given the adverse initial position of most program countries and the failures of policy in the pre-program period, it should be recognized that credibility and confidence may produce a different result from a given policy package within the context of a Fund program than without it; and (4) the effects of Fund programs, particularly on growth, should be assessed in the medium- to long-term (certainly over more than a year) rather than in the short run; an excessively short-run framework will almost inevitably exclude any positive growth effects of supply-side and structural measures in programs, and will make it very difficult to distinguish between the adjustment to a sustainable internal and external position and that sustainable position itself.
In short, by comparing what is versus what would have been, by considering the influence of the avail-ability of foreign exchange on import decisions in program countries, by accounting for the full range of policy changes in programs as well as for the contribution of programs to the credibility of those proposed policy changes, and by adopting a medium-term rather than a short-term true horizon, the Fund has come to a rather different assessment of program effects than some other observers. Since assessment of the domestic effects of Fund-supported programs is a logically prior input to the assessment of the global effects of programs, it follows that the Fund’s appraisal of the size and even the sign of such global effects of programs also often differs from that of some others.
Third, to the extent that Fund-supported programs do lead to changes in expenditure, output, and import volumes in program countries, one should expect these changes to induce changes in these same macroeconomic variables in the rest of the world, and in the same direction. After all, one should not expect cross-country expenditure multipliers to cease operating just because the initiating changes in expenditure or imports results from a Fund program. But perhaps the more telling point is that even with roughly 35 developing countries undertaking programs supported by the Fund, the size of these global trade and economic activity effects is likely to be relatively small. This is because: (1) program countries still account for rather small shares (7–8 percent) of world imports and of world exports; (2) relatively little (8–9 percent) of program countries’ trade is with other program countries; and (3) the average size of the initiating changes in import volumes in program countries has been rather small (3–6 percent) over the past decade. Indeed, even when the “later-round” effects of changes in the imports of program countries are considered, and when marginal trade propensities replace average propensities—as revealed in simulation exercises with several global trade models—the global effects of expenditure changes in program countries remain limited. To take a specific example, results from both the OECD Interlink Model and the Project LINK World Trade Model suggest that even a 15 percent decline in the imports of (1983) program countries—a figure twice as large as that recorded by program countries in 1983 and almost three times as large as the average for 1973–82—would lead after a year to perhaps a 0.3–0.5 percent decline in real GNP in the OECD countries. In the end, all analyses of the global effects of programs must face the twin realities that expenditure changes in the North (i.e., in the industrial countries) have by far the greatest impact on global economic activity, and that domestic expenditure multipliers are typically much larger than cross-country ones. This means that, for better or for worse, the primary impact of Fund programs falls on the countries themselves, and that serious, or potentially serious, global effects will be the exception rather than the rule. None of this of course denies that the external effects of Fund-supported programs on individual countries, industries, firms, or perhaps even regions, can be serious. Further, if the program-country group were to become considerably larger in the future, the global effects of programs would certainly need to be reassessed.
Fourth, the concern that the simultaneous exchange rate depreciations by program countries could have serious and deleterious effects on prices of the exports of program countries is certainly a reasonable one in theory. Indeed, economic theory suggests that for a homogeneous primary commodity, the ability of program countries to affect the prices of their exports depends (positively) on their ability to affect the world production or the world consumption of the good, and on the (absolute) value of their own price elasticities of supply and demand for the good. In practice, however, the risks of adverse aggregate effects on export prices are much reduced—for five reasons: (1) primary commodities now represent a much smaller share of the exports of non-oil developing countries than they did two decades ago; (2) non-oil developing countries now account for only about 30 percent of world exports of non-energy primary commodities; (3) supply-price elasticities for most primary commodities are rather low in the short run; (4) the share of program countries of world production of various primary commodities is much below the share for all developing countries; and (5) not all program countries change their exchange rates during the same period, and those that do will usually not export identical bundles of primary commodities. But a lower risk is not the same as no risk. There are some primary commodities (such as coffee, cocoa, natural rubber, or tin) for which world production is relatively highly concentrated in the top three or top five developing-country producers, and there are quite a few non-oil developing countries where exports are still highly concentrated in only a few primary commodities. In these countries with relatively high market power, aggregate price effects cannot be lightly dismissed. Also, “beggar-thy-neighbor” effects of exchange rate changes are not confined to primary commodities. For manufactured exports, simultaneous exchange rate depreciation by many producers can mean that each gets less of a gain in market share than expected, and that the devaluing group takes export volumes away from the passive group of manufacturing non-oil developing countries. For these reasons, the Fund needs to monitor closely the aggregation and interdependence effects of exchange rate changes by program countries. Equally important, the Fund needs to continue to urge the industrial countries to improve the access of developing countries to their large markets, for what-ever the effectiveness of devaluation in increasing export supply, the amount of developing-country exports actually sold will be the minimum of the demand for and the supply of these goods.
Fifth and finally, the raison d’etre of the Fund is precisely that the global, systemic, and third-country effects of the policies of individual countries do matter and that efforts must be made to ensure that countries with balance of payments problems seek solutions that are not destructive of the general welfare. Also, be-cause the Fund is offering advice on economic policies to almost all countries in the system, it must be concerned about the aggregate effects of that advice, as well as with its consistency and feasibility across countries. A number of procedures have therefore evolved in the Fund for the evaluation of such aggregation and interdependence effects. At the broadest level, these effects are analyzed by the staff and discussed by the Fund’s Executive Board at least twice each year during the World Economic Outlook exercises. Program countries are of course included in that analysis and discussion, albeit as part of more structural country-classification groups (major exporters of manufactures, low-income countries, or 25 major borrowing countries) rather than as a separate entity. At the level of individual consultation missions and program negotiations, cross-country effects are analyzed as part of normal mission preparation and of normal program design. The waivers and modifications pro-visions in Fund programs also represent a well-established mechanism for assessing, and if necessary, redressing the consequences of, inter alia, unforeseen interdependence effects. None of this means that the evaluation of the global effects of Fund-supported programs within the Fund staff itself is unerring. But it does suggest that serious efforts are being made to consider the repercussions of these Fund programs on other countries and on the operation of the world economy as a whole.