II. Methodological Problems and the Counterfactual Argument
- International Monetary Fund
- Published Date:
- September 1986
Any economic study confronts methodological issues which limit its application and the conclusions it can offer, but by its very nature this study must be more limited than most. The scope of the issue is as broad as it is complex, encompassing the welfare of entire populations in many diverse countries, while by its primary focus on developing countries the data base is extremely limited. An empirical study was regarded as the most desirable approach but, for the reasons outlined below and given in Appendix I, had to be rejected as impractical. Instead, the study reviews the measures which have been undertaken in the context of 94 Fund-supported adjustment programs during the period 1980–84 and examines their effects on various socioeconomic groups to determine resulting implications for internal income distribution (Appendix III). In this way a link is retained between macroeconomic theory and actual program experience without attempting to resolve intractable problems of quantification.
The section begins by reviewing the methodological problems which confront any study of income distribution and then provides some perspective for the extended discussions of economic policies in Fund-supported adjustment programs that follow by first describing the typical (composite) circumstances of countries which have a balance of payments disequilibrium and have approached the Fund to support a stabilization program. It then concludes with a discussion of the circumstances which are likely to arise (the counterfactual argument) should a country not undertake a program or adjustment.
Conceptual Problems in the Study of Income Redistribution
A variety of issues have to be overcome before a truly adequate assessment of the effects of a Fund adjustment program on the distribution of income in a country can be undertaken. Most of these individually are sufficient to preclude a thoroughly satisfactory examination of the issue, but collectively they severely limit the scope of any study on this important topic.
First, and most important, one needs to define what is the relevant “state of the world” with which the distributional outcome under the Fund program is to be compared. Is it the situation before the program, as it would have been in the absence of the program, or in the light of an alternative policy regime? The existing income distribution of a country seeking balance of payments support from the Fund is itself the product of the prevailing economic disequilibrium and is unsustainable over the longer term. Hence, simple comparisons between preprogram and postprogram distribution are not useful for evaluating the distributional impact of Fund-supported adjustment programs. However defined, the evolution of both the level of income and its distribution in response to an alternative set of events is the only relevant basis for comparison with the distributional situation that emerges under a Fund program. One cannot simply examine the effects of a Fund program in isolation from such a comparison. Implicit in such a comparison is the idea that the set of policies which constitutes the Fund program can be defined in contrast to those policies which would prevail in the relevant counterfactual situation. One would also need to isolate the effects of the policy regime negotiated by the country with the Fund in contrast to the effects of other factors or policies that occurred independent of the Fund program.
A related issue in the scope of the analysis is that no Fund-supported policies take place in isolation; numerous variables are changing simultaneously and most changes affect the size and direction of other changes. Is it analytically meaningful to view the impact of taxation in isolation from that of expenditure, or to analyze the effects of various individual components of taxation as if they operate independently of each other? In essence, the evaluation of every adjustment policy requires the solution of a general equilibrium model. This is simply impossible for most countries and certainly impractical for most countries with Fund-supported programs.
Second, one would need to explore not only the short-term effects but those that emerged over the medium term. By focusing on resource allocation rather than on income distribution per se, the impact of traditional Fund-supported adjustment policies necessarily has longer-term implications than any simple income transfer, just as the adverse effects of factor-price distortions may manifest themselves slowly. Given the adverse circumstances under which countries generally feel compelled to initiate negotiations with the Fund, it is likely that economic adjustment will have negative effects on real incomes in the short run. The relevant focus of examination should be appropriate distribution over a realistic time path of adjustment. As most Fund programs are short term (and extended facilities only focus on the medium term), this discussion must correspondingly focus on that time frame despite the relevance of longer-term factors.
Third, Fund programs seek to achieve adjustment in a country’s external accounts, but with the expectation that the policies pursued will have a favorable impact, over the medium term, on the level of output and the rate of growth. While the focus of most criticisms of Fund-supported adjustment programs has been directed at the effect on income distribution, the real issue is the effect of these programs on the sense of “well-being” among different socioeconomic groups in program countries. As such, issues relating to the income distribution should be viewed as an important factor in assessing the effects of Fund-supported adjustment programs, but not necessarily as an end in itself. Distributional issues also must be viewed in the context of achieving an efficient use of resources and satisfactory rate of economic growth. A more even income distribution may be unacceptable if it entails a substantially lower level of economic activity for the society as a whole.
Fourth, the assessment should provide some measure of the effects of the program on the well-being of different economic groups. No one distributional classification is fully adequate to encompass the concept of poverty or “the poor” for all purposes. The paper pursues distributional issues by looking at six main types of classification: (1) by each individual in the society; (2) by households; (3) by “institutional distributions,” especially between government, firms, and households; (4) by the division between the rural and urban sectors; (5) by the type of employment individuals undertake, such as for large and small farmers, or self-employed; and (6) by factors of production, for example, labor and capital. On the whole, the paper tries to specify more precisely those who constitute the poor (e.g., the rural poor, the urban unemployed) as this seems more useful than any general concept of poverty, which can quickly decay into disputes about relative poverty. Thus in what follows, the economic policies in Fund-supported programs are analyzed in the context of the six categories identified above. However, to contain the discussion, only the most relevant comparisons are selected in each case.
Finally, the analysis should be empirical. Ideally, one would like to quantify the effects of actual Fund programs in recent years, but when the data requirements for such an analysis are considered the statistical barriers become obvious. Heuristically, one would want to trace the effects of the various income and relative price effects associated with specific policies (price increases, tax increases, etc.) and need, at least, the data to evaluate the relative importance of such changes for different income groups. At an absolute minimum, a detailed income and expenditure survey for urban and rural households for the country in question would be required, referring to a time reasonably close to the point at which a Fund program was negotiated. Such a survey would need to include sufficient data to provide a picture of the preprogram distribution of income, both monetary and nonmonetary. To evaluate the effects of the particular policy instruments in a Fund program, such a survey would also need to contain disaggregated data on the pattern of consumption expenditure and the sources of income (including government subsidies) by income group.
Even such a data base would have enormous gaps. To appraise the effects of any cutbacks in government expenditure, one would need data on the consumption of public services by income group and the distribution of government and public sector enterprise employees by income group. Such a comprehensive data base is unlikely to be found in any country. Even the house-hold income and expenditure surveys that are periodically undertaken in developing countries are likely to be far less than adequate for the analytical task at hand. Such surveys are generally done by the national statistical office of a country with specific objectives in mind, such as the compilation of a consumer price index or of specific components of the national accounts; common problems in such surveys are the underrepresentative nature of the sample surveys used, the frequent underreporting of income (particularly property income and nonmonetary income derived from the subsistence sector), excessive aggregation in the consumption categories included in the survey, and skewed underreporting in low-income budgets. No data are likely to be contained on public service consumption or public employment by income group.2 As noted by the World Bank,
- [the] narrow focus [of such surveys] severely limited their usefulness as a basis for the analysis of levels of living, and attempts to use the data for that purpose quickly reached a point of diminishing returns.3
A fortiori, it is even more unlikely that comparable statistical survey data would also be available at several time points during and after the period of adjustment under a Fund program. For only a few developing countries are comparable income distribution data available at more than one point in time; even where there have been multiple surveys of this kind, it is rare that the statistical survey procedures used are sufficiently similar so as to allow comparisons.4 Even if such information were available, it would be difficult to determine how much of any observed change should be imputed to the effects of the Fund adjustment program and how much to the effects of other factors affecting the economy since the last survey.5 Finally, even if one could make such an imputation and observe that the Fund program had negative effects on the income distribution, the relevant issue would then be how these results differ from those that would emerge in the absence of a Fund program. The counterfactual difference is, of course, not an observable phenomenon yet without it a true appraisal of the effects of Fund programs on the income distribution cannot be made.
A final problem is that even with adequate data it may be impossible to make conclusive statements comparing different income distributions. It is well known (Champernowne (1974)) that the most common methods of comparing income distributions do not lead to unambiguous conclusions except in exceptional circumstances. Moreover, the movement in prices during the course of the program will affect the consumption possibilities available to each individual, leading to further ambiguities with index comparisons.
Sketch of a Country About To Draw on the Fund
Countries that undertake a balance of payments adjustment program face disequilibria from a variety of causes. The imbalances may be due to exogenous factors or domestic structural imbalances with the resulting balance of payments difficulties only one aspect of the stabilization problem. Usually, the balance of payments pressures originate from some combination of the two causes, and the authorities must undertake a wide-ranging economic program if an acceptable level of economic performance is to be resumed and the balance of payments is to be financed in the medium term.
The imbalances arising from the foreign sector are at times due to factors beyond the authorities’ control (particularly in the short run), such as a deterioration in the terms of trade, which may stem from depressed export prices or sharp increases in the prices of major imports, such as petroleum. In other cases, the imbalance in the current account is the result of inappropriate policies pursued by the authorities over an extended period, such as an overvalued exchange rate which encourages the use of imports at the expense of domestic production (and discourages worker remittances from abroad), or unrealistically low producer prices (often related to an overvalued currency), leading to a stagnant export sector. In some cases the problem may be related to the capital account, and again the imbalance may evolve from recent developments or longer-term policies. A sudden reduction in aid flows may create a financing gap which the authorities cannot easily close, while a major development effort with little export response may lead to burdensome debt-servicing requirements. Furthermore, foreign exchange reserves may be so low that any shortfall in the external sector has severe and immediate repercussions on import flows.
Domestic imbalances may arise also from factors outside the authorities’ control, such as those caused by unusual weather patterns which suddenly reduce domestic food output. Often, however, they arise from factors that are subject to the authorities’ control, as when inappropriate monetary and fiscal policies increase domestic demand to unsustainable levels.
In most cases domestic fiscal policies are the cause of, or at least exacerbate, the imbalances which emerge in the external sector. Large government deficits, normally financed mainly by the domestic banking system, lead to domestic absorption levels which may place excessive pressure on the indigenous productive capacity, resulting in increased imports and the rationing of available goods through higher prices or the creation of queues.
A country that has an economy in which some or many of the conditions identified above exist may seek to use Fund resources to help finance a short-run balance of payments gap while measures are being implemented that will move the economy toward a more sustainable medium-run balance of payments. Indeed, the majority of countries undertaking Fund-supported adjustment programs in recent years have exhibited, in varying degrees, a substantial number of these problems. If the balance of payments disequilibria result from terms of trade or other exogenous developments outside the authorities’ control (e.g., bad weather), which are expected to be temporary, access to Fund resources may be limited to drawings under the compensatory financing facility. More often the balance of payments imbalance is due to inadequate demand and supply policies and an adjustment program is warranted. The program may focus on either demand management issues or structural reforms, and frequently has elements of both. In either case, the primary objective of a Fund program is to provide for the resumption of a viable balance of payments position and a reduction of inflation, both of which are needed to ensure a more satisfactory and sustainable rate of economic growth over the medium term.
The Counterfactual Argument
It may be helpful to examine the likely developments in a country experiencing economic disequilibria that decides not to pursue an adjustment program supported by the use of Fund resources. The following counter-factual discussion provides some insight into the tradeoff between economic equity and growth. It should be stressed, however, that while some countries have slipped far down the spiral which is depicted here, many others have realized the self-defeating nature of poor economic policies and have arrested this decline by adopting more pragmatic goals. The counterfactual argument presented here should be viewed as a “worst case” scenario that illustrates the need for countries to adopt realistic economic policies and the likely consequences if they do not.
The Social Choice
Perhaps the most natural starting point is to assume that the authorities have, for what seem compelling reasons, determined that the government should intervene in the economy to implement social policies more in accordance with their objectives. In developing countries, these motives usually begin with the understandable desire to improve the material and social well-being of the citizens, but it may also originate from an intellectual debate over the future course of the economy, the desire to protect its citizens from a worsening in the terms of trade, or even the need to encourage popular support for the existing government. In any case, the decision to intervene in the market can be expected to result in some form of price controls and a larger public sector. If, as is usually the case, there is also emphasis on diversifying the economy and increasing the domestic value added, there will also be higher public sector investment to develop local industrial capacity.
These decisions may touch off a series of second round effects which can quickly exacerbate the existing economic disequilibria instead of containing them. Expansion of the public sector, increased public sector investment, and the regulation of prices can lead to an expansion of domestic absorption which the economy may not be able to satisfy. For example, expanding the civil service to monitor an inefficient system of prices is unlikely to stimulate production but will have the immediate effect of increasing government expenditure. An ambitious public sector investment pro-gram may in time produce a reasonable return if the projects are well conceived, well managed, and have access to the necessary inputs to operate efficiently, but these conditions are often absent. A complicating factor with respect to input availability is that often these projects are undertaken with external financial resources to protect existing domestic investment levels, but by the time they are completed the external position has deteriorated (for reasons discussed below) to the point that debt-servicing charges compete with the importation of required inputs for the increasingly scarce supply of foreign exchange. In regard to regulated prices in general, artificially low consumer prices will invariably stimulate the consumption of “cheap” commodities.
Need for Economic Control
The imposition of widespread price controls is not likely to cause severe economic disruptions immediately, but in the long term a pricing system which departs significantly from underlying market forces is likely to seriously misallocate resources. Normally, the thrust of these policies is to maintain consumer purchasing power and to suppress inflationary pressures, but the retention of prices below market-clearing levels either chokes off supplies, as producers do not receive adequate incentives, or exacerbates budgetary pressure, by subsidizing production. The desire to protect the consumer almost always extends to the external sector through an overvalued exchange rate. As long as foreign sources of capital are available, the demand for goods is met through a deficit in the current account, often rationed by quantitative restrictions.
The price distortions caused under these conditions can permeate the economy. Domestically, as agricultural incentives deteriorate, market agricultural activity may die in those markets where prices are artificially controlled by the government. Countries which have provisioned the market at subsidized market-clearing prices when resources are abundant resort to under-provisioning the official markets as imports dwindle and resources become scarce, with the result that consumers turn from the underprovisioned official market and pay market-clearing prices in unofficial markets. Commonly under these circumstances, exclusive facilities, open only to members of select groups (for example, civil servants, military, and workers in select industries), continue to be adequately stocked at a level consistent with price-clearing at the administered, subsidized price. Thus better-off consumers with employment in the official or formal sectors purchase products at subsidized prices while those in the informal sector are forced to purchase at “black market” prices.
Medium-Term Developments Under Economic Controls
Over time, the initial irregularities resulting from market intervention become chronic, and the policies to compensate for these first-order structural rigidities become more and more extreme. Usually they manifest themselves in increasingly expansionary fiscal policies as the need to subsidize normal economic activities increases and revenues stagnate. The overvalued exchange rate dampens exports, worsening the scarcity of foreign exchange and transmitting the external imbalance throughout the economy. Although short-lived improvements in the terms of trade may mask the adverse external sector developments temporarily, a widening gap between the official exchange rate and the underlying market-clearing rate eventually creates a balance of payments crisis which the authorities must address. If the decision is to avoid major exchange rate adjustments, the authorities may react to the noncompetitiveness of their country’s exports by offering export rebate schemes to stimulate domestic production and to accommodate producer price increases. Such direct subsidies may prove useful in expanding exports, but their implications for the domestic economy are substantially less favorable. Under a more appropriate exchange rate, the export sector could be expected to contribute resources to the budget rather than requiring subsidization. If the primary export commodities are also the main sources of domestic food supply, the final result may be still larger budgetary subsidies. The authorities may attempt to increase external borrowing for these purposes, which will alleviate the immediate shortage of foreign exchange at the cost of a larger debt-servicing problem in the future.
Usually, a shortage of foreign exchange precipitates a series of restrictions on imports, which works to equate supply and demand of traded goods by imposing a scarcity premium on imports and thus lowering demand. In addition to the interesting distributional issue of who benefits from the scarcity premium, this approach also has important budgetary implications, as import duty receipts reflect a constrained volume of imports but not the implicit price of these imports. Much of the reduction in import volume is likely to be in inputs for domestic production, the decline of which further restricts excise and income tax collections as it contributes to domestic demand pressures.
The budget thus bears the brunt of official attempts to avoid structural adjustment. This reliance is one of the principal reasons behind concerns with the fiscal position in Fund-supported programs to restore external stability. But the financial system also shares in the burden; the rising budget deficit necessitates additional bank financing, contributing to monetary expansion. The development of parallel markets to relieve pent-up demand substantially accelerates the underlying inflation rate, leading to increased wage demands and the public’s desire to hold currency (the only medium of exchange in these markets) rather than demand deposits.6 As inflation increases, real interest rates are lowered still further making deposit rates unattractive, access to credit very important, and increasing the opportunity for corruption.
If the gap between the official and a market-determined price system continues to widen, government intervention in official markets will become more pervasive. The unrealistic prices in these markets will ultimately discourage economic activity in the markets in favor of a flourishing system of parallel markets where goods are more readily available, albeit at a significantly higher price. The decline in business activity in official markets will erode budgetary revenues even further as many tax bases (e.g., for corporate income, excises, and sales taxes) contract. As individuals find the parallel market more and more attractive, they have less interest in reporting the true magnitude of their incomes for tax purposes or in channeling financial resources through official markets for fear of having their activities revealed. The combination of stagnant export levels and large external borrowing leads to a sharply higher debt-servicing ratio, with the likely development that the central bank does not have an adequate supply of foreign exchange to supply the foreign exchange needs of the country, and the country begins to incur external arrears. The growing imbalance between government revenue and expenditure needs may even induce the emergence of domestic arrears, although a more likely vent for the excess demand will be still larger domestic bank financing. Heavy recourse to domestic bank borrowing for budgetary financing serves to transmit the adjustment back to the private sector in the form of a currency which has been effectively devalued, whether the government chooses to recognize it or not.
With the slowdown in activity in official markets, the related price controls, and rapid inflation, hoarding goods or foreign exchange, which ultimately further constricts domestic supply, becomes a more profitable form of investment than investment in plant and equipment. In time, overall business activity shrinks, the public becomes reluctant to hold domestic currency, and, in the extreme, the country gravitates toward a barter economy.
Stability is achieved under these circumstances, if at all, through rapid migration of human, financial, and mobile agricultural resources. The initial effect of policies to support imports through an overvalued exchange rate and budget subsidies is to encourage urban migration. The resources for supporting the urban sector, both in domestic currency and foreign exchange, come from a combination of an explicit and implicit (exchange rate) tax on major, generally extractive, export industries with a strong comparative advantage. Where such industry exists the economy continues on a downward spiral and the agricultural sector is essentially abandoned; gradually the number of people that can be subsidized from the shrinking national product diminishes. Over time, the failure of these policies can lead to a second migration as individuals seek to maintain an acceptable standard of living by returning to a subsistence economy that at least provides food.
Adjustment has taken place, in the sense that the system the authorities attempted to maintain was untenable, but it is uncontrolled adjustment, and not the most efficient use of resources the economy could achieve. The income distribution cost through this adjustment may have been spread equally, but this is an equality of poverty achieved by great sacrifice in the overall level of economic activity, and it is likely that most of the population is worse off than it would have been under a program which would have permitted more growth. The policy to establish a predetermined pattern of income can thus be viewed as another type of expenditure program, albeit a very expensive one, and one which extracts a contribution from the poor as well as from the rich in the form of a lower absolute level of income.
Of course, as discussed at the beginning of this section, by no means do all countries manage their affairs as badly. There are many resting places on the downward spiral, but there are sufficient worldwide parallels to give substance to this description of adjustment without a Fund-supported program. It is against this alternative and with these possibilities in mind that actual Fund-supported programs and policies should be considered.