Policy, Research, and External Affairs Socio-Economnic Data International Economics Department The World Bank August 1990 WPS 479 Conversion Factors A Discussion of Alternate Rates and Corresponding Weights Michael Hee Time series of alternative conversion factors and of correspond- ing weights provides the framework for estimating overall conversion factors that are analytically relevant and meaningful. The Policy. Research. and External Affairs Complex distributes PRE Working Papers to disseminate the findings of work in progress and to encourage the exchange of ideas among Bank staff and all others interested in development issues. These papers carry the names of the authors, reflect only their views, and should be used and cited accordingly. The findings. interpretations, and conclusions are the authors' own. They should not be attributed to the World Bank, its Board of Directors, its managanent, or any of its member countries. Policy, Research, and External Affairs "WORKING PAPERS Socio-Economic Data WPS 476 This paper - a product of the Socio-Economic D3ta Division, International Economics Department - is part of a largereffort in PRE toward a more versatile approach to estimating conversion factors for the World Bank Atlas and operational purposes. Copies are available free from the World Bank, 1818 H Street NW, Washington DC 20433. Please contact Estela Zamora, room S7-136, extension 33706 (59 pages, including tables). The significant operational implications underly- other rates (implicitly) have zero weights. A ing the Bank's estimates of per capita GNP major component of this framework is the represent important considerations in systematiz- redistribution of weights among multiple ex- ing the use of official and other exchange rates in change rates. determining the exchange rate to be used in the Bank's Atlas methodology. Hee explores the Such a matrix of conversion factors and potential for a system of time series for various corresponding weights could provide the mecha- conversion factors and a corresponding set of nism (1) for a systematic approach to weighting time series for weights. alternative rates; (2) for estimating the effects of alternative weighting schemes; (3) for determin- The framework is useful for countries with ing the effects of incorporating parallel or black multiple exchange rates. It gives us a way to market exchange rates; (4) for providing the develop time series on parallel and black market basis for less erratic and unpredictable fluctua- exchange rates, purchasing-power parities, trade- tions in the data in the World Bank Atlas and related taxes and subsidies, and potentially more. World Tables; and (5) for improved and more transparent documentation of methods and The starting premise is that a single official "special" cases. exchange rate has a weight of 1.0 in all years; all The PRE Working Paper Series disseminates the findings of work under way in the Bank's Policy. Research, and External Affairs Complex. An objective of the series is to get these findings out quickly, even if presentations are less than fully polished. The findings, interpretations, and conclusions in these papers do not necessarily represent official Bank policy. Produced by the PRE Dissemination Center TABLE OF CONTENTS SUMMARY AND CONCLUSIONS ....................................... i I. INTRODUCTION ................................................. 1 A System of Conversion Factors and Weights ............................... 1 Scope of Paper ..................................................... 4 Magnitudes of Deviations as Signals ..................................... 6 II. CONVERSION FACTORS a..d RELATIVE PRICE MOVEMENTS ............. 8 Conversion Factor asaPrice ........................................... 8 Conversion Factor and Price Levels of Traded/NonTraded Goods ............... 9 III. CONVERSION FACTORS and MULTIPLE EXCHANGE PRACTICES ......... 13 Effects of Multiple Currency Practices on Conversion Factors ................... 13 Incidence of Multiple Currency Practices .................................. 15 Deviations of M ultipleRates ........................................... 15 "Composite" Conversion Factor ......................................... 18 Countries for Review and Adjustment for Multiple Rates ...................... 21 IV. BLACK or PARALLEL MARKET EXCHANGE RATES .................... 22 Formal and Informal Parallel Exchange Markets ............................ 22 Purchasing Power Parity and Factors Affecting Parallel Market Rates ............. 23 Parallel Market Rates and Market Information ............................. 25 Deviations of Black MarketRates ...................................... 27 Countries for Review and Adjustment for Parallel Rates ....................... 28 V. PURCHASING POWER PARITY as CONVERSION FACTORS ............... 32 Purchasing Power Parity (PPP) ......................................... 32 PPP and Data Conversion ............................................. 34 What Price Measure to Use in PPP Computations ........................... 34 VI. TRADE and EXCHANGE TAXES and CONTROLS ....................... 37 Trade Taxes on Gross Output and Value-Added ............................. 38 Effects of Tariffs and Subsidies on Conversion Factors ........................ 39 Incidence of Trade Taxes and Exchange Controls ............................ 40 "Trade Tax-Effective" Conversion Factors .................................. 45 Deviations of "Trade Tax-Effective" Exchange Rates .......................... 48 Countries for Review and Adjustment for Trade Taxes ........................ 49 VII. VOLATILITY in EXCHANGE RATES and TRADE VOLUMES .............. 50 Merchandise-Weighted Conversion Factors ................................. 50 Deviations of Merchandise-Weighted Rates ................................ 51 Countries for Review and Adjustment for Volatility . ........................ 52 VIII. USE of PERIOD AVERAGE EXCHANGE RATES in the BANK ............. 55 Shorter Periods to Longer Period Rates ................................... 55 Search for a General Purpose Conversion Factor ............................ 57 BIBLIOGRAPHY ................................................... 58 SUMMARY AND CONCLUSIONS The significant operational implications underlying the Bank's estimates of per capita GNP, represent important considerations in systematizing our use of official (and other) rates in determining the exchange rate to be used in the Atlas methodology. The paper explores the potential for a system of time series for various conversion factors, and a corresponding set of time series for weights (Table 1). The International Economics Department (IEC) has expanded the availability of the repertoire of exchange rates; a major task in this system, however, is the redistribution of weights among multiple rates.The starting premise is that a single official rate has a weight of 1.0 in all years; all other rates have (implicitly) zero-weights. We can examine the issue of how the "composite" conversion factor would look like if such rates have non-zero weights. IEC's practice is to take into account such multiple exchange rates and to the extent data are available, estimates an appropriate conversion factor from a multitude of exchange rates prevailing in a country (see the examples of Egypt and Syria). In this connection, the paper reviews the main concerns about the adequacy of the official exchange rate as a general conversion factor when in effect countries maintain dual or multiple exchange practices, or there are active foreign exchanges in the parallel or black markets, or that there is a high incidence of price controls and trade-related restrictions. ii One aspect receiving special attention is the compilation of a general purpose (composite) conversion factor from a set of multiple exchange rates for different types of international transactions. The paper identifies the kinds of evidence that are relevant in assessing the impac. of these practices on exchange rates; provides illustrative examples of IEC using the evidence to derive alternative conversion factors (Tables 4 and 5), and concludes by identifying countries where review, and adjustment if necessary, of the official exchange rate merit serious consideration. These countries are shown in the Summary Table below. It is of interest to note that the countries emerging from this paper as needing review of their official exchange rates cover some 70 percent of countries listed in the Fund's Annual Report on Exchange Arrangements and E%change Restrictions as having "cost-related import restrictions" and/or maintaining -.iultiple exch. we rates for exports and imports. It should be emphasized that adoption of approaches suggested in the paper will not necessarily lead to an unambiguous conclusion regarding the "right choice" of the conversion factor. It must also be recognized at the outset that any adjustments to the official exchange rate require intimate knowledge of the trade and exchange system of the country concerned. Nevertheless, these approaches improve the basis for judgments that will always be required in such a decision. More information from and regular discussions with Regional economists will, hopefully, lead to the assembly of adequate information to enhance the robustness of such judgments. iii Summary Table: Countries Recommended for Review/Adjustment of Exchange Rates (1) DEVIATIONS of MULTIPLE RATES >30% >20% < 30% >15% <20% >10% <20% # Egypt Bahamas # Yemen AR # El Salvador # Guatemala # Paraguay # Romania # Venezuela (2) DEVIATIONS of BLACK MARKET RATE >30% >20% < 30% >15% <20% >10% <20% # Afghanistan # Argentina Chile Bahamas Algeria Botswana # Jamaica S. Africa # Bangladesh Haiti Korea Yugoslavia # Egypt India # Sudan Zaire Ethiopia Thailand # Guyana # Hungary # Kenya Libya Morocco Myanmar # Nicaragua # Nigeria s- Paraguay # Peru # Poland # Romania # Syria Trinidad & Tobago # Identified in the Fund's Annual Report on Tunisia Exchange Arrangements and Exchange Restrictions # Venezuela as maintaining multiple exchange rate system. # Zambia Zimbabwe (3) TRADE TAXES ("Implicit Tariff Rates") >30% >20% < 30% >15% <20% >10% <20% Burundi Bangladesh * Belize Guinea-Bissau * Pakistan * Burkina Faso * India * Sudan Comoros Uganda * Egypt Gambia Ghana * Iran Myanmar * Peru Sierra Leone * Somalia Sri Lanka Togo Tunisia * Vanuatu Venezuela * Yemen AR (4) VOLATILITY in EXCHANGE RATES and TRADE VOLUME Select countries on basis of Volatility in quarterly exchange rates (e.g. Argentina, Brazil, Yugoslavia.) I. INTRODUCTION A System of Conversion Factors and Weights 1. This paper discusses issues and procedures to develop a time series of alternative conversion factors with corresponding weights for use in converting and presenting macro- level country statistics in a common unit of account, e.g. the U.S. dollar. The potentially significant operational implications underlying the Bank's estimates of per capita GNP for the Operational Guidelines represent important considerations in the search for an appropriate conversion factor, notably for use in the Atlas. The discussion is set within the context of IEC's present procedures for choosing exchange rates for use in the Atlas and the Operational Guidelines.' Departures from the use of an exchange rate that is not closely related to the rate at which international transactions actually t ' place are tolerated over a wide range. However, when the deviation between the official rate and the rates effectively applied to foreign transactions are egregiously large, any conversion of principal macroeconomic indicators based on the official exchange rate are rendered meaningless. 2. The Bank's practice in choosing conversion factors involves IEC staff systematically reviewing and occasionally estimating alternative conversion factors, often involving dialogue with country operational staff. When "conditions are egregiously different from those 1A review of the methodological issues, including choosing appropriate conversion factors, relevant to the calculation of per capita GNP for operational purposes is found in "Per Capita Income - Estimating Internationally Comparable Numbers," (International Economics Department, World Bank, January 13, 1989). See, in particular, Annex 4 which discusses the search for better conversion factors. 2 prevailing under free trade, alternative conversioi. factors are estimated that are deemed to reflect the actual rate at which foreign transactions take place. Such estimates seek to take into account the nature and restrictiveness of trading regime, information on relative prices, and the evolution of real exchange rates." 2 Estimating alternative conversion factors is a difficult process and, in part judgmental.In those highly egregious cases, the Bank attempts to estimate alternative conversion factors that permit meaningful calculations for intercountry comparisons. 3. The repertoire of exchange rates available has expanded. In addition to the official exchange rate' shown in the Fund's International Financial Statistics (IFS), secondary and tertiary rates are also available in the IFS. In addition, IEC has created an internal file on parallel or black exchange rates from various sources. These are weekly or monthly quotations covering the period 1960-89 from Pick's Currency Yearbook, International Report: Statistical Market Letter, and Africa Analysis. Other rates (e.g., purchasing-power- parities, rates accounting for trade taxes and subsidies) can also be estimated, as explained later in the paper. With the increasing availability of country-specific multiple exchange rate data, the paper suggests that these alternate conversion factors be included in a system so that time series of alternate conversion factors can be made by choosing a preference order among different sets of conversion possibilities. Obviously the choice 'f preference order varies from country to country and also from year to year within each country. 2 Ibid; pp. 2-3. ' In general, the official exchange rate is that reported as line rf in IMF,International Financial Statistics (IFS) (Washington D.C.; International Monetary Fund). 3 4. A major task of this system is the assembly of appropriate weights. The starting premise in the envisaged system of alternative conversion factors and corresponding weights is that the official exchange rate has a weight of 1.0 and all other rates implicitly assigned zero-weights. It is expected that further review and analysis of the trade and exchange system by IEC staff, and more information from and regular discussions with Regional economists on various exchange rates will, hopefully lead to the assembly of information on some distributional weights to be attachea to the various rates. Table 1 below illustrates the range of conversion factors and weights underlying this objective. It is expected that users and analysts can individually devise a set of simple rules to derive what may be called "the preferred" conversion factor. Table 1: EGYPT - System of Alternative Conversion Factors and Weights 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 Iternative Conversion Factors Principal (official) rate (IFS RF) 0.7000 0.7000 0.7000 0.7000 0.7000 0.7000 0.7000 0.7000 0.7000 0.8516 Secondary rate NA 0.7390 0.8317 0.8317 0.8317 1.3010 1.3503 1.5015 2.2128 2.5712 3. Tertiary rate NA 1.0100 1.1300 1.1453 1.2543 1.5488 1.8838 2.1459 2.5082 2.6452 4. Parallel/black market rate 0.7609 0.8772 1.0236 1.1039 1.2184 1.4588 1.8975 2.1894 2.5082 2.6452 5. Purchasing Power Parity Rate (1970) /1 0.5917 0.5503 0.5893 0.6133 0.6553 0.6914 0.7849 0.9079 11850 - 6. Purchasing Power Parity Rate (1980) /1 0.7000 (1.6510 0.6971 0.7255 0.7752 0.8179 0.9285 1.0740 1.4018 - 7. Purchasing Power Parity Rate (1987) /1 0.4562 0.4243 0.4544 0.4729 0.5052 0.5331 0.6052 0.7000 0.9137 - 8. Merchandise weighted conversion factor Wei2hts 1. Principal (official) rate (IFS_RF) 0.4350 0.4350 0.43.0 0.4350 0.4350 0.4350 0.4150 0.3690 0.3200 - Secondary rate 0.3150 0.3150 03150 03150 0.1400 0.1400 0.1830 0.2370 0.2880 Tertiary rate 0.2500 0.2500 0.2500 0.2500 0.4250 0.4250 0.4020 0.3940 0.3920 Parallew/black market rate Purchasing Power Parity Rate (1970) Purchasing Power Parity Rate (1980) Purchasing Power Parity Rate (1988) Merchandise weighted conversion factor Memo Items Nominal exchange rate index (1980= 100) 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 82.2 Real exchange rate index (1980= 100) /2 100.0 93.1 98.6 102.1 109.4 115.2 129.7 148.3 193.0 "Weighted Avg. Conversion Factor" 0.7152 0.7898 0.8490 0.8528 0.9540 1.1449 1.2949 1.4596 1.8445 /1 The purchasing power parity relationship is computed as R. Wd,/,r), where R. is the exchange rate in the base year "0". Pd, and P,,t are the domestic price index and US (or any other foreign country) price index, respectively, in time period "t" (see Chapter V below). /2 Nominal exchange rate index divided by relative domestic absorption deflator index; this shows the real exchange movement. 4 5. Such a matrix of conversion factors and corresponding weights could provide the mechanism for the following: (a) a systematic approach to the weighting of alternative rates; (b) estimating the effects of alternative weighting schemes, (c) determine what would be the effects of incorporating parallel or black market exchange rates; (d) provide the basis for less erratic and unpredictable fluctuations in the data shown in the Atlas, Worid Tables etc. and (e) improved and more transparent documentation of methods and "special cases." Scope of Paper 6. The focus is on conversion factors that are as representative as practicable of transaction or "effective" conversion factors. The choice of an appropriate conversion factor is based on the clear principle that the rate chosen must be that at which foreign transactions are effected. The Bank, like the Fund or any other international institution, would prefer to convert national currency data into a common numeraire at prevailing market exchange rate.4 Determining the prevailing market rate, however, is a matter of judgment, even for international transactions and in the absence of exchange and trade restrictions. It must be stressed at the outset that much is beyond the scope of this paper.In particular, the search here is not for an "equilibrium" exchange rate or that which is sustainable and compatible with reasonable economic growth and a viable balance of payments situation in the medium term. Suffice to --v, countries are seldom in equilibrium and the labelling of an exchange rate as "being in equilibrium" is usually associated, ' The technical issues of using the period average of the official exchange rate as a convenient and readily available conversion factor, and the primary requirements for moving from shorter period to longer period rates are discussed in Section VIII. 5 explicitly or implicitly, with a complex policy network and policy changes that bring the rate to the "equilibrium" level. 7. The paper emphasizes the implications for Bank estimates of per capita, the kinds of evidence relevant in assessing the underlying conversion factors, and identifies countries where review, and adjustment if deemed necessary, of the official exchange rate merits serious consideration. The paper provides illustrative examples in using the evidence to derive alternative conversion factors. 8. This paper takes the view that the exchange rate 1,onversion factor) is a pric used in international transactions. Restrictions introduced into the price system in international transactions impact on the domestic relative price level. The reliance, in general, on the officil exchange rate sometimes produces substantial discrepancies in the data. The paper reviews the main concerns about the adequacy of the official exchange rate as a conversion factor when in effect countries maintain dual or multiple exchange practices, or there are parallel or black market foreign exchange activities, or there is a high incidence of price controls and trade-related restrictions. One aspect receiving special attention is the compilation of a general purpose conversion factor from a set of multiple exchange rates for different types of international transactions. s A useful survey of this topic is given in G.G. Johnson, "Formulation of Exchange Rate Policies in Adjustment Programs," IMF Occasional 1. jer, No. 36 (Washington,D.C.: International Monetary Fund, August 1985). 6 9. The existence of q parallel or secondary market, where a portion of current transactions takes place at floating exchange rates that are more depreciated than the rate in the official market, is prima facie evidence that the official exchange rate may be inappropriate and therefore, requires further review. Also, a priori, it is reasonable to expect that countries with price controis and trade-related restrictions would exhibit large deviations between the transaction and official rates. Magnitudes of Deviations as Signals 10. Any large deviations of the respective conversion factors from the official rate should be taken as signals that further review of the adequacy of the official exchange rate as an appropriate conversion factor is warranted. When the official exchange rate is judged to diverge by an exceptionally large margin from the rate effectively applied to international transactions, that serves as to signal that an alternate conversion factor needs to be estimated if conversions are to be meaningful. Any decision on the preference order between alternative conversion factors would have to take into account relevant country-specific knowledge, not just the magnitudes of the deviations between the two rates. For example, where we know national compilers used the official exchange rate to assign a national currency value to international transactions, that rate must be used as the conversion factor, regardless of whether it was the rate actually applied to international transactions. Assume in this instance that the two rates (the official rate used by national compilers and the transaction rate) are significantly different. If a decision is made to use the known transaction rate as a conversion factor, say for GNP per capita computations, it is necessary that all the expenditure and value-added components of the national accounts 7 in local currency be "modified". This should be avoided. Notwithstanding that the official rate may be egregiously different from the transaction rate, the conversion factor to be used in the above instance should be the official exchange rate as used by the national compilers. 8 II. CONVERSION FACTORS and REIATIVE PRICE MOVEMENTS Conversion Factor as a Price 11. This paper focuses on the search for conversion factors that reflect market valuations (prices). The market-valuation principle serves to measure the economic value of resource transferred between two economies. External transactions should be viewed as the exchange of goods for money. In this connection, the role of the conversion factor is to define the price of money in terms of goods. Whenever governments impose price and non-price measures in international trade, valuation or price (i.e. conversion factor) adjustments should be made that will serve to bring the valuation closer in concept to market value. The probl.m, therefore, is to determine how far the price deviates from or approximates the market valuation. 12. The exchange rate is the relative price of two national monies and must bear a close relationship to price level developments in the two countries. Thus the exchange rate determines the relative prices of traded goods and, therefore, international competitiveness. Thus the conversion factor should be considered as a "price" and, like any other price, it can influence the environment in which domestic and international transactions are conducted. As a price, the conversion factor appears as exogenous to any individual transaction in the market but endogenous to the aggregated system of market demands and supplies because 9 the conversion factor is determined by interacting supply and demand for foreign exchange. It is shifts in these supply and demand schedules that give rise to price (i.e., exchange rate) changes. Thus it should be recognized that the exchange rate is a price that balances the desire to exchange one currency for another in order to effect trade in goods and services. The introduction of restrictions into the price system prevents the price from fulfilling a stabilizing function of equalizing supply and demand, thus affecting relative prices, the profitability of individual activities, and thus resulting in distortions in resource allocations. Conversion Factor and Price Levels of Traded/NonTraded Goods 13. The imposition of price measures (tariffs, specific taxes, import surcharges, advance deposits on imports, export taxes and subsidies, and multiple exchange rates) and non-price measures (quotas, licencing ) alter relative price levels between sectors, e.g., traded goods and nontraded goods. Here we are confronted with the issue of the distinction between "traded goods (sector)" and "nontraded goods (sector)" as well as between "tradables" and "nontiadables." In very general terms, traded goods can be regarded as those that can be imported (and exported), while nontraded goods are those without close, foreign-produced substitutes and must therefore be supplied by domestic producers. More specifically, traded goods can be regarded as "goods that are sufficiently substitutable with goods produced abroad that there exists an international market; nontraded goods are those that are not highly substitutable with foreign-produced goods so that international trade in these goods does not exist".' A very narrow dichotomy is that given in the foreign trade statistics in the 6 George A. MacKenzie and Susan M. Schadler, "Exchange Rate Policies and Diversification in Oil Exporting Countries," (Unpublished; International Monetary Fund, May 9, 1980) pg.3. 10 GDP accounts where GDP is divided into goods and services actually traded (exports and imports, i.e., "traded goods") and goods and services absorbed in the domestic economy ("nontraded goods"). 14. Provided a market exists through which trade could potentially be conducted, a good need not actually be traded to be termed a "traded" good: it becomes a "tradable good." If one classifies as tradable any good that either is internationally traded or could be traded at some plausible range of variation in relative prices, then the tradable category becomes very broad indeed. 15. However, given the close integration of the traded and nontraded sectors in most economies, in practice the distinction between industries producing traded (tradable) goods and those producing nontraded (nontradable) goods is rather tenuous. Rather, a country's output can be thought of as lying along a spectrum of greater or lesser exposure to foreign competition. Goldstein and Officer' suggest the use of both trade flows and market behavior, particularly the degree of independence between domestic and foreign prices, in identifying tradable and nontradable commodities or industries. They suggest three complementary criteria to distinguish between tradables and nontradables: (a) the degree of foreign trade participation should be substantially higher for tradables than nontradables; (b) cross-country correlations of price changes should be much higher for tradables than ' Morris Goldstein and Lawrence H. Officer, "The Measures of Prices and Productivity for Tradable and Nontradable Goods", Review of Income and Wealth, Vol. 4, December 1979, pp. 413-425. 11 nontradables; and (c) tradables should be closer substitutes for traded goods from other countries than are nontradables. 16. Policies adopted to protect the domestic traded goods sector, such as import tariffs, export subsidies, quotas, licensing, have the immediate effect of increasing the domestic price of the commodities in question. Consider the following simple examples: -- an importer pays the c.i.f. import price of $1.00 for a unit of product X. Converted at the official exchange rate of LC200.00 = $1, the price (i.e., the exchange rate) paid by the importer is LC200.00. A 10 percent tariff (or specific taxes, import surcharges) on X means that the importer effectively pays LC220.00 in the transaction. The price (i.e., conversion factor) increases from LC200.00 = $1 to LC220.00 = $1. A world market price of $1.00 now corresponds to a domestic price of LC220.00; -- in the meat export industry, meat is sold at world market prices. The meat exporter receives the world market price of $1.00 per unit. Converted at the official exchange rate of LC200.00 to the dollar, the exporter receives a price of LC200.00. An export subsidy of 25 percent on that unit of export means that the exporter effectively receives LC250.00 in the transaction. 17. Specifically, for a country facing a given world price of traded goods, the domestic price of traded goods is simply the world price adjusted for changes in the domestic currency value of foreign exchange, so that price measures introduced into the trade system 12 have the immediate effect of raising the relative price of traded goods to nontraded goods. The new, higher relative prices of traded goods in terms of nontraded goods are likely to induce consumers to substitute nontraded for traded goods. The increase demand for notraded goods puts pressure on the p. :ze of nontraded goods. The increase in the traded goods price raises the general price level. Higher wage demands aimed at restoring real wages drive up the cost of producing traded and nontraded goods. The pervasive effect on the real cost of domestic factors of production will impact on the competitiveness of the country's traded goods sector. 13 III. CONVERSION FACTORS and MULTIPLE EXCHANGE RATE PRACTICES 18. Multiple exchange rates arise directly when separate groups of foreign exchange transactions are conducted at different exchange rates; they also arise indirectly when a variety of taxes, subsidies, surcharges, or equivalent devices are applied to foreign exchange transactions. Multiple exchange rates are usually adopted to address external payments difficulties by attempting to relieve the pressure on different elements of the external account. The most common features of multiple currency regimes are the maintenance of a preferential rate for essential imports, often combined with debt service payments; separate rates based on a distinction between capital and current transactions;' and a separate rate for some important current invisible payments. Effects of Multiple Currency Practices on Conversion Factors 19. The operation of a system of multiple exchange rates leads to effects very similar to the intended results of a system of tariffs and subsidies. The percentage difference between * For a broad historical review of country experiences with multiple exchange systems, the reasons for resorting to such system, their outcome, and the significance for the Fund in terms of its jurisdiction and maintenance over the international exchange system, see the following unpublished documents of the International Monetary Fund: "Review of Experience with Multiple Exchange Rate Regimes", (SM/84/64), March 19, 1984; "Review of Multiple Exchange Rate Regimes - Background Information", (SM/84/65), March 20, 1984; and "Multiple Currency Practices Applicable Solely to Capital Transactions", (SM/85/19), January 19, 1985. " For a discussion of issues arising from dual exchange rate markets for capital and current transactions, see Anthony Lanyi, "Separate Exchange Markets for Capital and Current Transactions," IMF Staff Papers, Vol. 22 (Washington D.C.: International Monetary Fund, November 1975) pp. 714-49. 14 the exchange rate applicable in a particular transaction and the basic (official) rate can be regarded as a tariff on imports, or a subsidy on exports. The impact of these rates is on goods where relevant consumption and production decisions are sensitive to relative price changes brought about by the application of differentiated rates for traded goods. In principle, a set of differentially appreciated and depreciated rates should provide similar economic incentives and distortions to those resulting from a system of subsidies and tariffs. In terms of its effect on price and resource allocations, the application of a depreciated rate to "non-essential" imports to foster import substitution by domestic producers is similar to the imposition of a tariff on imports that compete with domestic production;'o essentially, the importer and the domestic consumer pay a domestic price that is higher than the world market price (i.e.; more local currency per unit foreign currency). 20. The choice between multiple currency practices and a system of taxes and subsidies may be based on administrative feasibility." On the other hand, multiple rates may be considered less expensive to administer than a system of taxes and subsidies that require complicated administrative machinery. 10 The very similar effects on prices and resource allocations of multiple exchange rates and a system of tariffs/taxes and subsidies are discussed in details in, John F. Laker, "Fiscal Proxies for Devaluation: A General Review," (Unpublished; International Monetary Fund, October 21, 1980); and George A. Mackenzie and Susan M. Schadler, "Exchange Rate Policies and Diversification in Oil Exporting Countries," (Unpublished; international Monetary Fund, May 9, 1980). " The basic problems in the use of fiscal proxies for exchange rate changes are discussed in John F. Laker, "Fiscal Proxies to Devaluation," (Unpublished; International Monetary Fund, October 21, 1980). 15 Incidence of Multiple Currency Practices 21. The incidence of multiple exchange rates among Fund members is shown in Table 2. The Fund's Annual Report on Exchange Arrangements and Exchange Restrictions has reported that developing countries have, in general, moved towards reduced reliance on multiple currency practices, or made progress in either simpl;fying multiple currency practices, or in reducing the distortions associated with them. A look at Table 2 reveals that virtually all countries with multiple currency practices employ different rates for different types of imports (e.g.; essential imports against payments on invisibles) as well as for exports. Multiple exchange rates for about 20 member countries are shown in the IFS. Deviations of Multiple Rates 22. The existence of a multiple exchange system provides indications that the use of the official rate as a conversion factor is inappropriate. The larger the deviations in the multiple rates relative to the ficial rate, the stronger the reasons for a careful and systematic review of the various rates. The focus on multiple rates should be determined on the basis of the magnitude of the premium as well as the size of the market, i.e., how much economic activity is performed through this channel. " For a detailed discussion of developments in multiple currency practices, see International Monetary Fund, Annual Report on Exchange Arrangements and Exchange Restrictions, (Washington,D.C.; International Monetary Fund, 1986); pp. 18-20. 16 Table 2: Incidence of Multiple Currency Practices in Developing Countries Different Rates Country for Exports Imports Exports and Imports More Than 1 Rate More Than 1 Rate *Afghanistan x x x * Argentina x x * Bangladesh x x x Bolivia x x Brazil x x x Colombia x x x Dominica x x x * Dominican Rep x x x Ecuador x x x *Egypt x x x * El Salvador x x x Gambia x x x Ghana x x x Grenada x * Guatemala x x x GuyanA x x Honduras x x x * Hungary x x x * Jamaica x x Kenya x x *Lao x x Mexico x x Nicaragua x x x Nigeria x x x * Paraguay x x x * Peru x x x * Poland x x x * Romania x x x * Somalia x x x * Sudan x x x Syria x x x Uruguay x x * Venezuela x x x Vietnam x x x W.Samoa x x * Yemen AR x x Zambia x x Total 34 31 32 Source: Extracted from IMF, Annual Report on Exchange Arrangements and Exchange Restrictions (Washington DC:1989) * Principal, secondary, tertiary rates, and their applicability to types of transactions are found in IFS. 17 23. Table 3 shows the distribution of deviations of multiple rates from the official rate of those countries for which multiple rates are available in the IFS." Table 3: Distribution of Deviations of Multiple Rates from IFS_RF Range (%) 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 0.0-5.0 2 1 2 2 4 2 4 3 5 3 5.1-10.0 1 3 1 1 0 1 1 3 1 0 10.1-15.0 0 2 2 1 4 1 0 1 0 0 15.1-20.0 2 1 1 1 2 2 1 1 0 0 20.1-25.0 1 1 1 2 0 1 1 0 1 1 25.1-30.0 1 2 2 2 0 2 3 1 1 0 30.1-40.0 2 1 1 1 1 0 0 1 2 1 40.1-50.0 1 2 3 1 0 0 0 1 0 0 50.1-60.0 1 0 0 1 1 0 1 1 0 0 60.1-70.0 2 1 0 1 2 0 1 0 1 1 70.1-80.0 0 0 0 1 1 0 0 0 0 0 80.1-90.0 0 1 0 0 1 2 0 1 0 0 90.1-100.0 0 1 0 0 0 2 1 0 0 0 >100 0 0 1 3 3 3 4 2 3 1 TOTAL 13 16 14 17 19 16 17 15 14 7 Source: Multiple rates are from IFS. 24. Almost half of them maintain multiple rates with the premium exceeding 30 percent. In recent years, the spread between multiple rates has narrowed for most countries. However, the spread has remained consistently high in Egypt (exceeding 100 percent), Paraguay and Venezuela (about 90 percent), Guatemala (averaging 60 percent), and Romania (about 40 percent). '3 As trade and exchange systems have been liberalized, and multiple exchange systems have been unified, the number of countries with multiple rates shown in the IFS has decreased. 18 "Composite" Conversion Factor 25. The proper use of multiple rates demands a "matrix" weighting approach, that is, a distribution -f weights among the multiple exchange rates. What is envisaged here is a system with time series for multiple rates and a corresponding set of time series for weights. Obviously, adjustments to the official and multiple rates require intimate knowledge of the applications of respective rates to specific transactions. Two illustrative examples of IEC's attempt to develop the envisaged system for composite conversion factors are shown in Table 4 for Egypt, and Table 5 for Syria. 26. Table 4 below illustrates the Egypt situation where the multiple rates are from the IFS but the weights are from the Fund and Regional staff. Notice that the premium enjoyed by each of the rates in the secondary and tertiary markets is very significant. Table 4: EGYPT -- Rates and Weights for Computing Average Exchange Rates from Multiple Rates 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 Alternative Conversion Factors 1. Principal (Central Bank) Rate 0.7000 0.7000 0.7000 0.7000 0.7000 0.7000 0.7000 0.7000 0.7000 0.8516 2. Secondary (Commercial Bank) Rate NA 0.7390 0.8317 0.8317 0.8317 13010 1.3503 1.5015 2.2128 2.5172 3. Tertiary (Parallel) Rate NA 1.0100 1.1300 1.1453 1.2543 1.5488 1.8838 2.1459 2.5082 2.6452 Wei2hts 1.1 Principal (Central Bank) Rate 0.4350 0.4350 0.4350 0.4350 0.4350 0.4350 0.4150 03690 0.3200 NA 1.2 Secondary (Commercial Bank) Rate 0.3150 03150 03150 03150 0.1400 0.1400 0.1830 0.2370 0.2880 NA 13 Tertiary (Parallel) Rate 0.2500 0.2500 0.2500 0.2500 0.4250 0.4250 0.4020 0.3940 0.3920 NA "COMPOSITE" Rate NA 0.7898 0.8490 0.8528 0.9540 1.1449 1.2949 1.4596 1.8445 NA [(1*1.1) + (2*1.2) + (3*13)] Memo Items GNP per capita (using Principal Rate) 573 553 637 690 771 847 925 1034 1274 NA GNP per capita (using "Composite" Rate) NA 1432 798 582 611 606 576 561 592 NA Ratio: "Composite" Rate/Principal Rate NA 1.13 1.21 1.22 1.36 1.64 1.85 2.09 2.64 NA "Composite" per cap/"IFS RF' per calNA 2.59 1.25 0.84 0.79 0.72 0.62 0.54 0.46 NA Source: Multiple rates are from IFS Weights: 1980-85: Unpublished data from the Fund's Information Notice System; 1986-88: Regional Staff. 19 27. The respective premiums for the commercial (secondary) rate and the parallel (tertiary) rate are extremely high. Given such a high degree of deviation among the different rates, it is to be expected that the average of the multiple rates is significantly different from the official (Central Bank) rate. Perhaps what is more significant from the Bank's operational point of view is that with the shift (and therefore weights) away from the official (Central Bank) rate in recent years, its is likely that Egypt's per capita GNP will remain within the threshold for Bank's IDA eligibility." 28. The Syria example (Table 5 below) is intended to illustrate the types of relevant information IEC staff can be expected to extract from various Bank and Fund documents to estimate a trade-weighted average of officially recognized exchange rates. The Fund's Annual Report on Exchange Arrangements and Exchange Restrictions lists Syria as maintaining multiple rates, but these rates are not available in the IFS. On the basis of information in the Fund's Annual Report on Exchange Arrangements and Exchange Restrictions, and unpublished Recent Economic Development on member countries, Table 5 below illustrates the procedures by which IEC staff, in consultation with Regional staff, derives a time series of multiple rates and a corresponding time series of weights. 29. As can be seen in Table 5, the impact of the 30 percent premium in the secondary market is dampened by the fact that transactions in this market is only about one-third of total transactions. Thus, the "composite" rate does not exhibit the wide margin of deviation "' The Bank uses the weighted ("composite") exchange rate in the computation of Egypt's GNP per capita for the Atlas and the Operational Guidelines. 20 seen in the example on Egypt. This does not detract from the fact that there is a substantial differential in the GNP per capita estimate for the Bank's Operational Guidelines. Table 5: SYRIA - Derivation of Composite Exchange Rate 1983 1984 1985 1986 1987 1988 Multile Rates 1. Principal Rate /1 3.95 3.95 3.95 3.95 3.95 11.225 2. Secondary Rate: /1 5.45 5.45 5.45 5.45 5.45 3. Tertiary Rate 5.70 7.06 8.80 9.81 10.00 Derivation of Weights 1. Principal (Offlcial) Rate for Public Sector Merchandise Trade Exports % Total GNFS 20.9 21.8 20.4 16.6 193 Imports % Total GNFS 48.9 493 44.9 36.7 33.0 % Share in Total GNFS 69.8 71.1 653 53.3 52.3 2. Secondary (Parallel) Rate for Private Sector Trade in Goods and NFS Exports GNFS % Total GNFS 7.4 6.4 6.8 13.1 14.2 Imports GNFS % Total GNFS 14.9 13.7 18.6 23.9 24.8 % Share in Total GNFS 22.3 20.1 25.3 37.0 38.9 3. Tertiary Rate for Tourism % Share in Total GNFS 7.90 8.88 9.37 9.64 8.82 Summary Weights 1.1 Principal 0.70 0.71 0.65 0.53 0.52 1.00 2.1 Secondary 0.22 0.20 0.25 0.37 0.39 3.1 Tertiary 0.08 0.09 0.09 0.10 0.09 *0 "COMPOSITE" Rate 0*** 4.4225 4.5269 4.7844 5.0701 5.0676 11.225 Computed as: [(1*1.1) + (2*2.1) + (3*3.1)] Memo Items GNP per capita (Atlas..principal rate) 1919 1849 1946 2053 2155 1954 GNP per capita (Atlas..composite rate) 1703 1629 1670 1688 1701 1711 Ratio: Composite rate to Official rate 1.13 1.15 1.22 1.29 1.29 1.00 "Composite" per cap/"Atlas" per 0.89 0.88 0.86 0.82 0.79 0.88 Source: IMF, "Recent Economic Development" (Unpublished); March 6, 1987; and March 9,1988. 21 Countries for Review and Adjustment for Multiple Rates 30. The time series multiple rates and corresponding time series weights derived in Tables 4 and 5 form the crux of the envisaged matrix of conversion factors and weights. 31. The examples of Egypt and Syria illustrate the operational significance of the need for careful review of the presence of multiple conversion factors. Table 2 lists countries that should come under the type of review shown for Egypt and Syria. Given the continued high spread between these rates from the official rate and, therefore, the potential operational impact of per capita GNP estimates, particular attention should be directed at Egypt, El Salvador, Guatemala, Paraguay, Romania, and Venezuela. 32. It should be emphasized that a definitive assessment would require more information from and continued dialogue with Regional staff. Emphasis is placed, wherever data are available, on the applicability of the respecti,e rates on the types of transactions. 22 IV. BLACK or PARALLEL MARKET EXCHANGE RATES Formal and Informal Parallel Exchange Markets 33. Lindauer suggests that "a parallel market is a structure generated in response to government interventions which create a situation of excess demand or supply in a particular product or factor market which is both dependent upon conditions in the official market and responsive to market forces."5 For example, the excess demand for foreign exchange in the official market, resulting from exchange controls, is satisfied at a premium price in a secondary, or parallel market. The official, and overvalued, exchange rate creates a parallel, usually called "black," market for foreign currency. A parallel exchange market is one of the by-products of restrictionist trade and exchange policies. Import and export duties, and import quotas generate parallel markets by creating excess demand for commodities at illegal, pre-tax prices. The removal of government interventions can unify an otherwise parallel market. 34. Are parallel markets always illegal? A formal parallel market typically is introduced when a government resorts to a secondary exchange system as a transitional measure to correct for external imbalance and, therefore, avoid a formal depreciation. Usually essential ' David L. Lindauer, "Parallel, Fragmented, or Black? Defining Market Structure in Developing Economies," World Development, Vol. 17, No. 12 (Cambridge, MA: Harvard Institute for International Development, December 1989); pp. 1871-1880. 23 imports and exports are conducted at the (relatively overvalued) official rate, while other transactions, including capital flows, are assigned to the secondary exchange market. On the other hand, the tightness and effectiveness of exchange controls result in transactions that operate outside the normal and regulated channels, that is, in an informal secondary exchange market which is characterized by risks and penalty structure. 35. In this paper, the terms "black market" and "parallel market" are used interchangeably, as well as "illegal" and "unofficial" to refer to "market structures which result from suboptimal government interventions which create a situation of excess demand."16 Purchasing Power Parity and Factors Affecting Parallel Market Rates 36. In his empirical paper on black markets for currency, Culbertson" postulates that given the official exchange rate, changes in the parallel market rate will occur due to shifts in either the supply or demand schedule for foreign exchange. One of the most likely sources of change stems from variations in the equilibrium rate. Since the equilibrium or market-clearing rate is unobservable, it is proxied through the ratio of domestic-to-foreign price level -- i.e., the purchasing power parity relationship is instrumentai in determining the (unobservable) equilibrium rate. A secondary consideration concerns the role of the government's holdings of international reserves, especially foreign exchange holdings. " Lindauer, Op.cit., pp. 1878. 17 W. Patton Culbertson, "Empirical Regularities in Black Markets for Currency," World Development, Vol. 17, No. 12 (Cambridge, MA: Harvard Institute for International Development, December 1989); pp. 1907-1919. 24 Consider the following exchange-control scenario in Culbertson's study: an overvalued currency produces a supply of foreign exchange; a black market price of foreign exchange emerges as the government allocates this quantity to the market (e.g. to importers, investors, travel abroad, etc.); consider further that the government may choose to add to its reserve holdings, in which case the supply of foreign exchange is reduced; or the government may choose to draw down its reserves, thus increasing the supply of foreign exchange. Consequently, changes in the level of foreign exchange holdings represent a potential influence on the parallel market rate. 37. Culbertson's study for 10 countries with long-existing exchange controls and black market activity lends support for the hypothesis that the black market rate depends on the level of the official rate, the foreign-to-domestic price level (which proxies the unobservable equilibrium rate), and government reserve-level policy. The strongest coefficients are those relating the black market rate to changes in the foreign-to-domestic price level. While the R' are relatively robust for most of the countries (see the following chapter for a discussion of purchasing power parities as conversion factors), the large degree of serial correlation is disturbing. This indicates the danger of inferring that only those variables with significant coefficients are important. 38. The supply of foreign exchange in this market may come from various sources. Over- and underinvoicing of exports and imports could potentially attract significant amounts of foreign exchange to the black markets over time. Also, if the premium on foreign currency is significantly higher than the official rate, a significant proportion of remittances would 25 eventually pass through the illegal market. Another potential source of generating a higher volume of foreign exchange in this market is smuggled exports, though conclusions on the level of smuggling is largely impressionistic. An important factor determining the level of the black market rate is the nature and extent of the risk and penalty structure faced by sellers who circumvent foreign exchange controls. Suppliers of foreign exchange in the black market. In general, the black market rate is likely to be more depreciated than the official (and restriction-free) rate. This is more so in a situation where there is strict imposition of foreign exchange controls and the emergence of expectations of depreciation. Proper caution should be exercised, however, in using the black market rate as an ineication of the appropriate exchange rate. 39. While the estimated coefficients in Culbertson's study should be treated with caution, the empirical results suggest that: "black market exchange rates are largely determined by and reflect fundamental economic developments... ...... the exchange rate is the relative price of two national monies and, while influenced by myriad factors, must bear a close relationship to price level developments, and hence monetary conditions in the two countries. Parallel Market Rates and Market Information 40. If the exchange rate is itself a price upon which economic agents must rely in making decisions about resource allocation, then a question in an exchange-control economy posed by Culbertson is whether the parallel market rate better reflects available information on price developments, including current expectations about the future price level, than the 18 Culbertson, Op.cit., p. 1917 26 official rate. Indeed, the very fact that a parallel or secondary market, where a portion of current transactions takes place at floating exchange rates that are usually more depreciated than the rate in the official market, exists is evidence of the inappropriate level of the official rate.19 Without the black market, there v. uld be little if any guide to indicate just how overvalued the official rate had become. The official exchange rate changes only at intervals and can hardly react to or reflect constantly changing economic conditions. Consequently, Culbertson's empirical study hypothesizes that it would seem that a prima facie case can be made that the black market rate is likely to be a better guide to efficient resource allocation than the official rate. 41. If black market rates fully reflect all historical price information, including current expectations about the future rate, then the current price will provide good prediction of next period's price. To assess how well black market rates seem to meet this criterion of "market efficiency", Culbertson regresses current period black market rates on previous period rates. The principal conclusion from the regression suggests that the behavior of black market rates is consistent with the broad principles of the efficient markets hypothesis. 42. The focus on the parallel exchange market should be determined on the basis of its size, that is, how much economic activity is performed through this channel. In addition, the " Culbertson, op.cit., pp. 1915-1916; see also Ahsan H. Mansur, "Determining the Appropriate Levels of Exchange Rates for Developing Economies," IMF Staff Papers, Vol. (December 1983), pp. 784-818; and G.G. Johnson, "Formulation of Exchange Rate Policies in Adjustment Programs," International Monetary Fund, IMF Occasional Paper, No. 36 (Washington, D.C.:International Monetary Fund, August 1985). 27 extent to which the legal parallel market rate is allowed to float and react to market forces, and the demand and supply elasticities of foreign exchange in the official market relative to the secondary market will help draw some tentative inferences about the appropriateness of the official rate as a conversion factor. 43. Although for many developing countries the illegal or black market can be quite sizeable and potentially influential in the trade and exchange system, it is rarely possible to assign (or agree on) a weight to the parallel market. The very nature of parallel markets makes it difficult to assess total quantities marketed, since many small traders are likely to be involved. Black or parallel market rates are seldom documented. Therefore, it is too large a leap to reach the conclusion that the exchange rate prevailing in the parallel (especially that in the illegal parallel) market is necessarily a reliable indicator of the (unobserved) equilibrium rate. Deviations of Black Market Rates 44. Table 6 shows deviations of the black or parallel rates from official rates. Again, caution must be exercised in reading the significance of the deviations from official rates. These do not necessarily provide a direct indication of the appropriate exchange rate. The divergence between the official and black market rates also depends very much on the penalty structure, that is, the rigor of its enforcement and the extent to which participants in black market exchange transactions are apprehended and prosecuted. Nevertheless, it is useful to know parallel market rates (and their deviations) if only to set upper limits for 28 what the Bank's Regional staff sometimes advocate as rates to replace official rates in those cases where IEC and the Region agree that official rates differ egregiously from transaction rates. Table 6: Frequency Distribution of Deviation of Parallel/Black Market Rates from Official Rates (percent) Deviation 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 L9 0-10 49.4 443 53.8 57.7 48.7 49.4 48.1 40.5 34.2 29.5 33.9 37.1 28.6 28. 11-20 10.1 13.9 9.0 6.4 12.8 12.7 10.1 10.1 11.4 14.8 13.6 9.7 163 17. 21-30 63 5.1 7.7 5.1 6.4 7.6 63 3.8 63 9.8 3.4 14.5 6.1 7.5 31-40 2.5 5.1 2.6 5.1 2.6 3.8 2.5 1.3 5.1 6.6 6.8 1.6 4.1 5.7 41-40 8.9 5.1 1.3 1.3 2.6 13 5.1 8.9 3.8 1.6 1.7 1.6 10.2 5.7 51-60 0.0 1.3 1.3 1.3 5.1 0.0 3.8 8.9 5.1 1.6 1.7 4.8 0.0 3.8 61-70 2.5 1.3 3.8 1.3 3.8 5.1 3.8 3.8 13 1.6 6.8 0.0 4.1 3.8 71-80 2.5 2.5 2.6 3.8 2.6 2.5 2.5 5.1 2.5 4.9 1.7 1.6 0.0 0.0 81-90 1.3 2.5 1.3 3.8 3.8 0.0 1.3 1.3 1.3 1.6 1.7 0.0 0.0 0.0 91-100 5.1 1.3 1.3 1.3 0.0 3.8 0.0 1.3 1.3 3.3 3.4 0.0 2.0 1.9 > 100 11.4 17.7 15.4 12.8 11.5 13.9 16.5 15.2 27.8 24.6 25.4 29.0 28.6 26. No. of Countries 79 79 78 78 78 79 79 79 79 61 59 62 49 53 Deviations 27 29 23 24 25 24 28 36 38 28 29 24 24 25 > 30% Sources: 1975-1983: Pick's World Currency Yearbook (various issues); 1984-1988: Financial Times International Reports: Statistical Market Letter (weekly issues) Countries for Review and Adjustment for Parallel Rates 45. Table 7 below shows countries with egregious (say, 30 percent or greater) deviations between the black market and official rates. They represent that set of countries for which the appropriateness of the official rate as a general conversion factor merit further review and adjustments, if deemed necessary. Also, a closer look at the deviations of the black or 29 parallel market rates from the official exchange rates reveals half of the countries with such egregious deviations are characterized by the maintenance of dual or multiple exchange systems (see Table 7 below). The significance of these relatively large deviations is Table 7: Ratio of Parallel/Black Market to Official and Atlas Exchange Rates Black to Official Rate Total Trade (Official = 1.00) % GDP 1987 1988 (Avg.1987-88) Algeria 7.64 na 28.7 * Romania /1 5.70 6.42 na * Syrian Arab Rep. 5.66 1.98 31.5 Burma 4.32 4.50 na * Afghanistan /1 3.95 na na Iraq 3.76 3.76 na * Poland /1 3.40 4.10 41.6 * Egypt /1 3.13 3.58 50.0 Zambia 3.06 3.04 64.0 * Bangladesh /1 2.91 2.84 23.2 * Nicaragua 2.81 na na Ethiopia 2.54 2.54 33.9 Zimbabwe 2.42 2.83 51.5 Libya 2.34 2.41 na * Peru /1 2.27 2.27 23.8 * Guyana 2.05 3.58 149.5 * Venezuela /1 1.93 2.32 46.8 Tunisia 1.68 1.62 78.2 South Africa /1 1.61 1.14 50.8 * Paraguay /1 1.46 1.69 54.6 Trinidad and Tobago /1 1.46 1.33 72.5 * Hungary /1 1.46 1.43 74.4 * Nigeria 1.44 1.58 53.0 Morocco 1.42 1.46 49.6 * Kenya 1.32 1.32 44.8 Zaire 1.31 0.86 73.9 * Countries identified in IMF, Annual Report on Exchange Arrangements and Exchange Restrictions as maintaining multiple exchange systems. /1 Principal, secondary or tertiary rates for these countries are available in IFS. 30 further heightened by the importance of total external trade in GDP. It is clear that the combination of large deviations and large trade shares in GDP provides prima facie evidence that adjustment of official exchange rate as a general purpose conversion factor should be accorded serious consideration. 46. IEC has expanded its collection of various exchange rates to include parallel or black market rates. For Operational Guideline purposes, black market rates have zero weights. They are used only to shed light on conversion issues and set outer limits on any possible adjustment of conversion factors. 47. Should "unofficial" exchange rates, in addition to officially recognized multiple exchange rates, be used to obtain reasonably meaningful estimates for per capita GNP? In principle, the conversion factor used by the Bank to transform major macroeconomic aggregates in local currencies into a common numeraire should reflect the effective transaction rates used in international transactions. In addition, for the resulting overall conversion factor to be analytically relevant and meaningful, the major macroeconomic variables, such as GNP, should reflect consistently the relative prices implicit in the effective transaction rates believed to apply. Therefore, the effective transaction rates should refer to all the various rates at which foreign transactions take place, including "unofficial" exchange rates. Failure to attach any importance to the unofficial rates which are known to prevail (because they are "unofficial" or "illegal") gives rise to both price and conversion distortions which will tend to show up in erratic trends in derived per capita GNP estimates. 31 It should be stressed, however, that consistency must prevail, that is, the effective transaction rate encompassing all (official and "unofficial") rates becomes a relevant and meaningful conversion factor only if the macroeconomic aggregates in local currencies reflect consistently the relative prices implicit in the effective transaction rates believed to apply (cf.para.10). 48. The Harvard Institute for International Development-sponsored workshop on parallel markets in developing countries provided clear indications that "increasing attention to parallel markets in general has arisen in response to price controls," .......that "although economists increasingly recognize the importance of parallel markets, many policy analysts treat them as an inconvenience to be avoided," ....... and, in the face of pervasive price controls and market interventions by governments in developing countries, "parallel markets can compromise the policy prescriptions that are part of most reform packages and can lead to different interpretations of the benefits of reform." 20 49. While recognizing the importance of parallel or black markets in many developing countries the Bank, however, should be mindful of the risks of using statistics pertaining to what are in many countries illegal activities. * A summary of the major contributions to the workshop on parallel markets in developing countries sponsored by the Harvard Institute for International Development (November 11-12, 1988) is found in Christine Jones and Michael Roemer, "Editors' Introduction:Modeling and Measuring Parallel Markets in Developing Markets," World Development, Vol. 17, No. 12 (Cambridge, MA: Harvard Institute for International Development, December, 1989); pp. 1861-1870. 32 V. PURCHASING POWER PARITY as CONVERSION FACTORS Purchasing Power Parity (PPP) 50. The purchasing power parity (PPP) theory of exchange rates argues that the foreign exchange value of a national currency is largely determined by the purchasing power of that currency relative to the purchasing power of foreign currencies.21 Two aspects of the PPP doctrine deserve attention. The more fundamental element focuses upon the manner in which the relation between national price levels leads to market-clearing exchange rates for national currencies. This argument hypothesizes that changes in relative price levels dominate exchange rates changes rather than vice versa. Thus PPP involves a relationship between a country's foreign exchange rate, on the one hand, and its price level or price movemcnt compared to the foreign price level or movement, on the other, that is, the exchange rate is determined by the level of prices in the domestic currency compared to that abroad, that changes in the exchange rate are determined by changes in these price levels, and that the percentage change in the exchange rate per month, quarter, or year is determined by inflation at home relative to that abroad over these time intervals. Thus in the PPP, relative inflation movements are offset by exchange rate changes. PPP may be " A comprehensive discussion of various aspects of the purchasing power parity doctrine is given in Lawrence H. Officer, Purchasing Power Parity and Exchange Rates (Greenwich, Connecticut: Jai Press, 1982); see a.. Alan C. Shapiro, "What Does Purchasing Power Parity Mean?" Journal of Money and Finance, Vol.2, No. 3, December 1983, pp. 295-318. 33 expressed in terms of price movements rather than levels, with price movements measured by price indices: Et = Pdt/PC where E, = the exchange rate expressed as the number of domestic currency units per unit of local currency at time t; Pd = domestic price index at time t; and Pt = foreign price index at time t. Alternatively, the purchasing power parity of the domestic currency with respect to the foreign currency can be expressed as: (P,t/Pt) * R. where Ro is the actual or "market" or current exchange rate in the base year "o" and not necessarily equal to the "equilibrium" rate defined in any manner. Purchasing power parities for Egypt are shown in Table 1 above. 51. PPPs are distinct from and not necessarily equal to the "equilibrium" exchange rate, defined in any manner. They do not measure the degree of overvaluation or undervaluation of a currency. In the absence of other influences PPPs can, however, serve as a guide to the levels of equilibrium exchange rates between the various currencies provided that the base period chosen is one in which equilibrium values prevailed.' PPPs become all the more ' William P. Culbertson, "Purchasing Power Parity and Black-Market Exchange Rates," Economic Inquiry, Vol. 13, June 1975, p. 287. 34 applicable if trade restrictions and exchange controls are sufficiently high and comprehensive to distort the official exchange rate. PPP and Data Conversion 52. Probably the most generally accepted use of PPP is as a conversion factor, given the simplicity of the PPP approach, the ease in calculating relative PPP measures, and the intuitive appeal of PPP as a valid theory. International comparison of real output in different countries requires that the conversion rate reflect relative price levels. The absolute PPP measure ( E, = Pd/PL founded on the GDP price level has been used for the currency conversion of the national accounts data at finely disaggregated levels of expenditure by Kravis.' What Price Measure to Use in PPP Computations 53. The individual prices used to construct the price level or price index for PPP must be presumed to reflect accurately the prices at vhich transactions occur in a free market. In a multicommodity world with relative price changes, differing consumption preferences across countries, nontraded goods, transportation costs, and trade barriers, the PPPs are likely to be distorted so that they do not reflect true relative purchasing power. ' Irving B. Kravis, Alan W. Heston, and Robert Summers, International Comparisons of Real Product and Purchasing Power (Baltimore,MD: Johns Hopkins University, 1978). 35 54. Suggestions to the appropriate price index to use in computing and testing PPPs have "ranged from the price of the least traded commodity -- labor -- to an index of nontraded goods only, to the broadest domestic index available, to an index comprised only of traded goods, to individual prices of traded commodities." 2 Comparison of unit labor costs is beset with measurement problems. Among these are measurements reflecting differences in types and quality of labor; and adjustment for differences in productivity (traded goods sector only, nontraded goods sector only, or combination of both?). Price indices based on internationally traded goods (export and import price indices) are limited to a small class of commodities, relative to all goods and services, and are therefore subject to variations that presumably would not be present in a broad-based price measure. Also, traded and nontraded goods are not unvarying collections of commodities. There is never a definite group of commodities that can be exported. Changes in the structure of the economy or changes in profitability may widen or restrict the group of traded/exportable goods.In the same vein wholesale price indices (WPI) are also heavily weighted with traded goods. The use of cost-of-living indices (COL) requires that the weighting pattern be the same for each country's price level, that is, taking a common basket of goods with a standard system of weighting. The PPP theory essentially refers to the internal value of the currencies concerned, and variations in this value can be measured best by general index figures representing as far as possible the whole mass of commodities (and services), traded and nontraded, marketed in the country. This is reflected in the GDP deflator which is the price ' Alan C. Shapiro, "What Does Purchasing Power Parity Mean?" Journal of International Money and Finance, Vol. 2, No. 3 , December 1983; pp. 295-318. 36 concept with a firm analytical foundation in PPP theory.' In essence, a PPP measure based on the GDP deflator is advocated since each country's own pattern of production provides the ideal source of weights to construct its price measure for PPP computation. " See Lawrence H. Officer, "The Relationship Between Absolute and Relative Purchasing Power Parity,* Review of Economics and Statistics, Vol. 60, November 1978, pp. 562-568 37 VI. TRADE and EXCHANGE TAXES and CONTROLS 55. Countries facing unsustainable external balance and payments difficulties, often react by imposing a variety of policy instruments and restrictions on the use of foreign exchange and/or to protect industry-specific industries. These include price measures such as tariffs, import surcharges, advance deposits for imports, export taxes, subsidies and multiple exchange rates, as well as non-price measures such as quotas, licensing, voluntary restraint agreements, and exchange controls. Over time and as a result of these ad hoc decisions rather than an overall policy design, the trade regulations become increasing complex." Price measures, in general, can be expressed in terms of tariffs. Specific taxes or import surcharges are added to tariffs; the tariff equivalent of advance deposits is given by the interest foregone on loans designed to make the deposits; export taxes and import subsidies are regarded as negative tariffs; and the premium in the secondary exchange market can also be regarded as a tariff. In the case of nonprice measures, permissible levels of imports are set directly in quantitative terms. By restricting the amount of imported commodities made available on the home market, these measures lead to a rise in the domestic prices of the commodities in question. The excess of domestic price over world price can be regarded as an "implicit tariff." 2 For a comprehensive description of exchange rate arrangements and exchange controls, and changes in these as they occur, see International Monetary Fund, Annual Report on Exchange Arrangements and Exchange Restrictions (WashingtonD.C.: International Monetary Fund, 1989). 38 56. The nature and extent of price measures and nonprice measure restrictions can alter relative prices and profitability between sectors, leading to distortions in resource allocations. Frequent increases in tariffs and import surcharges, the imposition or tightening of quantitative restrictions and resorting to multiple exchange rates for balance of payments purposes mean that the actual value of traded goods do not reflect the outcome of market forces. Tariffs and nontariff barriers and subsidy measures represent attempts by governments to offset the negative effects of the overvalued exchange rate by influencing domestic relative prices and, therefore, on import demand and export supply. Trade Taxes on Gross Output and Value-Added 57. The impact of taxes and subsidies on relative prices and the profitability of industry- specific activities (with its attendant resource allocation implications) should take note not only of the tariff (and subsidy) structure on the final product but also on material inputs. From the point of view of the producer, not only tariffs and other protective measures on the final output matter but also tariffs levied on material inputs because the latter would reduce the domestic value-added. In export industries the increase in costs due to tariffs on material inputs may be compensated by the provision of export subsidies. Balassa concludes that tariffs and subsidies on gross output and intermediate inputs permit domestic industries to operate with a value-added higher than under tariff-free trade." " Bela Balassa, "Industrial Protection in Developing Countries," Report No.EC-175, (Unpublished; The World Bank, June 1980). 39 Effects of Tariffs and Subsidies on Conversion Factors 58. The imposition of tariffs, surcharges and subsidies represents the use of "fiscal proxies" for exchange rate changes. Simply stated, such a policy may be designed to force imports competing with domestically traded goods to enter the country at a depreciated exchange rate (i.e.; at a higher domestic currency price) and/or to allow producers to export at a depreciated rate while preserving a higher exchange rate for other transactions. In essence, the imposition of import tariffs and export subsidies means that, effectively, importers pay more and exporters receive more local currency units per dollar than they would under free trade. It appears, then, that the price (i.e.; conversion factor) paid for or received under tariffs and export subsidies tends to overvalue the domestic currency as compared to the tariff-free trade situation and, in the process, raises the relative price of traded to nontraded goods.' On the other hand export taxes (and import subsidies) are regarded as negative tariffs. 59. The extent of the implications on international data comparability and on comparative analysis resulting from the reliance on the official exchange rate as a conversion factor depends on the incidence and magnitude of price measures and nonprice measures in international trade." How prevalent are import and export taxes and subsidies in developing countries? Are tariff structures and schedules prohibitive? This is examined below. 2 See para. 15 above for simple numerical examples illustrating this. 2 An overview of current developments in trade and exchange controls is presented in International Monetary Fund, Exchange Arrangements and Exchange Controls, (Washington,D.C.; International Monetary Fund, 1989); and International Monetary Fund, "Developments in International Exchange and Trade Systems," World Economic and Fir .ncial Surveys (Washington,D.C,: International Monetary Fund, September 1989). 40 Incidence of Trade Taxes and Exchange Controls 60. The Fund's Annual Report on Exchange Arrangements and Exchange Restrictions reports that, in general, most developing countries had adopted, or were edging towards more liberal trade policies at the end of the 1980, than they had before the external balance difficulties of the late 1970s, albeit from a considerably restrictive base. On the other hand the industrial economies, especially the EEC and the US, have intensified their recourse to quantitative controls. Given their varying forms and purposes and lack of transparency, it has been difficult to assess the incidence and quantitative impact of nontariff barriers on trade. They take many forms: import licensing, quotas, quantitative restrictions in the form of outright prohibition, voluntary export restraint arrangements, government procurement and state trading practices that may have some hidden discrimination against specific foreign suppliers, etc. But the end result of tariffs and quantitative controls is the same: they raise the price of imported goods in a particular market relative to prices without restrictions and relative to world prices. 61. Table 8 below shows the frequency distribution of developing countries by the size of the tariff rates, measured here as total, import, and export taxes as a percent of total, imports, and exports of goods and nonfactor services, respectively? It is clear that many developing countries still resort to tariffs and import surcharges to deal with acute balance of payments difficulties. Dudes on overall trade of 10 percent or greater are registered in about one-third of developing countries for which data are available. Other notable features 30 Since trade flows and taxes used here are at the aggregated level, the implicit assumption is made that the tariff barrier, which in fact applies only to certain product groups, affects all trade. 41 in Table 2 are: (a) most of the taxes on international trade and transactioas are imposed to curtail imports; (b) two-thirds of the developing countries for which data are available impose implicit import tariff rates of 10 percent or more; and (c) the duty rate on exports are relatively low though its incidence is still high. Table 8: Distribution of "Implicit Tariff Rates" "Implicit Tariff Rate" (Total Duties % Total Goods & Nonfactor Services) Range(%) 1980 (%) 1981 (%) 1982 (%) 1983 (%) 1984 (%) 1985 (%) 1986 (%) 1987 (%) 0.0-5.0 21 (263) 23 (28.4) 24 (31.6) 22 (28.6) 23 (28.4) 20 (26.3) 19 (29.7) 8 (27.6) 5.1-10.0 33 (41.3) 42 (51.9) 32 (42.1) 37 (48.1) 35 (43.2) 38 (50.0) 30 (46.9) 11(37.9) 10.1-15.0 17 (21.3) 11 (13.6) 14 (18.4) 12 (15.6) 16 (19.8) 10 (13.2) 9 (14.1) 8 (27.6) 15.1-20.0 7 (8.8) 5 (6.2) 5 (6.6) 4 (5.2) 4 (4.9) 5 (6.6) 3 (4.7) 0 (0.0) 20.1-25.0 1(1.2) 0 (0.0) 1 (1.3) 1 (1.3) 3 (3.7) 3 (3.9) 1 (1.6) 1 (3.4) > 25.0 1 (1.2) 0 (0.0) 0 (0.0) 1 (1.3) 0 (0.0) 0 (0.0) 2 (3.1) 1 (3.4) Total 80 81 76 77 81 76 64 29 "Impilcit ImBort Tariff Rate" (Import Duties % Imports Goods & Nonfactor Services) Range (%) 1980 (%) 1981 (%) 1982 (%) 1983 (%) 1984 (%) 1985 (%) 1986 (%) 1987 (%) 0.0-5.0 5 (7.6) 7 (10.4) 5 (7.8) 6 (9.5) 7 (10.3) 7 (11.1) 4 (7.5) 1 (4.3) 5.1-10.0 9 (13.4) 9 (13.4) 11(17.2) 7 (11.1) 7 (10.3) 7 (11.1) 6 (11.3) 0 (0.0) 10.1-15.0 11(16.4) 15 (22.4) 16 (25.0) 16 (25.4) 15 (22.1) 12 (19.0) 10 (18.9) 7 (30.4) 15.1-20.0 16 (23.9) 21(31.3) 14 (21.9) 19 (30.2) 17 (25.0) 17 (27.0) 17 (32.1) 5 (21.7) 20.1-25.0 12 (17.9) 8 (11.9) 11(17.2) 9 (14.3) 11 (16.2) 12 (19.0) 6 (11.3) 5 (21.7) > 25.0 14 (20.9) 7 (10.4) 7 (10.9) 6 (9.5) 11 (16.2) 8 (12.7) 10 (18.9) 5 (21.7) Total 67 67 64 63 68 63 53 23 "Implicit Export Tariff Rate" (Exnort Duties % Exports Goods & Nonfactor Services) Range (%) 1980 (%) 1981 (%) 1982 (%) 1983 (%) 1984 (%) 1985 (%) 1986 (%) 1987 (%) 0.0-5.0 40 (63.5) 48 (77.4) 48 (78.9) 46 (76.7) 53 (80.3) 50 (82.0) 40 (81.6) 18 (78.3) 5.1-10.0 16 (25.4) 11(17.7) 11(18.0) 9 (15.0) 8 (12.2) 6 (9.8) 4 (8.2) 2 (8.7) 10.1-15.0 0 (0.0) 1 (1.6) 0 (0.0) 3 (5.0) 1 (1.5) 2 (3.3) 2 (4.1) 1 (4.3) 15.1-20.0 3 (4.8) 1(1.6) 0 (0.0) 0 (0.0) 1 (1.5) 0 (0.0) 1 (2.0) 1 (4.3) 20.1-25.0 0 (0.0) 0 (0.0) 1 (1.6) 1 (1.6) 2 (3.0) 0 (0.0) 0 (0.0) 0 (0.0) > 25.0 4 (6.3) 1(1.6) 1(1.6) 1(1.6) 1 (1.5) 3 (4.9) 2 (4.1) 1 (4.3) Total 63 62 61 60 66 61 49 23 Source: Taxes on international trade are from the IMF Govt. Finance statistics file; Exports and Imports of goods and nonfactor services are from IEC data file. 42 62. Table 9 lists the countries in the Fund's Annual Report on Exchange Arrangements and Exchange Restrictions with "cost-related import restrictions" (import surcharges and advance import deposits) as features of the trade and exchange system, as well as other countries not included in the Fund's list but with import and export duties of at least 10 percent of imports and exports of goods and nonfactor services, respectively. The "implicit average import tariff rate" varies substantially among the countries, ranging from 1 percent in Greece to 45 percent in India. It should also be noted that these average taxation rates on imports often mask some fairly high tariffs protecting individual import-sensitive industries, although in practice the bulk of transactions may take place within a relatively narrow tariff schedule. 63. The data in Table 9 reinforce the features evident in Table 8: that most trade taxes are imposed on imports; their incidence is dispersed in both low- and mid-income countries; and the implicit tariff rate in low-income countries (e.g.,Bangladesh, Gambia, India, Myanmar, Pakistan, Sudan, Sierra Leone) averages 25 percent, which is twice that imposed by the middle-income economies. 43 Table 9: Incidence of Import and Export Taxes in Selected Developing Countries Import Duties Export Duties Total Duties Country % Imports GNFS % Export GNFS % Tai GNPs X Argentina 36.2 /1984 8.9 /1984 9.2 /1984 GFS Bangladesh 26.9 /1985 6.1 /1985 17.9 /1985 X Belize 21.7 /1985 1.0 /1985 11.2 /1985 GFS Botswana 17.0 /1986 1.0 /1985 7.4 /1986 X Burkina Faso 15.2 /1987 1.6 /1987 10.5 /1987 GFS Burundi 15.6 /1986 32.9 /1986 21.9 /1986 GFS Cameroon 19.0 /1987 2.5 /1987 9.3 /1987 X China na na na X Colombia 20.1 /1987 1.0 /1987 8.3 /1987 GFS Comoros 21.1 /1987 13.2 /1986 13.8 /1987 X Congo 7.7 /1980 1.0 /1980 3.8 /1980 X Costa Rica 20.8 /1986 5.5 /1986 7.4 /1986 X Cote D'Ivoire 31.3 /1985 5.6 /1985 9.7 /1985 X Cyprus na na na X Djibouti 2.6 /1986 1.0 /1986 1.6 /1986 X Dominican Rep. 20.5 /1986 3.2 /1986 9.9 /1986 X Ecuador 15.6 /1985 1.0 /1985 6.2 /1985 X Egypt 18.2 /1987 1.0 /1987 11.5 /1987 GFS El Salvador 19.1 /1987 10.6 /1987 6.9 /1987 GFS Fiji 15.8 /1986 7.7 /1986 GFS Gabon 19.3 /1985 1.1 /1985 6.7 /1985 GFS Gambia 25.5 /1987 1.0 /1987 13.6 /1987 GFS Ghana 10.1 /1987 17.4 /1987 13.5 /1987 X Greece 10 /1985 na na X Grenac:a na na X Guatemala 11.1 /1982 3.8 /1982 4.5 /1982 X Guinea na na na GFS Guinea-Bissau na na 24.1 /1987 X Haiti 11.9 /1987 1.3 /1987 6.8 /1987 GFS Honduras 20.2 /1981 6.4 /1981 8.1 /1981 X India 45.7 /1987 26.1 /1987 X Indonesia 3.7 /1987 3.5 /1987 3.6 /1987 X Iran Iraq na na 11.9 /1985 X Iraq na na na X Israel na 3.0 /1985 X Jamaica na na 1.6 /1981 X Jordan na na 7.2 /1986 X Kenya 17.5 /1986 2.6 /1986 7.4 /1986 X Lebanon na na na X Liberia 14.0 /1986 1.1 /1986 6.2 /1986 X Libya na na na X Madagascar 15.0 /1982 3.8 /1982 8.1 /1982 X Malawi 15.3 /1986 1.0 /1986 7.6 /1986 GFS Mali 14.6 /1987 3.6 /1987 8.5 /1986 X Mauritius 19.5 /1988 3.2 /1988 8.2 /1988 GFS Myanmar na na 14.6 /1985 X Nepal 12.0 /1984 1.0 /1984 7.2 /1984 X Netherland Ant. na na na 44 Table 9 (Cont): Incidence of Import and Export Taxes in Selected Developing Countries Import Duties Export Duties Total Duties /1 Country % Imports GNFS % Export GNFS % Total GNFS X Nicaragua na na na X Pakistan 29.5 /1986 1.6 /1986 17.7 /1986 X Panama 11.8 /1986 1.0 /1986 5.3 /1986 X Paraguay 3.4 /1986 1.0 /1986 2.1 /1986 X Peru 21.7 /1986 2.7 /1986 10.5 /1986 X Philippines 12.0 /1986 1.0 /1986 4.9 /1986 GFS Poland 14.8 /1987 1.1 /1987 6.4 /1987 GFS St. Lucia 14.9 /1986 1.0 /1986 7.6 /1986 GFS St. Vincent 16.6 /1986 1.0 /1986 8.6 /1986 GFS Senegal 15.8 /1984 1.0 /1984 8.8 /1984 X Sierra Leone 28.9 /1987 1.0 /1987 13.2 /1987 X Somalia 17.8 /1978 1.1 /1978 12.5 /1978 X South Africa 3.6 /1985 1.0 /1985 1.4 /1985 GFS Sri Lanka 21.1 /1988 3.9 /1988 10.7 /1988 X Sudan 25.1 /1981 5.8 /1981 17.5 /1982 X Suriname 16.8 /1986 1.0 /1986 8.5 /1986 GFS Swaziland 14.8 /1986 1.0 /1986 7.9 /1986 X Syria 8.9 /1987 1.0 /1986 5.8 /1986 X Thailand 11.3 /1987 1.0 /1987 5.3 /1987 GFS Togo 19.5 /1986 1.0 /1986 11.0 /1986 X Trinidad 7.8 /1981 na 3.5 /1981 GFS Tunisia 23.3 /1984 1.0 /1984 13.1 /1984 X Turkey 5.9 /1987 1.0 /1987 3.1 /1987 GFS Uganda 3.1 /1986 32.6 /1986 16.0 /1986 X Uruguay 18.1 /1987 1.0 /1987 7.5 /1987 X Vanuatu 18.2 /1986 1.6 /1986 12.5 /1986 GFS Venezuela 25.0 /1986 na 12.5 /1986 X Yemen AR 14.8 /1987 na 13.2 /1987 X Yemen PDR na na na X Yugoslavia 13.2 /1987 na 6.5 /1987 GFS Zaire 19.7 /1986 5.1 /1986 X Zambia 24.0 /1988 7.4 /1988 8.5 /1987 X Zimbabwe 16.4 /1985 na 8.0 /1985 Total 81 X = 56 GFS = 25 Sources: IMF, Annual Report on Exchange Arrangements and Exchange Restrictions, 1989; Import and exports duties are from the IMF file on Government Finance Statistics. 1/ " X "denotes countries identified in the Fund's Annual Report on Exchange Arrangements and Exchange Restrictions with "cost-related import restrictions" (import surcharges and advance import deposits) as features of the trade and exchange system. " GFS " denotes other countries not included in the X list with import and export duties of at least 10 % of imports and exports of GNFS, respectively 45 "Trade Tax-Effective" Conversion Factors 64. Given the high incidence and relatively high implicit tariff rates on imports (and heavy export taxes in a few countries) shown in Tables 8 and 9, there are strong reasons to examine conversion factors that reflect the impact of substantial tariffs and subsidies. A priori, it is reasonable to expect that countries with cost-related restrictions on international transactions would exhibit large deviations between the transaction rate and the official exchange rate. In situations where tariffs and subsidies are widely dispersed and substantial, the price (i.e.; conversion factor) paid for or received by transactors differs from the price denoted by the official exchange rate. In essence, we need to compute so-called "tax-effective" import and export prices (i.e.; conversion factors) as well as a conversion factor that represents an overall "trade-tax effective" con"ersion factor. Significant deviations of the overall "trade-tax effective" conversion factor from the official exchange rate shouid be taken as a signal that further review of the adequacy of official exchange rate as a conversion factor merits consideration. 65. The majority of developing countries (90 percent) for which data are available impose export duties considerably lefs than 10 percent (see Table 8); that is to say that the impact of export taxes on the transaction rate is small. Therefore the last column in Table 9 (total duties % total goods and nonfactor services) is a close proxy for the magnitude of the deviations of the "trade-tax-effective" conversion factors from the official rates. 66. Table 10 illustrates an approach towards computing "trade-tax-effective" conversion factors. Because data on subsidies are not available, the computational approach shown 46 below does not take into account the impact of subsidies. To illustrate the a priori reasoning that countries with high cost-related restrictions on international transactions would exhibit larger deviations between the transaction rate and the official rate, Table 10 illustrates the impact in two differing situations. India's trade system is characterized by high price- measure restrictions, particularly on imports; the implicit import tariff rate is over 40 percent since the mid-1980s. On the other hand, the Malaysian trade system is relatively restriction free. As seen in Table 10 (indicator K), there is a marked difference in the magnitude of the deviation of the "trade-tax effective" conversion factor (TIXR) from the official exchange rate between the two countries. In Malaysia, the conversion factor effectively used in trade is virtually no different from the official exchange rate; on the other hand, in India the conversion factor in international transactions is some 25 percent more depreciated than that represented by the official exchange rate. 67. It is apparently clear that where we know that a country's trade and payments system is very restrictive, the use of the official exchange rate as a general conversion factor needs farther examination. In such a situation, attempts should be made to estimate a conversion factor that takes into account the impact of the trade restrictions (e.g., the "trade tax- effective" conversion factor). It is suggested that this be done for countries shown in the Fund's Annual Report on Exchange Arrangements and Exchange Restrictions as having "cost-related import restrictions", supplemented with a country list (other than those listed in the above-mentioned Fund document) with high import and export duties. 47 Table 10: "Tax-Effective" Conversion Factors for Imports and Exports INDIA: HIGH TARIFFS/SURCHARGES 1980 1981 1982 1983 1984 1985 1986 1987 Basic Data A. Total Duties % Total GNFS /1 15.52 17.17 20.17 19.06 19.1524.24 27.56 26.09 B. Import Duties % Import GNFS /1 24.66 28.12 33.63 32.43 32.94 39.74 46.06 45.10 C. Export Duties % ExportS GNFS /1 1.33 0.60 0.63 0.64 0.52 0.53 0.56 0.43 D. Imports GNFS % Total GNFS 60.83 60.22 59.22 57.96 57.46 60.47 59.33 57.46 E. Exports GNFS % Total GNFS 39.17 39.70 40.78 42.04 42.54 39.53 40.67 42.54 F. IFS RF 7.8629 8.6585 9.4551 10.0990 113630 123690 12.6110 129D G. E)R_ATLAS 7.8930 8.9290 9.6280 103120 11.8870 12.2370 12.7870 13.0000 DERIVED INDICATORS H. ExRate."Import-Tax" Effective (ITXR) /1 9.8017 11.0928 12.6346 13.3738 15.1057 17.2844 18.4198 18.8075 I. ExRate."Export-Tax" Effective (ETXR) /1 7.7580 8.6066 9.3955 10.0345 11.3034 12.3029 12.5403 12.9068 1. ExRate.-Trade-Tax" Effective (TTXR) /1 9.0011 10.1038 11.3135 11.9699 13.4883 15.3150 16.0286 1629 K. Ratio: TTXR/IFS RF 1.1448 1.1669 1.1966 1.1853 1.1870 1.2382 1.2710 12573 L. Ratio: TTXR/EXR ATLAS 1.1404 1.1316 1.1751 1.1608 1.1347 1.2515 1.2535 1236 M. Index Real Effective ExRate 100.00 111.96 12435 128.09 137.27 150.20 151.20 14790 MAlAYSIA: LOW TARIFFS/SURCHARGES 1980 1981 1982 1983 1984 1985 1986 1987 Basic Data A. Duties % Total GNFS /1 7.71 7.00 5.83 5.93 5.64 537 4.17 3.54 B. Import Duties % Import GNFS /1 7.06 6.66 6.21 6.51 6.48 6.53 5.70 4.87 C. Export Duties % Export GNFS /1 8.34 7.38 5.40 5.28 4.84 4.32 2.80 2.50 D. Imports GNFS % Total GNFS 48.89 52.79 53.94 52.64 49.11 47.55 47.10 43.91 E. Exports GNFS % Total GNFS 51.11 47.21 46.06 47.36 50.89 52.45 52.90 56.09 F. IFS RF 2.1769 2.3041 2.3354 2.3213 2.3436 2.4831 2.5814 25L G. EXR_ATLAS 2.1769 23041 2.3354 23213 2.3436 2.4831 2.5814 2515 DERIVED INDICATORS H. ExRate."Import-Tax" Effective (ITXR) /1 23305 2.4576 2.4803 2.4724 2.4954 2.6419 2.7270 26 I. ExRate."Export-Tax" Effective (ETXR) /1 1.9954 2.1341 2.2093 2.1988 2.2301 23728 2.5077 245 J. ExRate."Trade-Tax" Effective (TTXR) /1 2.1593 23049 23555 2.3429 23604 2.5008 2.6110 253% K. Ratio: TTXR/IFS RF 0.9919 1.0003 1.0086 1.0093 1.0072 1.0084 1.0120 1074 L. Ratio: TTXR/EXR ATLAS 0.9919 1.0003 1.0086 1.0093 1.0072 1.0072 1.0115 1015 M. Index Real Effective ExRate 100.00 109.77 115.12 114.05 116.63 124.79 129.24 128.88 COMPUTATION PROCEDURES: /1 H. ExRate."Import-Tax" Effective = [Fx (1+B)I 1. ExRate."Export-Tax" Effective = [Fx (1-C)] J. ExRate."Trade-Tax" Effective = [(H x D) + (I x E)] /1 Because data on subsidies are not available, their impact on the respective conversion factors is not reflected in the computations shown above. 48 Table 10 above illustrates the typ of data IEC staff is expected to culP from various Bank and Fund documents to support the estimation of an alternative conversion factor that gives some recognition to price-measure restrictions. Deviations of "Trade Tax-Effective" Exchange Rates 68. Table 11 below shows the distribution of the deviations of the "trade tax-effective" rate from the official rate. The distribution sh(,wn here is very similar to that of total duties imposed on trade in goods and nonfactor services shown earlier in Table 8. In Table 11, some 30 percent of developing countries for which data are available show deviations greater than 10 percent in the "trade tax-effective" conversion factor; similarly, Table 8 shows that the "implicit tariff rate" of 10 percent or greater is imposed in some 30 percent of developing countries. Table 11: Distribution of Deviations of "Trade Tax-Effective" Rate from Official Rate Range (%) 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 M87 0.0-1.0 1 1 1 1 1 1 1 3 2 2 2 3 6 1.1-5.0 17 21 17 18 18 19 20 22 19 19 17 14 6 5.1-10.0 32 35 33 30 30 31 39 29 33 32 35 27 10 10.1-15.0 15 16 16 18 15 17 10 12 11 15 8 8 7 15.1-20.0 9 6 5 6 7 7 5 5 3 3 4 1 0 20.1-25.0 2 1 5 2 2 1 0 1 1 3 3 1 0 2r.1-30.0 0 0 1 3 1 0 0 0 0 0 0 2 1 > 1 1 1 1 2 1 1 1 1 1 1 1 0 Total: 77 81 79 79 76 77 76 73 70 75 70 57 30 49 Countries for Review and Adjustment for Trade Taxes 69. Table 12 below lists countries" with "implicit tariff rates" of at least 10 percent. In the context of the search for an appropriate general purpose conversion factor, the list represents the suggested countries for review, and adjustment if necessary, in the official exchange rate to take into account the cost-related restrictions in the and trade system. It should be emphasized that in the review process, Operations staff should be encouraged to provide data on subsidies if they form a significant part of the trade system. Table 12: Selected Countries for Review and Adjustment of Official Rate for Trade Taxes "Implicit Tariff "Implicit Tariff Country Rate" (%) Country Rate" (%) Bangladesh 17.9 /1985 * Pakistan 17.1 /1986 * Belize 11.2 /1985 * Peru 10.5 /1986 * Burkina Faso 10.5 /1987 Sierra Leone 13.2 /1987 Eurundi 21.6 /1986 * Somalia 12.5 /1978 Comoros 13.8 /1987 Sri Lanka 10.7 /1988 * Egypt 11.5 /1987 * Sudan 17.5 /1982 Gambia 13.6 /1987 Togo 11.0 /1986 Ghana 13.5 /1987 Tunisia 13.1 /1984 Guinea-Bissau 24.1 /1987 Uganda 16.0 /1986 * India 26.1 /1987 * Vanuatu 12.5 /1986 * Iran 11.9 /1985 Venezuela 12.5 /1986 Myanmar 14.6 /1985 * Yemen AR 13.2 /1987 Source: Table 9 * Countries identified in the Fund's Annual Report on Exchange Arrangements and Exchange Restrictions with "cost-related import restrictions." 31 Based on the Fund's 1989 Annual Report on Exchange Arrangements and Exchange Restrictions, and supplementary information from the Fund's Government Finance Statistics file. 50 VII. VOLATILITY in EXCHANGE RATES and TRADE VOLUMES 70. In addition to taxes and subsidies and the use of multiple exchange systems in international trade directly applicable to the conversion of local currency into foreign exchange, the conversion factor should also encompass volatility in exchange rate and volume of trade during the period in question. The following paragraphs of the paper examines the merchandise-trade weighted conversion factor (MWXR) that captures volatility in the exchange rates and traditional cycles or fluctuations in the volumes of trade transactions. Because of large variations in the official rate, the MWXR encompasses the recommendation stated in Section II, para. 13 above: that shorter period data (e.g., quarterly exchange rates and trade volumes) should be used and then summed to derive longer period data, e.g., annual exchange rates and trade volumes. Merchandise-Weighted Conversion Factors 71. Each average quarterly exchange rate from the IFS is weighted by the respective shares of the corresponding quarterly total merchandise trade (exports plus imports) in the annual total trade according to this relationship: 51 Q4 MWXR =L. RF, i= Q where MWXR = merchandise-weighted conversion factor; wj= merchandise trade weight in Quarter i (xi + mi) 2:(xi + in') x, = merchandise exports in Quarter i; mi = merchandise imports in Quarter i; RFj = official exchange rate in Quarter i; i = Quarter 1,2,3 4 Deviations of Merchandise-Weighted Exchange Rates 72. The magnitude of the deviations of merchandise-weighted rates (MWXR) from the average annual official rates for the period 1980-88 are presented in Table 13. A closer examination of the percent deviations of MWXR from the official rate reveals that for most countries the deviations are relatively minor. Consistently about 95 percent of the countries exhibit deviations of less than 5% over the period 1980-88, as shown in Table 13. Thus, it is evident from the results presented in Table 13 that the MWXRs, in general, do not reveal anything significant about the preference order between the MWXR and the official rate for most countries. 52 Table 13: Distribution of Deviations of MWXR from IFS RF Deviation (%) 1980 1981 1982 1983 1984 1985 1986 1987 1988 0 15 14 14 12 9 8 7 10 7 0.1-1.0 62 54 52 53 49 52 59 50 40 1.1-2.0 2 7 7 6 12 9 5 3 7 2.1-3.0 2 2 2 2 3 4 4 4 0 3.14.0 1 2 2 1 0 0 0 3 6 4.1-5.0 0 1 2 0 1 2 0 0 1 5.1-6.0 1 U 0 1 2 0 0 2 0 6.1-7.0 0 0 0 1 2 1 1 0 1 7.1-8.0 0 0 2 1 0 1 0 1 0 8.1-9.0 0 1 0 0 1 0 0 0 0 9.1-10.0 0 0 0 0 0 0 1 2 0 10.1-15.0 0 2 0 1 1 2 0 2 0 15.1-20.0 0 0 1 2 0 1 2 0 2 20.1-25.0 0 0 0 0 0 0 0 0 0 25.1-30.0 0 0 0 0 1 0 0 0 1 > 30.0 0 0 0 2 1 0 1 0 2 Total 83 83 82 82 82 80 80 77 67 Source: IFS Countries for Review and Adjustment for Volatility 73. Nevertheless, any large deviations should be taken as signals that the MWXR could be the preferred choice among the two rates. The results do suggest, however, that for three countries in particular the use of MWXR, instead of the official exchange rate, merits consideration. The three countries showing strong deviations are Argentina, Brazil, and Yugoslavia. Any decision on the preference order between MWXR and official rate would have to take into account relevant country-specific knowledge, not just the magnitudes of the deviations between the two rates. 53 74. Argentina offers some interesting observations which account for the relatively large percent deviations of its MWXR from the official rate. A closer look at the quarterly official exchange rate (Table 14 below) reveals significant volatility in the quarterly rates in the 1980s. TabLe 14: ARGENTINA --Fluctuating Quarterly Exchange Rates and Trade Weights IFS Quarterly Rates % Change 01 02 03 04 02 03 04 1980 0.00017 0.00018 0.00019 0.00020 7.0 5.2 3.5 1981 0.00022 0.00038 0.00051 0.00065 71.3 36.5 27.0 1982 0.00103 0.00137 0.00377 0.00420 33.2 176.0 11.4 1983 0.00575 0.00782 0.01073 0.01781 35.9 37.3 66.0 1984 0.02784 0.04118 0.06934 0.13224 47.9 68.4 90.7 1985 0.24952 0.55672 0.80050 0.80050 123.1 43.8 0.0 1986 0.80050 0.84896 0.97216 1.15051 6.1 14.5 18.3 1987 1.40493 1.61000 2.14627 3.41599 14.6 33.3 59.2 1988 4.37877 6.79771 11.18150 12.65240 55.2 64.5 13.2 Trade Weights Z Change 01 02 03 04 02 03 04 1980 0.2172 0.2240 0.2662 0.2926 3.1 18.8 10.0 1981 0.1263 0.2403 0.3309 0.3024 90.2 37.7 -8.6 1982 0.1439 0.1970 0.2455 0.4136 36.9 24.6 68.5 1983 0.1308 0.2002 0.2643 0.4048 53.1 32.0 53.2 1984 0.1071 0.1836 0.2831 0.4262 71.5 54.2 50.5 1985 0.0953 0.2654 0.3336 0.3057 178.3 25.7 -8.4 1986 0.1781 0.2427 0.2876 0.2916 36.3 18.5 1.4 1987 0.1408 0.1903 0.2613 0.4076 35.2 37.3 56.0 1988 0.0897 0.1838 0.3500 0.3765 104.9 90.5 7.6 % Deviations of taR from IFS RF 1980 0.7 198513.5 1981 10.2 1986 2.2 1982 18.9 198714.6 1983 17.4 198817.9 1984 27.9 Source: IFS 54 This is compounded by the fluctuations in Argentina's quarterly trade shares where, traditionally, a high proportion of total trade occurs in the third and fourth quarters. Together, they account for the large deviations of the merchandise-weighted conversion factor form the unweighted official exchange rate. A closer look at Table 14 reveals that the small deviation in MWXR in 1980 is the result of relative stability in exchange rates and the dispersion of total trade, by quarters, in 1980; similar observations are seen in 1986. The outcome of the high volatility in exchange rates and trade shares in the other years is obvious: large deviations between MTXR and official rate. It is obvious that while the official rate encompasses the unweighted exchange rate changes over the entire year, its suffers from the drawback of not encompassing the volumes of transactions at each of these fluctuating rates; hence the need for a weighted conversion factor based on shorter period (quarterly) data. 55 VIII. USE of PERIOD AVERAGE EXCHANGE RATES in the BANK 75. In principle, the Bank adopts the unweighted annual average of the official exchange rate (line rf in IFS) for conversions between national currency and U.S. dollar values, except in those instances where it was seriously considered that this rate was "inappropriate" and that consideration of an alternative conversion factor is warranted only when the difference between the official exchange rate and that applying to the average of foreign transactions is egregiously large. Shorter Periods to Longer Period Rates 76. Market rates, however, vary from day to day, but the various types of data series that need to be converted to a common currency (e.g., GNP, foreign trade, balance of payments, etc.) are compiled over a period -- usually quarters or years. Thus longer period data are the sums of the compilation period. It follows that the use of prevailing market rate conversion requires the use of averages. 77. The primary requirement for converting longer period data is that the conversion should be made from shorter period data and that longer period data be calculated as the sums of the converted short period data. For example, the monthly (quarterly) foreign trade statistics, if available, in national currencies should be converted with the respective 56 average monthly (quarterly) exchange rates and then summed to obtain the annual data instead of summing the monthly (quarterly) data and then converting with the average annual exchange rates. The major drawback of the latter instance is that the annual average exchange rates may not be in accord or conform with the known contributions of the annual data within the monthly (quarterly) time periods. Given country X's traditional, say, fourth quarter's trade or current account surplus together with fluctuating fourth quarter's exchange rate, any difference in the conversion procedure will result in noticeable effects on the longer period data expressed in the common numeraire. 78. The recommendation is obviously clear ... that whenever shorter period data are available these should be converted at the average of monthly (or even daily) rates and then summed to derive longer period data, e.g.,annual data. One should, however, draw the distinction between what is desirable and what is statistically feasible. Consideration of statistical feasibility would pr3clude deriving an overall conversion factor based on highly disaggregated levels involving various commodities at respective rates. More often than not many macroeconomic data (e.g., GDP, GNP) from developing countries are not available for the shorter periods but are usually available as annual observations. In the extreme circumstance where volatility in the exchange rates is accompanied by fluctuating volumes of international transactions, the adequacy of the official exchange rate as a conversion factor becomes questionable. An alternative conversion factor in such a situation may be in the form of weighted exchange rates at a highly aggregated level e.g., merchandise trade weighted exchange rate (discussed in Section VII above) which will capture traditional cycles not only in the data but also in exchange rates. 57 Search for a General Purpose Conversion Factor 79. Much of the Bank's analytical work on development in member countries and on the efficacy of policies require inter-country (region) comparisons which require, at some stage, the conversion of indicators expressed in national currencies into a common numeraire. While it is clear that the preferential method of conversion is to use the prevailing or market valuations, their determination is not without ambiguity, even for international transactions that are relatively free of exchange and trade restrictions. Furthermore, the national currency value of a country's GNP does not change at the moment its exchange rate changes; it is the measurement of its value in foreign currency that changes. Many value aggregates in GNP are not aggregates of international transactions and there is, therefore, no uniquely "true" conversion factor. They are converted to a common currency only for the purpose of international comparison. Thus the search for an appropriate conversion factor should be viewed as a search for a general purpose conversion factor that applies to major economic aggregates. 58 BIBLIOGRAPHY Balassa, Bela, "Industrial Protection in Developing Countries," Report No.EC-175 Unpublished; (The World Bank, June 1980). Culbertson, William Patton, "Purchasing Power Parity and Black-Market Exchange Rates," Economic Inquiry, Vol. 13, June 1975; pp. 287-296. Culbertson, W. Patton, "Empirical Regularities in Black Markets for Currency," World Development, Vol. 17, No. 12 (Cambridge, MA: Harvard Institute for International Development, December, 1989), pp. 1907-1919. Goldstein, Morris, and Lawrence H. Officer, "The Measures and Productivity for Tradable and Nontradable Goods," Review of Income and Wealth, Vol. 4, December 1979; pp. 413- 425. Gupta, Sanjeev, Black Market Exchange Rates (Kiel University Institute for World Economics, 1981). International Monetary Fund, International Financial Statistics (Washington DC: International Monetary Fund, various issues). International Monetary Fund, Annual Report on Exchange Arrangements and Exchange Restrictions (Washington, DC: International Monetary Fund, 1989). International Monetary Fund, Exchange Arrangements and Exchange Controls, (Washington, DC: International Monetary Fund, 1989). International Monetary Fund, "Developments in International Exchange and Trade Systems," World Economic and Financial Surveys (Washington, DC: International Monetary Fund, September 1989). International Monetary Fund, "Review of Experience with Multiple Exchange Rate Regimes", (SM/84/64), Unpublished (March 19, 1984). International Monetary Fund, "Review of Multiple Exchange Rate Regimes - Background Information", (SM/84/65), Unpublished (March 20, 1984). International Monetary Fund, "Multiple Currency Practices Applicable Solely to Capital Transactions", (SM/85/19), Unpublished (January 19, 1985). 59 Johnson, G.G., "Formulation of Exchange Rate Policies in Adjustment Programs," IMF Occasional Paper, No.36 (Washington, DC: International Monetary Fund, August 1985). Jones, Christine, and Michael Roemer, "Editors" Introduction: Modeling and Measuring Parallel Markets in Developing Countries," World Development, Vol. 17, No. 12 (Cambridge MA: Harvard Institute for International Development), pp.1861-1870. Koveos, Peter E., and Bruce Seifert, "Market Efficiency, Purchasing Power Parity, and Black Markets: Evidence from Latin America," Review of World Economics Vol. 2, No. 122, 1986; pp. 314-326 Laker, John F., "Fiscal Proxies for Devaluation: A General Review," Unpublished (International Monetary Fund, October 21, 1980). Lanyi, Anthony, "Separate Exchange Markets for Capital and Current Transactions," International Monetary Fund, [NIF Staff Papers, Vol. 22 (Washington, DC: International Monetary Fund, November 1975), pp. 714-49. Lindauer, David L., "Parallel, Fragmented, or Black? Defining Market Structure in Developing Economies," World Development, Vol. 17, No. 12 (Cambridge, MA: Harvard Institute for International Development), pp. 1871-1880. Mackenzie, George A., and Susan M. Schadler, "Exchange Rate Policies and Diversification in Oil Exporting Countries," Unpublished (International Monetary Fund, May 9, 1980). Mansur, Ahsan H., "Determining the Appropriate Levels of Exchange Rates for Developing Economies," International Monetary Fund, IMF Staff Papers, Vol. (December 1983), pp. 784-818. Officer, Lawrence H., Purchasing Power Parity and Exchange Rates: Theory, Evidence and Relevance (Greenwich, Connecticut: Jai Press Inc. 1982) Pick's Currency Yearbook (various issues). Shapiro, Alan C., "What Does Purchasing Power Parity Mean?" Journal of Money and Finance, Vol. 2, No. 3, December 1983; pp. 295-315 PRE Workina Paper Series Contact Mae ~ ALghor 12-12aper WPS455 A Formal Estimation of the Effect Junichi Goto June 1990 M. T. Sanchez of the MFA on Clothing Exports 33731 from LDCs WPS456 Improving the Supply and Use of S. D. Foster June 1990 Z. Vania Essential Drugs in Sub-Saharan Africa 33664 WPS457 Financing Health Services in Africa: Germano Mwabu June 1990 Z. Vania An Assessment of Alternative 33664 Approaches WPS458 Does Japanese Direct Foreign Kenji Takeuchi June 1990 J. Epps Investment Promote Japanese Imports 33710 from Developing Countries? WPS459 Policies for Economic Development Stanley Fischer June 1990 WDR Office Vinod Thomas 31393 WPS460 Does Food Aid Depress Food Victor Lavy July 1990 A. Murphy Production? The Disincentive 33750 Dilemma in the African Context WPS46' Labor Market Participation, Returns Shahidur R. Khandker July 1990 B. Smith to Education, and Male-Female 35108 Wage Differences in Peru WPS462 An Alternative View of Tax Anwar Shah July 1990 A. Bhalla Incidence Analysis for Developing John Whalley 37699 Countries WPS463 Redefining Government's Role in Odin Knudsen August 1990 K. Cabana Agriculture in the Nineties John Nash 37946 WPS464 Does A Woman's Education Affect Shoshana Neuman August 1990 V. Charles Her Husband's Earnings? Results Adrian Ziderman 33651 for Israel in A Dual Labor Market WPS465 How Integrated Are Tropical Timber Panos Varangis August 1990 D. Gustafson Markets? 33714 WPS466 Is There An Intra-Household Lawrence Haddad August 1990 J. Sweeney Kuznets Curve? Ravi Kanbur 31021 WPS467 Structural Adjustment and Living Nanak Kakwani August 1990 B. Rosa Conditions in Developing Countries Elene Makonnen 33751 Jacques van der Gaag WPS468 Does the Structure of Production Indermit Gill August 1990 M. Abundo Affect Demand for Schooling in Peru? 36820 PRE Working Paper Series Contact Authat for ape WPS469 Modeling Economic Behavior in J. Barry Smith August 1990 M. Abundo Peru's Informal Urban Retail Sector Morton Stelcner 36820 WPS470 What Do Alternative Measures of Alexander J. Yeats August 1990 J. Epps Comparative Advantage Reveal 33710 About the Composition of Developing Countries' Exports? WPS471 The Determinants of Farm Gershon Feder August 1990 C. Spooner Investment and Residential Lawrence J. Lau 30464 Construcion in Post-Reform China Justin Lin Xiaopeng Luo WPS472 Gains in the Education of Peruvian Elizabeth M. King August 1990 C. Cristobal Women, 1940 to 1980 Rosemary Bellew 3:640 WPS473 Adjustment, Investment, and the Riccardo Faini AuguF' -90 R. Sugui Real Exchange Rate in Developing Jaime de Melo 37951 Countries WPS474 Methods for Measuring the Effect Anne Maasland August 1990 P. Dixon of Adjustment Policies on Income 39175 Distribution WPS475 Does Divestiture Matter? A Ahmed Galal August 1990 G. Orraca-Tetteh Framework for Learning from Experience 37646 WPS476 Health Insurance in Sub-Saharan Ronald J. Vogel August 1990 K. Brown Africa: A Survey and Analysis 35073 WPS477 Private Participation in the Delivery Thelma A. Triche August 1990 M. Dhokai of Guinea's Water Supply Services 33970 WPS478 Interrelations Among Child Mortality, John Marcotte August 1990 S. Cochrane Breastfeeding, and Fertility in John B. Casterline 33222 Egypt, 1975-80 WPS479 Conversion Factors: A Discussion Michael Hee August 1990 E. Zamora of Alternate Rates and Corresponding 33706 Weights