Policy Research WORKING PAPERS Trade Policy Country Economics Department The World Bank May 1992 WPS 915 Trade Policy and Exchange Rate Issues in the Former Soviet Union W. Max Corden One approach to trade policy among the former Soviet republics is to have no trade policy - to have completely free trade with convertibility for current account transactions. Trade policy should be transparent. Any tariff and export tax structures should be simple. Quantitative controls should be avoided. And no barriers to existing trade between the republics should be intro- duced. PolicyRearchWo*ingPapau din aw the findings of workin progs and encouragethe exchangeof id -as amongBank staffand alloe itas ted in devdlopmatiw uesThesepspcs. distnbutedby theReseuch Advisay Staff.cany thenames of dieauthors,rdefct onlythitviews, and shm1 d beused and cited accordingly. hefindings, i etation,ndon clusions arethea uthozs'own NTheyshould na beulurted to the World Bank. its Board of Disctor, its management, or any of its member counties. Policy Resarch \ . ~~~~~~~~~~~~~~Trade Policy WPS 915 This paper -a product of the Trade Policy Division, Country Economics Department -was prepared under the auspices of the UNDP/ World Bank Trade Expansion Program. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington DC 20433. Please contact CECTP, room N10- 023, extension 37947 (May 1992, 53 pages). Taking the long view (assuming prices have The case for a free trade area is strong been liberalized), Corden reviews possible trade because the republics of the former Soviet Union and exchange rate policies of the republics of the are so highly specialized, but there will be former Soviet Union. problems if price controls remain and differ among the republics. There could be a free trade He considers alternative exchange-rate area even if there is no monetary union. A regimes, including a monetary union. For Russia, customs union, involving a common extemal a fixed but adjustable regime is most realistic. tariff, would prevent border controls and trade Frequent adjustment may be desirable, to prevent deflection but would make a common tariff the use of trade restrictions to achieve balance of policy (and hence a joint tariff commission) payments objectives. necessary. Trade intervention policies, if needed, should One approach to trade policy, Corden not weaken the fiscal situation. There may be a concludes, is in effect to t"ve no trade policy- case for export taxes or for a uniform revenue to have completely free trade with convertibility tariff, for example, but subsidies and quantitative for current account transactions. Some tariffs and restrictions should be avoided. If some domestic export taxes may be justified, at least as second- price controls remain, export taxes are needed. best policies. If so, however, Corden stresses that four principles be observed: Corden examines the need for transitional tariffs, including the argument for a temporary * Barriers to existing traWz between the uniform tariff that is higher than the long-run republics should not be set up. revenue tariff. The temporary uniform tariff * All quantitative control measures should be would be designed to prevent temporary over- avoided. shooting of the exchange rate - but the argu- * Tariff and export tax structures should be ment is subject to important qualifications. If very simple. there is any protection of infant industries, there * Trade policy should be transparenL must be a "hard tariff' path - the gradual decline of the tariff rate should be firmly pro- vided for in advance. ThePolicy Research Woking PaperSeriesdisseniinates thefindings of work under way in theBanlk Anobjectiveof 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 Policy Research Dissemination Center Table of Contents I. Preliminaries p2age 1. Introduction .............******** ....*..... .. 1 2. Definitions ..................................................... 4 3. Monetary Overhang, Fisjal Deficits and General Disequilibrium ........................... 7 4. Extreme Specialization ............................. 11 5. Disequilibrium Pricing .......................... 13 II. ExchanQe Rate Policy 1. Exchange Rate Policy for Russia ..................... 17 2. Relationship between Trade Policy and Exchange Rate Policy .......................... 21 3. Exchange Rate Policy for the Other RepuUl1ics ...... 23 4. Monetary Integration ......................... 25 III Trade Policies General Remarks ................................... 27 1. Trade Taxes for Revenue ........................... 28 2. Policy Instruments to Avoid . .................... 29 3. Tariffs for Balance-of-Payments Reasons ............ 31 i 4. Is There Need for Transitional Protection? ....* 33 5. Infant Industry Protection and a Hard Tariff Path .. 37 IV The Free Trade Area Issue General Remarks ***********e**...................... 38 1. PossiDle Arguments against a Free Trade Area ...... 39 2. Free Trade Area or Customs Union? ................... 43 3. Monetary Integration and Trade Integration ...... 45 Some General Princivles General Remarks ............ ........................ 47 Appendix . . ............ 50 Bibliography ................ *.... .................... 52 This paper was prepared under the auspices of the UNDP/World Bank Trade Expansion Program. ii PRELIMINARIES 1. INTROLUCTION This paper aims to discuss analytically a number of issues likely to arise when considering future trade policy and economic integration arrangements of the Republics of the former Soviet Union. Particular attention will be paid to the relation between trade and macroeconomic policies, especially exchange rate policy. In fact, policy towards the exchange rate and the related "convertibility" issue must be central to the discussion. We can distinguish those issues that are special to the former Soviet situation from those which are familiar in market economies. Naturally the former are more interesting and also less understood, so they will be - -cussed in some detail below. But the latter issues are so important and will become increasingly relevant as the Republics acquire more and more the characteristics of market economies. Not all the problems and choices facing the republics of the former Soviet Union are unknown. Some are familiar problems where there is much experience and analysis to draw upon from market economies, both developed and developing. Examples are the use of trade policy to encourage or ensure domestic competition, versus its use for I I am indebted to comments on an ear'lier draft from Jaime de Melo, John Nash and David Tarr. 1 job protection; the infant industry argument for protection; and the relationship between exchange rate and trade policy. Out of the Soviet Union has emerged one very large economy, Russia; one medium-sized economy, the Ukraine; and thirteen small ones, the latter being .lt very different stages of development. At the time of writing (January 1992) Russia was taking radical steps to move towards free market pricing, but it was not clear which of the other Republics would follow. One might expect the Baltic Republics and Georgia to move (or try to move) towards the market fairly rapidly, and four or five of the central Asian republics to move more slowly, if at all, while the matter is still uncertain with regard to the others. But all such remarks are just speculation and can soon be negated by events.2 In looking ahead it is useful to think in terms of three time periods or episodes. The first is the transition -eriod during which a Republic takes the essential steps to create a domestic market economy, freeing most domestic prices and creating incentives for state-owned and privately-owned enterprises to respond in a market way, notably by removing various restrictions, and providing the necessary legal framework, taxation system, and so on. Possibly this stage would require extensive privatization. There is no shortage of discussion of the characteristics and vast problems of such a transformation, and here it will be assumed that it takes place, 2 For background on the recent state of the Soviet economy, see International Monetary Fund, et al. (1991) and Aslund (1991). Many of the issues of this paper are discussed in International Monetary Fund, et al. (1991), in Williamson (1991), and in Havrylyshyn and Williamson (1991), the latter dealing particularly with monetary stabilization, with exchange rate policy, and with the question of the financing of bilateral payments imbalances. The last subject is not discussed here. See also Tarr (1991), Havrylyshyn and Tarr (1991), Cooper (1991), and Gros and Steinherr (1991). 2 whether in an orderly or a chaotic way. One might imagine this stage to take two years (but that is just a casual guess!). Se;ondly, there is the adjustment geriod. During this period the economy is mainly a market economy; the institutiors and rules of a new system have been (more or less) installed, but industries and the population are going through an extended adjustment process, involving major resource reallocation, income redistribution, and accumulation of new forms of physical and human capital. The economy is adjusting to the changes in institutions and rules. I would expect this stage to last perhaps ten years. The third stage is the normality stage when the economy is a fairly normal market economy, with normal (and familiar) problems, but the drastic adjustment process in response to the transition from socialism is clearly at an end. This paper will say only a little about trade policy during the first period. It will be mainly concerned with the second period and, to a lesser extent, with the third. Thus the paper really looks ahead and assumes a successful transition. Nevertheless, most of the issues discussed are currently relevant for those Republics, notably Russia, where there is already a partial transition--since the foundations for the second period are beinag laid. At various points, "first-best" policies will be proposed. But one cannot be under any illusion that such policies will necessarily be followed, no more than first-best policies from a national (as distinct from a special interest group) point of view are consistently followed in the major market economies. Hence, there has to be a good deal of "second-best" analysis. 3 one preliminary warning is perhaps obvious. A paper which is designed to clarify some issues cannct claim to have simple solutions to problems that everybody knows to be immense and possibly not solvable by means that are socially or politically acceptable. Furthermore, there is complete uncertainty about the starting point: what will the economies loolc like at the time when coherent trade and exchange rate policies will actually begin to be implemented (at some stage during tlh transition)? Nor do we know when this will be. 2. DEFINITIONS (a) The Meaning of Trade Policy One should first define "trade policy." In the socialist system it consists of the processes and decisions which determine quantities of various goods and services imported and exported, and the prices at which trade will take place. This involves negotiations with foreign suppliers and buyers and then central instructions to domestic suppliers of exports to ensure that the trade commitments are fulfilled. As is well known, Soviet trade took place through the intermediation of a limited number of foreign trade organization.; there was not a direct relationship between So -4et export prodi-ers or users of imported inputs and foreign purchasers or suppliers.3 In the market system the matter is different. Trade policy does involve government negotiations with foreign governments with regard to their trade restrictions and also commitments with regard to the country's own restrictions (for example, through membership in GATT), and to that extent there is some similarity. 3 For descriptions of the system, see Aslund (1991) and International Monetary Fund, et al. (1991, Chapter IV.3). 4 But this aspect will not be discussed here. For given foreign restrictions, quantities and prices in trade are determined in a decentralized way and do not necessarily involve "policy" at all. There is no need for centralized det-rmination of quantities traded or of the prices at which trade takes place. Of cotrse, there is still a role for official guidance about demand and' supply prospects--i.e., for the provision of information--and for the provision of infrastructure which is required for all economic activities. In this paper much attention must be paid to exchange rate and exchange control policies--though these are not conventionally included in the term "trade policy." The conventional usage defines trade policy as referring to various interventions in trade, with the exchange rate as giveni: it refers to tariffs, quantitative import restrictions, export subsidies and taxes and (in a broader definition), subsidies and taxes for domestic production and consumption that discriminate in favor or against specific tradable goods. If there is a single, unified exchange rate with full current account convertibility (discussed below), and there are no such interventions, then there is "free trade"--the limiting case of a passive trade policy. Much of this paper will be concerned with exchange rate policy and with various possible departures from free trade, especially in the form of tariffs and export taxes. (b) The Meaning of Convertibility In the Soviet Union the distinction was made between commodity convertibility and currency convertibility. The former referred to the ability to convert domestic currency (roubles) into domestic goods and services. Such convertibility is, of course, taken for granted in a market system. There was absence of commodity convertibility for two reasons. First, the savings of 5 enterprises in "enterprise accounts" in banks were not allowed to be spent, i.e., to be converted by the enterprises without special permission. They were blocked. Second, households and enterprises (notably cooperatives) that were allowed to spend their roubles were not always able to buy goods with them owing to generalized shortages. Here it will be assumed that during the transition to a market economy, full commodity convertibility is established. Currency convertibility refers to the ability to crnvert roubles freely into foreign currency. Outside the former Soviet Union the term "convertibility" refers just to this concept. But here a further distinction can be made between current account convertibility and unrestricted convertibility, the latter including also convertibility for capital account transactions.4 When there are exchange controls on capital account transactions, capital account convertibility does not exist or, at least, is not complete. Finally, in the former socialist countries of Eastern (or Central) Europe, the distinction is also made between current account convertibility for domestic residents, called internal convertibility, and convertibility for foreigners, called external convertibility, the latter concerning particularly remittances of interest and dividends, as well as the ability to repatriate capital (the latter being part of capital account convertibility). For the present paper the convertibility that is relevant is current account convertibility for domestic residents, and the simple term "convertibility" will henceforth refer to this. Two points need to be stressed about convertibility. First, 4 See Williamson (1991, pp. 376-380) for discussion of concepts of convertibility. 6 if A currency becomes convertible, this does not mean that it has to be convertible at a fixed exchange rate. The exchange rate can float, or be adjusted flexibly. It is often said that convertibility requires foreign financial support, implying that, because of speculation (indirect zapital movements, perhaps throuqh leads and lags) or other reasons, the Republic's foreign exchange reserves might run out unless confidence is established by the existence of an adequate reserve of available foreign currency. This would be true if there were a commitment to a fixed exchange rate. In the a' ;ence of such a commitment it would also be true if it were desired to avoid a sharp depreciation of the currency once convertibility was established. But if there is a willingness to accept exchange rate flexibility, and thus the possible overshooting ot the exchange rate in the short run, convertibility can be established without large foreign exchange reserves or foreign assistance. Second, when a currency is not completely convertible (for current account transactions by residenta), the effect is much the same as when import restrictions are imposed. Inconvertibility--or convertibility subject to controls--has much the same effect as trade policy interventions. Of course, administratively there may be differences, and it all depends on the details. But if the currency is not completely convertible (in the sense defined here), one cannot say that there is free trade e:en if there are no tariffs or quantitative import restrictions. It will now be assumed that complete current account convertibility for residents is established, this being an element of the transition tc a market economy. Trade policy thus takes the form of tariffs, import restrictions, export taxes and subsidies, and so on. In zxlition, there can be exchange controls 'or capital account transactions; it is suggested by man., commentators that such controls should be maintained, but this issue goes beyond the scope of the present paper. 7 3. MONETARY OVERHANG. FISCAL DEFICITS AND GENERAL DISEQUILIBRIUM The following remarks apply to the situation at the end of 1991. At that point (and still at the time of writing in January 1992), there was generalized excess demand resulting from a massive monetary overhang which had been rapidly increasing owing to out-of-control fiscal deficits. While the seeds were laid earlier--by financially irresponsible policies 1985 and 1986--the policies that led to the serious crisis of 1989 really began in 1988, with the Law of State Enterprises. Aslund (1991, Chapter 7) describes in fascirating detail and very clearly how the eycess demand situation and consequent crisis came abotut. In addition to the central government and enterprise deficits, the financial balances of the enterpises became more liquid (for several reasons), so that not only the money supply but also velocity increased. This is a simple explanation for the increasing open inflation in that part of the economy where prices have been flexible. But at the end of 1991 the fomer Soviet Union was not in a typical situation of high inflation. In the larger part of the ex-Soviet economy, prices were fixed, so that there was excess demand, generating the shortages and queuing wi-ch have dominated Soviet urban life. Since many goods have simply been unobtainable (except with much effort), the incentive to earn more roibles has been reduced, and this appears to have been a major element in the contraction of supply. This problem would disappear if prices were generally freed (or the relative size of the market sector were greatly expanded), in which case repressed inflation would be converted into open inflation. This process may now be underway, or would be if price liberalization were continued while excessive growth of the money supply to finance deficits continued. But there has been a further complication, unique to the Soviet system. As noted above, roubles credited to enterprises and 8 deposited in banks have not been available to them automatically to be spent freely. Thus, there have been two reasons for lack of commodity convertibility, and thus two reasons for a lack of incentives to earn roubles by producing and selling goods There may be other reasons also for the decline in supplies, given that the command-and-control system of determining supplies has broken down. But the monetary disequilibrium combined with the fixing of major prices provide sufficient explanation. In any case, the establishment of a domestic market system clearly requires both the freeing of most prices and the unblocking of enterprise accounts. Price increases could naturally eliminate the mcnetary overhang, though other methods are also possible, such as the conversion of currency and monetary deposits into longer-term bonds of some kind (possibly usable eventually for the purchase of privatized assets). In addition, it is clearly necessary for the fiscal and enterprise deficits to be ended or greatly reduced. In the absence of macroeconomic stabilization, the economy would convert into high or hyperinflation. As can be observed from the experience of various Latin American countries, notably Argentina and Brazil, high open inflation, while having adverse effects on growth and in extreme cases (such as Brazil recently) actually leading to declines in output, it need not lead to the kind of precipitous declines in output that have been seen in the former Soviet Union. Thus, the conversion of repressed into open inflation--arn hence the creation of commodity convertibility-- would in itself be beneficial for supplies. It should be added that Brazil (the major current example of extremely high inflation) has been able to adjust to very high inflation to some extent because of the availability of financial institutions and indexing arrangements that take time to evolve and that, in any case, are not available in the former Soviet Union. 9 Nevertheless, in high inflation (end even more, hyperinflation), relative prices also get highly distorted and prices fail in their signalling role. In other words, unless there is macroeconomic stabilization, in addition to domestic price liberalization, there is little point in attempting coherent trade policies of the kind to be discussed below. Hence, I shall now assume5 that the overwhelming macro problem has been dealt with. This does not mean that one has to assume zero inflation and zero fiscal deficits,'but they have to be moderate and under control. This, of course, is not a forecast, but is a starting point required for a coherent trade policy. Once stabilization has been brought about, the maintenance of fiscal balance and strict monetary control should be the first concern of all policies, and all other concerns should be secondary. It must never be forgotten that the transformation of the era of stagnation into the years of crisis is essentially explained by the failure of fiscal and monetary control.6 This has important implications for trade policy. The fiscal implications of policies become crucial. There may be a strong argument to impose trade taxes in order to raise revenue, if this is a convenient or politically acceptable way of taxing. Such an argument for tariffs or export taxes should outweigh a concern with the undesired protection of import-competing industries that would be a by-product. Proposals for subsidies should be ruled out since, indeed, so many existing subsidies have to be eliminated. The subsidy habit should be ended. In standard trade theory, a case is often made in favor of the use of subsidies rather than tariffs or import restrictions to deal with "domestic distortions." But in 5 Like the economist in the famous can-opener fable. 6 See Aslund (1991). 10 the current and prospective situation of the former Soviet Republics, either tariffs (or export taxes) should be used, or the distortions themselves must either be eliminated or their consequences lived with. Furthermore, apart from all other considerations, tariffs should always be preferred to import restrictions because of the favorable revenue effects. At the beginning of the paper it was noted that there are some issues that are special to the former Soviet Union, while others are familiar. Issues connected with open inflation are certainly not special, but those resulting from large-scale repressed inflation and hence commodity inconvertibility are. Thus, this is the first of the "special" issues. We now turn to two other's. 4. EXTREME SPECIALIZATION A special feature of the ex-Soviet economy is extreme specialization of production. It has resulted from a Soviet planning system which reflected an excessive belief in economies-of-scale and placed no value on the availability of product variety or competition. Reducing the number of enterprises to which the planners sent their instructions was also meant to make planning easier. The net result has been a degree of concentration of production that is quite extraordinary by Western standards.7 This extreme specialization has a number of implications. First, it means that a large volume of trade between the Republics will have to continue, even if they do have their 7 For details, see International Monetary Fund, et al. (1991, pp. 36-38). 11 separate trade policies and trade with the outside world is freed. In a free market situation, they are likely to do much more trade with each other than with the outside world. The imposition of restrictive import or export quotas on trade with each other could have severe costs and should be ruled out. Any measures that improve the convenience or smoothness of inter-Republic trade will be beneficial, so that, for this reason, there is a strong case for a free trade area. Opening up to outside trade may well reduce interdependence between the Republics somewhat, but this will be limited by the ability to generate exports to obtain alternative imports from abroad. Second, extreme specialization means that, in a free market, and in the absence of new competition from outside, there would be many situations of near or complete bilateral monopoly, leading to difficult bargaining situations that could get politicized and induce trade frictions. It is then particularly desirable to open external trade so as to increase competition. Such opening of trade to the outside world may not always lead to actual increases in imports from, or exportz to, the outside world, but the potential trade will provide a discipline for both buyers and sellers. The most common argument for free--or relatively free--trade is that it encourages international specialization of production on the basis of comparative advantage. This certainly applies to the Republics, as to all other countries. But a second, distinct argument is that free trade provides competition from imports for local producers, and competition from export demand for local monopsonistic buyers of domestic products. This argument is overwhielmingly important in the case of the former Soviet Union, where domestic production is so highly concentrated that there is an exceptional need for the discipline of competition to come from abroad. 12 Third, extreme specialization is likely to lead to strong pressures to reduce interdependence by pursuing policies of import substitution within the separate Republics. The argument will be made that import substitution is necessary to increase competition, and perhaps to avoid exploitation by monopoly suppliers in other Republics. On the other hand, it is quite likely that when the centralized bias to extreme specialization is removed and market forces are allowed to work, local production of various products, perhaps differentiated somewhat from imports from other Republics, will begin even without protection. Hence, there is not necessarily an argument here for protection. More important, it must be remembered that imports from outside the former Soviet Union will also provide alternative supplies. 5. DISEOUILIBRIUM PRICING It seems reasonable to expect that currency convertibility will be introduced as a crucial step in the transition process and that prices will generally be freed. In any case, this will be assumed here. Nevertheless, it is probable that some prices will continue to be kept below their market value by price controls or arbitrary decisions. This is likely for food and energy prices, for example. There would no longer be generalized excess demand, but there would be excess demand for particular goods and services. A dual economy would thus emerge--a dominant market economy and a smaller non-market economy. Up to 1991 the controlled economy dominated while the market economy was relatively small. In the new situation the market economy would dominate but probably there would still be a significant 13 non-market economy.8 Thus, the third special feature of the ex-Soviet situation is the likelihood that a significant non-market sector--much greater proportionally than in most or all developing market economies--would remain. There will thus be a relative price distortion--or, indeed, many distortions--which optimal trade policy will need to take into account. One branch of the theory of trade policy is concerned with the relationship between "domestic distortions" and trade policy. The standard recommendation is that first-best policy is to remove the domestic distortions, and then accompany this with free trade, unless there are other arguments for trade intervention. Free trade means here complete convertibility of the Republic's currency for current account transactions, and absence of trade policy interventions, such as tariffs, export taxes, and import quotas. But it may be that this first-best policy of removing "domestic distortions" cannot be or will not be implemented. A distortion (such as price control on some basic goods) may be justifiable for distributional or political reasons. In that case, trade policy has to take the distortion policy as given, and try to offset some of its otherwise adverse effects. One intervention (trade policy) then modifies the by- product distortions that result from another intervention (sectional controlled prices, in this case). It is likely that this consideration will be important in forming trade policy in 8 The theory of dual exchange rates is applicable to studying the implications of such dual economies. Many developing countries have had an official--i.e., controlled--exchange rate, and a free or black market rate, and over time the relative importance of the latter has increased. The economic implications of such arrangements have been much studied (for the theory, see Corden, 1971, pp. 87-92), and the analysis is relevant for the socialist transition societies. 14 Petce FD pi P2 It I , txP02TAU5 PI0bucT V) 14a the Republics of the former Soviet Union, at least during the early adjustment stage. The price of an exportable product--say, energy in Russia or grain in the Ukraine--may be underpriced to domestic industrial and household consumers. If suppliers were free to export as much as they wished, there would then be an incentive to export more than they would if prices were flexible. Indeed, if domestic prices are rigidly fixed below export prices (allowing for transport costs), all domestic production would be exported in the absence of controls. Controls on exports, or an adequate export tax, would then be needed. This case is represented in Figure 1. The world market price is OP1. If there were no price control, exports would be LK. But there is a controlled price of OP2 which applies only to sales in the home market. In the absence of export controls, everything would be exported (i.e., OK) at the price OP1. Further controls need to keep some of the good at home. There are three possibilities. (i) Controls keep OL at home. Exports would be LK, which is the non-intervention result. But there would be excess demand at home of LJ. (ii) Controls keep 03 at home. Excess demand at home would be avoided, but there would be under-exporting of LJ, in the sense that an opportunity to export products at a cost less than the price received would be foregone. (iii) An export tax of P,P2 is imposed. Hence, the net price received by exporters would be reduced to P2, and production would fall to OH and exports to JH. Even without any price control, the price to domestic consumers would fall to OP2, so that price controls would become redundant (or easily enforceable). There would be no excess demand at home but under-exporting would be greatest, namely, at LJ+ HK. 15 Policy (i) would avoid under-exporting; that would be its advantage, but it would lead to excess demand at home. Solution (ii) would avoid excess demand at home, but would lead to under-exporting. Solution (iii) would have two advantages: First, it would eliminate the need for price control, with its enforcement problem; it is surely easier to enforce an export tax than a control on domestic sales. Second, it would bring in revenue from the export tax represented by the shaded area, surely a major attraction. On the other hand, it would lead to under-exporting not only through the diversion of given output t( the home market but also through reduced production. There is a fourth possibility, namely, to subsidize home consumption by PtP2 per unit. But, as noted earlier, that is the kind of policy which has led to the current macroeconomic problems in the former Soviet Union because of the fiscal cost, and hence can be dismissed. The issue just discussed is highly relevant at the time of writing (January 1992) when Russia is liberalizing prices but the other Republics have not done so. In spite of various controls, there is a considerable degree of free trade between the Republics. The natural tendency is then for producers of price-controlled commodities in, say the Ukraine, to export to Russia and keep their home market short, possibly failing to supply it completely. When there is general price liberalization in Russia but not in Ukraine, supplies will naturally be diverted on a large scale from Ukraine to Russia, the only limits being transport problems, and controls or taxes that Ukraine is actually able to enforce. Unless Ukraine imposes export taxes or controls, and is able to enforce them, the only other alternative is to follow Russia with price liberalization. The current policy issue is thus very similar to the more long-term issue discussed above. (Thus, price liberalization in Russia creates 16 pressure for others to liberalize, a case of a good policy driving out a bad one!) Finally, another case of a controlled price to domestic consumers may be noted here. This time it concerna a product where there are, or would be, imports rather than exports. The price of a particular food product may be underpriced to domestic consumers because of price controls. This price-control policy will then lead to inadequate domestic production, something that is beyond the power of pure trade policy to deal with. But it will have two other effects. First (assuming the controls also apply to imports), it will lead to less imports of the controlled product than would take place in the absence of controls, and possibly none at all. Second, it will lead to excess demand that is satisfied by imports of a substitute where there are no controls. There will then be both a distortion in the pattern of imports and a net increase in total imports as the consumption of the imported substitute replaces consumption of the locally-produced product. One possible second-best approach (which cannot actually remove the distortion) would be to tax imports of the substitute and use the revenue to subsidize imports of the price-controlled product. The general conclusion is that disequilibrium pricing should be avoided, and certainly should not be part of the third period--the period of normalcv. But, for distributional and political reasons, it is likely to be unavoidable during the early part of the adjustment period, and thus some export taxes, and possibly also tariffs, may be needed. 17 II . EXCHANGE RATE POLICY 1. EXCHANGE RATE POLICY FOR RUSSIA Exchange rate policy and trade policy are closely related. We begin with an extended discussion of possible exchange rate regimes. Here it is necessary to distinguish Russia from the other Republics. For Russia there are essentially three alternatives, namely, a fixed exchange rate regime, a flexible peg regime--where the exchange rate is fixed by the monetary authorities at a point in time by intervention, but is frequently or occasionally adjusted--and a floating rate regime, where the rate is fully determined by the market, without intervention. In the fixed rate regime the rate would be fixed to an outside currency, such as the D-Mark or the dollar, or to a basket, such as the ECU, the SDR, or a special trade-weighted oasket calculated specifically for Russia. For the other Republics there are the further alternatives of fixing the rate to the (Russian) rouble, of forming or joining a monetary union with Russia, and of forming a monetary union with some Republics other than Russia. It is inconceivable that Russia would establish a fixed rate regime of the first kind, and it can certainly not be recommended. Of course, the advantages of such an arrangement are well known. Inflation would be kept down, provided domestic monetary policy were adapted to the exchange rate commitment. The exchange rate commitment might provide a discipline on monetary policy. A decline in the foreign exchange reserves would have to compel a 18 policy of monetary contraction. If the commitment were seen to be firm, and possibly were embodied in a constitution either of the nation itself or of its central bank, speculation on the exchange rate should be avoided and, in addition, domestic wage demands would not be based on inflationary expectations. But it is highly unlikely that Russia would allow its monetary (and fiscal) policy to be determined so completely by what would appear as external considerations. Given that trade with the outside (non-ex-Soviet) world is, and will remain, a small proportion of GDP, one cannot believe that the foreign trade tail would be allowed to wag the monetary (and hence fiscal) dog. Quite apart from this consideration, there is also the well-known disadvantage of a fixed rate regime. An instrument of policy is given up. There are circumstances when exchange rate adjustment can play a useful policy role, principally when a country suffers an adverse shock--a decline in the terms of trade, a politically unavoidable r1se in the general wage level, or a cessation of capital inflow. In such circumstances, a rise in the relative price of tradables to nontradables may be required, and this would be brought about more easily and quickly by a devaluation than by squeezing domestic demand with the hope that eventually domestic prices of nontradables and wages would fall sufficiently. In the case of a nation going through a radical transition and adjustment process, an additional consideration is very relevant. There is complete uncertainty about the way in which the economy will evolve--its capacity to produce once the central macroeconomic problem is dealt with, its success in foreign trade, its methods of wage determination, and so on. Hence, it will be quite uncertain what the right exchange rate should be when the new regime is instituted. If it turned out to be wrong and yet could not be altered, owing to a firm exchange rate commitment having been made, the domestic level of nominal wages 19 might need to be adjusted to turn the fixed nominal exchange rate into the "right" rate. This would be no prublem if the initial rate were undervalued, since wages could easily rise, but it would be a big problem if the exchange rate were grossly overvalued. If the exchange rate were finally devalued in spite of the initial commitment to keep it fixed, the government's credibility would suffer. Another possibility is a floating exchange rate regime. But this can also be dismissed. One can conceive of a floating exchange rate for a limited number of transactions--for example, for tourism and capital movements. But it is improbable that Russia would willingly choose a unified floating rate for all its foreign trade transactions for a prolonged period (i.e., other than during a brief transition). The people of Russia have been used to excessive price stability and have a great dislike of "speculation." Hence, they are hardly likely to accept a system where the exchange rate for important transactions--such as food imports and energy exports--fluctuates day by day, often in ways that mystify even the market participants themselves, let alone academic observers. Such a regime would certainly be a discouragement to trade with the outside world, something which it is desirable to expand. In addition, Williamson (1991, p. 393) gives other arguments against floating in the emerging market economies of Eastern Europe, and these also apply to Russia. First, he makes the point that a floating rate needs to be associated with a reasonably predictable monetary policy (defined by some sort of aggregate), in order to anchor expectations to some extent. But this is almost impossible when the monetary systems are so new. Second, a floating system can only operate efficiently when there is a well developed capital market. Finally, he argues that the experience of floating in the market economies has not been 20 particularly favorable, with large fluctuations in rates resulting, apparently, from speculative bubbles and sharp variations in expectations. In any case, for the reason given above, Russia is likely to find a pure floating regime an unacceptable gamble. One is thus left, inevitably, with a "fixed but adjustable" exchange rate regime for Russia. The adjustment might be frequent, in which case the exchange rate becomes a "crawling peg," or it might be occasional, in which case it becomes more like the Bretton Woods or the EMS system. The less frequontly it is adjusted, the closer it comes to the fixed rate regime. Compared with the completely (firmly) fixed rate regime, the barrier against inflation, and against speculation on the exchange rate itself, would be reduced, since an anti-inflationary commitment made by the monetary authorities would be less credible than when it was backed by an exchange rate commitment. But the main difficulty is that, while exchange controls on capital movements will no doubt be maintained, there are many ways of exporting capital, and the possibility of profitable speculation on the exchange rate--leading to foreign exchange crises followed by devaluation, and hence losses by central banks--would then arise. This is the familiar problem of "fixed but adjustable" exchange rate systems. The quicker the exchange rate is devalued when there is speculation against the rouble, the less these losses would be. But experience from many market economies, both developed and developing, shows that there is often a reluctance to devalue because of the potential .%nflationary effects of devaluation--raising the cost of living and thus reducing real wages, as well as raising the costs of domestic industries that 21 use imported inputs. One can expect all these problems .o arise in Russia eventually.9 2. RELATIONSHIP BETWEEN TRADE POLICY AND EXCHANGE RATE POLICY Before turning to exchange rate policy for the other Republics, the relevance of exchange rate policy for trade policy (narrowly defined as excluding exchange rate policy) needs to be discussed. The main point is simple and very important. When a country's foreign exchange reser-ves run down and it cannot readily borrow any more, yet it is committed to a fixed exchange rate, or, in a flexible rate regime, it makes only infrequent and reluctant adjustments of the exchange rate, a recourse to trade restrictions is almost inevitable. Such has been the experience of many market economies at many times in the past. As standard analysis teaches, if equilibrium is to be re-established, aggregate demand (absorption) has to be reduced. But, in addition, some switching of demand away from tradables towards nontradables and of domestic output in the opposite direction is usually desirable to minimize adverse employment effects domestically. If the exchange rate cannot be used as a switching instrument, or if there is a reluctance to use it, a recourse to restrictions on imports is highly likely. Such restrictions are often difficult to remove later, since interest groups grow up that benefit from them. Furthermore, their later removal may still require some devaluation of the 9 The effects of capital movements could be absorbed by a floating exchange rate applying only to capital (and, perhaps, tourist) transactions. In other words, as has been widely suggested, there could be a dual rate system. But this would still require exchange controls to keep the two markets separate--i.e., to prevent capital movements taking place at the official rate. 22 exchange rate, which is ruled out by a fixed exchange rate system. In the case of the Republics of the former Soviet Union, one would expect a ready recourse to restrictions since the idea of quantitative controls is, of course, quite familiar, and an awareness of the distorting cost of the restrictions is likely to be low. In any case, to repeat, it is quite likely that the use of trade restrictions would be determined to a great extent by the balance-of-payments situation. Such restrictions might Z±eld short-term benefits in maintaining domestic employment, but would be offset by longer-term losses from the distortions in resource allocation created, and especially from the adverse effects on exports. If a country were inflating at a rate faster than its trading partners, continuous or frequent depreciation of its currency would be needed to maintain the real exchange rate and thus competitiveness of its export and import-competing industries. But if the nominal exchange rate is fixed, this is not possible. Yet, trade restrictions could only be a short-term substitute since they could not maintain the competitiveness of export industries. Eventually a devaluation would come about. But the experience of many countries shows that, in the initial stage, trade restrictions are indeed often intensified, with adverse effects in distorting and reducing trade. The conclusion is that the case in favor of flexibility of the rouble exchange rate, involving, if necessary, quite frequent adjustments, is very strong. Such flexibility is necessary to forestall the use of import restrictions for balance-of-payments purposes, whether in the form of quotas or tariffs. A commitment of the government and the monetary authorities to non-inflationary monetary policies must be direct, rather than brought about via an exchange rate commitment. 23 3. EXCHANGE RATE POLICY FOR THE OTHER REPUBLICS All the issues just discussed also apply to the other Republics. But the smaller the economy, the stronger becomes the case for a fixed exchange rate.10 But the question rumains: fixed to which currency? To begin, let us consider a small economy, say Latvia, which is eventually likely to trade extensively with the European Community. Here the currency might be fixed to the ECU or, alternatively, the D-Mark. All the problems of a fixed rate just discussed would still arise. But since trade would be a much higher proportion of GDP than in the case of Russia, it is more plausible that domestic monetary policy be adjusted in the light of the foreign exchange situation: the foreign exchange tail would be a lot bigger relative to the monetary dog than in the case of Russia. Furthermore, the gains from exchange rate stability to the economy would be greater since these gains take the form of making foreign trade less costly, and foreign trade is a larger part of the economy than in the case of Russia."1 Hence, one might conclude that the case for Russia and Ukraine having flexible exchange rate regimes is very strong, but for the other thirteen economies--which are much smaller--the possibility of a fixed rate regime, or a regime of a fixed but infrequently adjusted rate, should be viewed more favorably. In all the latter cases there would still be the problem of determining the correct rate initially, given that prospects are so uncertain. It may be necessary to start with a fixed but 10 This is, essentially, an argument following from the theory of optimum currency areas. lt Perhaps even more important than the gains for trade are the gains for foreign investment. If the Latvian currency, for example, were fixed to the D-Mark, foreign investment from Germany and indeed, from other members of the Exchange Rate Mechanism of the EMS, would surely be encouraged. 24 adjustable rate, switching at some later stage to a firmly fixed rate. Furthermore, there would still be a danger that trade restrictions would be used at a time of foreign exchange crisis. To repeat, all the earlier doubts about fixing the exchange rate still apply, but the weight of various arguments is different than in che case of the large economies, namely, Russia and Ukraine. The possibility must now be considered that a Republic fix its exchange rate to the (Russian) rouble. The argument in favor, indeed the only argument, is that its trade with Russia may be very high relative to its trade with the outside world or other Republics that don't fix to the rouble. The reason for this is, of course, the extreme specialization that has been mentioned earlier. Here it is important to consider not only the current flow of trade with Russia--which in all cases is high because of the extreme specialization--but also the expected flow of trade in the future, during the normality period. Many Republics, notably Ukraine, the Baltic Republics and Moldava, will surely expect the pattern of their trade to change drastically in favor of trade with Western Europe. If they did wish to fix their exchange rates--in spite of various considerations noted earlier--they might be wiser to fix it to the D-Mark or ECU. Thus, the high level of trade with Russia provides a possible argument for fixing to the rolible, at least for those economies that do not expect their trade patterns to shift drastically towards the West. The crucial qualification to the case for fixing to the rouble is that there is no point in doing so unless Russia succeeds in macroeconomic stabilization and, furthermore, looks like being able to sustain this stabilization. Any Republic whose government believes that it could stabilize more successfully on its own should certainly do so. It could operate a fixed but adjustable exchange rate system, preferably adjusting frequently if required by its foreign exchange situation, and 25 fixing in the short run to the D-Mark or the ECU. At a later stage, when Russia has fully stabilized its economy, the Republic could switch from a D-Mark to a rouble peg. 4. MONETARY INTEGRATION Finally, for the non-Russian Republics, the possibility of monetary integration with Russia should be noted. This, of course, is an extreme version of a fixed exchange rate commitment. It is the present situation (January 1992). Each Republic is still using the rouble, which is the Russian currency.12 Here there are essentially two possibilities. First, a true monetary union embracing the fifteen Republics might be established, with a single central bank subject to control by all the member Republics' governments, or appointees of these governments. Seigniorage would be distributed to the member Republics in some agreed way. The central bank might be quite independent, and if its constitution required it to pursue a policy of price stability, it would be rather similar to the central bank envisaged in the recently agreed proposals for a European Monetary Union. The problem here is that such a union is bound to be dominated by Russia. Monetary policy--and hence the exchange rate relative to the outside world--would be determined by Russian considerations. It is not conceivable that Russia would accept anything else, given that its economy is so large relative to the others. But some Republics, notably Ukraine, would hardly accept such dominance. At the time of writing, the establishment of such a Union involving joint 12 In the case of Ukraine a process of monetary disintegration is underway, the details of which are not really relevant here. 26 decision making does not appear likely on political grounds, quite apart from economic implications. The second possibility is that a Republic operate a "currency board" system with a firmly fixed exchange rate to the rouble and no ability of the monetary authority to create credit. The Republic could have its own currency, but it would be fully backed by roubles. There would, of course, be no exchange controls between it and Russia, as well as other countries with similar currency boards. Any loss of foreign exchange reseives would lead to an automatic decline in the money supply. This is similar to the system that operates in the African franc zone--the CFA countries. In practice it differs little from the monetary union described above, except that in this case control over monetary policy is fully handed over to Russia. If the Republic received interest on its rouble reserves, the seigniorage from money creation would accrue to it; otherwise it would accrue to Russia. One cannot imagine a Republic agreeing to such a system unless it did receive interest. If a Republic is small; if the argument for fixing the exchange rate to the rouble is strong because trade with Russia is expected to be very high for a long time, dominating the country's trade; and if Russia is expected to succeed in macroeconomic stabilization; then the case for going all the way and monetarily integrating with Russia becomes quite strong. Perhaps this applies to the central Asian Republics. If it is intended to maintain a fixed exchange rate indefinitely, it is better to lock it in through an institutional arrangement so as to avoid any foreign exchange speculation. While accepting Russian dominance, the Republic might be represented on the central bank board so that its voice would at least be heard. 27 III TRADE OLICIES Trade policy for a single Republic will now be considered. The issue of whether the Republics should form, or maintain, a free trade area will be left aside for the moment. It will thus be assumed that all measures apply to exports to or imports from all countries with which the Republic trades. It will also be assumed that the Republic concerned has established what is primarily a market economy. There is both commodity and current account convertibility. Any trade restrictions take the form of tariffs, import quotas or export controls and taxes, and not exchange controls for current account transactions. As already discussed, export taxes or controls may be needed in particular cases where prices to domestic consumers are kept below market levels, taxes being far preferable to controls. There remain now a variety of issues concerning the use of trade policy interventions, issues which are similar to those discussed extensively in the existing literature of trade policy for market economies. 1. TRADE TAXES FOR REVENUE Tariffs and export taxes raise revenue. Provided the exchange rate is adjusted appropriately, a uniform tariff will have a similar effect to that of a uniform export tax. Both will tend to restrict trade. This is an important equivalence, though it does depend on exchange rate flexibility.13 During the adjustment period the fiscal problems of the Republics are likely 13 This is A.P. Lerner's "symmetry theorem." It is explained fully in Corden (1971, pp. 119-122). 28 to be such that any tax that is politically acceptable and can be readily collected will be desirable, provided iLt does not create excessive distortions. One has to remember, as noted earlier, the overwhelming importance of budget deficits as the sources of the monetary disequilibrium. Export taxes may well be the most convenient and politically acceptable way of raising revenue--at least as a supplement to more desirable consumption and income taxes. But tariffs to raise revenue--especially a uniform tariff--cannot be ruled out. If the tariff is uniform and at modest levels, it is unlikely to be too distortionary. A tariff structure that were "made-to-measure" to the needs of particular domestic industries or consumers (a concept to be discussed again later), and that were continually varied, would provide undesirable opportunities for rent-seeking and bureaucratic meddling. A simple uniform tariff, possibly with a few exemptions, is far preferable. Refunds ("drawbacks") to exporters using imported inputs might be provided. If a value-added tax were imposed on domestic production, with production for exports exempted (a method of taxation used in the European Community, and widely advocated for other countries), a complementary tariff is needed to turn the value-added tax system into a tax on domestic consumption. By the time of the normalcy period, any taxes that discriminate against trade should ideally have been replaced by consumption, value-added or income taxes as sources of revenue. 2. POLICY INSTRUMENTS TO AVOID Import quotas, exchange controls for current account 'ransactions, and made-to-measure tariffs should be avoided. The aim of opening a Republic's economy to trade would not only be to equilibrate demand and supply conditions in the domestic 29 market, and to obtain the familiar benefits of comparative advantage, but it would, above all, be to provide competition for local enterprises which would otherwise be able to exploit monopoly positions resulting from the extreme specialization of production. (This was the primary motivation for the ending of import controls in Russia in January 1992.) When it is impossible to generate domestic competitors in a short period, it is essential to introduce foreign competition. This objective is still compatible with moderate tariff levels, provided the tariff levels are -Zixed and predictable. But all quantitative restrictions on trade or on converting the local currency into foreign currency for trade transactions should be avoided. Exchange controls for current account transactions which discriminate between different uses of foreign currency have the same sorts of effects as quantitative restrictions on trade. The difference is more administrative than economic, except that exchange controls can also apply to the use of currency for trade in services, where quantitative restrictions on the actual trade are more difficult to apply. A tariff which is not so high as to exclude imports completely, and which is not deliberately varied to maintain the profitability of a local protected industry, will still allow competitive pressure from abroad. The price of the imported good in the domestic market (including the tariff and transport costs) sets an upper limit to the prices that can be charged for domestically produced goods that are close substitutes. The higher the tariff, the higher the price, but as long as there can be potential imports, some limit is set to the price a domestic producer can charge. By contrast, a quota shelters the domestic producer much more, even when some imports are allowed: the domestic price is insulated from the foreign price. 30 Insofar as tariffs are used for any reason, they should in general not be "made-to-measure" to meet the specific needs of particular protected industries. They should not be reduced when the industry is more successful and makes higher profits, nor raised when its profits fall. Some non-uniformity of tariffs may be unavoidable, and even desirable, but the general point of the need to avoid such adjustments should be borne in mind. It appears to be "common sense" to adjust tariffs (or other protective devices) in this way, and to vary the rate of tariffs protecting particular domestically produced products on the basis of need, prov:ding high tariffs for industries in difficulties and low or no tariffs at all for those that are doing well. But this "made-to-measure" method has a disincentive effect. It has similar effects to that of a profits tax, which--if at all--is better imposed at moderate rates directly. Furthermore, it introduces elements of arbitrariness, makes the tariff system unduly complex, and provides opportunities for rent-seeking, all features of the economic system which the former Soviet Republics are seeking to leave behind. It seems quite likely that the natural tendency will be to use tariff policy, and indeed other interventions, in this way, since it conforms with the traditional approach in the Soviet system. This approach has been well described by Litvak (1991, p. 82): Salaries, bonuses, and other decentralized funds are regulated on a discretionary basis, with the purpose of expropriating excess profits from organizations that reveal themselves to be more productive and guaranteeing 'normal' salaries and bonuses. ... The essence of this system is captured by a Russian word that has found its way into the vocabulary of all the Eastern European countries: uravnilovka, which translates as 'equalization' or 'levelization.' ... Inequalities are observed and subsequently leveled off. 31 A break with this approach is clearly required. Rules need to replace discretion, as far as possible, complexities in policy instruments should be avoided, and firms should be allowed to profit from success in import competition and exporting. 3. TARIFFS FOR BALANCE-OF-PAYMENTS REASONS It has already been observed that, if the exchange rate is fixed or is devalued only reluctantly, in the face of a balance-of-payments problem, there will be a tendency to use tariffs or import quotas (in conjunction with a decline in domestic expenditure) to equilibrate the balance-of-payments and maintain demand for domestic goods. This tendency provides a strong argument against fixing the exchange rate, or against devaluing only reluctantly. To avoid this tendency in a fixed exchange rate regime, nominal wages would need to be flexible downwards, so that a required improvement in the competitiveness of domestic industry would be brought about not by nominal depreciation, or by tariff increases, but by declines in wages. Yet, such downward wage flexibility is improbable. Nevertheless, there are some arguments in favor of fixing the exchange rate even when nominal wages are not flexible, these applying more strongly to small economies. In any case, "first-best" policies are often not followed. Hence, "second-best" issues remain. Given that trade restrictions will be used, which would be the best pattern and method of restrictions? The common instinct is to impose quantitative import restrictions and to vary them according to "essentiality." This is a non-price approach which may well appeal in the former Soviet Republics and, in the past, has appealed in many developing countries. But an alternative--which is preferable when there is a desire to move permanently away from quantity controls and towards market incentives--is to use tariffs. 32 one simple approach would be for the Republic to impose a uniform nominal tariff on imports which would normally (for revenue purposes) be at some modest level, say 10%, but which would be raised temporarily when there is a balance-of-payments problem. Imports of "essential" goods would not fall much, if at all, their demand elasticities being low, while consumers and producers would reduce imports of "non-essentials," these being the goods with the high demand elasticities. Thus the market would automatically discriminate on the basis of "essentiality" as perceived by buyers. Possibly exporters might get rebates on tariffs paid on imported inputs. For various reasons, such a flexible uniform nominal tariff would not yield a uniform effective tariff rate (tariff rate relative to value-added), and furthermore it would produce some bias in favor of import substitution against exporting. Hence, it would be a very imperfect substitute for a devaluation. But its simplicity and flexibility are attractive, and it should be considered if the use of the exchange rate is ruled out. 4. IS THERE NEED FOR TRANSITIONAL PROTECTION It may be argued that, if current account conv;ertibility is introduced and there is little or no protection, most or all of a Republic's industries would not be able to compete against imports from the outside world. It appears to follow from this argument that transitional tariffs during the adjustment period are thus needed to avoid drastic losses of jobs in some, and perhaps all, industries when the economy is opened up. Because of its low quality, much of domestic industrial output would be uncompetitive under free trade, so that massive unemployment would result from opening up the economies. Here the experience cf East Germany may be cited. 33 There is a weakness in this argument. Essentially, it appears to confuse absolute and comparative advantage. If a large part of domestic industry turns out to be uneconomic when the economy is opened up, the nominal exchange rate relative to the nominal wage level must have been overvalued. If nominal wages are taken as given, and if the exchange rate is available as an instrument of policy, there should be a devaluation, possibly a very substantial one. This would make many industries competitive again, to the point if necessary, where external equilibrium is restored. But there is a complication, which suggests that there could be some logic in the argument. During the adjustment period the industries of the former Soviet Union are likely to become gradually more efficient: hence, they would become more competitive at a given exchange rate and given wage level. In a free market the exchange rate would then tend to appreciate, or the domestic wage level to rise. The question is whether this likelihood might justify transitional tariffs. This issue of the need for transitional protection is certainly important, and it can be clarified with the help of a diagram. A number of simplifying assumptions are made, but removing them would not alter the main messages. In Figure 2 the vertical axis shows the real exchange, a movement upwards being a real devaluation. With a given nominal wage level and given constant foreign price level, this can be equated with the nominal exchange rate. Hence, the vertical axis shows the price of foreign currency in terms of domestic currency. The horizontal axis shows the value of exports and imports in foreign currency, and it will be assumed that they have to be equal. Inequality would result from capital inflow or outflow, use of foreign exchange reserves, and aid. S1 is the short-run supply curve of exports, showing that, in the short run, supply is fixed, so that the value of export income is fixed. S2 is the 34 ~eoQ E';diwts S VOttje OP EAOqsA1t, a4 i (o6Ac,l CuVthf,j 34c: 2 34a long-run supply curve: devaluation would increase exports. D1 is the deman'! curve for imports. It reflects both the domestic demand for importables (imports plus import-competing goods) and the domestic supply of import-competing production. Thinking of Russia, we can imagine the supply curve to refer primarily to energy exports and the demand curve to imports of manufactures of all kinds (an obvious simplification). Over time domestic manufacturing will become more efficient, so that at a constant exchange rate (constant nominal exchange rate and constant domestic nominal wage level, with world prices given) domestic output would increase, and so the demand for imports would fall. This is represented by a gradual movement of the demand curve to the left from D1 to D2. Such an efficiency improvement would no doubt also take place in export industries, hence shifting the supply curve to the right, a consideration that is ignored in this argument, an important and limiting assumption. Suppose convertibility is introduced, there are no tariffs at all, and the exchange is adjusted to maintain equilibrium in the balance of payments on current account (or is allowed to float). It will then, initially move to OK, with equilibrium at A. As the export supply curve becomes more elastic and the demand curve shifts to the left, the exchange rate will appreciate, until equilibrium at C is reached. Before convertibility the exchange rate might have been, say, at OH (the initial exchange rate at which exporters just covered their costs and at whieh there was excess demand for imports). Overshooting thus takes place: first, the value of the currency depreciates sharply, and then it gradually appreciates. Initially export industries will obtain a windfall gain. Since trade is determined by comparative and not absolute advantage, the fact that domestic manufacturing is very inefficient, at 35 least initially, does not alter the ability of the system to ensure balance-of-payments equilibrium and the maintenance of employment. An employment problem would arise only if the real exchange rate could not be altered from its pre-convertibility level, either because the nominal exchange rate were fixed or nominal wages increased to compensate for any devaluation. In that case the demand for imports would rise by HJ as a result of convertibility (inconvertibility having limited imports to MH). Where is the argument for transitional tariffs? This could be made in the following way. In deciding how much to produce, exporters and import-competing producers are likely to look at the exchange rate that becomes established directly after convertibility, and not at the rate that will eventually emerge. They will not practice "rational expectations." The same applies to domestic producers who use imported inputs. But the immediate post-convertibility highly devaluied exchange rate sends out a false signal. Not only would it give exporters a windfall (which could be temporarily taxed away) but it would lead them to over-expand, aiming for point F on the diagram. Similarly, as import-competing producers become more efficient, they would expand output on the basis of the initial exchange rate, moving towards point E. At the given exchange rate OK, a surplus would thus emerge. But this would lead to real appreciation that would take away some of the improvements in competitiveness for these tradable industries again. There will be a disappointment of expectations and a temporary misallocation of resources. The suggestion (or implication) is then that the exchange rate should initially be set at a more appropriate long-term, less depreciated, level. Thus, it might be set at the long-term equilibrium OL. This would encourage exporters to move directly to point C, while import-competing producers that foresaw their 36 improvements in efficiency would move to output levels that yielded a demand for imports at that level. Yet, in the transitional period import-competing producers would make losses. In the absence of a tariff, at the exchange rate OL, their output would fall, yielding demand for imports at G. It follows that an initial tariff equal to KL should be imposed, and this should be gradually reduced as the demand curve shifts downwards. If domestic producers are to plan for the correct output level, they should know firmly in advance that the tariff rate will be reduced, and will eventually be eliminated. They would then use the exchange rate OL as their guide for long-term output planning. There are some obvious problems here, and the policy proposal that has been expounded here is not necessarily one that should be supported. I have aimed here only at clarification. The proposal assumes that productivity improvements would take place primarily in import-competing industries, not in export industries. Furthermore, it is hardly possible to predict the equilibrium exchange rate (i.e., OL) for the end of the adjustment period on the basis of which current planning would take place. The most one can say is that extreme overshooting of the current exchange rate could be avoided by providing a temporary uniform tariff well above the basic (say 10%) rate appropriate for revenue reasons alone. But such a tariff does discriminate against exports, and it must of course be temporary. 5. INFANT INDUSTRY PROTECTION AND A HARD TARIFF PATH The general point of the preceding argument is that there can sometimes be a case for temporary protection by means of tariffs (quite apart from revenue tariffs). But it has to be remembered that protection of one industry is always at the expense of 37 other industries. In the example just given, import-competing industry is protected at the expense of export industries, and the assumption (not necessarily justified) was that gradual improvements in efficiency would emerge in the former and not the latter. If a particular industry seeks infant-industry protection to cover a period that would otherwise be loss-making in order to build up experience, markets, and so on, for a later profitable period when it would not need protection, it is usually suggested that two questions should be asked. First, why is it not able to obtain credit from banks or investors to cover its initial losses? Second, why are its problems and future prospects different from those of other industries? In the case of the ex-Soviet Republics, it can reasonably be assumed that the capital market, insofar as it will exist at all, will be imperfect and credit will not readily be obtainable. But the case for giving one industry special treatment relative to other industries has still to be made. Hence, the case for providing temporary infant-industry protection for specific industries appears weak, though the case for some general transitory protection, on the grounds outlined above, is stronger, at least if one expects efficiency improvements to be greater in import-competing than in export industries. But whether the temporary protection is general or specific, it is vital that a declining tariff rate path be established and committed to in advance. One might call this a hard tariff Path. A soft tariff path is one that is not clearly established in advance, or if a commitment is made, it is not firmly adhered to. In the latter case, if an industry fails to fulfil its promise, tariff making that responds to interest group 38 pressures will prevenit tariff rates declining predictably as they should. 14 IV THE FREE TRADE AREA ISSUE For a number of reasons, the Republics may be tempted to impose trade restrictions on trade with each other. Great dislocations might be caused by such restrictions. Because of the extreme specialization, the former Soviet Union is now a highly integrated area, and while this degree of integration may naturally decline when trade with the outside world is opened up, to force a further decline by means of trade restrictions would impose undue dislocations and costs. Hence, a strong case exists in favor of the Republics forming a free trade area and committing themselves to maintain it for a long period ahead. The issue was already discussed earlier. Yet, in practice, a free trade area may not be established, if only for nationalistic reasons. Furthermore, there are also some possible arguments on the other side, and some of these will be discussed here.15 14 McKinnon (1991) has noted that some industries in the ex-Soviet Union probably produce negative value added at world prices--i.e., the cost of energy inputs valued at world prices is greater than the value at world prices of gross output. (The concept of "negative value added" is expounded in Corden, 1971, pp. 51-55.) They have survived because of excessively low domestic energy prices in the Soviet Union. He suggests that, if efficiency improvements and reduced energy use would eventually turn them into (sufficiently) positive value-added producers, they should be protected. But this is just an extreme case of the broader issue of whether infant industry protection should be provided. Even industries that produce positive value-added may be uneconomic in the short run and yet economic eventually. 15 Gros and Steinherr (1991, pp. 33-38) have an extensive discussion of whether a Soviet customs union should be maintained. In the case of the non-Russian European Republics, they argue that they will eventually trade more with 39 1. POSSIBLE ARGUMENTS AGAINST A FREE TRADE AREA (a) Fixed Exchanae Rate One Republic may have committed itself to a fixed exchange rate (perhaps fixing its currency to the D-mark) for the reasons discussed earlier, and it may find that it has overvalued its exchange rate, at least in the short run. Even without a fixed rate commitment, it may have chosen to overvalue its exchange rate to avoid a temporary excessive devaluation (overshooting) of the rate. As also discussed above, it might then deal with the problem by imposing or raising a uniform tariff--which would be reduced as the country's industries become more competitive. Let us suppose that the problem does not arise for other Republics. The question then arises whether the uniform tariff should apply to all imports or only to imports from outside the former Soviet Union. On the one hand, if the competitiveness problem is really caused by an inflow of imports from outside owing to the sudden opening-up of trade, the problem might be dealt with by just imposing the tariff on these imports and preserving free trade with other Republics. On the other hand, the Republic concerned may be becoming much less competitive relative to other Republics, and, in addition, the problem of "trade deflection" (discussed below) may arise. If free trade with other Republics is maintained, imports may flood in from other Republics, whether originating there, or in transit from outside the former Soviet Union. the European Community than with Russia, so thmat it will be more in their interest to join a European rather than a former-Soviet customs union. It is implied, in this view, that the extreme specialization will be greatly modified during the adjustment period. They may well be right. But so far the Community has been notably unwelcoming to new members from its East (and their exports). 40 If the exchange rate becomes severely overvalued, the imposition of the uniform tariff on all imports, including imports from other Republics, may become difficult to avoid: the free trade area may be endangered. The moral is that establishing a firmly fixed exchange rate regime by individual Republics, or deliberate overvaluation of the rate in the short run, can produce real problems and endanger the continuance of a free trade area. Big problems for trade policy are created by a decision to forego the exchange rate instrument. (b) Revenue tariff A uniform tariff designed to raise revenue could be imposed either on imports from outside only--a free trade area being maintained--or on all the Republic's imports, including imports from other Republics. If it is imposed only on outside imports, the rate of tariff will have to be higher to yield the same amount of revenue. Taxing inter-Republic trade is clearly a convenient source of revenue, a source which is foregone if a free trade area is to be maintained. If a free trade area is maintained, trade will be less with countries outside the former Soviet Union, but restriction of trade with other Republics will be avoided. A bias will be introduced against trading with the outside world relative to "inside" trade, this being the phenomenon of "trade diversion" discussed in the theory of customs unions. Indeed, this bias against "outside" trade would exist even if the rate of tariff on imports from outside did not have to be increased when inter-Republic tariffs are foregone. Clearly an excessive bias against trade with outside brought about by a high tariff would be undesirable, because foreign trade would then be unable to increase competition in the domestic market substantially. Increasing competition will be a major benefit of such opening up, as has been noted above. A 41 supplier of a product from Republic B should face competition from imports from outside the former Soviet Union in the market of Republic A. But the starting point--the recent situation, before opening up--is an extreme bias against imports from outside, with no competitive imports. Even with tariffs of the order of 10-20% the bias would be reduced, even though some bias would remain. One might regard such a bias as acceptable, given the advantages of maintaining a free trade area for the sake of keeping going (and keeping down the costs of) the greater part of existing inter-Republic trade. Provided the levels of tariffs are not too high, the trade creation effects of keeping a free trade area while imposing revenue tariffs on imports from outside are likely to outweigh trade diversion effects. The aim should be to maintain the free trade area. (c) ariffs for Particular Industries Tariffs might be imposed to deal with the short-term employment problems of particular industries when the economies are opened up to trade with the outside world. This has not been recommended here, but may nevertheless happen. The pressure for such protection is likely to become significant once enterprise losses cease to be readily covered by subsidies from the banking system--i.e., by money creation. Proceeding then to second-best analysis, if frme trade with other Republics is maintained, such tariffs might conceivably lead to trade diversion, i.e., there might be extra imports from other Republics, replacing imlports from outside. While this effect is conceivable, it is unlikely to be an important effect owing to the high degree of specialization. Insofar as other Republics also produce the product., they may have similar short-term competitive problems. Thus, these special short-term tariffs--if they are used--are probably best imposed only on imports from outside while an inter-Republic free trade area is maintained. 42 (d) Price Controls and Export Taxes A Republic may choose to maintain some price controls and keep prices of certain goods to its domestic consumers below free market levels. This possibility was referred to above. As discussed there, to avoid these goods being exported at world prices, export controls or export taxes would be required to supplement price controls. The question then arises whether export controls or taxes should also apply to exports to other Republics. Clearly, from the point of view of preserving a free trade area and getting all its benefits, it is undesirable that exports to other republics be taxed or controlled. Furthermore, there would be no need for such export taxes or controls if similar controlled prices applied in the other Republics. But if free market prices (or much higher controlled prices) were maintained in some other Republics, a diversion of sales away from the domestic market to the other Republics would take place in the absence of controls or taxes. Thus, it would be difficult to avoid export controls or export taxes and so breaching the free trade aiea principle. Controlling prices of certain goods to domestic consumers would either have to be avoided completely, or would have to be coordinated between Republics, if a free trade area is to be preserved. This is an important issue at the time of writing (January 1992). Republics other than Russia are under pressure to follow Russia in price liberalization since export taxes or controls are difficult to impose, and also undesirable. 2. FREE TRADE AREA OR CUSTOMS UNION? So far no distinction has been made between a free trade area and a customs union. Two Republics that form a free trade area could 43 still have distinct policies influencing trade with the outside world. One Republic might impose a high tariff on goods imported from outside, and the other a lw tariff. By contrast, if they formed a customs union they would, in addition, need to have the same tariff rates (and export taxes) on external trade. From the point of view of trade policy, they would be more like a single country, with a common external tariff structure. It is also conceivable that they form a free trade area for some goods and a customs union--in effect, a partial customs union--for others. Given that free trade between the Republics is desired, the choice between free trade area and customs union involves a balance of considerations. A free trade area creates the problem of "trade deflection." If Republic A has a low or zero tariff on imports from outside, while Republic B has a high tariff, goods from outside will tend to be imported into Republic B via Republic A, so that B's high tariff would, in effect be evaded. If one believes that low tariffs or free trade are really desirable, one should welcome such evasion. Good policies would, again, have driven out bad policies. But the protectionist Republic is unlikely to accept this outcome. To avoid trade deflection, trade barriers between A and B would have to be set up to check the origins of goods: goods originating from outside would have to pay a tariff. This can get very complicated, bearing in mind that goods imported into high-tariff Republic B from low-tariff Republic A may have components originally imported from outside, but also have value-added in A. A "certificates of origin" system is usually used. Opportunities for arbitrary bureaucratic decision making then arise. The problem would be avoided if the differences in tariff levels were small, or if transport costs were high. Thus (unless tariffs are low in any case), from this point of view, a customs 44 union--where there is no incentive for trade deflection--is clearly preferable. The problem about a customs union is political: it would be necessary to get agreement between the Republics about the common tariff policy to be applied to outside trade, and this may be difficult. For example, the survival of an industry in one Republic may be greatly threatened by the opening up of external trade, even though, allowing time for adjustment, it can reasonably be expected to survive. Thus, this Republic may want to impose a temporary infant industry tariff. But other Republics, being net consumers of the product concerned, would benefit from free trade. The general conclusion is that all these problems would be avoided if the use of tariffs (and import restrictions) were minimized, and if such tariffs were generally kept low. Insofar as they are used, it would be better if a customs union were maintained and some collective arrangement among the Republics were established to determine the common tariff levels to apply to all their imports from outside. Possibly the members of the customs union might agree to establish a semi-independent Tariff Commission. This might have written into its Constitution some general rules about tariffs, for example, that any tariffs above a certain percentage should not be maintained for more than a defined number of years. 3. MONETARY INTEGRATION AND TRADE INTEGRATION An area of monetary integration need not coincide with an area of trade integration, i.e., a free trade area or customs union. Thus, two Republics might have separate moneys and exchange rate policies, and yet they might still commit 45 themselves to maintaining free trade with each other. This, indeed, has been the position in the European Community.16 It is also conceivable that two or more Republics form an area of monetary integration but not an area of free trade: i.e they maintain fixed exchange rates relative to each other, or literally have a common monetary policy, and yet still feel free to impose trade restrictions on each other. For example, for many years up to 1976, Mexico maintained a fixed exchange rate to the dollar and yet imposed restrictions on imports of US (and other) goods. And within Canada--an area of complete monetary integration--some restrictions have been imposed by the provinces on imports of goods or services from each other. But the latter is unusual. In any case, it is desirable that an area of monetary integration also be a free trade area. The reason is that the inability to make an exchange rate adjustment between two countries that form part of a single monetary union is likely to lead to excessive use of trade restrictions when an adverse shock that increases local unemployment occurs. There has to be some wage flexibility to take over the role of the exchange rate. If the Republics were still free to impose tariffs or quantitative restrictions on trade with each other while they were not able to alter exchange rates, a high level of restrictions might well result. Finally, while an area of monetary integration, should thus also be a free trade area, the case for a free trade area may also stand even if they do not form an area of monetary integration. 16 In the European Monetary System (EMS), exchange rates have frequently been realigned relative to each other eve,. though certain common rules have been followed and short-term fluctuations have been avoided. Hence, the EMS has not been an area of monetary integration. 46 v SOME GENERAL PRINCIPLES One approach to trade policy is, in effect, to have no trade policy. Currencies would be made convertible, at least for current account transactions, and all use of quotas, tariffs, export taxes and export subsidies would be foresworn. Credit would be made available for export development, as for other activities, on the basis of normal banking principles practiced in market economies. All enterprises, state-owned and private, would be free to engage in foreign trade themselves or use any intermediaries they chose, and certainly would not have to use state trading organizations. This would be true "free trade" and could w-ll be the most desirable policy, at least for the normal-.:v period. But, for the various reasons discussed, there are some reasonable arguments for taLiffs and possibly export taxes, especially in order to raise revenue; furthermore, strong pressure for trade intervention, particularly for protection from imports coming from outside the former Soviet Union, can be expected, and it is probable that the same forces and attitudes that prevent the attainment of complete free trade in market economies (and that, for example, have stood in the way of the rapid completion of the Uruguay Round of trade negotiations) will develop in the Republics of the former Soviet Union. In view of this, the following four general principles should be kept in mind. (1) As far as possible, barriers to existing trade between the Republics should not be set up. This trade should be encouraged to continue, though gradually it may change its character and be reducea for two reasons: supplies from outside may replace supplies from other Republics, and excessive and artificial specialization in production may be modified or ended 47 by more decentr".ization of production brought about by natural market forces. 16 But measures designed to deliberately reduce inter-Republic trade (possibly for nationalistic reasons) would, in the short run at least, cause severe dislocations. (2) All quantitative and control measures should be avoided so as to minimize bureaucratic decision-making of a detailed kind, rent-seeking, and the delays that go with such measures and thus raise the costs of trading. Only tariffs and export taxes should thus be used as instruments of trade policy. (3) Tariff and e*port tax structures should be very simple. It might be best if there were just one single uniform nominal tariff or export tax f-- revenue purposes. Proceeding to second-best possibilities, a uniform tariff might be varied over time, since some case for this can be made. Possibly it might not be applied tJ all imports. Perhaps it might be supplemented by a limited number of special tariffs, or there might be some exemptions (notably in the form of drawbacks for imports used as inputs in exports). If the uniform tariff rate is kept low--or, even it starts high but is gradually reduced to a low level--the distortions resulting from uniformity would not be great, and would be compensated by the administrative and other advantages of simplicity.17 16 Once transportation within the former Soviet Union is costed properly, some of this specialization may also be reduced. 17 17 A uniform nominal tariff, particularly if it does not apply to all imports, is unlikely to produce a uniform effective tariff structure (the effective tariff rate refers to protection in relation to value added). It is thus likely to yield somewhat uneven incentives for domestic production, favoring those industries that get high effective protection, possibly because they use imported inputs on which there are no tariffs, or inputs which are also exported, the prices of which are determined by export, not import, prices. 48 (4) Finally, transparency in trade policy is highly desirable. Of course, a simple uniform tariff is indeed transparent and presents no problem. But the tariff and export structure might get complicated as a result of the addition of various special tariffs or exemptions to the basic uniform tariff. An institution such as a Tariff Commission might be set up that advises on, and possibly actually determines, the tariff and export tax structure, and also analyzes its economic effects. It should be independent of the day-to-day political process. Its decision-making processes should be open. If some or all Republics agreed on forming a customs union, this Tariff Commission would have to be a joint one, with members appointed by t'Ae various Republican governments. Furthermore, it should be concerned with the interests of consumers, and not just producers. 49 APPENDIX This paper has been mainly about the adjustment period, not the transition, which is now in progress, and the success of which is still uncertain. Thus, it hardly deals with the immediate problems that go well beyond the issues discussed in this paper--such as the need for a legal framework for private activities and a market, not to speak of major redistributive effects. Indeed, it must be admitted that the paper deals with the easy issues, and hence it may seem somewhat unrealistic. But, like the other writings to which I have referred, it does try to look ahead, perhaps with an implied optimism. In this Appendix I shall use a simple diagram to highlight two problems of the transition, focussing on the question: given that there is price liberalization, what are the implications of opening up, i.e., introducing current account convertibility and removing any controls on imports? I consider the market for a single potentially import-competing product. When a curve is drawn, it should be understood that its position and slope will be quite uncertain to participants in the market, and the curves may rapidly change anyway. In Figure 3 the pre-liberalization price is P., the demand curve is DD, and the supply curve SS. Initially output is at A. There is, at this stage, neither price nor trade liberalization. If price liberalization were introduced without trade liberalization, and producers were competitive (I assume that buyers are), output would rise to B. Of course, the price would rise, with inevitable and problem-creating redistributive effects. But that might be accepted as the cost of inducing increased supplies on the market. 50 D PT I' e E Pa C/ I . . flcajr 3 50-a 50-a Now problem One. Supplies may be monopolized (as they very often are in the ex-Soviet Union). It will pay to keep supply below B so as to raise the price further. The profit-maximizing price might be at C, so that supply would actually fall. The objective of getting increased supplies would have been defeated. The monopolist, lacking experience of price-gouging, may not have judged the demand elasticity correctly, and may have raised the price so much that he settles on a point like E, where hardly anything is bought. Anecdotes reported in the press tell us that this has been happening. No doubt he will feel his way towards C. The apparent solution is to open up the economy. A price will eventually be set by the cost of imports at the prevailing exchange rate, also allowing, of course, for transport and other transaction costs. If this price is below P2 it will achieve the objective of limiting monopoly exploitation and if it is at or below P1 it will do so completely. That is thk standard argument (also given in this paper) for opening up, quite apart from the comparative advantage argument. But now we get problem Two. It is quite uncertain where the import price will be. It will depend on the pattern of world prices, on the degree of substitutability of foreign for particular domestic goods, and so on, and also on the exchange rate, which brings about a general equilibrium adjustment when many markets are being opened up. Perhaps the import price will be below P0, possibly well below. In that case, domestic output will fall, and perhaps cease. We now have the familiar unemployment problem. To cope with the monopoly problem we have created the possibility of a new short-term problem. 51 Bibliography for: Integration and Trade Policy Issues in the Ex-Soviet Union Aslund, Anders. 1991. Gorbachev's Struaale for Economic Reform. 2nd Ed. Ithaca, NY: Cornell University Press. Cooper, R.N. 1991. "Opening the Soviet Economy." In Merton J. Pack et al., What Is to Be Done? New Haven: Yale University Press. Corden, W.M. 1971. The Theory of Protection. Oxford: Clarendon Press. Gros, D. and A. Steinherr. 1991. "Economic Reform in the Soviet Union: Pas de Deux between Disintegration and Macroeconomic Stabilization." Princeton Studies in International Finance No. 71. Princeton, NJ: International Finance Section, Princeton University. Havrylyshyn, 0. and D. Tarr. 1991. Trade Liberalization and the Transition to a Market Economy. PPE Working Paper, WPS 700. Washington, D.C.: World Bank. Havrylyshyn, 0. and J. Williamson. 1991. From Soviet DisUnion to Eastern Economic Community? No. 35, Policy Analyses in International Economics. Washington, D.C.: Institute for International Economics. Litwack, J.M. 1991. "Legality and Market Reform in Soviet-Type Economies." The Journal of Economic Perspectives 5(4): 77-89. McKinnon, R.I. 1991. "Foreign Trade, Protection, and Negative Value-Added in a Liberalizing Socialist Economy." In R.I. McKinnon, The Order of Economic Liberalization. Baltimore and London: The Johns Hopkins University Press. 52 Tarr, D. 1991. "Problems in the Transition from the cMEA: Implications for Eastern Europe." Forthcoming in: Communist Economies and Economic Transformation. Williamson J. (ed.) 1991. Currency Convertibility in Eastern Europe. Washington, D.C.: Institute for International Economics. 53 Policy Research Working Paper Series Contact Tile Author Date for paper WPS891 Publie Institutions and Private Andrew Stone A4ril 1992 G. Orraca-Tetteh Transactions: The Legal and Brian Levy 37646 Regulatory Environment for Business Ricardo Paredes Transactions in Brazil and Chile WPS892 Evaluating the Asset-Based Minimum Antonio Estache April 1992 A. Estache Tax bn Corporations: An Option- Sweder van Wijnbergen 81442 Pricing Approach WPS893 The Evoling Legal Framework for Cheryl W. Gray Private Sector Activity in Slovenia Franpo D. Stiblar WPS894 Social Indicators and Productivity Gregory Ingram April 1992 J. Ponchamni Convergence in Developing Countries 31022 WPS895 How Can Debt Swaps Be Used Mohua Mukherjee April 1992 Y. Arellano for Development? 31379 WPS896 Achievement Evaluation George Psacharopoulos April 1992 L. Longo of Colombia's Escuela Nueva: Carlos Rojas 39244 Is Multigrade the Answer? Eduardo Velez WPS897 Unemployment Insurance for Daniel S. Hamermesh May 1992 S. Khan Developing Countries 33651 YWPS898 Reforming Finance in Transitional Gerard Caprio, Jr. April 1992 W. Pitayatonakarn Socialist Economies: Avolding the Ross Levine 37664 Path from Shell Money to Shell Games VVPS899 The Financing of Small Firms Christian Harm May 1992 W. Pitayatonakarn in Germany 37664 WPS900 The Relationship between German Christian Harm May 1992 W. Pitayatonakarn Banks and Large German Firms 37664 WPS901 Opening the Capital Account: James A. Hanson May 1992 D. Bouvet A Survey of Issues and Results 35285 WPS902 Public Sector "Debt Distress' Paul Beckerman May 1992 A. Blackhurst in Argentina, 1988-89 37897 WPS903 The Economic Effects of Minimum Federico Changanaqul May 1992 D. Ballantyne Import Prices (With An Application Patrick Messerlin 37947 to Uruguay WPS905 Investing in Allthe People Lawrence H. Summers May 1992 M. Fernandez 33766 WPS906 Bulgaria's Evolving Legal Framework Cheryl W. Gray May 1992 CECSE for Private Sector Development Peter lanachkov 37188 Policy Research Working Paper Series Contact Tritl Author Date for paper WPS907 Institutional Reform in Emerging Robert Pardy May 1992 Z. Seguis Securities Markets 37664 WPS908 Tax Incentives, Market Power, and Dagmar Rajagopal May 1992 C. Jones Corporate Investment: A Rational Anwar Shah 37669 Expectations Model Applied to Pakistani and Turkish Industries WPS909 Parallel Markets, the Foreign Janine Aron May 1992 V. Barthelmes Exchange Auction, and Exchange Ibrahim A. Elbadawi 39175 Rate Unification in Zambia WPS910 Policy Issues in Financial Regulation Dimitri Viltas May 1992 W. Pitayatonakarn 37666 WPS911 Does Exchange Rate Volatility Hinder Ying Qian May 1992 D. Gustafson Export Growth? Additional Evidence Panos Varangis 33714 WVPS912 Understanding the Investment Cycle Andres Solimano May 1992 E. Khine in Adjustment Programs: Evidence from 39361 Reforming Economies WPS913 The Women's Devebpment Program Maitreyi Das May 1992 L. Bennett in Rajasthan: A Case Study in Group 82772 Formation for Women's Development WPS914 Health Personnel Development In J. Patrick Vaughan May 1992 0. Nadora Sub-Saharan Africa 31091 WiPS915 Trade Policy and Exchange Rate W. Max Corden May 1992 CECTP Issues In the Former Soviet Union 37947