POLICY RESEARCH WORKING PAPER . I . W~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preserving the CFA Zone Macroeconomic Coordination 5 After the Devaluation V:i a Sbantayanan Devarajan AMichael Walton .n ge The World Bank PubficEonomics Dhion Jue1994 EiS POLicy RESEARCH WORKING PAPER 1316 Summary findings On January 12, 1994, the CFA franc - the currency of primary surplus requirement, and no borrowing to the thirteen African states of the CFA Franc Zone - was finance current spending). devalued 50 percent. The event had been expected for * Sanctions on errant states should be imposed some time, but the magnitude and one-shot nature of the through reduced access to borrowing. Central-bank and devaluation posed problems for members of the zone's at least some foreign borrowing should be conditional on two monetary unions. mecting the annually agreed-upon targets. Devarajan and Walton conclude the following, among * The central banks' ability to impose these sanctions other things, about what has happened: should be strengthened, possibly by channeling a portion - Inflation has been substantially lower than in most of foreign credit going to the zone through the central developing countrics, but the mechanisms of banks. Technical assistance may also help. macroeconomic discipline have been inadequate, * Insulation can be effected by ensuring that quasi- especially for Fscal discipline. The recent crisis has its fiscal deficits are explicitly financed by country budgets, roots in failures of fiscal discipline as much as in the reversing the recent trend to internationalize them by constraints on zestoring competitiveness because of the having the BCEAO finance part of the national banks' fixed parity. portfolio problems. * The transmission of inflation across states has not * The current size of the eluasi-fiscal deficit (and hence been a problem in the past, but could be more of one in the future earnings position) of the two central banks the future with the common nominal shock, the should be assessed early and put on the budgets of the temporary loss of the French franc as an anchor, and the various national governments, with allocation based on rising importance of supranational quasi-fiscal deficits. the original source of the problem. If necessary, * For macroeconomic coordination, it is appropriate additional measures should bc undertaken to secure a to continue relying on a mixture of rules and discretion strong capital base for the'central banks. and not on thc market, at least in the medium term. The * Exit from the zone is best discouraged by securing 20-percent rule has been inadequate in the past and the zone's credibility. It should also be clear that those should be supplemented by annual targets for fiscal that exit because of macroeconomic problems will not performance (including deficit-to-GDP ceilings, a have easier access to international sources of finance. This paper - a product of the Public Economics Division, Policy Research Department - is part of a larger effort in the department to study structural adjustment in Sub-Saharan Africa. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Carlina Jones, room N1O-063, extension 37699 (19 pages). June 1994. 7'bc Policy Research Wor6 Paper y thsenats re n " o f Res woarc D mi ton Cen the rba of ab dselopmwaiacn An obicaticof tecris ics Itoga mftthfin*p mqi4f.ev en if the prcstamoavckm tb fttilypdfisdm The papffs ay the names of Ike awors axd shoW be usdand cidaccodfgl. 7Ihc findhs, snwrtatons, and candlussons are the ardtor'own ad sbold not be amibsed to dlx VWB=k1s1 its E;sFmve Board of D;rct:trs oray of its mcnbercowd Produced by the PoLicy Research Dis seniination Ccnter Preserving the CFA Zone: Macroeconomic Coordination After the Devaluation Shantayanan Devarajan and Michael Walton* * We thank Abdoulaye Bio Tchane, Larry Hinkle, Christine Jones, Miguel Kiguel, Kouame Kouassy and Richard Westebbe for comments on an earlier draft. On January 12, 1994, tho CFA Franc-the currency of the thirten African states comprising the CFA Franc Zone-was devalued 50 percent'. While the event had been expected for some time, the magnitude and one-shot" nature of tie devaluation pose a new set of problems for toe members of the zone's two monetary unions. Unles properly managed, the change in the CPA Franc's parity could undemin the credibility of the zone's monetary and exchange rate standards, negating a major benefit of zone membership and possibly threatening tho very existence of the monetary unions'. Avoiding this outcome requires mechanisms that prevent problems in one country from endangering the monetary union-that is, mechanisms of macroeconomic coordination. While coordination is always important in a monetary union, it becomes especially relevant in toe wake of a devaluation. The CFA zone served its members well, with low iflation and relatively high growth, untl the tubulence and loss of competitiveness of the 1980s (the seeds of which were sown in the commodity booms of the 1970s). This was achieved in a world otherwise fill of capital controls and flexible exchange rates. The zone continued to confer low inflation in the tubulent 1980s, even while the loss in competitiveness and financial discipline contributed to a deepening economic, financial and human crisis. As the rest of the world moves towards open capital accounts and fixed exchange rates in the 1990s, the CPA zone's convertible currency and strong, independent Central Banlks should prove even more valuable commodities. A large, one-shot devaluation of the CFA Franc was the best means of preserving the benefits of the zone at a new but still fixed currency rate while restoring competiveness in the member countes. But secuing a large, real depreciation (many estimates suggest 50 percent in local currency terms is required) is tough enough for a single country. At a minimum it requires sound fiscal and monetary policy and measures to prevent the maxi-devluation from triggering a wage-price spiral. Is it more difficult in a currency union? After the devaluation, could the excesses of some members be exported to the rest of the union, jeopardizing the success of the whole enterprise? The answers to these questions, and the related policies to accompany a large devaluation of the CPA Franc, depend on three sets of issues: (a) how to keep members of the union on the macroeconomic straight and narrow; If and when this fails, (b) how to insulate the whole unon from the problems of errant states, so that fiscal (or other) problems in one county do not magnify problems in the central bank, and paripassu in other countries; and (c) how to do so in a manner that discourages exit from the union by states suffring macroeconomic difficulties. Not only would this be a permanent loss to the exiting state, it could damage the credbility of the zone as a whole. '100 pert in local currency terms. 2 Thmrughout this papa, we will refer to the CPA Zon as the thteen Afican counli whih use the CPA Fran as their acncy, recognizing that he countri belong to two diffrent monetry unuom, namedy, UMOA and the menber of BEAC. Whie most of the poles we advocate apply to the CFA Zone, ty ar aimed at preservg and strngteing each of the two unions. Most of the disousion and data will concrn BCEAO, UJMOA's centalb aloug a number of poin apply to BEAC. Fmally, note that Como also uscs a currency caled the CFA Frnc but sinc it has its own control bank, none of e coordinaion ues raised here apy to i 2 These are perennial questions that have always been relevant to the zone. They are particularly important in the aftermath of a devaluation, since the major price shock could spark union-wide inflation. The recent history of the CPA Zone suggests it is highly unlikely ta every country will follow exemplary macroeconomic policies after a devaluation. Some will be susceptible to added instability that could threaten the success of a devaluation. This paper reviews the history of the relationship between zone-wide controls and national macroeconomic management, assesses the risks of added instability after a devaluation and outlines solutions to reduce these risks. We draw on parallels with the debate on the European Monetary Union in our conclusions on coordination in the CPA zone.? Current Mechanisms for Enforcing Macroeconomic Disciplne The monetary unions in the CFA zone have used two methods to enforce macroeconomic discipline: (i) annu, country-specific credit allocations issued by the two union-wide central banks; (ii) the rule that central bank credit to any government cannot exceed 20 percent of that country's previous year's fiscal receipts. Thus, the zone uses a mix of discretionary and rule-based mechanisms. The credit allocation is the province of every central bank, and in the case of the CFA zne, is practiced dtough the bank's refnancing facility. The "20 percent rule is necessitated by the fixed parity of the CFA Franc. Without a rule like this, governments could finance their deficits by monetization at will. The ensuing inflation would dran the zone's foreign exdhange reserves, maing the fixed excange rate unsustinable. Note that the 20 percent rule links central bank credit to a government's revenue performance, thereby reducing the possibility of deficits in one country spilling over into another. How effecive have these methods been in practice? At one level, they have been remarkably successful at maintaining macroeconomic discipline in the zone. The fixed parity was preserved for over 45 years. The membership of the two monetary unions has changed verylitle. Maurtaniahas left and Mai has joined UMOA; Equatorial Gi,nea has joined ihe BEAC Zone. ilation in the CFA countries has been considerably lower than in their Afifcan neighbors! Looking more closely, however, we find that the discretionay mechanism has not always been applied to promote macroeconomic stability. Meanwhile, the 20 percent rule has proved an increasingly ineffective mechanism for overall fiscal discipline. Not only has the rule been openly broken by some countries, but several governments have found indect means of getting around the ceiling. That the zone found itself in a crisis in the 1990's is only partially attributable to adverse 3We ae concrned here with one set ofreons for suppoting cser inter-country relations: the potential fr th tw monetary unions to support macroeconomic stbiity and credibility. This is distinct frm other motvations for regionalism, including the case foram optimal currency urem and a customs union. The CFA countris do not have ny offthe characteristics of an opfimal currey area: they fac differnt tading strucus, and so hoe differential terms of td shocks; they trade much less with each other than th icrst of the world; and factor mobiliy amongst the countries is quite imperfect Reducing intrnal bariers will help economic efficiency, but the major i from trade will be from reduced barers to trade with the principal exernal trading partnes. Internal hbecalization should be designed to spur, and not substitut for, extral liberaizaon. Monetary igation makes sense in the zone counties icndecntly of inter-county liberaEIzatio, but can also form one part of an overU program of strengthened intgrin. 4 Sec Daraja and de Melo (1991), Boughton (1991). 3 external shocks. It is also testimony to the fact that the disciplinary mechanisms have been inadequate, especially in times of boom and bust. The central problem is the weakness of fiscal controls. The 20 percent rule restricts government borrowing from the central bank only; there are no limits on either foreign or other domestic borrowing by governments.' Furthermore, it applies only to di=et lending from the central bank to the governments. The limit does not apply tu govermnent guaranteed borrowing, nor to government loans which commercial banks rediscount with the central bank. CFA governments have made use of these unrestricted instruments to finance public expenditares. As Tables 1 and 2 show, in some years, net foreign borrowing exceeded net domestic borrowing (central bank plus other domestic borrowing) by a actor of two or three. More recently, the governments of Cote d'Ivoire, Senegal, Cameroon, Gabon and Congo have run wears with the domestic banking sector. Finally, as their revenues have fallen, many goverments (Cameroon, Congo, Cote d'Ivoire, Togo) have not been in a position to repay outstanding debt, resulting in a violation of the rule. In addition, some countries (Benin, Niger) have been granted exceptions to the 20 percent rule due to uparticularly difficult" circumstances. Foreign borrowing is limited by a country's creditworthiness or, in the case of low-income countries, by concessional aid flows. The debt crisis has taught us that both of these are fickle sources of discipline. In the CFA zone there appear to have been greater opporunities for larger countries to borrow abroad (often due to "loan-pushing' by commercial banks or the enthusiasm of export credit agencies) until the problems of the 1980s revealed how precarious the underlying position of these countries were. Note from Tables 1 and 2 that Cote d'Ivoire and Senegal borrowed from abroad the most, both in absolute terms and as a percentage of government expenditures. To the extent that runaway fiscal deficits in the large countries are more likely to have zone-wide consequences, this loophole in the 20 percent rule is particularly worrisome. Domestic borrowing (other than from the central banks) does not directly threaten price stability, but it has a perverse effect on the central bank's own liability position and income, and hence on the zone's coherence. The banking systems in COte d'Ivoire and Senegal are in a crisis. This is partly due to the decliningprofits of theprivate sector, squeezed between an overvalued exchange rate and domestic recession. It is also a consequence of the arrears which the govermments of these two countries have been running with the financial systems and private sectors. The governments have aen steps to restructure the banking systems by absorbing some of the bad debts, but the net effect has been to worsen the income position of the BCEAO. For example, in Cote d'Ivoire, the banks received govemment paper in exchange for their bad debts-which they could then rediscount with the BCEAO at an interest rate of only 3 percent. In effect the BCEAO has absorbed some of the costs of bad debts that have their roots in part in the fiscal deficit of a member country.' But the BCEAO liability is shared by all the member states. Thus, COte d'Ivoire's and Senegal's financing part of their fiscal deficits by arrears to the private sector has led to the export of these countries' deficit to the union as a whole. Note that these credits fiom the BCEAO to the governments do not come under the 20 percent rule, again demonstrating the rule's porosity. 5One example of the kate loophole is scasonalcaop crdis-whicb in some countries became a defacio form of public-setor finanfc as public-soctor marketng agencies accumulated substantial dcbL 'There ae, of course, othr sources of the banking sectors' potilio problems, including the low pmfitbiLy of an enterpris scctor hit by recession and dcelining compeliveness. 4 Having established that theo nule part of CFA discipline leaves much tD be desired, we consider now the "discretion" part, namey, credit allocation by tho centra bank. Credit is allocated to countries annually based on their estimated credit demanid and the central bank's reserve position. The credit demand, in turn, is built up from inflation and real-output targets. In this way, credit allocation in the CFA zone closely resembles the monetary pro ng exercises conducted by the International Monetary Fund. However, monetary programming exercises are undertaken for individual counes whereas each central ban's credit allocation is to members of a currenc union. If capital wert perfectly mobile across countries in the union, it would not be possible to sustain different credit and inflation targets (i.e., differt real interest rates) across the countries. Since capital is not perfectly mobile in the CFA zone, though, it is possible to view credit allocation as country-specific monetary progrmmng exerciseo. The recent reforms in the operation of the money market In UMOA permit the central bank to manage credit allocation on a week-by-week basis (using the interest rate as an instument). In practice, however, this market has not been widely used in the zone and some "coarse bming" of credit allocations is still necessary. The CFA zone's track record of low, long-term inflation would suggest that credit management bas been consistent with non-inflationary money growth over the long n. But the central banks bave not been able to prevent temporary bursts of inflation, especially those driven by expansionary fiscal policy following a commodity price boom. It is also noteworthy tha over 90 percent of the low-interest agricultural export credit has gone to C&te d'Ivoire and Senegal, further emphasizing the bias in subsidizing resources to the luger countries (Table 3). A similar problem has occurred with loans to fledgling public enterpises in UMOA countries. To the extent these have contributed to the portfolio problems of the banking systems of these countries, the burden has been intrnationalized by BCEAO, with the "tax" spread over the whole union. The Tran ion of Economic Problems Across States We now trn to the potential implications of the large devaluation. Will such a nominal shock strain the mechanisms of macroeconomic management and coordination? In low-inflation countries, large nominal devaluations have often led to large real devaluations with only a moderate and temporary acceleration of inflation.7 The absence of an inflationary history means an increase in the price level after a devaluation is less likely to get transmitted into an indexation- or expectaions-driven spiral. However, this is by no means guranteed: it depends on the underlying economic conditions supporting the return to low inflation. Monetary policy that does not acco such a permanent acceleration is a necessary condition. But tO recite the need for tight money is of little help. Underlying this are two more findamental factors: the fiscal position and wage management. Usually both fiscal contraction and real-wage reductions are necessary in the short rua to support a swift reun to low inflation with new relative prices. Low-inflation countries which neglect these precepts (Mexico in the early 1970s is a good example) find their devaluation triggering ajump first to moderate and then to high inflatio The CFA zone has thirteen countries. Even with the benefit of a history of low inflation, the effective management of a devaluation represents a macroeconomic and political chalenge to each of them. The probability of some going off course on either fiscal management or real wage policy must surely be high. This raises the two questions that are at the center of this note: Can the institutions of the two unions Sce Ktud (1992). be used to help reduce the probability of individual countries' going off course? And, if some do, can tho whole union be Insulated from such failures in some countries? In prindple the issues apply equally to each country in the zone. But failures in the core' countries in each union-COW d'lvoire and Senegal in UMOA and Cmneroon and Gabon in BEAC-would be parcularly costly, at a minimum in terms of the propordon of th population affected by filure, and potenially In terms of greater spillover effects. Furthernore, failure in the core countries would me that the whole exercise might have to be repeated. Fily, as we have seen, the larger countries in UMOA have experienced laer discipline and have contributed more than their share to the union's problems. We therefore now examine potential trasmision mechanisms of infladon between states, starting with a review of what past experience in the zone tells us about their importance. The next section proposes some solutions. As in individual countries, for inflationary acceleration to be sustained, it would have to be underwritten by the monetary policies of the central banks. Each of the central banks draws up a financial program that essentially accomm s nominal output growth in individual countries, and allocates money and credit growth accordingly. It is probably realistic to assume that the central bank lacks the power to force a severe credit crunch on individual countries: it can lean against the wind of ongoing inflation, but not go totally against the state of markets and wage and price movements. There are then two mechanism by which inflationary acceleration in one country can be transmitted to otiers, and so to the whole union: through wage and price movements in one country directly affecting changes in another (dtat are then largely accomdatd by money and credit policy) or through the transmish an of underlying problems, such as the ineationalization of fiscal or quasi-fiscal deficits. We look at each in turn. Diet trawmission of wage wadprce movements Are wage and price movements quickly equalized aross the member countries of the CFA? Completely free movement of goods and factors of production would assure this. In the absence of perfect factor mobility, the prices of tradeables would still be expected to converge (except for differences in each county's basket of tradeable goods and differential movements in the intemational prices of these goods) but those of noradeables would respond to domestic market conditions, so that overall inflation may not be synchronized. The historical evidence bears out this reasoning. There is indeed a common pattern in price movements: inflation has been in the range of zero to eight percent for most countries for most of the time in the past three decades, with a tendency to accelerate in the 1970s and decelerate in the 1980s. Yet, significant differentials remain, between France and the members of the two zones and among the zone members. Past work suggests that the best way to characterize inflation in the CFA zone is the following: * an imperfect overall association of inflation across countries, either between France and the CFA zone members or amongst zone members; in technicda terms. both Honohan (1990) and Booaa and Devarajan (1993) rejcted co-integation (dtat is evidene of inflaion following a common path to a signifcant extent) between Franc and CFA counrie. The la study aso rejoctod co-ingration bdwoen the two core countries and each odter and the smaller countres, wth the single exception of Cote dIvoire and Burina, Paso. 6 * a common core of Inflation between the CFA zone members and France, that can be Interpreted as a common tendency of zone members to converge to the French, or international inflation rate; * divergences from this core, sometimes substantial, explained by country-specific factors. Inflation patterns In 1976-a particularly turbulent year-illustrate. In UMOA, Cfoe d'lvolre experienced an inflationary burst from S to 10 percent per annum, when French Inflation was 4 percent and falling. This can largely be explained by the investmentboom that followed the coffee and cocoa price gains in COte d'lvoire. However, with the possible exception of Burkina Faso, thre is littlo evidence of this being transmitted: inflation was flat in Senegal and falling in Togo. Sinilarly in BEAC in 1976, Cameroon experienced a major deceleration of inflation whie Congolese inflation was rising. Analysis of the differential between inflation in zone countries and France also confirms that correlations of this differential are weak, variable and sometimes negative. In other words, an inflationary acceleration in one country-taking it above core inflation-has not, in the past, tended to spill over into inflation in another. This pattern is further confirmed by the behavior of wages. While labor institutions and wage policy in CFA countries are interrelated,' and there is a tradition of labor movement for both skilled and unskilled work, there are substantial wage differentials across states (Figure 1). Futhernore, there is litde evidence of wages moving together (Fable 4). These observations must bc qualified by the fact that our knowledge of this area is still quite limited. The need for further research is especially acute. Nevertheless, the evidence so far appears to be reassuring. There has been little direct transmission of price or wage inflation across frontiers. Divergences from the core rate are primarily due to local shocks, and a fuller analysis would show that macroeconomic conditions and the tightness of product and labor markets in individual countries were the principal factors in explaining such differences. Fiscal and quasi-fiscal deficits It is clear that fiscal and public-sector wage management is central to moderating inflation in individual countries. Fiscal policy is also undoubtedly important in explaining divergences of country inflation rates from the core inflation rate of France: increased spending puts pressure on the markets for domestic goods, and increased wages directly raises inflation. Does this raise a federalist issue? Can fiscal laxity in one country lead to the transmission of inflation? At one level, this comes back to the same issue as before: if fiscal difficulties lead to an acceleration of inflation in one country, there is some risk that others will follow. As just discussed, this appears not to have been irnportant in the past. Where the federalist issue enters is through the actual and potential quasi-fiscal deficits of the two central banks. A quasi-fiscal deficit is equivalent to the losses of the central bank. In the 1980s, economists came to realize that it was necessary to think of such losses as part of the government's deficit. Characteristic sources of such deficits were foreign exchange losses (in Yugoslavia) and banking losses that were taken over by the central bank Cm Chile). They are called 'quasi-fiscal' deficits since they are a public Iln C&t d'lvoirc, some salria-notably thosc of tcachs-ac linked to levels in France- 7 soctr liability that has, In the end, to be financed, just as a fiscal deficit. Thoy are financed either by monetary creation or credit creation, boti of which have inflationary consequences. The issue is gormae to the Franc Zone boeauso the central bank are supra-national while goverment budget are national10. In UMOA and BEAC ther are two potential sources of quasi-fiscal deficits, that could be of Importance In the wae of a devaluatLon; * foreign exchage losses; e banking system losses that are financed, in part, by the central bank. Foreign exchange losses have, of course, been negligible in the past because, outside the operations account (which is In French Francs), the central banks do not hold foreign currency11. They could potentially be large after a devaluation. The position is different for the two central banks, largely as a consequence of differing conventions on who bears the foreign exchange risk of IMP loans. In UMOA, the central bank is in principle responsible, and consequently faced significant losses immediately after a devaluation. Tn practice, it appears that BCEAO was able to get the member governments to support the losses. In BEAC the risk is passed on to governments, who would automatically bear the losses. The appropriate principle in this area is that the govenment bear the risk of the IMF purchases it made and such losses have to be budgeted for. That is, governments will need to have a larger primary surplus (or smaller deficit) to finance the larger liability as a consequence of these losses. The approach in BEAC has the advantage of making country-upecific obligations both transparent and automatic. Banking system losses, which have been steadily rising in the zone during the 1980's, form the second potential source of quasi-fiscal deficits. The losses flow from three factors: a worsening enteprise portfolio in the wake of the recession and the declining competitiveness of some firms; the rise of arrears-directly to the banks or to enterprises, especially from govenments (notably in COte d'Ivoire); and the weakening deposit base as capital flowed out in the expectation of a devaluation. As noted above, there has already been a tendency to deal with these losses by effectively passing them on in part to the centrl bank. In Cote d'Ivoire and Senegal, the governments absorbed the banks' bad debts, giving them goverment paper in reurn. In COte d'lvoire, the banks refinanced this government paper with BCEAO, at an interest rate of 3 percent. The net result is that BCEAO's income position could turn negative. The situation calls for a careful assessment of the central banks' capital position with a view towards solving their income problem, possibly by recapitalization. HOw wil these be affected by the devaluaion? Direct transmission of infationary accelerations across states has been weak in the past, but there is the contmiing and possibly worsening problem of supranational quasi-fiscal deficits. Both could be affected by the devaluation. 1MThc issues discussed in tIis sub-action arc procisely the mame as those being considcred in the contet of a fuU economic union of the UMOA and BDAC Zone stats. While hannonirtion of fiscal policies wiUl be necessary for the economic union, the point madc here is that it is essential even before that-to ensure the succcas of a devaluation. " AR CPA members are a tquired to have at kast 65 percent of their foreign assets in Frcnch Franca in te Opeations Account with the French Treasury. Most hold more than this, although thcre is some evidence that Cameroon put a large pa of its oil surplus in US assets. S In the aftermath of the devaluation, all tho countries are experiencing the same inital pricwt shock, and a temporary but large divergence from French inflation. The anchor of French inflation is temporarily gone-for both t. monetary authorities and the public. Everyone expects short-run inflation to be substantially above tho past. There may be a propensity to look at tih Initial change In tho exchange rate as a yardstick to adjust domestic prices. There may also be a tendency for the smaller countries to look to the core CFA countries in each Union for leadership In setting 'ages and prices. Under these condidon It Is imperative to flnd an alternatve anchor for the movement of nontraded goods prices. Tho best candidato Is changes In tivil service wages. Of course, beyond the short (or very short) run, it is important that the CPA countries move back to a situation where the French Franc Is the nominal anchor in the system. The impact on quasi-fiscal deficits would be mixed. On the foreign exchange side, it depends largely on who holds the liability for public loans (notably IMF purchases). As to financial setoDr losses (that will potentially be transmitted to the central banks) the situation is unclear. Some commercia banks have significant foreign exposure, but its value is unclear. With Ivorian external commercial debt trading at less than 10 percent of face value in the secondary market, it appears that market valuations of this exposure have already adjusted downwards substantially. Indeed, some recent studies indicate that for commercial enterprises in UMOA, foreign exchange assets exceed liabilities. Enterprise profitability will change depending on the activities of firms-producers if tradeables will do much better, importers and producers of noutradeables will suffer: It will also depend on the overall state of the economy. Finally, liquidity will depend crucially on whether capital retr. This strengthens the case noted above for exercising caution in any takeover of fincial sector problems by the central bank. solutions (I) To the macroeconomic coordintion problem The ouestion of how to achieve maonom ic coordination in a federal system is not unique to the CFA zone. There are parallels to the debate on mainning discipline in the fifty U.S. states as well as among the potential members of the European Monetary Union. In general, dtere are three possible mechanisms for imposing discipline: market-based - in which the market imposes discipline through increasing the cost of borrowing for profligate states; * rule-based - such as a balanced-budget requirement, or a ceiling on the deficit-tDGDP ratio; * discretionary - where macroeconomic perormance is monitored by an independent or supranational body. The second and third depend on the adminiistrative imposition of some sanction, such as a fine or a quantitative limit on borrowing. The U.S. method is largely based on the market, supplemented in some states by elf-imposed rules. The evidence indicates that the fiscal position does influence the cost of borrowing, so the market imposes some discipline (Box 1). The 'Maastricht' approach, as proposed for a fiture European Monetary Union, involves a blend of all three. Rules are set, but a supranational body has some discretion in agreeing on transition steps and deciding when sanctions are imposed. Discipline is expected to be imposed both by reduced access to public fimds (e.g. from the European Investment Bank) and from the anticipated consequences on borrowing costs for a state which is out of compliance (Box 2). p zis '1 I I . .~~~~~~~~~~~ mfleUbe fir?lints-kg .um.q,San o4f Myf&m ,;, sob of.! wWo nduot Mu luR poyHw b~ .W S) ~ldpi1ne mnoq lb. nmrber utalun4,uhmad epoLolsly olnoo thr. ma-n.*dl lamlympoue borwMJni.' en i . * SM.?; Par ons thing; its gomurnns cmnnorborow born do ufol banc.. For SOU a US. ' * mnmkmh«ibae bmitkuusnv,S the d&L o.4m 3utle bandi d. wt thn *eltm eCwhbo ae'aau ildm5.1; l.' .dufig wie qfbrowig pmvdi th hiosnel to gon't bapoodbla Sisl behavior. . tots dup. 'ON.. :Ai: r[6 Gii a YkW klihavion! T" 1i s.." w camwf thb isrrketbaod fluoil dhulpine wormn to ha" winod quhe iml level UtS uni~have valiantarl 1' hiioiudthicr w n-ii an thedr deihak. TheMoo'a.Rhd qbr i top A. i*dm.:' .: 31 Doosnber18'nlsuUaio fr pelroapkaeb.s th datme: ;- .. ...;, r;:j,>g. j: ,,i;'.r;l|1 ; I; ' ..1 ' J I *it : " ; *g;rgp 1 ; MoahOaolma ~* I *r! ' i '' , YI * |..,>1i,-< l * X" V9i | | ! , j ; | r AX 2o,I. 'I &, ,;rlrla .oT . .~ RSm '51 . ' ' ' : i:A Dnm.'Z'bin. . . Zailna, .; 1ij*2*:i:,j ,j POj f : . ,mm 11.: : .: ,:. . , I ;puaRJo r ie .. . . tR'a il . .... . .......... .'-.;. ;j.I -l' :j' L8uuhlaa ;,m * .; . ,a l = , ; , .- ,;,., -~ ~~~ : : :1 . . . ..i;.. ;;t *s axemmwwcv c (idmutcm maDd| twogiom [19911) ~oaDln th aoUn tht ta with tItu isa pobm enjoy l~owr borwoing cor: iem whiobh were ooe Macdinr dewain dghtr t ,ha nibsmrpl nmduniiltdmughlflS,2Or. . bidc pokaain. an o Ihr.ir generall obligation bondJ. Finally, ultcug which met oninhdoizil l;limb om'tUh& debtp.kI anbbt I hal ; Sipoint lhemthMiLESthathidid not.. : 0. *; . . *' ' S N1. For the CPA zone, a markcet-based system is essentially irrelevant in the near future. Capital markets in the zone are thiin and, withi a history of arrears;-financing, govenments are both unusually interest-rate insensitive and unlikely to face a hard internal credit constraint. The toCFA monetary unions now operate a blend of a rule-based and discretionary syste, as disscussed abovre, but this has had major loopholes in the past. The underlying approach however should be maintained. Apt!rt from the market element, the Maastricht model is therefiore directly applicable to te CPA zone. Improving discipline should inolve two areas of change: in the rules and in thie sanctions. The rules need to be modified in two way. linst, there is a need to broaden the coverage from centalbank nnce i.e., the 20 perce2t rule) to all sources of borrowing. While the Yamassoulro reforms of 1989-90 did his in large part for UMOA (y controlling public enteprise and crop credit), thie problems with arrears inicate that thiere is still room for furher tigening. It is recommended that, while the 20 percent rule be ket (see below) it should be supplemetd by a rule based on ovrerall fiscal perfonnance. as in the Maricht proposal. A rule has t enormousa ad n pren nd no egotiable. espite all e problems of differing sarting points and special conditions, it is recommended that this be the core of the system While the details ofthe rule will have to awaitDirther analysis of the fiscal patterns in the CPA zone, tine approah 10 fAlJ: PrmnluglIhu ln e. Moa4 r Vlor?' ..~ ~~~~~: ,I .~ , - i+ ;. * .e l 1991 iTeay ofaatoht provid:I1ho iaalte tatement to diti of bow iol daolplioo will bblmuauiu Ji theo forhoodming European Moncw Unioi (EMU). The tieaty .iAUmist b rof and diiou a mncUdi ofd:o presring fmnal hrnnony. Itlprciboua "my aof .muhl tla iurvidUane blied an th.aMbLewitha ub. iS.rVeilbJl pfttood ddlNtoneyFund. ;Spoeifla*.d .M Ttreypornul Uifoi4ij- UN . j u:LNjr ',..i A -, ; i j... j:, j i . ;. prooodurea .. I~~~~~~~~~~NIIX .. .. < ~ ww < il[[!-l3 {. ; - J * n - i*I t 4|4 *>j * :.(a) NSo f;lnaaiam* ofpnbli6:geotor ib bSI;hZ toi Buropcus Ihii ( l diorou .or coapuluoq p scof tommenmcWt acnurd.i by tbe Ceiiil aBiiik; - t s .nopamure:r monta eq -nion. (b) Avoidance of ca~uioba~elw hnaaceaive" is dcflned s: (i) aM~oIIcWDP! ri.reA'er" and: ii) a public dcbd bDP ml; thn 60 pO. :; . j t! >: :5 '. ti' W ::.Nhen a momber:mt Iiilaicm oneICI ofC tr he crije;rIa, de goveiiiijcdnca aOXCIOtdIe EurOPeanMoI,b Uni;on Eaddrcasea its reoontgieadaiiogim(baiNdoiiI repa lo in Moneliny Commiueowhieh eoknse. ally spoal cireumtanocs ..ing thie -a-nry n uuion) to tle slate injaiClienoco. 1f t istatc fl n dmltie onm ji Cm ml ;can make b reaomnmcndajiona publio :11 the Kiw still doom not mpond,, ou,mlznam r iape;oilo rcconiiimniani,dati. incldingi timeitble for. dricdi-ec6etioa, and' a deaild irport on'ajw bnmmt asuies to be undertliin. "If tebe recommndtaons arc not respected, the Council will . ' : , re..' .e"t.he.i to aom' 'ta crediors out it cure nt sturonL wh r k is tn ;s; boda .,',.nwA . - . n#lthe,,European Invreote Bto i gb iadnjt ,: the -;,,untry, .................................. ,,, , ::, g ui the counb to i u of monywt te'Hms Conimuniy';'',.................................'. r'.... . unpoa sc ar gncdnm. ,. ;. E . .hr. tveath concudes by owgtdembr .um o ied wi actois.cannot juif t xcesiv c procedural; juridicO l or nlzliutvkiak gmimdri Thin. whiicdiartdions ailloiwed at lte early ages oft idefives problem Ihc rrcazy ezphcidy mnarcises.he reianevi'on 'rueN frJr tiiitug will mimalci iit anrnbubsarla.i * . descnrbed below provides a starting point. The rule would be that each state should satisfy all three of the following: * a primary fiscal surplus; * positive savings, to ensure there is no borrowing for current spending; * an overall deficit target, such as Maastricht's 3 percent of GDP, to ensure an investment boom does not become a source of indiscipline. In each of these, the definition of the fiscal deficit should be the broadest possible, i.e., inclusive of those public enterprises whose debts are implicitly or explicitly guaranteed by the sktae. Furthermore, for countries receiving concessional loans which are earmarked for public investment, the definition of public expenditure should be exclusive of this investment (although the investment should be scruinized independently). The third target will require projection of nominal GDP. These projections should be consistent with those used by the central banks in their monetary progming exercises. Since few countries have projected inflation above out turrn, the risks of this becoming a source of indiscipline are low. Finally, as in Maastricht, there may be a case for allowing a transition period to the agreed ceilings. Second, it is recommended that the 20 perceat rule be kept, in view of the macroeconomic significance of central bank borrowing92. In addition to the 20 perceut rule, mechanisms need to be introduced to ensure compliance with the other mles. In particular, foreign borrowing by profligate governments needs to be curtailed. IThe international donor community can play a role in instilling discipline in this arena. For instance, there could develop regional sources of import finance which are jointly administered by the central bank and an international agency. It would also be desirable to provide disincentives to arars financing, something which cumently has zero financial cost to the governments (at least for domestic arrears). Reduced eligibility to central bank financing in line with any arrears would be a way of providing this discipline. Needless to say, designing new rules for the zone is easier than enforcing them, especially given the track record of existing rules. Some form of sanctions for non-compliance, as in the Maastricht treay, will be necessary, although the details will have to await firther study. In addition, in an analogous vein to Maastricht, access to credit from international financial institutions could be made conditional on the rules being obeyed. To carry out this role institutional strengthening of the two central banks should be provided, with technical assistance and training playing an important role. (i) To the insulion problem It was concluded above that the principal threat of inflation transmission within the zone lay in the internationalization of fiscal and financial problems, though with the acceleration of short-run inflation there is a greater risk of direct transmission of inflation. The main solution to this problem is to prevent the occurrence of quasi-fiscal deficits of the central banks caused by country difficulties. This implies reversing the recent practice of having portfolio problems of country fmancial systems being financed, in part, by the central banks (as in BCEAO foir Senegal and COte d'Ivoire). The central banks should remain strong, independent and income-earning institutions-this might require recapitalization-and the problems of the financial sectors in individual countries should be dealt with by budgetary action within each country under schemes supervised by the central banks. Any other sources of quasi-fiscal deficits, such as foreign exchange losses, should similarly be dealt with by budgetary allocations. 12The rue may tn out to be ovedy rsiie n the fit year because of the post-devaluation inaion surg Since tbe rui constrains borowing to 20 percent of the previous years rcceipts, it may prove to be difficult to maiantin the sae levd of rcal govcrment spending in the first year after the devaluation. Howevcr, it- should be noted that most estimtue indiate that government revenues will rise Cm real terms) aftr the devaluation, so that mantaining the same level of govenmeatborrowing from the central bank may not be necessay. In any event, relaxing the 20 percent rule even for one year will require chann the statutes of the cental banks, something which is clcarly undsirable, especially since it may not be necesay. The best strategy is for the monetary authorities to be aware of the possibly rsftictive nature of this rule in e first year, and tke appropiate deciions on a case-by-case basis. 12 There is also a need for a rule on the financing of existing and new quasi-fiscal deficits. The proposed principle is straightforward: budget finance should be provided to take over such deficits in accordance with their origial source. Thus foreign exchange liabilities due to foreign borrowing would be allocated according to the source (i.e. the debtor country), and losses due to portfolio problems in the financial sector in accordance with their country of origin. Where losses exceed what can be allocated in this manner, there are two possibilities: in accordance with GDP, or using proportions obtained from identifiable factors. While financing of quasi-fiscal deficits is imporant, it is equally important to avoid these kinds of losses in the future. To toe extent that tho losses stem from problems in the financial sector-themselves a result of the economic recession of the 1980's-the growth and improved competitiveness from the devaluation should lower the chances of recurrece. To the extent they arise from fiscal indiscipline among member governments, the approach suggested here-of relying on rules rater than discretion, and of plugging as many loopholes as possible-is a step in the right direction. (iii) To terisks of exit The proposed reforms involve a tightening of discipline in the CPA zone. There is a risk that, for countries unable to comply with the requirements, exit will be a seductive option, permitting them to shift to a laxer fiscal regime. There are two ways in which the incentives can be devised to discourage exit. First, and in the long run most important, are the gains from price stability and full convertibility conferred by the zone. But this will have to be earned again in the aftermath of the devaluation. Second, and of particular importance for the transitional period, international resources should be lnked to good performance in the zone, so that countries which suffer from indiscipline cannot do better outside. While they may be able to achieve temporary gains by resort to an inflation tax, they will quickly end up with both high inflation and slow growth. In other words, countries that feel constained by their membership in the union should be made to realize that they will face much the same constraints if they left the union. There will also oontinue to be strog political pressures to maintain the two unions, but this should come from the member- governments. Finally, it should be reterated that the strongest incnve not to exit will come from the success of the devaluation Conclusion Since it is highly desirable for the membership of the two monetary unons to be preserved, it was appropriate for the counies to undergo, jointly, a sharp devaluation. Gains from interal trade are small but membership of the unions brings one powerful attribute that is unrelated to economic structure or mobility of goods and factors: macroeconomic discipline. This discipline is intimately bound up with the credibility of the CFA zone. However irrational it may seem to have held on to an overvalued currency in the 1980s, the discipline provided by the institutions of the zone remains a valuable asset. Breaking up the zone makes it less credible. So all have an interest in staying in the union, and in maintaining the new link to the French Franc. This paper has looked at federalist aspects of macroeconomic management after a devaluation; how to maintain macroeconomic discipline in a monetary union; how to prevent mismanagemen in one state spilling over to another; and how to use the istitutions of the zone to make the devaluation a success- 13 Our conclusions can be summarized as follows: * Imlation performance has been substantially better than that of most developing countries in the past, but the mechanisms of macroeconomic discipline have been inadequate, especially with respect to fiscal discipline. The recent crisis has its roots in failures in fiscal discipline as well as the constaints to restoring competitiveness because of the fixed parity. * Transmission of inflation across states has not been a problem in the past, but it could be more of one in the fiture with the common nominal shock, the temporary loss of the French Franc as an anchor, and the rising importance of supranaonal quasi-fiscal deficits. * It is appropriate to condnue to rely on a mixture of rules and discretion, and not on the market, at least for the medium term. The 20 percent rule has proved to be inadequate in the past, and should be supplemented by annual targets on overall fiscal performance (mcluding deficit to GDP ceilings, a primary surplus and no borrowing to fince current spending). - Sanctions on sates should be imposed through reduced access to borrowing. Central bank and at least some foreign borrowing should be conditional on meeting the annually agreed targets. - There is a need to strengthen the cental banks' hands in imposing these sanctions, possibly through channeling a portion of foreign credit going to the zone through these institutions, and additionally through technical assistance. Insulation can be effected through ensuring that quasi-fiscal deficits are explicitly financed by country budgets, reversing the recent trend to intemrationalize them irough fiancing part of the national bank's portfolio problems by BCEAO. There should be an early assessment of the current size of the quasi-fiscal deficit (and hence the future eamings position) of the two central banks and this put on the budgets of the various national governments, with allocation based on the original source of the problem. If neceary additional measures should be undertaken to secure a strong capital base of the cental banks. * Exit from the zone is best discouraged by securig the credibility of the zone. This will have to be supplemented by ensuring that those that exit becmse of macroeconomic problems do not have easier access to international sources of finance. 14 Refereaes Boccara, Bruno and Shantayanan Devarajan, 'Determinnts of Inflation in ihe Franc Zone in Africa," World Bank Policy Research Working Paper No. 1197, 1993. Boughton, James, "The CFA Franc Zonez Currency Union and Monetary Standard," Working Paper No. WP/911133, Inteational Monetary Fund, 1991. Devarajan, Shantyanan and Jaime de Melo, "Membership in the CFA Zone: Odyssean Journey or Trojan Horse?" In Chibber, Ajay and Stanley Fischer, eds., Economic Reform in Sub-Saharan Africa, World Bank, 1992. Goldstein, Morris and Geoffrey Wiglom, 'Market-Based Fiscal Discipline in Monetary Unions: Evidence from the U.S. Municipal Bond Market, Working Paper, International Monetary Fund, 1991. Honohan, Patridck, 'Price and Monetary Convergence in Currency Unions," Working Paper No. 390, World Bank, 1990. Kignel, Miguel, "A Note on the Devaluation Experiences of Low Inflation Countries", processed, World Bank, 1992. Medhora, Rohinton, "The West Afiican Monetary Union: Institational Arrangemet and the Link with France," Canadian Journal of Development Studies, 1992. Table 1. External Financing of Government Expeudit (billion CFAF) (and as proportion of total gover_ expenditure %) Burkina Faso COte d'Ivoire Niger 1976 0.8 31.8 5.8 -(4.0) (14.0) (16.4) 1977 2.3 52.4 6.3 (7.6) (13.8) (14.8) 1978 2.5 82.0 9.9 (7.6) (16.4) (17.0) 1979 1.0 99.3 14.3 (2.5) (16.5) (20.2) 1980 - 113.2 21.6 (18.5) (21.9) 1981 4.3 169.6 (9.0) (27.3) 1982 3.5 190.6 (6.1) (29.8) 1983 4.5 143.4 (8.4) (22.5) 1984 5.4 193.0 (9.9) (29-3) 1985 - - - 1986 - 81.7 57.7 (15.4) (39.5) 1987 - 66.5 55.2 (12.1) (38.4) 1988 - 207.2 (23.5) Notes: 1. Data for Benin, Mali, Senegal and Togo not available. 2. t- implies not available. Source: BCEAO. Raport Annuel, various. Table 2. Govermnt Net Borrowing: Foreign and Domestic (billion CFAI) Burkina COte Benin Faso d'Ivoire Mali Niger Senegal Togo 1976 - 5.8 - - - - 4.4 - 1977 0.89 - - 6.3 - 31.4 -0.43 - - 2.9 - 8.9 1978 1.80 - - 9.9 - 55.4 2.07 - - 4.3 - 5.7 1979 2.46 - 134.7 14.3 - 23.2 0.15 - 33.6 1.0 - -1.3 1980 - 1.31 139.6 21.6 -16.9 4.2 -0.13 107.4 8.3 10.6 -0.05 1981 - 2.90 - - 6.47 6.6 4.55 - - 23.5 7.4 1982 - 4.03 - - 26.08 3.1 4.29 - - 49.6 5.4 1983 - 2.05 - - 41.7 5.6 0.76 - - 15.9 17.3 1984 - 3.35 - 32.5 - 40.6 - 4.31 - 2.5 - 23.9 - 1985 - -1.67 - 2.5 - - 2.8 -1.46 - 3.3 - 3.3 Notes: 1. In each year, first row is for foreign borrowing, second row for domestic borrowing. 2. [ - I implies data not avaiable. 3. Domestic borrowing does not include arrears by government nor public-enterprise borrowing. Source: International Monetary Fund. International Financial Statistics. Table 3. Net BCEAO Iandingto Governmnnt and efinancing of AgricUlral Credit (billion CFAF) Burkina Cote Government Beni. Faso d'Ivoire Mali Niger Senegal Togo 1975 0.70 0 5.4 2.0 0 0 1976 0 0 6.2 1.0 -4.9 0. 1977 0.9 1.1 -6.9 0.7 1.0 3. 1978 0.5 2.0 -7.9 3.2 -2.3 B. 1979 0.3 2.0 20.1 3.3 -12.6 8.9 1980 3.0 2.0 45.7 6.8 -3.4 9. 1981 2.4 3.4 64.0 8.5 14.2 11. 1982 5.5 7.5 68.8 11.9 43.8 10. 1983 8.2 8.0 61.0 10.6 37.7 10. 1994 9.1 8.2 44.6 7.7 7.6 41.7 6.6 1985 11.1 9.4 109.6 9.9 8.8 34.3 13.6 1986 11.6 10.8 118.2 14.3 12.9 39.9 14. 1987 14.2 13.5 123.7 21.2 12.9 39.8 16.7 1988 14.9 17.3 125.2 26.3 12.9 39.9 17.8 1989 15.2 18.0 181.0 34.2 13.0 21.2 16.7 Agriculture 1994 0 0 63.2 2.3 3.9 4.6 0 1985 0 0.2 79.0 3.6 2.0 8.5 0 1986 1.0 2.4 82.6 6.1 6.2 20.2 0.4 1987 5.6 5.2 130.3 5.4 9.9 42.1 0.2 1988 8.7 2.4 158.1 0 7.4 64.4 1.0 1989 2.3 2.2 135.9 0 5.3 46.2 0.9 Note: 1. Figures for net lendig to the agridcultd sector are availae only for 1984 and after. Source: BCEAO REM=o ASmtud and Notes dInfbrmation et Statistigmes. Table 4. Real wage increses in selected CFA countries, 1985-90 (in percent per annum) Laborers Primary school teachers Benin 4 -3 Burkina Faso 4 1 Chad 1 6 C6te d'Ivoire 0 14 Mali -1 -8 Figure 1 Wages in the CFA Zone (1990 CFAFImonth) 160000- 140000- j '100000-- 80000- 40000- - O | | tmTeachers Benin Burkina CA D 3bin M eLors Guinea Niger Source: ILO Yearbook Policy Research Working Paper Series Contact Title Author Date for paper WPS1292 Services as a Major Source of Growth William Easteidy Aprl 1994 C. Rolilson In Russia and Other Former Soviet Martha de Melo 84768 States Gur Oler WPS1293 Product Standards, Imperfect Glenn Harrison April 1994 N. Ards Competition, and Completion of the Thomas Rutherford 38010 Market In the European Union David Tarr WPS1294 Regulations, Instiutons, and Hadi Salehi Esfahani April 1994 B. Moore Economic Performance: The Political 35261 Economy ol the Philippines' Telecommunications Sector WPS1295 Why Higher Fiscal Spending Persists Bruno Boccara April 1994 M. Pfeiffenberger When a Boom in Primary Commodtlies Ends 34963 WPS1296 Earnings-Related Mandatory Salvador Vald6s-Prieto April 1994 H. Rizkalla Pensions: Concepts for Design 84766 WPS1297 How Relative Prices Affect Fuel Charles C. Guo May 1994 C. Jones Use Patterns in Manufacturing: James R. Tybout 37699 Plant-Level Evidence from Chile WPS1298 Capial Goods Imports, the Real Luis Serven May 1994 E. Khine Exchange Rate, and the Current Account 37471 WPS1299 Fiscal Policy in Classical and Klaus Schmidt-Hebbel May 1994 E. Khine Keynesian Open Economies Luis Serven 37471 WPS1300 Dynamic Response to Extemal Klaus Schmidt-Hebbel May 1994 E. Kline Shocks in Classical and Keynesian Luis Serven 37471 Economies WPS1301 Estimating the Health Effects of Bart Ostro May 1994 C. Jones Air Pollutants: A Method with an 37699 Application to Jakarta WPS1302 Sustainability: Ethical Foundations Geir B. Asheim May 1994 C. Jones and Economic Properties 37699 WPS1303 Conflict and Cooperation in Managng Scott Barrett May 1994 C. Jones Intemational Water Resources 37699 WPS1304 Informal Gold Mining and Mercury Dan Biller May 1994 D. Biller Pollution in Brazil 37568 WPS1305 Iniormation and Price Determination Nemat Shafik June 1994 A. Yideru Under Mass Privatization 36067 Pollay Resaeroh Working Paper Series Contact ltit Author Date for paper WPS1m=8 Capital Flows and Long-Term Ibrahim A. Elbadaw June 1994 R. Martin Equilbilum Real Exchange Rates Raimundo Soto 39085 In Chile WPS1307 How Taxation Affects Forelgn Direct Joosung Jun June 1994 S. Klng-Watson Investment (Country Specific Evidence) 31047 WPS1308 Ownership and Corporate Control In Brian Pinto June 1994 M. Kam-Cheong Pdand: Why State Firms Defied the Swedervan Wljnbergen 39618 Odds WPS1309 Is Demand for Polluting Goods GunnarS. Eskeland June 1994 C. Jones Manageable? An Econometric Study Tarhan N. Feyziogiu 37699 of Car Ownership and Use In Mexico WPS1310 China's Economic Reforms: Ponrters Justin Ylfu Un June 1994 C. Spooner for Other Economies In Transitlon Fang Cal 30484 Zhou U WPS1311 The Supply Response to Exchange Mustapha Roule June 1994 J. Schwartz Rate Reform In Sub-Saharan Africa Weshah Razzak 322.50 (Empirical Evidence) Carios Molinedo WPSI312 The New Wave of Private Capital Eduardo Femandez-Arias June 1994 R. Tutt Inflows: Push or Pull? 31047 WPS1313 New Estimates of Total Factor Vlkram Nehru June 1994 M. Colerdge- Producthivty Growth for Developing Ashok Dhareshwar Taylor and Industrial Countries 33704 WPS 1314 The Signiflcance of the Europe BartiomleJ KaminaNk June 1994 M. Patefa Agreementtr for Central European 37947 Industrial Exports WPS1315 Global Tradable Carbon Permits, Bjom Larsen June 1994 0. Jones Participation Incentives, and AnwarShah .37754 Transfers WPS1316 Preserving the CFA Zone: Macroeco- Shantayanan Devaralan June1994 C. Jones nomlc Coordination Aftar the Michael Walton 37699 Devaluation