Policy, R...oh, end Extmnal AfJW5 O&§I7 WORKING PAFPQRS Mearocoonomlc Adjustment L *nd Growth Country Economics Department The World Bank April 1991 WPS 659 Inflation and Growth in the Transition from Socialism The Case of Bulgaria Andr6s Solimano A fragile political-macroeconomic equilibrium is bound to re- sult in high inflation in the transition from socialism. The collapse of growth in Bulgaria is the result of cuts in oil deliveries from the Soviet Union and Iraq, domestic dislocation in the supply of inputs following the dismantling of central planning, and the contraction of the Soviet market. The Policy,Research, and External Affairs Complex distributea PRE Working Paper to disseainate the fndins of wok in progress and to entounge the exchange of ideas among Bank staff and all othes inimested in dewelopnent issues. These papers eany the names of the authors, rdect only their views, and should be uscd and cited accordingly. The findings. interpretations, and conclusions ae the authors'own. They should not be attributed to theWorld Bank, its Board of Directors, its managen ent, or anyof its member countries. Policy, Researh, and EYtlfl Maimr Macrownomkc Adjustment and Gtowth WPS 659 This paper -a product of the Macroeconomic Adjustment and Growth Division, Country Economics Department-is partofa largereffort in PRF, to conduct researchon reforming socialist economies. Copies are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Emily Khine, room N11-067, extension 37471 (51 pages, with tables). Bulgaria's shaky macroeconomic situation is a least two major problems at a micro level. First, serious obstacle for a smooth transition from the rules of operation for firms in the productive central planning to markets. It has to correct sphere are still dominated by enterprises operat- large current account deficits with the convert- ing under soft budget constraints - with little ible currency area. It also has to eliminate price responsiveness. Second, in a setting of inflationary pressures and large price distortions. monopolistic competition, where individual And it has to get into a path of suv-^inable firms have considerable market power, full price growth. deregulation may reduce output. The links between inflation, money velocity, Bulgaria's moves toward a market economy the money overhang, and the fiscal deficit are are likely to affect growth through several crucial for assessing probable inflationary trends channels. The correction of macroeconomic in Bulgaria. SQlimano shows that with con- imbalances - cutting imports and cooling trolled prices and financial reprcssion, low aggregate demand to dampen inflationary velocity keeps inflation at an artificially low pressures - will contract aggregate economic level despite large fiscal deficits. But as prices activity. Reforms of the incentive structure will are deregulated and the financial sector is make part of the capital stock economically reformed, velocity can be expected to increase obsolete, hampering productive capacity in the - due to expectations of higher inflation and short run. The response of private investment to financial innovation. the new incentives will be highly sensitive to macroeconomic stability and the perceiv~d Solimano uses cost-determined price equa- probability that the reform process will last and tions to explore the effects on domestic prices of consolidate. Otherwise, private investors will a devaluation of the leva and an increase in the wait before acting, delaying the resumption of price of foreign inputs imported from the Soviet growth. Given these impediments, extemal Union and other CMEA countries. The input support in the fonn of new financing and direct price shock has a bigger effect on intemal prices investment will play a major role in consolidat- than does an equivalent devaluation. ing the reform and in the resumption of growth. The supply response to changes in relative prices and market incentives is likely to face at The PRE Working Paper Series disseminates the findings of work under way in the Bank's Policy, Research, and Extemal AffairsComplex. An objective of the series is to get these findings out quickly, even if presentations are less than fuly polished. The findings, interpretations, and conclusions in these papers do not necessarily represent official Bank policy. Produced by the PRE Dissemination Center TABLE OF CONTENTS 1. Introduction . . . . . .... ...... ... .... ... ....... .... .... ....... 1 2. The Bulgarian Economy: A Brief Background . . . . . . . . . . . . . 3 3. Macroeconomic Imbalances and Stabilization in the Transition From Socialisms A Framework . . . . . . . . . . . . . . . . .. . . . .. 7 3.1 Inflation, Fiscal Deficit and Velocitys A Numerical Example . 12 3.2 The Monetary Overhange .................. 15 4. Exchange Rate Adjustment and Foreign Price Shockss Impact on Domestic Prices . . . . . . . . . . . . . . . . . . . . . . . . . . 17 4.1 Effects of Currency Devaluation . . . . . . . . . . . . . . . 21 4.2 Effects of an increase in the price of imported inputs from the CMEA Zone . . . . . . . . . . . . . . . .. . . 25 5. The Determinants of Growth in the 1980s, the Collapse of 1990, and the Transition to a HarKet Economy . . . . . . . . . . . . . . 27 5.1 Determinants of Growth in the 1980s . . . . . . . . . . . . . 28 5.2 Transitional Issues: .. . . . . . . . . 35 5.3 The Collapse of Economic Activity in 1990 and other Transitional Effects on Growth . . . . . . . . . . . . . . . 40 6. Final Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 References . . . . . . . . . . . . . . . . . . .. 46 APPENDIX * Helpful comments made by B. Balassa, R. Caballero, V. Corbo, 0. Havrylyshyn, M. Kiguel, N. Liviatan, C. Michalopoulos, R. Rocha and J. Wilton to a previous version of this paper are appreciated. 1. Introduction The transiti i from central planning to a market-based economy in Bulgaria will be a complex process. On top of the structural problems inherited from decades of running the economy over strictly orthodox central planning lines, the economy faces now also severe macroeconomic imbalances, a feature that obviously complicates the transition. Economic growth in Bulgaria has decelerated sharply in the last four years and started to turn negative in 1989 turning to a definite collapse in 1990 with an additional drop in output of the order of -10Z. External debt is a serious problem as Bulgaria more than doubled its foreign debt in a few years (1985-1989) and currently its servicing is suspended for both principal and interest payments; moreover, the economy is in a critical need of foreign exchange to keep it running, let alone the servicing of its foreign debt. Domestic inflation started to accelerate in 1990 as prices are being deregulated, though near 85 percent of prices remain still under government control. In last May the domestic currency -- the leva -- was devalued and a two tier exchange tate system was established for current account transactions. The fiscal situation is bound to deteriorate significantly as a consequence of the devaluation of the leva besides of the effect of pending increases in current spending. Furthermore, on the external front, Bulgaria is expected to face a deterioration in its terms of trade particularly with the Soviet Union -- Bulgaria's main trade partner -- as by January 1991 the subsidies given to oil and other products that the USSR exports to Bulgaria (and other CMEA countries) will be eliminated. The paper is organized as follows. Section 2 provides a background on the structural features of the Bulgarian economy and its current macroeconomic situation. In section 3 we provide a simple framework to link fiscal 2 deficits, the money overhang, inflation and the real exchange rate in the process of restoring internal and external balance and eliminating repressed inflation. Section 3.1 provides a numerical exercise related to the monetization of fiscal deficits, money velocity and the rate of inflation in a context where prices are being decontrolled. The sensitivity of inflation to different degrees of velocity is examined as well as the issue of setting fiscal targets for anti-inflationary purposes. There we caution on the impact of financial deregulation on money-velocity durin, disinflation. Then in section 4 we look at the inflationary process from the cost side and assess the impact of two shocks on domestic pricest a devaluation of the leva and an increase in the price of inputs imported from the CMEt. zone (particularly from the Soviet Union). For that purpose a cost-based model of price determination is constructed and subsequently parameterized using the information provided by an input-output table available for 1987. The simulation of the price effects of those shocks is carried out under alternative assumptions regarding the degrees of (formal or informal) wage indexation. Section 5 discusses the determinants of growth in the 1980s in Bulgaria looking both at supply (factor accumulation) and demand factors. Furthermore, we deal with the issue of the level and efficiency of investment and provide also a decomposition of growth from the demand side. Moreover, the factors behind the huge drop in output during 1990 are spelled-out. Then the discussion turns to the issue of supply response to incentives at micro level highlighting the role of the enforcement of budget constraints and the market structure in getting supply response from changes in incentives. The section ends-up discussing the impact on economic growth of supply shocks, demand restraint, private investment response, resource reallocation and external 3 financing during the transition from socialism. The paper concludes in section 6. 2. The Bulgarian Economy: A Brief Back%round The Bulgarian economy is in a process of transformation from central- planning, the system adopted after World-War II following the lines of the Soviet model. From an agrarian economy until the World War Ill Bulgaria industrialized under strong prominence of heavy industry. The pattern of industrialization was defined according to the needs of the Soviet Union and some criteria of *comparative advantages" within the CMEA zone.2 The importance of the industrial sector is clearly dominant. In the 1980s over 602 of GDP, on average, was generated in industry. The agricultural sector, dominant before socialism, shrunk sharply in relative terms under the socialist systems its share in GDP went down from 632 in the late 1930s to 132 in 1989. Public property is overwhelming in Bulgaria and near 97 percent of GDP is generated in the socialist sector. The allocation of resources under socialism was carried out through quantity targets and direct inputs allocation. Central planning involved a complex interaction of planning bodies, holdings of firms and economic councils. The Bulgarian economy was able to sustain a high rate of growth of the Net Material Product, NMP, of the order )f 7-8 percent per year between 1945 and 1975, according to the official statistics. This growth record was heavily concentrated in industry, mostly 1 See Solimano 1990(b) for a historical account of the economies of Eastern Europe before socialism. 2 Import competition from the West was completely curtailed through prohibitive tariffs. 4 Droducing c'pital goods; In addition, the sector of consumption durable and services-relatec ictivities lagged well behind the capital goods sector. Central planners -- as in other socialist economies -- seemed to have followed a policy of maximizing the rate of growth of output through maintaining a very high investment rate (over 35 percent of GDP in Bulgaria). Drawing welfare implications from the growth strategy followed in Bulgaria is not easy, though it is apparent that keeping such high investment ratios tends to depress the growth rate of per-capita consumption thus preventing further improvement in the standard of living of the population. Also the problem of the quality of the goods produced -- particularly in the produiction of consumption durable -- is certainly not captured in the figures on GDP or NMP growth. However, regarding social sectors Bulgaria shows indicators of high birth expectancy, low infant mortality, access to and completion of general education that are among the highest of Eastern Europe. 5 Table 1 Bulaaria -- Selected Macroeconomic Indicators GDP Investment Inflat..',n Honey Current External (growth (1 of GDP) (rate of Supply Account Debt rate, Z) change,Z) (M1,S (convo. .rr, Millions of Grt) of U.S dollars) 1981 4.90 35.40 2.82 38.41a 1985 2.70 32.20 0.24 36.38 49.00 3940.00 1986 4.20 35.90 1.34 39.67 -715.00 5218.00 1987 6.10 32.90 0.07 43.74 -773.00 7082.00 1988 2.60 34.80 2.29 45.59 -840.00 8892.00 1989 -1.90 29.50 4.41 47.66 -1306.00 10223.00 1990 -10.00 70.00 e/ Sources The World Babk. notess e/ preliminary estimates. Inflation is measured by the rate of change in the GDP deflator.The current account and external debt is with the convertible currency area. a:l982. The macroeconomic performance of Bulgaria deteriorated since the mid- 1980s. GDP growth slowed down, turning negative in 1989, and the estimates for 1990 are even worse than for 1989. Open inflation was very low, though shortages in goods and inputs markets, provides indication of the existence of repressed inflation in the system. Since 1989 signs of an acceleration of inflation started to appear and that tendency got sharper in 1990 (see next two sections). On the monetary side the data show very high --and increasing -- ratios of Ml over GDP; in 1988-89 that ratio hovered around 45-50 percent 6 of GDP, a number exceedingly high for the standards of aon-socialist economies. However, this is consistent with the alleged existence of a money owerhang. It is also worth noting that in the last five years the ratio of Ml over GDr has grown. Money has grcwn faster than output in this period due to money financing of widening fiscal deficits. On the external sector, one important development of the last five years is the increase in the current account deficlt with the convertible currency area reflecting growing trade delicits and increased interest payments on foreign debt.3 Those deficits were financed, mainly, by an increase in external borrowing. In fact, the external debt of Bulgaria increased 2.5 times between 1985 and 1989. This impressive increase in external liabilities concentrated mainly with commercial banks in West Germany, Japan, the United Kingdom and Austria. Finally, by mid-1990 the macroeconomic situation severely worsened. GDP is expected to fall this year by 10 percent, and open inflation is accelerating. The government devalued the Leva in May 1990 and the fiscal situation -- including the Quasi-Fiscal deficit of the Foreign Trade Bank and the National Bank of Bulgaria -- deteriorated sharply (on an accrual basis) as a consequence of the adjustments in the exchange and in other outstanding expenditures. Furthermore, on the servicing of external debt, the country is in arrears of both principal and interests since April and its level of international reserves is meager. 3 The foreign trade of Bulgaria with the convertible currency area amounts near 25 percent of its total external trade. 7 3. Macroeconomic Imbalances and Stabilization in the Transition From Socialisms A Framework The transition from central planning to a market based economy may involve hard policy choices on the macro side in an economy starting the transition with large macroecornomic imbalances. In the case of Bulgaria at least three macro imbalances need urgent corrections * The deficit in the current account with th; convertible currency area has to be reduced. * The fiscal deficit is too high, near 5.5 percent of GDP in 1990; in addition, the quasi-fiscal deficit was of the order of 4Z in that year. - There is still repressed inflation and a large money overhang. A reduction in the current account deficit is needed to adjust the economy to the level of external financing available for Bulgaria on a more sustained basis. Moreover, further adverse external shocks are looming. Starting in January 1991, the Soviet Union will charge its deliveries of oil, natural gad, and raw materials to Bulgaria in hard currency and at international prices. In addition, currently, oil prices are soaring because of the Gulf crisis. Fiscal adjustment is required for several reasons: to improve the current account of the balance of payments, to reduce inflationary pressures and to release resources to the traded gcods sector that need to expand. The latter is important, if th external constraint facing Bulgaria is to be relaxed on a more permanent basis. An important part of the adjustment process in the transition from a centrally planned economy, is the deregulation of controlled prices. Price deregulation is further complicated because of the existence of a money overhang resulting from chronic shortages in the goods market and repressed financial markets. From a macroeconomic perspectivte a eoasic concern is how to avoid an acceleration of inflation that may be of a big magnitude and protracted when prices are deregulated. Let us present a simple formal setting to deal with these issues4. The goods market equilibrium can be written as the equality of full capacity output, YP, and aggregate demand, AD, where AD is a function of the real exchange rate, er, and the ratio of fiscal deficit over GDP, d. (1) YP - AD( er, d) The current account of the balance of payments, CA, is a function if the real exchange rate, the fiscal deficit ratio and L vector, f, of foreign* variables like external terms of trade, foreign real interest rates, and net capital inflows. (2) CA - CA( er, d, f) The rate of inflation is assumed to depend on the ratio of fiscal deficit to GDP and on /elocity of circulation of money, V. This relationship is obtained from the a*sumption that all the fisral deficit is financed from the inflation tax. In a stationary equilibrium (zero growth)s (3) f(M/P) - dY where f is the rate of inflation,and M/P is the stock of real balances (the nominal stock of money, M, divided by the price level, P). Solving for the rate of inflation, equation (3) can be expressed ast (4) V - V d where V (- PY/M) is money-velocity. The syaL X provided by (1), and (2) is represented in figure 1 in tae plane real exchange rate and fiscal deficit 4 This model is adapted from Dornbusch (1982) and Solimano (1984) for the case of an economy with price controls and repressed inflation. 9 ratio. The upward sloping schedule CA represents combinations of er and d that are consistent with equilibrium in the current account.5 An increase in the fiscal deficit ratio increases domestic demand and imports. In order to restore equilibrium in the current account the real exchange ratt must depreciate, giving rise to an upward sloping schedule. The downward sloping r locus GG represent goods market equilibriums a real depreciation of the exchange rate generates an increase in net exports that require a reduction in absorption -- through a cut in d -- in order to equilibrate aggregate demand with full capacity output. Figure 2 represents a positive, non-linear, relationship between fiscal deficits and inflation, given money velocity (equation (4)]. Here a distinction is made between *virtual, inflation (that is, the rate of inflation ruling in the absence of generalized price controls) and *observed' (official) inflation represented by the horizontal line in Figure 2: the rate of inflation observed when generalized price controls are binding.6 In turn, th.e difference between 'virtual" inflation -- the non-linear schedule -- and the rate of inflation under price controls is termed as the degree of *repressed' inflation in the sletem. It is worth noting also that changes in velocity of circulation will shift the 'virtual' inflation schedule, and given a set of price controls will change the degree of repressed inflation in the system. / 5 External balance could be redefined also as a certain (financeable) positive deficit in the current account rather than as a zero deficit. 6 The rate of inflation need Aot be necessarily zero since not all prices are controlled and fixed. 10 Point A in figure 1 denotes an initial position where equilibrium in the goods market takes place along with a deficit in the current account deficit (internal and external balance is at E). In this case the restoration of external and internal b-lance, moving the economy from A to E, requires a combination of reduction in the fiscal deficit along with a real depreciation of the exchange rate. In Figure 2 the fiscal adjustment required to restore macro equilibrium will reduce "virtual" inflation in the medium run. Besides as the system of price controls at excess den.and levels is relaxed (as part of the transition to a market economy), inflation will go up. Prices will increase due to the validation of repressed inflation, hence eliminating the gap between virtual and observed inflation. The adjustment process may get a further inflationary bias due to tne behavior of money-velocity. A reduction in the demand for money (increase in velocity) may take place for two reasonst first, the public may anticipate an increase in inflation as a consequence of the abandoning of the system of price controls at excess demand levels7. Second, it can be expected that as part of the reforms to the market economy a wider menu of financial assets will be available as capital market start to be developed. Hence financial innovation will shift the demand for money and increase velocity. Nonetheless, the inflationary impact of an increase in velocity is basically a transitional problem: in the medium-run, the rate of inflation will be determined by the level of the fiscal deficit and the (sustainable) level of the demand for money (inverse of velocity). 7 Formally we may assume perfect foresight and make velocity a positive function of actual inflation, V=V(f), V'>0. 11 Figure 1. Internal and External Balance er real exchange rate GG d dE dA Inflation, Fiscal Deficit, and Price Controls I (vl) inflation (V) (virtual inflation rate / rate) // / tow' _ " repressed inflation inflation under price controls ex ~~~~~~~d dE dA fiscal deficit ratio goods market equilibrium GG: y - AD (d, er) current account CA: CA - CA (d, ern f) ('virtual') inflation rate r = exp {V, d) 12 3.1 Inflation, Fiscal Deficit and Velocity: A Numerical Example In this section we illustrate the conceptual framework. The exercise aims to provide a quantitative sense of the sensitivity of inflation to changes in the fiscal deficit ratio and the ratio of base money to GDP. As we shall see below the exercise suggests that a slide into very high inflation rates following price decontrol is not an unlikely outcome in an economy with a high fiscal deficit ratio of GDP and with changing velocity: Table 2. Annual inflation rate under different velocity assumptions.High fiscal deficit case (fiscal deficit ratio, FD/GDP- 0.10) Ratio of base money to GDP, 2 rate of inflation (velocity is under parenthesis) (X) 50.0 22.14 (2.00) 40.0 28.40 (2.5) 20.0 64.87 (5.0) 10.0 171.83 (10) 5.0 638.91 (20) Note: The approximation r=Vd for the rate of inflation will lead to misleading results at moderate and high rates of inflation. Thus the continuous time formulation r= exp{Vdt is used to generate Tables 2 and 3. 13 Assuming that, initially, the deficit financed by printing money is 10 percent of GDP and that velocity is 2.0 (associated with a ratio of MI to GDP of 50 percent a number relevant for Bulgaria, see Table 1). For these values of the parameters -- velocity and the fiscal deficit ratio -- the rate of inflation would be 22.14 percent per year. That number looks "low" provided a fiscal deficit of 102 of GDP fully financed by printing money. Clearly the impact of the deficit on inflation is dampened by a low level of velocity (associated with a high money GDP ratio). However, this situation is bound to change (for the reasons given above) as price controls are lifted. Table 2 shows that the rate of inflation is highly sensitive to reductions in the demand for money or increases in velocity. In particular, as the ratio of money to output declines (velocity increases) inflation goes up sharply. Incidentally, inflation will reach three digits levels when the ratio of Ml over GDP falls below 152 given a fiscal deficit ratio of 10. Two important conclusions for designing stabilization policies emerge: first, the increase in velocity may greatly complicate the task of stabilization and price decontrol and second, the fiscal deficit needs to be sharply reduced if a low, (virtual) inflation equilibrium is to be achieved in an economy with deregulated prices8. Table 3 performs the same exercise but assuming that the fiscal deficit is reduced to 3 percent of GDP. 8 Of course some prices may still be regulated by the state in a market economy: typically, the price of public utilities, minimum wages and very often the nominal exchange rate. 14 Table 3 Annual rate of inflation under different velocity assumptions. Lower fiscal deficit case (fiscal deficit ratio, FD/GDP - 0.03) ratio of base money to GDP,Z rate of inflation (velocity is under parenthesis) (X) 50.00 6.18 (2.00) 40.00 7.79 (2.5) 20.00 16.18 (5.0) 10.00 34.99 (10.0) 5.00 82.21 (20.0) This second exercise shows that the reduction of the fiscal deficit from 10 Z to 3X of GDP reduces sharply the (virtual) rate of inflation for a given level of velocity. Thus, fiscal adjustment is certainly a necessary condition for getting inflation down. However, as it could be anticipated, inflationary expectations and credibility are of key importance to decelerate actual inflation. In terms of our exercises a low confidence scenario would be associated with high velocity; conversely a high confidence scenario is associated with relatively low velocity. Besides, we mentioned before it is important to have in mind that velocity and the structure of the demand for money will depend also on the reforms being implemented in financial markets as part of the structural reforms towards the market economy. Particularly, ther0 is need for caution in the speed and timing of liberalization of capital markets and the rate at 15 which new assets, substitutes of money, are introduced in the system. The results here show that changes in velocity--resulting greatly from "financial innovation" -- can seriously jeopardize the stabilization effort. Deep financial liberalization in the wake of a stabilization effort could produce undesirable results. Better stabilize first. Another issue that deserve some discussion is that of defining the required fiscal adjustment, particularly the setting of the target fiscal deficit. That target will depend on: (i) the level of inflation that the authorities consider tolerable on efficiency and stability grounds and (ii) the value of velocity and the level of the demand for money (as a ratio of GDP) that is expected to prevail in the medium-run. The international evidence might be relevant at this point. In chronic inflation countries (typically Latin America and Israel), a rate of inflation in the threshold of 15-20 percent per year is often, implicitly, considered as "tolerable" by the authorities; moreover, below that levels the cost of reducing an extra percentage point of inflation, in terms of lower output, may be too large for governments to find worth trying that additional disinflation. On the other hand, in Ea^.t Asia and in OECD countries, inflation tends to be in the one digit range. What is going to be the long-run rate of inflation in former socialist economies is still an open question. 3.2 The Monetary Overhang The problem of the monetary overhang has to be tackled in some way. The risk of an explosion in prices following deregulation of price in conditions of excessive liquidity is a real r.:sk. The alternatives to soak excessive liquidity ares i) a monetary reform that reduces the stock of nominal 16 balances, 2) let prices to get rid of excessive real balances, 3) the issuance of a bond to reduce liquidity and 4) privatization. The alternative of letting prices to go up an wipe-out excess liquidity runs the risk of not stopping in a once and for all increase in the price level. In practice a protracted period of inflation because of monetary accommodation and/or indexation may follow an initial jump in prices. In general, the less the social support to a stabilization program, the more likely government will avoid paying the cost of a fall in activity levels associated with non- accommodative policies, and therefore the more likely a slide into protracted inflation. The issuance of a government bond to absorb liquidity has the difficulty that the real interest rate to be paid to induce asset holders to acquire the bond may be very high in a context of uncertainty and lack of credibility. In addition, if the size of the money overhang is large the debt burden for the government will be large. Besides, it is unclear that the government should start stabilization with additional public debt that will strain public finances as prospective primary surpluses will have to be generated to serve interest payments on the bond. The alternative of privatization is attractive as it serves the purpose of reducing the role of the state in the production process and also to eliminate excessive liquidity. However, privatization will be necessarily a slow process and the money overhang is a pressing problem whose solution can not be postponed over time. Finally, the alternative of monetary reform may take different modalities in practice. A crucial issue is that its implementation requires political support as it will entail a tax on part of the wealth of the 17 population.9 4. Exchange Rate Adiustment and Foreign Price Shocks: Impact on Domestic Prices This section extends the analysis on the effects of macro policies and external shocks on domestic prices. The focus now is on the cost side of the economy. The Bulgarian economy in 1990 devalued the exchange rate (the leva price of the US dollar increased substantially in early May) and its full impact on domestic prices seems have not yet occurred. In addition, Bulgaria faces the increase in oil prices related to the Gulf crises and the realignment of prices towards international levels in the CMEA countries, particularly in the Soviet Union. Thus, currency devaluation and the increase in the price of imported raw materials and intermediate inputs are likely to have an impact on domestic prices whose magnitude is worth exploring. For that purpose we construct cost determined price equations in terms of direct and indirect value added and imported inputs from the convertible and non-convertible currency areas. Moreover, the simulation of the impact on domestic prices of currency devaluation and the increase in the price of impcrted inputs are carried-out under different assumptions regarding wage indexation. Let us provide analytical background for these calculations by starting with a production function F(.) where the level of output, Y [=F(.)], is produced with domestic value added, VA, and two kinds of imported inputs which are not perfect substitutes and therefore cannot be aggregated into a single composite. Those inputs aret M1, namely inputs from the convertible 9 See Gurley (1953) and Dornbusch and Wolf (1990) for analysis on the historical experience on monetary reform in post-world wvr II Europe. 18 currency zone and M2 inputs from the CEMA area. (1) Y = F(VA, M1, M2) We will assume positive first partial derivatives and negative aecond partial derivatives with respect to variable inputs. Unitary costs, c(.), will be a function of the price of value added, Pv' the local price of input imported from the convertible currency area, Pml, where Pml is equal to the exchange rate (units of leva per dollar) times the international price of M1 in dollars, PMl* adjusted by tariffs (l+t), Pml=eP*ml(l+t). The local price of the imported input from the non-convertible currency area will be Pm2. Assuming that domestic prices are set as a constant mark-up, over unitary costs we can write down the price equation as:'0 (2) P = P( Pv, Pml' Pm2) Let us assume a Cobb-Douglas production function (1)' Y - VAsvMlsmlM, sm2 where sv is the share of value added in output, aml is the share of input Ml in output and sm2 is the share of input 2 in output. In addition, sv+ sml +sm2 = 1. Now the price equation can be written as a Cobb-Douglas index (2)' p = PvsvPmllpm2sm2 This expression shows that the effect (in percentage changes) of a currency devaluation on domestic prices will be proportional to the share of M1 in output, sml. In turn, the impact of an increase in the international price of the input brought from the CMEA zone will be proportional to sm2. As we show below those effects may be enlarged if the price of value added -- say 10 The mark-up is not explicitly included in the price equation since, by assumption, is a constant. 19 wages -- is indexed to prices. The empirical base for obtaining the direct and indirect cost components of prices is an input-output matrix of Bulgaria for 1987. The methodology and the algebra of the construction of the price equations from the input-output table are provided in the appendix. We shall focus on four price indexes: the price of food products, the price of non-food goods, the consumption deflator and the output deflator. Starting with the price of food and non-food products. The Price of food, P1 iss (3) P1 . Pv0762 PM,0086 pM20.151 Price of non-food, P2: (4) P2 , p 0.628 pm10-061 p 20.31 Notice that the direct and indirect intensity of imported inputs from the CMEA zone is higher -twice as large-in the production of non-food goods than in the production of food; on the other hand, the production of food is relatively more intensive in the se of imported inputs from the convertible currency area. To compute the consumption deflator, Pc,we use the shares of food and non-food spending in total consumption, derived from the input-output table. The consumption deflator expressed in terms of value added and imported inputs is: (5) Pc N[Pv0.762 PM10.086 PM20.151]O.29[p 0.628pm10.061pM20.31]0.71 In a more compact form: (6) Pc= PV 0666pM10.068pm2O.264 Aggregating the input -output table into one sector and following the same procedure used to compute P1 and P2, we can obtain an equation for the 20 output deflator, P, (the net material product in this case). The expression is (7) p = Pv0.68 Pm 0.066P=20.252 It is interesting to note that the relative intensities of value added and imported inputs from both the convertible currency area and the CMEA zone are quite similar for the consumption deflator as well as for the NMP deflator. Table 4, summarizes the relative shares of the different cost components for the different price indexes.1 Table 4. Bulgaria. Value Added and Imported Inputs Shares in Unitary Costs (percentages) Value Imports of Imports of added Intermediate Intermediate goods from goods from non- convertible convertible currency area currency area Food Sector 76.2 8.6 15.1 Non-Food 62.8 6.7 31.0 Sector Consumption 66.6 6.8 26.4 Deflator NMP deflator 68.0 6.6 25.2 11 The concept of output implicit in the construction of the price indexes from the input-output matrix is the net material product. Since the NMP excludes some services activities like education, health and other government services and these activities are relatively labor-intensive, the economy-wide share of labor in GDP would tend to be greater than the share of labor in NMP. 21 4.1 Effects of Currency Devaluation The need to correct the external imbalances of the Bulgarian economy led the authorities to devalue the exchange rate with respect to the US dollar. The exchange rate was adjusted in May 1990 from 2.5 leva per US dollar to 7 leva per dollar for the "auctioning rate", and 3 leva per US dollar for imports of basic necessities from this area. Estimates of the effect of a nominal devaluation of the leva of 20X on domestic prices is provided in Table 5 below, under the asoumption of a fixed price of value added. This amounts, basically, to assume that the price of the variable factor -- labor -- is held constant e.g., a constant nominal wages.12 In this case, the impact-effect of a 20X devaluation on prices is rather small, in the order of 1.2 to 1.7 percent, depending upon the price index used. This clearly reflects the small share of imported inputs from the convertible currency area in the production structure of Bulgaria. Turning to wage adjustment, it is not unrealistic to assume that workers will demand some increase in nominal wages in compensation for an increase in the price level following a devaluation of the currency. Moreover, if a process of sustained inflation is set in motion, either because of accommodative monetaly policy (or increases in the fiscal deficit financed by printing money), there will be pressures to establish some sort of wage indexation mechanisms in Bulgaria.13 12 Implicitly we assume that the remuneration of capital is constant. That assumption will remain thereof. 13 It is not a necessary condition that formal wage indexation be established for a dynamic of indexation be observed, since informal wage indexation may also take place in a process of sustained inflation. 22 Typically, wage indexation is tied to a consumer price index.14 The impact of a ZOX currency devaluation under different degrees of wage indexation is shown in the third and fourth columns of Table 5. The algebra of indexation is as follows. Let us first denote the rate of change in the variable x as g. - 6x/x , and the relative share in unitary costs of the factor j in sector i as sji (see Table 4). Equation (11), shows the rate of change in the price of good i as the weighted average of the rate of change in the price of value added and the rate of change in the price of imported inputs from both trade areas. (8) gPi - OvigPv + smligPml + sm2igPm2 The wage indexation rule tied to the consumption deflator (equation 6) yields: (9) gw - Xgpc - X(O.666 gw + 0.068gpm. + 0.264gpr2) where X is the degree of wage indexation. Special cases are: r.o-indexation e.g., fix nominal wages, t0O and full-wage indexation, X-1; partial indexation implies 0 < X < 1. Further assumptions are: (a) the rate of change in the price of value added is equal to the rate of change in nominal wages, gp = gw, (b) the rate of change in the domestic price of 45mported inputs from the convertible currency area is equal to the rate of devaluation gpml get (implicitly we are assuming both constant dollar prices from the convertible currency area and fix tariffs), (c) the domestic price of imported inputs from the CMEA zone is fixed, gpm2=0. Under these assumptions, we obtain the 14 For simplicity we assume that wages are indexed to the consumption deflator of the same period. Other modalities of indexation are lagged indexation (tied to the price index of previous periods) or forward-looking indexation, tied to an estimate of future consumer prices. 23 following expression for the effect of a devaluation of ge percent on the rate of change in the price of the (composite) good i, gpij (10) gpi 0 O.068svi/(l- 0.666X) + 5Pmli I ge Notice that when nominal wages are fixed, X - 0, the effect of the devaluation on the price index of (the composite) good i will be given by the share of imported inputs from the convertible currency area in unitary costs of producing the good i, the coefficient spmli. Moreover the inflationary impact of the devaluation will be large. the higher the degree of wage indexation.In fact columns three and four of Table 5 show that the impact on prices of the devaluation of the leva is highly sensitive to the degree of wage indexation. Take for example the consumption deflator: while a 20 percent devaluation produces an increase in PC of 1.365 percent in the absence of wage indexation, the corresponding price increase goes up to 2.04 percent with a coefficient of wage indexatioa of 502, 0X - 0.5), and up again to 3.4 percent with 90 2 of indexation, ( - 0.9). Similar results hold for the other price indexes. 24 Table 5. Bulgaria. Effect of a 20Z Devaluation of the Exchange Rato (percentages) Degree of No Indexation Intermediate High Wage Indexation (X 0) Indexation Indexation - 0.5) - 0.9) Effect on: Price of Food 1.72 2.49 4.04 P1 Price of 1.22 1.86 3.14 Non-food P2 Consumption 1.36 2.04 3.40 Deflator, PC Consumer Price 1.41 2.09 3.48 Index * NMP Deflator,P 1.32 2.01 3.39 Note: * In the consumption deflator obtained from the input -output table the share of food is 0.29 and the share of non-food is 0.71. In the household expenditures survey, HES, of Bulgaria the share of food, that includes a larger range of goods, is 0.38 and the share of non-food is 0.62 (for '.989). Assuming that the rate of change in the price of the food and non-food items of the HES, are equal to the rate of change in the price of food and non-food of the I-0 table, and using the shares of the HES, we can calculate the effect of the devaluation on a consumer price index based in the HES.Ir. general, the effect of the devaluation on the consumption deflator and the consumer price index differ very little. 25 Two main conclusions emerge from this exercise: First, given the relatively low share of imported inputs from the convertible currency area in the unitary costs of the Bulgarian economy, the effect of a currency devaluation on prices, -- as determined by costs of production -- may be not very large; in particular, under the assumption of fixed wages. Second, the introduction of wage indexation schemes will increase the inflationary effect of a nominal devaluation and hence reduce its effect on the real exchange rate, because the second round effects from increased wages as a response to the devaluation. 4.2 Effects of an increase in the price of imported inputs from the CMEA Zone As mentioned above Bulgaria will face an increase in the price of the goods imported from the CMEA zone, particularly from the Soviet Union. Thus, the price of oil, natural gas, coal and other raw materials that Bulgaria imports from the Soviet Union are the most likely candidates to increase. A similar methodology used for computing the effects of a devaluation on internal prices will be used to asses the effects of an increase in the price of imported inputs from the CMEA area. The only new assumption is that domestic prices of imported inputs from the convertible currency area will remain unchanged, gpml-o (this implies that the nominal exchange rate, international prices in the convertible currency area and import tariffs will remain fix). An expression that gives the effect on the price index of the good i of a percentage increase in the price of imported inputs from the CMEA zone is given by 26 (11) gpi ' 0.264Xsvi/(1-0.666X) + spm23gpi2 The magnitude of the effect of an increase in Pm2 on Pi will depend, basically, on the share of imports of intermediate goo-s from the CMEA zone in the unitary costs of production of the good i and the degree of wage indexation in the economy. Table 6 provides the estimates of the effect of a 202 increase in Pm2 on various price indexes under different degrees of wage indexation to the consumer price index. Table 6 Bulgaria--Effects of a 20X increase in the price of imported inputs from the CMEA area Degree of Wage No Indexation Intermediate High Indexation = 0) Indexation Indexation (X - 0.5) (X-0.9) Effect ons Price of Food 3.02 6.033 12.05 P1 Price of Non-food 6.20 8.67 13.61 P2 Consumption 5.27 7.90 13.15 Deflator, Pc Consumer 4.99 7.66 13.01 Price Index * Deflator 5.04 7.27 13.08 NMP, P Note: * see footnote in Table 3. 27 From Table 6 follows that the domestic inflationary effect of a 20 Z increase in the domestic price of imported inputs from the CMEA is between two and three times higher than the effect of an equivalent devaluation of the leva with respect to the dollar, for a given degree of wage indexation. This results stems directly from the much higher intensity of inputs and raw materials imported from the CMEA area as compared to the share of inputs imported from the convertible currency area. Remember that around 75 percent of total trade of Bulgaria is conducted with the CMEA, a proportion that also seems to be observed for imports of intermediate goods. Furthermore, the sensitivity of these results to the degree of wage indexation, is high. In particular in the case of high wage indexation, )-0.9, a 20 percent increase in Pm2 will come along with an increase in domestic prices in a range between 12 and 13.5 percent, certainly a sizeable effect on internal prices. 5. The Determinants of Growth in the 1980s, the Collapse of 1990, and the Transition to a Market Economy The transition to a market economy will ultimately succeed if the reforms bring sustainable growth and an improvement in the standard of living of the population. Economic growth under socialism was historically rapid (according to the official statistics) and it relied in the extensive use of natural resources, was capital intensive and a concern for efficiency of resource use and high-quality goode was replaced by a drive to meet quantity targets. This section first reviews the growth performance of Bulgaria in the 1980s disentangling supply and demand factors in the explanation of observed growth rates. 28 Then the issue of the effects on growth of the transition to capitalism is dealt highlighting micro and macro factors deemed to be important in this regard. 5.1 Determinants of Growth in the 19809 The contribution of capital accumulation and labor growth to output growth is assessed as well as the contribution of consumption, investment and exports on growth on the demand side. In addition, the efficiency of investment is reviewed. Finally, the impact of short term stabilization and the structural reforms towards a market economy on growth is discussed.15 15 The system of material balances in Centrally -Planned economies calculates the Net Material Product (NMP) which is net output generated in "material" activities. Transport, commerce and communication are included in productive activities but health, education and most services are generally excluded from the material sphere for calculating NMP. To convert NMP figures into Gross Domestic Product (GDP) the following adjustments have to be made:(i) the value added created in the non-material sphere is added to NMP, (ii) the value of non-material output consumed in the material sphere is subtracted to avoid double-counting, (iii) depreciation exclusive of capital repair, is added the latter being considered a current expenditure in the system of national accounts but a capital expenditure in the system of material balances. Typically, GDP to NMP ratios are in the order of 1.15 to 1.53 and tend to rise over time with the relative expansion of the service sector. 29 Table 7. Economic Growth in Bulgaria (rate of growth, percentages) GDP Agriculture Industry Services 1981 4.90 4.47 5.47 4.10 1982 2.30 5.79 9.3 - 12.66 1983 3.40 - 16.34 7.64 7.28 1984 3.40 11.68 6.13 - 7.44 1985 2.70 - 20.62 5.09 11.35 1986 4.20 22.04 5.19 - 6.50 1987 6.10 - 14.v4 5.37 20.48 1988 2.60 - 1.71 2.43 4.92 1989 - 1.90 - 2.36 0.33 - 5.44 1990* -10.0 annual averages 3.07 - 1.29 5.22 1.78 Source: World Bank. * Estimate. Table 7 shows that during the 1980s GDP decelerated its rate of growth. On average the annual rate of growth of GDP (1981-89) was 3.07; over one percentage point less than the 4.3 percent growth of GNP in the period 1971- 80. At a sectoral level, the driving sector behind growth was industry -- a sector with a share in GDP of 61 percent -- with an annual average rate of growth of 5.22 Z in the period 1981-89. The performance of agriculture in that period was dismal with negative average annual growth. The growth record of the service sector in the 1980s has been rather modest: its growth rate 30 fell below the growth rate of industry (and GDP) but it was above that of agriculture. To gain a better understanding on the sources of growth in Bulgaria let us calculate the contribution of labor and capital to measured growth. The following equation for the rate of GDP growth is useful in this respects Syly - h 514K + (1-h) 6LIL + e where, h is the share of capital in va'tIu added and 1-h is the share of labor, Ochcl. The rate of growth of GDP is denoted by 6y/y. the rate of growth of capital is given by 54KI and the rate of growth of labor is 6LIL. The term e represents the residual, namely the rate of growth of output that is not explained by capital accumulation and employment growth. This term is often imputed as technical progress but it could also reflect measurements problems both in the aggregation of output and in the quality of the factors of production. 31 Table 8. Economic Growth in the 1980,s the Role of Labor and Capital Accumulation10 (annual averages,percentages) rate of growth 2.370 of capital rate of growth 0.295 of labor contribution of capital accumulations percentage points 1.232 percent of total 40.93 contribution of labor growth : percentage points 0.141 percent of total 4.70 residual 1.63 rate of growth of GDP 3.07 Note: the share of capital is 0.52 and that of labor is 0.48. They were obtained from the input-output table. The results skow several interesting results regarding the sources of growth in Bulgaria in the 1980s. First, it is clear the overwhelming relative importance of capital accumulation, as compared to labor growth, in explaining GDP growth in this period. Around 40Z of GDP growth is explained by capital 16 The rate of growth of capital was calculated as the product of average net investment ratio over GDP divided by the capital-output ratio. The rate of growth of labor corresponds to the rate of growth of total employment. 32 accumulation. Second, the very low contribution of labor growth to total growth is due mainly to the very low rate of growth of employment , 0.295 percent per year (the annual rate of growth of population in Bulgaria is around 0.4 percent).17 Third, the size of the unexplained residual, 1.63Z per year, as a percentage of output growth explained near a half of real growth.18 However, we have to be careful in inferring that the rate of technical change is large in Bulgaria from this value of the residual. As mentioned before the residual could be substantially reduced if factor inputs could be measured more precisely. In particular the measure of aggregate capital conceal different vintages of capital equipment that may have different marginal productivities. Another extension of the analysis would be to distinguish between human capital and unskilled labor. Furthermore, the data on output may reflect index numbers problems since prices do not reflect relative scarcities of goods and productive factors. Given the importance of capital accumulation in the growth process it seems useful to examine some features of investment and particularly assess, in a first approximation, the efficiency of investment. 17 In the methodology of sources of growth another cause of low contribution of labor, besides a slow rate of growth of employment, would be a low marginal productivity of labor. However in a socialist economy, real wages can not be assumed to be equal to the marginal productivity of labor so this effect is hard to measure. 18 Empirical studies using the sources of growth methodology in non- socialist developing countries, have found that, on average, technical progress accounts between 1/2 and 1/3 of real growth. 33 Table 9. Bulgaria Investment rates 1979-88 (share of GDP, percentages) Net Fixed Capital Formation 10.80 Depreciation 13.90 Unfinished Construction 2.20 Gross Fixed Capital Formation 26.90 Changes in Inventories 6.80 Gross Investment 33.70 (Index) Incremental Capital-Output Ratio 3.51 Average Capital-Output Ratio 4.47 Notes the incremental -output ratio is calculated as the quotient between net investment and GDP growth. Table 9 illustrates some interesting features of the investment process in Bulgaria. Gross investment is very high, (33.7 percent on average for the period 1981-89 and it was even higher in previous periods), to sustain a rate of growth of GDP of around 3 percent19. Computing both ICOR's -- the ratio of net investment over the rate of growth of GDP -- and the capital-output ratio -- the inverse of average capital productivity -- both ratios yield high values, at least for the standards of non-socialist economies. This supports the notion of a rather low level of investment efficiency in Bulgaria. Moreover, it is interesting to mention the very high inventory accumulation 19 These high investment ratios are by no means unusual in socialist economies, see Solimano (1990a). 34 purported by the data. On average for a period of 9 years the accumulation of inventories represented near 7 percent of GDP. This needs to be explained. A possible explanation lies in the production of defects or low quality goods that cannot be sold in the market; it also may represent delays into putting capital goods in the production process. Given the rigidities of the planning system any self correcting mechanism to reduce the degree of inventory accumulation will operate in a weak way. To sum-up, discrepancies between investment expenditure and additions to the capital stock, an important factor behind poor investment efficiency, develop as many investment projects are never completed, new equipment is not brought into operation, and remains as inventoricw and imports of capital goods from the West do not represent the most up-to-date equipment (see CEPR, 1990). 35 Table 10. Economic Growth in the 19808: A Demand Decomposition (average annual rates of growth, percentages) 1981-1989 Total Consumption 3.29 -Private Consumption 2.40 -Public Consumption 5.10 Gross Investment 2.79 Total Exports 6.28 -to convertible area 5.25 -to non convertible area 6.86 Total Imports 5.44 -from convertible area 7.83 -from non-convertible area 2.02 GDP 3.07 On the demand side public consumption and exports to the non-convertible currency area drove growth in the 1980s. Investment grew at a slower pace than GDP leading to a reduction in the investment share, (though it is still very high). Imports, in dollars have been growing at a higher rate than GDP and exports thus enlarging the (convertible currency) trade deficit. Regarding trade with socialist countries, exports to that area grew almost three times more rapid than imports. 5.2 Transitional Issues: A key objective in the transition from central planning is to make relative prices and decentralized markets to play a major role in supply 36 decisions. At a macro level, getting output response from changes in relative prices is particularly useful as it would ease the costs of the transition. Two complications for a sizeable supply response are likely to arise at a microeconomic and sectoral level: the first is that just a (very small) fraction of firms operate in a regime of binding budget constraints and their objective function is to maximize profits --or minimize costs -- so that we can expect their supply decisions to be price-dependent. The second complication refars to the fact that price deregulation may lead to a drop ir. output when the market structure is characterized by firms which have market power e.g., monopollv#4. competition.20 The coexistence of a; A- ;et constrained (optimizing) firms with soft budget constrained firms. In Bulgaria large segments of the enterprise sector still operates under a regime of soft budget constraint in the socialized sector. In the regime of soft budget constraints, if the firm requires more inputs (including credit) to meet the targetr of physical production, the firms will get those inputs at no extre. cost (e.g., firms are not budget constrained). Thus, the level of output of the traditional socialist firm can be expected to be largely insensitive to changes in relative prices or in market profitability. On the other hand, private (or public) firms that maximize profits (or minimize costs) subject to a hard budget constraint are expected to react to prices.21 Formally, the aggregate level of output, y, will be a function of 20 I owe this point to Paul Krugman. 21 The price-elasticity of supply depends on technological parameters, the time period of the production process and in a context of uncertainty on the mean and variance of the distribution of relative prices faced by the firm. 37 the level of output produced by the set of optimizing-budget constrained firms, yo, whose supply function can be written as yo - pa where the coefficient a denotes the price elasticity of supply. In turn, the (aggregate) level of output of the set of soft-budget constrained firms is denoted by ys and is independent of p. Then, y - #(Yo# Ye) Assuming that the function # is Cobb-Douglas with shares b and 1-b, we can write the aggregate supply function of the economy as yu(pa)b(ye)l-b where 0 < b< 1. The price elasticity of aggregate supply is - 6logy/61ogp - ba Then the output response to a change in relative prices in the aggregate depends on two parameterst(i) the share of sectors in the economy that are subject to a binding budget constraint and optimize an objective function, the coefficient b, and (ii) the price elasticity of the aggregate supply of the hard budget constraint sectors. Clearly, the rules of operations cf public enterprises and the relative importance of the private sector as producers (in Bulgaria nowadays the private sector contributes just to 3 percent of GDP) are important elements behind the supply response to changes in relatives prices. Furthermore, it is worth to keep in mind that privatization has an important role to play in introducing market discipline to enterprises.. 38 Deregulating Prices in a Monopolistic Setting The market structure is also important for the output response to changes in relative prices. In centrally-planned economies the most frequent disequilibrium in the goods market is excess demand. Then the standard recipe is price liberalization supplemented by tight demand policies. Let examine the case in which price deregulation takes place in a setting of monopolistic competition, where every firm enjoys some market power. The market equilibrium is represented in Figure 3 where on the vertical axis we have the relative price Pi/P which represents the ratio between the price Pi the monopolistic producer charges over average prices P. In symmetric equilibrium Pi - P. On the horizontal axis we have the quantity of the good, Qi. The downward demand curve facing producer i is given by D(Pi/P, MIP) where M/P denotes the stock of real balances. The upward sloping curve MC is the marginal cost schedule of producer i and the negatively sloped schedule MR is the marginal revenue. The *old* equilibrium will be associated with a price Pj/P - 1 where firms cannot raise their prices. In turn, at that price ratio firms produces at QiO, wlth excess demand, (Qid > Qc)' generating a black market premium and rents. Now as part of the liberalization program, prices are deregulated and firms are allowed to rise their prices. The increase in the price level will reduce the stock of real balances shifting the demand curve down to the point where the firm setting its optimal price ends up charging P. Therefore at the new equilibrium -- under deregulated prices -- the new level of outpul Qin (given by the intersection of MC with the new marginal revenue schedule associated to Dn) which is lower than the initial level of output under 39 regulated prices. 0. course, the point must be not overstated since there were shortages at the "old' equilibrium and consumers paid an additional implicit price in terms of searching costs to get the goods they desire besides the "unproductive" rent-seeking involved in the initial situation characterized by shortages. Anyway a main message from this exercise is that free prices under monopolistic competition might not be welfare improving. 40 Figure 3. Price Deregulation under Monopolistic Cormetition D(M, P,) lRaN ~DN (H/PN) QiN Qio Qi 5.3 The Collapse of Economic Activity in 1990 and other Transitional Effects on Growth The large contraction in economic activity developed in 1990 was closely related to supply shocks like: (i) the disruption of deliveries of oil from the Soviet Union and Iraq; (ii) the shortages of domestic inputs because of the breakdown of distribution networks associated with the collapse of central planning; and (iii) the loss of export markets in the Middle-East and the contraction of the Soviet and CMEA Markets. Regarding the effects of the transition from central planning on the rate of growth in the short to medium run, several effects are relevant. A first channel is linked to the correction of large macroeconomic imbalances observed in Bulgaria. The running of large deficits in the current account with the convertible currency 41 area and the existence of serious inflationary pressures call for tight demand policies that may decelerate growth in the short run (since 1989 GDP growth is negative). Cooling down growth in aggregate spending tends to ease the inflationary pressures on the demand side.22 Then, the system is likely to pass from being supply constrained to become dema,.d constrained. A second channel through which the transition from central planning may affect growth is through changes in relative prices and incentives that make part of the existing capital stock (and skills of the labor force) economically obsolete. Then if Bulgaria under the CMEA arrangements was concentrated in producing, say, machinery for heavy industry and computers to be sold in the Soviet Union in exchange for oil and raw materials, now it will have to concentrate in developing the production with an outlet in western markets, e.g., wine, fruit and light industry. During the transition period it is clear that part of the capital stock used to produce the Hold" goods will be inappropriate to produce the 'new' goods. That will adversely affect the level of output and perhaps its rate of growth in the short-term.23 A third set of factors relates with the response of investment to the new structure of incentives that is expected to emerge from the reform process. The basic argument for expecting an initial weak investment response in the aggregate stems in two elements: first, the process of reform is characterized by systemic uncertainty as the final fate of the reform process 22 This should be qualified, however, since a reduction in real growth may worsen the fiscal deficit by reducing seigniorage. This can be an inflationary shock. 23 The improvement in efficiency of resource use may counteract part of this effect. The composition of investment may become more efficient and that may speed up growth in the medium term (see Easterly, 1990). 42 is unknown. Under these conditions, the rational investor has large incentives to wait or delay its investment decisions until the level of perceived uncertainty is reduced. A second element for a quantitatively small investment response in the aggregate is the fact that the private sector in Bulgaria is extremely small and probably inexperienced in the rules of operation of a market economy, something that can be reversed with privatization over the medium run. Additional complications for an adequate investment response relate tos (i) the lack of a legal setting that define and protect property rights, (ii) the absence of well-developed capital markets that provide adequate financing for p..ofitable investment opportunities, and (iii) the lack of modern infrastructure in terms of roads and telecommunications. Last, but not least, Bulgaria needs additional external financing to support sound investment and modernize its productive structure. A major obstacle in this regard is the foreign debt overhang. Then a solution to the foreign debt problem and the possibilities of getting new external financing are closely linked issues. Finally, direct foreign investment could be amply welcomed as it would bring new technologies and managerial capabilities and access to new external markets. 6. Final Remarks The shaky macroeconomic situation of Bulgaria is a serious obstacle for a smooth transition from central planning to a market based economy. The correction of large current account deficits with the convertible currency area, the elimination of inflationary pressures and large price distortions and a resumption of sustainable growth in the context of structural change are 43 the key issues in the policy agenda for Bulgaria. This paper examined both analytically and empirically several issues associated with stabilization and growth in the transition from socialism in Bulgaria. The links between inflation, money velocity the money overhang and the fiscal deficit are of crucial importance for assessing probable inflationary trends in Bulgaria. It is shown that in a context of controlled prices and financial repression, low velocity contributes to maintain inflation at an artificially low level in spite of large fiscal deficits. However as prices are deregulated and the financial sector is reformed, velocity may be expected to increase due to both expectations of increased inflation and/or financial innovation. In that context the rate of inflation may jump to high levels and perhaps a protracted process of inflation set-in with monetary accommodation, wage indexation and lack of fiscal adjustment in the background. Stabilization and the setting of fiscal targets eonsistent with low inflation becomes a complex task provided the inscability of the demand for money in a context of uncertain stabilization and financial innovation. Using empirical, cost-determined, price equations we explore the effects on domestic prices of a devaluation of the leva and an increase in the price of foreign inputs imported from the Soviet Union and other CHEA countries, two price shocks underway (besides also the impact of the Gulf crisis on oil prices in international markets). The empirical results show, that the input price shock has a more important quantitative effect on internal prices than an (equivalent) devaluation. This result arise from the greater (relative) reliance on inputs imported trom the CMEA zone than from the convertible currency area. The introduction of wage indexation in the wake of price 44 adjustment measures is shown to exacerbate sharply the effects on domestic prices from devaluation and/or the external price shocks. The supply response to changes in relative prices and the introduction of market incentives is likely to face at least two major problems at a micro level. First, the rules of operation of firms in the productive sphere is still largely dominated by the socialized sector of enterprises operating under a regime of soft budget constraints, with the ensuing little price responsiveness. Second, it is shown that in a setting of monopolistic competition. where individual firms have considerable market power, full price deregulation may bring about a drop in output. Economic growth in Bulgaria decelerated sharply in the 1980s and completely collapsed in 1990. A factor decomposition exercise shows that GDP growth was explained mainly by capital accumulation (40 percent) and the residual (53 percent) with a very little contribution of labor growth. From the demand side, growth has been driven mainly by public consumption and exports to the non-convertible currency area. Moreover, the collapse of economic activity in 1990 is closely related to the cut in oil deliveries from the Soviet Union and Iraq, the interruptions in tha supply of intermediate and capital goods associated with the dismantling of central planning and the loss in export markets in the Middle-East and the contraction of the Soviet market. The transition to a market economy is likely to affect growth through several channels. The correction of macroeconomic imbalances -- involving a cut in imports and a cooling down of aggregate demand to dampen inflationary pressures -- tends to generate a contraction in aggregate economic activity. The structural reforms that will alter the incentive structure will render part of the existing capital stock .-- shaped by the economic structure 45 of central planning and the division of labor imposed by the CMEA trade arrangements -- as economically obsolete therefore hampering productive capacity in the short run. The response of private investment to the new incentives will be highly sensitive to the degree of macroeconomic stability and the perceived probability that the reform process will last and consolidate. Otherwise it becomes rational for the private investor to wait - - an option with a high premium in an unstable environment -- therefore delaying the resumption of growth. Finally, external support in the form of new financing and direct investment will necessarily play also a major role in helping consolidate the reform process and help in the resumption of growth. 46 References Dornbusch, R. (1990) "From Stabilization to Growth" World Bank Annual Conference on Development Economics. April. - -- (1982) Stabilization Policies in Developing Countries in Developing Countriess What Have We Learned? World Development. ------- and Wolf (1990) "Monetary Overhangs and Reform in the 1940s N.B.E.R. Working Paper. CEPR (1990) Monitoring European Integration. The Impact of Eastern Europe. Gurley, J. (1953) "Excess Liquidity and European Monetary Reforms, 1944-1952", American Economic Review, Vol. XLIII, March. Easterly, W. (1990) wPolicy Distortions, Size of Government and Growth," N.B.E.R. Working Paper # 3214. International Monetary Fund (1990). "Data base on Bulgaria." C.A. Rodriguez'(1989) "Macroeconomic Policies for Structural Adjustment,' PPR Working Paper # 247. The World Bank. A. Solimano (1990a) "Macroeconomic Adjustment, Stabilization and Growth in Reforming Socialist Economies. Analytical and Policy Issues," PPR Working Paper # 339. The World Bank. ------- (1990b) "On Economic Transformation in East-Central Europe$ A Historical PerspectIve." mimeo. The World Bank. ------- (1984) wDevaluation, Unemployment and Inflationt Essays on Macroeconomic Adjustment," Chapter 3. Ph.D. thesis, M.I.T. (unpublished). L. Taylor (1979) Macro Models for Developing Countries McGraw-Hill. The World Bank (1990) Bulgaria Data. 47 APPENDIX Construction of Price Equations from the Input-Output Table As it is well known the input -output table provides a breakdown of both the intermediate and final demands for the goods produced by each sector of economic activity as well as the cost structure in terms of purchases of intermediate goods, value added and imports of intermediate goods used in each sector's production.24 Table A.1. provides an input-output, 1-0, table for Bulgaria at millions of 1987 leva (the local currency). The sectors of economic activity are aggregated in twos the "food sector, and the "non-food sector'. The food sector corresponds to industry 017 of food, beverages and tobacco; the non- food sector corresponds to the rest. In terms of demand for final consumption the food sector comprises 29 percent of total consumption expenditure in the economy. With this share (and that of non-food) we can construct a deflator for consumption. In turn an output deflator (for Net Material Product) will be also computed ix terms of its cost c,mponents from the input-output table. Reading across a row in the 1-0 table we can obtain the following material balances linking output supply with intermediate and final demands. In matrix forms (1) X - AX + p 24 The underlying technology in the input-output table is that of f ix- proportions. This is not an overly restrictive assumption for Bulgaria, still a centrally planned economy -though starting a period of transformation- and where the response of the factor mix to changes in factor prices is very weak if it does exist at all. 48 The column vector X denotes the level of production for the two sectorst the food sector X1 and the non-food sector X2 the aij elements of the square matrix A (of order 2x2) are the input-output coefficients showing the intermediate sales from sector i to sector j for ij -{1,2}. The components F1 and F2 are the final demands (consumption, investment and net exports) for food and non-food output. Reading across a column we can obtain the following decomposition of cost per unit of output in sector is (2) Pi - £ aji Pj + ViPv + mmliPm. + m2iPm2 The variable Pi is the cost determined output price in sector i, Pv is the price of value added , Pm, is the price in local cur:ency (leva) of intermediate goods imported from the convertible currency area. This price can be written as PM, - eP*(l+t) where e is the exchange rate (units of leva per US dollar), P * is the international price, in dollars, of inputs imported from the convertible currency area and t is the ad-valorem tariff rate. The price Pm2 is the domestic price of foreign inputs purchased in the CMEA zone. The coefticient vi is the amount of 'ralue added required to produce one unit of good i, the coefficient mli is the amour.t of intermediate imports from the convertible currency area required to produce one unit of good i, and m2i is the amount of inputs imported from the CMEA zone required to produce one unit of good i. 49 TABLE A.2. BULGARIA INPUT-OUTPUT COEFFICIENTS Food Non-Food Food all a12 (0.247) (0.118) Non-Food a2l a22 (0.069) (0.461) Value Added vl V2 (0.183) (0.303) imports from Convertible Mii m12 currency area (0.0210) (0.029) imports from Non-Convertible m21 m22 Currency Area (0.028) (0.120) Table A.2. provides the input-output coefficients required to parameterize equation (2) with the information provided by the input-output table. 50 Equation (2) as it stands, shows direct unitary costs in terts of value added and imported inputs (from both trade areas), and the costs of intermediate purchases, of producing a unit of the composite good i. To obtain the direct and indirect unitary costs in terms of primary factors and imported inputs we need some transformations. Restating equation (2) in matrix form, we get (see Taylor, 1979): (3) PT(I-A) = PvvT + P1m1T +Pmm2T The term PT is a row vector of the price of food and non-food output e.g., pT.1p1 P2]; vT, mjT,and m2T are row vectors of coefficients of value added- output and imported inputs-output. Inverting the matrix (I-A), the so-called Leontieff inverse, we obtain:25 (4) PT * (PvvT + Pm1m1T + Pm2m2T][I-A]-l Equation (4) shows the direct and indirect cost per unit of output of both sectors (food and non-food) in terms of value added and imported inputs from both the cornvertible currency area and the non-convertible currency zone. For example, in the case of food production, it adds-up the content of value added and imported inputs used directly in the production of one unit of food and the content of value added and imported inputs used in the production of domestic goods that enter as intermediate goods in the production of food (indirect component). 25 pT is conformable of order 1x2, the first right hand side row vector in (4) is of order lx2 , and the matrix (I-A] is of order 2x2. 51 Using the coefficients calculated in table 2, and inverting the matrix 1I-A] we can apply equation (4) and weite the price of food and non-food just as a function of Pv and Pm5 and Pm2 (direct and indirect unitary costs). Then the price of food is s (5) P1 - 0.301 Pv + 0.034 Pml + 0.06 Pm2 The price of non-food is (6) P2 - 0.628 PV + 0.061 Pml + 0.31 Pm2 Nevertheless, the parameters of Pv and the Pmi are inputs-outputs coeffici3nts and therefore they are not unit -free. To get the relative shares of each cost component we adopt the normalization Pv - PM, - PM2 -1 , then compute the sum of unitary costs and get the corresponding relative shares. Note that the resulting relative shares will be completely independent of the units at which factor prices, Pv and Pmj are measured in the normalization. PRE Working Papsr Series Contact Aiflb4 .Qap WPS632 The Macroeconomics oT the Public Carlos Alfredo Rodriguez March 1991 R. Luz Sector Deficit: The Case of Argentina 34303 WPS633 The Macroeconomics of the Public Virabongse Ramangkura March 1991 R. Luz Sector Deficit: The Case of Thailand Bhanupongse Nidhipralha 34303 WPS634 Trends in Developing Country Bela Balassa March 1991 WDR Office Exports, 1963-88 31393 WPS635 Exchange Rates and Foreign Trade Bela Balassa March 1991 C. Ctiskelly-Young in Korea 39413 WPS636 Economic Integration in Eastern Bela Balassa March 1991 C. Cuskelly-Young Europe 39413 WPS637 Poverty in Poland, 1978-88 Branko Milanovic March 1991 WDR Office 31393 WPS638 Researching the Trade-Productivity James Tybout March 1991 D. Ballantyne Link: New Directions 37947 WPS639 The High Cost of Protecting Wendy E. Takacs March 1991 D. Ballantyne Uruguay's Automotive Industry 37947 WPS640) -he Impac, of Policy in African William Jaeger March 1991 A. Yideru Agriculture: An Empirical Investigation 34663 WPS641 Intertemporal Substitution .n Karsten N. Ptedersen April 1991 M, Hileman Consumption: Evidencefor Some 31284 High- and Middle-income Countries WPS642 How a Change in Brazil's Sugar Brent Borrell April 1991 A. Kitson-Wahers Policies Would Affect the World 33712 Sugar Market WPS643 Regional Integration among Andras Inotai April 1991 D. Ballantyne Developing Countries. Revisited 37947 WPS644 Trade and Payments Arrangements Constantine Michalopoulos April 1991 M. Colinet in Post-CMGA Eastern and Central David Tarr 34898 Europe WPS645 Poverty, Policy, and Industrializa- Ben Polak April 1991 WDR Office tion: Lessons from the Distant Past Jeffrey G. Williamson 31393 WPS646 The Developmental Effectiveness Tony K.llick April 1991 S. King-Watson of Aid to Africa 33730 WPS647 Growth Rates and Aggregate Nanak Kakwani April 1991 B. Rosa Welare: An International Comparison 33751 ., PRE Working Pap.r Serias Contact B6 Atuthnr f g2= WPS648 Who Paid the Bill? Adjustment and M. Louise Fox April 1991 WDR Office Poverty in Brazil, 1980-95 Samuel A. Morley 31393 WPS649 An Observation on the Bias in Margaret E. Grosh April 1991 B. Diallo Clinic-based Estimates of Kristin Fox 30997 Malnutrition Rates Maria Jackson WPS650 Administrative Valuation of Soviet Karen Brooks April 1991 C. Spooner Agricultural Land: Results Using 30464 Lithuanian Production Data WPS651 Taxation of Financial Assets in Christophe Chamley April 1991 A. Bhalla Developing Countries 37699 WPS652 Demographic Response to Economic Kenneth Hill April 1991 WDR Office Shock 31393 YWPS653 The Effects of Option-Hedging on the Donald F. Larson April 1991 D. Gustafson Costs of Domestic Price Stabilization Jonathan ColemfAn 33714 Schemes WPS654 Reflections on Credit Policy in Mansoor Dailami April 1991 M. Raggambi Developing Countries: Its Effect on Marcelo Giugale 37657 Private Investment WPS655 Interest Rate Policy in Egypt: Its Mansoor Dailami AprW 1991 M. Raggambi Role in Stabilization and Adjustment Hinh T. Dinh 37657 WPS656 Relative Deprivation and Migration: Oded Stark April 1991 M. Felix Theory, Evidence, and Policy J. Edward Taylor 33724 Implications WPS657 Distributional Aspects of Debt Ishac Diwan April 1991 S. King-Watson Adjustment Thierry Verdier 33730 WPS658 Fiscal Policy with Fixed Nominal Christophe Chamley April 1991 Raquel Luz Exchange Rates: C6te d'Ivoire Hafez Ghanem 34303 WPS659 Inflation and Growth in the Transition Andr6s Solimano April 1991 E. Khine from Socialism: The Case of Bulgaria 37471