Romania - Economic memorandum : external stabilization and structural adjustment (Vol. 2 of 2) : Statistical appendix

Report Nb. 5832-R0 Romania: Economic Memorandum: External Stabilization and Structural Adjustment (In Two Volumes) Volume II: Statistical Appendix May 6,1986 Europe, Middle East and North Africa Regional Office FOR OFFICIAL USE ONLY -~~- - -, . - - .. 7 ~ . . Document of the World Bank This report has-a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disdosed without World Bank authorization.  Fiscal Year = Calendar Year Weights and Measures = Metric System Currency Equivalents Domestic Currency = Leu (L) (Plural: Lei) Average Rate 2' Lei per US$1.00 1975 18.56 1980 10.90 1981 13.56 1982 14.31 Commercial Rate July 1983 3' 17.50 1984 18.33 1985 17.79 1986 15.73 Non-Commercial Rate 1975 14.38 1980 12.00 1981 12.00 1982 11.00 1983 12.50 1984 14.13 1985 12.70 1986 11.23 Rates are those in effect as of January 1 unless otherwise indicated. Nominal effective average exchange rate for trade in convertible currencies. A unified exchange rate for commodity trade was introduced on July 1, 1983. At the same date, the commercial rate was pegged to a basket of currencies.  FOR OFFICIAL USE ONLY ROMANIA ECONOMIC MEMORANDUM Volume II: STATISTICAL APPENDIX AND ANNEXES TABLE OF CONTENTS Page No. ANNEX I: Statistical and Methodological Issues ...................... 1 ANNEX II: Relationship Between Growth of Investment and Growth of Productivity .*...*.......................... 16 ANNEX III: Analysis of Trends in Romanian Exports to Major Markets .... 18 LIST OF TABLES: I. POPULATION AND EMPLOYMENT 1.1 Population, 1975-84 ................. ,........................ 27 1.2 Demographic Indicators, 1975-84 ......... ........... 28 1.3 Labor Force, 1975-84 .................. ..* .. ........... 29 1.4 Occupied Population by Age and Sex, 1975-90 . 30 1.5 Labor Force Structure by Sector, 1975, 1980-84 31 1.6 Industrial Employment by Subsector, 1975, 1980-84 .....32 1.7 Average Monthly Wage, Net of Tax, by Sector, 1975-85 ....... 33 1.8 Money Income and Expenditure of Population, 1975-85 ....... 34 II. NATIONAL ACCOUNTS 2.1 GDP by Origin, 1975-84 . .............. ... ......... ........ 35 2.2 GDP by Expenditure, 197584 - ............ . ............... . 36 2.3 Gross Fixed Capital Formation by Economic Sector, 197584 4... 37 III. BAIANCE OF PAYMETS AND TRADE 3.1 Balance of Paymets, 1975-85 - ................. ...... . 38 3.2 Balance of Payments, Convertible Currency, 197585 39 3.3 Service Receipts and Payments, 197645 40 3.4 Service Receipts and Payments, 197685, Convertible Currency .. ........ ....... o ...... ..... 41 3.5 Comodity Composition of Expo-ts, 197584 ...4 42 3.6 Couodity Composition of Impc,ts, 197584 ........ 43 3.7a) ) 3.7b) Exports by Major Trading Partners, 1975-84 ...............) 44- 3.8a) ) 47 3.8b) Imports by Major Trading Partners, 197544 ..............) 3.9 Composition of Exports to NonSocialist Countries, 1976-84... 48 3.10 Composition of Exports to Socialist Countries, 1976-84 ..... 49 3.11 Composition of Imports from NonSocialist Countries, 1976-84. 50 3.12 Composition of Imports from Socialist Countries, 197684.... 51 3.13 Composition of Imports, Convertible Currency, 198084 ...... 52 3.14 Composition of Imports, NonConvertible Currency, 1980-84 .... 53 3.15 Exchange Rates, 1975, 1980-85 ............. ................. 54 IV. EXTERNAL DEBT 4.1 External Debt and Capital Flows, 1975-84 .........6......... 55 4.2 External Debt and Capital Flows by Type of Credit, 1976-84.. 56 T dacuiAmt lime rantud distrbution an may be mu by ncWaes only in the pefomane of tk offiil duhoi Its conent my not otherwism be diuoud without Wofld ulak autborirno  - ii - Page No. V. PUBLIC FINANCE AND INVESTMENT 5.1 Revenues and Expenditure of the State Budget, 1981-85 ........ 57 5.2 Investment Financing, 1980-85 ............................... 58 5.3 Credit and Deposit Interest Rates, 1975, 1980, 1983, 1985 .... 59 5.4 Sectoral Allocation of Actual and Planned Investments, 1981-85, 1986-90 ............................................ 60 5.5 Machinery and Equipment Investment by Branches of Economy, 1970-84 ..................................................... 61 5.6 Machinery and Equipment Investment by Branches of the Manufacturing Sector, 1970-84 ................ .. ............. 62 VI. AGRICULTURE 6.1 Trends in Production of Principal Agricultural Goods, 1975-84 63 Fig. 6.1 Trends in Production of Principal Agricultural Goods, 19/5-84 64 6.2 Principal Indicators of Agricultural Mechanization, Fertilization, and Irrigation, 1975-84 ...................... 65 6.3 Livestock, 1975-84 ....................................... 66 6.4 Average Yield of Main Agricultural Products, 1975-84 ........ 67 Fig. 6.4 Trends in Average Yields of Main Agricultural Products 68-69 6.5 Exports and Imports of Agricultural Products and Foodstuffs, by volume, 1975-84 ..... ......................... 70 6.6 Exports and Imports of Main Agricultural Products, in US Dollars, 1978-84 ...................................... 71 VII. INDUSTRY AND ENERGY 7.1 Gross Industrial Production, by Subsector, 1975-84 .. ........ 72 7.2 Industrial Investment, by Subsector, 1975-84 ....... ......... 73 7.3 Changes in Production Costs by Branch of Manufacturing, 1979-83 ................ .. .................... 74 7.4 Construction Output by Type, 1975-83 ..... ................... 75 7.5 Production, Consumption, Exports and Imports of Oil and Oil Products, 1975-85 .............. .. ................... 76 7.6 Installed Capacity and Electric Power Production, 1976-90 77 7.7 Domestic Production of Energy Products, 1976-90 . . 78 7.8 Trends in Domestic Energy Consumption, 1976-84 . 79 7.9 Trends in Electric Power Consumption, 1976-84 . .80 VIII. PRICES 8.1 Comprehensive Consumer Price Index, 1981-85 ................. . 81 8.2 Retail Price Indices for Goods and Services from the Socialist Sector, 1975-85 ......................... 82 8.3 Energy Prices, 1980-85 . ........................ ............ . 83 8.4 Producer Prices in Republican Industry and Agriculture, 1980-85 ..................................................... 84 IX. PRIVATE CONSUMPTION AND SOCIAL WELFARE 9.1 Retail Sales in Socialist Trade, 1981-85 .... ................. 85 9.2 Construction of Housing, 1981-85, 1986-90 ................... 86 9.3 Meat Consumption Per Capita, 1975, 1980-84 . .87  ANNEK I STATISTICAL AND METRODOLOGICAL ISSUES This Appendix highlights several methodological and statistical problems which have arisen in preparing the report. The following presentation is not intended to assess in detail the Rowanian statistical procedures and data. -This has been done to a large extent in separate studies and further discussion of the issues would be redundant and outside the mission's terms of reference. 1' The purpose of this note is simply to highlight additional problems not discussed in the literature and specific issues affecting this report. CONTENTS A. Production 1. Comparison of Changes in Gross Industrial Production by Branches with Changes in Output of Main Products, 1980-83 2. National Income Accounts and Implicit Price Deflators 3. 1984 Production and Expenditure Accounts 4. Transport Output in 1984 B. Pricins 1. New Products 2. Moving Price Base C. Foreign Trade Statistics and Exchange Rate Movements ' For more details see M.R. Jackson: Romanian National Accounts and Estimation of its Gross Domestic Product and Growth Rates, Washington, D.C.: World Bank, forthcoming as Working Paper, 1985. T.P. Alton, et al.: The Structure of Gross National Product in Eastern Europe (Derivation of GNP Weights for 1975-1979), New York, L.W. International Finance Research, Inc., 1981.  -2- A. Production 1. Comparison of Changes in Gross Industrial Production by Branch with Chanzes in Output of Main Products, 1980-83 1. The attached tables, drawn entirely from the Statistical Yearbook of the Romanian Central Statistical Office (CSO), permit a comparison of the reported increase in gross production of the engineering and metalworking branch, accounting in 1983 for almost 30S of total industry (Table A-1), with reported changes in output of what are described in the Yearbook as the main products of this branch (Table A-2). (The period covered by the 1984 issue of the Yearbook is 1980-83.) While not fully conclusive in the absence of weights, the tables show that whereas aggregate gross production of this branch is reported to have risen at the rate of 11.5S, as indicated in Table A-1, the last two columns of Table A-2 indicate a decline over the period 1980-83 in the output of the main product groups. The only product groups for which output was reported to have unambiguously increased more strongly than the purported average were batteries, fine mechanical and optical products, metalworking machines, bicycles, and gas stoves. Most products, including those most commonly associated with Romanian engineering, showed either smaller increases (e.g., tractors) or sizable declines (e.g. railway locomotives and cars, motor vehicles, and radio and TV sets). 2. The discrepancy between the two sets of statistics is also evident in the case of otner branches of industry. Production of electrical and thermal power is reported to have increased by 5.5S in Table A-1, whereas the listing of physical output in the Statistical Yearbook suggests an increase of only 4.1X in electrical energy (measured in kWh) and a decline of 5.4S in thermal power (measured in kcal). Crude oil extraction is reported to have inreased by 10.71 in Table A-1, whereas the volume of crude oil extracted rose only from 11.5 to 11.6 million tons according to the product listing of the Yearbook. Gross production of the ferrous metals branch is reported to have increased by 6.11 in Table A-1, whereas all of the iron rose and steel products (with the exception of specialty steels) showed sizable declines in the product listing. There is evidence of similar discrepancies in the case of most other branches of industry, although not always as conclusive, given the absence of appropriate weights for individual products. 3. The Romanian authorities explained the apparent discrepancies between the two sets of statistics on the grounds that gross industrial production (Table A-1) includes the value of repairs as well as the production of spare parts and unfinished products, whereas Table A-2 includes only the main final products. However, in order to account for the magnitude of difference in trends between the two sets of data, the weight of repairs and spare parts in total production would have to be very significant, and growing very rapidly. As indicated in Table A-1, however, for the machine-building and metal- processing branch alone, repairs only constituted 2.41 of total production (8.31 of this subsector's output). 4. It is possible that the apparent discrepancies in the data can be partly explained by quality improvements and by the existence of "new  - 3 - Table A-1: GROSS INDUSTRIAL PRODUCTION BY BRANCH, 1980-83 Growth, in comparable Structure in 1983 1/ Prices (in percent of total) (in percent) Electrical and thermal power 3.3 5.5 Fuel 9.7 3.8 Coal 1.1 10.5 Coke 0.8 35.3 Petroleum 7.0 -1.4 Crude oil extraction (0.8) (10.7) Processing (6.2) (-3.1) Methane gas extraction 0.8 7.9 Ferrous metals (including mining of non-ferrous ores) 7.0 6.1 Engineering and metal-working 29.0 11.5 Mechanical engineering 18.7 11.6 Electrical engineering 4.2 0.4 Metal products 3.7 11.7 Repair works 2.4 32.9 Chemicals 10.3 11.2 Non-metallic minerals 0.4 37.3 Building materials 3.5 4.4 Forestry and wood-processing 4.0 9.7 Forestry 0.6 16.7 Wood processing 3.4 9.2 Pulp and paper 1.3 3.7 Glass and chinawear 0.8 15.7 Textiles 7.3 11.2 Clothing 3.5 14.8 Leather and footwear 2.2 18.8 Food 11.5 2.1 Other branches 2.4 ... TOTAL INDUSTRY 100.0 8.7 Memorandum items: Producer goods 74.7 8.8 Consumer goods 2.5.3 7.3 - Calculated in current prices. Source: Directia Centrala de Statistica, Anuarul Statistic, 1984.  -4- Table -2: 1fl P C UD2 AIS LF Uhit of Puu&tim TIncrw DwcLiz "bumnmt in 198 (in ) Iiustrial atesm boilrs units 387 -24.9 n I PN mu ... 20.9 Stem_ tbium Ihits 13 - - n to '00 W 257 -48.3 Wdr.uic nbz& units 54 640A. to lo w 178 -16. mutin -u '1000 ado 228 -2.3 Eltric ar1 cat u1 qpaUrn '000 lei 926 -6.1 ELitric ad wtrwAuc .utcwtim equiPt -. '0 ei 8*261 -1.8 Eltric 'aFtor 00W 606 -45.5 B wttie oo low i 1,371 25.5 Electric ca ut cAbls '000 tICs 70 -29.3 Eltric ioudio1d qap a li1ii li 2,533 2.3 dio ts '00o uit. 54&2 -37.2 Thlwisii us '000 arits 390 -27.9 Fim dmuil ad optical po&ts Mlim I 10,398 2.2 %whiz for anel *&utti and 10 f1illizg 59,997 10 Vimrns, eaziutand instaflatimu fwe poll latio, wil ilirz ad 1sploiati '00D tm 138 10.6 ~du~acgc mhim, pp ad ztallaeia f ;,duutzy 'OOD ton 602 -12.0 ZwAtoc Ibits 791 -34.0 Tft=s IMit 77,142 8.8 Tructr-drive pjua 1t 12,748 -0.5 T ratO--dLO seda Ttiu 14,028 -27.5 Cin hwwseers units 5,094 4.2 Yin line lomtives units 168 -39.1 tMin line fit cms Ibt 11,298 -6.0 .in line pauw4pr en Liits 606 -32.8 tr vddclm '0 mits - 108 -15.6 of Uti&: tam cas '000 iits 7 -3.1 em frihaern coo'000 t&, 287 -40.3 lEw ba, sel-poplled lmi 20 -25.9 Ball bearizW Hilia unit 109 6.7 Bicycl '000 wits 248 15.9 1bindld seirg amchirm '00 Oii 49 -48.4 GM stawes =00 wits 390 18.5 Liquid gs cylindes '00oo units 83 10.7 Sm: Direie Caemla de Statistica, AmIul Stistic, 1984. 2141F  - 5 - products" not captured by physical output series.-' However, these factors are unlikely to play a significant part in the case of fuels and electric pover and, perhaps, other industr-al branches producing standard and homogenous commodities which are less amenable to quality upgrading and irnovation. A more Likely explanation is the methodology of computing the gross output series reported in Table A-2 (see item B.2 below, Moving Price Series). 2. National income accounts and implicit price deflators. 5. The Government supplied the mission with a table showing GDP figures by origin in current and constant prices. The constant price series were provided in 1981 prices rather than in terms of 1977 prices which were used previously. The table shows that the level of industrial output expressed on 1981 prices was lower in 1980, 1981 and 1982 than that expressed in terms of 1977 prices. This implies that the level of industrial prices in 1981 was lower relative to 1977. According to Government's own figures, however, industrial prices increased by 12.4S in 1981 relative to 1980. Even though no comparable figures are available for 1978-1980, it is unlikely that industrial prices were actually reduced in those years by that magnitude. 6. Similarly, in 1984, aggregate consumption increased by 5.6X in 1981 prices and by 4.1% in current prices as compared to the previous year. However, these changes are inconsistent with the change in the consumer price index which declined by only 0.2X in 1984. 3. 1984 production and expenditure accounts 7. According to the national income data provided by the Romanian authorities, net exports (including statistical discrepancy) in terms of lei increased by 143.4X in 1984. The actual data are shown in Table A-3. 8. However, the above net export figures are not consistent with those in the balance of payments. Net exports of goods and non-factor services in 1984 in the balance of payments amount to US$2,378 billion (Table 3.1, Statistical Appendix). If this figure is converted into lei at the average commercial exchange rate for the year of 21.28 lei = US$1 (cf. IFS), we obtain 50.60 billion lei, i.e, 19 billion lei less than the figure from the national income accounts. 9. The Romanian authorities have provided the explanation that the balance of trade surplus achieved in 1984 was US$2.3 billion, "excluding US$900 million representing commodities ready for the export as of December 31, 1984."A' The figure of US$900 million is equivalent to the discrepancy ' See M. Jackson, 'Romania's Economy at the End of the 1970s: "Turning the Corner of Intensive Development," East European Assessment; US Congress, Joint Economic Committee, 1981, Part 1, p. 238. Government"' Economic Memorandum, June 1985, page 1.  - 6- Table A-3: USE OF NATIONAL rNCOME, 1981-84 &' (in billions of lei) 1981 1982 1983 1984 bt 1985 c/ Consumption Fund 379.8 441.7 442.2 458.6 476.0 Population 337.1 396.8 392.7 405.2 420.7 Government 42.7 44.9 49.5 53.4 55.3 Accumulation Fund 149.9 164.1 186.8 180.8 194.0 Net investment 142.1 144.2 153.3 160.1 172.5 Increase in stocks 7.8 19.9 33.5 20.7 21.5 Net export and statistical discrepancy 1.0 23.0 28.6 69.6 109.9 National income 530.7 628.8 657.6 709.0 779.9 '' National income as defined by Romanian authorities. The concept and, consequently, the figures and growth rates differ slightly from the Western convention. b' Preliminary. C' Forecast. Source: Data provided by Romanian authorities. between the net export figure in the national income accounts and that in the balance of payments accounts (US$900 million x 21.28 lei:US$1=19.2 billion lei). 10. There are two important implications of the above discrepancy. First, the figure for net exports in lei in the national income accounts would be consistent with a balance of trade of US$3.1 billion, an amount which was actually announced in the 1984 Plan Fulfillment Commun.Lque. If we take the lower trade surplus figure of US$2.3 billion and calculate the corresponding level of national income used (Romanian definition) in that year, we obtain a figure of 688.3 billion lei, or about 3X less than the figure shown in Table A-3. To put it differently, the growth of national income used in 1984 after adjustment for the statistical discrepancy would be 4.71 in comparison to 7.8S as implied in Table A-3. 11. Second, to the extent that the comodities which are "ready for exports" have been actually sold and await loading, they will '"blow up" the export figures in 1985. If they are not sold and remain in warehouses, they should show up as "increases in stocks" in 1984. If treated as stocks, the figure of 19 billion lei would increase the existing "changes in stocks" to 39.9 billion lei (=20.7+19.2), i.e., by 931. The authorities explained that the goods in question were sold and delivered in 1985, but will not be double-counted as exports in 1985.  12. The discrepancy between figures for the net exports in the balance of payments and in the national income accounts seem to be limited only to data for 1984. There are also some discrepancies in 1981 and 1982 but they are not significant. To avoid such statistical problems in the future, it is recomended that actual net exports be distinguished from "statistical discrepancy" in the national income accounts. 4. Transport sector output in 1984 13. According to Plan Fulfillment figures, niet production value in industry increased by 8.41 in 1984 in comparison to 1983 (gross industrial supply increased by 71) and net agricultural output increased by 12.12 in the same year (gross agricultural supply increased by 13.3Z). These substantial increases in growth of agricultural and industrial output are not consistent with the volume of goods transported by means of "public means of transportation" which shows 4.21 decline in the same year. The mission was told that road transport did decline by design as traffic was shifted to rail and river conveyance. However, the data in question are inclusive of all means of transport, including those provided by producing enterprises from their own stock of transport vehicles. No indications were made available about the magnitude of the enterprises' own transport and about the implication for the value of industrial output which is now presumably distorted by the inclusion of own transport. Pari passu, "transport" figures are, therefore, downward biased. B. Pricing 1. New products 14. It is well known that introduction of new products may have significant influence on changes in the rates of growth of output as well as price indices. Since new products are typically introduced mainly by the manufacturing sector, it is the growth rates of the manufacturing output which are affected most. Biases in growth indices of centrally-planned economies (CPEs) come from two sources: (i) choice of indices, and (ii) over- estimation of quality improvement in the system of price formation. While issues arising from (i) are the result of the familiar index number problem, issues under (ii) are specific for CPEs. The effect of Romanian methodology of constructing index numbers is discussed in general terms in the 1980 CEM (Report No. 2757-RO), Appendix A. 15. It is less clear, however, what is the magnitude of the biases arising from the introduction of new products in Romania. The bias will depend on the rate at which new products are introduced into production lines. The attached Table A-4 gives, therefore, an indication of the importance of "new products" in industrial output. According to production plans for 1980, new products introduced between 1975 and 1980 amounted to about 451 of gross industrial output. The corresponding percentage for the period 1970-1980 was estimated in a separate source as 70-75S of the volume of output. In other words, the product assortment was fundamentally changed in 1980 in comparison to 1970; Romania produced in 1980 only about 251 of products which it produced in 1970.  - a - Tble A-4- : M Or F nIM Or W twa i I Uwaz. wsr 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1960 Total T,wtay - (1975 - ba mow) 2 Sir of am 2CS 15.5 Z.1 27.5 23.0 45.0 bW of r!l volam of tdumlIgiD O0 27 46 W.niatzy of _ iU9- (970 - bs yer) 2 Suof muPi'O 21.6 35.3 52.5 60.9 67.2 lliatay of ahidldid d Elacticsl Oiam - (19)0 u br vow) 2 Shauofam p. 20.0 32.4 4J2*J 6.3 AD0 Skwc: IL Jadru: _i ibtinl Ac_ _s and EAtima of La Count Dometi,c Ptu& a CMW& am, t,hii-. D.C.; l1d a. 1exW- Pa, FUMmb. 16. One of the effects of the measures introduced under the umbrelLa of the stabilization program in 1981 was that resources were reallocated more towards production of import substitutes. The policy was focussed primarily on substitution of fuels imports but it also included products of the manufacturing industry. As a result of import cuts introduc.3d at the same time, enterprises were requested to replace imports of machinery and equipment from domestic sources as many import plans had to be scrapped. It is believed that these commodities would be defined (and counted) in production statistics as new products. 17. The 1986-90 Plan puts major emphasis on the introduction of new and technologically improved products. The definition of "new products" is determined by the authorities on the basis of a set of parameters of product performance and other product features. For example. reduction of production costs resulting from savings of intermediate inputs may not lead to production of a qualititatively different final product but could permit the product to be identified as new. It is questionable that such commodities should be defined as new and it is even less clear how the costs savings are related to price formation of new products. Nevertheless, it is possible that the importance of new products is overestimated, leading to increases in their prices. 18. It is not possible to ascertain precisely the magnitude of the bias without detailed assessment of the price policy towards new products and without actual price data. The price coummission and, possibly, other central organs will typically provide guidelines such as "produse de talon" (i.e. reference products) which specify standards for production costs and use values. Both ministries and centrales, and even enterprises, have considerable discretion to set and change prices of many products within such guidelines.  -9- 2. Moving price base 19. According to the Statistical Yearbook, "the dynamics of industrial production ... is calculated on the basis of prices existing in the beginning of the year and adjusted for price interventions in the year immediately preceding it."-' For example, the value of output in, say, 1984 is derived from prices of January 1984 and it is believed that these were adjusted for price changes in 1983 and should correspond, therefore, to the prices existing at the end of 1982. A similar procedure was adopted in deriving the value of output in 1983 but with an adjustment for price changes in 1982 leading to adoption of prices existing at the end of 1981 and so on. In other words, the procedure corresponds to a derivation of a "constant" price series which is calculated on the basis of (continuously) moving price base. It should be noted further that how the adjustment is actually made for "price intervention in the preceding year" is not explained in the official sources. Here and in the following example it was assumed that the adjustment for price changes is made for actual prize changes rather than planned changes and that the changes refer to those which occurred during the whole year rather than any part of it. 20. The effect of the current methodology on the value of output measured by the "constant" (moving base) prices is shown in Table A-5. The table, which is based on a simple numerical example, shows that the procedure of using "moving" base price series generates an upward bias ii output series (compare rows 4 and 7). Table A-5: MOVING BASE SERIES: A NUMERICAL EXAMPLE 1979 1980 1981 1982 1983 1. Output Volume (tons) .. 100 102 104 106 2. Price (Lei/ton) 5 5 8 10 11 3. Value (Lei)=(lx2) .. 500 816 1040 1166 4. Index of Output Volume (1980 = 100) .. 100 102 104 106 5. Price Index (1980 = 100) .. 100 160 200 220 6. Index of Value of Output (=3)(1980 = 100) .. 100 163 208 233 7. Index of Output at "Constant" (Moving Base) Prices (1980 - 100) .. 100 160 204 229 ' Statistical Yearbook 1984, p. 67.  - 10 - C. Foreign trade statistics and the exchange rates movements 21. The analysis of Romania's foreign trade is subject to difficulties arising from peculiarities in the organization of foreign trade (and its implications on domestic price formation) and exchange rate policy; the determination of a particular level of the exchange rate and its changes are the results of administrative decisions and the exchange rate does not typically reflect the real scarcity value of foreign currencies. The recent revaluation of the leu (plural:lei) seems to complicate further the interpretation of Romanian foreign trade statistics. 22. The effect of the exchange rate on foreign trade statistics can be seen from Table A-6, which shows the total value of imports from socialist countries in lei terms and in terms of US dollars and the respective shares of socialist countries in total imports. The lei value of imports in 1980 refer to "valuta lei" and lei values of imports after 1980 refer to domestic lei."" The transactions are, therefore, affected by three changes in the exchange rate: by the introduction of a commercial exchange rate in 1981, devaluation in 1982 and revaluation in 1984. 23. As the table shows, the reported trade flows depend considerably on the units of measurement (i.e., domestic or valuta lei or US dollars). Both the growth rates and the geographical composition of imports vary considerably depending on whether they are based on lei or US dollars. The share of socialist countries in total imports, which was approximately the same in 1980 when measured in valuta lei in comparison to the US dollar share, increased considerably in 1981 and has remained consistently above the share measured in terms of dome3tic lei. 24. The effect of changes in the exchange rate does not appear to have been neutral with regard to the share of socialist countries in total imports. The introduction of a commercial exchange rate in 1981 is associated with a substantial increase in the share calculated in dollar terms, as indicated above. However, the share calculated in domestic lei remained constant. The revaluation in 1984 was accompanied by an increase in the dollar-based share of socialist countries and a substantially reduced share based on domestic lei. Only the devaluation of 1982 appears te have a uniform effect on both shares, slnce they show a similar increase. 25. At least part of the explanations for these dichotomies lies in the arbitrary changes in cross exchange rates between the US dollar,"' which is used in convertible currency transactions, and transferable rouble (TR), which is used in non-convertible currency transactions. The effect of revaluation of the leu in November 1984 on the dollar-rouble cross exchange rate is shown in the Table A-7. ~-L A "leu valuta" expressed the formal gold price of the leu, termed the "official rate"; it was used only to express the value of foreign trade in government statistics. The "domestic leu" is converted at the conmercial exchange rate and is used in actual transactions as well to express the value of foreign trade in official statistics (starting in 1981).  - 11 - Table A-6: ROMANIAN IMPORTS: TOTAL AND FROM SOCIALIST COUNTRIES IN US DOLLARS AND LEI 1980 1981 1982 1983 1984 1. Imports from socialist countries 4851 5555 5141 5422 6051 (mln. US dollars) 2. Imports from socialist countries 22315 61154 54430 65289 73190 (mln. lei) 3. Total imports 12685 12264 9745 9643 10344 (mln. US dollars) 4. Total imports 59006 164671 124851 130369 160816 (mln. lei) 5. Share of socialist countries in total imports 38.2 45.3 52.8 56.2 58.5 [Based on US dollar data, 5 = 1 as S 31 6. Share of socialist countries in total imports 37.8 37.1 43.6 50.1 45.5 [Based on lei data, 6 = 2 as X 4] 7. Ratio of rows 2 and 1 4.60 11.01 10.59 12.04 13.00 8. Ratio of rows 5 and 6 1.01 1.22 1.21 1.12 1.29 Source: Data in Statistical Appendix. Revaluation of the implicit cross exchange rate between the transferable rouble and US dollar 26. The effect of the November 1984 revaluation of the implicit cross exchange rate between the transferrable rouble and US dollar can be analyzed with the help of the following example. Let Us assume that exports to CMEA area, to OECD area and, consequently, total exports are fixed at given levels and the revaluation has no effect on the value of these exports. The revaluation of the leu is also assumed to generate a revaluation of 8.8S of the TR vis-a-vis US dollar. The impact of the changes in exchange rate are summarized in the following tables. Both tables are essentially identical except that Table A-8 shows data expressed in lei terms while Table A-9 shows the data in US dollar terms.  - 12 - Table A-7: ROMANIAN REVALUATION 1984 1. Via-i-Vis US dollar : 19.4S [17.50 as X of 21.50 7 2. Vis-i-Vis Transferrable Rouble (TR) : 12.62 [15.50 as Z or 17.50 1 3. Transferrable Rouble - US dollar exchange rate: a. Before revaluation: 1 TR = US$0.8140 b. After revaluation: 1 TR = US$0.8857 4. Implicit Revaluation of Transferrable Rouble: Vis-i-Vis US dollar : 8.82 27. One can draw the folloving conclusions from the tables. Assuming constant value of exports to CMEA in terms of transferrable roubles and constant value of exports to OECD countries expressed in terms of US dollars: i) the value of exports to CMEA area and the share of CMEA increase as a result of the revaluation; (2) the share increases by the same amount irrespective whether it is calculated in lei or US dollar terms. This must clearly follow from the application of the same implicit cross exchange rate between TR and US dollars in both cases; (3) the effect of devaluation of the implicit cross exchange rate between TR and US dollar will be exactly the opposite to the effect of revaluation. Statistical Anomalies 28. The dollar share of, say, socialist countries in total Romanian imports will be the same as the corresponding share expressed in lei provided that the dollar/leu rate implicit in data for trade with socialist countries is the same as the dollar/leu rate in trade with non-socialist countries. Since trade with socialist countries is conducted in transferable roubles, their share will depend on cross exchange rates between the US dollar and TR. As the figures in Tables A-6 and A-10 reveal, however, the implicit dollar/leu rates were different in trade with socialist countries in comparison to those implicit in trade with non-socialist countries, with the exception of 1980 (rows 7). This means, therefore, that import transactions with socialist and non-socialist countries expressed in terms of lei were calculated on the basis of TR/leu and dollar/leu exchange rates which implied a cross dollar/TR rate -' An indication of the arbitrariness of these changes is the transferrable rouble-US dollar rate as quoted by the Czechoslovak National Bank in 1983: 1 Rb US$1.2638, which compares to 1 Rb - US$ = 0.8857 quoted by Romanians. The latter figure refers to 1984 but is comparable to the Czech figure for 1983.  - 13 - Table A-8: STATISTICAL EFFECT OF NOVEMBER 1984 REVALUATION OF RONANIAN LEI: EFFECT ON LEI VALUES OF EXPORTS A CMEA OECD Total 1 2 3=1+2 1. Assumptions: Exports 100 TR US$ 100 II. Export Values: [leie A. Before revaluation 1750 2150 3900 B. After revaluation 1550 1750 3300 III. Export Shares: A. Before revaluation 44.9 55.1 100.0 B. After revaluation 47.0 53.0 100.0 ' Implicit cross exchange rate: I TR = US$0.8140 (before revaluation) 1 TR = US$0.8857 (after revaluation) Table A-9: STATISTICAL EFFECT OF NOVEMBER 1984 REVALUATION OF ROMANIAN LEI: EFFECT ON US DOLLAR VALUES OF EXPORTS i CMEA OECD Total 1 2 3=1+2 I. Assumptions: Exports 100 TR $100.00 II. Export Values (US$) A. Before revaluation: (1 TR:US$0.8140) lOOxO.8140=81.40 100.00 181.40 i .fter revaluation: (1TR:US$0.8857) 100x0.8857=88.57 100.0 188.57 III. Export (in percent) A. Before revaluation 44.9 55.1 100.0 B. After revaluation 47.0 53.0 100.0 J' Implicit commercial cross exchange rates: 1 TR = US$0.8140 (before revaluation) 1 TR = US$0.8857 (after revaluation)  - 14 - different from the cross exchange rate between dollar and TR which was actually used in the dollar computations, except in 1980, i.e., in deriving the dollar values of imports. 29. Moreover, the changes in import shares of socialist countries in dollar and lei terms will be the same, provided the change in cross exchange rate which is implicit in the transactions and the one actually used in the calculations is the same. To put it differently, the changes in lei and dollar shares of imports of socialist countries will be the same. provided the changes in cross exchange rate which is used in the conversion of imports from transferable roubles into dollars corresponds to revaluation/devaluation of leu vis-a-vis dollar and TR. This was also clearly not the case during the period of 1980-84. 30. An interpretation. It is not entirely clear which cross exchange rate was actually used in the calculations. It appears, however, that in 1980 both the trade transacticns and the cross exchange rates were based on the same, and most probably, valuta, exchange rate when the dollar/rouble rate stood at US$l.49/lTR. This practice was discontinued in 1981 when rouble and dollar transactions were recorded at new commercial rates but the cross exchange rate between dollar and rouble actually used in data conversions was the same as in 1980 when the rate was derived from the valuta exchange rate. Thus, the implicit cross exchange rate between the IR and US dollar changed as a result of devaluation in 1983 and revaluation in 1984, but the cross exchange rate used in the conversion of trade from TR into US dollars did not. Instead, a constant TR-US dollar valuta rate was used. 31. The effect of this methodology on the trade structure of Romania can be sumarized as follows. The dollar share of socialist countries in Romanian imports increased relatively faster than the corresponding lei share as a result of "revaluation" of the TR vis-a-vis the US dollar.,' "Devaluation" of the TR vis-a-vis the US dollar in 1983 increased the dollar share of socialist countries relatively less than the share in terms of lei. Moreover, the shares of socialist countries expressed in terms of US dollars have been generally higher in com_rison to those expressed in lei since 1981. The term "revaluation" ("devaluation") of transferable rouble vis-a-vis US dollar refers here to the change in the implicit cross exchange rate resulting from changes in the commercial exchange rates of lei vis-a-vis US dollar and transferable rouble.  - 15 - Table A-10: ROMMNIAN IMPORTS: TOTAL AND FROH NON-SOCIALIST COUNTRIES IN US DOLLARS AND LEI 1980 1981 1982 1983 1984 1. Imports from non-socialist 7834 6709 4604 4221 4283 countries (mln. US dollars) 2. Imports from non-socialist 36691 103517 70421 65080 87626 countries (mln. lei) 3. Total imports 12685 12264 9745 9643 10344 (mln. US dollars) 4. Total imports 59006 164671 124851 130369 160816 (mln. lei) 5. Share of non-socialist countries in total imports 61.8 54.7 47.2 43.8 41.4 [Based on US dollar data, 5 - 1 as S 31 6. Share of non-socialist countries in total imports 62.2 62.9 56.4 49.9 54.5 [Based on lei data, 6 = 2 as S 41 7. Ratio of rovw 2 and 1 4.68 15.43 15.30 15.42 18.58 8. Ratio of rows 5 and 6 0.99 0.87 0.84 0.88 0.76  - 16 - AN= Ir RELATIONSHIP BETWEEN GROWTH OF INVESTMENT AND GROWTH OF PRODUCTrVITY IN ROMANIA 1. The table in this Annex shows, in a summarized form, the results of the regression analysis of investment as a source of productivity growth. According to the null hypothesis, which was developed in the text, the growth of productivity in individual industrial sectors is dependent on investment (I) on year t (or t-1, that is in the year preceding t, or t-2). The numbers in columns identified as It, It-, and I-,_ are the estimated b-coefficients of the regression equation OILma+bI+u, where O/L is the output per worker, I is the fixed investment put into operation and u is the random term. The b-coefficient shows the strength of the relationship between investment and productivity. The higher the numbers are, the stronger is the relationship, i.e. the greater productivity growth was generated from one unit of investment growth. 2. Each sector's growth of productivity is related to investment in any one year, say t, while possible effects of fixed investment put into operation in other years, such as t-l, were not considered. As shown in the table, most results are reported in column It-2, vhich suggests that each year's growth of productivity vas related in most sectors to fixed investment put into operation 2 years earlier. Experiments for alternative years have been also carried out but they turned out to be statistically insignificant. The figures in the last three columns give the results of standard statistical tests of the regression. 3. The form of the equation is based on vhat the mission believes to be a strong argument of economic theory. The relationship between investment and productivity in market economies is probably subject to a great deal of simultaneity. In other words, investments (together with other factors) are believed to determine the growth of productivity while there may also be a strong effect of growth of productivity on growth of investments. Under conditions of central planning, however, the flow of causality from productivity to investment is likely to break down since investments are determined administratively and are subject to various constraints such as social objectives and price distortions. On the other hand, there are strong reasons to believe that the incentive schemes in centrally-planned economies induce enterprise managers to demand new investment as a "precondition" for growth of output and productivity, which contributes to what is sometimes called "investment hunger".  - 17 - Am H10 inUS OF U INU Mm ANLU Itm (I. - !t-) tt-1 't-2 a r D41 Actricity 0.035 - - - O.m2 o.02S 0.363 t64.221) MIS 0.41 - O _ O.1 0.511 OJi3 (0.057) uous Iha11 - - - OA55 O.86 47.429 0.05 (0.066) a umu 1ItaUllr - - - 0.569 0.698 23.115 0.938 (0.116) hdiiin ldig - - - 0.563 0913 1..306 1.038 (.053) dhu.ca1u o- - 0