Policy, Research, and ExternE. Affairs WORKING PtNPERS Agricultural Pollcier j Agriculture and Rural Development Department The World Bank April 1991 WPS 650 Administrative Valuation of Soviet Agricultural Land Results Using Lithuanian Production Data Karen Brooks New land tenure arrangements in the Soviet Union require those who farm the land to pay for its use. But Soviet land markets are constrained or ineffective, and land use fees must be set admin- istratively. The Policy,Research, and External Affairs Complex distributes PREWorking Papers todisseminate the findings of work in progress and to enoourage the exchange of ideas among Bank staff and all others interested in development issues. These papers carry the names of the authors, reflect only their views, and should be used and cited accordingly. The findings, interpretations, and conclusions are the authors' own. They should not be attributed to the World Bank, its Board of Diretors, its nanagement, or any of its member countries. Policy, Ftesearch, and External Affairs Agricultural Pollcles WPS 650 This paper - a product of the Agricultural Policies Division, Agricultural and Rural Development Department - is part of a larger effort in PRE to analyze the agricultural transition in former centrally- planned economies. Copies are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Cicely Spooner, room N8-039, extension 30464 (26 pages, with tables). New land tenure arrangements in the USSR from production data and can serve as an appro- require that agricultural producers pay for land priate initial value for users' fees. use. The current distorted pricing system and the absence of functioning land markets complicate Brooks estimates marginal value products land valuation, and slow the adontion of new for land for 1,032 collective and state fanns in property relations. Lithuania using farm-level data for 1986 and 1987 and compares the marginal value producLs In a market economy that functions well, derived from actual received producer prices agricultural land would earn its approximate with those derived from border prices with marginal value product in agricultural produc- altemative assumed exchange rates for the ruble. tion. This value can be measured empirically The PRE Working Paper Series disseminates the findings of work under way in the Bank's Policy, Research, and External AffairsComplex. An objective of the series isto get these findings out quickly, even if presentations are less than fully polished. The findings, interpretations, and conclusions in these papers do not necessarily represent official Bank policy. Produced by the PRE Dissemination Center Administrative Valuation of Soviet Agricultural Land: Results Using Lithuanian Production Data by Karen Brooks* Table of Contents Price Reform and Land Valuation 5 Lithuanian Producer Prices 5 Output Prices, Land Quality, and Profitability 7 Marginal Value Products of Land with Current and Reformed Prices 9 Border Prices at Which Exchange Rate? 12 Price Reform and Demand Side Constraints 14 Conclusions 16 Tables 17 References 25 Endnotes 26 * Agriculture and Rural Development Department, the World Bank and Department of Agricultural and Applied Economics, University of Minnesota, St. Paul, MN 55108. This work was supported by the exchange agreement between the Center for Agricultural and Rural Development, Iowa State University, the Lithuanian Scientific Research Institute for Agricultural Economics in Vilnius, and the National Science Foundation. Willis Peterson and Benjamin Senauer provided comments on an earlier draft. I particularly thank Robert Dorfman for his useful comments. ADMINISTRATIVE VALUATION OF SOVIET AGRICULTURAL LOD Results Using Lthuanian Production Dat Ever... throughout East and Central Europe and the Soviet Union since January, 1990 have clarifled the common tasks that these countrles face In their efforts to replace defunct institutlons of central planning with a more vital market mechanism (see Brooks, et.al., 1991). Each country has adopted a unique mix of policies to move forward on this broadly common agenda. In fall of 1990 It became apparent that the Soviet Unlon was diverging, perhaps temporarily, from the directions of her East and Central European neighbors. The Gorbachev government at the Union level In October, 1990 promulgated a set of guidelines that Imply continued administrative control of many aspects of the economy, including pricing and procurement, and continued state ownership of most productive assets. The guldelines seek to capture many gains of a market economy without allowing markets to function. They thus pose the need for Imaginative administrative behavior that mimics the market mechanism. The effort to ellcit such behavior through the piece-meal reforms of the past has not been successful. The hiatus in the Soviet transitlon, whether temporary or permanent, has profound Implications for agriculture and particularly for land tenure. The differences In land tenure In the Soviet Unlon and East/Central European countries throughout the post war period were evident, but were masked by managerial practices that muted the importance of formal property rights In land. The formalities remained, however, and throughout East and Central Eu-ope the state owned little agricultural land. Ev'3n in countries that were fully collectivized, Individuals retained formal title to much land, and where they had surrendered title, It was usually to the collective, rather than to the state. Throughout Eastern and Central Europe debates about land ownership over the past year have centered on the Issue of which competing private property claims hold priority and how they should be recognized, rather than the more elementary Issue of -2- the legItImac- of privnte ownership o! lard. Incomplete nationalizatlon of land In Eastern and Central Europe appears markedly to have altered the political economy of agricultural reform, and weakened the constituency of support for the Inherited system. In the USSR, private ownership of land Is at this writing not yet legal at the Unlon level, despite recent legislation that legalized private ownership of other assets. Republics are proceeding with their own land laws that recognize different forms of Individual tenure with varying restrictions on transferability. Although It Is difficult to predict the future course of policies, It appears le3s liKely that fully functioning land markets will develop In the USSR In the near term than In Eastern Europe. The state will retain ownership of most If not all agricultural land, but will allow a more diverse set of users to have contractual access to the land. The parties Involved In contractual agreements will be the state and collective farms, local councils, Individuals, and small cooperatives. Under the new tenurlal arrangements, those who farm land are to pay for the use of It. With land markets constrained or Inactive, land use fees must be set administratively. There Is no consensus In the Soviet economics literature yet on how best to value land administratively. The Union level land law of March, 1990 mandates a full scale land cadastre to record characteristics of land. Reliance on cadastral results would greatly postpone the Introduction of land use fees. Furthermore, there Is no clear methodology for transforming cadastral surveys Into land values. In a well functioning competitive market economy, agricultural land would earn Its marginal value product In agricultural production. This value can be measured empirically from production data and actual or proposed prices. The marginal value product will depend on the structure of output prices, the productlon technology, and managerial efficiency at the farm level. Valuation of land Is Intimately llnked to the -3- structure of output prices, and the impact of price llberalizatIon or administratl- pricek revislon on optimal users fees can be shown. This approach to land valuation Is approximate at best, and would not be necessary If Iand markets were permitted to functlon. It has the advantage of simplicity and speed, and allows tenure reform to proceed before the lengthy cadastre Is completed. It furthermore provides a benchmark with which to evaluate the claims of farm chalrmen that user fees (which they pay) should be very low but fees for Intra-fam leaseholds (whlch they collect) quite high. The calculatlon of marginal value products for land of differing quality Is a simple technical exercise that embodles no assumptlons about allocat?ve efficlency, market clearing, or prices that reflect scarcity values. To step further and argue that marginal value products are optimal user fees does require strong assumptions, In particular that other factors of production are priced according to their scarcity values, and that farm managers are ailocatively efficient. When these conditions do not hold, the administrator given the unenviable task of stimulating market behavior without a market has several choices. He or she can adjust user fees to reflect distortions elsewhere, or can attempt through administrative means to remove these distortions. The measurement of marginal value products, and particularly demonstration of their dependence on the structure of output prices, simply expands the set of empirical Information that can be used for administrative decision making. The Introductbon of a genulne market mechanism would greatly reduce the need for continued administratlon of the economy. Under several plausible paths of transition to a functioning market mechanism, however, initlal administrative valuation of agricultural land Is required. An empirical understanding of the marginal value products of land could therefore generate Informatlon useful for a genuine transition. In the following discussion I use farm level data from Lithuanian state and collective farms to esthnate the marginal value product of land under the 1986 and - 4 - 1987 output price structure, and what It would be If producers had recelved world trading prices but used unchanged technology and did not adjust to the dlfferent relathve prlf 9. The particular parameter estimates In this work do not generalize to the Sovlet Union as a whole. Producer prices for grains In 1986 and 1987 were higher In Lithuania than In the graln belt of the USSR, and prices for livestock products, particularly milk, were lower than In more easterly regions. Because the USSR Is so large and Its separate regions have not been connected through rationally Interlinked prices, It Is Inappropriate to genera!tir about the effects of reform on the USSR as a whole from analysis of a partIcular region. Conclusions based on analysls of average prices throughout the USSR do not capture the potential for better utilization of regional comparative advantage and Interregional trade within the c,)untry, and this woud be a major payoff of the reform. The methodological approach of this paper; Le.; estimation of marginal value products of land under alternative output price structures, could be used throughout the USSR as the methodology of valuation If land markets continue to be repressed. The following discussion Is based on data from the 1986 and 1987 annual accounts (godovye otchety) of 1032 state and collective farms In Lithuanla. The LlthuanIan data are additionally valuable because they Include measures of land quality rarely available with production data. Lithuanian researchers have made detailed analysis of different soil types on each of the state and collective farms. Each farm Is assigned a soil quallty Indicator (bonitet) based on the yield of a standard crop (standardized according to feed units) on comparable soil under conditlons of average management. The variable measuring land quality (bonitet) ranges from 27 to 66 with a mean value of 41. Price Reorm and Land Valuation The price structure affects the returns to operators working under alternative tenurlal arrangements. Valuation under the current pricing system, wlth a multpilcity of pricew for the same product, Is difficult and ybilds distor*ed asset values and poor resource use. The centrality of land valuation to the pr cess of tenurial reform makes It very difficult to pronote now property relations wlthout price reform. Tho price reform would Ideally constitute a full liberalization of prices, but would as a minimum correct distortions In relative prices and remove farm specific deviations from provaillng prices. Land has historically been offered to farms In the Sovlet Union wlthout charge. Impilclt land rents have been collectod through differentlated, farm s'eclfIc output prices. One obJective of the economic reform is to make the domestic price system more consistent with world prices. Ti9 can be attempted by setting output prics equal to border prices (adJusted for transportation costs), and deriving land use fees based on the marginal value product of land with existing factor productivity and technology. Translation of border prices Into domestic prices requires choice of an exchange rate, and no readlly defensible candidate Is avaQ Domestic relative prices can be allgned with border prices, however, and the price level made conditional upon the ehosen exchange rate. Lithuanian Producer Prices Prices recelved by Lithuanlan producers for sales to the state are shown In Table 1 and displayed to show the dispersion among farms. The prices are unit values, and Include all premia for quality, quantity, and farm specific differentials. For example, the one quarter of milk producers who received the lowest prices In 1987 were paid less than 335 rubles per ton of fluld milk, while the quarter of producers -6- who received the hlghest prices were pald between 352 and 410 rubles per ton. The base price for milk In Lithuanla at the tine was 310 rubles p'.,r ton. Comparison of Lithuanlan prices with those In Latvia, Estonla, and Belorussba Indicates some significant differences. Belorusslan prices for milk and meat are slgnificantly hlgher, as shown In Table 2. The price dispersion across repubiics within this relatively small region Illustrates the adjustments that reform would bring. These adjustments would facilltate better use of resources specific to each locality, and expanded gains from Interregional trade within the country. Procurement organizations have traditionally pald dlfferent prices to different farms for the same product. Price differentlation enters through zonal pricing and through the bonus system. Several bonuses can raise received prices above base prices. The most Important are quality differentlals, bonuses for sales In excess of a moving average of past years, and premia for farms In financial stress.1 The Lithuanian data show price differentiation, but the degree appears to be less than reported In other parts of the Soviet Union. In 1987 15 percent of Lithuanlan state and collective farms had profitability (rentabel'nost) of 10 percent or less (which would qualify them for bonuses of up to 75 percent of base prices In many parts of the country). The data do not show that these farms received high bonus prices for meat or milk, suggesting that special bonuses for financlally weak farms were not widely used In Lithuania. A strong negative correlation between meat prices and land quality In 1987 Is shown In Table 3, Indicating that farms on poor quality land were paid higher prices for livestock products. The 1986 data also show a strong negative correlation between meat and milk prices and land quality. Dlfferentiated prices for livestock products appear to be the mechanism through which Implicit land rents were ecollected In this region. Borc4ar prices convorted Into rubles at exchange rates of two and four rubles to the dollar are also shown In Table I for reference and comparison. The border prices for commercially traded products are prices c.l.f. northern European ports &djusted to farm gate product definitions.2 For example, the beef price Is beef c.l.f. northern European ports adjusted to live welght using the standard Soviet coefficlent of .55 live welght to carcass weight. The boruer price for mlik Is equlvalent to $10 per hundredwelght converted to rubles at the stated exchange rate.3 For little traded products, such as potatoes, the listed border price Is the marginal price of the third quarter of domestic producers; 75 percent of Llthuanlan producers recelved a price at or below the llsted price, In choosing this price for nontraded products, I assumed that the price (for average quality) after the reform would be based not on the hlghest current price, but on one that was In the high end of the current range. Relative prices In Llthuanla In 1987 differed from border prices; grains and mli were relatively low and meat high. Virtually all dalry producers received less than border prices even at two rubles to the dollar and the difference Is greater at four. Meat prices were not as low relative to world values as were milk prices. Output Prices, Land Quality, and Profitabillty Differentiated prices and meticulous measurement of land quality were Intended to capture rents assoclated with preferential access to land at no cost (Poshkus, 1P79). If the system worked weii, farF2a would show a range of profitabillty, but there would be no systematic link between land quality and profitabillty, unless some other factor contributing to profitability were strongly correlated with land quality. Table 3 shows palr-wise correlation coefficlents for prices, land quality, and profitability In 1987. For example, farms that recelved high prices for rye tended also to receive hlgh prices for barley, as Indicated by the positive, significant correlation coefficient of .28. Individual product prices tended not to be strongly correlated wlth lam profitability, except for poultry. Land quality, on the other lhand, was positively and slgnificantly correlated with profitability; the correlation coefficlent Is .36 between land quality (bonltet) and profitability (rentabelrnost'). Farms In the lowest quarter according to profitability had an average Indicator of land quality of 38.50, compared to 44.52 for farms In the hlghest quarter of profitability. Even though there Is a significant correlation between land quality and measured profitability (suggesting that not all implicit land rents are taxed away through prices), not all farms with hlgh quallty land are highly profit-,le. The range of land quallty within profitability groupings Is greater than the difference In mean quality across groupings. Regression analysis Indicates that varlation In land quality Is a slgnificant contributor to variation In net output when output is aggregated with actual unit values received by each farm. Regression coefficlents for log llnear production functions are shown In Table 4.4 The dependent variable In Table 4 Is farm level net output aggregated at prices actually received. Net output excludes all product used for feed, whether produced on the farm or purchased. Output prices vary among farms, and hence this dependent variable Is flawed as a measure of farm level output or productivity. It does allow, however, a view of the contributlon of factors of production to farm revenues under the prevailing distorted pricing structure. The Independent variables In this rogression are iand quallty (bonitet), land quantity (one unit of cropland Is assumed equivalent to two units of hay land), labor In reported hours worked, ruble expenditures on fertilizer, machinery In reported horsepower units, and an aggregate of animals with swine, sheep, and goats welghted at .3 of a cow. The table shows separate e_. :nates for the full samples of over one thousand farms In 1986 and 1987, and sub-samples In 1987 In profitability groupings. -9- Results for another set of regresslons are displayed In Table S. The dependent varlable In Table 5 Is net output aggregated at border prices Invarlant across farms and converted Into rubles at two rubles to the dollar. The Independent variables In Table 5 are the same as In Table 4. The difference In estimated coefficlents In Tables 4 and 6 thus shows the effect of the domestic price structure on returns to factors of productlon compared to returns using world prices (but keeping the same technology and physical productivity.) If price differentlation (In the domestic price structure) had succeeded In fully capturing rents associated with superior land quallty, the regresslon coefficient associated with land quality (bonitet) In Table 4 would presumably be small. It Is In fact rather large (.53 for 1986) and precisely measured (t-12.74). When net output Is aggregated with border prices Invariant across farms (Table 5), the contributlon of land quality to variatlon In net output Is even greater (coefficlent - .81, t-20.66). This suggests that the current prices differentiated a% the farm level capture some of the rent associated with hlgh quality land, but not all. Descriptive data characterizing the farms In the 1987 samples are displayed In Table 6. Table 6 Indicates that farms with different leveos of profitability show little difference In farm size (mean size Is about 2200 hectares of cropland and welghted hayland In each category). Jore profitable farms pay silghtly higher wages but emrioy the same number of hours, spend a little more on fertilizer, command more machinery power, and have a few more animals. The fams of different levels of profitability are remarkable for their similarity In average command of factors of production. They differ more In net output than In Inputs, and the difference Is more pronounced under the domestic price structure than under world prices. Marginal Value Products of Land With Current and Reformed Prices Marginal value products for all factors of production calculated from sample means of the 1987 regressions with actual prices are displayed In Table 7. Marginal - 10 - value products for land quantity and quality evaluated at sample means for all regressions are shown In Table 8. The estimated margInal value products (Table 8) range from approxbnately 40 rubles per hectare of average quality land to approximately 150, depending on year, olvislon of the samj; -, and prices used to aggregate net output. The full sample of 1032 farms In 1987 shows a low value for a hectare of average quality land (40 rubles). When the sample Is broken Into profitability groups, the dispersion of quallty withIn the groups Is reduced somewhiat, since quallty and profitability are positively correlated. In the profitabilIty based subsamples, the value of a hectare of average quality land (average for the group) Is greater and the value of a unitary deviation of quality from average Is reduced. The substitution of uniform border prices (at two rubles to the dollar) for actual received prices raises both the value of a hectare of average quality land and the value of deviations from average quality, since the contribution of quality Is not taxed away through the price system. This Is apparent In both 1986 and 1987. The estimated values for land quantity appear reasonable, although they span a wide range. One could conclude conservatively that a farm manager able to sell output at border prices (at two rub!as to the dollar) would be willing to pay approximately one hur ired rubles for use of a hectare of average quallty cropland, If payments to other factors of production do not exceed their marginal value products. Introducing payments for land while retaining the current price structure would be more problematical. The regressions Indicate that the user fee for land would be lower, perhaps by as much as a half, and the gradation according to quality would also be less than If output prices were revised. Introducing uniform user fees (differentiated by quality) wlthout output price revision would penalize livestock - 1 1 - producers on hlgh quality land. User fees without price reform or revislon could thus bring a negative Instead of positive supply response. Is 100 rubles per hectare a hlgh or low rental rate? Milnnesota's mixed dairy, feed, wheat, hogs, and sugar beet economy Is superficially similar to Lithuanla's. Cash rents In Minnesota have been declining since 1983 and were In 1989 $S0 per acre (240 rubles per hectare at 2 rubles to the dollar) for average quality land. Now can the measurement of the contribution of land quality best be used? This Is a nontraditional Input, measured on a per hectare basis. The attempt to value It directly as If It were a traditional varlable Input yields unrealistically high values for a unit of bonitet (e.g., see Table 7). If used for policy purposes this would Imply that land of slightly less than average quality should be offered without fee, and land of slightly higher than average quality should command a much higher fee. An alternative to dir"-ct valuation of a unit of land quality Is to exploit the measured contribution of bonitet to the marginal value product of a hectare of land.1 A farm that was average In other iespects but had land quality of 30 units Instead of the mean value of 41 earned a marginal value product (with actual 1987 prices) of 37 rubles per hectare, rather than 40. An average farm with bonitet equal to 50 earned a marginal value product of land of 47 rubles per hectare. These values understate actual contribution of land quality to physical productivity, since price distortions mute the Impact of land quality on farm revenue. The measured coefficlents of land quality when border prices, rather than domestic prices, are used are slgniflcantly higher, and knply a larger deviation of land's marginal value product associated with lower and higher quality land.5 I This strehes th teal tW preton of the elasty of fm outPut or rvenue with respect to land quawly, since we are evaluatg Xt contbution of land quality at a valwu ar from its mean. It offems appa reasonab vWaluafton of hcares o land of dfflerent quafi, however, and Is thus atmract for pratil purposes. I tank Robet Doulman for suggest ih uts measured conbution o bof L - 12 - The analysis further suggests that a share contract In Lithuania In which all Inputs bosides land are separately priced should give the parent fam about 10 percent of output, plus perhaps a risk premium.6 The share of revenue pald for land should not exceed the share of land's contribution to net output If land Is to be pald Its marginal value product, and costs of other purchased Inputs are not shared. The regression coefficlents for land quantity using actual prices recelved (Table 4) range from .07 to 11 In the subsamples, and are .045 and .13 In the full samples for 1987 and 1986, respectively. From these estimated elasticitles, It can be concluded that the share of land's contribution to variatlon In net output Is approxlmately 10 percent, and not greater than 15 percent. Land users asked to pay a higher share than this for use of land will adopt uneconcmically land-saving technology that will Increase society's costs of producing a given level of output. The fragmentary evidence on lease contracting reported from various regions of the USSR suggests that collective and state farm managers often charge share leaseholders a higher share than can reasonably be expected to correspond to land's contribution to net output (Brooks, 1990). Border Prices at Which Exchange Rate? Border prices at two rubles to the dollar represent relative prices slgnificantly dlfferent from producer prices In Llthuanla In 1W87. Border prices for milk and grain were higher and meat lower. Border prices converted at two rubles to the dollar, however, had a surprisingly small combined Impact on mean net farm output. Since most farms In Lithuania are diversified producers of milk, meat, and cash crops, the price changes Introduced with border prices (at two rubles to the dollar) were offsetting In their Implied Impact on aggregate farm Incomes. This Is shown In Table 6; there Is virtually no difference In mean net output calculated with actual prices and border prices converted at two rubles to the dollar. - 13 - The adoption of border prices would brlng galns and losses at the farm level even though the Impact on mean net output Is small. Nlnety percent of producers In 1987 would have had net farm output at border prices (at two rubles to the dollar) within a range of 83 percent to 116 percent of received net farm output (actual prices). The ninety percent Interval for 1986 Is bounded by .81 and 1.19. Thus the Introductlon of prices based on border prices converted at two rubles to the dollar would brlng short run gains and osses between 10 percent and 20 percent of farm Income for ten percent of farms. Long run changes would depend on the supply response to the new relative price structure. The finding that net farm output In Lithuanla would change llttle If border prices (at two rubles to the dollar) were substituted for current prices Is new and bnportant. Its Importance Is enhanced by the absence of any rlgorously Justifiable estImates of an equIlibrlum exchange rate for the ruble. Wlthout knowledge of an equlilbrIum exchange rate, and In the absence of a general price liberalIzation, those deslgning administrative adjustments of agricultural producer prices must use their Judgement about what exchange rates are feasible In a partial, sector specific price revision, considering the Impact on farm Incomes and consumer budgets. The LIthuanlan data Indlcate that an Implied exchange rate of fewer than two rubles to the dollar would bring a hlghly undesirable supply shock associated wlth the new price structure. In some parts of the Soviet Union where producer costs and prices are higher than In Lithuanla, a supply shock may be a necessary part of the adjustment process to remove marginal producers. There Is little Indication of a need for exit of many marginal producers In Lithuania, If they can operate with the current relatively low producer prices. Although 1987 levels of farm Income were roughly consistent with an exchange rate of two rubles to the dollar, and producers realized Implicit rents under the current system, It would be difficult to Introduce land payments wlthout raising the - 14 - general level of producer prices and Incomes. Farms retained impllilt land rents and reallocated them to cover other expenses and Investment. If required to pay expllcit rents wlthout augmented Incomes, many farms would have been pushed further Into financial difficulty or bankruptcy. Almost half of the fams In Lithuanla In 1987 fell In the range of profitability considered to Indlcate questlonable long-term financil viabIllty. There Is little economic justification for forcing exit of producers who cannot compete at an overvalued exchange rate, and most observers would agree that the ruble Is overvalued at two to the dollar. Producer prices have changed since 1987, but macroeconomic deterioration at the national level has probably Increased the gap between domestic and world agricultural prices. An Implied rate of three or four rubles to the dollar would In 1987 have allowed payments for land and higher producer Incentives for most products. If the chosen exchange rate with which to translate border prices Into ruble prices Is higher than the equilibrlum rate that emerges as the economy opens, agricultural prices can be expected to adjust upward over time. The constralnt In cholce of an exchange rate on which to base revised agricultural producer prices Is clearly what the demand side can absorb at the time of revislon. Price Reform and Demand Side Constraints Producer prices based on border prices converted at three or four rubles to the dollar run directly Into constraints Imposed by the demand side. Consumers will face significantly higher prices when the subsidy covering the difference between retail prices and current producer prices Is removed. The demand side adjustment would be even greater If subsIdles were removed and producer prices simultaneously revised upward, as would be the case with a ruble exchange rate of 3 or 4 to the dollar. Retail price policy has stymied revision or reform of producer prices In the past, and It continues to present difficult dilemmas. Retall prices are kept iow by - 15 - large payments from the state budget. This component of the agrieultural subs1dy cost 90 billion rubles at ths national level In 1989, or roughly 11 percent of GNP. The coe4 was estimated to Increase to 115 billion rubles In 1990, with additlonal increases In early 1991. In Lithuania the budgetary cost of the subsidy for food consumed In the repubilc In 1989 was 1.39 blillon rubles, or about 30 rubles per capita per monUt7 If the subsidy were distrlbuted as a compensatory per capita payment of thirty rubles per month, It would augment the money Income of a family of four with two earners on average 30 percent. The compensation would be a higher relative payment for poorer and larger families. Without detailed Information on !ncome distribution and family budgets, It Is difficult to Judge whether compensation of this magnitude would exceed or fall short of the Increment In food costs for most families. The subsides are known to go disproportionately to higher Income urban familles who have preferential access to food at subs!dIzed prices. It Is thus likely that removal of the subsidies and distribution of the total as equal per capita compensatory payments would overcompensate many poorer familles and undercompensate the wealthler. It would thus improve the income distribution without distorting wage payments by linking componsation to wages. If revised producer prices were based on an exchange rate of three or four rubles to the dollar, retail prices would be even higher, since the current subsidy would be removed and producer prices would simuitaneously be Increased by a factor of 1.5 or 2 on average. in that case per capita monthly payments of thirty rubies would leave many familles unable to cover their additional food costs, and arguments in favor of a targeted compensation program would Increase. Means tested and commodity specific compensation programs could be considered. Poorer people receive much of thelr subsidy through dairy products, which are widely avallable in Lithuania at subsidized prices. - 16 - Concluslons The old price system, along with constraints on marketing and Input supply, Ilmits the attractiveness of new tenurial relations to producers, and distorts the values of agricultural land. The multipllilty of prices complicates the contractual negotlations and leads to monetizatlon of current distorted asset values. Lithuanlan farm accounts for 1986 and 1987 suggest that farms at that time received on average between 40 and 100 rubles annually In producer rents for use of e hectare of average quality land. Producer rents on Individual farms varied according to the prices received on that particular farm. If prices were standardized and changed to border prices converted at two rubles to the dollar, the marginal value product of average quality land would be higher (146 rubles per hectare In the 1986 data, and 90 rubles In the 1987 data). Farms coWd be asked to pay that amount In rent without exceeding land's contribution to net output. If prices were aligned with world prices at two rubles to the dollar, however, most Lithuanlan farms would not have Increased their Incomes very much, and payment of 100 rubles per hectare would have worsened financial stress. If prices were standardized and Increased to border prices converted at three rubles to the dollar Instead of two, the marginal value product of average quality land would be approxhmately 150 rubles per hectare (1987 data). Producers could pay that much without additional financial stress, although movements In prices of other Inputs are also relevant. If all agricultural users pald land use fees and had access to the same prices for Inputs and output, state and collective farm managers would be more eager to offer land lor lease or private proprletorship or ownership. Agr!cultural price reform would thus facilitate changes In land tenure and land management. Producer prices corresponding to border prices converted at three rubles to the dollar would present a shock to retall prices. Consumers probably could not be fully compensated for the Increase, and targeted compensation would be appropriate. - 17 - Table 1: LUthuanian Producor Prices (Uit values, rubles per ton, 1987 V/) Grouped Accordhg to Prices Received Lowest Highest Base Border Border 25% Median 75% Prices 2 rubles/$ 4 rubles/$ Rye 162 169 186 170 Barley 135 150 172 130 260 520 Oats 115 136 171 Wheat 125 133 160 130 Potatoes 125 143 167 125 167 334 Beets 59 59 61 58 40 80 Beans 323 476 625 Fruit 254 334 400 398 796 Beef * 2,506 2,680 2,892 1,550 2,220 4,440 Mutton * 2,564 2,651 2,798 2,200 1,540 3,080 Pork * 2,447 2,685 3,000 2,100 1,540 3,080 Poultry * 2,000 2,079 2,323 1,540 3,080 Milik 335 343 352 310 440 880 Wool 4,626 5,000 5,667 5,000 10,000 * IJve weight Sources: (a) Godovye otchety, 1987, Lthuanian SSR. (b) Chursin, A.M. Tsenv i kachestvo sel'skokhoziaistvennol Produktsli, Moscow, Kolos, 1984. (c) 1987-88 Commodit Trade and Price Trends. John Hopkins University Press, 1988. Price ProsPects tor Maior Primarv ComffiMies, 1988-2000. The World Bank, 1989. - 18 - Table 2: Producer PRes I988 Av7e&e Unit Values (tubles per ton) _ Toal ReepWTdTomal T to State ESTONIA BELORUSSIA LATVIA LITHUANIM Collective State Collective State Collective State Collective 1987 Rye 221.5 220.9 190.1 187.2 214.3 221.3 180.1 Barley 171.1 151.5 176.5 169.5 160.7 166.4 155.3 Potatoes 244.1 230.5 153.3 154.6 208.2 183.2 155.4 Beef 2,586.4 Z50s.2 3,406.7 3,361.4 2,902.2 2,881.3 2,689.7 Pork 2,303.5 2,285.1 2,912.9 2,568.6 2,599.7 2,501.0 2,720.5 Mik 370.4 367.6 497.5 528.1 414.1 421.2 343.8 Sugar-beets 70.4 75.3 85.9 91.0 60.8 Source: Gcdovye Otchety, Svodnye, 1988, listed republics -19- Table 8: Correlation Coefficients for Price Len Qutlity, Farm Profitability Land Farm Rye Barley Beets Beef Milk Wheat Oats Pork Poultry Quality Profitability Rye 1.00 (800)* Barley .28 1.00 (745) (948) Beets 1.00 (260) Beef .21 1.00 (260) (1030) Milk .12 .2S .41 1.00 (798) (260) (1028) (1028) eh"t .12 .17 1.00 (S62) (734) (766) Oats 28 .21 1.00 (420) (488) (48S) Pork -.10 .25 .54 .2S 1.00 (988) (259) (1020) (1018) (1020) Poultry 1.00 (148) Land Quality -.55 -.20 -.47 1.00 (1027 (102S) (1017) (1028) Fam ProfTlt bT I Tey .11 .14 .10 .19 .23 .36 1.00 (PCb (948) (1028) (1020) (148) (1028) (1082) All coofficients significant *t .006 level. Price. are unit value. Land quality ts bonTt t. o( ) Indicates number of observations. Profitability (rentabel'nost) = (ernings-costs)/coas. -20- Table 4: Estimated Cofticient for Log Linear Production Functions Lithuania, Actual Received Pricaem ! 1986 1987 FullI FullI Sampl e Sample Prof it 1 S/ Profit 2 Profit 3 2 Profit 4 VY Intercept 3.69 3.42 3.37 3.04 4.01 3.35 (13.43)kI (11.89) (6.68) (6.26) (7.80) (8.33) Land Quality 21 .S3 .4S .24 .28 .20 .41 (12.74) (10.41) (2.82) (3.54) (2.62) (6.51) Land Quantity h/ .1J .045 .07 .11 .07 .10 (4.63) (1.50) (1.30) (2.08) (1.31) (1.97) Labor ±1 .29 .34 .45 .44 .38 .39 (10.68) (11.13) (7.98) (8.03) (7.70) (6.79) Fertilizer 1/ .04 .10 .18 .10 .12 .05 (2.62) (4.43) (3.56) (2.63) (3.37) (1.28) Machinery k/ .13 .18 .10 .14 .18 .17 (7.06) (8.13) (2.49) (3.81) (4.52) (3.82) Animals V .30 .2S .14 .18 .18 .24 (14.69) (12.05) (3.65) (5.66) (4.47 (8.28) R2 .73 .73 .74 .80 .78 .79 */ Dependent variablo is not output S/ bonit aggregated with actual prices recelved h/ Land cropland + .6 hayland b/ t-stat;ce I/ Labor 2 hours worked In agriculture e/ Profit < 15.70% J/ Fertilizer = total ruble expenditures d/ 16.69% < Profit < 23.58% k/ Machinery n total horsepowr */ 23.65% C Profit < 32.01% 1 Animuls = cow + .3 (hogs and pigs) . .8 (sheep *nd gots) f/ Profit > 32% Source: Godovyc otchety, Litovskala SSR, 1987, 1988. -21 - Table S: Estimated CeeftIele.nt for Log Lie.ar ProduCtion Fq lctown. Lithuania, 3ordw Priem at 2 rublev/lt YI' 1986 1987 full Full Saple Sample Profit 1 9/ Profit 2 Profit 3 Profit 4 Intercept 2.12 2.39 2.41 1.70 4.01 2.6s (8.22)k/ (8.79) (5.00) (3.57) (8.05) (5.22) Land Quality .81 .71 .62 .57 .20 .s8 (20.68) (17.21) (7.72) (7.33) (2.62) (8.41) Land Quantity .18 .10 .14 .15 .07 .14 (6.94) (3.60) (2.49) (2.79) (1.31) (3.13) Labor .33 .35 .45 .48 .38 .41 (18.05) (12.36) (8.28) (8.53) (7.70) (7.73) Fertilizer .06 .09 .08 .12 .12 .05 (3.70) (4.46) (2.28) (3.31) (3.37) (1.34) Machinery .14 .18 .10 .11 .16 .14 (7.59) (7.73) (2.57) (3.14) (4.52) (3.52) Animals .21 .20 .10 .14 .1S .21 (10.81) (10.41) (2.73) (4.61) (4.74) (8.00) R2 .78 .77 .77 .83 .78 .82 */ Dependent variable Is net output aggregated with border prices converted at 2 rubles/il. b/ t-statistics c/ Profit 1: Rentabel'not.t < 15.70% d/ Profit 2: 15.69X < Rentabel>nost < 23.85X */ Profit 3: 23.64% < Rentabel'nost < 32.01X fJ Profit 4: 32% < Rentabel'nost Source: Codovy. otchety, Litovskala SSR, 1987, 1988. -22- Table St Saple _ess, t# FullI Sample Profit 1 ! Profit 2 Y Profit a Profit 4 4/ No. tn samplo 1,032 260 269 259 257 Land Quality (bonitot) 40.92 38.60 a9.53 41.13 44.52 Land Quantity (hotaros) 2,199 2,236 2,171 2,189 2,200 Labor (hours) 638,798 538,335 532,525 651,741 622,447 wags (i*ublos/hour) 1.32 1.19 1.30 1.33 1.49 Fortl I izr Expenditure in rubles 130,608 129,492 126,799 130,324 13S,818 Machinery (Horepower) 13,639 12,412 13,194 13,859 16,104 Anmale(w.ighted aggreated cow = 1) 1,088 930 1,030 1,080 1,234 Net output (actual prices) 1,991,851 1,691,212 1,869,373 2,090,193 2,423,374 Not output (border prices) 2 rubles = 81) 1,971,870 1,690,417 1,838,058 2,071,387 2,396,070 Net output (border pricos 4 rubles = 81) 3,943,741 3,180,834 3,672,111 4,142,775 4,792,140 a/ Profit 1: RentabelOnost < 16.70% b/ Profit 2: 15.69% Rentabl nost < 23.85X c/ Profit 3: 23.64U < Rentabl'nost < 32.01% d/ Proftt 4: 32X ( M.ntabel'nost Source: Godovy* otchey, Llto*vskal SSR, 1987. - 23 - Table 7: MargIal Value Products, Actual Prices, 1987 Full Sample Profit 1 a/ Profit 2 Profit 3 c/ Profit 4 d/ Land Quality 9.99 4.41 6.11 4.69 10.30 Land QuantIty 40 52 94 64 109 Labor 1.22 1.31 1.52 1.43 1.76 Fertilizer 1.48 1.70 1.49 2.01 .90 Machinery 25.77 12.94 19.24 25.56 26.90 Animals 477 253 344 354 489 a/ Profit 1: Rentabel'nost < 15.70% b/ Profit 2: 15.69% < Rentabel'nost < 23.65% c/ Profit 3: 23.64% < RentabelPnost < 32.01% d/ Profit 4: 32% < Rentabel'nost Source: Tables 4 and 6, derived from Godovye otchety, Litovskals SSR, 1987. - 24 - Tabb Marghi VaMlo Prodcts of Lnd (Mbb)) Received ReceOed Border Border Prbes PrIs Pries 198 Prie 1987 1988 1987 (2 rubles = $1) (2 rubles = $1) 1 Hectae Average qualit Full sample 107.21 40.11 145.85 89.58 Profit < 15.70% 52.43 9B.02 15.69 < Profit < 23.65% 93.52 119.84 23.64% < Profit < 32.01% 64.41 110.O2 Proft > 32% 108.79 158.93 1 Unit of quaity (bontet) Full sample 10.79 9.99 15.85 15.36 Protd < 15.70% 4.41 11.52 15.69 < Profit < 23.85% 6.11 1.08 23.64% < Profit < 32.01% 4.69 10.30 Profit > 32% 10.30 14.31 Sources: Tables 4, 5 and 6. - 25 - References Avdllants, lu. P. and A. L. Meiendorf. Teenoobrazovanie v agropromyshiennom komplekse. Moscow, Agropromizdat, 1989. Bolsvert, Richard. "The Transiog Production Function: Its Properties, Its several Interpretations and Estination Problems." A.E. Res. 82-28, Department of Agricultural Economics, Cornell University, 1982. Brooks, Karen, "Lease Contracting In Soviet Agriculture In 1989," Comparative Economic Studies, Vol. XXXII, No. 2, 1990. Brooks, Karen, J. Luis Guasch, Avishay Braverman, Csaba CsaKi, 1991. "Agriculture and the Transliion to the Market.a Journal of Economic Perspectives, forthcoming. Cook, Edward C, Willam Liefert, and Robert Koopman, "Government Intervention In Sovlet Agriculture: Estimate of Consumer and Producer Subsidy Equivalents." Agriculture and Trade Analysis Division, Economic Research Service, U.S. Department of Agriculture. Staff Report, 1990. Poshkus, Bollus 1. Vyravnivanie ekonomicheskikh usiovil khozialstvovanlia. Vlinius: Mokslas, 1979. Andrew Schwab and Philip Raup, The Minnesota Rural Real Estate Market In 1988, University of Mlnnesota, Agricultural and Applied Economics, Economlc Report ER 89-3, July, 1989).22 - 26 - Endnotes 1. Avdilants, lu. P. and A.L. Melendorf. Tsenoobrazovanie v agropromyshlennomkomplekse. Moscow, Agropromizdat, 1989. 2. The border prices are not adjusted for transport differentials within the Soviet Unlon. Lithuania Is small and has a major Ice free port at Kialpeda linked to the rest of the republic by a road and rall system better than In many parts of the USSR. Analysis of Sovlet producer prices In areas more distant from borders would require adjustment for transport differentlals. 3. These reference prices are consistent with prices used In the SWOPSIM model for the USSR developed by the Centrally Planned Economies Division of the Economic Research Service, USDA, but were Independently derived (Cook, 1990). 4. Transiog production functlons yielded negative marginal value products for several factors, Including labor, and are not reported here. Use of the translog may not be justified If elasticitles of substitution are less important to the analysis than are marginal value products (Bolsvert, 1982, p. 32). The log llnear production functions reported here ylelded reasonable parameters, and the translog did not, even though several of the cross product terms In the translog had non-zero estimated coefficients. 5. Sale prices of high and low quallty agricultural land In Minnesota show a range of 40 percent around the price of land of average quality (Schwab and Raup, 1989). 6. As long as the budget constraint for the parent farm remains soft, Justification for a risk premium Is weak. 7. Interviews during August, 1989 with economists In the Lithuanian Council of Ministers. PRE Working Paper Series Contact Autho for pape WPS632 The Macroeconomics of 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 Nidhiprabha 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. Cuskelly-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 The Impact of Policy in African William Jaeger March 1991 A. Yideru Agriculture: An Empirical Investigation 34663 WPS641 Intertemporal Substitution in Karsten N. Pedersen April 1991 M. Hileman Consumption: Evidence for Some 31284 High- and Middle-income Countries WPS642 How a Change in Brazil's Sugar Brent Borrell April 1991 A. Kitson-Walters Policies Would Affect the World 33712 Sugar Market WPS643 Regional Integration among Andras lnotai April 1991 D. Ballantyne Developing Countries, Revisited 37947 WPS644 Trade and Payments Arrangements Constantine Michalopoulos April 1991 M. Colinet in Post-CMEA Eastern and Central David Tarr 34698 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 Killick 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 Paper Series Contact lilL ~~~~~~ ~~Author Qate for paper 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