Prices in the Two Germanies

Journal of Comparative Economics 27, 131–149 (1999) Article ID jcec.1998.1562, available online at http://www.idealibrary.com on Prices in the Two Germanies* Erik Dietzenbacher 1 University of Groningen, P.O. Box 800, NL-9700 AV Groningen, The Netherlands E-mail: E.Dietzenbacher@eco.rug.nl and Hans-Jürgen Wagener Frankfurt Institute for Transformation Studies, Europa–Universität Viadrina, P.O. Box 776, D-15207 Frankfurt/Oder, Germany Received October 14, 1997, revised October 19, 1998 Dietzenbacher, Erik, and Wagener, Hans-Jürgen—Prices in the Two Germanies Comparative evaluations across economic systems are often hampered by the typical valuation problem, that is, market prices have a different meaning than planned prices. This paper considers prices in the two Germanies and the widely accepted hypothesis that East German prices did not reflect relative scarcities as well as did West German prices. To test this hypothesis empirically, Seton’s eigenprices are adopted as the standard of valuation. On the whole, the hypothesis is rejected in the sense that West German prices deviated from this standard of valuation by just as much as did East German prices. Closer inspection of the prices at the sectoral level provides an explanation for this unexpected result. J. Comp. Econom., March 1999, 27(1), pp. 131–149. University of Groningen, P.O. Box 800, NL-9700 AV Groningen, The Netherlands; and Frankfurt Institute for Transformation Studies, Europa–Universität Viadrina, P.O. Box 776, D-15207 Frankfurt/Oder, Germany © 1999 Academic Press Key Words: prices; economic systems; Germany; transformation. Journal of Economic Literature Classification Numbers: P51, D46, C67. 1. INTRODUCTION Since the collapse of the communist planned economies a large amount of information and data has become available. This would seem to allow an * We thank two anonymous referees and the Editor of this journal for their helpful comments and suggestions. 1 To whom correspondence should be addressed. Fax: 131 50 363 3720. 131 0147-5967/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved. 132 DIETZENBACHER AND WAGENER assessment of the real comparative performance of market and centrally planned economies in the years preceding the collapse. The case of the two Germanies seems to be a model case for comparative evaluation, because a culturally and economically fairly homogeneous nation was split into two. West Germany, i.e., the former Federal Republic of Germany, FRG, pursued capitalism, while East Germany, i.e., the former German Democratic Republic, GDR, created a communist economic system. After 40 years of separate development, the socialist eastern part was in a desperate economic state and its political system collapsed. The nation was reunified and the western part accepted the task of recapitalizing the east. Comparative national accounting should now help to explain what happened. Unfortunately, comparative evaluations across economic systems are hampered by the typical valuation problem. That is, prices serve widely different functions in market economies and in centrally planned economies and they are formed by different methods. Yet, in both systems, prices are used to construct aggregate indicators. These are difficult to interpret if we do not have information on the precise nature of the prices that were used. Therefore, comparisons across economic systems on the basis of such indicators must be treated with caution. In Germany, the question of whether or not the database necessary for such comparisons can be reconstructed from the statistical material that has become available (Statistisches Bundesamt, 1993) has been hotly debated. The stumbling block recurrent in the arguments is the use of prices. “The dilemma of valuation cannot be solved. This is so because a planned economy, other than a market economy, has no logically consistent system of valuation” (Utzig 1993, p. 9; our translation). If this statement is interpreted as asserting that prices in a planned economy are arbitrarily set exchange ratios without any relation to relative scarcities or economic valuations and capitalist market prices are close to equilibrium levels, it is a hypothesis in need of empirical testing on both sides of the wall. The hypothesis has been widely accepted, although its empirical verification has not been attempted. A plausible explanation for this omission may be that all problems with distorted GDR prices were believed to be circumvented by reconstructing East German statistical aggregates using rational West German prices. This is an economically meaningless exercise because a different set of prices would have implied a different set of quantities (Lützel, 1993, p. 75). The fact that exactly the same procedure is frequently applied without hesitation when prices of a different base year are used in order to make comparisons over time provides no vindication. This paper attempts to test the hypothesis that West German prices reflected relative scarcities much better than did East German prices. To do so, we need a set of prices derived from a logically consistent system of valuation that can serve as a standard of evaluation. In order to assess the degree of distortion from this standard, it makes no sense to measure the deviation of East German prices PRICES IN THE TWO GERMANIES 133 TABLE 1 Original Input–Output Table Z V t* y — — x* y x v t from West German prices. A theoretically more satisfactory method should use the East German set of prices and quantities and the West German set of prices and quantities separately and see how much they deviate from a normative standard. To this end, we adopt Seton’s eigenprices (Seton, 1985, 1992) as a systemspecific set of rational prices derived from a logically consistent process of valuation. These prices reflect marginal revenue products and marginal costs; thus, they fulfill the equilibrium criteria for logically equivalent solutions in the cases of perfect competition and perfect planning. Seton’s eigenprices have been used to evaluate pricing behavior in different economies and different economic systems. The calculations have shown, in most cases, that communistadministered pricing leads to considerably greater distortions than pricing in capitalist competitive systems (Seton, 1985; Steenge, 1985). In the following section, the model of eigenprices is discussed briefly. Section 3 presents the data. Section 4 contains the results and Section 5 presents the conclusions. 2. THE CONCEPT OF EIGENPRICES In this section, we discuss briefly the concept of eigenprices (Seton, 1981, 1982, 1985, 1989, 1992), which is based on two transformations. First, factor prices are transformed into product prices following the typical price calculations in input– output analysis. 2 Second, by specifying the distribution of income to the primary factors, factor prices are determined from product prices. 3 Combining these two transformations yields an eigensystem from which the prices are solved as eigenvectors. Consider an economy with n production sectors and m primary factors. Our starting point is the input– output table (Table 1). Z is the n 3 n matrix of intermediate deliveries; its typical element z ij denotes how much of product i is bought by sector j, to be used as an intermediate input. y is the n 3 1 vector of 2 For example, see Miller and Blair (1985, pp. 351–357). Generalizations, which are based on the same considerations, include two- and three-channel prices (Bródy, 1970, Sekerka et al., 1970, Brown and Licari, 1977, Seton, 1977, 1993, and Fink, 1981). 3 In deriving the second transformation we follow an approach that differs from Seton’s. 134 DIETZENBACHER AND WAGENER final demands; its typical element y i denotes the purchases of product i for final use purposes, i.e., for private and government consumption, investment, and exports. x is the n 3 1 output vector; its typical element x i denotes the value of the output of product i. V is the m 3 n matrix of factor inputs, including imports; its typical element v kj denotes the payments of sector j for the use of factor k. t* is the 1 3 n row vector of surpluses (t j ). 4 Such a surplus may consist of indirect taxes, subsidies, and certain profits as parts of the operating surplus. When countries with different economic systems are compared, the exact contents of the surplus row will be different. Therefore, it is obtained as a residual row. The corresponding totals are denoted as y 5 e9n y, t 5 t*e n and v 5 Ve n , where e n is the n 3 1 summation vector, i.e., e9n 5 (1, . . . , 1). It should be noted that all entries in the input– output table record transactions in money terms, e.g., millions of German marks. Prices, however, depend on the measurement of physical flows, which cannot be obtained from the observed data. This adds to the usual problem of measuring the output of service sectors and the inputs of capital and heterogeneous labor. In order to circumvent the problem of unavailable physical data, it is common practice to interpret these entries in physical terms. This can be established by setting each current price equal to one. Implicitly, this means that the physical units of measurement are chosen so that the current price of one unit of product i, or one unit of factor k, equals 1. In what follows, we derive the eigenprices p i for the products and f k for the factors, according to a well-defined equilibrium concept. An eigenprice p i 5 1.2, for example, indicates that the equilibrium price is 20% higher than the current price. For a given situation as described by the input– output table, eigenprices are the prices that would occur if the sectoral surpluses were obtained according to a uniform rate t, which is called the eigensurplus ratio. Note that it is explicitly assumed that all quantities remain fixed. It is our aim to analyze the effects of this change on the prices, given the current physical structure of the economy as reflected by Table 1. Applying the new prices p i and f k to the quantities in Table 1 and using s* for the new surplus row yields a new, revalued input– output table (Table 2). The sectoral surpluses s j are obtained as a share t of the primary factor costs. That is, s j 5 t ( f 1 v 1j 1 . . . 1 f k v kj ), or s* 5 tf*V using matrix notation. Column j in Table 2 now yields p j x j 5 ¥ i p i z ij 1 ¥ k f k v kj 1 s j 5 ¥ i p i z ij 1 (1 1 t )¥ k f k v kj . Dividing both sides by x j and using the common definitions for input coefficients, i.e., a ij [ z ij /x j and b kj [ v kj /x j , 5 gives the price per unit of product j (5 1, . . . , n) as p j 5 ¥ i p i a ij 1 (1 1 t )¥ k f k b kj . In matrix notation, 4 Vectors are by definition column vectors. A prime is used to denote transposition, so that t* is a row vector. 5 The coefficient a ij gives the input of product i required per unit of output of product j and b kj gives the input of factor k required per unit of output of product j. Note that the coefficient matrices A and B are calculated from Table 1, which is given. PRICES IN THE TWO GERMANIES 135 TABLE 2 Revalued Input–Output Table p̂Z f̂V s* p̂y — — p*x̂ p*y p* 5 p*A 1 ~1 1 t !f*B. p̂x f̂v s (1) This equation may be rewritten as p* 5 ~1 1 t !f*C , with C ; B~I 2 A! 21 . (2) Equation (2) describes how the product prices are obtained from the factor prices and the uniform eigensurplus ratio. In order to derive a similar, but reversed, transformation we first consider the quantity side of the model. Using the definitions for the input matrices A [ Zx̂ 21 and B [ Vx̂ 21, it follows that the first n rows of Table 1 may be written as x 5 Ax 1 y or x 5 (I 2 A) 21y. The typical element (i, j) of the matrix (I 2 A) 21 describes how much of product i needs to be produced in order to satisfy a final demand of one unit of product j. For the next k rows, describing the factor use, in Table 1 we have v 5 Bx 5 B(I 2 A) 21y 5 Cy. So the element c kj denotes the total amount of factor k that is required for one unit of final demand for product j. The elements c kj are now used to distribute the income to the factors. Product j is sold on the market for final products at a price p j and in the amount y j , thus generating p j y j as income in sector j. Satisfying a final demand of y j units of product j requires c kj y j units of factor k. In distributing p j y j , each factor k is rewarded according to its own contribution to p j y j . 6 This yields c kj p j y j for factor k (5 1, . . . , m). Note that the remainder, p j y j (1 2 ¥ k c kj ), is allocated to surpluses. The total receipts of factor k are obtained by summing over j, so as to yield ¥ j c kj p j y j . Dividing this value by the total number ( v k ) of units of factor k that is used provides the factor price f k . Thus, f k 5 ¥ j p j c kj y j / v k , or in matrix notation, f* 5 p*N, with N ; ŷC*v̂ 21 . (3) Equation (3) shows how factor prices are obtained from product prices. Combining Eqs. (2) and (3) yields 6 Although this distribution seems appropriate, it is not the only possibility (Fink, 1982). Other types of income distribution lead to different equilibrium prices but these may create difficulties with the empirical implementation and/or computation. 136 DIETZENBACHER AND WAGENER ~1/1 1 t !p* 5 p*NC and ~1/1 1 t !f* 5 f*CN. (4) The vector p* of eigenprices for the products is calculated as the left eigenvector corresponding to the dominant eigenvalue 1/1 1 t of the matrix NC. The eigenprices for the factors are then obtained from Eq. (3) as f* 5 p*N. 7 Since eigenvectors are only unique up to a scalar multiple, a scaling technique is required to obtain the final results. We adopt Seton’s scaling procedure in which product prices are chosen such that the total value of final demand, which equals GDP for the case of a closed economy, remains constant. That is, p*y 5 e9n y or O p y 5 O y 5 y. i i i i i (5) A consequence of scaling Eq. (5) is that the product prices are on average equal to 1, which reflects their current price. This does not necessarily hold for the factor prices, however. Suppose that the rate t turns out to be lower than the overall surplus rate t/e9m v 5 t/( y 2 t) in Table 1. Using v 5 Cy and Eqs. (2) and (5), we have f*v 5 f*Cy 5 p*y/(1 1 t ) 5 y/(1 1 t ) 5 (t 1 e9m v)/(1 1 t ) . e9m v. Hence, t , t/( y 2 t) if and only if f*v . e9m v, implying that the weighted average of the factor prices is larger than 1. Eigenprices require that the surplus be distributed according to a uniform ratio. 8 This eigensurplus ratio t can be interpreted as an overall measure for the discrepancy between prices and the real costs of production, i.e., in terms of the payments for the factors. It is easily seen that the row s* in Table 2 vanishes completely if factor prices become (1 1 t ) f k instead of f k . In other words, if all surpluses were redistributed to the primary factors, the factor prices f k could be increased by 100t%, given the same product prices p i . Equivalently, leaving the factor prices f k the same, the product prices p i could be decreased by 100(t/1 1 t)% to become (1/1 1 t ) p i , to cover exactly the costs of the primary factors. In an empirical comparison between countries or between points in time, a larger eigensurplus ratio t indicates that a smaller part of the product price pays for the primary costs of production (see, for instance, Dietzenbacher and Steenge, 1985, or Steenge, 1985). This endogenously determined parameter t is inherent or eigen 9 for the economy under consideration. The numerical values for the prices p i and f k indicate whether the current prices are more or less consistent with the framework of a uniform distribution of the country’s eigensurplus. Consider an economy in which prices are determined by markets and an economy in which prices are set by the planning authorities. Suppose that, in the 7 An alternative procedure, which yields the same results, computes f* as the left eigenvector corresponding to the dominant eigenvalue 1/1 1 t of the matrix CN after which p* is obtained from (2) as p* 5 (1 1 t)f*C. 8 Dietzenbacher (1990) discusses the case where this ratio does not need to be uniform but is allowed to be sector-specific. This leads to generalized eigenprices. 9 “Eigen” in German and Dutch means “own” or “characteristic.” PRICES IN THE TWO GERMANIES 137 first economy, some markets are characterized by strong competition that drives profits to zero while, in a few other markets, monopolistic forces prevail. On average, the overall surplus is modest so that t will be relatively close to zero. The eigenprices, however, may be expected to differ from the current prices. In order to force the system into a uniform distribution of the surplus, prices on competitive markets will rise, i.e., p i . 1, and those on monopolistic markets will fall, i.e., p j , 1. For the centrally planned economy, suppose that a substantial part of the production costs is spent on the maintenance of the government body. This will result in a positive value for t. If the planning authorities set the product prices more or less equal to the costs plus a uniform markup to account for the government, we will find that p i ' 1 and f k ' 1. On the other hand, if the planning authorities decide that basic needs ought to be cheap while luxury goods should be expensive, we will find p i . 1 for basic needs and p j , 1 for luxury goods. As the model shows, eigenprices are standardized so that they can be compared directly with actual prices. An eigenprice below unity indicates the overvaluation of the respective product. The cause may be monopolistic practices, special taxes, levies, and tariffs, or random errors of the planner. An eigenprice above unity signals the opposite distortion caused usually by a severe form of overcompetition, subsidies, or, again, planners’ errors. It should be borne in mind, however, that a difference between eigenprices and actual prices may also indicate a distortion in the quantity system. It was assumed that the quantity system behind a given input– output table reflects valuations and constraints characteristic of the economy in its current state. An eigenprice above 1, indicating undervaluation of a given commodity, thus may also reflect the fact that the domestic production of this commodity uses too many resources compared either to its domestic substitutes or to production by international competitors. 3. THE DATA The starting point for calculating eigenprices is comparable input– output tables. Major difficulties arise because national accounting differed fundamentally between the two economic systems. We mention a few of the most serious differences. First, West German national accounting followed the System of National Accounts concept while East German national accounting followed the Material Product System concept, which allocated nonmaterial services to the sphere of consumption (United Nations, 1977, or Boda and Stäglin, 1990). Second, prices in both systems were influenced by government policy. In the west, market prices reflected taxes and subsidies and only in rare cases, e.g., public transport, were they directly set by political authorities. In the east, price setting was the responsibility of the central planner who, in principle, followed simple cost-plus pricing rules. However, costs were influenced by intricate 138 DIETZENBACHER AND WAGENER TABLE 3 Domestic Production a Zd L m* Yd — — x9d e958Y d xd Le 58 m*e 58 a Z d , domestic intermediate deliveries (a 58 3 58 matrix); Y d , final use of domestic production (a 58 3 5 matrix); x d , total domestic production (a 58 3 1 vector); L, primary cost categories (a 5 3 58 matrix); and m*, imports of intermediate goods by productive sectors (a 1 3 58 vector). financial flows, i.e., charges and subsidies, which were meant to supplement the quantity planning system. End-user prices thus became an instrument of economic policy that was used extensively in the sphere of consumer goods. Third, in a market system, economic data are confidential at the firm level. This difficulty may be overcome by laws on statistical reporting. However, in a planned system, some data were confidential at the national level. In the GDR, authorities were reticent to publish data on true economic performance and even tried to give an unrealistically positive picture of the economic state of affairs. Due to the continuous need for hard currency credit, foreign trade statistics were notorious for this manipulation and balance of payments statistics have never been published. This led observers to argue that data were falsified. Most specialists, however, share the opinion that basic statistical material in the GDR reflects the true situation of the economy and can be used for empirical purposes (Kockel, 1992). It is a cumbersome task to make East and West German input– output tables comparable or, in particular, to rearrange and recalculate the East German table according to West German standards. Ludwig and Stäglin (1993) prepared a GDR table for 1987 having 58 sectors, which includes a table of imports and which can be compared to the West German input– output table. The Statistisches Bundesamt publishes FRG tables for even years. For the purpose of comparison, it has also provided a table for 1987 which, however, is less detailed with respect TABLE 4 Imports a a Zm Ym m* e958Y m xm Z m , import of commodity i by sector j (a 58 3 58 matrix); Y m , import for final use (a 58 3 5 matrix); and x m , total imports (a 58 3 1 vector). PRICES IN THE TWO GERMANIES 139 TABLE 5 Domestic Production and Imports Zd 1 Zm L x9m Yd 1 Ym — — x9d 1 x9m e958(Y d 1 Y m ) xd 1 xm Le 58 x9m e 58 to imports. Therefore, we are using the 1987 table of East Germany and the 1988 table of West Germany (Statistisches Bundesamt, 1994) as the basic material for the computation of eigenprices. In open economies, foreign trade causes difficulties in the construction of input– output tables. While exports are an element of final use, imports can be treated in different ways. Tables 3–5 are usually available and the format is indicated in brackets. Input– output analysis and the concept of eigenprices start from the idea that the production structure is described by input coefficients. These are obtained by dividing each element of a column in the Z matrix by its corresponding column total. Looking at Table 3, an obvious problem arises. Input coefficients reflect only domestically produced inputs, imports are treated as a primary factor of production. It makes technologically little sense to differentiate between domestically produced and imported steel for the production of cars, and to compute input coefficients by relating domestic total car production to the use of domestically produced steel, since steel is assumed to be a homogeneous product. Table 5 does not resolve the issue. Indeed, the Z matrix now reflects total inputs and can be interpreted in a technologically meaningful way. However, the row and column sums do not represent production, but domestic production plus total imports. A nonnegligible part of imports goes directly into final use. Therefore, the input coefficients are distorted in this case also. The only compromise possible, given the data of Tables 3–5, is Table 6, which has Z d 1 Z m as its Z matrix but excludes all imports that are not used as intermediate products. In Table 6, x mp is the vector of imports that are used in the TABLE 6 Domestic Production with Imported Intermediate Goods Zd 1 Zm L x9mp Yd — — x*d 1 x*mp e958Y d x d 1 x mp Le 58 x9mp e 58 140 DIETZENBACHER AND WAGENER TABLE 7 The Input–Output Table Used for Eigenprice Calculations Z 5 Zd 1 Zm V t* y 5 Yde5 — — x* 5 x9d 1 x9mp y x 5 x d 1 x mp v t production process, or productive imports. The aggregate amount x d 1 x mp can be interpreted as the total output necessesary to produce Y d with productive imports included. Each sector j has a primary input (x mp ) j that supports total production. A second difficulty with the data set derives from the difference in economic systems between West and East Germany. Usually, western input– output statistics record labor income as income from hired labor, while the income of the self-employed, including family workers, is included in the income from entrepreneurial activities and capital. This is unsatisfactory, in general, but especially so when we make comparisons with socialist economic systems. For these systems, all labor is considered since self-employment does not exist. Hence, the published figures for labor income represent total labor income. In order to make the necessary corrections for West Germany, we followed the method of Bedau and Klatt (1992). We calculate the average income of hired labor, multiply it by the number of self-employed, including family workers, and add this amount to the row of income from hired labor. The necessary data can be found in Statistisches Bundesamt (1994, p. 316). Thus, we obtain rows for labor input and capital input that are fairly representative and comparable. Lack of separate data for self-employed persons in sectors 51, i.e., services of renting buildings and lodgings, and 55, i.e., other market-determined services, made it necessary to aggregate the two. A final remark on the data concerns the services of banks, i.e., sector 49, which are a notorious problem in input– output tables. The most important product of banks is interest margins, for which it is not clear who is the end user of the concomitant services. Hence, these are recorded as deliveries to the sector itself. In fact, the input– output data of this sector cannot be interpreted in any meaningful way. We follow the usual practice of not making an attempt to do so. 4. THE RESULTS The concept of eigenprices was discussed using Table 1 and the actual data underlying our empirical results were presented as Table 6. For the sake of clarity, both tables are integrated in Table 7. The matrix V records the following three rows; labor, capital, i.e., deprecia- PRICES IN THE TWO GERMANIES 141 tions, and imports, i.e., the row x9mp in Table 6. The row t* with surpluses covers indirect taxes minus subsidies and income from entrepreneurial activity and capital for West Germany and a row published as “differences” for East Germany. Calculating eigenprices from Table 7 yields the results presented in Tables 8 and 9. Table 8 presents the eigenprices p i for the products, the contribution ( p i y i ) of sector i to final demand measured in eigenprices, and this contribution as a percentage. We can now measure the overall deviance, d, as the weighted sum of the squared sectoral deviances. This yields d 5 D 1/ 2, where D 5 O iÞ49 $~ p i 2 1! 2 @ p i y i / O p y #%. iÞ49 i i Note that we have omitted sector 49, i.e., services of banks, from the overall deviance for the reasons indicated above. Since ¥ i p i y i 5 ¥ i y i from the normalization of eigenprices, dropping sector 49 causes a small inconsistency in the weighting scheme for which we do not correct. The results for d, based on Table 8, are d GDR 5 0.1647 and d FRG 5 0.1611. Comparing these values for overall deviances reveals that actual prices in East Germany at the end of the 1980’s did not deviate significantly more from eigenprices than did actual prices in West Germany. The data do not support the hypothesis under examination. In other words, on the whole, East German prices did not reflect relative scarcities significantly worse than did West German prices. This is an astonishing result because earlier calculations of eigenprices by Seton (1985) and Steenge (1985) and other measures of price deviations by Fink (1981) always derived substantial differences between market economies and planned economies. Indeed, one should be careful in comparing the overall deviance between input– output tables of different sizes. However, not all of the earlier calculations used tables with considerably fewer sectors than we have. For example, Seton (1985, p. 116) reports overall deviances of 0.2862 for the USSR in 1972 using a 56-sector classification, 0.1228 for Japan in 1970 using a 60-sector table, and 0.1229 for West Germany in 1970 based on a 44-sector classification. Closer inspection of the eigenprices in Table 8 yields some explanation for the surprising result. The first 8 sectors represent primary production and utilities. Under the conditions of West German capitalism, market forces do not play a dominant role here. Products of agriculture (sector 1) and mining (6 and 7) are undervalued or, given their subsidized prices, use up too many resources. Communally produced utilities, such as gas (4) and water (5), appear to be overvalued, suggesting some monopolistic practices. This segment of total production does not exhibit significant differences between east and west, but there is no reason that we should expect any differences. However, in sectors 9 through 40, coverering manufacturing, in which market forces should prevail, West German pricing behavior is considerably better than that of East Germany with few exceptions. The exceptions are petroleum products (10), ships (24), and tobacco products (40), of which the first and the last are characterized by huge excise taxes and the middle by state subsidies. Sectors 41 and 42, covering 142 DIETZENBACHER AND WAGENER TABLE 8 Eigenprices for the Products a GDR (1987) FRG (1988) i pi p iy i % pi p iy i % 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 0.8934 1.1024 1.0595 1.2478 0.8943 0.9971 0.9782 1.2647 0.8603 0.9753 0.9106 0.8300 0.8446 0.7788 0.9558 1.0592 1.1547 0.9795 0.9351 0.8263 1.0092 1.1632 1.0579 0.7417 0.9763 0.8876 0.9333 0.8118 0.8359 1.0865 0.9236 0.9505 0.7673 0.8333 0.9813 0.9371 0.9118 1.2718 0.5675 0.2004 0.9242 0.8782 0.6248 10643.8 461.9 2047.0 482.9 0.0 3969.5 1775.5 1005.5 16291.9 11702.7 1099.1 1326.3 2352.2 704.1 1708.0 5859.3 1817.4 327.2 4198.7 6635.4 35393.9 2677.7 12576.7 2256.9 153.3 21976.4 2308.0 4499.2 2431.7 637.8 6887.1 1095.9 860.2 577.5 4158.8 9452.7 7993.1 41340.6 4093.0 1054.7 20479.0 9339.1 5003.5 2.5 0.1 0.5 0.1 0.0 0.9 0.4 0.2 3.8 2.7 0.3 0.3 0.6 0.2 0.4 1.4 0.4 0.1 1.0 1.6 8.3 0.6 3.0 0.5 0.0 5.2 0.5 1.1 0.6 0.1 1.6 0.3 0.2 0.1 1.0 2.2 1.9 9.7 1.0 0.2 4.8 2.2 1.2 1.2753 1.1114 0.9923 0.8474 0.8264 1.4952 1.1776 1.1591 1.0076 0.7631 1.0104 1.0626 0.9625 1.0986 1.0156 1.1124 1.1302 1.0904 1.0055 1.0469 1.0770 1.0435 1.0646 1.1825 1.0649 1.0633 0.9829 1.0260 0.9262 1.1414 1.0668 1.0788 1.0251 1.0023 1.0905 1.0783 1.0341 1.0943 0.8965 0.2670 1.0125 0.9149 0.9515 15138.1 2499.6 24884.9 6073.1 13.2 2329.6 997.4 139.1 86116.8 19510.4 17698.5 7334.0 6529.3 2772.9 4273.8 20357.3 12518.3 2236.4 13111.8 19051.9 134751.0 18340.8 176061.7 4384.6 5600.4 105603.8 14029.3 27630.3 7094.4 2260.0 30763.9 7244.2 7516.1 2476.7 6764.7 19624.4 22599.3 117905.9 15306.9 4830.7 115258.8 56523.6 71086.9 0.7 0.1 1.1 0.3 0.0 0.1 0.0 0.0 3.9 0.9 0.8 0.3 0.3 0.1 0.2 0.9 0.6 0.1 0.6 0.9 6.1 0.8 7.9 0.2 0.3 4.8 0.6 1.2 0.3 0.1 1.4 0.3 0.3 0.1 0.3 0.9 1.0 5.3 0.7 0.2 5.2 2.6 3.2 PRICES IN THE TWO GERMANIES 143 TABLE 8—Continued GDR (1987) i 44 45 46 47 48 49 50 51155 52 53 54 56 57 58 Total a FRG (1988) pi p iy i % pi p iy i % 1.0359 1.1166 1.0351 1.1929 1.0661 3.3617 0.2905 1.0294 1.0690 0.7722 0.6657 1.1933 1.0913 1.2761 9395.0 7834.9 4326.6 1984.9 7732.3 463.9 336.5 8746.7 16401.9 2468.6 289.6 73262.9 13197.2 8264.1 2.2 1.8 1.0 0.5 1.8 — 0.1 2.1 3.9 0.6 0.1 17.2 3.1 1.9 1.0456 2.0122 1.0911 0.9788 0.9710 4.5773 0.8037 0.6006 1.0272 1.0358 0.7354 1.1332 0.9489 1.1530 126970.2 13833.6 8784.5 26341.6 44350.5 40770.4 23552.0 146411.6 39842.3 33958.4 11141.7 330319.5 132737.8 40377.1 5.7 0.6 0.4 1.2 2.0 — 1.1 6.6 1.8 1.5 0.5 14.9 6.0 1.8 426360 100 2256636 100 The sectoral classification is given in the Appendix. construction, indicate that West German performance is slightly better than East German. The rest of the sectors are services. If we neglect banks (49), the major deviations from unity in the west are rail transport (45), which is notorious for not covering its costs, and the aggregate of sectors 51 (renting of structures and lodgings) and 55 (other market determined services), which appears to be heavily overpriced with significant consequences for the overall deviance because of its weight of about 7% in West Germany. In these sectors the indicated overvaluation or underutilization of resources is due partly to incorrect capital stock estimates of houses and buildings, which constitute a major part of total national wealth. Also, the last 4 sectors representing health care, public authorities, and nonprofit organizations have a rather large weight but no scarcity-led pricing practice. To sum up, the unexpected similarity of the overall deviance between actual prices and eigenprices in East and West Germany does not warrant the conclusion that East German pricing was not so bad after all because it was guided, more or less, by relative scarcities. In those sectors where market forces prevail in the west, East German actual prices deviate significantly more from eigenprices. However, the similarity of the overall indicator testifies to the fact that a sizeable part of West German GNP is not produced under competitive market conditions. Obviously this is the case in sectors with predominant public ownership rights and strong government influence. This is a positive, not a normative, statement. 144 DIETZENBACHER AND WAGENER TABLE 9 Further Results Eigensurplus ratio t Eigenprices f k Labor Capital Import Actual cost ratio t/(y 2 t) Rescaled eigenprices f̃ k Labor Capital Import GDR (1987) FRG (1988) 0.3687 0.1497 1.0587 1.0306 1.0105 0.4178 1.1038 1.0285 1.0313 0.2433 1.0220 0.9949 0.9755 1.0206 0.9510 0.9536 The upper part of Table 9 reports the eigensurplus ratio and the eigenprices for the factors. As far as the eigensurplus ratio t is concerned, the West German system is expected to exhibit a smaller t than the East German system because we cannot compute eigenprices net of direct taxes and West Germany relies much more upon direct taxes than did East Germany. The eigensurplus ratio t represents, to some extent, the monopoly charge or value-added tax that an external authority, i.e., the government, is able to impose upon the system. If many of the taxes and levies are indirect, this charge will be higher than when direct taxes form a major share of total taxes. The results confirm this hypothesis. By implication, we see that a much larger share of total revenue accrues to productive factors in West Germany. Alternatively, in East Germany, the system managers, i.e., the government and the central planners, can dispose directly of a larger share of total product. This leaves aside direct taxes and the fact that in East Germany the government as owner of productive capital has a claim on capital rentals, which, as a rule, they extracted from the enterprises and added to the state budget. However, consumption need not be relatively smaller in the east than in the west because the government can use the extracted surplus for any purpose. In East Germany, the subsidization of consumer prices and housing rents was extreme and we note that eigenprices were calculated from input– output tables in producer prices. On the other hand, a smaller share of total product was subject to the competitive regime or, in Marxist terminology, to commodity–money relations in the East. Our system employs three primary factors of production, labor, capital, and imports. Their eigenprices ( f k ) are proportional to the factors’ productive contribution, i.e., its marginal revenue product. As we stated already in Section 2, the weighted average of the factor rentals need not be equal to 1. In both cases, all eigenprices of factors are greater than 1, which cannot be interpreted as undervaluation of factors of production. PRICES IN THE TWO GERMANIES 145 In explaining these large values for the factor eigenprices, recall from Section 2 that f*v . e9m v if and only if t , t/( y 2 t). Eigenprices are computed so as to distribute the surplus according to a uniform ratio. Factor payments as a share of total final demand amount to 1/1 1 t when eigenprices are used and to 1 2 (t/y) in the actually observed input– output table. So, t , t/( y 2 t) implies 1/1 1 t . 1 2 (t/y), indicating that a larger share of total final demand is available for factor payments. Thus, in comparing the eigenprices ( f k ) for the factors with their current prices, i.e., 1, a comparison of t with t/( y 2 t) is implicitly involved. For East Germany, we find that t 5 36.87% and t/( y 2 t) 5 41.78%; for West Germany we have t 5 14.97% and t/( y 2 t) 5 24.33%. Note that the eigenprices indicate that the gap between t and t/( y 2 t) is larger for West Germany than for East Germany, in the sense that the deviation from unity is larger. To allow for a proper comparison in terms of under- or overvaluation, we use rescaled eigenprices (f̃ k ) for factors. These are obtained from the eigenprices by applying a correction for the change in the total factor payments as a share of total final demand, that is, f̃ k 5 f k [1 2 (t/y)](1 1 t ). These rescaled eigenprices are on average equal to 1, i.e., f̃*v 5 e9m v. It is interesting to note that the eigenprices ( p i ) for the products and the rescaled eigenprices (f̃ k ) for the factors yield a uniform distribution of the current surplus t across the sectors. The results for the rescaled eigenprices are presented in the lower part of Table 9. They indicate that labor is undervalued to a similar extent in both countries. Capital and imports are overvalued, but significantly more in West Germany than in East Germany. 5. CONCLUSION Eigenprices can be used as a rational standard in order to evaluate actual pricing behavior and to compare structural data between different economic systems. Eigenprices cannot be used to compare levels of output or income between two systems. For such comparisons, the old index number problem remains. Varying the assumptions and the databases for the two Germanies, the results turned out not to be sensitive to minor changes. Eigenprices are a quite robust standard of valuation. Our starting hypothesis that prices in a socialist planned economy are less rational than prices in a capitalist market economy has not been supported unambiguously by the data for the two Germanies at the end of the 1980’s. The reason for this rather unexpected result is revealed by closer inspection of the calculated eigenprices. The West German economy is not a pure capitalist market economy; rather it is a mixed economy with some sectors, mainly manufacturing, subject to competitive market processes and others, mainly primary production and noncommercial services, under heavy government influence. Hence, actual 146 DIETZENBACHER AND WAGENER prices deviate significantly from a logically consistent system of valuation in the latter case. Regarding East German pricing behavior, our results conform more with intuitive expectations. The process of price planning, which was interactive between the firms, i.e., the so-called Kombinate, and the price-setting administration, did follow some uniform rules but, in the end, it resulted in major deviations from a rational system of valuation. These deviations are distributed more evenly over the whole profile of sectors of production than were the deviations in West Germany. The factors causing such deviations, such as deliberate policy choices, differences in monopoly power between individual sectors, and planners’ errors, can be detected only after closer analysis. How do our results apply to the reliability of using national accounting statistics to make comparisons? First of all, calculations in eigenprices will yield more useful comparisons over time and economic systems. If the comparisons are based on data in actual prices, reservations are in order. However, West German data are used for international and intertemporal comparisons without hesitation. The present study has indicated that, by implication, East German data deserve a similar treatment since, on the whole, they are not worse than West German data. APPENDIX List of the 58 Sectors of the Input–Output Table Used in This Study 1. Agricultural products 2. Forestry and fishery products, etc. 3. 4. 5. 6. 7. 8. Electric power, steam, hot water Gas Water Coal, products of coal mining Products of mining (excluding coal, crude petroleum, natural gas) Crude petroleum, natural gas 9. Chemical products (including nuclear fuel) 10. Refined petroleum products 11. Plastic products 12. Rubber products 13. Stones and clays, building and construction materials, etc. 14. Ceramic products 15. Glass and glass products 16. 17. 18. 19. Iron and steel Nonferrous metals, semifinished products thereof Foundry products Products of drawing plants, cold rolling mills, etc. PRICES IN THE TWO GERMANIES 20. 21. 22. 23. 24. 25. Structural metal products, rolling stock Machinery and equipment (excluding electrical) Office machinery, automatic data processing equipment Road vehicles Ships, boats, and floating structures Aircraft and spacecraft 26. 27. 28. 29. Electrical machinery, equipment, and appliances Precision and optical instruments, clocks, and watches Tools and finished metal products Musical instruments, games and toys, sporting goods, etc. 30. 31. 32. 33. 34. 35. 36. 37. Wood Wood products Pulp, paper, and board Products of paper and board Products of printing and duplicating Leather and leather products, footwear Textiles Wearing apparel 147 38. Food products (excluding beverages) 39. Beverages 40. Tobacco products 41. Building and civil engineering works, etc. 42. Installation and building completion works 43. 44. 45. 46. 47. 48. Services of wholesale trade, etc., recovery Services of retail trade Railway services Water transport services, services of ports, etc. Communication services Other transport services, n.e.c. 49. 50. 51. 52. 53. 54. 55. Banking services Insurance services (excluding social security funds) Real estate renting services Market services of hotels and restaurants, homes, and hostels Market services of education, research, culture, and publishing services Health and veterinary market services Other market services, n.e.c. 56. Services of central and local government 57. Services of social security funds 58. Services of private nonprofit institutions, domestic services 148 DIETZENBACHER AND WAGENER REFERENCES Bedau, Klaus D., and Klatt, Helmut, Die Entwicklung der gesamtwirtschaftlichen Arbeitseinkommensquote und ihre Einflußen [The Development of the Aggregate Rate of Earned Income and the Factors Influencing It], Discussion Paper 47. Berlin: German Institute for Economic Research, 1992. Boda, György, and Stäglin, Reiner, “Intersystem Comparison between the Federal Republic of Germany and Hungary on the Basis of SNA Type and MPS Type Input–Output Tables.” Vierteljahr. Wirschaft. 4, 363–377, 1990. Bródy, András. Proportions, Prices and Planning. Amsterdam: North-Holland, 1970. Brown, Alan A., and Licari, Joseph A., “Price Formation Models and Economic Efficiency.” In Alan Abouchar, Ed., The Socialist Price Mechanism, pp. 184 –230. Durham, NC: Duke Univ. Press, 1977. Dietzenbacher, Erik, and Steenge, Albert E., “Seton’s Eigenprices: Comparisons between Post-War Holland and Hungary.” In Anatoli Smyshlyaev, Ed., Input–Output Modeling, pp. 237–248. Berlin: Springer-Verlag, 1985. Dietzenbacher, Erik, “Seton’s Eigenprices: Further Evidence.” Econ. Systems Res. 2, 2:103–123, June 1990. Fink, Gerhard, “Price Distortions in the Austrian and in the Hungarian Economy.” Z. Nationalökon. 41, 1–2:111–132, 1981. Fink, Gerhard, “International Comparisons with Input–Output Models.” In Reiner Stäglin, Ed., International Use of Input–Output Analysis, pp. 231–258. Göttingen: Vandenhoeck & Ruprecht, 1982. Kockel, Klaus, “Ausgewählte Probleme der früheren DDR-Statistik” [Some Problems of Statistics in the Former GDR]. Allgemeines Statistisches Archiv 76, 1:1–14, 1992. Ludwig, Udo, and Stäglin, Reiner, “Problems in Comparing Input–Output Tables for East and West Germany.” Econ. Systems Res. 5, 2:137–147, June 1993. Lützel, Heinrich, “Sinn und Unsinn einer Umbewertung” [Sense and Nonsense of a Reevaluation]. In Statistisches Bundesamt, Rückrechnungen gesamtwirtschaftlicher Daten für die ehemalige DDR [Backward Calculations of National Accounting Data for the Former GDR], pp. 73– 82. Stuttgart: Metzler-Poeschel, 1993. Miller, Ronald E., and Blair, Peter D. Input–Output Analysis: Foundations and Extensions. Englewood Cliffs, NJ: Prentice–Hall, 1985. Sekerka, B., Kyn, O., and Hejl, L., “Price Systems Computable from Input–Output Coefficients.” In Anne P. Carter and András Bródy, Eds, Contributions to Input–Output Analysis, pp. 183–203. Amsterdam: North-Holland, 1970. Seton, Francis, “The Question of Ideological Obstacles to Rational Price Setting in Communist Countries.” In Alan Abouchar, Ed., The Socialist Price Mechanism, pp. 10 –39. Durham, NC: Duke Univ. Press, 1977. Seton, Francis, “A Quasi-Competitive Price Basis for Intersystem Comparisons of Economic Structure and Performance.” J. Comp. Econ. 5, 4:367–391, Dec. 1981. Seton, Francis, “Relative Prices across Space and Time—The Production of Commodities by Means of Commodities and Factors.” In Béla Csikós-Nagy, Douglas Hague, and Graham Hall, Eds, The Economics of Relative Prices, pp. 24 – 45. London: Macmillan, 1982. Seton, Francis. Cost, Use and Value—The Evaluation of Performance, Structure, and Prices across Time, Space, and Economic Systems. Oxford: Clarendon Press, 1985. Seton, Francis, “Eigenprices Revisited—Recapitulation and Apologia.” Econ. Systems Res. 1, 2:273– 283, June 1989. Seton, Francis. The Economics of Cost, Use, and Value, 2nd ed. Oxford: Oxford Univ. Press, 1992. Seton, Francis, “Price Models Based on Input–Output Analysis: East and West.” Econ. Systems Res. 5, 2:101–111, June 1993. PRICES IN THE TWO GERMANIES 149 Statistisches Bundesamt. Rückrechnungen gesamtwirtschaftlicher Daten für die ehemalige DDR [Backward Calculations of National Accounting Data for the Former GDR]. Stuttgart: Metzler–Poeschel, 1993. Statistisches Bundesamt. Volkswirtschaftliche Gesamtrechnungen, Fachserie 18, Reihe 2, Input– Output-Tabellen 1986, 1988, 1990 [National Accounts, Series 18, Part 2, Input–Output Tables 1986, 1988, 1990]. Stuttgart: Metzler–Poeschel, 1994. Steenge, Albert E., “Time-Series Analysis for the Netherlands and Hungary.” Special Annex in Seton (1985), pp. 129 –162. United Nations. Comparison of the System of National Accounts and the System of Balances of the National Economy, Part II, Series F, No. 20. New York: United Nations, 1977. Utzig, Siegfried, “Welche ökonomische Aussagefähigkeit enthalten Rückrechnungen ehemaliger DDR-Statistiken?” [What Is the Economic Meaning of Calculating Backwards Statistics of the Former GDR?]. In Statistisches Bundesamt, Rückrechnungen gesamtwirtschaftlicher Daten für die ehemalige DDR [Backward Calculations of National Accounting Data for the Former GDR], pp. 3–12. Stuttgart: Metzler–Poeschel, 1993.