Distortions, factor proportions and efficiency losses : Argentina in the Latin American scenario

World Bank Reprint Series: Number 384 Julio J. Nogues Distortions, Factor Proportions and Efficiency Losses: Argentina in the Latin American Scenario Reprinted with permission from Welftuirtschlaffliches Archiv, vol. 121, no. 2 (1985), pp. 280-303.  Distortions, Factor Proportions and Efficiency Losses: Argentina in the Latin American Scenario By Julio J. Nogues Contents: I. Introduction. - II. The Impact of Alternative Trade Strategies on Employment in the Argentine Manufacturing Sector. - III. Quantitative Importance of Factor Market Distortions. - IV. Domestic Factor Market, Distortions and Employment: A Simulation Exercise. - V. Labor Intensity per Unit of International Value Added. - VI. Argentina in the Latin American Scenario. - VII. Conclusion. I. Introduction Tr his paper reports some orders of magnitude of the employment and income gains from resource reallocation that could be expected from removal of policy-induced distortions in the Argentina manufacturing sector. The data used in the analysis is from the early 1970s when Argentina was following import-substitution policies. The paper will focus attention on policy-induced distortions in factor markets. The results complement previous analysis regarding the employ- ment effects of altemative trade strategies [Nogues, 1983]. It proves useful for the objectives of this paper to summarize these results. This is done in Section II. The framework which we use for simulating the effects of factor market distortions on factor proportions is presented in Section III. This section also discusses and presents estimates of factor market distortions. Section IV uses the results of Section III to simulate the employment effects of removing factor market distortions. Section V will present some estimates of income gains that could be obtained by removing policy-induced distortions quanti- fied in this paper. Section VI compares our results with those obtained for other Latin American countries, and finally Section VII presents some concluding comments. II. The Impact of Alternative Trade Strategies on Employment in the Argentine Manufacturing Sector In this section I summarize findings showings the employment effects of alternative commercial policies in the manufacturing sector. In order to Remark: Tne author is a World Bank staff member. The paper reflects the author's views alone and should not be interpreted to express those of the World Bank. I appreciate comments received from Professor A. 0. Krueger while writing my doctoral dissertation from which this paper draws [Nogues, 1980].  Efficiency Losses 281 estimate these effects, manufacturing industries were classified according to trade orientation between exportables ard import competing (IC)'. We have also classified IC industries according to whether they compete with imports coming from developed or less developed countries (DC's and LDC's)2. This is done because countries like Argentina, whose manufacturing capital-labor endowment lies in between extreme values corresponding to capital abundant and labor abundant countries (DC's and LDC's), might have their IC industries competing with either of these groups of countries3. Also for these types of countries, exportable industries are expected to require techniques whose factor intensity characteristics lie in between those corresponding to these two groups of IC industries. Table 1 shows estimates of employment per unit of domestic value added (L/DVA) and effective rates of protection (ERP's). We observe, not surprising- ly, that the trade regime was clearly biased in favor of IC industries. While this Table 1 - Employment per Unit of Domestic Value Added and Effective Rates of Protection: Manufacturing Industries (1973)a Employment per Unit Effective Rate of Trade Balance Trade Orientation of Domestic Value Protection (ERP) (1,000 U.S. $) Added (L/DVA) A. Exportables ....... 18.899 - 0.026 1,392 B. IC .............. 14.555 1.302 - 1,074 1. With DC's ...... 13.365 1.498 - 1,029 2. With LDC's ..... 21.873 2.274 130 3. Other IC industries 11.574 0.470 - 175 C. Total Manufacturing 15.329 0.841 318 a ERP's have been estimated by weighing 1969 industry-specific rates with 1973 estimates of value added. Source: Abridged from Tables 2 and 3 in Nogues [1983]. ' The theory which has facilitated the empirical analysis of the employment effect of alternative trade strategies is presented in Krueger [1983, Ch. 4]. This is essentially a two-factor trade model extended to include many commodities and many countries. 2 As a general rule, exportables (IC) industries were considered to be those having a positive (negative) trade balance, and/or low (high) effective protection. The group of other IC industries includes those where the origin of competing imports under a liberal trade regime is not straightfor- ward, as well as industries enjoying protection from relatively high transport costs. A more detailed discussion of the classification of industries as well as of the analysis summarized in this section can be consulted in Nogues [1983]. 3 In Nogues [1980, App. to Ch. 1], I present quantitative evidence showing that the Argentine manufacturing capital-labor endowment lies in between extreme values observed in capital and labor abundant countries.  282 Julio J. Nogues group of industries had a weighted average ERP of 130 percent, the correspond- ing figure for exportables was as low as -3 percent. From the standpoint of resource allocation and given the higher labor intensity of exportable indus- tries, the bias in the trade regime was clearly worsening labor opportunities for employment. We conclude that the adoption of an export promotion policy which is characterized essentially by relatively uniform incentives for domestic and export sales together with the adoption and maintenance through time of a realistic exchange rate would increase labor demand for a given level of manufacturing activity. In order to be more confident on this result, I have classified U.S. manu- facturing industries according to trade orientation of Argentina industries. U.S. factors and goods markets are significantly less distorted than Argen- tina's markets. Therefore, for this country the reduction in the importance of distortions is expected to result in a structure of factor proportions across industries more similar to that observed in the U.S. [Nogues 1980, pp. 126-129]. When estimating labor per unit of domestic value added for U.S. manufacturing industries I find that the group of industries which I have classified as Argentine exportables is 33 (26) percent more (less) labor intensive than the group of industries which in Argentina are classified as IC with DC's (IC with LDC's). Therefore, the factor intensity ranking according to these broad groups of industries is similar between the U.S. and Argentina. I take this result as an additional piece of evidence in favor of the predictions emanating from my findings on alternative trade strategies and employment in the Argentine manufacturing sector. Nevertheless, in Latin American countries, import substitution policies have been accompanied not only by distortions in goods markets, but also by severe distortions in factor markets, including as we shall see, those occa- sioned by the trade regime on the price of capital goods. The next section will assess the quantitative importance of these distortions for Argentina during the early 1970s. These estimates will then (Section IV) be used to simulate the impact of these distortions on factor proportions. M. Quantitative Importance of Factor Market Distortions This section is arranged in the following way. In Subsection 1, I present a simple partial equilibrium framework which allows to simulate the employ- ment effects of removing factor market distortions. Subsections 2 and 3 will then discuss and estimate policy-induced distortions in capital and labor markets. Several distortions have been pointed out in the literature, and some of them have been estimated for other countries'. For example, it has long since I See for example papers written for the National Bureau of Economic Research (NBER) project on Trade and Employment directed by Anne 0. Krueger. Country specific papers of this project are published in Krueger et al. [1981].  Efficiency Losses 283 been recognized that the trade regime may have important distortionary effects on the costs of capital goods. Quantitative restrictions, overvalued exchange rates, and an escalated structure of tariffs are all potential sources of distortions introducing a wedge between the actual cost of capital goods paid by different industries, and the one that would prevail in a more open economy. Financial policies can also have such effects if they create a wedge belween the real market interest rate and the opportunity cost of financial capital. These topics are analyzed in Subsection 1 below. Subsection 3 presents a brief discussion of some relevant characteristics of urban labor markets and shows estimates of distortions which have been introduced by policy measures in this market. Specifically attention is focused on old-retirement plans, and other earmarked taxes on wages. In both, Subsections 2 and 3, special attention is paid to differentiating effective as opposed to legal distortions'. 1. Framework for Simulating the Effects of Factor Market Distortions on Employment In this paper, I use a very simple framework for analyzing the employment effects of factor distortions in a partial equilibrium context2. This framework assumes that production functions in manufacturing industries are of the Cobb-Douglas type. In a two-factor world, labor (L) and capital (K), and assuming constant returns to scale3, labor productivity in any industry can be represented by: (1) ()O = A (L) where Q is output, is capital's share in output, and A is a scale factor. Under the assumption of profit maximization the capital-labor ratio for a given factor-price ratio is given by: I The estimates of the extent to which policy-induced distortions have affected the price of productive factors should be regarded as lower bounds. For example, and because we could not differentiate legal as opposed to effective incentives, the effects of tax deductions for investment purposes are not discussed. In many cases, these incentives have been of major importance in deciding whether or not to invest. For a discussion of legal as opposed to effective tax subsidies on the price of capital goods in the Argentine context see Nogues [1980, p. 206]. 2 It is obviously recognized that in a general equilibrium model, changing reiative factor prices would affect not only factor utilization but also the composition of output and the vector of equilibrium prices. Due to these and other assumptions presented in the text, the estimates to be presented in Section IV should be regarded as possible orders of magnitude. The reader interested in general equilibrium estimates might consult Henderson [1982]. 3 Econometric evidence supporting the assumption of unitary elasticity of substitution between labor and capital and constant returns to scale is presented in work done for the NBER project by Behrman [1982]. Weltwirtschaftliches Archiv Bd. CXXI. 6  284 Julio J. Nogues () K a P (2) (L-)