“The Tyranny of Numbers”CAUSES OF LABOR PRODUCTIVITY IN ARGENTINA, 1993-2004

ARIEL COREMBERG

Joensuu

Congreso; 29th General Conference of the International Association for Research in Income and Wealth Joensuu, Finland, August 20-26, 2006; 2006

International Association for Research in Income and Wealth (IARIW)

The main purpose of the paper is to show some evidence that the main cause of labor productivity growthin Argentina during 1993-2001 would have been capital intensity growth, instead of the Solow´s residual orTotal Factor Productivity (TFP).This result is analogous to the findings of Alwyn Young (1992) (1994) (1995) (1997) and Timmer and VanArk (2000), spread by Krugman (1994) for the Southeast Asian countries , but are different from otherprevious results on the Argentine experience, for example Kydland and Zarazaga (2002), which found thatTFP was the principal cause of GDP growth during the 1990s based on the calibration of a real businesscycle model.The discrepancies arise as a result of the methodological approach and the macroeconomic consistencyanalyses of the series, that in the case of this paper belong to National Accounts System of Argentina.According to the recommendations of the economic literature on productivity measurement, the sources ofeconomic growth are analyzed through the economic theory of index numbers. This allows the extractionof substitution and composition effects in production, labor input and capital services from the residualTFP.The paper evaluates the possible main causes of labor productivity growth: capital intensity (capital-laborratio) vs. TFP. Following the OECD Productivity Manual, this paper analyzes the main components of TFP;not all TFP is technological change and not all technological change is TFP. The paper discusses thepossible wrong identification of residual TFP –or the so call Solow’s Residual– as exclusively shifts in theproduction function.It is also considered the main problems of measurement of the capital stock and their possible impact onTFP growth, due to: biases due to the assumptions of the Perpetual Inventory Method on the level andperformance of the capital stock; changes in the quality of capital assets; econometric estimation of thefunctional form of depreciation; and other measurement problems in the capital stock figures, partiallycorrected by the author in recent estimations of the capital stock series in National Accounts of Argentina.As a result of the application of this methodology, taking into account the National Accounts series andtheir macroeconomic consistency, the paper shows that labor productivity growth and TFP growth had aremarkable procyclical behavior during the 1993-2004 period, being its trend significantly reduced for bothindicators, after subtracting the changes in capacity utilization and substitution and composition effects inproduction and factors from the Solow Residual.During 1993-2001, there was an important increase of the capital intensity of the Argentine economy as aresult of the high labor cost due to the low real exchange rate and high openness of the Argentineeconomy to external trade during the Convertibility Plan. One of the main results is that during 1993-2001,capital intensity reacts to changes of factors’ relative costs (fall of relative price of capital goods), showinga high elasticity of factors’ substitution in the Argentine economy.During the post devaluation period (2002-2004), Argentina reduces the capital intensity as a result of theincrease in the labor-GDP elasticity associated to increases in the capacity utilization, and some influenceof the reduction of its relative cost originated in the 2002 devaluation, displaying cyclical increases in laborproductivity and the residual TFP. The apparent cyclical increase of the residual TFP (and its trend) wouldbe reduced remarkably as a result of the discount of the Solow Residual by capacity utilization.The main conclusion is that the Argentine economy during the 1993-2001 period would present anextensive growth profile, based on factor accumulation rather than positive shifts in the production function(similar to the case of Southeast Asian NICs in the 80’s). Technological change would have been of anembodied type, stimulated by the fall in the relative price of capital goods as a result of the appreciation ofthe domestic currency and the increase in the openness to external trade during the 1990’s, but withouthaving a substantial positive effect on the organization of the argentine economy in the long run.During the post mega devaluation and crises period (2002-2004), GDP growth would have been explainedby the reduction of capital intensity originated in a higher relative labor demand, as a consequence ofincreases in the capacity utilization previously underused during five years of deep economic depression(1999-2001) and some influence of the high decrease of unit labor costs due to the high increase of realexchange rate. It is possible to ask if the deep changes of factor prices stemming from the 2002devaluation will produce higher labor-GDP elasticity sustainable in the long term.As a consequence of these stylized facts, some doubts arise about the capacity of the Argentine economyto generate the necessary gains in strictly TFP (regardless of the changes in relative prices of goods andfactors) that allow sustainable growth in the long run.