Argentina's labor markets in an era of adjustment

Policy, Research, and Extemal Affairs WORKING PAPERS Macroeconomlc Adjustment and Growth Country Economics Department The World Bank March 1990 WPS 386 Argentina's Labor Markets in an Era of Ad justment Luis A. Riveros and Carlos E. Sanchez Changing Argentina's structure of production requires funda- mental reform of the labor market with regard to wage policies and the extent of government intervention. It will alsc require adequate financing during the transition period to compensate the potential losers from structural adjustment who might other- wise prevent its successful implementation. Tle Policy. Research. and Etenmal Affairs Complex distributes PRE Working Papers to disseminate the findings of work in progress and to ancourage the exchange of ideas among Bank staff and all others interested in development issues. These pers cary the names of the uhors, rdelect only their views, and should be used and cited accordingly. The findings, interpretations, and conclusions au the 'authore' ownx They should not be attributed to the World Bank, its Board of Directors, its mnnagement, or any of its member countries.  Policy, Research, and Extemd Affair lMlacrooconomlo Adjustment and Growth This paper - a product of the Macroeconomic Adjustment and Growth Division, Country Economics Department - is part of a larger effort in PRE aimed at identifying typical labor market policies in LDCs, specifically those affecting wage flexibility and the inteisectoral labor mobility. Copies are available free from the World Bank, 1818 H Street NW, Washtington DC 20433. Please contact Raquel Luz, room NI 1- 061, extension 39059 (48 pages with figures and tables). The current economic crisis in Argentina is only ity is going to allow for an adequate labor partly the result of inappropriate domestic market response, certain institutional changes policies to cope with recent extemal shocks. are necessary - including the decentralization Years of inappropriate policies have damaged of wage bargaining and the elimination of Argentina's economy. Even if no extemal traditional wage policies and general govem- shocks had occurred, the country would still ment intervention. have to change the structure of production. * If resources are to be shifted among regions Argentina has had trouble sustaining a and industries, rigidities and restrictions on labor program of structural adjustment. Its experi- mobility must be eliminated. More labor ences provide policymakers with some lessons mobility means less pervasive govemment inter- in designing a sustainable program to achieve vention in the form of restrictive regulations and price stability and change the incentive system: spending pattems. It also means that the govem- ment must stop the use of deliberate policies of Macroeconomic and trade policies must be employment absorption. consistent. Trade policy in particular must mesh with fiscal reforms, because cuts in public * Prospective losers in adjustment programs spending exert downward pressure on the real can stop their implementation, so it is important exchange rate. that adjustment programs provide them with benefits. Public spending must be profoundly Labor relations and labor market institu- changed to reduce social costs during the tions must be changed. If resources are to shift transition period. This will require extemal among industries and regions and wage flexibil- financing and debt alleviation. Pe PRE Working Paper Series disseminates the findings of work under way in the BankPs Policy, Research. and ExteDal Affairs Complex. An objective of the series is to 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 at the PRE Dissemination Center  The research program of the Macro, onomic Adjustment and Growth Division is looking at how the structure of key markets affects the effectiveness of adjustment programs. In particular, the research program on labor markets aims at identifying typical labor market policies in LDC s. specifically those affecting wage flexibility, labor mobility and institutional rigidities. Along these lines, the Division is preparing several country studies that will be later on integrated in a set of common findings allowing to develop a comparative study. This paper is part of this research effort.  Table of Contents Page I. Introduction 1 II. Long Run Economic Trends and Short Run Adjustment Policies 2 1. Macroeconomic Policies, Economic Organization and Growth 2 2. Political and Economic Developments of the 1970s 4 3. The Crisis of the 1980s 9 4. A Frustrated Process of Adjustment 11 III. The Structure and Trends of Labor Markets 15 1. Wage Trends 15 2. Regional and Sectoral Labor Allocation 18 3. Labor Market Segmentation and Increasing Informality 22 4. Labor Market Institutions 24 IV. The Role of Labor Markets in the Adjustment Process 26 1. The Performance of Labor Markets in the Adjustment 26 2. Level and Composition of the Aggregate 29 3. Unemployment and the Public Sector 35 4. Women, Labor Markets and Adjustment 37 5. Income Distribution and Poverty 38 6. Political Economy, Labor Markets and Adjustment 40 V. Conclusions 41 Endnotes 44 References 46 Comments on a previous draft by M. Faig, R. Newfanmer, S. Horton, R. Paredes and the participants of Seminars at The World Bank, the University of Chile and the University of Warwick, as well as the efficient research assistance of 0. Giordano are deeply acknowledged.  Page List of Tables Table 1 Different Strategies and Growth 2 Table 2 Argentinat Economic Ir.dicators 5 Table 3 Selected External Accounts - 1980/1984 9 Table 4 Changes in Relative Prices 13 Table 5 Manufacturing Wage Exchange Rate, and Product Wage 13 Table 6 The Evolution of Relative Wages (1962-1987) 18 Table 7 Average Annual Rate ,rf ropulation Growth 20 Table 8 Characteristics of the Sectoral Employment 23 Table 9 Argentina: Relative Prices and GDP, 1962-87 28 Table 10 Indicators for the Main Urban Labor Markets 32 Table 11 Labor Participation and Unemployment 33 Table 12 Public Sector Employment (1985) 35 Table 13 Public Sector Employment Growth 1960-87 36 Table 14 Unemployment, Sectoral Income Distribution and Poverty 39 List of Figures F'ig ra ' FT, .RZ, CDPpc and Real Wages 6 Figure 2 Real Wage and Product Wage in Manufacturing 14 Figure 3 Real Wage and FTT 17 Figure 4 The Evolution of Relative Wages 19 Figure 5 Labor Participation and "Equivalent' Unemployment 34  1 I. INTRODUCTION Poor growth and macroeconomic imbalances have characterized Argentina's recent history. In combination with long run stagnation, the country has suffered chronic high inflation and deep cyclical fluctuations compounded by intractable balance of payment crises. Attempts to stabilize the economy and achieve a structural adjustment to restore sustained growth and basic balances have failed due to both a fragile political climate and inconsistent policies. The lack of adjustment has been partly due to anticipated negative short run labor market outcomes and, in turn, the existence of persistent economic imbalances over the long run has negatively affected wages, employment and the income distribution. This vicious circle transforms labor markets in the Argentinean economy into a crucial factor regarding both the need for and the feasibility of any significant macro reform. This paper analyzes the performance of Argentina's labor markets in recent years. It aims at pointing out both the form in which labor market outcomes have been affected by a poor economic performance and the form in which reform efforts have been hindered by anticipated short run costs. We begin by providing economic background on the Argentine economy, particularly in recent years. In section III we discuss the structure and trends of labor markets and their connection with long term economic policies. In Section IV we analyze the role of labor markets in the adjustment process, and discuss the long run stop-and-go cycle of real wages growth prompted by the role of unions and government intervention. In section V we present a summary and conclusions.  2 II. LONG RUN ECONOMIC TRENDS AND SHORT RUN ADJUSTMENT POLICIES I1.1 Nacroeconouuic Policies, Economic Organization ard Growth Trade and macroeconomic policies followed after the Great Depression are crucial to explain Argentina's poor economic performance. During 1860- 1929 an export-led growth strategy was followed, which included almost-free trade and appropriation of the benefits of trade according to the country's c=p3rativa advai"Lges. After the Great Depression, adoption of an import substitution strategy was a key policy shift. Post-1945 government policies aimed at expanding domestic markets through overvalued exchange rates and high import tariffs, distorting the resource allocation and thwarting exports over the long run. Economic results of inward-oriented policies were deficient relative to those seen under the export-led strategy (Table 1) Table 1 Different Strategies and Growth (Real Growth in per capita GDP) Yearly Average growth rate Export-led strategy 1900-1929 1.5 Inward-oriented strategy 1929-1958 0.9 1958-1987 0.7 The entire period 1900-1987 1.0 Source: IEERAL Data Base. Macroeconomic policies affecting variables like the share of goverrment consumption, public debt and the money stock in total income, as well as commercial (export taxes and import controls) and exchange rate policies, were  3 mostly responsible for the increase in the effective exchange rate for imports relative to that for exports. The resulting price increase of import goods relative to exports made import substitution activities much more attractive for investment decisions, thus prompting an inefficient specialization of production. In addition, macro policies als-i affected relative prices of productive factors, thus resulting in a distorted capital/labor mix in production affecting sectoral and regional resource allocation.1 The Argentinean economy can be divided into three sectors: a rural and two urban sectors (Llach & S&nchez, 1984). The rura: sector is a net ey c:. -of *wage-goods (mainly agricultural). Therefore, real agricultural wages are inversely related to the incentives to produce exportable goods for a given exchange rate. The two urban sectors are net importers. They consist of an import substituting sector (which under the prevailing inward-oriented strategy is both a marginal exporter and a net importer of inputs and capital goods),. and a sector producing non-tradables. This economic structure led to a trade off between the trade balance situation and the prospects for domestic growth. Long-term economic growth crucially required a stable and high real exchange rate. But a high exchange rate implied higher food prices and thus lower urban real wages. Since urban real wages were a key variable for policy makers, there was a tendency to hold down incentives to export. Thus, due to the implementation of a deliberate policy of overvalued exchange rates--i.e. a policy mix aimed at bearing high wages in combination with low real exchange rates-- relative prices attained two simu _neous roles: a mechanism for resource allocation  4 and a distributive device. This led to contradictory long run economic targets: inefficient organization of production or improved well-being of urban workers, a conflict generally resolved in favor of the former. As a result of the deliberate policy of overvalued exchange rates, Argentine'r economic history during recent decades can be described as a series of redistributive periods (domestic currency overvaluation as well as increased activity and reai wpges) followed by periods of stabilization (devaluation and reductions in the levels of activity and real wages). 1I.2 Political and Economic Developments of the 1970s Argentina experienced crucial political and economic changes in the second half of the 19709. tDusually high growth rates of per capita GDP at an average of 3.9 percent p.a. had been observed in 1963-73. However, growth did not result from specific domestic policies, but fror favorable foreign terms of trade (FTT), especially at the beginning (1964-66) and the end (1971-73) of that period. Nonetheless, economic growth allowed for policies aimed at overvaluing the real exchange rate and at increasing both the domestic absorption of goods and real wages (Table 2 and figure 1). A new government (Peronist) that took over in 1973 inherited the combination of a satisfactory economic performance with persistent imbalances associated with the structural organization of the economy. The FTT were changing still favorably in 1973-74 and the government continued using redistributionist policies. Achievement of increasing real per capita GDP and real wages at the cost of a deteriorating real exchange rates was still feasible. However, the hike in oil prices and the ensuing global recession revealed the fragility of this policy. Imported inflation  5 and a sharp decline in FTT in 1975 made it impossible to maintain the overvaluation of the domestic currency without creating a sharp external deficit. When the Peronist administration was overthrown by the military in 1976, there was an acute balance of payments crisis and a huge fiscal deficit of more than 13 percent of GDP. Table 2 Argentina% Economic Indicators (Index, 1970 - 100) FTT RER GDP pc w 1962 94 125 78 83 1963 101 120 75 81 1964 115 100 81 92 1965 114 95 87 98 1966 112 90 87 98 1967 107 112 88 96 1968 106 106 90 91 1969 101 103 96 96 1970 100 100 100 100 1971 114 94 103 104 1972 125 109 105 98 1973 142 88 109 107 1974 114 82 114 120 1975 95 82 111 115 1976 90 117 109 74 1977 94 105 114 69 1978 92 80 107 72 1979 100 58 113 80 1980 117 46 113 93 1981 125 57 104 85 1982 105 100 97 74 1983 102 123 98 90 1984 111 107 99 99 1985 95 133 93 81 1986 81 106 96 76 1987 77 107 97 67 FTT- Price of Exports relative to Imports; RER- Real exchange rate estimated according to the purchasing power parity definition (nominal exchange rate adjusted for changes in domestic and foreign prices) GDP pc-per capita GDP; w-Real wages (nominal wages deflated by the CPI). Source: IEERAL Data Base.  6 Figure 1 FTT, RIR, GDPpc and Real Wages (Index, 1970 - 100) l.a. Real Wage and Gross Domestic Product 130 - 120 t2:0 1 117 1174 * 131 1 O bi Vi + R ol CP i.b. FTT auld RER 110 105 73. 1M low I= 1171 174 1177 1_ t 1_ n ftdWr + Red GO  7 The military government's objective was to reduce drastically inflation and to initiate a longer term strategy of switching expenditures and productive resources to encourage gust:ained growth and full employment. A central part of that strategy w& a two-stage trade reform in 1976-81. During the first stage (1976-1978), an unannounced tariff cut was implemented, along with elimination of export t&zes, and placement of quotas by tariffs in the cases of ali commcdities excepr steel and aluminum. During Lhe second stage (1979-81) a preannounced schedule of quarterly tariff reductions was achieved. Even though the program was supposed tD have continued until January 1984, trade reforms were reversed in 1981 due 1J - sharp balance of payments crisis. Completion of the trade liberalization was jeopardized by excessive gradualism in the second stage and also because a uniform tariff was not the end target. Strikingly, the final result on overall nominal protection -- including tariff and nontariff barriers -- was contradictoryt in 1979-80 the unweighted average nominal tariff rate decreased slightly (from 51.9 to 49.2 percent), but its dispersion increased greatly.2 Another problem with the trade liberalization program was the use of an accompanying inconsistent mix of fiscal/monetary and exchange rate policies. In 1979 and 1980 use of an active-crawl-reduction scheme based on a preannounced schedule of future devaluations resulted in an especially severe currency overvaluation. Although tax collection increased, the fiscal policy produced a large budget deficit because of the increase in both current spending and public investment. Thus overall deficit went up from 13 percent of GDP in 1976 to 16 percent in 1981 and 17 percent in 1982. Since a basic aim of the macro policy was to curb inflation, the government reduced the monetary financing of the deficit and it began to rely heavily on domestic and external  8 borrowing, thereby pushing up market interest rates and significantly increasing the public external debt. The persistence of a large fiscal deficit in combination with a severe overvaluation caused pressures towards balance of payments disequilibrium, accompanied by high real interest rates as well as low employment, production and investment. The economic policy of the late 1970s did not produce any positive result in terms of domestic balances. Although stabilization r3duced the yearly CPI inflation from a peak of 441 percent in 1976 to 101 percent in 1980. inflation remained high: in 1981 it was still 104 percent p.a.. Likewise, the active-crawl reduction scheme implemented after 1978 and the external financing of the government's excess demand caused the real exchange rate to appreciate: taking 1976 as a base year, its level was 39 in 1980 and 49 in 1981. Financial policies used after 1978 also resulted in volatile real interest rates. In January 1979 the preannouncement of nominal devaluations and the elimination of most restrictions on capital mobility were implemented. During the first eight months of 1979, when the policy still had certain credibility, real interest rates were low but reached negative values at times. Later on, uncertainty increased and risk premiums became high, thus raising real interest rates as high as to 2 percent per month at the lower end, and 6 percent at the upper end. The increased uncertainty observed in 1980-81 was closely related to both the existing gap between inflation and the rate of devaluation,3 and to observed change3 in external accounts (Table 3). The long-term structural adjustment program was barely implemented and its final result was a shift in incentives in favor of nontradable activities. The short-term stabilization program was unsuccessful; in the  9 presence of a persistent budget deficit, financing via capital markets (which replaced simple money creation) produced a crowding-out effect and was strongly deflationary (Mann & Sanchez, 1984,1985). Table 2 also shows the significant variability in real output between 1975 and 1980, before the economy moved into another recession in 1981, which exerted pressures on the labor market. 11.3. The Crisis of the 1980. In 1981 new economic authorities had to address the external and internal imbalances that resulted from overvaluation and the fiscal/monetary mismanagement. A drastic program of exchange rate devaluations was instituted to deal with the most urgent policy problem. From 1981 to 1983, the real exchange rate (RER) depreciated by 115 percent. 1981 and 1982 were years of macroeconomic adjustment when the RER sharply depreciated and both the real per capita GDP and real wages experienced large reductions (Table 2). The current account deficit in 1982 was substantially smaller than that observed in 1980 and 1981 (Table 3), thus probably making policy makers to believe that further adjustment was not necessary. In fact, they opted for a new shift in policies during 1983 and 1984. Table 3 Selected External Accounts - 1980/84 (millions of US dollars) 1980 1981 1982 1983 1984 Exports 8,021 9,143 7,623 7,838 8,100 Imports 10,540 9,430 5,336 4,505 4,600 Trade balance -2,519 - 287 2,286 3,331 3,500 Interest payments 956 2,925 4,400 4,983 5,273 Current Account Balance -4,769 -4,714 -2,357 -2,461 -2,492 Sources D. Cavallo (1986), table 1.  10 During the last year of the military government in 1983, the policy aimed at recovering real wages, thus reinstating the traditional deliberate policy of overvalued exchange rates. As a result, real wages rose 22 percent in 1983, while the ratio wage/exchange rate increased more than 40 percent during the last three quarters. The overvaluation was accompanied by active fiscal policies and a resurgence of inflation. Long term adjustment was relinquished and traditional populist policies returned to guide policy making. A civilian government (Radical Party) took office in December 1983. A very weak economic situation confronted vast expectations on better social welfare resulting from the restoration of democratic institutions. Activity levels and wages continued to grow in 1984 accompanied by a high fiscal deficit, growing inflation, low public utility rates, and exchange rate overvaluation (Table 2 shows that the real exchange rate declined 13 percent between 1983 and 1984, while real wages increased 10 percent). At the end of 1984 the government signed an agreement with the IMF which initiated an external sector adjustment based on demand reduction. During the last quarter of 1984 and the first half of 1985 the economy suffered a drop in real wages and activity levels, a depreciation of the exchange rate, and rising inflation. In June 1985, after sharp increases in public utility rates and a drastic devaluation, the Austral Plan was introduced. It contained both heterodox and orthodox ingredients to deal with inflation. The former included a wage and price freeze and a deindexation of debts. The latter included long-term measures such as a high exchange rate and fiscal restraint.  11 Inflation declined rapidly in 1985 due to price and wage controls. However, since the fiscal problem remained unsolved, monetary policy continued to be very active. At first, the demand for money increased substantially, but inflation returned later on due to expectations on the unsustainability of the program. In August 1986 and February 1987 there were two other attempts to reduce inflation by means of a tight monetary policy and control on wages and prices. However, inflation remained high because its primary source -- lack of fiscal discipline -- was not eliminated. In addition, in August 1986 a period of overvaluation begans The RER averaged 113 (index December 1976-100) during the first seven months of 1986 and then dropped to 106 during the period August 1986 to August 1987. II.4 A Frustrated Process of Adiustment (1987 - 1988) During September and October 1987 the government began to implement some new policies much more in line with a program of structural adjustment. However, no positive result has been yet observed and no structural adjustment has taken place. The relevant question is as to why a government politically committed to structural adjustment ended up with quite different results. The period can be divided into two phases. During the first one, from September 1987 to July 1988, a devaluation followed by a crawling peg adjustment of the exchange rate was implemented. During the second one, after July 1988, the real exchange rate again appreciated importantly. During the first phase, non-agricultural wholesale prices experienced very significant increases of 49 and 23.6 percent with respect to private services and the cost of living, respectively. Cereals and oilseed prices increased 84 percent with respect to private services with the help of the  12 increase in international grain prices. During this phase, the exchange rate policy provided substantial incentives to the tradable sector (Table 4). As a result, export activities expanded and a reallocation of resources towards export-oriented activities began. The effects of this policy on the volume of exports and the trade surplus were significant. Exports increased 43.6 percent and the trade surplus 607 percent in 1988. In the specific case of manufacturing -- a potentially exportable sector in Argentina -- the change in relative prices (wages, exchange rate and domestic terms of trade) led to a pattern of increasing profits and remarkable exports growth. Inflation accelerated, mainly due to the failure in reducing the fiscal deficit. After falling from 25 percent in October 1987 to 3 percent in December, the monthly inflation rate reached more than 20 percent in July 1988. At the time that price and exchange controls were imposed the government announced long-term measures aiming at shifting resources to tradable activities, improving the x-efficiency of the public sector, and freeing-up rigidities in factor and output markets. During the second phase, -- after July 1988 -- real wages in manufacturing increased and profits began to decline (Table 5). Table 4 also indicates that non-agricultural wholesale prices (i.e. mainly manufacturing prices) deteriorated vis-a-vis private services (-0.5 percent), the cost of living (-7.4 percent) and construction (-4.2 percent). In other words, the domestic terms of trade turned in favor of non-tradables. When the increase in wages and in the price of non-tradables is accompanied by overvaluation and elimination of import restrictions, the manufacturing sector probably faces falling profit margins (the squeeze on tradables) and lower domestic market  13 shares. The poor timing and lack of coordination between short-term and long- term economic policies hindered the achievement of structural adjustment. Table 4 Changes in Relative Prices (percentages) Relative Prices June '85/ Aug '86/ Sept '87/ July '88/ Aug.'87 Aug '87 July '88 Oct '88 Exch.rate /private services -40.1 -5.2 25.0 -12.7 Non-agric WPI/irivate services -44.7 -8.9 49.0 -0.5 Non-agric WPI/cost of living -32.8 -6.9 23.6 -7.4 Non-agric WPI/cc. struction cost -7.2 1.0 15.1 -4.2 Cereals-oilseeds/private services -37.6 11.0 84.0 -21.8 Source: IEERAL Data Base. Table 5 Manufacturing Wage, Exchange Rate, and Product Wage Index, 1988/III - 100 Wage/Exchange Product-wage/ ratel gains2 1987 III 121 125 IV 111 116 1988 I 114 125 II 109 109 III 100 100 IV 111 116 1. Manufacturing nominal hourly wage divided into the excbtnge rate at which imports are traded. 2. Ratio of wage cost (ratio between the nominal hourly wage paid in manufacturing and the corresponding wholesale product price) to productivity (output per man-hour). Source: IEERAL Data Base.  14 Figure 2 Real Wage and Product Wage in Manufacturing Index, 1988/11I - 100 124. 12U Ile 1146 1126 1140 10U 104 102 100 */6 lv V/U o h I r * Rti gos + W^wEp/"haigm r* o Zb 4uim  15 III. T STRUACT NRE AD TRISIDS OF LABOR MARKETS 113.1. MaM Trends Observed labor market results are paramount in analyzing the import- substituting policy followed by Argentina. A wage setting mechanism based on a deliberate policy of overvalued exchange rates, and restricted by the *trade balance/ domestic growth trade-off, was the driving force of labor markets. The basic policy tool was government intervention supported by urban-based unions and political groups. The main observed outcomes were a mild real wages growth over the long run, large short-run economic fluctuations and distorted relative wages among productive sectors. Between 1940 and 1985, real wages neither rose nor fell for more than three consecutive years. In all but one case (1969-1971), periods of growth in real wages were followed by periods of sustained decline. By 1985 real wages were only 61 percent higher than in 1940, thus implying an average yearly growth rate slightly higher than 1 percent (Sanchez, 1987; Riveros, 1989). If the shorter period 1962-1987 is considered, the evidence more than confirms the wage deterioration over time (See Table 2 and figure l.a): after a growth period in 1962-74, real wages declined much more than real per capita GDP. In addition, the magnitude of short term fluctuations increased. For instance, in 1962-74 real wages rose at a yearly rate of 3.1 percent, to subsequently drop at an even higher rate (-4.4 percent). Over the whole period 1962-87, real wages fell by an average of 0.8 percent per year. The growth in real per capita GDP seen in the period 1962-1974 was positively associated with high and growing FTT. In the context of the inward-oriented growth strategy, this made it possible to devalue the exchange rate at a pace lower than inflation, without concurrently causing a balance of  16 payments crisis. This was the usual mechanism used to enlarge domestic markets, and therefore to increase real GDP and real wages in the short run. As soon as FTT growth began to decline, as in 1973, a balance of payments crisis arose making unavoidable a devaluation and a decline in wages. Thus, in the context of a deliberate policy of overvaluation, FTT can be seen as a determinant of the RER (see Figure 3).4 The observed trend in relative wages is an outcome of the inward- oriented growth strategy. Domestic market-oriented growth required relative prices favorable to urban activities and high purchasing power for wage- earners. Argentine population is mostly concentrated in a few urban centers, while services, construction, and import-substituting industries produce and sell most of their output in these markets. Thus, the evolution of relative wages from the 1940s to the 1980s has clearly favored labor in non-tradable activities (Sanchez, 1987; Riveros, 1989). Moreover, observed wage changes did not reflect changes in labor productivity in non-tradable activities (Sanchez, 1987) and did it only mildly in manufacturing. Table 6 and figure 4 show the evolution of relative wages over the period, 1962-87. In order to avoid problems of interpretation associated to the peculiar behavior of public sector wages after 1985 (sharp wage cuts due to government stabilization programs), they are not included in the group of non-tradables. Relative wages seemed to change in favor of tradable activities only in 1984. Over the entire period, however, the stability in relative wages is notable, probably reflecting more government intervention than a relatively stable average labor productivity.  17 Figure S Real ExchaeI Rat. (RUR) and FT? to 120 110 100 90 70- 50 40 1962 1965 1980 1971 1974 1977 1980 1963 19S 0 R + FTr Following the lines of the overall wage strategy, the wage determination system has been a crucial determinant of the observed stability in relative wages. In fact, collective bargaining takes place at national level and with significant government intervention. Thus, it has been generally simple to bias the results in favor of political clients, and to promote wage growth regardless of the situation of the firms. Given its effects on relative wages, this wage determination system has also been a source of poor inter-sectoral labor mobility (Riveros, 1989).  18 Table 6 The Evolution of Relative Wases (1962-1987) Year VTIWU WTIVW Year VTIWU WPT/VN (1) (2) (1) (2) 1962 0.83 0.99 1974 0.93 1.05 1963 0.79 0.95 1975 0.92 1.02 1964 0.82 0.96 1976 0.93 1.04 1965 0.88 1.01 1977 0.94 1.04 1966 0.90 1.03 1978 0.91 1.01 1967 0.89 1.05 1979 0.93 1.04 1968 0.88 1.03 1980 0.89 0.99 1969 0.88 1.04 1981 0.91 1.03 1970 0.89 1.04 1982 0.91 1.03 1971 0.91 1.04 1983 0.95 1.05 1972 0.91 1.03 1984 0.96 1.10 1973 0.88 1.00 1985 0.97 1.10 1986 0.96 1.12 (1) Wages in Tradable Industries to Wages in Nontradable industries. Tradables include agriculture and manufacturing, Non-tradables include construction and services excluding government. (2) Wages in Potentially Tradable industries to wages in Non-tradable industries. Potentially-tradables include manufacturing. WN is defined as in (1). Source: IEERAL Data Base. ri1.2 Regional and Sectoral Labor Allocation Important population concentration in a few urban centers is an outcome from both observed trends in public investment and social expenditures, as well as distorted relative output prices and the skilled- unskilled wage gap among regions (which, in turn, resulted from the centralized nature of the wage-setting process). Although population growth has been traditionally low in Argentina, the urban population has grown much more rapidly, and rural population growth rates have been negative for many years (see table 7).  19 Figure 4 The 3volutios of Relative WAges 'a us -ur t_r mm| tin, tm4 1u t_r t_ t1_ (a) WPT/WN (b) WT/WN Source: IERRAL Data Base. According to 1980 figures, the urban population represented 83 percent of the total population. In 1988, of the about 33 million inhabitants of the country, 42 percent lived in Greater Buenos Aires, C6rdoba, and Rosario, and 35 percent in the two former cities. This pattern of demogrsphic concentration is likely a consequence of the deliberate policy of overvalued exchange rates, (and of the accompanying commercial and macroeconomic policies) which affected relative prices and wages and biased labor allocation against tradables. Any increase in the average wage of the urban economy (such as government mandated across-the-board increases, or union demands leading to similar results) increased wages in non-tradable sectors more than  20 labor productivitw. Thus, prices in this sector had to grow at a higher rate relative to that of tradable sectors in order to maintain relative wages at their targeted levels. The final outcome was that the change in relative prices necessary to obtain a glven increase in the average wage of the urban economy increased over time. Table 7 Average Annual Rate of Population Growth (percentages) Census Greater Periods Total Urban Rural Buenos Aires Cordoba Rosario 1914-1947 2.1 2.5 1.4 m.a. m.a. m.a. 1947-1960 1.8 2.8 - 0.5 6.1 3.6 2.4 1960-1970 1.6 2.5 - 1.4 3.8 3.0 2.0 1970-1980 1.8 2.6 -0.03 2.5 2.2 1.6 Source: C. Stnchez (1986) tables 1 and 2. In addition to the bias in favor of non-tradable sectors derived from aggregate economic policies and government intervention investment in education, health, housing, transport, communications and culture was also concentrated in urban areas. This allocative pattern was reinforced by social policies which were characterized by subsidies that instead of supplement.ng the income of the poor regardless of where they lived, provided free-of- charge public services to persons living in the largest cities, regardless of their income (SAnchez, 1986,1987). Legal regulations on nominal wage setting also contributed to the concentration of resources in urban areas. Nominal wages, whether determined by legal procedures for bargaining or established by the government in periods when union activity was prohibited, were compulsory for all workers and firms.  21 Therefore, differences in the relative availability of labor in different regions did not affect the level of wages. At the same time, this wage policy discriminated against small firma and the marginal investor, thus also restraining the creation of modern activities in areas of the country apart from the largest cities and producing a one-way flow of migrants from regions to urban centers. This latter result is consistent with economic theory. A disaggregation of labor into two groups, skilled and unskilled, reveals that the regions have excess demand for the former and excess supply of the latter, while the opposite occurs in the wealthiest urban areas. Wage differentials among regions should move accordingly and provoke a flow of skilled workers towards the regional markets, and a flow of unskilled workers in the opposite direction. For this to occur, wage differentials would have to be sufficiently large in order to compensate for the cost of moving in either direction. Thus, labor flows did not occur as expected because: a) Even though wage drifts existed, they were not large enough to fully compensate for the rate of institutional wage intervention in equalizing wages among regions. Therefore, wage differentials were lower than those that would have prevailed in a decentralized and more competitive wage-setting process; b) Welfare policies and related social expenditures increased the cost of moving for the skilled, and reduced it for the unskilled. If the former decided to move, they would lose the relatively inexpensive or free-of-charge provision of social services in the cities. But unskilled workers did not have to bear such a cost: if they moved they would have been able to receive larger subsidies in urban areas.  22 The resulting sectorai labor allocation was as expected. The share of employment in commerce and services in total employment was almost 50 percent in 19t0 and more than 50 percent in 1980 (Table 8). The non-tradable sector accounted for 69 percent of all the new employment created between 1947 and 1980 (85 percent in 1960-70 and 87 percent in 1970-80). Finally, the share of the non-tradable sector in total employment was 64.1 percent in 1980 compared with 49.6 percent in 1940. In contrast, the share of agriculture in total employment declined by more than half during 1940-80, from 27.1 to 12.9 percent, and the share of manufacturing declined steadily since 1947. 111.3 Labor Market Se2 entation and Increasina Informality Economic growth resulted in a notable employment expansion in services and construction relative to that in tradable sectors, and a higher increase in non-wage than wage employment (Table 8). Following a standard approach to analyzing labor markets in LDCs, urban employment can be classified into the broad categories of formal and informal. This classification depends on the existence of protection (or lack of protection) of certain sectors with regard to the coverage and enforcement of labor regulations. Thus, formal employment corresponds to the public sector and to private sector firms employing five or more persons. Informal employment corresponds to unorganized private sector firms employing fewer than five persons. Since not all the workers in the second category earn a "low" income, informal employment may be further classified on the basis of income levels: those earning high incomes due to advanced skills, high capital intensity or an oligopolistic market environment are classified in the 'quasi-  23 formal sector. The rest having low incomes are classified in the informal sector (Sinchez, Palmieri & Ferrero, 1981). Table 8 Characteristics of the Sectoral ZmDloyment (percentages) Sectoral Distribution Average Employment Growth p.a. Agri. Manu. Total Agri. Maniu. v nw N-T w nw w nw 1940 27.1 23.3 49.6 na na na na na na 1947 26.1 27.2 46.7 0.1 na 7.2 na 3.8 na 1960 20.8 27.1 52.1 -1.6 1.3 1.2 2.2 1.9 3.6 1970 16.2 23.3 60.5 -0.6 -2.1 0.4 -3.6 2.6 2.8 1980 12.9 23.0 64.1 -0.9 -0.6 1.1 2.6 1.7 3.2 Agri-Agriculture; Manu-Manufacturing; Cons-Construction; Comms- Commerce; N-T- Non-Tradables; w- wage employment; nw- non-wage employment. Data on sectoral distribution are expressed as a proportion of total employment. Source: Taken from C. Sanchez (1984), table 8. The proportion of quasi-formal and informal workers in Argentina-- which is equivalent to the concept of informal sector used elsewhere -- is of about 302 of the urban work force. Therefore, in relation to other Latin American countries where informality accounts for 40 percent or more of the labor force, informality is not as important in urban labor markets of Argentina. Nonetheless, given that formal wages are protected vis-a-vis the informal sector through several labor regulations, as well as government and union intervention, adjustment policies have increased the formal-inforral wage gap (Lopez & Riveros, 1989). This result is associated with the presence of a relatively higher wage rigidity in formal activities and low labor mobility. In addition, the increase in the formal-informal wage gap as a  24 result of a typical macroeconomic adjustment program; probably is a negative factor affecting its political sustainability. Lack of macroeconomic adjustment has probably been at the root of relatively high income levels observed in the informal sector, which is mostly a producer of non-tradables. For example, statistics for 1984 and 1985 show that self-employed workers with less than 8 years of schooling (a reasonable proxy for the unskilled) earned incomes that were approximately 18 percent higher than those earned by wage-earners with the same level of education. Indeed, average self-employed income was only 7 or 8 per-ant less than the national average wage.5. However, if one uses the concept of total labor costs, to account for all incomes received by wage labor in the formal sector, a different conclusion may arise, particularly in considering that in 1985 non-wage labor costs (considering fringe benefits, social security and regular bonuses) were about 51 percent of total wage costs (Riveros, 1989). III.4 Labor Market Institutions The role o' labor market institutions is important in leading to the formal/informal distinction which is, in turn, a key factor with regard to the labor market response to macro policies. The Argentinean economy is characterized by politically strong oligopolistic firms producing for domestic markets that have created powerful protectionist lobbies. Their counterpart is an equally powerful labor union which is organized as corporations, whose influence on government policies has been paramount.  25 Although some changes in the structure of power in the labor movement have occurred during the last decade in Argentina, unions are still extensive, disciplined, and politicized. The law allows workers to freel.y organize unions. But in order to obtain the legal right to bargain with employers, a credential of 'trade- union representativeness' (personeria gremial) is required. This is granted by the government to the most important union (mainly in terms of membership) in each area of activity, although exceptionally, and mostly due to political reasons, trade union representativeness has been granted to more than one union. Even though the right of a worker in order to join or not to join a union is protected by the law, the terms of the labor contract agreed to by the representative union and the employer are compulsory for all workers in the activity. In turn, unions may join federations and confederations which, once obtained legal representativeness can negotiate wages at a very aggregate level. In this manner, wages and working conditions are centrally determined by negotiation between national unions and entrepreneurs, process also characterized by strong government intervention. In periods when collective bargaining was not in practice, wages have also been centrally determined by government decrees and resolutions. Job security regulations are considered an important factor hindering interindustrial labor mobility. The law states the right of the worker to keep his job regardless of the circumstances, although it does not establish anything about income maintenance. The law also states that in case of dismissal, a severance payment equal to the highest monthly wage or salary  26 earned by the employee in his current job, multiplied by the number of years worked has to be paid. However, this highest wage cannot be higher than three times the legal minimum wage. Mandated minimum wages are another important institution exerting several effects upon labor market outcomes. First, they set a limit for severance payments, thus affecting the normal rate of job turnover. Second, they affect prevailing equilibrium wages due to the effect of minimum wages in shifting the entire wage structure up (Paldam & Riveros, 1988; Sanchez & Giordano, 1988). Third, given the positive effect of minimum wage changes on average wage changes, the former are also related to existing inflationary pressures (Paldam & Riveros, 1988). Fourth, evidence suggests that relatively high minimum wages positively affect the formal-informal wage gap and negatively the employment level in formal activities (Sanchez & Giordano, 1988; Lopez & Riveros, 1989).6 IV. THE ROLE OF LABOR HARKETS IN THE ADJUSTMENT PROCESS IV.l. The Performance of Labor Markets in Adiustment As discussed above, the two most recent attempts to introduce structural adjustments in Argentina were unsuccessful. Interactions among policies and relative prices affecting labor markets operation can be seen in Tables 2, 6, and 9. Despite a series of nominal devaluations, the wage- exchange rate policy mix followed systematically used between 1962 and 1983 resulted in wage rigidity and increasing non-tradable prices relative to the price of exportables.  27 As discussed earlier in section II, there has been a sequence of devaluation-recession periods followed by periods of exchange rate appreciation. During the latter, wage indexation normally hindered the intended adjustmeat process initiated through an usually severe devaluation. Implementation of the deliberate policy of overvalued exchange rates increased urban prices and activity levels relative to tradables, affecting mostly rural production. Union activity and minimum wages, were the basic driving forces used for the maintenance of high urban wages. Real wage increases were usually attained first in the manufacturing sector (where unions acted as the leaders of the entire urban labor movement), and then propagated to services and construction.7 Nominal devaluations were always the final outcome of periods of policy-induced urban expansion. Both the length and the intensity of periods of overvaluation and urban expansion were highly correlated with observed hikes in FTT (see, for example, Figure 3). When the FTT were at a low level, it was not possible to maintain an overvaluation anymore without causing an external imbalance. The strong relationship between the exchange rate policy and relative wages can be more clearly seen in the following averages, estimated from tables 2 and 6 (Index 1962172 - 100). Reform period Wt/Wn Wn/Wn RER 1973-1975-1976 105 101 83 1976-1980 106 101 80 1981-1983 106 103 92 1984-1987 113 111 112 Wts wages in agriculture and manufacturing. Wn: wages in non-tradables. RER: real exchange rate.  28 Table 9 Argentina: lelatlve Prices and GDP, 1962-87 Pu/Pt Pu/Px Pn(pm W/e Wh/Wh GDP 1962 91 98 78 57 0.99 69 1963 89 92 78 58 0.95 67 1964 84 82 73 79 0.96 74 1965 87 92 72 90 1.01 81 1966 97 106 87 98 1.03 82 1967 98 106 90 78 1.05 84 1968 102 109 97 80 1.03 87 1969 103 110 100 90 1.03 95 1970 100 100 100 100 1.04 100 1971 96 89 99 114 1.04 105 1972 87 79 186 97 1.03 108 1973 95 88 99 149 1.00 115 1974 102 104 102 211 1.05 122 1975 108 138 95 219 1.02 121 1976 99 123 83 103 1.04 121 1977 101 111 94 114 1.04 128 1978 110 118 112 166 1.01 122 1979 111 114 116 289 1.04 131 1980 125 134 135 487 0.99 133 1981 127 150 131 391 1.03 125 1982 103 129 89 196 1.03 118 1983 101 127 85 199 1.05 121 1984 106 134 92 254 1.10 124 1985 113 164 95 174 1.10 118 1986 124 138 130 199 1.12 125 1987 126 137 138 177 1.17 127 Pu =urban price, or price of goods and services produced by the import substitution sector (manufacturing) and the service (including commerce) and construction sectors; Pt =price of tradables, i.e. price of goods produced by the exportable and by the importable sectors; P. -price of exportables, i.e. goods produced by the rural sector and by the food and beverage sectors; Pm -price of importables, i.e. the price of goods produced by the import substitution sector and the price of imported goods; Pn -price of non-tradables, i.e. the price of goods and services produced by the service and construction sectors; Wh -wages and salaries paid in the import substitution or home goods sector; Wn -wages and salaries paid in the non-tradable sector. e -nominal exchange rate. Source: IEERAL Data Base.  29 In spite of the structural adjustment program announced in 1976, the relatively low RER prevailing from 1973 to 1980 held the ratio of wages in tradables and nontradables practically constant. This suggests that the effects of the wage-exchange rate policy mix were more important than those of the trade liberalization and other adjustment measures announced in mid- 1970s. Later on, the 1987 adjustment program resulted in an immediate increase in relative prices of urban tradables. It was apriori expected that in response to this change in relative prices wages in non tradables (W.) would have increased relative to manufacturing (Wh), thus creating incentives for labor reallocation. As discussed in section II, this did not happen due to a series of policy reversals. The conclusions to be drawn from these two adjustment attempts suggest that labor markets variables achieved a negative role regarding expected structural changes. While the main policy target was a change in relative prices of tradables to nontradables to allow a shift of the economy towards production for external markets, a wage indexation mechanism supported by government and union intervention created substantial rigidities. Since such a mechanism implies an increase in the relative price of urban goods relative to exportable goods -- and in the price of non-tradables relative to importables -- a loss in international competitiveness inevitably resulted. Likewise, each of these adjustment attempts worsened inflation, the resource allocation across industries and the poverty situation. IV.2. Level and Composition of Aagregate Employment Traditionally, Argentina did not have employment problems resulting from an excess of labor supply. The situation was traditionally one of labor  30 scarcity, especially in the least skilled segment of the labor force (Llach, 1978). Nevertheless, urban labor markets (specially in the largest cities) experienced successive periods of scarcity as well as of relative abundance of labor, due to fluctuations in domestic and foreign immigration to large urban centers, and in high labor force participation.8 A situation of relative labor scarcity in large urban centers Was seen in the second half of the 1970s. Migratory flows changed after 1970 when the largest urban markets were no longer the recipients of large number of workers in search of better job opportunities. Rural-urban migration continued in the second half of the 1970s but from rural areas of each province to its capital city as the final destination (Sanchez, 1984,1986). In addition, a series of institutional and economic developments, particularly the adjustment program implemented from 1976 to 1981, reinforced the falling trend observed in labor force participation: the real wage was probably below the reservation wage of many labor market participants and a considerable number of women, young adults, and even 20-59 year old males abandoned the labor force. The period from the early 1970s to the early 1980s not only witnessed declining labor force participation rates and employment (table 10), but the sectoral allocation of labor tended to diminish the supply of wage labor to the goods-producing sectors. Workers were transferring to construction and service activities and to non-wage occupations thus affecting the availability of labor in the industrial sector (see table 8). Although relative sectoral wages did not chow a well defined trend (see table 6), employment creation in manufacturing and agriculture was poor.  31 Due to these economic trends, and to the withdrawal effect, open unemployment was not a significant problem in the urban labor markets until the early 19809. But the situation has been changing during the 1980s and an open employment problem has arisen. It was stated above that labor force participationi rates moved in the same direction as economic activity in the short-term, but there have also been long run trends: considering the period from 1974 until May of 1988, table 11 and figure 5 show that participation r&tes decreased through 1983, while open unemployment and underemployment remained low compared to their historic levels (table 10).9 Labor force participation has ircreased over the last five years despite high variability in output growth and the recent slowdown in activity levels (Table 11, Fig. 5). The reason is that the labor force response to the existing labor market situation is now different than prevailing until the late 19709. The 'discouraged worker' effect was predominant in those years as declines in employment and real wages were accompanied by withdrawal of people from the labor force (including migrants from nearby countries). Currently, however, the dominant effect derives from the loss of family income and shows up in the increase in the net inflow of secondary workers into the labor market. Of course, if the probabilities of finding a job are low in the case of experienced workers (for example males of 20-59 years of age), they are even lower in the case of secondary workers. The result in those trends has been an increase in open unemployment and underemployment in association with the laddedm worker effect.10  32 table 10 AraentinatIndlcators for Main Urban Labor Markets LFP L U LFP L U 1950 46 44 4.3 1968 44 42 5.3 1951 46 44 3.5 1969 44 42 4.4 1952 45 43 4.0 1970 44 42 5.0 1953 45 43 4.8 1971 43 41 5.9 1954 46 43 6.2 1972 42 40 6.7 1955 46 43 5.8 1973 41 39 5.5 1956 46 43 6.8 1974 40 39 3.9 1957 46 43 6.3 1975 40 39 3.2 1958 46 43 7.8 1976 39 38 4.7 1959 44 42 5.5 1977 39 38 3.2 1960 45 42 5.6 1978 39 38 3.0 1961 45 42 7.2 1979 39 38 2.2 1962 45 42 7.3 1980 39 38 2.5 1963 45 41 8.9 1981 39 37 4.7 1964 45 42 6.3 1982 39 37 4.9 1965 45 43 5.3 1983 38 36 4.4 1966 45 42 5.8 1984 38 36 4.2 1967 46 41 6.4 1985 39 36 5.6 LFP: Labor Force Participation expressed as a percentage of total population. L : Employment expressed a percentage of the economically active population. U : Unemployment as a percentage of the labor force. Source: Taken from C. Sanchez (1987). 1950-1962: Computed from population and census data. Employment data were taken from Llach & Sanchez (1984). 1963-1985: Data from the PHS.  33 Table 11 Labor Partclipation and UnemPloyment (Greater Buenos Aires, 20 capital cities, and 5 non-capital cities) Labor Force Open Participation Unemployment Underemployment Equivalent Rate Rate* Rate* Rate* 1974 40.4 4.2 5.0 6.7 1975 39.9 3.7 5.3 6.4 1976 39.3 4.8 5.3 7.5 1977 38.7 3.3 4.0 5.3 1978 38.9 3.3 4.7 5.7 1979 38.3 2.5 3.8 4.4 1980 38.4 2.5 5.2 5.1 1981 38.4 4.8 5.5 7.6 1982 38.3 5.3 6.6 8.6 1983 37.3 4.7 5.9 7.7 1984 37.9 4.6 5.7 7.4 1985 38.1 6.1 7.4 9.8 1986 38.9 5.2 7.3 8.9 1987 39.2 5.6 8.1 9.9 1988 39.4 6.1 7.9 10.1 * As percentage of the economically active population Source: Permanent Household Survey (PHS). See definitions in endnote no. 7.  34 Figure 5 Labor Participation and lauivalent' Une.loyment 5.a. Labor Force PartLcipation Rate (as percentage of total population) *1 4-a Mi - -A - 5.b. Equivalent Unemployment Rate (as percentage of econto ically active popla-tlon) 4_ r I t1  35 IV.3. Unemploment and the Public Sector The previous analysis suggests that construction and services of low productivity are characterized by substantial hidden unemployment. Another sign of hidden unemployment are observed public sector employment trends.In fact, estimates made in 1985 indicate that total public sector employment was 2,009,000 people representing as high as 17.5 percent of the labor force and 25 percent of all wage-earners (Table 12). Table 12 Public Sector Dmloyment (1985) Relative Share Cfl Public Total Sector Labor Wage (in thousands) employment force earners National Administration 605 30.1 5.3 7.5 Public Enterprises 398 19.8 3.3 4.9 Local Government 1006 50.1 8.9 12.6 Total 2009 100.0 17.5 25.0 Source: Sanchez & Giordano, Proyecto Gobierno Argentino/PNUDIOIT, ARG/87/003 (1988). Rapid growth of public sector employment in the provinces has been notable in Argentina. Table 13 shows that public employment grew by more than 40 percent during the period 1960-1987. Most of this growth was associated to a huge increase in employment in local governments (more than 200 percent), which took place in periods when unsuccessful adjustment programs were underway. Thus, at the time that overall adjustment failed, partly as a consequence of labor market intervention policies and the implementation of the wage-exchange rat.3 policy mix, resulting low employment growth demanded active job creation in the public sector.  36 In addition to the political commitment of keeping low open unemployment, high employment growth in local governments was associated with changing migratory patterns. Domestic migratory flows changed after 1970 as economic incentives to move to the largest cities declined. Instead, population began to flow from rural areas to the capital of the provinces, where the probability of finding employment was higher. Mostly due to the resulting demands for more employment, local governments started active job creationl1, thus also being char&cterized by the presence of hidden unemployment. Another problem with public employment refers to itu internal composition. In 1985, about 20 percent of publi_ employment was in public firms (in many cases with very low levels of labor productivity, as in the case of railroads). In contrast, less than 30 percent is employed in areas providing services like health and education. Thus, the relative amount of workers in purely bureaucratic activities is high, possibly demanding a thorough public sector reform program. Table 13 Public Sector Employment Growth; 1960-87 (Index, 1960=100) -----------------------------------------------------..----__----------------_ 1960 1965 1970 1975 1980 1985 1987 National Administration 100.0 93.7 100.1 110.3 97.6 106.3 104.5 Public Enterprises 100.0 81.1 76.6 101.4 73.9 80.5 79.1 Provincial Governments 100.0 105.7 131.7 178.1 205.7 239.0 259.7 City Governments 100.0 116.0 134.4 181.6 179.2 202.4 220.0 Total 100.0 94.4 103.2 129.2 122.0 136.9 141.7 Source: 1960/1985, see table 15 1987, estimated from figures published in El Cronista Comercial, April 27, 1988, p. 15.  37 IV.4 Womn. Labor Markets and Adlustmeat A sustained increase in the rate of female labor force participation has been observed in the 1980s. The most notable increase was observed in women in the age group 35-49. The drop in family income due to the economic crisis and the need to compensate for the associated loss in welfare is a possible explanation for the higher participation of secondary labor force. As in the data presented below, the main sources of employment for the rising female labor supply were nontradable activities: Sectoral Composition of Female Employment (2) 1960 1970 1980 Total 100 100 100 Tradable 32 25 21 Agriculture 5 4 3 Industry 27 21 18 Nontradable 68 75 79 Sources Estimated from census data. Examination of the composition of female employment reveals a concentration in relatively few occupations. In general, increasing female employment has concentrated in nontradables, possibly in occupations of low productivity. In 1970, approximately 58 percent of the total female employment was in clerical positions, sales and other services (mostly domestic services). In 1980 this figure was 63 percent. As expected, the female average wage was much lower than the male wage: according to PHS figures, it was approximately 56 percent of the male wage.  38 IV.5. Income Distribution and Povert, Absence of structural adjustment in Argentina has been associated with an increasing deterioration in income distribution. This deterioration derives from poor economic growth in turn resulting from the disarray of policies that have produced higher relative prices of non-tradables and protective regulations favoring urban workers in the formal sector. Examination of the poverty situation by economic sectors is less satisfactory than analyses based on geographic or functional characteristics or the formal/informal sector distinction. Unfortunately, however, data (Permanent Household Surveys: PHS) are not available to study the latter disaggregation. In addition, data are limited to the cities of C6rdoba and Buenos Aires.12 Finance, a nontradable sector, and Industry, a tradable sector enjoy the highest income levels per worker (higher than average), the lowest unemployment rates (lower than the overall rates for each city), and the lowest levels of poverty (Table 14). Commerce, Services and Construction, all nontradable, not only reveal relative incomes lower than the other two sectors, but also a significantly higher level of unemployment and underemployment. The result is a large incidence of poverty in each sector. An important characteristic of nontradable sectors as a whole are incomes slightly lower than the average, high unemployment rates, and consequently a greater than average incidence of poverty. This is concordant with a model of the labor market where the informal sector is mostly a producer of non- tradables and an employer of unskilled labor. If this is a neoclassical labor  39 market, a succession of failed nominal devaluations combined with wage rigidity in formal activities must have produced low incomes in informal activities and relatively higher unemployment. Table 14 Unemploynent. Sectoral Income Distribution and Poverty Equivalent X of Poor in Relative income Unemployment Rate 2 Total Employment Sector BUE COR BUE COR Poor Indigents Total 1.0 1.0 9.4 8.9 14.6 3.3 Finance 1.4 1.6 7.1 5.1 9.6 2.7 Industry 1.1 1.1 7.1 5.5 11.7 1.8 Commerce 1.0 1.0 11.7 11.1 13.0 4.3 Services 0.9 0.9 11.2 10.6 15.5 4.5 Construction 0.8 0.8 15.3 14.5 25.2 4.6 Tradables 1.1 1.1 7.1 5.5 11.6 1.6 Nontradables 0.9 0.9 10.2 10.0 15.6 4.3 Definitionst BUE: Greater Buenos Aires; COR: C6rdoba. Relative income includes wages and the income of the self-employed for C6rdoba, and only wages for Buenos Aires. Equivalent unemployment rates see endnote 9. t of poor: data refer only to C6rdoba. The steps followed to identify the poor were: a) Size and composition of each household were identified; b) Each pmember of the household was transformed into a 'standard adult' according to a calorie intake table (see IPA-INDEC, 1988); c) Average per capita income was estimated and compared with the 'poverty line; d) The poverty line was defined as the income level necessary to buy a basic basket of food (IPA-INDEC, 1988) and to cover other basic needs; e) Household members with per capita below the poverty line were classified as poor. Indigent are those with income below the amount required to buy the basic basket of food only. Source: Own estimates based on PHS data: October 1986 for Cordoba, and April 1987 for Buenos Aires. The rising poverty in Argentina during the last 15 years is mostly a result of failed adjustment. In addition to rigid wages and distorted labor allocation, open unemployment and underemployment have been growing for the last S years. Young males and females in the age group 35-49 are the most  40 significant group of secondary workers swelling the labor supply to the informal sector. Most likely, this surplus labor usually takes refuge in construction, commerce, and services, activities which have very low productivity rates and where unemployment and poverty are higher. At least in the short and medium term, any adjustment process will inevitably increase poverty due to higher transitional unemployment and the drop in real wages. Adjustment will also imply declining public sector employment an wages, as well as expenditure reduction in current social welfare programs. Until now, however, Argentina has been suffering the social cost of not-adjusting and trying to keep an inappropriate productive structure combined with intervention in the labor market and use of a deliberate policy of overvalued exchange rates. IV.6 Political Economy, Labor Markets and Adiustment Hyperinflation is now forcing the government to attempt a thorough adjustment program, that includes profound changes in the functioning and institutional organization of the labor and goods markets. Deregulation. privatization, reduction in public expenditures, trade liberalization, a shift of factors of production against the non-tradable sector, and other similar actions are not without cost. Therefore, a set of labor market policies are required in order to facilitate an efficient change at a minimum cost. Severe difficulties will be faced in shifting labor among sectors. Reforms in the labor market should be designed to facilitate the movement at minimum cost, which would need deregulation policies. Therefore, a set of policies should aim at decentralizing collective bargaining and the structure of labor unions. Another set of policies should aim at making labor contracts  41 more flexible in order to attain higher labor mobility and to implement feasible programs designed to assist the unemployed. Finally, the application of macroeconomic policies should be done in combination with a dismantling of labor market intervention, which is probably going to diminish labor market segmentation and increase the trickle down effect of growth. The program of economic reforms needs to be sustainable and credible. In Argentina, those conditions require strong political support and a clear commitment on the part of the government. In turn, this requires a clear engagement of unions with the process of economic reforms, or & substantial change in traditional political attitudes of unions. This is not going to be an easy task not only because the enormous lags that may be involved in changing institutions, but also because unions are aware that structural adjustment would imply, one way or another, a loss in their power. V. CONCLUSIONS Although severe difficulties in implementing and sustaining a program of structural adjustment, change is inevitable in Argentina. The current economic crisis is not only the result of inappropriate domestic policies to cope with recent external shocks. Even if no external shocks had occurred, the country would still have to face the urgent need to shift the structure of production in appropriate directions. The current economic crisis is a result of wrong policies applied for years that have damaged the economy. Frustrated adjustment experiences provide some lessons that can assist policy makers in designing a sustainable program to achieve price stability and to change the system of incentives.  42 1. The consistency of macroeconomic and trade policies has to be ensured. In particular, trade policy must be combine with fiscal reforms since reducing public spending would exert downward pressures on the real exchange rate. 2. The importance of implementing changes in labor relations and labor market institutions also has to be emphasized. Institutional changes, such as the decentralization of wage bargaining snd the elimination of both wage indexation and government intervention at large, are necessary if resources are to going to shift among sectors and regions. 3. Shifting resources among regions and industries requires elimination of rigidities and restrictions on labor mobility. Restrictions resulted from the regional allocation of public investment and social welfare expenditures, as well as from regulations on public services and subsidized urban utilities. Thus, more mobility will require less pervasive intervention. 4. Another important lesson from the past is that the prospective losers in an adjustment program may have sufficient power to stop its implementation. It is therefore very important to try to incorporate these losers into the benefits of the program. In addition to implementing the changes in policies and institutions previously mentioned, a profound switching of public expenditures is necessary to reduce the social cost of transition period. This will require external financing and debt alleviation.  43 5. In uuumary, the main lesson from the past is that the nature and extent of the changes require a deep reform program which will probably not be feasible without government credibility and the commitment of all concerned groups. This challenge for the Argentine economy require deep labor market reforms, and probably a high transitional cost.  44 NOTES 1. See Nogues (1981), Cavallo & Mundlak (1982), Cavallo (1986), Cavallo & Cottani (1986), Sanchez (1987) and Cavallo & Domenech (1988). 2. The standard deviation increased from 28.6 to 39.3 percent and the coefficient of variation from 0.55 to 0.88 (Cavallo and Cottani, 1986). 3. 1980 quarterly figures show that inflation was 17 percent while the rate of devaluation was only 6 percent. 4. Cavallo & Domenech (1988) have modelled the behavior of the RER dependirg upon the FTT, the trade policy (taxes on imports and exports), the income level, and the macroeconomic policy. 5. It must be noted that the average income of the self-employed as a whole was approximately 23 percent higher than the average wage of the economy (Permanent Household Survey- PHS). Evidence obtained from the PHS indicates that the relative income of the average self-employed is higher than that of the wage earner, and that the differential tends to increase when the exchange rate lags behind wages. 6. Workers receiving a wage not higher than 1.5 times the legal minimum represented in 1981-1982 less than 20 percent of the labor force. However, in construction, commerce and services the specific proportion varied from 21 percent to 28 percent. The proportion of women, 27 percent, was also higher than that of men, 18 percent. See Sanchez and Giordano (1988), table 6. This suggests that minimum wages are more binding in non-tradables and in informal sector activities. This probably affect prices of non-tradables and spurs illegal methods of hiring in formal activities. 7. Llach (1987) has estimated sectoral productivity figures and concluded that urban sector development is based much more on agricultural productivity than on its own. 8. This adjustment mechanism in urban labor markets has been extensively studied since 1979 when the first paper by Sanchez, Ferrero & Schulthess was published. See for example Beccaria (1980), Beccaria & Orsatti (1985), Mann & Sanchez (1984), Dieguez & Gerchunoff (1984), Llach (1980), Sanchez (1982), SAnchez (1987) and Riveros (1989). 9. Open unemployment is defined as the ratio of those identified as unemployed in the Permanent Household Survey (PHS) to the economically active population. Underemployed are those working less than 35 hours per week and seeking for more hours of work. If the convention is accepted that two underemployed are equivalent to one unemployed, and these 'converted' unemployed are added to open unemployed, "equivalent unemployment rate" can be calculated.  45 10. Analysis of figures for Greater Buenos Aires reveals that from October 1983 to October 1988, the overall labor force participation increased more than 7 percent and that, when the statistics are disaggregated by age and sex, they show a substantial increase in female participation, especially in the 35-49 year age group. Increases in male labor force participation were small, with the exception of relatively younger groups. Thus, this reveals a vital change in the participation of secondary labor force. (see Riveros (1989)). 11. For a detailed analysis of this mechanism of labor absorption in the provinces, refer to Sanchez (1986) and Lindenboim (1985). 12. The conclusions obtained cannot be extended to the rest of the country unless it is assumed that these two cities are representative of the overall situation. However, given the large concentration of population in the urban areas, and the large share of these cities in the urban total as was discussed in section III.2.1. this seems an acceptable assumption.  46 REFERENCES W. Baumol (1967) "Macroeconomics of Unbalances Growth: The Anatomy of Urban Crises' American Economic Review, 57 (415-426). L. Beccaria (1980) "Los Movimientos de Corto Plazo en el Mercado de Trabajo Urbano y la Coyuntura 1975-78 en la Argentina" Desarrollo Econ6mico, 20:78. L. Becaria and A. Orsatti (1985) "La Evoluci6n del Empleo y las Salarios en el Corto Plazo. El Caso Argentino 1970-1983." Documento de Trabajo N