Mexico - Earnings inequality after Mexico's economic and educational reforms (Vol. 2 of 2) : Background papers

Report No. 19945-ME Mexico Earnings Inequality after Mexico's Economic and Educational Reforms (In Two Volumes) Volume II: Background Papers May 16, 2000 Mexico Country Management Unit Poverty Reduction and Economic Management Division Latin America and the Caribbean Region Document of the World Bank  CURRENCY EOUIVALENTS Currency Unit - Mexican Peso (mxp$) MP$ 1.0=$0. 105 WEIGHTS AND MEASURES Metric System FISCAL YEAR July 1 -June 30 MAIN ABBREVIATIONS & ACRONYMS AMCM: Metropolitan Area of Mexico City (Area Metropolitana de la Ciudad de Mexico) ANUIES: National Association of Universities and Institutions of Tertiary education (Asociaci6n Nacional de Universidades e Instituciones de Educaci6n Superior) CBTA: Center of Technological Agricultural Baccalaureate (Centro de Bachillerato Tecnol6gico Agropecuario) CBTIS: Center of Technological Industrial and Services Baccalaureate (Centro de Bachillerato Tecnol6gico Industrial y de Servicios) CBTF: Center of Technological Forester Baccalaureate (Centro de Bachillerato Tecnol6gico Forestal) CECYT: Center of Scientific and Technological Studies (Centro de Estudios Cientiricos y Tecnoidgicos) CENEVAL: National Center of Evaluation for Tertiary Education (Centro Nacional de Evaluacid6n para la Educaci6n Superior A.C.) CETAC: Center of Technological Studies ofContinental Water (Centro de Estudios Tecnol6gicos de Aguas Continentales) CETIS: Center of Technological Industrial and Services Studies (Centro de Estudios Tecnol6gicos Industrial y de Servicios) CETMAR: Center of Technological Studies of Sea (Centro de Estudios Tecnol6gicos del Mar) COMIPEMS: Metropolitan Commission of Public Institutions of Upper Secondary Education (Comisi6n Metropolitana de Instituciones Publicas de Educaci6n Media Superior) CONAPO: National Council of Population (Consejo Nacional de Poblaci6n) CONALEP: National College of Technical Professional Education (Colegio Nacional de Educaci6n Proresional Tecnica) COSNET: Council of the National System of Technological Education (Consejo del Sistema Nacional de Educaci6n Tecnol6gica) DGAIR: General Direction of Accreditation, Incorporation and Revalidation (Direcci6n General de Acreditaci6n, Incorporaci6n y Revalidacion) DGETA: General Direction of Technological Agricultural Education (Direcci6n General de Educaci6n Tecnol6gica Agropecuaria) IBRD Vice President David de Ferranti Chief Economist: Guillermo Perry Country Director: Olivier Lafourcade Lead Economist/Manager: Marcelo Giugale Task Manager: Marcelo Giugale Operation's Lead Specialists: Steven Webb Fernando Rojas William Dillinger Team Production Support: Michael Geller This operation was prepared by a World Bank team composed of MessrslMmes. Dillinger, Rojas (LCSPS); Webb (LCSPE); Marquez (LCSHH): Brizzi, Giugale, Nguyen, Velez, Everhart, Draaisma, Ordonez, Duval, Urbiola, Geller, Toxtle (LCCIC); Genta-Fons (LEGLA); Sherman and Vetter (consultants). The team was led by Mr. Giugale (Lead Economist, LCC1C), and worked under the general guidance of Mr. Olivier Lafourcade (Director, LCCIC).  DGETI: General Direction of Technological Industrial Education (Direcci6n General de Educaci6n Tecnol6gica Industrial) DGPPP: General Direction of Planing, Programing and Budgeting (Direcci6n General de Planeaci6n, Programaci6n y Presupuesto) EGCP: General Test of Profesional Quality (Examinationen General de Calidad Profesional) ENIGH: National Household Survey of Income and Expenditures (Encuesta Nacional de Ingresos y Gastos de los Hogares) ENEU: National Urban Employment Survey (Encuesta Nacional de Empleo Urbano) FOMES: Fund for Modernize the Tertiary Education (Fondo para Modernizar la Educaci6n Superior) IPN: National Polytechnic Institute (Instituto Polit6cnico Nacional) ITESM: Technological Institute of Higher Studies of Monterrey (Instituto Tecnol6gico de Estudios Superiores de Monterrey) OECD: Organization for Economic Cooperation and Development (Organizaci6n para Ia Cooperaci6n y Desarrollo Econ6mico) PROMEP: Program of Improvement of Professors (Programa de Mejoramiento del Profesorado) SEIT: Vice Ministry of Technological Research and Education (Subsecretaria de Educaci6n e Investigaci6n Tecnol6gicas) SEP: Ministry of Education (Secretaria de Educaci6n P6blica) SESIC: Vice Ministry of Tertiary education and Scientific Research (Subsecretaria de Educaci6n Superior e Investigaci6n Cientifica) SNTE: National Union of Education Workers (Sindicato Nacional de Trabajadores de la Educaci6n) UAEM: State of Mexico Autonomous University (Universidad Aut6noma del Estado de M6xico) UAM: Metropolitan Autonomous University (Universidad Aut6noma Metropolitana) UECyTM: Educational Unit of Science and Sea Technology (Unidad Educativa de Ciencia y Tecnologia del Mar) UNAM: National Autonomous University of Mexico (Universidad Nacional Aut6noma de Mexico)   Earnings Inequality after Mexico's Economic and Educational Reforms Part II: Background Papers CONTENTS Background Papers 1. Earnings Inequality and Education Attainment after Mexico's Economic Reforms Gladys Lopez-Acevedo, Lauro Ramos, Angel Salinas, and Monica Tinajero 2. The Financial Crises Impact on the Income Distribution in Mexico Gladys L6pez-Acevedo and Angel Salinas 3. The Evolution and Structure of the Rates of Returns to Education in Mexico 1987-1997): an Application of Quantile Regression Gladys Lopez-Acevedo, Lauro Ramos, and Angel Salinas 4. Educational Public Expenditure in Mexico: Some Distributional Issues Gladys Lopez-Acevedo, Angel Salinas, and Monica Tinajero 5. Marginal Willigness to Pay for Education and the Determinants of Enrollment Rates in Mexico Gladys LJopez-Acevedo, Angel Salinas, and Monica Tinajero 6. Educational Policy for Intermediate and Tertiary Level of Education in Mexico Gladys Lopez-Acevedo, and Angel Salinas   Earnings Inequality and Educational Attainment after Mexico's Economic Reforms Background Paper # 1 Relevant topics: Distribution consequences of crises and reforms. Gladys Lopez-Acevedo (LCSPE), Lauro Ramos (IPEA), Angel Salinas (LCC1C) and Monica Tinajero (UNAM) " 2 Abstract Even though the educational attainment levels expanded very rapidly, Mexico also experienced a pronounced increase in the degree of income and earnings inequality over the period of analysis. Most of the worsening of income distribution happened in the mid-eighties. The early nineties display little variation in income and earnings inequality except for a small trend towards deterioration. Income distribution improved between 1994 and 1996, an interval of time that entailed a severe financial crisis in the Mexican economy. Three broad hypotheses are frequently advanced to explain the similar increases in earnings inequality experienced in Mexico and other countries. These link the increase in earnings inequality to i) the increased openness of the economy, ii) institutional changes in the labor market and iii) skill-biased technological change. Unlike other papers based on these hypotheses, this paper deals explicitly with the changes in the distribution of education, as well as the interaction between educational policies, which induced such changes, and the working of the labor market. One interesting finding is that education is by far the variable that accounts for the largest share of the variation in earnings inequality in Mexico, both in terms of its gross and marginal contributions. This result indicates that as the Mexican economy progresses, education becomes even more important in determining the choices of sectors and occupations. That is, the marginal contribution of education by itself remains the same, but the gross contribution increases. In addition, the gross contribution of age to inequality has been going up and at the same time its marginal contribution has been decreasing. In other words, differences in both educational attainment and distribution among cohorts have become pronounced in recent times, leading to a higher correlation (negative) between education and age. Another important result is that the educational contribution to inequality in Mexico is the second highest in Latin America, next to Brazil. Moreover, what seems to be particularly interesting in the Mexican experience is the fact that the significance of education has been increasing over time. This paper also shows that education has the highest gross contribution to the explanation of changes in earnings distribution. Moreover, the income effect is always the prevalent one, and the significance of changes in the distribution of education remains high even if one controls the changes in other relevant variables such as age, economic sector, region and labor market status. The significance of education as an explanatory source for inequality changes seems to be a common pattern in Latin American countries. In fact, the relevance of the income effect over the allocation effect is also a trait shared by all countries where a similar analysis was carried out. What deserves attention in the Mexican case is the fact that the figures are well above those for other countries over a shorter period of time. This means that changes in the structure of supply of labor were particularly relevant for the earnings distribution. This type of analytical policy support is central to the social development objective of the World Bank Country Assistance Strategy. It is also part of a comprehensive work meant to build a poverty and inequality strategy for Mexico. This research was completed as part of the "Earnings Inequality after Mexico's Economic and Educational Reforms" study at the World Bank. We are grateful to INEGI and SEP (Ministry of Education) for providing us with the data. These are views of the authors, and need not reflect those of the World Bank, its Executive Directors, or countries they represent. 2 gacevedo@world bank.org and asalinaseworldbank.org.  INTRODUCTION This paper reviews the factors and mechanisms that have been driving inequality in Mexico, particularly in terms of educational policies. More specifically, the paper relates the recent evolution in earnings inequality to the changes in the distribution of education. In addition, this study analyses the way labor market interacts with the distribution of schooling in the labor force. This is done aiming at (i) establishing an analytical framework that permits the analysis of interaction between education and labor market, and (ii) examining the evolution of earnings inequality after the macro economic and educational policies followed in the 80's and 90's in Mexico. The paper is organized as follows: Section 1 describes the evolution of total current income inequality, between 1984-96 based on the National Household Income and Expenditures Survey (ENIGH), and using the household per capita income as the unit of analysis. Section 2, focuses on the evolution of individual earnings inequality, using the information of the National Urban Employment Survey (ENEU). Section 3, investigates how much of the earnings inequality can be explained by education, as well as other control variables, both in gross and marginal terms.3 Then, one takes a closer inspection of the evolution of educational attainment and distribution in recent times. Section 5, relates the changes in the distribution of education to the changes in earnings inequality. The last section has the concluding remarks. 1 THE EVOLUTiON OF TOTAL INCOME INEQUALITY The first evaluation of income inequality evolution in Mexico is based on the information available in the ENIGH4. The reason for doing so is that this survey captures total current income of the households, including non-monetary income, besides earnings and other sources of monetary income. The unit of analysis is the household, and the concept of income is household per capita income.5 The main results of this evaluation are shown in table 1. It indicates that a very sizable deterioration in the income distribution has taken place between 1984 and 1996. While the poorest 20% of the population lost almost one seventh of their income share (0.6 percentage points), the richest 10% increased theirs by something close to one seventh (5.2 percentage points). Moreover, this last group was the only one that gained over that period, as not only the poorest, but also those in the middle lost in relative terms. Looking at the results of this comparison, one can say that the 1984-1996 period in Mexico was marked by a series of regressive income transfers from almost the entire population spectrum to the richest stratum. Accordingly, the most commonly used inequality index points to a worsening in income inequality over this span of time. The Gini coefficient, which is more sensitive to changes in the middle of the distribution, rises from 0.473 in 1984 to 0.519 in 1996. On the other hand, the Theil T index, which is extremely sensitive to changes in the upper and lower tails, goes up from 0.411 in 1984 to 0.524 in 1996. Even though the worsening of the distribution is indisputable, there are, nevertheless, two points that must be stressed. The first one is that, according to the ENIGH survey, most of the worsening of the income distribution happened in the mid-eighties (1984-1989). The early nineties display little variation in earnings inequality except for a small trend towards deterioration. From 1989 to 1994, the income share accruing to the 20% poorest decreased slightly (it went down from 3.9% to 3.8%), whereas the richest 10% were the only ones that 3Educational attainment has not only a monetary impact but can also affect other outcomes, which are important for individual's well being, but that are not necessarily measured in monetary terms. This study, however, will not consider the non-monetary impacts of education. An interesting methodology for the estimation of these impacts can be found in Wolfe, Barbara and Samuel Zuvekas (1997). 4 See annex I for a brief description of this survey. SThis means that total current income of the household divided by its number of household members. That is, we are considering the household as a unit characterized by a flow of income transfers and disregarding aspects related to equivalence scale. 2  increased theirs (by one percentage point), and, therefore, those in the middle also experienced losses. Table 1. Lorenz Curves for Total Current Income" (Accumulated Income Share - %) Population Share 1984 1989 1992 1994 1996 10 1.66 1.39 1.32 1.39 1.39 20 4.47 3.88 3.68 3.76 3.89 30 8.19 7.29 6.92 6.98 7.29 40 12.85 11.65 11.09 11.08 11.63 50 18.76 17.05 16.26 16.28 17.08 60 26.15 23.78 22.83 22.79 23.86 70 35.51 32.25 31.13 31.10 32.39 80 47.64 43.12 42.14 41.93 43.44 90 64.53 58.75 58.32 57.68 59.33 92 68.79 63.06 62.81 62.03 63.61 94 73.73 68.03 68.03 67.26 68.68 96 79.38 73.82 74.47 73.70 74.95 98 86.68 81.60 82.81 82.49 83.32 100 100.0 100.0 100.0 100.0 100.0 Bottom 20% 4.5 3.9 3.7 3.8 3.9 Middle 40% 21.7 19.9 19.2 19.0 20.0 Mhigh 30% 38.4 35.0 35.5 34.9 35.5 Top 10% 35.5 41.3 41.7 42.3 40.7 Gini 0.473 0.519 0.529 0.530 0.515 Theil T 0.411 0.566 0.550 0.558 0.524 Source: Own calculations based on ENIGH. "Based on household per capita income. The second fact to be emphasized is very surprising and hard to explain: the observed improvement in the income distribution between 1994 and 1996, an interval of time that entails a severe financial crisis in the Mexican economy.6 Usually one would expect inequality to go up during recessive times, as it seems plausible to admit that the rich have more ways to protect their assets than the poor do. Especially when it comes to labor which is basically the only asset of the poor (the labor-hoarding hypothesis). The fact, however, is that the 10% richest experienced relative losses (their income share dropped 1.6 percentage points) and, accordingly, inequality went down: the Gini coefficient came down from 0.534 in 1994 to 0.519 in 1996, whereas the drop in the Theil T was from 0.558 to 0.524. In principle, it could be argued that the richest experienced severe capital losses due to the crisis, in such a way that their total current income was affected compared to the poor. This hypothesis, however, is not supported by the data shown in table 2, as monetary income other than wages and salaries, and financial income as well, increased their share in total income in that time interim, particularly so for the urban areas. Therefore, the fall in inequality remains somewhat puzzling. 6 In 1994, current account deficit was 30 billion dollars, about 7 percent of GDP. The main effects of the financial crisis were i) GDP and domestic demand felt 6.2 percent and 14 percent respectively each; ii) the unemployment rate rose from 3.7 percent in 1994 to 6.2 percent in 1995; and, iii) the GDP per capita decreased 7.8 percent and workers experienced a significant reduction in their real wage, nearly 17 percent in 1995. 3  Table 2. Share of Total Income by Source (%) 1994 1996 Source Total Urban Rural Total Urban Rural Monetary Current Income Total Labor Earnings 47.12 49.01 32.07 44.51 46.08 33.75 Property (Business) Income 16.96 16.23 22.75 17.74 17.11 22.07 PropertyIncome and Rents 1.10 1.13 0.87 1.35 1.47 0.51 Income from cooperative firms 0.22 0.24 0.12 0.06 0.03 0.32 Monetary Transference 5.44 4.72 11.23 6.55 5.89 11.11 Other Current Income 0.64 0.67 0.36 0.69 0.66 0.91 No Monetary Current Income Self-Consumption 1.44 0.81 6.46 1.20 0.69 4.72 Non Monetary Payment 1.55 1.58 1.28 2.25 2.32 1.82 Gifts 5.04 4.73 7.57 6.07 5.86 7.55 Housing Imputed Rent 16.02 16.60 11.39 13.76 14.28 10.20 Financial Income 4.46 4.28 5.91 5.80 5.62 7.04 Total Income 100.00 100.00 100.00 100.00 100.00 100.00 Source: Own calculations based on ENIGH. In table 3 the results for the Gini and Theil T are displayed for urban and rural areas using total current income. There, one can see, for both indices that inequality in rural areas was lower than in urban areas and remarkably stable until 1992. After a small decrease in 1994, it increased in 1996, contrary to the aggregate result discussed before. In light of these outcomes, it seems pertinent to state that the leading force behind the behavior of current income distribution in Mexico is in the urban areas. Table 3. Inequality Measures for Total Current Income Gini Coefficient Theil T Index Year National Urban Rural Year National Urban Rural 1984 0.473 0.442 0.448 1984 0.411 0.356 0.375 1989 0.519 0.498 0.444 1989 0.566 0.526 0.361 1992 0.529 0.498 0.434 1992 0.550 0.483 0.353 1994 0.534 0.508 0.419 1994 0.558 0.499 0.325 1996 0.519 0.493 0.452 1996 0.524 0.470 0.390 Source: Own calculations based on ENIGH. 2 THE EVOLUTION OF EARNINGS INEQUALITY How much of total income inequality is due to earnings inequality? Table 4 presents the results of total current income inequality for each of its components: earnings7, monetary income excluding earnings and non-monetary income by urban and rural areas (see annex 2.1 and table I in annex 3).8 Earnings are the source of income that contributes to most of the overall inequality, being responsible for almost half of it at national level. It is clear that these figures may be affected by a possible underreporting of capital gains, but it seems valid to state that if one better understands the mechanisms that lead earnings inequality, this will be a large step towards understanding the 7Earnings as defined in the ENIGH survey include salaries and wages, paid over-time, tips, contract workers' earnings, Christmas or New Year gifts and other gifts, and other monetary compensations (non-regular earnings). Earnings as defined in the ENEU survey include salaries and wages, self-employed workers' earnings, contract workers' earnings, and implicit firm owners' salaries, as well as non-monetary earnings. s Though the results are shown for the Gini coefficient, these could have been obtained for the Theil T index, as both of them satisfy the six propositions listed in Shorrocks (1982). 4  total inequality behavior. Besides, as long as labor is the main, if not the only, asset of the poor, a better knowledge of earnings inequality is a valuable input for the assessment of poverty and welfare issues. Table 4. Decomposition of Total Current Income (Percentaze Share in Overall Gini) Income Source Earnings Monetary income No monetary TOTAL Excluding earnings Current income National 1984 46.0 32.9 21.0 100.0 1989 41.0 36.0 23.0 100.0 1992 42.9 31.9 25.2 100.0 1994 50.2 25.9 23.9 100.0 1996 46.7 29.4 23.9 100.0 Urban 1984 45.6 32.2 22.2 100.0 1989 38.6 37.3 24.1 100.0 1992 41.4 33.1 25.5 100.0 1994 50.0 26.0 24.0 100.0 1996 46.1 29.8 24.1 100.0 Rural 1984 30.7 49.5 19.8 100.0 1989 35.7 43.5 20.8 100.0 1992 29.6 42.2 28.2 100.0 1994 31.9 43.8 24.2 100.0 1996 35.7 41.2 23.1 100.0 1996 35.7 41.2 23.1 100.0 Source: Own calculations based on ENIGH. To examine the behavior of inequality in recent times the household survey ENEU and ENIGH were used. However, in this paper the ENEU results are analyzed, while the ENIGH results are presented in the Annex 4. The ENEU survey is collected quarterly and for the purposes of this study the third quarter of each year was chosen, in an attempt to avoid the influence of seasonal factors that could make non-comparable results to those obtained from ENIGH.9 By examining the results shown on table 5, one also reaches the conclusion that the distribution of earnings has become more unequal in recent times. The Gini coefficient jumps from 0.395 in 1988 to 0.442 in 1997, after reaching a peak in 1996 of 0.464. Similarly, the Theil T index went up from 0.327 in 1988 to 0.372 in 1997, with 0.474 in 1996. Another index, the R10/20,'0 increased from 4.48 to 6.04 over the period, reaching a maximum of 6.74 in 1996. Table 5. Inequality Indices for the Distribution of Earnings (1988-1997) Population Earnings Share (%) Share (%) 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 Bottom20% 7.54 7.62 7.19 6.84 6.47 6.13 5.98 5.91 5.72 5.95 Middle 40% 25.23 24.45 23.86 23.41 23.37 22.86 22.36 22.59 22.09 23.01 MHigh 30% 33.44 34.15 33.96 33.77 33.52 33.37 32.94 33.42 33.61 35.13 Top 10% 33.78 33.78 34.98 35.98 36.64 37.63 38.72 38.08 38.58 35.91 Gini 0.395 0.398 0.414 0.426 0.434 0.447 0.458 0.455 0.464 0.442 Theil T 0.327 0.328 0.350 0.380 0.396 0.414 0.470 0.427 0.474 0.372 Rio/2o 4.48 4.43 4.87 5.26 5.66 6.14 6.47 6.44 6.74 6.04 9 In order to reduce the heterogeneity of the samnple and also aspects related to self selection, the population under analysis are individuals living in urban areas, between 16 and 65 years old, working 20 hours a week or more and no seasonal workers. Also, the two highest observations were dropped from the sample, as there was clear evidence of the presence of outliers in some years. 10This index is the ratio of the income share accruing to the 10% richest and 20% to the poorest. 5  Source: Own calculations based on ENEU (3d quarter). There are two main differences in the pattern shown by the earnings and total current income distribution. First, the gains are not limited to the richest 10%. As those in the seventh, eight, and nine tenths of the distribution also improved their relative earnings over the period by almost two percentage points; the biggest losers were the middle 40%, who lost more than two percentage points of their income share. Second, there is a clear worsening in the earnings distribution in the present decade throughout 1996. On the other hand, the inequality associated with the total current income was moderately stable in the nineties, displaying an improvement in 1996. The different behavior between total current income and earnings inequalities from 1994 to 1996 gives support to the idea that the poor, who mostly rely on labor as a source of income, were the least able to protect themselves during the recession. However, the substantial drop in earnings inequality from 1996 to 199711 is, once more, a surprising finding. It is true that the Mexican economy as a whole had a strong and impressive performance in 1997. The aggregate growth rate was around 7%, real investment grew by 24% and exports by 17%, industrial production increased 9.7%, and the civil construction sector, which is highly intensive in less skilled labor, experienced a growth close to 11%. Under such a scenario, an improvement in distribution of earnings itself is not unlikely, but the magnitude and quickness of the recovery calls for a detailed inspection of the mechanisms responsible for it. Three broad hypotheses are frequently advanced to explain the similar increases in earnings inequality experienced in Mexico and other countries.'2 These link the increase of earnings inequality to (i) the increased openness of the economy, (ii) institutional changes in the labor market, and (iii) skill-biased technological change. The first of these hypotheses argues that as trade barriers are reduced, an economy is placed under increased competitive pressures to specialize along its lines of comparative advantage. A developed country that is relatively high skilled-abundant, like the United States, will be induced to specialize in high skill- or educational-intensive activities as its low-skilled industries come under increased competitive pressure from low skilled-abundant, low-wage countries. Hanson and Harrison (1995) examined the impact of the Mexican trade reform on the structure of wages using information at firm level. They tested whether trade reform had shifted employment toward industries that are relatively intensive in the use of skilled labor force [the Stolper-Samuelson-Type (SST) effect]. Their main conclusion was that the wage gap was associated to changes within industries and firms, which cannot be explained by the SST effect. Thus, the increase in wage inequality was due to other factors.'3 Hanson's (1997) paper examined a trade theory based on increasing returns, which has important implications for regional economies. Hanson's conclusion is that employment and wage patterns are consistent with the idea that access to market is important for industry location. This first hypothesis has several problems when applied to the United States, and becomes even less persuasive when applied to Mexico. Mexico greatly liberalized its trade regime since 1984. However, the reduction of its trade barriers has mostly been vis-a-vis imports from the developed countries, notably the United States and Canada, whose share in total Mexican merchandise imports increased from 68 percent in 1985 to 73 percent in 1993 and to 77.5 percent in 1996. Since Mexico is a low skilled-abundant country compared to its two northern neighbors, it would be expected that the liberalization of trade would have induced a specialization pattern that would raise the relative demand (and hence wages) for the lesser-educated members of the labor force. This did not happen. Instead, the increase in earnings inequality observed in Mexico follows the same pattern as that observed in the United States: less educated workers experienced real wage declines, while highly educated workers experienced real wage improvements. The trade-based explanation may still be relevant, however, to the extent that greater openness "The R0/20 index, for instance, was 6.74 in 1996 and went down to 6.04 in 1997. 12 See, for example, the "Symposium on Wage Inequality" (1997) and the "Symposium on How International Exchange, Technology and Institutions Affect Workers" (1997). " The Stolper-Samuelson effect was also examined under NAFTA in Burfisher, Mary E., Robinson, Sherman, and Thierfelder (1993). 6  facilitates the transfer of ideas and technology, which is identified below as the more persuasive explanation of increase in earnings inequality. A variant on the globalization/technology nexus explanation, advanced by Feenstra and Hanson (1994), involves outsourcing behavior where multinational enterprises in the developed country relocate their lower skilled-intensive activities to the less skilled-abundant developed countries. However, what is referred to as a low skilled activity in the United States may be a high-skilled activity in Mexico, which could explain the similar evolution of earnings inequality in both countries. The second explanation revolves around institutional changes such as reductions in the minimum wage, the decreasing strength of trade unions and the declining share of state-owned enterprises. The existence of a binding minimum wage, for example, truncates the lower end of the wage distribution. As the minimum wage is allowed to erode away, say through inflation, it becomes less binding by moving further down the low end of the wage. distribution, with the result that, ceteris paribus, a higher share of wages will lie below the previous minimum wage level. This translates into an increased dispersion in wages and earnings. A look at institutional developments in Mexico since the early 1980s indicates that these developments have not exerted a significant influence on the earnings distribution.'4 The distribution of real wages, for example, does not reveal any significant distortions around the minimum wage, which suggests that it is not a binding constraint. The fact that this minimum wage has continued to erode in real value, therefore, seems to have been irrelevant. Similarly, the distribution of union wages is not significantly different from that of non-union wages, once differences in education levels are accounted for. That also renders any erosion of union power irrelevant for the distribution of earnings. In conclusion, while the influence of institutional factors cannot be rejected entirely, it also does not appear to be the principal cause of the increase in earnings inequality. A persuasive explanation, both for the United States and Mexico, seems to be one which links earnings inequality to skill-biased technological changes that raise the relative demand for higher- skilled labor. Cragg and Epelbaum (1996) examined the demand shift in Mexico. They pointed out that the major source of rising inequality is a biased demand shift rather than a uniform demand growth when there are different labor supply elasticities. Meza (1998) also investigated demand shifts. The author 's hypothesis is that the demand shifts for more educated labor force, "within" economic sector, explains the increase in their premium when compared to the demand shift for less educated workers "between" economic sector. Tan and Batra (1997) studied the skill-biased technical change hypothesis as a plausible explanation of wage inequality using data at the firm level for Colombia, Mexico, and Taiwan. They obtained the following results: i) firm investments in technology have the largest impact on wage size distribution for skilled workers. ii) It had the smallest impact on wages paid to unskilled workers. iii) Lastly, a decomposition of wage effects by sources of technology revealed that wage premiums paid to skilled workers are led primarily by firm investment in R&D and training. Such conclusions seem to support the skilled-biased technological change hypothesis.'5 According to the typology used by Johnson (1997), the type of technological change that drives wages up for the more highly skilled workers and drives wages down for the less skilled workers (as occurred in both the United States and Mexico) is extensive skilled-biased technological change. Under this type of technological change, skilled workers become more efficient in jobs that were traditionally performed by unskilled workers. 14 See Hernandez Laos et al. "Productividad y Mercado de Trabajo en Mexico", 1997 15 Note that these results should be considered carefully, since the analysis is based on data at the firm level and just for the manufacturing industry. 7  Figure Ia Mexican Economy Openness Degree and Inequality 65 55 45 35 25 GATT A eetNAFTAAgreement 15 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 - Oppeness Degree - - R10/20 Source: Own calculations based on ENIGH and INEGI data Figure lb Conditional Median Real Hourly Earnings by Educational Level 20.00 1988=100 18.00- 16.00 ._- 4.00- - - - - 1200 10.00 - 8.00 - _-_-_-_-- .. 6.00 - . .4= = = : 4.00 - 2.00 0.00 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 Primary Incomplete --- Primary Complete ----- L. Secondary Comp. - U. Secondary Comp. - -- --University Source: Own estimations based on ENEU survey. I/ Medians were calculated conditional on experience, experience squared, gender, economic sector, labor market status, and region As it is shown in figure Ib6, all series have the same trend for all period. However, from 1990 conditional real earnings for University increased substantially, while conditional real earnings for lower educational levels remained steady, up to 1994. After that, it seems that the earnings differentials among all educational levels remained constant. This suggests that other factors rather than supply of new basic educational comers drove earnings differentials among level of schooling. 16 Median real hourly earnings were estimated using quantile regression models (0=0.5) and conditioned on experience, gender, labor market status, economic sector and region (see annex I for groups definition). 8  In sum, demand and supply, interacting within a context of economic modemization and globalization generate the trend towards greater wage disparity. It should be noted, however, that none of these explanations deal explicitly with the changes in the distribution of education, as well as the interaction between the educational policies that induced them and the workings of the labor market. 3 STA Tic DECOMPOSITION This section aims at evaluating the contribution to earnings inequality in Mexico of a set of variables, either related to individual attributes, as schooling and age, or form of participation in the labor market, as number of hours worked or status, for selected years from 1988 to 1997. The idea is to measure the reduction in inequality that results from excluding the differences in average earnings among workers in different groups fonned by those variables. When the exercise is conducted for a single variable, this reduction is said to be the gross contribution of such a variable to the overall wage inequality. When a variable is added to a model that contains all the remaining ones, the change in the gross contribution of these two models is called the marginal contribution of the added variable. In other words, the gross contribution can be regarded as the uncontrolled explanatory power of a given variable, and the marginal contribution as its explanatory power controlled by a set of other seemingly relevant variables. 3.1 SHORT REVIEW Before proceeding to the decomposition exercise, it is worth to review the conclusions of other recent studies in relation to the evolution of earnings inequality and some variables that are important in the process of eamings formation. Cragg and Epelbaum (1996) show that both average wage and education skill premium, which is defined as the percentage increase in wages over the primary schooling group, have increased substantially for more educated workers. In other words, the higher the level of education the larger the increase in average wage is, which in tum leads to an increase in inequality. They also examined whether the high demand for skilled labor is industry specific, task specific or simply general education. In order to assess the marginal contribution of other factors that are not related to education, these factors are controlled by a set of dummy variables that describe the industry and task specific effects. The authors concluded that the industry-specific effect was small and that the task-specific effect (occupational variable) explained half of the growing wage dispersion from 1987 to 1993. This conclusion, however, may not be correct, as occupation might be considered an endogenous variable, which is determined by education. As shown on table 2 in annex 3, educational level and occupational variables are highly correlated. In contrast, the correlation between education and other variables are low. Hence the occupation variable should be carefully handled in any kind of analysis. 3.2 METHOLOLOGY The decomposition analysis is a useful tool for assessing the impact of certain factors on the evolution of the income distribution. In general, the different decomposition methods follow two definitions (Fields, 1996): * Inequality in the population can be decomposed into different elements such as the sum of the parts is equal to total inequality; * Inequality in the population can be decomposed as a weight sum of inequality within and between groups. The Gary S. Fields (1996) and Bourguinon, et al (1998) papers employed the first decomposition method. Fields decomposed total population inequality in a sum of different 9  variables or elements, each being the explanatory variable in the earnings function. This will help us to answer two questions: how much income inequality is explained by each right hand side variable in a given point in time? And, how much of the difference in inequality between groups or dates is explained by each variable? Notice that this technique assumes that we know the correct model specification. Formally, the above methodology can be written as Y= Z'B where: Y = In (W) is the vector of the logarithm incomes Z = (1, X, ...X, X ) is the matrix of explanatory variables and error term B = (a, 8,....f, I) is the regression coefficient vector. Then, cov(Pj3 ZJ, Y) P Ia(Z, )corr(Z., Y) a I2(y) a(Y) where: s, is the relative factor weight and s, = R2 (determination coefficient) The contribution of factor] to the change in the inequality measure I() between time 0 and time 1 is: Ai [I( ) = St - T ) ' ( .) where: s* is the relative weighted factor for vear 0 s , is the relative weighted factor for year 1. Fields also proposes a change break down in the factor's contribution into the following: the change in the coefficient of the factor or variable, the change of the standard deviation of the variable and the change in the correlation between the variable and earnings. Bourguignon et al. (1998) carried out a decomposition of the effects of changes in an entire distribution, rather than on a scalar summary statistic. This methodology was originally proposed by Almeida dos Reis and Paes de Barros (1991) and Juhn, Murphy and Pierce (1993) and later generalized by Bourguignon et. al. The methodology, by means of micro simulations, decomposes the changes in income distribution into different effects. Bouillon, et. al. (1998) used this technique in the case of Mexico decomposing the change into the return effect, the population effect, the error term effect and the residual effect. This can be expressed as follows, let D(y)=D66, X, E) be the income distribution measure and define: y=Xf + E, where: X is the set of demographic variables, f3 is the set of prices and E the error terms. If y is the income in year 0 and y' in year 1, it can be show that the change in income distribution can be expressed as: A = D(y )-D(y) =,f(X', ) + X6, E) + cf6', X) + (-66, X) -E(6', X)j where: ,f(X'. c') =D6/3',X'. ') - D6f. X', &) is the return effect X6/, E) = D(66, X;' E) - D66. X 6) is the population effect f66 ', X) = D(f, X, 6') - D6',X', X ) is the error term effect 10  ft(s, X') -E 66', X')} is the residual effect Notice that the analysis makes the following assumptions: * Income is correctly expressed as a linear combination; * In order to compute Df6, X: s) the residuals in the second year are re-scaled to the second year of reference by a constant such that the variance in that year is the same as the variance of the residuals in the first year. This in turn implies the assumption that the distribution of Ey -'just differs by the variance. Cesar Bouillon, Arianna Legovini and Nora Lustig (1999) and Cesar Bouillon, Arianna Legovini and Nora Lustig (1998) used this methodology. In these documents, although the assumption of unchangeable dispersions, of the regression error terms, does not significantly restrict the model's results, it is questionable to use the variance instead of a proper inequality index. That means that one measure for the within inequality is used and another for the between inequality. Miguel Szekely (1995), in order to explain the inequality changes between two points in time applied the following formula: C,6 (C )T'B (r)-TB(lr) T'-T where: 7r is the partition or division of the population T'B(7r) is the Theil index between group in year I TB(7r) is the Theil index between group in year 0 CB(7r) is the percentage of the change in inequality explained by the variables in ir T' is the Theil index in year I T is the Theil index in year 0. It is important to note that this methodology does not allow us to separate the income from the allocation effect. The second approach, which is used in this study, uses inequality measures known as "generalized entropy indices". Bourguinon (1979), Cowell (1980) and Shorrocks (1980, 1984) have shown that only such measures satisfy all the desirable properties for any inequality measure and are additive decomposable. Assume that the population is divided into g groups (according to education, for instance). Then, a measure of inequality is said to be additive decomposable (see Shorrocks (1980)) when it can be written as: 1 = I(Pglag ,lg ) = IB(Pg,ag) + :W(Pg ,a9)I9 ( I) g where Pg is the fraction of the labor force employed in the groupg, ag is its relative mean income, and Ig represents the wage dispersion within this group as measured by the index I . The term IB on the right side of (1) corresponds to the inequality between groups (i.e. the amount of inequality that would be observed in the case of an earnings redistribution within each group, in such a way that, at the end, all workers in a group would receive the same earnings). The second term in the right-hand side (from here on called Iw ) reflects the inequality within groups; i.e., the share of overall inequality that is associated to factors other than those involved in the particular partition under study. It represents the degree of inequality that would be observed if all groups had the same average earnings. Notice that Iw is a weighted average of the internal inequalities, the weights, w(I3g,ag), being a function of the population share and average earnings of each group. One can thus estimate the contribution of a given variable(s) to the overall earnings inequality at a given point in time as the fraction of this inequality that would be eliminated if the average wage of all groups formed by that (those) variable(s) were equalized, while keeping the internal 11  dispersions unchanged. The rationale behind this exercise is that the effect of this (these) variables is captured by differences in average earnings at group level. Amongst the most commonly used inequality indices, the Theil T is one of the few that is additive decomposable."1 The general statistics needed for the decomposition by age, sector, level of schooling, hours worked and status from 1988 to 1997 are shown on table 6. Table 6 General Statistics for the Static Decomposition 1988 1992 1996 1997 Variable Beta Alfa Theil Beta Alfa Theil Beta Alfa Theil Beta Alfa Theil Schooling Primary Incomplete 0.185 0.70 0.220 0.147 0.65 0.234 0.129 0.57 0.283 0.127 0.57 0.207 Primary Complete 0.277 0.81 0.257 0.259 0.72 0.207 0.244 0.65 0.270 0.237 0.67 0.207 Lower Secondary Comp 0.241 0.88 0.228 0.264 0.80 0.281 0.257 0.74 0.264 0.263 0.76 0.229 Upper Secondary Comp 0.189 1.09 0.234 0.205 1.07 0.300 0.216 1.04 0.278 0.221 1.05 0.259 University Complete 0.107 2.10 0.343 0.124 2.32 0.359 0.154 2.30 0.430 0.151 2.22 0.289 Total 0.327 0.395 0.464 0.372 Age 16-25 0.320 0.74 0.202 0.323 0.68 0.201 0.280 0.64 0.239 0.282 0.66 0.217 26-34 0.278 1.07 0.259 0.276 1.07 0.334 0.279 1.02 0.332 0.274 1.05 0.320 35-49 0.282 1.17 0.364 0.293 1.24 0.441 0.323 1.26 0.541 0.327 1.21 0.374 50-65 0.119 1.13 0.475 0.108 1.14 0.521 0.119 1.08 0.589 0.117 1.13 0.496 Total 0.327 0.395 0.464 0.372 Sector Primary Sector 0.019 0.99 0.508 0.016 0.99 0.667 0.014 1.20 0.976 0.012 1.20 0.621 Manufacturing Industry 0.274 0.97 0.323 0.242 0.96 0.379 0.221 0.94 0.559 0.227 0.92 0.371 Non Manuf Industry 0.058 0.91 0.224 0.064 1.06 0.409 0.060 0.91 0.382 0.057 0.88 0.331 Commerce 0.178 1.01 0.415 0.196 0.92 0.415 0.188 0.90 0.484 0.180 0.89 0.407 Finance Services/Rent 0.030 1.39 0.230 0.027 1.77 0.384 0.024 1.90 0.407 0.027 1.79 0.332 Transp./communication 0.066 1.12 0.191 0.069 1.12 0.310 0.064 1.03 0.344 0.068 1.06 0.255 Social Services 0.253 1.10 0.280 0.261 1.12 0.380 0.294 1.23 0.373 0.293 1.25 0.317 OtherServices 0.122 0.73 0.385 0.125 0.70 0.291 0.136 0.58 0.274 0.136 0.58 0.269 Total 0.327 0.395 0.464 0.372 Hours Worked 20-39 0.201 0.89 0.278 0.174 0.87 0.391 0.174 0.84 0.399 0.172 0.86 0.333 40 - 48 0.581 0.96 0.280 0.566 0.95 0.332 0.525 0.96 0.421 0.540 0.98 0.331 49-+ 0.218 1.20 0.438 0.260 1.20 0.483 0.301 1.16 0.535 0.288 1.13 0.444 Total 0.327 0.395 0.464 0.372 Status Employer 0.046 2.32 0.549 0.048 2.44 0.463 0.048 2.18 0.561 0.046 2.15 0.428 Selfemployed 0.158 0.97 0.338 0.149 0.89 0.354 0.174 0.75 0.377 0.167 0.79 0.340 Informal salaried 0.122 0.58 0.210 0.140 0.54 0.158 0.147 0.47 0.174 0.150 0.48 0.175 Formal salaried 0.609 0.99 0.240 0.602 1.03 0.342 0.558 1.14 0.412 0.567 1.13 0.311 Contract 0.064 0.98 0.230 0.062 0.96 0.297 0.072 0.77 0.302 0.070 0.79 0.268 Total 0.327 0.395 0.464 0.372 Source: Own calculations based on the ENEU (3d quarter). Note: The sample includes only those who reported information on schooling level, age, economic sector, and labor market status simultaneously. '7For the decomposition of the Theil T, see Ramos (1990) and annex 2.2. 12  3.3 RESUL TS The results for the exercise of static decomposition are shown on table 7'18 Education (the result of the interaction between demand and supply) is by far the variable that accounts for the largest share of earnings inequality in Mexico, both in terms of its gross and marginal contributions. The gross contribution, i.e., its explanatory power when it is considered alone, amounts to one fifth of total inequality in 1988 and one third in 1997.'9 The marginal contribution, i.e., the increase in the explanatory power when it is added to a model that already has the other variables, is remarkably stable and meaningful, staying around 21% throughout the whole period. It is worth pointing out that the difference between the twvo contributions has been increasing over time, indicating that the degree of correlation and other variables has been going up, i.e., the "indirect" effects are becoming more important. Table 7. Contribution to the Explanation of Earnings Inequality(%) 1988 1992 1996 1997 Variables Gross Marginal Gross Marginal Gross Marginal Gross Marginal Education 20.2 20.8 26.9 21.6 29.3 21.2 32.6 21.2 Age 5.4 8.3 7.2 6.1 6.6 6.2 7.3 5.4 Economic Sector 2.3 8.1 4.0 5.2 6.8 5.2 8.6 4.4 Status 12.8 11.2 13.7 8.9 13.7 7.4 15.6 7.5 Source: Own calculations based on ENEU The other variables considered seem to be much less important. The three of them, and particularly the economic sector and status in the labor market, display an upward trend in their gross contribution, and a declining one in their marginal contribution. This can be interpreted, as evidence that the interaction between these variables and education has become more intense. That is, the workers' skills are becoming increasingly more relevant to the determination of their type of participation in the labor market, as well as for their position across different economic segments of the economy. Note that the same pattern holds when hours worked instead of sector is considered (see table below). Table 8. Contribution to the Explanation of Earnings Inequality 0 1988 1992 1996 1997 Variables Gross Marginal Gross Marginal Gross Marginal Gross Marginal Education 20.2 20.2 26.9 22.3 29.3 22.6 32.6 24.5 Age 5.4 5.4 7.2 4.8 6.6 4.6 7.3 4.5 Hours worked 1.7 3.8 1.9 3.3 1.3 4.0 1.2 3.5 Status 12.8 7.6 13.7 71 13.7 6.0 15.6 6.5 Source: Own calculations based on the ENEU (3' quarter) The analysis of these results leads to the conclusion that education is a key variable for the understanding of earnings inequality in Mexico.202' Even though this is to some extent a IB Since this exercise is very intensive in the number of observations (which constitutes its main handicap) the variable "hours worked" was dropped in order to avoid the problems with cells with too few observations. The decision was made through the comparison among different combinations of variables, where hours worked ended up being the least relevant. 19 In most earnings equations for any country, the set of measurable observable variables explains at most 60% of the total variance. In the United States, education accounts for 10% of the total variance. 20 Additional evidence is that the explanatory power of the complete model was 42.5% in 1988, 45.0% in 1992, 45.5 in 1996. and 48.3% in 1997. This means that the marginal contribution of education is almost equal to the joint 13  remarkable finding, it comes as no surprise in the Latin American context. The results for some countries in the region, where similar exercises were carried out. are reported on table 9. Mexico stays on the average range for Latin American countries, and displays a situation close to that observed in Colombia and Peru. However, education seems to be more important for inequality in Brazil, and much less important in Argentina and Uruguay. It is important to stress the fact that this is a comparison in relative terms. Given that in Colombia and Peru, where education has a similar explanatory power, there is a lower degree of inequality compared to Mexico, the absolute contribution of education is higher in Mexico. As a matter of fact, in absolute terms, the contribution of education to inequality in Mexico is the second highest in Latin America, next only to Brazil. Moreover, what seems to be particularly interesting in the Mexican experience is the fact that the significance of education has been increasing over time. Therefore, the inspection of the evolution of the educational distribution and the income profile associated to it, as well the link between changes in this distribution and changes in earnings inequality will be addressed in the next sections. Table 9. Contribution of Education to Earnings Inequality. International Comparison Country Author(s) Period Gross Contribution (%) Latin America Altimir and Pifiera (I 982) 1966/74 17-38 Argentina Fiszbein (1991) 1974/88 16-24 Brazil Ramos and Trindade (1992) 1977/89 30-36 Vieira (1998) 1992/96 30-35 Colombia Reyes (1988) 1976/86 29-35 Moreno (I 989) 1976/88 26-35 Costa Rica Psacharapoulos et alt. (1992) 1981/89 23-26 Peru RodrHguez (1991) 1970/84 21-34 Uruguay Psacharapoulos et alt. (1992) 1981/89 10-13 Venezuela Psacharapoulos et alt. (1992) 1981/89 23-26 4 THE EVOLUTION OF EDUCA TIONAL A TTAINMENT Educational attainment levels increased rapidly in most developing countries since the 1950s Schultz (1988). While Mexico also partook in that development, earlier studies had identified a significant lag in its educational indicators. Londofno (1996) for example, points to an "education deficit", according to Latin American countries in general, and Mexico in particular, have approximately two years less of education than would be expected for their level of development.22 Elias (1992) found that education was the most important source of labor quality improvement in Latin America between 1950 and 1970, but points out that such improvements did not take place to the same extent in Mexico as in other countries in the region. This changed dramatically in the 1 980s, figure 2 shows that although Mexico's educational attainment increased steadily since the 70's, it continued to remain below the international trend line23. In contribution of age, economic sector, and status in the labor market. 21 Miguel Szekely (1995) applied the static decomposition of the Theil to the ENIGH for the years 1984, 1989 y 1992. He used education, occupation, region, economic sector, and job status as control variables. The main finding is that this set of variables explain 55%, 58% y 64% of income dispersion for each respective year with education and job status being the relevant variables. 22 On the other hand, Behrman (1987) classifies Mexico as an overachiever in what comes to the relation between economic development and educational progresses in the context of developing countries. 23 The scatter diagram is based on 317 observations from five different years. The trend line represents the least square regression line given by: S= -13.17 + 2.28 Ln(GDPcap) Adj.R2=0.68 (-18.7) (26.0) t-values in parentheses The application of Ramsey's RESET test to this regression equation failed to detect a specification error; unlike with 14  the 1980s, however, the growth of educational attainment in Mexico accelerated, permitting it to catch up with international standards by 1990; where its placement in figure 2 is slightly above the trend line. In the closure of Mexico's education gap vis-a-vis the rest of the world was hastened in part by the country's economic stagnation. Mexico's real GDP per capita in the mid-1990s was roughly the same as it had been in the first half of the 1980s. Nevertheless, the preceding observation should not detract from the remarkable increase in schooling that occurred during the 1980s. While the level of average schooling in Mexico increased by roughly a year per decade during 1960-1980 (from 2.76 to 4.77 years), it increased by two years in the 1980-1990 decade. The acceleration in schooling during the 1980s, in turn, was the product of concerted efforts to increase basic education coverage combined with advances made in the reduction of primary school repetition and dropout rates. Figure 2 Cross-Country Relation between Education Attainment and GDP 12* 10 - Mexico 1990 * * 86 /; 4o ~ 4* eico 1980 < - t g w Mexico 1970 2- Mexico 1960 0 5.00 5.50 6.00 6.50 7.00 7.50 8.00 8.50 9.00 9.50 10.00 Ln (GDP per capita) The observations pertaining to Mexico, ordered by date, are as follows: Year Average Schooling (years) Ln (GDP per capita; 1980 US$) 1960 2.76 7.95 1970 3.68 8.29 1980 4.77 8.71 1985 5.20 8.63 1990 6.72 8.67 With respect to changes in the distribution of schooling by socioeconomic groups, there are several aspects to be considered. In particular, three of them are examined here: the changes in this distribution related to gender, economic sector and age. Table 10 shows the schooling distribution by gender from 1988 to 1997. There one can see that, even though there were clear improvements for both males and females, which translates to an upgrade of educational attainment, women achieved a better performance during that period, especially at the top of the distribution. Improvements for males, on the other hand, were more evenly spread over the entire distribution. Nevertheless, in 1997 it is possible to state that women the alternative specification of type: S= a + bX + cX2. 15  were undoubtedly more educated than men, as their cumulative distribution dominates that of men (see figure 3).24 Table 10. Evolution of Educational Distribution by Gender (%) Educational Group Primary Primary Lower Secon. Upper Secon. University Incomplete Complete Complete Complete Complete 1988 Male 19.0 30.1 24.5 14.6 11.8 Female 17.3 22.2 23.2 29.1 8.2 Total 18.5 27.7 24.1 18.9 10.7 1997 Male 13.0 25.7 28.4 18.0 14.9 Female 12.2 20.0 22.3 30.1 15.5 Total 12.7 23.7 26.3 22.1 15.1 Source: Owin calculations based on the ENEU survey (3rd quarter). Figure 3 Cumulative Educational Distribution by Gender, 1997 80 60 , 40 20 0 Universitv Upper Sec Lower Sec Prim Com Prim Incom Educational Level --- -Male - Female Source: Own calculations based on ENEU data With respect to the distribution of schooling by economic sector, table II shows that there has been a significant upgrade from 1988 to 1997. Three points, nonetheless, deserve to be stressed. First, financial and social services industries became relatively more intensive in the use of high- skilled labor. Second, the primary sector, together with non-manufacturing industry and other services, were characterized by more intensive use of low-skilled labor. Third, in a surprising way, the manufacturing industry, in contrast to what seems to be the common wisdom, cannot be characterized as a sector that intensively uses high-skilled labor. 24 This remark is true for the 1997 overall distribution relative to the 1988 one. 16  Table 11. Evolution of Educational Distribution by Economic Sector (%) Educational Group Primary Primary Lower Secon. Upper Secon. University Incomolete Comolete Complete Complete Complete 1988 Primary Sector 41.1 21.0 13.3 14.3 10.3 Manufacturing Industry 16.2 33.3 27.8 14.7 8.0 Non Manufacturing Industry 36.6 28.5 14.7 9.0 11.2 Commerce 18.0 28.7 28.8 18.7 5.8 Finance Services/Rent 4.8 6.1 19.5 47.1 22.5 Transportation/communication 14.4 35.7 26.0 18.9 5.0 Social Services 11.3 17.6 21.7 28.2 21.2 Other Services 32.8 36.6 20.2 8.1 2.3 Total 18.5 27.7 24.1 18.9 10.7 1997 Primary Sector 28.1 27.4 17.7 10.9 15.9 Manufacturing Industry 11.0 29.5 32.7 18.2 8.7 Non Manufacturing Industry 28.6 31.7 18.4 10.0 11.4 Commerce 12.4 23.4 30.6 24.1 9.5 Finance Services/Rent 2.7 5.4 16.1 40.3 35.6 Transportation/communication 9.1 26.8 32.2 23.9 8.0 Social Services 6.0 13.2 21.1 29.6 30.0 Other Services 26.2 35.7 24.6 11.1 2.4 Total 12.7 23.7 26.3 22.1 15.1 Source. Own calculations based on the ENEU (3d quarter). Another relevant observation is that the age groups also experienced upgrades in their educational attainment, as the distribution by educational level in 1997 is above the one in 1988 (table 12). In an attempt to reach a better understanding of this event, it is interesting to contrast the time and cohort effects.'5. In order to do this, one can look at the first age groups, 16-25 and 26-34, like synthetic cohorts. Namely, the 26-34 age group in 1997 can be directly compared to the 16-25 in 1988, and, in to a lesser extent, the 35-49 in 1997 to the 26-34 in 1988. From 1988 to 1997 the percentage of those in the primary incomplete level decreased, this reduction was higher than that experienced by the 16-25 age group (later being the 26-34 in 1997). The opposite took place for the highest level of instruction. In other words, it seems that the improvements throughout the educational process in Mexico are significant, both for those entering the system (higher coverage) and for those already in there (higher efficiency). 25 The time effect refers to the comparison of the same age group in two different points of time. 17  Table 12. Evolution of Educational Distribution by Age Groups (%) Educational Group Primary Primary Lower Secon. Upper Secon. University Incomplete Complete Complete Complete ComMlete 1988 16-25 8.5 26.5 - 36.7 23.7 4.6 26-34 12.6 23.7 23.1 22.5 18.2 35-49 24.0 33.3 16.8 14.3 11.6 50-65 46.1 27.2 9.9 9.0 7.8 Total 18.5 27.7 24.1 18.9 10.7 1997 16-25 5.8 23.8 38.7 25.5 6.2 26-34 6.9 19.5 28.1 27.0 18.5 35-49 14.8 25.8 19.5 19.1 20.7 50-65 37.3 27.6 11.5 10.6 13.0 Total 12.7 23.7 26.3 22.1 15.1 Source: Own calculations based on the ENEU (3rd quarter). Also concerning the interaction between age and education, one can argue that the effect of developments in the educational system is more important for the new generations than for the elderly. To investigate this, it is necessary to contrast the behavior of inequality between different age groups to that of inequality within synthetic cohorts and in relation to education. As seen above, the younger cohorts are in fact better educated. At the same time the "within" income dispersion for the youngest cohorts seems to increase over time, compared the internal Theil in 1997 and 1988 (see table 6). Thus, it becomes easier to understand why the gross contribution of age to inequality has been going up and at the same time its marginal contribution has been decreasing. In other words, differences in both educational attainment and distribution among cohorts have become pronounced in recent times, leading to a higher correlation (negative) between education and age. 5. THE DYNAMIC DECOMPOSITION S.1 ANSTYLIZED VIEW In order to address the relationship between education (the result of the interaction between supply and demand) and earnings inequality it is necessary to explain the role of the labor market, since the way it works determines the earnings differentials among workers with different educational attributes. Thus, this relationship can be viewed as being determined by two elements: (i) the distribution of education itself; and (ii) the way the labor market rewards educational attainment. The first element reflects a pre-existing social stratification that already entails some inequality, due to reasons other than the workings of the labor market itself. The second is associated to the degree of growth of this pre-existing inequality into earnings inequality due to the performance of the labor market (i.e. demand behavior). The diagram below shows the distribution of education in the horizontal axis (in, is an indicator of the average schooling of the labor force and i, represents its dispersion) while the vertical axis has the distribution of earnings. The first quadrant depicts the interaction between the pre-existing conditions (the distribution of education) and the workings of the labor market, through the steepness s,of the income profile related to education. Therefore, at a point of time: (i) the higher m, is, the larger the average earning will be; (ii) the lower i, is, the smaller the earnings inequality will be; and (iii) the higher s, is, the bigger the growth of pre-existing disparities. and, accordingly. the higher the earnings inequality will be. As these indicators change over the time. there are going to be alterations in the income distribution induced by them: changes in i,, assuming s, constant, will change earnings inequality due to changes in the composition of the labor force (the so-called 18  allocation/population effect), whereas changes in s, will produce alterations in the earnings differentials (the income effect). Figure 4. An Stylized View Between Education and the Labor Market Interaction Barros and Reis (1991) developed three synthetic measures for the indicators m, (average schooling), it (schooling inequality), and s, (income profile), based directly on the definition of the Theil T index (see annex 2.3). The figures for Mexico from 1988 to 1997 are presented in the table below. As it can be seen, there was some improvement on average schooling, but the inequality of the distribution of education has deteriorated over the period studied, whereas the income profile, which is related to the returns to schooling, has become much steeper. Meaning that, there was a shift in demand towards high skilled labor that was not met by the increase in supply probably due to the increased rate of skill-biased technological change, whose transmission to Mexico may be facilitated by the increased openness of the economy. The same pattern observed for the overall sample holds for the 16-25 years old age group: the m, goes up from 0.561 to 0.574 in 1988 through 1997; the i, increases from 0.0196 to 0.0218, whereas the s, doubles going from 0.0196 to 0.0383. Table 13. Synthetic Indicators of Schooling Distribution and Income Profile Year 1988 1992 1996 1997 mt 0.476 0.491 0.511 0.510 it 0.066 0.069 0.076 0.075 St 0.066 0.102 0.122 0.111 Source: Own calculations based on the ENEU survey (3rd quarter). 5.2 METHODOLOGY The dynamic decomposition analysis is a suitable tool for translating this stylized view in quantitative results, giving one a better understanding of the socioeconomic transformations responsible for changes in the earnings distribution. Besides permitting the identification of the relevant individual variables, it also helps in understanding the nature of the contribution of each variable to the evolution of earnings inequality over time. It has been shown by Ramos (1990), following Shorrocks (1980), that it is possible to break down the change in inequality between two points in time. This is done according to whether it can be attributed to changes in the socioeconomic groups relative to incomes, to group sizes or in their internal inequalities, through the use of the Theil T index. In generic terms, as shown before in a slightly different way, for a given partition of the population, the inequality indices of this class can be written as: I = I(ag, f3g ig) 19  where: ag is the ratio between the average income of group g and the average income of the whole population, Ig is the proportion of the population in group g, and Ig is the internal dispersion of incomes in group g. Of course the a' s are related to the indicator s in the previous picture, as well as the ,B's refer to m, and i,. In this context, the population or allocation effect corresponds to the variation induced in the inequality index I by modifications in the allocation of the population among the groups (changes in the ,B's), with no direct changes in the group's relative incomes (a' s).26-27 The income effect corresponds to the changes in I induced by changes in group incomes (a's), without changing the group population shares ( ,'s), and the internal effect is the change in the inequality caused only by modifications in the dispersions at group level (the Ig's).2 The expressions corresponding to the Theil T index are derived in annex 2.2. 5.3 RESULTS The results of the decomposition of the variations in the Theil T index for different intervals of time are shown in the table below. The first point to highlight is the fact that, when the variables are considered alone, education has the highest gross contribution to the explanation of changes in earnings distribution. Second, both the allocation and the income effect were positive in all periods. This means that the changes in the distribution of education and in the relative earnings among educational groups were always in phase with the alterations in the earnings distribution. Namely, when the income profile related to education became steeper and the inequality of education increased, the earnings distribution worsened (as in the 1988-1992, 1992-1996, and 1988-1997 periods), and vice-versa (as in the 1996-1997 period). Third, the income effect is always the prevalent one. If one considers, for instance, the 1988- 1997 period, the changes in the relative earnings among educational groups alone, would have generated deterioration in the earnings distribution higher than the one observed. To lesser extent, the same holds true for the other periods.29 Even the decrease in inequality observed between 1996 and 1997 is partially explained by the changes in relative earnings (it is possible to see that in table 13 the income profile related to education became less steep in this period). Therefore, it seems reasonable to conclude that the income effect is the leading force underlying the increase in inequality, and that, in turn, suggests that the workings of the labor market, and its interaction with the educational policies, should be thoroughly examined. Fourth, it is worth pointing out that the significance of changes in the distribution of education remains high even when one controls for changes in other relevant variables.03' As a matter of fact, with the exception of the 1996-1997 transitional period, the marginal contribution of age, economic sector and status in the labor market are usually negative. This means that the changes in these variables contributed to reduce the effects induced by changes related to education, as 26 The difference benveen this and what Knight and Sabot (1983) call the "compression" effect is that in the present exercise we are including the indirect change induced in Ithrough the variation in the weights ofthe 18.s. 2 Of course the individuals a's change as the P's change, since the overall average income is altered. This indirect impact is also computed in the composition effect (see the annex 2.2). 28 The methodology applied by Fields (1996) and Bouillon, et. al. (1998) makes important assumptions as it was pointed out in section 111.2 . On the other hand, Sz&kely (1995), in order to explain the inequality changes between two points of time, applied a methodology that differs drastically from the dynamic decomposition since he does not control for the effects that arise from changes in the population distribution and from changes in the relative income groups earnings considered in the partition of the population (see annex 2.4). 29 Of course the explanation for such a phenomenon is that the changes in the other variables worked in the direction of attenuating the changes in the rewards to education. 30 Szekelv (1995) concludes that for the 1984-1989 period, the variables that highly contributed to explaining inequality were education and economic sector, while education and job status were significant in the 1984-1992 period. The selected variables were education, occupation. region, economic sector, and job status. Bouillon, et al. (1998). applying Bourguignon methodology to the ENIGH, found that the retum effect to the household characteristics (age/gender. education/age. assets) explained 49% of the increase in the Gini between 1984 and 1994. education being the most important explanatory variable. The region effect (urban/ rural) was 9%, the south effect 15%. And the population effect 23%. 20  most of the time they work in the direction of reducing inequality after the influence of education is accounted for. Table 14. Results of the Dynamic Decomposition Period Variable Allocation Income Gross Marginal Education 11.4 58.8 70.2 30.5 1988-1992 Age -1.8 21.9 20.2 -5.2 Sector -0.6 7.8 7.1 -17.7 Status 3.9 15.1 19.0 -7.4 Education 23.9 32.8 56.7 27.6 1992-1996 Age 11.1 10.5 21.6 10.5 Sector -5.4 25.4 20.0 10.5 Status 1.2 12.4 13.6 -4.2 Education 2.2 15.5 17.7 24.2 1996-1997 Age -0.4 5.9 5.5 12.5 Sector 0.4 1.0 1.4 18.4 Status 1.4 6:1 7.5 7.8 Education 35.8 108.4 144.1 33.7 1988-1997 Age 7.4 32.7 40.1 -19.9 Sector -6.6 43.2 36.6 -40.6 Status 9.0 20.2 29.2 -35.6 Source. Own calculations based on the ENEU (3`= quarLer). The last period, from 1996 to 1997, deserves special comment. First because inequality was substantially reduced. Secondly because, once more, there were alterations associated with education, now working in the other direction, and such alteration appear to be the main factor responsible for the reduction in inequality. As it can be seen from the synthetic indicators, there was a small improvement in the distribution of schooling during the period and, a sizable decrease in the steepness of income profile related to education. All other variables, as observed for other periods, also contributed to an improvement in earnings inequality. The next table shows the results of the same kind of decomposition for Brazil, Argentina and Peru. The significance of education as an explanation of changes in inequality seems to be a common pattern in Latin American countries. Moreover, the relevance of the income effect over the allocation (population) effect is also a trait shared by all countries where a similar analysis was carried out. Interestingly, in the Mexican case the figures are above those for other countries (in a shorter period of time length. one should stress). That means that the changes in the structure of supply and demand for labor, which are greatly affected by the educational and macroeconomic policies followed by the country and/or their interaction with the workings of the labor market, were particularly relevant for the earnings distribution. Table 15. Education and Inequality Variation: Brazil, Argentina and Peru Country Author(s) Period Explanatory Income Power (%)* Effect (%) Brazil Ramos and Trindade (1992) 1977/1989 6-20 10-17 Argentina Fiszbein (1991) 1974/1988 54-56 38-46 Peru Rodriguez(1991) 1970/1984 32-47 34-43 *The explanatory power is the income plus the allocation/population effect. 21  6 COA'CLUSIONS Even though the educational attainment levels expanded very rapidly, Mexico also experienced a pronounced increase in the degree of income inequality over the period of analysis. Most of the worsening of income distribution happened in the mid-eighties and early nineties, display little variation in earnings inequality except for a small trend towards deterioration. There are two main differences in the pattern shown by the earnings and total current income distribution. First, the gains are not limited to the richest 10%, as ihose in the seventh, eight, and nine tenths of the distribution who also improved their relative earnings over the period by almost two percentage points. Second, there is a clear worsening in the earnings distribution in the present decade throughout 1996. On the other hand, the inequality associated with the total current income was moderately stable in the nineties, displaying an improvement in 1996. The different behavior between total current income and labor earnings inequalities from 1994 to 1996 gives support to the idea that the poor, who rely the most on labor as a source of income, were the least able to protect themselves during the recession. However, the substantial drop in earnings inequality from 1996 to 1997 is a surprising finding. This paper shows that education is by far the variable that accounts for the largest share of earnings inequality in Mexico, both in terms of its gross and marginal contributions. In addition, the gross contribution of age to inequality has been going up and at the same time its marginal contribution has been decreasing. In other words, differences in both educational attainment and distribution among cohorts have become pronounced in recent times, leading to a higher correlation (negative) between education and age. The education contribution to inequality in Mexico is the second highest in Latin America, next to Brazil. Moreover, what seems to be particularly interesting in the Mexican experience is the fact that the significance of education has been increasing over time. The increase in earnings inequality, however, does not appear to be the result of a worsening in the distribution of education, whereas the income profile, which is related to the returns to schooling, has become much steeper. Meaning that there was a shift in demand towards high skilled labor that was not met by the increase in supply probably due to the increased rate of skill- biased technological change, whose transmission to Mexico may be facilitated by the economy's increased openness. 22  ANVNEX 1. DA TA SOURCES The National Household Income and Expenditure Survey (ENIGH) and the National Urban Employment Survey (ENEU) were used in this study. 1.I ENIGH The National Household Income and Expenditures Survey is collected by the Instituto Nacional de Estadistica, Geografia e Informatica (INEGI). This survey is available for 1984, 1989, 1992, 1994 and 199632. Each survey is representative at the national level, urban and rural areas. For 1996, the ENIGH is also representative for the states of Mexico, Campeche, Coahuila, Guanajuato, Hidalgo, Jalisco, Oaxaca and Tabasco. For each year the survey design was stratified, multistage and clustered. The final sampling unit is the household and all the members within the household were interviewed. In each stage, the selection probability was proportional to the size of the sampling unit. Then, it is necessary to have the use of weighs 3 in order to get suitable estimators. The table below shows the sample size for each year. Table 1. Sample Size by Year Year Number of Number of households persons 1984 4.735 23.756 1989 11,531 56,727 1992 10,530 50,378 1994 12,815 59,835 1996 14,042 64,359 The available information can be grouped into three categories: - Income and consumption: the survey has monetary, no monetary and financial items. - Individual characteristics: social and demographic, i.e., age, schooling attendance, level of schooling, position at work, sector, etc. e Household characteristics. Category Selection For the purpose of the analysis, the individuals in the sample were classified according to their educational level, position in occupation, sector of activity and geographical region in the following categories: a) Educational level i) Primary incomplete: no education and primary incomplete (one to five years of primary) ii) Primary complete: primary complete and secondary incomplete (one or two years) iii) Secondary complete: secondary complete and preparatory incomplete (one or two years) iv) Preparatory complete: preparatory complete and university incomplete v) University complete: university complete (with degree) and postgraduate studies 32 The sample in a given year is independent from another. 33 The weights should be calculated according to the survey design and corresponds to the inverse of the probability inclusion. 23  b) Position in occupation i) Worker or employee ii) Employer iii) Se If employed c) Sector of activity i) Agriculture ii) Manufacturing iii) Construction iv) Commerce v) Services vi) Other (utilities, extraction, transports, financial services, communications, etc) d) Geographical regions i) North: Baja California, Baja California Sur, Coahuila, Chihuahua, Durango, Nuevo Leon, Sinaloa, Sonora, Tamaulipas and Zacatecas ii) Center: Aguascalientes, Colima, Guanajuato, Hidalgo, Jalisco, Mexico, Michoacan, Morelos, Nayarit, Puebla, Queretaro, Sari Luis Potosi and Tlaxcala iii)South: Campeche, Chiapas, Guerrero, Oaxaca, Quintana Roo, Tabasco, Veracruz and Yucatan iv) Distrito Federal. Group Selection The labor force was limited to individuals who are: i) working as employee, employer or self employed34; ii) between 12 and 65 years old; iii) living in urban areas; iv) working 20 hours or more per week; v) with positive income; vi) having the attributes of interest defined. The number of persons in the survey that belong to the labor force is shown in the next table. Table 2. Sample size for the labor force Year Number of % of the total persons sample 1984 3,892 16.4 1989 10,401 18.3 1992 8.752 17.4 1994 10.982 18.4 1996 12,996 20.2 According to the groups mentioned we have that, 34 The respective categories: workers without payment and cooperative members were excluded because of the sample size. 24  Table 3. Sample size by variable and year Variable 1984 1989 1992 1994 1996 Education Level Primary Incomplete 1,246 1,951 1,879 2,387 2,736 Primary Complete 1,299 3,006 2,501 2,975 3,411 Secondary Complete 803 2,875 2,489 3,014 3,734 Preparatory Complete 389 1,614 1,168 1,617 1,915 University Complete 245 955 715 989 1,2C0 Position in Occupation Employee 3,175 8,604 7,188 8.843 10,207 Employer 126 311 393 450 610 Self employed 681 1.486 1.171 1,689 2,179 Total 3,982 10,401 8,752 10,982 12,996 1.2. ENEU The National Urban Employment Survey is also a micro-level data set collected by INEGI and contains quarterly wage and employment data over the last ten years (1987-1997). Currently, the data is representative of the 41 largest urban areas in Mexico, covering 61% of the urban population following the 2500 inhabitants or more criteria, and 92% of the population who live in metropolitan areas with 100,000 or more inhabitants. In 1985, the ENEU included 16 urban areas: Mexico City, Guadalajara, Monterrey, Puebla, Leon, San Luis Potosi, Tampico, Torreon, Chihuahua, Orizaba, Veracruz, Merida, Ciudad Juarez, Tijuana, Nuevo Laredo and Matamoros, covering 60% of the urban population for that year. In 1992, 18 more urban areas were included in the survey: Aguascalientes, Acapulco, Campeche, Coatzacoalcos, Cuernavaca, Culiacan, Durango, Hermosillo, Morelia, Oaxaca, Saltillo, Tepic, Toluca, Tuxtla Gutierrez, Villahermosa, Zacatecas, Colima and Manzanillo. In 1993 and 1994, Monclova, Queretaro, Celaya, Irapuato and Tlaxcala entered to ENEU. Finally, Cancun and La Paz joined the survey in 1996. As can be seen in the previous description, the ENEU has always covered about 60% of the national urban population. Therefore, the results deduced from this survey allow one to know and assess the socioeconomic and employment characteristics of the national urban areas. The data is from household surveys, which fully describe family composition, human-capital acquisition, and experience in the labor market (the variables contain information about social household characteristics, activity condition, position in occupation, unemployment, main occupation, hours worked, earnings, benefits, secondary occupation, and searching for another job). As the ENIGH, the sampling design was stratified, in several stages (where the final selection unit was the household), and with proportional probability to size3. So this statistical construction allows us to make comparisons among different years. Moreover, this survey is structured to generate a panel data set which has the characteristic to conform to a rotator panel (a fifth of the total sample goes out and a new one comes in every quarter). Hence, the panel data follows the same household through out five quarters. 35 For this it was necessary to use weights or expansion factors. 25  Category Selection The individuals in the sample were classified according to their educational level, age, sector of activity, position in occupation, hours worked and geographical region in the following categories: a) Educational level i) Primary incomplete: no education and primary incomplete (one to five years of primary) ii) Primary complete: primary complete and secondary incomplete (one or two years) iii) Secondary complete: secondary complete and preparatory incomplete (one or two years) iv) Preparatory complete: preparatory complete and university incomplete v) University complete: university complete (with degree) and postgraduate studies b) Age i) 12 to 25 years old ii) 26 to 34 years old iii)35 to 49 years old iv) 50 to 65 years old c) Sector of activity i) Primary sector (Includes Agriculture, Forestry, Fishing, and Mining). ii) Manufacturing industry iii) Not manufacturing industry (Includes Construction, and Utilities) iv) Commerce v) Finance services and Rent vi) Transportation and Communication vii) Social services (Tourism, education, health, public administration, embassy) viii) Other services d) Labor Market Status i) Employer ii) Self employed iii) Informal salaried: people that work in an enterprise with 15 workers or less and do not receive social security (IMSS, ISSTE, private, etc.) iv) Formal salaried: people that work in an enterprise with 16 workers or more or receive social security (IMSS, ISSTE, private, etc.) v) Contract e) Hours worked i) 20 to 39 hours per week ii) 40 to 48 hours per week iii)At least 49 hours per week f) Geographical regions i) North:, Baja California, Baja California Sur, Coahuila, Chihuahua, Durango, Nuevo Leon, Sinaloa, Sonora, Tamaulipas and Zacatecas ii) Center: Aguascalientes, Colima, Guanajuato, Hidalgo, Jalisco, Mexico, Michoacan, Morelos, Nayarit, Puebla, Queretaro, San Luis Potosi and Tlaxcala iii)South: Campeche, Chiapas, Guerrero, Oaxaca, Quintana Roo, Tabasco, Veracruz and Yucatan iv) Distrito Federal. 26  Group Selection Analogous to the EN1GH, the sample is: i) between 16 and 65 years old; ii) living in urban areas (localities with at least 2,500 inhabits); iii) regular workers (non-seasonal workers); iv) working 20 hours or more per week; v) with positive earnings36; vi) having the attributes of interest defined. The table below shows the sample size and labor force. Table 4. Sample Size by Year Number of persons N'ear Total Labor force 1988 124,322 45.870 1989 125,820 47,630 1990 127,387 48,109 1991 126,262 48,080 1992 235,696 91,279 1993 239,394 90,860 1994 246,906 102,105 1995 252,563 100,838 1996 262,478 108,159 1997 272,356 116,559 36 In this survey an additional adjustment had to be made: if the worker got a benefit at the end of the year ("aguinaldo"), then the wage was expanded (wve assumed that this benefit as equivalent to 30 days of wages a year). 27  ANNEX 2. METHODOLOGICAL NOTE 2.1 GIAN IANDEX The Gini Index is defined by GI = 2 cov[Y, F(Y)] (X) where: Y is the distribution of per capita income Y=(yj,. .., yn) where y, is the per capita income of individual i, i=l,...,n 11 is the mean per capita income F(Y) is the cumulative distribution of total per capita income in the sample (i.e. F(M=V(Yd),...f(Y)] where f(yd is equal to the rank of ), divided by the number of observations (n)) . Gini deconrposition Equation (1) can be re% Titten and expanded into an expression for the Gini coefficient that captures the "contribution to inequality" of each of the K components of income (see Leibbrandt (1996)). K G =. ERkGkSk (2) k=1 where: Sk is the share of source k of income in total group income (i.e. Sk=PkVP) Gk is the Gini coefficient measuring the inequality in the distribution of income component k within the group 38 Rk is the Gini correlation of income from source k with total income The larger the product of these three components, the grater the contribution of income from source k to total inequality. 3 Both the covariance and cumulative distribution are computed using the household weights 38 Rk is defined as: R = cov[Yk, F(Y)] cov[Yk, F(Yk )] 28  2.2 THEIL T INDEX" This index was calculated as follows40: T=(n) EJ(J InCY ) (3) where: Y, is the income of the i-th individual Y is the average income n is the population size. Static Decomposition of the Theil Index If the population is divided into G groups with ng observations each, it is then possible to write (3) as: T=E ( i ) E ( Y,g ) In( Y'g ) (4) where: Y,g is the income of the i-th individual of the g-th population subgroup. If we now define 13g = X and Zg = k where Yg is the average income of the g-th group and k is a reference income, it is possible to show, after some algebraic manipulation, that T can be expressed as: T - log Zg InZg - Ink + k Zg Tg (5) where: k=Y1PgZg Tg is the Theil index for the g-th group. The first two terms on the right hand side of (5) correspond to the between group inequality, and the third one to the within group inequality. 3 Theil's T is sensitive to changes at the bottom and the top tail of the distribution. 40 The mathematical notation from in annex 2.2 through 2.3 2 follows Ramos (1990). 29  Choosing the mean income as the reference income, i. e., Zg = ag = , expression (5) simplifies to: G G T=Eag Pg Inag + Yag Pg Tg (6) g=1 g=) The first term in (6) is said to be the between group inequality, and the second term the within group inequality. Dynamic Decomposition Analysis By totally differentiating (6), we have: dT=z aT d ,g+ E aT dag + E-d Tg (7) * The first term on the right hand side is the population allocation effect (changes in T caused exclusively by population shifts); * The second term is the income effect (changes in T induced exclusively by changes in standardized mean incomes), and; * The third one is the internal effect (changes in Tcaused by changes in internal dispersion). It can be shown that: G G - = ag In ag -ag Yag Pg (I+ Inag) + ag Tg - cg ag Pg Tg (8) ag g=I g=I 89T G G =3,g(I+lnag)-Pg Eag Pg (IJ+lnag)+ Pg Tg -fg agg Pg Tg (9) aag g=1 g=I dT_ UT =ag ,Bg (I10) a Tg Replacing (8), (9), and (10) into (7) and simplifying we obtain G G G dT = 2ag( Inag + Tg - T - I) d ,Bg + Pg( Incag + Tg - T) d ag + Y,(cg P3g) d Tg (I 1) g=I g=l g=l The three terms on the right hand side of (10) correspond to the allocation, income, and internal effects, respectively. For estimation purposes, equation (I1) must be approximated. The convention used in the empirical exercises was to evaluate the expression at the middle points. 30  2.3 LEVEL, INEQUALITY AND THE INDICATOR OF STEEPNESS OF THE INCOME PROFILES IN EDUCATIONA L LEVEL Ramos (1990) used three synthetic measures for the indicators m, (average schooling), i, (schooling inequality). and s, (income profile), based directly on the definition of the Theil Index. The calculations of the principal parameters cg, Pg, and Tg (see equation 5), could determine the changes in the distribution by level of education (g groups in this category). These parameters allow us to analyze the trend in educational income differentials, the distribution of the population in each educational level and the inequality among them. Three synthetic measures are used to summarize the changes related to education: m, is the average level of schooling for the year t. i, is the degree of inequality in the distribution of education for year t. s, is the variations in the income ratios associated with education for year t. These measures can be calculated as follows: m,= c3aig g Eag,Bg log(a,) g i, = Sag',B r log ag pg ) g Ea' 5 ; log(ag sr=g log L a' P.' g where: ag is the standardized income of educational category g for the reference year P3g is the fraction of the labor force in the g-th educational category in year t pg is the value 3g in the reference year. s, can be understood as an indicator of the relative steepness of the income profiles related to education. If one fixes the fraction of the labor force in each educational group, it follows that the steeper the income profile the larger the between group inequality. i, corresponds to the Theil T index that would prevail in a population with no inequality within the educational groups. and where the group incomes were proportional to the group average incomes in the base year. 31  ANNEX3. EVOLuTIONOFINEQUALITY Table 1. Decomposition of Total Current Income Income Source Gini coeficient Share in Gini correlation Contribution to Gini Percentage by income Total with total income coeMcient of total share in overall source income rakings income Gini 1984 Earnings 0.6428 0.4688 0.7249 0.2184 46.0 Monetary income excluding 0.7568 0.3191 0.6470 0.1562 32.9 earnings No monetary current 0.6067 0.2120 0.7750 0.0997 21.0 income Total 0.4744 1.0000 1.0000 0.4744 100.0 1989 Earnings 0.6128 0.4635 0.7562 0.2148 41.0 Monetary income excluding 0.8185 0.3109 0.7410 0.1886 36.0 earnings No monetary current 0.6541 0.2256 0.8187 0.1208 23.0 income Total 0.5242 1.0000 1.0000 0.5242 100.0 1992 Earnings 0.6440 0.4541 0.7790 0.2278 42.9 Monetary income excluding 0.8129 0.2848 0.7316 0.1694 31.9 earmings No monetary current 0.6079 0.2611 0.8449 0.1341 25.2 income Total 0.5313 1.0000 1.0000 0.5313 100.0 1994 Earnings 0.6690 0.4932 0.8123 0.2680 50.2 Monetary income excluding 0.7948 0.2550 0.6827 0.1384 25.9 earnings No monetary current 0.6051 0.2518 0.8365 0.1274 23.9 income Total 0.5338 1.0000 1.0000 0.5338 100.0 1996 Earnings 0.6514 0.4725 0.7870 0.2422 46.7 Monetary income excluding 0.7924 0.2802 0.6884 0.1529 29.4 earnings No monetary current 0.6026 0.2472 0.8325 0.1240 23.9 income Total 0.5192 1.0000 1.0000 0.5192 100.0 Source: Own estimates based on ENIGH 32  Table 2. Pearson Correlation among explanatory variables 1988 Education Occupation Econ. Sector Status Education 1.00 Occupation 0.64 1.00 Econ. Sect. 0.08 0.10 1.00 Status 0.05 0.06 -0.04 1.00 Spearman's rho"' 0.58 1992 Education Occupation Econ. Sector Status Education 1.00 Occupation 0.63 1.00 Econ. Sect. 0.06 0.02 1.00 Status 0.08 0.08 -0.04 1.00 Spearman's rho' 0.60 1997 Education Occupation Econ. Sector Status Education 1.00 Occupation 0.64 1.00 Econ. Sect. 0.09 0 04 1.00 Status 0.11 0.09 -0.06 1.00 Spearman's rho" 0.62 Source Own calculation based on ENEU Survey. I/ It refers to Spearmnan's correlation betreen education and occupation. 33  ANNEX 4. DECOMPOSiTION RESUL TS USING ENIGH DATA Table I General Statistics for the Static Decomposition" 1984 1989 1992 . 1994 1996 Between Within Between Within Between Within Between Within Between Within Education2' Primary Incomplete -0.10 0.07 -0.07 0.04 -0.08 0.06 -0.08 0.03 -0.07 0.03 Primary Complete -0.03 0.08 -0.06 0.09 -0.08 0.05 -0.08 0.05 -0.07 0.04 L-Secondary Complete 0.04 0.05 -0.02 0.08 -0.02 0.08 -0.04 0.09 -0.04 0.08 U-Secondary Complete 0.07 0.04 0.07 0.07 0.10 0.08 0.09 0.08 0.09 0.08 University Complete 0.11 0.04 0.20 0.11 0.24 0.09 0.30 0.11 0.26 0.11 Total 0.08 0.28 0.12 0.39 0.16 0.36 0.20 0.36 0.17 0.35 Position in Occupation2' Employee 0.01 0.25 -0.05 0.32 -0.03 0.33 0.00 0.41 0.00 0.36 Employer 0.05 0.04 0.08 0.08 0.13 0.10 0.06 0.06 0.07 0.08 Self-employed -0.03 0.05 -0.01 0.09 -0.04 . 0.04 -0.04 0.06 -0.04 0.04 Total 0.02 0.34 0.03 0.48 0.06 0.46 0.02 0.53 0.03 0.48 Sector2l Agriculture -0.03 0.05 -0.02 0.02 -0.01 0.06 -0.02 0.02 -0.01 0.04 Manufacturing 0.00 0.06 -0.03 0.08 -0.01 0.11 -0.03 0.09 -0.02 0.09 Construction -0.01 0.02 -0.01 0.07 -0.02 0.03 -0.02 0.03 -0.02 0.04 Commerce 0.01 0.07 0.02 0.12 -0.01 0.09 -0.02 0.08 -0.02 0.08 Services 0.02 0.10 0.03 0.17 0.04 0.17 0.09 023 0.06 0.19 Others 0.03 0.04 0.02 0.04 0.03 0.05 0.04 0.07 0.03 0.06 Total 0.01 0.35 0.01 0.50 0.01 0.51 0.03 0.52 0.01 0.50 Source: Own estimates based on ENIGH I/ Labor force was limited to individuals who are i) working as emplovee, employer or self employed; ii) between 12 and 65 sears old; iii) living in urban areas; iv) working 20 hours or more per week: v) with positive income: vi) having the attributes of Interest defined. 2/ Annex I presents how the group categories are defined Table 2. Synthetic Indicators of Schooling Distribution and Income Profile Year 1984 1989 1992 1994 1996 ml 0.468 0.525 0.516 0.527 0.538 it 0.083 0.073 0.075 0.075 0.072 si 0.083 0.124 0.161 0.203 0.175 Source: Own estimates based on ENIGH 1/ Labor force was limited to individuals who are: i) working as employee, employer or self emplo)ed, ii) between 12 and 65 years old; iii) living in urban areas: i%l) Aorking 20 hours or more per sseek. %lI with positise income. vi) having the attributes of interest defined. 2/ Annex I presents how the group categories are defined 34  Table 3. Results of the Dynamic Decomposition Period Variable Allocation Income Gross Marginal Education -9.8 52.1 42.4 26.6 1984-1992 Pos. in Occupation 21.1 18.2 39.3 13.8 Sector -17.7 2.5 -15.2 -19.5 Education -3.6 63.0 59.5 46.0 1984-1994 Pos. in Occupation 6.2 -1.2 5.1 -2.8 Sector -7.8 11.5 3.7 -11.4 Education -7.4 59.4 51.9 34.5 1984-1996 Pos. in Occupation 14.6 1.3 15.9 -9.7 Sector -27.2 3.7 -23.5 -15.1 Education 5.7 68.4 74.1 76.6 1994-1996 Pos. in Occupation -15.8 -12.4 -28.2 29.5 Sector 21.8 25.0 46.8 21.0 Source Own estimates based on ENIGH I/ Labor force was limited to individuals who are. i) working as employee, employer or self employed; ii) between 12 and 65 years old; iii) living in urban areas, iv) working 20 hours or more per week; v) with positive income; vi) having the attributes of interest defined. 2/ Annex I presents how the group categories are defined 35  REFERENCES Altimir, 0, Pinera, S. (1982), "Analisis de descomposici6n da las desigualdades de ingresso en la America Latina". El Trimestre Econ6mico, vol. 49. num. 196, pp. 813-860. Barros, Ricardo and Jose Guilherme Almeida Reis (1991), "Wage Inequality and the Distribution of Education: A Study of the Evolution of Regional Differences in Inequality in Metropolitan Brazil". Journal of Development Economics, 36: 117-143. Bouillon, Cesar, Arianna Legovini and Nora Lustig (1999), "Can Education Explain Income Inequality Changes in Mexico?" Mimeo. Bouillon, Cesar, Arianna Legovini and Nora Lustig (1998), "Rising Inequality in Mexico: Returns to Household Characteristics and the Chiapas Effect". Mimeo. Bouillon, Cesar, Arianna Legovini and Nora Lustig (1999), "Can Education Explain Income Inequality Changes in Mexico?" Mimeo. Bourguignon, Francois, Martin Fournier, and Marc Gurgand (1998), "Labor Incomes and Labor Supply in the Course of Taiwan's Development, 1979-1994". Mimeo. Bourguignon, Francois (1979) "Decomposable Income Inequality Measures". Econometrica, vol. 47, num. 4, pp. 901-920. Burfisher, Mary E., Robinson, Sherman, and Thierfelder (1993) "Wage Changes in a US-Mexico Free Trade Area: Migration Versus Stolper -Samuelson Effects". Working paper 645, Department of Agricultural and Resources Economics, University of California, Berkeley Cragg, Michael and Mario Epelbaum (1996), "Why has the wage dispersion grown in Mexico? Is it the incidence of reforms or the growing demand for skills? ". Journal of Development Economics, vol. 51, pp. 99-116. Cowell, F. (1980), "On the Structure of Additive Inequality Measures". Review of Economic Studies, vol. 47, pp. 521-53 1. De la Torre, Rodolfo (1997), "Desigualdad, Pobreza y Polarizaci6n Social en Mexico". Mimeo. Elias, Victor J., 1992, "Sources of Growth: A Study of Seven Latin American Economies". International Center for Economic Growth. Feenstra, Robert C. and Gordon H. Hanson, 1996, "Globalization, Outsourcing, and Wage Inequality". AEA Papers and Proceedings, vol. 86, num. 2, pp. 240-45. Fields, Gary S. (1996), "Accounting for Income Inequality and Its Change". Cornell University. Paper prepared for presentation at the American Economic Association annual Meetings, New Orleans, L.A. Fiszbein (1991), "A Essays on labor markets and income inequality on less developed countries". Berkeley, Ph.D. Dissertation. University of California. Hanson, Gordon and Ann Harrison (1995), "Trade, Technology, and Wage Inequality". Working paper, num. 5110. National Bureau of Economic Research Working papers Series. 36  Hanson, H. Gordon (1997), "Increasing Returns, Trade and the Regional Structure of Wages". The Economic Journal, vol. 107, num. 440, pp. 113-133. Hemandez Laos, Enrique, Nora Garro Bordonaro and Ignacio Llamas Huitr6n (1997), "Productividad y Mercado de Trabajo en Mexico", processed. Knight, J. T. and Sabot, R. H. (1983), "Educational Expansion and the Kuznets effect". American Economic Review, 73, 1132-6. Johnson, George E. (1997), "Changes in Earnings Inequality: The Role of Demand Shifts". Journal of Economic Perspectives, vol. I1, num. 2 (Spring), pp. 41-54. Juhn, Chihui, Kevin Murphy and Brooks Pierce (1993), "Wage Inequality ad the Rise in Returns to Skill". Journal of Political Economy, vol. 3, num. 3, pp.410-448. Leibbrandt M. V., Chrisopher D. Woolard and Ingrid D. Woolard (1996), "The Contribution of Income Components to Income Inequality in South Africa". Living Standards Measurement Study, Working Paper No. 125. The World Bank. Londofno, Juan Luis (1996), "Poverty, [nequality and Human Capital Development in Latin America, 1950-2025". World Bank Latin American and Caribbean Studies, (June). Lustig, Nora and Miguel Szekely (1996), "Mexico: Evolucion Economica, Pobreza y Desigualdad". Mimeo Moreno, A (1989), "La distribuici6n del ingresso laboral urbano en Colombia: 1976-1988". Desarrollo y Sociedad, Bogota, num. 24. Panuco-Laguette, Humberto and Miguel Szekely (1996), "Income Distribution and Poverty in Mexico", in Victor Bulmer-Thomas, editor, The New Economic Model in Latin America and its Impact on Income Distribution and Poverty, St. Martin's Press, New York, pp. 185-222. Psacharopoulos, G., Morley, S., Fiszbein, A., Lee, H., Wood, B. (1992), 'Poverty and income distribution in Latin America: the story ot the 1980s". The World Bank. Psacharopoulos, George (1993), "Returns to Investment in Education: A Global Update". World Bank Policy Research Paper No. 1067, (January). Ramos, Lauro (1990), 'The Distribution of Earnings in Brazil: 1976-1985", Ph.D. Dissertation, University of California, Berkeley. Ramos, L., Trindade, C. (1991), `EducacAo e ddesigualdade de salarios no Brasil: 1977/89". Perspectivas da Economia Brasileira - 1992. IPEA, Rio de Janeiro. Reis, J., Barros, R. (1989), "Um estudo da evolucao das diferencas regionais da desigualdade no Brasil". Rio de Janeiro, IPEA. Reyes, A. (1988), "Evoluci6n de la distribuici6n del ingresso en Colombia". Desarrollo y Sociedad, num.21. Rodriguez, J. (1991), "Desigualdad salarial y educaci6n en Lima Metropolitana: 1970 a 1984". Masters Dissertation, PUC/RJ. Rio de Janeiro. 37  Schultz, T. Paul (1988), "Education Investments and Returns", in H. Chenery and T.N. Srinivasan, editors, Handbook of Development Economics, vol. 1, Elsevier Science Publishers B.V., pp. 543-629. Shorrocks, A.F. (1980), "The Class of Additively Decomposable Inequality Measures". Econometrica, vol. 48, num. 3, pp. 613-625. Shorrocks, A.F. (1984), "Inequality Decomposition by Factor Components". Econometrica, vol. 52, numn. 6, pp. 1369-1384. Shorrocks, A.F. and Foster, J.E. (1985), "Transfer Sensitive Inequality Measures". Review of Economic Studies, 5: 105-38. Shorroks, A.F. and Mookherjee, D. (1982), " A decomposition Analysis of the Trend in UK Income Inequality". The Economic Journal, 92, pp. 886-902. Szekely, Miguel (1995), "Aspectos de la desigualdad en Mexico". El Trimestre Econ6mico, vol. LXX (2), num. 246, pp. 201-243. Tan, Hong and Geeta Batra (1997), "Technology and Firm Size-Wage Differentials in Colombia, Mexico, and Taiwan (China)". The World Bank Economic Review, vol. 11, num. 1, pp.59-83. "Symposium Issue on How International Exchange, Technology, and Institutions Affect Workers". The World Bank Economic Review, 1997, vol. 11, num. I (January). "Symposium on Wage Inequality". Journal of Economic Perspectives, 1997, vol. I1, num. 2 (Spring). Vieira, M.L.P. (1998), "A relagao entre educagao e desigualdade no Brasil". Masters Dissertion, EPGE. Rio de Janeiro. World Development Report (1996). The World Bank. Oxford University Press. World Bank (1999). Government Programs and Poverty in Mexico, Green Cover. World Bank (1998). Country Economic Memorandum, "Factor Productivity and Growth". 38  The Financial Crisis Impact on the Income Distribution in Mexico Background Paper # 2 Gladys Lopez-Acevedo (LCSPE) and Angel Salinas (LCC IC) 1.2 Abstract After the financial crisis of 1994 income and labor earnings distribution improved in Mexico. Usually one w ould expect inequality to go up during recessive times, as it seems plausible to admit that people at the top decile have more ways to protect their assets than those at the bottom decile do. Especially when it comes to labor which is basically the only asset of the poor (the labor-hoarding hypothesis). It is true that the Mexican economy as a whole had a strong and impressive performance in 1997. The aggregate growth rate was around 7%, real investment grew by 24% and exports by 17%, the industrial production increased by 9.7%, and the civil construction sector, which is highly intensive in less skilled labor, experienced a growth close to 11%. Under such a scenario, an improvement in distribution of income and labor earnings itself is not unlikely, but the magnitude and quickness of the recovery calls for a detailed inspection of the mechanisms responsible for it. According to the National Household Income and Expenditure Survey (ENIGH), most of the worsening of the total current income distribution in Mexico happened in the mid-eighties (1984-1989). The early nineties display little variation in total current income inequality except for a small trend towards deterioration. From 1989 to 1994, the total current income share accruing to the 20% poorest decreased slightly (it went down from 3.9% to 3.8%), whereas the richest 10% were the only ones that increased theirs (by one percentage point), and, therefore, those in the middle also experienced losses. From 1994- 1996, a period of time that entails a severe financial crisis, the 10% richest experienced relative losses (their total current income share dropped 1.6% points) and, accordingly, total current income inequality went down. The Gini coefficient came dcwn from 0.534 in 1994 to 0.519 in 1996, whereas the drop in the Theil T was from 0.558 to 0.524. In principle, it could be argued that the richest experienced severe capital losses due to the crisis (1994- 1996), in such a way that their total current income was affected compared to the poor. This hypothesis, however, is not supported by the data as monetary income other than wages and salaries, and financial income as well, increased their share in total income in that time interim, particularly so for the urban areas. Nonetheless, this paper shows that financial income is a growing source of inequality in Mexico. This paper investigates the financial crisis impact on income inequality in Mexico. i) It analyses the fall in income inequality after the crisis; ii) provides an analysis of the contribution of the various income sources to the evolution of income inequality; and, iii) investigates the factors and mechanisms that have been driving inequality in Mexico. This type of analytical work is central to the social development objective of the World Bank Country Assistance Strategy. It is also part of a comprehensive work meant to build a poverty and inequality strategy for Mexico. This research wvas completed as part of the "Earnings Inequality after Mexico's Economic and Educational Reforms" study at the World Bank. We are grateful to INEGI and SEP (Ministrv of Education) for providing us with the data. These are views of the authors, and need not reflect those of the World Bank, its Executive Directors. or countries they represent. 2 gacevedoqr!lworld bank.org and asalinas,wvorldbank.org.  1. INTRODUCTION From 1994 through 1996 income distribution improved in Mexico at a time of a severe financial crisis in the Mexican economy. According to our results from the National Household Income and Expenditure Survey (ENIGH), the top decile experienced relative losses, their total current income share dropped 1.6 % points, while the other deciles increased their share in total current income. The Gini coefficient came down from 0.534 in 1994 to 0.519 in 1996, whereas the drop in the Theil T was from 0.558 to 0.524. Usually, one would expect inequality to go up during recessive times, as it seems plausible to admit that people at the top decile have more ways to protect their assets than those at the bottom decile do. Especially when it comes to labor which is basically the only asset of the poor (the labor-hoarding hypothesis). In principle, it could be argued that the richest experienced severe capital losses due to the crisis, in such a way that their total income was affected compared to the poor. This hypothesis, however, is not supported by the data as monetary income other than wages and salaries, and financial income as well, increased their share in total income in that time interim, particularly so for the urban areas. This paper is organized as follows: Section 2 discusses the evolution of income inequality and the income share within income groups in Mexico. Section 3 measures the impact of various income sources on inequality, for the period 1994-1996. Sections 4 and 5 examine the factors and mechanisms driving inequality. Section 6 relates the fall in income inequality to the observed economic sector activity. Section 7 presents the concluding remarks. 2. EVOLUTION OF INCOME INEQUALITY Achieving sustainable economic growth with a more egalitarian income distribution is at the core of Mexico's development challenge. Yet, the country does not performr well in terms of equity when compared with other Latin American countries. According to a recent study developed by the IDB (1998), Mexico has the sixth most unequal overall household income distribution (and the third worst in urban areas). In the broader intemational context, Mexico's ratio between the income share accruing to the 10 top percent to the bottom 40 percent of the population is higher than what is observed for the high-income countries and for the vast majority of the low-income countries (see table A5.1 in Annex 5). The evaluation of the income inequality evolution in Mexico is based on the information available in the ENIGHs. This survey captures total current income of the households, including non-monetary income, besides labor earnings and other sources of monetary income. The unit of analysis is the household, and the concept of income is the household per capita total current income.3 The main results of this evaluation are shown in table 1. It indicates that a very sizable deterioration in the income distribution has taken place between 1984 and 1996. While the poorest 20% of the population lost almost one seventh of their income share (0.6 percentage points), the richest 10% increased theirs by something close to one seventh (5.2 percentage points). Moreover, this last group was the only one that gained over that period, as not only the poorest, but also those in the middle lost in relative terms. Looking at the results of this comparison, one can say that the 1984-1996 period in Mexico was marked by a series of regressive income transfers from almost the entire population spectrum 3 Total current income of the household divided by its number of household members. That is, we are considering the household as a unit characterized by a flow of income transfers and disregarding aspects related to equivalence scale. 2  to the richest stratum. Accordingly, the most commonly used inequality index points to a worsening in income inequality over this span of time. The Gini coefficient, which is more sensitive to changes in the middle of the distribution, rises from 0.473 in 1984 to 0.519 in 1996. On the other hand, the Theil T index, which is extremely sensitive to changes in the upper and lower tails, goes up from 0.411 in 1984 to 0.524 in 1996. Even though the worsening of the distribution is indisputable, there are, nevertheless, two points that must be stressed. The first one is that, according to the ENIGH survey, most of the worsening of the total current income distribution happened in the mid-eighties (1984-1989). The early nineties display little variation in total current income inequality except for a small trend towards deterioration. From 1989 to 1994, the total current income share accruing to the 20% poorest decreased slightly (it went down from 3.9% to 3.8%), whereas the richest 10% were the only ones that increased theirs (by one percentage point), and, therefore, those in the middle also experienced losses. Table 1. Lorenz Curves for Total Current Income" (accumulated income share %) Population Share 1984 1989 1992 1994 1996 10 1.66 1.39 1.32 1.39 1.39 20 4.47 3.88 3.68 3.76 3.89 30 8.19 7.29 6.92 6.98 7.29 40 12.85 11.65 11.09 11.08 11.63 50 18.76 17.05 16.26 16.28 17.08 60 26.15 23.78 22.83 22.79 23.86 70 35.51 32.25 31.13 31.10 32.39 80 47.64 43.12 42.14 41.93 43.44 90 64.53 58.75 58.32 57.68 59.33 92 68.79 63.06 62.81 62.03 63.61 94 73.73 68.03 68.03 67.26 68.68 96 79.38 73.82 74.47 73.70 74.95 98 86.68 81.60 82.81 82.49 83.32 100 100.0 100.0 100.0 100.0 100.0 Bottom 20% 4.5 3.9 3.7 3.8 3.9 Middle 40% 21.7 19.9 19.2 19.0 20.0 Middle high 30% 38.4 35.0 35.5 34.9 35.5 Top 10% 35.5 41.3 41.7 42.3 40.7 Gini 0.473 0.519 0.529 0.534 0.519 Theil T 0.411 0.566 0.550 0.558 0.524 Source: Own calculations based on ENIGH. " Based on household per capita income. The second fact to be emphasized is very surprising and hard to be explained: the observed improvement in the income distribution between 1994 and 1996, an interval of time that entails a severe financial crisis in the Mexican economy.4 Usually one would expect inequality to go up during recessive times, as it seems plausible to admit that the rich have more ways to protect their 4 In 1994, current account deficit was 30 billion dollars, about 7 percent of GDP. The main effects of the financial crisis were i) GDP and domestic demand felt 6.2 percent and 14 percent respectively each; ii) the unemployment rate rose from 3.7 percent in 1994 to 6.2 percent in 1995; and, iii) the GDP per capita decreased 7.8 percent and workers experienced a significant reduction in their real wage, nearly 17 percent in 1995. 3  assets than the poor do,.especially when it comes to labor which is basically the only asset of the poor (the labor-hoarding hypothesis). The fact, however, is that the 10% richest experienced relative losses (their total current income share dropped 1.6% points) and, accordingly, total current income inequality went down. The Gini coefficient came down from 0.534 in 1994 to 0.5 19 in 1996, whereas the drop in the Theil T wvas from 0.558 to 0.524. In principle, it could be argued that the richest experienced severe capital losses due to the crisis, in such a way that their total current income was affected compared to the poor. Tables 2 and 3 show the shares by income source within income groups and the shares by income source within income source, respectively. Some interesting results are: i) labor earnings is the largest income share for all deciles. ii) The share of total labor earnings within income group decreased substantially for the top decile (13.7%) compared to the other income groups. And, iii) the largest increase within the financial income share was for the top decile (from 7.1 % to 10.0%). Table 2 Income share by source within income groups 1994 1996 Source Bot.20% Mid.40% M.H.30