Year-to-year economy-wide measures of income distribution, such as the Gini coefficient, are rarely available for long periods except in a few developed countries, and as a result few analyses of year-to-year changes in inequality exist. But wage and earnings data by industrial sectors are readily available for many countries over long time frames. This paper proposes the application of the between-group component of the Theil index to data on wages, earnings and employment by industrial classification, in order to measure the evolution of wage or earnings inequality through time. We provide formal criteria under which such a between-group Theil statistic can reasonably be assumed to give results that also track the (unobserved) evolution of inequality within industries. While the evolution of inequality in manufacturing earnings cannot be taken as per se indicating the larger movements of inequality in household incomes, including those outside the manufacturing sector, we argue on theoretical grounds that the two will rarely move in opposite directions. We conclude with an empirical application to the case of Brazil, an important developing country for which economy-wide Gini coefficients are scarce, but for which a between-industries Theil statistic may be computed on a monthly basis as far back as 1976 [ Full Text ]

This preliminary report asks whether there exist systematic relationships between changes in economic inequality and levels of state violence in countries around the world. The question is, of course, quite natural. Entire lexicons exist that describe economic relationships in terms that evoke violence; such words and phrases as exploitation, dependency, unequal exchange and class struggle are but prominent examples. And the case histories of war, revolution, state terrorism and coups d are certainly loaded with analyses of what seem transparently to be efforts either to rectify gross inequalities, or else to impose them.

Yet from the standpoint of an empiricist, interested mainly in the search for patterns in data, substantial obstacles stand in the way of definite observations. There is first of all the difficulty that reliable measures of change in economic inequality, measures that are both consistent and consistently available, have not existed. Second, there is the problem of arriving at a consistent categorization of types of violence, so that one may predict the effect of each type on economic inequality and vice versa. Third, there is the problem of developing consistent and comparable data across countries and through time on levels and types of violence. [ Full Text ] [IIC*]

We use industrial data to derive estimates of the pattern of change in wage inequality in Mexico and Brazil. Using the group decomposition of Theil's T statistic we present monthly series of measurements of change in the dispersion of industrial wages for Brazil (1976 through 1995) and for Mexico (1968 through 1997). Both countries show increases in wage dispersion over time, and we find a strong negative correlation with the rate of real economic growth, suggesting that real per capita income growth is important in the determination of movements in inequality. Heterodox plans seem to reduce inequality in the short-run. [ Full Text ] [IIC*]

This paper presents a procedure for studying industrial performance and related issues such as changes in the wage structure. This procedure combines cluster analysis and discriminant analysis as a package, and applies this package to time series data. This enables us to organize industrial data into groups with similar wage or performance histories and then to extract summary time-series showing the main pattern of variation in performance between groups. [ Full Text ]

This paper provides a summary of information in the UTIP data set on the evolution of industrial earnings inequality in the global economy. At present the data set covers 66 countries, with annual observations going back to 1972 in most cases and to 1963 in many. Our measure of changing inequality, based on the group-wise decomposition of the Theil statistic across industrial categories, appears to be a sensitive barometer of political and economic conditions in many countries, and the percentage change in this index appears to be meaningfully comparable across countries. We also measure and detect regional patterns of similarity in the movement of inequality through time. [ Full Text ] [IIC*]

In recent work one of us has presented measurements of the evolution of inequality in the U.S. manufacturing sector, from 1920 to 1992. This paper updates and revises those estimates, using a monthly data set for wages and employment of production workers in 18 sectors, for which continuous data are available back to January, 1947. The main findings of the previous study are confirmed: there is a close connection between the dispersion of hourly wage rates and unemployment. But the previous series erred in bridging a gap in the data between 1947 and 1958 by assuming that inequality in manufacturing in that period tracked the movement of a Gini coefficient for household incomes, which was fairly stable during this time. In fact, in the 1950s manufacturing wage rate inequality rose sharply, reaching the extreme levels of the 1930s. An implication is that inequality in manufacturing hourly wage rates in the late 1970s and 1980s, previously thought to be lower than during the Great Depression, was in fact much higher. The new series also shows that wage rate inequality began declining again in 1994, and has now fallen to just below the peaks of the inter-war period. The data are current to the end of 1998. [ Full Text ] [IIC*]

We employ the UTIP data set on the evolution of earnings inequality in manufacturing in the global economy to illuminate two questions. First, do regional patterns of similarity in the movement of large macroeconomic aggregates, such as real GDP, imply underlying similarities of industrial structure, so that knowledge of one national economy in a GDP cluster can reasonably be assumed to convey useful information about the others? We show that this is not generally the case. Particularly, regional co-movement of GDP in Asia, which is very strong, masks deep dissimilarities in underlying employment structures -- and, we argue, a range of potential sources of transmissible financial crisis. Second, what are the consequences of crisis for inequality? We show that crises typically generate increases in inequality, but more so in less developed countries, and more so in regions that are more liberal in their policy regimes. [ Full Text ] [IIC*]

For most of Latin America the 1970s were a decade of growth, though with political upheaval in Argentina and Chile. The 1980s were a disaster. The 1990s have seen economic reform, liberalization, a return to democracy and financial turmoil. This study attempts to explain the three decades as one piece, through an analysis of the evolution of earnings inequality from year to year in eight major Latin American countries and one Caribbean nation. We find that changes in earnings inequality are a sensitive indicator of slump, repression, political turmoil, civil war, natural disaster and -- on the positive side -- occasional periods of growth and stability in Latin America. Indeed almost the whole recent history of Latin America can be summarized in the movement of industrial inequality statistics. [ Full Text] [IIC*]

In this paper we show that inequality and unemployment are related positively across the European continent, within countries, between countries and through time. This contradicts the often-repeated view that unemployment in Europe is attributable to rigid wage structures, high minimum wages and generous social welfare systems. In fact, countries that possess the low inequality such systems produce experience less unemployment than those that do not. Moreover, large inter-country inequalities across Europe aggravate the continental unemployment problem. There is no paradox in low American unemployment. It stems in part from that country’s continent-wide programs of redistribution, including the Social Security System, the Earned Income Tax Credit, the federal minimum wage, and a uniform regime of monetary policy geared toward full employment, all of which reduce inter-regional inequality and all of which we recommend for adoption by the European Union. [ Full Text ] [IIC*]

This paper focuses on two questions. First, how did inequality in the industrial wage structure of Brazil evolve from 1985 to 1995? Second, what is the relationship between these dynamics and economic policy? We display the evolution of wage inequality in Brazil and relate this evolution to changing macroeconomic conditions. Our analysis suggests that there is a strong relation between rising inequality and the restructuring of the Brazilian economy that occurred in the middle 1980’s. [ Full Text]

This paper presents an analysis of the evolution of industrial wages in a selection of OECD countries, using data drawn from the Structural Analysis Database and a sequence of techniques that apply cluster and discriminant analysis to time-series of wage change by industry. The principal finding is that a small number of well defined groups of industries usually exist, whose cross-group differences account for almost all inter-industry wage variation. While the specific structure of groups varies according to patterns of natural resources, comparative advantage and trade union organization within each country, the between-group variation across time usually reflects the movement of macroeconomic variables, some of them internal and other external, such as inflation and exchange rates. In other words, individual countries appear to be able to control their internal institutional structures, perhaps best understood as pattern bargains, wage contours, or industrial sectors distinguished by type and degree of exposure to international trade. But they do not exercise internal control over the evolution of wage differentials across these groups, except insofar as they can manipulate the macro conditions to which the groups are differentially sensitive. [ Full Text ] [IIC*]

This paper discusses the implications of the decomposition property of the Theil index in sequences of nested and hierarchic grouping structures, formalizing general results applicable to a generic sequence of grouping structures. A specific application to data on wages and employment by industrial classification to measure the evolution of wage inequality through time will be explored, analyzing the links between Theil indexes computed at different levels of n-digit SIC codes. A dynamic analysis shows the extent to which a between group Theil statistic tracks the evolution of inequality within industries, and estimations are provided as to the amount of information gained by using ever more disaggregated grouping structures to assess the dynamics of overall inequality. The empirical illustration provides a monthly time-series for industrial earnings inequality in the US is computed at 2, 3 and 4-digit SIC codes from January of 1947 to March of 1999. [ Full Text]

We provide an historical survey of Chinese economic reform, a discussion of the current major problems, and measures of the evolution of inequality in China through time, 1979-1996.We argue that China's most pressing reform needs now are in the social sphere, specifically the creation of an adequate social security system. The paper concludes with maps showing measures of inequality within each province of China for the years 1989 and 1996, and the change over this period. The maps reveal a marked regional pattern of high and sharply rising inequality especially in North and West China, with a milder situation in the South. We suggest that this pattern is consistent with the Kuznets conjecture relating inequality to economic growth; the problem of rising inequality in China is in part the uneven character of growth and development across the country. [ Full Text]

In recent literature the famous Kuznets relationship between inequality and income has been reformulated in terms of levels of inequality and subsequent rates of growth. In this paper we criticize the World Bank data set on which these studies have been based, and present contrasting evidence on pay differentials derived from the 2000 release of the UNIDO Industrial Statistics, a rich source of information on inter-industry pay rates. Our evidence supports the original Kuznets formulation relating levels of inequality to levels of income (or changes in inequality to changes in income). We find that in modern data most countries are to be found on the downward-sloping portion of an inverted Kuznets U-Curve, and we find some support for an "augmented Kuznets curve" in which a few of the very highest-income countries experience rising inequality as their incomes rise. [ Full Text]

This short paper revisits the relationship between wage flexibility and unemployment. The conventional view of a trade-off between equality and employment suggests that the relationship should be negative. Using a panel of data from ten OECD countries across a twenty-two year period, we find a positive relationship between unemployment, when the latter is measured both nationally and OECD-wide -- and wage flexibility, measured directly as the coefficient of variation of inter-industrial wage change. We suggest that the evidence available through the early 1990s never supported the conventional view. [ Full Text]

This paper discusses the dynamic properties of generalized entropy measures of inequality, deriving an explicit expression linking changes in inequality with changes in the growth income. This establishes a time-wise decomposition for the family of generalized entropy measures of inequality, showing that the rate of change in the distribution of income can be divided into two parts. One is purely a function of the levels and rates of change of macroeconomic variables; a second part depends exclusively on the micro nature of the distribution process. We then focus on a subset of generalized entropy inequality indexes, focusing on Theil measures. By assuming that the distribution process follows a specific rule, we further simplify the time-wise decomposition formula, establishing a direct relationship between the rate of income growth and the rate of change in inequality as measured by Theil measures. A specific application of this formula to the simulation of an economy with a constant average positive growth rate produces behavior consistent with the Kuznets hypothesis. [ Full Text]

This paper exploits the decomposability properties of the Theil index to present new evidence on the evolution of earnings inequality in China by sector and province, for the years 1987-2000. The official data sources have rich possibilities for interpreting China’s "Retreat from Equality." However careful attention must be paid to changes in category schemes and other matters that affect the interpretation of the data.[ Full Text]

This paper discusses two issues in the relationship between inequality and economic growth: the data and the econometrics. We first review the inequality data set of Deininger and Squire, which, we argue, fails to provide adequate or accurate longitudinal and cross-country coverage. We then introduce our own measures of the inequality of manufacturing pay, based on the UNIDO Industrial Statistics. In our view, these provide indicators of inequality that are more stable, more reliable, and more comparable across countries than those of Deininger and Squire. Turning to the relationship between inequality and development, we diagnose several common econometric problems in the literature, including measurement error, omitted variable bias, serial correlation in longitudinal data, and the possible persistence of lagged dependent variables. By taking steps to account for these problems, we seek more reliable inferences concerning the relationship between inequality, national income and economic growth. We find evidence that generally supports Kuznets’ specification for industrializing countries: inequality tends to decline as per capita income increases. However, after 1981 two problems emerge. First, per capita GDP growth slows dramatically in most countries. Second, there is a worldwide trend toward rising inequality in our data, independent of GDP or its changes. The timing and geographic pattern of these increases suggest a link to the high real interest rates and global debt crisis of the period beginning in 1982. [ Full Text]

The deficiencies of the Deininger and Squire data set on household income inequality are well known to include sparse coverage, problematic measurements, and the combination of diverse data types into a single data set. Yet many studies have relied on this data due to the lack of available alternatives. In this paper we show how the UTIP-UNIDO measures of manufacturing pay inequality can be used, with other information, to estimate measures of household income inequality. We take advantage of the systematic relationship between the UTIP-UNIDO estimates and those of Deininger and Squire. The residuals from this exercise provide a map to problematic estimates in the Deininger and Squire data, and the estimated coefficients provide a way to construct a new panel data set of estimated household income inequality. This new data set provides comparable and consistent measurements across space and through time that Deininger and Squire's data do not pass. [ Full Text]

The collapse of the Soviet Union and the acceleration of economic reforms in the People’s Republic of China were hallmark events of the 1990s. The Soviet collapse had adverse consequences for many parts of the post-Soviet population -- including sharply rising mortality rates -- even as the country underwent a transition to apparent multiparty democracy. Meanwhile the Chinese experience produced a continuing rise of average living standards, with political change (mainly at the local level) only within the framework of continuing rule by the Chinese Communist Party. Thus the experiences of the two countries are widely viewed as having been polar opposites. Nevertheless, in both Russia and China, economic inequality rose sharply. In both countries, regional inequalities rose more sharply than inequalities across sectors but within regions. In particular, major urban centers gained dramatically, relative to the hinterlands. In both countries, moreover, there was a considerable reorientation of sectoral advantage, in both cases toward those sectors exercising the largest degrees of monopoly power. In both countries, the relative position of finance improved sharply, while that of agriculture declined. However the decline of agriculture in China was not as precipitous in China as in Russia, and certain sectors, such as education and science, maintained their position in China in a way that was not possible for them in Russia. [ Full Text]

A substantial body of literature has uncovered a robust relationship between institutions-including unionization, political democracy and economic inequality. This paper examines the effect of military spending on inequality controlling for the size of armed forces, GDP growth, per capita income and other possible determinants. Using a panel regression with country level observations from 1987-1997, we obtained consistent estimates that there is a positive effect of military expenditure on pay inequality. Given the close relationship between pay and income this result suggests that a country’s reduction in military spending could reduce income inequality. [ Full Text]

This paper reconsiders the problem of unemployment in Europe at multiple geographic levels and through time from 1984 to 2000. We employ a panel structure that permits us to separate regional, national and continental influences on European unemployment. Important local effects include the economic growth rate, relative wealth or poverty, and the proportion of young people in the labor force. As part of this analysis, we assess the relationship between pay inequality and unemployment in Europe, following the insight of Harris and Todaro (1970) that pay inequalities influence job search. With our own panel of inequality measures derived from Eurostat’s REGIO data set, we find that higher pay inequality in Europe is associated with more, not less, unemployment, and the effect is stronger for women and young workers. There are modest country fixed effects for the UK and Spain, but large effects are found only for small countries. These are all negative, a fact that may be due partly to large past emigration in some cases, and partly to strategic wage bargaining in others. Apart from this, distinctive effects at the national level are few, perhaps indicating that national labor market institutions are not the decisive factor in the determination of European unemployment. Changes in the European macro-environment are picked up by time fixed effects, and these show a striking pan-European rise in unemployment immediately following the introduction of the Maastricht Treaty, though with some encouraging recovery late in the decade. [ Full Text]

This note explores the relationship between the penetration level of Information and Communication Technologies and earnings inequality in Chile. The purpose of the note is to check whether income distribution significantly differs among ICT users and non users. I find that in addition to having a higher average income, the group of ICT users presents a broader dispersion of earnings than the group of ICT non users. In addition I present the results of a logistic regression showing that the most important factors facilitating or inhibiting Internet access are income, education, area of residence and gender. [ Full Text]

This paper explores the relationship between the between-groups component of Theil’s T Statistic measured across U.S. counties using Local Area Personal Income Statistics, and the information technology bubble of the 1990s. Our examination yields a predictable result: the technology boom had a major effect on the distribution of income in the United States. The surprising fact is that higher incomes in a mere handful of counties influence aggregate measures so dramatically. [ Full Text]

This paper presents the trend of changes in pay inequality in the manufacturing sector of India, by regions and sectors, for the years 1979-1998. The decomposability property of Theil index enables us to show that manufacturing pay inequality in India has risen both across sectors and across regions, though more strongly across sectors. We also show that the rise in inequality accelerates in the period following the introduction of reforms, after controlling for changes in the level of real per capita income. It appears that a large part of rising manufacturing pay inequality in the post-reform period can be attributed to rising relative pay in the electricity sector. [ Full Text]

(UTIP No.30)
Pay Inequality in Europe 1995-2000: Convergence Between Countries and Stability Inside

The decline in manufacturing employment relative to service sector employment in most OECD countries has reopened debate over the relationship of employment growth to pay inequality. A widely-held view presumes that a trade-off exists; in this view labor markets matching the supply and demand for skill will generate new jobs only at the expense of greater inequalities. This paper examines the actual relationship between employment growth and pay inequality in services at the regional level for 14 European countries from 1995 to 2000. Our evidence does not support the hypothesis of a tradeoff.
In this period, the most rapid employment expansion in services occurred in real estate, in renting and business activities, in wholesale and retail trade, and in repair of motor vehicles, motorcycles and personal and household goods. All of these sectors have average wages well above the bottom of the pay distribution. Almost twice as many jobs were created in these sectors than in all the others, with the result that pay inequalities in services declined as employment grew. Overall we find a striking pattern of declining pay inequality across Europe as employment expanded in the service sector during the period of the introduction of the euro. [Full Text]

This paper examines the role of religion in the 2004 US election. Using data from a recent Pew survey and the University of Texas Inequality Project, the paper shows that inequality counterbalances the oft-remarked tendency for richer societies to become increasingly secular. The paper suggests that globalization, by increasing inequality, has contributed importantly to the recent worldwide resurgence of religion. The analysis also points up flaws in “market” models of religion developed by Barro and McCleary and others.
The paper then develops a model of state-level voting in the 2004 presidential election. Using spatial regression, the paper finds that states with high percentages of evangelicals and Mormons were indeed more likely to cast more votes for Bush and Cheney. But the results also show that worshipers of the Golden Calf (“Are you better off today than you were four years ago?”) were also highly influential in determining the outcome, as was the decay of voting turnout in states between 1968 and 2000. A particularly striking result is that states that witnessed lesser changes in inequality (as measured by Census Bureau Gini Coefficients) were far more likely to vote for Bush in 2004. In sharp contrast, states such as Massachusetts, California, New York, or Connecticut, which topped all others in their increases in income inequality, went almost monolithically for Kerry. The paper concludes with an analysis of the effects of campaign financing, and particularly the “527s”, on state-by-state outcomes.
[Full Text]

In this paper we use a previously neglected, high-quality data source to generate consistent annual measures of income inequality by state, for the fifty United States and the District of Columbia from 1969 to 2004. We use the estimates in a model of presidential election turnout and outcomes at the state level from 1992 to 2004. In recent elections, we find that high state inequality is negatively correlated with turnout and a positively correlated with the Democratic vote share, after controlling for race and other factors.
[Full Text]

In this paper we analyze the distribution of pay and changing trends of inequality in Argentina and Brazil, illuminating the specific winners and losers, by region and by economic activity (sector). In both countries we find that inequality rose in the neoliberal period, but that it declined following the severe crises of neoliberal policy, in 1993 in Brazil and in late 2001 in Argentina. This period of post-neoliberalism is characterized in both countries by a decline in the economic weight of the financial sector and a recovery of the position of the civil service. In both countries, the rise in inequality leading to the crisis produced an increase in the relative position of the major metropolitan centers; this positional advantage also declined modestly in the post-crisis recovery period.
[Full Text]

(UTIP No.35)
Structural Change, Inequality and Growth in Mexico

This paper examines the relationship between income inequality and economic growth in Mexico. We first review changes in industrial trade, production, and investment patterns over the liberalization period and how those changes led to the creation of a relatively high-wage, economic enclave of industries producing capital-goods for export. We then compare annual changes in manufacturing pay inequality and annual GDP growth, finding that the previously stable, negative relationship predicted by Kuznets broke down at the height of the period of structural reform in Mexico, giving way to a positive relationship after 1989. The paper finds that reform fundamentally altered the relationship between inequality and growth as benefits accrued to an increasingly small number of firms. The findings support the hypothesis of an “augmented” Kuznets Curve according to which some developed countries are found on an upward-sloping addendum to Kuznets’ original formulation.
[Full Text]

(UTIP No.36)
Patterns of Wage Inequality in Costa Rica During the Structural Change, 1976-2004

This paper presents new measures of inequality for Costa Rica for each year from 1976 to 2004, using data on payroll and number of salaried workers by sector and province, provided by the insurance records of the country’s social security offices. Overall, after a long period of decreasing inequality from 1976 to 1985, wage inequality in Costa Rica has been more volatile during the last two decades. The behavior of inequality and real wages during the period 1976-1985 reflects the wage policies of the time before the application of the free market model. Reforms in the financial and health sector seem to be among the important factors influencing wage inequality since that time. Unionization in activities controlled by the state, and electoral cycles are also apparently important. Finally, the successful attraction of high technology firms to Costa Rica has been a key factor accounting for increasing average wages in manufacturing industries in the last decade.
[Full Text]

(UTIP No.37)
Economic Equality and Victory in War: An Empirical Investigation

This paper tests a simple hypothesis: that given the occurrence of war between two countries, the country that is more egalitarian at the moment of military decision is likely to emerge the victor. First, we examine cases where comparative economic inequality can be measured directly, using the nearly comprehensive global data-sets of the University of Texas Inequality Project for the years 1963-1999. Second, we examine cases where reasonable inferences about comparative economic inequality may be drawn by analogy to UTIP measurements or from other political and economic evidence, including both bi-national wars and larger wars where there existed clear pair-wise fronts. Third, we discuss selected cases where inferences may be drawn from literary or historical sources. We find, all in all, that the evidence for an egalitarian victory proposition is remarkably strong. [Full Text]

(UTIP No.38)
Pay Inequality in Cuba: the Special Period and After

This paper analyzes the evolution of pay inequality in Cuba from the early 1990s through 2004, during what was known as the “Special Period in Times of Peace” and after. We measure pay inequality across sectors and regions, using the between-groups component of Theil’s T statistic, and we map the changing components of that statistic in order to provide a compact summary of structural change in Cuba. This method helps us to observe the transition of the Cuban economy from one based fundamentally on sugar to one based largely on services, especially tourism, but also others with greater growth potential, such as information technology, pharmaceuticals, and biotechnology. Regionally, we observe that a main dividing line between winners and losers is the presence of tourist attractions: the recent increase of regional pay inequality is associated primarily with changing incomes in the city of Havana and the province of Matanzas. [Full Text][Espanol]

(UTIP No.39)
Maastricht 2042 and the Fate of Europe: Toward Convergence and Full Employment

This paper presents a strategy of economic convergence for Europe . European principles and ideals require convergence, but the pan-European economic policy of “labor market reform” imposes divergence, in the hope that greater inequality in European pay will bring Europe closer to the dynamism and employment performance of the United States . We resolve this European paradox by showing that in fact the (inter-regional) pay structure of the United States is substantially more egalitarian than Europe ; convergence toward American inequality levels will therefore require the systematic reduction of inter-regional pay differentials across Europe . We present quantitative targets for a strategy of egalitarian growth and pay convergence across the regions of Europe through 2042, the fiftieth anniversary of the Maastricht treaty. A theoretical section explains why such a strategy, following the experience of the American New Deal, should work to reduce the scourge of European unemployment. [Full Text]

IIC*. This paper has now been published as part of "Inequality and Industrial Change: A Global View," edited by James K. Galbraith and Maureen Berner and available now from Cambridge University Press. In order to protect the publisher's best interest, it is accordingly that certain features like printing, text selecting, etc. are disabled.