Thesis (Ph. D.)--University of California, Berkeley, 1999. / Includes bibliographical references (leaves 111-117).

The decline in private sector wages in the aftermath of the Great Recession reopened a longstanding debate about whether public sector employees make more than private sector employees. However, much of this debate has only focused on the difference in wages over the past few years. This paper uses the Current Population Survey from 1995-2013 to examine how the federal-private wage differential has evolved over time. Wage regressions are estimated by year for federal and private sector workers. I then use these estimates to calculate the federal-private wage differential. This is augmented with selectivity bias corrections for each year. Probit estimates of the probability of receiving employer-provided health insurance and a pension plan are also estimated for each year. The findings suggest that the federal pay differential is invariably positive, but fell during the 1990s, began to rise in the early 2000s, and has continued to rise to the end of the sample period. In this paper, I also examine the difference in wage and fringe benefit between state/local government employees and private sector employees. For the analysis, this paper uses the American Community Survey from 2012-2014 to examine how the state/local-private wage gaps vary by state. Probit estimates of the probability of receiving employer-sponsored health insurance are also estimated. The findings present a wide range of the wage differentials between state/local government employees and private sector counterparts. On the other hand, public employees enjoy higher probability of receiving health insurance through a current employer.

O objetivo desta dissertação é analisar a diferença salarial existente entre indivíduos que trabalham como professores do ensino fundamental e indivíduos que trabalham em outras ocupações. Busca-se compreender se tal diferença está na própria profissão docente ou na formação desses profissionais. Pretende-se identificar se é mais vantajoso trabalhar na ocupação docente ou em outras ocupações, dependendo da formação do indivíduo. Para tanto, calculam-se diferenciais de salário controlados e não-controlados. Os principais resultados para os diferenciais não-controlados revelaram que trabalhar como docente do ensino fundamental é mais vantajoso para indivíduos que ainda não possuem ensino superior. Para aqueles que são docentes e possuem curso superior ou estão em curso (graduação ou pós-graduação), independentemente dos controles, a ocupação docente não é vantajosa. Já no caso dos diferenciais controlados, ser professor do ensino fundamental quase sempre é vantajoso quando se considera o salário por hora. / The objective of this dissertation is to identify the wage differential between primary school teachers and individuals working in other occupations. We try to comprehend if the differential is on the teacher profession itself or on the career chosen by these individuals. In other words, we analyze if the differential is on the education of these professionals. We intend to identify through controlled and non-controlled wage differentials if (depending on the education of the individual) it is more advantageous working as a teacher or in another occupation. The main results for non-controlled differentials showed that working as a primary school teacher is more advantageous for individuals who did not complete the undergraduate level. For teachers with higher education completed or ongoing (undergraduation or graduation), independently of the controls, the teacher profession is not advantageous. For the controlled differentials, being a high school teacher is almost always advantageous when the hourly wage is considered.

Due to there is no record of migration in China, we estimate migration by Johnson¡¦s (2003) way. Estimates of migration among the provinces of China area were made by comparison of the provincial population from the 1994 to the 2003 censuses. The estimates were made by comparing every year between 1995 and 2004 population of each province with what it would have been if population had increased solely due to national growth-the excess of births over deaths. Unfortunately, the estimate of the increase in provincial populations due to migration was much greater than the estimate of the loss of population by migration. Possible reasons for the difference in the estimates due primarily to the underestimation of immigrants. Because this difference amounts to less than 3% of the 1995-2004 population of China, the data is still useful to us. Because wage is a kind of price, we adopt the method by Engel and Rogers (1996), and use the law of one price to examine the wage differential in China area. We find that high migration gap, long distance and the presence of coast all lead to an increase in wage differential in China area. We also find that distance and border reduce differential among provinces. Note that the distance has a positive effect on wage differential, and the square of distance has a negative effect. This means the distance relationship were concave. Finally, we examine the tendency in the standard deviation of wage differential and migration gap among the provinces of China during the period 1994-2003. The standard deviation of wage differential widened, and the standard of migration gap did not display strong increasing or decreasing. Although the growth rate of wage differential was positive, it reduced in recent years. The average growth rate of migration gap was close to zero, there is not significant tendency.

As the female workforce is increasing by year in the Taiwan labor market, it¡¦s visible that wage gap between male and female workers exists in the Taiwan labor market. No doubt that it is irony for people who strive for the gender equality in Taiwan. An empirical analysis of gender wage differential is presented by gaining labor force survey of 1981,1986,1991 and 1997, which has been conducted every year by Directorate-General of Budget, Accounting and Statistics Based on the concept and theory of human capital brought up mainly by Gary S. Becker (1975), several multiple regression functions are conducted in this research on purpose to understand how large the wage gap is, where the gap is stemmed from and what cause brings about this gap. The results of this research indicate that: 1. Femal wage is lower than male wage about 60%~70%. The gap is getting larger from 1981 to 1991; however, smaller after 1991. 2. Male workers get more wages by every year than female workers. 3. Except 1981, which female workers gain more wage by investing human capital by every year than male workers suggests that female workers hold larger invest-rate of education than male and invest-rate of education increases by time. 4. Female workers with more than high-school education degree have less wage differential compared with male workers than those who are with less than high-school education degree. This indicates that female workers investing more human capital have less wage gap compared with male workers in the labor market. 5. Wage gap stemmed from gender discrimination is ascending from 45% of 19981 to 69% of 1996; however, wage differential led from the characteristics of individual worker is descending by time. From the results of above, there is surely gender wage gap in Taiwan labor market, which is the outcome of interacting from diverse individual characteristic and subjective gender prejudice. It¡¦s very obvious that education enhances competitive advantage of female workers so as to decrease the wage differential, nevertheless, discrimination toward to female workers in the Taiwan labor market heavily wears down the endeavor of investing in human capital for them. How to motivate female workers to become more competitive and improve work place for female workers so as to make the best use of female human resources is the significant issue of Taiwan government in the near future.

O objetivo desta dissertação é analisar a diferença salarial existente entre indivíduos que trabalham como professores do ensino fundamental e indivíduos que trabalham em outras ocupações. Busca-se compreender se tal diferença está na própria profissão docente ou na formação desses profissionais. Pretende-se identificar se é mais vantajoso trabalhar na ocupação docente ou em outras ocupações, dependendo da formação do indivíduo. Para tanto, calculam-se diferenciais de salário controlados e não-controlados. Os principais resultados para os diferenciais não-controlados revelaram que trabalhar como docente do ensino fundamental é mais vantajoso para indivíduos que ainda não possuem ensino superior. Para aqueles que são docentes e possuem curso superior ou estão em curso (graduação ou pós-graduação), independentemente dos controles, a ocupação docente não é vantajosa. Já no caso dos diferenciais controlados, ser professor do ensino fundamental quase sempre é vantajoso quando se considera o salário por hora. / The objective of this dissertation is to identify the wage differential between primary school teachers and individuals working in other occupations. We try to comprehend if the differential is on the teacher profession itself or on the career chosen by these individuals. In other words, we analyze if the differential is on the education of these professionals. We intend to identify through controlled and non-controlled wage differentials if (depending on the education of the individual) it is more advantageous working as a teacher or in another occupation. The main results for non-controlled differentials showed that working as a primary school teacher is more advantageous for individuals who did not complete the undergraduate level. For teachers with higher education completed or ongoing (undergraduation or graduation), independently of the controls, the teacher profession is not advantageous. For the controlled differentials, being a high school teacher is almost always advantageous when the hourly wage is considered.

This thesis looks at the effect of industrial and occupational segregation on the wage differential between resident and migrant workers in China. It extends the work of Meng and Zhang (2001) by considering the possible employment segregation of resident and migrant workers by both industry and occupation. I contend that industry segregation is at least as important as occupational segregation for Chinese migrant workers, as most migrant workers in China have come from the countryside to fuel the booming labor-intensive manufacturing and construction industries in the cites. Due to the hukou policy (a household registration system) in China, migrant workers normally face more constraints in searching for jobs in other sectors. My empirical study confirms that the proportion of the resident-migrant worker wage differential that is explained by industrial segregation is much larger than that explained by occupational segregation. Taking both industrial and occupational segregation into account explains the substantial wage differential between resident and migrant workers, which indicates the influence of industrial and occupational barriers on the wage differential in China.

Η αύξηση των ποσοστών της παχυσαρκίας, σε παγκόσμιο επίπεδο, ακολουθεί αυξανόμενους ρυθμούς κατά τη διάρκεια των τελευταίων δεκαετιών, συγκεντρώνοντας το ενδιαφέρον ενός σημαντικού αριθμού επιστημόνων στον τομέα των οικονομικών της υγείας. Η παρούσα διατριβή έχει σκοπό να καλύψει τέσσερα ερευνητικά ερωτήματα που σχετίζονται με τα οικονομικά της παχυσαρκίας. Συγκεκριμένα, στα πλαίσια του πρώτου δοκιμίου, η παχυσαρκία μοντελοποιείται ως μια στοχαστική διαδικασία, διερευνάται ο δυναμικός της χαρακτήρας καθώς και οι πηγές της παρατηρούμενης εμμονής. Για αυτό το σκοπό εκτιμώνται κατάλληλα δυναμικά υποδείγματα διακριτών επιλογών, χρησιμοποιώντας διαστρωματικά στοιχεία χρονολογικών σειρών για την περίπτωση ενηλίκων στις ΗΠΑ (NLSY79). Στα πλαίσια του ίδιου κεφαλαίου, επιχειρείται επίσης και μια μελέτη της εμμονής στο σύνολο των κατηγοριών του σωματικού βάρους (φυσιολογικού βάρους, υπέρβαροι, παχύσαρκοι) εφαρμόζοντας δυναμικά διατεταγμένα υποδείγματα. Τέλος, παρουσιάζονται οι απορρέουσες προτάσεις πολιτικής. Στη συνέχεια, χρησιμοποιώντας διαχρονικά στοιχεία από την ίδια βάση δεδομένων (NLSY79), διερευνάται η εμμένουσα συμπεριφορά της διάρκειας της παχυσαρκίας καθώς και οι προσδιοριστικοί παράγοντες των πιθανοτήτων εξόδου από την παχυσαρκία και επαν-εισόδου στην παχυσαρκία. Εφαρμόζοντας μη-παραμετρικά μοντέλα καθώς και πολυ-μεταβλητά διακριτά υποδείγματα κινδύνου, που λαμβάνουν υπόψη τους τη μη-παρατηρούμενη χρονικά αμετάβλητη ετερογένεια των ατόμων, καταδεικνύεται ότι η πιθανότητα εξόδου από την κατάσταση της παχυσαρκίας χαρακτηρίζεται από μια αρνητική εξάρτηση από τη διάρκεια (negative duration dependence). Όσον αφορά στις επιπτώσεις πολιτικής, η ύπαρξη «πραγματικής» εξάρτησης από τη διάρκεια υποδηλώνει ότι οι παρεμβάσεις των πολιτικών δημόσιας υγείας πρέπει να στοχεύουν πρωτίστως στους «πρόσφατα παχύσαρκους». Το τρίτο εμπειρικό δοκίμιο πραγματεύεται τις επιπτώσεις της παχυσαρκίας στην αγορά εργασίας των ΗΠΑ και ειδικότερα στην περίπτωση των μισθών. Εκμεταλλευόμενοι τα ευρήματα των προηγούμενων αναλύσεων σχετικά με την εμμένουσα συμπεριφορά της παχυσαρκίας, μπορεί να θεωρηθεί ότι η παχυσαρκία αποτελεί μια σχετικά σταθερή κατάσταση που χαρακτηρίζεται από μειωμένη κινητικότητα. Ως εκ τούτου, γίνεται εφικτή η ανάλυση των μισθολογικών χασμάτων μεταξύ παχύσαρκων και μη-παχύσαρκων εργαζομένων, εφαρμόζοντας κατάλληλα υποδείγματα διάσπασης μισθών κατά Oaxaca-Blinder. Τα σχετικά ευρήματα καταδεικνύουν την ύπαρξη μισθολογικών χασμάτων εις βάρος των παχύσαρκων λευκών ανδρών και γυναικών. Τα εν λόγω μισθολογικά χάσματα μπορούν να ερμηνευθούν -σε ικανοποιητικό βαθμό- από τις διαφορές στα παρατηρούμενα χαρακτηριστικά μεταξύ παχύσαρκων και μη παχύσαρκων εργαζομένων και, ειδικότερα, από τις διαφορές στο επίπεδο εκπαίδευσης. Τέλος, το τελευταίο δοκίμιο της παρούσας διατριβής διερευνά τους προσδιοριστικούς παράγοντες των αποτελεσμάτων σωματικού βάρους στην περίπτωση της Ελλάδας χρησιμοποιώντας -για πρώτη φορά- πρόσφατα δεδομένα από την Εθνική Έρευνα Υγείας (GNHS-2009). Για τις ανάγκες της ανάλυσης εφαρμόστηκε μια σειρά από οικονομετρικές τεχνικές εκτίμησης διαφορετικής φύσης (γραμμικά υποδείγματα ελαχίστων τετραγώνων, διατεταγμένα μοντέλα και μη-δεσμευμένες παλινδρομήσεις ποσοστιαίων σημείων). Τα σχετικά αποτελέσματα δείχνουν ότι οι εκτιμήσεις μη-δεσμευμένων παλινδρομήσεων ποσοστιαίων σημείων οδηγούν σε σημαντικές ποσοτικές και ποιοτικές διαφοροποιήσεις στις επιπτώσεις των διαφόρων προσδιοριστικών παραγόντων του σωματικού βάρους, σε σχέση τόσο με τα διατεταγμένα μοντέλα όσο και με τις γραμμικές παλινδρομήσεις ελαχίστων τετραγώνων. Επομένως, οι υπεύθυνοι χάραξης πολιτικής για την αντιμετώπιση του αυξημένου σωματικού βάρους στην Ελλάδα θα πρέπει να βασίζονται σε εμπειρικά μοντέλα και τεχνικές εκτίμησης που λαμβάνουν υπόψη τους το σύνολο της κατανομής του ΔΜΣ και όχι συγκεκριμένα τμήματά της. / Recently economists pay a lot of attention on the analysis of obesity since the prevalence of obesity is widespread. This dissertation examines four research questions associated with the Economics of Obesity. Specifically, with respect to the first empirical chapter, obesity is modeled as a dynamic stochastic process and patterns of obesity experiences are investigated. Longitudinal data from US adults (NLSY79) and appropriate dynamic models were utilized. Strong persistence was detected in obesity and the body-mass-index (BMI), albeit with different properties. While controlling for initial conditions (early-life endowments and family background), the identified obesity persistence was decomposed into unobserved heterogeneity and genuine state dependence. Results from the sensitivity analysis show that obesity has a lasting effect, which however is not stable across the BMI distribution, exhibiting a steeper decline at BMI levels close to the clinical definition of obesity. Moreover, in the same chapter, we also explore the dynamic behavior of alternative fatness states (normal weight, overweight and obese) and the corresponding persistence, utilizing appropriated dynamic ordered models. Relevant policy implications are discussed. The second chapter investigates the dynamic patterns the duration of obesity and identifies the determinants of obesity-spell exits and re-entries. We utilize longitudinal data from the NLSY79, as before. Non-parametric techniques are applied to investigate the relationship between exit from obesity and spell duration. Multivariate discrete hazard models are also estimated, taking into account duration dependence and observed and time-invariant unobserved heterogeneity. In all cases, the probability of exiting obesity is inversely related to the duration of the obesity spell. Without controlling for unobserved heterogeneity, the probability of exit after one wave in obesity is 31.5 per cent; it is reduced to 3.8 per cent after seven or more waves. When time-invariant unobserved heterogeneity is taken into account, the estimated probabilities are slightly larger and broadly similar (36.8 and 10.3, respectively), which suggests that the identified negative duration dependence is not primarily due to composition effects. The obtained results indicate that public health interventions targeting the newly obese may be particularly effective at reducing incidence of long durations of obesity. In the third empirical chapter we investigate obese/non-obese wage differentials using data for white individuals from the 2000 wave of the NLSY79. Based on the results from the previous chapters, we assume that obesity is a rather permanent characteristic that associated with relatively small mobility. Hence, typical Oaxaca-Blinder wage decompositions could be applied in order to identify the proportion of the observed gender-specific wage differential between obese and non-obese. Based on numerous specifications and alternative sub-samples the results provide strong evidence for the existence of wage differentials in favor of non-obese individuals, which can be mostly explained by differences in early human capital investments and especially schooling investments. Finally, an attempt was made to identify the determinants of body-weight in Greece, utilizing individual-level data from the National Health Survey of 2009. BMI is treated as both, a cardinal and an ordinal measure of body-weight, while different estimation techniques are applied (OLS, ordered probit and unconditional quantile regressions). We employ a wide range of demographic, socio-economic, lifestyle, health-related and regional characteristics. The unconditional quantile regression estimates indicate that the impact of several correlates across the BMI distribution is distinctive. This differentiation concerns the effects of age, education, family income, physical activity, employment status, smoking, health-related impairments and regional characteristics. Thus, examining the entire BMI distribution and targeting specific segments of the Greek population can improve the efficiency of public health policies against obesity.

In this thesis, I study the impact of minimum wage policy and city agglomeration on wages and employment in local labour markets. This is an important topic because having a better understanding of the determinants of regional wage differentials and employment offers insights into: the roles played by local production, consumption and city structures; the standard of living enjoyed by workers with different human capitals; and policy recommendations for the future minimum wage law and city planning regulations. I use local occupation and geographic information to assess how highly productive occupations and local consumption amenities sort workers and generate local wage differentials. I also use this information to construct instruments that enable the accurate estimation of the effects of policy interventions. After an introduction in chapter 1, chapter 2, The Impact of the Minimum Wage on the Wage Distribution: Evidence from China, provides an empirical estimation of the effects of minimum wages using a Chinese household survey. I introduce new instrumental variables, relating to transport costs and local productivity, to control for the potential median wage endogeneity. The instrument variable regressions indicate that the effective minimum wage, defined as the ratio between the minimum wage and the median wage, significantly reduces the lower tail wage inequality — measured by the wage differential between the 50th and the 10th percentiles— by up to 0.3 per cent. In chapter 3, The MinimumWage and Its Impact onWage and Employment, joint work with Coen Teulings, we propose a novel framework for estimating the effects of minimum wages by considering the neoclassical wage and labour participation equations at the same time. To estimate the non-linear censored model with correlated error terms, we provide a five-step procedure and use maximum likelihood estimation. After correcting the bias using occupation information and city size, we find that effective minimum wage correlate significantly with the proportion of workers earning below minimum wage. I study the structure of city and commuting in chapter 4, Consumer City and the Sharing Economy. Based on the international trade literature, I develop a theoretical model with multiple cities, which have different amenities and productivities. In equilibrium, the unobservable parameters are estimated using local employment, wage, and commuting information. Cities show strong agglomeration effects in both productivity and consumption amenities. A counterfactual technological improvement, providing a cheaper transportation for workers and consumers, leads to a more concentrated employment distribution, commuting pattern and higher utility. In the final chapter, Agglomeration and Sorting, joint work with Coen Teulings, we show that agglomeration externalities are strongly related to the occupational structure. At the same time, regional differences in house prices offset these externalities. We develop a multi-region model with regional heterogeneity in workers and jobs, tradable versus non-tradable commodities, consumption amenities, regional house prices, non-homothetic utility, and interregional labour mobility. The model fits the regional data on the fixed wage effects, the return and mean level of human capital, land prices, and the city-rural area distinction well. We use land values to calculate the value of agglomeration.

Made available in DSpace on 2015-05-08T14:44:49Z (GMT). No. of bitstreams: 1 Arquivototal.pdf: 1836085 bytes, checksum: eae0ae0278a6c2c7af0f833a8ec29fac (MD5) Previous issue date: 2012-06-26 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The main objective of this study is to analyze the effect of experience of interstate migration in Brazil on the productive reinsertion and on the income of the employee that returned to the region of origin. The data used in this study were derived from the National Household Sample Survey (PNAD) of 1999 and 2009. This study is divided into three sections. The first is reserved for the investigation of the migratory pattern inside the Brazilian territory and for the comparison of the socioeconomic attributes of return migrants with those observed for non-migrants and non-returned migrants. In the 2000s, the evidences show that most migrants of return comes back to the Northeast and, also, that Southeast has sent more people who received, indicating a loss of attractiveness. The again migrated worker can be characterized as an individual male, nonwhite race, urban person and head of family, being, on average, younger, more educated and better paid than non-migrant. On the other hand, the migrated worker again is older, less educated and receives lower income than non-returned migrant. In the second part are analyzed the determination of the condition of migration and salary differentials between categories: interstate migrants of return, interstate non-returned migrant and non-migrant. The methodology consisted in estimating a structural model of joint determination of migration, remigration and wages, with the application of the method of Lee (1978) for the correction of possible sample selection bias. Moreover, wage differentials were calculated using the technique of Oaxaca-Blinder-Oaxaca and Ramson. The empirical findings show that male workers, more educated, employed as self-employed and entrepreneurs were more inclined to interstate migration and remigration to the state of birth. And, in relation to profits, show that male workers, more educated, employed as civil servants and entrepreneurs, profited higher values. But, the non-returned migrants and the migrants of return received higher incomes than non-migrants. These differences are mainly explained by unobserved characteristics, indicating a positive selection of migrants on attributes not observed in relation to non-migrants. Finally, the objective of the third part was to investigate the impact of the accumulation of savings and/or wealth in the decision to occupation of interstate migrants of return. The methodology used in that investigation was to estimate a probit model with endogenous regressor, correcting the bias this endogeneity in wealth accumulation. Empirical evidences enabled to identify that accumulation of wealth had a positive impact on the choice to employ themselves as self-employed and entrepreneur. Furthermore, the return migrants more favorable to self-employment or entrepreneur are men, white, aged older, with low schooling and that living in the North, Northeast and Midwest regions. In relation to the accumulation of wealth, were also verified that workers who returned before 2000 have accumulated more wealth than workers of recent decade and the accumulation of wealth was generated during the period of youth of the workers. / O principal objetivo desse estudo é analisar o efeito da experiência de migração interestadual no Brasil sobre a reinserção produtiva e os rendimentos do trabalhador retornado à região de origem. Os dados utilizados neste estudo foram provenientes da Pesquisa Nacional por Amostra de Domicílios (PNADs) de 1999 e de 2009. Esse estudo está dividido em três ensaios. O primeiro é reservado à investigação do padrão migratório dentro do território brasileiro e à comparação dos atributos socioeconômicos dos migrantes de retorno com aqueles observados para os não migrantes e migrantes não retornados. Na década de 2000, as evidências mostram que a maioria dos migrantes de retorno voltou para a região Nordeste e também que a região Sudeste emitiu mais pessoas que recebeu, indicando perda de capacidade de atração. O trabalhador remigrado pode ser caracterizado como um indivíduo do sexo masculino, de cor não branca, residente em área urbana e chefe de família; sendo, em média, mais jovem, mais escolarizado e melhor remunerado que o não migrante. Por outro lado, o remigrado é mais velho, menos escolarizado e aufere menor rendimento que o migrante não retornado. No segundo ensaio se analisam a determinação da condição de migração e os diferenciais de salários entre as categorias: migrante interestadual de retorno, migrante interestadual não retornado e não migrante. A metodologia utilizada consistiu na estimativa de um modelo estrutural de determinação conjunta de migração, remigração e salários, com a aplicação do método de Lee (1978) para a correção do possível viés de seleção amostral. Ademais, os diferenciais de salários foram calculados a partir da técnica de Oaxaca-Ramson e Oaxaca-Blinder. Os achados empíricos apontam que trabalhadores homens, mais escolarizados, ocupados como autônomos e empreendedores, foram mais propensos à migração interestadual e à remigração ao estado de nascimento. Ademais, em relação aos rendimentos, mostram que os trabalhadores homens, mais escolarizados, ocupados como funcionários públicos e empreendedores, auferiram maiores valores. Porém, os migrantes não retornados e os migrantes de retorno receberam maiores rendimentos do que os não migrantes, sendo estas diferenças explicadas principalmente pelas características não observadas, indicando uma seleção positiva dos migrantes em atributos não observados em relação aos não migrantes. Por último, o objetivo do terceiro ensaio foi investigar o impacto da acumulação de poupança e/ou riqueza na decisão de ocupação do migrante interestadual de retorno. A metodologia usada na referida averiguação foi a estimação de um modelo probit com regressor endógeno, corrigindo o viés de endogeneidade presente na acumulação de riqueza. As evidências empíricas permitiram observar que a acumulação de riqueza impactou positivamente na escolha por empregar-se como autônomo e empreendedor. Ademais, também foi constatado que os trabalhadores migrantes mais prováveis de se empregar como autônomo ou empreendedor são os homens, de cor branca, mais velhos, com baixa escolaridade e que residem nas regiões do Norte, Nordeste e Centro-Oeste. Em relação à acumulação de riqueza, também foram constatados que os trabalhadores que retornaram antes do ano 2000 foram os que acumularam mais riqueza do que os trabalhadores da década recente e os acúmulos de riqueza foram gerados no período de juventude dos trabalhadores.