The main purposes of this study are: To estimate the wage regression in Vietnam, To examine the existence of gender and urban/rural wage gap, and to decompose these wage gaps to clarify whether there are wage discrimination in Vietnam throughout the wage distribution.
MINISTRY OF EDUCATION AND TRAINING UNIVERSITY OF ECONOMICS OF HO CHI MINH CITY TRẦN THỊ TUẤN ANH QUANTILE REGRESSION DECOMPOSITION OF THE WAGE GAP IN VIETNAM DISSERTATION SUMMARY HO CHI MINH CITY, 2015 MINISTRY OF EDUCATION AND TRAINING UNIVERSITY OF ECONOMICS OF HO CHI MINH CITY TRẦN THỊ TUẤN ANH QUANTILE REGRESSION DECOMPOSITION OF THE WAGE GAP IN VIETNAM Major : Probability and Statistics Major code : 62.46.01.06 DISSERTATION SUMMARY SUPERVISORS: ASSOC PROF PH.D LÊ VĂN PHI PH.D BÙI PHÚC TRUNG ii The research is completed at University of Economics Ho Chi Minh City: Supervisors: Assoc Prof Ph.D Lê Văn Phi Ph.D Bùi Phúc Trung Examiner 1: Examiner 2: Examiner 3: The dissertation will be defended at dissertation councils, meeting at: University Of Economics Hồ Chí Minh City at You can find more information about the dissertation at: National Library or the Library of the University of Economics Ho Chi Minh City iii INTRODUCTION The necessary of the topic Wage is one of the most important factors in motivating employees Because wage depends on a variety of determinants, the existence of the wage gap is inevitable According to economic theory, the wage gap can be decomposed into two main components The first component is due to the difference in endowments of the workers The second one is due to the difference in the coefficients or due to market returns to the endowments The second component is statistical evidence of discrimination that can lead to inequality in society Therefore, the main purposes of this study are (1) to estimate the wage regression in Vietnam, (2) to examine the existence of gender and urban/rural wage gap, and (3) to decompose these wage gaps to clarify whether there are wage discrimination in Vietnam throughout the wage distribution These are the reasons that lead to this doctoral dissertation which is titled as “Quantile regression decomposition of the wage gap in Vietnam” Research objectives This dissertation aim to fulfill the following objectives: 1) Briefly summarizing the background of quantile regression and decomposition method based on quantile regression to analyze the wage gap 2) Applying advanced quantile regression which account for sample selection bias and the endogeneity of explanatory variables to estimate wage equations for men/women and urban/rural groups in Vietnam across their wage distribution 3) Determining the gender wage gap in Vietnam and decomposing this gap into the explained and unexplained components during the period from 2002- 2012 4) Determining the urban/rural wage gap in Vietnam and decomposing this gap into the explained and unexplained components during the period from 2002-2012 5) Examine the change of wage distribution over the years by comparing quantiles of wage in 2002 with that in 2012 This difference in wage is also decomposition into two components: the one that caused by the change in labor force’s characteristics and the other due to the change in the return of these characteristics The theoretical and empirical contributions Along with these research objectives this dissertation have some following theoretical and empirical contributions: (a) This dissertation briefly introduces the definition and features of quantile regression method which was first suggested by Koenker & Bassett (1978) and has been used widely around the world but still not popular in Vietnam There is a few of studies in Vietnam applied quantile regression techniques, especially in the area of wage inequality None of them cover fully features of quantile regression (b) Using the advanced quantile regression, this study estimates the wage equations in Vietnam which help examine the determinants of real hourly wage in domestic labor market The quantile regression techniques applied in this studies was adjusted to account for the problem of sample selection bias and endogeneity that leads to unbiased and consistent estimators (c) This study constructs the wage equations across the quantiles for each following groups: men, women, urban, and rural These results are compared in pairs to clarify the difference in their wage structures (d) This study confirms the existence and estimates the magnitude of gender wage differentials in Vietnam (for the entire sample and for each male/female and urban/rural group of workers) In addition, this study also shows the declined trends of gender wage gap over time in Vietnam (e) After showing the existence of gender wage differential, this research use Machado – Mata method to decompose this gap into two components: the first component represents difference in average characteristics between men and women; the second component represents differences in returns to these characteristics which may be interpreted as possible gender discrimination (f) This study demonstrates the urban - rural wage differential and the change of this gap over time by comparing the estimation in the year 2002 with that in the year 2012 (g) This dissertation decomposes the urban/rural wage differential in order to determine the proportion of this disparity which caused by the difference in endowments between urban and rural workers and the proportion of this gap caused by the difference in the market returns to the endowments (h) After all, this research illustrates in details the change in wage equation over time and shows the decreasing trends in these wage gap over time CHAPTER THE BACKGROUND OF QUANTILE REGRESSION AND MACHADO – MATA DECOMPOSITION 1.1 Mincerian wage model and some extensions The Mincerian wage equation may be written as ln wt s z z , where w: real hourly wage; s: years of schoolings, and z: worker’s years of experience Card (1994) extended the standard Mincer’s wage equation as ln wt s z z X u, (1.1) where X represents for control variables such as gender, race, region, marriage status, and so on After Card D (1994), many studies also augmented the Mincerian wage model by including various explanatory variables into the equation to examine the determinants of compensation and to conduct the ceteris paribus analysis of partial effects on wage 1.2 Quantile regression Quantile regression which was introduced by Koenker & Bassett in 1978 is a method for describing the causality relationship variables at different points in the conditional distribution of the dependent variable Considering the linear regression model Yi X i ui , the quantile regression estimator for each quantile τ ϵ (0, 1) minimizes the objective 1 n V ( ) (Yi X i n i 1 n ˆ In the other word, arg (Yi X i ) n i 1 R k function The quantile regression for quantile τ is written as Q (Yi | X i ) X iˆ (1.14) 1.3 Sample selection bias correction The problem of sample selection bias correction for linear regression with the pioneering work of Heckman (1979) has been extensively studied in econometrics and in labor economics Buchinsky (1998a and 2001) was the first to consider the difficult problem of estimating quantile regression in the presence of sample selection and to propose the correction for this bias in the quantile regression 1.4 Endogeneity and the method of two - stage quantile regression (2SQR) Chevapatrakul et al (2009) suggested the method named 2SQR (two-stage quantile regression) in order to account for the problem of endogeneity in the quantile regression 1.5 The decomposition method based on quantile regression A decomposition analysis is a standard approach to examine the wage differential between male and female workers According to Oaxaca - Blinder (1973)’s approach, the mean wage differential is decomposed into one component capturing differences in characteristics and another component referring to different returns using the estimates of male and female wage equations) Analogous to the linear regression case, Machado and Mata (2005) proposed a similar decomposition which combines a quantile regression and a bootstrap approach in order to estimate counterfactual density functions CHAPTER LITERATURE REVIEW 2.1 Previous studies around the world Some representative studies in investigating the determinants of wage and the wage gap decomposition before the appearance of quantile regression are Edgewort (1922); Becker (1957); Dunlop (1957); Slichter (1950); Cullen (1956); Dalton & Ford (1977); Long & Link (1983); Dickens & Katz (1987); Krueger & Summers (1988); Groshen (1991); Ferber & Green (1982); Lindley, Fish & Jackson (1992); Blackaby et al (2005) Buchinsky (1994) initiated the application of quantile regression in estimating wage regression This led to a trend of using quantile regressions in order to decompose the gender wage gap at different points of the wage distribution It can be listed some noticeable studies as Fortin and Lemieux (1998); Ajwad et al (2002); Albrecht et al (2003); Machado & Mata (2005); Melly (2006); Gunawardena (2006); Arulampalam et al (2007); Nestic (2010); Del Río, Gradín & Canto (2011) 2.2 Previous studies in Vietnam Very few studies in Vietnam applied quantile regression to investigate wage differentials as well as decompose these wage differentials into explained and unexplained parts The typical studies can be listed are Hung et al (2007a) and Hung Ho et al (2007b) However, these studies which used the VHLSS 2002 did not account for the problem of endogeneity CHAPTER DATA AND METHODOLOGY 3.1 Data This study uses the VHLSS 2002 and 2012 to estimate the wage equation in Vietnam labor market and conduct an empirical investigation of wage differentials between the male and female workers as well as the urban and rural areas In order to dispose of the wage change due to inflation, the data was deflate to obtain the comparable real wages Extraction of Table B.2: Wage equations for men and women on 2012 Men’s wage equation in 2012 Independent variables Primary Secondary High school Vocational Colleges Postgraduate Women’s wage equation in 2012 2SQR 2SLS 0.0788*** [2.690] 0.121*** [4.013] 25% 0.0780** [1.963] 0.132*** [3.238] 50% 0.0797*** [2.672] 0.107*** [3.488] 75% 0.0116 [0.338] 0.0475 [1.349] 90% 0.0273 [0.572] 0.0845* [1.719] 0.212*** [5.884] 0.306*** [9.123] 0.636*** [15.590] 1.047*** [12.302] 0.233*** [3.588] 0.275*** [4.533] 0.580*** [7.862] 0.934*** [6.074] 0.199*** [4.072] 0.233*** [5.106] 0.542*** [9.785] 0.969*** [8.384] 0.172*** [4.678] 0.251*** [7.340] 0.530*** [12.748] 0.925*** [10.661] 0.148*** [3.519] 0.283*** [7.213] 0.562*** [11.776] 1.066*** [10.705] Yes Yes Control Yes Yes Yes variables t-stat in brackets *, **, *** : significant at 10%, 5%, 1% 2SQR 2SLS 10% 0.126** [2.385] 0.169*** [3.099] 0.138*** [3.631] 0.179*** [4.497] 10% 0.0948 [1.102] 0.169* [1.878] 25% 0.141*** [3.128] 0.194*** [4.110] 50% 0.166*** [4.568] 0.183*** [4.800] 75% 0.155*** [3.871] 0.175*** [4.174] 90% 0.0524 [0.869] 0.122* [1.925] 0.203*** [3.461] 0.404*** [7.375] 0.700*** [10.513] 1.193*** [8.589] 0.294*** [6.285] 0.288*** [5.843] 0.532*** [9.823] 0.778*** [7.424] 0.198* [1.869] 0.218* [1.949] 0.476*** [3.878] 0.888*** [3.733] 0.242*** [4.373] 0.274*** [4.690] 0.537*** [8.365] 0.816*** [6.564] 0.268*** [5.971] 0.305*** [6.449] 0.511*** [9.836] 0.757*** [7.519] 0.259*** [5.257] 0.340*** [6.564] 0.547*** [9.593] 0.735*** [6.663] 0.310*** [4.167] 0.296*** [3.782] 0.576*** [6.680] 0.649*** [3.889] Yes Yes Yes Yes Yes Yes Yes Source : Author’s calculations Extraction of Table B.4: Wage equations in the urban and rural areas in 2012 Urban wage equation in 2012 Components Primary Secondary High school Vocational College Postgraduate 2SQR 2SLS 0.000577 [0.012] 0.0556 [1.125] 0.176*** [3.317] 0.242*** [4.636] 0.484*** [8.419] 0.766*** [8.911] Rural wage equation in 2012 10% -0.0693 [-0.773] 0.0402 [0.440] 0.0512 [0.521] 0.0799 [0.826] 0.349*** [3.278] 0.736*** [4.622] 25% 0.0173 [0.280] 0.0755 [1.196] 0.159** [2.345] 0.153** [2.297] 0.380*** [5.162] 0.656*** [5.974] Control yes yes yes variables t-stat in brackets; *, **, *** : significant at 10%, 5%, 1% 2SQR 2SLS 50% 0.0761 [1.431] 0.116** [2.132] 0.218*** [3.735] 0.269*** [4.680] 0.431*** [6.831] 0.686*** [7.265] 75% 0.000257 [0.004] 0.0509 [0.805] 0.137** [2.017] 0.328*** [4.908] 0.518*** [7.033] 0.851*** [7.729] 90% -0.0479 [-0.506] 0.0308 [0.318] 0.235** [2.264] 0.394*** [3.856] 0.765*** [6.795] 0.994*** [5.909] yes yes yes 0.148*** [5.585] 0.190*** [6.878] 0.301*** [8.732] 0.345*** [10.249] 0.591*** [14.053] 10% 0.176*** [3.113] 0.294*** [4.972] 0.343*** [4.656] 0.331*** [4.606] 0.577*** [6.416] 25% 0.183*** [5.242] 0.227*** [6.236] 0.290*** [6.394] 0.326*** [7.347] 0.569*** [10.271] 50% 0.143*** [5.535] 0.153*** [5.665] 0.217*** [6.462] 0.282*** [8.597] 0.479*** [11.687] 75% 0.102*** [3.558] 0.148*** [4.953] 0.245*** [6.568] 0.313*** [8.605] 0.530*** [11.650] 90% 0.0604 [1.544] 0.101** [2.467] 0.279*** [5.495] 0.355*** [7.149] 0.574*** [9.257] yes yes yes yes yes yes Source : Author’s calculations 12 4.2 Decomposition results Now we turn to the Machado – Mata technique to decompose the urban/rural wage gap across quantiles into two components – one due to urban – rural differences in the distributions of covariates and the other due to urban-rural differences in the distributions of returns to those covariates The decomposition of the gender wage gap based on Machado – Mata method is reported in Table C.1 As we can see from Table C.1, real hourly wages seem to be always greater for men than for women at all considered quantiles This wage gap is smaller at higher wage The largest gap is found at quantile 0.1 The gender wage differential declines over the time However, in each year, using the male wage structure as a reference, the gender wage gap is totally due to the differences in returns, which are traditionally interpreted as discrimination Table C.1 Decomposition of gender wage differential All sample Components Raw differential Due to endowments Due to returns Raw differential Due to endowments Due to returns 2002 0.2947*** [18.04] -0.0858** [-3.16] 0.3805*** [14.22] 0.2306*** [30.73] -0.075*** [-5.51] 0.3059*** 2012 By areas In urban 2002 Quantile 0.1 0.2173*** 0.1760*** [17.48] [8.87] -0.070*** -0.0348 [-2.92] [-1.46] 0.287*** 0.2109*** [21.81] [8.83] Quantile 0.25 0.1690*** 0.1595*** [19.89] [11.65] -0.076*** -0.046*** [-5.11] [-2.77] 0.2453*** 0.2055*** In rural 2012 2002 2012 0.1516*** [7.45] -0.0503** [-1.52] 0.2046*** [8.61] 0.3941*** [22.98] -0.071*** [-2.88] 0.4655*** [20.82] 0.2854*** [13.44] -0.061** [-1.16] 0.3465*** [12.77] 0.1589*** [9.29] -0.051*** [-2.44] 0.2101*** 0.3312*** [29.21] -0.064*** [-4.91] 0.3957*** 0.2071*** [18.20] -0.068*** [-2.75] 0.2755*** [23.49] [33.98] [11.65] Quantile 0.5 Raw differential 0.1569*** 0.121*** 0.1565*** [30.37] [15.70] [14.18] Due to endowments -0.084*** -0.085*** -0.073*** [-8.000] [-5.81] [-4.47] Due to returns 0.2416*** 0.207*** 0.2295*** [22.81] [23.65] [15.58] Quantile 0.75 Raw differential 0.0912*** 0.086*** 0.1590*** [17.30] [9.10] [11.30] Due to endowments -0.119*** -0.098*** -0.071*** [-9.36] [-5.73] [-4.07] Due to returns 0.2107*** 0.1849*** 0.2305*** [16.24] [10.66] [17.31] Quantile 0.9 Raw differential 0.0726*** 0.089*** 0.1314*** [8.31] [4.73] [6.05] Due to endowments -0.114*** -0.079*** -0.063*** [-6.57] [-2.98] [-2.89] Due to returns 0.1874*** 0.1688*** 0.1952*** [12.20] [6.25] [11.07] t-stat in brackets []*, **, *** : significant at 10%, 5%, 1% [12.19] [29.70] [16.96] 0.1477*** [8.10] -0.033* [-1.70] 0.1813*** [9.57] 0.2167*** [35.87] -0.053*** [-6.27] 0.2702*** [35.91] 0.1471*** [19.48] -0.063*** [-4.01] 0.2106*** [14.34] 0.1413*** [7.05] -0.004 [-0.23] 0.1463*** [7.28] 0.1287*** [16.75] -0.067*** [-6.32] 0.1965*** [23.80] 0.1076*** [9.46] -0.095*** [-4.28] 0.2031*** [11.95] 0.1760*** 0.0590*** 0.0569*** [5.87] [4.21] [3.11] 0.029 -0.101*** -0.128*** [0.81] [-5.56] [-3.5] 0.1469*** 0.1604*** 0.1850*** [4.66] [10.12] [7.63] Source : Author’s calculations The decomposition reveals very different patterns in the urban/rural wage gap depending on the considered quantiles Table C.2 contains the decomposition results of the wage gap between urban and rural areas The results show that wage in urban areas are always higher than in rural areas and the differentials are heterogeneous across quantiles Different in both covariates and the coefficients contribute to the urban/rural wage gaps, and both their effects are significantly different from zero at all estimated quantiles Nevertheless, the covariate effect is intuitively more important than the coefficient effect at five estimated quantiles Although the urban/rural wage gap declined over the year and the unexplained gaps 14 are larger in 2012 than in 2002, the unexplained gaps caused by coefficient effect are quite large, positive, and statistically significant across the distribution Table C2 Decomposition of the urban/rural wage differential All sample Components 2002 By gender Men 2012 2002 Women 2012 2002 2012 Quantile 0.1 Raw differential Due to endowments Due to returns Raw differential Due to endowments Due to returns Raw differential Due to endowments Due to returns Raw differential Due to endowments Due to returns Raw differential Due to endowments Due to returns 0.9817*** [102.45] 0.3786*** [19.40] 0.6031*** [39.30] 0.6692*** [87.32] 0.2974*** [26.48] 0.3717*** [41.86] 0.5373*** [80.77] 0.2726*** [37.58] 0.2647*** [52.12] 0.5523*** [52.05] 0.3134*** [29.13] 0.23.88*** [41.43] 0.6083*** [32.25] 0.3231*** [18.50] 0.2852*** [34.02] 0.2113*** [12.41] 0.1579*** [7.66] 0.0533*** [4.18] 0.9089*** 0.1629*** 1.1145*** 0.2992*** [45.48] [7.67] [56.19] [12.01] 0.3460*** 0.1354** 0.4248*** 0.1999*** [16.29] [8.16] [18.56] [5.11] 0.5628*** 0.0274 0.6897*** 0.0992*** [23.88] [1.60] [24.41] [3.34] Quantile 0.25 0.2199*** 0.6314*** 0.2148*** 0.7770*** 0.2638*** [21.4] [45.48] [11.88] [79.26] [12.15] 0.1520*** 0.2681*** 0.1466*** 0.3671*** 0.1922*** [11.04] [22.70] [10.41] [16.11] [7.11] 0.0678*** 0.36.33*** 0.0681*** 0.4099*** 0.0715*** [6.13] [21.73] [5.86] [19.97] [3.64] Quantile 0.5 0.2973*** 0.5486*** 0.3113*** 0.5800*** 0.3117*** [23.81] [51.85] [17.93] [45.82] [13.87] 0.1807*** 0.2486*** 0.1719*** 0.3377*** 0.2133*** [16.66] [28.44] [11.91] [18.70] [10.88] 0.1165*** 0.3000*** 0.1393*** 0.2422*** 0.0984*** [15.13] [43.80] [16.31] [26.28] [8.02] Quantile 0.75 0.4176*** 0.5883*** 0.4455*** 0.5342*** 0.4101*** [20.02] [46.82] [21.25] [27.09] [17.94] 0.2468*** 0.2917*** 0.2483*** 0.3729*** 0.2642*** [18.00] [25.96] [13.82] [20.40] [9.46] 0.1707*** 0.2966*** 0.1463*** 0.1613*** 0.1459*** [24.50] [49.37] [17.80] [13.84] [10.51] Quantile ị 0.9 0.5014*** 0.6531*** 0.5534*** 0.5400*** 0.4345*** [20.95] [34.37] [19.29] [19.04] [18.19] 0.2827*** 0.3100*** 0.3170*** 0.3488*** 0.2352*** [13.25] [24.17] [9.33] [13.80] [7.41] 0.2186*** 0.3431*** 0.2363*** 0.1911*** 0.1993*** [20.53] [44.68] [18.53] [11.13] [8.94] t-stat in bracket ; *, **, *** : significant at 10%, 5%, 1% Source : Author’s calculations 15 In this section, dissertation analyzes change of wage distribution between 2002 and 2012 by decomposing the increase of wage into change due to observed skill effect and unobserved effect The growth rates of wage at bottom quantiles is higher than at the top quantile The real wage in the urban areas grows faster than that in the rural Our decomposition reveals that both covariate effect and return effect contribute to the wage increase between 2002 and 2012 However, the return effect, which is known as unexplained part, is substantially greater than the covariate effect, approximately times on average Table C.3 Decomposition the wage differential between 2002 and 2012 By groups Components Raw differential Due to endowments Due to returns All sample 1.367*** [145.577] 0.123*** [3.672] 1.244*** [91.118] Gender Men Quantile 0.1 1.342*** [101.656] 0.119*** [3.238] 1.223*** [77.006] Areas Women Urban Rural 1.428*** [61.42] 0.373*** [14.528] 1.055*** [60.778] 0.777*** [42.271] -0.24*** [-1.734] 1.021*** [71.903] 1.548*** [103.601] 0.211*** [4.353] 1.337*** [101.6] 1.103*** [69.949] 0.302*** [16.807] 0.801*** [70.197] 0.719*** [51.248] -0.09*** [-1.073] 0.817*** [79.79] 1.165*** [140.305] 0.115*** [3.244] 1.051*** [140.38] 0.892*** [68.769] 0.273*** [19.941] 0.619*** [129.139] 0.673*** [62.444] 0.008*** [0.137] 0.665*** [102.848] 0.914*** [112.226] 0.105*** [6.152] 0.808*** [230.859] 0.749*** [56.869] 0.28*** 0.643*** [63.457] 0.089*** 0.779*** [95.239] 0.139*** Quantile 0.25 Raw differential Due to endowments Due to returns Raw differential Due to endowments Due to returns Raw differential Due to endowments 1.057*** [178.314] 0.071*** [3.759] 0.986*** [132.765] 0.867*** [127.94] 0.117*** [10.341] 0.75*** [176.573] 0.747*** [92.739] 0.179*** 1.038*** [104.586] 0.073*** [2.433] 0.965*** [125.565] Quantile 0.5 0.861*** [101.255] 0.102*** [4.914] 0.758*** [190.854] Quantile 0.75 0.746*** [61.825] 0.158*** 16 Due to returns Raw differential Due to endowments Due to returns [14.015] 0.568*** [117.008] 0.684*** [53.736] 0.216*** [10.547] 0.469*** [58.661] [7.758] [18.286] [1.296] [8.655] 0.588*** 0.47*** 0.554*** 0.64*** [106.025] [80.697] [58.209] [138.61] Quantile 0.9 0.692*** 0.679*** 0.582*** 0.69*** [31.866] [44.782] [34.397] [41.903] 0.208*** 0.281*** 0.134*** 0.159*** [8.581] [13.077] [2.147] [7.091] 0.484*** 0.398*** 0.448*** 0.531*** [42.468] [38.794] [30.975] [76.635] t-stat in brackets; *, **, *** : significant at 10%, 5%, 1% Source :Author’s calculations CHAPTER CONCLUSION AND POLICY IMPLICATIONS 5.1 Conclusion 5.1.1 The estimation of wage equation By gender The results of data analysis showed that qualifications seriously affect wages in all considered quantiles The higher level of qualifications, the higher wage In 2002, the impact of qualifications on wage is stronger at lower quantiles In 2012, at lower level of qualification, Female’s regression coefficients are greater than male’s; but the opposite pattern is found at higher level of qualification By zones In both urban and rural areas, workers who have higher qualification receive higher real wages across the wage distribution This is consistent with the previous results In 2002, urban workers’ wage changes among qualifications tend to be smaller at higher quantiles However, in rural areas, the opposite pattern is found In 17 particularly, wage changes among qualifications tend to greater at higher quantiles By time Returns to education in 2002 are mostly higher than that in 2012 across considered quantiles This implies that there is a structure change in wage between 2002 and 2012 There is a similar pattern for returns to education in urban areas Nevertheless, it is quite different in rural areas where returns on lower qualifications in 2012 is smaller than in 2002; meanwhile, returns on higher education in 2012 are higher than in 2002 5.1.2 Decomposition of wage differential Decomposition of gender wage differential In both 2002 and 2012, all quantiles of men’s hourly real wage is higher than women’s Differences in characteristics between male and female not explain the gender pay gap Negative sign of the characteristic effect implies that the gender wage gap is totally due to the differences in returns, which are traditionally interpreted as discrimination This is statistical evidence that shows the existence of gender inequality in Vietnam The gender wage differential tend to decrease over time at almost considered quantiles except quantile 0.9 The decrease occurred primarily in rural areas Gender wage gap in urban areas seems to be unchanged over the years Decomposition of urban/rural wage differential 18 Wage is greater in urban areas than in rural areas along the wage scale The characteristic effect participates in explaining the wage gap between the two zones and play a more important role than the coefficient effect across all considered quantiles However, there is still significant evidence of inequality between urban and rural zones However, the urban/rural wage gap sharply declined over the year Decomposition the wage differential between 2002 and 2012 Real hourly wage in 2012 is higher than in 2002 along the wage distribution Lower quantiles of wage distribution increases faster than higher quantiles The change due to labor force’s characteristics is statistically significant, but quite small in magnitude Most of the change is due to returns to those characteristics, which reflects the change in the wage structure in Vietnam within 10 years Both men’s and women’s wage distribution shift to the right which indicate a growth over time The general trend is that the higher quantiles of wage increase less than the lower quantiles Most of the wage change in both groups due to changes in returns to covariates In addition, real hourly wages increased over time in both urban and rural zones but rural areas have higher growth rates The improvement in the labor characteristics involved in explaining the wage increases at all the considered quantiles but its percentage is very small Most of wage changes are caused by change in the regression coefficients 19 5.2 Policy implications 5.2.1 Suggestions that may help to enhance workers’ wage Employees need to recognize the importance of education The compensation they receive is highly correlates with their level of education This implies that employees should have a plan to improve their own education level In addition to have higher qualification, they also enhance their skills through training process Employers should allow and give their employees chances to improve employees’ level of education, thereby enriching their skills and enhancing labor productivity Government policies not only concentrate in improving the concurrent labor force but also prepare for the future workforce That can be carried out by invest intensively and efficiently in education system from pre-school to graduate The result acquired by quantile regression analysis indicates that government policies that encourage people choose to become more educated usually have a significant impact on society, especially the low income group Additionally, if the high-wage groups of workers achieve a high level of qualifications such as college or graduation, they have wages increased faster than low-wage group Thus this thesis recommends that policies which promote education should account for the different impact on different groups in order to be more realistic and efficient 20 5.2.2 Suggestions to narrow the wage gap To improve the gender pay gap in the short run and long run, we should concentrate on following suggestions: - It is necessary to focus first on the rural areas in the progress of narrowing the gender discrimination on wage, especially for the low income group - The government should improve the regulation in order to minimize the wage inequality between men and women or at least to create an environment for fair competition environment for them in labor market - The government and local authorities should enhance people’s awareness about the gender equality for parents, employers and policy makers - In addition to protect women’s right and promote gender equality, the authorities also need to pay attention to the problem of occupational segregation - In the long term, gender equality can be integrated into the lessons at school from elementary to high school through stories, pictures, games, extracurricular activities to help mitigate thought gender discrimination - The authorities should forecast and orient the labor demand on the market demand in the long term and there are regulated by appropriate policies to balance the labor market in the future 21 - We need to conduct more researches to have an overall view about the gender discrimination in Vietnam, especially in education, employment and payment To shorten the urban – rural wage gap but ensure the growth of living standard for both zones, this study implies some suggestions - We should continue to implement the policies such as national poverty reduction policy, lending capital policy from social funds as we has done in recent years Because these policies have resulted in a considerable decrease in poverty rates Moreoves, these help to improve people living standard and narrow the urban – rural income gap - We should pay more attention to the economic solutions instead of the administrative solutions in retaining skilled labors in rural zones Orienting the progress of urbanization in a reasonable manner will help to attract the highly qualified workforce to the rural areas - In the rural areas, we improve not only the labor skills but also the infrastructure, compensation policy, administrative procedures in order to reduce urban – rural wage differential 5.3 Thesis’s contributions 5.3.1 Theoretical contributions This study briefly and systematically introduces the background of quantile regression which is quite new and has not been used widely 22 in the empirical studies in Vietnam There have been a few of studies in Vietnam that applied quantile regression, so this method has not been fully explained So this study might become a suitable reference for researchers, data analysts and students who are interested in quantile regression Additionally, this study reviews the development of wage gap decomposition method which have been becoming more complete over time, especially the Machado – Mata decomposition This study has also summarized the literature review of wage gap studies in the world as well as in Vietnam Consequently, his help building up a theoretical framework for finding empirical evidence for wage gap in Vietnam 5.3.2 Empirical contributions Firstly, this study estimates wage regressions in Vietnam during the period from 2002 to 2012 by quantile regression which allow to demonstrate the impact of the qualifications, occupations, ethnic groups, and regions on real hourly wage This study also compares the estimated wage regression between urban and rural, between male and female workers to see how the wage structure is different among these groups In addition, the wage regression in 2002 is compared with that in 2012 to clarify the change of wage structure over time 23 Secondly, this study provides strongly empirical evidence of the existence of gender wage differential in Vietnam There is also wage differential between urban and rural areas However the gender and urban – rural wage gap in Vietnam have reduced substantially over the years Thirdly, this study applied the Machado Mata technique to decompose wage differences between men and women, between urban and rural workers on the entire wage distribution The decomposition terms vary across the quantiles of the wage distribution In general, the result indicates that the gender wage differential is not explained by observable differences in male and female characteristics This implies that there is an estimated effects of gender discrimination Moreover, based on statistically evidence, the estimated result reveals that approximately half of the urban/rural is due to differences in observed characteristics between urban and rural workers The remains is explained by differences in the returns to observed characteristics across these two areas Thereby, the empirical evidence for inequality in wages between men and women, between urban and rural areas are significant This study also break down the wage gap between 2012 and 2002 to determine the proportion of wage changes due to the improvement in the labor force’s characteristics and to obtain the percentage of wage changes caused by the return of these characteristics 24 Finally, according to those analysis the study gives some suggestions to improve the wage structure in Vietnam and enhance the equality in wage between men and women or urban and rural workers From a policy point of view, the main findings of this study suggest a more effective role of anti-discrimination labor market policies aimed at reducing gender and urban/rural wage differentials throughout the wage distribution in the Vietnam economy 5.4 Limitations and further extensions The subject of this thesis can be extended in directions below: - Decomposing the wage gap into parts by groups of explanatory variables in place of all of them - Using quantile regression for panel data in order to have more convincing results 25 ... decomposing this gap into the explained and unexplained components during the period from 2002- 2012 4) Determining the urban/rural wage gap in Vietnam and decomposing this gap into the explained... examine the existence of gender and urban/rural wage gap, and (3) to decompose these wage gaps to clarify whether there are wage discrimination in Vietnam throughout the wage distribution These... factors in motivating employees Because wage depends on a variety of determinants, the existence of the wage gap is inevitable According to economic theory, the wage gap can be decomposed into two