Impact of income inequality on health from middle and high income countries in 1991 2010

89 5 0
  • Loading ...
1/89 trang

Thông tin tài liệu

Ngày đăng: 03/01/2019, 00:09

UNIVERSITY OF ECONOMICS HO CHI MINH CITY VIETNAM ERASMUS UNVERSITY ROTTERDAM INSTITUTE OF SOCIAL STUDIES THE NETHERLANDS VIETNAM – THE NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS IMPACT OF INCOME INEQUALITY ON HEALTH IN MIDDLE AND HIGH INCOME COUNTRIES IN 1991 - 2010 BY PHAM DANG XUAN ANH MASTER OF ARTS IN DEVELOPMENT ECONOMICS HO CHI MINH CITY, November 2016 UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES HO CHI MINH CITY VIETNAM THE HAGUE THE NETHERLANDS VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS IMPACT OF INCOME INEQUALITY ON HEALTH IN MIDDLE AND HIGH INCOME COUNTRIES IN 1991 - 2010 A thesis submitted in partial fulfilment of the requirements for the degree of MASTER OF ARTS IN DEVELOPMENT ECONOMICS By PHAM DANG XUAN ANH Academic Supervisor: DR NGUYEN VAN NGAI HO CHI MINH CITY, November 2016 Declaration I hereby declare that this thesis has been exclusively the original work of myself and the result of my own research, except where due reference has been made in the content, and free from plagiarism of the work of others I also certify that this master thesis has not been accepted in any degree or not under submission for any other degree or qualification, other than that of the degree of Master of Arts in Development Economics at Vietnam - Netherlands Programme Abstract: The income inequality and average heath of population level relation is tested in this paper with panel data of 48 high and middle income countries over 20 recent years Evidence of significantly negative impact of income distribution on life expectancy at birth and positive impact on infant mortality rate has been found Moreover, GDP per capita also has similar impact on heath in opposite directions Even though the marginal effects are quantitatively small, results are found to be quite robust when controlling for endogeneity concerns and other issues JEL: I14, I15, O15, C33, C36 Key words: Income inequality, life expectancy, infant mortality rate, health, human development, GDP per capita, secondary schooling, health spending, panel data Table of Contents Page Declaration i Abstract ii Tables of contents iii Abbreviations iv List of Figures v List of Tables vi Chapter 1: Introduction 1.1 Problem Statement .1 1.2 Research Objectives 1.3 Research methods and expected outcome 1.4 Thesis Structure Chapter 2: Literature review 2.1 Theoretical Background .5 2.1.1 Income and effects to health 2.1.2 Income inequality hypothesis 2.2 The conceptual framework 2.3 Empirical Studies Findings Chapter 3: Data and Model Specifications 14 3.1 Empirical Model .14 3.2 Data sources and Description 17 3.3 Estimation Method 28 3.3.1 Panel Data Model 3.3.2 Tests and Control for robustness of results Chapter 4: Results and Discussion .32 4.1 Descriptive Statistics 32 4.2 Result Interpretation 38 Chapter 5: Conclusions 48 5.1 Concluding Remarks 48 5.2 Policy implication 51 5.3 Limitations and further researches 52 References .54 Appendix .60 Abbreviations OECD - Organisation for Economic Co-operation and Development WHO – World Health Organisation UNESCO - United Nations Educational, Scientific, and Cultural Organization AGOA - African Growth and Opportunity Act WIID – World Income Inequality Database UNU-WIDER – United Nations University-World Institute for Development Economics Research CME – Child Mortality Estimates LE – Life Expectancy IMR – Infant Mortality Rate IV – Instrumental Variable GDP – Gross Domestic Product GNI – Gross National Income PPP - Purchasing Power Parities FE – Fixed Effects RE - Random Effects GLS - Generalized Least Squared FGLS - Estimator or Feasible Generalized Least Squared OLS – Ordinary Least Squared 2SLS - Two-stage Least Squares LM – Lagrange Multiplier GMM - Generalized Method of Moments List of Figures Page Figure 2.1: Life expectancy at birth and real GDP per capita in 48 countries, 19912010 Figure 2.2: Possible channels income inequality might affect health Figure 3.3: Gini ratio estimation by Lorenz curve .19 Figure 4.4: Life expectancy and infant mortality rate versus GDP per capita 33 Figure 4.5: Life expectancy and infant mortality rate versus Gini index 34 Figure 4.6: Life expectancy and infant mortality rate versus Health spending per capita 35 Figure 4.7: Life expectancy and infant mortality rate versus Secondary schooling enrolment ratio 36 List of Tables Page Table 3.1: Summary of data resources and denotation used in models 17 Table A.2: Summary of hypotheses testing of model effects selections 60 Table 4.3: Descriptive statistics for the explanatory variables, 48 countries 19912010 32 Table A.4: Model selection and Tests for life expectancy 60 Table A.5: Model selection and Tests for IMR .61 Table A.6: Correlations of variables in models .62 Table A.7: Regression of Gini with impact of IVs 63 Table 4.8: Comparison of OLS regressions and panel effect regressions - Life Expectancy 39 Table 4.9: Comparison of OLS regressions and panel effect regressions – IMR 39 Table 4.10: Effects of income inequality using fixed-effects and random-effects 42 Table 4.11: Effects of income inequality using instrumental variables on life expectancy 42 Table 4.12: Regressions on life expectancy with interactions of Gini and GDP per head - Trade Openness instrument .44 Table 4.13: Regressions on life expectancy with interactions of Gini and GDP per head – Investment Ratio instrument .45 Table 4.14: Regressions on infant mortality rate with interactions of Gini and GDP per head 46 Table A.15: Regressions with system GMM on life expectancy and infant mortality rate 64 Table A.16: List of countries in research according to World Bank 64 Table A.17: List of STATA output of empirical results .65 Chapter 1: Introduction 1.1 Problem Statement In the recent years, the researches on health and its surrounding relationships has been on the rise Explanatory factors affect health as the whole population is point of interest of many authors The outcomes of studies are among most controversies, not only in the conclusions, but also in the discussions and criticism of limitations regarding the methodologies, data, underlying channels of mechanisms Health, as definition, is “a state of complete physical, mental, and social well-being and not merely the absence of disease or infirmity” as from the World Health Organization (WHO) The concerns of health are one of the most significant matters in modern societies With the advances in technology and health care, all aspects of health have been considerably improved in almost every country, especially in life expectancy and infant mortality rate Life expectancy and mortality rate don’t necessarily reflect the quality of life in term of the income metrics, but in the most popular studies in this filed, the connection between the these two major metrics of life quality, and other income based measurements, has been investigated and hence, established (Lynch et al., 2004; Ellison, 2002) Health, at individual or population level has exposed some degrees of relationships to inequality according to Rodger (1979), Preston (1975), and Deaton (2001) Besides that, there is recently increase in studies regarding health and population health and its nexus with income, and especially, inequality (Gravelle et al., 2002; Torre & Myrskylä, 2014) Even though the measurement of inequality is itself hardly intuitive (Lynch et al., 2004), many economists tried to quantify it through numbers of metrics Therefore, the relationship between income inequality and the health are becoming important In other aspect, the association between economic growth in terms of income distribution and quality of life metrics are ongoing topic in economic studies The quality of life can only be raised if growth and standard of living go together Among determinants of a highly developed society, health and education are key opponents Apart from education attainment, which is a proven factor interacting with wealth distribution, health at aggregate level such as life expectancy and infant mortality rate has exposed some degrees of connections to income inequality according to Rodger (1979), Preston (1975), Deaton (2001) Alternatively, there are empirical works of researches on the connection between human capital and economic growth, in terms of income level As results, there is recently increase in studies regarding health and population health and its nexus with income, and more extending, income inequality (Gravelle et al., 2002; Torre & Myrskylä, 2014) Equally important, the mutual effect of health and income inequality is a source of debate in many papers In one hand, some papers have been indicated that the part of the income inequality hypothesis On the other side, the effects of health outcomes on income conception and vice versa have been investigated for long time (Leigh et al., 2009) However, the connections between three concepts: economic inequality, health progress and their interactions with income driving mechanism are not easily established or observe with solid evidences The consistent results of researches of this interest are still exceptionally unconvincing because of conflicting conclusions 1.2 Research objectives Due to the rising health concerns in welfare, especially when it comes to child mortality reduction and prolong human longevity, many studies has been accelerating the knowledge and connections of health policies in terms of income distribution instruments such as Gini or Robin Hood Indexes Exploring the pattern of Gini coefficient linking to life expectancy and IMR, with control of some insightful factors such as level of income, health spending, etc… are the main purposes of this income inequality on health indexes study, and are largely to contribute to literatures The longevity and quality of life are essential to modern societies, but lacking of understanding of how income inequality could impact health, lacking convinced evidences, particularly combined with controversies in underlying patterns of pathways in evidences in groups of countries, making perspectives become distorted Therefore, the proposed objectives of this research are to: Estimate the effect of income inequality on life expectancy at birth and infant mortality rate in some of middle and high income countries in period of 1991-2010 10 xttest1 Tests for the error component model: lifeexpc[id,t] = Xb + u[id] + v[id,t] v[id,t] = lambda v[id,(t-1)] + e[id,t] Estimated results: Var lifeexpc e u sd = sqrt(Var) 18.33578 6063904 7.014585 4.28203 77871075 2.6485063 Tests: Random Effects, Two Sided: ALM(Var(u)=0) = 2720.61 Pr>chi2(1) = 0.0000 Random Effects, One Sided: ALM(Var(u)=0) = 52.16 Pr>N(0,1) = 0.0000 3.76 Pr>chi2(1) = 0.0526 Joint Test: LM(Var(u)=0,lambda=0) = 3239.21 Pr>chi2(2) = 0.0000 Serial Correlation: ALM(lambda=0) = hausman fe re Coefficients (b) (B) fe re gini gdppc healthpc schoolenro~e -.0403224 0001713 0006826 0331182 -.0382807 0001527 0007534 0330098 (b-B) Difference -.0020418 0000186 -.0000708 0001084 sqrt(diag(V_b-V_B)) S.E .0018361 6.72e-06 0000252 0005321 b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(4) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 8.95 Prob>chi2 = 0.0623 75 xtreg imr gini gdppc healthpc schoolenrolshare, fe Fixed-effects (within) regression Group variable: id Number of obs Number of groups = = 670 47 R-sq: Obs per group: = avg = max = 14.3 16 within = 0.3361 between = 0.5768 overall = 0.5494 corr(u_i, Xb) F(4,619) Prob > F = -0.5155 = = imr Coef gini gdppc healthpc schoolenrolshare _cons 1588347 -.0005857 0010187 -.1420931 32.15258 0359751 0000622 0003509 0150997 2.011257 sigma_u sigma_e rho 7.6190296 2.895912 87376849 (fraction of variance due to u_i) F test that all u_i=0: Std Err F(46, 619) = t 4.42 -9.41 2.90 -9.41 15.99 52.68 P>|t| 0.000 0.000 0.004 0.000 0.000 78.34 0.0000 [95% Conf Interval] 0881867 -.0007078 0003295 -.171746 28.20287 2294827 -.0004635 0017078 -.1124403 36.10229 Prob > F = 0.0000 xttest3 Modified Wald test for groupwise heteroskedasticity in fixed effect regression model H0: sigma(i)^2 = sigma^2 for all i chi2 (47) = Prob>chi2 = 77354.49 0.0000 76 xtreg imr gini gdppc healthpc schoolenrolshare, re Random-effects GLS regression Group variable: id Number of obs Number of groups = = 670 47 R-sq: Obs per group: = avg = max = 14.3 16 within = 0.3324 between = 0.5956 overall = 0.5670 corr(u_i, X) Wald chi2(4) Prob > chi2 = (assumed) = = imr Coef gini gdppc healthpc schoolenrolshare _cons 1702108 -.0004425 0004814 -.1425683 29.34275 0347011 0000511 0003261 014881 2.133157 sigma_u sigma_e rho 5.7737312 2.895912 79899693 (fraction of variance due to u_i) Std Err z 4.91 -8.66 1.48 -9.58 13.76 P>|z| 0.000 0.000 0.140 0.000 0.000 374.14 0.0000 [95% Conf Interval] 1021979 -.0005426 -.0001578 -.1717346 25.16184 2382237 -.0003423 0011206 -.1134021 33.52366 xttest0 Breusch and Pagan Lagrangian multiplier test for random effects imr[id,t] = Xb + u[id] + e[id,t] Estimated results: Var imr e u Test: sd = sqrt(Var) 106.0995 8.386306 33.33597 10.30046 2.895912 5.773731 Var(u) = chibar2(01) = Prob > chibar2 = 1898.37 0.0000 77 xttest1 Tests for the error component model: imr[id,t] = Xb + u[id] + v[id,t] v[id,t] = lambda v[id,(t-1)] + e[id,t] Estimated results: Var imr e u sd = sqrt(Var) 106.0995 8.386306 33.33597 10.30046 2.895912 5.7737312 Tests: Random Effects, Two Sided: ALM(Var(u)=0) = 1489.45 Pr>chi2(1) = 0.0000 Random Effects, One Sided: ALM(Var(u)=0) = 38.59 Pr>N(0,1) = 0.0000 26.87 Pr>chi2(1) = 0.0000 Joint Test: LM(Var(u)=0,lambda=0) = 1925.24 Pr>chi2(2) = 0.0000 Serial Correlation: ALM(lambda=0) = hausman fe re Coefficients (b) (B) fe re gini gdppc healthpc schoolenro~e 1588347 -.0005857 0010187 -.1420931 1702108 -.0004425 0004814 -.1425683 (b-B) Difference -.0113761 -.0001432 0005373 0004752 sqrt(diag(V_b-V_B)) S.E .0094891 0000355 0001296 0025607 b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(4) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 57.87 Prob>chi2 = 0.0000 (V_b-V_B is not positive definite) 78 xtserial lifeexpc gini gdppc healthpc schoolenrolshare Wooldridge test for autocorrelation in panel data H0: no first-order autocorrelation F( 1, 46) = 236.397 Prob > F = 0.0000 xtserial imr gini gdppc healthpc schoolenrolshare Wooldridge test for autocorrelation in panel data H0: no first-order autocorrelation F( 1, 46) = 2836.817 Prob > F = 0.0000 xtreg lifeexpc gini gdppc healthpc schoolenrolshare, re robust Random-effects GLS regression Group variable: id Number of obs Number of groups = = 669 47 R-sq: Obs per group: = avg = max = 14.2 16 within = 0.6246 between = 0.6247 overall = 0.6289 corr(u_i, X) Wald chi2(4) Prob > chi2 = (assumed) = = 72.20 0.0000 (Std Err adjusted for 47 clusters in id) Robust Std Err lifeexpc Coef z gini gdppc healthpc schoolenrolshare _cons -.0382807 0001527 0007534 0330098 67.99819 0233347 0000483 0002032 0107689 1.546485 sigma_u sigma_e rho 2.6485063 77871075 92043139 (fraction of variance due to u_i) -1.64 3.16 3.71 3.07 43.97 P>|z| 0.101 0.002 0.000 0.002 0.000 [95% Conf Interval] -.0840159 000058 0003551 0119031 64.96714 0074546 0002474 0011517 0541164 71.02925 79 xtreg imr gini gdppc healthpc schoolenrolshare, fe robust Fixed-effects (within) regression Group variable: id Number of obs Number of groups = = 670 47 R-sq: Obs per group: = avg = max = 14.3 16 within = 0.3361 between = 0.5768 overall = 0.5494 corr(u_i, Xb) F(4,46) Prob > F = -0.5155 = = 8.31 0.0000 (Std Err adjusted for 47 clusters in id) Robust Std Err imr Coef gini gdppc healthpc schoolenrolshare _cons 1588347 -.0005857 0010187 -.1420931 32.15258 0941394 0001387 0003657 0436419 6.602331 sigma_u sigma_e rho 7.6190296 2.895912 87376849 (fraction of variance due to u_i) t 1.69 -4.22 2.79 -3.26 4.87 P>|t| 0.098 0.000 0.008 0.002 0.000 [95% Conf Interval] -.0306582 -.0008649 0002826 -.2299398 18.86278 3483276 -.0003065 0017548 -.0542465 45.44238 xtscc imr gini gdppc healthpc schoolenrolshare, fe lag(2) Regression with Driscoll-Kraay standard errors Method: Fixed-effects regression Group variable (i): id maximum lag: imr Coef gini gdppc healthpc schoolenrolshare _cons 1588347 -.0005857 0010187 -.1420931 32.15258 Drisc/Kraay Std Err .0451758 0000293 0003188 0296809 2.980118 Number Number F( 4, Prob > within t 3.52 -19.97 3.20 -4.79 10.79 of obs of groups 46) F R-squared P>|t| 0.001 0.000 0.003 0.000 0.000 = = = = = 670 47 124.01 0.0000 0.3361 [95% Conf Interval] 0679006 -.0006447 0003769 -.2018378 26.15391 2497688 -.0005266 0016604 -.0823485 38.15125 80 xtreg gdppc gini healthpc schoolenrolshare investmentshare tradeshare, fe robust Fixed-effects (within) regression Group variable: id Number of obs Number of groups = = 670 47 R-sq: Obs per group: = avg = max = 14.3 16 within = 0.7856 between = 0.8578 overall = 0.8136 corr(u_i, Xb) F(5,46) Prob > F = 0.7572 = = 47.27 0.0000 (Std Err adjusted for 47 clusters in id) Robust Std Err gdppc Coef t gini healthpc schoolenrolshare investmentshare tradeshare _cons -24.38816 3.927062 32.33735 147.3332 59.47592 6123.816 35.35491 4492532 18.42666 33.44948 15.43631 2423.913 sigma_u sigma_e rho 11230.659 1582.6127 98052851 (fraction of variance due to u_i) -0.69 8.74 1.75 4.40 3.85 2.53 P>|t| 0.494 0.000 0.086 0.000 0.000 0.015 [95% Conf Interval] -95.5539 3.022763 -4.753586 80.00289 28.40424 1244.731 46.77758 4.831362 69.42829 214.6635 90.5476 11002.9 xtreg lifeexpc gini gdppc healthpc schoolenrolshare investmentshare tradeshare, re r > obust Random-effects GLS regression Group variable: id Number of obs Number of groups = = 669 47 R-sq: Obs per group: = avg = max = 14.2 16 within = 0.6257 between = 0.6162 overall = 0.6206 corr(u_i, X) Wald chi2(6) Prob > chi2 = (assumed) = = 117.97 0.0000 (Std Err adjusted for 47 clusters in id) Robust Std Err lifeexpc Coef z gini gdppc healthpc schoolenrolshare investmentshare tradeshare _cons -.0383525 0001384 0007828 0328515 0088031 0030622 67.82052 0233636 0000426 0002109 0107015 019067 0090016 1.541029 sigma_u sigma_e rho 2.5086778 77864249 91212953 (fraction of variance due to u_i) -1.64 3.25 3.71 3.07 0.46 0.34 44.01 P>|z| 0.101 0.001 0.000 0.002 0.644 0.734 0.000 [95% Conf Interval] -.0841443 000055 0003696 011877 -.0285675 -.0145807 64.80016 0074394 0002219 0011961 0538261 0461737 0207051 70.84089 81 xtreg imr gini gdppc healthpc schoolenrolshare investmentshare tradeshare, fe robust Fixed-effects (within) regression Group variable: id Number of obs Number of groups = = 670 47 R-sq: Obs per group: = avg = max = 14.3 16 within = 0.3760 between = 0.5745 overall = 0.5492 corr(u_i, Xb) F(6,46) Prob > F = -0.2047 = = 11.20 0.0000 (Std Err adjusted for 47 clusters in id) Robust Std Err imr Coef gini gdppc healthpc schoolenrolshare investmentshare tradeshare _cons 1596326 -.0003468 0005134 -.1386419 -.1366209 -.0459075 34.1106 0907058 000142 0003772 043062 0545942 0324247 6.551349 sigma_u sigma_e rho 6.7712854 2.8121308 85289589 (fraction of variance due to u_i) xtivreg t P>|t| 1.76 -2.44 1.36 -3.22 -2.50 -1.42 5.21 0.085 0.019 0.180 0.002 0.016 0.164 0.000 [95% Conf Interval] -.0229487 -.0006326 -.0002459 -.2253212 -.2465133 -.1111751 20.92342 342214 -.0000609 0012728 -.0519627 -.0267285 0193601 47.29778 lifeexpc gini healthpc schoolenrolshare ( gdppc = tradeshare), re G2SLS random-effects IV regression Group variable: id Number of obs Number of groups = = 669 47 R-sq: Obs per group: = avg = max = 14.2 16 within = 0.6231 between = 0.6157 overall = 0.6178 corr(u_i, X) lifeexpc Coef gdppc gini healthpc schoolenrolshare _cons 0002014 -.036168 0005209 0303537 67.46121 sigma_u sigma_e rho 3.0666311 78377151 93868381 Instrumented: Instruments: Wald chi2(4) Prob > chi2 = (assumed) Std Err .0000389 0097694 0001958 0045518 7939808 z 5.18 -3.70 2.66 6.67 84.97 P>|z| 0.000 0.000 0.008 0.000 0.000 = = 1006.75 0.0000 [95% Conf Interval] 0001251 -.0553156 0001371 0214324 65.90503 0002776 -.0170203 0009047 0392751 69.01738 (fraction of variance due to u_i) gdppc gini healthpc schoolenrolshare tradeshare 82 xtivreg lifeexpc gini healthpc schoolenrolshare ( gdppc = investmentshare), re G2SLS random-effects IV regression Group variable: id Number of obs Number of groups = = 669 47 R-sq: Obs per group: = avg = max = 14.2 16 within = 0.6242 between = 0.6177 overall = 0.6203 corr(u_i, X) Wald chi2(4) Prob > chi2 = (assumed) = = lifeexpc Coef gdppc gini healthpc schoolenrolshare _cons 0001932 -.0382287 0005706 0314538 67.53321 0000444 009621 0002183 0045775 9497314 sigma_u sigma_e rho 4.5367962 77948985 97132606 (fraction of variance due to u_i) Instrumented: Instruments: Std Err z 4.35 -3.97 2.61 6.87 71.11 P>|z| 0.000 0.000 0.009 0.000 0.000 1007.10 0.0000 [95% Conf Interval] 0001062 -.0570855 0001428 0224821 65.67178 0002803 -.019372 0009984 0404256 69.39465 gdppc gini healthpc schoolenrolshare investmentshare reg lifeexpc gini gdppc healthpc schoolenrolshare Source SS df MS Model Residual 7889.43457 4358.86607 664 1972.35864 6.56455733 Total 12248.3006 668 18.3357794 lifeexpc Coef gini gdppc healthpc schoolenrolshare _cons -.0607617 0000513 0012452 0396939 69.69965 Std Err .0145679 0000141 0001507 0068733 9751057 Number of obs F( 4, 664) Prob > F R-squared Adj R-squared Root MSE t -4.17 3.63 8.26 5.78 71.48 P>|t| 0.000 0.000 0.000 0.000 0.000 = = = = = = 669 300.46 0.0000 0.6441 0.6420 2.5621 [95% Conf Interval] -.0893664 0000235 0009492 026198 67.78499 -.032157 0000791 0015411 0531899 71.61431 83 reg imr gini gdppc healthpc schoolenrolshare Source SS df MS Model Residual 45468.0667 25512.4984 665 11367.0167 38.3646592 Total 70980.5651 669 106.099499 imr Coef gini gdppc healthpc schoolenrolshare _cons 5142051 -.0001073 -.0010086 -.1006145 8.073983 Std Err .0351514 0000342 0003643 0166159 2.354513 Number of obs F( 4, 665) Prob > F R-squared Adj R-squared Root MSE t P>|t| 14.63 -3.14 -2.77 -6.06 3.43 0.000 0.002 0.006 0.000 0.001 = = = = = = 670 296.29 0.0000 0.6406 0.6384 6.1939 [95% Conf Interval] 4451841 -.0001744 -.001724 -.1332405 3.450808 5832261 -.0000402 -.0002932 -.0679884 12.69716 xtreg lifeexpc gini gdppc healthpc schoolenrolshare hi higdppc higini, re robust Random-effects GLS regression Group variable: id Number of obs Number of groups = = 669 47 R-sq: Obs per group: = avg = max = 14.2 16 within = 0.6547 between = 0.6238 overall = 0.6312 corr(u_i, X) Wald chi2(7) Prob > chi2 = (assumed) = = 169.17 0.0000 (Std Err adjusted for 47 clusters in id) Robust Std Err lifeexpc Coef gini gdppc healthpc schoolenrolshare hi higdppc higini _cons -.0324822 0003326 0008217 0236519 1.373233 -.00021 0114881 67.50106 0238124 000063 0002197 0098354 2.050601 000075 0425485 1.23819 sigma_u sigma_e rho 2.5823365 74843057 92250954 (fraction of variance due to u_i) z -1.36 5.28 3.74 2.40 0.67 -2.80 0.27 54.52 P>|z| 0.173 0.000 0.000 0.016 0.503 0.005 0.787 0.000 [95% Conf Interval] -.0791537 0002091 0003911 0043749 -2.645871 -.000357 -.0719054 65.07425 0141892 0004562 0012524 0429289 5.392337 -.000063 0948816 69.92787 84 xtreg imr gini gdppc healthpc schoolenrolshare hi higdppc, fe robust note: hi omitted because of collinearity Fixed-effects (within) regression Group variable: id Number of obs Number of groups = = 670 47 R-sq: Obs per group: = avg = max = 14.3 16 within = 0.4694 between = 0.2272 overall = 0.2260 corr(u_i, Xb) F(5,46) Prob > F = -0.2220 = = 11.47 0.0000 (Std Err adjusted for 47 clusters in id) imr Coef gini gdppc healthpc schoolenrolshare hi higdppc _cons 1088636 -.0015878 0004255 -.0863804 0012354 28.38582 sigma_u sigma_e rho 9.158136 2.5910165 92588862 Robust Std Err .0773407 000294 0003462 0335111 (omitted) 0003029 4.481255 t P>|t| [95% Conf Interval] 1.41 -5.40 1.23 -2.58 0.166 0.000 0.225 0.013 -.0468152 -.0021795 -.0002714 -.1538349 2645424 -.000996 0011223 -.018926 4.08 6.33 0.000 0.000 0006257 19.36552 001845 37.40612 (fraction of variance due to u_i) 85 xtreg imr gini gdppc healthpc schoolenrolshare hi higini, fe robust note: hi omitted because of collinearity Fixed-effects (within) regression Group variable: id Number of obs Number of groups = = 670 47 R-sq: Obs per group: = avg = max = 14.3 16 within = 0.3394 between = 0.6138 overall = 0.5862 corr(u_i, Xb) F(5,46) Prob > F = -0.6684 = = 7.16 0.0000 (Std Err adjusted for 47 clusters in id) imr Coef gini gdppc healthpc schoolenrolshare hi higini _cons 18973 -.000586 0010269 -.1394172 -.1600483 33.69399 sigma_u sigma_e rho 8.3634263 2.8909924 89326524 Robust Std Err .1162311 0001384 0003523 0441364 (omitted) 1273683 6.008265 t P>|t| [95% Conf Interval] 1.63 -4.23 2.91 -3.16 0.109 0.000 0.005 0.003 -.0442311 -.0008645 0003177 -.2282592 4236912 -.0003074 0017361 -.0505752 -1.26 5.61 0.215 0.000 -.4164274 21.59998 0963307 45.788 (fraction of variance due to u_i) xtivreg lifeexpc gini healthpc schoolenrolshare hi higdppc ( gdppc = tradeshare), re G2SLS random-effects IV regression Group variable: id Number of obs Number of groups = = 669 47 R-sq: Obs per group: = avg = max = 14.2 16 within = 0.6472 between = 0.6062 overall = 0.6122 corr(u_i, X) Wald chi2(6) Prob > chi2 = (assumed) lifeexpc Coef gdppc gini healthpc schoolenrolshare hi higdppc _cons 0004569 -.0255298 0006567 0172763 2.370363 -.0003062 66.77758 000098 0102326 0001371 0065158 3.88475 0000836 2.914548 sigma_u sigma_e rho 13.414869 75902815 99680879 (fraction of variance due to u_i) Instrumented: Instruments: Std Err z 4.66 -2.49 4.79 2.65 0.61 -3.66 22.91 P>|z| = = 0.000 0.013 0.000 0.008 0.542 0.000 0.000 1092.74 0.0000 [95% Conf Interval] 0002648 -.0455853 0003881 0045056 -5.243607 -.0004701 61.06517 0006489 -.0054743 0009254 030047 9.984333 -.0001424 72.48999 gdppc gini healthpc schoolenrolshare hi higdppc tradeshare 86 xtivreg lifeexpc gini healthpc schoolenrolshare hi higdppc ( gdppc = investmentshare > ), re G2SLS random-effects IV regression Group variable: id Number of obs Number of groups = = 669 47 R-sq: Obs per group: = avg = max = 14.2 16 within = 0.6472 between = 0.6338 overall = 0.6411 corr(u_i, X) Wald chi2(6) Prob > chi2 = (assumed) = = lifeexpc Coef gdppc gini healthpc schoolenrolshare hi higdppc _cons 0002339 -.0352637 0009258 0292089 1.019794 -.0001258 67.93498 0000591 0097259 0001061 0049781 1.023294 0000523 8524632 sigma_u sigma_e rho 2.852436 75856653 9339491 (fraction of variance due to u_i) Instrumented: Instruments: Std Err z 3.95 -3.63 8.73 5.87 1.00 -2.40 79.69 P>|z| 0.000 0.000 0.000 0.000 0.319 0.016 0.000 1092.95 0.0000 [95% Conf Interval] 000118 -.054326 0007179 019452 -.9858264 -.0002283 66.26419 0003498 -.0162014 0011337 0389658 3.025414 -.0000232 69.60578 gdppc gini healthpc schoolenrolshare hi higdppc investmentshare 87 xtivreg lifeexpc gini healthpc schoolenrolshare hi higini ( gdppc = tradeshare), re G2SLS random-effects IV regression Group variable: id Number of obs Number of groups = = 669 47 R-sq: Obs per group: = avg = max = 14.2 16 within = 0.6240 between = 0.5892 overall = 0.5938 corr(u_i, X) Wald chi2(6) Prob > chi2 = (assumed) = = lifeexpc Coef gdppc gini healthpc schoolenrolshare hi higini _cons 0001997 -.0424753 0005382 0309435 -2.291822 0203916 68.56537 0000378 0107565 0001878 0044326 1.332786 0240636 8828464 sigma_u sigma_e rho 3.1331971 78461682 94099011 (fraction of variance due to u_i) Instrumented: Instruments: Std Err z 5.28 -3.95 2.87 6.98 -1.72 0.85 77.66 P>|z| 0.000 0.000 0.004 0.000 0.086 0.397 0.000 1013.63 0.0000 [95% Conf Interval] 0001256 -.0635577 0001701 0222558 -4.904035 -.0267723 66.83502 0002739 -.0213929 0009064 0396312 3203915 0675555 70.29571 gdppc gini healthpc schoolenrolshare hi higini tradeshare 88 xtivreg lifeexpc gini healthpc schoolenrolshare hi higini ( gdppc = investmentshare) > , re G2SLS random-effects IV regression Group variable: id Number of obs Number of groups = = 669 47 R-sq: Obs per group: = avg = max = 14.2 16 within = 0.6247 between = 0.5930 overall = 0.5978 corr(u_i, X) Wald chi2(6) Prob > chi2 = (assumed) = = lifeexpc Coef gdppc gini healthpc schoolenrolshare hi higini _cons 0001922 -.0423132 0005713 0312309 -2.150685 019656 68.58198 0000453 0108411 0002234 0046607 1.3158 024167 8236333 sigma_u sigma_e rho 2.6941829 77998277 92266766 (fraction of variance due to u_i) Instrumented: Instruments: Std Err z P>|z| 4.24 -3.90 2.56 6.70 -1.63 0.81 83.27 0.000 0.000 0.011 0.000 0.102 0.416 0.000 1005.98 0.0000 [95% Conf Interval] 0001033 -.0635613 0001334 022096 -4.729605 -.0277104 66.96769 000281 -.0210652 0010092 0403657 4282354 0670224 70.19627 gdppc gini healthpc schoolenrolshare hi higini investmentshare xtreg imr gini gdppc healthpc schoolenrolshare r1#c.gdppc r2#c.gdppc r3#c.gdppc, fe > r Fixed-effects (within) regression Group variable: id Number of obs Number of groups = = 670 47 R-sq: Obs per group: = avg = max = 14.3 16 within = 0.3709 between = 0.4136 overall = 0.4018 corr(u_i, Xb) F(7,46) Prob > F = -0.5006 = = 6.29 0.0000 (Std Err adjusted for 47 clusters in id) Robust Std Err imr Coef t P>|t| [95% Conf Interval] gini gdppc healthpc schoolenrolshare 1508534 -.0012577 0009666 -.1265426 0820084 0007399 0004127 0413413 1.84 -1.70 2.34 -3.06 0.072 0.096 0.024 0.004 -.014221 -.0027471 0001359 -.2097584 3159277 0002318 0017973 -.0433269 r1#c.gdppc -.0001649 0009575 -0.17 0.864 -.0020922 0017624 r2#c.gdppc 0007247 0007328 0.99 0.328 -.0007503 0021998 r3#c.gdppc 0008051 0007184 1.12 0.268 -.0006411 0022512 _cons 31.91209 5.919215 5.39 0.000 19.99732 43.82685 sigma_u sigma_e rho 8.847045 2.8258722 90742005 (fraction of variance due to u_i) 89 ... effect of income inequality on life expectancy at birth and infant mortality rate in some of middle and high income countries in period of 1991- 2010 10 Estimate the effect of GDP per head in conjunction... conjunction of income inequality on health with consideration of differentiating high, upper middle and lower middle income countries Because the main objective of this research is to re-examine the... Because Gini construction in WIID is a composite calculated one from sources of surveys and works of income inequality, it contains some assumptions of concepts and methodologies A selection of criteria
- Xem thêm -

Xem thêm: Impact of income inequality on health from middle and high income countries in 1991 2010 , Impact of income inequality on health from middle and high income countries in 1991 2010

Gợi ý tài liệu liên quan cho bạn

Nhận lời giải ngay chưa đến 10 phút Đăng bài tập ngay