Appendices for Posting on Web Only

Appendices for Posting on Web Only

Appendix 2 – Ordered Probit Regressions

Dependent variable is ordinal, categorical response to question: in general, how good would you say your health is? The possible answers are: very bad, bad, fair, good and very good.

Variable / 2003 / 2004 / 2005 / 2006
Aged >64 / -0.339 / -0.382 / -0.315 / -0.305
(0.053)*** / (0.038)*** / (0.036)*** / (0.036)***
Female / -0.055 / -0.001 / -0.013 / -0.036
(0.036) / (0.026) / (0.025) / (0.026)
Log income / 0.120 / 0.183 / 0.169 / 0.152
(0.027)*** / (0.023)*** / (0.022)*** / (0.022)***
Intermediate Ed / 0.405 / 0.311 / 0.229 / 0.290
(0.045)*** / (0.033)*** / (0.032)*** / (0.032)***
Leaving Ed / 0.545 / 0.474 / 0.450 / 0.482
(0.047)*** / (0.035)*** / (0.033)*** / (0.034)***
Third level / 0.506 / 0.456 / 0.448 / 0.462
(0.045)*** / (0.034)*** / (0.032)*** / (0.032)***
In Education / 0.273 / 0.347 / 0.343 / 0.377
(0.070)*** / (0.050)*** / (0.049)*** / (0.050)***
Unemployed / -0.396 / -0.389 / -0.302 / -0.362
(0.073)*** / (0.058)*** / (0.057)*** / (0.059)***
Retired / -0.437 / -0.423 / -0.522 / -0.537
(0.062)*** / (0.045)*** / (0.042)*** / (0.043)***
Home Duties / -0.337 / -0.342 / -0.404 / -0.332
(0.051)*** / (0.037)*** / (0.036)*** / (0.037)***
Other / -1.326 / -1.414 / -1.511 / -1.407
(0.064)*** / (0.049)*** / (0.046)*** / (0.048)***
Married / -0.166 / -0.096 / -0.071 / -0.133
(0.036)*** / (0.027)*** / (0.026)*** / (0.026)***
Separated / -0.328 / -0.467 / -0.386 / -0.352
(0.078)*** / (0.054)*** / (0.052)*** / (0.054)***
Widowed / -0.309 / -0.278 / -0.223 / -0.235
(0.061)*** / (0.046)*** / (0.043)*** / (0.043)***
N / 6099 / 10897 / 11932 / 11347

Standard errors in parentheses.

* significant at 10%

** significant at 5%

*** significant at 1%

Appendix 3: Sensitivity of Results to Different Poverty Lines

Poverty Measures Using Poverty line of 50% of Median Income/Health

Table 1: Summary of Self-Assessed Health (%)

Very Bad / Bad / Fair / Good / Very Good
2003 / 0.97 / 2.98 / 15.61 / 33.66 / 46.79
2004 / 1.04 / 3.17 / 15.38 / 36.15 / 44.27
2005 / 0.89 / 3.31 / 16.30 / 36.54 / 42.97
2006 / 0.82 / 3.18 / 17.13 / 38.23 / 40.64

Table 2: Income and Health Poverty Correlations, 2003-2006

Area under ROC Curve (SE in brackets) / Pearson
Corr
Coef / Spearman
Corr
Coef
2003
(N=6099) / 0.7176
(0.0101) / 0.3988 / 0.5456
2004
(N=11059) / 0.7194
(0.0075) / 0.4059 / 0.5831
2005
(N=11934) / 0.7133
(0.0073) / 0.3857 / 0.5901
2006
(N=11347) / 0.6712
(0.0091) / 0.3240 / 0.5496

Table 3: Unidimensional Pα Indices, 2003-2006 (SE in brackets)

Income / Health
α =0 / α =1 / α =2 / α =0 / α =1 / α =2
2003
(N=6099) / 0.115
(0.004) / 0.033
(0.002) / 0.017
(0.001) / 0.095
(0.004) / 0.020
(0.001) / 0.009
(0.001)
2004
(N=11059) / 0.106
(0.003) / 0.022
(0.001) / 0.010
(0.001) / 0.056
(0.002) / 0.016
(0.001) / 0.008
(0.001)
2005
(N=11934) / 0.099
(0.003) / 0.021
(0.001) / 0.010
(0.001) / 0.062
(0.002) / 0.022
(0.001) / 0.011
(0.001)
2006
(N=11347) / 0.082
(0.003) / 0.017
(0.001) / 0.007
(0.001) / 0.067
(0.002) / 0.021
(0.001) / 0.011
(0.001)

Table 4: Bidimensional Union and Intersection Headcount Ratios, 2003-2006

Intersection / Union
2003
(N=6099) / 0.025
(0.002) / 0.184
(0.005)
2004
(N=11059) / 0.017
(0.001) / 0.144
(0.001)
2005
(N=11934) / 0.017
(0.001) / 0.147
(0.003)
2006
(N=11347) / 0.013
(0.001) / 0.138
(0.003)

Table 5: Bourguignon-Chakravarty Indices, b=1 (equal weights on income and health poverty), 2003-2006 (SE in brackets)

θ =1 / θ =5
α=1 / α=2 / α=1 / α=2
2003
(N=6099) / 0.052
(0.002) / 0.028
(0.002) / 0.049
(0.002) / 0.024
(0.001)
2004
(N=11059) / 0.037
(0.001) / 0.017
(0.001) / 0.036
(0.001) / 0.015
(0.001)
2005
(N=11934) / 0.042
(0.001) / 0.021
(0.001) / 0.040
(0.001) / 0.018
(0.001)
2006
(N=11347) / 0.039
(0.001) / 0.020
(0.001) / 0.037
(0.001) / 0.018
(0.001)

Table 6: Bourguignon-Chakravarty Indices, b=0.33 (three times higher weight on income poverty), 2003-2006 (SE in brackets)

θ =1 / θ =5
α=1 / α=2 / α=1 / α=2
2003
(N=6099) / 0.038
(0.002) / 0.020
(0.001) / 0.037
(0.002) / 0.018
(0.001)
2004
(N=11059) / 0.026
(0.001) / 0.011
(0.001) / 0.025
(0.001) / 0.010
(0.001)
2005
(N=11934) / 0.026
(0.001) / 0.012
(0.001) / 0.025
(0.001) / 0.010
(0.000)
2006
(N=11347) / 0.023
(0.001) / 0.011
(0.001) / 0.022
(0.001) / 0.010
(0.001)

Table 7: Bourguignon-Chakravarty Indices, b=3 (three times higher weight on health poverty), 2003-2006 (SE in brackets)

θ =1 / θ =5
α=1 / α=2 / α=1 / α=2
2003
(N=6099) / 0.091
(0.004) / 0.049
(0.003) / 0.086
(0.004) / 0.042
(0.003)
2004
(N=11059) / 0.070
(0.003) / 0.034
(0.002) / 0.067
(0.003) / 0.031
(0.002)
2005
(N=11934) / 0.088
(0.003) / 0.046
(0.002) / 0.085
(0.003) / 0.042
(0.002)
2006
(N=11347) / 0.084
(0.002) / 0.044
(0.002) / 0.082
(0.003) / 0.041
(0.002)

Poverty Measures Using Poverty line of 70% of Median Income/Health

Table 1: Summary of Self-Assessed Health (%)

Table 2: Income and Health Poverty Correlations, 2003-2006

Area under ROC Curve (SE in brackets) / Pearson
Corr
Coef / Spearman
Corr
Coef
2003
(N=6099) / 0.8008
(0.0065) / 0.3988 / 0.5456
2004
(N=11059) / 0.8167
(0.0045) / 0.4059 / 0.5831
2005
(N=11934) / 0.8059
(0.0045) / 0.3857 / 0.5901
2006
(N=11347) / 0.7765
(0.0051) / 0.3240 / 0.5496

Table 3: Unidimensional Pα Indices, 2003-2006 (SE in brackets)

Income / Health
α =0 / α =1 / α =2 / α =0 / α =1 / α =2
2003
(N=6099) / 0.300
(0.005) / 0.083
(0.003) / 0.037
(0.002) / 0.236
(0.005) / 0.064
(0.002) / 0.025
(0.001)
2004
(N=11059) / 0.303
(0.003) / 0.075
(0.002) / 0.028
(0.001) / 0.210
(0.003) / 0.053
(0.001) / 0.020
(0.001)
2005
(N=11934) / 0.294
(0.004) / 0.071
(0.002) / 0.027
(0.001) / 0.222
(0.003) / 0.058
(0.001) / 0.024
(0.001)
2006
(N=11347) / 0.284
(0.004) / 0.064
(0.001) / 0.022
(0.001) / 0.229
(0.003) / 0.061
(0.001) / 0.025
(0.001)

Table 4: Bidimensional Union and Intersection Headcount Ratios, 2003-2006

Intersection / Union
2003
(N=6099) / 0.156
(0.005) / 0.375
(0.006)
2004
(N=11059) / 0.139
(0.003) / 0.370
(0.005)
2005
(N=11934) / 0.143
(0.003) / 0.372
(0.004)
2006
(N=11347) / 0.135
(0.003) / 0.378
(0.004)

Table 5: Bourguignon-Chakravarty Indices, b=1 (equal weights on income and health poverty), 2003-2006 (SE in brackets)

θ =1 / θ =5
α=1 / α=2 / α=1 / α=2
2003
(N=6099) / 0.144
(0.003) / 0.083
(0.003) / 0.120
(0.003) / 0.054
(0.002)
2004
(N=11059) / 0.125
(0.002) / 0.064
(0.002) / 0.103
(0.002) / 0.041
(0.001)
2005
(N=11934) / 0.129
(0.002) / 0.068
(0.002) / 0.107
(0.002) / 0.045
(0.001)
2006
(N=11347) / 0.126
(0.002) / 0.065
(0.002) / 0.106
(0.002) / 0.044
(0.001)

Table 6: Bourguignon-Chakravarty Indices, b=0.33 (three times higher weight on income poverty), 2003-2006 (SE in brackets)

θ =1 / θ =5
α=1 / α=2 / α=1 / α=2
2003
(N=6099) / 0.101
(0.002) / 0.055
(0.002) / 0.089
(0.002) / 0.040
(0.001)
2004
(N=11059) / 0.089
(0.002) / 0.042
(0.001) / 0.079
(0.001) / 0.030
(0.001)
2005
(N=11934) / 0.088
(0.001) / 0.043
(0.001) / 0.077
(0.001) / 0.030
(0.001)
2006
(N=11347) / 0.081
(0.001) / 0.039
(0.001) / 0.072
(0.001) / 0.027
(0.001)

Table 7: Bourguignon-Chakravarty Indices, b=3 (three times higher weight on health poverty), 2003-2006 (SE in brackets)

θ =1 / θ =5
α=1 / α=2 / α=1 / α=2
2003
(N=6099) / 0.270
(0.006) / 0.149
(0.005) / 0.235
(0.006) / 0.100
(0.003)
2004
(N=11059) / 0.228
(0.004) / 0.116
(0.003) / 0.199
(0.004) / 0.079
(0.002)
2005
(N=11934) / 0.247
(0.005) / 0.132
(0.004) / 0.219
(0.004) / 0.093
(0.003)
2006
(N=11347) / 0.253
(0.005) / 0.129
(0.004) / 0.228
(0.004) / 0.094
(0.003)

Appendix 4: ROC Curves and Kernel Densities of Health and Income Measures

ROC Curve 2003

ROC Curve 2004

ROC Curve 2005

ROC Curve 2006

Appendix 5: Bi-Dimensional Poverty Indices

Appendix 6: Kernel Densities

Kernel Density for Log Equivalised Income, 2003

Kernel Density for Log Equivalised Income, 2004

Kernel Density for Log Equivalised Income, 2005

Kernel Density for Log Equivalised Income, 2006

Kernel Density for Scaled Health, 2003

Kernel Density for Scaled Health, 2004

Kernel Density for Scaled Health, 2005

Kernel Density for Scaled Health, 2006