UNDP‘s Gender-Related Human Development Measures: Problems, Issues, and a Constructive Proposal Stephan Klasen Universität Göttingen Amie Gaye, HDRO HDRO Workshop March 4, 2013 1 Table 1: Categorizing Gender-Related Development Indices Gender-Sensitive Well-being Empowerment Hybrid UNDP’s GDI Welfare Loss of Gender inequality GDI/HDI, HDI-GDI UNDP’s GEM (?) UNDP’s GII Average Gender Gap Djikstra’s RSW, Klasen/Schüler‘s GGM, Permanyer‘s GRS1 Klasen/Schüler’s GEM3 Social Watch’s GEI, World Economic Forum’s GGI, Djikstra’s SIGE, UNECA’s AGDI 2 Current Situation • UNDP‘s GDI and GEM never very successful: – GDI often misinterpreted, problematic earned income component; cumulation of gaps in opposite directions; – GEM driven by income levels (not gender shares) and penalty for inequality complicated and intransparent (plus too few countries); • Alternative gender gap measures: – WEF, Social Watch, OECD‘s SIGI; – All rather complicated and intransparent; – Still room for a good Gender-related indicator; • 2010 abandons GDI/GEM and creates GII; – Switches concept to welfare loss of gender inequality; 3 UNDP‘s Gender Inequality Index • Measures welfare penalty due to gender inequality – 5 components: labor force participation, secondary education, teenage pregnancy maternal mortality, parliamentary seats; • Some serious problems: – Very complicated; – Intransparent (welfare penalty with respect to ‚equality index‘ which is not reported); implied ethical judgements of welfare penalty not transparent; – Hard to understand and interpret drivers of GII; – Mixes well-being and empowerment (well-being versus agency concerns); – Mixes achievements (in maternal mortality and teenage pregnancy) with gaps (in other components); Poor countries cannot do well on GII regardless of gender gaps! – No link to HDI; 4 Proposal • Replace GII by reformed GDI and GEM; – Keep well-being and empowerment concerns separate; • GDI (called GGM): Geometric mean of f/m ratios of achievements in life expectancy, education (years and school life expect.), and labor force participation; – Option to cap GGM at 1 (to focus on gaps affecting women negatively and reduce ‚compensation‘); – Classic gender gap measure; – Clear link to HDI; – Easy to measure and interpret; – Labor force participation problematic but better than earned incomes (or employment or unemployment); – Substantial change in rankings compared to GII (esp. Transition countries and Africa versus OECD and Middle East); 5 Country Russian Federation Lithuania Kazakhstan Latvia Ukraine Barbados Moldova (Republic of) Mongolia Finland Sweden Slovenia Norway Iceland Bulgaria Poland Armenia France United States Slovakia Israel Denmark Uruguay Australia Portugal Hungary GII GII Rank GGM GGM Rank Ratio LE Ratio Ed Ratio LF Capped GGM Rank 0.326 0.188 0.331 0.204 0.333 0.372 41 26 42 30 44 52 1.046 1.044 1.041 1.028 1.019 1.017 1 2 3 4 5 6 1.236 1.190 1.207 1.146 1.213 1.035 1.028 1.045 1.018 1.057 1.016 1.129 0.900 0.915 0.919 0.896 0.858 0.901 0.966 0.971 0.972 0.964 0.950 0.966 8 6 5 9 18 7 0.287 0.409 0.075 0.047 0.160 0.073 0.103 0.241 0.191 0.343 0.094 0.297 0.194 0.145 0.052 0.364 0.137 0.137 0.218 38 59 6 2 21 5 9 34 28 46 8 39 29 19 4 50 16 18 32 1.015 1.009 1.008 0.995 0.989 0.989 0.988 0.973 0.973 0.970 0.969 0.966 0.964 0.963 0.962 0.960 0.957 0.956 0.956 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 1.076 1.092 1.034 0.980 1.044 0.989 0.969 1.053 1.091 1.038 1.036 1.003 1.072 0.990 0.990 1.050 0.990 1.028 1.076 1.011 1.069 1.041 1.065 1.035 1.045 1.096 1.018 1.053 1.049 0.999 1.053 1.044 1.036 0.980 1.107 1.051 0.982 1.008 0.960 0.881 0.951 0.943 0.895 0.936 0.909 0.859 0.801 0.839 0.879 0.852 0.801 0.872 0.916 0.761 0.842 0.867 0.806 0.987 0.958 0.983 0.974 0.964 0.974 0.959 0.951 0.929 0.943 0.957 0.948 0.929 0.952 0.962 0.913 0.941 0.948 0.931 1 13 2 4 10 3 12 17 32 21 14 19 33 15 11 41 24 20 30 6 Country United Kingdom Viet Nam Croatia New Zealand Jamaica Thailand Belgium Rwanda Romania Netherlands Brazil Czech Republic Kyrgyzstan Switzerland Spain Argentina China Ireland Austria Luxembourg Namibia Cyprus Peru Ghana Venezuela (Bol. Rep.) Lao PDR Japan Tanzania Tajikistan Cambodia GII GII Rank GGM GGM Rank Ratio LE Ratio Ed Ratio LF Capped GGM Rank 0.215 0.297 0.170 0.190 0.452 0.357 0.107 0.446 0.339 0.046 0.455 31 40 23 27 70 48 12 66 45 1 71 0.955 0.952 0.951 0.949 0.948 0.947 0.946 0.942 0.941 0.938 0.936 26 27 28 29 30 31 32 33 34 35 36 0.983 0.974 1.051 0.979 1.006 1.047 1.013 0.884 1.061 0.979 1.054 1.053 0.962 0.996 1.023 1.068 0.977 1.009 0.925 1.005 0.981 1.037 0.842 0.921 0.821 0.853 0.793 0.832 0.829 1.023 0.782 0.859 0.751 0.939 0.952 0.935 0.942 0.926 0.933 0.939 0.942 0.921 0.938 0.909 26 16 28 23 34 29 25 22 37 27 43 0.137 0.374 0.050 0.110 0.375 0.183 0.175 0.103 0.168 0.468 0.118 0.394 0.523 17 53 3 14 54 25 24 10 22 74 15 57 86 0.936 0.936 0.930 0.930 0.930 0.924 0.924 0.923 0.910 0.910 0.905 0.905 0.904 37 38 39 40 41 42 43 44 45 46 47 48 49 1.031 1.104 0.993 1.027 1.064 0.960 0.994 1.007 1.004 0.865 0.985 1.006 0.899 1.017 1.023 0.929 1.012 1.080 0.934 1.019 0.926 0.956 1.036 0.915 0.932 0.827 0.782 0.727 0.872 0.774 0.699 0.879 0.778 0.843 0.786 0.841 0.822 0.790 0.995 0.921 0.899 0.930 0.918 0.887 0.924 0.918 0.921 0.909 0.899 0.905 0.903 0.904 36 51 31 39 53 35 40 38 42 50 44 46 45 0.452 0.478 0.107 0.603 0.355 0.448 69 77 13 98 47 68 0.904 0.903 0.901 0.900 0.899 0.899 50 51 52 53 54 55 1.024 0.933 1.041 0.867 1.049 0.919 1.104 0.781 0.965 0.863 0.936 0.896 0.653 1.010 0.729 0.975 0.741 0.884 0.868 0.903 0.889 0.900 0.885 0.899 60 47 52 48 55 49 7 Country Burundi Greece Italy Bolivia Philippines Korea, Rep. Kenya Guyana Lesotho El Salvador Cuba Panama Paraguay Chile Botswana Malawi Costa Rica Colombia Malaysia Belize Qatar Mexico Indonesia Honduras Sri Lanka Guatemala Malta Mauritania Mozambique Senegal Swaziland Benin Cameroon Togo Algeria GII 0.439 0.148 0.104 0.460 0.431 0.089 0.566 0.506 0.545 0.472 0.332 0.498 0.468 0.364 0.493 0.542 0.362 0.479 0.275 0.497 0.543 0.438 0.448 0.502 0.408 0.517 0.222 0.530 0.528 0.525 0.559 0.546 0.601 0.466 0.377 GII Rank 65 20 11 72 61 7 95 83 92 76 43 81 75 51 79 90 49 78 36 80 91 63 67 82 58 84 33 89 88 87 94 93 97 73 55 GGM 0.887 0.883 0.882 0.881 0.878 0.875 0.875 0.874 0.874 0.869 0.864 0.863 0.863 0.853 0.849 0.839 0.839 0.837 0.822 0.817 0.814 0.808 0.802 0.799 0.796 0.795 0.792 0.792 0.791 0.787 0.780 0.765 0.759 0.728 0.721 GGM Rank 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 Ratio LE 0.862 0.999 1.008 0.982 1.051 1.039 0.878 1.039 0.587 1.129 0.977 1.005 0.978 1.026 0.680 0.765 0.997 1.063 0.985 0.950 0.878 0.997 0.952 0.992 1.032 1.060 0.995 0.935 0.820 0.886 0.638 0.944 0.832 0.913 0.948 Ratio Ed 0.781 0.984 0.982 0.900 1.035 0.894 0.873 1.117 1.245 0.934 1.048 1.057 1.000 0.990 0.980 0.804 1.022 1.015 0.991 0.980 1.159 0.968 0.878 1.001 1.019 0.853 0.939 0.723 0.610 0.757 1.023 0.551 0.801 0.564 0.859 Ratio LF 1.036 0.701 0.693 0.773 0.623 0.721 0.873 0.576 0.914 0.622 0.631 0.605 0.657 0.610 0.918 0.960 0.580 0.543 0.569 0.585 0.530 0.547 0.618 0.513 0.479 0.556 0.532 0.735 0.990 0.726 0.728 0.862 0.657 0.748 0.460 Capped GGM 0.887 0.883 0.879 0.881 0.854 0.864 0.875 0.832 0.813 0.834 0.851 0.846 0.863 0.845 0.849 0.839 0.833 0.816 0.822 0.817 0.775 0.808 0.802 0.798 0.783 0.780 0.792 0.792 0.791 0.787 0.775 0.765 0.759 0.728 0.721 Rank 54 56 58 57 63 61 59 71 75 69 64 66 62 67 65 68 70 74 72 73 85 76 77 78 83 84 79 80 81 82 86 87 88 89 90 8 Gender Empowerment Measure • Measures inequality in economic and political participation and power; • Some Problems: – – – – Data availability; Focus on elites? Compensation issue? Serious problem with income component: genderinequality adjusted levels of incomes; levels, rather than gaps drive results! • Last problem can be corrected (using income shares rather than rates). 9 GEM • Same proposal as made in Klasen and Schüler (2011); – Use indicators of GEM (parliamentary representation, gaps in skilled employment, and earned incomes); – Use income shares instead of harmonic mean of income levels; – Straight-forward geometric mean of ratios; • Again one could cap at 1; – Rankings fundamentally different; – Issues: • Elite indicator? • Use post-secondary education? • Poor country coverage; 10 11 12 Conclusion • UNDP has not yet succeeded in producing a gender-related development indicator that is clear, transparent, easy to interpret, and linked to HD concept; • Alternatives are also weak, opportunity remains to propose a sound measure; • GII on net not an improvement over GDI/GEM; • Reformed GDI and GEM might be better way to go; • Would drastically change country rankings (in interesting ways); 13 Fixing the Gender Inequality Index: Strengths and weaknesses of new proposals Second Conference on Measuring Human Progress New York HD Measurement Conference: March 4-5, 2013 14 • The GII unique-- incorporates indicators of reproductive health which are a result of entrenched gender discrimination: – Females have no autonomy over their body – Reproductive health services not sensitive to adolescent needs At the same time the GII suffers some conceptual and empirical limitations • Indicators measuring the reproductive health dimension—maternal mortality ratio and adolescent fertility--no male equivalents • Hard to differentiate between gender inequality and poor overall conditions (poor countries cannot score high on GII); • Equality benchmark differ by dimension; • Labour force participation rate neither accounts for gender segmentation of the labour market nor gender wage gap nor opportunity to be employed • The functional form-too complex to allow for easy policy interpretation 15 HD Measurement Conference :: March4-5, 2013 Proposed GGM and GEM Strengths • Simple, transparent and easy to interpret • Distinguishes between the concept of well-being and empowerment • Dimensions of the GGM are closer to the HDI than the GII Weaknesses GGM – issue with LFPR remains (data and conceptual issues) – Important dimensions are missing e.g. gender based violence, care economy, reproductive health issues GEM 3 – not addressing some of the limitations of GEM 16 HD Measurement Conference :March 4-5, 2013 Issues for discussion – Are there ways to improve the GII, or is a new start required? – Is the separation well-being versus empowerment useful? – How do we address the limitations of the proposed measures GGM and GEM3? • Different indicators (esp. employment and empowerment, post-secondary education?, reproductive health?); – What to do about limited country coverage (esp. a problem for GEM)? – How to improve employment data? 17 HD Measurement Conference :March 4-5, 2013 18 19 20