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