Guide on Gender Analysis of Census Data Full Draft of 6 December 2012 Contents


Table 42: Liberia (2008) – Causes of disability by sex



Download 3.92 Mb.
Page26/32
Date09.08.2018
Size3.92 Mb.
1   ...   22   23   24   25   26   27   28   29   ...   32

Table 42: Liberia (2008) – Causes of disability by sex
Males Females
From birth 4,708 4,174

Polio 2,288 1,878

Stroke 1,162 982

Epilepsy 1,195 1,125

War 7,634 6,131

Occupational injury 2,633 1,407

Transportation accident 1,731 923

Other types of accident 4,116 2,303

Ageing process 4,224 5,675

Other causes 11,147 12,476

Other diseases 15,824 16,624
Total 56,562 63,698
541. What this shows is that overall disability numbers are slightly higher for women than for men, but that this difference is entirely accounted for by the last three causes of disability. The greater incidence of disabilities associated with ageing is due, to some extent, to the larger number of elderly women in the population, but this does not explain everything. The female population over age 60 is 6.1 per cent larger than the equivalent male population, but the number of disabilities related to ageing in women is 34.4 per cent larger. Women, therefore, do appear to suffer disproportionally from disabilities associated with old age. Strangely, a category of causes of disability not included here is that of disabilities associated to childbirth. That this category can be important is shown by Stubbs and Tawahke (2009), in their study on Samoa, which shows that 7.7 per cent of women with disabilities acquired their disability through child birth complications. Another important category from a gender perspective is disability caused by spousal violence. This is one of the 7 categories (congenital/prenatal, disease/illness, injury/accident, spousal violence, other violence, unknown, other) included in the 2010 census of Zambia.

542. Investigating a specific topic or factor, such as education, may serve as a useful indicator of disability and gender issues. For instance, Possi’s (1996) study of Tanzania indicates that while the gender parity in education is not an issue at the primary school level, the number of enrolled girls with disabilities decreases as the number of years in secondary school increases, similarly to the pattern for non-disabled girls in primary and secondary schools in Tanzania. Further, very few women with disabilities have a college or university education. (This analysis is relatively meaningless if it doesn’t make a comparison with the statistics for boys).


5. Indicators
543. Using data from subsequent censuses can provide another gendered measure of disability status. Using census data from Ireland, it was found that there were more males than females living with a disability in 2000 (among the disabled population enumerated in the 2000 census, 20,576 were males and 20,214 were females; the sex-ratio was 101.8). Nevertheless, although the sex ratios indicated more males than females among the disabled population (both in 1990 and 2000), the decline in the sex ratio in 2000 denoted that the female disabled population has been growing at a faster rate than the corresponding male population. In order to correctly interpret this trend, however, one would need to separate the components of this increase that are due to the actual increase of the age-specific prevalence of disabilities from the age effect, which is due to the fact that the female population is ageing faster than the male population. A solution to this type of problem is by using age standardization (see the Methodology Box below).
Methodology Box 9: Age Standardization
The 2011 census of Montenegro found that 54 per cent of the people in the country living with a disability were women and 46 per cent were men (UNECE, 2012 b). If the objective of computing this statistic is to know whether the care for persons with disabilities should be organized predominantly to attend to female or to male patients, it is an appropriate indicator. However, if the objective is to establish whether women are more or less prone to suffer from disabilities than men, it is flawed by two intervening factors, namely:

1. There are more women than men in the population; and



2. The excess of women over men is concentrated in the oldest ages, where disabilities are most common.
To take care of the first problem, one may compute a different statistic, namely the percent-age of men and women that suffer from disabilities. Unlike the previous indicator, this one is not affected by the total number of men and women. The result is 11.7 per cent for women, compared to 10.2 per cent for men, which still suggests a higher incidence of disabilities among women. However, when one age-standardizes this difference, i.e. when one computes the percentages by age and then applies them to the same age distribution (in this case, the age distribution for both sexes combined), the difference disappears and both percentages become 11.0 per cent. The incidence of disabilities in men is higher early in life, whereas for women it is higher at older ages, but given the age distribution for both sexes combined, the overall incidence is about the same.
The following example from the 2010 census of Mexico illustrates how to carry out the various steps to standardize the age structure.





Population

Percent Disabled

Number Disabled

Idem Standardized

Male

Female

Total

Male

Female

Male

Female

Male

Female

0-4

5,346,943

5,181,379

10,528,322

0.87

0.70

46,259

36,323

91,086

73,807

5-9

5,604,175

5,443,362

11,047,537

1.87

1.35

104,718

73,380

206,431

148,928

10-14

5,547,613

5,392,324

10,939,937

2.00

1.59

110,882

85,523

218,660

173,509

15-19

5,520,121

5,505,991

11,026,112

1.81

1.43

99,954

78,713

199,652

157,628

20-24

4,813,204

5,079,067

9,892,271

1.89

1.37

91,053

69,806

187,135

135,958

25-29

4,205,975

4,582,202

8,788,177

2.08

1.42

87,530

65,043

182,890

124,746

30-34

4,026,031

4,444,767

8,470,798

2.45

1.67

98,726

74,296

207,720

141,593

35-39

3,964,738

4,328,249

8,292,987

2.78

1.97

110,259

85,073

230,627

163,001

40-44

3,350,322

3,658,904

7,009,226

3.59

2.85

120,248

104,229

251,571

199,668

45-49

2,824,364

3,104,366

5,928,730

4.77

4.14

134,690

128,499

282,733

245,408

50-54

2,402,451

2,661,840

5,064,291

6.40

6.04

153,815

160,884

324,237

306,090

55-59

1,869,537

2,025,828

3,895,365

8.50

8.30

158,881

168,187

331,044

323,399

60-64

1,476,667

1,639,799

3,116,466

11.44

11.49

168,883

188,340

356,423

357,943

65-69

1,095,273

1,221,992

2,317,265

15.32

15.56

167,839

190,138

355,096

360,559

70-74

873,893

1,000,041

1,873,934

20.64

21.29

180,328

212,863

386,687

398,875

75-79

579,689

665,794

1,245,483

27.54

28.39

159,620

189,031

342,949

353,615

80-84

355,277

443,659

798,936

35.39

36.63

125,739

162,497

282,758

292,623

85+

298,739

404,556

703,295

46.01

48.40

137,461

195,801

323,612

340,388

Total

54155012

56784120

110939132

4.17

4.00

2256885

2268626

4761313

4297736

Source: INEGI, Tabulations from the 2010 census


The percentage of persons with disabilities who are women is 2,268,626 / (2,256,885 + 2,268,626) = 50.1 per cent. The percentage of persons who have a disability is 100 * 2,268,626 / 56,784,120 = 4.00 per cent for women, compared to 100 * 2,256,885 / 54,155,012 = 4.17 per cent for men. The age standardization consists in applying the percentage of persons with a disability, separated by sex, not to the corresponding male or female population, but to a common population, in this case made up of all individuals, of both sexes. This yields the hypothetical results displayed in the last two columns, which vary between the sexes because of the different proportions of disabilities among men and women, but not because of the different numbers of men in the base population. The age standardized percentage of women with disabilities is now 100 * 4,297,736 / (54,155,012 + 56,784,120) = 3.87 per cent, compared to 100 * 4,297,736 / (54,155,012 + 56,784,120) = 4.29 per cent for men. Again, the male rates tend to be higher until age 60, whereas the female rates are higher at older ages.

544. In the case of Nicaragua, where the overall percentage of persons with disabilities is 9.1 for men and 11.3 for women (Mont, 2007), male disability rates are higher than female rates until adulthood; men and women are similar until age 39; after that, the rates diverge to a 10 percentage point differential in favour of men. This suggests that there is something about events during the life course that differentiate the experiences of women and men in how and when they acquire disabilities. The research by Stubbs and Tawake (2009) in Somoa showcased just above might suggest that this mid-life higher propensity to develop a disability for women could be associated with child birth or lower income. In order to understand the Nicaraguan case better, the researcher may tabulate having a disability by sex, age group, and poverty status.


545. The following table provides unstandardized and standardized indicators from the 2010 census of Mexico, separated by type of disability.
Table 43: Mexico (2010) – Unstandardized and standardized prevalence of selected types of disabilities


Type of disability

Percentage Female

Prevalence

Standardized

Male

Female

Male

Female

Walking or moving

53.3

2.10

2.29

2.19

2.20

Seeing

52.2

1.14

1.19

1.18

1.15

Hearing

45.2

0.50

0.40

0.53

0.38

Speaking or communicating

43.0

0.42

0.30

0.42

0.30

Personal care

52.6

0.20

0.21

0.21

0.20

Paying attention or learning

45.9

0.21

0.17

0.21

0.17

Mental disabilities

43.8

0.47

0.35

0.47

0.34

Total

50.1

4.17

4.00

4.29

3.87

546. Disability-free life expectancy. For addressing the interrelationships of ageing, gender and disability, the disability-free life expectancy measure may be useful. This concept provides an indicator of elderly persons’ health condition in order to help plan adequate services and facilities. The method to calculate this disability-free life expectancy was first presented in a report of the US Department of Health Education and Welfare (Sullivan, 1971) and is often referred to as ‘Sullivan Health Expectancy method’. The Sullivan method makes use of a life table and the age-specific proportions of persons with a disability. The age-specific proportions of persons with a disability are multiplied by the corresponding number of person years (Lx) lived between ages x and x + n in the life table. On the basis of these calculated Lx values, the total after lifetime (Tx) in the disabled state can be calculated. By dividing these age-specific Tx values by the total life table survivors (lx), one obtains the life expectancy in the disabled (and non-disabled) state.


Table 44: Life table for Aruba 2010-2011 (males and females) life with and without disability


Males

 

 

 

 

 

% Dis-

e(x) Not

e(x)

% e(x)

Age

l(x)

D(x,n)

L(x,n)

S(x,n)

T(x)

e(x)

abled

disabled

Disabled

Disabled

0

100000

710

99354

0.99290

7387957

73.9

0.4

67.9

6.0

8.1

1

99290

0

397160

0.99987

7288603

73.4

0.4

67.4

6.0

8.2

5

99290

0

496450

0.99934

6891443

69.4

2.2

63.4

6.0

8.7

10

99290

131

496123

0.99868

6394994

64.4

2.3

58.5

5.9

9.2

15

99159

131

495469

0.99569

5898871

59.5

2.5

53.7

5.8

9.7

20

99028

724

493332

0.99148

5403402

54.6

2.8

48.9

5.7

10.4

25

98305

957

489131

0.99084

4910069

49.9

3.0

44.4

5.6

11.2

30

97348

834

484652

0.98936

4420939

45.4

2.6

39.9

5.5

12.1

35

96513

1228

479494

0.98715

3936287

40.8

3.9

35.4

5.4

13.2

40

95285

1236

473332

0.98499

3456793

36.3

3.7

31.0

5.3

14.5

45

94048

1606

466226

0.97506

2983461

31.7

5.6

26.6

5.2

16.3

50

92442

3046

454597

0.96145

2517234

27.2

7.4

22.3

5.0

18.2

55

89396

3964

437072

0.94223

2062637

23.1

9.6

18.3

4.8

20.6

60

85432

6136

411823

0.89999

1625566

19.0

10.5

14.5

4.5

23.6

65

79297

10339

370636

0.84687

1213742

15.3

16.6

11.0

4.3

28.0

70

68958

12364

313879

0.76874

843106

12.2

21.1

8.2

4.0

33.0

75

56594

16671

241291

0.67552

529227

9.4

30.5

5.6

3.8

40.1

80

39923

14647

162997

0.55034

287936

7.2

39.2

3.7

3.5

48.1

85

25276

14671

89703

0.34571

124939

4.9

56.2

2.0

3.0

59.8

90

10605

8806

31012

0.13622

35236

3.3

67.9

1.0

2.3

68.9

95

1799

1799

4224

 

4224

2.3

76.9

0.5

1.8

76.9




Females

 

 

 

 

 

% Dis-

e(x) Not

e(x)

% e(x)

Age

l(x)

D(x,n)

L(x,n)

S(x,n)

T(x)

e(x)

abled

disabled

Disabled

Disabled

0

100000

1591

98552

0.98259

7975199

79.8

0.2

70.8

8.9

11.2

1

98409

150

393276

0.99892

7876647

80.0

0.3

71.0

9.0

11.3

5

98259

0

491296

1.00000

7483371

76.2

1.5

67.1

9.0

11.9

10

98259

0

491296

0.99929

6992075

71.2

1.8

62.2

9.0

12.6

15

98259

139

490948

0.99831

6500779

66.2

2.8

57.3

8.9

13.4

20

98120

192

490120

0.99817

6009831

61.2

2.6

52.5

8.8

14.3

25

97928

167

489222

0.99772

5519712

56.4

2.7

47.7

8.6

15.3

30

97761

278

488108

0.99481

5030490

51.5

3.1

42.9

8.5

16.6

35

97482

735

485573

0.99230

4542382

46.6

3.6

38.2

8.4

18.0

40

96747

761

481834

0.99200

4056809

41.9

4.6

33.7

8.3

19.7

45

95987

782

477978

0.98990

3574975

37.2

5.9

29.1

8.1

21.8

50

95205

1149

473151

0.97807

3096997

32.5

7.6

24.7

7.9

24.2

55

94056

3002

462773

0.96220

2623846

27.9

10.4

20.3

7.6

27.2

60

91053

3994

445282

0.94484

2161073

23.7

11.7

16.4

7.3

30.8

65

87059

5831

420719

0.91657

1715791

19.7

18.4

12.7

7.0

35.8

70

81228

8210

385617

0.86853

1295072

15.9

23.9

9.3

6.6

41.4

75

73019

12069

334921

0.79378

909454

12.5

33.4

6.4

6.1

48.8

80

60950

15558

265855

0.67355

574533

9.4

46.7

4.0

5.5

57.8

85

45392

19158

179066

0.50989

308678

6.8

63.9

2.2

4.6

67.4

90

26234

15946

91305

0.41956

129612

4.9

69.4

1.4

3.6

72.2

95

10288

10288

38308

 

38308

3.7

78.9

0.8

2.9

78.9

Source: Population and Housing Census Aruba 2010; Helder (2012)




Figure 13: Percentage of life expectancy at age (x) spent with at least one disability, by sex

547. Table 44 shows an example of the Sullivan method for Aruba. For each age-category, the percentage of total remaining life expectancy with a disability was calculated. Differences between males and females in the percentage of remaining life spent with a disability are depicted in Figure 13. Life expectancy for men on Aruba at age 60 is 19.0 years. At this age, an average man can expect to live 14.5 years disability-free and 4.5 with at least one disability. Women live longer. Their life expectancy at age 60 is 23.7 years, which is 4.7 years more than men. Of these years, they spend 16.4 years in the disability-free state and 7.3 years in the disabled state. Compared to men, most of the extra years women live are spent in the disabled state. At age 60, a man can expect to live with a disability for 23.6 per cent of his remaining years, for a woman this is 30.8 per cent.
548. A study from Thailand (Jitapunkul et al., 2003) showed similar results. Although women had a longer life expectancy than men, they spent more years in the disabled state. At age 60, women had a life expectancy of 23.9 years, and on average could expect to spend 18.2 years free from long-term disability, leaving 5.7 years (or 24 per cent of their remaining life expectancy) of years lived a disability. By contrast, men aged 60 had a remaining life expectancy of 20.3 years and a disability-free life expectancy of 16.4 years, resulting in 3.9 years (or 19 per cent of remaining life expectancy) spent in a disabled state. Women, therefore, had more years to live, both in disabled and in disability-free states. Similar conclusions were obtained in a study from the City of São Paulo, Brazil (Camargos, Perpétuo and Machado, 2005). In 2000, 60-year-old men could expect to live, on average, 17.6 years, of which 14.6 years (83%) would be free of functional disability. Women of the same age could expect to live 22.2 years, of which 16.4 years (74%) would be free of functional disability. Men would have a functional disability and be dependent on others for 1.6 years (9%), while the comparable period for women would be 2.5 years (11%).
549. Using the Sullivan method in censuses, one can at least make some predictions about gender differences in the incidence of disability on the basis of prevalence data, provided that one can control for the duration of disability. One potential solution is to compute what percentage of their remaining life men and women are expected to live with disabilities. In the example of Aruba in Table 44, this is 30.8 per cent for 60 year old women and 28.9 per cent for 60 year old men. But this result is still biased by the fact that women, because they live longer, survive to more advanced ages, where disabilities are more common. A better alternative is therefore to standardize using the same life table (i.e. the one that characterizes the mortality experience for both sexes combined) for both men and women. In this way one can see how the results would change if men and women survived equally. Based on the female life table, women aged 20 should expect to spend 8.8 years of their remaining life with a disability, compared to 5.7 for men (see Table 44). But when the life table for both sexes is used, this result changes to 7.5 years for women, compared to 6.5 for men. This implies that, although the prevalence of disabilities beyond age 20 is indeed higher for women than for men, a substantial part of the difference is also accounted for by the fact that women live longer and are therefore more likely to survive to higher ages, where disabilities are very common.
550. The Sullivan method provides an insight into gender differences in levels of disability. However, there are also some shortcomings in using this method. There are at least two methodological problems linked to this approach. The first one is that the results for both sexes may be biased because persons with disabilities may have higher mortality than the general population. This means that the number of years lived with a disability will be over-estimated in an approach that assumes the same life table for those with and without disability.
551. The second shortcoming has to do with the nature of the data used in the Sullivan method. In general, each health condition can be described in terms of its prevalence and incidence. Incidence refers to new cases of a health condition in a given period and prevalence to the number of existing cases at a certain point in time. The Sullivan method uses data on prevalence of disability, but to answer the question whether women are more or less prone to disabilities than men would also require data on the incidence of disability. These data are not available in censuses, as it would require a question on the timing when the respondent became disabled. However, prevalence and incidence are closely linked. One of the basic formulas in epidemiology states that prevalence = incidence x duration(avg). The study of Oman et al. (1999) sheds some light on the relationship between prevalence and incidence of disability. In their cohort study of 2,025 residents 55 years of age and older in Marin County, California, they found that the incidence rates for lower body physical disability were not significantly different between men and women. However, age-specific and age-adjusted prevalence rates were consistently higher among women. They attributed this difference to the longer duration women live with a disability, due to the lower recovery and mortality rates among females vis à vis males.

6. Multivariate and further gender analyses
MAYBE SOME OF THE STUFF THAT IS NOW IN INDICATORS COULD ALSO BE PUT HERE.
552. Persons with disabilities generally run a greater risk of being excluded from formal education. The following example indicates how in Vanuatu children and adolescents between the ages of 5 and 20 with a disability are less likely to attend education than others of their age. Table 45 shows the results of a logistic regression in which the dependent variable was whether a child was going to school (full time or part time – value 0) or not (i.e. left school or never attended – value 1). Answers to the four questions on disability used in the population census were included, namely:

Does this person have difficulty in:

a) Seeing, even wearing glasses ?

b) Hearing, even if using a hearing aid ?

c) Walking or climbing steps ?

d) Remembering or concentrating ?



In addition to these four disability conditions, sex, age, rural/urban residence, and citizenship were included as control variables.
553. The results of the logistic regression clearly show that, for each of the four disability variables, children with some difficulties have higher probabilities of not attending than those without difficulties. Children who cannot hear, walk or remember/concentrate at all score much higher than those who have some difficulty. For instance, a child who cannot walk has more than 6 times higher odds of not attending school. A child with some difficulty has about 90 per cent higher chance to remain without schooling. Somewhat unexpectedly, the effect is less pronounced (but present) for children who are visually impaired. Finally, girls have a slightly higher probability of not attending school than boys, but the difference is only 3 per cent. This is more or less the same as the difference in the general population, without controling for intervening factors in a multivariate analysis: among all boys aged 5-19, 72.2 per cent attend school against 71.5 per cent of girls. Note, however, that the education of the mother has a profound effect on the chance of a child to attend school. The odds of children to be in school, with a mother who has more than primary education is 2.5 times as high (1/.406) than among children whose mother has less than primary education.
Table 45: Vanuatu (2009) - Logistic regression of school attendance by children and adolescents aged 5-20, by type of disability and other explanatory variables


Variable

Category

B

exp(B)

Sex

Male










Female

0.032

1.033

Age

5 - 7 yrs.










8 - 10 yrs.

-0.755

0.470




11 - 13 yrs.

-0.374

0.688




14 - 16 yrs.

0.894

2.445




17 - 19 yrs.

2.420

11.248

Urban/Rural

Urban










Rural

0.358

1.431

Difficulty seeing

No difficulty at all










Some difficulties

0.361

1.435




Cannot do at all

0.528

1.696

Difficulty hearing

No difficulty at all










Some difficulties

0.170

1.185




Cannot do at all

1.256

3.512

Difficulty walking

No difficulty at all










Some difficulties

0.657

1.929




Cannot do at all

1.867

6.469

Difficulty remembering

No difficulty at all










Some difficulties

0.420

1.522




Cannot do at all

1.319

3.739

Citizenship

Vanuatu by birth










Vanuatu by naturalisation

-0.095

0.909




Other countires

-0.287

0.751

Education mother

Less than primary










Primary

-0.506

0.603




More than primary

-0.902

0.406

Constant

 

-1.274

0.280

Source: Population and Housing Census of Vanuatu (2009)


554. Noteworthy studies that utilize census data to profile persons with disabilities within their populations are Zambia and Israel. These studies can be viewed online at the following URLs: Zambia (http://www.statssa.gov.za/census01/html/Disability.pdf), and Israel (http://www.cdc.gov/nchs/washington_group.htm). This latter study was discussed in the 10th Washington Group Meeting in Luxembourg in 2010.
7. Interpretation, Policy and Advocacy
555. Negative stereotypes about persons with disabilities may result in lowered expectations of their abilities and social policies that do not allow them to realize their full potential. Negative images and portrayals of persons with disabilities, especially older women with disabilities, influence society’s view of them, and consequently, their ability to integrate and participate in society. Negative stereotypes also increase their vulnerability to abuse and discrimination.
556. According to Rousso (2003), boys may have the advantage in obtaining assistive devices and other rehabilitation services needed to get to and participate at school. Women receive only one fifth of the rehabilitation in the world and, particularly in developing countries, men have greater access to rehabilitation services and to prosthetic and orthotic devices than women. Gender bias in access to rehabilitative services and devices is in itself a barrier to education for girls with disabilities. He recommends that more reliable data should come from increased research on such basics as the number of girls with disabilities who are of school age, their school enrollment levels, and their educational outcomes. This requires developing a consistent definition of disability as well as disaggregating data on children who are disabled by sex, and disaggregating data on girls by disability status.
557. Country examples for successful advocacy for data collection include an NGO-alliance led by North India Cerebral Palsy Association that campaigned in Punjab, India for women and men not to hide their disabilities during the census.


Download 3.92 Mb.

Share with your friends:
1   ...   22   23   24   25   26   27   28   29   ...   32




The database is protected by copyright ©sckool.org 2020
send message

    Main page