Skip navigation links Sitemap | About us | FAQs

UN Programme on Disability   Working for full participation and equality
[_headerthemes.htm]
Theme: Statistics, Data and Evaluation, and Monitoring
Programme Monitoring and Evaluation; The Disability Perspective in the Context of Development

previousPrevious | Contents | Nextnext

IV. DEVELOPMENT OF "GLOBAL DISABILITY INDICATORS"

B.  Issues in the construction of indicators (1/2)

1. A case studies approach

As a country plans to develop long-range indicators, it is useful to look at historical data to discern important issues in their construction. In this section, data from three countries are explored to demonstrate some of these issues.[170] Criteria for choosing these countries for this case studies approach is not the quality of their data or the representativeness of the countries. Rather, countries are selected that have some data over time for persons with disabilities on socio-economic variables. Among the set of country experience, one, Canada is considered to be developed, and two others, Tunisia and Zambia, are developing countries. In each case, data are available on education and employment issues for at least two time points and data on prevalence rates for persons with disabilities are available from at least three points in time.

Such characteristics are promising for establishing indicators. However, a closer look at the data over time points to several potential problems in establishing indicators. Problems relate to changes in two broad areas - the population that is considered disabled and the measurement of the outcome variables. This section will explore issues related to the disabled population and then issues related to two potential outcome indicators - economic activity and education.

The stress on measurement of the disabled population may seem to contradict previous arguments in this paper that the focus should be placed on outcome variables. Unfortunately, however, if the population with disabilities is not defined and operationalized in the same manner over time, observed changes in outcome variables may be misleading. An initial focus on measurement of the disabled population is considered appropriate and provides as well important information related to the Programme objective of disability prevention.

2. Indicators of the prevalence of disability

a) The definition of disability employed

The purpose of this analysis is not to present definitive information on how countries should define disability. Rather, the point being stressed here is that changes in the definition over time have the potential to undermine the usefulness of indicators. Even if a country maintains either an Impairment or a Disability approach over time and maintains its data collection approach, there are many ways that definitions can be changed over time.

Both Zambia and Tunisia have employed the census approach to collect disability data over three points in time. They have tended to use an Impairment-based approach and have collected information in all three years on blind, deaf and/or mute and physical categories and usually on a mental category. However, as demonstrated in figure 7, even with these similarities, the categories changed for every census in Zambia and between the second and third census in Tunisia. For its final census, Tunisia collected data on the same broad groups, but expanded the category detail. For its second census, Zambia added the mentally retarded and a combination category.

Figure 7. DISABILITY CATEGORIES EMPLOYED IN THE LATEST THREE CENSUSES OF TUNISIA AND ZAMBIA

Country Census Categories Employed
Tunisia 1975 (1) Blind, (2) Deaf and Mute, (3) Motor Impairment, (4) Mental Handicap, and (5) Other
1984 (1) Blind, (2) Deaf and Mute, (3) Motor Impairment, (4) Mental Handicap, and (5) Other
1994 (1) Handicaps of Upper Body Limbs, (2) Handicaps of Lower Body Limbs, (3) Totally Handicapped, (4) Blind, (5) Mute, (6) Deaf, (7) Deaf - Mute, (8) Developmentally delayed, (9) Mentally Retarded, and (10) Undeclared
Zambia 1969 (1) Blind, (2) Deaf and/or Mute, (3) Loss of Limb, (4) Sick
1980 (1) Blind, (2) Deaf and/or Mute, (3) Crippled, or Loss of Limb, (4) Mentally Retarded, (5) Sick, (6) Combination of two or more categories, and (7) Not stated
1990 (1) Blind, (2) Deaf - Dumb, (3) Crippled, (4) Mentally Retarded, and (5) Multiple Disabilities

Sources: Republic of Zambia, Census of Population and Housing 1969, Final Report Volume I-Total Zambia (Lusaka, Central Statistical Office, 1973, pp. 11-14, Table 4 and p. 36, Table 18 and Tunisia, Institut national de la statistique, Recensement general de la population et des logements, 8 Mai 1975, Volume III, Caracteristiques demographiques, tableaux et analyses des resultats du sondage au 1/10eme75 (Tunis, Author, 1975) and Recensement general de la population et de l'habitat, 30 Mars 1984, Volume IV, Caracteristiques demographiques (Tunis, Author, 1984).

The most extreme change in approach is probably the dropping of the sick category by Zambia between its second and third census. If one only examines the total disability rate for Zambia over time, one would conclude that there has been a steady decline in the prevalence of disability between 1969 and 1990. While this may have been true between 1969 and 1980, as documented by figure 8, the observed decline between 1980 and 1990 was totally caused by the elimination of the category for sick in 1990. In 1969 and 1980, those counted as sick accounted for 50.8 per cent and 37.8 per cent of the reported population with disabilities. Even if one assumes some that: (a) the prevalence rates for sickness dropped between 1980 and 1990 in a similar manner to the drop between 1969 and 1980 and (b) some persons who may have been reported as sick were reported in other categories in 1990, the elimination of that reporting category has an effect on how the data are interpreted. If the trend between 1980 and 1990 is examined employing the 1990 disability reporting structure, the prevalence rate per 100,000 did not decline from 1,615 to 936 but actually increased from 610 to 936 per 100,000. Here is a case in which the categories reported as disabilities may actually influence the resulting trend.

Figure 8. PREVALENCE RATES FOR DISABILITIES, ZAMBIA,
1969, 1980 AND 1990 AND FOR SICK AND DISABLED
PERSONS, ZAMBIA, 1969 AND 1980

Bar chart. See source data below.

1969

1980

1990

Disabled

1.06 0.61 0.94

Sick

1.03 1.01 0

Sources: Republic of Zambia, Census of Population and Housing 1969, Final Report Volume I-Total Zambia (Lusaka, Central Statistical Office, 1973, pp. 11-14, Table 4 and p. 36, Table 18.

The focus here is not on whether Zambia should or should not have included the "sick" category as a reported disability category. The important point is that any change in this important decision has important ramifications. Zambia simply dropped the category, which made it possible to examine the changes in the other categories between 1980 and 1990 (assuming the other wording changes did not influence the resulting trend).

b) The changing disabled population

In the case of Canada, a Disability approach has been employed in three recent surveys over an eight year period. For example, the "Canadian Health and Disability Survey of 1983-84" counted children as disabled if they were reported as (a) using an aid or prostheses, (b) having a long-term health condition limiting activities, (c) attending a special school or class because of a condition, (d) having a specified impairment, (e) having vision or hearing trouble or (f) having a long-term health condition expected to last more than six months. The 1986 and 1991 Health and Activity Limitation Surveys employed similar definitions for children with disabilities.

From the three points in time for Canada, the data would suggest that the prevalence of disability is increasing. Confidence in that conclusion is increased by the consistency of the increase for both men and women, as shown in figure 9. Indeed, disability appears to be more prevalent in women, which, along with age, is strongly associated with the observed increase. In developed countries, disability rates increase with age, life expectancies are increasing and life expectancies for women are greater than for men. If disability rates increase with age and life expectancies are increasing, given no other change, the disability prevalence rate will increase over time. However, in Canada, the prevalence of disability is increasing within the age groups, as demonstrated by figure 10. Indeed, the rate is increasing among persons 65 years of age and over and for both men and women, as documented in figure 11. As more women enter the 65 years and over age group where women are more numerous than men and where female disability rates are higher than for males, overall disability prevalence will rise.

Figure 9. PREVALENCE RATES FOR DISABILITIES BY GENDER,
CANADA, 1983-84, 1986 AND 1991

Bar charts.  See data table below.

1983-84

1986 1991

Total

38.6 45.5 46.3

Males

37.9 43.7 43.4

Females

39.2 46.8 48.4

Sources: Statistics Canada, Health Division and Department of the Secretary of State, Social trends analysis directorate, 1986; Report of the Canadian health and disability survey, 1983-84 (Ottawa, Author, 1986).

Figure 10. PREVALENCE RATES FOR DISABILITIES BY AGE GROUP,
CANADA, 1983-84, 1986 AND 1991

Line charts.  See data tables below.

Age Group

1983-84

1986 1991

0 to 4

4.4 3.4 4.5

5 to 9

5.9 5.8 7.3

10 to 14

6.7 6.4 9.0

15 to 34

4.6 5.7 8.0

35 to 64

14.1 15.7 17.2

65 and over

38.6 45.5 46.3

Sources: Statistics Canada, Health Division and Department of the Secretary of State, Social trends analysis directorate, 1986; Report of the Canadian health and disability survey, 1983-84 (Ottawa, Author, 1986).

Figure 11. PREVALENCE RATES FOR DISABILITIES BY GENDER
FOR PERSONS 65 YEARS OF AGE AND OVER,
CANADA, 1983-84, 1986 AND 1991

Bar charts. See data table below.

1983-84 1986 1991
Total 38.6 45.5 46.3
Males 37.9 43.7 43.4
Females 39.2 46.8 48.4

Sources: Statistics Canada, Health Division and Department of the Secretary of State, Social trends analysis directorate, 1986; Report of the Canadian health and disability survey, 1983-84 (Ottawa, Author, 1986).

 

The trends that can cause the rise have been noted earlier, including an actual rise in disability, the increased life expectancies of persons with disabilities and the survival of persons who might have otherwise died and now survive with a disability. However, there are probably socio-cultural explanations for the rise. As disability advocates stress the human rights of persons with disabilities, persons may be more willing to report themselves as disabled. Note also, as shown in figure 10, the prevalence rates reported for children 10 to 14 years of age are higher than for persons 15 to 34. This suggests a social issue: children may be identified as disabled in the school situation but not in adult situations. Perhaps as learning disabilities gain more attention, persons may be more willing to identify themselves as disabled throughout their lives. Another important explanation of the rise in Canada for adults with disabilities may be found within the age categories. For example, just as the Canadian population overall is aging, there may also be a trend towards aging within some categories (i.e., as life expectancy increases, those 85 and over constitute a greater proportion of the population 65 years of age and over). This within-category aging would then contribute to an increase in disability for the 65 and over category.

As indicators are examined, caution is advised even in a time series of the quality of Canada's. For example, an examination of the gender composition of children over time raises some concerns. When gender ratios are calculated by dividing the number of boys by the number of girls and multiplying the results by 100, the results as given in figure 12 show a great deal of volatility over time within the five year age groupings. Not only does this volatility occur across time but also across cohort. As the children under 5 years of age grew older between 1986 and 1991, their gender ratio increased. In 1986, girls constituted the majority of disabled children under 5 but by 1996, boys constituted a majority of these children when they were 5 to 9. However, as the children who were 5 to 9 in 1986 became 10 to 14 in 1991, their gender ratio decreased. If indicators are at all tied to gender, then changes that occur over time could be due to changes in the sex composition of children with disabilities.

This problem is dramatized by the Tunisian data for adults 15 years of age and over, as well. Despite the relative similarity of categories, the number of women reported with disabilities in 1975 was extremely low. Between 1975 and 1994, while the number of disabled men reported more than doubled, the number of disabled women reported grew by many more times. As shown in figure 12, this resulted in the percentage of women in the disabled population tripling between 1975 and 1994. (Actually, the data demonstrate this change primarily occurred between 1975 and 1984.) There are many reasons to speculate as to whether disabled women were undercounted in 1975. Whether the change is real or a statistical artifact, it can clearly influence outcomes that are associated with gender.

Figure 12. SEX RATIOS FOR PERSONS WITH DISABILITIES
FOR CHILDREN UNDER 15 YEARS OF AGE,
CANADA, 1983-84, 1986 AND 1991

Age
Group

Year of Survey

1983-84 1986 1991
0 to 4 129.4 91.1 128.7
5 to 9 157.5 170.6 157.9
10 to 14 128.3 121.3 157.3

Note: Sex ratios are calculated by dividing the number of boys by the number of girls and multiplying the result by 100.

Sources: Statistics Canada, Health Division and Department of the Secretary of State, Social trends analysis directorate, 1986; report of the Canadian health and disability survey, 1983-84 (Ottawa, Author, 1986); Avard, Denise, Children and Youth with Disabilities in Canada: the 1986 health and activity limitation survey (Ottawa, Statistics Canada, 1996), Appendix A, p. 58, Table 2; and Children and Youth with Disability, Chapter 8 in The Health of Canada's Children: a CICH profile (Ottawa, Canadian Institute of Child Health, 1997), Table VIII.1.

Figure 13. PERCENTAGE MALE AND FEMALE AMONG DISABLED
PERSONS 15 YEARS OF AGE AND OVER,
TUNISIA, 1975 AND 1994

Two pie charts. See data table to the right.

  1975 1994
Females 2,660
10.4%
33,274
37.8%
Males 22,900
89.6%
54,736
62.2%

Sources: Tunisia, Institut national de la statistique, Recensement general de la population et des logements, 8 Mai 1975, Volume III, Caracteristiques demographiques, tableaux et analyses des resultats du sondage au 1/10eme75 (Tunis, Author, 1975).


Notes:

[170] See United Nations Disability Statistics Data Base …., Table A.1, "Countries or areas that collected national disability statistics …", pp 27-29.

previousPrevious : Nextnext


Home | Sitemap | About us | News | FAQs | Contact us

United Nations, 2003-04
Department of Economic and Social Affairs
Division for Social Policy and Development