|Theme: Statistics, Data and Evaluation,
Programme Monitoring and Evaluation; The Disability Perspective in the Context of Development
III. OPTIONS FOR IMPROVING PROGRAMME MONITORING TRENDS IN POLICIES AND PROGRAMMES FROM THE DISABILITY PERSPECTIVE
C. Issues in monitoring
1. Basic considerations in planning monitoring activities
The use of information on disability and rehabilitation for planning and evaluation involves obtaining information for initial planning, as well as for monitoring and evaluation. For initial planning, information is used to analyze the current situation, set objectives and plan activities. For programme evaluation, information can be analyzed for different components, such as effectiveness, progress, impact, efficiency and relevance.
The basic goal of monitoring activities is to evaluate the progress being achieved towards Programme goals and objectives. In relation to this determination, as discussed earlier, the Long-term Strategy urges the establishment of concrete targets and measures of those targets. Data may provide the measures, but targets would be driven by Programme goals. Thus, targets must be established first and then the measures of target achievement can be derived.
In making decisions in regard to disability monitoring, the Statistics Division has noted that four basic trends should be considered. These are:
When these trends are considered, three questions in regard to information for monitoring assume critical importance:
Once the characteristics of the information desired for monitoring are determined, the Statistics Division suggests the following points be considered when making choices related to gathering the information:
Applying these suggestions specifically to establish a monitoring programme to assess progress towards Programme goals, introduces four additional considerations. First, while comprehensive monitoring of all environmental facets as they contribute to promoting or hindering prevention, rehabilitation or equalization of opportunity is critical, information on the environment has been systematically collected in few countries. Likewise, data have not been routinely collected on those life areas where the environment prevents equalization of opportunity, such as independence, use of time, social integration, economic self-sufficiency and transition. Third, disability data collection efforts are constrained by resources. Finally, the fact that some data collection efforts in resource-scarce environments are successful implies that priorities should be established for data collection systems.
The first two issues are quite different from the last two, however. Indeed, the last two often hinder the establishment of evaluation criteria for the first two. Resource constraints combine with the fact that some data can be collected quite separately to support prioritization away from the consideration of monitoring the environment and life areas where environmental constraints Handicap people with disabilities.
All too often, information collected on disability has reflected issues where the data are considered to be reliable and not those areas, which may be critical but data are considered unreliable. Often, the criteria for determining reliability reflect a social welfare, rather than a social development perspective, because data on prevention and rehabilitation have been considered to be more accurate than data relevant to equalization of opportunities. For example, data on the functions of the body related to Impairment may be considered to be objective, whereas data related to whether people are independent because they have access to choice is viewed as culturally bound. Thus, given resource constraints, it is easier for all concerned not to develop evaluation criteria for independence. As such, the collection of such information has appeared to confirm the social welfare viewpoint, while neglecting issues that need to be considered to promote social change. It is critical to separate the priorities for collecting data from the priorities for social policy. Rather, as countries establish human rights and universal design policies, as well as programmes that empower disabled persons as consumers, these polices should drive the decisions on evaluation criteria for monitoring.
Data considerations and resource constraints should be separated from the establishment of evaluation criteria. Once the criteria are established, then the resource constraints can be recognized and priorities established. Such an approach then clearly identifies the situation and allows for advocates to press for resources to effectively monitor progress towards Programme goals - one that allows for consideration of the environment and critical life areas.
2. Establishment of evaluation criteria
a) Identifying monitoring dimensions
As data considerations and resource constraints need to be distinguished from evaluation criteria, so, too, the different dimensions along which criteria are to be elaborated need to be clearly elaborated, as well. This is not merely the identification of variables, such as housing or income, but, rather, the delineation of broad conceptual planes critical for achievement of Programme goals. The ICIDH framework attempted to do this in distinguishing Impairment, Disability and Handicap but has been criticized for not giving proper focus to the non-medical, social dimensions of disability. This criticism points out the reason why different conceptual planes should be identified - to be certain that the monitoring programme takes into account as many differing perspectives as feasible. Such a broad base increases the likelihood (but does not guarantee) that the monitoring programme will reflect the full range of progress or lack thereof toward Programme goals and objectives. These changes increase the likelihood that information obtained will reflect as many different types of variables as possible. Six potential broad dimensions for monitoring are: (a) target areas for outcomes, (b) environmental variables, (c) disability-specific information, (d) target areas for accessibility, (e) target areas for opportunity and (f) programme outputs and inputs.
While these dimensions will be discussed shortly, a distinction between outcomes and outputs may prove useful here. This distinction relates to the distinction noted earlier between measuring the situation of disabled persons as opposed to evaluation of disability programmes. Outcomes represent those results desired by programme activities in terms of positive changes in the lives of its intended beneficiaries. These results constitute the reasons for which a programme is established and should closely relate to the ultimate goals and shorter-term objectives of those who are served. Outputs are the direct results of a programme but may or may not accomplish the goals. For example, a training programme may have the goal of increasing the employment of disabled people. An ultimate desired outcome might be reduction in the unemployment rate of disabled people in a community to a level on a par with others in the community. Since that outcome is ultimate, a short-term objective might be set to reduce the unemployment rate of disabled persons by some percentage in a fix period of time (i.e., 1 per cent in a year). A training programme output could then be manifested in terms the numbers of persons served by the programme. It is then possible that the programme could have a positive output of training several persons with disabilities but have no impact on reducing their unemployment rate. This disparity between outputs and outcomes would not necessarily mean that the programme is unsuccessful. After all, many factors exogenous to the programme could be influencing the unemployment rate. If, however, a programme showed no impact and outcomes over a longer period of time and if it were established no other exogenous factors were present, its mix of activities need to be reviewed and evaluated.
These issues are similar at both the programme or the country level. For example, a country might establish a long-range goal that the educational attainment of persons with and without disabilities should be equal. Thus, education and training programmes would be targeted to that ultimate goal. Such a goal, however, may not be readily achievable so the country might set a short term objective for educational attainment - perhaps to raise the graduation rate to a specific level by a certain date. That would be an outcome measurable by data. Outputs could be measured by accommodations provided to students with disabilities to enhance their likelihood of graduating. Inputs would be measured by the resources brought to bear to provide those accommodations (i.e., money spent on accommodations).
Clearly, efforts to obtain data should not ignore the goals of both United Nations and of country-specific programmes in the field of disability. Any strategies to monitor progress towards the goals of the World Programme of Action must assess both outcomes and outputs. If the measures of outputs appear to demonstrate success but desired outcomes are not achieved, assessment of the determinants of outcomes is critical. This requires the clear delineation of outcomes from outputs, with the logical first step being the consideration of outcomes.
b) Socio-economic outcomes
Traditionally, programmes for people with disabilities would focus on reducing functional or activity limitations to achieve goals. To assess the situation of disabled persons, programmes must now measure outcomes related to the areas of equalization of opportunity cited in the World Programme of Action or the target areas for equal participation in the Standard Rules and reflect areas where outcome measures might be derived. Output measures of policy and programme activities related to these target areas must also be obtained and these tend to be related to socio-economic variables. The discussion begins with outcomes, because they reflect the ultimate quality of life goals of the World Programme.
One positive aspect of measuring socio-economic outcomes is that some basic guidelines for evaluation criteria have already been established. As cited earlier, the Statistics Division has long recommended that the measurement of both demographic attributes and socio-economic status for persons with disabilities be conducted in the same manner as for the entire population. To carry out this recommendation for certain areas, countries are urged to follow the recommendations of the United Nations for data collection, in particular those related to the year 2000 round of population and housing censuses. The recommendations relevant to outcomes cited in the World Programme and the Standard Rules are summarized in figure 3.
Figure 3. SUMMARY OF UNITED NATIONS PRINCIPLES AND RECOMMENDATIONS FOR POPULATION AND HOUSING CENSUSES APPLICABLE TO EQUALIZATION OF OPPORTUNITY AREAS CITED IN THE WORLD PROGRAMME OF ACTION CONCERNING DISABLED PERSONS AND TO TARGET AREAS FOR EQUAL PARTICIPATION CITED IN THE STANDARD RULES ON THE EQUALIZATION OF OPPORTUNITIES FOR PERSONS WITH DISABILITIES
Sources: United Nations General Assembly, World Programme of Action concerning Disabled Persons (A/37/51/Add.1 and Add.1/Corr.1, annex), pp. 30-34 and The Standard Rules on the Equalization of Opportunities for Persons with Disabilities (General Assembly resolution 48/96, annex) and United Nations, Principles and Recommendations for Population and Housing Censuses (United Nations publication, Sales No. E.80.XVII.8, 1980), Supplementary Principles and Recommendations for Population and Housing Censuses (United Nations publication, Sales No. E.90.XVII.9, 1990) and Principles and Recommendations for Population and Housing Censuses (ST/ESA/SER.M/67Rev.1, 1997).
For educational and economic characteristics, criteria for measurement have been identified by the United Nations system. These two areas stand out for monitoring for three reasons, as previously noted. First, along with social questions, these two areas were highlighted in the World Programme of Action as key factors in furthering equalization of opportunities of persons with disabilities. Second, UNESCO and ILO have already established systems to measure programmatic outputs in these areas. Finally, criteria for measurement of these two areas are relatively advanced; data have been collected in many countries for several years.
The measurement of certain other outcome areas has not advanced as far but there are technical resources to support this work. For instance, United Nations recommendations in the areas of income, marital status, household and family relationships and households make it possible to compare the situation of persons with and without disabilities. Moreover, DISTAT contains information on urban and rural residence, household information and family composition. Thus, these data can serve as proxies for information related to Standard Rule 9 on family life and personal integrity. Information on residence can also be compared.
Other areas, such as recreation, culture and sports may require the establishment of criteria for measurement, however. These may be more difficult, as information for these areas are usually obtained on surveys. If these are surveys only for disabled persons, comparative data cannot be obtained. Hence, countries may need to consider incorporating a disability question in surveys that obtain information on these areas to measure progress towards equality.
c) Environmental variables
Beyond measures of these target areas, measures of environmental change are also needed to ascertain the conditions that must occur in order to reach established targets. If outcomes represent the end of the process, the environment represents the beginning. As shown in Figure 4, environmental factors not only have an impact on the situation of disabled persons but may also reflect the actual identification of disability. For example, the prevalence of reported disability may actually increase in response to successful CBR services. Although the figure does not reflect this, it should be noted that disability can also influence the environment, for instance an increase in the number of disabled persons may cause governmental priorities to shift.
While information on the environment may not be systematically available, some countries have collected information on this issue. DISTAT contains tables on social isolation, community attitudes towards disability and persons with disability, transportation and problems encountered in the home. These data correspond well to Standard Rule 5 on accessibility. For the most part, these data are only available from surveys. However, these data are not just limited to developed countries. For example, over fifteen years ago, a survey of disabled children in Ethiopia obtained information from community leaders on reasons why children were not immunized, major problems facing disabled children, efforts to alleviate conditions and attitudes towards disability, as well as family opinions on the problems of their children with disabilities.
Figure 4. FACTORS INFLUENCING ACHIEVEMENT OF EQUALIZATION OF OPPORTUNITY.
In considering the physical environment, the World Programme considered barriers in the following areas of the built environment: human settlements including those in rural areas, public buildings, public housing, public transport and support services. Standard Rule 5 divides the environmental accessibility into the physical environment and information and communications services. Accessibility affects all, not just persons with disabilities, so the environmental frameworks used by the United Nations Conference on Human Settlement (Habitat II) in its Istanbul Declaration on Human Settlements and the Habitat Agenda, as well as frameworks elaborated in environmental forums at other levels, may provide useful guidance for monitoring this issue. If the revision of the ICIDH framework results in a "Classification of the Environment", such a taxonomy should be considered as the basis to assess progress towards accessibility.
d) Impairment and disability
The discussion thus far has not focused on Impairment and Disability variables for monitoring. Clearly, such variables are important for monitoring. Indeed, an important determinant of population socio-economic outcomes may be a dramatic increase in the number of persons with disabilities. As noted in figure 4, the environment clearly influences the societal perception of Disability in individuals.
To measure the disability dimension, the United Nations has recommended that a person with disability should be defined as a person who is limited in the kind or amount of activities that he or she can do because of ongoing difficulties due to long-term physical condition, mental condition or health problem. Only disabilities lasting more than six months should be included. The Statistics Division recommends that, due to the limited space available in a census, a Disability-oriented question be asked (as opposed to an Impairment or Handicap question).
While individual Impairments and Disabilities clearly can have particular impacts on equalization of opportunity, in the past the societal perception of these impacts have often driven the determination of evaluation criteria. Thus, the discussion here has focused on variables related to outcomes and the environment to encourage a balance in such criteria. It is anticipated, however, that basic levels of Disability should be monitored in accordance with the recommendations of the Statistics Division cited earlier for countries to use a Disability-oriented question.
e) Target areas for empowerment and accessibility
Like environmental determinants, measurements of critical life areas for equalization of opportunity, such as whether persons with disabilities are empowered to independently make decisions in their lives, to have control over their use of time, to have control over economic resources and to prepare to be ready for major changes, are needed. This is because results in these areas often determine if the outcome targets will be achieved. Because each individual's situation is unique and a function of many factors, such as age or culture, it is important to understand the particular encounter of each disabled person with his or her environment. If the interactive nature of the concept is not assessed, then understanding of a critical aspects of barriers to reducing disadvantage may be lost.
Often, the areas related to Handicap are confused with the target areas of the Standard Rules. In an unpublished paper, Dr. Mary Chamie has argued that there are two basic approaches to the evaluation of Handicap: (a) estimation of the gap between persons with and without disabilities and (b) estimation of specific reductions in life roles. The first type of estimates have been considered as the ultimate desired outputs from the World Programme of Action and the target areas in the Standard Rules. The second type has been discussed in Section II.B, above. The question is considered in greater detail in Annex I.
As demonstrated in figure 4, it is through its impact on these dimensions that the environment can enhance or block achievement of Programme goals. Thus, all of these dimensions must be considered in whatever goal is being considered. If both education and employment are target areas, then independence, use of time and the other dimensions in education must be evaluated and independence, use of time and the other goals in employment must be evaluated. For example, laws outlawing discrimination may allow all children to go to school, but students may be ignored by their fellow classmates or not be given the time they need to complete tasks. Thus, students may be in a situation where they may have independence and mobility in school, but their use of time and social integration experiences are not the same as other students. Others in the school situation may be doing fine but are not being prepared for transitions from school to work. If the situation is not systematically addressed along all of the Handicap dimensions, monitoring may indicate that environments are being changed while students are doing poorly in school, without allowing for an identification of why the situation is occurring.
While these concepts may appear to be abstract, operational questions have been provided for these dimensions. If these dimensions are measured properly, not as individual abilities, but rather as the actual circumstances in which people find themselves and may place them as at a disadvantage, then programmes can target those life areas to have their greatest impact.
f) Programme output and input measures
Output measures might be numbers of laws passed or country funds expended on activities. The implementation measures in the Standard Rules are reflective of areas where outputs might be measured. Input measures might be comprised of the number of United Nations staff that are assigned to disability issues or the amount of funds devoted to United Nations initiatives on disability. Programmes seek to change the environment or change the impact that the environment has on persons with disabilities. Thus, if one reexamines figure 4, programme variables could impact directly on the environment or disability or somewhere along the path from the interaction of disability and the environment to one or more of the opportunity and access life areas, ultimately affecting the outcome.
Countries may wish to consider examining several different dimensions to organize programme output measures. For example, in establishing measures, it may be desirable to ascertain whether all aspects of the environment must have been addressed, all target areas for the Standard Rules and all life area dimensions. At a minimum, data should be gathered to assess progress towards short- and medium-term goals for the Long-term Strategy. Data that have been gathered by UNESCO, ILO and the Special Rapporteur on Disability of the Commission for Social Development can be examined as to relevant models for programme data.
Traditionally, in some countries, there may be some tension between producers and users of programme output data and producers and users of data from population-based surveys. Those involved with administrative records data may believe population-based surveys do not obtain data on the critical questions for programme evaluation. Those involved with population data often feel the representativeness of administrative records data can be called into question, as well as its orientation. For instance, producers and users of administrative data may believe in the particular programmes for which the data are relevant, and bias can become an issue.
It is important to recognize that administrative records data are critical for monitoring programmes. Two types of administrative record data for disability programmes are (1) service records and (2) registries. Together, with population based censuses and surveys, such data can provide a comprehensive portrait of progress towards Programme goals. Strengths and weaknesses must be candidly addressed. Recognition of these caveats will allow for all data to be employed to their maximum potential as an important part of a programme monitoring.
g) Outline of a comprehensive monitoring programme
It is clear that there are several different sources of data for monitoring disability programmes. Regardless of source, there are six different types of information that need to be collected: (1) socio-economic outcomes, (2) environmental variables, (3) disability-specific information, (4) target areas for empowerment and accessibility, (5) programme outputs and (6) programme inputs. All of these are important individually and how they fit together into a comprehensive monitoring programme must be clearly delineated as well. A hypothetical example may provide an indication of how these pieces could fit together.
Let us assume a developing country targets employment as a priority area for implementation of the World Programme of Action. The country might set basic long-term goals of no difference between persons with and without disabilities in terms of their economic activity dimensions, including no differences in labour force participation, occupational, industrial and class of worker structures, unemployment and income. In other words, all aspects of economic activity would be the same, with the caveat that there might be legitimate differences due to age, residential and other population structure variables. Since the country is developing, the policy would recognize that such long-term goals might occur in the context of other goals, such as increasing employment opportunities for all.
The country would review the Programme, the Long-term Strategy and the Standard Rules to determine medium-term goals for implementation by 2002, as recommended by the Strategy. For example, instead of eliminating the difference in unemployment, perhaps the country would set a goal of reducing unemployment overall by some amount but reducing unemployment for persons with disabilities by a greater amount, thus narrowing the gap in unemployment between the disabled and total populations. From the Long-term Strategy, the country perhaps decides to implement ILO Convention Number 159 concerning employment of disabled persons and to adopt the Standard Rule 7 on Employment by 2000. The appropriate legislation is passed, but, in order to meet its medium-term unemployment goals, the government would work with organizations of workers and disabled people to target particular areas for environmental change. A systematic monitoring programme might produce measures in all of the areas shown in figure 5. The measures presented in the figure are examples and are not necessarily comprehensive or preferable. These measures might include the current baseline, the desired goal by 2002 and the actual figure that occurs in 2002.
Notice how the unit of analysis may change depending on the monitoring concept and target area. For outcomes, the overall situation of the disabled population as compared to the total (or non-disabled) is measured, with individuals as the unit of analysis. For the environment, persons are not the unit of analysis. Rather, the units may be characteristics of places of employment or other environmental constructs. For Impairments and Disabilities, the prevalence of particular conditions might be measured. When the opportunity target areas for access are examined, individuals are examined. While comparisons to non-disabled persons can be made here, the critical variable to examine is the particular circumstances of the individual's situation with a focus on access. For programme outputs, individual results may or may not be examined but, if so, only for individuals in the programme. If programmes are targeted not at individuals but directly at the environment, individuals would not be the unit of analysis. Hence, programme output measures are targeted towards the direct subjects of programme interventions, whether they are people or things. Programme inputs are primarily constituted of the resources available to the programme, such as finances, personnel, time and materials.
As soon as data are available, the 1999 to 2002 trends can be compared. If the desired progress is not achieved, the country would know whether movement had not occurred at the outcome, environment, life area, output or input level. It could then develop strategies to target the appropriate level where change is required.
In summary, a comprehensive monitoring programme consists of setting long-term goals and specific objectives and establishing the strategies required to meet those targets, as well as the strategies that will be employed if targets are not achieved. In such a programme, the goals and strategies drive data collection.
Figure 5. EXAMPLES OF MEASURES FOR A SYSTEMATIC EMPLOYMENT PROMOTION PROGRAMME, BY MONITORING CONCEPT AND TARGET AREA
3. Resource constraints
Resource constraints are an important consideration. Many countries may not have the resources to engage in a monitoring programme as described in the preceding section. Strategies to improve programme monitoring must consider several dimensions: the most appropriate information on programme implementation, the most appropriate assessment tools to obtain the information, the most appropriate frequency for obtaining the information, country-specific facilitators for programme monitoring and resource constraints. Different options involve placing resources in different combinations of these dimensions.
In making their plans, countries need to consider the particular strengths inherent in their situation that could facilitate information collection. The most influential institutions in societies should be encouraged to provide input to the most appropriate mechanisms to obtain the needed information and be enlisted to assist the collection effort as well. Organizations of persons with disabilities play a key role in the success of the information collection efforts. Such organizations should be provided the resources needed to ensure their involvement in all phases of data collection efforts: determination of the questions to be asked, implementation of the data collection, and analysis of the results.
Resource constraints should not prevent the participation of persons with disabilities in the data collection effort. In fact, consumers may be able to assist in the determination of questions that people are not likely to answer accurately. They may also suggest the most appropriate priorities for obtaining information for policy purposes. Determination of priorities for information gathering can be driven by the determination of policy and programme priorities.
The important consideration here is that resource constraints need to be considered after evaluation criteria are established. If the decision is made to prioritize the collection of information on Impairments, at least such a collection is not automatically the default. When decisions are made to implement certain criteria, persons are aware of the role that resource constraints are playing in the determination. This allows for the setting of monitoring priorities, based on programme priorities, subject to resource constraints.
4. Options for monitoring and setting priorities
There is no strict formula for setting priorities between monitoring options and evaluation criteria. When circumstances are especially difficult, the Statistics Division has recommended using rapid, low-cost methods, employing any existing counts for sampling frames, relying on direct interviews and observations rather than the "key informant" approach and prioritizing the obtaining of information on etiology, availability of CBR resources, barriers and indicators of Handicap. Handicap should be assessed in terms of integration, self-sufficiency, participation and opportunities.
The key point is that programme priorities should drive the monitoring priorities and not the reverse. With the key points of information in mind, there are several different options for monitoring, including a decision not to monitor. Since the World Programme calls for monitoring, options include the following alternatives:
In establishing a monitoring programme prioritizing the collection of at least some information from each dimension to obtain differing perspectives, countries are urged to make special effort to collect some information relevant to environmental variables and target areas. These dimensions have been ignored in the past and are critical to equalization of opportunities.
If resources are scarce, countries may wish to consider limiting the collection of Disability information to the United Nations recommendations so that some resources are available to collect environmental information.
Another option for countries is to establish priorities for some outcome variables over others. For example, if overall improvements in employment and education outcomes are likely to improve participation in sports and recreation rather than the reverse, countries may wish to prioritize the collection of education and employment information. This is difficult at a time when there is growing realization of the interrelation between many facets of life. However, given the scarcity of resources and other priorities that have been placed on education and employment, such a choice may be increasingly necessary.
Along with the goals and objectives, a third option for prioritizing information gathering involves identifying strategies that a country will employ if targets are not met. If the monitoring programme is to make a difference in the lives of persons with disabilities, a country should be prepared to make changes if monitoring information reveals objectives are not being met. If a country is prepared to do things differently in one area as opposed to another, that preparation may suggest that the first area obtains priority for monitoring. The only caveat is that if a country has a history of placing greater resources in the health ministry than in the educational and employment ministries, there may be a built-in bias towards monitoring disability prevention instead of equalization of opportunities.
A fourth option may be simply to rely on information currently gathered but not try to change data systems based on monitoring decisions. While this option would appear to be close to doing nothing - or to let data resources drive the monitoring decision - such an option may be necessary if the census is the only source of population data or if a country is unlikely to shift resources in response to monitoring information. If this is the case, at least it will be clear that the option was chosen and is not being viewed as a more activist monitoring effort.
Of course, there may be other options. This is recognized in the final option, which recognizes that countries may wish to engage in several different combinations of monitoring activities. For example, if a country is at a stage in development where it is effectively designing its transportation system but does not have the resources to enforce equal opportunity in employment policies, a priority might place a higher short-term priority on monitoring mobility in travel, rather than on employment. Country issues will determine the choice of options.
These issues pose a dilemma for the United Nations system as it engages in international monitoring activities. As countries set different priorities, the comparability of monitoring information becomes an issue. This creates a tension between two competing desires: to have a true set of international standards for all people; and to respect country and cultural differences. To some extent, these problems are not completely different from problems faced when the United Nations monitors other development instruments and programmes. In such cases, technical co-operation activities can play productive roles in balancing the substantive concerns and country-specific issues while making monitoring a critical element in efforts to improve the lives and well being of persons with disabilities and their families.
 See "Using statistical information to plan and evaluate rehabilitation programmes," Chapter IV in Manual for the Development of Statistical Information for Disabiltiy Programmes and Policies (United Nations publications, Sales No. E.96.XVII.4).
 Manual for , p 5.
 Ibid., pp 46-48.
 Ibid., pp 102-104.
 Review and Appraisal of Progress (CSDHA/ DDP/GME/CEP.4)" p 30.
 United Nations, Principles and Recommendations for Population and Housing Censuses (United Nations publication, Sales No. E.80.XVII.8); Supplementary Principles and Recommendations for Population and Housing Censuses (United Nations publication, Sales No. E.90.XVII.6); and Principles and Recommendations for Population and Housing Censuses; revision 1 (ST/ESA/SER.M/67/Rev.1).
 World Programme of Action , paras 120-133.
 See United Nations Disability Data Base, 1975-1989: technical manual (United Nations publication, Sales No. E.88.XVII.12), pp 3-4; and Manual for the Development of Statistical Information for Disability Programmes and Policies (United Nations publication, Sales No. E.96.XVII.4), pp 14-15 and table I.3.
 Manual for the Development , p 104.
 Ibid., p 14, table I.3.
 These data are contained in DISTAT.
 World Programme of Action , para 114.
 Ibid., para 6; "Standard Rules ", paras 17-21; Manual for the Development , pp 9-12.
 "Demographic and Social Statistics; 2000 world population and housing census programme; report of the Secretary-General (E/CN.3/1997/14)" para 29.
 Chamie, Mary, "A perspective for the classification of handicap" (Zoetmeer, the Netherlands, 1992) (unpublished paper presented to Second meeting on updating the International Classification of Impairments, Disabilities and Handicaps, Zoetemeer, 17-19 March1992).
 Chamie, Mary, Report of the sub-committee on the conceptual harmonization of statistics for the study of disability-free life expectancy (Strasbourg, 26-27 November 1990) Reseau de Vie en Santé (REVES ) paper no. 41.
 Manual for the Development , pp 108-111.
 United Nations, Principles and Recommendations for Population and Housing Censuses; revision 1(United Nations publication, Sales No. E.98.XVII.8), para 2.261.