UN Convention on the Rights of Persons with Disabilities: a call to action on poverty, discrimination and lack of access

Menu:

Click on the image to download the PDF Version

Cover page of Conference Report (Click here to download the PDF Version) 

6. Count me in: pre-conference workshop on data and statistics


… Statistics are essential for sustainable economic, environmental and social development, and public trust in official statistics is anchored in professional independence and impartiality of statisticians, their use of scientific and transparent methods and equal access for all to official statistical information - Principles Governing International Statistical Activities, United Nations – Statistics Division

Research is not the end. The end is a fully inclusive society with no discrimination and people not being excluded from life. Statistics and research are tools. They can be particularly effective used by governments to figure out how to allocate their resources. - Daniel Mont, Senior Economist, Disability and Development Team, the World – Bank

Stand up and be counted: a familiar rallying cry for many social justice movements, not least among those who lobby for full human rights for disabled people. Yet being counted, both physically and metaphorically, is precisely what many people with disabilities have not been able to achieve. Effective disability and development policy relies on accurate statistics about disability prevalence, but there is significant debate about the reliability of almost all the disability statistics from developing countries.

Many pressing questions need answers, and speakers and participants throughout the conference have had to rely on statistics for many of their arguments and to inform their work. At the same time, some very basic questions, such as how many people with disabilities live in developing countries, and how many disabled children in Africa attend school, cannot be easily answered based on existing data.

How to best collect the data that can change this situation was the focus of the pre- conference workshop’s presentations, which covered the issue from a wide range of different viewpoints.. Perhaps surprisingly, this statistics workshop was not primarily about numbers, but instead about how definitions of disability influence how it is perceived both by the public and those trying to measure it.

Speaker Daniel Mont, for instance, argued that previous census questions had unwittingly reinforced the medical model of disability, by concentrating on physical conditions rather than on what a person can or cannot do. Jean-Francois Trani echoed this in his contribution on the capability approach to disability measurement, which is based on a philosophy that considers what an individual can achieve or not achieve within the society around him or her.

The importance of cohesion and comparability arose throughout the presentations. In discussing the World Health Organization’s International Classification of Functioning (ICF), and the work of the UN Washington Group on Statistics, the participants recognized the need for a worldwide set of measures that allow statistics to be understood and believed across cultures. They also pointed out that the ICF and the Convention define disability as a continuum and that, as such, measurements must capture a situation that is in constant flux.

Since not all governments carry out regular national censuses, it was also thought to be vital to develop tools that can be used in many different kinds of sample surveys. These include UNICEF’s Multiple Indicators Cluster Survey (MICS) system, and other surveys specifically on disability or looking at broader demographic and economic issues, defined and conducted by national statistical offices, which can complement other kinds of economic and social research projects.

Questions of methodology are not at all theoretical; they have very real implications for data collection and resulting policies. As Marguerite Schneider showed in her presentation, different kinds of surveys can lead to extremely different counts of people with disabilities. With disability prevalence estimates in one area of South Africa varying from 2% to 12%, the importance of methodology in gathering accurate results becomes paramount.

In summation, Workshop Chair Professor Nora Groce reminded participants that we stand in the middle of an evolving process. One of its goals is to make sure that disability is considered in all spheres of public life. Coming from such a “famished atmosphere,” some stakeholders may feel as if we can only have one disability question in a survey.

“The goal,” Professor Groce reminded everyone, “is not to have one project after another collect fragments of information, but rather to ensure that major population surveys and censuses include disability as a key component in their regular work. When disability-specific data collection efforts are undertaken, it is also vital that there is collaboration and coordination among researchers, organizations and agencies. This would allow us to gather, use and share such information as effectively and efficiently as possible. Unless, and until, we have such data, it will be much harder to engage in meaningful discussion with policy makers in health and development circles, or to plan effective programs or policies for people with disabilities at local or national levels.”

To stress this point to the international disability and development community, and to underline the need for accurate statistics in policy making and service development, the workshop participants came up with a unifying call for action on disability data and resulting development programs.

We call for:

  1. Collection of disability statistics using current developments in thinking on disability, measurement and statistics, such as the International Classification of Functioning and the United Nations City Group (Washington Group on Disability Statistics)
  2. The collection of more and better statistics for all groups of people with disabilities, particularly focusing on the need for the development of national disability statistics pertinent to children and older people with disabilities
  3. Development of national disability statistics pertinent to children with disabilities
  4. Medical and therapeutic evaluations following household surveys, to address issues around false positives and false negatives.

6.1 Monitoring child disability via household surveys: the MICS experience
Speaker: Claudia Cappa, Strategic Information Section, Division of Policy and Planning, UNICEF USA

As the world’s main safeguard for children’s rights, UNICEF recognizes that the incidence of child disability is relevant to many areas of its work. These include monitoring the impact of:

  1. Improvements in child survival, and the potential that these are accompanied by increases in the prevalence of child disability
  2. Childhood exposures to nutritional deficiencies, environmental toxins, infectious diseases (such as malaria, HIV, encephalitis), and trauma
  3. Interventions designed to improved child health and development.

In addition, population-based information on the status of children with disabilities is needed to:

  1. Identify risk factors, causes and strategies for prevention
  2. Facilitate planning of services
  3. Monitor and improve participation and quality of life.

To supply the most useful statistics through household surveys, UNICEF uses MICS, or Multiple Indicators Cluster Survey. Its disability module is based on asking parents or caregivers ten questions designed to identify children aged two to nine years old in any cultural and social setting, who have congenital and developmental impairments. The MICS disability module can provide data on the type of impairment (seeing, hearing, language production, etc.) and actual health conditions, such as epilepsy, and activity limitations, such as not walking or playing.

Accurate data collection supports development of services for children with disabilities. © Leonard Cheshire Disability.So far, MICS has been used in 53 countries and measured 101 indicators, including nutrition, mortality, child protection, HIV and others. The data can be broken down by background characteristics such as ethnicity, wealth and education and at the sub-national level, and can be used for monitoring progress towards the Millennium Development Goals and other international commitments.

Like the World Bank’s methods50, MICS follows the social model of disability, and looks specifically at activity limitations and participation restrictions within the ICF framework. Data analysis can then assess risk factors, and explore associations between existing impairments and participation in activities such as education, which point toward social barriers to inclusion.

With a reliance on universal human abilities, MICS is cross-culturally comparable. It also relies on caregivers’ ratings of children’s development and behavior, which ensures that answers are relative to their own cultural norms. Analysis has also shown MICS not to be gender biased, making it equally valid for girls and boys.

Under the MICS definition, a child is considered disabled if he or she:

  1. Has delay in sitting, standing or walking
  2. Has difficulty seeing, either in the daytime or at night
  3. Has difficulty hearing
  4. Has difficulty in understanding instructions
  5. Has difficulty walking or moving arms, or has weakness or stiffness of limbs
  6. Has fits, becomes rigid, loses consciousness
  7. Does not learn to do things like other children of his/her age
  8. Does not speak at all
  9. Speaks differently from normal or cannot name at least one object
  10. Appears mentally backward, dull or slow.

MICS works through a two-phase design, meaning that if the questions indicate a possible impairment, then this must be verified by a clinical evaluation. This leads to several limitations in MICS, namely that it can be hard to provide a diagnosis in countries where access to health care resources are limited, and that it tends to pick up more severe impairments such as intellectual, motor and seizure disabilities, but can miss milder visual and hearing impairments. Its validity for evaluating autism and attention deficit disorders is also not yet established.

MICS has recently been used by a global study on child disability by UNICEF and the University of Wisconsin in the USA. The results are based on data from 20 of 53 countries participating in MICS3, with 205,674 children screened. Findings show the correlation of disability to a wide range of variables, including gender, nationality, national region and health interventions, such as iodization of salt and supplementation with vitamin A.

Some interesting challenges to prevailing theories have emerged from this data, such as finding no consistent relationship between maternal education/household wealth and screening positive for disability across countries. In addition, salt iodization is not correlated with the percentage of children screening positive, and neither is more violent methods of discipline.

This study has shown other factors related to disability. For instance, children from minority populations within countries often have a higher rate of disability. Children who have never been breastfed have a higher rate of disability in every country, as do children who have never taken vitamin A supplements. In almost every country, the number of children with disabilities who did not attend school outnumbered those who did.

The next steps for data analysis of disability among children in developing countries include recognizing the need for medical follow-up to validate data, and guaranteeing treatment for children at risk. It is also important to keep addressing issues of participation and discrimination, and to create partnerships that can bring about synergies in data collection.

To learn more about MICS and UNICEF’s role in disability data collection, please visit www.childinfo.org.

6.2 Measuring disability: from concept to operation
Speaker, Daniel Mont, Senior Economist, Disability and Development Team, the World Bank

Questions of how best to collect data strike at the heart of one of the most potent debates within the disability movement: whether concentrating on measuring and quantifying impairments promotes the medical, rather than the social model of disability. This is the thesis of Daniel Mont, who pointed out that taking a medical model approach when designing surveys and censuses, keeps the focus on individuals themselves rather than how they interact with their environment. This can lead to interventions being done at the personal level, including medical care and individual adaptation, instead of considering the overall physical, cultural, and policy circumstances.

One problem with medical model type data collection is its focus on diagnoses. Since not everyone is aware of or has a diagnosis for impairment, the survey will never capture all the relevant information. Even if a person can accurately report a diagnosis, this does not necessarily provide information on conditions that have a wide variety of effects. Interventions such as rehabilitation, which cannot cure but can improve a person’s physical functioning, and thus impact on his or her social interaction, are also not taken into account. Finally, medical model questions will also miss socioeconomic factors that are either a cause of, or caused by, the impairment.

Even using the word disability can miss people in a census, Dr. Mont argued. For instance, some people may not consider themselves or their family members to be disabled, particularly if the impairment is associated with age, and others might not understand what the word means. People may interpret disability differently to the people who designed the questionnaire, and the word itself can carry a stigma and so be avoided for personal and social reasons.

Partly in response to these issues, the World Health Organization (WHO) developed the International Classification of Functioning (ICF). This system is an attempt to capture the social model of disability, which is based on a person’s ability to function within society rather than a specific impairment, and considers a person’s bodily function and structures, activities and participation. It does not measure disability and is a classification rather than a measurement system, meaning that disability is not considered as separate from the environment.

The three elements of the system are designed to interrelate and be considered simultaneously:

Body function and structures consider the physiological and psychological function of body systems. The ICF allows specific recording of detailed functional abilities and impairments that are not linked to a cause. For example, issues of “fuency and rhythm of speech functions” could result from stuttering, stroke, or autism.

Activities describe someone’s functioning in his or her environment as a whole person, as opposed to focusing on the function and structure of body parts. The activities considered range from basic (such as walking) to complex (for instance, work and schooling), and cover tasks an individual can do that require multiple body functions.

The participation element allows for inclusion of higher order activities that involve integration in the community.

When considering how to put these principles into use, remember that good survey and census questions capture specific concepts in a uniform way, and that the purpose of measurement determines the method. The first question to consider, therefore, is why do you want to identify disabled people? From this flows the issue of whether it is disability or functioning that is most important in each particular research initiative.

An effective census measures opportunities and abilities as well as impairments. © Jenny Matthews/ Leonard Cheshire DisabilityResearchers can then decide which section of the ICF model is most pertinent – body function, activity, or participation. Does the survey want to consider health conditions and their causes? What attention is to be paid to the environment, both general and personal (for instance, are assistive devices available)? What is a feasible survey length and sample size, and what is to be measured compared to what will be identified through analysis?

In practice, a census may not allow much space for questions on health and disability. The option to compare results across countries, across demographic groups and across time then becomes highly desirable, so questions should be chosen carefully.

The UN Washington Group on Statistics (a temporary organization set up by the UN in 2001, through which representatives from national statistical agencies come together to address issues around disability and statistics) decided that measurement within a census should determine “Equalization of Opportunities.” This approach leads to basic activity questions, because relatively few questions can yield some important information and opportunities for cross-referencing. Examples of these include:

Core questions:

  • Do you have difficulty seeing, even if wearing glasses?
  • Do you have difficulty hearing, even if using a hearing aid?
  • Do you have difficulty walking or climbing steps?
  • Do you have difficulty remembering or concentrating?

Additional questions:

  • Do you have difficulty (with self-care, such as) washing all over or dressing? Using your usual (customary) language, do you have difficulty communicating, for example understanding or being understood?

Response categories:

  • No – no difficulty
  • Yes – some difficulty
  • Yes – a lot of difficulty
  • Cannot do at all.

Analysis can then make links that show, for example, that if people with activity limitations (such as difficulty walking) have lower employment, education and income, then there are barriers that lead to unequal opportunities. That is, disability. The prevalence rate generated must be understood in those terms, and this combination of measurement and analysis is consistent with the social model of disability.

The World Bank has chosen to follow this approach, and to support others to do so as well. Their first initiative in this area is to support the work of the UN Washington Group on Disability Measurement, both with census questions and other more specialized measures. They are also developing indicators for evaluating the impact of development activities on people with disabilities, and collecting primary data for use in planning, monitoring and evaluation.

Recently they have used these techniques in evaluating a microfinance program for disabled people in Andhra Pradesh, India, and in doing needs assessment for an earthquake relief project in Bangladesh. They are also experimenting with using a “Ten Question Screening Instrument” (TQSI) in collaboration with the OECD in Bangladesh, Ethiopia, Cambodia and other countries.

To support governments in collecting disability data, the World Bank has started an online training course51 and video for Latin America, and held regional training sessions in South and Central Asia. Disability surveys are another focus, and they have helped with projects in Bangladesh, India, Vietnam, Uzbekistan, Indonesia, Bosnia, Ecuador, Bolivia and Yemen, as well as collecting data for educational management information systems in Cambodia, Vietnam, Indonesia and Mongolia.

6.3 Review of surveys in Africa that capture disability
Speaker: Oumar Sarr, Social Statistics Expert, African Centre for Statistics, UNECA

Every African country, like all countries, has a national statistics office (NSO) to collect and analyze social and economic data, following internationally accepted methods and standards. Much effort has gone into empowering African NSOs and making them autonomous agencies, with more power to coordinate their national statistical system regardless of national political climates. African NSOs, however, may still lack some of the resources, staff and knowledge of methodology they need to obtain and disseminate the most accurate information.

Many NSOs make economic statistics a priority and just measure some common social indicators, including those on education, health, poverty and living conditions, as these are often used to support poverty reduction strategies. Gender, environment and good governance are new areas for African NSOs and they need support to strengthen their data collection practice in these subjects.

Disability information would be commonly collected along with other social data during population and housing censuses, which are the primary source of information about the number and characteristics of a given group of people. They are carried out every ten years and are often the only statistical operation that covers 100% of a country’s population. Such censuses remain the sole source of socioeconomic characteristics in many African countries, and also provide a sampling frame for other statistical surveys.

The UN principles and recommendations for population and housing censuses stressed that social issues be included in the 2010 round of censuses. However, NSOs must check the sensitivity and appropriateness of the survey framework, and determine whether specific questions are applicable for the census. Because of stigma within societies and cultures, whole sets of questions could be jeopardized. As other speakers also noted, there may not necessarily be a common conceptual framework for data or for definitions of disability, and countries use different approaches to capture disability information.

In common with the previous speakers, Mr. Sarr also recommends that censuses use the “diffculty” rather than “disability” approach. Questions can measure limitations in basic activity functioning, and then be cross-referenced with questions on performance of, or participation in, organized activities such as education or work. Persons who are at greater risk than the general population for experiencing restrictions in performing specific tasks, or participating in important activities, can then be identified, and a larger picture of the impact of impairments and social barriers can be formed.

For international comparison, it is considered appropriate to have questions on four different domains:

  1. Walking
  2. Seeing
  3. Hearing
  4. Cognition.

Additional domains to be considered include:

  1. Upper body functioning of the arms, hands and fingers
  2. Mental/psychological functioning, trauma.

Mr. Sarr then discussed the many different kinds of surveys currently being used in Africa, and the kind of information on disability that they yield. Disability information from other surveys typically covers individuals and households and collects data by disability type, cause and severity. The typical response rate is 70–100%, but it is worth noting that most socioeconomic surveys other than full censuses do not include questions on disability. Survey types include:

  1. A Demographic and Health Survey (DHS) will consider common health indicators and usually sample 3,000 to 5,000 households. The cost of such surveys is often high. National governments and partners such as USAID usually sponsor the survey, with a periodicity of four to five years. It does not always include disability questions.
  2. A Multiple Indicator Cluster Survey (MICS) (Please see the above report of Claudia Cappa’s talk on this subject) focuses on women and children’s health. Samples range from 3,000 to 5,000 households and sometimes include questions on disability. The cost of a MICS survey is high, so they are usually sponsored by countries, and the main partner is UNICEF.
  3. Household Living Conditions Surveys (HLCS) examine poverty and living conditions, with samples varying from 1,600 to 8,000 households. It is an expensive survey because of the length of data collection (almost 12 months), and is sponsored by multilateral and bilateral organizations such as the World Bank, European Union and DFID. Such surveys are typically done every four to five years and sometimes include disability as an issue.
  4. Core Welfare Indicator Questionnaires (CWIQ) also focus on poverty and living conditions, with samples varying from 3,000 to 15,000 households. It is a light version of HLCS and the data collection duration varies between two and four months. This kind of questionnaire is cheaper to implement, but also attracts sponsorship from organizations similar to those that sponsor the HLCS. These surveys usually yield some data on disability, but it is not disaggregated by type of impairment.
  5. An employment survey considers employment indicators only, and usually samples approximately 3,000 households. It can be expensive and is often sponsored by governments and the International Labour Organization. Such surveys are not common in Africa, but do sometimes include data about disability.
  6. Agricultural surveys are more common, and focus only on agriculture, with a sample of approximately 2,000 to 5,000 households, giving a picture of a large part of the active population (more than 80% of the active population in the region work in the agricultural sector). The cost of agricultural surveys is high and the Food and Agriculture Organization often sponsors them every three to four years. Disability questions are rarely covered in this kind of survey.

Given the wide range of types of surveys done in Africa, and the relative lack of funds available for them, including for the recommended population and housing censuses, Mr. Sarr concluded that disability stakeholders must advocate for the collection of data on disability in as many surveys as possible. He also argued for harmonization of definitions and methodology so that data can be compared across subject areas and across countries.

6.4 Capability and disability: an approach for a better understanding of disability issues
Jean-Francois Trani, Researcher, Leonard Cheshire Disability and Inclusive Development Centre, University College London

Previous speakers at this workshop discussed the medical and social models of disability, and the challenges that both of these present to accurate collection of data on disability. The speaker in this presentation, Jean-Francois Trani, introduced another perspective – moving from the social model to the capability approach.

Data collection in developing countries presents many challenges, Dr. Trani commented, not least of which is choosing the right measurement tools. Another complication is how best to measure not just disability prevalence, but also living conditions. The capability approach framework can help with this, by defying disability and impairment and also clarifying the best methods of assessment in a given situation. This compares favorably with other approaches, which all have both strengths and limitations.

The medical model (or the individual model) compares impairment to a physical norm, and the danger is that tools can end up measuring only a few categories of impairment, leading to a low prevalence rate. Stigma can also lead to underreporting and underestimates of disability prevalence.

The social model has helped achieve great progress within the disability movement in terms of advocacy, but also presents certain limitations for measurement. First, looking at levels of functioning in a given context can imply that society needs to be redesigned, rather than considering the implications of a specific impairment. In some cases, such as in Australia and the UK, society has already been somewhat redesigned, and there are disability prevalence rates of up to 20%. Some people in these societies are considered to have a disability in order to benefit from certain policies, yet from the point of view of the social model they are not actually disabled, because they are largely integrated into society.

The social model also does not consider a disabled person’s functioning in his or her environment. This is why Dr. Trani suggests moving towards the ICF model, which looks not only at impairment, but also at personal factors and the environment, thus taking interaction with society into account. Yet a survey that uses the ICF must ask many questions, and right now there is no single data collection tool that includes all its aspects. Some researchers also feel that the ICF needs to be adapted according to the cultural context in which it is used.

The capability approach, as developed by the economist Amartya Sen, gives a wider view – it considers not only disability issues, but also the needs of people with disabilities. Constraints, choices and values, as well as functioning, are all part of its analysis, and it moves from considering what a person can do physically to looking at the range of possibilities open to someone in a given context. It then progresses to exploring the potentialities of how an individual develops his or her place in society.

The philosopher Martha Nussbaum takes the system further, by proposing a list of ten central capabilities, including life preservation, good health and body integrity. This takes physical limitations into account, and also explores the ways in which societies can compensate or allow people with disabilities to benefit from central capabilities. In this way, she can address situations such as severe mental health problems and intellectual disabilities.

There are limitations to the capability approach, however. First, it needs to look at an individual’s potential, so in terms of measurement it leads to questions on both activities, and on the person’s perception of opportunities or limitations within society. It must also look at vulnerability, which may be defined as a risk of falling into a lower level of wellbeing, and this too necessitates further questions. Finally, studies using this approach should also consider the opportunities and constraints offered by a given environment. All of this leads to a complex and long measurement instrument, which raises concerns about how it can be practically applied.

A further complication is that the capability approach concentrates very much on individuals, both disabled and non-disabled, but does not include the possibility of collective action. Nor does it examine how people can connect with society and what responsibilities society has towards its citizens. Dr. Trani and the co-author of his paper on this subject, Professor Jean-Luc Dubois, suggest complementing the capability approach with a phenomenology approach. 52

This would allow for a systematic view and includes all behaviors and interactions inside a given cultural context. It also permits analysts to look at rights and obligations, the intentions behind actions and the full range of emotions and beliefs that constitute the reality of human life. In this context, the freedom exercised by the person with disabilities becomes a specific phenomenon to be observed and analyzed. The negative implication of combining these two approaches is that it would make measurement even more complex.

The first issue to consider within this new framework is how social and political collective capabilities could emerge from the consideration of several different individual capabilities. For the statistician and social scientist, this creates a problem of aggregation. How do you go from considering attributes of an individual that you can measure with an instrument, to understanding collective capabilities such as how people interact? The next task is to measure whether such interaction increases or decreases each person’s capabilities. A study done by P.B. Anand in India in 2007, started to tackle these issues by looking at capabilities for using metered water collectively, and considering the model of social relations in terms of public health and development. 53

The second issue is to look at responsibility. Two types of responsibility have been identified: that of an individual after an event happens (retroactive) and a prospective dimension – the feeling of each person in a group or society of being responsible for the wellbeing of everyone else. Will individuals limit their own wellbeing to improve the wellbeing of society? This issue has been well explored by the philosopher Emmanuel Lévinas, who considers wellbeing as being each person’s individual responsibility in society. 54

Dr. Trani tried to carry out an exercise of this kind in Afghanistan, and can offer a few examples of how to integrate this kind of approach. For example, they asked questions on whether people found education useful, which can uncover if people value education. To examine opportunities, they asked what people believed their situation would be like in five years. If it will be better, then what new opportunities will people have? If it will be worse, then what are society’s restrictions? If people with disabilities have been mistreated, then what are the motivations of the people who mistreated them?

Defining an instrument for data collection in this context is extremely complex. More variables are needed and the volume of information collected is much higher than in other kinds of surveys. This has consequences for both measurement and analysis, and researchers must consider issues such as choice, agency, capability and behavior and the local context.

To analyze the data collected, you could use micro modeling, predictive modeling and other techniques, to estimate people’s potential capabilities. Extrapolating from all potential capabilities that exist allows the creation of a typology of possibilities. Overall recommendations from such a study would look at ways to increase every person’s capabilities, particularly those of people with disabilities.

Time is another variable to be considered. To analyze it, researchers must look at how people’s situations improve or deteriorate through time in terms of achievements, potentialities and restrictions. One way to do this is through national surveys over a period of time, such as repeating the 2005 Afghanistan survey in 2008. This can be expensive, and dangerous in some circumstances (including Afghanistan), and so following a control group of people with disabilities through smaller surveys over a certain period can be easier. Both of these kinds of tools exist, and have been set up in several countries to tackle different issues. More research is needed to apply these methodologies to the case of disability, though, and to develop the most efficient tools for this kind of data collection.

6.5 Disability measurement and statistics – the state of the notion
Speaker: Marguerite Schneider, Research Manager, Human Sciences Research Council of South Africa

A. The issues to consider

There is growing interest and an increase in research into disability statistics, contends Marguerite Schneider, but progress in this area must accelerate even further if the comparability of statistics at regional and global level, required by the UNCRPD, is to be reached. Agreeing with the other speakers on this point, Ms. Schneider commented that we are now not sure if we are measuring the same thing, and cannot readily compare statistics across countries.

Generally, there are three levels of factors affecting disability statistics:

  1. Population – well understood
  2. Individual – not well understood
  3. Methodology – reasonably well understood.

From the population point of view, the world’s current aging population means that the prevalence of disability is rising. The level of development of health care services in a country has a direct influence on this trend, because of the varied capabilities in managing injuries and illnesses. Curable health conditions can also sometimes lead to disability if not treated, for instance, untreated middle ear infections can result in permanent hearing loss. In more developed countries, higher levels of industrialization and car use have led to higher rates of accidents and injuries.

At the individual level, a person’s overall sense of independence and identity, social inclusion or exclusion and overall disadvantage experienced (for instance, limited access to education and employment), all contribute to disability and need to be recognized and measured. Pertinent issues include:

  1. Poverty resulting from the impairment
  2. Access to health care services and having a diagnosis to report
  3. Age
  4. Cultural beliefs and notions of health and functioning
  5. Racial, ethnic and gender identities
  6. Level of education Socioeconomic status
  7. Access to knowledge and resources.

Methodology becomes particularly important when considering the wording of questions – for instance using the terms “have” versus “suffered” and “disabled/ disability” versus “difficulty.” The response option is also crucial for getting the best data, for instance, only offering the option of a yes or no response does not include enough opportunity for giving details of partial impairments, and results in fewer people reporting disabilities.

Increased and graded response options, from “no difficulty” to “extreme difficulty/unable to do” make people with mild difficulties more comfortable saying yes to a question. On the other hand, using the word serious within a question can create problems, as people have very different ideas of what levels of difficulty may be interpreted as serious.

As always, the social stigma of disability can influence answers. One survey question was worded, “God forbid someone has a disability, but if they do are they…?” Answers to such a question can hardly be expected to reflect true prevalence rates. Other questions that label people as deaf, blind, crippled or mentally retarded may also lead to a lower response rate about impairments.

It is better to consider difficulty as a complex, multidimensional phenomenon covered by a range of different questions as is done in the World Health Organization’s International Classification of Functioning. Since there is no single definition of disability, there can be no single measurement or statistic – we need to understand what we are measuring to understand the data.

B. The impact of different methodologies in South Africa

In South Africa, several studies have looked at qualitative and quantitative information on how people understand and view the notion of disability, and how this affects their answers to different questions on the subject. It also considered what frame of reference people use in answering questions on disability. The research examined three different sets of questions on how people rate their own situation, which were:

  1. “Are you disabled?” (Yes/no/sometimes)
  2. South African Census 2001: “Do you have any serious disability that prevents your full participation in life activities (such as education, work, social life)? MARK ANY THAT APPLY – None, Sight, Hearing, Communication, Physical, Intellectual, Emotional” (with yes/no response options)
  3. Washington Group (WG) Short Set, which asks about “diffculties” in doing range of activities, including:
    1. Do you (or does the person) have difficulty in doing any of the following (because of a health condition)?
             1= No difficulty, 2 = Some difficulty, 3 = A lot of difficulty, 4 = Unable to do,
             5 = Don’t know
    2. Seeing (with glasses if he/she wear(s) them)?
    3. Hearing (with a hearing aid if he/she wears one)?
    4. Walking a kilometer or climbing a fight of steps?
    5. Remembering?
    6. Concentrating?
    7. With self-care, such as washing all over or dressing?
    8. In communicating in his/her usual language, including sign language (i.e. understanding others or being understood by others)?
    9. Joining in community activities (for example, festivities, religious or other activities) in the same way as anyone else can?

In assessing the results, the study considered how easy it was to answer questions about difficulty and differences between responses for the three question sets. Specific issues included determining what sectors of the population are counted in or out for each of the three question sets, and the effects of having different response options – for example binary versus four response options.

One of the studies involved six focus groups across urban and rural areas and among different language groups. Twenty-one included adults answering questions about themselves, with nine groups calling themselves disabled, six groups unsure and six groups, non-disabled. Five groups were made of parents, with three groups having disabled children and two groups having non-disabled children.

Each participant completed a standard questionnaire before the group met, and then the group discussed the definition of disability and difficulty and reactions to questions. Sessions were recorded and analyzed thematically using Atlas-ti, a qualitative data analysis tool.

The national survey covered 1,508 Enumeration Areas (EA), with four households in each EA, and was a population-based survey of 6,032 households, of which 4,705 yielded useable data. Over 11,000 individuals aged 15 and older answered three questionnaires:

  1. a household questionnaire: head or most knowledgeable member of the household
  2. an adult questionnaire: all household members 15 years and older who could respond for themselves
  3. a Living Standards Measure (LSM) questionnaire: respondent on the household questionnaire.

Interesting differences in the way respondents viewed difficulty and disability became clear from the responses to this survey. Participants concluded that disability is:

  1. Permanent Visible and physical
  2. Not “curable” or “solvable”
  3. “Cannot do anything by themselves” versus “can’t do” only in impaired domain of functioning
  4. “Them” and “us” across all groups
  5. Not a positive connotation.

Responses to the idea of difficulty were very different, however:

  1. Difficulty is less severe than disability and can be solved
  2. Disability includes difficulty, but difficulty does not include disability.

In general, people thought that the questions were not too long, easy to answer and reflected their own lives. They also liked using terms such as “usually” or “normally” when talking about functioning with and without assistive devices.

As is shown in the chart below, responses to disability questions using the three different sets of methodologies, led to very different measures of disability within the community. For instance, out of all the people who said that they were unable to do one or more of the activities mentioned in the Washington Group survey, only 61% had defined themselves as disabled in the national census, meaning that the census had potentially counted 39% of disabled people as non-disabled. Among the people who reported that they had a lot of difficulty with certain tasks in the Washington Group survey, only 47% had defined themselves as disabled in the national census, potentially leaving 53% of disabled people out of official statistics of disability.

Such differences have a huge impact on official figures and the policies that are based on them. In the Washington Group survey, the response on one or more activities out of seven activities was:

  1. 67.7% of adults = no difficulty
  2. 19.9% of adults = some difficulty
  3. 9.9% of adults = a lot of difficulty
  4. 2.5% of adults = unable to do.

Using these measures, 32.3% of the population over 15 years old has some sort of disability. When the stricter definition of the disabled population is used (a lot of difficulty or unable to do) then 12.4% may be considered disabled. Even this figure is much higher than official estimates in South Africa.

Based on these findings, the questions for the national census in 2011 are being redesigned to use the difficulty model. The aim is for this census to use much clearer measures based on a better understanding of what we are measuring. They will also be inclusive measures that are more mainstreamed across social groups, and allow extensive analysis that leads to more effective disability policy and inclusion.

6.6 Data collection, gaps and issues: statistics relating to poverty and disability
Speaker: Imran Nazir Rana, National Coordinator, Leonard Cheshire Disability and Development Programme, Pakistan

Considering the situation for disability statistics in Pakistan, Imran Nazir Rana believes that there is a lack of reliable data. The most recent census in Pakistan was in 1998, which reported a disability prevalence rate of 2.49%, a figure thought by most people working in the disability field to be far too low.

The census included only one question on disability, which was phrased, “do you have a disability?” As other speakers have noted, this kind of wording commonly leads to underreporting, since people consider “disability” to be very serious and may not report less severe impairments. Disability may also be shameful in certain societies and so people may choose not to describe themselves in this way. This wording can also lead to undercounting of older people who have difficulties carrying out some tasks, since these problems may be seen as a normal part of the aging process.

National policies and programs have been designed based on these possibly inaccurate statistics, leading to insufficient resources being available for people with disabilities. For example, there is a 2% quota for employing people with disabilities in government departments, and funds allocated for rehabilitation programs also work on a basis of providing service to 2% of the population. The national budget for education for children with disabilities is now actually falling, because of overall cuts in government spending, but may have been insufficient to cover the true numbers of disabled children even before the budget cuts.

If we accept the contention that poverty is both a cause of and caused by disability, there is a growing need to conduct a systematic and scientific survey to uncover accurate data on disability prevalence in Pakistan. This could take the form of a single, reliable tool or method for data collection that could be used by all stakeholders.

The problems in both counting and providing services for disabled people in Pakistan was further complicated by the devastating earthquake that hit the North West Frontier Province in October 2005. Among the earthquake’s survivors, 4.35% of men and 4.28% of females became disabled, but none would have been covered by the 1998 census. These statistics are also not completely reliable, because different agencies used different measures to count the number of people affected.

There may be much to learn from the measurement process at grassroots level, where a number of local, national and international agencies in the earthquake-affected areas are trying to identify and count people with disabilities. Some of the common methods used to collect the data related to disability are:

  1. Working with community members on information gathering and reporting
  2. Collecting and analyzing registration forms used by organizations to register people with disabilities
  3. Checking referrals of people with disabilities from other organizations and individuals for help and services
  4. Female health workers/visitors reporting disability after household visits
  5. Persons with disabilities who directly reach out to the organization’s office for assistance
  6. Field workers working for organizations at grassroots level to collect the data and information about the PWDs in working areas
  7. Survey method, using questionnaires; however, the validity and reliability of the identification tools is under consideration.

Using these approaches in addition to other improvements in census taking and surveys, discussed by other speakers in the workshop, a single, reliable measurement program could be devised. Policy makers and other disability stakeholders could then use one single method for collecting data in all parts of the country, leading to poverty reduction and disability programs that are uniform from region to region and reach all Pakistani citizens equally.

Some workshop participants wondered whether waiting for accurate statistics to be gathered might hinder the development of effective disability and development programs in the short or medium term. Others even questioned whether a more accurate knowledge of disability prevalence has actually had a positive impact on practice or service delivery. The response to this came from participants from Brazil, whose experience is that better statistics have led to more inclusive policy.

There was also discussion as to whether it is better to measure disability in national surveys, such as the census, or if disability specific surveys are more effective. Participants again referred to the Washington Group recommendations, and suggested that census takers and other researchers develop longer versions of their questions. The information gained, they believe, will lead to inclusive measures.

How much does a particular impairment interfere with participation in the community? How many disabled children attend school in sub-Saharan Africa? Does the association between poverty and disability vary by country or region in Asia?

Unfortunately, nobody knows. And without answers to these and other questions like them, implementing effective disability and development programs and policies remains a challenge. Since understanding a situation is a prerequisite to changing it, addressing the worldwide lack of clear data about the prevalence and consequences of disability must become a priority.

6.7 Survey research in Afghanistan and Cambodia on youths with disabilities
Speaker: Jean-Francois Trani, Researcher, Leonard Cheshire Disability and Inclusive Development Centre, University College London

In countries as far apart in geography, history and culture as Afghanistan and Cambodia, participatory research in which people with disabilities studied the needs of other disabled people, has yielded similar valuable results, asserted Jean-Francois Trani.

Several things made the design and implementation of these studies similar to each other, and also unusual among disability research projects. Both were informed by the ICF and looked at a range of environmental factors. Both also used the capability framework inspired by Amartya Sen to consider disability within the wider spectrum of human development and the wider participation of people with disabilities in society.

Each study had a different purpose: in Cambodia, Handicap International (HI), which was providing physiotherapy services to children, wanted a better understanding of children with disabilities’ rehabilitation and other needs. The country had faced a long conflict that destroyed its infrastructure, which is now being rebuilt, and a socioeconomic survey had established a disability prevalence of 4.7%. The survey was thus designed to explore how HI should develop their programs to best meet disabled people’s needs. In Afghanistan, a range of stakeholders needed information to tackle the needs of people with disabilities as they themselves expressed them.

The studies began by organizing those involved into qualitative and quantitative study groups, to come up with objectives to assess: prevalence, livelihoods, education, service access, income, self perception, and social participation.

For both teams, the idea was to expand choices of people with disabilities in their cultural, economic and social environments. Questions in the surveys considered impairment, activities, participation, and, in Afghanistan, a control group of non-disabled people were included.

The survey teams had similar training, which was one month long and based on the inclusion of people with disabilities in the research teams, and sensitizing interviewers to disability issues. Differences included the fact that the survey in Afghanistan was a national exercise, so researchers went all over the country and interviews could range in length from 30 minutes to two hours. The Cambodian survey just targeted children, was held in only a few provinces and used shorter interviews.

The studies uncovered some striking similarities and some striking differences between the two countries. For instance, exclusion of children with disabilities from school in Cambodia is 45%, while it is 64% in Afghanistan. In both countries, however, more girls than boys are excluded. Again in both countries, when girls do go to school, they progress further in the education system. To a certain extent, if by comparing contexts in which surveys reveal similarities, you can define recommendations for policy makers that could be very similar across continents.

For instance, certain activity limitations are associated with higher levels of school exclusion in both countries. In Cambodia, children who have difficulties remembering things are most excluded from school, and children with trouble concentrating are most included. Sensory limitations and mental difficulties are also associated with a higher level of inclusion in both countries.

Access to health services in both countries is limited by a lack of equipment, inadequate staff training in disability issues, high costs to service users and the remote locations of health centers. All of these findings can help disabled people’s organizations lobby their governments to address these issues, and also inform direct recommendations to policy makers.

The participation of people with disabilities in the whole research process was key to the success of both of these studies. Both used the ICF, and so measured the same typology between difficulties, impairments and health conditions. As stated in Dr. Trani’s other presentation, this approach helps to identify or screen difficulties, but doesn’t help to look at policy issues, unmet needs, aspirations, values, capabilities and increasing freedom of choice among people with disabilities. To take this forward most effectively, people with disabilities will need to participate in the research analysis, and the promotion of its findings within governments and civil society.

6.8 Effect of the SAFOD living conditions surveys on disabled people’s lives.
Speaker: Alexander Phiri, Director General, SAFOD

“Are you not embarrassed?” This is how Alex Phiri, Director General of SAFOD, the Southern Africa Federation of the Disabled, and speaking from the floor of the conference room, began his workshop presentation. “Can you answer me, are you not embarrassed that I’m not up there and I had to come down?”

Mr. Phiri was talking about the fact that the podium at the UNECA Conference Centre in Addis Ababa, where the other speakers were sitting, was not wheelchair accessible. This is not the case for the toilets, which are. “I think that this meeting, or rather our engagement here, will not be complete if we do not send a message to the UN,” he continued, “that it is not only toilets that should be made accessible for disabled people, but it’s all features in a building.”

These comments served as a good introduction to Mr. Phiri’s talk about how human rights for people with disabilities are advanced by SAFOD’s research program. SAFOD coordinates disability organizations in ten Southern African countries: Angola, Botswana, Lesotho, Malawi, Mozambique, Namibia, South Africa, Swaziland, Zambia and Zimbabwe.

Human rights promotion is a core area of their work, which is the reason why Mr. Phiri feels so strongly about pointing out issues such as the inaccessible podium. SAFOD also has programs in:

  1. Raising awareness about disabled people’s living conditions
  2. Poverty alleviation
  3. Building capacity of disabled people’s organizations
  4. Development of disabled women, children and youths with disabilities.

Most recently, SAFOD has embarked on a major research program that includes building partnerships with researchers.

The philosophy and founding principle of SAFOD is that disabled people must be at the forefront of their own development, including research. This is critical, Mr. Phiri believes, because when people represent themselves, such as demanding the improvement of this building, they raise awareness of disability issues more effectively than others who speak on their behalf. This leads directly to becoming part of governments, at which point mainstreaming begins to be understood and implemented. This, he thinks, is also a goal of the UNCRPD.

SAFOD chooses not to define disability. People who have a disability, Mr. Phiri contends, do not have a problem with it and do not need to spend their time defining it. “It is only those who are afraid of becoming disabled who want to define what it is and how you can measure it against other people,” he added.

Instead, SAFOD has run a series of representative studies on living conditions among people with disabilities, which are the result of international cooperation between SAFOD and the Norwegian Federation of Organizations of Disabled People. The Norwegian partner has mobilized money from the Norwegian government and encouraged researchers to work with them. This collaboration is noteworthy, because disabled people have the money and power to engage or disengage the researchers, something unusual even in other research on disability.

The studies collect and analyze data for the specific purpose of improving the circumstances of people with disabilities, primarily using it with governments and planners to influence their policy and practice. Without such findings, Mr. Phiri believes, DPOs can go on advocating and advocating without achieving their desired results. “At the end of this meeting,” he proposed, “we can do some quick research. Look behind the podium and see how many people here are not able to go on it. Then present this to the UN.”

SAFOD studies have so far been sponsored in Zimbabwe, Namibia, Zambia, Malawi, Mozambique, Lesotho and Swaziland, with the ongoing support of the Norwegian government. In each country, the studies give an overview and comparison of the situation for people with disabilities and non-disabled people. Personally, Mr. Phiri does not believe that people with disabilities have activity limitations, and so the studies have not been framed in this way.

Rather, he contends, activity limitation is about those who create an environment that limits disabled people. “So far,” Mr. Phiri continued, “the other presenters have not talked about what limits the disabled person – such as not being able to reach the podium. This is not my problem, it is the problem of the people who designed the building with limited thinking.” Following this orientation, the studies looked at specific questions relating to the day-to-day lives of people with disabilities. For instance, why do fewer children with disabilities than non-disabled children attend school?

Equality is a hallmark of the SAFOD research – people with disabilities participate equally as researchers, managers and subjects of the research. There may be a question of user control here, contends Mr. Phiri, who believes that people with disabilities are not truly helped by being used only as subjects for other people’s research. For this reason, he congratulated the previous speaker, Jean-Francois Trani, for also using people with disabilities as researchers.

The impact of the research program on SAFOD itself has been noticeable. First, it has given them and their partner organizations valuable experience that will empower them to conduct further research. The long-term engagement of researchers has also helped raise awareness of how other researchers can work in partnership with disabled people. SAFOD conducted training courses to promote a basic understanding of research methodology and application, so that their DPO partners begin to understand what research is about and how to apply it in various situations. The disabled researchers also earned money through their activities, which helped them economically. “A lot of capacity has been built through this process,” believes Mr. Phiri.

Practical examples of how the research results have been used include the case of Namibia, in which the government’s planning commission has asked people with disabilities to help translate the research results for the planning process. In doing this, people with disabilities are being included in planning the country’s development. In Malawi, the research has helped the government come up with inclusive policies in many areas. One negative side to the research has been raised expectation among people with disabilities that have not yet been met.

These developments, and the research results themselves, have also attracted media coverage of disability issues. Previously, newspaper stories about disability were often tucked away near the sports column, but now they are coming closer to the front pages. In Malawi, SAFOD and its partners met with the media and were asked all sorts of questions about disability, showing the journalists’ lack of understanding. One reporter, for example, asked if it is true that disabled people are short tempered. SAFOD has seen this kind of ignorance change slowly, however, and can credit the research projects and their implications for it.

The research program has also helped SAFOD and its partners inform their positions on a number of issues. Moving from research to practical issues adds credibility, Mr. Phiri feels, because they can argue on the basis of something that has been found to be true. For instance, they have used study results to raise awareness among teachers and in rural areas, head teachers have come to understand why schools do not enroll disabled children, and that children with disabilities deserve an equal education. Anyone interested in the specific results of the SAFOD studies can get figures from study reports for each country from the SAFOD website, www.SAFOD.org.

                                                                                                                                Top