Select Page
Your Perfect Assignment is Just a Click Away
We Write Custom Academic Papers

100% Original, Plagiarism Free, Customized to your instructions!




International Journal of

Environmental Research

and Public Health


Measuring Disability Inclusion: Feasibility of Using ExistingMultidimensional Poverty Data in South Africa

Marguerite Schneider 1,* and Helen Suich 2


Citation: Schneider, M.; Suich, H.

Measuring Disability Inclusion:

Feasibility of Using Existing

Multidimensional Poverty Data in

South Africa. Int. J. Environ. Res.

Public Health 2021, 18, 4431. https://

Academic Editor: Manjula Marella

Received: 9 February 2021

Accepted: 13 April 2021

Published: 22 April 2021

Publisher’s Note: MDPI stays neutral

with regard to jurisdictional claims in

published maps and institutional affil-


Copyright: © 2021 by the authors.

Licensee MDPI, Basel, Switzerland.

This article is an open access article

distributed under the terms and

conditions of the Creative Commons

Attribution (CC BY) license (https://


1 Department of Psychiatry and Mental Health, Alan J Flisher Centre for Public Mental Health,University of Cape Town, Cape Town 7700, South Africa

2 Crawford School of Public Policy, Australian National University, Canberra 2601, Australia;

* Correspondence:

Abstract: This paper presents a framework for measuring disability inclusion in order to examinethe associations between disability severity and levels of inclusion, provides an example of itsoperationalization, and assesses the feasibility of using an existing dataset to measure disabilityinclusion using this framework. Inclusion here refers to the extent to which people with disabilitiesare accepted and recognized as individuals with authority, enjoy personal relationships, participatein recreation and social activities, have appropriate living conditions, are able to make productivecontributions, and have required formal and informal support. Indicators for the operationalizationwere drawn from the Individual Deprivation Measure South Africa country study and were mappedon to the domains of inclusion (where relevant), and the Washington Group Short Set of questionswere used to determine disability status (no, mild, or moderate/severe disability). The analysisindicates that individuals with disabilities experience generally worse outcomes and a comparativelack of inclusion compared to individuals without disabilities, and broadly that those with moderateor severe disabilities experience worse outcomes than those with mild disabilities. This analysisalso provides insight into the limitations of using existing datasets for different purposes from theiroriginal design.

Keywords: measurement; multidimensional poverty; disability inclusion; Washington Group onDisability Statistics; South Africa

1. Introduction

Social inclusion is a standard and value that underpins all disability policies and pro-grams to reverse the marginalization and related disadvantage of people with disabilities.Social inclusion has been described in the context of disadvantage generally as being com-plex and multidimensional [1], though descriptions range from a narrow conceptualizationof simple economic inclusion [2] through to a broader conceptualization of social, cultural,economic, and political inclusion where conditions are such that individuals and groupsare able to take part in society [3]. This literature on social inclusion is extensive but doesnot seem to effectively address disability as a factor that leads to exclusion. Furthermore,the disability inclusion literature is generally focused on small-scale studies, which areoften qualitative in nature, generating a gap in how we can measure inclusion at scale—forexample, to provide national-level statistics.

The themes in the broad social inclusion literature are, nevertheless, reflected to somedegree in the literature on inclusion of people with disabilities. Hall sets out three keyelements of social inclusion for people with disabilities, on the basis of her meta-analysis ofthe use of social inclusion in qualitative studies on disability. These are involvement in ac-tivities, maintaining reciprocal relationships, and a sense of belonging [4]. Ikäheimo furtheridentifies institutional status and interpersonal status as separate features of personhood.Institutional status refers to the obligations of states towards their citizens, with respect

Int. J. Environ. Res. Public Health 2021, 18, 4431.

Int. J. Environ. Res. Public Health 2021, 18, 4431 2 of 22

to enabling the right to life, health, and education, and being counted as a human being.One clear example is whether a person has been issued with a formal identity documentor birth certificate, which can enable them to access public services necessary to accessingequal opportunities. Interpersonal status refers to ‘being seen as a person by others’ ([5]p. 79), through being seen to have authority and a recognized claim to happiness, and tobe contributing something worthy. Being seen as person ensures a person with disabili-ties is included into the ‘us’ by society and not dismissed as ‘them’—a group apart frommainstream society.

It is well documented that people with disabilities are often excluded from educationaland employment opportunities [6], and that accessible transport and public buildings, andavailability of assistive devices and person assistance play key roles in facilitating inclu-sion [7]. The UN’s flagship ‘Disability and Development Report’ [6] highlights the commonunderstanding that stigma and discrimination are key factors in hindering inclusion forpeople with disabilities. What is less extensively documented is the extent to which otheraspects of inclusion are experienced, such as being respected and valued, having sufficientaccess to food and accommodation, and having a voice through being able to vote freely.

Hall identified six themes relating to inclusion from her review of 15 primary qualita-tive studies: (i) being accepted, (ii) having relationships, and (iii) involvement in activitiessupported by (iv) adequate living accommodations (e.g., including accessible dwellings, ac-cess to water, energy, and sanitation) and (v) support systems (e.g., formal service providersupport and less formal support from family, friends, and community) and reflected in(vi) employment [4]. The ability to make choices and being part of and being seen in arange of social contexts (spaces and events including in employment and education) wasalso noted as being key to a person with disabilities feeling included [8,9]. Being givenopportunities to reciprocate, being valued, and being expected to contribute to, for example,community events, foster agency and recognition and contribute to making a person feelreal [9,10]. Having a voice involves being able to vote, being part of decision making at apersonal level and within local activities, and being heard [8,10,11]. While a number of thestudies in Hall’s review were conducted in high-income countries, these concepts remainrelevant for any context, although the way they are realized may differ by cultural andgeographical context.

In their review of social inclusion and people with disabilities in West Africa, Jolleyet al. conclude that measures of disability and social inclusion need to be developed andadopted to allow for a more coordinated monitoring of social inclusion [12]. Collectingdata through surveys is an important component of monitoring of inclusion as it providesan indication of the extent of inclusion across the broad target population and can recordpositive or negative changes over time. However, these data are only valid if measurementaccurately reflects the notion of inclusion.

On the basis of this review of the disability studies literature, we propose a frameworkof the key domains of inclusion that should be included in large-scale studies measuringlevels of inclusion of people with disabilities. We then provide an example of the oper-ationalization of elements of this framework examining associations between disabilityseverity and levels of inclusion using existing data from the Individual Deprivation Mea-sure (IDM) South Africa Country Study. Given the financial and time costs associatedwith large data collection exercises, it is useful to identify existing datasets that allow us toidentify indicators of inclusion to monitor these at the population level, for example forreporting on the Sustainable Development Goals, and we thus use the IDM South Africadata to assess the feasibility of using existing datasets to measure inclusion.

By using existing data to measure inclusion, we can gain useful insights not only onthe associations between inclusion and disability severity in South Africa, but also on issuesaround utilizing data originally designed to measure something different (in this case,levels of individual deprivation). Poverty is broadly understood as an inability to achieve asocially acceptable standard of living across multiple dimensions. The IDM was designedto measure individual-level deprivation, capable of revealing gender disparities, across 14

Int. J. Environ. Res. Public Health 2021, 18, 4431 3 of 22

economic and social dimensions. Inclusion is a different concept, addressing the extentto which people are included in various aspects of life (or not), and is somewhat morerelational, explicitly incorporating issues around belonging and personhood, and includingwider social, political, cultural, and attitudinal forms and arenas of exclusion [13,14].Poverty is often described as an outcome of exclusion but also as a cause, and poorpeople with disabilities can be included or excluded as much as non-poor people withdisabilities. The two concepts are overlapping but not the same [14,15], which is why thecoverage in the IDM dataset of some domains and indicators of importance to inclusion isincomplete. Thus, while the IDM dataset includes many aspects that are relevant to themeasurement of inclusion, our analysis can also inform us about what data remains missingand how measurement tools could be designed to measure inclusion more effectively andcomprehensively at a large scale.

2. Materials and Methods2.1. A Framework for Measuring Disability Inclusion

The framework presented in this paper was developed and adapted primarily fromthe key elements of inclusion as identified in the literature review—8 domains derivedfrom the literature review and a further 3 domains that we can include specifically becauseof the availability of data from the IDM that is known to be of importance to peoplewith disabilities.

Interpersonal status refers to being seen as a person by others, as worthy of respect,being treated with dignity, and seen as contributing value to the household (this domainis called ‘being accepted’ by Hall and ‘interpersonal status’ by Ikäheimo). Personal rela-tionships are about relationships with family and friends and being able to reciprocate as asign that what one offers is of value (i.e., contributing to agency and recognition), free fromstigma and discrimination (Hall’s ‘having relationships, also overlapping with Ikäheimo’sinterpersonal status). Being involved entails having wider social relationships with thebroader community, beyond friends and family (Hall’s involvement in activities). Livingconditions refers to the basic needs of individuals being met with dignity and withoutshame (Hall’s adequate living accommodations). Economic opportunity and contributionsinclude a range of economic and other contributions (i.e., paid and unpaid activities) madeby the individual to the household and its operation. This framework includes a widerrange of contributions than is typically recognized, and which is usually measured as‘employment’ only (e.g., in Hall). Support systems include formal support received fromservice providers and is usually provided at a cost, and informal support which is usuallyunpaid and provided by family, friends, and/or community (Hall’s support systems). Insti-tutional status refers to the formal recognition of the individual by the state/government,which may affect whether states’ obligations towards their citizens can be met (e.g., socialassistance/welfare, health, education) (Ikäheimo’s domain of institutional status). Forexample, in South Africa a social assistance grant is only provided to people with a formalidentity document. Voice is about participation in decision making processes affecting theindividual (Ikäheimo’s domain of interpersonal status and being seen as worthy).

Three additional domains are included because they can be populated with data fromthe IDM South Africa Country Study, and they reflect domains known to be of importanceto inclusion generally, as well as to people with disabilities more specifically, underpinningindividuals’ ability to be involved in activities and employment (for example). Educationentails the formal education received and its quality, while healthcare access refers toboth the services accessed (if needed) and their quality. Finally, personal safety refers toindividuals’ safety from threats or hazards while undertaking household activities, andtheir perceptions of their own safety in and around their home.

The individual indicators measured are presented in Section 2.4.

Int. J. Environ. Res. Public Health 2021, 18, 4431 4 of 22

2.2. The IDM South Africa Country Study

The IDM recognizes that poverty is experienced at an individual-level, is gender-sensitive, and is multidimensional, and as such, measures it in this way. The 14 dimensionsof poverty that the IDM measures were identified largely by participatory work undertakenacross 6 countries [16]. Data collection for the IDM South Africa Country Study wasundertaken in 2019, and the data from the main national sample, of 8652 individuals,were used in this analysis. For this sample, enumeration areas (EAs) were randomlyselected across all 9 provinces, stratified by rural/urban locality. All dwellings within theselected EAs were identified by using satellite imagery to remote sense roofs and were thenrandomly sampled [17].

Individuals eligible to be interviewed were all of those living in a sampled dwellingwho were 16 years and older, able to communicate for themselves, and who were competentto give informed and ongoing consent during the interview. Each individual interviewwas undertaken with an enumerator of the same gender, in privacy, and in the preferredlanguage of the respondent (choosing from any of the 11 official spoken languages of SouthAfrica). The individual interviews collected data on 14 dimensions of deprivation, ondemographic characteristics, and on functioning difficulties, and took, on average, 44 min.Suich et al. provides detailed information about the country study implementation andresults [18].

2.3. Disability Definition and Measurement

The IDM individual survey tool used the questions in the Washington Group onDisability Statistics Short Set on Functioning (WG SS) of as the measure of disabilitystatus [19]. These questions identify people at risk of experiencing disability and its relateddisadvantages by measuring difficulties people have in various basic activities of seeing,hearing, remembering and concentrating, walking and climbing stairs, communication,and self-care. In this paper, 3 categories of disability severity are determined. The firstcategory is people with no disabilities, who reported ‘no difficulty’ on any functioning or‘some difficulty’ for only one activity—6952 individuals. The second category is peoplewith ‘mild disabilities’—812 individuals reporting ‘some difficulty’ for between 2 and 4of the 6 activities but not reporting ‘a lot of difficulty’ or ‘cannot do at all’ for any. Thethird category is people with moderate or severe disabilities (‘moderate’)—888 individualsreporting ‘a lot of difficulty’ or ‘cannot do at all’ for at least 1 of the 6 activities, or ‘somedifficulty’ for 5 or more of the 6 activities. This is a slightly modified version of therecommended cutoff for disability statistics reporting [20]. We have included the categoryof 5 or more ‘some difficulty’ responses as indicating a moderate to severe disability.

The demographic characteristics of the sample are presented in Table 1 for each ofthese 3 disability severity categories and the whole sample. These data were not weightedto be fully representative of the South African population, and therefore caution should beexercised in making inferences from these results to the wider South African population.

Int. J. Environ. Res. Public Health 2021, 18, 4431 5 of 22

Table 1. Demographic characteristics in terms of disability status from the main study (%).

None Mild Moderate Overall

Total cases 6952 812 888 8652


Male 46.6 34.6 24.3 43.2

Female 53.4 65.4 75.7 56.8


16–24 28.2 9.7 6.9 24.3

25–64 66.1 57.3 59.1 64.6

65+ 5.7 33 34 11.2

Population group

Black African 84.0 83.9 83.3 83.9

Coloured 12.7 12.9 14.6 12.9

Indian or Asian 1.1 1.1 0.9 1.0

White 2.2 2.1 1.1 2.1

Educational completion

Matriculation orhigher

45.7 23.6 20.4 41.0

Some secondaryschooling

39.0 33.9 30.6 37.7

Primary or less 15.3 42.5 49.0 21.3

Few individuals interviewed stated that the functioning difficulties they reportedwere present at birth (Table 2), in particular for those over 65 years—most of whom appearto be experiencing age-related functioning difficulties.

Table 2. Age of onset of disability for those with mild and moderate disabilities in terms of genderand age (%).

Male Female

16–24 25–64 65+ 16–24 25–64 65+

At birth 11.3 4.3 0 3.4 2.8 1.0

Childhood/school age 64.2 13.2 1.2 63.2 9.2 0.7

Early adulthood (±18–29) 22.6 17.4 2.5 32.2 14.7 3.2Later (30+) 1.9 65.1 96.3 1.1 73.2 95.1

Total cases 53 281 161 87 706 409

2.4. Selection of Indicators to Populate the Framework

The proposed framework describes the domains of inclusion identified as importantfor the measurement of inclusion, and Table 3 describes these domains and the indicatorsselected within each of the domains, as drawn from the IDM South Africa Country Study.The questions used in the IDM individual survey tool were developed specifically tomeasure deprivation at the individual level and therefore do not overlap exactly withmeasures of inclusion; thus, some domains can be more comprehensively illustratedusing the IDM data than others in the reported results. For example, in the interpersonalstatus domain, 2 indicators were selected, referring to respect and value associated withunpaid domestic and care work. A specifically designed inclusion measure would includeadditional aspects of this domain, such as being addressed directly rather than through

Int. J. Environ. Res. Public Health 2021, 18, 4431 6 of 22

one’s personal assistant or carer and being listened to when contributing to a discussion,which are not available in the IDM data. Other domains include a range of indicatorsrarely measured elsewhere. For example, the living conditions domain includes basicclothing and footwear ownership and quality, as well as ownership of bedding materials,which have rarely been measured elsewhere, but can make important contributions toindividuals’ dignity. As noted above, we expanded the conceptualization of what istypically identified as the domain of ‘employment’, which typically refers only narrowlyto whether an individual is in paid employment. A wider range of economic and othercontributions can be determined using the IDM data—including contributions necessaryto the running of a household and which may enable other members of the householdto undertake paid employment or income-generating activities [21,22], and which arerecognized as an important component of household economies. As a result, we renamedthis domain ‘economic opportunities and contributions’.

Table 3. Domains of inclusion and indicators used to measure aspects of those domains (from the Individual DeprivationMeasure (IDM) dataset).

Domain of InclusionIndicator (Drawn from the IDM South

Africa Country Study)Concept Measured

Interpersonal statusUnpaid domestic and care work humiliation

Whether the respondent was subject to humiliatingtreatment while doing unpaid domestic and care work

Unpaid domestic and care work valueWhether household members of the respondent value

the unpaid domestic and care work they do

Personal relationships Ability to reciprocate support Frequency and ability to reciprocate support received

Being involved Community event inclusion (#)Frequency and ability to participate in community

events (e.g., religious activities, ceremonies, or festivals)

Living conditions

Food security (#) *Degree of food (in)security in not having sufficient and

nutritious food

Drinking water source and reliability (#)Type (improved/unimproved) and reliability of

drinking water source

Domestic water source and reliability (#)Type (improved/unimproved) and reliability of

domestic water source

Cooking energy source and reliability (#)Type (clean/polluting) and reliability of cooking

energy source

Lighting energy source and reliability (#)Type (clean/polluting) and reliability of heating

energy source

Heating energy source and reliability (#)Type (clean/polluting) and reliability of heating

energy source

Home toilet facilities (#) Type of toilet facility (improved/unimproved)

Toilet modificationsIf toilet facility is (partially) modified to accommodate

physical needs

Basic clothing and footwear ownership (#) *Ownership of two complete sets of basic clothing

and footwear

Basic acceptability and protection (#) *Acceptability and protection of basic clothing

and footwear

Bedding ownership Ownership of sufficient bedding materials

Eviction concern Fear of eviction from accommodation

Int. J. Environ. Res. Public Health 2021, 18, 4431 7 of 22

Table 3. Cont.

Domain of InclusionIndicator (Drawn from the IDM South

Africa Country Study)Concept Measured

Economic opportunityand contributions

Labor force status (#) Labor force status (all respondents)

Unpaid domestic and care work Unpaid domestic and care work undertaken at home

Fuel collection responsibilityResponsibility for collecting fuel sources from outside

the home (if necessary)

Water collection responsibilityResponsibility for collecting water sources from outside

the home (if necessary)

On call time (#)Responsibility for caring for a child under 13 and/or a

sick, elderly, or disabled person (the previous day)

Public transport availability andaffordability (#)

Availability and affordability of public transport

Support systems(formal and informal


Support availabilityNeed for help to meet basic needs and support and

frequency of receiving it

Old-age pension Receipt of an old-age pension (if relevant)

Disability grant Receipt of a disability grant (if relevant)

Institutional statusIdentity document

Current possession of a South Africanidentity document

Birth certificate Current possession of a birth certificate

VoiceLocal decision-making inclusion (#)

Frequency and ability to participate in localdecision making

Voting inclusion Freedom to vote and vote freely


Educational completion Level of education achieved

Basic literacy (#)Functional literacy (ability to read and to write in an

official language)

Basic numeracy (#)Functional numeracy (ability to complete two simple

mathematical calculations)

Healthcare access

Healthcare access (#) Accessed healthcare in South Africa (or reason why not)

Healthcare communicationCommunication difficulties associated with

health treatment

Respectful treatment Respectful treatment from healthcare workers

Personal safety

Fuel collection hazardsExperienced hazards/threats while collecting fuel

outside the home (for those responsible)

Water collection hazardsExperienced hazards/threats while collecting water

outside the home (for those responsible)

Safety in the neighborhoodPerceived safety of walking alone in the neighborhood

after dark

Safety at home Perceived safety of being at home alone after dark

(#) indicator constructed from more than one survey question. * an indicator constructed using the IDM scoring methodology [23].

2.5. Data Analysis

The results below are presented as the proportion of the sample that falls into eachresponse category. All data in the tables are reported only for those who provided a relevantresponse to each question, excluding those for whom a question may not be relevant (e.g.,those who did not do any unpaid domestic and care work were not asked whether that workwas valued by household members). Those few individuals who refused to answer anyone question were also excluded; of the 75 survey questions used to create these indicators,71 had refusal rates of less than 1%. The four questions with refusal rates exceeding

Int. J. Environ. Res. Public Health 2021, 18, 4431 8 of 22

1% were those assessing functional literacy and numeracy, which asked respondents tocomplete reading and writing tasks and 2 mathematical calculations, for which refusal toanswer ranged between 3.8 and 7.1%. Exact p-values from chi-squared tests are reported inthe Results section tables in order to indicate where statistically significant differences occurbetween subgroups, using Benjamini–Hochberg adjustments for multiple simultaneouscomparisons to control for the false discovery rate (i.e., the number of false-positive resultsincreasing with the number of tests).

3. Results3.1. Interpersonal Status

The two indicators in the interpersonal status domain relate solely to respect associatedwith unpaid domestic and care work (and are only asked of those individuals who reporteddoing this type of work). As shown in Table 4, there were no statistically significantdifferences found between the three groups in the levels of respect received.

Table 4. Interpersonal status domain indicators in terms of disability status.

Variable Level None Mild Moderate p-Value Overall

Humiliating treatmentwhilst carrying out unpaid

domestic and care workNo 97.2 96.2 95.9 0.08942 97.0

Yes 2.8 3.8 4.1 3.0

6022 655 708 7385

Unpaid domestic and carework valued by household

membersYes 87.0 87.6 88.7 0.41056 87.2

No 13.0 12.4 11.3 12.8

6039 654 708 7401

3.2. Personal Relationships

The data for the personal relationships domain, presented in Table 5, show the ex-pected pattern—the proportion of individuals who were able to reciprocate ‘only sometimes’and ‘never’ was higher for those with mild and moderate disabilities than for individualswith no disabilities.

Table 5. Personal relationships domain indicator in terms of disability status.

Variable None Mild Moderate p-Value Overall

ReciprocationCan reciprocate always

or most of the time43.5 39.9 37.0 0.00031 42.5

Can reciprocate onlysome of the time

or never56.5 60.1 63.0 57.5

6918 809 881 8608

3.3. Being Involved

Table 6 shows the results for the being involved domain. Relatively low levels ofall three groups were able to attend community events always or most of the time. Theproportion of individuals who reported feeling excluded from these events increased withthe severity of disability.

Int. J. Environ. Res. Public Health 2021, 18, 4431 9 of 22

Table 6. Participation in community events domain indicator in terms of disability status.

Variable Level None Mild Moderate p-Value Overall

Participation incommunity events

Always or sometimes attendedcommunity events OR there

was no event to attend56.7 58.3 46.6 0.00001 55.9

Rarely attended OR did notattend because individual was

too busy, too sick, or notinterested in doing so

32.6 27.8 34.4 32.3

Rarely or never attendedbecause they were prevented

from doing so or excluded10.7 13.9 19.0 11.8

6885 803 879 8567

3.4. Living Conditions

The living conditions domain had the largest number of indicators of any domainpresented in this paper, and the results for each can be seen in Table 7. For all but three ofthe indicators, the worst outcomes were most frequently realized by those with moderatedisabilities, followed by those with mild disabilities and the best outcomes most likely forindividuals with no disabilities—across food, water and energy security, basic clothing andfootwear, and bedding.

Table 7. Living conditions domain indicators in terms of disability status.

Variable Level None Mild Moderate p-Value Overall

Food security Food secure 39.1 26.6 22.2 0.00001 36.2

Mild or moderate food insecurity 20.9 18.3 14.5 20.0

Severe food insecurity 39.9 55.0 63.3 43.7

6952 812 888 8652

Drinking waterImproved drinking water and enough to meet

needs always or most of the time80.8 72.0 69.7 0.00001 78.8

Improved drinking water and enough to meetneeds some of the time or never OR unimproveddrinking water and enough to meet needs always

or most of the time

17.2 24.8 27.3 19.0

Unimproved drinking water source and enoughto meet needs some of the time or never

2.0 3.2 3.0 2.2

6950 812 888 8650

Domestic waterImproved domestic water and enough to meet

needs always or most of the time82.2 72.0 68.4 0.00001 79.8

Improved domestic water and enough to meetneeds some of the time or never OR unimproved

domestic water and enough to meet needsalways or most of the time

17.6 27.8 31.5 20.0

Unimproved domestic water source and enoughto meet needs some of the time or never

0.2 0.1 0.1 0.2

6948 812 887 8647

Int. J. Environ. Res. Public Health 2021, 18, 4431 10 of 22

Table 7. Cont.

Variable Level None Mild Moderate p-Value Overall

Cooking energyClean cooking energy and enough to meet needs

always or most of the time68.6 59.2 50.1 0.00001 65.8

Clean cooking energy and enough to meet needssome of the time or never OR unclean cooking

energy and enough to meet needs always or mostof the time

24.4 31.2 42.5 26.9

Unclean cooking energy and enough to meetneeds some of the time or never

7.0 9.6 7.4 7.3

6948 812 887 8647

Lighting energyClean lighting energy and enough to meet needs

always or most of the time78.9 66.4 58.0 0.00001 75.6

Clean lighting energy and enough to meet needssome of the time or never OR unclean lighting

energy and enough to meet needs always or mostof the time

18.2 29.6 40.1 21.5

Unclean lighting energy and enough to meetneeds some of the time or never

3.0 4.1 1.9 3.0

6947 812 888 8647

Heating energyClean heating energy and enough to meet needs

always or most of the time57.7 54.3 43.0 0.00001 55.9

Clean heating energy and enough to meet needssome of the time or never OR unclean heating

energy and enough to meet needs always or mostof the time

23.3 25.7 33.9 24.6

Unclean heating energy and enough to meetneeds some of the time or never

19.0 20.0 23.1 19.5

6924 806 883 8613

Toilet facility (athome)

Improved toilet facility 85.9 82.9 83.6 0.0166 85.3

Unimproved toilet facility 12.3 15.8 14.8 12.9

No toilet facility 1.9 1.4 1.6 1.8

6951 811 885 8647

Modified toiletfacility

Modified to accommodate physical needs 73.8 60.6 71.4 0.00001 69.0

Partly modified to accommodate physical needs 5.9 11.2 6.5 7.7

Not modified 20.4 28.1 22.1 23.3

904 747 802 2453

Ownership ofbasic clothingand footwear

Own two changes of clothes and two pairs offootwear

86.8 79.2 73.5 0.00001 84.7

Do not own two changes of clothes and two pairsof footwear

13.2 20.8 26.5 15.3

6952 812 888 8652

Basic clothingand footwear


Basic clothing is acceptable always OR most ofthe time AND protection is excellent OR good

71.7 58.5 52.0 0.00001 68.5

Int. J. Environ. Res. Public Health 2021, 18, 4431 11 of 22

Table 7. Cont.

Variable Level None Mild Moderate p-Value Overall

Basic clothing ownership is always OR most ofthe time AND provides some or no protection;

basic clothing is acceptable some of the time ORnever AND protection is excellent OR good

17.9 24.6 25.8 19.3

Basic clothing is acceptable some of the time ORnever AND provides some or no protection

10.4 16.9 22.2 12.2

6952 812 888 8652

Ownership ofsufficientbedding

Own sufficient bedding to sleep comfortably 78.9 70.3 65.3 0.00001 76.7

Do not own sufficient bedding to sleepcomfortably

21.1 29.7 34.7 23.3

6943 811 887 8641

Fear of eviction Did not fear eviction 90.5 90.1 87.3 0.00926 90.2

Feared eviction 9.5 9.9 12.7 9.8

6944 808 888 8640

There were three differences from this pattern—higher proportions of those withmoderate disabilities than those with mild or no disabilities feared eviction, there wereno differences between the three groups with respect to the type of toilet facilities usedat home, and a smaller proportion of those with mild disabilities had a modified toilet(compared to those with moderate and no disabilities).

Several issues can be observed from these data. The first is the high levels of foodinsecurity amongst all three groups—of the overall sample, only 36.2% are food-secure,and severe food insecurity was found to worsen significantly with increasing severityof disability. What is not available in these results is a breakdown of food security forindividuals within a household to see if there are intra-household differences which couldbe explained by disability severity.

Ownership and quality of clothing and footwear are rarely measured, but speakto issues of dignity, respect, and opportunity, and these data illustrate that as disabilityseverity increases, there is a declining proportion of those who own the most basic levelsof clothing and footwear, and a decline in the quality of that basic clothing and footwear.

The outcomes for bedding ownership were also found to decline with increasingseverity of disability—one in five of those with no disabilities reported not having enoughbedding to sleep comfortably (21.1%), compared to 3 in 10 of those with mild disabilities(29.7%) and slightly more for those with moderate disabilities (34.7%).

3.5. Economic Opportunity and Contributions

The results for each of the indicators in the economic opportunity and contribution do-main, including labor force status and the various indicators demonstrating contributionsto the household, can be seen in Table 8.

Int. J. Environ. Res. Public Health 2021, 18, 4431 12 of 22

Table 8. Economic opportunity and contributions domain indicators in terms of disability status.

Variable Level None Mild Moderate p-Value Overall

Labor forcestatus

Employed 39.2 24.7 20.8 0.00001 35.9

Unemployed 27.0 15.8 13.8 24.6

Not in the labor force (choice) 24.0 45.7 40.7 27.8

Not in the labor force (exclusion) 9.8 13.7 24.7 11.7

6915 809 886 8610

TransportPublic/mass transport available and affordable

always or most of the time43.5 32.6 38.3 0.00001 41.9

Public/mass transport available and affordablesome of the time OR some public/mass transportwhich is affordable always or most of the time

35.2 37.0 33.0 35.2

Some public/mass transport which is affordablesome of the time or never OR no public/mass

transport available21.3 30.5 28.8 22.9

6945 811 886 8642

Unpaiddomestic and

care workPerformed unpaid domestic and care work 87.2 80.9 80.3 0.00001 85.9

No unpaid domestic and care work 12.8 19.1 19.7 14.1

6942 812 884 8638

On-callSpent time responsible for a child under 13and/or a sick, elderly, or disabled person

46.6 51.5 50.8 0.00404 47.5

Did not spend time responsible for a child under13 and/or a sick, elderly, or disabled person

53.4 48.5 49.2 52.5

6952 812 888 8652

Fuel collectionresponsibility

Collected fuel from outside the dwelling 69.7 68.2 59.7 0.00001 68.5

Did not collect fuel from outside the dwelling 30.3 31.8 40.3 31.5

6939 812 885 8636

Water collectionresponsibility

Collected water from outside the dwelling 58.5 62.1 60.0 0.12471 59.0

Did not collect water from outside the dwelling 41.5 37.9 40.0 41.0

6947 810 887 8644

In understanding employment status using these data, one must remember that allrespondents—regardless of age—were given a categorization for this indicator. That is,even those 65 years and older were included, which is unlike most other statistics dealingwith employment status and labor force participation. The category ‘not in the laborforce (by choice)’ includes all individuals of legal working age (i.e., younger than 65) whoreported being in full time education, individuals of any age who reported they did notneed or want to work, and individuals 65 years and older who were retired. The categorynot in the labor force (exclusion) included individuals who reported not being in the laborforce because they were too busy with domestic/caring duties, were unable to work, orwere ‘discouraged workers’ (i.e., had given up looking for work because it was too hardto find).

The results show the anticipated pattern—with the lowest rates of exclusion from thelabor force amongst those without disabilities and the highest rates amongst those withmoderate to severe disabilities. It is not clear why higher proportions of those without

Int. J. Environ. Res. Public Health 2021, 18, 4431 13 of 22

disabilities are unemployed, with the proportion declining progressively with disabilityseverity; perhaps those with disabilities are more discouraged (and thus categorized asnot in the labor force by exclusion). For the transport indicator, the worst outcomesare experienced by those with mild disabilities, and the best outcomes by those withno disabilities.

The indicators assessing contributions to the household examine whether individualsdo unpaid domestic and care work within the household by whether they spent timethe previous day on-call (i.e., responsible for a child under 13, or for caring for a sick,disabled, or elderly person) and whether they collected fuel sources and/or water forthe household (for those that relied on sources not delivered to the dwelling). Overall,these indicators show the important contributions made by people with disabilities to theday-to-day running of households.

Though the patterns are different for each indicator, it can be seen that significantcontributions to the household are being made by all three disability status groups—withapproximately half or more of each group undertaking these four activities. Those with nodisabilities were found to have the highest proportion doing unpaid care and domestic workand collecting fuel/energy from outside the dwelling, with slightly lower proportions forthose with mild and moderate disabilities. The pattern reversed for on-call time (with thehighest proportion being those with moderate disabilities), and there were no differencesbetween the three groups in the collection of water from outside the dwelling.

3.6. Support Systems (Formal and Informal Support)

The first two indicators of the support systems domain can be seen in Table 9. Thethird—which examined individuals’ receipt of the old-age and disability social assistancegrants—were found to be further disaggregated by disability status and age group, as seenin Table 10.

Table 9. Support systems domain indicators in terms of disability status.

Variable Level None Mild Moderate p-Value Overall

SupportDo not need support OR do need

support, but get enough always ormost of the time

89.2 80.1 74.4 0.00001 86.9

Need support, and have enough onlysome of the time or never

10.8 19.9 25.6 13.1

6931 809 884 8624

Carer in household Yes 32.6 40.4 43.6 0.00001 37.8

No 67.4 59.6 56.4 62.2

1360 812 888 3060

Just over 1 in 10 individuals without disabilities require support but do not receiveenough of it (10.8%), compared to 2 in 10 with mild disabilities (19.9%), and one-quarter ofthose with moderate disabilities (25.6%). Just under half of those with mild and moderatedisabilities reported having a carer in the household (40.4% and 43.6%, respectively).

Table 10 highlights the relatively low proportion of individuals reporting functioningdifficulties using the WG SS but receiving a disability grant. Note that eligibility rules forsocial grants mean that all respondents over 60 should receive an old-age pension ratherthan a disability grant, which suggests a far more effective distribution.

Int. J. Environ. Res. Public Health 2021, 18, 4431 14 of 22

Table 10. Social assistance domain indicator in terms of disability status and age group.

None Mild Moderate

16–24 25–64 65+ 16–24 25–64 65+ 16–24 25–64 65+

Old-age grant

No 100 96.2 9.9 100 81.5 8.6 100 84.1 5.6

Yes – 3.8 90.1 – 18.5 91.4 – 14.9 94.4


1961 4596 395 79 465 268 61 525 302

Disability grant

No 99.6 98.8 99.5 98.7 94.2 98.5 93.4 85.1 99.0

Yes 0.4 1.2 0.5 * 1.3 5.8 1.5 * 6.6 13.9 * 1.0


1961 4596 395 79 465 268 61 525 302

* It is likely that these individuals (who are not eligible for a disability grant) misreported the specific type of social assistance they received.On turning 60 years of age, recipients of the disability grant are automatically transferred to an old-age grant.

3.7. Institutional Status

The indicators of the institutional status domain are current possession of a SouthAfrican identity document (ID) and/or of a birth certificate, as shown in Table 11. The mostaccurate interpretation of these indicators is representing ‘current’ institutional personhood,as the questions asked whether individuals possessed an ID document or birth certificate atthe time of the survey (i.e., not whether they had ever had one, even if they did not possessone in 2019). The two indicators for this domain move in opposite directions—those withdisabilities are more likely to have an ID document, but far less likely to currently be inpossession of a birth certificate. This is partly age-related (i.e., older people are less likely tocurrently have a birth certificate) and likely also because of difficulties with secure storageover the long term.

Table 11. Institutional personhood domain indicators in terms of disability status.

Variable Level None Mild Moderate p-Value Overall

Identity documentCurrent possession of a South African

identity document89.0 95.7 97.0 0.00001 90.4

No current possession of a South Africanidentity document

11.0 4.3 3.0 9.6

6950 811 887 8648

Birth certificate Current possession of a birth certificate 66.2 48.6 44.8 0.00001 62.4

No current possession of a birth certificate 33.8 51.4 55.2 37.6

6761 774 849 8384

3.8. Voice

The two indicators of the voice domain are shown in Table 12, and the results forboth indicators ran in the opposite direction to virtually all other indicators and domainsconsidered. Unexpectedly, individuals with mild and moderate disabilities were found tohave better outcomes in both indicators than do individuals with no disabilities. However,there were similar proportions (close to one-quarter) of each group that were excludedfrom participating in local decision making and from voting freely. This could be a particu-larity of the South African party political context and the strength of the disability rightsmovement in South Africa that ensured that disability was specifically included in theSouth African Constitution [24].

Int. J. Environ. Res. Public Health 2021, 18, 4431 15 of 22

Table 12. Voice domain indicators in terms of disability status.

Variable Level None Mild Moderate p-Value Overall

Participating in localdecision making

Participated in local decision making 26.6 36.7 33.6 0.00001 28.3

Did not participate because too busy; notinterested; no process to participate in

46.3 37.3 41.1 45.0

Did not participate because excluded (notinvited; afraid/uncomfortable; do not

trust the leaders; not appropriate for me)27.1 26.0 25.3 26.8

6872 799 881 8552

Voting freely Voted and free to choose who to vote for 53.3 66.7 69.6 0.00001 56.2

Did not vote, because not interested, notold enough to vote, OR not a citizen

17.4 9.2 6.7 15.5

Voted but not free to choose who to votefor OR did not vote, for all other reasons

29.3 24.2 23.8 28.3

6915 807 887 8609

3.9. Education

The differences between the three groups in education levels can be seen in Table 1—those with no disabilities were found to be more likely to have higher levels of education,whilst there were very large proportions of those with mild and moderate disabilities whowere found to have completed only primary school or less. The remaining indicators of theeducation domain are functional literacy and functional numeracy, as shown in Table 13.

Table 13. Education domain indicators in terms of disability status.

Variable Level None Mild Moderate p-Value Overall

Functional literacyAble to read and write to a basic level in

an official language77.9 50.9 48.3 0.00001 72.3

Able to read or write to a basic level in anofficial language

16.1 22.7 24.5 17.6

Not able to read or write to a basic levelin an official language

6.0 26.4 27.2 10.1

6542 770 845 8157

Functional numeracyAble to correctly answer two mathematics

problems67.0 41.8 41.8 0.00001 62.2

Able to answer onemathematical problem

20.8 22.8 23.2 21.3

Unable to answer mathematical problems 12.1 35.4 35.0 16.5

6506 732 801 8039

Far higher proportions of individuals with no disabilities are classified as functionallyliterate or functionally numerate than those with mild or moderate disabilities. Note thatone of the eligibility criteria for participating in the IDM survey was to be able answerquestions for themselves, which is likely to bias this sample toward those with less ratherthan more severe disabilities. Note also that there are important (negative) correlationsbetween functional literacy and numeracy and age, which are related—at least in part—tothe lack of education and educational quality received by older citizens.

Int. J. Environ. Res. Public Health 2021, 18, 4431 16 of 22

3.10. Healthcare Access

The indicators used in the healthcare access domain are shown in Table 14, whichrelate to accessing healthcare (excluding prenatal and birthing care) in South Africa inthe 12 months prior to the survey (or the reasons why not), and measures of respectfultreatment and communication difficulties.

Table 14. Healthcare domain indicators in terms of disability status.

Variable Level None Mild Moderate p-Value Overall

Sought healthcare inprevious 12 months

(in RSA)No 46.9 26.0 27.7 0.00001 43.0

Yes 53.1 74.0 72.3 57.0

6944 812 887 8643

Reason for not seekinghealthcare

Did not need or want 98.4 91.0 93.9 0.00001 97.7

Excluded 1.6 9.0 6.1 2.3

3249 211 245 3705

Respectful treatment Received respectful treatment 92.6 92.7 88.7 0.00352 92.1

Did not receive respectful treatment 7.4 7.3 11.3 7.9

3686 600 639 4925


No communication difficulties withhealthcare provider

93.8 91.7 91.4 0.0212 93.2

Communication difficulties withhealthcare provider

6.2 8.3 8.6 6.8

3686 601 639 4926

The differences between the three groups were found to be significant for three of thefour indicators. Far higher proportions of those with mild and moderate disabilities soughthealthcare than those without, and of those who sought healthcare, those with moderatedisabilities were most likely to not receive respectful treatment, although the numberswere small. There were no significant differences found between the groups with respectto problems with communication with healthcare professionals, though a small minoritydid experience this problem.

Of those who did not access healthcare, individuals with mild and moderate disabil-ities were more likely to have reported feeling excluded from accessing healthcare thanthose with no disabilities—whether because healthcare was too costly, was too far away,there was no transport to get there, the respondent was too embarrassed to seek healthcare,the provider refused to treat the individual, or they felt vulnerable to discrimination. Thenumber of excluded respondents were too small to analyze separately.

3.11. Personal Safety

The four indicators in the personal safety domain are shown in Table 15. For threeof the four indicators in this domain, outcomes deteriorated with increasing severity ofdisability, as for most other domains, with those with moderate disabilities experiencingoverall the lowest levels of personal safety (there was no difference between the threegroups for fuel collection threats). Note that the figures reported for those facing hazardswhile collecting water and fuel from outside the dwelling are reported only for thoseindividuals who reported being responsible for this activity.

Int. J. Environ. Res. Public Health 2021, 18, 4431 17 of 22

Table 15. Personal safety domain indicators in terms of disability status.

Variable Level None Mild Moderate p-Value Overall

Fuel collection hazardsNo hazards while collecting fuel outside

the dwelling90.2 89.9 89.4 0.89499 90.0

Faced hazards while collecting fueloutside the dwelling

9.8 10.1 10.6 10.0

2092 257 357 2706

Water collectionhazards

No hazards while collecting wateroutside the dwelling

94.9 95.8 91.2 0.01 94.6

Faced hazards while collecting wateroutside the dwelling

5.1 4.2 8.8 5.4

2852 306 352 3510

Walking alone in theneighborhood after

darkVery safe 6.0 3.6 3.6 0.00001 5.5

Safe 30.2 25.4 18.4 28.6

Unsafe 40.3 41.6 33.1 39.7

Very unsafe 23.5 29.3 44.9 26.2

6920 798 873 8591

At home by yourselfafter dark

Very safe 20.6 12.2 16.4 0.00001 19.4

Safe 54.9 51.7 46.6 53.7

Unsafe 17.6 26.1 22.9 18.9

Very unsafe 7.0 10.0 14.1 8.0

6945 811 884 8640

4. Discussion4.1. Associations between Disability Severity and Inclusion

An objective of this paper was to conduct an initial exploratory analysis to opera-tionalize the proposed framework to assess disability inclusion and to examine associationsbetween important domains of inclusion and disability severity. Across the 11 domainsdescribed, we identified 40 indicators measuring aspects of inclusion. Of these, only sixshowed no statistically significant differences between those with no disabilities, mild dis-abilities, and moderate disabilities. These included the two indicators in the interpersonalstatus domain (asking about being humiliated while doing unpaid domestic and care workand such work being valued), water collection responsibility, communication problemswith healthcare worker, fuel collection hazards (if required to collect fuel), and having atoilet facility at home.

Of the remaining 34 indicators, by and large those with no disabilities had the bestoutcomes, followed by those with mild disabilities, and the worst outcomes were reportedby those with moderate and severe disabilities. In a few cases, there were no differencesbetween those with mild and moderate disabilities, but those with no disabilities had betteroutcomes. This analysis therefore indicates a negative relationship between disabilityseverity and inclusion—those with disabilities experienced a comparative lack of inclusion(worse outcomes) compared to those without disabilities—even though, in some cases,those with no disabilities also had poor outcomes.

However, there were some unusual or unexpected patterns, where the outcomeswere better for those with disabilities than those with no disabilities. One example is inthe voice domain, measured using two indicators—the ability to vote freely and withoutcoercion, and participation in local decision making. While there was generally a low level

Int. J. Environ. Res. Public Health 2021, 18, 4431 18 of 22

of participation in local decision making by all groups, it was lowest for people with nodisabilities—people with disabilities were found to be slightly more likely to participatein such processes. The strong presence of disabled people’s organizations in South Africamay contribute to people with disabilities feeling part of the activities and running ofthese organizations [24]. The higher levels of voting, and voting freely, may be linked tothe activities of the South African Independent Electoral Commission to increase votingparticipation by people with disabilities. While these suggest positive trends in inclusion,further research is needed to better understand the meaning and drivers of these results.

There are some results that are of particular interest, such as those for food insecurityand for access to clothing and footwear. These are not commonly reported outcomes in re-lation to disability at the population level and highlight an important area of disadvantage.The results show very high levels of food insecurity for all three groups, but outcomes werefound to be worse for those with mild disabilities and worst for those with moderate orsevere disabilities. More than half of those with mild disabilities, and almost two-thirds ofthose with moderate disabilities experienced severe food insecurity, with these data beingcollected before COVID-19 lockdowns and associated economic impacts. Studies in theUSA, Canada, and South Korea show that people with disabilities are more likely to be orlive in households that are food-insecure [25–27]. While there seems to be little literaturereporting on this in low- and middle-income countries, there is some emerging evidence ofhigh food insecurity among people with severe mental illness in Ethiopia [28].

A very rarely measured indicator in relation to disability (or indeed with respect topoverty more generally) is that of access to adequate clothing and footwear. The IDMmeasured whether individuals owned at least two complete changes of clothing andfootwear or not. There are high proportions from all three groups experiencing limitedownership of clothing and footwear, including one in five of those with mild disabilitiesand one in four of those with moderate or severe disabilities. The standard applied islow, and a lack of even this level of clothing and footwear has important implicationsfor people’s ability to move around in public with dignity and without shame, as wellas for employment possibilities, among other things. A scan of the published literatureconfirms the lack of reporting of access to clothing and footwear as a factor in experiences ofpeople with disabilities. A scoping review on the role of clothing on participation of peoplewith physical disabilities shows that clothing design is also an important determinantof participation [29]. Clothing that is not designed to accommodate a physical disabilitywill limit participation. However, other factors, such as poverty or neglect, may limit thenumber of changes of clothing a person with disabilities owns.

In the literature, one of the main indicators of inclusion is paid employment, but theIDM dataset provides us with an opportunity to look beyond employment and to consideradditional contributions which are often ignored in conventional statistics because theseactivities are typically unpaid. As anticipated, employment levels are lower for thosewith disabilities than those without, though all three groups have poor outcomes of lowemployment levels. However, significant contributions are made by all groups in termsof unpaid activities—for example, in unpaid domestic and care work, and in fuel andwater collection (from outside the dwelling). While there are often higher proportions ofpeople without disabilities that do these activities, this is not true for every activity, andthere are substantial proportions of people with mild and moderate disabilities makingthese types of contributions to their households. If being seen as ‘productive’ is a markerof inclusion, we can start to describe productivity more broadly than employment andincrease the visibility of these contributions that are so often ignored. This would be in linewith global trends of including unpaid work (usually performed by women) as part of theglobal economy [21,22].

This initial analysis does not consider a range of other factors that could provide otherexplanations for these significant differences—for example, the role of household povertyin accessing food and adequate clothing for everyone in the household and not just theperson with disabilities in that household; the role of age in providing unpaid care where

Int. J. Environ. Res. Public Health 2021, 18, 4431 19 of 22

older people, who are more likely to have disabilities, are required to provide care for theirgrandchildren—a common occurrence in South African households [30]. The IDM dataare individual level outcomes and the analysis conducted to date does not yet provide acomparison of levels of intra-household differences between inclusion of a person withdisabilities in relation to non-disabled members of their own household. Such analysescould improve our understanding of the potential effect of these variables on the measuresof inclusion.

The cut-offs chosen for each indicator, determining whether an individual is cate-gorized as included or not, were described for each variable in a way that seemed todifferentiate between a strongly positive (inclusion) and a strongly negative outcome (ex-clusion) as set out for each indicator. However, these cut-offs have not been tested andwould benefit from further scrutiny.

As noted above, the IDM data used are unweighted and only those people who wereable to answer for themselves and give informed consent were recruited into the IDMsurvey (i.e., excluding people with more severe disabilities who were not able to respondfor themselves); thus, caution should be exercised in making inferences from these results tothe wider South African population with disabilities. Furthermore, given the high levels ofpoverty and extremely high levels of inequality in South Africa, which are atypical featuresof middle-income countries, these results are not thought to be widely generalizable toother middle-income countries.

4.2. Limitations of Using Existing Datasets for Measuring Inclusion

One of the objectives of this analysis was to assess the feasibility of using an existingdataset for an allied but different purpose—the domains of inclusion identified in theframework proposed for measuring disability inclusion were populated with IDM data,where the data measured aspects of these domains. Given the high costs of collecting dataon sufficient numbers of respondents with disabilities, it makes sense to economize on datacollection activities where possible.

This study highlights a number of limitations associated with using existing datasetsfor a different purpose than they were originally designed, as described using exam-ples from our use of IDM data, designed to measure individual-level multidimensionaldeprivation and repurposed to populate the framework for measuring disability inclusion.

The first is that a number of the domains of inclusion were not comprehensively mea-sured because the IDM survey tools do not include questions covering the measurement ofall aspects of these domains. Examples include the limited data that can be used to popu-late the domains of interpersonal status, personal relationships, and being involved, andwhich therefore make it difficult to come to meaningful conclusions about the relationshipsbetween disability severity and inclusion in these domains.

The second limitation occurs when some data have to be excluded because they areinsufficient or incomplete. For example, the IDM dataset contains information aboutwhether individuals own their dwelling and/or the land on which it sits, but this couldonly be used as a measure of inclusion if it was supplemented with information on whetherindividuals choose to own their dwelling, or are somehow prevented from doing so (e.g.,renting may be a valid choice, and thus not indicative of exclusion). IDM data are alsoavailable regarding the use of assistive devices for those with functioning difficulties.However, these data would need to be supplemented with information about whetherindividuals had an unmet need for an assistive device to determine whether not using anassistive device was an indicator of a lack of inclusion.

Finally, some of the IDM indicators could potentially be allocated to more than onedomain. For example, while the interpersonal status domain should incorporate issues ofrespect, the healthcare domain also includes an indicator assessing whether individualswere treated respectfully by their healthcare provider.

Int. J. Environ. Res. Public Health 2021, 18, 4431 20 of 22

Each of these limitations would be addressed by specifically designing measures ofinclusion—and the necessary survey tools—in order to ensure comprehensive data capturewith appropriate coverage of indicators within each domain measured.

4.3. Future Directions—Considerations for Monitoring Inclusion for People with Disabilities

It is difficult to use existing data that have been collected for different purpose, butthe IDM data do provide evidence of some important trends. The first is that it highlightsindividual-level information across a range of indicators that are rarely (if ever) available,such as access to adequate clothing and footwear, as well as contributions to the house-hold. Secondly, it shows broad trends (as expected) in the associations between lack ofinclusion and severity of disability. Given that the IDM survey tools were designed tobe broadly standard across contexts, this analysis could potentially be replicated usingother IDM datasets, and the feasibility of using other non-IDM datasets containing relevantinformation (including an appropriate measure of disability status) using this frameworkfor measuring inclusion could also be examined.

However, survey tools designed specifically to measure inclusion should include notonly more detailed questions on many of the domains such as being respected and havingdignity, but also measures of the person’s own sense of their level of inclusion. Thus, whilethis analysis has also demonstrated that it feasible to operationalize this framework formeasuring disability inclusion utilizing an existing dataset, attempts to do so would bebetter served by specifically designed tools.

Analysis of the current domains disaggregated by factors such as age, age of onset,type of disability, and gender is possible, though not undertaken in this paper, and wouldbe recommended to yield a more nuanced understanding of these results on the basis of theIDM data and the factors driving levels of inclusion. Importantly, any specifically designedinclusion measure should ensure that intersectional analyses such as these (and othersidentified as important) are possible.

Two additional lessons that are relevant to the design of tools to measure inclusionspecifically can be drawn from the design process of the IDM. One is the importance of de-signing a gender-sensitive measure—for example, the IDM collects data on various aspectsof menstruation, including the non-attendance of activities because of a lack of sanitaryproducts and/or shame or stigma associated with menstruation, which are collected onlyfor women. A gender-sensitive measure of inclusion should also include data about thoseissues in each domain that affect women and men differentially.

A second lesson can be drawn from the process of selecting the dimensions of depriva-tion that are measured by the IDM. The 14 dimensions measured in the IDM were selected,in large part, on the basis of participatory work conducted with poor men and women toidentify the dimensions of poverty that they themselves prioritized. To maximize relevancefor people with disabilities and their supporting organizations, we recommend that anyfuture work to design specific tools to measure disability inclusion should also includeparticipatory work with people with disabilities designed so they identify and prioritizethe key domains and indicators of inclusion that are important to measure. Such a processwould be likely to highlight important indicators (and perhaps domains) of inclusion thathave not been considered in this analysis.

5. Conclusions

The objectives of this paper were threefold: the first was to propose a framework of thekey domains of inclusion that should be included in large-scale studies measuring levels ofinclusion of people with disabilities. The second was to operationalize the framework usingan existing dataset, from the IDM South Africa Country Study, and thirdly, to assess thefeasibility of using existing datasets to measure inclusion. The purpose of operationalizingthe framework is to improve our understanding of the associations between disabilityseverity and levels of inclusion.

Int. J. Environ. Res. Public Health 2021, 18, 4431 21 of 22

Do the IDM data allow us to measure inclusion on the basis of the proposed frame-work? This analysis suggests that there are important and interesting elements that showpromise and that could be developed further to create better measures of inclusion. Overall,the data show the expected (negative) relationship between social inclusion outcomes anddisability severity. However, because the IDM was designed to measure deprivation, it isnot as comprehensive with respect to indicators of social inclusion as measures designedspecifically to understand inclusion would be. That means we are missing data on someimportant issues, for example, about the recognition of people with disabilities by others,as well as on issues around individuals’ social life.

This analysis shows that the use of data collected for one purpose can have utility foralternative purposes, though there are limitations of the approach. These limitations wouldbe overcome by using specifically designed tools, but in the absence of such specificallydesigned tools and given the high costs of data collection, it makes sense to examine thefeasibility of using existing datasets containing relevant data for alternative purposes.

Author Contributions: Conceptualization, M.S. and H.S.; methodology, H.S. and M.S.; writing—original draft preparation, M.S. and H.S.; writing—review and editing, H.S. and M.S.; visualization,H.S. All authors have read and agreed to the published version of the manuscript.

Funding: This research was funded by the Australian Department of Foreign Affairs and Trade.

Institutional Review Board Statement: This paper undertook a secondary analysis of existing data.The original IDM South Africa Country Study gained several ethics approvals in Australia and SouthAfrica, from the Australian National University Human Research Ethics Committee (Protocol No.2016/355), the South African Human Sciences Research Council Research Ethics Committee (ProtocolNo. REC 5/21/11/18) and the Limpopo Provincial Research Ethics Committee (Clearance CertificateNo. LPREC/35/2018: PG).

Informed Consent Statement: Informed consent was obtained from all individuals involved in theIDM South Africa Country Study surveys.

Data Availability Statement: The data presented in this study have been lodged with the AustralianData Archive.

Acknowledgments: The 2016–2020 Individual Deprivation Measure (IDM) Program was a partner-ship between the Australian National University (ANU), the International Women’s DevelopmentAgency (IWDA), and the Australian Government, through the Department of Foreign Affairs andTrade. Special thanks Sharon Bessell, Janet Hunt, Trang Pham, and Mandy Yap. Thanks also go toikapadata Pty Ltd. who undertook the data collection in South Africa, and to all the fieldworkerswho helped make the data collection such a success.

Conflicts of Interest: The authors declare no conflict of interest.

References1. Levitas, R.; Pantazis, C.; Fahmy, E.; Gordon, D.; Lloyd-Reichling, E.; Patsios, D. The Multi-Dimensional Analysis of Social Exclusion;

Department of Sociology and School for Social Policy, Townsend Centre for the International Study of Poverty and Bristol Institutefor Public Affairs University of Bristol: Bristol, UK, 2007.

2. Gingrich, L.G.; Lightman, N. The empirical measurement of a theoretical concept: Tracing social exclusion among racial minorityand migrant groups in Canada. Soc. Incl. 2015, 3, 98–111. [CrossRef]

3. Huxley, P. Introduction to ‘Indicators and measurement of social inclusion’—Editorial. Soc. Incl. 2015, 3, 112–126.4. Hall, S.A. The social inclusion of people with disabilities: A qualitative meta-analysis. J. Ethnogr. Qual. Res. 2009, 3, 162–173.5. Ikäheimo, H. Personhood and the social inclusion of people with disabilities. A recognition-theoretical approach. In Arguing

about Disability. Philosophical Perspectives; Kristiansen, K., Vehmas, S., Vehmas, S., Eds.; Routledge: London, UK, 2009.6. UN. Disability and Development Report. Realising the Sustainable Development Goals by, for and with Persons with Disabilities; United

Nations: New York, NY, USA, 2018.7. WHO; World Bank. World Report on Disability; WHO: Geneva, Switzerland, 2011.8. Hay, K.; Pascoe, K.M. Disabled people and disaster management in New Zealand: Examining online media messages. Disabil. Soc.

2019, 34, 253–275. [CrossRef]9. Milner, P.; Kelly, B. Community participation and inclusion: People with disabilities defining their place. Disabil. Soc. 2009, 24,

47–62. [CrossRef]

Int. J. Environ. Res. Public Health 2021, 18, 4431 22 of 22

10. Sherlaw, W.; Lucas, B.; Jourdain, A.; Monaghan, N. Disabled people, inclusion and policy: Better outcomes through a publichealth approach? Disabil. Soc. 2013, 29, 444–459. [CrossRef]

11. Virendrakumar, B.; Jolley, E.; Badu, E.; Schmidt, E. Disability inclusive elections in Africa: A systematic review of published andunpublished literature. Disabil. Soc. 2018, 33, 509–538. [CrossRef]

12. Jolley, E.; Lynch, P.; Virendrakumar, B.; Rowe, S.; Schmidt, E. Education and social inclusion of people with disabilities in fivecountries in West Africa: A literature review. Disabil. Rehabil. 2018, 40, 2704–2712. [CrossRef] [PubMed]

13. Rohwerder, B. Disability Inclusion: Topic Guide; GSDRC, University of Birmingham: Birmingham, UK, 2015.14. Lister, R. Poverty; Polity Press: Malden, MA, USA, 2004.15. World Bank. Inclusion Matters: The Foundation for Shared Prosperity, New Frontiers of Social Policy; World Bank: Washington, DC,

USA, 2013.16. Wisor, S.; Bessell, S.; Castillo, F.; Crawford, J.; Donaghue, K.; Hunt, J.; Jaggar, A.; Liu, A.; Pogge, T. The individual deprivation

measure. In A Gender Sensitive Approach to Poverty Measurement; International Women’s Development Agency: Melbourne,Australia, 2014.

17. ANU IDM. ANU IDM Survey 2019. Fieldwork Report, Unpublished Report Prepared for the ANU IDM Team; Ikapadata Pty Ltd.: CapeTown, South Africa, 2019.

18. Suich, H.; Pham, T.; Hunt, J.; Yap, M.; Hasan, M.; Bessell, S. The Individual Deprivation Measure South African Country Study;ANU, Crawford School of Public Policy: Canberra, Australia, 2020. Available online: (accessed on 15 April 2021).

19. WGDS, ud. The Washington Group on Disability Statistics Short Set on functioning; Washington Group on Disability Statistics: Wash-ington, DC, USA. Available online: (accessed on 15 April 2021).

20. Madans, J.; Loeb, M. Measuring disability and inclusion in relation to the 2030 Agenda on sustainable development. Disabil. Glob.South 2017, 4, 1164–1179.

21. Hirway, I. Unpaid work and the economy: Linkages and their implications. Indian J. Labour Econ. 2015, 58, 1–21. [CrossRef]22. Hoskyns, C.; Rai, S.M. Recasting the Global Political Economy: Counting Women’s Unpaid Work. New Political Econ. 2007, 12,

297–317. [CrossRef]23. Suich, H.; Pham, T.; Hunt, J.; Hasan, M.; Yap, M.; Bessell, S.; Bexley, A.; Crawford, J.; Fisk, K.; McInerney, C.; et al. Individual Depri-

vation Measure Index Construction and Dimension Scoring Procedures, South Africa Country Study Version 2; ANU: Canberra, Australia,2020. Available online: (accessed on 15 April 2021).

24. Howell, C.; Chalklen, S.; Albert, T. A history of the disability rights movement in South Africa. In Disability and Social Change: ASouth African Agenda; Watermeyer, B., Wartz, L., Lorenzo, T., Schneider, M., Priestley, M., Eds.; HSRC Press: Pretoria, South Africa,2006.

25. Heflin, C.M.; Altman, C.E.; Rodriguez, L. Food insecurity and disability in the United States. Disabil. Health J. 2019, 12, 220–226.[CrossRef] [PubMed]

26. Park, J.E.; Kim, S.Y.; Kim, S.H.; Jeoung, E.J.; Park, J.H. Household Food Insecurity: Comparison between Families with andwithout Members with Disabilities. Int. J. Environ. Res. Public Health 2020, 17, 6149. [CrossRef] [PubMed]

27. Schwartz, N.; Tarasuk, V.; Buliung, R.; Wilson, K. Mobility impairments and geographic variation in vulnerability to householdfood insecurity. Soc. Sci. Med. 2019, 243, 112636. [CrossRef] [PubMed]

28. Tirfessa, K.; Lund, C.; Medhin, G.; Hailemichael, Y.; Fekadu, A.; Hanlon, C. Food insecurity among people with severe mentaldisorder in a rural Ethiopian setting: A comparative, population-based study. Epidemiol. Psychiatr. Sci. 2017, 28, 397–407.[CrossRef] [PubMed]

29. Esmail, A.; Poncet, F.; Auger, C.; Rochette, A.; Dahan-Oliel, N.; Labbé, D.; Kehayia, E.; Billebaud, C.; de Guise, É.; Lessard, I. Therole of clothing on participation of persons with a physical disability: A scoping review. Appl. Ergon. 2020, 85, 103058. [CrossRef][PubMed]

30. Makiwane, M.; Gumede, N.A.; Makoae, M.; Vawda, M. Family in a changing South Africa: Structures, functions and the welfareof members. S. Afr. Rev. Sociol. 2017, 48, 49–69. [CrossRef]

  • Introduction
  • Materials and Methods
    • A Framework for Measuring Disability Inclusion
    • The IDM South Africa Country Study
    • Disability Definition and Measurement
    • Selection of Indicators to Populate the Framework
    • Data Analysis
  • Results
    • Interpersonal Status
    • Personal Relationships
    • Being Involved
    • Living Conditions
    • Economic Opportunity and Contributions
    • Support Systems (Formal and Informal Support)
    • Institutional Status
    • Voice
    • Education
    • Healthcare Access
    • Personal Safety
  • Discussion
    • Associations between Disability Severity and Inclusion
    • Limitations of Using Existing Datasets for Measuring Inclusion
    • Future Directions—Considerations for Monitoring Inclusion for People with Disabilities
  • Conclusions
  • References

How it Works

  1. Clіck оn the “Place оrder tab at the tоp menu оr “Order Nоw” іcоn at the bоttоm, and a new page wіll appear wіth an оrder fоrm tо be fіlled.
  2. Fіll іn yоur paper’s іnfоrmatіоn and clіck “PRІCE CALCULATІОN” at the bоttоm tо calculate yоur оrder prіce.
  3. Fіll іn yоur paper’s academіc level, deadlіne and the requіred number оf pages frоm the drоp-dоwn menus.
  4. Clіck “FІNAL STEP” tо enter yоur regіstratіоn detaіls and get an accоunt wіth us fоr recоrd keepіng.
  5. Clіck оn “PRОCEED TО CHECKОUT” at the bоttоm оf the page.
  6. Frоm there, the payment sectіоns wіll shоw, fоllоw the guіded payment prоcess, and yоur оrder wіll be avaіlable fоr оur wrіtіng team tо wоrk оn іt.

Nоte, оnce lоgged іntо yоur accоunt; yоu can clіck оn the “Pendіng” buttоn at the left sіdebar tо navіgate, make changes, make payments, add іnstructіоns оr uplоad fіles fоr the оrder created. e.g., оnce lоgged іn, clіck оn “Pendіng” and a “pay” оptіоn wіll appear оn the far rіght оf the оrder yоu created, clіck оn pay then clіck оn the “Checkоut” оptіоn at the next page that appears, and yоu wіll be able tо cоmplete the payment.

Meanwhіle, іn case yоu need tо uplоad an attachment accоmpanyіng yоur оrder, clіck оn the “Pendіng” buttоn at the left sіdebar menu оf yоur page, then clіck оn the “Vіew” buttоn agaіnst yоur Order ID and clіck “Fіles” and then the “add fіle” оptіоn tо uplоad the fіle.

Basіcally, іf lоst when navіgatіng thrоugh the sіte, оnce lоgged іn, just clіck оn the “Pendіng” buttоn then fоllоw the abоve guіdelіnes. оtherwіse, cоntact suppоrt thrоugh оur chat at the bоttоm rіght cоrner


Payment Prоcess

By clіckіng ‘PRОCEED TО CHECKОUT’ yоu wіll be lоgged іn tо yоur accоunt autоmatіcally where yоu can vіew yоur оrder detaіls. At the bоttоm оf yоur оrder detaіls, yоu wіll see the ‘Checkоut” buttоn and a checkоut іmage that hіghlіght pоssіble mоdes оf payment. Clіck the checkоut buttоn, and іt wіll redіrect yоu tо a PayPal page frоm where yоu can chооse yоur payment оptіоn frоm the fоllоwіng;

  1. Pay wіth my PayPal accоunt‘– select thіs оptіоn іf yоu have a PayPal accоunt.
  2. Pay wіth a debіt оr credіt card’ or ‘Guest Checkout’ – select thіs оptіоn tо pay usіng yоur debіt оr credіt card іf yоu dоn’t have a PayPal accоunt.
  3. Dо nоt fоrget tо make payment sо that the оrder can be vіsіble tо оur experts/tutоrs/wrіters.


Custоmer Suppоrt

Order Solution Now