“I Don’t Know Whether It’s Priority, or Capacity, or Both”: Implementation Gaps in Employment Policies Targeting People with Disabilities in Kenya and Bangladesh

Abstract

Disabled people are often excluded from employment, particularly in accessing waged employment. This exclusion persists despite enabling policies and legislature, indicating a need to address gaps in implementation. To improve policy implementation, it is crucial to better understand the gap between policy intention (what was planned) and effect (what happens), including reasons that widen this gap (what explains what happened). This study explores this in the context of Kenya and Bangladesh, drawing on key informant interviews and analysis of policies and programmes related to employment, social protection, and disability. Using the policy triangle framework, we analysed the policy context, process, actors, and content of relevant policies and programmes. The study identified key policy provisions on employment quota, social assistance, job placement services, tax incentives, and training opportunities, all of which were designed to promote employment for disabled people. However, implementation is limited in both countries, for reasons including ambiguity in policies, unavailability of data for monitoring, and lack of transparency among implementers. The data suggest that addressing these aspects would promote accountability among duty-bearers, advocacy power among rights-holders, and ultimately improve implementation and inclusion of disabled people in employment.

1. Introduction

An estimated 16% of the population across the world—over one billion people—are living with one or more disabilities [1]. This figure is believed to be increasing, particularly in low- and middle-income countries (LMICs) and resource-poor settings. This is significant because disability and poverty are known to be linked in a causal and consequential relationship—poorer people are more likely to experience impairments, and disabled persons and their households experience multi-dimensional poverty [2,3,4].

Exclusion of disabled people from the labour force is a dominant driver of multi-dimensional poverty [5], generating hardship for households with disabled members. Even when disabled people are engaged in employment, they are more likely than nondisabled people to experience a range of factors that exacerbate poverty and insecurity, such as lower earnings, self-employment, and work in the informal sector, where employment is less stable [2].

Disabled people are often excluded from employment, particularly in accessing waged employment. Key barriers include discriminatory practices among employers, limited reasonable accommodations, and lack of appropriate and inclusive policies [6]. Countries that have ratified the United Nations Conventions on the Rights of Persons with Disabilities (UNCRPD), like Bangladesh and Kenya, are expected to have established national laws and policies supporting disabled people to access employment. However, studies suggest that the employment of disabled people is especially challenging in LMICs because of the underlying socioeconomic and political reasons affecting the employment market and social welfare policies [7], indicating that a myriad of factors may be contributing to this inequity.

Wickenden and colleagues identified three broad categories of “vital ingredients” needed to accelerate disability-inclusive formal employment: (1) legal, policy and regulatory mechanisms (e.g., inclusive employment policies); (2) supply-side interventions (e.g., skills development); and (3) demand-side interventions (e.g., incentives for employers) [8]. Applying this to the contextual focus of this paper, Bangladesh was among the first 10 countries to ratify the UNCRPD in 2007 [9], and provisions for non-discrimination and the inclusion of disabled people is evident throughout the legislative and policy environment, including the Constitution [10]. Similarly, the Kenyan government ratified the UNCRPD in 2008, and provisions protecting and promoting the rights of disabled people are included in the 2010 Constitution and the 2003 Persons with Disabilities Act and reiterated in the Disability Bill of 2019, amongst others [11,12]. In terms of supply-side interventions, both countries have committed to and invested in promoting the education and training of disabled people in order to promote their inclusion in the workforce [13]. As demand-side interventions, the International Labour Organisation (ILO) Business and Disability Networks have national networks in both Kenya and Bangladesh, through which national-level businesses and employers are supported via technical advice, peers, sharing of good practices, and connection to local disability organisations [14].

Yet, disabled people are still falling behind: data from the 2015/2016 Kenya Integrated Household Budget Survey revealed that over half of Kenyans with disabilities had “difficulties engaging in economic activities” [15], and that in Bangladesh, disabled adults have a higher rate of unemployment than non-disabled adults (1.9% vs. 1.5%) [16]. So why is there a gap despite seemingly fulfilling vital ingredients?

An implementation gap is the space between intention (the objectives of a policy) and effect (what happens in practice) [17], and it is this gap that we scrutinise in this study. Past studies in Bangladesh have commented upon the lack of progress in promoting access to employment among disabled people [18] and more have highlighted how policy commitments do not translate into practice [19]. In Kenya, too, there have been reports of ineffective policies and the need for enabling environments through the enforcement of protective laws [20]. However, as Gilson and Raphaely point out, far too many studies only concern themselves with “what happened” instead of also investigating “what explains what happened” [21]. Using Kenya and Bangladesh as case studies, we examine how the policies and programmes that promote waged employment among disabled people are being implemented (i.e., “what happened”) and the reasons underlying any implementation gaps (i.e., “what explains what happened”).

In doing so, we aim to shed light on whether it is “non-implementation” or “unsuccessful implementation,” as distinguished by Hogwood and Gunn [22]. They defined non-implementation as the situation that arises when a policy is not put into effect as intended, which, in our case, would mean that implementation mechanisms would need to be strengthened in order to achieve the desired objective of having more disabled people in waged employment. In contrast, “unsuccessful implementation” means that the policy was implemented in full as intended but has failed to produce the desired outcome, in which case policymakers and stakeholders must revisit the specific policy instruments (e.g., a different approach to engaging employers in establishing inclusive practices). Thus, in addition to examining why the presence of supportive policies and programmes is not yielding changes in these two settings, this study aims to highlight what is needed to improve policy implementation.

Background

Before embarking on study methodology and results, we situate the concepts and the context of this study.

Disability is defined in the UNCRPD as including people “who have long-term physical, mental, intellectual or sensory impairments which in interaction with various barriers may hinder their full and effective participation in society on an equal basis with others” [23]. There have been several models of disability that explain the way disabled people have been regarded by governments, service providers, researchers, and communities. These have included the medical model that identifies disability as a medical problem that resides in the body and the social model that argues that people are disabled by society [24], calling for “barrier removal, anti-discrimination legislation, independent living and other responses to social oppression” [25]. For example, a person with a spinal cord injury may have limited employment opportunities not because they use a wheelchair but because most workplaces are not accessible to wheelchairs. However, the social model has been criticised for failing to acknowledge the role of impairments [25]—e.g., the person in our example may experience pain as a result of their injury and miss work, even though their workplace is fully accessible. The International Classification of Functioning, Disability and Health (ICF) by the World Health Organisation (WHO) attempts to combine components of the social and medical model, suggesting that disability results from an interplay of impairments (e.g., mobility impairment), activity restrictions (e.g., difficulty walking), and participation restrictions (e.g., participation in work), all of which are mediated by personal (e.g., being able to afford a wheelchair) and environmental factors (e.g., supportive policies, existence of ramps) [26].

The measurement of disability, too, is varied and can lead to different estimates of disability prevalence. For example, being asked directly to self-identify (e.g., “do you have a disability?”) may lead to under-reporting due to stigma associated with disability. In Zambia, using this approach led to a disability prevalence of 1% compared to the 13% figure when using the Washington Group (WG) short-set questions that screened people’s functioning on six domains (seeing, hearing, walking, remembering/concentrating, self-care, communicating) [27]. Although not without limitations, the WG questions are used extensively in many countries, and enable governments to track disability-disaggregated development indicators, as required by the Sustainable Development Goals (SDGs) as well the UNCRPD [23,28]. Inaccuracies in disability prevalence figures were key challenges in both Kenya and Bangladesh.

Bangladesh has a population of 165 million [29], with 24.3% of the population living below the poverty line in 2019 [30]. In the 2016 Household Income and Expenditure Survey, disability prevalence (among those aged 5+ years) was estimated to be 6.9% [31]. However, disability prevalence in Bangladesh has been a contested figure, with inaccuracies attributed to factors such as the lack of routine data collection or a surveillance system that includes disability [32]. Until 2019, a key source of prevalence data was the Ministry of Social Welfare’s Disability Detection Survey that, crucially, depended on self-reporting, which indicated the number of disabled people as 2.4 million (1.4% of the population) [33]. In contrast, the alternative CRPD report of 2019 based on NGOs’ micro-studies estimates this figure to be 24 million, which is 14.5% of the total population [34]. Since then, the Bangladesh Bureau of Statistics has conducted the first standalone household survey on disability, the National Survey on Persons with Disabilities (NSPD) 2021 [35]. It reported that according to government-defined categories, 2.8% of the population had at least one form of disability, while this figure was 7.14% according to the Washington Group questions on functional limitations. The data thus vary extensively depending on disability definition and data collection methodology, which then has implications for monitoring proportions of disabled people registered for benefits.

Kenya has a population of nearly 50 million, with an estimated 20.9% living in extreme poverty [36]. According to the 2019 Census, the prevalence of disability in Kenya among people 5 years and older is 2.2% [37]. This disability prevalence figure is likely to be an underestimate, with the more probable prevalence being possibly as high as 15%, according to the WHO/World Bank estimates [6]. As elsewhere, there is a lack of reliable data on disability in Kenya, including employment rates, which are often incomplete or poor quality [8]. Evidence suggests that the lack of access to employment opportunities is a significant contributor to higher poverty rates among disabled people [20].

2. Materials and Methods

This policy analysis was conducted as formative research for the Innovation to Inclusion (i2i) programme to improve access to waged employment among disabled people in Kenya and Bangladesh. As such, the selection of these two country sites were programmatic decisions beyond the control of the authors. However, in conducting the analyses, it was evident that similar and contrasting elements between the two countries highlighted important lessons.

This research was guided by the following research question: what are the reasons underlying any gaps in implementing policies and programmes promoting waged employment for disabled people in Kenya and Bangladesh? Based on past research (e.g., [18]) we hypothesise that implementation gaps are indeed present. Nonetheless, a portion of our results establish the extent of implementation before delving into the underlying reasons. In examining these reasons, we selected the policy triangle framework by Walt and Gilson, originally developed for health policy analysis, though it has been applied and proven relevant in many other types of policies [38,39]. In addition to its agility in application, this framework was chosen because it allows scrutiny of not just the contents of the policies but also contextual factors and policy actors that can strengthen, weaken, or mediate implementation. This, in turn, enabled us to pinpoint where implementation gaps lie, and why.

Two methods were used. We first conducted a document review and analysis of relevant policies and programmes regarding employment, social protection, and disability. Then, in-depth qualitative interviews were conducted with relevant stakeholders to provide insights into the process, context, and implementation gaps. Key informants were identified through the document review, with input from local partners, and through snowball sampling.

In Bangladesh, 22 key informant interviews were conducted in Dhaka in December 2018. Key informants included policy and programme staff from the government (n = 9) as well as representatives from international non-governmental organisations (INGO) (n = 1), representatives from local NGOs and organisations of people with disabilities (OPDs) (n = 10), and private actors, including experts and employers (n = 2). In Kenya, 18 key informant interviews were conducted in Nairobi in January 2019. Key informants included policy and programme staff from the government (n = 8) as well as INGOs (n = 2), NGO/OPD representatives (n = 4), and private actors, including employers (n = 4).

Interviews were guided by semi-structured interview guides. Participants were provided information sheets with details about the study, and written informed consent was obtained prior to commencing the interview. Interviews ranged from 30 to 60 min and were conducted in a mutually agreed-upon location. All interviews were conducted in person by SH, and audio-recorded where participants consented. Some government officials in Bangladesh declined to be audio-recorded—in these instances, the analysis was based on detailed interview notes.

Coding and analysis were managed using NVivo (qualitative data analysis software by Lumivero, Denver, CO, USA) and guided by the policy triangle framework [38] and examined the following dimensions:

• Context: to examine the social and political climate and identify contextual influences on policy or programme design;
• Process: to examine how relevant policies and programmes were developed, identify challenges, and capture the extent of consultations with OPDs;
• Actors: to examine the actors involved in developing and implementing the policies, the roles they play, and opportunities for engagement;
• Content: to review the relevant policies and programme, identifying strengths and weaknesses as well as barriers and facilitators to implementation.

3. Results

The results of the analysis are presented in three main parts. The first provides analyses of the policy context and process, followed by that of policy actors, highlighting how they affect policy implementation. The third pertains to the policy content, within which we highlight key provisions, implementation gaps, and reasons contributing to these gaps.

3.1. Policy Context and Process

3.1.1. Bangladesh

The first National Policy for the Disabled was formulated in 1995, followed by the comprehensive Bangladesh Person with Disability Welfare Act of 2001. The Rights and Protection of Persons with Disabilities Act (2013) moved away from the welfare-based model of preceding laws, and included a disability definition aligned with the UNCRPD.

Other relevant legislature and policies include the Neuro-Developmental Disability (NDD) Protection Trust Act of 2013 focusing on autism spectrum disorders, intellectual disability, Down syndrome, and cerebral palsy [40]. That the NDD Act and the Disability Act exist in parallel provides unique opportunities for lessons about what works in policy advocacy and engagement with the government. The Labour Act 2006 and the National Skills Development Policy 2011 both include provisions for disabled people. However, both reflect disability definitions from the repealed Disability Welfare Act 2001 [41,42].

Key informants described the policy process in Bangladesh as slow. Though the Disability Act was enacted in 2013, the action plan for its implementation was drafted in 2015 but had not been approved by 2018. The data further indicate that disability rights advancements are heavily dependent on the political climate, and key informants remarked upon explicit support for disability rights among politicians.

3.1.2. Kenya

The 2003 Persons with Disabilities Act predates the UNCRPD. As such, revisions have led to the 2019 Persons with Disabilities Bill being more in line with the UNCRPD, defining disability as:

“…a physical, sensory, developmental, or other impairment, including any visual, hearing, learning or physical incapability which impacts adversely on social, economic, or environmental participation.”

[43]

The Kenyan legislative framework is regarded as supportive of inclusive employment and contains non-discriminatory clauses providing the basis for equal access to labour and employment. For example, discrimination on the grounds of disability in employment in both the public and private sectors was outlawed by the Employment Act 2007. Other relevant legislation includes the 2016 National Employment Authority Act, the 2003 Public Officers’ Ethics Act, and the 2018 Disability Policy and Guidelines for the Public Service [44].

Like Bangladesh, the policy process in Kenya was described as slow, with many key informants remarking upon the slow pace of the Disability Act 2003 revision coming into effect. Several key informants described the need for powerful allies and capitalising on key events such as the Global Disability Summit as a way of prompting action. In contrast to Bangladesh, commitment to advancing disability rights was less dependent on changes to leadership. The relationships between government, lawmakers, and non-government actors seemed ingrained, including unofficial connections and alliances that were evident from the key informant interviews.

3.2. Policy Actors

3.2.1. Bangladesh

The Disability Act 2013 details two inter-Ministerial committees responsible for policy formulation, coordination, and implementation of national-level activities. However, key informants report that these Committees are rarely convened, if at all. The Act also outlines similar committees at the district, upazila (sub-district), and urban levels, with responsibilities that include approving registration cards and granting social assistance. Analysis indicates difficulties with accountability and role definition among the various committees, as there are reportedly over 1200 at the different levels.

“[The] problem is that the Committees don’t have their roles defined yet. There is no secretariat or office for the Committee. Even if a letter comes, who will sign for it? If there are suggestions, who will they contact?”

(Civil society key informant, Bangladesh)

The Ministry of Social Welfare (MoSW) holds the mandate for disability-related programmes. Within it, the Department of Social Services (DSS) manages the social protection programmes such as the Disability Allowance. Both are heavily criticised by key informants as being inflexible on two issues: (1) failing to allocate to disability programming beyond the Disability Allowance and (2) failing to involve other government authorities in implementing disability-related programmes.

I: Do you think that [sole assignment of disability to MoSW] is effective?

R: Not effective at all.

I: How come?

R: Because we need to involve all the Ministries. And we are trying to do that. Like the Disaster Management. Their Law is now inclusive, their policies are now inclusive. Like the Education Ministry, they try that if they make any policy they make it inclusive. But if you… rely everything on the Ministry of Social Welfare, they will not be able to do that. I think one of the biggest failures in the disability sector is because of the Ministry of Social Welfare. Because everything goes to them… it is their responsibility, so they are not capable to do that.

(Government key informant, Bangladesh)

The Jatio Protibondhi Unnayan Foundation (JPUF) (National Foundation for the Development of Disabled Persons) is the implementation arm of the Ministry of Social Welfare and is similarly criticised as lacking capacity. Formed in 1999, their flagship programme is the one-stop service centres, of which there are 103 throughout the country. Most key informants doubt the impact and efficiency of these centres, with some explaining that they were designed and better suited for the Ministry of Health but are now thinning JPUF resources and stretching their capacity.

The National Forum of Organisations Working with the Disabled (NFOWD) is a national network of over 300 OPDs and NGOs working on disability in Bangladesh. Although it played key roles in the CRPD ratification process and development of the Disability Act, key informants described frequent in-fighting.

3.2.2. Kenya

The Ministry of Gender, Children and Social Development leads government disability inclusion policies and programmes. The Ministry of Labour runs disability-related employment programmes, managed by the Disability Division. The Labour Ministry works closely with the Federation of Kenya Employers (FKE) and the Central Organisation of Trade Unions—Kenya (COTU-K). All three parties describe a good working relationship.

The National Council for Persons with Disabilities (NCPWD) is a semi-autonomous body also within the Labour Ministry. Many key informants suggested that the NCPWD have an oversight role, as opposed to one of implementation, for which many felt they lacked the capacity. Many NCPWD programmes were described as piecemeal, though good for publicity. Most key informants believed that addressing the NCPWD’s role was crucial to better implementation of the Disability Act and other commitments, and to increase accountability of other government agencies. One key informant provides an example:

“…NCPWD is given the authority to issue Adjustment Orders on buildings that are not accessible. They have never issued a single order, to anybody. They have never gone out to access the buildings, and I don’t know whether its priority, or capacity. Or both.”

(NGO key informant, Kenya)

The United Disabled People of Kenya (UDPK) is an umbrella organisation of OPDs, so designed to present a united voice in engagements with the government. The UDPK co-chairs the Disability Caucus for the Implementation of the Constitution (DCID) with government counterparts and reports to have a collaborative relationship. Key informant interviews indicated that there are some non-governmental actors whose input carries more weight than others and are well connected with government actors. Other outspoken advocates include the Kenya Disabled Parliamentary Association (KeDiPA) within the Parliament and the Kenya National Commission on Human Rights.

3.3. Policy Content

Table 1. Summary of policy content analysis.

Provision	Implementation Details	Gaps and Reasons
(1) Registration as a person with a disability	All legal provisions and support (including for employment) are dependent on being registered as a disabled person. By application, including a medical assessment to verify disability.	- Disability prevalence figures vary, thus making it hard to verify the proportion registered. - Data suggest that likely a large proportion of disabled people are excluded from registration and other legal provisions. - Registration is challenging. Reasons include medical assessment difficulties, accessibility constraints, and financial barriers.
	Bangladesh: submitted online or at district committees. There are 1.6 million estimated to be registered, but at least 11 million estimated as having a disability [45].
	Kenya: submitted at central office in Nairobi and renewed every 5 years. There are 500,000 estimated to be registered, but 6.6 million estimated as having a disability.
(2) Disability allowance/benefit scheme	Bangladesh: Allowances for the Financially Insolvent Persons with Disabilities: disability-targeted benefit of BDT 750 (USD~8) monthly. By application. Means-tested and has several eligibility criteria. Restrictions to household income and receiving other benefits. Must be registered (see 1). Coverage reported as 0.75 million in government report [46], but this was challenged by civil KIs.	- Disability prevalence figures vary, thus making it hard to establish coverage of benefits. - Lack of clarity about eligibility criteria. - Eligibility assessed by household, not individuals. - Beneficiaries forced to choose between disability benefit and other benefits.
	Kenya: Cash Transfer for Persons with Severe Disabilities (PWSD-CT): disability-targeted benefit of KES 2000 (USD ~20) every 2 months. Not by application: community members identify and verify “needy” households. Given to households with disabled members requiring full-time caregiver support. Restrictions to household income and receipt of other benefits. Must be registered (see 1). Government KIs report coverage as 47,000 households.
(3) Employment quota	Bangladesh: 10% quota in public-sector jobs, shared between orphans and registered disabled people. No evidence of quota being met or monitored. Unclear whether it was resumed after halt in 2018.	- Ambiguity in phrases used (e.g., “shared quota,” “progressive implementation”) makes implementation and monitoring difficult. - Unwillingness of government bodies to publish relevant figures, even when it is collected. - Without figures, limited accountability among duty-bearers, and no advocacy power for rights-holders.
	Kenya: 5% quota for registered disabled people in all public elective and appointive bodies and in the private sector. No evidence of quota being met. Public institutions required to report figures to the National Council for People with Disabilities, but these are not published.
(4) Tax incentives for employers	Tax incentives for private employers who employ registered disabled people.	- Lack of clarity about eligibility and application process to receive this tax relief. - Without guidelines, raising awareness is difficult. - Application process is considered complicated and the amount too low to be worth the effort.
	Bangladesh: Implementation “upon framing proper guidelines” [47], but no guidelines have been published and exemptions are granted on a case-by-case basis.
	Kenya: Private employers can apply for (1) a tax deduction equivalent to 25% of salary of disabled employees; and (2) a tax deduction equivalent to 50% of direct costs of the modifications made as reasonable accommodation.
(5) Income tax exemption	Kenya only: Income tax exemption for registered disabled people with annual income not exceeding KES 150,000. By application.	- Must be registered (see 1). - Other issues unclear, not investigated in this study.
(6) Job matching services	Bangladesh: Job fairs organised by government and private sector, linking disabled jobseekers with potential employers. Government-led job fairs reported as numerous.	- No follow-up beyond initially linking a disabled jobseeker with a potential employer. No evidence of jobs being secured or retained. - No coordination between different implementers to increase effectiveness.
	Kenya: Job placement services where CVs of disabled jobseekers are received by the NCPWD and forwarded to potential employers. Implementation dependent on the motivation of the NCPWD staff.
(7) Training opportunities	Bangladesh: 5% enrolments at training institutions to disabled people. KIs report that the target is not met, attributing to the insufficient number of applications.	- Key underlying reason for insufficient applications and poor readiness is low educational levels among disabled people who have been excluded from education systems. - Many trainings offered are outdated and not industry-ready.
	Kenya: 60% of enrolments in the Vocational Rehabilitation Programme for disabled students. Other public institutions also target disabled applicants. KIs report that the target is not met and that there is poor readiness among disabled students.

below summarises the key findings of the study, highlighting (1) key provisions as identified in document analyses, (2) implementation details reported in document review and key informant interviews, and (3) reasons for implementation gaps, drawn from key informant interviews. Further evidence and analyses are detailed after the table.

3.3.1. Registration as a Disabled Person

Although not directly linked to employment, this aspect is crucial because unless they are registered, disabled people are excluded from all provisions listed below, in both countries [11,47].

In Bangladesh, registration is by application to the Upazila (sub-district) Committee or the Urban Committee and includes disability assessments by designated medical professionals. A concurrent route for registration seems to be through the Disability Detection Survey, a database maintained at the Department of Social Services (DSS) that is also based on an application and a medical assessment. Although it has been reported that door-to-door registration was piloted in 2012/2013, the current application process is through a website [33]. A 2021 news article reported that DSS personnel are present at the upazilas, including health professionals, to support this registration [48].

Successful applicants are provided with a national disability card, making them eligible for government support systems that include a means-tested allowance, reserved seats on public transport, and an educational stipend. However, key informants indicated that additional registration and verification may be required for disabled students to access the stipend.

In Kenya, too, all supports for disabled persons are dependent upon them registering with the NCPWD, after which they are provided with a disability registration card. Unlike in Bangladesh, disabled Kenyans need to renew this registration every five years.

R: With that card, you’re able to access, for example… you cannot access social assistance. Without that card, you cannot get exempted from tax. Without that card, you cannot import your car, you cannot get assistive devices from NCPWD, you cannot get bursaries, without that card. You cannot be appointed in these appointments without that card.

I: So, you’re not counted towards that 5% or anything without it?

R: Yes, yes, not without that card.

(OPD key informant, Kenya)

Implementation gaps and potential reasons: Key OPD informants in Bangladesh reported numerous barriers to registration, particularly for people not affiliated with OPDs or NGOs that would assist with the process. Barriers echoed in other sources include financial and access difficulties visiting the field offices and accessing the internet, and health professionals not being trained enough to recognise all impairment types [48]. Barriers to registration would lead to fewer people being registered as disabled, subsequently excluding more disabled people from government support systems. In contrast to civil society key informants, government interviewees expressed satisfaction with the proportion of people registered.

However, the contention between stakeholders regarding the country’s disability prevalence (described earlier) makes it difficult to establish the extent of registration. As a result, little is known about how many disabled people are excluded from provisions entitled to them by law. This limits government accountability and the advocacy power of OPDs and activists. Similarly, an underestimation of need (of how many disabled people require support) could lead to reduced budget allocation for disability programming.

In Kenya, three major barriers to registration were identified. The first barrier relates to challenges with the medical assessment required as part of the application. These include health practitioners failing to characterise mental and psychosocial disabilities as a disability, and failing to waive medical fees despite clear regulations in the legislative framework [49]. Second is the bottleneck in the approval stage at the Director of Medical Services, a single office in Nairobi. This contributes to months of backlog of applications needing approval, and applicants face accessibility and financial challenges in reaching the office, especially because they need to renew this registration every five years. The third challenge was the long duration between submitting the application and being approved. This duration has been reported to be as long as one year, far exceeding the six weeks promised in the NCPWD’s Service Charter [50].

Key Kenyan government informants confirmed that the number of disabled persons who are currently registered with the NCPWD has not been made public and speculated that these data are not analysed or used for programming.

3.3.2. Disability Allowance and Benefit Schemes

In Bangladesh, the cash transfer “Allowances for the Financially Insolvent Persons with Disabilities” is the only disability-targeted benefit provided under the National Social Security Strategy (NSSS) [51]. The NSSS consolidated over 140 social assistance programmes in a life-cycle approach: Disabled children benefit only through a stipend provided to schools (approximately 100 Bangladesh Taka (BDT) per child), and working-age (19–59 years) people with severe disabilities receive the allowance and then transition to the Old Age Allowance (BDT 300 per month) (where 100 Bangladesh Taka (BDT) is ~USD 1; BDT 300 is ~USD 3.50; BDT 36,000 is ~USD 420; BDT 24,000 is ~USD 280) [51].

The allowance is means-tested and by application. An applicant’s household must not earn an annual income exceeding BDT 36,000. Other criteria include being over six years of age, being a local resident of the area, and, notably, being approved by the selection committee in the area. In addition to the eligibility criteria, there are several “priority criteria” that include being elderly, landless, or homeless; having multiple impairments; having a psychiatric condition; and being autistic or blind [52]. Beneficiaries may not receive any other assistance from the government. At the time of the interviews, approved applicants received BDT 750 (~USD 8) every month. Some government documents mentioned plans to increase it to BDT 1500 (~USD 17) and limit the income criteria to the individual’s income rather than that of the household [51].

In Kenya, there are 18 social assistance programmes, of which the Cash Transfer for Persons with Severe Disabilities (PWSD-CT) is the only disability-targeted programme, targeting disabled children and adults requiring full-time caregiver support [53].

Entry into this programme is primarily through recruitment, not by application. Key informant interviews indicated that this cash transfer programme is considered a community-based programme where community members and village meetings (baraza) play a key role in identifying and verifying households that are eligible. At barazas, the community receives information about the programme, and are asked to go through a questionnaire Household Targeting Form [54] to identify households deemed most “needy”. Key government informants described this as “a whole process that keeps changing to fit the time”.

Eligibility is considered by household: One household may receive only one benefit, even if more than one member of the household has a disability. Crucially, each household may not receive another benefit from the government—for example, upon reaching 70 years of age, the disabled beneficiary must choose to receive either the disability benefit or the old persons benefit. In contrast, the Older Persons Cash Transfer (OPCT) beneficiaries are individuals, so two individuals in the same household may both receive this. Lastly, no member of the household may be a government employee (which contradicts the Social Assistance Act 2013 that stipulates “no known source of income or support” [55] (p. 13)—this is an aspect that is checked and verified by the community members. When asked whether this eligibility criteria posed a barrier to employment, a key informant responded that these beneficiaries were “not the type [of people who] would be employable”. At the time of interviews, approved applicants receive 2000 Kenyan shillings (KES) (~USD 20) that are dispensed through county banks every two months.

Implementation gaps and potential reasons: In Bangladesh, the 2018 NSSS Action Plan reported the number of disabled persons receiving this benefit as 0.75 million people [46]. As described earlier, stakeholders disagreed about the number of disabled people in Bangladesh, making it challenging to gauge the coverage of this programme. Nonetheless, key government informants stated confidence that there was good coverage of the programme, while other stakeholders described it as an overstatement.

“[The] government is now saying […] 50% of disabled people are under Social Safety Net. When in fact, it’s not even 2%. It can’t be. It can’t be!”

(OPD key informant, Bangladesh)

Key government informants reported increases in government budget allocations to disability-related programmes. This is evident in documentation showing an increase from BDT 6.93 billion in 2017/1018 to BDT 16.20 billion in 2020/2021 [56]. However, many civil society interviewees responded with caution, noting that this growing budget is solely with the Ministry of Social Welfare, when more advancements can be made if disability were to be recognised as a cross-sectoral issue. Moreover, they were wary of this growing budget being branded as evidence of success and fulfilment of the needs of disabled people.

“If you hear the government people talking, they’ll talk about the number of this, and the amount has increased to such and such amount, and this has also increased, and the allocation of that school and that place has increased, but these are all safety net programmes. But what about rights? What about other development areas?”

(NGO/OPD key informant, Bangladesh)

In Kenya, the coverage of the disability allowance programme was clearer. Key government informants estimated the coverage at about 47,000 households in 2019, nearly reaching the reported target of 47,200 households. Although this is far fewer when compared to the 500,000 beneficiaries receiving Older Persons Cash Transfers, key government informants claimed that there is good coverage of those in need. They reported that in some counties, no eligible households could be found. This is likely linked to the stigma around disability, as well as a lack of transparency in eligibility criteria, as noted by a key informant from a Kenya human rights organisation:

“Generally, most complaints we received often are about the Article 28 of the Convention on social protection and social assistance. Because the Government has begun that programme of giving an amount of money for some families with persons with severe disabilities, but it’s not very well understood. So, many times you have people with disabilities coming and saying, ‘I have a disability, but I’m not getting the allowance.’ So, there’s a lack of information about the eligibility criteria, so we receive many, many complaints falling on that.”

(Human rights organisation key informant, Kenya)

3.3.3. Employment Quota

An employment quota is an obligation, sometimes required by law, to set aside a percentage of jobs for a particular group of people [57]. It is a common tool used in many countries to promote employment among disabled people, but with varying success, as discussed later [57,58].

In Bangladesh, there is a provision for a 10% quota in public-sector jobs to be shared between orphans and disabled people. The Disability Act 2013 does not specify this quota, and key informants noted that it was a government circular that specified the 10%.

A similar provision exists in Kenya and is clearer—the Constitution in Kenya provides for a 5% quota for disabled persons in all public elective and appointive bodies [12]. In addition, the Disability Act (2003) extends this to any “casual, emergency, and contractual position in public and private sectors” [11]. This provision fits into the broader disability mainstreaming activities described by the NCPWD as efforts to integrate disability into all government policies, plans, and programmes by providing disability awareness training and establishing committees for oversight [59].

Implementation gaps and potential reasons: According to key Bangladesh informants, no guidance has been provided regarding how the 10% quota for public-sector jobs is split between orphans and disabled persons. Moreover, no documentary evidence could be found regarding how this quota system is being implemented or how it is (or will be) monitored. Interviews indicated a range of issues that include disregard for the provision.

I: So, how does it work? Nobody is following this 10%?

R: No.

I: Nobody?

R: No. No one. Even the government.

(OPD key informant, Bangladesh)

This contrasted with government key informants who reported implementation but a reluctance to monitor or report it.

R: Implementation is a problem because people are reluctant, Government is reluctant. It’s difficult.

I: Why are they reluctant? But there is support from the top [interrupted]

R: That’s why, you know, if you compare with government and NGO, the government is the highest employer for particular disabilities. Last 10 years there’s a huge number of persons with disabilities included in education sector, Social Welfare Ministry, health sector, of course, of course, it’s a big number.

I: Where can I find that number, has somebody done any studies?

R: No, nothing because… we know because we are involved with the government. But there are no statistics.

(Government key informant, Bangladesh)

More crucially, the quota system had been discontinued in 2018, in the run up to the elections. This was reportedly in response to student demonstrations that claimed over half of the available government jobs were being reserved for protected groups, including disabled persons, women, and indigenous communities [60]. Civil organisations protested this abolition because despite the lack of clarity, the existence of the provision enabled them to advocate for their members to be considered for employment. Others said the quota for disabled people need not be as high as 10% if it is clear and better implemented.

“We don’t want 10%. We want a small percentage but with perfect implementation. You have [inaudible] percent but there’s no implementation of even 1%. So, we don’t want that. If there’s 1% [quota], this needs to be implemented.”

(OPD key informant, Bangladesh)

Key informants described anticipation of a new policy for the quota system. However, at the time of this paper’s submission, there has not been a new policy in the quota system, and the issue was the focus of protests in the country [61].

Implementation of the quota system in Kenya, too, was difficult to ascertain. It is among the disability mainstreaming indicators that each county agency is required to report to the NCPWD every quarter:

“Ensure progressive realisation of attaining the 5% on elective, appointive, contractual of all recruited personnel in appointments, employment/promotion for persons with disabilities.”

[59]

However, all key informants agreed that these figures were not being analysed by the NCPWD and were critical of the fact that no figures had been published. Without these figures, advocates are unable to monitor the fulfilment of the 5% quota provision. One key government informant implied that not all public institutions obliged with this quarterly reporting, revealing that for most who did report, less than 1% of employees were disabled persons. This statistic could not be verified, as no public record could be found. NCPWD representatives stated that their procedure was to provide feedback on an institution’s quota figures only to that institution “directly”.

According to key informants, another reason for gaps in implementation was the ambiguity in the Constitution: “…ensure the progressive implementation of the principle that at least five percent…” (emphasis added) [12]. They noted that this wording prevents the NCPWD or watchdog organisations from instructing public organisations to increase their efforts, because all can claim to be “progressively implementing” it, despite making no meaningful effort or advances.

In the Kenyan private sector, there were good examples of inclusive practice. However, some key informants speculated that these private companies may be seeking persons with certain disabilities that they described as less severe or “easy”. According to a key informant from one private company, they employed 2.1% (117) of disabled employees at the time of the interviews. Though their narrative indicated employees with various disabilities, a breakdown by impairment type was not ascertained during the interview.

3.3.4. Tax Incentives for Private Employers

Tax incentives are used as a tool to incentivise private employers to hire disabled people. These may be in the form of tax abatements or a tax credit, and may be to offset employers’ expenses in implementing accommodations or workplace modifications [57].

In Bangladesh, the Disability Act (2013) provides for tax incentives for private employers. However, these are “upon framing proper guidelines” and “subject to prior approval from the National Board of Revenue” [47]. According to key informants and documents reviewed, no guidelines have been made available to date, and the National Board of Revenue grants these on a case-by-case basis.

Likewise, the Disability Act (2003) in Kenya provides tax incentives for private employers where they can apply for (1) a tax deduction equivalent to 25% of the salary of disabled employees and (2) a tax deduction equivalent to 50% of the direct costs of the modifications made as reasonable accommodation.

Implementation gaps and potential reasons: Findings highlight three key reasons contributing to implementation gaps. The first reason echoes the limitations with registration—tax incentives are available only if the disabled person employed is registered as a disabled person [11,47]. In Kenya, they also need to be “…accredited with the [National] Council [for Persons with Disabilities] as to their disabilities, skills and qualifications” [11]. As discussed earlier, there are several barriers to registration in both countries, leaving many disabled people unregistered.

Secondly, there was a lack of clarity among many key informants from both Kenya and Bangladesh about the details of the tax incentives regarding eligibility, process, and the amount that can be claimed. An interviewee from the Federation of Kenya Employers surmised this lack of awareness as widespread among most private employers. This is echoed by key informants in Bangladesh who reported that most private employers remain unaware of this provision, adding that sensitisation is difficult given the absence of guidelines about the amount and eligibility.

I: Are you finding that it’s enough incentive?

R: No, not at present, because, as you know, it’s still a bit vague. If it could be further clarified in terms of detail, then we could probably make a bit more informed opinion about it. Right now, there’s a lack of awareness about it and a lack of clarity.

(Private sector key informant, Bangladesh)

Lastly, in Kenya, the rebate application and process were described as complicated, and the amount too low to be worth the effort. This was confirmed by key informants from private employers known for inclusive practices. Other key informants remarked on how bigger private companies can afford to forego this rebate, indicating the usefulness of this provision to smaller organisations.

In sum, with the current lack of clarity and added bureaucracy, the tax rebates did not offer sufficient incentives to employ disabled persons, and there is continued need to appeal to goodwill and corporate social responsibility.

3.3.5. Income Tax Exemption

In Kenya, the Disability Act (2003) provides for disabled persons to apply for and obtain exemptions from income tax [11]. A 2016 policy analysis reported that this is applicable for disabled persons earning an annual income not exceeding KES 150,000 Kenyan shillings [62], and this was corroborated by key informants. Experts involved in formulating the current act described how the original draft did not include this cap but that it was instated during negotiations among policy actors. Therefore, at present, disabled persons earning an annual income exceeding KES 150,000 are not eligible for this income tax exemption. There is no such provision in the Bangladesh policy framework.

Implementation gaps and potential reasons: Since this study did not collect interview data from disabled beneficiaries, evidence on this provision’s implementation is limited. Some key informants in Kenya were disabled employees themselves. When asked to reflect on this income tax exemption, they described it as a welcome relief. Further inquiry is necessary to detect whether the KES 150,000 threshold influences disabled persons to remain within a particular income bracket, indirectly contributing to limiting their career progression.

3.3.6. Job Placement Services and Job Fairs

The NCPWD in Kenya is expected to maintain a database of registered disabled persons, along with their skills and qualifications, which are used to identify suitable job vacancies. NCPWD staff interviewed as key informants described this service as ongoing, reporting receiving an average of four CVs a day, which they pass on to potential employers.

In Bangladesh, there is no legal requirement of a job placement service. However, many key informants described job fairs as a commonly used tool to link disabled job seekers with potential employers. The JPUF (National Foundation for the Development of Disabled Persons) noted conducting one annually and the Ministry of Social Welfare one every six months. Job fairs are also held by private-sector organisations such as the Bangladesh Business Disability Network (BBDN).

Implementation gaps and potential reasons: According to some key Kenyan informants, the jobseekers’ database maintained by the NCPWD has been dormant, noting that the service’s effectiveness was dependent on the motivation of the NCPWD officer managing it at the time. In addition, NCPWD staff interviewees agreed that the current service is limited in that it is Nairobi-based, and that they are unable to verify whether those disabled persons whose CVs are provided to potential employers are eventually employed. Notably, they were unaware of a similar job-matching service provided for the general population by the National Employment Authority, which may provide opportunities to integrate or share responsibilities.

In Bangladesh, most key informants were sceptical about the reach of job fairs reported by the Ministry. They noted weaknesses in government-led job fairs such as poor follow-up of whether job fairs led to employment uptake or job satisfaction, mentioning high attrition rates among those offered jobs at these fairs. Implementers of private-sector job fairs also noted limited success in conducting job fairs in districts outside of Dhaka, attributing this to misperceptions in rural communities about disabled people. Overall, interview data from Bangladesh indicated that job fairs were frequently conducted by various organisations, though there was little evidence of coordination or follow-up to monitor effectiveness as a tool to promote the employment of disabled people.

3.3.7. Training Opportunities

In Bangladesh, the National Skills Development Policy (2011) states a target of 5% of enrolments at government training institutions be for disabled persons [42]. This is identified as a quota in some documents [63]. This policy highlights apprenticeships and recognition of prior learning, both of which could facilitate employment pathways for disabled persons [42]. The Recognition of Prior Learning initiative provides formal recognition of skills gained through previous learning, training, or work experience, which earns them credits towards their skills assessment and ultimately bolsters the chances for employment [60]. Other notable strategies in Bangladesh include engaging the Industry Skills Councils (ISCs) to promote industry-driven skills development, and disability inclusion metrics being part of the performance metrics at the Directorate of Technical Education (DTE) within the Ministry of Education.

In Kenya, key government informants described 12 institutions established as part of the Vocational Rehabilitation Programme, with a target of 60% of students who are disabled persons. These are residential institutions throughout Kenya, managed through the Social Development offices (under the Ministry of Labour), and delivered by trainers with Special Educational Needs (SEN) certification. Upon completion, students undergo the government trade test and are provided with tools under the Tools for Trade programme by the NCPWD. Additionally, through the Ministry of Education, the vocational and technical training programme oversees four publicly funded institutions throughout Kenya for disabled persons.

Implementation gaps and potential reasons: According to key Bangladesh informants, vocational training institutes are numerous, and the extent of and commitment to inclusive practices are varied. The DTE oversees about 100 institutions and colleges, among which many staff have typically undergone short trainings on inclusive education. Additionally, about 80% of training institutions are private, with 270 registered training organisations with various levels of adherence to the 2011 National Skills Development Policy. At the time of the interviews, the National Skills Development Council had recently been granted more regulatory and oversight power, which was anticipated to improve the monitoring adherence to the Skills Development Policy.

Nearly all relevant key informants in Bangladesh report that the 5% quota (or target) for disabled persons is not met, citing insufficient numbers of applications. As many pointed out, it is insufficient to establish a quota and expect to reach it without addressing gaps in education that have persisted.

“If you haven’t prepared the ground, but you have given the quota, it doesn’t work. If the primary schools and secondary schools are not accessible, and people are not motivated enough to join them, then there’s a problem in having this 5% quota in technical education. So, if people are not literate and the persons with disabilities don’t have access to education, and you’re thinking ‘I will be punished if I don’t hire them’. Sorry. Demand and supply have to match. That’s where the biggest glitch is.”

(INGO key informant, Bangladesh)

Key informants in Kenya report similar issues—training institutions report that though their training is certificate-level, they often need to provide basic education when some disabled students first join. A key informant from the DTE described additional challenges when inclusive educational regulations were introduced in 2018, where non-disabled students were added to these public institutions previously meant only for disabled students. This led to difficulties in integrating them, and further research is required to investigate this in more depth.

In Bangladesh, the various trainings offered under the Vocational Rehabilitation Programme (e.g., electrical, welding, hairdressing, and IT) were described as “archaic” because the tools and machinery taught in training are no longer used in factories, indicating a gap in industry-driven skills development.

4. Discussion

The results show encouraging progress in promoting employment opportunities for disabled people in Bangladesh and Kenya. Provisions for inclusive opportunities are framed by progressive legislative and policy frameworks and international commitments. The study has identified key policy provisions on employment quota, social assistance, job placement, tax rebates for employers, and training opportunities for disabled people in both countries. However, the study has also revealed that despite the efforts, implementation of these policies in both Bangladesh and Kenya is limited.

Implementation gaps in disability policies are not unique to these two countries, nor are they challenging only in LMICs. For example, while disability mainstreaming is a core principle of the EU Disability Action Plan, member states such as France and the UK experienced difficulties in consistently implementing this for reasons that include inadequate disability knowledge among practitioners and the inaccessibility of facilities [64]. One systematic review found that even for programmes targeting disabled people as key recipients, their participation was below the need in many countries, often due to difficulties in disability assessment and application processes [65]. This is indicative of complexities in implementing even disability-targeted programmes, and how it is persistent in several countries.

This study has highlighted a key bottleneck in access to support services for disabled people—they are able to access all legal provisions only if they have been registered as a disabled person. Thus, inefficiencies and gaps in registration result in excluding a significant proportion of disabled people from supports to which they are entitled. These findings echo those in other developing contexts. For example, in the Maldives, too, disabled people can access their entitled support only after undergoing disability registration, a process complicated by misperceptions of eligibility, stigma, and mistrust in the system [66]. This means that only 25.6% (i.e., one in four) of disabled people had registered as a disabled person, which in turn affected their access to the disability allowance and employment opportunities [66].

Quota systems have been frequently used to encourage diversity in the workforce. This has applied to women [67], minority ethnic groups [68], and disabled people [69]. Establishing employment quotas is widely used in many countries around the world as an approach to promote the employment of disabled people [70]. Results of this study indicate that the quota systems in Bangladesh and Kenya cannot claim adequate implementation, with some key informants postulating whether introducing levies (or the “stick” approach) could improve this. Examining the quota system in China, Liao [70] highlights how their implementation is also plagued by challenges, such as low compliance and a preference for paying levies rather than hiring disabled people. This provides essential learning for Bangladesh and Kenya, whose quota systems are still under-formed in comparison. Implementation requires effort, and monitoring mechanisms are crucial.

So, what is affecting implementation? Is it low awareness of rights and provisions, leading to low demand for services and provisions? A 2008 survey of disabled women in Bangladesh found that the vast majority were unaware of services, laws, or policies for disabled people [71]. Indeed, key informants from both countries described not receiving enough applications from disabled people to allow them to meet the quotas for employment or training, thereby displaying a gap in implementation. We posit, however, that it is too simplistic to offload the responsibility to rights-holders. As another key informant pointed out, the failure occurred earlier, in excluding disabled people from education, which is often a result of inaccessible facilities and learning materials, lack of trained educators, and negative attitudes in communities [20,44,72]. Gaps in education contribute to gaps in employment—again, a phenomenon not specific to these two countries, but a persistent challenge in developed countries as well [64].

Turning to Hogwood and Gunn’s concepts, we argue that there is policy non-implementation in both settings. That is, the policies promoting employment for disabled people have not produced the desired outcome because it had not been put into effect as intended. For example, key provisions such as the disability allowance and employment quotas have not been implemented in full because of several blocks in steps prior, such as disability registration.

Further, our findings indicate that the reason behind failing to achieve the desired policy outcome is overly optimistic expectations—one of four reasons for policy failure documented by Hudson and colleagues [73]. The five interacting factors contributing to over-optimism were initially identified by the UK National Audit Office [74], four of which are evident in this study’s findings. One, it was based upon a weak evidence base: The number of beneficiaries (i.e., disabled people) was not factored in adequately in either country, and there were limited monitoring mechanisms preventing timely information of costs and coverage. Two, the complexity was poorly understood: The challenges faced by disabled people in accessing the provisions were underestimated, as were the challenges in delivering programmes in accessible ways. Three, policies and linked programmes are approved despite gaps in planning to secure political gain: Employment quotas were introduced (and reversed in Bangladesh) without establishing implementation processes or clarity for employers regarding incentives and responsibilities. Four, there was over-optimism about the ability to align views of different stakeholders: Policymakers were invested in establishing the provisions, government implementers were keen to demonstrate some action, and private employers were encumbered, while advocates wanted to see impact. As a result of these interlinked factors, employment policies for disabled people in Kenya and Bangladesh have failed, mainly because they have not been put into effect as intended—i.e., bad execution [22,69].

As with any study, there are several limitations. First, this study relied on key informant narratives and reports to assess implementation. Quantitative assessment of coverage of these programmes would provide important evaluative information to strengthen their implementation. Second, the sample did not include beneficiaries or mid-level implementers. This would have deepened the understanding of the reasons behind implementation gaps, highlighting areas for future research.

This paper highlights the complexity of current efforts to improve employment opportunities for disabled people. Experiences from Bangladesh and Kenya generate important data to inform the process of developing and implementing disability-specific policies in other countries, particularly those targeting poverty alleviation and social protection in LMICs [75]. Indeed, although several human rights instruments and government policies should be commended for their initiative in promoting the rights of disabled persons, there is an implementation gap between the positive legislative picture and the situation on the ground [76], showing that there is still a long way to go before everyone understands disability to be a human rights issue, not a charitable issue [77,78].

5. Conclusions

The study identified key policy provisions on employment quota, social assistance, job placement services, tax incentives, and training opportunities, all of which were designed to promote employment for disabled people. However, implementation is limited in both Kenya and Bangladesh, for reasons including ambiguity in policies, unavailability of data for monitoring, and lack of transparency among implementers. The data suggest that addressing these aspects would promote accountability among duty-bearers and advocacy power among rights-holders, and ultimately improve implementation and inclusion of disabled people in employment.