Disability inclusive elections in Africa: a systematic review of published and unpublished literature

Abstract The United Nations Convention on the Rights of Persons with Disabilities underscores the equal right of persons with disabilities to participate in political life. However, in Africa they are often unable to exercise their right to vote. This study sought to systematically review available evidence on inclusive elections in Africa. Findings showed that although most African countries ratified disability-focused legislation and proclaimed equal opportunities, the implementation of the legislation varies across the continent. Barriers to political participation can occur at any electoral stage and can be broadly categorised into three groups: lack of education and financial resources; stigma and negative social attitudes; and inaccessible physical infrastructure.


Introduction
Globally over one billion people experience disability, the majority of whom live in low and middle-income countries. The United Nations Convention on the Rights of Persons with Disabilities (UNCRPD) describes persons with disabilities as those 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 (International Foundation for Electoral Systems 2014b; United Nations 2011). The 2011 World Report on Disability reported that although people with disabilities have the same needs as non-disabled people, they often experience limited access to services including health, education and economic opportunities (World Health Organization and World Bank 2011).
Political participation is a fundamental aspect of democratic governance, the rule of law, social inclusion and human rights approaches aimed at eliminating marginalisation and discrimination (United Nations 2011). Political participation is often defined as the actions of private citizens by which they seek to influence government and politics (Inclusion International 2015a). Voting is one way that individuals may exercise their voice but participation in electoral processes is more than just voting: it includes the ability to take part in the conduct of public affairs, the opportunity to register as a candidate, to campaign, to be elected and to hold office at all levels of government (Inclusion International 2015a). Without opportunity to participate in politics, an individual is effectively denied citizenship and potentially rendered inconsequential to those who hold power.
The UNCRPD, adopted by the United Nations General Assembly in December 2006, is an international human rights treaty, which promotes, protects and ensures the full enjoyment of human rights by people with disabilities (International Foundation for Electoral Systems 2012a). The UNCRPD made a large step towards transforming people with disabilities from passive recipients of aid to fully empowered citizens who enjoy equal rights and protections under national and supranational laws (National Democratic Institute 2012). Article 29 of the treaty focuses on participation in political and public life. It calls on States to: ensure that persons with disabilities can effectively and fully participate in political and public life on an equal basis with others, directly or through freely chosen representatives, including the right and opportunity for persons with disabilities to vote and be elected (United Nations 2007, 21). This means that election laws and processes should not contain provisions that restrict the right to vote, while voting procedures, facilities and materials should be appropriate, accessible and easy to understand and use (International Foundation for Electoral Systems 2012a, 2014bNational Democratic Institute 2012). It also means promoting an environment in which people with disabilities can fully and equally participate in public affairs as well as form and join organisations concerned with public and political life at various levels.
In Africa, not-for-profit organisations such as the International Foundation for Electoral Systems and the Electoral Institute for Sustainable Democracy in Africa support citizens' rights to participate in free and fair elections, including people with disabilities (Electoral Institute for Sustainable Democracy in Africa 2017; International Foundation for Electoral Systems 2017). The National Democratic Institute, which is also a not-for-profit non-governmental organisation, has supported democratic institutions and practices in Africa for over 30 years (National Democratic Institute 2017). Information resources including ElectionAccess.org serve as a global clearinghouse of resources related to political participation and disability rights (Election Access 2017).

Conceptualising disability
Disability is a complex phenomenon which can be hard to define. A multitude of theoretical models exist that attempt to describe what exactly is disability, conceptualising it through nuanced lenses that are influenced by factors such as geographic location, religion, culture, medical knowledge, welfare policy, advancements in human rights and time. While this article does not aim to examine or critique these theories, it is important to acknowledge the diversity of thought that is an important driver in the lack of global consensus on how disability is understood and defined. While a number of models focus uniquely on either the individual's biological impairments or the socio-political environment as to 'blame' for the disability, the definition used by the UNCRPD and described in the first paragraph of this article relies heavily on the relational model which describes a complex interaction between an individual's impairments and the environment in which they live (Goodley 2016). This approach identifies a mismatch between an individual's impairments and the prevailing environment as a dynamic interaction that results in the disability that hinders their full and effective participation in society on an equal basis with others.
The prevailing environment in which a person lives includes not only the physical features but also the political, economic and social features. The World Health Organization's (WHO) International Classification of Functioning, Disability and Health (ICF) classifies the environmental factors that shape people's lives into five groups that may promote or hinder how an individual may experience disability: products and technology; natural and built environment; support and relationships; attitudes; and services, systems and policies (WHO 2017). It further identifies an individual's personal characteristics such as gender, education, social background, self-efficacy and so forth as potential mediators to their experience. These characteristics, operating within the prevailing environment, interact with an individual's physical impairments to hinder or promote their ability to participate effectively in political processes.

This review
Although participation of people with disabilities in political processes is increasingly recognised as a human rights issue, little is known about how the key UNCRPD principles are translated into day-to-day practice, particularly in low and middle-income countries (World Health Organization and World Bank 2011). It is also unclear how factors such as the prevailing local environment and individual characteristics that interact with impairments to create a disability may promote or hinder an individual's political life.
In response to this knowledge gap we conducted a systematic review of published and unpublished literature with the aim to understand, first, what experiences and practices have already taken place in countries to support political inclusion of people with disabilities and, second, what effect they have had on the participation of people with disabilities in public and political life at local and national levels. This review was conducted as an integral part of the Political Participation project implemented by the international non-governmental organisation Sightsavers and focused specifically on countries in Africa.
The questions we sought to answer are as follows: (1) What laws and policies exist in countries in Africa to ensure the participation of people with disabilities in national, district and local elections? (2) What barriers to participation of people with disabilities in electoral processes in African countries have been documented? (3) What documented approaches and interventions have been employed in Africa to increase the participation of people with disabilities in elections? (4) What impact has been documented in the literature on the interventions applied and how has it been measured?
The review aims to identify and include as much evidence as possible that may contribute to building a robust evidence base. Evidence will not be excluded on the basis of how it conceptualises or defines disability, although such definitions will be referenced to interpret and position the data and to understand how theoretical models are being used in this field. Where possible we sought to understand how factors that promote or hinder participation fit within the environmental framework suggested by the WHO's ICF as described earlier.

Search strategy
We searched for peer-reviewed literature from electronic databases including EBSCO, Ovid, Science Direct and EPPI Centre, drawing on thesaurus and non-thesaurus words as appropriate. We further supplemented these data with grey literature through a search of the electoral commission (EC) websites, service providers and donor organisations working on related issues. We also contacted authors and organisations, such as the African Union, The European Union External Action and Disabled Peoples' International, for relevant reports. We also reviewed reference lists of included reports and papers to check for any missed sources.

Sifting and selection
Titles and abstracts of all identified sources were reviewed by one author against the predefined inclusion criteria. Documents containing information on legislation, electoral guidance and political participation of people with disabilities in Africa, written in English, French and Portuguese, and published from 2006 onwards were included. The Africa region was defined as 54 countries of the African continent, including Eastern, Middle, Northern, Southern and Western Africa. 1 The year 2006 was chosen as a cut-off point, as this is the year when the UNCRPD was signed, and when most governments initiated efforts to improve the political participation of their citizens with disabilities. Reports written in languages other than English, French or Portuguese, those that were not focused on disability inclusive political participation, those that were from outside Africa and those published before 2006 were excluded. Searching and screening was undertaken in March 2016.

Data extraction and quality appraisal
Two authors independently extracted data and assessed the quality of the documents that met the inclusion criteria. The data extraction and appraisal form was developed by the authors and was organised in a qualitative manner according to the research questions.
Critical appraisal is the process of systematically reviewing evidence to examine its trustworthiness, and relevance within a certain context. Judgement is made on the methods used by the author(s) to collect data, taking into account its source, representativeness, coherence and so forth. Evidence shows that biased results due to poorly designed and reported studies can indeed mislead decision-makers. Therefore, appraisal allows decision-makers and policy-makers to use evidence reliably and efficiently (Mhaskar et al. 2009). For non-research papers/reports, there are generic critical appraisal tools (Cottrell 2011;Woolliams et al. 2009); however, it is important to identify or develop a tool specific to the type of literature being critiqued (Aveyard 2011).
For the purpose of this systematic review, which included non-research and research reports, we developed a critical appraisal tool. The appraisal checklist was based on three validated tools: the CERQual approach (Lewin et al. 2015), the Critical Appraisal Skills Programme tool (Singh 2013) and the Authority, Accuracy, Coverage, Objectivity, Date, Significance tool (Tyndall 2017). Quality assessment of each report was based on the following criteria: reliability, accuracy, methods, relevance and coherence. Based on these criteria we were able to attribute a level confidence to the conclusions of the report.
A document was attributed 'high confidence' if the authors reported details for all of the criteria mentioned earlier and therefore it was highly likely that the review finding was a reasonable representation of the phenomenon of interest. 'Moderate confidence' was attributed if the authors reported the methods used to collect the data, if the findings were applicable to the context specified within the study and if the document discussed how findings compared and contrasted to other research in the field. Moderate confidence implied it was likely that the review findings were a reasonable representation of the phenomenon of interest. 'Low confidence' was attributed to a document if the authors did not report most of the criteria noted, implying that it was not clear whether the findings were a reasonable representation of the phenomenon of interest. It should be noted that non-research papers often exclude a comprehensive description of methods and so they may be attributed low confidence due to the lack of description rather than on the basis of the methodology itself. Therefore, it is important to stress that an attribution of low confidence may not necessarily indicate poor methods used to collect the data, but rather a lack of clear description of those methods within the report.
It should also be noted that the legal documents included in the review were not critically appraised.

Data analysis
Following extraction, the data were analysed using a narrative synthesis approach where the papers were read and the data were extracted into the key themes broadly corresponding to the research questions. Data relating to barriers and facilitators of participation were analysed using the ICF environmental factors as a framework. As the sources included in the review were heterogeneous and contained few quantitative data, no statistical meta-analysis was possible. Pertinent findings were presented alongside each other to allow for comparison and contrast.
When analysing data from included reports, we divided country findings across income and regional classifications, following the World Bank's (1998) World Development Report and policy research working paper (The World Bank 1998). Income classifications included high-income countries, upper middle-income countries, lower middle-income countries and low-income countries (The World Bank 2017). 2

Search results
The search retrieved 54 unique documents that met the inclusion criteria. Of these, 28 documents were identified through databases, 24 documents were identified through the review of references and two were received from the organisations contacted for grey literature.

Characteristics of documents identified
Documents identified were diverse and included legal documents, electoral codes and guidance, position papers, case studies and peer-reviewed articles describing or evaluating the participation of people with disabilities in electoral processes. All documents included in the critical appraisal were attributed low confidence using the criteria described earlier. The majority of the research reports did not provide sufficient information on the methods used to collect the data; it was not clear whether the findings were applicable to the setting specified or a wider context; and there was no discussion on how the findings compared and contrasted to other research. The non-research papers (e.g. Frequently Asked Questions documents) were attributed low confidence because the authors did not mention the sources of information included in the report (see Table 1).
Documents identified covered various issues and many described more than one thematic area. In total, 30 sources contained information on strategies to support inclusive political processes, 29 of the documents focused on legislation and policies related to political participation and disability, 14 documents detailed barriers faced by people with disabilities in public and political life while five of the documents described specific interventions to make political processes more inclusive. Most of the data contained within the documents were descriptive with few (if any) data on the impact of interventions on the participation of people with disabilities in elections.
Most documents contained information from low-income countries, followed by lower middle-income countries. We found most sources on Ghana (eight documents) and Nigeria (seven documents). Five documents reported information from Kenya; three sources each focused on Sierra Leone, Liberia and Tanzania; two documents each were on Zimbabwe, South Africa and Mali; and two documents reported information for more than one country in the Africa region. One of these documents referred to Kenya and Zanzibar; the other document described experiences from Burkina Faso, Burundi, Cameroon, DRC, Mauritius, Mauritania, Cote D'Ivoire, Ethiopia and Uganda. Nine documents focused on global approaches and practices relevant to the Africa region. No documents were identified that described the impact of any intervention on the participation of people with disabilities in political processes.
The majority of documents refer to people with disabilities as one homogeneous group. However, where distinctions between sub-groups were made these have been reported here.

Global legislation
As of July 2017, 46 of 54 African countries (85%) have ratified the UNCRPD and an additional three (6%) have signed it (Cameroon, Chad and Libya). It has been neither signed nor ratified by five countries (9%): Botswana, Equatorial Guinea, Eritrea, Somalia and South Sudan. The optional protocol has been ratified by 28 countries (52%) and signed by an additional 12 countries (22%). This has been neither signed nor ratified by 14 countries (26%)

National and local legislation
The reviewed documents identified national legislation referring to the participation of people with disabilities in 18 African countries. Most of these were low-income countries, including Burkina Faso, Burundi, DRC, Guinea, Liberia, Mali, Niger, Sierra Leone, Uganda, Tanzania and Zimbabwe. Six of the 18 countries were lower middleincome, including Cameroon, Cote D'Ivoire, Ghana, Kenya, Mauritania and Nigeria, and one country was upper middle-income (Mauritius).
Some countries referred to the protection of human rights for all, while 15 countries made provisions specific for the political participation of people with disabilities. Overall, the documents proclaimed equal opportunities for people with disabilities to elect and be elected, and referred to specific adjustments, mainly personal assistance, accessible communication and prioritisation at the polling stations. The majority of low-income and lower middle-income countries, including Kenya, Sierra Leone, Nigeria and Uganda, and one upper middle-income country (Mauritius) included such provisions in their Constitutions, while only one lowincome country (Tanzania) referred to specific legislation such as a Disability Acts.
For example, the Constitution of Kenya, Article 38, states that every citizen has the right to be involved in any political matter (Inclusion International 2015a). In Sierra Leone, the Abridgement Act 2010 states that any voter with physical disability can be accompanied by another voter of their choice, who must follow their will and keep their vote confidential (International Foundation for Electoral Systems 2012c). In Nigeria, the 2010 Electoral Act states that people with disabilities may request support from the registration officer to complete the registration form (The Commonwealth 2015). In Mauritius, the amendments of the National Assembly Elections Regulations and Municipal Election Regulations enable disabled voters to be accompanied by a close relative (Permanent Mission of the Republic of Mauritius to the United Nations 2011).
In Tanzania, Article 51 of the Persons with Disability Act 2010 states that all people with disability aged 18 years and older are entitled to vote, hold public office and otherwise participate in the political rights and opportunities without any form of discrimination. The Act prescribes the Responsible Minister to ensure that: voting materials, facilities and polling stations are accessible; relevant training is provided for all poll workers; people with disabilities have free expression of will and can be assisted by a person of their choice; people with disabilities are encouraged to stand for elections; and there are criteria and procedures for appointing qualified people with disabilities to represent (International Foundation for Electoral Systems 2015d; 'Tanzania: Persons with Disabilities Act' 2010). In Nigeria, the Independent National Electoral Commission's procedures explicitly state the need to communicate with people with disabilities in a targeted manner (Independent National Electoral Commission 2016b); and visually impaired registered voters may, where available, use assistive tactile facilities (The Commonwealth 2015). The 1997 Local Government Act in Uganda recognises the importance of the use of sign language for people with hearing impairments (Abimanyi-Ochom and Mannan 2014).
A few documents explained how these legal provisions were translated into practice. In low-income countries, legislation recognises representation of people with disabilities in the government. For example, in Uganda the 1997 Local Government Act recognises the representation of males and females with disabilities at all levels of the local government. In Kenya, 12 seats in the National Assembly were reportedly reserved for 'marginalised' and historically underrepresented groups (including people with disabilities). In the Senate, two seats were reserved to represent people with disabilities (International Foundation for Electoral Systems 2013b). In Zimbabwe, a National Disability Board was established to fulfil the provision of self-representation for people with disabilities following the Disabled Persons' Act adopted in 1992 (Permanent Mission to the United Nations and other International Organisations at Geneva 2011). In Mauritania, Article 11 of the 2006-043 ordinance recognises the importance of Disabled People's Organisations (DPOs), while Article 13 of the same ordinance notes that the State accords the designation of Public Utility to DPOs (Ministère des Affaires Sociales de l'Enfance et de la Famille 2012).
A few documents, however, referred to restrictions in political participation for certain population groups, including people with particular types of disabilities. Thus, Article 40 of law #012-2010/AN in Burkina Faso states that people with disabilities have the same civil and political rights, including the right to vote and stand for office, unless they fall into the category legally declared as 'incapacitated' (Permanent Mission of the Republic of Burkina Faso to the United Nations 2011). The Burundian Electoral Code 2014 states that all Burundian citizens aged 18 or older are eligible to vote; however, individuals with mental illness or other extenuating circumstances are deemed ineligible, along with citizens with a criminal history (Electoral Institute for Sustainable Democracy in Africa 2007; International Foundation for Electoral Systems 2015a). Article 83 (1) (b) of the Kenya Constitution, the Election Act 2011, which regulates the national elections, and the Local Government Act 2010, which regulates local elections, state that a person of 'unsound mind' may not vote (Inclusion International 2015a). Additionally, Article 99 (2) (e) of the Constitution states that a person of 'unsound mind' may not be elected as a Member of Parliament, and if the President becomes mentally or physically incapacitated they may be removed from Office. Similar provisions are made in the National Land Commission Act 2012 and the Constitution Act 2010. The terms 'unsound mind' , 'mental infirmity' and 'mental capacity' and how they are determined are not defined in these documents.

What barriers to participation of people with disabilities in electoral processes in African countries have been documented?
The review identified only one document that reported the levels of participation of people with disabilities in elections from a lower middle-income setting. A small-scale study among 120 participants in two districts in Ghana found that 58% of people with disabilities showed an interest in serving in the Unit Committees in their electoral areas; however, only 4% actually contested and won the Unit Committee elections in 2006 (Voice of People with Disability Ghana 2014). About 36% of participants were interested to become an EC registration officer or polling assistant; however, only 4% did serve as an assistant in both the 2006 and 2010 District Assembly elections (Voice of People with Disability Ghana 2014).
Most documents reviewed described the barriers to political participation faced by people with disabilities as complex and multi-faceted. The International Foundation for Electoral Systems developed a framework describing the three dimensions leading to political exclusion as societal stigma, discriminatory legal frameworks and infrastructure and positive rhetoric unsupported by political action (tokenism) (International Foundation for Electoral Systems 2014b).
The same document suggested that barriers to political participation may occur at any stage of the election cycle, including pre-election, election and post-election stages (International Foundation for Electoral Systems 2014b).

Products and technology
Inaccessible formats were identified as barriers to participation in both the pre-electoral and electoral periods. Information regarding securing identification and disability cards and registering to vote as well as communication from political parties was identified as inaccessible to many people with seeing and hearing impairments.

Natural and built environment
Physically inaccessible and unsafe buildings were identified as major barriers to registering to vote mainly in low-income countries (DRC, Mozambique and Malawi), and polling itself mainly in lower middle-income countries (Ghana and Nigeria). In countries with a history of security problems around elections, restrictions placed to reduce potential crime could also hinder participation of people with disabilities.
In DRC, election observers reported access to the registration centres as problematic due to the poor status of the local roads and the voters' need to travel as far as 20 kilometres to their nearest centres (Electoral Institute for Sustainable Democracy in Africa 2010). In Mozambique, mobile registration centres were deployed in rural areas with the objective that no voter should travel more than 10 kilometres to the registration point. However, this standard was not always met and some voters had to travel 30-40 kilometres to register (Electoral Institute for Sustainable Democracy in Africa 2010).
In Ghana and Malawi, all voters were required to appear and register in person, but there were no registration provisions in place for people with physical or sensory disabilities (International Foundation for Electoral Systems 2013b). In Liberia, some polling stations and polling precincts were reported to be located on the first floor, making them inaccessible to people with physical impairments (Electoral Institute for Sustainable Democracy in Africa 2012).
In Ghana, over 50% of respondents said that the registration centres and polling stations were not accessible to accommodate needs of people with physical disabilities, although these needs were not described (Voice of People with Disability Ghana 2014). An accessibility audit undertaken in the country in 2007 showed that the National House of Parliament was not accessible to people with mobility problems. Similarly, there is no assistive device to adapt parliamentary proceedings for people with hearing impairments. It is important to reiterate that while the National Disability Act and the UNCRPD, which Ghana has ratified, make provision for such services, the laws have not been implemented (Sackey 2015).
In Nigeria, security measures placed restrictions on vehicle movement on the day of the elections. However, this prevented many people with physical disabilities from voting (The Commonwealth 2015).

Support and relationships
Support for people with disabilities from their families, friends, communities and professional bodies often plays an important role in whether and how they participate in political processes. A study from Kenya (Inclusion International 2015b), classified as a lower middle-income country, reported that the majority of people with intellectual disabilities did not have identity cards. In many cases families or other community members had to apply on their behalf, as people with intellectual disabilities were not viewed to be equal citizens and their right to have an identity card was questioned.

Attitudes
Negative attitudes towards people with disabilities often pervade all sections of society, including people who work in registration and polling stations as well as the general population.
In Ghana, people with disabilities reported the attitude of registration officials and polling staff towards people with disabilities was not friendly and in some cases was hostile (Sackey 2015). In Kenya, some people with intellectual disabilities were accused of dishonesty when they applied for an identification card; others were denied a priority right at the polling stations, as their disability was not obvious and they were forced to wait for hours in a queue (International Institute for Democracy and Electoral Assistance 2014; Inclusion International 2015a).
General social attitudes were also thought to be strongly associated with political participation. For example, over 95% of people with disabilities in the study in Ghana reported that stigma against people with disabilities played a major role in their political marginalisation (Sackey 2015). It was explained that the negative attitudes towards disability were often rooted in cultural and social norms, where physical and sensory impairments were associated with punishment from God. In many Ghanaian communities, for instance, people with disabilities were not allowed to become chiefs, even if they were part of the social elite. Those, who did dare to contest in political elections were often mocked or treated unfairly. As a result, many people with disabilities were discouraged from engaging in politics even if they were educated, aspired to contest and had resources to finance their election campaign (Osman et al. 2008;Sackey 2015).

Services, systems and policies
In the pre-electoral period potential barriers may include discriminatory election laws, and the lack of capacity of DPOs to advocate for political rights. In the post-election period challenges occur when people with disabilities are not involved in feedback and lesson learnt reviews.

Personal factors and political participation
Personal factors also impeded participation in political processes in lower middle-income countries. Lower levels of education and limited access to financial resources among people with disabilities were reported in one study conducted in Ghana (Sackey 2015). The authors argued that many people with disabilities dropped out of school and were functionally illiterate, which impacted on their awareness of their rights and their willingness to participate in politics. The same study argued that political campaigns were expensive.
Significant resources were required for mobilisation campaigns, which included television and radio advertisements, printing of banners and flyers, and organising meetings. It was argued that many people with disabilities could not meet such costs. As a result, 90% of study participants considered the lack of financial resources to be a major challenge obstructing their participation in local politics (Sackey 2015).

What documented approaches and interventions have been employed in Africa to increase the participation of people with disabilities in elections?
The review identified a number of strategies deployed to increase participation of people with disabilities in political processes, although the effectiveness of such interventions in increasing participation (Question 4) was not discussed. Information identified on the different strategies was mainly from low-income settings including DRC, Zimbabwe, Malawi, Liberia, Niger and Burkina Faso. Different strategies were also identified for two lower middle-income countries (Ghana and Cameroon) and for two upper middle-income countries (Mauritius and South Africa).
These strategies included introducing equal opportunity legislation, awareness-raising activities and education, supporting inclusive infrastructure, and strengthening capacities of DPOs and other civil society organisations (CSOs). Some interventions focused on electoral reforms and political participation more generally, while others targeted specifically disability inclusion. Most projects included a wide spectrum of activates and many were funded by international organisations, such as the UK Department for International Development (DFID), the US Agency for International Development (USAID), the European Commission (EU) and international non-governmental organisations.

Products and technology
Strategies relating to products and technologies were only reported in Ghana. In 2000, the EC developed and piloted a tactile ballot guide to allow visually impaired people to vote in confidence. During the 2004 election, the Ghana Federation of the Disabled worked with the EC to raise awareness on the right of people with disabilities to vote as well as on the inclusion of the sign language interpretation. In the same year, members of the Ghana Association of the Blind were trained to use the tactile ballot guide and 117 people with disabilities were trained as election observers (Osman et al. 2008).

Natural and built environment
Measures to improve physical accessibility of the elections were reported for two upper middle-income settings (Mauritius and South Africa) and for one low-income setting (DRC) and one lower middle-income setting (Cameroon). In Mauritius, voting premises were adapted by ramps and by adjustable height booths to accommodate voters using wheelchairs (Permanent Mission of the Republic of Burkina Faso to the United Nations 2011).
In South Africa, people with disabilities could request a home visit from their registration officer in order to register or amend their registration form (Electoral Institute for Sustainable Democracy in Africa 2010). In DRC, only the ground floors of schools were used for registration to avoid problems with using stairs by people with physical disabilities. Assistance was also made available for older people with disabilities and illiterate voters (Electoral Institute for Sustainable Democracy in Africa 2007).
In Cameroon, a project called Accessible Elections for People with Disabilities was developed with the support of Sightsavers and several local CSOs (Republique de Cameroun 2011) and included refurbishment of 12 pilot voting booths in six regions of the country. This included building ramps, improving lighting and lowlevel tables that would allow people with disabilities to better access the facilities.

Support and relationships
Low-income countries, such as Zimbabwe and Malawi, reported investing in education activities to increase political participation of people with disabilities. Whereas in Cameroon classified, as a lower middle-income country, awareness-raising and education were reported as strategies to increase political participation. Much information targeting of people with disabilities was done through DPOs, making them important conduits of information.
In Zimbabwe, as part of the country support strategy, the DFID provided a grant for education of people with disabilities on their right to vote (Permanent Mission to the United Nations and other International Organisations at Geneva 2011). In Malawi, the EC developed specific target messages and education materials for the general public. The messages focused on the importance of voting, the new register, places to register and registration requirements (Electoral Institute for Sustainable Democracy in Africa 2010). Voter education also took place in Cameroon where DPOs were trained on the legal aspects of the presidential election and community and radio/television programmes encouraged people with disabilities to vote (Republique de Cameroun 2011).

Attitudes
The reviewed documents did not report strategies to specifically tackle negative public attitudes and stigma.

Services, systems and policies
Strengthening of the democratic function to promote inclusive elections was reported for lower middle-income countries including Ghana and Cameroon, and one low-income country (Liberia).
An inclusive election programme in Ghana funded by the DFID in 2011-2013 (Strengthening Transparency, Accountability and Responsiveness) focused on a number of initiatives (unspecified) to strengthen the democratic function including support to various civil society groups to play their role in promoting inclusive elections. In 2012, the Voice of People with Disability Ghana (Voice Ghana) established a Situation Room to monitor accessibility of polling stations on the day of the election. The project contacted 60 people with disabilities at random through telephone calls and engaged presiding officers to address any issues raised (Strengthening Transparency Accountability and Responsiveness in Ghana 2015). In Cameroon, a project called 'Accessible Elections for People with Disabilities' was developed with the support of Sightsavers and several local CSOs (Republique de Cameroun 2011). Project activities led to advocacy with the Election and Referendum Management Institution, which led to the inclusion of information about an individual's disability status in the electronic electoral register.
In Liberia, the USAID supported inclusive electoral and political processes through the implementation of a four-year multi-agency project, which focused on increased civic participation and accountability and included public education, training of the National Electoral Commission staff and inclusion observations (Electoral Institute for Sustainable Democracy in Africa 2012). No further information on inclusive observation was provided.
A number of documents mostly from low-income settings and one from an upper middle-income country (Mauritius) focused on the representation of people with disabilities in different government structures with a remit for either specific disability inclusion or, more broadly, inclusive political participation.
In Niger, the National Committee for the Promotion of Rights of People with Disabilities was established in 2011. Two National Federations of Disabled Persons served within the Committee and were in charge of monitoring and implementing the UNCRPD (Permanent Mission of Niger to the United Nations 2011). The government of Burkina Faso established a multi-sectoral Committee for Rehabilitation and Equality of Opportunities. The Committee included representatives of different ministries, civil society and technical and financial partners. It was noted that the inclusion of DPOs allowed the Committee to reflect on the issues related to the rights for people with disabilities, and their involvement in decisionmaking processes (Permanent Mission of the Republic of Burkina Faso to the United Nations 2011). In Zimbabwe, the development of the National Disability Board enabled the formulation of policies which ensured that people with disabilities lived independently and had access to all social services; it also supported an initiative to estimate the costs of the welfare and rehabilitation services targeting people with disabilities and made provision for people with disabilities to elect two senators to represent them (International Foundation for Electoral Systems 2013d; Permanent Mission to the United Nations and other International Organisations at Geneva 2011). A paper on the work of the National Democratic Institute reported on its efforts in building coalitions with other non-governmental organisations in Sudan, Lebanon, Kenya and Mali. The broader coalitions were involved in drafting policies and regulations, educating voters and monitoring the election processes (National Democratic Institute 2012). In Mauritius, disability activists and local non-governmental organisations represented people with disabilities at the National Implementation and Monitoring Committee on the UNCRPD (Permanent Mission of the Republic of Mauritius to the United Nations 2011).

Conclusion
This systematic review focused on four interrelated elements: policies and laws to ensure participation of people with disabilities in political processes in Africa; barriers to political participation with a specific focus on elections; approaches and interventions applied in Africa to increase political participation; and the impact of the interventions applied. We exhaustively searched a wide range of databases as well as grey literature and identified 54 unique documents that met our inclusion criteria. However, despite identifying a reasonable amount of relevant literature sources, we found much of it to be highly descriptive in nature, with little -if any -data on the effects of the legislation, policies or programmes applied and data on the impact of those interventions.
Overall, the documents identified were diverse in their scope and content. Nine documents focused on the global level; the remainder were country specific, with the majority of documents providing information on Ghana and Nigeria.
With regards to policies and legislation, the review shows that the majority of African countries ratified important disability-focused legislation, including provisions for improved participation of people with disabilities in political institutions. The documents proclaim equal opportunities for people with disabilities to elect and be elected, and refer to specific measures, mainly personal assistance, accessible communication and prioritisation at the polling stations. However, the review also found that the implementation of the legislation varies greatly between countries, as such it was very difficult to draw conclusions based on income classifications; and despite progressive policies and laws being in place, people with disabilities continue to suffer from discrimination and exclusion in many contexts.
There are various classifications of the barriers to political participation and most sources agree that the barriers faced by people with disabilities are complex and multifaceted. The challenges can occur at pre-electoral, electoral and postelectoral stages and often relate to inaccessible infrastructure and communication and negative social attitudes. Other factors, which often correlate with disability, such as low educational levels and poverty, further undermine participation and inclusion. Registration to vote was reported as a main barrier in low-income countries including DRC, Mozambique and Malawi; whereas accessibility challenges were reported as a main barrier in lower middle-income countries such as Ghana and Nigeria. In addition, negative attitudes towards people with disabilities and personal factors impeding participation of people with disabilities was reported as barriers in lower middle-income countries. Many documents had information on the strategies to support inclusive political processes. These included inclusive legislation and policies, adjustments of physical infrastructure of the registration and polling stations, and mobilising and empowering people with disabilities using the structures and capacities of local CSOs and DPOs. Many strategies have been implemented with strong financial and technical support from international development partners. Due to the heterogeneity of included documents and due to the lack of information it was difficult to draw any conclusions on trends by each income classification.
Although the reviewed documents described a wide range of programmes and strategies, the impact of these interventions remained unclear. The documents identified were largely descriptive, with little articulation of the intended effect and limited opportunities for measuring outcomes. In addition, all documents included in the critical appraisal were attributed 'low confidence' and had a high degree of bias, as they mainly consisted of grey literature reports, case studies, press releases and normative manuals.
Most documents did not address any specific exclusion issues, such as gender or ethnicity, and did not identify any strategies relevant to specific population subgroups. As a result, while a thorough search of the literature has been conducted and a range of examples from different countries has been identified, the review concludes that political inclusion activities in Africa are poorly documented and it is impossible to draw any definitive inferences on which interventions work, where and for whom. Most of the sources identified were of low quality, which suggests the need for high-quality research using explicit theory of change and rigorous measurements of the effect of the tested interventions.

Limitations and strengths
This systematic review is based on comprehensive searches of the literature. We conducted a search of relevant databases and websites for published documents. In addition, as part of the search strategy, we contacted experts and searched references of included documents. The review conclusions are reliable, as they are based on a rigorous methodology in terms of searching, data extraction and appraisal of documents. However, this study cannot exclude the presence of publication bias, which is that larger programmes and initiatives were more likely to be published and therefore the sources included in this review are not representative of all information available on this subject. This review may be prone to language bias as it only included documents written in English, French and Portuguese, although no documents written in other languages have been identified. In addition, the limited amount of literature found to answer Questions 2 and 4 compared to Questions 1 and 3 should be recognised. Therefore, a systematic review approach may not be the most appropriate approach to answer these objectives.
Overall, the majority of the documents included in this review did not report the impact of the approaches and interventions described. Several papers reported output level data, such as the number of education materials distributed, the number of tactile ballot jackets designed or the number of election observers trained, but provided no indicators at the outcome or impact levels. In some cases, such as in Ghana, the authors reported that the levels of participation of people with disabilities in the elections had increased but no comparative data to quantify the observed changes were provided (Voice of People with Disability Ghana 2014). In addition, the majority of documents described a wide range of activities aiming at increased political participation with no indication of the role and contribution of each individual activity to the overall programme outcome.
A number of documents noted that empowering people with disabilities led to increased political participation. However, the concept of empowerment was not defined and the outcome of empowerment was not measured. Some authors described intermediate outcomes of their interventions, mainly improved collaboration between different agencies or improved transparency and fairness of the elections. However, how these intermediate measures led to the ultimate goal of the interventions, political participation of people with disabilities, was not articulated. Overall, the reviewed documents were weak in describing their theory of change, scalability, replicability and sustainability.