Policy Development: An Analysis of Disability Inclusion in a Selection of African Union Policies

Contemporary debates in international development discourse are concerned with the non&#8208;tokenistic inclusion and participation of marginalized groups in the policy&#8208;making process in developing countries. This is directly relevant to disabled people in Africa, which is the focus of this article. The United Nations Convention on the Rights of Persons with Disabilities delineates the principles of inclusion in society. Furthermore, the African Union (AU) plays a key role in advising its Member States about disability issues, and this advice should be reflected in disability&#8208;inclusive policies. This article analyses nine policy or strategy documents produced by the AU, covering the policy domains of education, health, employment and social protection that are crucial to the inclusion of disabled people in international development. These were analysed according to seven discrete elements (rights, accessibility, inclusivity, implementation plans, budgetary allocations, enforcement mechanisms or disaggregated management information systems) using a rating scale of one to four, with four being the highest level of inclusion. The process (for example, level of consultation), the context (for example, the Sustainable Development Goals) and actors involved in the policy development were reviewed as far as was possible from the documents. None of the policies reached even 50% of the total possible score, indicating poor levels of genuine inclusion. Rights scored a highest rating, but still at a low level. This suggests that there is recognition of the rights of disabled people to inclusion, but this is not generally integrated within inclusive implementation plans, budgetary allocations, enforcement mechanisms or disaggregated management information systems for monitoring. The limited socio&#8208;economic inclusion of disability within AU policies is a lost opportunity that should be reviewed and rectified. The findings have broader ramifications for the non&#8208;tokenistic and genuine involvement of poor and marginalized groups in the international policy&#8208;making arena.


| INTRODUCTION AND CONTEXTUAL BACKGROUND
This article critically analyses the extent to which disability issues are included as a key component in the development and implementation of a range of policies developed by the African Union (AU). Inclusive policies are a critical component in the realization of the rights of disabled people as set out in the United Nations Convention on the Rights of Persons with Disabilities (UNCRPD) (Fernandez, Rutka, & Aldersey, 2017;Katsui, 2013;Mittler, 2015;UN, 2006). As of April 2017, 173 countries have ratified the UNCRPD, including most countries in Africa (UN, 2017). However, ratification is only the start of an ongoing process that reviews existing laws, policies and related programmes to determine their compliance with UNCRPD principles (Aldersey & Turnbull, 2011) and inclusion of disabled people. Given the role of the AU in guiding national policies of its Member States, analysing the extent to which its policies reflect the principles of the UNCRPD provides insight into the status of people with disabilities throughout the continent (Raymond & DeNardis, 2015;Ruggie, 2014).
The central argument presented in this article is that, with regard to disability policy-making in Africa, there is an apparent and discernible "disconnect" between the rights of disabled people and their inclusion enshrined and guaranteed through the ratification of the UNCRPD, and application of its principles in the actual process of policy development and implementation. The extent to which there has been an active and non-tokenistic involvement of disabled people as actors in the policymaking process is an important indicator of inclusion of disabled people and the inclusion of disability issues within the content of the policy.
This policy analysis was limited to reviewing AU disability-relevant policy and strategy documents. The analysis concentrated on the evidence available within these documents on the rationale for the policy or strategy subject to analysis, and the process and involvement of the actors participating in this process. It did not include face-to-face interviews with, for example, disabled people's organizations (DPOs) or their involvement and inclusion in the policy development process. This of the policies reached even 50% of the total possible score, indicating poor levels of genuine inclusion. Rights scored a highest rating, but still at a low level. This suggests that there is recognition of the rights of disabled people to inclusion, but this is not generally integrated within inclusive implementation plans, budgetary allocations, enforcement mechanisms or disaggregated management information systems for monitoring. The limited socioeconomic inclusion of disability within AU policies is a lost opportunity that should be reviewed and rectified. The findings have broader ramifications for the non-tokenistic and genuine involvement of poor and marginalized groups in the international policy-making arena.

K E Y W O R D S
Africa, disability, implementation, international development, participation, poverty eradication, public poverty, SDGs would indeed be a very profitable and insightful study in its own right, and one that legitimately warrants further research. Historically, there is strong evidence to suggest that disabled people have been systematically excluded from being active in the formulation and implementation of disability policy and practice, particularly in developing countries (Dube, 2006;Lang & Murangria, 2009). This is in direct breach of the maxim that underpins the international disability movement of "nothing about us without us" (Albert (Ed.)., 2006; Lockwood & Tardi, 2014;MacLachlan & Swartz (Eds.), 2009).
The analysis, findings and subsequent recommendations presented here are an integral element of a larger research programme: Bridging the gap: Examining disability in four African countries, managed by the Leonard Cheshire Research Centre, University College London (Leonard Cheshire Research Centre, n.d.). This three-year programme is funded by the UK's Department for International Development (DFID) and the Economic and Social Research Council (ESRC). The overall objective of this programme is to develop an in-depth, nuanced understanding of how disabled people are increasingly at risk of being excluded from participating in social and economic development. The research specifically focuses on four sub-Saharan African countries: Kenya, Sierra Leone, Uganda and Zambia, all of which have ratified the UNCRPD. Furthermore, this research primarily, but not exclusively, focuses on four policy domains: education, health, labour markets and social protection.
The fundamental hypothesis underpinning this entire research programme is that, in general, as the development status of low-and middle-income countries (LMICs) increases, there is a corresponding increase in the socio-economic status of the general population. Part of this commitment to socioeconomic development is reflected in the increasingly recognized importance of upholding the rights of women and of a range of other recognized marginalized groups, through their empowerment and social integration as a critical element for achieving the post-2015 development agenda (Oldekop et al., 2016). However, during this process, if there is no corresponding commitment to full inclusion and non-tokenistic participation of disabled people in all development policies and programmes, there is a risk that the socio-economic status of disabled people will remain static while that of other members of society progresses, thereby causing a "disability and development gap" (Groce & Kett, 2013). Poverty reduction and proactively decreasing inequalities is a key aim of international development, underpinning the foundations of the Sustainable Development Goals (SDGs). In general, for those who are poor in developing countries, inequality can be reduced through initiatives such as the SDGs, but this reduction almost certainly will not include, nor have a positive impact upon, a corresponding increase in the socio-economic status of disabled people (Buse & Hawkes, 2015;Lockwood & Tardi, 2014;Norheim et al., 2015;Mtra et al, 2017).
The policy analysis component undertaken for the Bridging the Gap programme, presented in this article, is contextualized in the broader contemporary development studies discourse and mainstream international development policy and practice-particularly in the substantive policy areas of good governance, poverty, inequality, participatory development, human rights and inclusion (Atkinson, 2015;DFID, 2007;Levi-Faur (Ed.), 2012;Marmot et al., 2008;Piketty, 2014;Pogge, 2008;Rhodes, 2012;Smith, 2007;Weiss, 2013). These concepts are all relevant to the attainment of full disability inclusion and reducing poverty and inequality through reduction of the "disability and development gap" described above.
Good governance has also become an increasingly important policy issue in developing countries, particularly in Africa, over the past 30 years, with a movement away from the traditional, hierarchical "top-down" Weberian government structures (whereby government is responsible for the development and implementation of all public policy), to an increasing recognition of the value in creating "policy networks," whereby civil society, together with the private sector, work with governments in policy development and service delivery (Kjaer, 2004;Messner, 2013;Osborne & Gaebler, 1992;African Union, 2015a). Furthermore, the very nature of developing social and economic policy has now become so complex that it is no longer feasible for the state to undertake all the stages of policymaking alone. Hence, Jacob Torfing, echoing the earlier insight of Jan Kooiman, states, "Globalisation, the fragmentation of social and political life, the growing number of 'wicked problems,' and the new ideas of how to govern through 'regulated self-regulation' make it clear that government agencies cannot govern alone. Indeed, no actor has the knowledge, capacity, and authority to regulate society and the economy single-handedly" (Torfing, 2012, p. 100;Kooiman, 1993, p. 4).
Therefore, non-state entities, such as civil society institutions, think tanks, academia and the private sector have a legitimate role to play. Multi-stakeholder participation in the policy-making process adds an additional level of accountability for implementing public policy, as well as drawing on their in-depth knowledge of specialist policy subject areas and working with clients or beneficiaries of public services at ground level.
Linked to this broader notion of governance is the rise of participatory development, spearheaded by the work of Robert Chambers among others (Chambers, 1997;Fischer, 2000). The principles of participatory development argue for the inclusion of poor and marginalized groups in the policymaking process, and are closely linked with the fundamental axiom of the international disability rights movement (Charlton, 1998;Driedger, 1989;Kumtiz, 2016;Yeo & Moore, 2003).
The promotion of human rights for all is premised on the United Nations' International Bill of Rights, which arose from the 1948 Universal Declaration of Human Rights (Ishay, 2008;Mégret, 2008;Mertus, 2009;Pogge, 2008) and upon which all human rights initiatives spearheaded by the United Nations have been framed (Mertus, 2009). This includes the UNCRPD, which forms the basis of the analysis presented in this article.
Human rights have been categorized in a number of different ways, not all of which can be reviewed here. A common distinction is made between negative and positive rights (Shelton, 2015;Tomuschat, 2014), both of which are reflected in the 50 Articles of the UNCRPD. Examples of negative rights include freedom of speech, freedom of religion and freedom of association, all of which can be negated if curtailed by other people or legislation and policies. Conversely, positive rights principally relate to economic, social and cultural rights that should be provided to individuals, such as the right to education, health, a minimum wage and social protection, which are aligned with the four policy domains encompassed in this study. Importantly, as will be discussed in this article, enforcing positive rights requires potentially significant expenditure by governments and donor agencies (Donnelly, 2013;Griffin, 2008). Policies and programmes that explicitly set out detailed information on how their aims, objectives and activities are to be financed and monitored, either in the short or the long term, are those that will successfully address positive rights.
Given the importance of good governance, participatory development, inclusion and achieving human rights in the reduction of poverty and inequality for disabled people, analysis of the policies of the AU must assess the extent to which these principles are embedded in the documents reviewed. The analysis of these policies must in addition reflect key stages of the policy-making process and explicitly demonstrate the extent to which all actors (including civil society institutions) have been involved in this process. Hence, if applied successfully, such an approach to policy-making will hopefully ensure that policies will be implemented in a manner that genuinely meets the explicit needs and aspirations of target beneficiaries.
In 1994, Walt and Gilson set out a clear policy analysis framework and methodology that goes beyond reviewing only the content of the policy documents (Walt & Gilson, 1994) to encompass a review of the process, context and actors involved in the development of the policies. For example, policy development and subsequent implementation that includes detailed consultation of a wide range of actors suggests the application of concepts of good governance; one that is participatory in nature and that fosters a genuinely inclusive policy. In this way, the realization of disability and human rights is more likely to be achieved.
A growing body of research shows that there is a symbiotic correlation between disability and poverty, with each being a cause and consequence of the other (Awasthi, Pandey, Dubey, & Rastogi, 2017;Graham, Moodley, & Selipsky, 2013;Grech, 2016;Groce, Kett, Lang, & Trani, 2011;Mitra, Posarac, & Vick, 2013;Owens & Torrance, 2016;Palmer, 2013;Trani & Loeb, 2012). It is also becoming increasingly clear that poverty, in relation to disability, must be conceptualized in a multidimensional manner (Mitra et al., 2013;Trani & Loeb, 2012). Therefore, poverty is not exclusively concerned with the lack of monetary income, but is a much broader holistic concept, that embraces, for example, the lack of access to public services and the lack of ability to participate in society (Comim & Nussbaum, 2014;Robeyns, 2005;Sen, 2001;Sen, 2011;Trani & Cannings, 2013;Trani, Biggeri, & Mauro, 2013;Trani, Bakhshi, Myers Tlapek, Lopez, & Gall, 2015). Thus, inclusive access to healthcare, education, employment and social protection are all imperative to realizing the rights of disabled people. Notwithstanding the increasing understanding of the basic correlation between disability and poverty, it remains the case that the intricacies of the dynamics that drive this correlation remain unclear and warrant further research (Groce & Kett, 2013).
Finally, it is important to consider the applicability of intersectionality in the context of this research. This concept has become increasingly salient in international development research (Larson, George, Morgan, & Poteat, 2016;Cross & Hadjar, 2016). Although, some academic research has been undertaken on intersectionality and disability, this is an embryonic field (Erevelles & Minear, 2010;Ben-Moshe & Magana, 2014).
In the international context, Olena Hankivsky has defined intersectionality in the following manner: Intersectionality promotes an understanding of human beings as shaped by the interaction of different social locations (for example, 'race'/ethnicity, indigeneity, gender, class, sexuality, geography, age, disability/ability, migration, status, religion). These interactions occur within a context of connected systems and structures of power (for example, laws, policies, state governments and other political and economic unions, religious institutions, media). Through such processes, interdependent forms of privilege and oppression shaped by colonialism, imperialism, racism, homophobia, ableism and patriarchy are created (Hankivsky, 2014, p. 2).
This resonates with the issues that are addressed in this article. Disabled people constitute a heterogeneous category of individuals, each with their own unique characteristics. Furthermore, as noted by Larson et al. (2016), intersectionality highlights the interdependency of interrelated factors, and these are not necessarily causally hierarchical. Rather, "intersectionality considers how individuals can simultaneously experience and embody privileges and disadvantage as different social hierarchies combine in varied ways across time and diverse locations" (Larson et al., 2016, p. 965). It also seeks to analyse the complexities of multiple compounding exclusions that marginalized groups encounter, considering the power relations that exist between different stakeholders. This is very much aligned with the multiple identities of disabled people encountered in developing countries, including geographical location, gender, income, age and so forth. Furthermore, intersectionality explicitly explores institutional exclusion, that again has much to do with the marginalization and exclusion of disabled people from participating in the policy-making process in Africa.

| METHODOLOGY
In order to evaluate the importance and impact of policy in the field of disability from a continental Africa perspective, this study reviewed nine relevant mainstream AU policy and strategy documents to determine the extent to which disability was considered and included in the documents.
A set of predetermined inclusion criteria were developed by the authors for selecting the policy documents that were reviewed. The principal aim of these criteria was to identify and analyse published AU policy or strategy documents published in English that related to the domains of education, health, labour markets and social protection for the general population. Given that this was a desk study of the available literature, it was not possible to review documents that were not available online. In addition, AU continental-level action plans that addressed poverty in relation to the assessment and implementation of the earlier Millennium Development Goals (MDGs) and the new Sustainable Development Goals (SDGs) were included. Preference was given in the selection process to policies published in or after 2010. This was because the UNCRPD was enacted at the United Nations (UN) level in 2008 and policies developed after 2010 were deemed more likely to reflect the principles of this Convention.

| The analysis model
A comprehensive policy analysis framework that specifically addressed the four elements of content, context, process and actors was undertaken. This framework followed the model for health policy analysis described by Walt and Gilson (Walt & Gilson, 1994) as set out above. The strength of adopting the methodology developed by Walt and Gibson is that it provides an insightful, critical, analytical lens to look at how policy-making happened in practice, encompassing the power relations that exist between all stakeholders involved, and the relative impact of a multi-stakeholder approach to the policy-making process. Moreover, this methodological approach has been widely used in a multiplicity of countries (Bennett et al., 2014;Cassels, 1995;Koch & Weingart, 2016;Martineau et al., 2015;Shearer, Abelson, Kouyate, Lavis, & Walt, 2016), primarily in the health sector, and is practically relevant to international disability policy analysis. Our analysis, however, focused on analysis of the documentary evidence only without any additional information collected through interviews with the actors, including DPOs.
A careful review of all AU documents allowed us to identify nine documents between 2009 and 2017 that fitted the inclusion criteria outlined above.

| Content analysis
A template was specifically developed for each policy domain to analyse the content of the document. The content component included seven elements each rated according to the level of disability inclusion reflected in the document. The ratings ranged from one to four, with higher ratings reflecting greater disability inclusion. The seven elements and the rating structure are presented in Table 1. A separate template was designed for each of the four domains covering the same seven elements.
A comparison across the four policy domains (education, health, employment and social protection) provides an indication of which is most responsive to the inclusion of disabled people. Similarly, the comparison across the seven elements (i.e. rights, accessibility, inclusivity, implementation, enforcement mechanisms, budgets and finance, and management information systems) for a given domain indicates which are the most reflective and accommodating regarding disability inclusion.

| Context, actors and process components of the analysis
The research examined the political, economic and social contexts in which the policy was developed, and analyzed, and how this influenced the policy-making process. For the actors, we examined the different key stakeholders and the role they played in policy development as set out in the documents reviewed. For example, we analysed any references to Disabled Peoples' Organisations (DPOs), private sector or civil society organizations and other government departments as an indication of them being involved in the development/negotiation of the policy. The "process" component reviewed any available information regarding the level of consultation during the development of these documents and the main groups or actors included in the consultation process. Of particular interest was any evidence of consultation with DPOs. As noted above, this analysis was limited by the availability of such information in the documents themselves. Given the limitation of only reviewing the actual documents, the context, actors and process components of the Walt and Gilson framework were analysed in less depth than would have been ideal, in comparison with the content analysis.
The ratings for each policy and strategy for each domain were summarized and the mean total calculated for all policies within one domain if there were more than one policy analysed. The total possible score for any one policy was 28 (i.e. seven elements with a maximum possible rating of four for each). The mean range of ratings for each policy was calculated as the mean for each domain.

| RESULTS
Nine policy or strategy documents were collected and analysed or reviewed. The full list of the policies and strategies is presented in Table 2.
The documents included range from plans of action, assessment or meeting reports, strategies and the Draft declaration on employment, poverty eradication and inclusive development in Africa. The rationale for the inclusion of this latter "Draft declaration" in the analysis is that, although still in the process of being negotiated and finalized, it is nevertheless indicative of the AU's current thinking on the relationship between employment, poverty eradication and inclusive development. It also provides an insight into how  Table 3 presents the findings of the total scores and the range in the ratings that were given for each of the documents, and this is disaggregated by policy domain.  Table 3 and Figure 1, the highest ratings were scored for the General Development and Social Protection domains and the lowest was for the employment domain. Each of the individual policies and strategies in the Education and Health domains received almost identical scores, thereby demonstrating that they were, overall, equally disability inclusive. Of note is that not one of the policies reached even 50% of the total score, reflecting a de facto low level of disability inclusion. Figure 2 presents the mean ratings and total scores across all domains for each element within the content analysis. The reference to the rights of disabled people was the most included with a total score 66% of the possible total score of 28 and had a mean rating of just over two out of the highest possible rating of four. This indicates that the concept and importance of rights was perceived to be important, in comparison with the other six categories upon which each policy or strategy was rated.

| Context, actors and process components of policy analysis framework
The analysis of the actors and process revealed little information in the documents themselves regarding the context within which these documents were developed or the players involved. Context information was provided in most documents in the form of a preamble. Information on the development and negotiation process and actors who were involved was very limited. The global context of most of the documents was dominated by hegemonic status given to the MDGs and SDGs for AU documents, particularly the ones published after 2010. If documents mentioned the process and actors involved at all, this was usually with reference to the AU being the main institution drafting the document, with input from various groups such as women's groups and refugees. However, no document made explicit reference to the inclusion of disabled people as proactive and important actors in the process. This is somewhat surprising, given that most African countries have ratified the UNCRPD, and therefore are legally bound to implement its obligations, as delineated in its 50 Articles.

| Reference to and definition of disability
Most of the documents made some reference to disability or disabled people, but none provided a clear definition of what was meant by "disability" or a "disabled person." These references occurred in the context of other vulnerable groups, or with reference to chronic health conditions and related disabilities (i.e. morbidity). Disability was also referred to and recognized, in the majority, but not all of document reviewed, as a cross-cutting issue. Moreover, from a critical reading of the documents, it was evident that disability was not clearly understood from a human rights perspective. Rather, a strong impression is given that disability issues and disabled people are often seen by policy-makers as an afterthought, when compared with other minority groups, such as women and children. The following four extracts illustrate this:   Health Strategy: 2007 The alarming rate of growth of the burden of both death and disability from noncommunicable diseases in Africa is ever more recognised, with chronic diseases becoming ever more prevalent, linked to demographic, behavioural and social changes and urbanisation … Injuries from violence, wars, traffic accidents and other mostly preventable causes result in widespread death and physical disability, while the impact of mental ill-health has previously been underestimated [emphasis added] (African Union, 2007, p. 2).

Common African Position (CAP) on the post-2015 Development Agenda
The following two extracts are from the Common African Position (CAP) on the post-2015 development agenda (African Union, 2014c); the first about the eradication of poverty and the second about universal and equitable access to healthcare.
This will require the empowerment of all people, including those living in vulnerable situations (including women, children, the elderly, youth, people with disabilities, rural populations, displaced persons and migrants), through inclusive growth that creates decent jobs, improved access to social protection and through the promotion of measures that ensure that no individual remains below the poverty line. In this regard, we commit to ensure that no person -regardless of ethnicity, gender, geography, disability, race or other status -is denied universal human rights and basic economic opportunities [emphasis added] (African Union, 2014c, p. 10).
We must improve the health status of people living in vulnerable situation such as mothers, newborns, children, youth, the unemployed, the elderly and people with disabilities by: reducing the incidence of communicable diseases, non-communicable diseases (for example mental health) and emerging diseases; ending the epidemics of HIV and AIDS, tuberculosis and malaria; reducing malnutrition; and improving hygiene and sanitation [emphasis added] (African Union, 2014c, p. 10).

| DISCUSSION
This section draws out the key points from the analysis presented above, with a view to making some strategic observations that will influence disability and international development policymaking in the future. It will therefore highlight the need to adopt a rights-based approach to the policy-making process in this increasingly important policy arena, as well as highlighting the imperative of including, in a non-tokenistic manner, disabled people and their representative organizations at every stage in this process. It will be recalled that one of the principal objectives of undertaking this policy analysis research was to determine the extent to which disabled people were involved in the policy-making process at continental level in Africa, and the extent to which disability rights were reflected in the reviewed policy documents and strategies published by the AU. Table 3 and Figure 1 clearly show that none of the documents analysed reached even 50% of this maximum score. These policies and strategies are not disability inclusive as they do not address disability issues directly or in a way that reflects an inclusive rather than separate and specialized approach to providing services for disabled people. The mainstream development documents (i.e. the Common African Position (CAP) on the post-2015 development agenda (African Union, 2014c) and Agenda 2063: The Africa we want (African Union, 2015c) achieved the highest overall ratings and total score. Despite neither study explicitly mentioning disability, both had frequent, detailed and nuanced references to the importance of human rights. The mean scores for the individual content analysis elements rated each document for education, health, employment and social protection (see Figure 2), and clearly show that the category of rights to inclusion for disabled people have the highest mention (i.e. highest average score) for the seven policies and strategies that were reviewed. Indeed, many of these documents-but not all-explicitly mentioned disability in relation to rights to inclusion in education, health, employment, social protection and overall development. This is to be expected, given the increasing political profile that is gaining momentum because of the importance of UNCRPD and the growing recognition of the rights of all people as reflected in the SDGs. The elements which are not addressed that support the realization of these rights are financial data, monitoring, enforcement and implementation. These identified deficiencies were not developed to any extent in any of the domains.
Furthermore, the AU documents reviewed in the domains of education, health, employment, social protection and general development, are not comprehensively disability inclusive, and disability issues are not perceived by policy-makers as a key policy priority in the context of Africawide social and economic policies. None of the documents analysed warranted a rating that was indicative of a good, viable model of disability inclusion. Of particular concern is the apparent lack of impact of the UNCRPD regarding the development and subsequent implementation of these documents, despite the AU having a clear commitment to recognizing disability as an important issue, as reflected in the two continental plans of action for the African Decade of Persons with Disabilities (1999Disabilities ( -2009Disabilities ( and 2010Disabilities ( -2019 developed jointly with the African Disability Alliance (African Union, 2002Union, , 2012.
The lack of or scant references to any definition of "disability" or "disabled people" is problematic, as it provides no objective criteria or guidance on who should be targeted as disabled and who is responsible for implementing programme elements to address disability issues. Furthermore, the conflation of disabled people with other "vulnerable groups" obscures the specific needs of each group. Without a clear elucidation of the specific needs of individual groups, the needs of all groups are not likely to be addressed effectively, if at all.
However, it is duly recognized here that while this is a problematic area, it is a global issue rather than one confined to Africa. Indeed, during the seven-year negotiation of the UNCRPD, there was a great deal of debate, even over basic issues, such as what "disability'' really means. During the negotiation process, it was not possible to attain consensus around the definition of the term 'disability' itself. This is because "disability" is culturally defined (Lang, 2009). Notwithstanding this, a working "definition" and framework, especially for the AU, is the Preamble of the UNCRPD, which states that: The State Parties to this Convention, … [Recognize] that disability is an evolving concept and that disability results from the interaction between persons with impairments and attitudinal and environmental barriers that hinders their full and effective participation on an equal basis with others. (UN, 2006, Preamble (e)).
Within this research, disability is recognized as an important issue by the African Union and its constituent Member States, but is not given the required status to ensure the realization of the rights of disabled people. Of particular concern is that the document on the Agenda 2063: The Africa we want (African Union, 2015c) does not make any explicit reference to disability, despite being the strategic plan for the continent for the next 50 years and the high profile that this document gives to other groups, such as refugees, women and children. Furthermore, this lack of reference to disability is in direct conflict with the spirit of the SDGs, which clearly and explicitly refer to disability as a key factor to consider in achieving these goals (Madans, Loeb, & Altman, 2011;Olsen et al., 2014;Tangcharoensathien, Mills, & Palu, 2015;Waage et al., 2015).
On a positive note, it is evident that many of the policies and strategies reviewed addressed key ideas and concepts that are fundamental to international disability rights. For example, there are many references to human rights, although not specifically to the rights of disabled people. Furthermore, "inclusion" and, perhaps more importantly, "inclusive development" are often mentioned. However, it is far from clear that these ideas of inclusion and inclusive development are understood in the way those engaged in international disability policy and practice understand them. "Inclusion" is often associated with the notion of inclusion of poor and marginalized or vulnerable groups as described above, but with seemingly little understanding of "inclusive development" as it relates specifically to disability as a legitimate area of international development (Albert (Ed.), 2006;Yeo & Moore, 2003). Furthermore, even within international development policy, practice and research, the concept of "inclusive development" and "inclusive aid" is ill-defined (Groves & Hinton (Eds.), 2013; Hickey, Sen, & Bukenya (Eds.), 2015). Indeed, these terms have multiple meanings, contingent upon precisely who is using them. Therefore, because of this confusion there is a dissonance between what mainstream development policy-makers and practitioners mean by "inclusive development" on the one hand, and by disability practitioners on the other (Grech, 2009;Hilbert, 2016;Soldatic & Grech, 2014).
An additional finding in this review was that far more references to disability were made in the Africa Health Strategy 2007(African Union, 2015 than in the later Africa Health Strategy -2030(African Union, 2016b. Irrespective of the other innovative attributes, in terms of raising the profile of disability issues, the later Africa Health Strategy seems retrograde when compared with its predecessor. This is very surprising, given the high profile attributed to the UNCRPD and the importance ascribed to disability within the SDGs (Madans et al., 2011;Olsen et al., 2014;Tangcharoensathien et al., 2015;Waage et al., 2015).
From the analysis presented for the seven policies and strategies addressing health, education, labour markets and social protection, the following conclusions can be made. First, education has the highest average rating of the four policy domains and also has the highest individual score (see Figure 1 and Table 3). In the light of this, we can tentatively conclude that education accords a higher priority to disability inclusion than do the other policy domains. However, the nature of the rating (2 or lower) reflects that this mention is generally in relation to provision of "special education" for children with disabilities, rather than advocating for "inclusive education." The "inclusive education" approach to educating children with disabilities has gained hegemony in the last 20 years. It refers to the notion that children with disabilities should be educated in mainstream schools alongside their non-disabled peers. Indeed, this was endorsed in the UNCRPD. Article 24 (Education) states that "States Parties recognize the right of persons with disabilities to education. With a view to realizing this right without discrimination and on the basis of equal opportunity, States Parties shall ensure an inclusive education system at all levels and lifelong learning" (UN, 2006). A further explanation for these higher ratings is that, globally, education for children with disabilities has received relatively greater attention by bilateral and multilateral donor agencies and other international organizations, spearheaded by the Global Partnership on Education and the publication of the 2013 State of the world's children report by UNICEF, which focused on children with disabilities (UNICEF, 2013).
Additionally, with respect to the employment and social protection domains, it was only possible to identify one document for review in each of these. Furthermore, both domains included policies that were in preparation at the time of writing (i.e. the Draft declaration on employment, poverty eradication and inclusive development in Africa (African Union, 2014b) and the Report of a meeting on Social Protection and Inclusive Development: social protection for inclusive development (African Union, 2015d). The analysis presented here provides some useful recommendations for ensuring disability inclusion in the finalization of the policies and strategies related to these reports and the draft declaration.
There are also some important issues identified in the analysis that are causes for concern. None of the seven documents had any budgetary/financial information and very few had robust management information systems that could monitor how effectively these were being implemented. However, Agenda 2063: The Africa we want was very strong in its commitment to develop a robust monitoring and evaluation framework to monitor its future effectiveness. In the absence of robust financial allocations and management information systems, it is very difficult to foresee the effective and efficient implementation of social and economic policy, not least in the field of disability. The low score for the inclusion of accessibility factors reflects a lack of awareness of the importance of this component for effective inclusion of disabled people. This accords with the lack of explicit references to disability and the needs of disabled people in most of these documents.
There are fundamental challenges in ensuring that public sector budgeting and management is effectively instituted in many developing countries, not least in Africa (Booth & Cammack, 2013;Hickey et al. (Eds.), 2015). One possible explanation for this gap identified in our analysis could be the continental remit of these documents, which set out the principles for individual Member States to adhere to, but which leave it to the individual countries to develop in the detail. However, an argument can easily be made that such data should be provided to guide the Member States. This is in alignment with the strategic goals of the AU, which include an obligation "to encourage international cooperation, taking due account of the Charter of the United Nations and the Universal Declaration of Human Rights" and "to promote democratic principles and institutions, popular participation and good governance" (African Union, n.d.). Furthermore, none of the documents reviewed provide any detailed budgetary or financial information or guidance on how their respective strategic goals, aims, objectives and related activities should be implemented. In the absence of such data, it is very difficult to foresee how any of these can be implemented in a meaningful manner.
None of the nine documents reviewed had specific, measurable, achievable, relevant or time-bound (SMART) indicators by which to assess the extent to which each of these policies and strategies are implemented effectively. This omission could have a significant detrimental impact on the future progress of disability policy and practice regionally, especially in the global context of the SDGs and the ongoing implementation of the UNCRPD. In the absence of SMART indicators, civil society institutions, including DPOs, will not have the necessary tools and benchmarks to hold their respective governments to account over their disability rights and policy commitments. Neither will it be possible to assess the extent to which the UNCRPD has been implemented. It can be argued that, without these indicators, a "democratic deficit" is created, which compromises the principles of democracy, transparency, accountability and the rule of law (Gaventa & McGee, 2013;Lang, 2009;Lang & Murangria, 2009;van Zyl, 2014). This further entrenches and intensifies the growth of the disability and development gap. If policies are not clear regarding the importance of disability inclusion and how to address it, SMART indicators will not be developed.

| CONCLUSIONS AND RECOMMENDATIONS
The analysis of the AU policies and strategies presented in this article gives rise to some final observations and recommendations for the future development of disability policy and practice in Africa, and the sustained reduction of poverty, defined in its broadest sense, for disabled people on the continent.
First, it is important to emphasize that some progress has been made at policy level with respect to disability inclusion, but much remains to be done before full inclusion is achieved. Policy-makers and development practitioners, at least as these are represented in major AU documents, do not seem yet to fully comprehend the importance of addressing disability issues as an inherent component of social and economic policy at the national level, and also at the continental level, particularly as an integral component of international development. However, it is anticipated that this will change with the progressive implementation of the SDGs over the next 15 years, as these give greater prominence to disability issues than did the MDGs.
While some of the documents reviewed for this article referred to disability and the importance of addressing disability issues, many did not make any reference to disability. This is despite the fact that most AU Member States have ratified the UNCRPD. This was very clear for example, when reviewing the health policies. It will be recalled that the Africa Health Strategy 2007(African Union, 2007 made more progressive references to disability than did the later Africa Health Strategy -2030(African Union, 2016b. This is a regressive and disappointing development, and every effort should be made to rectify it. A more consistent, ongoing engagement by the AU and its Member States with continental and national DPOs would help address this gap.
Given the limited explicit mention of disability and the needs of disabled people in many of the documents analysed, the lack of budgetary and human resource provision, the limited plans for monitoring and evaluation and the scarcity of mentions of disability disaggregated data collection in management information systems, much work is required to raise awareness and develop disabilityinclusive policies and strategies.
Three key messages arise from this analysis which warrant significant attention by policy-makers and implementers, DPOs and other development organizations: Key message 1: Disabled people as key actors in mainstream policy development The active and non-tokenistic involvement of DPOs as key actors from the start of any policy, strategy or implementation plan development is crucial to ensuring that the voices of disabled people are reflected in the intent and content of these documents. This appears to have been largely missing in existing AU policies, but only if this takes place will mainstream public policy and strategies become genuinely inclusive. Relegating disability issues to disability-specific policies and strategies contradicts the fundamental principle of inclusion.
Key message 2: Use of consistent and clear definitions of disability The definition of disability must be clearly stated in all policy documents and should reflect the important role that environmental factors (for example, physical, social, attitudinal, policy and legal barriers) play in disability. As stated above, the UNCRPD provides a good working definition of disability, while also recognizing that it is an evolving and culturally contingent concept. This provides a good, viable catalyst for developing a countrywide or continentally consistent definition, but should not be perceived as a panacea. However, the challenges of achieving this must not be underestimated, as evidenced by the impasse encountered when attempting to define disability during the negotiation of the UNCRPD.
Key message 3: Explicit mention of disabled people as a targeted group The explicit mention of disability and the needs of disabled people is important to raise awareness of disability issues and provide policy-makers and development practitioners with a deeper understanding of how to realize these needs and develop appropriate indicators to monitor and evaluate implementation of policies and programmes.
In the light of the analysis presented above, it is apparent that while the AU and its constituent Member States have in recent years made concerted efforts to address disability issues at both continental and country level, this is not consistently or comprehensively reflected in recent key documents pertaining to international development initiatives. This is surprising, particularly in light of the increasing attention to disability globally through the UNCRPD, now ratified by most African countries, and the explicit and implicit inclusion of disability throughout the SDGs.
In addition, there is a need to ensure that robust monitoring and evaluation frameworks are developed, so that in future the AU and its constituent Member States are able to effectively evaluate the extent to which AU policies and strategies have been implemented, and the impact-positive or negative-that these have on raising the socio-economic status of disabled people throughout Africa.
The limited inclusion of disability within AU policies is a lost opportunity that should be reviewed and rectified. Regional and global development initiatives and the progressive implementation of the SDGs provide a new impetus and a strong catalyst for the effective inclusion of disability, particularly in the field of policy development and implementation. Renewed attention to and inclusion of disability in regional efforts would be an important step in this direction. Finally, the analysis and findings presented in this article aim to contribute to the ongoing development of effective and genuinely inclusive social and economic policy throughout Africa, spearheaded by the AU and its constituent Member States.