Social Protection as a path out of poverty

Social Protection as a path out of poverty - A study about policy strategies for Kenya Hanna Norman Development Studies C (Bachelor Thesis) Department of Government Uppsala University, Spring 2017 Supervisor: Joakim Palme Word count: 13372 Number of pages: 40

Table of Contents 1 Introduction... 3 1.1 Purpose... 4 1.2 Research Question... 5 2 Theory: social policy strategies... 5 2.1 Social Policy and Social Protection... 5 2.2 Gender and Class... 7 3 Methodology... 12 3.1 Case selection and Choice of Actors... 12 3.1.1 Kenya... 12 3.1.2 Choice of Actors... 14 3.2 Material... 15 3.3 Preconceptions, Validity and Reliability... 17 3.4 Analytical Framework... 17 4 Analysis... 19 4.1 The United Nations Development Programme (UNDP)... 19 4.2 The World Bank (WB)... 23 4.3 The Swedish International Cooperation Development Agency (SIDA)... 27 4.4 The Government of Kenya... 29 5 Discussion... 33 6 Conclusion... 35 7 References... 37 2

1 Introduction Poverty reduction and gender equality are two central concepts within development discourses and have also been included in the eight Millennium Development Goals (MDG) drafted by world leaders at the United Nations (UN) in year 2000. The goals came with a vision to fight all dimensions of poverty worldwide. Eradicate Extreme Poverty and Hunger was introduced as goal number one and Promote Gender Equality and Empower Women was goal number three (UN, 2015:4f). Although improvements have been done, human deprivation is still widespread and even with a significant decline, at least 836 million people live in extreme poverty (UN, 2015:4), and according to the United Nations Development Programme (2015: 4f), 795 million people still suffer from chronic hunger. Big gaps between rich and poor households still persist where the mortality rate for children under five are about twice as high for children in poor families (UN, 2015: 8) and access to basic services such as food, education, health care, sanitation facilities and work opportunities is a big concern (ibid: 8f). Inequalities regarding class and gender still remain, with discrimination against women in all phases of life, including private life and both the formal and informal labour force. After the MDGs, the international community set a new agenda of post-2015 goals to implement new sustainable thinking. The new Sustainable Development Goals (SDG) takes up where the MDG s left off. The resolution (General Assembly, 2015:1), states that extreme poverty is one of the greatest challenges of the global community and needs indispensible work in order to eradicate poverty and create larger freedom for all people including the achievement of equality. UN also declares that in these goals, social policies like social protection (SP) is of importance to assure mental, physical and socialwell being to people worldwide (ibid: 3). SP systems are also mentioned in the resolution as creating basic living standards, which all people should have the right to enjoy by 2030 (General Assembly, 2015: 7). Even though many see SP as a fundamental right, only around 27 percent of the world population is covered by it (UNDP, 2015:19). SP is used by organizations and nations as means to reach certain goals and the way it is constructed may vary between them. This may be shown in diverse strategies and ways policies handle inequalities, which in turn can affect gender and class structures in a society. SP is becoming more recognizable in Africa with programmes being put in 3

place as a tool to fight poverty around the continent. Organizations together with governments are moving from focusing on emergency humanitarian relief projects to create more stable social protection systems to avoid emergency actions (Ellis, Frank et.al, 2009:3). The nation of Kenya is one example of many countries in Africa that have introduced SP programmes as a tool in their development discourse and to improve well being of its people (Davis et.al, 2016: v). Kenya has an ambitious agenda when it comes to the formation of SP. They have manifested in their newest constitution, drafted in 2010 that SP is a basic right for its entire population (Republic of Kenya, 2010:31). Because of the right that is laid out in the constitution, the Kenyan Government has put in place a social protection sector that includes social security, social assistance and social health insurance programmes. SP has for some time been and still is a part of Kenya s National Development Policy, where the government is working together with different actors like the United Nations, the World Bank, and the Swedish International Development Cooperation Agency to develop SP strategies to reach the goal set out in the constitution. The goal to ensure that all Kenyans should be given the capability to live a life they have reason to value without social or economic fear (MGCSD, 2011:3ff).). 1.1 Purpose After being anchored in the SDG s and with the important challenge of ending world poverty, the focus on social protection has become stronger and so follows the interest in global organizations and strategies developed around the issue. The idea behind this study comes from a general interest in understanding social protection systems in a developing context and how they are formulated as a tool in the effort to reduce poverty as well as diminishing inequalities. The purpose of this thesis is to look at global actors on a local level to identify strategies by actors involved with SP in Kenya. The method of the thesis is to systematically analyse policy documents to see if diverse ideas about central elements like gender and class inequalities are given specific focus in strategies conducted by stakeholders involved in the process of designing and implementing social protection in Kenya. Therefore, the main point is to investigate alternative definitions of 4

social protection to see what ways the concept is operationalized and to see if different actors put emphasis on diverse means and use different models to reach their goal with SP. Research shows that strategies by different actors can create possibilities but can also be an obstacle for development. Therefore it is interesting to identify if the strategies differ and later see in what ways this can be seen in the practical work of social protection. 1.2 Research Question The research question of this study is: What strategies with regard to gender and class inequalities can be identified in policies formulated by stakeholders of the Social Protection Programme in Kenya? 2 Theory: social policy strategies In this section theories are outlined to give the reader a concise understanding of social policy and social protection. Previous research is presented to give a view of what scholars have done within this area of study and also to get an idea of the main concepts and findings, thus giving the reader a better understanding into the research findings and results. Additionally, this study will start in the research advanced by central theories about social protection, in an attempt to identify indicators of inequalities in the work of poverty reduction. 2.1 Social Policy and Social Protection There has been a long debate on how social policies should be designed to reduce poverty and inequality in society as efficient as possible also in development contexts. Social protection has grown in developing countries as a way to reduce poverty and vulnerability and has earned a place both in social and economic development strategies (Barrientos and Hulme, 2008:3). It is a way for governments to help vulnerable and poor 5

citizens to live a better and more well-of lives. However, the approaches towards advancing social protection diverge. The debate around different types of transfer programmes often finds itself discussing the central concepts of targeting versus universalism. There is a division between actors promoting universal policies and those promoting more selective and targeted provision of social protection. Thandika Mkandawire (2005:iii) explains it as a continuum of two extremes with nearly no policy regime that is purely one or the other. However, a shift towards using selective policies can be identified in developing countries as a tool in combating poverty. The difference between universalism and targeting is in a way self-explanatory. The idea of universalism is that the whole population is beneficiaries and has social protection as a basic right, while targeting is a selective method that provides social protection to the ones that need it the most. To determine eligibility, one uses means testing, however the way countries collect information on assets and incomes of individuals may vary (Mkandawire, 2005: 1). One could see the shift from universalism to targeting in developing countries being grounded in macroeconomics and policies regarding aid. This in turn could be determined by ideological thoughts and how the polity s ideas are constructed around the responsibilities of its members (Mkandawire, 2005:1). Mkandawire, who is quite negative towards the more selective method, debates how the shift towards targeting came together with the ideological shift towards the right and with a more limited role of the state. In addition, new beliefs about how to deal with fiscal constraints surfaced where people saw the selective method as more efficient and commonsensical to use when there is a budgetary restraint (ibid: 2, Ferrarini et.al, 2016:25). The argument that one could fight poverty with less money however comes with the complication regarding the social contract that regimes have with their entire populations since everyone won t be covered. The weakness of universalism in welfare states is therefore, according to Mkandawire (2005: 4) the failure of providing a universalistic proclamation to the actual politics. Instead of creating an all-covering politics, a type of stratified universalism (ibid.) was created and gave room for a right-wing populism that meant that the privileges that were provided through these policies created a rent-seeking community where instead of creating new wealth, one was manipulating the economic activities. 6

When it comes to the provision of aid and its policy discourses, one also see the shift from the ideas of helping developing countries to having the idea of only helping the poor. This created a new way of thinking which could be seen as forming the shift to more targeted methods in order to make sure that the help reached the poor and vulnerable and not the ones that were more well-off (Mkandawire, 2005: 3). Thus, there is still a question surrounding whether universalism or targeting is the best method for redistributing effects. Some mean that targeting is more redistributing per unit of money since it in fact targets the people on the bottom of the pyramid. Others debate that through a universal redistribution one can create a smaller inequality gap between the ones that are more well off and the ones on the bottom line (Mkandawire, 2005:6p). However, in the quantitative research constructed by Ferrarini et al (2016:38f), they discuss the fact that scholars, stakeholders and policy makers are still focusing on targeting as an important design of SP however their results shows that targeted or non-targeted systems are not as important for global poverty reduction as the size of the income transfers. However, one need to keep in mind that targeting methods can create both stigmatization and under-coverage, which comes down to the capacity of the state and its way of relating equally to all citizens without creating systematic inequalities (Mkandawire, 2005:9pp). Moreover, to identify if systematic inequalities are given specific focus in social protection the thesis will now continue with theories more specific to classify indicators that show unequal relations in a society. 2.2 Gender and Class Social policy and social protection deal with different dimensions of inequality and in the paper, Faces of Inequality: Gender, Class, and Patterns of Inequalities in Different Types of Welfare States, Walter Korpi makes an attempt to tackle the complexity regarding gender and class inequalities within different dimensions of the society and in the process of redistribution. The idea is that one could with his theories enlighten structures and processes regarding gender and class within SP policies in a developing context. Korpi s theory is that in countries where social rights are less developed, it might contribute to a minimized chance of legislations that outlaws gender and class discrimination (Korpi, 7

2000: 148) and this study is trying to see if this can be identified in SP strategies for Kenya. Moreover, Korpi has categorised welfare states in a typology to analyse inequalities concerning both gender and class, thus he believe that they are closely linked and the explanation to the relation they have in a society will not be completed without the other (Korpi, 2000:130). An important demarcation for this thesis is to understand that Korpi s typology, just as many theories about social politics and gender is elaborated for, and applied on developed western welfare states with already existing functional institutions. As the purpose of this thesis is to examine an issue in a developing context, some elements of Korpi s theories might not be as relevant. In his article, Korpi explains that social policies may provide people with capabilities to make their own choices for how they want to live. These capabilities are affected by inequalities through standards of living and through aspects of agency. Walter Korpi writes: In analysis of inequality with respect to gender as well as class, we should consider not only material standards of living but also different aspects of agency reflecting the capability to achieve functionings, that is, the freedom to have well-being (Korpi, 2000: 133). Korpi s perspective is in many ways connected to Amartya Sen s capability approach. Sen s (1999:3) perspective on development is focusing on the means that can create freedoms that all members of a society can enjoy. These freedoms can be connected to different elements that can be provided through social policies and further social protection, such as education and health care. To be able to provide people with capabilities to live the life they have reason to value, one needs to improve both social and economic arrangements (ibid.). Something that social protection could provide. Agency inequalities can be explained as on a continuum with basic level and top level of agency, this can be explained differently, based on different spheres. Within the cultural-educational sphere the basic level of agency could describe inequalities like access to basic education whereas top-level agency could be access to higher education (Korpi, 2000:135). Gendered agency inequalities in a society can also be understood through examples like division of labour, regarding paid and unpaid work (ibid: 140f). The idea is that traditionally women are involved with more unpaid work than men and therefore not included in the paid workforce and thereby excluded from social rights that are only given to people within economic activities. Criticism has been directed toward 8

previous social services and policies for not including unpaid work in gender relationships (Korpi, 2000:142), so the question is how social protection is dealing with this indication, if these perspectives of agency are given any focus in the policies and to what extent ideas within social protection programmes are developed regarding differences in the labour force. Korpi (2000:167) explains that the strength of political tendencies can be transmitted from institutions within areas such as social insurance and gender policies. This then can affect patterns of gender and class inequalities in different ways. However, one needs to remember that these patterns do not necessarily have to be simple or clearcut but rather diverse and based on context. Even though Walter Korpi may be focusing on actually existing institutions, this study make use of his theories in order to analyze policy frameworks constructed for social protection systems with the idea that political tendencies like those made by institutions might be identified in documents regarding social protection, which then can contribute to either the reduction or reinforcement of gender and class inequalities. Korpi s typology is a combination of gender policy institutions and social insurance institutions, which creates an inclusive approach to explain overall patterns of relative inequalities in terms of gendered agency poverty in the labour force as well as inequalities in income distribution and poverty (ibid.174). Additionally, the typology focuses on social rights or what Korpi describes as claim rights that citizens have in relation to the government. This connects to the idea of how SP can be seen as a social contract between governments and its citizens. The gender typology can be explained as an ideal typical model of gendered policy institutions. Here one can identify three models: general family support, dualearner support and market-oriented policies. Something that goes through the models is if the formation of gendered agency inequalities is left to markets and families or if public policies are developed to grant citizens their claim rights. As an indicator to see which model that can be identified, Korpi looks at the support given to nuclear families. The importance is to see whether policies interpret women as the caregivers of the family first, only to enter the paid workforce as a secondary recipient, or if the family model helps to encourage women to enter the paid labour force (2000:143f). In the general 9

family support model, indicators that are directed to housewives can accordingly to Korpi encourage homemaking and increase gender inequalities (ibid: 146). In the dual-earner support model women are given incentives to join the paid labour force and policies have a special focus on redistribution of social care work, both within the family and in the society (Korpi, 2000:143). Indicators that can be identified as describing the dual-earner support model can be seen as day care services for small and older children and maternity as well as paternity benefits. Another example can be help for elderly since in many cases caring for the elderly often is considered a woman s work (ibid: 146). To identify class inequalities Korpi uses ideal types of welfare state institutions. In doing so he studies the distribution of household income to see the material standards of living for citizens in a country (Korpi, 2000: 161), which basically can give the idea of how capabilities are provided to the population. The idea is that the structure of the institutions is relevant for later redistributive consequences. One can see the institutions as intervening variables that contribute to the distributive outcomes in the same way as the redistribution of SP. The ideal types that are first developed by Walter Korpi and Joakim Palme are called: targeted model, state corporatist, basic security and the encompassing model. The targeting model gives minimum or flat-rate benefits on the basis of need assessed via a means test (Korpi, 2000:162). The state corporatist model only covers economically active citizens, thereby the model exclude different categories such as housewives. The basic security model provides benefits on a flat-rate or with only a modest relation to previous earnings. The model is universal per se and is either based on citizenship or individual payment of contributions. The last model called encompassing can be seen as a combination between earnings-related benefits from the state corporatist model and universality from the basic security model (ibid.). On account of these models, one could explain how they contribute to configurations within a society dependent on risk and resource distribution (Korpi, 2000:162). What we could read in previous research about SP, the targeting model is becoming more mainstream in developing countries however this paper have tried to identify strategies linked to the other models as well. Together, the ideal types of gender policy institutions and welfare institutions create Walter Korpi s institutional typology of welfare states. The focus is that 10

institutional structures relates to patterns of inequalities and through using the typology these structures and patterns can be identified. The typology, according to Korpi (2000:171) is the first one that creates a base for analysis inequalities with respect to gender as well as class. With his typology the ranging of inequality could be explained that one should find high inequality within targeted and basic security models, medium level could be found in state corporatist models and the lowest inequality should be found where encompassing models can be identified (Korpi, 2000:169). Regarding gender inequality, the dual-earner model, according to Korpi, is the one that contributes to the lowest levels of gender inequality where as the general family support contributes to traditional thinking and more inequality between genders. The third model, marketoriented is somewhere in between these two (ibid.). What it comes down to is how diverse ideologies and strategies can be found among actors and organizations and that global actors contribute to policy frameworks on national levels where they contribute to the politics around SP. Diverse agendas could determine what kind of instruments are adopted by a specific nation (Devereux and White, 2010: 53). Bob Deacon and Paul Stubbs also describe the effect of international actors, both nations and organizations, within global social policy where global actors can have policies reflecting international interests that can create both limitations and opportunities on a national level. Within SP, international actors can also share ideas and results from projects around the world to create new policies and systems (Deacon, Stubbs, 2013: 11ff). Additionally, international institutions have the opportunity to create new paths within discourses that could contribute to effective policy changes (ibid: 16). One can for instance observe how strategies for SP differ between nations. The majority of strategies for SP both in Africa and in Latin America, focus on cash transfer. A big difference though, is that in Latin America many of the programmes use conditional cash transfer (CCT) where the beneficiaries need to meet specific conditions in order to get money. These programmes have been criticized for contributing of framing women as the prior caregiver of the family since one condition to receive money could be that the women have to make sure that the children go to school. However, the argument is that the CCT contributes to breaking intergenerational cycles of poverty where they create new 11

capabilities and give agency to the younger generation. Although, scholars are debating weather it s the conditions or just the cash that breaks the cycle (World Bank, 2012: 26). What one can see is that strategies and ideals can be diverse which contributes to policies conducted on a national level. The question is how this can be identified in Kenya. 3 Methodology For this thesis, a qualitative text analysis has been used with a focus on document and policy analysis. The aim was to identify strategies within policy frameworks regarding the same concept and this was done through skimming, reading and interpreting documents that I could later systematically organize with the help of the advanced analytical framework. Text analysis can vary with diverse focus but analysing documents and ideas are one common form. The methodology part is divided into 4 different sections starting with Kenya as a case study and the choice of actors. The second part describes the choice of materials and thereafter comes a short part regarding preconceptions, validity and reliability to explain the perspective of the researcher and strengthen the upcoming results. Lastly, the developed analytical framework will be provided to explain the basis of the analysis. 3.1 Case selection and Choice of Actors Kenya As mentioned, social protection is shown to take diverse forms within different nations regarding ideas, context and capabilities. Therefore, it is interesting to look at a developing nation that is in a phase of advancing sustainable social protection systems. Economic development in Kenya has had a steady increase during the last decades but the country still faces many challenges in regards to social divisions, including different dimensions of poverty, covering rights for all citizens as well as reducing vulnerability and shocks. Improvement on poverty reduction has been made through healthy economic growth but problems with income distribution are still recognized (UD, 2009:13). Households with women as head of the family with many children are often poorer than 12

the average Kenyan households and class distinctions are wide spread where many of the most poor and vulnerable are excluded from the countries resources. Many governments in Africa have been hesitant to create legal commitments to its people regarding social protection despite agreeing to such rights in the international forum (Devereux and White, 2010:68) and with this in mind, Kenya is an interesting case because of their emphasis on social security and protection. As mentioned in the introduction, the new constitution drafted in 2010 as well as the Kenyan Vision 2030 initiative shows ambitions to help the country become more equitable with a comprehensive view of the future in line with the SDG s. Both the Constitution and the Vision 2030 stress the importance of rights for all citizens in Kenya. The development policy for Kenya Vision 2030 aims to provide all citizens with a high quality of life in a clean and secure environment and in the Constitution one can read in Article 43, Economic and Social Rights: 1. Every person has the right (a) To the highest attainable standard of health, which includes the right to health care services, including reproductive health care; (b) to accessible and adequate housing, and to reasonable standards of sanitation; (c) to be free from hunger, and to have adequate food acceptable quality; (d) to clean and safe water in adequate quantities; (e) to social security; and (f) to education 2. A person shall not be denied emergency medical treatment. 3. The State shall provide appropriate social security to persons who are unable to support themselves and their dependants (Republic of Kenya, 2010:31) Additionally, in article 21, the Constitution (Republic of Kenya, 2010:20) outlines that the state and its organs have the fundamental duty to observe, respect, promote and fulfil the rights and fundamental freedoms in the Bill of Rights. These include the rights stated in Article 43. The state should according to the article achieve the progressive realisation of the rights guaranteed under Article 43 (ibid.). In an effort to guarantee these rights the Kenyan Government has introduced a Social Protection Policy to help provide its citizens with safety net services (MLEAA, 2016). Today, Kenya has four nationwide Cash Transfer Programmes. These are: Cash Transfer for Orphans and 13

Vulnerable Children (CT-OVC), Older Persons Cash Transfer (OPCT), Persons with Severe Disabilities Cash Transfer (PWSD-CT) and The Hunger Safety Net Programme (HSNP). Currently the numbers show that between these programs, over 500,000 households get the benefits that these Cash Transfer Programmes provide (MLEAA, 2016:2). Picking only one case instead of carrying out a comparison between different empirical units could contribute to limitations in the analysis and to make generalisations. However, the intention is to look at Kenya as one example of many in Africa that has introduced SP as a tool in their work on poverty reduction as well as being a one of many nations building social policies from a colonial legacy. Kenya as well as South Africa, Zimbabwe and Namibia is presented to be part of something called labour reserve economics (Mkandawire, 2016:2) where regimes were setting up policy s preferable for white people. Being part of a common history and by being one of many introducing social protection as a tool in the discourse of poverty reduction, looking at Kenya as a case could give us an idea of social protection in a developing context. 3.1.1 Choice of Actors The chosen actors for the analysis of this paper are: the United Nations Development Programme (UNDP), the World Bank (WB), and the Swedish International Development Cooperation Agency (SIDA). The choice of actors is based on the criteria that they need to be involved in the global fieldwork of SP programmes and also involved with SP in Kenya. They are also chosen because of their involvement in numerical constructions and implementing of SP around the world either financially or just as an influential actor. All three actors are involved in the Social Protection Floor Initiative where actors have come together to promote SP and together support countries on a local level to establish social services, transfers and to ensure as many as possible with basic social protection (SPF, 2012). Also, because of using a qualitative design and a text analytical method where intensive reading of the material is required, analyzing too many units would not be feasible to get through given the amount of time for a bachelor thesis (Esaiasson.et al, 2012: 46,211). Therefore only three actors were chosen. However, a risk with the 14

choosing of cases comes with the potential absence of actors that could give the analysis more depth with complementary ideas for a more comprehensive analysis. In an attempt to prevent this, the selection of actors tried to identify stakeholders both as individual cases in the social protection discourse but some were also picked for being part of the UN-family that includes many other important actors in the global development discourse. The idea is that chosen actors could then contribute to a wider range of ideas within the discourse. 3.2 Material The materials gathered for this thesis are first and foremost policy documents advanced by chosen stakeholders. Additionally, to enhance a greater knowledge about the field of research regarding social protection and social policy, relevant articles by different scholars have been chosen. In order to escape biased selectivity when choosing the material, hence when the collection of documents are incomplete or that the collection of documents creates a subjective image (Bowen, 2009: 32), it was important to select policy documents from specific stakeholders to find principal ideas from the organisations alone. The policies are also selected to give an reflection of strategies on a global level as well as national level, therefore policy frameworks advanced by actors regarding general social protection ideas and policies focusing on Kenya, if possible, have been chosen. What is important to bring up here is that the Kenyan policy framework is advanced together with other stakeholders and not only by the government itself. 1 This could contribute to frameworks being similar to one another. An overall weakness with the material chosen for this paper is that the majority of them are not as new as I would have wanted them to be. Therefore, it is important to know that ideas can look different today than how they were formulated in earlier years. However, since many of the documents are developed for a specific time-period that are up-to-date, they should still be seen as central and fruitful for this thesis. 1 When visiting the government office of Social Protection I met people from both the Kenyan Government and officials from the World Bank where they were working together to develop new ideas (April, 2016). 15

To analyse UNDP the policy document Leaving no one behind: A social protection primer for practitioners is used. This document is described as a primer practical resource for the global community of Social Development (UNDP, 2016: 4) and thereby contributing with firsthand ideals by the UNDP for the analysis. To get more information of their work in Kenya, the document Mid Term Review of the Kenya United Nations Development Assistance Framework was analysed. The document gave a short contribution to their idea of social protection but at least it provided a bit more in depth thinking of how the UNDP promoted their ideas of social protection in a development context. Thereafter, the policies Resilience, Equity, and Opportunity: The World Bank s Social Protection and Labor Strategy 2012-2022 and Kenya Country Partnership Strategy 2014-2018 by WB are analysed. Both are polices conducted by the organization and can contribute with ideals they favour. The first policy contributed to many ideas for social protection with a focus on labour strategy, and the second policy provided a good foundation of what ideas the WB emphasise on in the development discourse for Kenya. However, the country strategy focused little on SP per see and more on creating prosperity for Kenya, which SP is a part of. Although the policy document as still useful in the sense that it described a development discourse for Kenya. To analyse SIDA, their position paper Social Security Systems is scrutinized. This paper aims to describe SIDA s position within the field of Social Protection (SIDA, 2008: 5). However, the paper describes that it should be used for this subject until the year of 2012 thus, no other policy paper with the same purpose could be obtained which is very unfortunate. The decision was made to still use it as a material for this study though it still contains main ideals of concepts regarding SIDA s position within social protection. The second document used to understanding SIDA s ideas on social protection for Kenya is the document, Strategy for Sweden s development cooperation with Kenya 2016-2020. For Kenya, their official social protection policy has been analysed, Kenya National Social Protection Policy from June 2011. This policy gave an idea of the origin of Social Protection in Kenya. 16

3.3 Preconceptions, Validity and Reliability With regard to the method chosen for this paper it is important for the reader to understand that the interpretations from analysing a text are to an extent based on the researchers prior experience and belief. This can in turn contribute to the researchers conclusion on the studied subject. The distance between a text and its interpretations is also based on the researchers knowledge of the context where the material is drafted, including cultural or social matters (Esaiasson et.al, 2012: 222). Therefore it is important to be intersubjective, in other words to be open and transparent with the way that conclusions are drafted and try to have an objective position. As the researcher for this paper, my interpretations may be colored by my education and prior knowledge about Kenya. I am studying international development where I have learnt to look at development ideas and projects with a critical eye and try to see the latent message. As for Kenya, my social and cultural knowledge comes from living in Nairobi for two short occasions. It is also worth mentioning that I am a woman from a western liberal society, which undeniably can contribute to the results based on specific contextual familiarities. However, as mentioned above the analysis and drafted results will try to be as intersubjective as possible to give the reader insight and to see what the researcher sees (Esaiasson et.al, 2012: 26). 3.4 Analytical Framework I have developed a framework guided mainly by Korpi s perspective on gender and class inequalities but altered by including previous research on social protection in a development context. I have taken a critical stance towards Korpi s conceptual apparatus in terms of their relevance for a development context. By developing a clear framework the idea is to strengthen the validity as well as the reliability of the research thus it creates transparency when measuring what one is supposed to measure (Beckman, 2005:21). The intention with the analysis is to distinguish strategies that might not be clear-cut when just reading the policy documents. Data will be compared and scrutinized in an attempt to organize the contents and identify concepts in the selected documents with a focus on gender and class. After an attempt of clustering concepts in the documents, asking 17

questions will then make the comparison. The questions will be: In what way is this framework similar to the others and what are the differences? and Is there any ideas or expressions mentioned in all policies?. By doing this one can identify similarities, differences and general patterns (Bowen, Glenn A. 2009: 37) and this will be done later in the discussion part of this paper. By giving a detailed explanation of the analytical framework and questions used, the ambition is also to prevent from further problems regarding validity and reliability. Because of the challenge around operationalizing the typologies in this context, the hope is that by describing how the researcher have addressed the text it will still give the reader an understanding of what was captured in the analysis. For it is important that the theoretical definition and the operational indicators are closely linked. If the gap is to wide one might experience lack of validity and the conclusions will not convince by the reader (Esaiasson et.al, 2012: 57ff). The analytical conceptual framework When analyzing the text I will look for this strategic reasoning: Universalism are strategies supporting a universal coverage as a strategy for social protection? Targeting are the strategies describing and favoring targeting methods? General family support model are any strategies giving incentives of backing the traditional idea of women as the primer caregiver, where they are responsible for the household and where entering the paid workforce comes second? Dual-earner model are any of the strategies showing a shift in focus for gender relations where women are given incentives to join the paid workforce and where the care work is more divided between men and women? I will also keep in mind other strategies and theories regarding social policies that might be found in the documents, knowing that they might not be find in documents within this context. Some of those strategies are the State corporatist model where focus is given only to economically active citizens were unpaid labour is excluded. The Basic security model represents a flat-rate beneficial system that is universal and often based on 18

citizenship. The Encompassing model is a combination between earnings-related benefits and the universality of the basic security model and in the absence of formal models like the ones above or if the government leaves it up to markets to decide or if it is up to the community and the family to support, one can identify a market- oriented model or a Community/family model. To continue, the analysis will be structured after four dimensions, created with respect to social protection and issues that are seen as important regarding the thesis question. The dimensions are as follow: Focus within social protection, approach to social protection, approach to poverty reduction, and approach to gender equality. 4 Analysis The analysis is divided into sections where actors are analysed separately. This is done to create balance between the findings regarding each actor and to make it easier when comparing them in the discussion part. 4.1 The United Nations Development Programme (UNDP) One can read early in the policy framework conducted by UNDP in 2016 that the focus of social protection is to reduce poverty inequality, increase human capital and protect men, women, girls and boys from risks (UNDP, 2016:8). The policy also show a focus that goes beyond the idea of just income redistribution, rather they explicitly say that it is important to face other factors including discriminations and inequalities. Here one gets the sense that they have a wider focus that includes creations of capabilities for both gender and class relations. They write: UNDP has a long history and ample experience of supporting inclusive social protection that goes beyond income redistribution to include reforms that tackle social exclusion. These include legal and policy reforms to change disempowering and discriminatory social norms and practices and to create social environments that enable and encourage the most marginalized to register, access, and benefit from social protection (UNDP, 2016: 27). 19

The focus and goal of social protection for UNDP is also described as an investment in the people with effects that helps creating a more sustainable environment, which can contribute to economic growth and social inclusion (UNDP, 2016: 72). This argument can also be found in the assistance framework advanced for Kenya, where the UNDP mention social protection within their four strategic results areas (UNDAF, 2014: 1) and emphasise on the idea that social protection can be seen as a tool in the goal of human capital development. For Kenya, UNDP put focus on how the mortality rate in the country is still high, education and learning needs to be improved as well as creating a more cohesive society with a healthy and skilled population (UNDAF, 2014: 6). Their approach to social protection can be identified as universal where they state that UNDP is a proponent of universal social protection (UNDP, 2016: 41) and they emphasis on universal protection schemes in their framework (ibid: 31). Universalism is also brought up early in a statement in the assistant framework for Kenya where they describe that the Kenyan Cabinet are working to create at least universal access regarding education from early childhood till secondary school. Linking to education and learning, the UNDP also mention that it should be gender responsive (ibid.). It is also evident from their strategies that people are affected differently when exposed to chocks. Therefore they emphasise on means to improve maternal healthcare and childcare in Kenya. The healthcare should also be universal and accessible for everyone (UNDAF, 2014: 7). Another argument linking to their universal approach is that through means of a more comprehensive system one can address more dimensions of deprivation. UNDP discuss the importance of reaching dimensions of poverty, such as the vulnerability and time poverty that results from unequal allocations of unpaid care and domestic work (UNDP, 2016: 41), an argument that can be linked to the idea of the dual-earner model, where they want to change the traditional way of gender relations. The framework also emphasize on how implementation and design of social protection can address constrains regarding gender relations and especially unequal distribution of domestic work. They give the example that benefits regarding childcare and maternity can create new capabilities for women and contribute to economic advancement for women (UNDP, 2016:22). At the same time as this argument present the sense of wanting to change the unequal gender relations it still leaves the question if it really change the relations within 20

the household where it only brings up changes for women. According to Korpi (2000: 146) it is also important to discuss paternity benefits to create equal gender-relations. UNDP does not offer any example of providing paternity benefits to show that men can also do domestic work. One can read strategies regarding maternity leave as a way to create agency and empower women in many paragraphs (UNDP, 2016:34,39,49) but only twice in the primer do they bring up paternity benefits as having an impact on gender-relations (ibid: 10, 35). They do however state that they support national partners efforts to increase the gender-responsiveness of social protection measures so that women as well as men can participate in the economy and manage their household (UNDP, 2016: 29). To link this to Korpi, they do bring up indicators like day care services for children as a mean to balance intra-household relations and to encourage women to join the paid labour force as well the importance to create insurance schemes for unemployment and old age (ibid: 34ff). UNDP address the issue of how women are seen as the primer caregiver for everyone in a family, including the elderly. This issue also increases social exclusion for women, therefore by stating the importance of creating sustainable insurance schemes for all people of risks (UNDP, 2016: 49), UNDP covers many of the indicators that Korpi link to the dual-earner support model. By providing social protection for all members in the family, the reliance of the woman as the caretaker reduces. And even if they do not develop the importance of paternity benefits they still state the idea of how it can give impact on gender equality. UNDP also strengthen their argument of going beyond the traditional family model by discussion how the conditional cash transfer (CCT) programs that are common social protection programmes in Latin America, are rather reinforcing a traditional gender role where targeting women in order to carry out unpaid work and not include men s responsibilities with domestic work are common. This also shows that UNDP are aware of the consequences that policy structures can lead to. Regarding SP in Kenya, UNDP underline issues of ineffective system and that interventions made for SP needs to be improved to create a well coordinated, effective and a sustainable system on a country level as well as on a national level. Their idea is that the Government of Kenya needs to strengthen their commitment and ensure a leadership role (UNDAF, 2014:8). Here we can see that UNDP emphasise that SP is 21

primarily something that should be provided by the government. Their aim is to offer help for the government to ensure good quality social services to all Kenyans, which contributes to the idea of universality. However when describing the goals of coordination of social protection in Kenya they write county Governments & partners have adequate technical & financial capacity to set up a functional social protection coordination mechanism at national & country levels, which are inclusive & ensure harmonization, efficiency complementary & synergies of interventions (UNDAF, 2014:17). In this statement they prevent to write out that the systems should be universal. Even if they write national & country level one can think that they do not really follow through with their idea of social protection being universal in Kenya. Following that they also emphasise on social inclusion for everyone in order to prevent deprivation due to economic setbacks or incapacity to work (UNDP, 2016: 12). To sum up, the goal of UNDP with social protection is to reduce poverty inequality with an increased protection of vulnerable people and increase human capital and social inclusion. Their approach is clearly supporting a universal model of coverage, with small hesitations regarding coverage in Kenya due to ineffective systems. Moreover, one of their ideas shows encouragement to go beyond redistribution of income and create more social inclusion, which then includes changing paths of inequality and disempowerment, by creating reforms of including the most marginalized groups so social protection can benefit everyone (UNDP, 2016: 46) which illustrate strategies going beyond the state corporatist model where some people are excluded. It also links us to how they in many ways show the importance of empowering women as a main topic where they both bring up paternity benefits, promoting childcare and better protection for elderly to change household relations. All these indicators show that UNDP are much closer to a dualearner model than supporting a traditional family model where they promote diverse actions to improve capabilities for women and encourage them to join the paid labour force. 22

4.2 The World Bank (WB) The World Bank (2012: 98) state that the key role and the focus of social protection is to provide help in risk and vulnerable situations as well as being a tool for a pro-poor economic growth to reduce poverty. Escaping poverty and SP could also improve productivity and reduce negative outcomes from shocks (World Bank, 2012: xi). Moreover, WB put emphasis on the goal of creating new jobs and sees this as a critical topic (World Bank, 2012: 27) as well as enhancing productivity where providing young children with good nutrition could increase productivity later on (ibid: xvi). This focus on social investment can also be found in the Kenyan strategy where they see it as critical to develop peoples potential to increase wealth and end poverty (World Bank, 2014: vi). The importance for Kenya to invest in healthcare, education and housing can be identified throughout the strategy, especially for vulnerable communities, children and women (ibid: iv). Investing in all these spheres of life will enable better participation in both economic and social life and later lift the county out of poverty (ibid: 4). In their policy document, WB presents a more inclusive coverage of social protection however they do not state that they have an universal approach. When addressing the coverage gap they write SPL (Social Protection and Labour) programs need to be available to all, but lack of coverage remains a pressing issue (World Bank, 2012: 22). WB is pushing for inclusion of rural areas and informal sector jobs (ibid.) similar to UNDP and their argument of introducing means that have and inclusive approach to reduce risks of disempowerment and discrimination. However, WB links the inclusive approach to how larger-scale programs often comes with bigger problems including being cost ineffective, badly coordinated and not responding to shocks in an effective way (ibid: 15). The question is if the argument about supporting targeting methods comes because a more universal one can lead to more problems? They do write that in fragile and the lowest income countries it could be good to start with systems that only cover the most vulnerable and then progressively try to extend the SP programmes (World Bank, 2012: 30). In the country strategy for Kenya, the WB shows that targeting methods are favoured. They state that the social protection policy that Kenya have now with targeting systems are fruitful which is something that they support and the policy should continue 23