The politics behind social protection in Nepal

Introduction

Social protection in a broad sense consists of policies and programs that provide access to essential social services such as health and education and ensure an adequate level of security under multiple contingencies of life related to unemployment, sickness, etc. While many programs could be labelled as social protection in Nepal (one mapping exercise identified over 70), in practice many Nepalese bureaucrats consider the government’s cash transfers as social protection. However, the official definition of social protection comes from the Government of Nepal’s draft social protection framework: ‘Social protection is defined as a set of policies and actions aiming at reducing poverty and multidimensional deprivations to ensure a basic minimum livelihood for all citizens. It includes all non-contributory and contributory measures’ (NSPF, 2012: 20).11 ¹

There are certain incentives and disincentives that build social protection commitment and capacity in fragile states (Harvey et al., 2007). These incentives and disincentives are hidden behind the generic term ‘political will’ in the literature (Schüring, 2012: 164). Schüring and Lawson-McDowall (2011) argue that a more nuanced understanding of what shapes ‘political will’ in each country and how this then affects the social protection landscape is necessary if social protection is to be strengthened. This is important because ‘the greater the need for social protection, the lower the capacity of the state to provide it’ (Devereux, 2000: 2005).

It is often assumed that social protection gains political party’s votes and that this is a major cause of its popularity. However, the way social protection is utilised may be different in countries transitioning to democracy and those that are conflict-affected. The balance of power in these societies is less likely to be with voting citizens and more likely to still be with powerful interest groups and elites that are accustomed to using government institutions to distribute rents ² and favours (Barrientos and Hulme, 2009: 14). So why would social protection exist in such states at all? The way different political parties use and rely upon policies for their survival and legitimacy, versus clientelism or economic growth may determine why certain social protection policies are adopted (Lavers and Hickey, 2015: 11). This article develops a Nepal-specific understanding of the politics of social protection.

Nepal presents a useful proposition to study the politics of social protection, as it is a conflict-affected state transitioning from an exclusive monarchy to an inclusive federal democratic republic and has a range of social protection programs. While there is considerable literature on social protection in fragile and conflict-affected states heavily reliant on donor assistance, such as in Africa and elsewhere, these countries do not have Nepal’s caste system and have very different historical experiences of political reform and state building. India has a powerful influence over Nepal and also has the caste system, but one thing the literature on social protection stresses is its context-specific nature (Devereux and McGregor, 2014).

The article is situated during a changing political settlement. A 10-year Maoist insurgency ended in 2006 with the signing of the Comprehensive Peace Agreement (CPA). The settlement reached to end the conflict was inclusive of many traditionally excluded population segments. ³ Thus, many new interest groups that had never had this kind of opportunity before were given a stake in the maintenance of power. The Maoists won a surprising victory in the 2008 Constituent Assembly elections and formed their majority with Madheshi ⁴ parties. After a politically turbulent four years, the Maoists lost power in 2012.

This article’s structure starts with the methodology, and introduces Nepal and the theoretical framework and then the historical context of social protection. The attitudes and manifestos of the main Nepalese political parties (the Communist Party of Nepal (CPN-UML), the Nepali Congress (NC), the Maoist Party (UCPN) and the Madheshi parties) are examined for their social protection content. The comparative analysis highlights the differences between parties with regard to social protection as a political tool and development approach. It identifies discrepancies between party programmes, party discourse and statements of individual party representatives and implementation. Political parties are networks of power and their position can change quickly with opportunities to augment power. The role of third sector actors is introduced for the pivotal role they play in Nepal’s political settlement, before concluding.

Study design

Grounded theory provides useful tools and allows for a mix of methods to be used in order to learn about individual perceptions and feelings, and assists researchers to make connections between events, processes and perceptions. It focuses on identifying key components and categories of a social phenomenon and how they relate to each other. In a grounded theory, data analysis and collection occurs concurrently because analysis of the data informs the next cycle of data collection. Corbin and Strauss (2008: 145) call this ‘theoretical sampling’. This process allows categories and themes to be refined again and again until a theory emerges (Lingard et al., 2008), thus making it ‘a process rather than a product’ (Glaser and Strauss, 1967: 40).

The Constructivist version of grounded theory (Charmaz, 2002; Charmaz and Mitchell, 2001) used in this research has its ontological roots in relativism and also takes a reflexive stance on the modes of knowing. It pays close attention to empirical realities and to people’s collected renderings of them, and in fact locates itself within these realities and builds its theory from it (Charmaz, 2006). With the grounded theory approach the researcher must ‘stand within the research process rather than above, before, or outside it’ (Charmaz, 2006: 180).

Theoretically oriented forms of research are useful when studying issues where there can be multiple viewpoints. Since Nepal’s 10-year conflict ended in 2006, overcoming centuries of discrimination based upon caste, ethnicity, religion, gender and region is necessary but the direction and extent of the need for change is contested by various interest groups. Nepal is in a situation of high politics where rhetoric, political wrangling and hidden intentions abound. Understanding why certain policy decisions are taken or why manifestos are written as such may not be general knowledge in such a context, and reading such documents may not offer the ‘truth’. In particular, political parties are considered the least trustworthy institution on Transparency International’s 2014 Corruption Index for Nepal.

‘The truth’ may be an elusive concept in highly divided post-conflict societies. Therefore, the researcher would ask more challenging questions than in the interview guide to deeply probe the rhetoric or ideology espoused during the interview. Bernard (2002: 214–215) describes this kind of probing as ‘baiting’. This technique worked well in terms of building rapport. Many elites would often smile when the researcher ‘baited’. By showing elites that the researcher understood the political dynamics and their game, the researcher often heard phrases such as ‘you have obviously been in Nepal for a while so I will tell you the real issue…’. Interviews and the interview process are ‘a game of positionalities’ (Ward and Jones, 1999: 304). The position of the researcher opened some doors and some ‘truths’ and possibly closed others that researchers with different backgrounds may have opened.

The subjectivity of the researcher is overcome in grounded theory by testing the theories identified from the data analysis with participants and allowing them time to reflect on the researcher’s observations, or strategically in a group setting through presentations. The researcher does not just ‘follow their nose’. The use of historical evidence, case studies, manifestos, policies and secondary data and comparing this to the primary data collected further helps to reduce the potential for bias and improves the quality of the data. Political settlement typologies were used to provide a comparative framework and independent variable to test the validity of the data.

Evidence was collected in phases over time to ensure credibility and to reduce data collection gaps that may have arisen at the end if the data had not been subjected to theory testing along the way. From July 2012 to February 2014, 66 key informants were interviewed (50 Nepalese and 16 foreigners, 41 males and 25 females), predominantly residing in Kathmandu. Key informants included elite decision-makers and policy influencers, political parties and politicians, journalists, academics, non-government employees, private sector employees, trade unions and development partners (foreigners or Nepalese). Development partners play a prominent role in social protection in low-income states, often as policy drivers and funding bodies.

The interviews were then sorted and coded for emerging themes. ‘Open coding’ is when ‘the researcher breaks down and categorises the data into manageable segments’ (Ary et al., 2010 [2006]: 553). In the current research this was achieved by the researcher asking the following questions:

What does the data suggest/pronounce (Charmaz, 2006: 47)?

The unit of analysis for coding tended to be sentences or phrases, but incident and opinion coding also proved useful. Axial coding then followed which involves ‘a set of procedures whereby data are put back together in new ways after open coding, by making connections between categories’ (Corbin and Strauss, 2008: 96). Selective coding was then undertaken, which is ‘the process of selecting the core category, systematically relating it to other categories, validating those relationships, and filling in categories that need further refinement and development’ (Corbin and Strauss 2008: 116).

Deviant cases were managed through scope conditions, the grounded theory method, case studies and typologies. Using ‘scope conditions’ (see Ragin, 2000: 61–62; Walker and Cohen, 1985) involves comparing data across issues. For example, this may show that the political settlement appears to operate in a similar way across the results, or it may reveal certain consistent aspects of state functioning, suggesting that deviant cases have been accounted for. Deviant cases are also tested by other respondents. If the outlying case is deemed relevant to respondents, then a rethinking of the theory follows. Rather than ignoring deviant cases, grounded theory considers them to be of substantive interest.

Nepal

Nepal is still on the World Bank’s Fragile State List. ⁵ Nepal is categorised as a LDC with a per capita GDP of US$721 per annum (MoF, 2014: 1). Nepal ranks 145th out of 187 countries on the 2014 Human Development Index, with a score of 0.540. ADB (2014) estimated 2014 GDP growth at 5.4 per cent due to a favourable monsoon, an increase in the services sector and increased remittance flows. However, this growth has since been eroded by the 2015 earthquakes.

Nepal is a heterogeneous country with 125 caste/ethnic groups all with differing levels of human development. Since the Rana Dynasty unified Nepal as a country in 1768, a Hindu based social stratification system and its legacy has excluded and discriminated against Dalits ⁶ and indigenous ethnic groups known collectively as Janajatis. ⁷ Colonial strategies of constructing administrative and political classes to manage the empire involved selectively recognising the ‘rights’ of organisationally powerful groups as a method of political stabilisation and as part of divide-and-rule strategies (Khan, 2009). In Nepal’s case the intermediate class consists of the Brahman and Chhetri castes that sit at the top of the caste hierarchy, just under the Rana ‘ruling class’.

Pro-democracy movements began after 1940 and were organised by Brahman–Chhetri groups (upper castes) seeking access to Rana power (Hangen, 2007). A slight reduction in the Rana hegemony was achieved in 1951 when a new government was formed with predominantly Nepal Congress (NC) party members (Whelpton, 2005). In 1959, under King Mahendra (1955–1972), a party-less local government system known as the Panchayat system was imported from India. Respected (usually wealthy) elders were chosen by the local community to settle disputes between individuals and villages (Whelpton, 2005). The Panchayat system augmented upper caste dominance in local areas by undermining the many types of indigenous governance structures operating at the local level. This system remained in place (despite a 1979 referendum) until 1989 when the first Jana Andolan (people’s movement) allowed all citizens of Nepal the opportunity to participate in changing government structures.

A parliamentary monarchy was established in 1990 with the Rana king as the head of state and a PM as the head of the government (a multi-party democracy) (Whelpton, 2005). However, this period had mixed results for Nepal’s excluded groups. The two major political parties at the time, the Nepali Congress (NC) and the Nepal Communist Party–Unified Marxist Leninist (NCP–UML), both had a majority of Brahman and Chhetri members and kept collapsing due to inter- and intra-party bickering (see Bleie, 2011; Parajulee, 2010). As Lawoti (2007: 29–30) explains, ‘almost all state power was enjoyed by the top leaders of the ruling party’ and even opposition parties could not influence policy. The Maoists, who had the third largest contingent in the 1991 parliament, were repeatedly ignored by the government and went underground in frustration (Lawoti, 2007: 30).

The 10-year People’s War led by the Maoists fought for the abolition of the monarchy and for the inclusion of traditionally excluded groups. Nepal’s Comprehensive Peace Agreement (CPA) was signed in 2006. The 2008 elected Constituent Assembly led initially by a Maoist and Madheshi coalition declared Nepal a federal democratic republic and began developing a new constitution. The Consistent Assembly was a unicameral parliament, consisting of 601 members – many first-time parliamentarians, a quarter of whom were illiterate due to affirmative action measures initiated through the inclusive state building project (World Bank, 2010).

Having finalised the constitution in September 2015, two Constituent Assemblies later, Nepal finally emerged from its formal peace process. Nonetheless, social unrest continues, as many excluded groups believe that the rights they were promised after the conflict ended were not upheld in the new constitution. The importance of social inclusion (as well as that of social protection) ended when the Nepali Congress regained power in 2013, and reasons for this are explained below. Moreover, social protection in Nepal cannot be separated from the broader struggle to achieve a more inclusive naya ‘new’ Nepal.

Regimes and power distribution

Different types of regimes shape opportunities for contestation and rent seeking (Tilly and Tarrow, 2007). According to Tilly and Tarrow (2007: 45), ‘regimes consist of regular relations among governments, established political actors, challengers, and outside political actors, including other governments’. The nature of the regime determines how citizens engage with the state and the degree of rights enjoyed (Cornwall et al., 2008). Even when formal institutions guarantee incomes and rights, these can be effectively challenged through informal processes in clientelistic regimes (Khan, 2010: 54). The source of legitimacy of regimes may be associated with the backing of powerful interest groups and elites and this involves bargains and negotiations and the distribution of rents to buy their support, rather than the views of the public through elections (Khan, 2010).

A more recent term used in the literature about developing countries to understand politics is the ‘political settlement’. Castillejo (2014: 1) defines a political settlement as ‘a dynamic bargain (primarily between elites) on the distribution of power and resources that is subject to changes and re-adjustments over time’. According to Khan (2010: 57), ‘all developing countries have clientelist political settlements, but they differ across countries, and change over time’.

In Khan’s (2010: 65) typology, Nepal best fits the situation of competitive clientelism whereby stability is maintained by cycling power between competing elite factions. The downside is short-term policymaking and weak implementation capacity resulting from the bargains made with other factions to maintain power or privilege (Khan, 2010). In this settlement, the ‘subjects’ ⁸ are less involved because the nature of the agreement is between ‘principals’ (elites, interest groups and political leaders) (Levy, 2014: 18). In a competitive clientelist state:

the number of potential factions is so great or they are so fragmented that the inclusion of all of them in a ruling coalition would not work but neither would a strategy of keeping excluded groups out by legal or military mechanisms. (Khan, 2010: 67)

An additional variable to consider is the distribution of power, as this impacts upon how decisions are made. In a ‘horizontal distribution of power’, the power of excluded coalitions is weak or fractured relative to the ruling coalition, giving them a longer time horizon to govern. Horizontal power distributions are more exclusive than vertical distributions as legitimacy comes from elites. Thus, elite interests and keeping the elite happy is the priority. A ‘vertical distribution of power’ will work better for the long term public interest because the self-interest of those with power aligns with what the public wants (Khan, 2010: 65). Vertical distributions are more inclusive as legitimacy comes from looking after the well-being of citizens.

However, in a vertical distribution of power, the more powerful that lower-level factions become, ‘the greater the number of points at which the enforcement of particular rules can be blocked’ (Khan, 2010: 65). Given the pyramidal structure of patron–client organisations, the distribution of rents to many or all of the lower-level factions will be required to ensure their cooperation (Khan, 2010: 65). This will obviously be difficult to achieve in a poor country (because it is costly and distribution channels are weak) and may involve a number of adverse trade-offs and a new regime (Khan, 2010: 67). Very often, all the hands involved in distributing rents incurs a fee. The results below will show that the Maoists have a vertical distribution of power and the NC a horizontal distribution of power. As the UML are an older party but also debatably communist, they have a hybrid distribution of power. The Madheshi parties are too new to predict at this stage.

Additionally, in vertical distributions of power, citizens and third party actors – like cooperatives and unions – can play an important role. Di John and Putzel (2009) label these groups/institutions ‘third party enforcers’, as they have ‘veto powers’ that come from their ability to generate revenue from outside the state, such as from a paying membership base. So while traditional elites may still use the state’s resources to maintain their power, new ‘political entrepreneurs’ can emerge as leaders of third sector organisations and challenge this power. In this way, vertical power distributions are more inclusive as they involve many more actors. This article finds that whether the distribution of power is horizontal or vertical has great relevance to understanding social protection reforms.

Results: Historical evidence

This section examines the historical evidence surrounding the introduction of Nepal’s cash transfers. Nepal’s first cash transfer was announced in 1994 by the Unified Marxist Leninist party (UML). The PM’s ⁹ adviser at the time explains: ‘Touching people directly was the essence of UML’s approach. A universal transfer appeared the fastest way to reach as many people as possible’. ¹⁰ The UML used social protection to capitalise on the opportunity to differentiate itself from the old ways of politics and make a quick impact as a new party with a new agenda. It also had to prioritise ending opposition-spread rumours, as the PM’s adviser explains:

There was a strong rumour that the communists [UML] did not care about the old people and they would be killed because they were not productive. Different contesting parties were pushing this kind of slogan and the rumours had to be killed off. The whole idea was basically about going directly to the people and becoming popular. ¹¹

Social protection was not the only approach used to achieve political gain in the mid-1990s, however. The UML introduced a social democratic approach called ‘let’s build villages ourselves’. This method of reaching the people and encouraging participation in local planning decisions was seen to strengthen democracy and development. A bureaucrat involved at the time explains:

I think of it as something that promoted democracy. People came together to deliver on what is important to them rather than it being decided by the central government or even the VDC chair. It directly involved households deliberating on certain issues and promoted democracy by showing that everyone’s voice is important. It was more important than looking for how many kilometres of roads were constructed or how many water taps were built. ¹³

Initially the senior citizen allowance was meant to be for all over 70, but the budget funds available at the time resulted in it being raised to a 75-year age eligibility. A senior bureaucrat involved at the time explains:

we worked out the details – if we have this size of population crossing 75 years of age and if this is the amount, this is the budgetary liability, which was not even one per cent of the budget then. So, we thought that it could be afforded. We also thought that it could be covered with additional tax measures. ¹⁵

The theory of political settlements can help to understand how the state functions in Nepal and the role of social protection within it. New social protection announcements are made when parties take power because rents are available to buy support with and access to the state budget is easier. New parties have certain internal dynamics and models of leadership that can override existing institutional arrangements of clientelism that may later block the use of the budget (Natali and Rhodes, 2004: 6). After a period of time in power, fewer rents are available, a client’s loyalty changes or their support comes at a higher cost, support wanes, reform ideas are blocked and the ruling party loses power.

The senior citizen allowance was, and still is, immensely popular. Despite its association with the UML, the PM’s adviser explains: ‘every subsequent government has kept the old age allowance going because it has social merit and carries the message to society that the state is taking care of its citizens’. ¹⁶

In 1994 the eligibility age was 75 years and the benefit amount was NRP100 (US$1) per month. In 1996, the NC announced a widow’s and a disability allowance for the same amount. Towards the end of the conflict, in 2005 the NC increased amounts to NRP150 (US$1.50) per month, and then in 2007 the Koirala-led NC raised it to NRP200 (US$2) per month. In 2009 the UCPN–Maoists increased most allowances to NRP500 (US$5) per month and lowered the age eligibility as well as introducing new transfers. As one MP noted, it is ‘very hard to convince the people of policies because citizen consciousness for democracy is low’. ¹⁷ Thus, all political parties consider the role cash transfers (a direct, quick and visible form of assistance) can play in maintaining their support.

Social protection’s popularity with constituents can also influence the shape of social protection. As one NC party informant explained:

consolidation is very hard. Of course we have to consolidate. But once we have already given it, it’s very hard to withdraw unless you add more things to justify it. I was talking to the party about this free healthcare proposal and I proposed at the beginning that maybe we can manage with just health security and food security. But the question is who is going to withdraw the aged pension? If you go to the old people and say, ‘Do you want health insurance or 1000 rupees?’ they will say 1000 rupees because that’s direct. So you cannot do that. ¹⁸

There are multiple incentives for choosing to introduce cash transfers and fewer incentives for reforming them in a weak state. The cash transfers suffer from misappropriations as all the budget is spent on the actual transfer and none on monitoring and accountability. If beneficiaries do not receive their rightful cash transfer entitlement then this should influence their vote and make parties care more about social protection implementation, but it does not. As an NC party member lamented, ‘distributing money but not achieving the right result is not the proper way to distribute the money. But what can we do?’. ¹⁹ Thus, even when political parties understand the situation they do little to improve it. The Maoists are a notable exception as they instigated a National Social Protection Framework in 2009 to reform and consolidate all social protection programs, suggesting that they do care about the quality of social protection. There is a difference in political parties in the way they maintain power and value citizen votes and in their approach to social protection, which is discussed in the next section.

Political parties, their manifestos and social protection

Through examining the manifestos of the four main political parties and in-depth interviews, it seems that the political parties use and understand social protection differently. As one bureaucrat stated ‘all three major parties have a strong commitment to provide social security. There may be some difference in ideas but at the root they want to provide at least some protection to needy people’. ²⁰ But the question remains, at what cost? Clearly passing reforms or making new announcements involves political deals and wrangling. A union member explains:

Social security in every country has a high political benefit. All parties want to do something new with social security when they are in power because it is more popular than managing a budget or economic growth. Yet parties in opposition don’t want the ruling government to introduce more schemes that are popular so they raise concerns about corruption and affordability and such things. ²¹

It was commonly heard that party positions are developed as needed and by those at the top and may contradict the manifesto if an opportunity to gain popularity or form alliances arises. Although central-level members had a reasonably good understanding of social protection and what was written in the manifestos, this decreased considerably at the local level. In fact, some local-level parties described the positions of other parties without realising it when they answered questions. The political party manifestos are explored in more detail to better understand the differences between the parties.

The Unified Communist Party of Nepal–Maoist

The UCPN–Maoist manifesto is the most detailed and comprehensive approach to social protection. It speaks of developing a National Social Security Strategy that is funded by a National Social Security Fund and a 180-day employment guarantee to families in need, based on the belief that ‘if we fail to give employment to the people then we should give them an unemployment allowance with skills training’. ²² According to the party representative, they have always considered social protection to be a core pillar of their approach:

We initiated a national pension scheme and the Social Security Act when we had power. But we failed to convince other traditional parties. They don’t like social security, or welfare or pro-people programs; that’s the problem. And we failed to rightly communicate our vision to the people.

The Maoist 2013 manifesto promises to double most cash transfers. Equality appears to be a central objective of the UCPN–Maoists, as explained by the party representative:

Our party supports a progressive taxation system because this will make the people equal. Those who earn must have to contribute to the national pension scheme and distribute to all. If you earn, and if you get back it is not welfare. If you earn and if you share with me or others who are not involved in earning that’s welfare.

According to a Maoist party representative, redistribution can be used to correct inequality in Nepal:

An accountable government and state takes care of all the people. To do this they need a plan to protect them. There is a gap between the ‘haves’ and the ‘have nots’ in Nepal and this caused the conflict. So the state needs to bridge this gap. Only after equal footing and equal protection will the state be people friendly, welfare oriented and accountable, peaceful and stable.

A number of respondents suggested that delivering social protection as a peace dividend was a convenient power-building tool for the Maoists. An international adviser who worked for a development organisation in Kathmandu from 2005 to 2009 explains:

at the time, I heard that because the Maoists were planning to introduce allowances for ex-combatants in the cantonments they felt that doing all kinds of other social protection grants would pave the way for that. But it might also be that they felt that they should do something that people see right away. ²³

According to respondents, the Maoists suffered from disputing internal factions when in power and this also affected social protection. According to one interviewee from the bureaucracy, the Baidhya-led Maoist faction that left the party in 2013 considered social protection ‘a capitalist supporting tool’ that ‘prevents revolution’. ²⁴ A bureaucrat explained this line of thinking, ‘when people have full stomachs then they have no need to be radical. They will become lazy’. ²⁵ Yet, the Maoists have a broad support base to keep loyal, and cash transfers are a cost-effective way to reach large numbers of formally excluded citizens. The tension between factional ideology and the reality of operating in a competitive clientelist settlement was not explored by the research but merits further investigation. Internal factions were considered the reason for the contradictions in the Maoist approach to social protection by one respondent.

A bureaucrat explained that when the UCPN–Maoists were first in power they were conflicted: ‘they wanted to give a stronger role to the state but also to recognise the market’, but the state could not afford much and the market is under-developed in Nepal, so they brought in third sector organisations:

There are also other actors in our country with huge power: the community, the cooperative sector and the non-government sector. I call these the third actor. They have so many schemes – microcredit or cooperatives, community health insurance, community crop insurance – they are so powerful in creating jobs, distributing income and in providing social protection and semi-formal social protection. The third actor is so important to reduce poverty and also to widen and deepen social protection if not social security. ²⁶

The Communist Party of Nepal–Unified Marxist Leninist

The UML’s manifesto covers social protection but does not dedicate as much space as the Maoist manifesto to outlining their position. Their 2013 manifesto mentions developing policies and mechanisms to improve the implementation of an integrated social security system and implementing the contribution-based Social Security Fund and health insurance for those in employment. Special priority will be given to women and people with disabilities. The promises it makes to various groups include different cash transfers along with health, education, skills development and employment assistance. It also states that the amounts of all existing cash transfers will be increased, but amounts are not specified. However, a respondent explained that during the 2013 campaign they promised to increase the age pension to NPR2000 (US$20) per month. This appears to be a legitimate intention. A news article from 2014 reveals the UML putting pressure on the NC to increase the senior citizen allowance, as promised during their 2013 campaign (Republica, 2014). The UML manifesto explains that taxation will cover lifelong social security for targeted groups and that social support programs and other forms of relief will be provided to ‘backward’ regions and senior citizens so that they can have ‘respect and a productive lifestyle’. A UML informant was clear that social protection funding should come from increasing revenues and taxes and not from development partners. There was a clear understanding of the interrelationship between social protection and democracy, inclusion and state building, by UML. A UML member explained:

By promoting social justice our democracy will be strengthened. Our society is very diverse and there is a big gap between groups. By planning to uplift all those who are discriminated against and excluded from politics and government organisations, and by giving them some political, economic or social rights, the democracy will be strengthened.

The Nepali Congress

According to the mix of programs highlighted in the NC manifesto, ²⁷ social protection is considered a form of charity for the destitute that can also be used to increase productivity. The manifesto confirms a focus on cash transfers, but also includes measures to improve health and education for people who cannot take care of themselves, plus social security and pensions. It promises to double the age pension to NRP1000 (US$10) per month and to review it every two years; to provide free food, to create aged residences and to provide subsidised health treatment for all senior citizens over the age of 80; to increase the birth assistance allowance to NRP10,000 (US$100) for women who give birth at health centres; to provide an allowance to unemployed youths; and to improve labour rights and income-generating programs for landless squatters. The NC manifesto discusses poverty targeting and identification and assistance for those below the poverty line.

An interviewee explained that although the NC has promised to increase all allowances as the cost of living has increased, they would not be too generous:

one should not think that they will get an allowance their whole life and not work. Just getting an allowance from the government is not good…Many here [in the party headquarters] believe that people should solve their own problems as an individual. Each individual has the responsibility to make their own life. ²⁸

The NC promotes economic growth first and other measures such as social security when they align or if they are needed to reach the people. The social protection sections of their manifesto appear well thought out. An informant ³¹ explained that an ex-Finance Secretary developed the social protection aspects of the NC’s 2013 manifesto, suggesting little consultation occurred. As the NC has not increased cash transfer amounts as promised since they took power in 2013, their conviction for social protection even as a form of charity remains questionable. The NC has a horizontal distribution of power that mainly involves elites. Their need for citizen support (and maintaining promises) is less important.

The Madheshi parties

A brief history of the Madhesh and Madheshi parties is provided to contextualise the Madheshi party responses. The history of the fertile Madhesh/Terai region that borders Nepal and India is one of ‘entangled territories, recurring disputes and conflicted claims to land, taxes and political authority’ (Bernardo, 2010: 20–24). A large number of former Maoist supporters from the Madhesh/Terai broke away to start their own movements after the Maoists sidelined many high ranking Madheshi officials (Miklian, 2008). A series of bandhs ³² were called by Madheshi and Janajati activists, starting in 2007, and these continued ³³ even after the Madheshi parties had formed a coalition with the Maoists in the historical 2008 Consistent Assembly election. Prior to this, janajati and Madheshis were prevented from forming political parties.

Three Madheshi parties were interviewed; two in Kathmandu (the Terai Madheshi Party and the Madheshi Jana Adhikar Forum), and the Sadbhavana Party in Sarlahi. These parties comprise the coalition, United Democratic Madheshi Front. Madheshi parties have a peripheral influence on social protection and do not consider it a priority. One Madheshi interviewee explained, ‘This topic is not a priority for us now because we are still fighting for the constitution, for recognition, and for citizenship. First we need to be recognised as Nepali, and then later we can get social protection’. ³⁴ The Madheshis are largely ignored in the media when it comes to social protection; the news article ‘Parties make tall social security promises’ fails to mention any Madheshi parties. ³⁵ Despite this, Madheshi party respondents had thought about social protection and felt that increasing all cash transfer amounts was important because they strengthen democracy:

It is the duty of the state to deliver the allowance. It will strengthen the democracy if the government takes care of disabled, women, widows…The government might change from time to time, but receiving allowances is possible only when there is democracy. ³⁶

Most of the Madheshi parties did not see social protection as the means to reduce inequality, like some of the larger parties did. For Madheshi informants, the caste system and other forms of exclusion would only end with economic growth and an educated nation:

We would like more education for Dalits. Some reservation should be offered, free education for Dalit, even free meals for Dalit students so they will be more encouraged to go to school. Then they will get a job. If they will get more work, their lifestyle will automatically change. ³⁹

Discussion of political parties

In many cases the manifestos appear to be developed for the sake of regulations and have little bearing on party views. Political parties are allowed to make promises that do not align with their manifestos or their actions because political and democratic accountability is weak. ⁴⁰ It is not unusual globally for political parties to say things when campaigning and then to do different things when in power or opposition. The difference in Nepal is that citizens have a low understanding of democracy and what political parties can achieve. As a UNDP (2014: 1) report notes, ‘there has been no systematic study on the role of party manifestos in shaping election outcomes and the degree to which the parties adhere to the manifestos thereafter’. If there is very little public awareness of policy options and people cannot understand the net implications for them personally or even what it means for national budget allocations, it leaves them with little real choice about who to vote for (Moore, 2005: 269) unless they see a direct benefit such as a cash transfer.

Social protection in Nepal incorporates older patronage models. ⁴¹ Many of Nepal’s informal systems support patron-client relationships. As a respondent explained, the patronage system is very strong in Nepal:

Individual goods in a patronage system now substitute for good public policy in Nepal. Parties get elected on the basis of the flow of individual patronage to people. Individual favours, not taxation policy or good education policy or better roads and things like that. ⁴²

Different parties understand the purpose and value of social protection differently. The way in which the UCPN–Maoists introduced social protection programs to Dalits and endangered indigenous groups suggests alignment with their social inclusion agenda. However, the amounts are too small to be transformative and as such are token gestures that may serve the purpose of generating loyalty, rather than alleviating poverty. However, the Maoists, the UML, the Madheshis and some NC members saw the value of cash transfers to democracy. It is not clear how the parties interpret democracy given the prevalence of patronage. Is their desire to consolidate democracy real or is it a cover for the continuation of power maintenance through clientelism? More research is needed on the disconnect between practice and rhetoric because both must serve a purpose.

In a competitive clientelist settlement, ruling coalitions will be more likely to use social protection as a form of patronage to prevent factions from diverting their loyalty to opposition groups (Lavers and Hickey, 2015: 11). As a form of patronage, cash transfers can buy the support of end recipients and provide local political parties with funds (by allowing them to capture transfers), and favoured ministers can be given large cash transfer programs from which to manage and extract rents. This is why Gough and Wood (2004: 323) rationalise that ‘the transformation of informal security regimes into welfare state regimes entails a subtle and complex process of de-clientelisation’.

The NC may suggest that individuals need to fend for themselves, and yet they make social protection promises and do not reform social protection in line with their ideology because it would be too unpopular. Ferguson (2007) identified similar findings during South Africa’s income grant introduction. He found that while at times there were recognisably ideological moves that could be labelled ‘neo-liberal’, ‘communist’ or ‘welfare’, they were combined with other moves of a different orientation and with different alliances in order to achieve various personal gains or political ends (Ferguson, 2007: 80). The next section explains how the way power is distributed determines social protection programs more than ideology does. Ideology is a luxury in a competitive clientelist settlement.

The role of the third sector

Political parties can only afford to align with ideology at the beginning of their time in power. New social protection announcements are made when parties take power because rents are available to buy support with and access to the state budget is easier. New parties have certain internal dynamics and models of leadership that can override existing institutional arrangements of clientelism that may later block the use of the budget (Natali and Rhodes, 2004: 6). After a period of time in power, fewer rents are available, a client’s loyalty changes or their support comes at a higher cost, support wanes, reform ideas such as the National Social Protection Framework are blocked and the ruling party loses power. The theory of political settlements provides a dynamic understanding of how the state functions in Nepal and the role of social protection within it.

However, an additional variable aside from the political settlement is the power distribution. A case study on the establishment of a new ministry and the move to poverty targeting for Nepal’s cash transfers is provided to illustrate the role of the third sector in the political settlement and the way the distribution of power can determine their involvement, as well as the way this involvement may shape social protection. In the absence of a sufficient formal budget (or the ability to access it), political elites need to generate and distribute sufficient off-budget rents and benefits to keep or bring powerful actors and groups, like cooperatives, into the political settlement (Khan, 2010; Lavers and Hickey, 2015: 6).

In countries like Nepal, support for reform may ebb and flow depending on which elites hold the balance of power. As the theory section explained, third party actors are able to increase their power quickly in transitioning states because their membership base and hence revenue generation is not tied to the state and the time-consuming political bargains occurring in that space. Thus, they are independent of the traditional elite structure and have ‘veto’ powers that can limit the actions undertaken by the executive authority and can ensure that the executive plays by the rules of elite bargains (Di John and Putzel, 2009: 16; Tsebelis, 2002). These ‘third party enforcers’ may be brought into the political settlement at different times for different reasons. Voters too can keep parties in power but the balance of power has not completely shifted to citizens in states transitioning from monarchy to democracy (Khan, 2010: 62).A new Ministry of Cooperatives and Poverty Alleviation was formed in 2012 and has taken over the Poverty Alleviation Fund’s (PAF) poverty identification card process. ⁴⁴ PAF has support from the World Bank and is trialling new technology to better target the poor. A new Ministry of Cooperatives and Poverty Alleviation was formed in 2012 and has taken over the PAF’s poverty identification card process. The Maoist leader Dr Bhattarai made the new ministry’s announcement under a weak consensus government largely designed to please the National Cooperative Federation of Nepal (NCF Nepal, 2012). This is significant because the Maoists’ manifesto and party representatives support caste and ethnic targeting – not poverty targeting. It shows a backflip on their ideology in order to maintain power.

In competitive clientelistic settlements, weak governments can still drive reform even when the political settlement is transitioning. In the absence of a sufficient formal budget (or the ability to access it), political elites need to generate and distribute sufficient off-budget rents and benefits to keep or bring powerful actors and groups, like the Nepalese Cooperative movement, into the political settlement (Khan, 2010; Lavers and Hickey, 2015: 6). The Maoists may not have been able to access the budget at the time but could announce new ministries. New ministries that control new technologies will have sufficient budgets from which to generate rents. This is how the Maoists can still ‘buy’ support when they are losing power, even without access to the state’s budget.

The UML opposed the move at the time as political manipulation (Badal, 2012). The decision to create a new ministry that manages the implementation of new technologies for targeting was more about maintaining factions and power than Maoist ideology for reaching the poorest. This example may typify most political announcements and not just those that concern welfare or targeting. Nonetheless, it shows that other actors aside from political ones may influence the will for, and ability to achieve, social protection reform (Lavers and Hickey, 2015: 6; Pierson 1993). It also suggests that third party actors are important to the Maoists’ support base. Moreover, it highlights the role of social protection in Nepal’s democratisation and the limitations imposed by a clientelist structure. Despite a new regime and an inclusive settlement, party-based patronage politics continues to dominate the political landscape.

The Maoists need non-elite actors to make bargains with because it is very costly to reach people directly in a weak state. Moreover, the Maoists need a broad support base to beat the traditional parties that elites support. Thus, the Maoists utilise third party actors, whereas the NC and UML have more traditional powerful families supporting them. These wealthy families would include some Indian elites who are accustomed to dominating the intermediate class that administers the state. While the NC and UML would equally transfer rents to their support base, the support base is narrower and already owns considerable amounts of property and is more likely to need favours rather than money.

Some of these wealthy families temporarily supported the Maoists and Madheshis to bring an end to the conflict. After the conflict, Nepal was working towards an ‘egalitarian welfare social state’. ⁴⁵ Political parties were in alignment on the need for change (see also Kohler et al., 2009). Rights were becoming more important as citizens were raising their voices, in line with their expectations for naya Nepal. Charity was not enough, and people were demanding jobs, security and protection. However, as the social inclusion state-building project was implemented some of these powerful families realised what was at stake in creating an inclusive state. The cost of this change became higher than the risk of more conflict (Drucza, 2016). They began to support the NC once again and this reduced the power of the Maoist and Madheshi parties. The political settlement has changed, as have social protection trajectories and priorities.

Chopra (2014: 207) argues that populist gains are considered the most important motivator for social protection across South Asia and that the feedback effect in a multi-party democracy has motivated political parties to develop a welfare agenda. Certainly, democratic consolidation has contributed to Nepal’s social protection approach, however the evidence presented here suggests that populist gains are only one part of what influences social protection in Nepal (because citizens’ votes do not count as much as those of the elites in a horizontal distribution). The UCPN–Maoists’ rise to power depended on factions of dispersed citizens predominantly from excluded groups. In such a vertical distribution of power, populist gains and the quest for gaining legitimacy align (but in a horizontal distribution of power this may not necessarily be so). The distribution of power will determine how necessary cash transfers and other forms of social protection are.

The reason the NC support social protection is because they learnt from losing power to UML in 1994 that it was popular with poorer social segments. While they can rely on the rich to manoeuvre covert deals to secure power for them, the 2008 election, like the 1994 election, taught the NC not to only rely on the rich. Citizens’ votes are becoming increasingly important as Nepal’s democracy consolidates. Cash transfers use the government distribution system, and are legal, widespread and official. Other forms of gaining legitimacy in Nepal such as rents and favours are covert and outside the law. Citizens see little benefit in a weak state from a horizontal distribution of power. The NC have a horizontal distribution of power but they are being forced to consider more vertical players. Social protection is not just about what is popular with constituents but is also about the distribution of power and how political leaders learn to play the democratic game.

Lavers and Hickey’s (2015: 12) prediction holds true in Nepal: social protection can be considered ‘a political strategy for maintaining regime stability and legitimacy, rather than as a means of achieving development per se’. However, these motivations – development, stability and legitimacy – may not necessarily be in conflict, nor have equal weighting. Moreover, what ‘legitimacy’ in a democratic sense even means to parties with a horizontal distribution of power is unclear given that citizens play a less significant role in the regime’s maintenance. The evidence suggests the parties need citizens’ votes to different degrees, depending on whether they have a vertical or horizontal distribution of power, and that this will affect social protection. More research is needed to understand the relative weight or significance of these motivations for different power distributions.

Conclusion

While the literature on social protection policies emphasises state capacity, ideology and economic growth for understanding under what conditions social protectionist measures may be put in place, this article argues that the political settlement and the distribution of power can be more significant in determining policy outcomes. More specifically, whether the power distribution is horizontal or vertical is a key determinant of the shape social protection takes in post-conflict, transitioning contexts.

Transitioning, post-conflict states have complex operating environments and it can be difficult to understand how regimes function during periods of rapid change. Thus, only applying regime theory would not have given such a complete picture of Nepal’s state – understanding how regimes function is necessary for understanding the political settlement. Nepal is transitioning from a pre-capitalist to a capitalist society and is also transitioning political systems from a monarchy to a federal democratic republic. During transitioning periods in competitive clientelist settlements, former power holders ‘can only capture benefits and sustain their power by using informal organisations and/or informal rules such as patron-client relationships to distribute benefits’ (Khan, 2010: 64).

States like Nepal that are historically exclusive, weak and fragmented use cash transfers to reach citizens to meet their needs, buy their votes or demonstrate that they are included in the state and are important. Cash transfers are easily approved as they do not need policies or wider approval, just access to the budget. The budget is more readily accessible in the first couple of years of a new political party taking power. This is when the political settlement is most stable because rents are being allocated to reward key supporters. People wait to see the net benefit the regime offers them before changing their loyalty. Policy can also be passed more easily during this time (if the political party in power supports it). As time passes, there will be fewer rents in the state coffers and support will wane unless creative ways of distributing rents are found (such as creating new ministries). The Maoists had some honest intentions to reform social protection, along with ulterior motives for doing so.

The review of party platforms and interviews with party members confirms Khan’s (2010) theory that a vertical distribution of power will work better for the long-term public interest than a horizontal distribution because the self-interest of those with power aligns with what the public want, at least in the initial phase after gaining power. Distributing ‘rents’ (which in this case can be interpreted as cash transfers) can be a way to gain the support of lower-level factions. Vertical power distributions are more inclusive as they involve many more actors. Yet, maintaining support to a variety of actors is difficult in an historically exclusive state founded upon horizontal distributions of power. The analysis of third sector organisations reveals how these organisations can become more important than citizens in holding onto power in a clientelist settlement. Further research should be done on how to maintain the primary importance of citizen votes given democratic ambitions and the need for inclusion in Nepal.

The way political parties use and understand social protection can change dramatically during the course of their time in power. The clientelist settlement necessitates that social protection must, at some point, be used to augment power. Social protection is more than the mix of social policy measures carried out by the state: it refers to a particular political settlement in society as a whole (Davis, 2001: 81–82). It is about how power is shared and held and how much citizens matter. Overtime social protection should become more about redistribution to the poor and society’s protection. Every time a new party gains power, Nepal’s welfare regime and the rules of the game shift slightly.

Although Kohler and Khatiwada (2014) suggest that Nepal’s welfare system is nascent but is becoming something more substantial in terms of acknowledging universal rights (and this was true under the Maoists), Nepal’s welfare regime has reverted to its former post-conflict exclusive ways. Thus, critical junctures for social protection reform in Nepal include democracy and emerging communist political parties gaining power and seeking to augment their power through third-party actors and citizen votes. Di John and Putzel (2009: 13) rightfully predict that regime changes and transitions present ‘windows of opportunity’ to institute fiscal and welfare reforms. For development partners, this window should be capitalised upon because this same window is often used to repay those who supported their rise to power and it will not last long in clientelist states. Donor funding for large cash transfer programs at such a critical junction could have a big impact on building inclusive states. Once citizens have experience with rights and receiving state-led benefits such as social protection, they are more likely to demand/expect the same treatment in the future.

Currently in Nepal, the ability of social protection to achieve the various goals listed by political parties (consolidate democracy; protect and include poor and excluded citizens; bring peace and social cohesion) is constrained or enabled by the political settlement, the power distribution and by Nepal’s development phase. Social protection coverage is low and shallow because the political settlement still largely involves the elite. Elites have few incentives to see social protection become more substantial. Social protection is more important to Nepal’s excluded and marginalised groups and political parties with a vertical distribution of power. It can be used as a political tool and different parties may see it this way, but reducing it to a purely instrumental function does not do it justice (Hall, 2008). Social protection exists in Nepal because it offers an alternative to the traditional form of exclusive governance and this is popular for parties with a vertical distribution of power.

The main agents of the distribution of political power are party leaders and civil society/third sector organisations in Nepal. Consequently, aid agencies must engage these actors, and understand the distribution of political power, if they are to effectively promote social protection in transitioning polities. This article has developed a Nepal-specific understanding of the politics of social protection that may be helpful to other post-conflict countries, along with those seeking to ‘work politically’. The research has shown the value in understanding the local politics in which social protection programs are embedded. Moreover, the data shows potential entry points for inclusive state building – even if the pace of change is slow, at least an inclusive state is being built.