Representative and responsive bureaucracy in Nepal: a mismatch or a realistic assumption?

Abstract

Purpose

The purpose of this paper is to investigate, firstly, to what extent has Nepal’s bureaucracy become representative in terms of reflecting the country’s demographic composition, and secondly, has the bureaucracy become more responsive to citizens since the implementation of a quota policy in 2007.

Design/methodology/approach

This paper relies on factual and perceptual data in analysis. In order to analyze and interpret representative bureaucracy, this paper adopts factual data derived from the secondary sources, especially data generated by the Government of Nepal. Second, the perceptual set of data was collected through two rounds (2008, 2014) of a country-representative survey in Nepal.

Findings

The findings suggest that in terms of representativeness, the bureaucracy is still dominated by high-caste Hindus, while other ethnic communities, except the Newars, are utterly under-represented. Surprisingly, Dalits are represented in higher posts as per their percentage in the population, but they are still underrepresented in the civil service in general. Women’s representation has also increased through participation in the civil service, but they still mostly hold junior or non-gazetted posts. Citizens’ evaluations regarding responsiveness and processes of service provision are also mixed.

Originality/value

This paper is a unique attempt to understand the aspects of representativeness and responsiveness in relation to Nepalese Civil Service.

Citation

Jamil, I. and Baniamin, H.M. (2020), "Representative and responsive bureaucracy in Nepal: a mismatch or a realistic assumption?", Public Administration and Policy: An Asia-Pacific Journal, Vol. 23 No. 2, pp. 141-156. https://doi.org/10.1108/PAP-03-2020-0016

Introduction

‘Representative bureaucracy’ has a positive impact on public administration due to its fostering of social equity and inclusion in policy performance. As a concept, it reflects the extent to which a nation’s bureaucracy represents different segments of society (Kinsgley, 1944; Meier, 2019). The absence of representativeness can make a country’s bureaucracy elitist and compromise neutral decision-making (Meier and Capers, 2014). A study by Jamil and Dangal (2009) shows that afno manchhe (favor towards one’s own people) in the Nepalese Civil Service (NCS) discriminates between people on the basis of family, kinship, caste, and social relations. To address this problem and to ensure that people from diverse sections of the country are represented in the NCS, Nepal introduced an affirmative action or quota policy in 2007. The ‘People’s Movement of 2006’ brought down the monarchy. One of its demands was to introduce inclusive and representative governance. Following this, the newly-formed government introduced an affirmative action policy. The second amendment of the “Civil Service Act, 2007” introduced 45 percent reservations for disadvantaged groups (Dong, 2016; Paudel, 2013). The affirmative policy has thus been functional for over a decade but remains controversial. This paper intends to analyze two aspects of this affirmative policy: the extent of representativeness and the degree of responsiveness in the NCS.

Another much-debated issue is whether, after the Civil Service Act of 2007, Nepal’s bureaucracy has become more responsive in terms of service provision and in the processes involved. In the quest for a modern, effective, and professional bureaucracy, responsiveness in combination with representativeness has become crucial, even more so in the context of democracy and social inclusiveness. In Nepal, demands for an inclusive government have been high, especially when the country experienced a decade-long Maoist movement (1996–2005) that resulted in the termination of monarchism and the establishment of democracy. The question to be addressed in this paper is: to what extent has Nepal’s bureaucracy become more representative and responsive since 2007? The paper draws from facts and figures related to the previous unitary system of government.

Representative and responsive bureaucracy

According to the theory of representative bureaucracy, increased representation should ensure social equity and inclusiveness in accordance with the demographic features of a country. The argument is that when bureaucracy reflects a country’s demographic composition, there is a likelihood that citizens will identify more with bureaucrats. This may, in turn, enhance citizens’ trust in bureaucracy and their inclination to cooperate with bureaucratic initiatives (Atkins and Wilkins, 2008, cited in Meier, 2019). Representative bureaucracy can also promote fairness in public policy making and implementation and therefore enhance public organizations’ legitimacy and acceptance in society.

While the concept of representative bureaucracy was initially understood in terms of power-sharing, it has now expanded to incorporate inclusiveness and responsiveness in line with democratic governance and New Public Management reform initiatives (Andrews et al., 2016) to work for the betterment of society. Representativeness initially meant racial and ethnic inclusion, but it has now expanded to include women as a response to concerns for legitimacy.

Research on representative bureaucracy mainly focuses on passive and active representation. Passive representation points to a bureaucracy’s similarity to the demographic composition of a country, while active representation indicates one group (the civil servants) representing the interests of another group (the citizens) (Mosher, 1968). According to Pitkin (1967, cited in Bishu and Kennedy, 2019), passive representation is about “being something rather than doing something.” Here, doing something means acting on behalf of the people they are representing, in other words, being responsive to the need of those people. However, a representative bureaucracy may not always be responsive to those it represents, and an unrepresentative bureaucracy can be relatively more responsive; it all depends on the structural mechanisms for ensuring discipline and control, and the results-orientation of bureaucrats, leading to more socialization and the internalization of a public service ethos. All these factors play a role in making members of a bureaucracy think, feel, and act on behalf of citizens, regardless of their religion, caste, and so on.

An appropriate scenario is to have a representative bureaucracy which is also responsive to the needs of the people. An absence of this may make people feel excluded. Both representativeness and responsiveness are fundamental tenets of democracy. Representativeness ensures inclusion and entitlement of different groups in the administration of the country, while responsiveness makes different groups active participants in the process of governance, such as in policy formulation and implementation.

Representative bureaucracy was introduced in Nepal in 2007, and it has aspired to include an assortment of groups. In this paper, the analysis is two-fold. First, the authors look at the extent to which Nepal’s bureaucracy has become representative in terms of reflecting the country’s demographic composition. Second, whether the bureaucracy is responsive to citizens is analyzed. Responsiveness is analyzed based on policy performance, that is, the extent of citizens’ satisfaction with the services that are provided, and the extent to which citizens perceive the processes of service provision as either problematic or pleasant. This paper seeks to discover the level of representativeness and responsiveness in Nepal’s bureaucracy, as outlined in Figure 1.

Figure 1 presents four scenarios that combine degrees of representative and responsive bureaucracy. In a democratic context, situation 4 is ideal: bureaucracy is both highly representative and highly responsive. Scandinavian bureaucracies may reflect such a scenario, given that women are well-represented in civil service. Scandinavian bureaucracies also attract a high degree of trust, which reflects their responsiveness and, thus, also legitimacy in society (Rothstein and Stolle, 2008). Situation 1 is the least desirable. Examples of it may be found in public administrations that operate on the ‘spoils’ or patronage system, or where civil servants are loyal to the kings and the nobility rather than to ordinary citizens. Situation 3 represents an elitist but responsive bureaucracy, as represented by the civil service in certain East Asian countries (Japan, Korea, Singapore), where most civil servants have graduated from top universities and display a high level of policy responsiveness and efficiency. This type of bureaucracy may be mostly responsible for the rapid economic growth that has transformed these formerly poor countries into affluent and developed ones (Evans, 2012). Situation 2 may be exemplified by several South Asian countries where, even though affirmative action policies have been introduced, the bureaucracies have remained elitist in their attitude and behavior, thus reflecting a legacy of the colonial past. This observation on South Asian bureaucracies is further assessed based on the Nepalese bureaucracy (Jamil et al., 2013).

Research methodology

This paper uses two types of data: factual and perceptual. First, in order to analyze and interpret representative bureaucracy, the authors draw on factual data derived from secondary sources, especially data produced by the Government of Nepal. These data were collected and updated from various studies, relevant laws, policy papers, and official reports. Second, the perceptual data were obtained from two rounds (in 2008 and 2014) of a country-representative survey in Nepal known as the ‘Governance and Trust Survey’. Data from this survey has already undergirded published research such as Jamil (2019) and Baniamin et al. (2020). The first round of the survey was conducted right after the adoption of the quota policy, and the second was conducted six years later. Comparative analysis of these data will, therefore, reveal longitudinal change over time. The details of the survey’s methodology can be found in the book chapter by Jamil (2019).

The perceptual data fall into two categories: a) policy performance and b) processes of service delivery. Under policy performance, the data highlight key policy areas such as health, education, energy supply, agriculture, law and order, and utility services. These indicators are measured on a 5-points scale, where 1 represents ‘Very bad’ and 5 ‘Very good’. Higher values on these indicators signify higher perceived policy performance.

Along with policy performance, various procedural aspects of service delivery are also measured. For this variable, service-enabling factors are measured through three indicators: promptness and efficiency, friendliness, and degree of equal treatment. Besides, the study also measured service-impeding factors through three indicators: corruption, public officials serving their own interests, and difficulty in getting access to public officials. The processes of service delivery are measured on a 4-point scale, where 1 represents “Strongly agree” and 4 “Strongly disagree”.

Nepal and the NCS

Nepal was never colonized – unlike its South-Asian neighbors. Its bureaucracy, however, is heavily influenced by its neighbors, particularly India. In 1951, the Butch Commission was headed by an Indian administrative expert, N.M. Butch, and his recommendations helped build the structure of Nepal’s current civil service (Dhakal, 2013). In 1956, Prime Minister Taka Prasad Acharya promulgated the Civil Service Act with the help of the Indian lawyer, G. Murdeshowr. This act stands as the cornerstone of Nepal’s current civil service, the modern NCS. Murdeshowr also formulated other regulations such as the Nepal Administrative Service Rules (1956) and the Nepal Administration Level Classification and Recruitment Rules (1956) (Shrestha and Paudel, 2019).

In Nepal, caste or Jat plays a decisive role in a person’s status in the social hierarchy and in choosing a profession (Jamil, 2019). Bhattachan et al. (2003) see the Hindu caste system and feudal rulers’ discretionary laws as the main elements undergirding exclusion in today’s Nepali society and the NCS. For example, the Muluki Ain or Country Code of 1854 brought all Nepalese people under a single legal system and created a hierarchy based on Hindu social stratification. Non-Hindus were also treated as subject to this hierarchical system (Dhakal, 2013). Bahnus (Brahmin), Chetri, and Newar are still today identified as ’pure’, placed at the top of the hierarchy, and considered a higher class. Muslims and foreigners are considered ‘water-unacceptable’ (i.e., pure caste people cannot accept water from them), and Dalits are considered ‘untouchable,’ both groups stigmatized low and impure (Bennet et al., 2006) (Figure 2).

One of the key events creating social exclusion in Nepalese society happened in the 14^th century, when King Jayashiti Malla introduced 64 vertical occupational caste groups among the Newar community. This went against Newar traditions and customs because they did not belong to any caste (Bhattachan, 2009, cited in Dong, 2016). Inspired by this caste system, King Prithivi Narayan Shah, when establishing the Kingdom of Nepal in the mid-18th century, adopted a policy whereby only certain castes would receive administrative employment: the Panta, Bohora, Khanal, Aryal, Pandey, and Rana. The first four are Bahunss, and the last two are Chhetris. This effectively excluded Janajatis, Madhesis, and Dalits from public employment (Bishwakarma, 2008, cited in Dong, 2016). Social exclusion was further nurtured during the Rana rule (1846–1951), when Nepal’s first National Legal Code of 1854 was adopted. This legal code ended up confining people in a single Hindu legal caste system (Dong, 2016).

Along with the caste system, languages were treated as grounds for social exclusion. In 1920, the government declared that in cases where languages other than Nepali were used for indenture/written agreements, such agreements could not be submitted as evidence in a court. The Nepali language became the only official language and the only one used in the education sector. It thus became a powerful instrument for subordinating diverse groups of people under Hindu norms, rules, and values (Hachhethu, 2009, cited in Dong, 2016). Still today, the exclusionary system affects Nepal’s society. Report on Gender and Social Exclusion Assessment (GSEA) of 2006 (Bennett et al., 2006) identifies six dimensions of social exclusion in Nepal. According to this report, Tagadhari (denotes high caste and religious status in Hindu religion (Riaz, 2010) such as Brahmin and Chettris who still comprise the dominant class in Nepal, while Dalit is the subordinate class. As Table 2 shows, women, no matter what their class, are more marginalized than men.

Table 3 indicates the dominance of Bahuns and Chhetris in the NCS. Although Bahuns comprise only 12.74 percent of the population, they occupy 72 percent of posts in the NCS; similarly, although Chhetris only make up 15.80 percent of the population, they have 15.89 percent of NCS jobs. These statistics also extend to upper-level (special and gazetted category) NCS jobs.

Structure of the NCS

The current NCS is divided into three categories: gazetted, non-gazetted and classless (Dong, 2016). The word ‘classless’ itself may indicate that people who belong to this category are treated in an indifferent manner. Gazetted jobs, which are higher-level jobs and enjoy higher prestige, are divided into four subgroups: special, gazetted first, gazetted second, and gazetted third. Non-gazetted jobs also fall into four categories from gazetted first to gazetted fourth (Dong, 2016).

The NCS has 13 categories of services, for instance, General administration, Agriculture, Auditing, Education, and Health services. Nearly half of the country’s civil servants work in the General administration and about one-fifth work in the Health services (Paudel, 2018).

Quota system in Nepal

To ensure the participation and representation of various marginalized groups in the civil service, Nepal introduced a quota policy in 2007. In the budget speech of 1995, affirmative action was introduced, and 16 social groups were categorized as oppressed (Dong, 2016). In 2003, the then-prime minister Surya Bahadur Thapa planned to reserve 35 percent of civil service posts: women would constitute 20 percent, Dalits 10 percent, and Janajatis 5 percent. Thapa also formed a committee to implement the reform, but due to political and governmental instability, it was unsuccessful (Dong, 2016). In 2006, after the fall of the monarchy, the newly formed government introduced affirmative action. The Interim Constitution of Nepal (2007) created various inclusive policies to ensure the participation of different groups in national life. Article 21 of this constitution states that women, Dalits, Adivasi Janajatis, the Madheshi community, downtrodden classes, poor farmers, and workers who are underprivileged shall have the right to participate in the state structure based on the principle of proportionate representation (Paudel, 2013). Later that year, through the second amendment to the “Civil Service Act, 2007”, the government introduced 45 percent reservations for disadvantaged people (Dong, 2016; Paudel, 2013). An overview of the Nepalese quota system is given in Table 6:

Discussions on quota/representative policy in Nepal

Women’s share in the public sector

With the introduction of a quota system in Nepal, the percentage of women in the public sector has gradually increased. Overall, however, it has not yet hit the expected 50 percent mark. Figure 3 shows that in 2003, before the introduction of the quota policy, women’s participation was around eight percent and increasing to 15 percent by 2010. By 2015, it further increased to 18 percent.

Table 7 indicates that women’s increased participation in the NCS is mostly at the ‘non-gazetted’ level. One fifth (21%) of female civil servants belong to this category. The presence of women in higher-level positions is still paltry, even after the introduction of the quota system. Table 7 also shows that in 2003, the percentage of women’s participation was 2 percent for ‘Special Class’, and one percent for ‘Gazetted I’ class; ‘Special Class’ participation only increased to three percent by 2015. In the ‘Gazetted I’ class, this has increased from one percent in 2003 to seven percent by 2015. At the Gazetted III level, the percentage of women doubled, that is, from six to 12 percent. In the long run, these women are likely to be promoted to higher-level positions, thereby forming a critical mass.

Table 8 shows that the percentage of women who applied for public jobs and passed the examination is increasing. In the year 2007–08, there were 39 percent female applicants; by 2012–13, it increased to 47 percent. However, there is a considerable mismatch between female applicants and the percentage that passed the civil service examinations. Although the percentage of women who passed the exams has increased from nine percent in 2007–08 to 12 percent in 2012–13, this is remarkably disproportionate to the percentages of both female and male applicants. Paudel (2018) attributes the increase in female applicants to the inclusion policy, which encourages women to apply for civil service employment. However, since few women passed the civil service exams, this may indicate that they still face a range of challenges. According to the Public Service Commission (PSC), around 250 posts could not be filled due to the lack of suitable candidates (Paudel, 2018).

Geographically-based quota

In Nepal, there are nine so-called “backward” districts for which some civil service positions are reserved. According to the Civil Service Act of Nepal, these are the Accham, Kalikot, Jajarkot, Jumla, Dolpa, Bajhang, Bajura, Mugu, and Humla districts. This may be a more effective strategy to target and include people from backward areas than the population-based geographical quota that used to be followed in Bangladesh. Nevertheless, the Nepalese quota policy is yet to achieve the intended representation from different areas. Table 9 shows that most of the civil servants come from the Eastern, Central, and Western regions while there is less representation from the Mid-Western and Far-Western regions. Further, if looking at the population size of these regions, it was observed that the Mid-Western and Far-Eastern regions are less represented in the NCS, even though all the backward districts fall inside these two regions.

Attribute/Ethnicity-based quota

Interestingly, Table 10 indicates that since the introduction of the quota system, the representation of backward caste/ethnic communities at the higher official level has declined. For example, in the year 2006, there was 17.49 percent Janjati at the official level, but this decreased to 12.28 percent in 2012. During the same period, it was observed the increase of the category ‘Others’, which includes high-caste Hindus and Newars, in official positions. Some researchers have explained the discrepancy as being due to the downsizing of bureaucracy and the excessive time the recruitment process took (Paudel, 2013). Shrestha and Paudel (2019) blame the quota system’s ineffectiveness on the lack of human resource development activities.

Given that top-level bureaucrats (e.g., secretary, joint secretary) are important for policy formulation, implantation, and the allocation of resources, it becomes clear that the presence of different groups in these positions is a key to ensuring a representative and inclusive bureaucracy that may contribute to a balanced development of the country. Table 11 indicates that Brahmin and Chhetri still dominate these higher positions, and least represented by ‘indigenous nationalities’ and ‘others’ (including Muslim).

Some key concerns regarding the Nepalese representative policy

In terms of representation, Bahuns and Chhetris still dominate the NCS. This indicates that Afno Manche – that is, the cultural practice of discriminating against or favoring someone based on family, relatives, caste, and social relations – is still dominant (Dong, 2016; Riaz and Basu, 2010). Based on the current quota system, people generally perceive that 55 percent of jobs are reserved for Bahun/Chhetri and that the remaining 45 percent is reserved for others (Bishwakarma, 2008, cited in Dong, 2016). It is probably due to these cultural factors and the dominance of the higher caste that people from other ethnicities are still struggling to join the civil service. The quota policy does not address other forms of structural discrimination. For example, many marginalized people’s education is affected by poverty and language barriers, since the Nepali language is the main medium of study in educational institutions (Dong, 2016). Bennett et al. (2006) states that 52 percent of the population do not speak Nepali as their mother tongue. This makes them disadvantaged in terms of access to education. Moreover, civil service exams (both written and oral) are conducted in Nepali. This report also indicates that the major cause of school dropout among the Janajati and Tarai people is that the curriculum is non-representative of Nepal’s diverse ethnolinguistic population. Language, therefore, directly or indirectly affects many ethnolinguistic groups when they seek access to public services in Nepal (Dong, 2016; Lawoti, 2014). Due to this, people from marginalized groups struggle to succeed in public service examinations.

The recruitment procedure takes about 18 months (Paudel, 2018), so it is difficult for women and backward ethnic groups to wait for public service examinations. Furthermore, many women get married and never return to their job in the public sector. Similarly, marginalized ethnic groups get involved in other economic activities because they cannot afford to hold out with such a lengthy recruitment process. For these groups, it is also challenging to apply for the examination. However, if it became possible to apply online, this might ease the problem.

Due to the caste system, different groups face different challenges that may limit their possibilities to be promoted in the civil service; Dalits are seen as untouchable, while Buddhist Janjatis, due to their different religious beliefs, may receive less respect than Hindus and be discriminated against when they are up for promotion. Women may fail to be nominated for promotion due to existing patriarchal norms and values (Dong, 2016).

The objective of the ‘lateral entry system’ (horizontal recruitment) was to promote efficiency in the civil service. However, as of today, the provision is used mostly by those who are already in service. The provision of lateral entry fails to attract the best candidates from outside because the tenure of past service is not taken into consideration, and past-service tenure is essential when seeking promotion to higher-level positions and for retirement benefits (Shrestha and Paudel, 2019).

Jamil (2019) and Limbu (2009) identify several of the challenges some groups face when trying to join the civil service: the internal colonization of Hindu based-culture and language, the curriculum, the lack of political access, the centralized examination system, policy implementers’ ignorance regarding the inclusive policy, some ethnic groups’ lack of required education, insufficient interest in the civil service, and more interest and motivation among some ethnic groups to join foreign military services (i.e., the British Army and the Indian Army). The groups with a history of army recruitment are the Gurung, Magar, Rai and Limbu (Dong, 2016). Like any quota policy, the Nepalese policy is criticized for undermining the merit-based recruitment system. It is also criticized for promoting elite applicants from disadvantaged groups and for creating a dominant class within different ethnic communities (Dong, 2016).

The status of responsive bureaucracy in Nepal

After analyzing the representativeness of bureaucracy in Nepal, attention is focused on the bureaucracy’s responsiveness to citizens. Responsiveness is measured based on citizens’ perceptions of two groups of variables: a) processes of service delivery; and b) policy performance.

First, citizens’ trust in the NCS for two time periods was analyzed. Table 12 indicates that trust in the NCS in general increased from 2008 to 2015, as M = 2.56 for 2008, and M = 2.77 for 2015. Such differences are statistically significant, as the p < .001. This improvement is also visible for trust in the Chief District Officer (CDO). It cannot be claimed that such an increase in trust is due to the introduction of the quota policy, but since the first round of the survey was conducted immediately after the introduction of quota policy, we may get an idea of how the citizens tended to perceive the NCS at that time, and then how that perception changed or remained stable during the second round of the survey in 2015.

Processes of service delivery

In the second step, the authors analyze these two groups of variables in two time periods to see variations in citizens’ perceptions. Table 13 presents citizens’ perceptions of the processes of service provision, in other words, of their interactions with the bureaucracy. It was observed that there has been an improvement from 2008 to 2015 in the indicators for ‘Service enabling’ processes. Even so, such ratings are relatively low, as the mean scores in both periods are below 2.5. On a four-point scale, any score below 2.5 is considered below average. In this regard, it was observed that in both periods, the scores on these variables are below 2.5, indicating that citizens evaluated the service-enabling process as below average, despite slight improvement by 2015.

In regard to ‘Service impeding’ processes, it was observed that citizens’ perceptions on whether bureaucrats are ‘corrupt,’ ‘serve their personal interests,’ and whether they are ‘easily accessible’ are somewhat negative. The scores on these variables are above 2.5 (on a scale from 1 to 4); in other words, they are above average, indicating that the bureaucrats are corrupt, unfriendly and not easily accessible. This means that citizens negatively assess service provisions.

Policy performance

In the third step, citizens’ evaluations of several public services in the two time periods were mapped in order to see whether these evaluations have increased over time; in other words, whether the introduction of representative bureaucracy in Nepal has made bureaucrats more service-oriented as assessed by citizens. The findings in Table 14 are interpreted in the following manner: on a five-point scale, a mean score of 3 and above for a service indicates a positive evaluation by citizens; a score below 3 for a particular service is considered a negative evaluation by citizens. In this regard, it was observed that education, health, and agricultural services are positively evaluated in both periods, but that the overall perceived performance in these areas decreased between the two periods. The services that scored below 3 in both time periods are energy and utility services (water supply, garbage collection, street lighting, etc.). The maintenance of law and order received a more positive evaluation in 2015 than in 2008. The mean differences between these time periods indicate that there have been significant changes in citizens’ evaluations of services, either positive or negative.

These data may not give a conclusive picture of the role of representative bureaucracy in policy performance, or even the actual effects of representative bureaucracy on service delivery processes, but we may get an idea of how these aspects changed after the introduction of the quota system. Ideally and theoretically, more representativeness would be assumed to have a positive effect on policy performance and service delivery processes. If different groups are properly represented, then they should receive due justice and share in policy formulation and implementation. However, from the surveyed data, it appears that in most cases, policy performance still needs to improve. Without improved policy performance, representative bureaucracy may not be meaningful and citizens’ trust may evaporate. Public service provision also needs improvement. If civil servants remain corrupt, are preoccupied with personal concerns, or if they are difficult to get hold of, then increased trust in the civil service will be short-lived.

Conclusion

The hype around the concept of representative bureaucracy probably contributes to increased trust in the NCS, but such increased trust may not last long if the it fails to respond to people’s needs. To ensure a higher level of trust, the NCS needs to be genuinely representative, perform well in different policy areas, and provide services to the people in fair and efficient ways.

From this study, it is apparent that fine tuning of the Nepalese bureaucracy will ensure the representation of different ethnic groups, religions, regions, and genders in all the services and at all levels within it. To warrant this, it needs to address the problems in the recruitment and promotion processes. Some of these can be addressed more easily than others; for example, it may be reasonably easy to set up an online portal for people to submit their applications, or to reduce the length of the examinations. Other problems, however, are deeply rooted in society. One of these is Afno Manche, which favors or discriminates against people based on their family, relatives, caste, and social relationships. As long as this system is in place, it will continue to favor those who already wield power within the system. The caste system further complicates these problems and needs to be addressed as part of more substantial social reforms.

Apart from these problems, the other issue is that the bureaucracy needs to perform well and to respond to citizens’ expectations. If the civil service remains unresponsive to the needs of the people, then the ultimate objective of representative bureaucracy will remain unattained. It, therefore, appears that the mere introduction of representative bureaucracy is insufficient; it needs to be reflected through policy performance and the way public services are provided.

Figures

Figure 1

Representative and responsive bureaucracy

Figure 2

The Nepalese caste pyramid according to the Muluki Ain of 1854.Source: Bennett et al. (2006)

Figure 3

Percentage of women in the public sector of Nepal. Sources: Adapted from Paudel (2018); Paudel (2013)

Table 1

Quick overview of the Governance and Trust Survey in Nepal

Waves	1st Wave	2nd Wave
Year of survey	2008	2014
Sample size	1,836	2,404
Selection of sample	By using ecological and five development regions, 17 districts were selected out of 75 districts. From these districts, households were selected from municipalities and village development committees (VDC). For randomization and representative sample, every fifth household was selected from each area.	Used voter list; (20% of constituencies, i.e. 48 constituencies were selected to conduct survey on the respondents; one polling booth was randomly selected and from there, 50 voters were selected from each polling booth)

Source: NORHED Survey Documents (unpublished), cited in Baniamin (2019)

Table 2

Dimensions of Exclusion in Nepal

Social status	Gender	Caste	Ethnicity/Race	Language	Religion	Geo-Political
Dominant	Men/Boys	Tagadhari: Bramhin, Chettris	Caucasoid	Nepali	Hindu	Parbatiaya: Hill Dweller
Subordinate	Women/Girls	Dalit	Janajati/Mongoloid	Others	Non-Hindu	Madheshi: Plain Dweller

Source: Bennett et al. (2006)

Table 3

Caste/ethnic representation in special and gazette class of the NCS

No.		Caste/Ethnicity	Population	Representation in the civil service
1		Bahuns	12.74	72.00
2		Chhetris	15.80	15.89
3		Newar Janajatis	5.48	7.14
4		Non-Newar Janajatis	30.83	1.64
5		Madheshis	12.32	1.17
6		Dalits	14.99	0.67
7		Muslims	4.27	0.1
8		Others	3.57	1.39
		Total	100	100

Source: Ministry of General Administration, cited in Dong (2016)

Table 4

Composition of the Nepalese Civil Service and its recruitment policy

Types	Subgroups	Composition	Recruitment policy
Gazetted	Special class	Chief secretary, secretary, and others	The secretary and chief secretary are to be recruited by the Government of Nepal.
	Gazetted I	Joint secretary, joint zonal commissioner, director general, and others	10% are to be recruited through open competition, 35% through performance evaluation, 20% through internal competition, and the remaining 35% through seniority and performance evaluation.
	Gazetted II	Undersecretary, director, deputy director, assistant zonal commissioner, chief district officer, land administrator, land reform officer, district education officer, and others	Same as Gazetted I
	Gazetted III	Section officer, panchayat* and development officer, administrative officer of district office, and others	70% are to be recruited through open competition and the remaining 30% through seniority and performance evaluation.
Non-Gazetted	Class I-IV	Naib Subba, Kharidar, Mukhiya, Bahidar, and others	Different recruiting mechanisms such as open competition, internal competition, seniority, and performance evaluation.
Classless	-	Office assistant/messenger	Recruited on contract

Source: Shrestha and Paudel (2019)

*Note: This was during the Panchayat regime (1960–1990).

Table 5

Structure of the NCS

Categories	Percentage
General administration	46.30
Health	20.52
Engineering	9.95
Agriculture	6.16
Forestry	6.04
Others	11.03
Total	100

Source: Paudel (2018)

Table 6

Quota distribution in the NCS

Category	%
Merit	55
Quotas:	45
1. Women - 33%,
2. Adibasi Janjati - 27%,
3. Madheshi - 22%,
4. Dalit - 9%,
5. Persons with disability - 5%
6. People from the backward areas - 4%
Total	100

Source: Dong (2016)

Table 7

Women’s share in the public sector of Nepal

Rank	2003			2010			2015
	Male	Female	Female (%)	Male	Female	Female (%)	Male	Female	Female (%)
Special class	41	1	2	67	2	3	63	2	3
Gazetted I	304	4	1	416	13	3	429	29	7
Gazetted II	1399	52	4	2474	113	4	3119	163	5
Gazetted III	4909	304	6	7129	593	8	12,451	1698	12
Non-gazetted	38,596	4224	11	34,813	7632	22	32,686	8774	21
Classless	23,885	1432	6	20,366	1446	7	19,175	1518	7
Total	69,134	6017	8	65,265	9799	15	67,923	12,184	18

Source: Paudel (2018)

Table 8

Success of female applicants in the public service exams

Year	% of eligible applicants			% of applicants who passed the examination
	Male	Female	Total	Male	Female	Total
2007–08	61	39	100	91	9	100
2008–09	62	38	100	94	6	100
2009–10	53	47	100	89	11	100
2010–11	56	44	100	89	11	100
2011–12	58	42	100	89	11	100
2012–13	53	47	100	88	12	100

Source: Annual report of Nepalese Public Service Commission (PSC) (2008–2014), cited in Poudel (2018)

Table 9

Bureaucratic representation in the top level of the NCS, based on regions (Percent distribution, N =3,805)

Development regions*	Secretary (%)	Joint Secretary (%)	Undersecretary (%)	Population (%)
Eastern	18	19	23	22
Central	42	39	34	36
Western	29	36	34	19
Mid-Western	8	3	5	13
Far-Western	3	3	4	10
Total	100	100	100	100
N	65	458	3,282	26.5 million

Source: Adapted from Ministry of General Administration (2014)

Note: * The five development regions are now regrouped into seven provinces called Pradesh.

Table 10

Representation of the different ethnic groups in the NCS (higher official levels)

	2006		2012
	Numbers	%	Numbers	%
Janjati	1,416	17.49	1,593	12.28
Madhesis	805	9.94	1,068	8.23
Dalits	74	0.92	117	0.90
Others	5,801	71.65	10,193	78.59
Total	8,096	100	12,193	100

Source: Awasthi and Adhikary (2012)

Table 11

Bureaucratic representation in the top level of the NCS, based on caste/ethnicity. Percent distribution

Caste/Ethnicity	Secretary	Joint Secretary	Undersecretary	Population size
Brahmin (Bahun)	65	68	66	12.74
Chhetri	20	17	20	15.80
Indigenous nationalities	-	2	1	48.63
Dalit	15	13	13	14.99
Others (including Muslim)	-	0.4	0.4	7.84

Source: Adapted from Ministry of General Administration (2014)

Table 12

Changes in trust in the NCS (Independent t-test)

Institutional trust	Survey Year	N	Mean (1-4 scale)	Mean differences	Std. Deviation	Std. Error Mean
Civil Service	2008	1,585	2.56	.21***	.775	.019
	2015	2,151	2.77		.686	.015
Chief District Officer (CDO)	2008	1,700	2.75	.22***	.685	.017
	2015	2,272	2.97		.653	.014

***p < .001, ** p < .01 and * p < .05 (two-tailed tests)

Table 13

Citizens’ assessment of public service delivery processes (Independent t-test)

Processes of service delivery	Round	N	Mean (1-4 scale)	Mean differences	% above mean value	Std. Deviation	Std. Error Mean
Service-enabling processes (low-high)
Civil servants: prompt and Efficient	2008	1,643	2.28	.05*	38	.767	.019
	2015	2,191	2.33		42	.756	.016
Civil servants: friendly	2008	1,667	2.30	.10***	37	.774	.019
	2015	2,260	2.40		43	.774	.016
Civil servants: treat all equally	2008	1,671	1.90	.26***	61	.909	.022
	2015	2,260	2.16		80	.829	.017
Service-impeding processes (low-high)
Civil servants: corrupt	2008	1,666	2.80	.08**	66	.872	.021
	2015	2,268	2.88		69	.826	.017
Civil servants: serve personal interests	2008	1,755	3.09	-.09***	34	.844	.020
	2015	2,289	3.00		27	.789	.016
Civil servants: difficult to get access to	2008	1,712	2.80	.00	66	.864	.021
	2015	2,288	2.80		65	.814	.017

***p < .001, ** p < .01 and * p < .05 (two-tailed tests)

Table 14

Citizens’ assessments of policy performance in different areas in two time periods (Independent t-test)

Policy areas (low-high)	Round	N	Policy performance mean (1-5 scale)	Mean differences	% above mean value	Std. Deviation	Std. Error mean
Maintenance of law and order	2008	1,521	2.74	.26***	60	1.030	.026
	2015	2,128	3.00		74	.904	.020
Education index	2008	1,591	3.48	-.03	56	.75752	.01899
	2015	2,273	3.45		51	.78485	.01646
Health index	2008	1,631	3.51	-.11***	44	.77708	.01924
	2015	2,224	3.40		35	.79314	.01682
Energy index	2008	1,732	2.22	.28***	49	1.03140	.02478
	2015	2,309	2.50		55	.96365	.02005
Utility services index	2008	1,605	2.70	-.18***	43	.94626	.02362
	2015	2,269	2.52		38	.92842	.01949
Agriculture index	2008	1,517	3.25	-.24***	50	.93552	.02402
	2015	2,207	3.01		34	.91705	.01952

***p < .001, ** p < .01 and * p < .05 (two-tailed tests)