Nepal’s Economic Situation: A State of Perpetual Poverty, Stagnation, and Regional and Ethnic Disparities

Ram Acharya, Canada Forum for Nepal, Ottawa, Canada
and Prem Sangraula, America Nepal Society, Washington D.C., USA

Proceedings of Unfolding Futures: Nepalese Economy, Society, and Politics
Friday-Sunday, October 5-7, 2007, Ottawa, Canada


This paper compares Nepal’s economic performance with its neighbours India and China, and examines regional and ethnic, language, and caste (ELC) group disparities, and intra-region and intra-ELC group disparities in the economic standing and educational attainment of Nepal. The regions and ELC groups considered in the paper are the same 12 regions and groups previously identified as building blocks for a federal Nepal (Acharya, 2007). Results show that Nepal’s productivity has stagnated and that it contains alarming levels of poverty. There are wide disparities in economic standing by region and ELC group. The wealthiest region and group is approximately four times richer than its poorest counterpart. However, these disparities come mainly from the income differences between people at the highest strata. The poorest are equally poor across all regions and ELC groups, but the share of people living in poverty varies greatly by region and group. This makes poverty a national and, to some extent, a regional and group issue. Nowhere in the country is the rate of illiteracy less than 46% (in some cases it is up to 59%), indicating that basic education is also mostly an outcome of economic disparity rather than of regional and ELC group division. However, there are wide regional and group disparities at the higher levels of education attainment. Furthermore, the female illiteracy rate is at least 20% more than that of males in all regions and ELC groups. Taken together it appears that while poverty, illiteracy, and gender inequality have certainly been influenced by regional and ELC group disparities, they penetrate across all regions and groups in Nepal. One of the causes of these problems is the alarming amount of intra-region and -group income disparities, especially in the wealthier regions and groups. The current state of the Nepalese economy indicates that there is a failure of past economic and social policies. This failure warrants an immediate and new national strategy with appropriate regional and group dimensions in order to address these problems.

Nepal is one of the poorest countries in the world, and its economic growth performance in the last half a century has been disappointing. Not only does it lag behind the global, and fastest-growing, economies of its neighbours India and China, it lags behind many other countries in the world. There are actually very few developing countries that are growing at a slower pace than Nepal in this period. Equally appalling is the fact that the economic opportunities in the country are asymmetrically distributed across regions and ethnic, language and caste (ELC) groups. The wrong economic and social (e.g. education) policies of the past have locked the country into perpetual poverty and economic stagnation. In order for Nepal to be a peaceful, stable, and prosperous society it is essential for there to be economic policies that unleash the forces of growth and uplift the poor.

Although though there has been much talk of political restructuring, there is no discussion of whether different economic and social policies are needed for a federal Nepal, let alone how they should be packaged. Thus, it seems likely that most of these policies will continue. Considering the piteous state of the economy today, it is obvious that no matter what political restructuring takes place, past policies neither can maintain peace in Nepal nor unleash the forces of growth. Continuing past economic policies is a recipe for furthering the socio-economic disparities among people in different regions and ELC groups.

Nepal is currently at the crossroads political restructuring. Conducive political restructuring is necessary for economic growth and equitable opportunity, but that alone is not sufficient. There must be a set of complementary economic policies to go along with political restructuring. There economic policies should address the root causes of poverty and stagnation. Thus, we need to understand the state of the economy at the national and regional levels. Since there are large differences in the economic standing of various ELC groups, which are the vital forces behind political restructuring, we must understand this dimension as well. There are no papers on this topic as far as we know. Thus, the objective of this paper is to fill in the gaps and provide an overview of the state of stagnation, poverty, and economic disparity across regions, ethnic groups, and gender in economic possession and education attainment. Although the paper stops short of outlining the economic and social policies needed to break the poverty perpetuation and stagnation, it will provide a vital and novel analysis that will be necessary for successful economic and social policies in a federal Nepal. The specific objectives of this paper are to answer to the following questions:

1. What has been Nepal’s developmental trajectory in the last three decades‌
2. How does Nepal lag in comparison to its neighbours India and China in terms of economic growth, education attainment, and fiscal prudence‌
3. How equitable is Nepalese society‌ How deep and widespread is poverty in Nepal‌
4. How great are the regional, ethnic, and gender disparities in Nepal‌
5. What is the root cause of the current state of the nation‌ Is it due to regional, ethnic, and gender disparities, or to the poverty across the board‌
6. What will it take to turn the nation from a state of gloom to a state of bloom‌

We have used the data from the Population Census of 2001 and Nepal Living Standard Survey II of 2003/2004 to answer these questions (for survey details see the Central Bureau of Statistics, 2004). The regions and ELC groups considered in the paper are the same 12 regions and groups that are identified as building blocks for a federal Nepal (Acharya, 2007). The main findings of the paper are as follows.

Nepal’s productivity is stagnated and the level of poverty is alarming. It will take 49 years or one generation, for per capita income double in Nepal compared to only eight and half years in China and 14 years in India. Nepal’s per capita gross domestic product (GDP) is at 20% of China and 41% of India. In 2006, Nepal’s per capita GDP of US$ 234 was only 36% of India’s per capita GDP and 15% of Chinese per capita GDP. Poverty has penetrated all regions and ELC groups. About 31% of people live below the poverty line as conservatively defined at 7,696 Nepalese Rupee (NR) per person.

There are large disparities in economic plight both by region and ELC group. The wealthiest region and group is approximately four-times richer than its poorest counterparts. However, these disparities come from the income differences of people at the higher strata of the income distribution – the poor are equally poor across all regions and ELC groups. This fact makes poverty mostly a national, and to a limited extent, a regional and group issue. There are intra-region and -group income disparities, wider in richer regions and richer groups. Nowhere in the country is the rate of illiteracy less than 46% (in some cases it is up to 59%), indicating that basic education is also mostly an outcome of economic disparity rather than of regional and ELC group division. However, there are wide regional and group disparities at the higher levels of education attainment. For example, about one-quarter of highly educated people live in a region with only 7% of the nation’s population. Furthermore, the female illiteracy rate is at least 20% more than that of males in all regions and ELC groups. Taken together it appears that while poverty, illiteracy, and gender inequality have certainly been influenced by regional and ELC group disparities, they penetrate across all regions and groups in Nepal. One of the causes of these problems is the alarming amount of intra-region and -group income disparities, especially in the wealthier regions and groups. The current state of the Nepalese economy indicates that there is a failure of past economic and social policies. This failure warrants an immediate and new national strategy with appropriate regional and group dimensions in order to address these problems.

The rest of the paper is organized as follows. In the following section, we describe the data used in this paper. We then compare Nepal’s achievement with that of India and China. We will describe how the regions used for this study are formed and which ethnic, language, and cultural groups are considered for comparison. We will analyse poverty by region and group, with the educational situation of those regions and groups is provided later. Finally, we will conclude the paper by providing policy suggestions.


The paper uses macro level data compiled by the World Bank (the database called World Development Index) using different sources, data from 2001 Nepal’s Population Census and Nepal’s Living Standards Survey (NLSS II), conducted in 2003/2004. The Population survey data are used to identify the regional blocks and ethnic, language and cultural groups that we want to study as regional and group economies. Once the regions and groups have been identified, we use the individual level data from NLSS (II).[1] The survey used a two-stage stratified sampling approach: a nationally representative cross-section survey to estimate trends and levels of socio-economic indicators in the country and its different geographic regions; and the second component was a panel survey to track exact changes experienced by those previously enumerated households during last eight years. In the paper, we will use the data based on cross-section survey only. A brief discussion of how the survey was designed in provided in Appendix A.
We use NLSS II extensively because this is the only survey that provides information on consumption, incomes, assets, housing, education, health, fertility, migration, employment, and child labour by different ELS groups in Nepal. This survey can be used not only to assemble information based on the regions but also based on different ethnic, language and caste groups.

Nepal and Its Neighbours

In 2006 Nepal was an economy of 583 billions NR, which converted at the official exchange rate of NR 73 per US$, is an economy of US$ 8 billion. With purchasing power parity (PPP) conversion factor to NR vis-à-vis US$ official exchange rate of 0.182, Nepal’s GDP in 2006 was US$ 44 billion in PPP.[2] In terms of US$, Nepal’s GDP is about one-fifth of the province of Manitoba. And in terms of PPP, Nepal’s GDP is 10% higher than that of Manitoba or Saskatchewan. However, in terms of population, Nepal’s 28 million populations is 24 times the population of Manitoba and 28 times of Saskatchewan.[3]

Nepal, which is bordered with India and China, has been a lost case in terms of both economic growth and equity. For the last 40 years, Nepal’s GDP per capita increased only by 1 percent annually whereas that of India and China increased by 3% and 6.8% respectively. Taking decade as a sub-period, as the Chart 1 shows, Nepal’s average annual growth rate has been substantially lower than those of two neighbours in each decade Chart 1). For example, in the last decade (1996-2006), Nepal’s per capita income increased only by 1.4%, whereas that of India increased by 5% and that of China by 8.4%. In this rate, it will take 49 years, a generation, to double the per capita income of a Nepali, whereas for an Indian it will take 14 years and for a Chinese it will take less than nine years. By the time the per capita income of a Nepalese doubles it will increase by 4-folds for an Indian and by 6-folds for a Chinese.

As a result of growth stagnation, per capita income of a Nepalese lags far behind that of its neighbours. Nepal is one of the poorest countries in the world with per capita GDP of US$

234 (at constant 2000 US$), ranking at 163 out of 179 countries in the world, all 16 countries with less than Nepal’s GDP are sub-Saharan countries. As shown in chart 2, in 2006 Nepal’s per capita GDP of US$234 was only 36% of India’s per capita GDP of US$ 634 and 15% of Chinese per capita GDP of US$ 1,595. In terms of PPP, Nepal’s per capita GDP of US$ 1,397 is 41% of India and 20% of China. In other words, income of a Chinese is equal to that of five Nepalese and income of an Indian is equal to two and half Nepalese.
That is the story of overall growth. What about the social development in terms of education and health‌ Nepal’s record in terms of its neighbours is at least as bad as its record in economic growth. For example, in education, still more than half of Nepalese are completely illiterate compared to 40% in India and 9% in China (Chart 3).
In terms of gender equality, even though in primary and secondary education, female to male enrolment ratio in Nepal is about 90%, the situation in tertiary education is alarming. Female which slightly out number male in total number represents only 40% of tertiary education. Both India and China fair far better than Nepal in female enrolment ratio. This is just the macro national picture; we will have much more on education while we discuss regional, group and gender inequalities in education attainment in the coming section.
There has been no structural change in Nepal from agriculture to other sectors of the economy; it still remains predominantly an agrarian economy. However, India is emerging as a global service centre, whereas China is transformed into a global manufacturing powerhouse. With such emerging economies in its doorstep, a tremendous opportunity for Nepal’s prosperity is missed because of Nepal’s wrong economic and social policies.

One could argue that the fiscal structure of Nepal perpetuates poverty and discourages growth. The data on Table 1 shows the following. First, share of taxes on income, profits and capital gains in total government revenue is very low (11.4%) in Nepal compared to 35.4% in India. This means that there are basically no taxes on income and other property. Second, taxes on international trade are very high (21.9%) and that comes mainly from taxes on customs and other import duties. As most of the necessities are imported, the incidence of tax on trade falls on ordinary people. Third, one third of Nepal’s revenue is contributed by grants and other revenue (which is mainly grants as the aid to central government budget is 40%) from outside. This should have reflected at lower tax to GDP ratio, but that is not the case. This ratio of 9.7% in Nepal is comparable to 10.2% in India which does not have any aid money at all. Fourth, despite this heavy grants, the debt to GDP ratio in Nepal is not low, it is as high as in India.

Table 1. Tax structure in 2004


Nepal India Canada
Tax revenue (% of GDP) 9.7 10.2 14.2
Central government debt, total (% of GDP)
64.2 65.3 48.7
Aid as % of Central government budget 40.0 0.6 0.0
Revenue structure (in percent) 100.0 100.0 100.0
Taxes on income, profits and capital gains 11.4 35.4 52.1
Taxes on goods and services 30.1 31.4 17.6
Taxes on international trade 21.9 13.8 1.2
Other taxes 4.3 0.0 0.0
Grants and other revenue 32.2 19.0 6.2
Social contributions1 0.0 0.0 22.8
Tax structure (in percent)
Taxes on income, profits and capital gains 16.9 43.9 73.5
Customs and other import duties 31.0 17.0 1.7
Taxes on exports 1.1 0.1 0.0

1Social contributions include social security contributions by employees, employers, and self-employed individuals, and other contributions whose source cannot be determined. They also include actual or imputed contributions to social insurance schemes operated by governments.
Source: World Bank

This all boils down to the bottom line that the rich and affluent in Nepal are exempted from the tax network as landlord in rural area and as property owner and income earner in urban area. Major part of the tax is paid by indirect tax which is regressive in nature as it is collected mostly from necessities.[4]

Despite the fact the smaller countries are more trade oriented, Nepal’s trade orientation is smaller than that of China and India, and it has fallen over the years. As is shown in Chart 5, share of exports of goods and services in its total GDP has declined to 18.6% in 2006 from its peak of 23.3% in 2000. As China and India are taking increasingly larger international market, Nepal’s trade orientation has declined. It is not only been able to rapidly growing neighbouring markets, its neighbours in export markets are rather replacing it. The potential boon has turned into a challenge.
Thus Nepal’s economy is in serious structural problem in terms of economic, social and fiscal policies. The country has alarming state of poverty. According to World Bank’s data, poverty head count ratio at $1 a day (PPP) is 24% of population, whereas the ratio at $2 a day (PPP) is 69%.[5] In order to look the potential causes of Nepal’s stagnation, we look at the regional, ethnic and gender disparities in Nepal. However, before starting our analysis, we provide a rationale of grouping countries in 12 regions and one territory and making our analysis based on 12 distinct groups.

Forming the Unit of Analysis

Since the focus of the rest of the paper is to look at the region-wise and group-wise comparison of economic condition, what regions and groups to consider for unit of analysis is an important issue. So far for the regions, most of the studies have used ecological belts and/or (Himalaya, Hill and Terai), rural-urban dichotomy (see ; Pradhan and Shrestha, 2005; World Bank, 2006a; World Bank 2006b). For groups the prevailing studies use rather aggregate ethnic groups by lumping several of them into a few groups. In this paper, we follow a systematic approach where we identify the clustering of each ethnic, language and caste (ELC) groups that are more than 1% of Nepal’s population. Then combining all districts that are home of an ELC group, we define a region. The analysis is then carried out across such regions and across such ELC groups that define regions. This structure is taken from Acharya (2007), where he has proposed political restructuring by making these regions as election constituents.

According to population Census 2001, there are 23 groups, which constitute at least one percent of Nepal’s total population.[6] However there are some common threads (language, ethnicity, socio-economic conditions etc.) that some of these groups could be combined without compromising the effectiveness of the study. The results are presented in Table 2.[7] In a nutshell, the population of Nepal can be divided into the Indo-Aryan caste groups and Tibeto-Burmese ethnic groups. Caste is defined as social group within the Hindu caste-system, and ethnic or nationality (Janajati) is defined as a social group within its mother tongue, native area and religious tradition. Besides, the so-called high and medium caste, Brahamin, Chhetri, Thakuri and Sanyasi (BC, hereafter), there are 12 groups (as shown in Table 1) that are more than 1% of the population, not counting “other”.

Six of them (Limbu, Rai, Tamang, Gurung, Newar and Magar) are mountain and hill ethnic groups who have their own mother tongues.[8] The remaining group in the hill is hill “lower caste” the so-called untouchables whom we refer by “Dalit”.[9] Three other groups based on language (Maithali speaking, Bhojpuri speaking and Awadhi speaking) live in the Terai (the plain southern part of Nepal). A common name for people living in terai (these three language groups and other smaller groups whose mother tongues are different) is Madhesi, and they could be separated in three different groups along the line of caste and ethnicity (janajati): upper and medium caste, lower caste and janajati. In the Terai, there exists another group, Tharu, whose culture and language is not Nepali and is different from Madhesi people. Then there are Muslims who are mainly inhabited in the Terai.

Table 2. Nepal’s population composition by broader caste/ethnicity in 2001

Ethnic, language and caste groups Share in total population Number of districts
Majority (& plurality)
2nd place
Brahman, Chhetri, Thakuri, Sanyasi 30.9 19 (46) 19
Total Madhesi (excluding Tharu) 22.5 9 (11) na
Maithali speaking 12.3 5 (5) 1
Bhojpuri speaking 7.5 3 (4)
Awadhi speaking 2.5 1 (2)
Hill Dalit 7.2 0 (0) 15
Magar (hill janajati) 7.1 1 (3) 9
Tharu (terai janajati) 6.7 1 (2) 5
Tamang (hill janajati) 5.6 1 (4) 5
Newar (hill janajati) 5.5 1 (2) 1
Muslim 4.3 0 (0) 6
Rai (hill janajati) 2.8 0 (3) 4
Gurung (hill janajati) 2.4 1 (2) 4
Limbu (hill janajati) 1.6 0 (2) 1
Others 3.3 0 (0) 3
Total 100 33 (75) 72

In column 3, we have total of 72 districts at the last row because due to lack of data on language, I am not able to identify which group is in second position in the district of Jhapa, Morang and Siraha. The three districts where other (mountain and hill janajatis) are in second place are districts of Mustang, Humla and Mugu.
Although the Madhesi group is 22.5% Nepal’s population, the three language groups add only to 22.3 indicating that other remaining Madhesi people speak other than these three languages.
na: not available

The largest group, BC makes 30.9% (column 2) of Nepal’s population. The other groups in descending order of their shares are Maithali speaking, Bhojpuri speaking, Hill Dalit, Magar, Tharu, Tamang, Newar, Muslim, Rai, Awadhi speaking, Gurung and Limbu. And “others”. BC constitutes the majority (more than 50%) in 19 districts (first entry in column 3) and the plurality (largest fraction but less than 50%) in another 27 districts, with a combined districts of 46 where they are either in majority or in plurality (entry inside the parentheses in column 3). The distribution of other ELC groups is rather spread thin across nation. Out of 75 districts in Nepal, the other ELC groups are in majority only in 14 districts, and they are in plurality in additional 15 districts, with a total of 29 districts where they are either in majority or in plurality (instead of 46 districts for BC).

Acharya (2007) looks at the population distribution of these groups across 75 districts to identify the natural homelands of these groups. An existing district is considered a natural homeland of a group if that group is at least the second largest fraction after BC in the district. Hence, if that group is in majority (more than 50% of the population) among all groups, or in plurality (with highest fraction but less than 50% of population) in the district, then the district will be the natural homeland of that group. According to this criterion, there might be majority or plurality of BC population in a district, but still that district will be considered a natural homeland of another group if that group is largest among other ELC groups. Indeed that is the case for 46 out of total of 75 districts. All districts that qualify to be a natural homeland for a given group are combined to make a focus region for that group. Since BC group is either in majority or in second place in most of the districts, the exercise of identifying natural homeland was not applied to BC.

By this mechanism, 68 districts were combined among eleven regions, as natural homelands for eleven groups except for Muslim and others. Understandably the latter category could have a natural homeland because it combines several very small groups. The Muslim group, however, represents 4% of the population and does not have a natural settlement because they are quite scattered. One of other groups given in Table 1 is higher in number than Muslims in all districts.

These eleven groups and their respective homeland regions are: Limbu (Kanchenjunga), Rai (Sagarmatha), Tamang (Gaurishanker), Newar (Kathamandu), Maithali speaking (Mithila), Gurung (Annapurna), Magar (Ridi) Bhojpuri speaking (central terai), Dalit (Khaptad), Awadhi speaking (Lumbini) and Tharu (western terai). Of the seven districts which could not be identified as natural region for ELC group, in three districts, there is no group that could be defined as natural dweller of that area, and majority or plurality of people are of Nepali mother tongue, mostly BC (note that Nepali is mother tongue of Dalit as well). These three districts compose a region without any focus group, thereby giving 12 regions. The remaining four districts are located in the most remote area of Nepal. Most of the people are BC and there is no other distinct group that could be considered a focus of that region. We propose it to be a territory.[10]

Based on this criterion, Acharya (2007) has the following Table which is reproduced here as Table 4. Our regional analysis will be based on the same 12 regions and one territory. These regions include different number of districts as described in the note below the table. We would like to continue the analysis with 12 groups (BC and the 11 groups that are considered as focus groups for the regions defined above, but that was not possible. Since we were not able to identify some individuals from survey data from the Terai by their mother tongues, we could not form the three groups (Maithali speaking, Bhojpuri speaking and Awadhi speaking). We put most of them together into the category “other”. But we put Yadav which belongs to these three language groups as a separate group. We also carry Muslim as a separate group, thereby making the number of groups twelve. Neither Yadav nor Muslim was a separate group in our regional formation. To sum, out of 12 groups here, only nine of them are the same as set previously. The reaming three, Yadav, Muslim and “Other” appear as separately but cannot be broken down into language profile.

Table 3. Summary by regions and provinces

Regions Focus group Number of existing districts* Population (%) Region’s share in focus group’s population (%) Focus group’s share in region’s population (%) BC’s share in region’s population (%)
Kanchenjunga region Limbu 3 2.0 49.5 39.5 26.6
Sagarmatha region Rai 8 7.0 63.2 25.2 29.7
Eastern terai region 3 9.2 8.3 27.7
Gaurishanker region Tamang 9 10.6 59.7 31.8 33.1
Kathmandu region Newar 3 7.2 46.8 35.4 37.8
Mithila region Maithali 5 13.2 82.8 77.2 5.4
Annapurna region Gurung 5 3.8 37.4 23.4 37.9
Ridi region Magar 13 14.0 51.1 26.0 41.2
Center terai region Bhojpuri 4 9.1 62.7 51.8 15.3
Khaptad region Dalit 11 8.0 17.8 18.0 68.0
Lumbini region Awadhi 3 6.9 91.7 32.6 19.6
Western terai region Tharu 4 8.1 46.0 38.4 34.7

Rara territory

4 0.7 1.6 70.7
NATIONAL TOTAL 75 100 61.4 30.9

Note: The sum of last two columns (92.3%) do not add to 100 because the share of Muslim population (4.3%), and other mountain and hill janajatis (2.3%) and others (1%) is not included in either of the two column.
* The district names for different regions are as follows. Kanchenjunga includes, districts of Taplejung, Terhathum and Panchthar. Sagarmatha region includes eight districts: Ilam, Sankhuwasabha, Solukhumbu, Bhojpur, Khotang, Okhaldhunga, Dharan, Udayapur. The region of eastern terai includes Jhapa, Morang and Sunsari. Gaurishanker region has nine districts: Dolakha, Sindhuliplachok, Rasuwa, Ramechap, Kaverpalanchok, Nuwakot, Dhading, Sindhuli and Makawanpur. Kathmandu region includes, Kathmandu, Patan and Bhaktapur districts. Mithila region includes Saptari, Siraha, Dhanusa, Mahottari and Sarlahi. Further to the west is Annapurna region with five districts, Gorkha, Lamjung, Manang, Mustang and Kaski. Ridi region has 13 districts: Tanahu, Syanja, Parbat, Myagdi, Palpa, Gulmi, Baglung, Arghakhanchi, Pyuthan, Rolpa, Rukum Salyan and Nawalparasi. Central terai has districts of Rautahat Bara, Parsa and Chitwan. Khaptad region has 11 districts: Surkhet, Jajarkot, Dailekh, Kalikot, Achham, Bajura, Bajhang, Doti, Darchula, Baitadi and Dadeldhyra. Lumbini region includes Rupandehi, Kapilbastu and Banke. The four districts Dang, Bardia, Kailali, and Kanchanpur are in western terai region. Finally, Rara territory includes Dolpa, Jumla, Mugu and Humla.

Poverty and Income Inequality

We know from the above discussion that Nepal is one of the poorest countries in the world with per capita income of US$234. Besides, very large segment of the population live below poverty line whether based on conservatively defined national level or measured as $1 day (PPP) by the World Bank. In this section, we look at the regional and group dimension of Nepal’s per capita expenditure and provide detail inter-regional, inter-group and intra-region and intra-group comparison.

In developing countries, a measure based on expenditure is considered more accurate than that based on income, as income tends to fluctuate more in these countries. This is true for Nepal as well where income depends, among other unstable factors, on volatile whether conditions. Therefore, we make expenditure per capita rather than income per capita as the basis of our analysis. However, in general, since relatively poor people have higher expenditure than their income, the expenditure-based measure will be upward biased (showing them relatively better off than they actually are). The first set of results are provided in Table 4 where average annual per capita expenditure for all 12 regions and one territory is shown. Column 1 shows the number of sample households that were covered in the survey, the total of 3,912.

At the national level, per capita mean annual expenditure is NR 15,707 (column 2).[11] The median per capita annual expenditure, however, is only NR 10,427 in column 3, implying that half of the population in Nepal have per capita annual expenditure less than NR 10,427. The 50% higher mean compared to median indicates very skewed distribution of expenditure toward very rich individuals. In other words, relatively very small number of very rich people have pulled the mean value. Although the level of per capita income series is a lot smaller using median than mean but the relative position of regions and groups is not much different whether we use mean or median. Since our focus is to look at relative position of the regions and the groups (rather than absolute values), without any loss of generality, we base our analysis on mean expenditure.

The magnitude of mean annual per capita income ranges from nine thousands in Khaptad region to, four times higher, 37 thousands in Kathmandu region. The second richest is Annapurna region, whereas the second poorest is Gaurishanker, with 10 thousand annual average per capita expenditure. The other two regions with higher than national average per capita are Eastern Terai and Center Terai. The richest region, Kathmandu, has more than double and Annapurna region has almost double the per capita income of national average. Indeed, the per capita income of Kathmandu is not much less than India’s average GDP per capita.

Table 4. Annual average per capita expenditure in Nepalese Rupee

Regions Number of Sample


Mean of the lowest 20%
Mean of the highest 20%
Ratio of top 10% to bottom 10%
Kanchenjunga 120 11,225 9,609 5351 23669 6.8
Sagarmatha 300 11,698 8,143 4775 39097 14.5
Eastern Terai 372 17,728 13,122 5588 34840 11.1
Gaurishanker 432 10,094 7,936 4706 32461 13.2
Kathmandu 504 37,031 24,627 5240 50780 14.4
Mithila 420 12,940 10,727 5369 29544 10.7
Annapurna 228 29,359 15,757 5166 53417 17.7
Ridi 503 13,774 10,356 4966 33450 10.8
Center Terai 300 15,984 10,934 5421 35678 13.0
Khaptad 264 9,058 7,847 5204 31148 10.3
Lumbini 156 15,707 9,418 4965 40296 10.9
Western Terai 265 13,131 8,906 5464 42749 13.9
Rara Territory 48 9,972 8,957 5455 21649
National 3,912 15,707 10,427 5,101 40,163 13.5

>Source: Authors’ calculation based on Nepal Living Standard Survey II, 2003/2004

Columns 2 and 3 show the income disparity among regions, however, they do not explain about income distribution pattern within a region, which is done in the next three columns. Column 4 shows the average annual mean per capita expenditure of the poorest 20%. Interestingly, the level of expenditure of this group is not that different across regions, implying that the poorest are equally poor in all regions. That is, their economic plight of the poorest is almost the same in every region. Their per capita expenditure ranges from NR 4,706 to 5,588. The lowest regional per capital expenditure of the bottom 20% (in Gaurishanker region), is at least 84% of the highest regional per capita expenditure of this group (in Eastern Terai). Hence, the issue of poverty is not a regional issue; it is a national issue that penetrates into all regions.

Column 5 is similar to column 4 but provides per capita expenditure of the richest 20% of the population. The expenditure of this group ranges from the lowest NR 23,669 in Kanchenjunga (lowest after Rara Territory) to the highest NR 53,417 in Annapurna. The mere fact that the per capita expenditure of the top 20% group in the former region is only 44% of the per capita expenditure by its counterpart in the latter region is a clear indication that the people in the higher strata of income are not equally rich in each region. So the findings that the poor are almost equally poor across regions, and the rich are not equally rich across regions, meaning that there are inter-region income disparities and within region income disparities, and the latter vary by regions.[12]
The within region income disparities are higher for richer regions is confirmed by the last column where the quotient of top 10% (richest) to bottom 10% (poorest) population’s per capita expenditure is computed. At the national level, the top 10% (the richest) group has 13.5 times more expenditure than the bottom 10%, the poorest.[13] The region with the highest level of within region disparity (shown by highest number in column 6) is Annapurna followed by Kathmandu and Sagarmatha. On the other hand, the most egalitarian regions are Kanchenjunga, Khaptad and Mithila. Expenditure per capita of top 10% is about 7-fold higher in Kanhenjunga and about 18-fold higher in Annapurna region, compared to that of the bottom 10% of the same region. So, the richer regions are more unequal than the poorer ones. This is due to the fact that the level of expenditure for the poorest groups (either looking at the bottom 10% or the bottom 20%) does not differ that much but the expenditure of the top 10 or 20% differs a lot by region.

Although Table 4 reveals a lot about the status of expenditure by regions, it does not provide explicit count of how many people falls below poverty line and what is the magnitude of between and within region income disparities. That information if provided in Table 5, where the first column shows that at the national level 31% of people are below poverty line (defined at income of NR 7,696). The regions with highest shares of people below poverty line in regional population are Khaptad (48%), Sagarmatha (47%) and Gaurishanker (47%). In Kathmandu, only 4% of the population are below poverty line, the lowest in any region.

The poverty measure provided in column 1 measures the incidence of poverty (percent of poor), but it does not tell us about how the overall distribution of expenditure is across individuals in the region. For that we need to calculate, Gini coefficient, which is given in the second column. Gini coefficient measures the degree of inequality within each region’s population by taking account of all population in the region.[14] The higher the coefficient, the higher is the inequality among individuals in the region. The highest inequality is seen in Annapurna region followed by Lumbini and Kathmandu. The first two are the only regions that have higher Gini coefficient than that for the nation (more unequal than the national average). The regions with the lowest inequality are: Khaptad, Rara territory and Gaurishanker.

So far, columns 1 and 2 provided the share of people under poverty and the overall poverty measure among individuals respectively but they did not provide the depth of the poverty which is provided in column 3, the poverty gap. Here too, the higher the number the higher is the depth of poverty; that is, relatively large number of poor people are further below from the poverty line.[15] In this count, the poor in Sagarmatha regions are the most destitute ones followed by those at Gaurishanker and Khaptad.

As reported in column 4, the largest number of poor people (16% of the nation’s population) resides in Gaurishanker region. The next column shows the ratio of regional share in nation’s poor to the regional share in nation’s population. The factor with more than unity means that the region has disproportionately more poor than at the national level, and the region with less than unity means that the region has disproportionately less poor than at the nation level. Note that this column is directly related to column 1 (each region entry over the national average should have given the same number as in this column).

Table 5. Measure of poverty and income inequality by focus regions

Population below poverty line (%)


Poverty gap


Regional share in Nepal’s poor (%)
Ratio of share of poor to share of population (%)
Kanchenjunga 38 28 7.5 2.4 1.2
Sagarmatha 47 42 15.1 10.6 1.5
Eastern Terai 14 37 2.8 4.1 0.4
Gaurishanker 47 35 14.2 16.0 1.5
Kathmandu 4 44 0.8 0.9 0.1
Mithila 26 31 5.6 11.0 0.8
Annapurna 17 54 4.4 2.1 0.6
Ridi 28 36 6.3 12.6 0.9
Center Terai 26 41 5.1 7.6 0.8
Khaptad 48 26 11.8 12.4 1.6
Lumbini 41 46 10.4 9.1 1.3
Western Terai 40 40 8.5 10.4 1.3
Rara Territory 41 26 10.3 0.9 1.3
National/Total 31 44 7.6 100 100

Note: Column (3) is obtained by multiplying regions’ entry in Column (1) with region’s population that is below poverty line and dividing that product by total number of people below poverty line in Nepal.
Source: Authors’ calculation based on Nepal Living Standard Survey II, 2003/2004

With this, the discussion on between and within regional inequalities is complete. Next we want to look at the between and within inequalities for the major ELC groups. According to the average per capital expenditure presented in Table 6 (columns 2 and 3) the group with the highest per capita expenditure is Newar, whose main settlement is in Kathmandu region. Gurung, which has the second highest per capita expenditure, has only 68% per capita expenditure of Newar. The per capita expenditure is as low as NR 8,828 for Tamang, just 20% of the Newar. The BC has per capita expenditure is NR 19,370, less than 60% of Newar.

Again, the per capita expenditure of the bottom 20% is not that much different across groups. Surprisingly the third lowest per capita income for this bracket is recorded for Newar group. The top 20% of Newar has the highest income compared to their counterpart from any region. Income disparity within group is shown in the last column where the share of bottom 10% expenditure in respective group’s top 10% expenditure is computed. The larger the number in this column, the larger is within group inequality. For example, the entry in the first column of 10.3 indicates that, the richest 10% Limbus have 10.3-fold larger expenditure compared to the poorest 10% Limbus. The highest within group inequality is reported for Rai and Newar. On the other hand, the lowest within group inequality is reported for Yadav and Tharu.

Table 6. Annual average per capita expenditure in Nepalese Rupee by focus groups

Focus groups

Sample size

Mean(2) Median

Mean of the lowest 20%(4) Mean of the highest 20%

Ratio of top 10% to bottom 10%
Limbu 64 12,766 9,977 5,285 32,097 10.3
Rai 139 11,753 7,989 4,953 39,890 17.9
Tamang 225 8,828 6,541 4,440 29,752 9.7
Newar 411 32,696 18,937 4,837 53,806 17.6
Yadav 105 13,897 11,707 5,457 27,297 7.9
Gurung 124 22,182 13,316 5,448 46,901 14.4
Magar 248 13,192 9,686 5,100 32,907 9.7
Muslim 304 10,838 8,123 5,064 33,830 11.8
Dalit 188 10,472 9,256 5,385 34,643 9.6
BC 1247 19,370 13,166 5,332 40,858 12.5
Tharu 168 11,827 8,618 5,378 24,755 8.7
Others 689 9,816 5,199 32,235 12.0
National 3912 15,707 10,427 5,101 40,163 13.5

Source: Authors’ calculation based on Nepal Living Standard Survey II, 2003/2004

In Table 7, we provide share of population below poverty line and Gini coefficient for the groups. The share of people who are below poverty line is highest for Tamang at 61% followed by Rai at 49% and Dalit at 46%. On the other hand, this share is lowest for Newar at 14% followed by BC and Yadav at 18% each. Column (2) shows that the most unequal group, compared to any other, is Newar whose Gini coefficient is highest at 66. The most equal group, on the other hand, is Yadav (Gini coefficient of 14) followed by Muslim. For the BCV group, 18% of them are below poverty line and income distribution for this group is the third unequal after Newar and Gurung. The largest number of people below poverty line are in BC, wheras the largest share in their own population are from Tamang.
In the last column, the entry of “2” for Tamang indicates that their share in people below poverty line is double their share in total population. Therefore, Tamang is the group where the poverty is most concentrated followed by Rai and Muslim. The lowest level of poverty concentration is found in Newar, Yadav, BC and Gurung.

Table 7. Measure of poverty and income inequality by focus groups

Focus groups Population below poverty line (%)
Gini coefficient
Poverty gap

Group share in Nepal’s poor
Ratio of share of poor to share of population (%)
Limbu 41 22 7.5 2.0 1.3
Rai 49 30 14.6 5.9 1.6
Tamang 61 24 22.1 11.8 2.0
Newar 14 66 3.5 3.4 0.5
Yadav 18 14 3.3 1.9 0.6
Gurung 19 45 4.7 1.3 0.6
Magar 34 22 6.9 6.9 1.1
Muslim 41 16 11.1 10.9 1.5
Dalit 46 18 7.5 9.2 1.1
BC 18 34 4.0 16.4 0.6
Tharu 35 17 9.3 8.7 1.3
Others 34 7.7 21.7 1.1
National 31 44 7.6 100.0 1.0

Source: Authors’ calculation based on Nepal Living Standard Survey II, 2003/2004

To sum up, at the national level, the average annual per capita expenditure is NR 15,707. In other measure, half of the Nepalese people have annual per capita expenditure of less than NR 10,427. However, the situation is more appalling for some regions and groups. In regional terms, Khaptad is the region with the lowest per capita expenditure, whereas Kathmandu is the region with the highest per capita expenditure. Measured by expenditure factor of top 10% to bottom 10% expenditure, the highest level of within group inequality is in Annapurna and the lowest is in Kanchenjunga. Data shows that the poorest have the same economic standing throughout the region and income differs by region based on how better are middle and rich income groups. In other words, poverty is spread throughout the region but there are region differences among people who are different regions.

At the national average, 31% people are below poverty line in Nepal and the region with the highest share of people below poverty is found in Khaptad again with 48%. The lowest share of people below poverty line is in Kathmandu at 4%. According to Gini coefficient, the most unequal distribution is in Annapurna with the lowest unequal in Khaptad. In terms of poverty gap, the deepest poverty is in the region of Sagarmatha. In tems of number in the country, the highest number of people below poverty line is residing in Gaurishanker (16%).
Looking at the same information by ELC groups, the group with the highest per capita expenditure is Newar, while the one with lowest is Tamang. The share of expenditure of the 10% poorest people in expenditure of top 10% in its own group is lowest in Newar (making it most unequal within group income distribution) and highest for Yadav. TAmang has the highest share of people below poverty line and Newar has the lowest. Based on Gini, as shown in top 10% to bottom 10%, the most unequal distribution is among Newars. Poverty is deepest for Tamangs, Tamangs share of people below poverty line in Nepal is double the share of Tamang in Nepal’s population.

Education by Region, Group and Gender

In this section, we look at the education attainment of regions and groups. There is more or less consensus that the early education has generates very high rate of social returns, private return plus positive return that is spilled to others in the society. In that respect, the country acquires high rate of return by investing on education. However, in Nepal’s case, education, even the most basic primary one is a very luxury commodity that the majority of Nepalese cannot afford. Nationally, 51.9% of the Nepalese is illiterate (Table 8, column 1).

Table 8. Education status of population above age five by focus region (in percent)

Regions Illiterate
Grade 1-4
Grade 5-10
Grade 10+
Grade 1-4 for household head
grade 5
(6 = 1 + 5)
Kanchenjunga 49.6 38.1 9.6 2.8 22.8 72.4
Sagarmatha 55.0 31.9 10.2 3.0 21.9 76.9
Eastern Terai 50.1 30.5 13.1 6.3 21.5 71.6
Gaurishanker 53.0 37.7 8.2 1.1 20.5 73.5
Kathmandu 46.1 22.9 20.8 10.2 17.4 63.5
Mithila 59.2 33.2 6.4 1.1 9.4 68.6
Annapurna 53.4 27.5 14.7 4.4 17.9 71.3
Ridi 49.7 38.8 9.4 2.2 22.0 71.7
Center Terai 53.1 33.8 8.5 4.6 14.4 67.5
Khaptad 53.0 40.4 5.7 1.0 18.2 71.2
Lumbini 48.9 39.1 9.3 2.7 22.8 71.7
Western Terai 47.6 36.3 13.6 2.5 19.5 67.1
Rara Territory 62.8 30.3 6.9 0.0 24.5 87.3
National 51.9 34.2 10.6 3.4 74.1

Source: Authors’ calculation based on Nepal Living Standard Survey II, 2003/2004

In Mithila region 59% of its population is illiterate, the highest in any region (after Rara Territory). The region with the second highest number of illiterate is Gaurishanker, and not surprisingly, the lowest share of illiterate is in Kathmandu at 46%. The other regions with worse than national standing in illiteracy are: Sagarmatha, Annapurna, Center Terai, Gaurishanker and Khaptad. The share of illiterate people between the region with highest share of illiterate and the region with lowest share of illiterate differs by thirteen percentage points (from 46% to 59%), a wide disparity. However, compared to the illiteracy rate of 52% at the nation, this difference is rather small, spreading six percentage points below and above average each. What this reveals is that illiteracy is not a regional issue; it is a national issue. The high illiteracy rate is not mainly because of the failure of regional balance (it certainly is a factor), but is mainly because of the failure of nationwide education policies. Note that even if all regions acquire the same level of literacy as the region with the highest level of literacy, still the level of illiteracy would have been 46%, a daunting number.

In column 2 we show the share of population with grade 1-4 education. Now, since this data are for everybody above the age of five, this number includes both who are of school age for this grades and those who have dropped out of the schools and are not going to get more than this level of education (note the unusually high number for Khaptad; that represents drop out rate). Since the latter type can be counted as illiterate for working life, we would like to see the share of the latter type separated from the total. However, data are not readily available for this separation.

However, there is another piece of information in the data, which gives us some handle on it. There is information on household head with grade 1-4 education, which is given in column 5. Since the household heads are adult, we don’t expect them to have further education. In other words, these are school dropouts who constitute about 22% of the household head. By adding the share of illiterate (column 1) and the share of grade 1-4 education household head (column 4) we have column 6. The share of people given in column 6 can be considered as effectively illiterate. It shows that an astonishing 74% of Nepal’s population are illiterate in effective sense. The share of effectively illiterate people is as high as 77% for Sagarmatha.

Column 3 shows the share of people who have education from grade five to grade ten, and Column 4 shows the share of people with above grade 10 level of education. As we go to the higher level of education, the regional differences start to surface. For example, the share of people with grade 5-10 level of education is double in Kathmandu than found at the national level. And this share of Khaptad is half that of the national share. At the highest level of education attainment reported in the survey (above grade 10), there are only 3.4% people in Nepal who have obtained this (column 4). In other words, in a country of 22.7 million, only 730 thousand people have above grade 10 level of education. The share of people with this level of education is three times higher in Kathmandu than at the national level, two times higher in Eastern Terai. The other two regions that fairs better than the national level are Annapurna and Center Terai.

Among the people with above grade 10 level of education, about one-quarter of them lives in Kathmandu, a region with just 7% of the population. The other hilly regions which has about half of the nation’s population has only 29% of above grade 10 level educated people.[16] Particularly alarming disparity is in the region of Khaptad, where only 1% of the population has education above grade 10. As a result, a region with more than 8% of nation’s population, Khaptad has only 2.5% people with education above grade 10. Put differently, the region with 1.8 million population has only about 18 thousand people who have attained education higher than grade 10. Similarly, Gaurishaker has 11% of nation’s population but only 3.6% of people with education above grade 10. The odd for Mithila is also the same with 14% population but only 4.5% with this level of education. The regional disparities seen at the illiteracy level even widens at the higher level of education.

Next we look at the education attainment by groups. The level of illiteracy among groups varies from 48% for Newar to 57.5% for Gurung (Table 9). The other groups, which have more incidence of illiteracy than the national average, are Limbu, Rai, Tamang, Yadav, Magar and Muslim. The range of group disparities, ten-percentage point, is smaller than the 13-percentage point seen in regional disparities. It means that the disparities in literacy among groups are smaller than among regions. Here too, each group has substantial level of illiteracy making it as national phenomenon, but there are substantial group wise variation as we move to the higher level of education.

In the highest level of education category, that is above grade 10, Dalit and Tamang are the worst achievers, with only 0.8% and 0.9% of their respective population. Out of 1.35 million population of Tamang, only about 12 thousands have above grade 10 education. And with 1.83 million populations, only 14.6 thousand Dalits have above grade 10 level of education.
The groups that fear the best in education are BC and Newar, with the shares of illiterate at 48% for both. Newar supplies 15% of the nation’s population with above grade 10 education (with 5.5% share of the population), almost three times its share of nation’s population. BC supplies 43% of the high school graduates (with 28% of the population share). Newar and BC combined supplies more than 58% of highly educated people in Nepal. To conclude, each group has almost the same share of people who are illiterate, but the share of people with higher level of education varies by group. Therefore, there is not much group disparities in basic education but there are group inequalities in higher level of education.

Table 9. Education status of population above age of five by focus groups (in percent)

Focus groups Illiterate
Grade 1-4
Grade 5-10
Grade 10+
Number of people with 10+ education (in thousand)
Limbu 57.2 30.3 8.0 4.5 15.2
Rai 54.4 30.8 11.2 3.6 30.7
Tamang 54.2 39.0 5.9 0.9 12.1
Newar 48.1 26.6 18.4 6.9 118.0
Yadav 54.0 38.0 5.4 2.6 20.0
Gurung 57.5 32.2 7.4 3.0 14.7
Magar 52.7 36.5 9.4 1.5 21.2
Muslim 51.5 40.2 6.3 2.0 33.5
Dalit 50.1 35.1 14.0 0.8 14.6
BC 48.0 32.6 14.2 5.2 329.0
Tharu 53.3 39.9 1.8 5.0 73.7
Others 57.4 33.6 7.3 1.8 81.2
National 51.9 34.2 10.6 3.4 22,739.2

Source: Authors’ calculation based on Nepal Living Standard Survey II, 2003/2004

Next we look at he gender inequality in education and type of schooling the school goers are attending by regions and groups in Table 10 and 11 Respectively. At the national level, 40% of male and 62.5% of female are illiterate. In every region, the number of illiteracy is higher for female than for male. However, the discrepancy between male and female illiteracy in each region is almost like the discrepancy at the national level. This makes the gender inequality also a national problem. The difference between female and male illiteracy ranges from low of 20% in Eastern Terai, Gaurishanker and Kathmandu to 33% in Khaptad. So even though the level of gender discrimination is particularly high in Khaptad and center Tarai (27%), it is not less than 20% anywhere.

About 17% of the school-goers are in private school and that share varies significantly across regions from as low as 2% in Khaptad to 58% in Kathmandu. And there is nobody attending private school in Rara Territory. The lack of public education and for profit private schooling will have serious consequences as it create a pool of educated people who have attended private school and school dropouts who cannot get anywhere through private schooling in the future. However, this subject deserves a complete paper and will not be discussed here much further in this article.

Table 10. Gender inequality in education and types of school attended (percent) by region

Male illiteracy rate 5 years +
Female illiteracy rate 5 years +
Share in currently school attending population by types of schools
Government/ community
Kanchenjunga 39.4 60.1 93.9 6.1
Sagarmatha 42.2 63.2 91.8 8.2
Eastern Terai 29.6 49.2 75.9 20.5
Gaurishanker 51.9 71.7 92.0 8.0
Kathmandu 13.0 33.4 41.0 58.4
Mithila 56.9 81.8 88.7 9.0
Annapurna 22.5 45.2 79.8 19.7
Ridi 25.2 50.7 92.3 7.7
Center Terai 44.9 72.0 73.6 21.3
Khaptad 33.3 66.7 98.1 2.0
Lumbini 36.5 59.7 63.9 26.9
Western Terai 44.5 64.2 87.1 12.2
Rara Territory 39.4 60.1 66.5 0.0
National 39.9 62.5 81.5 16.7

Note: the sum of columns 3 and 4 does not add to 100 in some regions, as there are reported cases that people were attending “other” school. Howeve, it is not clear what types of school they are. The share of “other school” was very high at 33.5% for Rara Territory, 9.2% for Lumbini, 5.1% for Center Terai, 3.6% for Eastern Terai, 2.4% for Mithila, and less than 1% to zero for remaining regions.
Source: Authors’ calculation based on Nepal Living Standard Survey II, 2003/2004

By groups also the inequalities among male and female education attainment are equally alarming.

Table 11. Gender inequality in education and types of school attended, by groups (percent)

Male illiteracy rate 5 years +

Female illiteracy rate 5 years +

Share in currently school attending population by types of schools
Government/ community

Limbu 44.0 65.3 87.3 12.7
Rai 35.3 56.2 89.7 10.0
Tamang 57.3 78.0 89.5 10.0
Newar 21.4 42.6 57.7 41.7
Yadav 43.8 70.8 83.4 14.1
Gurung 32.8 57.4 71.1 23.6
Magar 29.9 54.9 86.5 13.4
Muslim 43.7 63.2 91.9 5.6
Dalit 42.5 66.1 93.7 6.3
BC 22.6 46.5 81.6 18.2
Tharu 60.2 82.9 49.5 20.0
Others 51.3 78.4 84.6 15.2
National 39.9 62.5 81.5 16.7

Source: Authors’ calculation based on Nepal Living Standard Survey II, 2003/2004
The country has equally unsuccessful educational policy. Still 51.4% people in Nepal are illiterate. If we consider at least grade 5 educations to be educated, the number of illiterate in Nepal is 70%. Again, there are regional and group-wise inequalities that are shocking. For example, the region with the highest shares of illiterate in Mithila (69%). On the other hand, Kathmandu has the lowest share of illiterate at 31%. Only 4% people at least grade 10 level of educatioin, making only less than a million people in Nepal who have this level of education. More than one-third of these people are in Kathamandu. The region of Khaptad with 1.8 million people, it has only 16 thousand who has grade 10 and above education attainment.

By group, the most illiterate are Tamangs; seven in 10 are illiterate; nine in 10 are either illiterate or less than grade five level of education, thereby making the effective level of illiteracy at 90%. The highest share of educated are in Newar; 61% of them has some level of education. With only 5.5% of nation’s population, they have 27% of nations with grade 10 or above education. Dalit with 1.8 million population has only 1800 people who have higher than grade 10 level education. In similar vain, in 1.35 Tamangs, only 4000 have received grade 10 and above degree.

There is also gender divide in education. In each region, there is at least 20-percentage points difference at the level of male and female illiteracy rate, higher for the female. The gender gap is highest in Khaptad with 33%, where women’s illiteracy is as high as 81.8%. Female illiteracy rate is lowest in Kathmandu with 33% (with male rate at 13%). The richer the region is, the higher is the percentage of people attending private school. In total, 17% of school goers are in private school. The rate is very high for Kathmandu value at 58% and lowest for Khaptad at 2%.

Looking at gender disparities in education by group, we find that Tharu has the highest share of illiterate in both male and female. All groups has 20-27% gap between male and female illiteracy, higher for the female. In terms of types of school, 42% of Newar go to private school whereas less than 6% of Muslim and about 6% of Dalit attends private school. BC group is in the middle of the pack in terms of sending children to the private school at 18%.

As a final component, we look at the access to school, hospital and paved road (the table are provided in Appendix B, Tables B1 and B2). The access to school, hospital is farthest for people in Gaurishanker region. It takes about an hour to reach a primary school in this region wheras it takes only 0.13 hours in Kathmandu. In Gaurishanker region, it takes about 1.4 hours to reach a health post whereas it takes 0.28 hours in Kathmandu, about four times less. Paved road is 3.22 hours away for Khaptad, 3 hours for Ridi and slightly less than 3 hours for Gaurishanker. Here too, the shortest time taken for reach pave road is in Kathamandu, 0.32 hours.
Among groups, Tamang travel farthest to reach a primary school, health post and paved road. By comparison, Newar are nearer to primary school by about half an hour, nearer to health post by about an hour and nearer to pave road by about an hour compared to Tamang.


We have provided a comparative picture of Nepal vis-à-vis India and China using data from the last four decades. Nepal’s situation is disappointing compared to its neighbours. We have also provided a novel analysis based on the major population groups (i.e. those that are more than 1% of the total population) and their settlement regions in Nepal. This analysis is vital for policy makers that are genuinely concerned with tackling the poverty, stagnation, regional and ethnic, language, and caste (ELC) group disparities. We have also analysed the state of educational attainment in Nepal and have shown that the neglect of public education and rapidly increasing number of private schools is creating systematic segregation in Nepalese society. Nepal will fall into a deeper crisis if these issues are not dealt with.
Political restructuring alone is not sufficient for Nepal’s peace and prosperity. Without companion economic restructuring political restructuring will not provide much-needed growth or poverty alleviation for Nepalese society. This paper has provided a blueprint for designing successful economic and social policies using the most recent surveys from 2001 and 2003/2004 (Central Bureau of Statistics, 2004).


Acharya, R.C. 2007. A Model for Political Restructuring and Electoral System of Federal Nepal: Building on the Strength of Ethnic Diversities and Regional Complementarities, Draft.

Central Bureau of Statistics. December 2004. Nepal Living Standard Survey 2003/2004, Statistical Report, Volume 1

Pradhan, R. and S. Ava. June 2005. Ethnic and Caste Diversity: Implications for Development, Working Paper Series No. 4, Asian Development Bank.

The World Bank. 2006a. Unequal Citizens: Gender, Caste and Ethnic Exclusion in Nepal.

The World Bank. June 2006b. Nepal Resilience Amidst Conflict: An Assessment of Poverty in Nepal, 1995-1996 and 2003-04.

Appendix A: Nepal Living Standard Survey

Using 2001 Population Census of Nepal the size of each ward (as measured by number of households) was taken as a unit of sample frame, where bigger wards were divided and smaller one were appended into neighbouring wards. The resulting sampling frame consisted of 36,067 enumeration areas (wards or sub-wards) covering all 3,914 Village Development Committees (the smallest political entities) of the country. For the cross-section part of the sample, first 334 Primary Sampling Units (PSU ) were selected from six explicit strata (Mountains, Kathmandu valley urban area, Other Urban areas in the Hills, Rural Hills, Urban Terai and Rural Terai) of the country. Then 12 households in each of 334 PSU were selected randomly, with total of 4008 households.

The allocation of cross section PSUs by districts were as follows. Taplejung 4; Morang 16; Sunsari 11; Solukhumbu 3; Panchthar 3; Dhankuta 4; Bhojpur 4; Okhaldhunga 2; Ilam 5; Tehrathum 2; Khotang 3; Siraha 8; Jhapa 13; Sankhuwasabha 5; Saptari 9; Dhanusa 11; Udayapur 6; Makwanpur 8; Ramechhap 3; Lalitpur 10; Mahottari 8; Rautahat 7; Dolakha 5; Bhaktapur 6; Sarlahi 9; Bara 8; Sindhupalchok 10; Kathmandu 6; Sindhuli 5; Parsa 7; Kavrepalanchok 7; Nuwakot 5; Dhading 6 Kathmandu 35; Chitwan 9; Tanahun 5; Syangja 7; Gorkha 5; Kaski 14; Gulmi 5; Lamjung 4; Myagdi 3; Palpa 4; Manang 1; Parbat 2; Rupandehi 10; Nawalparasi 8; Baglung 4; Arghakhanchi 4; Kapilbastu 8; Banke 7; Bardiya 5; Doti 4; Pyuthan 4; Surkhet 7; Kalikot 3; Kailali 8; Rolpa 3; Dailekh 4; Mugu 2; Kanchanpur 5; Rukum 2; Jajarkot 1; Bajura 3; Dandeldhura 2; Salyan 3; Dolpa 1; Bajhang 4; Baitadi 4; Dang 6; Jumla 1; Achham 4; Darchula 3; Humla 1.

In addition, this survey interviewed 1160 households from 95 panel PSUs (962 out of 1160 households were panel households that were also interviewed in 1995/96), and the remaining 198 households were new households from panel PSUs.

Appendix B: Access to facilities

Table B1. Average hours to access the nearest facilities, by regions

Primary school
Health post
Paved road
Kanchenjunga 0.30 0.78 4.48
Sagarmatha 0.43 1.13 2.95
Eastern Terai 0.23 0.53 0.93
Gaurishanker 0.55 1.38 2.55
Kathmandu 0.13 0.28 0.32
Mithila 0.22 0.37 1.13
Annapurna 0.18 0.88 1.32
Ridi 0.32 1.03 3.02
Center Terai 0.18 0.45 1.10
Khaptad 0.42 1.28 3.22
Lumbini 0.28 0.63 0.87
Western Terai 0.27 0.75 0.92
Rara Territory 0.55 0.62 0.00
National 0.30 0.78 1.88

Table B2. Average hours to access the nearest facilities, by groups

Primary school
Health post
Paved road
Limbu 0.35 0.93 2.50
Rai 0.47 1.40 2.27
Tamang 0.52 1.42 2.53
Newar 0.23 0.48 1.95
Yadav 0.20 0.47 1.13
Gurung 0.23 1.17 0.95
Magar 0.38 1.17 2.92
Muslim 0.30 1.10 2.43
Dalit 0.23 0.52 1.52
BC 0.33 0.80 2.12
Tharu 0.22 0.45 1.18
Others 0.23 0.50 1.32
National 0.30 0.78 1.88

[1] This was the follow up survey of NLSS (I) conducted in 1995/1996 by Central Bureau of Statistics, Nepal following the methodology of Living Standards Measurement Survey methodology developed at the World Bank and applied in more than 50 developing countries.
[2] Official exchange rate is US$/NR, and conversion factor is International $/NR, the ratio of these two yields International $/ NR and multiplying the GDP at NR by this factor yields Nepalese GDP at International $, which is called GDP in PPP. The ratio of the PPP conversion factor to the official exchange rate (also referred to as the national price level) makes it possible to compare the cost of the bundle of goods that make up gross domestic product (GDP) across countries. Official exchange rate refers to the exchange rate determined by national authorities or to the rate determined in the legally sanctioned exchange market. Purchasing power parity conversion factor is the units US dollar that is required to buy the same amount of goods and services in the domestic market as a U.S. dollar would buy in the United States. Put simply, conversion factor of 0.18 means that 18 US cents is required to buy a product in Nepal that cost 1 dollar in the US.
[3] In terms of PPP, Nepal’s GDP is slightly smaller than that of Rhode Island (with about US $46 billion), the seventh smallest state in terms of gross state product and with a population of about 1.1 million.
[4] If a serious tax incidence study is conducted, we will not be surprised to see that Nepal’s fiscal structure actually siphons money to the relatively affluent to the society form those who are at the lower level of income distribution.
[5] Poverty headcount ratio at $1 a day (PPP) (% of population is defined as Population below $1 a day is the percentage of the population living on less than $1.08 a day at 1993 international prices. As a result of revisions in PPP exchange rates, poverty rates cannot be compared with poverty rates reported previously for individual countries. Data showing as 2.0 signifies a poverty rate of less than 2.0 percent.
[6] Data in Table A1 are presented in the same way that the Census data are reported. The number of groups that represent more than 1% of the population would have been even more than 23 had the Census data were collected for every single ethnic/language caste group. Notice in some cases, data have been reported for a combined group. For example, “Koiri, Kurmi, Kanu, Haluwai, Hajam and Thaku” are reported under one group (see Table A1 in the appendix).
[7] Hill Dalit includes three groups,(Kami, Damai, and Sarki reported in Table A1. Others include mountain Janajati Sherpa, Bhote, Walung, Byansi, Hyolmo, hill Janajatis Gharti, Bhujel, kumal, Sunuwar, Baramu, Pahari, Thakali, Yakkha, Chhantal, Jirel, Dura, Thami, Lepcha, Chepang, Hayu, Raute, and Kusunda and others such as Panjabi/Shikh, Jaine and Undefined/Others.
[8] They are considered hill janajati, or hill ethnic groups, or hill nationalities.
[9] It is not pleasant to identify the region as focus for “untouchable” or Dalit, the very concept that should be eliminated. However, since the Dalits are the poorest and excluded from all walks of life, it still makes a lot of economic sense to have this as a focus region. One might change the name from Dalit focus to something else, but a focus region considering the future of this group is essential.
[10] The difference between region and territory is how they are treated in federation, is clarified in Acharya (2007).
[11] At the conversion rate of NR 75 per US$ (for average of years 2003 and 2004 years that the survey was conducted), this amount is equal to US$209 (US$ 0.57 for a day). At PPP conversion rate of 0.16 of the official exchange rate, this amount is equal to PPP international $1,306 ($3.6 per day).
[12] Note that neither the region with the lowest per capita income (Khaptad) is the one with the lowest per capita expenditure for the bottom 20% (Gaurishanker) nor the region with highest per capita expenditure (Kathmandu) is not the one with highest per capita income of the 20% bottom population.
[13] Intuition tells the factor would have been much greater had we use the income rather than expenditure.
[14] Gini index measures the extent to which the distribution of income (or, in some cases, consumption expenditure) among individuals or households within an economy deviates from a perfectly equal distribution. A Lorenz curve plots the cumulative percentages of total income received against the cumulative number of recipients, starting with the poorest individual or household. The Gini index measures the area between the Lorenz curve and a hypothetical line of absolute equality, expressed as a percentage of the maximum area under the line. Thus a Gini index of 0 represents perfect equality, while an index of 100 implies perfect inequality.
[15] Poverty gap is the mean shortfall from the poverty line (counting the nonpoor as having zero shortfall), expressed as a percentage of the poverty line. This measure reflects the depth of poverty as well as its incidence. Data showing as 0.5 signifies a poverty gap of less than 0.5 percent.
[16] The other hilly regions are Kanchenjunga, Sagarmatha, Gaurishanker, Annapurna, Ridi, and Khaptad.

Leave a Reply