Net Versity

 Poverty trends in South Asia

^. 44% of the population of India lives below the international US$1/day poverty line.

^. In Nepal, Pakistan and Bangladesh the figures are also relatively high (at 38%, 31% and 29% respectively).

^. In Bhutan and Afghanistan, where data is unavailable, the proportion of people living on US$1/day is likely comparable and much higher, respectively.

^. Internationally, South Asia has the worst indicators of stunting and female illiteracy, and very poor rates of child mortality and female illiteracy.

^. The headcount ratio for the chronically poor has been declining in many parts of the region – particularly in southern and western India, and in Bangladesh.

^. Most human development indicators also have improved over the past two decades, although in Afghanistan years of war have obstructed almost all potential progress.

How many people are chronically poor in South Asia?

The number of recent, high quality, representative and comparable panel surveys available to determine the extent of chronic poverty is very limited. Best estimates suggest that about one-third of the poor population in South Asia is chronically poor – between 135 and 190 million people, of whom 110–160 million are Indians. Bangladesh and Pakistan account for the majority of the remainder.

A survey of rural Bangladesh suggests that close to one-third of the rural population was poor in both 1987/8 and

2000.² In India, two national sample surveys suggest that in the late 1960s³ and between 1970 and 1981⁴, almost half the rural poor were chronically poor.

A third survey, collected only in semiarid rural Andhra Pradesh and Maharashtra, found that over one-fifth of the population was poor in all nine years between 1975/6 and 1983/4, while 60 % were poor in at least five of nine years.⁵ Further analysis of this dataset suggested that even relatively affluent households are highly vulnerable to long spells of poverty when severe crop shocks occur.⁶

The Indian National Sample Survey reported that the number of poor people increased by 13 million between 1987 – 88 and 1993–94, while data from 1999 – 2000 shows a very large reduction in the second half of the decade. This finding is intensely disputed, however, due to changes in the way the national figures have been calculated, and as such it remains difficult to estimate the absolute numbers of chronically poor people today. Due to the very nature of chronic poverty, however, it is unlikely that the proportion of people in chronic poverty has declined at anything like the rates of poverty in general.

For instance, village-level research in Rajasthan, where headcount poverty has unambiguously declined, suggested that about 18% of the total population was poor both 25 years ago and in 2002. This figure ranged from 8% to 31 % across districts, and was highest among scheduled tribes, more than two-thirds of whom had stayed in poverty over the past 25 years.⁷

For Pakistan, a significant amount of analysis has been undertaken using one particular dataset.⁸ Different approaches to defining chronic poverty and the poverty line have led to a wide range of estimates of chronic poverty. The best all-

Table 7.1 Summary of poverty indicators for South Asia

Pakistan estimate of rural chronic poverty, based on mean income over five years, is 26% – this represents about 50% of households classified as poor in the first year of the survey, and about 6% of households classified as non-poor in the first year. Table 7.2 presents a summary of these different approaches and estimates, and includes another survey that is more recent, but also contains fewer households, fewer waves and is confined to a single province.

There are no panel data from which to determine the numbers of chronically poor in Sri Lanka. It is clear, however, that although per capita GDP passed the US$800 hurdle in 1999, poverty persists. The proportion of the population living on less than US$1/day, and the nutritionally ‘ultra poor’,⁹ both seem stable at just above 5% of the population.

The extent to which the 40% of Sri Lankans who survive on between US$1 and US$2/day are likely to be chronically poor is an empirical question, and further research is needed to understand the poverty dynamics of the ultra poor, poor and non-poor in Sri Lanka.

Who are the chronically poor in South Asia?

The chronic poor in South Asia are disproportionately made up of excluded minorities, including tribal peoples; people belonging to perceived low status castes; and casual and migrant labourers. Women and girls also tend to be particularly vulnerable to chronic poverty in the region. Many chronically poor live in persistently poor Indian states and/or less favoured or remote areas.

The working poor

Contrary to the common perception that the chronically poor are ‘unproductive’ – unable or unwilling to work – the working poor actually constitute a significant proportion of the chronically poor. The largest group of chronically poor people in rural India are casual agricultural labourers; cultivators, the second largest group. Most of the chronically poor are either landless or near-landless, and highly dependent on wages.Agricultural wages have been rising slowly in much of the sub-continent, and this is probably the best single explanation for the slow but steady reduction in the depth of consumption poverty. However, getting work does not always translate into exiting poverty. In agrarian economies with large casual labour markets, the number of days of work obtained in a given period, is almost as important as the wage level.

Migration is often part of a broader set of livelihood strategies employed by poor wage labourers. Chasing scarce, short-term, insecure, and low-paid wage labour from area to area, migrant labourers often find themselves in a constant battle to repay debt and maintain household consumption levels. In some cases this can result in people becoming more vulnerable to exploitative employment (see Box 7.1). Much migration for work undertaken by the poor in South Asia is this rural-rural, temporary and seasonal movement,¹¹ although migrants are also often among the urban chronically poor. This is not to say, however, that all migrants are chronically poor. For some, migration has proved to be an effective means of escaping poverty.

Excluded minorities

Excluded minorities, including ‘tribals’, people of ‘low’ caste and religious minorities, find it more difficult to marshal the necessary social, political and economic resources to progress, and are much more likely to experience longterm and absolute poverty. As touched upon in Chapter Two, both Scheduled Castes (SCs) and Scheduled Tribes ( STs ) are stigmatised groups, within which many suffer extreme discrimination although the harsh oppression associated with untouchability has been banned.¹²

In rural India, for example, a SC or ST household was more likely to be poor in both 1970–71 and 1981–82 than other caste households. Scheduled Caste women have one of the lowest levels of literacy of all groups in India – in the 1991 Census more than 80% rural SC women were found to be illiterate. STs have literacy rates of just 40%, compared to 54% national average, with

only a quarter of ST women being literate.¹³ This varies greatly from state to state, with female literacy ranging from about 88% to just 9% in 1991.¹⁴

While per capita incomes are lowest among SCs followed by STs, tribal status is more significant than caste status in determining poverty persistence.¹⁵ STs in India are often located in isolated areas where opportunities to diversify income earning strategies is low.

The chronic poverty dimension of tribal status is most pronounced in the context of social movements and conflict. Indigenous peoples of south-eastern Bangladesh, for example, have only recently emerged from years of struggle against Bengali in-migration cum colonisation. Agitation for separate states in parts of India has taken root partly in response to rising resentment within deprived regions and tribes.

Poor women, older women, disabled women and widows

Poor women feature prominently as a group of the chronically poor in South Asia. They are generally less educated (see Table 7.3), triply burdened¹⁶, less well connected and informed, and often unable to ensure that they benefit from husbands’ income.¹⁷ Gender divisions within labour markets restrict the employment opportunities for women, though the demand on women to work is strong within poor and chronically poor households.

The position of women is particularly vulnerable to continued poverty when they reach old age and/or are widowed and/or become disabled. In India, widows represent 6.5% of the total female population – 30 million in absolute terms, perhaps three times the number of underweight children.¹⁸ Property and inheritance laws are highly gender discriminatory across the South Asian region, and ignorance and misapplication of these laws often mean that women do not even enjoy the minimal protection that they can afford.¹⁹ In much of northern India and Pakistan, for example, strong patriarchal traditions of ownership and inheritance continue to dominate despite legal provisions to protect women’s ownership rights. In Nepal, recent constitutional changes that ensure equal property rights for women present a significant and positive opportunity for poor women and their children to avoid slipping further into deep, inescapable poverty.

Since women usually move to their husband’s village on marriage, they do not have strong support systems if they are widowed. Although not always the

tend to spend a large proportion of their

ments. Families facing chronic food insecurity are caught in a hunger trap. The inadequacy and uncertainty of their food supply make it difficult for them to take advantage of any development opportunities that might emerge.

improvement between 1986 to 2001 – the absolute numbers of stunted and wasted Pakistani children have grown.²²

Breadwinner illness is a major cause of the financial deterioration for poor households – almost one-fifth of all deterioration in Bangladesh, for example.²³ The costs are direct (medical fees and treatments) and indirect (lost wages or production, care, withdrawal of children from school, asset depletion and longterm indebtedness). Chronic diseases such as TB have particularly devastating results.²⁴ Severe or prolonged illness or accidents are more likely in very poor households. Clean water, and good household and community sanitation, are increasingly recognised as factors in determining not only the health of children but also of adults.²⁵

The despair caused by the combination of long term hunger, ill-health and poverty, responsibility for older people and other dependants, lack of employment opportunities or any hope in the future for children, further debilitates the chronically poor. Multiple deprivations and starvation are reported to have culminated in suicides by skilled powerloom weavers in India.²⁶ Such reports highlight the hopelessness and despair often experienced by the desperate, facing the prospect of chronic poverty.

Although hypertension and heart disease are commonly considered problems of the middle class, they also are significant problems for the long-term poor (Box 7.1). Studies warn about heart disease and diabetes reaching epidemic proportions in India.²⁷ The choices chronically poor people are forced to make in order to survive can be highly detrimental to their health. Some of these decisions may have high physical and psychological costs, such as heart attacks and high blood pressure.

Where are the chronically poor in South Asia?

Chronic poverty in South Asia has both macro and micro-level features. At a regional level, most indicators show a swathe of poverty cutting across eastern and southern Pakistan, central India, western Nepal, and northern and southeastern Bangladesh. Within this general ‘poverty tract’, however, there are pockets of improvement, lower levels of poverty and even relative prosperity – sometimes urban areas, sometimes areas

dependent on remittances or strong NGO programmes. Similarly there are pockets of deprivation in otherwise welloff regions – areas, both rural and urban, less-favoured by nature and/or man.

Most poor South Asians still live in rural areas, and it is likely that the proportion of chronic poor is greater in rural areas, given the greater opportunities in towns and cities. However, in India the proportion of severely poor people in rural and urban areas is similar at about 15%, indicating that urban chronic poverty may be greater than supposed.

In Bangladesh, spatial inequalities in human development are considerable, with the central and south-western regions doing relatively well (see Figure 7.2). However, modest reductions in spatial inequalities have occurred, during the late 1990s in particular. The northwest and southeast are beginning to catch up, based upon two main factors: a better-integrated national market, and decreased conflict. The construction of the Jamuna bridge – representing a massive public investment – helped to integrate long-neglected northern and western districts with the rest of the country, while the peace process in the Chittagong Hill Tracts removed some obstacles to improvement in that region.

There are also pockets of poverty in areas much smaller than districts, due to variations in agro-ecological vulnerability, or the presence of minority populations. And, as Bangladesh is characterised by the highest population density in the world,²⁸ even small pockets of severe distress can affect a very large number of people. Panel data for 1987–88 and 2000 indicate that 15% of households that had descended into poverty had experienced a shock related to a natural disaster, suggesting that poor geographic capital at the most local level played a role. Poverty rates are highest in extremely low-lying areas that are frequently flooded, including chars (river-islands that seasonally disappear; see Box 3.2), and in tribal areas where social and geographical disadvantage overlap.²⁹

In India, there is significant but incomplete overlap of areas with the highest poverty rates and those with the lowest human development indicators, and of poor regions, states and districts (see Figure 7.4). At the regional level, the marginality of central and eastern India is explained largely by adverse agrarian relations, and poverty has persisted in these regions despite a good endowment of natural resources and a relatively strong focus of Indian development planning on ‘backward areas’. State, district and rural indicators broadly follow this general regional sketch, with one or two exceptions. Urban indicators show a markedly different trend.

Over 70% of India’s poor reside in six states: Uttar Pradesh, Bihar, Madhya Pradesh, Maharashtra, West Bengal and Orissa.³⁰ In four of these states – Bihar, Orissa, Madhya Pradesh and Uttar Pradesh, plus Assam, persistently high levels of poverty in excess of 30% have occurred for several decades.³¹ As most central Indian states are the size of large countries – Uttar Pradesh would have the world’s sixth largest population if it were a country – numbers of people suffering persistent poverty and deprivation are huge.

In Assam, both income poverty and human development performance declined strongly in the 1990s, from already low levels. In the mid-1990s, 46% of rural households in the lowest expenditure class could not access two meals per day throughout the year, compared to an all India average of 15%.³²

At the micro-level, severe deprivation is remarkably concentrated in India. District-level multidimensional indices have been developed combining indicators of literacy and enrolment, infant mortality rate, agricultural productivity, and infrastructural development – low levels of which can reflect persistent deprivation.

Figure 7.3                Pakistani districts

Out of 379 districts in fifteen states, the same 52 to 60 districts are consistently identified as the most deprived, despite computing nine different indices with different combinations of indicators and methodologies (see Map 4 in Figure 7.4). 80% of the districts identified are located in one of the five states with high persistence of poverty.³⁵

% of the most deprived districts according to the multidimensional indices (including one of the seven districts suffering extreme deprivation) are in Rajasthan. This north-western state is something of an anomaly in the pattern. Poverty rates are significantly below the all-India average, and have been declining much faster than average in the late 1990s. Rajasthan does not show up at all on the National Sample Survey list of regions (clusters of districts) with the highest rates of poverty and severe poverty (see Figure 7.4, Map 2). At the same time, the state’s HDI is significantly below the all-India average, although in the late 1990s some improvement in this index has also been noted, in part due to enormous progress on education indicators. Yet it contains one-fifth of the most deprived districts in India.

Comparing Figure 7.4 Maps 2 and 4 , it is clear that even within the core five persistently poor states, overlap is sketchy, and that there are several regions that the National Sample Survey identifies as poorest that do not contain any of the most deprived districts. As has been found in Vietnam,³⁶ there is not the expected near-universal or exact correspondence between changing levels of income poverty and other dimensions of deprivation. The reasons for this are likely to relate to differing patterns of economic growth and socio-economic inequality.

Many remote rural areas in India are largely populated by scheduled tribes, who face extreme marginalisation and discrimination. In general, two types of area are viewed as less-favoured on the basis of agro-ecological and socioeconomic conditions. These areas also exist in less poor states.

^. First, large tracts of dryland characterised by frequent crop failure and sporadic opportunities for employment.

^. Second, forested regions, especially in hilly regions with predominance of tribal populations, with limited access to natural resources, information and markets.³⁷

These areas are not only persistently income poor, but are generally much less well-endowed with human capabilities. Tribal populations living in forested areas affected by consecutive years of drought, such as south-western Madhya Pradesh, face extreme deprivation.³⁸ Geography is only part of the reason why access to resources may be limited. See Box 7.2 for a discussion of the effects of some government lease oriented policies on traditional access to resources in Orrisa.

There is significant variation in the degree to which Indian states have mitigated the effects of drought. On the face of it, drought-related chronic poverty is most likely in arid areas in poorly governed states. However, many dryland populations have been able to develop coping strategies to facilitate their resilience to drought, including groundwater development, economic diversification with infrastructural development, drought relief safety nets, and migration. The latter is especially significant. Forest-based regions have few of these possibilities. Migration is more likely to be from distress, since regions of economic growth are often further away, and markets function less well so that investments at home have less effect.³⁹

AP (Andhra Pradesh); KA (Karnataka); MA (Maharashtra); MP (Madhya Pradesh); RA (Rajasthan); TN (Tamil Nadu); UP (Uttar Pradesh);

WB (West Bengal).

Urban poverty and hunger, particularly urban hunger, do not conform to the broad notion that persistent and absolute poverty is concentrated in central and north-eastern India. The southern states of Karnataka, Andhra Pradesh and Tamil Nadu have above average rates of urban poverty and urban hunger, while Kerala – India’s showcase state in terms of high levels of human development – has the highest and third highest urban and rural hunger rates in

India.

Andhra Pradesh suffers a low and declining HDI in contrast to its low levels of income poverty. This may suggest that growth and public investment have been less than pro-poor, with particularly adverse effects on the urban population. On the other hand, Karnataka, and in particular Kerala and Tamil Nadu have strong HDIs and governance is relatively pro-poor. Urban poverty is clearly a specific and complex problem.

In Pakistan, available evidence suggests that chronic poverty exists in several areas, and is harshest where ecological and social deprivation overlap (see Figure 7.3). First are the harsh environments – the mountainous Northern Areas, and arid parts of Balochistan and Sindh in the west and south. Second, areas dominated by oppressive tribal and/or feudal agrarian and gender