Why Don’t ‘the Poor’ Make Common Cause?
The Importance of Subgroups
Anirudh Krishna
Associate Professor of Public Policy and Political Science
DukeUniversity
Box 90245
Durham, NC27708-0245
(919) 613-7337
(919) 681-8288 (fax)
Abstract
Analyses that regard ‘the poor’ as a sociological category need to take account of recent studies quantifying the extent of flux within these ranks. Frequent movements into and out of poverty regularly refresh the pool of the poor. Large numbers of poor people were not born poor; they have descended into poverty, some quite recently. Concurrently, many formerly poor people have escaped from poverty.Distinct subgroups are defined by these divergent trajectories. Members of different subgroups have diverse economic needs, political interests and mobilisation potential, making cohesive action as a political force unlikely (and certainly uncertain) among all of ‘the poor’.Policies to assist poor people will be more effective, and political analysis will yield more fruitful results,if instead of working with any generic category of ‘the poor’ heed is taken of subgroup-specific experiences and demands.
1. Introduction: Why Not a Party or a Politics of the Poor?
Where the poor constitute a majority or near-majority of the population, why don’t they vote themselves to power in democracies? In countries such as Madagascar, Mozambique, Mali, Guatemala, Honduras, Kenya and Bangladesh, where the poor constitute, respectively, 71 percent, 70 percent, 64 percent, 56 percent, 53 percent, 52 percent, and 50 percent of the population, why don’t parties of the poor emerge and take power democratically? Even in countries such as India, Philippines, and Ecuador, where the poor form a smaller but still sizeable part of the population – 29 percent, 37 percent and 35 percent, respectively – why are the politics of poverty not more emphatic, potent and visible?[1]
Lower political participation by poorer people can provide a possible explanation. Likened by Marx to sacks of potatoes, the rural poor have not been considered particularly active politically (Bates, 1981). Empirical studies have repeatedly affirmed lower participation rates among the poor (for example,Verba, Nie and Kim, 1978; Rosenstone and Hansen, 1993; Jackman and Miller, 1995; and Lijphart, 1997). A ‘culture of poverty’ is claimed, reflecting apathy and submission among the poor (Lewis, 1963).
Yet, these explanations are hardly sufficient to justify why large majorities of people, more than two-thirds of the population in some cases, are unwilling or unable to act collectively. Participation rates may be lower among poorer people, but the sheer weight of numbers can handily compensate for this difference, reported to be no more than a few percentage points. Recent evidence also shows that participation rates are not uniformly low among all of the poor. In many cases poorer people participate as actively as others (Yadav, 1999; Bratton and Mattes, 2001; Mattes, et al., 2003; Krishna, 2006; Krishna, forthcoming), and some sections of the poor can actually have higher-than-average participation rates, as shown below.
In contexts where they are quite numerous – which includes vast swathes of South and Southeast Asia, Sub-Saharan Africa, and segments of Central and South America, or nearly one-half of all countries in the world – the rural poor should present to political entrepreneurs a natural constituency for effective organisation. As inequality is rising in many countries (Wade, 2004; World Bank, 2006), and as older forms of social organisation and their associated norms and customs are eroding (Griffin, 2000; UNDP, 2000), the organisational potential of the rural poor should be growing, rendering them more attractive as a constituency for political organisers. Whether to restore a ‘subsistence ethic’ (Scott, 1976) or to obtain better deals from markets and state bureaucracies (Popkin, 1979), movements of the rural poor should therefore be on the rise.
So why are efforts to organise the poor so infrequent, scattered, and localized? Divisions among the poor on account of caste, ethnicity, and religion have been advanced as a possible explanation for lack of organisation (for example, by Burnell, 1995; Alesina, et al., 1999; Bates, 1999; Good, 1999; Keefer and Khemani, 2003; and Varshney, 2005), but while answering one question these explanations evade a more fundamental one: Why are caste or ethnicity so much more commonly the currency of political organisation – even where poverty is a more widely shared feature? Why do political parties in developing countries more frequently exploit cleavages drawn along ascriptive lines and less often assemble broad-based coalitions of the poor?
In this paper, I present an additional explanation, supplementing the ones provided earlier. ‘The poor’ does not constitute a valid category for analysis or action: it is no more than an article of speech, I will contend.
Recent studies showthat significant differences in identities and material interests exist across distinct subgroups of poor people. There are those who have fallen into poverty recently, others who are on the verge of escaping from poverty, and still others who have remained persistently poor. In Section 2, I present evidence from recent studies that quantify the extent of movements into and out of poverty. Large numbers, and in many contexts, the majority, of those who are poor at the present time were not poor some years ago. Conversely, large numbers of formerly poor people have escaped from poverty, and others, still poor but upwardly mobile, are making their ways out fromthis statistical pool.
Different reasons are associated, respectively, with escaping poverty and falling into poverty (Section 3). As a result, different needs are experienced and different demands are expressed by members of different subgroups. Those who have recently fallen into poverty are most directly encumbered by one set of reasons, and they tend to demand related public policies. Others, on the verge of escaping from poverty, are assisted by a different set of reasons. They see a different set of needs and opportunities. The persistently poor constitute yet another subgroup. Neither recently fallen into poverty, nor experiencing any significant upward mobility, members of this subgroup face a different opportunity set; they tend to make a different set of demands upon the state.
Different experiences, different identities and different material interests tend to make collective action uncertain among all subgroups of the poor. Evidence collected in 36 villages of Andhra Pradesh, Indiaprovides some initial support for this proposition, showing how members of different subgroups demand very different things from the state.
Coming together under a common umbrella of action is hardly automatic. Looking within separate subgroups provides a better starting point for policy design, political analysis, and coalition building. Some consequences of this categorical advance are examined in Sections 4 and 5.
2. Not a Rising nor a Falling but a Rising-Falling Tide
Analysts referring to the poor implicitly assume a stable and homogeneous category.[2]However, ‘the poor’ is a very approximate and possibly a misleading category of analysis. For a category to be robust and useful for analysis it must have sufficient discriminating power (Sartori, 1970: 1039). It must be stable, with clear boundaries, and have defining properties shared by all members (Collier and Mahon, 1993: 845). ‘The poor’ fulfills none of these conditions. Considerable differences exist within the ranks of the poor, and considerable similarities are observed across the categorical dividing line. ‘The poor’ is also an ephemeral category with highly porous boundaries. Many who form part of the poor at a previous point in time are not included among them at a later time, and many others who were not poor at a previous time become part of the poor in the future.
Considering the poor as a fixed category is therefore like using a freeze-frame snapshot to depict a vast churning tide. It captures the peaks and troughs of the moment but it is instantly reconfigured by movement. The moving picture rather than a snapshot is both more truthful and more productive in terms of analysis, but it provides no support for any well-defined category of ‘the poor’.
Table 1 presents results from a geographically diverse selection of recent studies that examined poverty in dynamic context. These studies consider different sample sizes, ranging from a small group of 347 households in a few communities of Egypt to over 6,000 households in one part of India (respectively, Haddad and Ahmed, 2003; and Krishna, 2004). Statistically representative samples for entire countries are included (Deininger and Okidi, 2003; Bhide and Mehta, 2004), alongside studies of particular regions or groups of communities (Sen, 2003; Krishna, et al., 2004). Different periods of time are considered, ranging from a short span of three years to long periods of 25 years. Commonly, however, all studies illustrate the extent to which there is flux within the ranks of the poor.
Considering the magnitude of these movements both into and out of poverty has the effect of substantially changing our imagery of the poor.
-- Table 1 about here --
The first row of Table 1 shows that over the 13-year period, 1987-2000, 26 percent of a panel of 379 Bangladeshi households considered by Sen (2003) escaped from poverty (Column 5). Simultaneously, however, another 18 percent of households fell into poverty (Column 6). Movements out of and into poverty were both large. Fifty seven percent of all households were poor at the start of this 13-year period (Column 7), and 49 percent were poor at the end of this period (Column 8). However, not all those who were poor at the beginning of this period were also poor at the end. In fact, 46 percent of those who were poor at the beginning were not poor by the end of this period (Column 9). Conversely, 37 percent of those who were poor at the end had not been poor at the beginning of this period (Column 10). Because movements out of poverty were large (26 percent) and movements into poverty were also large (18 percent), the composition of ‘the poor’ changed considerably.
These results are hardly confined to Bangladesh. All studies in the sample reported in Table 1 present a similar account of flux among the poor. Other studies not reported here also bear out a similar conclusion.[3] New poverty is being created even as some old poverty is destroyed. Large numbers of people are entering poverty even as large numbers escape from poverty.
Take, for example, the study by Krishna et al. (2004) of households residing in 20 Western Kenyan villages. Eighteen percent of these households escaped from poverty in the 25-year period examined by this study. At the same time, another 19 percent of households fell into poverty. A total of 55 percent of the poor at the beginning of this study period were not poor any longer by the end of this period (Column 9). Conversely, 56 percent – the majority – of those who were poor at the end were not part of ‘the poor’ at the beginning of this period (Column 10).
Studies that consider a shorter time horizon also have similar results to show. A study of 1,171 households in KwaZulu-Natal, South Africa, shows that over a five-year period, 1993-1998, 10 percent of households moved upward out of poverty (Carter and May, 2001). During the same time period, however, another 25 percent of households fell into poverty. The ranks of the poor were considerably refreshed: 60 percent of all the poor at the end of this five-year period consisted of the newly impoverished. Results from a nationally representative sample of 1,300 Ugandan households similarly show that of all households who were poor in the starting year (1992), 54 percent – the majority – were no longer in poverty by the ending year, 2000 (Deininger and Okidi, 2003).
In Asia, Africa and Latin America, everywhere household and individual poverty has been examined over time, similar results have emerged:Movement reconstitutes the profile of people in poverty. No matter how long or short is the time period studied – or what measure or threshold of poverty is employed[4] – the results are the same: Escapes from poverty occur concurrently with descents into poverty. A sluggish pace of net poverty reduction does not occur because there is no movement out of poverty. It is a resultant of two large and frequently offsetting trends.
These movements, into and out of poverty, are not marginal or temporary events.[5] Thus, people do not fall into poverty only to escape in a later period, nor indeed are these results confined to borderline households fluctuating around the poverty line.
Only one-third of households that fell into poverty during the 15-year period, 1979-1994, were able to escape from poverty over the next 10 years in Uganda. The remaining two-thirds who had become poor 10 years ago were still poor when investigations were conducted in the summer of 2004. Meanwhile, an additional 11 percent of households fell into poverty, further refreshing the ranks of the poor (Krishna, et al., 2006a). Many who were formerly poor have climbed a considerable distance out of poverty. Many others who were previously well-to-do have fallen deeply into poverty. A total of 368 of all 2,631 households in these 36 Ugandan communities (14 percent) came out of poverty over the ten-year period, 1994-2004. Among these households, most of whom could barely afford to purchase food and clothes 10 years ago, 26 percent now possess land of their own and another 21 percent have advanced even further, owning concrete houses and businesses in addition to farm lands. Concurrently, another 325 households (12.5 percent) fell into poverty, and of these freshly impoverished households as many as 24 percent can no longer afford to buy clothes and food; another 29 percent have pulled their children out of schools. Investigations in other countries also show how descents and escapes are mostly not temporary, marginal or fringe occurrences.[6]
It makes little sense thus to speak of the poor as a stable, consistent or homogeneous category. Not all who are poor today have always remained poor. Neither have all who are not poor today always kept safely away from poverty.
Distinct subgroups are given rise by these movements into and out of poverty. Different interests and identitiesare associated with each separate subgroup. Members of different subgroups commonly wish to get out of poverty, it is true, and to that extent there is, indeed, a shared aspiration. But the means that each subgroup regards as most relevant – the howof combating poverty – differs notably among members of different subgroups.
3. Demands from the State by Different Subgroups
Escaping poverty and falling into poverty are not symmetric in terms of reasons. One set of reasons is associated with movements upward, out of poverty, while another and different set of reasons is associated with movements downward, into poverty. Two different sets of public policies are required, therefore: one set to address and reinforce the reasons associated with escaping poverty, and another set of policies to block or subdue the reasons for descent. Different subgroups of the poor are differently attracted toward these different policy sets.
Ill health and healthcare expenses constitute commonly the single largest reason associated with descents into poverty. Studies show how for as many as 73 percent of the newly impoverished in Western Kenya, 77 percent in Uganda, 67 percent in Peru, and 88 percent in Gujarat (India), ill-health and healthcare expenses constituted a principal reason for descent. Large majorities of households that fell into poverty in other countries have also suffered from ill health and high healthcare expenses. Evidence from Cambodia, Ethiopia, Haiti, Kenya, Peru, Sierra Leone, Senegal and Vietnam points unambiguously to the deleterious effects of healthcare costs upon households’ welfare.[7] Health-related factors operate commonly across all regions to drive households into poverty. Households in China facing one serious health situation suffer average income drops of 17 percent. When two or more health incidents are experienced back-to-back a household’s slide into poverty becomes steeper and more assured (Gan, et al., 2005). More than half of all personal bankruptcies in the United States are attributable to medical costs (Himmelstein, et al., 2005).
‘Rises in out-of-pocket costs for public and private healthcare services are driving many families into poverty and increasing the poverty of those who are already poor. The magnitude of this situation – known as “the medical poverty trap” – has been shown by national household surveys and participatory poverty alleviation studies’ (Whitehead, et al., 2001: 833). Data from extensive household surveys undertaken in 59 countries of Asia, Eastern Europe, and South America shows that catastrophic payments on healthcare – those that far exceed households’ capacity to pay – are unfortunately on the rise and ‘common in middle-income countries, countries in transition, and in several low income countries. This negative impact of health systems on households that can lead to impoverishment has long been ignored’ (Xu, et al., 2003: 115). A very large part of debt incurred by poor families in India and elsewhere arises on account of large healthcare expenses (Dilip and Duggal, 2002). In rural Vietnam, 60 percent of poor households were found to be in debt, and more than one-third of these households cited medical expenses as the main reason (Ensor and San, 1996).