POLITICAL PARTICIPATION, REPRESENTATION AND THE URBAN POOR: FINDINGS FROM RESEARCH IN DELHI. John Harriss

Transcription

1 POLITICAL PARTICIPATION, REPRESENTATION AND THE URBAN POOR: FINDINGS FROM RESEARCH IN DELHI John Harriss Introduction: Research Questions It is widely believed that an historic shift is taking place in forms of political representation available to the poor, from the classic patterns of the earlier 20 th Century, based on social relations forged in workplaces, organised in trades unions and other mass organisations linked to programmatic political parties, and concerned with the achievement of social and economic rights, to new patterns based on other sorts of social networks and goals, and involving, for example, social movements rather than unions, voluntary associations rather than political parties, and local rather than national concerns. Political parties, though they were once linked to society through ideology and mass organisations, are increasingly only the loosely organised followings of populist and charismatic leaders and/or rely on marketing themselves to voters through mass media, tapping popular sentiments through public opinion polls and focus groups (see Roberts 2002: 20). It seems possible, as Castells suggests in his three volume study of the Information Age ( ), that as a result of these shifts it is becoming increasingly difficult for many segments of the poor to build organisations and participate in effective channels of popular representation capable of exerting political pressure. He argues that the failure of proactive movements and politics (for example, the labour movement, political parties) to counter economic exploitation, cultural domination, and political oppression, had left people [by the 1980s] with no other choice than either to surrender or to react on the basis of the most immediate source of self-recognition and autonomous organization: their locality. But, he said, while urban social movements do address the real issues of our time, they do so on neither the scale nor on the terms that are adequate to the task (Vol II, 1997: 61). State responsiveness to social claims, and the ability of existing structures of representation to provide poorer social groups influence over policy always historically limited in many low and middle income countries may, therefore, be on

2 the decline despite the recent wave of democratisation. It is ironic that, alongside this wave, there should be evidence of a crisis of popular representation in many of the low and middle income countries 1, and that poorer social groups appear to have limited capacity to present a reform agenda that addresses issues of basic rights and ensures livelihoods. Somewhat oddly, pessimism about the prospects for progressive programmatic politics in the party and union arenas stands alongside optimism amongst some researchers and development policy actors for the success of direct popular participation. 2. Analysts point to a flowering of new forms of communitybased associations and see these not only as replacing parties and unions in representing popular interests, but also as being more responsive to their constituencies and autonomous from external influence and control. The World Bank, for example, has argued that In most societies, democratic or not, citizens seek representation of their interests beyond the ballot as taxpayers, as users of public services, and increasingly as clients or members of NGOs and voluntary associations. Against a backdrop of competing social demands, rising expectations and variable government performance, these expressions of voice and participation are on the rise (World Bank 1997, p.113). The research reported in this paper treated this proposition as a hypothesis - to be tested against evidence on how people generally, and poorer people in particular, are trying to pursue their social and economic rights and to tackle collective social problems, and on their relationships with political parties and with formal and informal associations. The paper is based on data from a sample survey of 1401 of Delhi s citizens, conducted in , that was intended to describe and to explain patterns of political participation, focussing in particular on the ways in which poorer social groups organise, obtain political representation and try to solve collective social problems. The survey was part of a wider comparative study of Rights, Representation and the Poor (see Houtzager et al. 2002), carried out according as far as possible, given the widely differing availability of basic data to a common methodology, in Brazil (Sao Paulo) and Mexico (Mexico City) as well as in India (Delhi, Bangalore and Coimbatore). The methodology of the Delhi survey is described in an appendix to this paper 4 ; and it was designed so as to address a series of questions about the politics of the present. Is it the case, in Delhi (and in Bangalore, in Sao Paulo and Mexico City), that collective action is now based on collective 2

3 actors such as associational networks, perhaps facilitated by NGOs, capable of aggregating local groups? Is there evidence that lifespace social networks (neighbourhood-based, religious etc) are now producing the most politically efficacious collective actors? Or is there evidence, on the other hand, either of reliance on self-provisioning rather than political claim-making, or of persisting reliance on clientelistic relations? Where do political parties come into the picture? The generalised account of global trends that served to frame our research on Rights, Representation and the Poor is, at best, partially applicable to India where trades unions have only ever had a membership of a small fraction of the labour force and where political parties have according to the views of most political scientists long suffered from being only very weakly institutionalised. They may indeed be better seen as the followings of particular leaders who have been all to ready to sacrifice programme to the compulsions of electoral politics in order to secure the loaves of office for themselves. Electoral politics in the 1990s have been marked by what Yogendra Yadav (1996) has described as the second democratic upsurge (the first having followed on India s independence from colonial rule). The democratic upsurge of the 1990s has brought political leaders from some of the historically lower and more backward castes to the fore, and it is associated with a pattern of electoral participation that is the reverse of that which is commonly found, because participation rates are higher amongst poorer, less well-educated and lower caste/class people. Poor people are in this way apparently somewhat more strongly represented politically than they were previously. But if no longer mobilised as much as they were through clientelistic relations, poor people are mobilised by the populist appeals of charismatic political leaders whose performance, in office, rarely offers much hope of social and economic transformation. In India, as is true elsewhere in the world, there is at least some expectation that the needs and interests of poorer people may now be met more effectively through the new politics of social movements, rather than those of the old trades unions, and of the burgeoning NGOs and of associations in civil society, rather than those of the political parties. This paper in fact reaches what may seem to be pessimistic conclusions about the extent to which these trends are actually taking effect in Delhi, finding in line with the wider trends depicted in the idea of the second democratic upsurge - that poor people, in particular, most commonly seek to represent themselves and to tackle their problems through political 3

4 parties. There is little indication, in fact, of the existence of any other significant collective actors in the worlds of poor people. We also find, contrary to what has been argued, say, in regard to Latin America (e.g, Roberts 2002), that party political mobilisation does reflect major social cleavages. Conceptualising Political Participation Analytically the paper is organised around the problem of political participation. Our survey of Delhi s citizens enables us to assess political participation in various ways: by describing people s involvement in attempts to solve collective problems; through measures of support for political parties, of electoral participation, and of participation in political activities of different kinds; and through measures of associational practices. Political participation is defined, generally, in the words of Verba, Schlozman and Brady as: activity that is intended to or has the consequence of affecting, either directly or indirectly, government action (1985: 9). These authors also recognise, however, in their classic work on the United States, that there are many social activities that are non-political in themselves but which do have political consequences (the overtly non-political is also political). They argue, therefore, that political participation should be understood as including much...non-electoral activity (that) takes place outside official channels. This includes a vast number of official contacts and communications with government officials as well as a large volume of informal, problem-solving activity among friends and neighbors in local communities (1985: 8, emphasis mine, JH). Rosenstone and Hansen, similarly, in their more recent work on political participation in America offer what they describe as an expansive definition : Political participation is action directed explicitly toward influencing the distribution of social goods and values (1993: 4). The understanding of political participation in this paper follows this wider definition, and it is concerned with that large volume of informal problem-solving activity to which Verba, Schlozman and Brady refer, as well as activity intended directly to affect government action. 4

5 Delhi and its politics Here Delhi refers to the National Capital Territory, which constitutes a separate political unit governed by its own Vidhan Sabha or Legislative Assembly, with 70 constituencies. At the time that this study was carried out the Congress was in office in the Delhi government, and the Congress was successful (against the general trend at the time) in retaining office in the elections of April-May The BJP has held office in Delhi, and the Delhi BJP is a relatively well organised and a powerful political force. The left parties have never had much of a presence in the capital. The population of Delhi has grown very rapidly over the last thirty years, with large in-flows of migrants from the nearby states of Punjab, Haryana, Rajasthan, Bihar and (especially) UP 5. Although it was originally and is still primarily an administrative city, Delhi has also developed as a centre of service industries (including, latterly, some development in the IT sector), and it has large numbers of small and medium as well as some large-scale industries. According to the Economic Census of per cent of Delhi workers were in manufacturing and 23 per cent in trade and commerce. So though the city did not grow up, as did Kolkata and Mumbai, as trading and industrial cities, it is nonetheless fairly described as an industrial city. Important developments in the recent past have been the passing of legislation to move out polluting industries from residential areas, which has had the effect of closing down some small and medium industrial units. Industrial restructuring since the inception of India s economic reforms in 1991 has also led to factory closures and to the loss of industrial jobs. In June 2003, according to information put out by the Delhi government (on its web-site) there were as many as one million unemployed workers in the capital territory. In regard to local politics a significant development under the Congress government over the past five years has been the establishment of the Bhagidari Scheme. This means something like Partnership Scheme and it is said to be a mechanism for an active, effective and target-oriented citizen-government partnership. Involving partnerships between Residents Welfare Associations and Market Traders 5

6 Associations and the Delhi Government, the Bhagidari Scheme is intended to develop joint ownership by citizens and the government of the change process in the city. In practice it involves Residents Welfare Associations, in particular, in such tasks as securing payment and collection of water bills, electricity meter reading, house tax collection, the supervision of sanitation services, and the maintenance of community parks and community halls. It remains to be seen whether the scheme really does represent partnership or whether it rather involves the off-loading of tasks by city government onto local associations. But the scheme has certainly given particular prominence to Residents Welfare Associations which are found, however, mainly in Planned Colonies and other higher income areas, and not in Jhuggi Jhopris (slum clusters) and other poor parts of the city. It is the intention of the Delhi Government to extend the scheme to these areas, but that had certainly not happened at the time that our research was being conducted. There is no doubt that the Delhi Government is inspired by the example of city government in Brazil, and it is intended to introduce a scheme of participatory budgeting. PATTERNS OF POLITICAL PARTICIPATION AND THEIR DETERMINANTS PROBLEM SOLVING Problem Definition The first part of the survey asked questions about the extent to which people considered a range of public issues as a problem, both for themselves and for the city or for the country as a whole. The results were as follows: Table 1: Assessments of Collective Problems % of sample defining as a big problem + one of the biggest problems Problem For themselves For the city/ country All Poor** All Poor** Air Pollution Basic needs Violence & Crime Public services* n/a Medical care *total for water, sanitations, roads and combinations of these, as being the services with which people say they have most problems 6

7 ** In this and subsequent tables this second set of figures refers to the 603 respondents (43% of the total sample) whose average household incomes are said to be up to Rs6000 per month, and who can therefore be considered to be the poorer people 6. Note that the comparison of the two sets of figures those for the whole sample and those for the poorer sections - systematically under-states the difference between the poor and the rest, because of the inclusion of the poorer sections also in the figures for the whole sample. There is often a clear difference between the figures for the two sets of observations, and it is important to note that this difference represents an under-statement. There was little variation between different groups of respondents, though as can be seen poorer people identified basic needs, violence and crime, public services and medical care as important problems, especially for themselves, to a greater extent than others. It was then very striking that most people attributed responsibility for action on these problems to government (whether local or national was not specified in the question), suggesting rather strongly as Neera Chandhoke has argued that people continue to repose hope in the state, rather than on action in civil society : Table 2: Responsibility for Problems % of sample attributing responsibility for action to: Air Pollutio n Basic needs Violenc e & crime Public services Medical care Governmen t Individual s Association s Marke t All Poo r All Poo r All Poo r All Poo r <1 <1 < <1 < <1 2 1 The evidence from the survey does not suggest that there is very much difference between people, whether in terms of gender and age, religion, their residential status (owner occupier or tenant), their income, education or labour market status, in the 7

8 attribution of responsibility. As might be expected there may be a slight tendency for wealthier and better educated people to attribute responsibility to individuals; and there is also a slight tendency both for those who are sympathetic to the BJP and for those who are active in associations to attribute responsibility to individuals, but the variations shown up in the data from the survey are all very small and cannot be considered statistically significant. How then do people in Delhi go about trying to tackle such social problems? Problem-Solving The survey distinguished between direct approach to government, taking legal action, going through political parties, or through big men, taking out demonstrations or petititions, or self-help, as approaches to problem-solving 7. The attempt was in made in the way the question was asked to distinguish between going to a political party as an organisation and going to an influential person associated with a particular political party, but it is to be expected that there is some overlap in the responses to the question 8. The general findings of the survey are these: Table 3: Problem Solving (%) Proportion of the sample reporting different channels Channel All Frequency* Poor Government Legal action Political Party Big Man Demonstration Self-Provisioning Other 5 Any Channel One Channel only More than one Channel * Frequency refers to the proportion of respondents who have used the Channel on more than one occasion These data suggest that the citizens of Delhi are quite active problem solvers, certainly by comparison with those of Sao Paulo, and by comparison with those of Bangalore 9. The possibility reflected in their generally higher scores in this table 8

9 (noting that the difference between the poor and the rest is under-stated in the comparison here) that poorer people in Delhi are more active problem-solvers, except as might be expected through the channels of the law and of self-help, and that they are somewhat more likely to try to solve problems by pursuing more than one channel, is striking 10. It is then of considerable interest to ask whether people undertook problem-solving action alone, or with the representatives of an organisation (as might be expected if NGOs or civil society organisations are involved in a major way in assisting people to represent themselves) or with other individuals such as acquaintances or family members. The last possibility is shown in the following table under the general heading of friends : Table 5: Mode of Problem-Solving By Channel (%) Channel Friend * Organisation Alone All Poor All Poor All Poor Government Legal action Pol Party Big Man Demonstration Self-Provisioning *here friend = acquaintance or family member; org = with a representative of an organisation It is very striking that problem solving is undertaken principally with acquaintances or family members and, it turns out, these are neighbours, in particular. There is little organisational involvement in any problem solving activity, saving for demonstration; and it is quite unusual for people to attempt any kind of problem solving alone, except in the (small in number) cases of formal legal action. In other words there appears to be rather a high level of collective action in Delhi - at least in the sense of people getting together to support each other in trying to tackle collective problems - but this does not involve much in the way of organisation, at all. Table 6: Social Support in Problem-Solving By Channel (%) Channel Family Neighbour Other All Poor All Poor All Poor 9

10 Government Legal action Pol Party Big Man Demonstration Self-Provisioning It is clear that people especially poorer people - most commonly undertake problemsolving together with their neighbours. In those cases where people did cite the support of an organisation of some kind, those most frequently mentioned were neighbourhood associations (130 citations), followed by political parties (30 citations), then religious groups (12), unions (11), others (5) and lastly NGOs (cited only 5 times in total). The near-invisibility of NGOs in the ways in which people go about problem-solving, in view of popular arguments about the role and potential of these organisations, is especially striking. Who Are The Active Problem Solvers? 11 The results of logistic regression analysis (see Appendix 3) suggest the following conclusions: (i) There is a strong neighbourhood effect. Those living in jhuggi jopris are very much more likely than the residents of other types of colonies to be active problem solvers. (ii) Women in general, and all those who are outside the labour force (inactives) are much less likely to be active in problem-solving than are others. (iii)those without education are less likely to be active problem solvers. (iv) Richer people are less likely to be active problem solvers (though the income effect is small). (v) Those who are permanent workers are more active in problem-solving than are casual workers. Bearing these major analytical conclusions in mind, let us look in more detail at the empirical findings. In general, scrutiny of patterns of problem-solving in relation to basic demographic characteristics of Delhi citizens does not show up very marked 10

11 variations. On the whole it is older men who are home-owners and long term residents of Delhi who are those most likely to be active in problem solving (perhaps suggesting that engagement in problem-solving draws on social capital deriving from local social networks built up over longish periods of time, and from which short-term residents and tenants may be excluded). The possibility noted earlier, that poorer people are generally more active problem solvers, is confirmed; and there are also some suggestions in the data that the poorer people, disadvantaged in terms of income are relatively more inclined than others to problem solving through political parties and participation in demonstrations, while those who are disadvantaged by sex (women) or by religion (Muslims) show a relatively stronger tendency to try to problem solve through approach to big men with possible implications in terms of dependency. With regard to gender (see Table A1) the differences between the sample of the poorer citizens and that of the population as a whole are not significant. In both cases, save only for the big man channel where women are as active (or more active even, amongst the poor) as men - men are considerably more active (50 per cent or more) than women. There is little consistent difference between age groups. The youngest people (aged between 18 and 25) are generally those least likely to be active in problem solving, and the oldest the most likely. Older people, particularly amongst the poorer sections of the Delhi population, are especially more active than younger people in problem solving through approach to government or to political parties. There is a statistically significant correlation between the number of years that citizens have been resident in Delhi, and their involvement in problem solving activity, both in the population as a whole and amongst the poorer sections. Home owners, too, are much more likely to be politically active than are tenants, both in the population as a whole, and amongst the poorer sections, save in regard only to participation in demonstrations (possibly reflecting recent political mobilisation, led by the former Prime Minister, V.P.Singh, over homelessness in Delhi). 11

12 In the Delhi population as a whole there is not a great deal of difference between the members of different religious groups, in terms of problem solving activity (see Table A2). Muslims may be less inclined than Hindus and Sikhs to try to tackle problems by going directly to government, and more inclined than Hindus to go to big men though not as much so as Sikhs (who, however, constitute only a small share of the sample population). But the fact that poorer Sikhs do not go to big men at all indicates that the kind of big men to whom wealthier Sikhs may go for help in solving problems are in a higher social class than those from whom most Hindus and Muslims are able to seek help. Amongst the poorer sections similar trends are found to those observed in the population as a whole, though it appears that the poorer Muslims are rather less inclined than others, including wealthier Muslims, to go to political parties or to get involved in demonstrations. There may be a hint here of a degree of alienation from political parties and regular politics amongst poorer Muslims. The most important findings about the significance of basic demographic factors the salience of gender, age, length of residence and (less clearly marked though it is) of religion seem to reflect the persistence in Delhi society of hierarchical values. Socio-Economic Status and Problem Solving There is often a correlation between income and education; and this is the case in Delhi, too. There is a correlation between continuous variables for household income and years of education that is significant at the 1 per cent level (Spearman coefficient; and see Table A3). This general correlation has to be borne in mind as we examine the relations between education and income, and problem solving, separately. Table 7: Problem Solving By Channel & By Education Level Completed (%) Channel None Primary Incomplete Primary Complete Secondary Incomplete Secondary Complete* University All Poor All Poor All Poor All Poor All Poor All Poor Freq. In sample Government Legal action Pol Party Big Man Demonstration

13 Self-Prov Any Channel x 1 channel x >1 Channel *Secondary schooling completed (10 [=SSLC] completed+study to 11 or 12). The most striking finding is that at most levels of education, but especially at the levels of primary schooling completed, or secondary schooling completed, and via most problem solving channels, citizens in the poorer sections are more active problem solvers than the general population. Poorer people with these levels of education are notably more active in problem solving by means of approach to government, or to political parties or through participation in demonstrations. It is striking that those who are most active in problem solving amongst the poorer sections, via all channels (except for legal action), are those who have only completed their primary education. The more educated (above primary) amongst the poorer sections, however, are not notably more active in problem solving by approach to big men; and they are at least comparably active in self-help as those with the same education in the general population. In other words it appears that being poor but relatively well-educated quite strongly disposes people to being active in problem solving, especially through approach to government or through demonstrations, and that education amongst the poor is associated with less reliance upon the intermediation of big men. It also appears possible that those at lower levels of education, both in the general population and in the poorer sections are more inclined than those with tertiary education, to try to solve problems through political parties. Table 8: Problem Solving By Channel & By Household Income, Monthly (%) Channel < Rs > Freq in sample Government Legal action Pol Party Big Man Demonstration Self-Prov Any Channel One Channel More than One Channel

14 It seems quite clear that except when it comes to self-help the poorer people in Delhi are more active in problem solving than wealthier people. The gaps between those with household incomes of Rs and more on the one hand, and those with incomes of up to Rs 4000 on the other are most clearly marked in regard to problem solving through political parties, through big men, and participation in demonstrations. Poorer people, and especially it seems (from the discussion of Table 7 above), those with little education, are especially likely to problem solve through political parties, and in, common with those who are disadvantaged by gender or religion - through big men and participation in demonstrations. But they are also more likely to try to solve problems through more than one channel. Wealthier people, and those amongst the poorer sections who have some significant education are more inclined to direct approach to government. Considering all these observations together, therefore, there is evidence that there is in Delhi a group of people who are distinguished in a sense against the overall correlation between education and income by being relatively well educated, in spite of being poor, and that these people are distinctively active in problem solving. The relations that are observed between education and income, and problem solving activity, are cross cut by employment and labour status. As expected people in the poorer sections are almost invariably a little more active than those in the population as a whole; and as we might have expected those who are labour force participants are much more active in problem solving than those who are outside it (see Table A4), except in regard to approach to big men. Again it appears that this mode of problem solving may be the resort of those with least means (by virtue of sex, religion, income and occupation). What might not have been expected is that the unemployed are as active in problem-solving (see Table A4) as those who are active in the labour force. Table 9: Problem Solving By Channel and By Labour Status (%) Permanent Casual Self-Employed Channel All Poor All Poor All Poor Freq in sample Government Legal action <1 <

15 Pol Party Big Man Demonstration Self-prov Any Channel Note: The categories of Permanent, Casual and Self-Employed, referring to different labour market positions are usually well understood by respondents, and the data come from the answers to the question Which of the following work situations is closest to yours? Are you a: Permanent worker (entitled to Provident Fund)/Casual worker/self-employed? The frequencies shown here are percentages of those who are economically active. Given that those who are employed as casual workers are usually less well paid than permanent workers and also likely to be less well paid than a significant fraction of the self-employed, it is quite striking (see Table 9) that the casual workers appear often to be less active than others, though low income people generally, as has been shown, have a tendency to more active involvement in problem solving than the population as a whole. The puzzle here is answered when we examine the relationship between educational levels and labour status, amongst people in the poorer sections (see Table A5). Those who are better educated in the poorer sections of the Delhi population are more likely to be in permanent work or to be self-employed, and it is evidently these people who are also the ones who are distinctively more active in problem-solving. It does not appear, however, that workplaces are very significant sites of action. Only 23 per cent of permanent workers amongst the poor and 29 per cent of casual workers report having got together with work mates to improve wages or conditions of employment; and only very small numbers 5 per cent - of those who are self employed reported having taken any similar action. Given the low incidence of workplace-based activity it is not surprising to find that some categories of citizens who are not participants in the labour force are active, nonetheless, in problem solving. Homemakers and students are less active than the average citizen, but retired people and the unemployed are somewhat more active, especially through approach to government and, in the case of retired people, through political parties. Finally we may examine the relationships between some measures of political participation, and of associational practices, and problem solving activity. 15

16 Political Participation and Problem Solving Activity Two-thirds (64%) of citizens surveyed in Delhi expressed sympathy for a particular political party, 39 per cent of them expressing sympathy for Congress and 23 per cent of them for BJP (and less than 2 per cent of them for any other parties). The proportions of the party sympathisers pursuing problem solving by different channels, by comparison with those who specifically state that they do not sympathise with any party was as follows: Table 10: Problem Solving By Channel and Party Identification (%) Congress BJP None Channel All Poor All Poor All Poor Freq sample Government Legal action Pol Party Big Man Demonstration Self-prov Any Channel There is little evidence that party sympathy has any particular connection with propensity to be active in problem solving, except that Congress sympathisers may be a little more inclined than others to problem-solve by going through a political party, and BJP supporters (both amongst the poorer sections and in the population as a whole) rather more inclined than others to problem solve by means of approach to government, and less likely to go about through a political party. This follows from the fact that BJP supporters are more commonly found amongst those who are somewhat better off, who are generally more likely to go to government reflected in the facts that whereas only 39 per cent in the population as a whole expressed sympathy for Congress and 23 per cent for B JP, the same figures amongst the poorer sections are 48 per cent (Congress) and 16 per cent (BJP). Evidence on party voting in the last Legislative Assembly elections (when 72 per cent of those interviewed in the whole sample say they voted) gives much the same result. The two distinctive points, again, are that Congress supporters (by comparison with 16

17 the sample as a whole) seem more inclined, and BJP supporters less inclined to go about problem solving via political parties; and the BJP supporters more inclined to go to government. These points seem easily explained, as already suggested, as being a reflection of the fact that Congress is more strongly supported amongst poorer people than is the BJP. Amongst those in the poorer sections who voted in the last Assembly elections (and the turnout was a little higher at 74 per cent than in the population as a whole), 64 per cent voted for Congress (as compared with 53 per cent in the whole population) and only 25 per cent for BJP (compared with 37 per cent). Given this then the other remarkable feature of the electoral data is that a particularly high proportion of BJP supporters amongst the poorer sections are active problem solvers, suggesting that support for BJP amongst such people is a very deliberate choice. The other possible connection that might exist between political activism and active problem solving may be observed in data on participation in political rallies and demonstrations. Just as poorer people in Delhi are evidently more active in problem solving than are their wealthier fellow citizens, so it appears that they are more active politically. In the sample for the population as a whole only 13 per cent said that they had participated in political rallies or demonstrations, whereas 21 per cent of those in the poorer sections said that they had. Exactly as we should expect, those who are politically active are also particularly active problem solvers. This is shown in the significant correlations that are observed between political activism (measured by participation in demonstrations and in work for a political party) and problem solving through any channel, and between political activism and problem solving through government, through political parties and through demonstration. Table 11: Correlates of Political Activism Problem solving via: Correlation coefficient Government 0.21* Legal action 0.03 Political party 0.37* Big man 0.15* Demonstration 0.40* 17

18 Self-provisioning 0.11* Any channel 0.25* There is also very little difference apparent here between the population as a whole and the poorer sections, suggesting that those in the wealthier part of the population who are politically active are as likely to be active in problem solving as are poorer people, against the general trend. Associational Practices and Problem Solving Finally we may examine the relationships between active involvement in associations of different kinds, and problem solving activity (Tables A6 and A6a). Exactly as we should expect it is clear that those who are active in associations are distinctly more active than is usual, in problem solving, except in regard to problem solving through big men. The connection between associational activism and problem solving is reflected in the significant correlations (though the coefficients are small) that are observed between participation in any kind of associational practice and problemsolving through all channels except that of going through a big man: Table 12: Correlates of Associational Activism Problem solving via: Correlation with Any Associational Practice Government 0.22* Legal action 0.10* Political party 0.21* Big man 0.04 Demonstration 0.17* Self-provisioning 0.16* Any channel 0.23* The connection is especially marked in the case of those who are members of neighbourhood associations. The differences between the associational activists and others when it comes to problem solving are particularly strong in regard to approach 18

19 via government, demonstrations and self-help less so in connection with problem solving through approach to a political party. Those in the poorer sections of the population who are active in associations are even more active problem solvers than are their equivalents in the general population, but, on the whole relatively fewer of them are active in associations than is the case in the population as a whole. Summing Up This analysis gives rise to a few strong conclusions. *First, there is evidence from Delhi that, contrary to what has been observed in western countries, there is some tendency towards an inverse relationship between household income and problem solving activity. This is shown up most strongly in the statistically very strong effect that we observe between living in a jhuggi jhopri and problem-solving activism. Given that, in circumstances where there is a positive correlation between income and education, the same inverse relationship is not in evidence when it comes to the relationship between education and problem solving - except in so far as those with no education at all are less likely to be problem solvers - it seems clear that those who have low incomes but who have had a fair amount of education (they have completed primary or secondary schooling) are notably active problem solvers. This suggestion is confirmed by an examination of the characteristics of the 87 persons in the sample who can be described as those who are super-active, being politically active, active in associational life, and amongst those who problem-solve through more than one channel. Of these 87 people 57 (66 per cent) are in the poorer sections, but almost half of them (47 per cent) have completed secondary education or have at least some tertiary education. These are people who are more likely to be permanent workers or to be self-employed than they are to be casual, daily-paid workers. It is such people who are those amongst the poorer sections who are most likely to be politically active, and active in associations - who are in turn especially active problem solvers. *Second, there is evidence that poorer people with no education, and some others who suffer from particular disadvantage such as women and Muslims are more likely to try to solve problems through political parties, through participation in demonstrations or (in the case of women and Muslims) through big men. Problem solving through big 19

20 men is often the resort of the poorest and more disadvantaged people and though some relatively wealthy people, as we noted amongst Sikhs, may also problem solve through big men it is likely that the kind of social actors who are involved are different from those (like the pradhans of slum areas) resorted to by poorer people. Poorer people are more likely than the rest to be Congress party supporters, but those amongst them who are supporters of the BJP are likely to be more active in problem solving. *Third, though there is evidence that there is more activity around neighbourhoods and neighbourhood associations than there is around workplaces and unions, there is not much evidence of the existence of vibrant associational activity in Delhi. People rarely report that associations of any kind have been involved in their problem solving efforts, though when they do it is most likely to have been a neighbourhood association. Yet there is strong evidence of informal collective action because people are very commonly involved in attempts to solve collective problems with their neighbours. There does appear to be quite a lot of collective action in Delhi, but it is fluid and certainly fragmented. *Fourth, and most important, it is very clear that political parties remain extremely important in the ways in which poorer people, in particular, go about problem solving. In regard to the core hypothesis stated at the beginning of this paper, therefore, even though unions and programmatic political parties are not major players in Delhi politics, nor vehicles for the representation of poor people and it is doubtful whether they ever have been neither does it appear that new social movements, local associations, or NGOs are particularly well established as alternative vehicles. Certainly neighbourhood associations have a visible role, but NGOs play almost no part in the way in which people try to solve collective social problems. Amongst the poorest people, and those who are disadvantaged in other ways, political parties and especially the Congress party and big men (and the two the big men and the political parties are not always very clearly distinguished) are especially significant in the ways in which they go about problem-solving. Political parties, no matter what their limitations and weaknesses, therefore, remain particularly important for poorer people; and there is little indication in these data of the existence of any other significant collective actors in the worlds of poor people in the capital city. We have 20

21 some evidence, too, that many people do find political parties quite effective. Poorer people in Delhi are especially likely to be supporters of Congress, and 29 per cent of respondents in the whole sample (36 per cent in the poorer sections) said that Congress is most concerned with the problems that are most important to you (in the words of the question that was asked). This compares with 52 per cent who said that no political party is concerned with their problems (47 per cent amongst poor) and only 13 per cent (10 per cent amongst poorer sections) who said that the BJP was most concerned with their problems. Of those in the poorer sections who identified with Congress 72 per cent said that the party had done a lot or had done something to solve problems important to them. This compares with 6 per cent who said the same thing of unions, the mere one per cent of religious groups, and the 11 per cent who said it of ethnic associations. Only neighbourhood associations, of which 33 per cent (though 41 per cent in the whole sample) said that they had done a lot, or something, even remotely approached the political party in terms of popular assessment of their effectiveness in regard to problem-solving. PARTICIPATION IN ELECTORAL POLITICS Our observations from the survey on participation in electoral politics confirm the most important conclusions that we reached in the analysis of problem solving activity. Just as we found that on the whole poorer people are more active in problem solving, we find most strongly in the case of those who have been participants in political events, less strongly so in the case of electoral participation that it is the poorer and also the less well educated who are more active politically. Logistic regressions (see Appendix Three) point to the following conclusions about the determinants specifically of political activism (measured by participation in political events and in party work): (i) Those living in jhuggi jopris are much more likely to be politically active than those living in the (upper income) planned colonies and unauthorised, regularised colonies. 21

22 (ii) The poor and lower middle class people (those with household incomes of up to Rs6000 per month) are significantly more likely to be active than those with higher incomes. (iii) Tertiary education may have a small but significant negative effect. (iv) Women, those who are labour force inactive, and possibly permanent workers, are less likely to be politically active than others. Poorer people are also much more likely to be Congress supporters than they are supporters of the BJP. The following analysis of more formal political participation and of patterns of support for the two major parties in Delhi politics the Congress and the BJP suggests that the social cleavage, between poorer, less well educated people in more vulnerable kinds of jobs, and wealthier and better educated people in good jobs, is reflected quite strongly in party politics. The two parties are not very clearly distinguished programmatically (both supporting similar economic policies, for instance) but they are evidently rather sharply distinguished from each other ideologically, and through these differences apparently have strong links with different fractions of the Delhi electorate. Electoral Participation Here we discuss political participation in the narrow sense of electoral participation, party support and party activism. We noted earlier that the turn out in the last Delhi Legislative Assembly elections evinced shown in the sample was high at 72 per cent in the sample of the whole population and 74 per cent in the poorer sections. It makes most sense, therefore, to examine who did not participate in those elections. The data show that they are more likely to have been women than men, in the youngest age groups, Muslims rather than Hindus, tenants rather owner occupiers, relatively recent settlers, and perhaps more commonly found amongst the better educated (a pattern which corresponds, except in the last particular, concerning education, with that for those who are less likely to be involved in problem-solving). On the whole the least well-educated seem less likely to be non-participants in elections, and this appears to be most marked amongst the poorest who also appear, except in the youngest age group to be rather less likely than those who are wealthier not to participate in elections. Otherwise, there was no discernible trend in regard to household income. 22

23 Those who are self-employed were especially unlikely not to have participated in the election. In sum, electoral participation is high, and there are some indications albeit not strong ones that participation may be a bit higher amongst poorer and less-well educated people, especially if they are also self-employed, broadly in line with what has been observed in Indian election studies more widely, where the phenomena associated with the second democratic upsurge have been identified. Patterns of Party Politics As we have seen 39 per cent of the Delhi citizens identified with Congress, and 23 per cent with BJP; and in the last Assembly Elections 53 per cent reported themselves as having voted for Congress (64 per cent amongst the poorest sections) and 37 per cent for BJP (25 per cent amongst the poorest sections). What can we say further about patterns of identification with the two dominating parties in Delhi politics? We have already noted some evidence of a stronger preference for Congress amongst people in the poorest sections, by comparison with the population as a whole. Indeed, detailed analysis of our data confirms that poorer and less well educated people, as well as Muslims, younger people and men rather than women are drawn to support of Congress, while older, wealthier, better educated people, and quite a lot of those outside the labour force are drawn to the BJP. The clear picture that emerges of the relationship between support for the Congress party by people of lower incomes, lower levels of education, and those who are in vulnerable positions in the labour market as casual workers, or as unemployed, is further reflected in the evidence we have that BJP supporters are more active in associations. This is also an indication of the important observation that is elaborated upon later in this paper, that it is those who are wealthier and who are better educated who are most active in associational life in Delhi: Table 13: Associational Practices by Party Identification Assoc. Activity Congress BJP None Religious group Cultural/Sports Union/Professional