Cover Image
close this bookCommunity Leadership and Self-help Housing (HABITAT, 1988, 50 p.)
View the document(introduction...)
View the documentForeword
View the documentI. Why analyse community leadership
View the documentII. The nature of community leadership and decision making in low-income settlements
View the documentIII. Why do leaders emerge and how do they relate to society at large?
View the documentIV. The impact of leadership upon low-income community development
View the documentV. Leaders and external agents: cooperation or conflict?
View the documentVI. Reaching the poor through community leaders; an agenda of points for consideration
View the documentReferences

(introduction...)

United Nations Centre for Human Settlements (Habitat) Nairobi, 1988

HS/132/88E ISBN 92-1-131059-8

Foreword

International discussion on shelter strategies has been brought about by the alarming trends in the field of human settlements. It emphasizes the need to mobilize coordinated and widely-based efforts on the part of governments, the private sector and communities to act in concert, if there are to be visible improvements in the shelter and neighborhoods of the poor and disadvantaged by the year 2000. Effective strategies for the development of shelter, infrastructure and services at the community level have many facets and components, yet they all underline the necessity of involving communities in decision-making, planning, construction activities and the maintenance of shelter, infrastructure and services.

This publication, prepared under the Community Participation Training Programme of the United Nations Centre for Human Settlements (Habitat), discusses the role of community leadership within low-income urban settlements. Prevailing or emerging leadership patterns have to he fully understood by local authorities, non-governmental organizations and international agencies who are considering alternative strategies of working in and with communities. More often than not, efforts by the established governmental authorities. supported by international assistance, have failed as they have underestimated or misjudged the role of community leadership.

This report, based on extensive research work carried out by Peter M. Ward and Sylvia Chant, analyses the impact of leadership on the development of low-income settlements, and suggests an agenda for cooperation with local leaders for the purpose of reaching the poor in their urban settlements through various forms of participation.

Arcot Ramachandran
Under-Secretary-General
Executive Director

I. Why analyse community leadership

1. The importance of community participation in many rural and urban development schemes is widely accepted by planners and governments alike. Participation is viewed as a fundamental right. It improves the extent to which community-development programmes meet people's needs, and it raises the sense of commitment to both the implementation and on-going use of a project (see references, 1, 40, 68, 83). A key issue is how to encourage community action at different phases of a development project. (95) Belatedly, perhaps, the importance of gender is gradually being acknowledged: women are widely involved in these processes but often in different ways from men. (96) Successful project planning in future will demand greater recognition of the different roles and contributions that women, men and children respectively fulfil within community participation.

2. Community leadership is one important area of decision-making that has been almost entirely neglected. In the academic literature relating to developing countries there is no specific study that examines the various types of community leadership that emerge; the nature of its links to outside bodies and to the community members it represents; the characteristics, motives, skills and legitimacy of individual leaders; and so on. Virtually nothing is known about the relative contribution that different forms of leadership make to the nature of community participation and to successful project development. Interventions in established communities may ignore the structure of leadership that already exists, while in newly-established settlements it is often assumed that active leadership will evolve spontaneously or that it can be mechanistically prescribed as part of the local organizational structure.

3. An adequate understanding of the issue of leadership in decision-making is paramount for several reasons. First, the influence that individual leaders may have in encouraging participation, and in successfully channeling it to the benefit of all members, is an important variable in its own right. As one author remarked more than 10 years ago:

"anyone who does extended fieldwork in squatter settlements or related types of low-income zones in a Latin American city cannot fail to be impressed by the importance of leadership differences in accounting for the developmental trajectories of each settlement. There appear to be strong relationships between leadership performance and differences in the outcomes of demand-making experiences, the length of time needed to accomplish certain developmental objectives, the quality of relationships maintained between a settlement and political and governmental agencies, and the level of internal organization within a settlement over time."

Despite the truth of this statement its implications have largely been ignored.

4. A second reason for eliciting greater awareness of leadership is that it is not a "passive" variable in any development scheme. Where leaders already exist, any external intervention, no matter what its aims, will be perceived by those individuals as a potential threat, regardless of how benign or apparently valuable the assistance being offered. These fears must be recognized and, with sensitive handling, can usually be allayed. In those circumstances where leadership has to emerge in a newly-established community, leadership is conferred upon some individuals but not on others. This may generate jealousies and conflicts that are obstructive rather than conducive to community action.

5. Thirdly, multiple and overlapping leadership often already exists or may emerge. In some cases this is not a problem, but past research and experience suggest that it can lead to divisiveness and conflicts of interest between individuals and their supporters. Competition between rival leaders is as much a stumbling-block to community participation as internal heterogeneity of the community, in terms of tenure, income and so on.

6. A final reason for being aware of the importance of leadership is that insensitive action by an the whole viability and future of a development project. The attitudes of authorities are often patronizing, sexist and unilateral. In addition, apparently simple matters of protocol, such as where to meet leaders, on what terms, and in whose territory, if handled ineptly or if they are not considered carefully, may kill an initiative before it has begun.

Aims of this report and its organization

7. The primary target group for this report consists of professionals working in community development projects in human settlements. The sort of projects envisaged are those of irregular-settlement upgrading programmes, sites-and-services, the extension of primary health-care facilities and so on. This is likely to include professionals such as the project staff of development agencies and non-governmental organizations (NGOs), planners, engineers engaged in the provision of public utilities, doctors and paramedics, as well as a whole host of individuals with an interest in becoming involved in a local community (church people, political parties, academic researchers etc.). In short, it is envisaged that this material will enjoy a wide readership.

8. In addition, field experience in several Latin American countries has been used. The drafters of this document also talked with other researchers who had worked with local leaders, either directly or indirectly, in Africa and Asia and who made comments which have assisted in making genuine cross-cultural comparisons. Unfortunately, given the limitations of space, detailed case studies have not been included. Readers are encouraged to consult original sources where they require more detailed accounts related to a particular region.

9. It is felt that this review has developed sufficiently for meaningful generalizations to be made about the nature, dynamics and impact of community leadership upon urban development schemes that involve self-help, upgrading and settlement improvements. While many of the recommendations about ways of establishing liaison and working with local leaders are tentative and unproved, it is felt that they offer a realistic agenda for future action. Practitioners (both leaders and agency staff) who are at the "sharp end" of these proposals and who have to implement them are asked to provide UNCHS (Habitat) with feedback about their effectiveness, and with information and suggestions about ways in which they ought to be improved. This feed-back can be communicated to the address given at the end of the report.

10. The report contains five substantive chapters. In chapter II the nature of community leadership is described and the characteristics of leaders in different parts of the world are examined, along with their motivations to take on what often appears to be an onerous task with little to commend it. Chapter III analyses the nature of leaders' links to the wider socio-political system and attempts to identify the key determinants of different leadership styles and practices, the reasons for factionalism and for changing patterns of representation. The influence that leadership exercises over community development is examined in chapter IV. Leadership outcomes are evaluated not only in terms of settlement improvements achieved, but also from the point of view of the impact that leaders may have upon socioeconomic inequality, as catalysts for participation, and as so-called "opinion leaders". Finally, chapters V and VI seek to offer advice about some of the major stumbling blocks to successful collaboration and about how those who work with leaders might best overcome such difficulties. Specifically, some of the key points of conflict likely to emerge between leaders and external agents and the rationale underlying those conflicts are identified. Many aspects of project implementation are extremely sensitive and the report shows how different types of agency have responded to the challenge. The "agenda of points" in chapter VI answers many of the practical questions such as how to identify leaders, how to establish contact and work with them, where and how to make a judgement about when to abandon attempts to work within an existing leadership structure and when to create a new one.

11. To repeat: the aims in this report are to raise the level of awareness about the nature and importance of leadership in community development, and to encourage external agencies to be sensitive in the way in which they engage with local leaders and residents. There already exists a local leaders and residents. There already exists a substantial "graveyard" of projects which have failed, or the success of which was seriously impaired, through external agencies paying inadequate attention to the interests and influence of community leaders. It is hoped that this document will give cause and pause for greater thought.

II. The nature of community leadership and decision making in low-income settlements

12. This chapter attempts to identify the various sorts of leaders and leadership functions common in low-income settlements and to show how different types of leader emerge in different cultural contexts. First it asks who the leaders are, and examines both traditional and modern patterns of local leadership selection in order to account for and describe the principal characteristics of leaders in contemporary urban areas. The second section asks how leaders derive support, by identifying the main types of leadership legitimacy found in low-income settlements, and by examining their influence on leadership patterns and community organization. Finally, it asks why certain people become leaders, and identifies the various sources of motivation to take on leadership positions.

A. Characteristics of leaders

1. Traditional patterns of leadership selection

13. Before identifying the characteristics of leaders of low-in come urban communities, it is important to examine traditional patterns of leadership selection in rural areas in order to develop a benchmark by which contemporary structures may be evaluated. In this section, data from both historical accounts of traditional leadership and contemporary patterns in remote rural areas have been drawn upon.

14. Traditional forms of leadership selection vary greatly according to cultural context and to the functions that leaders are expected to fulfil. They may also differ significantly over local areas depending on the particular characteristics of the individual settlement. On the whole, the primary determinant of traditional leadership is ascription, i.e., leaders owe their position in local society to who they are, and not to what they do. This is particularly the case in many African and Asian countries, where appropriate leadership qualities include age, status and affiliation to the dominant kinship group or lineage.

15. A leader's functions mainly revolve around representing the community to the outside world, resolving disputes within the village and offering protection to villagers. In many parts of Africa, for example, villages are headed by tribal chiefs or clan elders. (43, 64) Age is a critical factor because political power, high social status and the ability to command trust and respect are prerogatives of the elderly in traditional African society. (12, 30, 106) Ethnicity, that is racial and cultural background, is also an important consideration, and the local headman is likely to come from the dominant tribe or clan in a given area. (8) Wealth too, while generally less important than age, kinship and prestige, is similarly a key factor, not only because it tends to go hand-in-hand with these other criteria, but also because large landowners control resources on which the local population may be dependent for its livelihood, and are an important source of employment in village communities. Both historically, and in many present-day villages in the rural areas, there is a widespread tradition of the "Big Man" phenomenon, whereby a paternalistic leader of a village offers patronage and protection for his community.(8) While a "Big Man" is not necessarily a male leader, and indeed in many West African tribes such as the Igbo, the Hausa and the Yoruba in Nigeria, women and men have traditionally held important positions in local political matters, women have tended to become excluded with the penetration of colonialism and accompanying male- dominated administrative systems.

16. In Asia, respect for age and experience, along with membership of a high-status lineage, is similarly an important criterion in traditional patterns of leadership selection. In South-East Asia, particularly, Confucian ethics, which emphasize a fundamental relationship between wisdom and old age, have led to a system whereby villages were traditionally governed by a small, elderly elite. Traditional respect for age continues to be even more important than gender in terms of status in South-East Asia, meaning that some women have the opportunity to occupy leadership roles within their local communities.(4)

17. In Latin America, traditional leaders in rural areas conform to a slightly different mould than in other continents, possibly because indigenous forms of organization were wiped out by the Iberian conquest. In this case, leaders in rural areas are often either self-imposed, authoritarian leaders, generally known as caciques, who derive their support from a combination of force and the exploitation of a "brokerage" role between the community and local government, or they are elected as municipal leaders - a task which is seen as onerous, but nevertheless a duty.

18. The above-mentioned factors are important in the following discussion of contemporary leadership patterns in urban areas, for the affective or ascriptive dimension in leadership selection may continue to be considerable where identity with, and commitment to, one's tribe or kin group is deemed more important than specific developmental objectives. However, it is critical not to mistake "traditional" values as the primary basis for local-level organization and leadership with the fact that those values may be a veneer for more "modem", political or instrumental associations. For example, although tribalism per se has been heralded by some as an important political factor in local development in the United Republic of Tanzania, in Kenya, Zimbabwe and elsewhere, it is likely that ethnic symbolism is merely employed as a strategy to mobilize support for movements which are not first and foremost tribally-oriented. One must also be careful not to equate "modem" values with "Western" conceptions of appropriate forms of community organization. In several respects, the qualities associated with traditional leadership are also found in urban areas, albeit under a slightly different guise.

2. Contemporary patterns of leadership in urban areas

19. In the following discussion of contemporary leaders in urban communities, examples from "modem" rural settlements which have been the object of national development programmes have also been included. A critical difference between the analysis of traditional and contemporary leadership structures is that leaders in the latter category are more likely to be exposed to a wider range of influences and also to more bureaucratic decision- making machinery than those living in remote villages. Therefore, when considering contemporary leaders one must not only be aware of the cultural backcloth against which different types of leaders emerge and the various functions they fulfil, but also the political context in which housing and community-development projects take place.

20. As far as leadership functions are concerned, urban environments provide a wide diversity of opportunities for neighbourhood representation. Community organization and leadership may spring up around specific issues such as upgrading, (43, 58, 100) or the promotion of different group activities such as social and welfare associations. For example, in urban African neighborhoods a multiplicity of organizations can be found, such as traditional councils of elders, religious affiliations and landlord associations; or groups may be established for the arbitration of disputes within the settlement and for the protection of the community. Mbithi and Rasmusson, in their analysis of the harambee (self-help) movement in Kenya, maintain that different leaders emerge for separate phases of local development projects - initiation, organization and implementation.(55) Moreover, attention has been drawn to the fact there are two main types of contemporary leader - those whose role is "social-emotional" and those whose function is "task-related". (77) The former are likely to derive their support on the basis of more traditional values; the latter either from skills acquired through education or experience, or from key contacts with government officials and influential politicians which are used for acquiring specific community benefits. For example, in a low-income community in Madras, India, the religious leader was a high-caste elder, while the "political" leaders were younger and tended to have contacts with influential people.(66) Indeed, it is quite common in certain contexts for there to be several leaders who may or may not be in competition with each other. Despite the diversity of leadership functions in low-income settlements and the various different challenges posed by membership of urban communities, many types of contemporary leader share common characteristics.

3. Principal characteristics of contemporary local leaders

21. In the following breakdown of the key characteristics of urban leaders it is important to note that many of these factors are interdependent. For example, gender often determines the type of employment one is likely to obtain, which in turn affects income and social status. Obviously the extent to which these variables overlap depends on the cultural and economic context of different countries. This report does not attempt to demonstrate the likely interrelationships within specific regions, although their possible existence should be borne in mind. The first three categories of leadership characteristics appear to have more universal significance than the other five, so they have been dealt with in slightly greater depth in order to show the range within each type.

Age

22. The first significant point about age is that contemporary leaders tend to be younger than their traditional counterparts. For example, in a Copperbelt mining town in what was then Northern Rhodesia (now Zambia), the tribal elders who traditionally presided over the local urban administration were gradually replaced by a group of younger, better-educated men whose basis of legitimacy stemmed from their ability to negotiate with so-called "modern" urban institutions.(30) Similar processes have been observed in India (98) and the Republic of Korea where, despite the traditional respect accorded to age in village communities, the Government of President Park in the early 1970s insisted that the local leaders of the New Community Movement should be younger than the traditional leaders and preferably in their 30s or 40s.(11)

23. Despite this trend towards relative "youth" of contemporary leaders, most tend to be in their early to late middle-age, and older than the majority of other community residents.(24,77) In Rio de Janeiro, most favela squatter leaders (65 per cent) are over 40 years old, (26) and similarly, in Kenya, the majority of harambee settlement leaders are aged between 40 and 49 years.(55) In urban communities in advanced industrial economies, there are similar patterns.

For example, in a study of institutional leadership in a housing estate in the United Kingdom, the median age of leaders was 44 years, and in a small group of up-and- coming potential leaders in their late 20s and early 30s, the local influentials tended to be older.(67)

24. The fact that most leaders in third-world communities are slightly older than non-leaders suggests that age is an important criterion in leadership selection. Comparative seniority furnishes an advantage for leaders in a number of ways. First, their age is likely to command respect from the community.(11, 18) Secondly, it will probably mean that they are more experienced and therefore more effective in dealing with external agents. Thirdly, older leaders are likely to be better-off economically which may allow them to devote their free time to community activities rather than to additional income-earning.(24)

Sex

25. Another leadership characteristic relevant to most areas of the world is gender. The great majority of contemporary leaders is male.(12, 21, 41, 55, 65, 98) This is probably due to four main factors. First, in several countries there is a long tradition of the management of public affairs by men. Secondly, men are likely to be more successful in negotiation with public bodies because they may have more experience in the "outside world", and because the personnel of bureaucratic institutions, especially in male-dominated societies, might well be male and give greater recognition to men rather than women leaders. Thirdly, there may be social stigmata attached to women working outside the home or alongside men.(58) Finally, women may be greatly restricted in their associational mobility and lack opportunities for building up a support base.(4)

26. However, in certain areas, women seem to have gained a significant foothold in urban leadership positions. For example, in Latin America, where there are comparatively high rates of female migration to cities in comparison with other regions of the world, women often occupy key roles within their communities.(21, 58, 99) This may, in part, be a reflection of the fact that the proportions of women residents are relatively higher than men in women residents are relatively higher than men in "self-help settlements", either because of high rates of female migration or because of the comparatively high percentages of women-headed, single-parent families in the region.(l7,99) Generally, women who are single have greater freedom than those who are married.(17,65) Female leaders are also frequently found in areas where women have traditionally maintained distinct, though complementary, organizations from men, for example, several leaders of harambee groups in Kenya are female, (49) although Muslim women in Kenya rarely play conspicuous leadership roles.(55)

27. However, there is no clear global pattern of variation in gender and leadership selection. In certain areas within countries as diverse as Ecuador, India, the Philippines and Spain, it is expected of women to participate in community improvement, either because they are "closer to the community"; or have more "free time", and/or participation in neighbourhood organization is seen as a "natural" extension of their domestic duties.(15,58,70,98) In other parts of these same countries, however, women are excluded. Overall predominance of men in leadership positions corresponds with the fact that most cultures are heavily male-dominated, and that only in certain cases, where the local political context allows or demands it, will women also take on leadership responsibilities.(41)

Education, occupation and income

28. The majority of contemporary leaders tend to be more educated than the other members of their communities, and to have completed at least a primary school education.(26,34,100) However, there are exceptions. For example, in the Mathare Valley, a squatter settlement in Nairobi, leaders tended to be less educated than other residents, with most of them having had no formal education.(77) Generally, however, literacy is a prerequisite of occupying a leadership position since leaders will inevitably be involved at some time in written communication, drafting and drawing up contracts with government officials or representatives from an external agency.(85,98)

29. One may also conclude that higher educational levels often go hand-in-hand with higher status occupations, for although leaders' economic activities vary widely they tend to be of higher status and better-paid than jobs held by other community members. Many leaders are either artisans or in middle- to lower- level service occupations such as government employment. (15,26,109) They are also frequently self-employed, which presumably gives them greater flexibility to engage in community affairs.(21,34) In some cases they are involved in activities in which they exercise a role of dispensing patronage: for example, in India, money-lending is often the basis on which potential leaders build their support base.(66)

30. Relatively higher status occupations tend to make leaders better-off than the majority of residents.(20,28,34,100) This is an advantage since they are often "unpaid" for their leadership duties.(11,24) Certain types of leader, however, also use their leadership role itself to enrich themselves.(21,34,98)

Skills and experience

31. Leaders are generally "skilled" in a number of ways, including the capacity to organize,(20) to articulate people's needs,(77) to negotiate with government officials and other institutions(12) and the willingness to listen which may sometimes also be considered an important skill.(11,32) Many of these skills will be acquired through earlier experience, such as involvement in other development and community-related activities.(11,109) Ethnic associations in West African cities, for example, often serve as training grounds for participation and leadership in neighbourhood politics.(36)

Resources and contacts

32. Leaders often have control over physical or economic resources within their communities, such as land, which are used either as sources of enrichment or as means of exercising control over residents.(17,20,100) Alternatively, they may have a monopoly over the political resources at the community's disposal. For example, leaders often engage in reciprocal relationships with key figures in government or political patrons whereby the leader secures certain promises and benefits in exchange for loyalty and for mobilizing residents on behalf of the patron when required.(8,20,21,43,77)

Social status

33. "Social status" may refer to inherited status or prestige derived from tribal, family or lineage affiliations. High status is an extremely common characteristic of contemporary leaders in Asia and Africa, especially for those leaders who fulfill "social-emotional" roles within their communities, such as local religious figures.(l2,66) In Africa and India, leaders are likely to be drawn from the dominant tribe or highest caste in a given area, respectively.(32,36,54) High social status tends to go hand-in-hand with wealth.(37)

Migrant status and length of urban residence

34. In urban settlements, contemporary leaders tend to be urban- born and/or long-term residents of the city.(l2,20,21) For example, 73 per cent of the leaders in Mathare Village 2 had lived in Nairobi for over 10 years.(77) Similar findings emerge for leaders in a British housing estate.(109) Length of residence obviously provides an advantage in terms of familiarity with the way in which the city system works, and experience of local issues which may prove useful in a demand-making context. Urban residence is also clearly correlated with educational level achieved which, as has been described, is frequently an important feature of leadership.

Personal dynamism, popularity and charisma

35. "Charisma" is probably the most difficult aspect of leadership to define, but it is one that is sometimes seen as critically important to the success of development efforts.(11,43,76,77,98) Essentially it consists of an individual possessing personal magnetism, impact and popular appeal. Certain authors maintain that without charismatic leadership, many development projects are doomed to failure, and although this may, perhaps, be somewhat overstated, "charisma" is obviously an added bonus to other more tangible leadership skills.(11,75)

Conclusion

36. There are certain characteristics of leadership common to many parts of the world. Leaders are likely to be middle-aged, male, relatively well-educated, and holders of comparatively "high-status" jobs with better than average remuneration. Wealth is often accompanied by high social status. They will probably have good access to resources and contacts, posses certain skills and experience relevant to the community in question, be long-term residents and have charismatic appeal. However, in spite of this similarity of their personal qualities, the way in which leaders acquire legitimacy varies markedly.

B. Leadership legitimacy and community organization

37. How do leaders derive support from their local community? In an attempt to answer this question it is proposed to identify the principal types of leaders in low-income communities and relate them to their basis of local legitimacy. This will also allow an examination of the degree to which popular participation for community-level improvements will be encouraged under different leadership systems.

38. There are four main types of leadership categories in low-income settlements which are briefly discussed below (see also table 1). It is important to note that these categories are not necessarily distinct from one another. For example, authoritarian leadership may overlap with traditional leadership if there is a history of narrow popular participation in rural areas, such as caciquismo in Latin America or the "Big Man" phenomenon in African villages. Similarly, political leaders may overlap with authoritarian ones. In low-income settlements in Mexico it is common for local caciques to be affiliated to, and representative of, the dominant political party.(7,20)

39. "Traditional" leaders spearhead a form of community organization common in rural areas which has sometimes more or less been replicated in an urban setting. For example, traditional councils of elders are common in inner-city areas of Ibadan, Nigeria, although they have gradually acquired new skills and contacts in order to extend their influence in an urban context.(12) However, "new skills" are not always necessary for continued support since the basis of legitimacy, as indicated in table 1, is who the leaders are and not what they do.

Table 1. Leadership legitimacy

Type of leader

Legitimacy

Degree to which leader seeks to mobilize action and support for community level improvements

Traditional

- Ascriptive

Variable, but often low


- Affective


Authoritarian

- Self-imposition

Moderate/low


- Pseudo-elections (if at all)



-Some derivative support acquired through contacts


Positional

- Placed or nominated by supra-local organization

Varies, but usually low


-Derivative support insofar as community perceives link with organization as useful


Community representational

- Elected (usually) democratically

Moderate/high

40. Sometimes traditional-type leadership may exist alongside other forms of leadership as a result of the function the leader performs. For example, in a Madras housing project, a high-caste, 57 year-old man fulfilled the function of religious leader as well as settling land disputes, whereas other types of leaders in the community played political roles.(66)

41. Under "traditional" leadership, participation is likely to be low if the leader is paternalistic and handles community affairs alone. Furthermore, since his support does not depend on what he achieves for the community, he may not be disposed to promoting community improvements unless they will benefit him personally. However, as table I suggests, this is variable. In many traditional communities there is a custom of donating a certain amount of unpaid collective labour each year for the good of the community, such as the tradition of shramadana in Sri Lanka and the faena in Mexico.(31) Yet although the community may in reality participate quite actively in voluntary work efforts, the key point here is that it is a tradition, and not necessarily the result of the leader's efforts.

42. Authoritarian leaders, such as the widely-described cacique in Latin America, are self-imposed and gain their leadership through domination and coercion. They also generally derive a certain amount of external support in urban areas through their contacts with patrons. Such "derivative" support may take the form of powerful patrons visiting the community and publicly congratulating the leader for his efforts.(20) Self-imposed leaders may maintain their control through the strategic monopoly of local resources (political and economic) or through force exercised by a small coterie of "strong men".(17) Pseudoelections are sometimes held in order to reinforce the leader's legitimacy (see table 1). Under caciquismo, rates of community participation in demand-making for services and so on are usually low and strictly controlled, for if it were too successful the cacique would undermine his basis of support. However, on occasions when the leader needs to demonstrate a turnout of support or voters for his patron in order to secure the promise of a benefit for the community, then, for a while at least, mobilization of the community may be high.(21)

43. Positional leaders are attached to an external institution, be it a political party, a trade union, a religious organization, or faction of any of these, where the aim of that organization is to foster grass-roots support among the urban poor.(28, 45, 66, 78, 108) "Positional" leaders derive their legitimacy either through the status conferred by the supra-local organization, or through the support of residents who perceive the leader's links with the organization as useful.(20) Where leadership legitimacy is mostly dependent on the local community, the desire to mobilize residents for community benefits may be high, since it will increase the popularity and credibility of the leader. However, where legitimacy is weighted, as it often is, towards the external institution, then the leader will have to comply with the orthodoxy and ground rules of that organization. In- this case he will seek to reduce the level of mobilization and the demands that are directed towards the external body.(20,28)

44. Then there are freely-elected/community representational leaders. The term "freely-elected" does not mean that the whole community has participated in elections, nor does it imply that the leadership is representative of all groups within the settlement. However, this category of leadership is probably the only one where the needs of community residents are important criteria for selection. The source of legitimacy is the community itself and not inherited status, force nor political contacts. "Freely-elected" leaders are chosen for instrumental reasons, namely to achieve improvements for their settlements. Therefore they are likely to involve residents in mobilization efforts in order to press the relevant agencies for goods and services. However, there is a progressive reduction of mass mobilization as services are gradually acquired due to the fact that the leader's role becomes less critical to the community. Leaders may combat this by deliberately dragging their feet in order to extend their usefulness as representatives.(100, 103, 108)

C. The motivation to take on leadership positions

45. Having discussed the characteristics of leaders in low- income settlements and their bases of support, it is time to ask why certain individuals are motivated to take on leadership positions in their communities. Certainly leaders are often subject to physical, verbal and slanderous abuse, and it is often a thankless task. So why do they become leaders? The review suggests that the reasons for taking on leadership of low-income settlements generally involve a combination of one or more of the following factors, although overall, motivation is generally weighted towards advancing one's personal prestige and/or personal economic gain.

1. Leadership as obligation

46. Certain leaders may feel obliged to become community representatives, either because of their widespread popular support, or because privileged status in certain cultural and ideological contexts carries with it the obligation to provide for, or protect those who are less fortunate. As an example of the first point, in Sirswadi Village, Maharashtra, India, a woman who was very popular among the residents was made leader, despite her reluctance to take on this responsibility.(98) As an example of the second, in low-income communities in Bangkok, leadership positions are almost exclusively dominated by better-off residents, mainly because it is expected that richer people should make sacrifices for the sake of their communities.(24) In some societies, every senior male citizen is obliged to take on leadership duties at regular periods in his life, even though most are not eager to do so.(50)

2. Leadership as a means to political and/or social advancement

47. Another key motive for adopting leadership positions is to acquire prestige, social status, or a springboard for future political careers.(6, 20, 21, 27, 28, 32) Many leaders exploit their broker's role between the community and the outside as a mechanism by which to develop their own interests or to advance socio-economically.(27, 35,57, 69) Full-time political careers often materialize for leaders as a result of either cooption or incorporation into the wider political machinery, or by making formal and informal contacts which serve to further their personal political interest.

3. Leadership for altruistic reasons to quicken the local development process

48. Some leaders genuinely seek to improve the conditions of their communities and adopt leadership positions in order to hasten the arrival of goods and services for their followers. Gender may be important here. For example, in Guayaquil, Ecuador, men take on leadership functions primarily for economic reasons while women do so mainly for non-pecuniary motives.(58) Obviously, a leadership position is in part desirable because it allows women to get out of the home, hut women are also generally more altruistic because they are most concerned about the plight of their children. Certainly, it is, in part, women's "incorruptible" image which facilitates their participation and/or helps to get them elected in the first place.(57.5X)

49. The motives for taking on leadership positions usually overlap, hut it is important to note that the process of external intervention in low-income communities itself can modify a leader's originally altruistic motive to one more concerned with individual gain. This is because it is generally the leaders and the better-off residents who benefit most from development programmes.(16,37) After evaluating leadership and community-development programmes in various parts of the world, Dore concludes:

"The relevant question is the balance between self-interest on the one hand and, on the other, the desire to do a good job to benefit others, a balance which all too often gets tipped decisively towards the first."(27)

This survey suggests that this is probably an accurate reflection of the motive for most leader involvement. Yet, as is shown in the following chapter, pressure to fill positions of leadership often does not emanate from the community, but from outsiders.

III. Why do leaders emerge and how do they relate to society at large?

50. Although the motive for taking on the role of community leader varies considerably, this fact, by itself, fails to explain why leadership adopts particular forms both between societies and within societies. Why, for example, do different systems of caciquismo exist in Latin America and why do caciques become prominent at different periods? Similarly, in the Republic of Korea, why should parallel structures of leadership exist within the New Community Movement'? (11) Leadership motivation, by itself, fails to explain adequately the changing patterns of leadership that one observes both at a city-wide level as well as within individual communities.

51. Perhaps the most important feature of leadership is its role as intermediary between local residents and supra-local institutions, he they governmental, political or private. This chapter specifically identifies the nature of the link between these two levels as mediated by leaders, and analyses the relative importance of the community and external bodies in shaping leadership structures.

52. There are several reasons why this is regarded as a key issue. First, the review of the literature demonstrates that one cannot divorce analyses of leadership from the historical and cultural settings in which they emerge (76) nor, more specifically, from the particular political and economic context of a society at any given moment in time. As Bartra (7) argues, leadership forms part of the power structure of a whole region. One needs to understand, therefore, how leadership both affects and is affected by the local political economy. Leadership cannot be analysed in a social or political vacuum. It is necessary to discover why and how leadership structures operate and this can only be achieved by examining the nature of engagement between leaders and society. Secondly, and arising from the first point, all leaders act as mediators between external institutions and the grassroots, but this mediation operates in a variety of ways with varying degrees of autonomy. Sometimes neither leaders nor institutions are what they seem: both have covert as well as overt rationales which affect their behaviour.(34) If these different rationalities are not appreciated then it is likely that the real nature of leadership functions will be totally misconstrued. Thirdly, leadership is a dynamic phenomenon. Thus, local leadership structures change as perceived "needs" shift, as socio-political structures alter, and as competitors emerge. It is important to recognize the dynamics of leadership and to seek to identify the variables which are most likely to lead to change. Finally, leadership overlaps spatially and functionally so that both the causes of, and likely outcomes associated with, factions and their formation need to be understood.

A. The nature of leaders' links upwards to external institutions and down towards those whom they represent

1. The need for representational structures

53. All societies have some sort of formal or informal structure through which information, demands, resources and so on may be channelled both upwards and downwards. The extent to which these channels successfully represent the interests of the originating group and the degree to which the traffic is one-way varies considerably. Community representational structures may be highly formal such as the Saemaul Undong in the Republic of Korea or Acciomunal in Colombia. (11,34) Alternatively, they may comprise very loose arrangements with little formal direction or internal rules to govern activity. In such cases, leadership may operate simultaneously on a variety of fronts (religious, emotional, petitioning, arbitrating) but without having to constitute itself formally as a representational structure.

54. The formal establishment of a community association is usually a result of initiatives from outside rather than an expression of the local population's needs. For example, in addition to those cases already cited, the Alliance for Progress during the 1960s required the creation of community participation organizations throughout Latin America;(92) British colonial governments did likewise in Africa,(53) as did Community Action Program in the United States of America during the 1960s.(15) In all cases there were several reasons why it was considered desirable to encourage community participation and the formation of representational structures: to empower and raise people's consciousness; to promote democracy; to facilitate the implementation of community-development programmes; maintain social control, to reduce unrest and to harness political support.(83) The important point to recognize at this stage is that the initiative for formal arrangements usually came from outside the settlement.

2. Types of leader-State relationship

55. There are four inroad types of relationship between leaders (and their followers) and external institutions. In most cases this relationship is unequal with control over leaders being exercised by the State in a variety of ways (see table 2).(14,57)

Patron-client

56. The identifying characteristics of a patron-client link are that it: (a) is informal and not legally binding; (b) comprises personal. face-to-face relationships; (c) involves an exchange of valued resources; (d) is between actors of unequal status; and (e) persists through time.(35) Each actor in the equation seeks to fulfill certain aims or goals by offering resources that he controls or has access to in exchange for resources that he does not control. The classic example is where a politician or government bureaucrat promises public utilities in exchange for votes or political support.(19) Sometimes mobilization around a political boss leads to the formation of an "urban political machine" which develops a whole network of obligations and favours in order to extend a support base.(25,43) The patron- client relationship is extremely widespread and numerous examples can be found in the literature of Africa,(8,77) on the South Asian subcontinent and elsewhere in Asia,(37,62,96,98) and in Latin America.(20,44,51)

57. In actual fact, leaders are rarely either patrons or clients but intermediaries or brokers.

"A broker does not directly command the resources relevant to an exchange but instead maintains a personal relationship both with an actor who does control the needed goods and with one who desires to acquire them".(35)

The relationship of the broker to his superior is a crucial one, but it is also two-way. If the leader fails to conform to what the patron expects of him (e.g., fails to deliver enough support, demands too many services too fast, or does not keep his followers in line), then he may be punished by withdrawal of the patron's support. By the same token, the patron must deliver some of the rewards desired by the leader's followers or they may seek an altemative patron.(21)

Table 2. Methods of control exercised by different forms of community-State relationship

Nature of linkage

Way in which external control is exercised

Patron-client

-Manipulation


- Ad hoc concessions

Cooption/ incorporation

- Occasional minor concessions


- Compliance with external orthodoxy

"Routinization"

- Queuing and concessions


- "Gatekeepers" allocating resources


- "Red tape" in order to delay and obstruct

Autonomy/ independence

- Concessions made in order to pacify


- Manipulation


- Repression; assassination of leaders etc.

Cooption and incorporation

58. Cooption occurs where a leader affiliates with a supra-local organization thereby becoming subject to its orthodoxy, procedures and rules. A leader may do this because he or she believes that affiliation is likely to help win benefits for those whom he or she represents (or himself/herself). Thus affiliation to an influential political party may be perceived (usually mistakenly) as a prerequisite for successful demand-making.

59. Sometimes a leader seeks affiliation for private gain, but does so secretly, recognizing that his/her local support would be undermined if the link became common knowledge. "Incorporation" is the same as cooption except that it involves the whole group rather than just the leader.

60. Usually cooption and incorporation lead to a reduction in influence that the leader and local group are able to wield in their negotiations with external institutions.(28) This is because it is primarily a device used by institutions to extend their influence and control downwards. rather than by a mobilizing agency geared to meeting community needs.

61. Cooption is also widely practiced by party political and State institutions. In Guayaquil, local communities allow themselves to be coopted by political parties in the hope that they would receive services,(58) while in Mexico, many groups joined the governing party (the Institutional Revolutionary Party (PRI)) with the same expectations. Cooption seems most likely when there is a dominant political or power group which, despite its influence, actively seeks to widen its control still further.

"Routinization "

62. This is a form of integration achieved by government agencies when leaders are recognized and accommodated by the institution in order for this to allow it to fulfil its goals efficiently. "Technical" criteria are largely used by the agency to decide its priorities and procedures for the allocation of services to settlements.(34) Leaders are accommodated within these "routines" insofar as they will assist in the implementation of agency goals. Where leaders prove intransigent or uncooperative, agency staff may seek to undermine their power base either by negotiating with competing factions, or by using other mechanisms to drive a wedge between leaders and their followers. Tactics to achieve this include the direct signing of contracts between the agency and individual, thus bypassing the leadership, or by dropping the community from its works programme, thereby encouraging resident alienation from the leadership.(101,104) "Routinization" usually has the twin goals of overtly implementing actions with which the agency is charged, and (covertly) achieving social control through mediation and appeasement.(l05)

"Autonomy and independence"

63. Not all leaders succumb to any of the above forms of external mediation. Whether born of cynicism, or political radicalism, or because they genuinely believe that the best tactic in winning positive response from government is independence, some leaders have insisted upon maintaining distance from the State and have rejected overtures that might lead to any loss of autonomy. Radical leaders within the Citizen Movement of Madrid, and in urban social movements in Mexico (15,57) and in Rio de Janeiro (44,97) all resisted being formally associated (and potentially coopted) with any political party - including those on the political left. In the Philippines, too, various community groups forged themselves into an independent citywide association in 1977 and integrated the issue of land tenure into a broader struggle for a better life.(l) Although relatively rare, "autonomous" structures offer an interesting insight into the true nature of State-community relations. The position of independence and class mobilization that they adopt represent a considerable threat to the State which may respond in one (or several) ways: to coopt, to "divide and rule", to "starve" into submission or, ultimately. to repress.

64. It must be recognized that these relationships are not watertight categories that describe leader or more of the forms of linkage may overlap: as where a broker, "Big Man" or cacique is also coopted by a political party, or where several different types of link exist within the same settlement to different leaders. Also, as has already been observed, both leaders and agencies may not appear what they seem: there is often a spectrum of overt and covert rationales that governs the behaviour of institutions as well as leaders.(34) Nevertheless, the typology presented here does offer a broad insight into the ways in which leadership structures and external institutions engage with one other.

B. The determinants of leadership

65. Three broad sets of factors which act to shape the nature of leadership in different contexts can be identified: first, those which relate to the attributes of the individual leader; secondly, those that stem from characteristics of the community and, thirdly, those determinants that derive from the wider socio-political structure of the city or region. In the past, the little discussion that existed about leadership tended to focus upon community factors, or upon personal qualities of the individual leader. The argument here is that while these factors are important, their significance is secondary compared with the overriding influence of external pressures such as government and/or political parties.

1. Individual qualities of leadership

66. Chapter II has already dealt with leadership qualities in some depth and this will not be repeated here. Briefly, the argument is that leaders emerge because of their intrinsic qualities. They may have certain skills such as being able speakers, or possess charismatic personalities around which people mobilize.(27,98) Individual status may also provide an important basis for leadership. Such status may be earned. as when individuals have worked their way into positions of power, or economic influence.(6,8) Alternatively, it may be "ascribed" to certain individuals through inheritance and lineage - often the case for tribal leaders.(36,54,94) It was also observed, that in some societies leadership is reserved for the elders, and particularly for male elders. Age and sex are, therefore, key determinants of who exercises leadership in a community.

2. Settlement conditions and leadership type

67. It is argued here that leadership types are an outcome of settlement conditions. Thus, heterogeneity of social or economic groups is likely to lead to a greater variety of leadership opportunities and behaviour.(27) Generally, the larger the settlement the more difficult it is for one leader to maintain control, or to represent adequately the interests of all residents. Diversity in large settlements may derive from the different social and economic groups they accommodate as well as from the different tenure arrangements enjoyed by different sections of their populations (compare full owners, squatters, purchasers, "community land" holders and so on). Some settlements which expand their frontiers accretively may also contain groups whose arrival at different times sets them against each other: pioneers feel that their efforts may be threatened by later arrivals, particularly if the latter draw adverse publicity. In each case, competing groups can spawn different representational structures.

68. The opportunities for coercion and self-enrichment presented by different types of settlement, particularly during the earlier phases of their development, encourage the presence and activity of authoritarian leaders.(21) Once the basis of their control is removed, through the legalization of land tenure, and people have recourse to the law in settling disputes, the influence of such leaders is substantially reduced. Indeed, as settlement needs change through time so resident participation in community affairs and demands made of the representation structure also change.(61) Caciques are replaced by democratically-elected leaders who, in turn, give way to low-key "positional" leaders representing the interests of political parties, unions, religious organizations, and so on (see chapter II). In Kenyan self-help programmes, harambee leaders are selected according to their ability to perform different roles during the various stages of a grassroots project (initiation, organization and implementation); so settlement or project needs shape the emergence of leaders.(55)

3. External determinants of leadership

69. Individual and settlement attributes are obviously significant, but their importance is probably secondary to the following factors which shape the propensity for different types of leadership emergence. Primary determinants of leadership appear to be derived externally: from the regional culture; from the party-political structure; from national and local government; and from the actions of NGOs.

70. The regional culture has traditionally been identified as a key variable that determines leader-follower relations, particularly in the context of tribal society. Yet, although tribal elders were traditionally accorded leadership status in rural areas, the cultural basis of leadership has largely broken down in urban areas since different ethnic groups have been thrown together.(30) Even where ethnic group or lineage remains an important feature of leadership and local organization, there is some feeling that this is because of its value as a channel for mobilization rather than intrinsic traditional values that are attributed to it.(36) While traditional Confucian values of age and experience as determinants of leadership were respected by the New Community Movement in the Republic of Korea, this did not prevent the creation of a parallel structure of "new-style" leaders more geared to the needs of the Movement.(11)

71. Other "externally imposed" factors that determine leadership and which derive from the regional culture include traditional attitudes towards the selection of leaders from different castes, and upon the basis of gender. As has been seen, women do not generally figure as leaders in traditional society, especially among Muslim sects,(55) although there is some evidence that women held important leadership positions in Africa prior to the colonial period.(65) Caste in India is important: although leaders are not exclusively drawn from higher castes, the latter are much more likely to assume leadership functions.(32,37) Generally, though, traditional determinants of leadership appear to have less importance in contemporary rural and urban society than in the past.

72. Structural determinants emanating from party-political needs exercise enormous influence over the type of leadership to emerge in different situations. Party needs may be expressed in the desire to mobilize populations through settlement leaders. The latter are expected to act as mobilizers of national or party ideology such as the UNIP development of its "humanist" philosophy through its branch and constituency structures in Zambia.(79) Here party leaders and local leaders are often one and the same.(72) If ideological needs are not uppermost then the need to mobilize the vote may be. Political parties in many countries seek to establish leaders in local communities who will activate the vote on their behalf.(28,58,66,73,108) In Mexico, party militants continued to use local caciques to mobilize the vote on their behalf despite government and party headquarters directives to desist. The alternative was to invest resources and time in developing a grass-roots political organization in social contexts that were largely unfamiliar to the party militants.(21) "Political machines" may be established whereby local leaders form part of a wider network (usually built around an individual city hall politician) in which votes are traded for city services or local improvements.(21,25,79) Parties may also use the arena of settlement politics as a proving ground on which up-and-coming leaders can cut their political teeth.(87)

73. In addition to their attempts at mobilization, political parties may also seek to demobilize communities and local leadership structures which threaten the political system. Thus in Mexico, when various "independent" leaders broke with the normal practice of individual and vertical links with the political system and formed a horizontal urban social movement, the State acted to resolve the dispute through appeasement, and subsequently coopted or isolated breakaway leaders.(34)

74. Government needs for social stability and control are also important determinants of the type of leadership that emerges in low-income communities. Again the aim may be to mobilize the poor for ideological purposes or to assist in the implementation of a local development initiative. Examples of the former include the New Community Movement in the Republic of Korea and Ujamoa in the United Republic of Tanzania both of which sought to inculcate a nationalist ideology through carefully wrought leadership structures that were imposed upon local communities - a common feature also in societies such as China and the Union of Soviet Socialist Republics.(48) Governments have also often encouraged the creation of community representational structures that mobilize local residents to participate in community development. (34, 72, 77)

75. Governments also use leadership structures to undermine mobilization in those circumstances in which it poses a threat. This may be achieved through the manipulation of different factions: supporting one group in order to weaken another; the setting of two powerful local constituencies against one another. A classic tactic in this respect is one of differential servicing whereby adjacent settlements are partially accommodated in different ways to ensure that they no longer have common aims. Another tactic is to be seen to spread benefits equally, but thinly, over a number of settlements in order to pacify the population and to prolong the period that the population is dependent upon politicians.(20,108) The processes of cooption and incorporation described earlier are key mechanisms whereby leaders are used to demobilize the poor.(8.44) Repression of local leaders who do not conform is not unknown. but it is usually adopted as a last resort.(1,13,15,57)

76. Government action sometimes has the opposite effect. Where residents perceive a strong threat from outside this may have a powerful integrating effect upon the community. "Out-group" hostility thereby creates "in-group" solidarity.(21)

77. As well as the government's need to mobilize and to demobilize low-income populations, the external control of local leadership presents an additional advantage. By creating a complex and internally divisive structure for channelling community-State interaction, the government promotes the appearance that it is busy and concerned on behalf of the poor. In fact, the outflow of central resources can be carefully controlled and slowed down.(20) There are several ways of achieving this: either through the classic patron-client relationship or through a more formalized system of competitive bargaining for limited resources at a local municipal or sub- municipal level. In both cases the issue is one of how to divide the cake rather than how much cake is put on the table in the first place.(102)

78. One may take the view, therefore, that community-development programmes, in many parts of Latin America at least, offer governments a prime medium for political mediation and the exercise of social control.(105) As Burges notes (14):

"in general the settlement organization is converted into an intermediary for the organization of construction processes in which ideological control, political manipulation and the subordination of spontaneous forms of organization is the rule".

79. Local leadership is one of the key elements in this process and the emergence of local structures is primarily an outcome of political and governmental needs rather than a result of the particular attributes of leaders or settlement. The latter are secondary to the structural factors which shape the propensity for community participation and leadership to emerge.

80. Non-governmental organizations may also influence the nature of leadership in different situations. In chapter V of this report different types of NGO are identified but generally their aims can he stated as developing the welfare of the poor in one or more ways: moral and spiritual if it is a church; developmental in the case of institutions like OXFAM, Save the Children and, often, church organizations as well; recreational in the case of some social organizations. However, in all cases it is important to recognize that NGOs are under pressure to achieve results. Community leaders may be "placed" in a community to represent and to extend the base of an NGO. Sometimes their role may be to develop a basis for organized struggle within a neighbourhood (see, for example, the pioneering work by Alinsky in the United States during the early 1940s and subsequently by his supporters).(15) Elsewhere NGOs may use cooption to involve local residents in work on their behalf.

81. The extent to which NGOs will play a major role in influencing leadership structures appears to relate to the vacuum left by the absence of governmental or political structures. In exceptionally repressive societies, internationally supported and staffed agencies may be the only vehicle for local organization. The Church, too, may be especially influential in these circumstances. One of the features of so-called "bureaucratic-authoritarian" governments is their denial of working class participation and organization.(71) In Brazil during the early 1970s the Roman Catholic Church was the only effective outlet through which local residents could negotiate for services. Priests were, therefore, at the forefront of local community leadership throughout this period.(10,25)

C. Multiple representation: the significance of factionalism

82. Many writers have commented upon the importance of "cohesiveness" in determining successful community participation and the effectiveness of leadership,(2, 27, 38) yet many pressures exist which reduce the likelihood of a cohesive community and enhance the emergence of different factions of interest within a community, many of which will be represented by different (and often competing) leaders.

1. The determinants of factions

83. An important basis of factionalism is the variety of intra-settlement interests and their relative strengths. Circumstances vary between societies but examples include different tribal factions and different ethnic groups or castes.(32, 60, 108) Socio-economic heterogeneity exists within most poor settlements, especially where they have been established for several years; only at the outset are age, family structure, incomes, migrant status, employment category and so on likely to be broadly similar. Once different economic and social groups emerge they are likely to have different priorities about such issues as settlement upgrading, attitudes towards newcomers, and willingness to participate in mutual aid versus self-help. People also have different needs for leaders, and this results in a multiplicity of leadership structures some of which fulfill task-oriented functions, while others meet the spiritual or emotional needs of the residents.(77) There are important differences, too, between those who rent land or dwellings and "owners" (whether or not they have full legal title to land). Clearly the two groups have different investment stakes and these may conflict. Owners are likely to favour formal land title, the provision of services and rising land values; renters, while also benefiting from improved servicing and security, will almost certainly have to pay more in the form of rent. Among owners there are also often important divisions associated with the various ways in which land has been acquired. Purchasers from real-estate companies are in conflict with those who have squatted, purchased from government, bought from holders of "common land rights", inherited from the latter, acquired from a third party who previously squatted, bought, inherited and so on. If forms of land acquisition are complicated, so are the processes of "regularization" or "legalization".(101) The route towards legalizing irregular settlements depends upon the way in which land was alienated and implies both differences of procedure and, ultimately, of cost. Therefore it is not surprising that different factions emerge, particularly where the financial pickings for leaders are high.

84. The size and spatial organization of the settlement may also facilitate factionalism. Larger settlements are more difficult to control, and disaffected or separate interest groups have a greater opportunity of establishing themselves. In these areas, residents even appear less able to identify leaders' names than they are in smaller neighbourhoods.(34) The same applies where settlements are divided by a physical barrier such as a road or railway tracks; or where one area suffers a problem not shared by the whole community (such as flooding, landslips, or fire hazards). In such cases a would-be leader can build a support base around that particular issue.

85. Past performance of leaders is also likely to shape the propensity for factionalism. Where they have proved to be highly effective, non-exploitative, and amenable to all groups, then incipient attempts from rival leaders are less likely to be successful. However, where disillusion and cynicism are rife, allegiance may be switched, provided supporting another leader is not likely to result in intimidation against oneself or one's family.

86. Factionalism will also emerge where it is required or given some formal status by external organizations. In circumstances which were described earlier in this chapter where governmental or political advantage may be derived from links forged with leaders, the propensity for factions to exist is heightened.(5,100) A similar conclusion was reached by a reviewer of urban politics in African towns; he suggested that factionalism is most likely to occur when "the external environment offers positive inducements for active intercourse".(6) Certainly, expansion of the number of departments with responsibilities for irregular settlements in Mexico City between 1970 and 1976 led to competing and conflicting functions within the bureaucracy. At the same time, each department forged links with local leaders in individual settlements for which it was responsible and often for settlements where a rival department had jurisdiction. Thus legitimacy for rival leaders in the same settlement was conferred by rival agencies. Significant, too, was the fact that conflict observed on the ground between factions mirrored conflicts operating between rival political groups within the Government. To a certain extent, therefore, conflict was transferred down to the settlement level.(103)

2. The significance of factionalism

87. Factionalism and settlement heterogeneity usually detract from successful community development.(46) As one writer comments "fragmentation of a low-income community into competing political factions can seriously inhibit the propensity to participate".(21) It diverts effort and attention away from the key pressure points of demand-making. It weakens leader and agency authority in carrying out negotiations. The accusations and counter-accusations between competing factions sow confusion and cynicism in residents' minds and act to discourage participation and collaboration. In one settlement in the United Republic of Tanzania, poor community-development performance was an outcome of tribal and inter-generational rivalries which resulted in a high turnover of leadership.(60) Similarly, inter-caste rivalries were found to lead to weak and ineffective leadership in a study of an Indian village.(32)

88. While arguing that factionalism is generally counter- productive to community development, it is necessary to recognize that factions are only a symptom of an underlying structure which encourages their existence. Thus, any attempt to head off factionalism at a settlement level without confronting the underlying rationale for the vitality of the factions will have only a limited effect. Nevertheless, where several parallel leadership structures exist in a single settlement or village it is imperative that these be identified, along with their external links to supra-local organizations. Only in so doing can appropriate decisions and judgements be made about those with whom one might collaborate, on what terms, how they should be approached, in what order, and so on - points which will be dealt with in chapter VI.

D. The dynamics of leadership

89. By now it should be apparent that leadership is a complex phenomenon. It has been shown how leadership structures are largely determined by external socio-political processes and that they may comprise multiple and overlapping structures. Nor is leadership a static phenomenon and in the final section of this chapter the main factors which determine changes in leadership are identified.

1. Political and/or ideological changes

90. Under conditions of competitive party politics, changing leadership positions will reflect the vitality and ideological persuasion of individual parties. A return to democratic government after military rule furnishes new opportunities for leaders - as is happening in Brazil since the "political opening" (abertura) in 1983. Elsewhere, too, one observes an increasing tendency to open the political franchise, albeit in carefully constrained ways and often largely motivated by a desire for greater legitimacy for the ruling party. Where leadership positions are linked directly to the governing party, political fortunes will shape which leaders are "in" and which are "out" at any one point in time: a feature described for countries as different as Pakistan and Venezuela.(73, 108) As one local leader in Valencia, Venezuela, commented wryly after his party had won the national elections "For five years we have sat with our arms crossed; now it is the turn of the Adecos (the opposition party) to do the same".(34)

91. It is not only when parties change that leaders rise and fall. important ideological shifts or the onset of more liberal attitudes can have the same effect. In Mexico it is argued that caciquismo has declined in rural areas although there is debate about why: as an outcome of government action to eradicate it, or as a result of the penetration of capitalism into traditional subsistence economies.(7,21,57) Changes in the predominant forms of political mediation (patron-client, cooption and repression) will also change quite dramatically what is expected of leaders and their behaviour.(104)

2. Changes in government policy

92. It is not possible here to analyse what processes or motives inform significant shifts in State intervention.(34) Nonetheless, there appears to be a tendency for an increasingly technocratic commitment to underpin bureaucracy behaviour. In Latin America, at least, some Governments are finding that they can achieve social control effectively by becoming more efficient and technocratic in matters of urban management.(5,105) If this is a broadening trend, then it will substantially change the format of State-community relations in the future. These are likely to become more "routinized" and less open to particularistic patterns of decision taking. "Gatekeepers" rather than caudillos will exercise control over the distribution of resources. There is considerable evidence to suggest that both the structure of leadership and leaders themselves are responsive to these changes, and that behaviour alters accordingly.(34)

3. Changing settlement form and changes of leadership

93. There is also a close link between the dynamics of settlement development and the type of leader who emerges at different points in time.(55) Although the broad political and governmental structural features described earlier condition the overall operation of State-community relations and leadership, local variables also determine which leaders emerge in the community. In squatter settlements, for example, when initial tenure insecurity is high, the financial pickings for leaders are potentially large, and the community is on the defensive, there is a greater likelihood leaders will be male, tough, intimidatory, and aggressive in the tactics they employ with government and would-be competitors. As the fledgeling community wins acceptance, and the possibilities for intimidation decline through formal registration of plots, so the possibilities for more representational forms of leadership emerge. There is concern to speed up the introduction of services and the tendency for leaders to "drag their feet" in order to sustain a leadership position is less prevalent. Finally, once most services are won then public participation and interest atrophies. Leadership committees become defunct or are replaced by positional leaders who represent external organizations.(103) While this cycle of participation and involvement has been widely described, its association with a leadership dynamic has not, although the importance of different types of legitimacy obtained from irregular settlements at different periods of their physical integration with the city have been recognized elsewhere.(20)

94. Caciques, who also serve a useful purpose by delivering the vote for the governing party, cease to be an asset to the party if they become increasingly despotic and arbitrary in their treatment of the local community, and will find themselves isolated: the government will shift its support to others.(7) Residents also change their allegiance to particular leaders thereby assisting in a turn-over. Sometimes they shift their support for good reason; but followers can also prove very fickle. Even in those cases where evidence has been found of hardworking, honest and efficient leaders, they have often been denounced publicly. Residents may come to envy the successful leader and, in gradually distancing themselves, leave the leader isolated.(S8)

95. Thus, having analysed what factors shape the emergence of leaders, their role, internal conflicts, and the processes leading to their removal and displacement, a position has been reached to turn to an analysis of their impact. The next chapter begins to evaluate the ways in which the patterns and changes so far described affect community development.

IV. The impact of leadership upon low-income community development

1. The impart of leaders upon settlement improvement

96. Several authors have emphasized that leadership is a key variable in upgrading.(11,20,21,24,27,42,77) One reason for this is that effective leadership may get the settlement on to the servicing or legalization agenda (or prevent it from being dropped from it). Often this is done through the leader enlisting the help of patrons or contacts in government agencies.(2,20) Alternatively, if the leader is perceived by the authorities to be doing a good job of maintaining order and control within the community, then they may implicitly legitimize the settlement's activities, even if they do not formally recognize its legality, as was found in a study of Mathare Village 2, Nairobi.(77) A second reason is that effective leadership can maintain a momentum of pressure on agencies, and/or speed up the acquisition of goods and services through "jumping up" the servicing queue. This happens either through leaders soliciting help from their patrons,(20) or by mobilizing their communities to demonstrate an obvious show of support for powerful political parties. Thirdly, certain leaders may win concessions such as a reduction in taxes, legalization fees or service costs.(104) In the Baldia Township Regularization and Improvement Scheme in Karachi, Pakistan, local leaders fought for lower lease rates and achieved a slight modification in the prices originally recommended by the Metropolitan Corporation. They also succeeded in getting a clause of non-transferability in their lease documents amended.(108) Fourthly, effective leadership can minimize the need for widespread and ongoing mobilization by the majority of residents.(21,27,28,78) Indeed, participation in efforts to pressurize public agencies for services and infrastructure can be very time-consuming and to some extent it is desirable to channel community objectives through local leaders, especially where personal contact is the conventional mode of bargaining. However, a monopoly by a leader over the making of demands is not always benign. As observed in the preceding chapter, it may be a mechanism by which leaders further their personal interests and subordinate the community's objectives to those of the government or a political party.(7,28) Finally, "good" leaders have the ability to stimulate and heighten participation where required, such as during the implementation phases of a project.(11,21,55) For example, the most successful villages within the Saemaul Undong Movement in the Republic of Korea were those in which leaders were able to encourage cooperation and activity in building and improvement works among residents.(11)

97. Despite the alleged importance of local leaders in the physical aspects of community development, experience suggests that even "effective" leadership often achieves only marginal returns over those settlements where leadership is either non- existent or ineffective. It may well be that the relative success of development projects depends primarily on the following three factors, each of which is largely unrelated to leadership.

(a) The government's commitment to providing resources to low-income communities. The scale of resources dedicated to the improvement of low-income communities is often very limited and fluctuates over time.(75,103) Furthermore, a "development programme" itself may not have been established as the result of a genuine desire to improve low-income settlements. Often upgrading and improvement programmes are only a "front" to give the appearance of activity, while covertly the intention is to reduce social unrest.(l0,34) For example, the Baldia Township Regularization and Improvement Scheme was initiated as a means of securing grass-roots support for the Government during a period of intense activity by an opposition political party. However, the programme got no further than that of a policy statement before the Government was over-thrown.(l08)

(b) The nature of the socio-political system and the rationale that underpins leadership-government relations also act to limit the effectiveness of strong leaders (see table 3). Where the political context is characterized by personalism and the patron-client relationship, authoritarian leaders may be more effective than others in securing benefits for their communities.(28,35) Alternatively, in situations where service allocation is established on a technical, "routinized" basis, everyone waits their turn and leaders may only be effective insofar as they sustain momentum, coordinate residents and stimulate the process of implementation.(34) In instances like these, elected leaders who do not derive their legitimacy from patrons in government positions may be equally if not more successful than those who rely on personal contacts with officials to maintain their support base. Under a situation of competitive party politics, "positional" leaders who are either placed by, or affiliated to, the governing party are likely to achieve greater returns than those who are in opposition.(66) This may also apply to "traditional" leaders where the party in power has an important ethnic or tribal dimension. The period immediately before elections may result in the quickening of improvements, but the effects are usually shortlived.(5,28,57)


Table 3. The impact that local leaders are likely to have in securing of community improvements according to the conditioning relationship between State and leader

(c) The internal characteristics of the settlement itself can also influence success in physical development, and specifically whether or not the settlement has a relatively homogeneous population in terms of ethnicity, class, tenure and income. Several investigations have shown that when communities are fairly homogeneous and well-integrated, leadership is usually very effective.(11,21,22,27,38,60) Alternatively, somewhat ironically, those settlements which really need effective leadership, in order to mediate between different social or ethnic groups, often have weak or divided leadership and low rates of community participation.(27,46,103) For example, in communities in Kamataka in India, inter-caste rivalries led to intense factionalism and multiple, opposing leadership.(32) Similarly, leadership was ineffective in the village of Lukenge in Morogoro District in the United Republic of Tanzania, owing to bitter political struggles between the dominant tribes in the community.(60) Therefore, internally homogeneous settlements are likely to give rise to strong leaders with a wide support base, but since these communities are probably united in their objectives for, and committal to, community development in the first place, the leader himself may not be the critical catalyst of participation.

98. So it may he concluded that in terms of tile physical impact of leadership upon the settlement, the effects of strong leadership may he negligible, despite conventional wisdom to the contrary. Indeed, leadership is more likely to have a far greater influence in terms of its impact upon the shaping of inequalities within the community.

2. The impact of leaders upon socio-economic inequalities

99. Socio-economic inequality may increase as a direct result of the control by leaders over development programmes, with resources being channelled to favour the better-off segments of the community. This is especially likely where leadership legitimacy is not widely-baled; for example, in cases of authoritarian or traditional leadership structures.(l6,32,37,76) Here resources may be directed to those of the same class, ethnic group of lineage and, in some cases, to the leaders themselves. In Uttar Pradesh, India, for example, wealthy village leaders are usually high-caste Thakurs or Brahmins who generally succeed in capturing the benefits of development programmes for themselves and their immediate followers.(37) In Mathare Valley, Nairobi, local landlords monopolized a development programme which was assisted by the National Christian Council of Kenya and took advantage of the opportunity to develop open plots at the expense of the poorer groups within the settlement. They ended up creating a sharply stratified community composed of rich landlords, vulnerable squatters and exploited tenants.(88)

100. Radical and collective readerships may succeed in lowering inequality, such as was the case with the leaders of Movement of the Revolutionary Left in Chile between 1970 and 1973.(15,38) In some cases this may reflect the ideology of the development programme itself. For example, one of the main objectives of the Saemaul Undong Movement in the Republic of Korea was to reduce class and income inequalities by stressing joint participation in national development. Within this programme strong emphasis was placed on dual-leadership whereby leaders were to work alongside the traditional village eiders. One of the main selection criteria for leaders was an ability to mediate between the very rich and the very poor.(11) In situations where a political party seeks support and legitimacy, leaders may have an incentive to spread the benefits of a programme to more rather than fewer people in order to capture potential votes.(15) However, examples of leaders who have genuinely or consciously attempted (let alone succeeded) to reconcile class differences are few and far between.

101. Inequality of gender roles is another area in which local leadership may have a considerable impact. As was pointed out in chapter II, the majority of local leaders are male, and their position may he reinforced through active cooperation by an external agency, thereby deepening gender-ascriptive roles and ideology. Since most agencies choose to work with existing leaders, women do not get an opportunity to participate. For example, in several development projects, women are assigned to passive, non-participatory roles and rarely occupy leadership positions. Male control of projects not only reinforces the status quo of gender inequality, but also means that women are often excluded from the benefits of development programmes.(14,59,76)

102. In circumstances where women become leaders, this may lead to a questioning of gender roles and to an opening of opportunities for women.(58) For example, women who are trained as voluntary health workers often end up taking responsibility for the broader aspects of community-development programmes and in many cases assume leadership of elements of the project.(70,91) However, female leadership poses several problems. Married women not only face opposition from their partners, but also from other women. The latter is especially true where the leader in question has children, since it is interpreted as a sign of neglect, and they are often alienated as a result. In many respects the resentment felt by other women conceals a deep-rooted envy of the greater freedom and mobility gained by women leaders as a result of their involvement in the public life of the community.(58)

103. Finally, inequality may be heightened by the existence of leadership factions in which the less successful leaders (and their followers) miss out.(21) For example, the Farmers Service Cooperative Society Programme in Kamataka, scarcely aided the low-caste groups in the village of Seshagirihalli where leadership consisted of five people from three different castes. Although one leader was from a low caste, he was very unpopular, and exploited the Harijans and the lower castes. As a result they received no benefits from the Programme. In Uttar Pradesh, an educated, low-caste leader who was an obvious rallying point for other low castes, did not stand as a candidate in the village elections for fear of reprisals from the local elite.(37)

3. The impact of leaders as catalysts of participation

104. The broad argument in this section is that the propensity for participation, like the nature of leadership itself, is an outcome either of external processes, such as the nature of the government's relationship to low-income groups, or internal settlement characteristics such as socio-economic homogeneity. To some extent leadership exercises an independent influence. This may be achieved through the ability of leaders to organize and the closeness of their relationship with residents. For example, the most successful development villages under the Saemaul Undong Movement programme in the Republic of Korea were those where the leadership was most able to mobilize participation. This was particularly the case where there had been a tradition of communal labour long before the Movement had started.(11) Similarly, charismatic leaders who inspire trust and confidence are well-equipped to inculcate the belief that participation is necessary or desirable, especially where the community is relatively homogeneous.(27,76) The chairman of the committee of Mathare Village 2, Nairobi, was a very effective and dynamic leader; he not only had good contacts and actively participated in community work efforts, but also had great personal integrity.(77)

105. Strong leaders may also shape the nature of community participation in accordance with their main goals. Where their covert motive is to demobilize residents to reduce pressure on government, or they want to mobilize their communities for political purposes, the outcome is likely to be nominal and ineffective participation, accompanied by high levels of cynicism and disillusionment from residents.(l7,21) Alternatively, where the leader has genuine overt goals for community improvement, then if residents perceive him as honest they may raise their level of participation. (See table 2 for the extent to which different types of leader are likely to raise or reduce participation.) It should be borne in mind, however, that the success of any community development programme will also depend on: (a) whether the programme is relevant to people's needs; and (b) the community's understanding of the programme. (27) Indeed, Olesen (66) suggests that convincing leaders of the need and appropriateness of an upgrading scheme offers relatively small advantage. Ideally the people themselves should be reached.

4. The impact of leaders on opinion and political awareness

106. It has been widely acknowledged that the urban residential context is an important forum for political learning.(6,8,20,21) In low-income neighbourhoods local leaders may act as key opinion leaders and exert considerable influence over the "consciousness" of residents. For example, they frequently play a major role in shaping people's perception of the political/governmental system and actors within it, including agencies.(21) The critical significance of local leadership in interpreting policy to low-income groups was recognized and capitalized upon by the Government of the Republic of Korea in the Saemaul Undong Movement. A National Saemaul Training Institute was established by the Ministry of Home Affairs which ran a compulsory programme for potential Saemaul Undong leaders. The training course comprised a great deal of ideological indoctrination of the principles of self-reliance and diligence - the comer-stones of the new Movement. It was also intended to increase grass-roots support for President Park.(11) Leaders are also important in terms of the extent to which they identify a class struggle of the poor and raise consciousness of key political issues. For example, the female leader of the San Judas Barrio Project in Managua, Nicaragua, attempted to demonstrate the fundamental interrelationship between class and gender inequalities during the implementation phase of a post-revolutionary, "bottom-up" housing project which raised major problems of cooperation between men and women.(99) However, many leaders do not try to raise consciousness of wider issues among residents. Instead, they focus attention on specific problems and personalized, local struggle. This perpetuates their legitimacy, and helps to take pressure off the State. which thereby further demonstrates their allegiance to their patrons.(7,20,28)

107. Leaders may also play an important role in the community's interpretation of events, by relaying their perceived reality of the situation, either in support of, or against, agencies or individuals. For example, they might exaggerate the significance of past claims and demands for services that have been received by the community in order to win support for themselves and their contacts. Alternatively, and particularly in cases where there are rival leadership factions, they may play down the achievements of other leaders and their patrons.(21) For example, in Santo Domingo Los Reyes, a low-income community in Mexico City, rival leaders frequently denounced each other, threatened the supporters of opposing factions and vilified the outside agencies with whom the other had connection.(103)

108. To conclude this analysis of the impact of local leaders on community development, although it is argued that leadership only exerts a marginal effect upon the actual physical improvements to low-income settlements, its importance is critical in many other ways and cannot therefore be ignored.

V. Leaders and external agents: cooperation or conflict?

109. Most low-income housing projects, whether initiated "top-down" or "bottom-up", involve intervention from an external agent in one form or another. Intervention may range from financial, technical or advisory assistance, to supervision and/or control of programme formulation, implementation and management. The nature and potential success of intervention depends on several factors operating both at the level of the agency and the level of the community. Particularly important is the issue of leader-agency interaction, since local leaders may be key mediators between the two. Indeed, because of the likely differences in approach to settlement improvement on behalf of community members, the local elite and the external institution, community development workers are often employed by the implementing agency to harmonize the interests of the sets of actors.(63) In some cases, these workers are trained outside the community; in others they are recruited from within the settlement and trained specifically for their role in negotiation, information dissemination and mobilization.

110. This chapter examines the characteristics of various types of external agency likely to intervene in the field of low-income housing programmes, looking specifically at the way in which different types of outside institution are likely to respond to self-help communities and their leadership. In many respects the type of organization will influence the type of programme adopted and degree to which community participation is an integral and important project component. In some cases, "participation" is restricted to implementation and is equated with the community providing labour for a project which has been devised and structured outside the settlement with little reference to community needs. In these cases the external agency's interest in leadership is likely to be limited to identifying those leaders who are well disposed towards officials and implementation. In other cases, "participation" will have broader significance and embrace the notion of community control over projects, in which case the agency may play a more minor, advisory role.(40,84) Here leaders are key representatives of their communities and, therefore, occupy a critical position in the organization of community development.

111. The chapter then looks at the degree of covergence between the aims of the agency and those of community leaders, especially as they concern the underlying rationale for participation and the methods by which programme objectives are to be achieved. The extent to which the priorities of agencies and leaders, as well as those of the community, accord with each other will be a critical element in determining the extent of cooperation, and thus the outcome of the programme. Finally, the specific operational issues relating to leader-agency interaction which may determine the success of a development project will be examined. These include decisions about whom to approach, in what order, the timing of intervention, where to meet and so on.

A. External agents involved in low-income housing projects

112. In this section the different types of agents likely to be involved in low-income housing projects are examined, particularly the aspects of scale of institutions, the resources at their disposal, their reasons for and modes of intervention, types of projects, and provision for community participation (see table 4). It should be pointed out that in many cases, different types of external agent work in conjunction with one another. For example, university teams often work with city councils or metropolitan corporations as has been the case in Karachi (105) and Bangkok.(24) Moreover, projects devised by central government are often administered through the local administrative machinery, such as the Saemaul Undong Movement in the Republic of Korea,(11) and the Kampung Improvement Programme (KIP) in Indonesia.(84)

113. National government ministries or agencies are frequently involved in the establishment of country-wide housing programmes. They usually have access to considerable financial and technical resources and employ highly-trained personnel (see table 4). Programmes are generally large-scale and implemented "top-down" through specialized agencies and local government, with little direct contact between central government officials and community representatives. The likelihood of widespread popular participation in the establishment of priorities and design of the project is small.(46) This is, in part, a function of the scale of programmes and the bureaucracy surrounding them, leading to inflexibility and administrative bottle-necks.(75) Governments may be more concerned with efficiency and technocratic considerations than with involving the poor in the planning and upgrading of their settlements. However, as was mentioned in the previous chapter, there may be other motives for intervention: covert aims of government upgrading and improvement schemes are often linked to the need for maintaining social and political control. While nominal participation may serve this purpose, it is unlikely that effective participation would achieve these ends, since government may not he able to meet the community's demands. Therefore, the usual mode of intervention is "top-down" and leaders compliant with governmental objectives are likely to be favoured.(75) The reluctance to incorporate community needs into housing programmes may well account for the failures found in many governmental schemes, such as those experienced by KIP.(85)


Table 4. Types of external agents likely to intervene in low-income settlements

114. Local government refers to municipal governments, city councils and metropolitan corporations. Resources vary according to city size and the town's relationship with national government. Where urban areas are relatively wealthy, city-wide programmes may be implemented; where they are not, attention is likely to be focused upon specific settlement improvement projects. The scope for community participation in project design and implementation is likely to vary depending on the size of the programme but, as in national schemes, participation is usually restricted to implementation. For example, the Lusaka urban upgrading project, funded jointly by the Lusaka City Council, the Government of Zambia and the World Bank, avoided involving the community in the initial decision-making for fear of unrest if the programme failed to take off. This decision was also made for the sake of speeding up project implementation, since extensive dialogue with the communities would have been time- consuming.(45,68,72) However, even where the community plays a minor role, settlement leaders often become heavily involved. Local branch representatives of UNIP, for example, were vested with considerable responsibility to mobilize support for the Lusaka project, along with trained community development workers.(78) The overt aims of local government upgrading schemes may be to improve housing conditions in the city, but they will also be concerned with reducing social unrest; therefore contact with, and cooption of, settlement leaders is likely to be a key tactic of local authorities.

115. Political organizations, such as political parties, as well as individual politicians are likely to become involved in low- income housing projects where they have access to government resources or seek to use local settlement issues as a basis for building up support. Political organizations include a wide range of personnel ranging from top-level politicians to local branch leaders and activists. Here, contact with communities is likely to be forged "top-down" either through a politician's personal links with settlement leaders or through the involvement of local parties in community affairs. Rather than implementing programmes across the board, political parties will probably become involved in particular settlements where they see most political advantage to be won - either in votes or in reducing dissent. As such, the leadership structure is likely to be important. If the aim is to coopt, it may be limited to certain leaders vested with considerable responsibility for shaping political opinions within the settlement. Alternatively, radical opposition parties may seek to spread an ideology of popular participation in which case broad-based community involvement and collective leadership may be high.

116. Large-scale international development agencies, such as the World Bank, act similarly to national government organizations (see table 4). They generally have access to considerable financial and technical resources and employ highly-trained specialists from a number of countries. Programmes are likely to be devised not only outside the communities, but also largely outside the countries concerned. Given the large and often inflexible nature of international development flexible nature of international programmes, community participation is likely to be minimal in the formulation stage, but higher during implementation, with widespread self-help and mutual aid being encouraged to reduce costs. However, even implementation is likely to be organized in a "top-down" fashion with large numbers of outside workers and advisers employed to supervise the community. This may then threaten the interests of local leaders.

117. Small/local NGOs, as the term suggests, are likely to have limited resources and to engage in small-scale local projects. They are more able (and probably more willing) to respond to local needs and initiatives than are larger, more bureaucratic institutions.(46) Their rationale for intervention is usually to assist low-income communities to help themselves, although project staff, recruited both locally and from abroad, may sometimes adopt paternalistic attitudes.(75) NGOs usually devise or support projects where the scope for broad-based popular participation is high in all project stages from formulation through to management. Thus the issue of local leadership may be especially problematic if it is not representative of community interests. Owing to the likelihood of NGOs having greater knowledge and sensitivity in dealing with local situations, projects are likely to achieve high success rates, although they may have limited replicability.

118. Academic institutions and their members often become involved in low-income housing and community development projects, either independently or in conjunction with an agency. Reasons for intervention vary and may include a genuine desire to assist in the physical and social improvement of low-income communities, training for their own professional development workers, and research into community-development techniques, as well as student-initiated "action" research which seeks to blend technical assistance with political consciousness-raising of residents. Examples of the latter include student groups from the self-governing school of architecture at Mexico City's National Autonomous University who were closely involved in many locally- initiated upgrading schemes throughout the city during the early 1970s and encouraged residents to identify with, and participate in, wider social struggles. Examples of the former include the African Regional Health Education Centre in Ibadan, Nigeria, which organized groups of students to assist communities with health and upgrading schemes as part of their training as professional health workers.(l2) The Vicos Experiment in Peru was carried out by a team of social anthropologists from Cornell University with the aim of creating local decision-making organizations among 300 Quechua-speaking tenant farmers to enable them to manage a former hacienda (estate).(42) Academic institutions usually work on individual projects which include a strong emphasis on popular participation. Community leaders who are governed by motives of personal gain rather than more altruistic objectives may not be persuaded of the desirability of such external involvement.

119. Church groups and voluntary sector organizations have traditionally been a rallying point for low-income communities in many areas of the world. The organizations are either locally-based, such as Co-Clube in Recife, Brazil,(88) or part of a national organization such as the National Christian Council of Kenya,(88) or the Christian Family Movement in the Philippines.(7) Like NGOs and academic institutions, they are likely to become involved in local projects and to respond to "bottom-up" initiatives with a strong measure of popular participation. The rationale for intervention, especially with religious charities, is often to facilitate an improvement in the social and moral welfare of low-income groups, although they may be guilty of paternalism in their aims and methods. These groups usually enter settlements where there is a leadership "vacuum", but one could foresee difficulties in those situations where existing local leaders do not share the ideology or religion of the external group.

B. Aims and priorities. convergence or divergence

120. Having identified the main types of external agents involved in housing and upgrading projects, it is now possible to pinpoint the range of issues which are likely to prove conflictive in pursuing their goals. Project success depends to a large degree on whether there is convergence or divergence between both the aims of external agencies and those of local leaders, and in the methods adopted to fulfill objectives. In some cases, the objectives of the external agency may be totally alien to the needs and priorities of the community it wants to "assist"; in others the aims of both sets of actors might converge, but the methods may not. For example, the agency might see its role as an initiator, and regard the community's contribution primarily as one of providing labour with which their schemes can be implemented, yet residents might feel resentment at contributing unpaid labour when project staff are salaried.(42) In considering convergence and divergence of aims, one should attempt to examine the relevant issues from the perspective of ordinary community members as well, given that certain leaders may not be particularly representative of their followers and have personal interests uppermost.

1. Possible sources of conflict over aim

Participation

121. The scope and nature of popular participation in a development project is likely to be a major source of tension between the community, local leaders and the external agency. In addition to the often cited fundamental disagreement over the extent of community participation in decision-making and formulation as well as implementation, there may also be disagreement over who "controls" participation in the project. This latter point is less widely documented. Specifically, local leaders may not wish to see their basis of legitimacy eroded through the agency deciding to supervise the community directly. The aims of the external agency may be to develop consciousness and solidarity which might run counter to the aims of local leaders who wish to maintain the community's "need" for a hierarchical organizational structure. There may also be disagreement over the spill-over effect of participation into other non-community development areas such as party political demonstrations. While leaders and residents may regard the project as a vehicle for achieving greater political autonomy and strength, the external agent may wish to do precisely the opposite. Sometimes it works the other way of course. Agencies may wish to use community development as a pretext for party political mobilization about which leaders are at best cynical but tolerant, and at worst hostile and in opposition.

Timing

122. The timing of a project may also result in conflict. For example, while the external agency generally wants to get a programme underway as soon as possible, this may not be appropriate for leaders or residents. The actual process of identifying and approaching local leaders is very time-consuming but if an agency blunders at the outset, initiatives may be killed before they have had a change to get off the ground. The length of time needed was one main reason why project formulation in Lusaka was kept almost entirely out of residents' hands and led to failure to mobilize effective support and participation in later stages of the programme.(45,68) The timing of implementation may also cause conflicts. If agencies have borrowed money to finance the project, and require the community to work solidly over a period of a few months, residents may object if it means they have to take time off work and lose income. Leaders may also wish to drag out discussion, negotiation and implementation in order to extend their indispensability. Another issue is whether staff, either external or community-trained, and leaders, should be full- or part-time. In this instance, much depends on previous experience of the development agency and which approach has worked best in the past, whether local leaders have full-time jobs, and pragmatic considerations such as cost.

Priorities/selection of development goals

123. A critical issue here is "who decides?" Planners may feel they know what is best for a community, which often leads to conflict with the community itself over the establishment of priorities.(44) Leaders will know best what their followers need, but their needs may be impracticable in technical or financial terms. Planners will tend to go for expediency and residents may lose interest if their demands are not met. Leaders for their part may have different priorities from those of community members. For example, in Baldia Township, Karachi, both government officials and local leaders prioritized regularization, while residents prioritized services. Government wanted regularization for the purposes of generating income to provide services. Leaders, who often owned several plots within their settlements, wanted legalization to increase land values and to raise rents. Residents felt that if services arrived, then legalization would be a fait so they preferred to receive services rather than commit themselves to paying higher rates for land.(108) Another conflictive issue is how priorities are assessed: how much weight should be given to the views of planners, leaders and residents, respectively?

Cost recovery

124. Cost recovery is usually one of the most problematic areas of housing projects: loans commit residents for several years and there is often a high default rate on repayments.(3,59) Conflicts of aims and interests may arise over the following issues. First, the amount the community is supposed to provide relative to that of the external agency (i.e., the level of subsidy and the opportunities for subsidy). Secondly, the issue of repayment terms. The agency may have borrowed money to finance the project and so needs to recover it rapidly. Leaders, however, may wish to extend repayment terms, not only for personal financial reasons, but also to maintain their key negotiating role in the community. Residents may find it difficult to repay over a short period of time. Another problem is actually setting repayment terms which fit the payment capacity of all groups within the community. This will depend very much on the occupational structure of the settlement. Problems might arise over whether rates should be fixed as an absolute value or as a percentage of mean monthly income, and over what should be done about those groups in the settlement who are unable to pay on the same terms as the majority. A particular source of strain is where the agency explicitly or implicitly expects leaders to exercise pressure on recalcitrant non-payers. Not surprisingly, leaders resent a rent collector role.

Administration

125. Finally, administration of the project is likely to be controversial, especially in terms of local leadership. For example, who should staff the project? It may be more expedient for the agency to employ its own community-development workers, rather than use valuable time to train local residents. However, if the agency's aims are to increase the skills and managerial capacity of low-income communities, local leaders may resent the training of ordinary residents to positions alongside their own. What should be the existing leader's role? The aims of local leaders and project staff may differ widely depending on their respective motives and expectations for control of project administration. How should staff liaison be established? Should the agency report to the local community organization's headquarters or vice versa? Should contracting for specialized labour be done on an individual household or community-wide basis? Finally, who should supervise on-going and future maintenance of the project?

C. Sensitive issues in agency involvement with community leadership

126. Given the examples of converging and diverging aims cited in the previous section, one cannot make the assumption that development agencies will be successful in their interventions in low-income communities. Obviously the success of a development programme depends on a number of different factors. Rifkin has identified these factors as falling into two main groups termed "passive" and "active", respectively.(76) The first group refers to certain structural elements in a society which are likely to play a major role in the relative success of development programmes, but which cannot effectively be changed - for example, the political and cultural background against which the project takes place. The second group refers to more specific, lower-level issues over which some influence can be exerted by the key actors. These "active", or dynamic, factors include such issues as community organization and programme management and form the basis of the following discussion.

127. Given a high degree of cultural specificity and the wide range of socio-economic and political situations in which development programmes take place, one should not over-generalize about appropriate or inappropriate behaviour on behalf derstand why the issues identified are sensitive and how different types of agency have responded to the challenge in different cultural contexts.

1. Initiation

128. A critical issue in leader-agency interaction is who precipitates intervention, since to a large degree the actor making the approach will probably have the clearest idea about what the project should entail. Initiators may have a great deal of influence on the subsequent progress of the project, since they are usually the most committed to its success. Certainly, there seems to be a clear correlation between the source of initiation and control over project decision-making. For example, in a "bottom-up" housing project in Managua, Nicaragua, the Housing Ministry played a low-key role throughout the project and agreed with most of the leaders' suggestions regarding formulation and implementation. This was also a response to the ideological background against which the project took place, where much emphasis was being laid on grass-roots involvement in building post-revolutionary Nicaraguan society.(99) Conversely, the Lusaka project which was initiated "top-down" by various external agencies, deliberately avoided consulting community members in the early stages of project formulation. This resulted in inappropriately planned plot arrangements and a lack of enthusiasm and commitment on the part of residents.(45, 68, 72)

2. Approach and communication

129. The approach of the agency in terms of protocol, attitudes and timing is also an important operational issue in leader-agency interaction, since through insensitive behaviour, external institutions often alienate community leaders from the goals they wish to achieve. Agencies often enter communities without respecting traditional channels of communication, or paying due attention to the process of identifying leaders.(83) Their attitudes are frequently patronizing, paternalistic, authoritarian or sexist. Residents in a "top-down" community development project in Valle de la Esperanza, El Salvador, for example, felt they were being treated as "social inferiors" rather than as responsible citizens, and their resentment was reflected in very low rates of community participation.(42) Similarly in Santo Domingo de los Reyes, Mexico City, the failure of one agency to discuss its proposals with all the important leaders led to that agency losing credibility with large sections of the settlement.(103)

130. Major benefits can result from the agency making an overt display of less arrogant, more sensitive opinions. For example, in Bangkok, the National Housing Authority and Mahidol University team genuinely felt that their upgrading programme would be better if the residents were actively involved - an attitude which ultimately gained the active support and cooperation of the community in that project.(24)

131. With regard to protocol, in Ibadan, Nigeria, students from the African Regional Health Educatation Centre (ARHEC) made a potentially disastrous mistake by trying to set up a local improvement scheme without the consent or approval of the traditional council of elders. Fortunately, however, they rectified this by contacting the elders and discussing their ideas at length with them as soon as they realized their error.(12)

132. As far as timing is concerned, agencies often forgo the need to spend a lengthy period trying to identify local leadership structures, or they may withdraw from the project before enough momentum has been established within the community to sustain improvement efforts. The problems associated with this have been recognized by certain institutions. For example, a recent policy recommendation for United Nations Children's Fund projects has suggested that the normal three-year cycle of agency involvement and evaluation should be raised to 10 years; otherwise there is insufficient time to build up loyalty, trust and commitment within the community.(91) The team of specialists from Cornell University, who spent several years closely working in the Vicos experiment in community development in Peru, demonstrated that a sensitive, gradualist approach was the only effective method to gain acceptance and to be able to hand over managerial skills to the community.(42)

3. Response to community structure

133. Related to the above, how have agencies proceeded in setting up local organizations for the implementation of projects? This is a very problematic issue and perhaps the single most important in determining whether a project is successful. The key question here is whether agencies have achieved higher success rates working in conjunction with, or independently of, traditional community structures. It is a critical question, since bypassing these structures may lead not only to failure of the agency in teens of achieving its objectives, but also even prevent the implementation of a project altogether by arousing opposition from existing leaders. In addition, they frequently exclude certain groups from participation and decision-making (sometimes disadvantaged minorities whom the agency may most want to reach). Therefore, if the agency chooses to work with existing leaders, it must recognize that it is to some degree legitimizing and reinforced traditional patterns of inequality.(27) Yet newly- created structures, especially if they are imposed from outside via community development workers, may not be respected by local communities who have traditional patterns of leadership selection. Both approaches have pitfalls: one may threaten the interests of the community elite thereby jeopardizing the programme; the other may further their interests, thereby perverting the agency's original objectives to the extent of misrepresentation of institutional aims and alienation by the community.(42,72,76) As is demonstrated below. the wholesale adoption of either approach is potentially conflictive depending on the local political, cultural and economic background. Ideally, responses to the community organizational structure should be drawn up only after a thorough appraisal and evaluation of the existing settlement situation.

134. Experiments in creating new structures have frequently proved unsuccessful. In India, for example, where trained workers from outside the community were brought in to run a health project, the resentment felt by the community elite was so intense that a paramedic was murdered.(76) However, a similar method had a positive outcome in an Iranian project, where the introduction of outside community-development workers successfully reduced the control exercised over a health project by existing male-dominated leadership.(76) In certain cases, existing leadership has been imaginatively incorporated into new forms of community-development organization. For example, in the Republic of Korea respect was paid to traditional leaders who were invited and encouraged, by means of incentives and privileges, to work alongside younger, more dynamic leaders from the Saemaul Undong Movement.(11) In Bangkok, existing leaders, who were culturally obliged to protect and represent their communities, became key promoters of a housing and community improvement programme jointly set up by the National Housing Authority and a team of specialists from Mahidol University.(24)

135. Even where leaders are sympathetic and compliant with agency aims, too much overt cooperation with them to the exclusion of direct dialogue and contact with the rest of the community may alienate residents from the outside institution.(12) It can also work to the detriment of disadvantaged groups within the settlement. For example, in the Nairobi Mathare Valley project cited earlier, a "bottom-up" initiative assisted by the National Christian Council of Kenya, a great deal of freedom was given to existing leaders. The endproduct of relative community autonomy was that better-off groups benefited to the extent that they intensified socio-economic divisions within the settlement.(88)

4. Decision-making and control over resources

136. The degree to which different sets of actors are empowered to make decisions over the type of project needed by the community and appropriate methods of project implementation is also a critical issue in leader-agency interaction. If community leaders and their followers are excluded from certain aspects of project execution (particularly key phases such as formulation and design), then high participation by the community in implementation, maintenance and cost recovery phases is unlikely.(84) Neither is the project likely to sustain itself once the external agency has withdrawn.

137. A classic example of conflict around this issue is the Lusaka project, where key decisions on project design were taken outside the community in Washington, D.C., and the Zambian Government headquarters. The outcome was low rates of resident commitment and participation.(45,72) With regard to control over skills, in the community of Oke Foko in Ibadan, Nigeria, ARHEC students had neither the time nor forethought to pass on crucial negotiating and lobbying techniques to traditional leaders. This had severe repercussions since local leaders were unable to make overtures to the relevant authorities by themselves.(l2) It has been argued that the greater the degree of community involvement in local decision-making, the better the fit between community needs and upgrading programmes.(40) However, since most development agencies do not work under such ideal conditions, there are bound to be problems. This will obviously depend to a large extent on the local political factors. In repressive regimes, community residents are unlikely to be able to exercise effective power and influence over the relevant authorities. However, in more progressive regimes, the community may be positively encouraged to take responsibility in local decision- making. For example, in Managua, the Housing Ministry played a very low-key role in housing project implementation, and transferred substantial responsibility for resource allocation to the community.(99)

5. Programme organization, management, staffing and training

138. This issue was raised briefly in the section on aims. Here concentration is focused on the operational aspects of leader- agency interaction with regard to programme management. Decisions over project administration in some respects links in with the discussion of agency responses to community organizational and leadership structures. The degree to which community administration is controlled by the outside institution will also reflect the latter's attitude towards the beneficiaries.

139. An important determinant of the success of the Guayaquil project in Ecuador was that the women were salaried. A major obstacle to successful community-agency relations is often posed by the fact that external bodies assume local residents will donate their labour free-of-charge. Cooperation is far more likely to be achieved if they are paid. For example, the Co-Clube project in Coque, Brazil, assisted by a French NGO, The Brothers of Men, and the municipal government through the Ministry of Health, recognized the importance of involving the community closely in project activities and from the outset employed local recruits to work alongside professional staff.(90)

140. A final point to note is that, despite the resentment existing elders might feel at the active involvement of other community members in the programme and direct contact between them and the external agency through training, there are often positive effects for groups within the settlement who do not normally have the opportunity to become involved in local affairs. The katiwalas (locally-trained women health workers) in the Philippines often come to assume a broader role in their communities through the skills and experience derived from their involvement in health programmes.(7) Similarly, female involvement in a pre-school feeding programme in Guayaquil led to increased participation of women as leaders.(91)

141. Having examined a range of problems faced by external agents in their interaction with leaders, it is now possible to discuss ways in which certain of the conflicts arising from intervention may be circumvented or overcome.

VI. Reaching the poor through community leaders; an agenda of points for consideration

142. Potential sources of conflict in leader-agency interaction having been identified in chapter V, the present chapter addresses the issues relating to leadership that institutions involved in housing and upgrading schemes should bear in mind when intervening in low-income settlements, and suggests ways of dealing with the obvious questions and uncertainties that arise in the course of that intervention. These include: methods of identifying leaders and different types of leadership structures; making contact with leaders and on-going liaison; how to make a decision about when to work alongside existing leadership structures or to create new ones; how to respond to changes in the community and leadership structure engendered by intervention and, finally, how to react to overtures made by other outside institutions once the community is opened up to external pressure.

A. Low to identify leaders and different types of leadership

143. As pointed out in the previous chapter, spending sufficient time to identify key leaders in the community is vitally important. Yet it is also difficult to achieve since agencies usually have limited time and finance. However, short-cuts at this stage may result in alienation from leaders, a subsequent lack of cooperation and major obstacles to project implementation.

144. The process of identifying leaders can be difficult for three main reasons. First, agency staff often have preconceived notions as to who are likely to be community leaders. For example, in the Kampung Improvement Programme in Indonesia the implementing agency felt that only those people with an academic education were capable of adopting leadership roles.(85) Rigid preconceptions held by agency personnel may lead them to approach the wrong people in the settlement and to alienate those in power. Secondly, there are considerable difficulties associated with making enquiries within the community. In many cases community members are reluctant to identify their leaders to outsiders, and, further more, leaders may resent the fact that preliminary enquiries have been made without their approval or permission. Thirdly, in some settlements there is more than one leader and the agency may miss the most influential one.

145. Where a project has arisen from a "bottom-up" initiative, it may be slightly easier to identify community leaders since the key personnel of the settlement are usually, though not always, involved in making the overtures to the agency. The way in which the approach is made may provide pointers to the type of leadership it is. For example, if leaders come accompanied by committee members with a formal petition signed by residents then leadership is more likely to be community representational rather than authoritarian. If the leader makes contact through personal introduction from a patron, then he is more likely to be either a positional or cacique-type figure. One must also be wary that the leader of a small or marginal faction may have made the approach.

146. When the approach is "top-down", the process of leader identification is more difficult, except where there is an immediate and obvious contact. On the whole, it is recommended that the agency research into local leadership structures from a distance in order that an inappropriate move is not made at a stage which might jeopardize the programme altogether. Agencies should probably work with a combination of the following methods.

147. Newspaper scans are probably the best method of identifying local leaders without setting foot in the settlement.(103) Coverage of issues relating to local irregular settlements are likely to yield pointers to leaders and other influential people within the community, and their contacts with outside authorities, and to highlight sensitive issues that have been important in the past. The advantage here is that the external agency runs no risk of compromising itself until actual contact is made with relevant personnel.

148. Consultation with other agencies which have been involved with the community in the past may be useful to find out who the leaders are, provided that the discussions do not become public knowledge. This approach is most appropriate where there is easy informal access to an agency, especially if it has had a good relationship with the settlement. However, there is a danger that the external institution will become tainted if it is seen to associate with a body which has had problems with the community in the past. Hence the need for discretion in making this sort of approach.

149. Consultation by means of socio-metric tests among local influential people who are relatively "neutral" such as the local priest, ward chief, municipal delegate, mayor, or public works engineers could prove useful. Sociometric techniques were used in a survey of local power structures in rural communities in Mexico. Groups of people were asked two questions: (a) "Who do you believe are the five most important people in the region/settlement in order of importance'?", and (b) "Of the people whom you know, who is the most capable and respected so that at any given moment they could represent the region/settlement to the general public/city?" Replies were presented in terms of a "sociogram" displaying a series of lines between the respondent and the nominee.(7) Obviously the technique has limitations, not least because of the subjective nature of the responses, but it is probably useful as a first step in identifying potential leaders. As a last resort, sociometric techniques could be used among residents within the community, especially in cases where there is an apparently weak organizational structure - often the case in a recently established sites-and-services scheme for example. However, as mentioned previously, leaders may resent direct contact at this stage between the agency and the community and the other methods should probably be tried first.

B. Making contact with leaders and on-going liaison

150. No two settlements are exactly alike and the final decision about whom to approach, the form in which contact is made, on whose territory and so on can only be taken by the agency bearing in mind local conditions. The following check-list is offered as a guide to the points that should be considered. The "types of leader" refer to those identified earlier in chapter 11 - authoritarian, positional, traditional and community-representational.

151. There are several ways in which to meet local leaders. Making contact is a sensitive issue, however, and none of the methods described below is completely satisfactory. Selection of the most appropriate method will depend on the type of leadership structure in the settlement and local socio-political conditions.

152. One of the least problematic methods is to contact the leader through a locally-respected figure, such as a teacher or settlement priest. This may work best where the leader is community-representational or political, since a "neutral" contact may be the most appropriate starting point for building up honest dialogue and cooperation.

153. Another method is to use an "honest broker" to make the introductions. This might be more suitable where the community has a traditional leadership structure, and someone who has the ability to relate each side's point of view might be desirable, especially where there are marked cultural and ethnic differences between the agency and the residents. The broker method was used in a project to introduce white change agents into Black communities in the southern United States.(67) However, the range of criteria specified for a successful broker, such as acceptability to a wide range of political and social interests, personal integrity, respect, being above individual gain, and compatibility with the aims of the change agents are probably very difficult to fulfil.

154. Finally, another method more applicable for authoritarian leaders where direct access may be very difficult, is to use the leader's patron. Naturally, there will be considerable difficulties with this approach, not least because the agency may also have to justify its objectives to the patron. If the patron happens to follow a particular political ideology which is not compatible with the aims of the agency, then he, too, is likely to present obstacles to the programme.

155. Whichever method of initial contact is used, it is important that it be made personally. Letters, messages or telephone calls are likely to alienate and to threaten. It is important that the initial contact, therefore, be made by someone with authority from the agency, who can satisfy the leader's initial curiosity and who is able to alleviate any disquiet that might become apparent.

156. In cases where there are multiple leaders, it is probably advisable to seek out the strongest first, since he or she will have the greatest capacity to influence community opinion and support for the aims of the programme.(103) However. the aim is to avoid alienating the leaders of minority factions within the settlement since they might represent the feelings of those who are normally excluded from community-level decision-making. If there is discernible tension between rival groups, it may be better to hold individual, and as far as possible, concurrent meetings with the different leaders initially, but to inform them that the others have also been contacted. If factions are not particularly competitive, it may be better to invite all the leaders to meet at one time which could even provide a platform for future cooperation between them.

157. The most appropriate locality for an initial meeting is probably on "neutral" territory within or just outside the settlement. Outside locations, such as a cafe or a local church hall, may be preferable in cases where individual meetings are being held with different leaders; in/M-settlement sites, such as a church, school or social hall may be more appropriate where there is a collective meeting of all leaders or in situations where there is only one leader or committee. Meetings at the external agency's headquarters are better avoided since they might give the impression that the agency is unwilling to recognize that the community already has an adequate organizational structure. Moreover, if the agency takes the trouble to come to the settlement it shows its personnel are willing to make an effort to cooperate and to compromise from the start.

158. The agency should avoid dictating the contents and aims of the meeting in advance since detailed issues should only be discussed after opinions about the project have been expressed and some level of trust has been built up. The institution should merely state that it wishes to hold informal discussions with the settlement leaders and committee and to invite their possible participation in a scheme. Some detail must be given, but it should not be presented in such a way as to suggest: (a) a fait accompli; or (b) that proposals are decided and immutable. In cases of multiple leadership a policy of honesty, discretion and confidentiality is probably best. Agency staff should avoid becoming involved in discussions about other leaders where they may accidentally be drawn into making a value judgement or stating an opinion, thereby undermining their image of professional integrity. Neither is there any point in lying.

159. Agencies should consider carefully the issue of selecting personnel to make the approach to the leader, specifically whether to use male or female staff. This will depend on both the local cultural context and the type of leader.

160. This issue will also depend on the sexual division of labour and responsibilities in the agency. A key point is that the person should be someone in authority, so that leaders do not feel slighted. Where the agency wishes to present an "informal" profile, it may be better to use women rather than men, since in many cultures women will be seen as less "threatening" by vested interest groups. This may apply particularly in cases of authoritarian or positional leaders who might have to be approached very cautiously. Traditional leaders, however, may interpret a visit from a woman agency representative as an insult or lack of respect, so it is probably better in this case to use men. As far as community representational leaders are concerned, either men or women could be used, although if the leader is a woman she may achieve a more sympathetic dialogue with a female rather than a male representative.

1. On-going liaison

161. Having established initial contact with leaders, how should on-going liaison be organized? Walton mentions the need to have both formal and informal in-depth meetings with leaders.(102) Indeed, extensive contact is required to build up trust and the possibility of meaningful dialogue over programme aims, especially where the agency decides to work with the existing structure. In this case, it is probably advisable for agency staff to meet with leaders as regularly and as often as possible. This is, first, because the leader will feel he is actively included in the programme and may develop a commitment to it, and secondly, because it will allow the implementing agency to maintain a watch on the leader's activities with regard to the handling of funds, accuracy of announcements to the community, and so on. To a certain extent it is also important to build up some kind of identity with leaders, since the scope for cooperation in certain cultural contexts may be affected by the size of the socio-economic divide between leaders and agency staff.(27) This problem was recognized in the Saemaul Undong Movement in the Republic of Korea, where leaders were given certain privileges such as free postage and the opportunity to travel to Movement conferences and projects in an attempt by the State to indicate their appreciation of leaders' efforts and responsibility in programme implementation.(11) However, there are likely to be problems over balancing the need for the agency to stimulate some kind of identification with leaders and going so far that the leader becomes resented by the community, especially where community members do not feel they have been sufficiently involved. For example, in Ibadan, Nigeria, students from the African Regional Health Education Centre nearly alienated residents from an incinerator project established with the cooperation of the local Parent-Teacher Association. They spent too much time discussing project initiatives with teachers, rather than with ordinary community members.(12)

162. Again, the location of these meetings should normally be on "neutral" territory near or within the settlement. Regular meetings within the agency headquarters or district office will give the appearance of a "summons" from on high and would be counter-productive.

2. Leadership compatibility with project aims

163. It is not always possible at the start of a project to predict the likely outcome of leader involvement, but over time it should become more apparent whether leaders are assisting the community-development process. It is up to the agency to look out for warning signs and to react to them sensitively. Relevant issues here would be to evaluate whether the programme is benefiting all sectors of the community, and whether the agency's objectives are being upheld over and above the personal interests of the leader. It might be useful to set up an on-site project "suggestions" department staffed by lower-level agency personnel where initiatives and grievances can be registered confidentially. If leaders are working to the detriment of programme objectives, the agency should devise means to reduce the impact of that individual's influence on the project. This might, for example, take the form of facilitating greater collective responsibility among ordinary residents by setting up sub-committees for specific branches of the programme.

C. Working with existing structures or creating new ones

164. In sites-and-services schemes, when residents have had no previous contact with one another, external agencies have to encourage the creation of a community organizational structure. However, in most upgrading schemes where leadership patterns already exist, the agency has to decide whether to work with, or independently of, the existing structure. This section identifies the principal problems and advantages of different approaches and recommends appropriate steps for implementing agencies to take in various contexts. The point is that it is not the role of the agency to deform existing leadership, but it is reasonable for them to facilitate a more community representation/collective style of leadership if existing leadership poses serious obstacles to programme success. Indeed, it has been argued that formal leadership patterns are perhaps more easily changed than most other aspects of the local social structure.(27)

165. On balance, working with existing leaders is recommended except where they are: (a) highly exploitative; (b) where the majority of residents desire a change in leadership; or (c) where leadership may abuse project resources to the extent that certain groups within the community are excluded from programme benefits. How the agency reaches a decision will depend on its evaluation of leadership which itself requires close observation and analysis.

1. Working with existing leaders

166. The previous chapter drew attention to the pitfalls and advantages of working with existing leadership. Drawbacks of working with self-interested authoritarian, traditional or political leaders include delays in programme implementation, subjugation of the community in favour of personal priorities, and/or limited representation of minority groups within the community.(76,102,108) When the agency works with these kinds of leaders it is in effect legitimizing and reinforcing traditional patterns of inequality. However, despite these difficulties, several authors advocate working with and through existing leaders.(76,89, 109) Not only are existing leaders often popular, but they may also be highly motivated for communal activity.(109) In many respects, strong, representative leaders make the job of the external agency easier since responsibility for stimulating interest and encouraging support and commitment amongst residents will lie with someone who can draw upon trust and loyalty built up over a number of years.(106) However, where leaders abuse the trust and responsibility vested in them by the agency, staff should act as quickly and decisively as possible in order to save the programme from further deterioration.

167. One response aimed at reducing leader monopoly is to encourage leaders to accept other forms of representation within the community.(41) The agency should try to promote a shift towards more collective responsibility for the project without initially arousing the suspicion of leaders. For example, it might ask the leader's approval for a training programme in management and administration for community residents. A move like this would indicate that the agency still respects the leader since it requests his or her permission to go ahead with it, but at the same time it will initiate a process whereby other residents could gain a foothold in programme management. This would have to be played down to an extent that existing leaders would not feel they were being betrayed or ousted. Another way is to introduce development staff from outside the community. Professional community development workers were introduced successfully into the Lusaka project.(45) In Botswana, a woman agency worker was placed in a rural community development project and had a fairly harmonious relationship with the existing council of elders.(106)

168. However, if introducing more collective forms of leadership or outside control proves very difficult in certain settlements, other methods of reaching those residents excluded from the project should be devised. For example, to cope with the fact that women were normally excluded from community decision-making in Oke Foko, Ibadan, students from the African Regional Health Education Centre visited women in their homes to find out their opinions about the project.(12) If traditional leadership patterns are markedly hierarchical and/or exploitative, the agency should also carefully consider whether the project itself in its existing form is likely to modify those structures or reinforce them. If the latter is true, it might he hefter to change the project radically, if not abandon it altogether and start again with an alternative structure.

169. Another, if rather extreme, action would he to shift agency support firmly behind a rival leader and to encourage residents to support him or her. Once again, full reasons for the move should be given publicly and openly by the agency in order to minimize unfounded rumour and speculation. This approach represents a considerable risk for the agency and should usually only be employed as a last resort.

2. Creating, new structures where other structures already exist

170. The drawbacks of creating new structures alongside existing ones are mainly that they usually arouse opposition from existing leaders.(103,108) Furthermore, where leadership patterns are determined by cultural and historical influences, residents may not recognize or respect new leaders.(76) In Awgu, East-Central Nigeria, for example, local people only respected project workers who were "sons of the soil".(64) When leaders come from outside the settlement and have attachments elsewhere, the programme may collapse when they leave the project (76) - although this can also happen when charismatic leaders come from within the community.(80) However, the main advantage of setting up a new structure for the specific purposes of the project is that it may facilitate a more equitable distribution of benefits to the rest of the community.

171. In order to ensure that resistance from existing leadership structures is not too intense, a low-key role should be adopted when setting up an alternative structure. It is also perhaps wise to establish a collective or committee-based leadership structure rather than using individuals, so existing leaders do not feel they are being personally slighted or replaced by other people.

172. One "discreet" tactic for introducing new project-specific community structures is to work through those groups in the settlement which normally have only a limited role in local decision-making, and therefore do not greatly undermine existing leaders. For example, in some cases, women could be used to implement agency objectives without stimulating an undue amount of hostility from existing (male) leaders. In many traditional communities, there is a custom of separate but complementary men's and women's organizations.(55,65) Not only are women "closer to the community" than men and thus probably better- equipped for involvement in local development projects, but involvement also provides them with skills that can be used elsewhere, for example, in income generation. Over time, the experience derived from women's active inclusion in specific project-related activities can develop their willingness and capacity to occupy leadership positions, as has been the case with the katiwalas (primary health care workers) of the Philippines,(7) and in an integrated community development programme in Guayaquil, Ecuador.(90) The key point here is that, at the outset, existing leaders do not feel threatened since they will not perceive any competition from what are regarded as weaker groups.

3. Creating community organizational structures where none exist

173. Creating community organizational structures in sites-and-services projects is often problematic since there is little "community spirit" among beneficiaries and "natural" leaders will not have had a chance to come to the fore. The three main issues that agencies have to confront in this instance are:

(a) What are the optimum leadership structures in the project context?
(b) How should one identify potential leaders?
(c) How should community organization and leadership for the project become operational?

174. With regard to deciding which kind of leadership structure would be most appropriate for the project, agency staff must draw on both previous experience and the opinions of community members. People will have different needs and expectations of leadership in different cultural contexts and it is the job of the agency to find out what kind of structure would be most effective in that situation. On the whole, agencies should opt for creating a collective and/or rotational leadership structure whereby responsibility is shared between many members of the community and particularly by key representatives of minority factions. While a collective, committee-based style of leadership may have drawbacks in terms of either slowing down the decision-making process (in order to account for extensive discussion and reconciliation of a wider number of opinions), or reducing the positive impact of individual psychological factors, such as charisma and dynamism, at least the excesses of a monopoly by an individual leader, such as personal financial gain and unrepresentiveness, may be minimized. Sociometric techniques, such as those described in the section on identifying leaders, could be used to single out local people who might make a success of leadership or committee work.

175. The actual establishment of a community organizational structure should probably be largely left to the residents themselves; otherwise it may seem that the structure is being imposed from outside and therefore not be a good basis for future community participation. However, the agency should provide some kind of building which could be used as headquarters and also oversee elections to ensure fair play and prevent the emergence of self-imposed leaders.

D. How to respond to changes in leadership resulting from intervention

176. How should agencies respond to the side-effects of project intervention on leadership structures? Several difficulties can arise from leader involvement in community development projects, but probably the two most important are: (a) the reinforcement of socio-economic differences between leaders and the majority of residents; and (b) progressive alienation of the leader from the rest of the community, which is in part related to the above. The example of Mathare Village 2 in Nairobi is an illustration of the first point. Here leaders ended up with considerable financial benefits by using the support of the agency effectively to exploit the rest of community.(88) With regard to the second, leaders may become distanced from other residents because of a psychological divide created by their position itself. Individual leaders are a priori in some sense different from other residents. The leadership position demands a code of conduct that is possibly paternalistic, disciplinary or authoritarian.

177. Over time, residents' attitudes towards the leader can become distorted by their perception of what the leadership role means rather than who the person occupying that role is. This latter problem is very difficult for agencies to deal with directly, although if they deliberately seek to gain leader identification with themselves rather than the community, they must recognize that they are likely to intensify what is perhaps anyway an inevitable process of alienation. One way of combating this tendency may he to institute a system of rotational leadership, whereby responsibilities are shared by different individuals over time. However, this may have equally unfavourahle repercussions in teens of continuity. With regard to direct abuses of leadership positions. such as financial gain, internal "checks" may have to be established to prevent this leading to project failure and the alienation of the community from the institution. The agency must not play a formal disciplinary or watchdog role, but it could try to help the local community members to supervise leadership themselves. To an extent the dual system of traditional and New Community leaders achieved a degree of balance and control "from below" in the Saemaul Undong Movement in the Republic of Korea.(11) In Mexico, areas of communally-farmed agrarian land, known as ejidos, consist of two committees: one, an executive committee (Comisariato Ejidal) which is charged with overall management and decision-making on the ejido, and a second watchdog committee (Comite Vigilancia), whose job it is to oversee the former. This is a quite commonly accepted dual structure of organization in Latin America.

E. How to respond to other external institutions which try to contact leaders

178. Finally, there are severe risks involved in the process of opening up communities to intervention in the sense that other external institutions may try to take advantage of the opening created by the agency. For example, a political party or rival agency might try to foster support within the settlement through the leader, resulting in confusion for residents. Sometimes this may be solved by direetor-director negotiations, but it is difficult where institution heads represent different power groups. Other, more practicable alternatives include the agency following a policy of honesty and widespread dissemination of information from the outset. If people are kept informed of what the external institution is, or is not, doing, at least there is a chance of minimizing the risks associated with misinformation. Furthermore, if the agency adopts a policy of contracting staff from within the community then it is more likely that ordinary members will believe reports from them rather than feeling they have to rely on less than fully trusted agency personnel or leaders. Of course. in some instances, a leader may be coopted by another institution to the extent that the project is unworkable. In this case, agency staff should make sure that they have kept channels open for approaching leaders of other factions who may be more willing to cooperate on behalf of the community.

F. Some concluding remarks

179. It is important to note that the issues of agency-leadership interaction which have been raised in this report are points for consideration rather than policy prescriptions. Institutions involved in community development should take into account the issues addressed in chapters II to VI, attempt to understand the structure and function of leadership in different cultural contexts, and make informed decisions about how to proceed on the basis of chapters V and VI. and decision-making in low- income communities. One of the main problems in preparing this document was that there was little specific information about leadership in the literature, and even less analysis and interpretation of leadership functions, or accounts of past experiences. Therefore, UNCHS (Habitat) would benefit from, and appreciate, feed-back to this report.

181. The hope is that a wide range of agencies will give greater consideration to the question of the nature of leadership structures in different cities in terms of its emergence, its links to the wider governmental and political apparatus, its basis for legitimacy - in short, the various issues that have been discussed in this report. UNCHS (Habitat) would also welcome written evaluations of how agency staff have fared in their interaction with leaders; of techniques that worked well; and, equally important, of the failures and errors that were made; the outcomes of those failures; attempts to rectify them; and again, success or failure. UNCHS (Habitat) will not divulge information that will allow identification of these agencies, unless specifically authorized to do so.

182. The DANIDA/UNCHS Training Programme for Community Participation in the Improvement of Human Settlements is engaged in establishing courses and training activities for project staff, managers, community leaders and residents of upgrading settlements and sites-and-services schemes. In order to carry this out effectively a clear understanding of the communities and the role of their leadership is essential. This report is part of a more general attempt to reach such an understanding. The feedback which is sought can only make this attempt more successful.

Contact address:

DANIDA/UNCHS Training Programme for Community Participation,
United Nations Centre for Human Settlements (Habitat),
P. O. Box 30030,NAIROBI,Kenya

References

1. Abion, Z., "The potential influence of people's organizations on government policies affecting the distribution of land", in S. Angel and others (eds.), Land for Housing the Poor (Singapore, Select Books, 1983), pp.501-510.

2. Algara Cosio, I., "Community development in Mexico", in R. Dore and Z. Mars (eds.), Community Development (Paris, UNESCO, and London, Croom Helm, 1981), pp.336-432.

3. Angel, S., "Upgrading slum infrastructure: divergent objectives in search of consensus", Third World Planning Review, vol.5 (1983), pp.5-22.

4. Asian and Pacific Centre for Women and Development, Participation of Women in Decision-making: Some Guidelines (Bangkok, APCWD, 1980).

5. Banck, G., "Poverty, politics and the shaping of urban space: a Brazilian example", paper presented at the 45th Congress of Americanists, Bogota, 1-7 July 1985.

6. Barrows, W., "Comparative grassroots politics in Africa", World Politics, vol. 26 (1974), pp.283-297.

7. Bartra, R., and others, Caciquismo y Poder Politico en el Mexico (Mexico D.F., Siglo XXI, 1975).

8. Bienen, H., Kenya. the Politics of Participation and Control (Princeton, Princeton University Press, 1974).

9. Boaden, N., Goldsmith, M., Hampton, W., and Stringer, P., "Planning and participation in practice: a study of public participation in structure planning", Progress in Planning, vol. 13 (1980), No.1/2, pp.1-102.

10. Boschi, R., and Valladares, L., "Movimentos asociativos de camadas populares urbanas: analise de comparative de seis casos", in R. Boschi (ed.), Movimentos Colectivos no Brasil Urbano (Rio de Janeiro, Zahar Editores, 1982), pp.103- 143.

11. Brandt, V., and Lee, M., "Community development in the Republic of Korea", in R. Dore and Z. Mars (eds.), Community Development (Paris, UNESCO, and London, Croom Helm, 1981), pp.49- 136.

12. Brieger, W., and Adeniyi, J., "Urban community health education in Africa", International Quarterly of Community Health Education, vol. 2 (1983), No. 2, pp.109- 121.

13. Burgess, R., "The politics of urban residence in Latin America", International Journal of Urban and Regional Research, vol. 6 (1982), pp.465-480.

14. Burgess, R., "The limits to state self-help housing programmes", Development and Change, vol. 16 (1985), pp.271-312.

15. Castells, M., The City and the Grassroots (London, Edward Arnold,, 1984).

16. Chambers, R., Managing Rural Development: Ideas and Experience from East Africa (Uppsala, Scandinavian Institute of African Studies, 1974).

17. Chant, S., "Las olividadas: a study of women, housing and family structure in Queretaro, Mexico", Ph.D. dissertation (Department of Geography, University College, London, 1984).

18. Cohen, E., "Persistence and change in the Israeli kibbutz", in E. Kamenka (ed.), Community as a Social ideal (London, Edward Arnold, 1982) pp.123-146.

19. Collier, D., Squatters and Oligarchs (Baltimore, Johns Hopkins University Press, 1976).

20. Cornelius, W., "Contemporary Mexico: a structural analysis of urban caciquismo", in R. Kern (ed.), The Caciques. Oligarchical Politics and the System of Caciquismo (Alburquerque, University of New Mexico Press, 1973), pp.135-191.

21. Cornelius, W., Politics and the Migrant Poor in Mexico City (Stanford, Stanford University Press, 1975).

22. Dasgupta, S., "Associational structure and community development: a comparative study of two communities". International Review of Community Development, Nos. 31 -32 (1974), pp.213-226.

23. Davidson, F., "Ismailia: combined upgrading and sites and services projects in Egypt", in G. Payne (ed.), Low-income Housing in the Developing World (Chichester, John Wiley, 1984), pp.125-148.

24. Dehavalya, N., Bunnag, A., Prachuabmoh, P., and Nokyoongthong, M., A Study of Four Improved Congested Areas in Bangkok (Bangkok, Institute of Population Studies, Chulalongkorn University, and Department of Health, Bangkok Metropolitan Administration, 1983).

25. Diniz, E., Vote e Maquina Politica: Patronagem e Clientalismo no Rio de Janeiro (Rio de Janeiro, Paz e Terra, 1982).

26. Diniz, E., "Favela: associativismo e participa social", in R. Dore and Z. Mars (eds.), Movimentos Colectivos no Brasil Urbano (Rio de Janeiro, Zahar Editores, 1982), pp.27-74.

27. Dore, R., "Introduction. Community development in the 1970s", in R. Dore and Z. Mars (eds.), Community Development (Paris, UNESCO, and London, Croom Helm, 1981), pp.13-46.

28. Eckstein, S., The Poverty of Revolution: the State and the Urban Poor in Mexico (Princeton, Princeton University Press, 1977).

29. Eisenstadt, S., and Roniger, L., "Patron-client relations as a model of structuring social change", Comparative Studies in Society and History, vol.22 (1980), pp.42-77.

30. Epstein, A., Politics in an Urban African Community (Manchester, Manchester University Press, 1985).

31. Fernando, M., Women's Participation in the Housing Process the Case of Kirillapone, Sri Lanka, Gender and Planning Working Paper No. 3 (London, Development Planning Unit, University College, London, 1985).

32. Gaikwad, V., "Community development in India", R. Dore and Z. Mars (eds.), Community Development (Paris, UNESCO, and London, Croom Helm, 1981), pp.254-344.

33. Gilbert, A., and Ward, P., "Community action by the urban poor: democratic involvement, community self-help or a means of social control?", World Development, vol. 12 (1984), No. 8, pp.769-782.

34. Gilbert, A., and Ward, P., Housing, the State and the Poor. Low-income Housing in Three Latin American Cities (Cambridge, Cambridge University Press, 1985).

35. Grindle, M., Bureaucrats, Politicians and Peasants in Mexico. a Case Study in Public Policy (Berkeley, University of California Press, 1977).

36. Gugler, J., and Flanagan, W., Urbanization and Social Change in West Africa (Cambridge, Cambridge University Press, 1978).

37. Hale, S., "The politics of entrepreneurship in Indian villages", Development and Change, vol. 9 (1978), pp.254-275.

38. Handelman, H., "The political mobilization of urban squatter settlements: Santiago's recent experience and its implications for urban research", Latin American Research Review, vol. 10 (1975), pp.35-72.

39. Harpham, T., Vaughan, P., and Rifkin, S., Health and the Urban Poor in Developing Countries, Evaluation and Planning Centre for Healthcare, Publication No. 5 (London School of Hygiene and Tropical Medicine, 1985).

40. Hollnsteiner, M., "People power: community participation in the planning of human settlements", Carnets de l'Enfance, vol. 40 (1977), pp.11-47.

41. Hollnsteiner, M., "People-powered: thoughts for urban planners, administrators and policy makers", paper presented at the Regional Congress of Local Authorities for Development of Human Settlements in Asia and the Pacific, Yokohama, Japan (9-16 June 1982).

42. Huizer, G., "Community development, land reform and popular participation", in T. Shanin (ed.), Peasants and Peasant Societies, (Harmondsworth, Penguin, 1971), pp.389-411.

43. Huntington, S., and Nelson, J., No Easy Choice: Political Participation in Developing Countries (Cambridge, Massachusetts, Harvard University Press, 1979).

44. Jacobi, P., "Movimentos populares urbanos y resposta do estado: autonomia e control vs. coopta e clientalismo", in R. Boschi (ed.), Movimentos Colectivos no Brasil Urbano (Rio de Janeiro, Zahar Editores, 1982), pp.145- 179.

45. Jere, H., "Lusaka: local participation in planning and decision-making", in G. Payne (ed.), Low-income Housing in the Developing World (Chichester, John Wiley, 1984), pp.55-68.

46. de Kadt, E., "Community participation for health: the case of Latin America", World Development, vol. 10 (1982), No. 7, pp.573-584.

47. Kaufman, R., "The patron-client concept and macropolitics: prospects and problems", Comparative Studies in Society and History, vol. 16 (1974), pp.284-308.

48. Kamenka, E., and Ehr-Soon Tay, A., "Community and commune in the USSR and China", in E. Kamenka (ed.), Community as a Social Ideal (London, Edward Arnold, 1982), pp.99-122.

49. Kayongo-Male, D., "Helping self-help groups help themselves: training of leaders of women's groups", Journal of Eastern African Research and Development, vol. 13 (1983), pp.88-103

50. Kearney, M., The Winds of Ixtepeji. World View and Society in a Zapotec Town (New York, Holt, Rinehart, Winston, 1972).

51. Leeds, A., "The significant variables determining the character of squatter settlements", America Latina, vol. 12 (1969), pp.44-86.

52. Manghezi, A., Class, Elite and Community in African Development, (Uppsala, Scandinavian Institute of African Studies, 1976).

53. Mayo, M., "Community development: a radical alternative?" in R. Bailey and M. Brake (eds.), Radical Social Work (London, Edward Arnold, 1975), pp.129-143.

54. Mbithi, P., and Barnes, C., Spontaneous Settlement Problem in Kenya (Nairobi, East African Literature Bureau, 1975).

55. Mbithi, P., and Rasmusson, R., Self-reliance in Kenya: the Case of Harambee (Uppsala, The Scandinavian Institute of African Studies, 1977).

56. Mghweno, J., "Tanzania's surveyed plots programme", in G. Payne (ed.), Low-income Housing in the Developing World (Chichester, John Wiley, 1984), pp. 109-123.

57. Montano, J., Los Pobres de la Ciaudad de Mexico en los Asentamientos Humanos (Mexico, D.F., Siglo XXI, 1976).

58. Moser, C., Residential Level Struggle and Consciousness. the Experiences of Poor Women in Guayaquil, Ecuador, Development Planning Unit, Working Paper No. 15 (London, University College, London, 1984).

59. Moser, C., and Chant, S., The Role of Women in the Execution of Low-income Housing Projects (training module) (Nairobi, United Nations Centre for Human Settlements (Habitat), DANIDA/UNCHS Training Programme for Community Participation in Improving Human Settlements, 1986).

60. Mushi, S., "Community development in Tanzania", in R. Dore, and Z. Mars (eds.), Community Development (Paris, UNESCO, and London, Croom Helm, 1981), pp.139-242.

61. Nelson, J., Access to Power: Politics and the Urban Poor in Developing Nations (Princeton, Princeton University Press, 1979).

62. Nientied, P., Meijer, E., and Van der Linden, J., Karachi Squatter Settlement Upgrading: Improvement and Displacement? (Amsterdam, Geografisch en Planologisch Institut, Vrije Universiteit, 1982).

63. Oakley, P.. and Marsden, D., "Approaches to participation in rural development", preliminary paper prepared for Inter-Agency Panel on People's Participation, March 1983 (Geneva, International Labour Office, 1983).

64. Odokara. E., "Supervised training in community development field practice in Awugu, East Central State of Nigeria", International Review of Community Development, Nos. 2728 (1972), pp.149- 160.

65. Okonjo, K., "Sex roles in Nigerian politics", in C. Oppong (ed.), Female and Male in West Africa (London, George Allen and Unwin, 1983), pp.211-222.

66. Olesen, A., "A study of slums and slum improvement in Madras. Report from a study period in Madras" (Aalborg, Institute of Development and Planning, 1984) (mimeo).

67. Ottaway, R., "Developing the white, moderate leadership of a community in racial conflict", in P. Ward (ed.), International Review of Community Development (London, Mansell, 1982), pp.117- 139.

68. Pasteur, D., The Management of Squatter Upgrading (Farnborough, Saxon House, 1979).

69. Payne, G., "Self-help housing: a critique of the Gecekondus of Ankara" in P. Ward (ed.), Self-help Housing: a Critique (London, Mansell, 1982), pp.117-139.

70. de la Paz, T., "The Katiwala: an experience in primary health care", paper presented to a Joint UNICEF/WHO Consultation on Primary Health Care in Urban Areas, Guayaquil, Ecuador, 15-19 October 1984 (Geneva, World Health Organization, 1984).

71. Portes, A., "Housing policy, urban poverty and the State: the favelas of Rio de Janeiro", Latin American Research Review, vol. 14 (991979), pp.3-24.

72. Rakodi, C., "Participation in the Lusaka project", in C. Rakodi and A. Schlyter (eds.) Upgrading in Lusaka: Participation and Physical Changes (Lund, The National Swedish Institute for Building Research, 1981), pp.65-76.

73. Ray, T., The Politics of the Barrios of Venezuela (Berkeley, University of California Press, 1969).

74. Redl, F., "Group emotion and leadership", Psychiatry, vol.5 (1945), pp.573-596.

75. Rifkin, S., "Community involvement in primary health care among the urban poor" (Geneva, World Health Organization, 1983) (unpublished paper).

76. Rifkin, S., Community Participation in MCH/FP Programmes: an Analysis Based on Case Study Materials (Geneva, World Health Organization, 1984).

77. Ross, M., The Political Integration of Urban Squatters (Evanston, Northwestern University Press, 1973).

78. Schlyter, A., and Schlyter T., George, The Development of a Squatter Settlement in Lusaka, Zambia (Lund, the National Swedish Institute of Building Research, 1979).

79. Scott, I., "Party functions and capabilities: the local UNIP organization during the first Zambian Republic (1964-73)", African Social Research, vol. 22 (1976), pp.107-129.

80. Seabrook, J., The Idea of Neighbourhood: What Local Politics Should Be About (London, Pluto, 1984).

81. Shah, K., "People's participation in housing action: meaning, scope and strategy", in G. Payne (ed.), Low-income Housing in the Developing World (Chichester, John Wiley, 1984), pp.199-208.

82. Skinner, R., "Community organization, collective development and politics in self-help housing: Villa El Salvador, Lima (1971- 1976)", unpublished doctoral dissertation (University of Cambridge, 1982).

83. Skinner, R., "Community participation: its scope and organization", in R. Skinner and M. Rodell (eds.), People, Poverty and Shelter: Problems of Self-help Housing in the Third World (London, Methuen, 1983), pp.125-150.

84. Skinner, R., "Community participation in third world housing. Potential and constraints", Cities, vol. 1 (1984), No. 6, pp.564-574.

85. Silas, J., "The Kampung Improvement Programme of Indonesia: a comparative case study of Jakarta and Surabaya", in G. Payne (ed.), Low-income Housing in the Developing World (Chichester, John Wiley, 1984), pp.69-87.

86. Thornley, A., "Theoretical perspectives on planning participation", Progress in Planning, vol. 7 (1977), No. 1, pp.1-57.

87. Ugalde, A., and others, The Urbanization Process of a Poor Mexican Neighbourhood (Austin, University of Texas, Institute of Latin American Studies Special Publication, 1974).

88. United Nations Children's Fund, Urban Examples for Basic Services Development in Cities (UE-20) (New York, 1979).

89. United Nations Children's Fund, The Management of the Delivery of Basic Services (New York, 1981).

90. United Nations Children's Fund, Urban Examples for Basic Services Development in Cities (UE-4) (New York, 1983).

91. United Nations Children's Fund, Reaching Children and Women of the Urban Poor, UNICEF Occasional Papers No. 3 (New York, 1984).

92. United Nations, Social Change and Social Development Policy (New York, ECLA, 1970).

93. United Nations, The Social Impact of Housing (New York, 1977).

94. United Nations Centre for Human Settlements (Habitat), Survey of Slums and Squatter Settlements (Dublin,Tycoolylnternational,1982).

95. United Nations Centre for Human Settlements (Habitat), Community Participation in the Execution of Low-income Housing Projects (Nairobi, 1983).

96. United Nations Centre for Human Settlements (Habitat), Women and Human Settlements (Nairobi, 1985).

97. Valladares, L., "Associativismo e representa comunitaria no Rio de Janeiro", in Second Congress Humaniza das Ciudades (Rio de Janeiro, Escola Nacional de Habitaca e Poupuanca, 1983).

98. Valsan, E.H., Community Development Programs and Rural Local Government. Comparative Case Studies from India and the Philippines (New York, Praeger, 1970).

99. Vance, I., Women's Participation in Self-help Housing: the San Judas Barrio Project, Managua, Nicaragua, Gender and Planning Working Paper No. 4 (London, Development Planning Unit, University College, London, 1985).

100. Van der Linden, J., "Actors in squatter settlement upgrading: their interests and roles", in J. Schoorl, J. Van der Linden and K.S. Yap (eds.), Between Basti Dwellers and Bureaucrats. Lessons in Squatter Settlement Upgrading (Oxford, Pergamon, 1983), pp.249262.

101. Varley, A., "Ya somos duenos: Ejido land development and regularisation in Mexico City", Ph.D. dissertation (London, Department of Geography, University College, London, 1985).

102. Walton, D., "The role of international consultants", in G. Payne (ed.), Low-income Housing in the Developing World (Chichester, John Wiley, 1984), pp.173-196.

103. Ward, P., The Impact of Local Leaders upon Squatter Settlements in Mexico City, UK Social Science Research Council, Final Report, (London, British Library-Lending Division, 1978).

104. Ward, P., "Political pressure for urban services: the response of two Mexico City administrations", Development and Change, vol. 12 (1981), pp.379-407.

105. Ward, P., Welfare Politics in Mexico (London, George Allen and Unwin, 1986).

106. Wass, P., "A case history: community development gets established in Botswana", International Review of Community Development, Nos. 21-22 (1969), pp.181-198.

107. Wegelin, E., "From building to enabling strategies in Asia: institutional problems", in R. Skinner and M. Rodell (eds.), People, Poverty and Shelter. Problems of Self-help Housing in the Third World (London, Methuen, 1983), pp.106-124.

108. Yap, K.S., Leases, Land and Local Leaders: an Analysis of a Squatter Settlement Upgrading Programme in Karachi (Amsterdam, Geografisch en Planologisch Institute, Virje Universiteit, 1982)

109. York, A., "Leaders of voluntary associations in a difficult housing estate", International Review of Community Development, Nos. 37-38 (1977), pp.273-281.