—l (D (A) _| O _s O) 01 _| m \l a ? This is to certify that the thesis entitled EVALUATING SQUATTEE RESIDENCE IN URBAN ZAMBIA: THE EFFECTS OF SOCIAL CLASS AND URBANIZATION presented by Anne Louise Mason has been accepted towards fulfillment of the requirements for Master of Arts degreein Community Service M Kéém Major professor Date October 3, 1980 0-7 639 NOV 33 7??? OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MTERSALE: Place in book return to mm charge from circularmn recor I I EVALUATING SQUATTER RESIDENCE IN URBAN ZAMBIA: THE EFFECTS OF SOCIAL CLASS AND URBANIZATION By Anne Louise Mason A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OR ARTS. Department of Family and Child Ecology 1980 6m: V3; '7‘ ABSTRACT EVALUATING SQUATTER RESIDENCE IN URBAN ZAMBIA: THE EFFECTS OF SOCIAL CLASS AND URBANIZATION By Anne Louise Mason This secondary analysis of housing data from Lusaka, Zambia focuses on squatters' evaluations of their residential situation. A housing adjustment theory proposed by Morris and Winter (1975) served as a conceptual framework for the selection of variables. Indices of six dimensions of resi- dential situation were constructed by combining male and female household heads' responses to open-ended questions on their residential likes and dislikes. Correlation coeffi- cients were computed to test relationships between these indices and the education and urban experience of household heads and household income. Findings indicated little support for hypothesized rela- tionships. Examination of the responses which comprised the indices revealed that squatters generally liked aspects of economic conditions of residence, location, and interpersonal relations; but disliked aspects of house and compound quality and services. Findings suggested possibilities for further research including the development of better measures for evaluating residence and the selection of other variables and relationships for analysis. ACKNOWLEDGEMENTS Many persons deserve recognition for their contributions to this research experience. As my academic advisor and chairperson of my thesis guidance committee, Linda Nelson provided consistent support and offered constructive sugges— tions and advice. She was always available to share in both my achievements and my frustrations and offered the encourage- ment I needed to bring this research to completion. The other members of my guidance committee, William Kimball, Beatrice Paolucci, and David Wiley, each offered unique in- sights and experience as work on this thesis progressed. They helped me to clarify and organize my ideas, and their confidence in my abilities pushed me to gain more from this study than I had thought possible. I am particularly grateful to David Wiley for making the data and a vast amount of relevant information available for my use. His personal enthusiasm for and knowledge of the subject matter helped to stimulate my own thinking and interest. Appreciation is also extended to the institutions which supported the research on which this study is based: the Sociology of Economic Change Training Program at the university of Wisconsin, Madison (funded through a training grant from the National Institute of Mental Health), and ii the graduate student members of that program; the Midwest University Consortium for International Activities, Inc.; the Ford Foundation; and the Rockefeller Foundation, Social Science Division- I am also grateful to the College of Human Ecology at Michigan State university for making it possible for me to have access to the MSU computing facili- ties. As I pursued this graduate program I also benefited from the steady emotional support and encouragement that Kageche Kamiri provided while at the same time fulfilling his own academic responsibilities. I also wish to thank Nancy Lee Radtke not only for her assistance in typing the final draft of this thesis, but especially for the training and guidance in writing that she gave me in other work that we have done together. iii TABLE OF CONTENTS Page LIST OF TABLES ......................................... v LIST OF FIGURES ........................................ viii Chapter I. INTRODUCTION .................................... 1 Conceptual Framework ............................ S Purposes ........................................ 9 Assumptions ..................................... 10 Definitions ..................................... 10 Hypotheses ...................................... 11 II. REVIEW OF RELATED LITERATURE .................... 15 Housing Adjustment Behavior ..................... 15 Squatter Housing ................................ 21 III. PROCEDURES ...................................... 29 Collection of Data .............................. 29 Description of the Survey Instrument ............ 31 Measurement of Variables ........................ 33 Method of Analysis .............................. 44 IV. FINDINGS ........................................ 47 Residential Evaluations ......................... 47 Hypothesis Testing .............................. 60 Other Analyses .................................. 64 V. SUMMARY, DISCUSSION, AND IMPLICATIONS ........... 65 Summary of the Study ............................ 65 Limitations of the Study ........................ 68 Discussion and Conclusions ...................... 7l Implications .................................... 76 APPENDIX: ADDITIONAL TABLES ........................... 83 BIBLIOGRAPHY ........................................... lOl iv LIST OF TABLES Table 1. Responses to Evaluation Questions Not Included in Evaluation Indices ...................... 7 ........ 2. Evaluations of Residential Situation (Distributions of Scores on Indices) ............................ 3. Evaluation of Economic Conditions of Residence ..... 4. Evaluation of Location ............................. 5. Evaluation of House Quality ........................ 6. Evaluation of Compound Quality ..................... 7. Evaluation of Services ............................. 8. Evaluation of Interpersonal Relations .............. 9. Evaluation of Residential Situation——Summary ....... 10. Relationships of Education and Urban Experience of Household Heads and Total Household Income to Evaluations of Residential Situation (Pearson Product Moment Correlations) ..................... A. Likes and Dislikes of Residential Situation (Distri- bution of Scores on Indices) ..................... B. Male Heads' Likes and Dislikes of Residential Situ- ation (Distribution of Scores on Indices) ........ C. Female Heads' Likes and Dislikes of Residential Situation (Distribution of Scores on Indices).... D. Years of Education of Male and Female Heads of Household ........................................ E. Years of Urban Experience of Male and Female Heads of Household ..................................... F. Total Household Income ............................. Page 36 4O 49 50 52 53 56 57 58 61 83 84 85 86 86 87 Relationships of Education and Urban Experience of Household Heads and Total Household Income to Likes and Dislikes of Residential Situation (Pearson Product Moment Correlations) ............ Relationships of Education and Urban Experience of Male Head and Total Household Income to Male Heads' Likes and Dislikes of Residential Situa- tion (Pearson Product Moment Correlations) ....... Relationships of Education and Urban Experience of Female Head and Total Household Income to Female Heads' Likes and Dislikes of Residential Situa- tion (Pearson Product Moment Correlations) ....... Relationships of Education (Square) of Household Heads and Total Household Income (Log) to Evalua- tions of Residential Situation (Pearson Product Moment Correlations) ............................. Relationships of Education (Square) of Household Heads and Total Household Income (Log) to Likes and Dislikes of Residential Situation (Pearson Product Moment Correlations) ..................... Relationships of Education (Square) of Male Head and Total Household Income (Log) to Male Heads' Likes and Dislikes of Residential Situation (Pearson Product Moment Correlations) ............ Relationships of Education (Square) of Female Head and Total Household Income (Log) to Female Heads' Likes and Dislikes of Residential Situation (Pearson Product Moment Correlations) ............ Effects of Education and Urban Experience of Male and Female Household Heads and Total Household Income on Evaluations of Residential Situation (Regression Analysis) ............................ Effects of Education (Square) and Urban Experience of Male and Female Household Heads and Total Household Income (Log) on Evaluations of Residen- tial Situation (Regression Analysis) ............. Effects of Education and Urban Experience of Male 'Household Heads and Total Household Income on Male Heads' Likes and Dislikes of Residential Situation (Regression Analysis) .................. vi 88 89 91 92 93 94 95 96 97 Q. Effects of Education (Square) and Urban Experience of Male Household Heads and Total Household In- come (Log) on Male Heads' Likes and Dislikes of Residential Situation (Regression Analysis) ...... 98 R. Effects of Education and Urban Experience of Female Household Heads and Total Household In- come on Female Heads' Likes and Dislikes of Resi- dential Situation (Regression Analysis) ......... 99 S. Effects of Education (Square) and Urban Experience of Female Household Heads and Total Household Income (Log) on Female Heads' Likes and Dislikes of Residential Situation (Regression Analysis)... 100 LIST OF FIGURES Figure Page 1. Relationships among Variables in a Housing Adjust- ment Model ........................................... 7 viii CHAPTER I INTRODUCTION Many countries in Africa, Asia, and Latin America are experiencing high population growth rates in urban areas. Although some of this expansion may be attributed to natural increase in population, most of the growth is a result of migration from rural areas to urban centers. Where a decline in agricultural productivity is coupled with urban economic expansion and industrialization, urban areas are particularly attractive to potential migrants. People migrate to urban areas for economic reasons primarily, e.g., to seek employ- ment, although other factors nay contribute to this trend. Zambia in southern Africa illustrates this phenomenon. Lusaka Urban District, Zambia's capital city and surrounding area, grew 93% between 1963 and 1969 (Sinmmnce,l972, p. 20). tremendous drain has been placed on the housing market in urban centers such as Lusaka by this rapid population growth. Ownership of housing in Zambia is dominated by the national government and private companies, who provide for civil servants and private employees respectively, and local housing authorities, although some private landlords partici- pate in the housing market. Occupancy is, to a large extent, restricted by place of employment, and housing supply has not 1 2 kept pace with demand. This situation reflects a lack of resources needed to increase available housing stock as well as poor planning, lack of understanding of the problem, and conflicts between local and national housing authorities over the type and design of houses to be built (Simmance, 1972, pp. 30-37) . Contrasted with this situation is the relative success with which people have been able to meet some of their most pressing needs through the establishment and expansion of high density residential settlements around urban centers. Such areas are variously referred to as squatter areas, shanty towns, unauthorized areas, or spontaneous settlements. The term "squatter" will be used in this study with the under- standing that many of the residential areas to which it applies are not necessarily considered to be illegally occu- pied and may, in fact, be recognized as permanent and viable residential areas. It has been estimated that between 1969 and 1972 offi- cial housing in Lusaka increased by 4,000 but about 22,000 units were provided in the squatter areas. Further it has been estimated that of the total Lusaka population of 381,000 in 1972, over 150,000 were residents of these areas (National Housing Authority, 1972, p. 4). Squatter areas are more than a collection of dwelling units; they embody a number of different dinensions associ- ated with housing such as security, access to jobs, and access to some urban services. Although squatter areas may serve 3 as receiving points for newly arrived mdgrants to the city, they are also the residential locations of many peOple who have lived in urban areas for much, if not all, of their lives. Most of the housing in squatter areas is free or in- expensive, a positive feature both to those who have little money available for housing and to those who prefer to invest their money in other things. Although squatter areas may lack clean water, adequate sanitation and refuse disposal systems, and improved roads, the houses within them are usually well-maintained and fairly permanent structures. They are often constructed of burnt brick, Kimberly brick (bricks made by pressing mud in molds), or concrete blocks, with thatched or corrugated metal roofs. The political party UNIP (United National Independence Party) usually controls the construction and arrangement of houses with the result that squatter areas are orderly although sometimes dense in appearance. Natural vegetation and small gardens are found in many squatter areas. Squatter areas in Zambia are frequently located close to places of employment which minimizes transportation costs and time for those who are employed. Informal economic activity flourishes in these areas, thus the rate of unemploy- ment among squatters is much lower than might be expected as many are self-employed. Flexibility is permitted in the construction and improvement of residences in squatter areas. Occupants make improvements when they are both needed and economically feasible rather than when mandated by a housing 4 contract. To a large extent, it appears that residences in squatter areas are at least as good as those which people had in rural areas and possibly better because of the potential access to urban services. In different places and at different points in time, official policies regarding these housing areas have varied, ranging from.elimmnation of high density areas to providing such areas with basic services such as clean water. Although the tendency of national and local authorities in Zambia was to ignore squatter areas and to fail to provide sufficient alternative housing, there has been a change in emphasis. The Second National Development Plan for the period 1972- 1976 recognized that these areas "represent assets" and "require planning improvement" rather than demolition (Sim- mance, p. 33). The New Housing (Statutory and Improvement Areas) Act (Republic of Zambia, 1974) gave power to local authorities to provide residents of these areas security of tenure and improved facilities (Tipple,l976, p. 186). In the literature dealing with housing policy and pro- grams for urban areas of Africa, Asia, and Latin America, there appears to be growing advocacy for recognition and offi- cial support for the various dimensions of squatter residential areas, i.e., the needs that they can and do fulfill. Housing policy-makers and program planners are challenged to formulate and implement relevant responses to these needs. Conceptual Framework One basis for the formulation of housing policies and programs may be an understanding of residential preferences, decisions, and behaviors at the level of the household unit. The theory of housing adjustment proposed by Morris and Winter (1975), which generally suggests that families adjust their housing to meet changing needs, provides a conceptual framework for considering housing at this level. According to this framework, household members evaluate their current housing conditions in terms of both family and cultural norms which represent different dimensions of housing. Housing needs derive from these evaluations rather than from basic needs for shelter and safety. When evaluation reveals that a housing deficit or need exists, i.e., current housing deviates from household memr bers' norms in a dimension of importance, feelings of dissat- isfaction may arise. This produces a desire to eliminate the discrepancy between current conditions and norms in order to meet the need. Household members may undertake one of several possible adjustment behaviors including residential mobility, residential adaptation, and family adaptation. However, for an adjustment to take place, constraints on the particular behavior must be overcome. These constraints include intra— familial strengths and weaknesses; economic, social, and political factors; and attractive features of the current dwelling which counteract its negative features. 6 The goal of the adjustment process is to maintain a bal- ance between housing norms and housing conditions. Given changing norms and changing conditions, this is a continuous process. It is also complicated by the fact that household members may consider only one aspect of their housing at a time while other aspects serve as constraints. The aspect under consideration at one time may later serve as a con- straining factor. In general, this theoretical framework suggests a num- ber of questions. Morris and Winter (1975) ask: 1. What are the determinants of the perception of a housing deficit? 2. What are the determinants of the kinds of deficits that are salient to families? 3. How does housing satisfaction respond to the presence and magnitude of housing deficits? 4. What are the determinants of preferences for the various housing adjustment behaviors? 5. What are the correlates of the intrafamilial constraints that deter family housing adjust- ment? 6. What are the correlates of the extrafamilial constraints that impinge on housing adjust— ment? (p. 86) Such questions may lead to the development of testable hypo- theses to seek support for the theory. When applied specifically to the context of squatters in Zambia, the housing adjustment theory leads to questions such as (1) How do squatters feel about their residential situation? (2) Can squatters be differentiated from each other or from other urban residents in terms of their feelings and behaviors related to their residential situation? (3) What 7 aspects of their residential situation do squatters consider to be the most important to them? (4) Why do squatters prefer particular residential locations? (5) How do squatters allo- cate their resources with respect to their residential situ- ation? and, (6) What actions are or can be undertaken by squatters in terms of changing or improving upon their resi- dential situation? Such questions imply a number of relation- ships which mdght be diagrammed as follows: Acde ," Residential : Cbndfljbns I n llwelcm : at is faction with 13:38 : Residential Goals : Cbndflfions : Characteristics ' 1 I ”a Of Household Social , Political : Mamers muiEommmmc ,, : Constraints DeCIsions a and : Bdmndons ' u l ___________________________________________________ _} Lukpauhnt Inmamedume therwafing Ikpamhnt variable variable variable variable Figure 1. Relationships ammg Variables in a Housing Adjustment Model These questions and this model directed the selection of var- iables for analysis from available data. Literature related to housing adjustment, particularly that related to residential mobility, emphasizes the influence of characteristics of household members such as stage in the family life cycle,on housing preferences, decisions, and 8 behaviors. For example, family expansion may lead to dissat- isfaction with house size and, subsequently, to a decision to adjust housing by enlarging the structure or by moving to a larger one. This research has been conducted primarily in the continental United States. In the context of squatters in Zambia, it seems possible that the effects of life cycle changes on housing preferences, decisions, and behaviors may be confounded by influences of rapid urbanization and social changes. Members of families at different stages of the life cycle may have had vastly different experiences with respect to the educational and income-generating opportunities available to them and to their urban living experience. This notion has led to the consider- ation of basic characteristics of household members that relate to social class and urbanization. Such characteristics may be related to squatters' housing preferences, decisions, and behaviors. For example, as the number of years that people have lived in a city increases, the importance that they place on urban services such as piped water may also increase. Evaluation of current residential situation is a funda- mental component of the housing adjustment process as desribed by Morris and Winter (1975). Evaluations may reflect both basic dimensions of residential situation, i.e., areas of need which may be important to household members, and attitudes about these dimensions, i.e., potential sources of residential satisfaction or dissatisfaction. 9 The theory of housing adjustment links these evaluations to people's residential decisions and behaviors. Hence, examination of these evaluations and the factors related to them may contribute to an understanding of why and how people make specific decisions regarding their residential situation. With such information policy-makers or program planners may be in a position to be sensitive to potential changes in housing patterns in the context of larger social change and to plan and implement relevant policies and programs. Purposes Housing adjustment is a complex process involving many components which vary from one context to another. It is beyond the scope of this study to test a full model of housing adjustment; this study will be limited to consideration of evaluations of residential situation by male and female heads of household and selected characteristics of these individuals. The study attempts through secondary analysis of data col- lected on housing in Zambia to contribute to an understanding of relationships between evaluations of residential situation by male and female heads of squatter households and their education and urban experience and total household income. Findings of this study may be useful to future analyses of household members' residential preferences, decisions, and behaviors. The following research questions have guided this study: 10 1. What dimensions are identified by male and female heads of household in their evaluations of their residential situation? 2. What relationships exist between education and urban experience of male and female heads of household and total household income and house- hold heads' evaluations in these different dimensions of residential situation? Assumptions This study is based on secondary analysis of data col- lected in a study in which the author was in no way involved. It is assumed that the procedures followed in the original study relating to the design and administration of the survey instrument and the coding and tabulating of data have yielded information that is both reliable and valid. Definitions Co ound. Refers to a residential area or neighborhood in which residents may share facilities such as schools, mar- kets, and clinics; and participate in political and social institutions. Head of Household. Refers to a person having the respon- sibility of looking after other members of the household. In this study, the male head of household has economic responsi- bility for other household members; the female head of house- hold is the spouse or partner of the male head. In cases where no male head is present, the female head of household corresponds to the woman who bears responsibility for other household members. 11 Household. Refers to a group of people who reside together and form an economic unit. In this study population, a household may include male and female heads of household, their children, and possibly other children and adults who are dependent on the heads of household. Evaluation of Residential Situation. Refers to male and female household heads' assessments of their residential or housing situation. In this study, residence refers to not only an actual dwelling unit but also its surrounding neigh- borhood. It therefore encompasses several different dimen- sions. The measure of evaluation used in this study is based on male and female heads' mentions of likes and dislikes of the residential situation. More frequent mention implies greater importance of a dimension of residential situation. Squatter. Refers to a resident of a high density resi- dential area which may not be sanctioned officially or managed by city government, but which may be recognized as a permanent and viable residential area by housing and government offi- cials. Hypotheses Possible relationships betweencharacteristics of house- hold heads and housing needs, which may be reflected by eval- uations of residential conditions, are implied by the housing adjustment model. Several characteristics were selected for consideration in this study including education of male and female heads of household, urban experience of male and 12 female heads of household, and total household income. It was felt that such characteristics could have similar im- pacts on residential evaluations. Literature on squatters suggests that the different needs fulfilled by squatter compounds vary in importance to household members as their characteristics change. It appears, for example, that for those who are newly arrived in urban areas or those with a very low level of income, the low cost of squatter residences and their proximity to places of em- ployment are their most important features. However, resi- dents' concern with these features may be replaced as other dimensions of these compounds such as quality of houses or compound as well as services take on more importance. Such shifts may occur as people change social class and develop the means and aspirations to improve upon their residential situation. For this study, measures of evaluation were developed by combining the male and female household heads' responses to questions conerning their likes and dislikes of their residential situation. Evaluation in a particular dimension of residential situation represents, therefore, the frequency with which responses were provided in that dimension. Given the implications of the literature on squatters, it was generally hypothesized that as characteristics of those indi- viduals change, the frequency of theirresponsesin.some dimen- sions will decrease as it increases in others. 13 The following specific hypotheses have guided this analysis: .111. Frequency of mention of economic conditions of residence is negatively related to: (DC-LOUD) education of male household head . urban experience of male household head education of female household head . urban experience of female household head total household income Frequency of mention of location is negatively related to: mantra: . education of male household head . urban experience of male household head education of female household head . urban experience of female household head total household income Frequency of mention of house quality is positively related to: (DD-.0693 education of male household head . urban experience of male household head education of female household head . urban experience of female household head total household income Frequency of mention of compound quality is positively related to: (DO-00"“! education of male household head . urban experience of male household head education of female household head . urban experience of female household head total household income Frequency of mention of services is positively related to: (‘DCLOO‘DJ . education of male household head urban experience of male household head education of female household head . urban experience of female household head . total household income In the original survey, respondents frequently mentioned as- pects of interpersonal relations in their evaluations of their residential situation. In the literature reviewed there l4 appeared to be no basis for relating evaluation in this dimension to the characteristics of household heads that have been selected for this analysis. Therefore the following hypothesis was derived: H6. Frequency of mention of interpersonal relations is not related to: education of male household head . urban experience of male household head education of female household head . urban experience of female household head total household income (DO-O 0‘03 CHAPTER II REVIEW OF RELATED LITERATURE The research and related literature will be reviewed under two major headings: housing adjustment behavior and squatter housing. Housing Adjustment Behavior Various theoretical papers and reports of empirical analyses on housing adjustment were reviewed as background to this study. None were directly related to the topic under study, i.e., residential evaluations and characteristics of members of squatter households in Lusaka, Zambia. However, they have provided some insights into some of the components of housing decisions and behaviors. Specifically, studies contributed to the selection of variables included in this study and suggested some avenues for the interpretation of study findings . In 1955 Rossi published a landmark study on why short- distance residential shifts take place. His analysis was performed on three levels: the level of mobility in urban areas, mobility of different types of families, and motiva- tions underlying residential shifts. From his findings Rossi concluded that mobility is "the process by which families 15 l6 adjust their housing to the housing needs that are generated by the shifts in family composition that accompany life cycle changes" (p. 9). Rossi created a Mobility Potential Index from character- istics of the household such as life cycle and tenure status variables which represented housing needs. He also devised a Complaints Index from residents' attitudes about their housing environment which represented whether present housing met the needs of families. Analysis of the Complaints Index assessed the importance of different kinds of dissatisfaction to residential mobility. Out of the 14 questions used to construct the Complaints Index, six dimensions were abstracted including dwelling unit space, utilities, distance, physical environment, social environment, and housing costs. The two indices were found to be good predictors of resi- dential mobility which Rossi operationalized as desires and plans concerning mobility in the future. The relationships between the indices and mobility desires and plans are thus determined by household needs, dissatisfaction, and aspira- tions. Sabagh, Van Arsdol, and Butler (1969) described a con- ceptual framework for the study of residential mobility which evolved out of Rossi's mobility accounting scheme. They found that some of the factors related to mobility are changes in family life cycle, social mobility aspirations, changes in residential environment, and patterns of social and locality participation. They also discussed some intervening factors 17 which may facilitate or impede residential mobility including availability of residential opportunities, availability of resources, and availability of residences sought by the family. Wolpert (1966) promoted a theoretical stress-threshold model of mobility which has paved the way for further research. According to this model, which described the relationships between individuals and their environment, stressors are in- troduced from the environment at different stages of the family life cycle. Stress in the environment or action space of household members may cause strain, and it affects their decision behavior. Household members undertake adjustment processes to reduce this strain. Sources of stress vary for different households as do people's responses to it. Brown and Moore (1971) further developed the concept of household decision-making in the process of residential mo- bility using this stress-threshold model. They suggested that if place utility, a measure of individual satisfaction with a residence, diverges too much from.needs, household members may decide to move in order to bring their residence into adjustment with their needs. Stresses in the environment derive from.this disparity between needs and characteristics of a residence and may be reduced not only by moving but also by adjusting present housing or by adjusting the house- hold. Brown and Moore suggested that an understanding of the decision-making process at the individual or household level 18 may indicate variables to be included in formulations at the aggregate level. An understanding of the nature of individual responses to environmental conditions will provide a basis for evaluating decisions related to planning for the growth add development of urban areas. Clark and Cadwallader (1973) proposed a model which takes into account the overall environment of the household as the context in which decisions are made. The decision to move is a function of the household members' present level of sat- isfaction and the level of satisfaction they perceive may be attained elsewhere. The difference between these levels is a measure of stress, and decision to move is an adjustment to this stress. Possible stressors include size and facilities of dwelling unit, kind of people living in the neighborhood, proximity of household location for interaction with relatives and friends, proximity to work place, and amount of air pol- lution. Speare (1974) also worked with the stress-threshold model to analyze data from Rhode Island residents. The study was designed in an effort to answer the question of why people move and thus to help explain the relationship between volun- tary mobility and other variables. He found that residential satisfaction functions as an intervening variable between individual and residence variables such as age of household head, duration of residence, home ownership, and room crowding, on the one hand, and mobility on the other. The more satis- fied individuals are with their residences, the less likely 19 they are to consider moving. Newman and Duncan (1979) used longitudinal data on a national sample of families to explore the incidence of per- ceived housing and neighborhood problems, thearelationship between these problems and satisfaction with house and neigh- borhood, and the impact of these problems on actual mobility. This study, also set in the stress-threshold framework ex- plored by others, found links between problems and discontent, but there appeared to be no strong links between problems and actual moves. Morris and Winter (1975) provided a conceptual and theo- retical framework for the study of housing adjustment behavior. Family members are viewed as evaluating their housing in terms of cultural norms and family norms. When housing does not meet norms, it tends to give rise to dissatisfaction producing a normative deficit. Residential mobility, residential adap- tation, and family adaptation are modes of adjustment. Vari- ous constraints operate to impinge on household members' ability to engage in successful adjustment behavior. It is the com— bination of theseconstraints, current housing conditions and norms which influence adjustment behavior of household members. Few studies on housing adjustment appear to have been undertaken outside of the continental United States. One exception is a study by Okraku (1971) in San Juan, Puerto Rico, which examined the appropriateness of the family life cycle concept as it related to residential mobility in a developing area. Okraku found that the life cycle concept may be of 20 value under different cultural and socioeconomic conditions, although careful consideration must be given to whether the basic assumptions of the life cycle concept (such as family nuclearity) are satisfied. In some respects these studies have raised questions concerning the viability of housing adjustment as a model for the study of housing behaviors in non-Western societies, the appropriateness of the model in a context characterized by limited resources and a lack of housing alternatives, the relevance to other contexts of variables found to be related to housing adjustment in these studies, and the potential significance of hypothesized relationships using data from other societies. Much further research needs to be under- taken. The purpose of this study is not to test a full model of housing adjustment which includes many different components. Rather, this study considers the evaluation of residential situation that occurs in the housing adjustment process and possible relationships between evaluations made by male and female heads of household and characteristics of those in- dividuals. Taken together the studies reviewed indicate that housing adjustment is a complex but essentially rational process on the part of household members. Housing adjustments may take place when a certain level of stress or dissatisfaction is reached provided that there are no constraints on this action. Dissastisfaction may arise with respect to various aspects of housing as a result of changes in either characteristics 21 of household members or their residential situation. These studies suggest that several groups of variables need to be considered in an effort to understand housing adjustment behavior. These include characteristics of the household and its members; characteristics of the residential situation of the household; household members' feelings with respect to that situation; and housing plans, decisions, and behaviors of household members. Data to investigate all of these aspects of housing adjustment were not readily available. Given this constraint, but also an interest in limiting the sc0pe of this study, only some variables related to charac- teristics of household members and their feelings about their residential situation were selected for analysis. Squatter Housing Recent literature on squatters and squatter housing with particular attention to the African context was reviewed in order to gain a better understanding of the problem.under study and to determine what relationships might be usefully tested given the available data. Most of the studies that were reviewed provided discus- sion related to the establishment and growth of squatter settlements. African countries have experienced high rates of urban growth, particularly in their post-independence periods, that can be attributed to rural-urban migration. Simmance (1972) noted that in Zambia a decline in the rural economy has been accompanied by economic growth and industri- 22 alization in urban centers. In the 19603 the gap between levels of real income in rural and urban areas widened as incomes in cities rose rapidly and those in rural areas in- creased very little. Similarly, the promise of economic Oppor- tunities in urban areas has attracted many pe0ple. Bates (1974) cited lack of private investment in rural areas and failure of policy-makers to counteract the tendency for capital accumulation in urban areas as the primary cause of the rural- urban disparity in Zambia. With the growth in population in cities, a tremendous demand has been placed on the housing sector. Governments have been unable to provide for the construction of a suffi- cient number of dwellings for city inhabitants. Simmance (1972) cited Lusaka as an example. In 1969 the Lusaka City Council had a waiting list for housing for 16,345 families and 3,636 single persons, but completed only 668 houses during the year (p. 29). People have had to provide dwellings for themselves, and squatter settlements have grown rapidly. Although such self-help activity has met people's immediate needs, squatter areas lack basic services making their existence problematic for authorities. Stren (1975) and others pointed out that this expansion of squatter areas is not purely a technical problem of providing dwelling structures, but must be viewed in a larger socioeconomic context. For example, residents perceive distinct economic advantages to living in squatter compounds and might prefer to continue living there 23 even if they had alternatives. Much of the literature reviewed indicated that inappro- priate housing standards have further aggravated the problem. Payne (1977) suggested that many standards have omitted the criterion of acceptability as perceived by urban residents, particularly low income groups. Standards have also promoted housing that low income groups simply cannot afford. Simi- larly, Mabogunje (1978) emphasized that standards should evolve from people's biological, psychological, and social needs rather than reflect a middle class technocratic percep- tion of what these needs are. Policies towards squatters and squatter housing have changed over time. Tipple (1976) described how the First National Development Plan, 1966-1970, of Zambia proposed that 76 percent of the houses built in that period should be in site and service schemes. In site and service schemes, local housing authorities lay out plots; provide roads, water, and sanitation; issue roofing loans; supervise construction; and charge modest rents. Residents construct their own dwellings (Simmance, 1972, p. 32). This type of housing is based on the belief that independence had established the right of every citizen to own urban land and that site and service schemes would assure this right in an orderly fashion. The govern- ment officially intended to move the growing number of squatters to site and service areas. However, the policy and the efforts were not successful, and new directions have been taken. Through the Housing (Statutory and Improvement Areas) 24 Act of 1974, the government has accepted the view that squatters have rights to security of tenure and improved facilities. This change in policy has paved the way for the promotion of local participation in the provision and improve- ment of housing. Martin (1977) wrote in support of such policy changes. He suggested that the Zambian government and housing authori- ties probably could not afford to construct enough housing for all, that people could not afford such housing if it were available, and that initial outlays for site and service schemes were too expensive. He concluded that upgrading of existing squatter areas is the only alternative. Perceptions of squatter areas and their residents as presented in the literature reviewed lend support to a policy that promotes upgrading and calls for involvement of people in the provision of housing. In the context of squatting, many authors applied the concept of housing as a process of meeting certain needs (as Opposed to being a dwelling unit) and considered housing in terms of a dwelling environment including not only the dwelling structure itself but also the surrounding community. Turner and Fichter (1972) emphasized that housing should be viewed in terms of the various functions that it performs such as shelter, location, and security. The priority that people accord to different functions changes as their needs change. For example, people with low income have low priority for "modern" shelter but have high 25 priority for proximity to jobs. These priorities reverse as incomes rise. Mabogunje (1978) also linked the housing pro- cess in the context of low-cost housing to people's changing socioeconomic conditions and stressed that housing can serve as a means to other ends. Although it was stressed by a number of authors that characteristics of squatter areas vary across cultures and even within cultures, some common features seem to exist. Herbert (1978) described the functions of squatter housing areas as providing shelter at affordable rates, serving as receiving areas for new arrivals to the city, serving as sites for employment in commerce and cottage industries, providing shelter close to employment, providing a context for strong social linkages and support, and encouraging and rewarding small-scale private investment in building. To this list of basic functions Martin (1977) added the notion that because of less rigid controls or standards than those often applied to "official" housing, squatter housing pro- vides residents with the opportunity to make improvements in their housing when they both desire and can afford it. Andrews, Christie, and Martin (1973) described several squatter areas in Lusaka, Zambia, in terms of both achieve- ments and needs. They emphasized that most of the areas are well-situated in relation to work places, that the organi- zation among individual houses both permits privacy for house- holds and facilitates interaction among neighbors, that come munity organization is maintained by the political party UNIP 26 (United National Independence Party), and that residents have demonstrated initiative in meeting what needs they can. Nevertheless, most squatter areas lack clean water supply, well-maintained roads, and sufficient number of schools and clinics. The Zambian National Housing Authority (1972) prepared a report for the Ministry of Local Government and Housing which summarized available information on squatters in Lusaka. Findings indicated that there is a fairly high employment rate ambng squatters but that their educational level is low (most of those with higher levels of education are in jobs which provide for housing). Few of the houses have piped water, most peOple use pit latrines, few roads in squatter areas are maintained, and residents are generally within reach of schools and clinics. There is evidence of much self-help in construction of clinics, schools, and com- munity organizations; in maintenance of roads; in care of elderly people; and in improvement of houses. Residents are generally satisfied; they mention liking the social environ- ment, the physical environment, and the convenient location. Services is the most frequently mentioned area for improve- ment. Some of the literature reviewed was concerned with spec- ific attributes or behavior patterns of squatters in parti- cular or new urban residents in general. Gugler (1969) discussed the phenomenon of urbanization and emphasized that mere presence in an urban setting can have important impacts 27 on a person because of contacts with different people, jobs, and authorities. Lloyd (1979b) suggested that the perceptions of resi- dents of urban settlements may be based on their experiences. He offered a decision-making approach for understanding squatters and their behavior. Squatters act in order to improve their quality of life, but their goals are dictated by their social position and their resources, particularly economic ones. What may be considered an unaffordable lux- ury at one point in time may be desirable and attainable at another. Many housing behaviors in this context may be-a result of planned choices, but some reflect a lack of alter- natives. Lloyd (l979a) suggested that in order to develop policies and programs for squatters it is important to under- stand their perceptions of opportunities available to them and to recognize their aspirations. In the literature reviewed, squatter housing was gen- erally considered to be an inevitable response to an urban housing shortage created by rural-urban migration. Further it was viewed as offering a viable alternative to counterpart "official" housing. Although squatter housing meets many needs, others remain unmet. A policy of upgrading is in order. Squatters were presented as people with initiative and a willingness to help themselves. However, they are constrained in their activities by a lack of resources. As conditions change (such as an increase in their income or experience in an urban area, some needs are met or become 28 less crucial whereas others become more important. The literature reviewed seems to indicate that for those with the fewest resources, i.e. those with little income, education, or experience living in urban areas; squatter areas meet immediate needs for low-cost housing which is close to place of employment. These economic and locational dimensions of residence may be replaced in importance to squatters by other dimensions, such as house and compound quality and provision of services, as characteristics of squatters change. In a society undergoing rapid social changeznulurbanization as is Zambia, characteristics that relate to social class and urbanization, such as income, education, and urban experience, may bear some relationship to people's housing needs. The nature of and changes in these relationships may be reflected in people's evaluations of their residential situation. CHAPTER III PROCEDURES This study was a secondary analysis of data collected in a survey of residents of households in urban Zambia. In the original study, the household was the unit of analysis with the male head as the point of entry. This study was under- taken to determine the content of household heads' evaluations of their residential situation and to test relationships be- tween these evaluations and selected characteristics of these individuals. The characteristics selected included years of education of male and female heads of household, years of urban residence of male and female heads of household, and total household income. Collection of Data The data used in this study were collected in 1973 for a housing survey of three Zambian cities (Lusaka, Kitwe, and Ndola). The survey, Zambia Study of Urbanization and Housing, was directed by David S. Wiley, then of the Department of Sociology, University of Wisconsin, Madison.* *Major funding for the project was provided by the Sociology of Economic Change Training Program of the Depart- ment of Sociology. University of Wisconsin, Madison, under a training grant from the National Institute of Mental Health. Supplemental funding was provided by the Midwest University 29 30 Data collected for the total study involved two ran- domly selected samples. The first sample focused on the household as the unit of analysis whereas the second focused on the individual as the unit of analysis. The first sample was used in this study. Cluster sampling was used to select compounds from which households were then selected. The compounds were of three general types determined by the type of management: compounds managed by the city council; compounds managed by particular companies; and squatter compounds under no official manage- ment but which have, in many cases, been recognized by local housing authorities and in government policy as permanent residential areas. A sampling design was developed which would make it possible to draw a random sample of households yet would not scatter these households over large areas. This would facili- tate collection of data by interviewers who walked from house to house. Therefore, sampling units were created of approxi- mately 65 houses each (identified primarily from aerial photo- graphs). Ten households were randomly selected from each sampling unit for detailed structured interviews with both the male and female heads of each household. The total number of households from all compound types in all three cities from which interviews were taken was 3,270. Consortium for International Activities, Inc. Funding for Zambian participation in the research process was provided by the Ford Foundation. Support for data coding and analysis was given by the Rockefeller Foundation, Social Science Division. 31 For the purposes of this study, a sub-sample of all 1,068 households in squatter compounds in Lusaka was used. The sub-sample was selected because it was large enough to permit confidence in findings without necessitating the inter- pretation of the influence of city size or compound type on the variables and relationships under study. This is not to suggest that such influences are not important but rather that they will be left to future analyses. Responses from both male and female heads of household were included, thus responses were obtained from 2,136 indi- viduals. In the case of single-headed households, the male or female head present responded to both male and female head sections of the questionnaire. These responses were included in the present analysis. In the sample of 1,068 households involved in the present analysis, the sections of the ques- tionnaire intended for male heads of household were responded to by 954 male heads and 107 other respondents (7 missing cases). The section of the questionnaire intended for female heads of household was reSponded to by 942 female heads and 121 other respondents (5 missing cases). Description of the Survey Instrument The survey on which this study was based was undertaken to examine characteristics of households and household members which might affect evaluations, attitudes, and decisions con- cerning housing and other social services in various urban compounds or residential areas. Questions in the survey 32 instrument were designed to elicit information on social and demographic characteristics of household members including mortality and fertility, life cycle stage, health, education, occupation, and ethnic origin; housing and compound evalua- tions; and personal aspirations and plans. The instrument was comprised of 150 questions in several sections. Nine questions answered by the household member with whom the interviewer had initial contact covered names, ages, languages, educational status, and length of residence in household of all household members. Six questions answered by the interviewer, based on observation, covered characteristics of the house including availability of elec- tricity; size and shape of house; wall, roof, and door com- position; and the state of repair of the house. Thirty- three questions answered by all respondents covered residen- tial history, ethnic background, religion, occupation, income, and health. Female heads of household responded to another 19 ques- tions on their evaluations of their residential situation. All female respondents answered five questions on their fer- tility and infant mortality. Male heads of household answered 37 questions on their evaluations of their residential situ- ation and on other residential experiences, decisions, and aspirations. Respondents selected randomly for the second part of the study answered another 41 questions on residential history, experiences with crime, personal aspirations, occu- pational history, friendship patterns, perceptions of urban 33 and rural life, and organizational affiliations. However, these were not utilized in the current analysis because of the interest in using households rather than individuals as the unit of analysis. The questions were originally written in English and then translated into both Bemba and Nyanja, two languages widely spoken in Zambia. Measurement of Variables Two categories of variables, residential evaluations and characteristics of household members, were derived from ques- tions in the original survey instrument and analyzed in this study. Evaluations of Residential Situation Both male and female heads of household in the survey sample answered questions concerning evaluations of their residential situation. The following questions from.the female head section of the instrument were selected for analysis in this study: 57. What things do you like best about this compound? 58. What things do you dislike the most about this compound? The following questions from the male head section of the in- strument were also used: 86. What things do you like best about this compound? 87. What things do you dislike the most about this compound? 34 Each of these four questions was Open-ended and permitted an unlimited number of responses. Detailed coding was developed for the primary analysis of the data; 100 different response categories were recorded for the entire study population with 88 appearing from the sample of Lusaka households.* For this study the responses were assigned to six dimen- sions. To a large extent these groupings had been determined by the original coders who had sorted and combined all of these into five dimensions. This researcher independently sorted the response categories into groups which were nearly identical to the five created by the original coders. The only exception was the creation of a dimension of location by this researcher; this dimension had been incorporated into compound quality by original coders. Because of the high level of agreement between the response groupings made by this researcher and those made by original coders, and because all of the original responses were not available to the researcher, it was decided not to use item analysis or factor analysis to determine dimensions of residential situ- ation. The categories thus generated appear to conform to those identified in literature on housing adjustment and on squatters. They represent basic dimensions of residential situation and include: economic conditions, location, house * O O O C A number of coders were involved in this process until a level of 1% error between coders and supervisors was achieved. 35 quality, compound quality, services, and interpersonal rela- tions. The complete lists of responses grouped into each of these dimensions are found in Tables 3, 4, 5, 6, 7, and 8. Several of the responses to the evaluation questions were not incorporated into any of the six dimensions. These included the responses "nothing, everything,’ and "do not know." Because the intent of this study was to examine rela- tionships between evaluations of particular dimensions of residential situation and characteristics of household members, it was decided not to analyze these responses directly. However, these responsesdfiklhave an effect on the indices of evaluation of residential situation that were created because the indices were developed out of a single pool of responses in which "nothing, everything,‘ and "do not know" were included. These findings do indicate, however, generally negative feelings about residential situation. Nearly 50 percent of the female heads of household mentioned no likes of their residential situation. The same was true of nearly 40 percent of the male heads of household. The distribution of these responses is provided in Table l. The measures of evaluation were based on the notion that residential situation is comprised of several basic dimensions which encompass both positive and negative features. Mention of a particular dimension, in terms of either a like or a dislike, was an indication that the dimension was important to the respondent. The more frequently a dimension was men- tioned, the more important it was. On this basis, the likes 36 Table 1. Responses to Evaluation Questions Not Included in Evaluation Indices Response Number of Responses3 Femalg Female Male b Male Total c Likes Dislikes Likes Dislikes Responses "Nothing" 431 103 258 100 892 "Everything" 0 0 2 4 6 Missing "Do not know" 2 2 4 1 9 Did not reSpond 84 77 125 121 407 Totald 517 182 389 226 1,314 Percent of Column N 48 17 36 21 31 aIncludes those responses which were not combined into the six evaluation indices. bN=1,068 CN=4,272 dColumn totals do not equal N because the interview questions were open-ended. Some respondents did not provide these particular responses. 37 and dislikes of both male and female heads in each household were combined to create indices indicating the relative levels of importance of these six dimensions of residential situa- tion to the household heads. All of the responses were used with no consideration given to the order in which they were provided by respondents. The six indices thus created on the basis of frequency of mention were: 1. Evaluation of economic conditions of residence Evaluation of location Evaluation of house quality Evaluation of compound quality mbwm Evaluation of services 6. Evaluation of interpersonal relations Each household unit received a score on each of the six indices. The score on a particular index for a household was determined by counting the total number of responses pro- vided by the male and female heads that could be grouped into the dimension measured by the index. For example, if the female head of household liked the neighbors (in the dimension of interpersonal relations) and the male head of household disliked them and no other responses were given in this dimen- sion, then a score of two would be registered in the index interpersonal relations for that household. To eliminate missing cases in the creation of the indices, those cases where the first response to any one of the evalua- tion questions was missing or marked "do not know" were elimi- nated. However, in some cases where "do not know" had been given as the first response, other responses had been given 38 as the first response, other responses had been given subsequently. These responses were lost in the creation of the indices. There were some problems in creating indices in this manner. Because the indices were based on open-ended ques- tions, it was not certain that respondents gave all relevant responses to interview questions or that coders correctly understood the intended meaning of some of the responses given. It was not known whether respondents provided valid responses or whether they were influenced in some way by the interviewers or the interview situation. There was no one question contributing to the creation of the indices with a sufficient number of responses which could be used to test the internal validity of the indices. Nor were there other questions within the survey instrument which could be used for external validation. The validity of the indices depended on the validity of the basis on which they were created, i.e. that mention of. a dimension indicates its importance to the respondent. Nevertheless, it was felt that such indices based on respon- dents' original responses to questions could offer some in- sights into the importance of different dimensions of the residential situation of squatters in Lusaka, Zambia. Table 2 provides the distribution of cases on the six indices which were created. The majority of households scored "0" on all of the indices except interpersonal relations and services. This reflects the dominance of these categories in 39 the responses given by household heads to the open-ended evaluation questions. The large number of cases scoring "0" on the indices may also be attributed, in part, to the high frequency with which "nothing" was given as a response. Once the indices were created, the frequency and distri- bution of individual responses which had been incorporated into the indices were examined. The findings of these analyses are indicated in Tables 3, 4, 5, 6, 7, and 8. Exami- nation of all responses to the evaluation questions provided an insight into the potential difficulty of using these in- dices as analytical tools. The indices were based on the notion that each of the six dimensions of residential situa- tion consisted of counterbalanacing positive (likes) and negative (dislikes) features which, when combined, would constitute a "neutral" dimension of residential situation. However, it was evident that aspects of economic conditions of residence, location, and interpersonal relations appeared most frequently as likes,whi1e house quality, compound quality, and services appeared as dislikes. It was decided to create other indices in addition to the six originally proposed which would include only likes or dislikes depending on which seemed to dominate in each dimen- sion. The indices were constructed in a manner similar to the original indices, i.e. by counting in each dimension the number of combined likes of male and female heads of household or the combined dislikes of male and female heads. The 40 .wcwvnpou mo museums 00H cu Hmuou uoc ewe mucmouom n .mcowmaoBHp xwm on» mo oco oucH pompouw on panoo msowumosU cowumsam>m Hmwuomvwwou osu ou uncommon mac unmoa um Scene Mom mpaoaomsos mo Henson onu mH z Hmuoem OOH O O N O OO Om HO OOOHOOHOO OOO O O OH OO OOO OOO HON HOcoOuOOHOOOH OOH O H O O O HH OH ON HO O O OOO O HH NO NO OO OO OOH OOH OHH OH OO OOOH>HOO OOH O H O OH HO OOHHOOO OOO O H OH HOH OOO Oaaoaaoo OOH O O OH HO Om OOO O OO HOH OOO OOO OOHHOOO mason OOH O H O HO OOO O O OO HOO OOHOOOOH OOH O H O HO OO OOO O OH OO OOH OOO OOOHOHOOOO owfioaoom Am 2 O z N z N z N z N z e z N z N z N z N z e z OHOOOH. OH O O O O O O O .O. H O GowumsHm>m ouoom 30mm um mpaonomnom mo unmouom mam Honssz .mo cowmamfiHn AmmowccH co monoom mo COODSAOMumHQV coHumSHOm Hmwuaovwmom mo maOOumnHm>m .N OHOOH 41 indices thus created included: 1. Likes of economic conditions of residence Likes of location Dislikes of house quality Dislikes of compound quality Dislikes of services GUI-PLAN Likes of interpersonal relations The distribution of cases on these indices is indicated in Table A in the Appendix. In order to better understand the composition of the original indices, it was decided to examine separately the responses of male and female heads. It was observed that they demonstrated slightly different response patterns to the evaluation questions. Indices were therefore created which would represent the responses in the six dimensions of male and female heads of household separately. Following the finding that each dimension was dominated by either likes or dislikes, only likes or dislikes were involved in the indices, depending on which dominated the responses in that dimension. These in- dices included: 1. Male likes of economic conditions of residence Male likes of location Male dislikes of house quality Male dislikes of compound quality Male dislikes of services Male likes of interpersonal relations \JO‘U‘IJ-‘UDN Female likes of economic conditions of residence 8. Female likes of location 42 9. Female dislikes of house quality 10. Female dislikes of compound quality 11. Female dislikes of services 12. Female likes of interpersonal relations The distribution of cases on these indices is indicated in Tables B and C of the Appendix. Characteristics of Household Members Some characteristics of household members, particularly those relating to social class and urbanization, may be asso- ciated with the extent to which household members may under- stand and be able to manage their residential situation and their eXpectations of this situation. Characteristics of household members selected for this analysis were number of years of formal education of both male and female heads of household, urban experience of both male and female heads of household measured by the number of years that they have lived in urban areas, and total household income. These three characteristics were selected mainly because they were easily measured from available data. It was also felt that they represented characteristics which may differen- tiate p0pulation groups into social classes in a society undergoing such changes as rapid urbanization, increased oppor- tunities for formal schooling and increased participation of citizens in the paid labor force giving them more income. Other characteristics of household members may also be relevant to a study of residential evaluations and behaviors. These include personal and structural attributes of households, 43 social linkages of household members, and material posses- sions of household members. These were not considered in this study primarily because they could not be easily derived from the available data and because it was felt that they were not as relevant totflmapurposes of this study as those that were selected. Education. One question was selected from the original instrument pertaining to years of formal education of heads of household and was answered by both male and female heads. 20. Up to what grade or standard did you reach in your education? (Response recorded as total years of education.) The distribution of education of heads of household is indi- cated in Table D of the Appendix. Urban Experience. One question was selected from the original instrument pertaining to the urban experience of household members measured by years of residence in an urban area. This question was answered by both male and female heads of household. 16. How many years of your life have you lived in town? (Response recorded in total num- ber of years.) The distribution of urban experience of heads of household is indicated in Table E of the Appendix. Household Income. Five questions in the original instru- ment were used to derive a measure of household income in kwacha (Zambia's basic currency). Three questions were taken from the section to which both male and female heads of 44 household responded: 26. How much money do you receive at the job you are working presently? (Response recorded in kwacha per month.) 28. Do you get a house allowance for your job? (If yes,) how much do they give you per month? (Response recorded in kwacha per month.) 29. Do you have any jobs that you do to earn extra money apart from your full-time job -—jobs like selling beer or food, as a driver, as a builder? Tell me each of the things you do or make. How much do you make from each of these other jobs? (Response recorded in kwacha per month for each job.) Two questions from the section to which male heads of house- hold responded were also used: 98. Do you have anybody in this house who pays rent to you? (If yes,) how much rent all together do they pay you each month? (Response recorded in kwacha per month.) 99. Not including the mbney that you and your wife earn per month, are there any other people in this house who bring in extra money for food and other things needed in the house? (If yes,) how much per month? (Response recorded in kwacha per month.) The distribution of total household income is indicated in Table F of the Appendix. Method of Analysis Two stages of analysis were undertaken in this study. The purpose of the first stage was to determine the content of the evaluations of residential situation made by male and female heads of household. Such analysis permitted an under— standing of the composition of the indices of six dimensions 45 of residential situation created for each household on the basis of these evaluations. This involved examining the frequencies of combined positive (likes) and negative (dis- likes) responses by male and female heads of household in each response category grouped by dimensions. The positive and negative responses and the responses made by male and female heads of household were considered separately as a sub-analysis in order to gain a better understanding of the combined responses. The second stage of analysis was undertaken to determine relationships between male and female heads' evaluation of their residential situation and their education, urban ex- perience, and total household income. Pearson Product Moment correlations were used to assess the degree and direction of these relationships. Correlation analysis requires that variables be of an interval level of measurement. Both the variables representing characteristics of household members and the evaluations indices satisfy this requirement. Correlation coefficients index two properties of a relationship, the magnitude and the direction. The magnitude indicates the degree to which variables vary together. The direction indicates whether the variables vary together (positively) or inversely (negatively). The coefficient has a range from +1.0 (perfect positive correlation) through 0.0 (no correlation) to -l.0 (perfect negative correlation) (Williams, 1979, p. 122). In order to interpret the magnitude of the correlation coefficients that were found, one-tailed 46 tailed tests of statistical significance were used. The level of significance for acceptance of the hypotheses being tested was set at p<.05. To test the research hypotheses of the study, correlation coefficients were calculated on relationships between the six indices of evaluation of residential situation and male and female household heads' education, urban experience, and total household income. Additional correlation analyses were performed on relationships between these characteristics of household heads and the additional indices that were created. All of the analyses in this study were performed using the Statistical Package for the Social Sciences (SPSS) (Nie, Hull, Jenkins, Steinbrenner, and Bent, 1975), Version 8a, on the Control Data Cyber 750 computer at the Michigan State University computer center. CHAPTER IV FINDINGS In this chapter the results of the analyses that were undertaken are reported. In the first section are the fin- dings of the examination of the content of evaluations of residential situation made by male and female heads of house- hold. The second section includes the results of the testing of hypothesized relationships between variables representing education and urban experience of male and female heads of household and total household income, and the indices of the evaluations in six dimensions of residential situation. The third section covers the results of additional analyses of relationships between those characteristics of household mem- bers and other indices that were constructed from responses to evaluation questions. Residential Evaluations Frequencies of all responses to residential evaluation questions were examined to determine how responses were dis- tributed. This analysis contributes to an understanding of the composition of the indices which were created for each of the six dimensions of residential situation. It also con- tributes to an understanding of the dimensions of residential 47 48 situation that are important to members of the study sample. Economic Conditions of Residence The distribution of the 11 different responses grouped into the general category of economic conditions are provided in Table 3. Observation of the frequency of responses in this dimension indicates that the issues of rent cost and tenure dominate the responses. Sub-analysis of likes and dislikes spearately reveals that positive evaluations (likes) tend to appear in this category more frequently than negative evaluations (dislikes). The only economic condition about which more dislikes were expressed was the cost of goods. Responses of male and female heads of hosuehold fell in simi- lar patterns in this dimension, although male heads mentioned relatively more likes than did females. Location Table 4 provides the complete distribution of the three responses grouped into the dimension of location. Both location with respect to work and location with respect to town were important. Positive evaluations (likes) of resi- dential situation appear more frequently than negative evalua- tions (dislikes) in this dimension. Location was mentioned relatively more frequently by male household heads than by female household heads. Patterns of response seemed slightly different for the two groups. The most frequent response in this category of male heads was location with respect to work. For female heads location with respect to town was most frequent. This difference may be explained in part by 49 Table 3. Evaluation of Economic Conditions of Residence *— t Response Number of Responsesa Residential Female Female Male Male Evaluation Likesb Dis likens Likes b Dis likesb (N=982) (N=989) (N=939) (N=946) Rent cost 138 61 7 68 2 Tenure 134 55 4 73 2 Standard of living 32 18 3 8 3 Personal plans 30 14 0 l6 0 Cost of goods 23 2 13 l 7 Job or business 20 4 l 11 4 Money 2 3 3 0 Land cost 7 5 0 2 0 Plot fees 2 0 1 Not own house ' 4 1 2 l 0 Risky situation 3 l 0 2 0 Totalc 403 165 33 186 19 Percent of total 100 41 8 46 5 aIncludes only the responses which could be classified into this category. bN is the number of respondents providing at least one response to the general Open-ended interview questions. cColumn totals do not equal N because the interview questions were open-ended. Some respondents did not provide responses in this category; others may have provided more than one response. 50 Table 4. Evaluation of Location Response Number of Responses'1 Residential Female Female Male Male Evaluation Likesb Dis lik s Likesb Dis likesD (N=982) (N=989) (N=939) (N=946) Location, near work 54 7 l 45 1 Location, near town 45 17 0 27 1 Location, general 1 1 0 0 0 Totalc 100 25 1 72 2 Percent of total 100 25 l 72 2 aIncludes only the responses which could be classified into this category. bN is the number of respondents providing at least one response to the general open-ended interview question. cColumn totals do not equal N because the interview questions were open-ended. Some respondents did not provide responses in this category; others may have provided more than one response. 51 different roles of male and female heads. In this study population, 92% of male heads of household had jobs whereas only 19% of female heads did. House Quality Table 5 provides the complete distribution of the 15 responses grouped into the dimension of house quality. Clearly, the issue of toilets dominates this category with the general quality of the house as the second most frequent response given. When likes and dislikes are considered separately, it is evident that negative evaulations (dislikes) of residential situation are found in this category more frequently than are positive evaluations (likes). Response patterns for male and female houehold heads were very similar in this category. Compound Quality Table 6 provides the complete distribution of the nine responses grouped into the dimensions of compound quality. Both issues of general compound appearance and dust and dirt were important. Consideration of likes and dislikes sepa- rately reveals that negative evaluations (dislikes) of resi- dential situation appeared in this dimension more frequently than did positive evaluations (likes). However, if the com- pound was perceived to be quiet, this was considered to be a positive quality; no one rated quiet as a negative factor. Male heads of household gave relatively more responses in this category than did female heads of household, particularly 52 Table 5. Evaluation of House Quality Response Number of Responses'i Residential Female Female Male Male Evaluation Likesb DislikeDs Likes b DislikesD (N=982) (N=989) (N=939) (N=946) Toilets, gen- 362 2 182 1 177 House, gen. 135 4 74 5 52 Toilets, type 80 O 42 l 37 Plot, size 35 3 7 1 24 Toilets, cleanliness 29 0 l4 0 15 House, size 25 0 ll 1 l3 (Rude: ll 4 l 5 Plot, gen. 8 0 2 1 No plot allocated 4 2 l 0 1 Plot, cleanliness 4 0 3 0 thmy'qne 2 0 2 0 Odmm'hamn: qualities 2 0 l 0 l Kiuflen,g§m~ l 0 0 0 l Baduxem,g$n. l O 0 0 l lkflls,type l 0 0 0 l Totalc 700 15 340 15 330 Penxmm of total 100 2 49 2 47 aIncludes only the responses which could be classified into this category. bN is the number of respondents providing at least one res- ponse to the general open-ended interview question. CColumn totals do not equal N because the interview questions were open-ended. Some respondents did not provide responses in this category; others may have provided more than one response. 53 Table 6. Evaluation of Compound Quality Response Number of Responsesa Residential Female Female Male Male Evaluation Likesb Dis likens Likes b Dislikeos (N=982) (N=989) (N=939) (N=946) Appearance 74 6 31 5 32 Dust and dirt 63 l 33 3 26 Quiet 36 7 0 29 0 Noisy 33 2 7 3 21 Compound, general 30 9 5 l3 3 Congestion 26 0 ll 0 15 Mud or wet S O l 2 Heat . 4 0 l 2 1 Cold 1 0 l 0 0 Totalc 272 25 91 56 100 Percent d of total 100 9 33 21, 38 aIncludes only the responses which could be classified into this category. bN is the number of respondents providing at least one res- ponse to the general open-ended interview question. cColumn totals do not equal N because the interview questions were open-ended. Some respondents did not provide responses in this category; others may have provided more than one response. dPercents do not total 100 because of rounding. 54 with respect to likes of compound quality. Services Table 7 provides the complete distribution of the 38 responses grouped into the dimension of services. This issue of water,including availability, quality, quantity, and dis- tribution was of overwhelming concern to both male and female heads of household. Over 40 percent of all responses in this dimension dealt with water. Issues of clinics or hos- pitals, road quality, and quantity of schools were also mentioned frequently. When likes and dislikes were con- sidered separately, it was evident that negative evaluations (dislikes) of residential situation appeared more frequently in this category than did positive evaluations (likes). Only the issues of distribution of schools, quantity and quality of shops, sports facilities, religious centers, and beer halls, were liked. Although overall response patterns for male and female heads of household are similar in the dimension of services, some slight differences in specific responses are evident. Both male and female heads of household dislike their situa- tion in terms of availability of water. Female heads of household mentioned general aspects of water relatively more frequently in their negative evaluations than did male heads. Male heads, on the average, mentioned dislikes of road quality, the electricity situation, and clinics and hospitals more frequently than did female heads of household. In general, the quantity of schools appeared in negative 55 evaluations by both male and female heads. Distribution of schools appeared in positive evaluations by some female heads. Welfare centers came up more frequently in negative evalua- tions made by female heads of household but more frequently in positive evaluations by male heads. Interpersonal Relations Table 8 provides the complete distribution of the nine responses grouped into the dimension of interpersonal rela- tions. The issues of interpersonal relations in general fighting, and neighbors dominate the responses in this cate- gory. Examination of likes and dislikes separately reveals that positive evaluations (likes) appear more frequently in thiscategory than do negative evaluations (dislikes). How- ever, issues of crime and security and of political con- flict both appeared in negative evaluations in this category. Separate consideration of male and female heads' responses indicate that male heads of household gave relatively more responses in this category than did female heads, with the most outstanding difference being the greater frequency with which male heads mentioned general aspects of interpersonal relations in their positive evaluations. Summary Table 9 provides a summary of all of the responses to the evaluation questions. The content of the evaluations of residential situation was dominated by concerns with services and interpersonal relations. Responses that could not be grouped into any of the specific dimensions were also very 56 Table 7. Evaluation of Services Responses Number of Responses‘. Rank'> Residential Female Female Male Male Evaluation LikesC Dislikqp Likes Dislikes (11-982) (N=989) (u-939)° (N-946)C Beer halls: 20 24 8 3 9 4 Buses/taxis: general 13 44 12 14 12 6 availability 20 24 9 6 6 3 Clinics/hospitals: general 3 380 26 157 17 180 quality 17 30 2 12 4 12 Electricity: general 10 100 1 38 l 60 quantity 37 1 0 1 0 0 Health: 18 27 2 15 l 9 Police protection: general 23 20 0 3 4 13 quality 34 3 0 0 0 3 Recreation facilities: 33 5 0 2 1 2 Religious centers: 27 13 5 0 8 0 Roads: general 23 20 3 5 0 12 quality 4 374 0 154 3 217 quantity 16 31 1 10 0 20 Schools: general 27 13 4 2 3 4 quantity 5 197 8 93 5 91 ualit 29 11 1 4 2 4 istri ution 14 43 13 7 16 7 other 37 l 0 0 1 0 Services: 34 3 0 2 0 1 Sewage/sanitation: 22 21 0 10 0 11 Shops: general 29 11 7 1 2 1 quantity 11 97 49 16 20 12 quality 19 26 12 7 6 1 Sports facilities: 15 35 7 7 16 5 Street lights: 23 20 0 9 0 11 Transportation: 8 104 16 29 21 38 Trash: general 31 7 0 4 0 3 collection: 9 103 0 59 0 44 Water: general 2 388 11 200 9 168 availability 1 507 8 248 7 244 quality 7 128 1 56 2 69 uantity 6 174 l 95 l 77 istribution 12 73 4 34 2 33 Welfare centers: 26 14 2 6 5 1 Other services: 32 6 1 3 l 1 State of repair: 36 2 0 1 O 1 Tot-ti" 3,080 214 1.313 185 1,368 Percent of Total 100 7 43 6 44 ‘lncludes only the responses which could be classified into this category. bRanked by frequency of response of residential evaluation. cN is the number of respondents providing at least one response to the general open-ended interview question. dColumn totals do not equal N because the interview questions were open-ended. Some respondents did not provide responses in this category; others may have provided more than one response. 57 Table 8. Evaluation of Interpersonal Relations Response Number of Responsesa Residential Female Female Male Male Evaluation Likesb Dis likets Likesb Dis likeg (N=982) (N=989) (N=939) (N=946) Interpersonal relations, general 559 208 10 324 17 Fighting 158 43 14 70 31 Neighbors 147 74 1 68 4 Crime and security 92 10 23 18 41 Friends 78 29 0 49 0 Political conflict 33 3 7 6 17 Freedom 26 6 l 18 l Relatives 9 5 0 3 1 Drinking 5 0 l 3 1 Totalc 1,107 378 57 559 113 Percent d of total 100 34 5 50 10 aIncludes only the responses which could be classified into this category. bN is the number of respondents providing at least one res- ponse to the general open-ended interview questions. cColumn totals do not equal N because the interview questions were open-ended. Some respondents did not provide responses in this category; others may have provided more than one response. dPercents do not total 100 because of rounding 58 Table 9. Evaluation of Residential Situation——Summary gziggggg Number of Responseséi Evaluation Female Female Male Male Likes b Dislikes Likes b Dislikeg (N=1068) (N=1068) (N=1068) (N=1068) Economic conditions 403 165 33 186 19 Location 100 25 l 72 2 House quality 700 15 340 15 330 Compound quality 272 25 91 56 100 Services 3,080 214 1,313 185 1,368 Interpersonal relations 1,107 378 57 559 113 Other c responses 898 431 103 260 104 Totald 6,560 1,253 1,938 1,333 2,036 Percent of total 100 19 30 20 31 8Includes all responses to residential evaluation questions. bN is the number of respondents in the study sample. cSee Table 2. dColumn totals do not equal N because the interview questions were open-ended. Some respondents did not provide responses in some of the dimensions; others may have provided more than one response. 59 frequently provided. When likes and dislikes of residential situation were considered separately, it was clear that the positive (likes) and negative (dislikes) responses were not distributed evenly throughout the six dimensions. More negative responses fell into the dimensions of house quality, compound quality, and services, whereas more positive responses related to economic conditions, interpersonal relations, and location. The other responses to the evaluation question, "nothing," "everything,' and "do not know,‘ appeared most frequently in response to the interview question, "What do you like best about this compound?" In general, more female heads of house- hold provided responses in the category of other responses than did male heads. "Nothing" was the most frequently given response by female heads of household to the question about likes. In fact, it was given more frequently by female heads to the likes question than their combined responses in any of the six dimensions. "Nothing" was the second most frequently given response to the likes question by male house- hold heads, being surpassed only by general aspects of inter- personal relations. Beyond this point, however, these other responses were not analyzed because they were not included in the indices of residential evaluation. When responses made by male and female heads of house- hold were considered separately, it appeared that both gave more negative (dislikes) than positive (likes) responses to the residential evaluation questions. 0n the average, male 60 heads of household gave more responses than did female heads, although differences were slight. Overall response patterns for the two groups were similar. Hypothesis Testing In this section, the results of the testing of the hypo- theses using Pearson Product Moment correlation analysis are reported. Table 10 provides a summary of the correlation coefficients of tested relationships. Economic Conditions of Residence The following hypothesis of the relationship between evaluation of economic conditions of residence and selected characteristics of household members was tested: H1. Frequency of mention of economic conditions of residence is negatively related to: education of male household head urban experience of male household head education of female household head urban experience of female household head total household income mQOUm Of these tested relationships, only Hla and Hle achieved an acceptable level of significance. Relationship Hla was found to be in the hypothesized negative direction. However, the magnitude of the coefficients in both cases is very slight. Such findings fail to provide strong support for H1. Location The following hypothesis of the relationship between evaluation of location and selected characteristics of house- hold members was tested: 61 OO.VO Om uOOOHOHcOHO % HHo.n mNH. one. omo. HNo. omo. ofiooaw « s pHonomsos annoy Hoo.n ono. moo. HHo.u moo.s oHo.u poo: onEom mo a mocofiuoaxo amen: mHo.u moo. coo. mwo. moo. mHo.u new: onEom % mo cowumospm wmo. wmo.u NH0.1 «00.: mao.s Hmo. poo: mama mo a monowummxo omen: mmo. mma. Hoo. oqo. wmo. Hmo.u mom: came a s a mo Gowumoapm mGOHumHmm huwamso zuwamno maowuwmaoo HmcomuomuoucH moofi>uom vasomaoo omsom nowumooq oflaocoom cowumsHm>m mo cowmamfiwn memo: paonomzom mo wofiumwhmuomhmso nonmoum acmnmomv cowumSuflm Hmwucopwmom mo moowumoam>m Amcowumamuuoo ucoaoz ou oaoocH paonomnom Hmuoa mam memo: pHozoODom mo oocmfiuoaxm amen: mam cowumoswm mo wafinmaowumaom .oH manna 62 , H2. Frequency of mention of location is neg - tively related to: . education of male household head . urban experience of male household head . education of female household head . urban experience of female household head . total household income (DOOOU'DD None of the relationships achieved an acceptable level of significance. These findings fail to provide support for H2. Housegguality The following hypothesis of the relationship between evaluation of house quality and selected characteristics of household members was tested: H3. Frequency of mention of house quality is pgsitively related to: . education of male household head . urban experience of male household head education of female household head . urban experience of female household head total household income (DD-00“” Only H3c achieved an acceptable level of significance and the relationship appeared to be in the hypothesized direction. However, the magnitude of the relationship is very slight. These findings fail to provide strong support for H3. Compound Quality The following hypothesis of the relationship between evaluation of compound quality and selected characteristics of household members was tested: 33. Frequency of mention of compound quality is positively related to: education of male household head urban experience of male household head . education of female household head urban experience of female household head . total household income (DD-DUO) 63 Only relationship H4a achieved an acceptable level of signi- ficance and the relationship appeared to be in the hypo- thesized direction. However, the magnitude of the relation- ship is very slight. Such findings do not provide strong support for H4. Services The following hypothesis of the relationship between evaluation of services and selected characteristics of house- hold members was tested: Hé, Frequency of mention of services is Epsi- tively related to: education of male household head . urban experience of male household head education of female household head . urban experience of female household head total household income (DO-O o‘m All but H5c achieved acceptable levels of significance and, of these, all but H5b were in the hypothesized direction. The magnitude of the correlation coefficients was slight in all cases. On the basis of these findings one might conclude that there is some support for hypothesized relationships between education of male head of household and total house- hold income and evaluations of services. Interpersonal Relations The following hypothesis of the relationship between evaluation of interpersonal relations and selected character- istics of household members was tested: 56. Frequency of mention of interpersonal rela- tions is not related to: 64 education of male household head urban experience of male household head education of female household head urban experience of female household head total household income (DCLOU‘OJ None of these tested relationships achieved an acceptable level of significance even though the correlation coefficients were very low. These findings fail to provide support for H6. Other Analyses The original intent of this study was to consider only the evaluations of six dimensions of residential situation consis- ting of the combined responses to questions about likes and dislikes made by both male and female heads of household. However, it was decided to examine likes and dislikes as well as male head and female head responses separately in order to gain a better understanding of the combined responses. When this examination revealed differences in the distribution of likes and dislikes and in male and female head response pat- terns, additional indices were constructed which reflected these differences. Correlation analyses were undertaken to determine possible relationships between these new indices and education and urban experience of male and female household heads and total household income. The results of these anal- yses are reported in Tables G, H, and I of the Appendix. In gen- eral , the results of these correlation analyses were not very dif- ferent from those obtained in the initial analyses. Few rela- tionships were found to be significant and of those that were, the magnitude of the correlation coefficients was very slight . CHAPTER V SUMMARY, DISCUSSION, AND IMPLICATIONS Summary of the Study Given the general problem of the rapid growth of high density, low-cost housing in African, Asian, and Latin Ameri- can cities, this study was undertaken to identify dimensions of residential situation which are important to members of squatter households in Lusaka, Zambia, and then to determine relationships between squatters' evaluations in these dimen- sions and selected characteristics of these individuals. Such information could contribute to an understanding of how and why decisions regarding housing are made and implemented at the level of the individual household, and thus provide in- sights for the formulation of relevant housing policies and programs. Although set in the framework of a model of housing adjustment proposed by Morris and Winter (1975), this study did not attempt a full application of the model. The study involved evaluations of aspects of residential situation by male and female heads of squatter households and selected characteristics of those individuals. Because housing adjust- ment provided a conceptual basis for the study, findings could suggest possible directions for further research 65 66 in this subject area. Secondary analysis of data collected in a 1973 study of housing in three Zambian cities was undertaken to serve the purposes of the study. Male and female heads of squatter households in Lusaka, the capital city, comprised the study sample. Measures of education for both male and female heads of household, years of urban residence of male and female heads of household, and total household income were the char- acteristics of those individuals selected for analysis. Their evaluations of their residential situation were indi- cated by indices of six dimensions of residential situation. The evaluation indices were created on the basis of the frequency of responses in each'dimension. The six dimensions were economic conditions, location, quality of house, quality of compound, services, and inter- personal relations. These dimensions were derived by grouping male and female household heads' responses to two open-ended questions: "What do you like best about this compound?" and "What do you dislike the most about this compound?" A house- hold's score on each of the indices was calculated by counting the total number of responses given by male and female heads of household in the dimension it measured. Thus the higher the score on the index, the more important the dimension it measured was assumed to be. It was generally hypothesized that frequency of responses in dimensions of economic conditions of residence and location would be negatively related to the characteristics of 67 household heads selected for analysis. It was also hypo- thesized that frequency of responses in dimensions of house quality, compound quality, and services would be positively related to these characteristics. It was hypothesized that there would be no relationship between frequency of response in the dimension of interpersonal relations and these char- acteristics. Pearson Product Moment correlation coefficients and one- tailed tests of statistical significance were calculated for these relationships. The correlation analyses provided little or no support for these hypothesized relationships. The few significant relationships, although very weak ones, that were found involved relationships of male head education to fre- quency of mention of economic conditions, compound quality, and services; of female head education to frequency of mention of house quality; of urban experience of male and female heads to frequency of mention of services; and of household income to frequency of mention of economic conditions of resi- dence and services. No significant relationships were found between characteristics of household heads and frequency of mention in the dimensions of location and interpersonal rela- tions. These findings seem to imply that other factors than education, urban experience, and household income may have some relationship to evaluations of residential situation made by squatters in Zambia. In order to understand better the composition of the in- dices, the frequency and distribution of responses comprising 68 them.were examined. It was observed that likes of resi- dential situation were strongly concentrated in the dimen- sionscflfeconomic conditions of residence, location, and inter- personal relations. Dislikes were concentrated in the dimen- sions of house quality, compound quality, and services. This seems to indicate that residential needs of squatters are being satisfied in some areas (likes of residential situation) and not in others (dislikes of residential situation). In general, the response patterns of male and female heads of household were very similar. Of the 88 different response categories identified for the study population, several individual responses appeared with much greater frequency than all others. These included aspects of tenure, rent cost,generalinterpersonal relations, toilets, roads and transportation, water, schools, electri- city, and clinics. One other response, "nothing," which had not been incorporated into any of the indices and therefore was not involved in the correlation analyses, was provided frequently by respondents. In fact it was the most frequent individual response given by female heads of household to the question about likes and the second most frequently given response by male heads of household to this question. Limitations of the Study This study was based on secondary analysis of data gathered in a study in which the author was not involved. Consideration of the design and implementation procedures 69 used in the original study led to the assumption that avail- able data were reliable and valid. However, some problems were encountered in the conceptualization of the study pro- blem and in the selection of specific variables for this analysis. Ideally, empirical studies are designed such that data collected represent or measure concepts or variables encom- passed by a particular conceptual or theoretical framework. In the design process, attention can be given to particular relationships to be tested as well as to relevant methods of analysis to be employed. Careful design may thus facilitate analysis and interpretation of findings. In this secondary analysis, a model of housing adjust- ment was selected as a basis from which to select variables and relatonships for analysis. Although it was felt that this model served as a useful point of departure, it was also found that the fit between the model and the available data was not as close as would have been desirable. This problemwas somewhat complicated by the fact that few applications of the model seem to have been undertaken outside of the conti- nental United States. Consequently, the study lacks strong theoretical or empirically tested underpinnings for hypo- thesized relationships or general study design. Of perhaps more significance as a problem to the study was the author's unfamiliarity with study data. There was no assurance that the data analyzed accurately measured the concepts that they were used to measure. Because the 70 distribution of cases on the selected variables was not known, the methods of analysis that were undertaken turned out to be problematic. Responses to open-ended questions were used as the basis for the construction of indices. In general, open-ended questions must be approached with caution; although one assumes Open—ended responses to be accurate and complete, there is no guarantee that they are. At the point where the author was constructing the evaluation indices, it was not known that each index would consist of either primarily likes or dislikes rather than a "neutral" combination of the two. This finding about the nature of the indices made the inter- pretation of analyses involving the indices somewhat more complicated than was originally intended. Despite these methodological difficulties, the use of open-ended questions as the basis for this study provided detailed information about the evaluations of residential situation made by heads of squatter households. A limited amount of research has been undertaken previously in this field of study in an African context. The use of open- ended questions may therefore be justified in terms of the new insights that they generate. Such information could be incorporated into later studies of squatters, their evaluations of their residential situation, and their behaviors and decisions related to housing. A household's score on an index of a dimension of resi- dential situation was incremented by one each time that a 71 response was given in a particular dimension. If no res- ponse was given by household members in a dimension, the score on the index of that dimension was "0". It was discovered in the process of analysis that two dimensions, inter- personal relations and services, dominated the responses. As a result, the distribution of scores on these two indices showed some variation, but because the indices were developed out of the same pool of responses, most of the cases on the other indices were at the "0" level. The skewed distri- bution of scores on indices may have had some influence on the low correlation coefficients and the lack of significance that were revealed by analysis. Overall, the study permitted useful observations pertaining to dimensions of residen- tial situation but the interpretation of findings of correla- tion analyses was difficult. Discussion and Conclusions Evaluations of Residential Situation To serve the first purpose of this study which was to identify dimensions of residential situation that are impor- tant to members of squatter households, responses to open- ended questions were examined. This examination indicated that respondents' concerns center in a few areas. Aspects of residential situation which were most frequently liked had to do with aspects of interpersonal relations, location, and economic conditions such as rent cost and tenure. Dis- likes focused on house quality, compound quality, and 72 services, particularly water, schools, and clinics. Such findings were not unexpected as previous studies of squatters (Andrews et al., 1973; National Housing Authority, 1972) yielded similar results. Unanticipated, however, was the degree to which respondents seemed to agree in their evaluations. Aspects of interpersonal relations dominated the things liked, and services dominated the things dis- liked in nearly all cases. This distribution of responses seems to indicate that squatter housing is meeting some needs for most residents, i.e. those aspects of residential situation that are liked, and not meeting others, i.e. those things disliked by household members. Some slight differences were found between responses provided by male and female heads of household. For example, male heads appeared to be relatively more concerned with is- sues related to interpersonal relations whereas female heads of household indicated more concern with some of the services to households. Such differences might be attributed in part to gender role differences. For example, in many African households it is women and female children who are responsible for carrying water to the house for use by all household mem- bers. Concern about distribution of water may be a response to this role. A large percentage of the respondents also stated that they liked nothing. There is a need to understand better the meaning of this response. There might be a segment of the population that is completely dissatisfied with its 73 residential situation. Such an interpretation seems plausible given that respondents mentioned more dislikes than likes about their residential situation. If this is the case, it is important to determine the factors which differentiate these people from others. There is also the possibility that some of the respondents, because of the dynamics of the interview situation or other factors, were not inclined to respond and gave "nothing" as an easy answer. Relationships between Evaluations and Characteristics of Household Members In most cases where relationshps between education and urban experience of male and female household heads and total household income and evaluations of residential situation were tested, no significant relationships were found. Of those that were found to be statisically significant, the coefficients were very low. Most of the significant rela- tionships seemed to exist within the index of services, with education of male head of household and total household income showing the strongest relationships. Such findings seem to indicate that very few relation- ships exist between household heads' education, urban experi- ence, and income and their evaluations of residential situa- tion. These findings are not consistent with literature which suggests that there will be differences in residential priorities indicated in people's evaluations and that dif- ferent personal characteristics, particularly those related to social class and urbanization, may account for variations. 74 There are several possible explanations for this incon- sistency. It has been suggested previously that the lack of statis- tically significant relationships or low correlation coeffi- cients may be attributed, at least in part, to a methodo- logical problem, i.e. the skewed distribution of scores on the indices. It is also possible that higher scores on in- dices may indicate a greater capacity on the part of the res- pondent to verbalize opinions rather than the relative importance of particular dimensions of residential situation. If this were true, one might find, for example, a strong positive relationship between education and index scores. There is the possibility that those characteristics which were selected for analysis were not those most relevant to the study. Urban experience might have been measured dif- ferently including, for example, only years of urban resi- dence at a certain stage of one's life cycle or encompassing more specific urban housing experiences. Some measure of socioeconomic status which combines effects of education, occupation, and income, and which may be related to different aspirations or expectations, might be more relevant to this kind of analysis than simple measures of income or education. It might be useful to study the effects of gender on evalua- tions. 75 Additional Analyses Previous analyses with the same set of data on which this study is based revealed that some of the characteristics selected do not have a completely linear effect on housing behaviors. Effects of equal increments of some are different at different levels. Transformations of these variables can be made in order to linearize relationships. Earlier analysis of data from this population indicated that useful transfor- mations were found to be square of education and log of in- come.* These transformations were applied to the vari— ables used in this study, and correlation coefficients were calculated of the transformed variables with the indices. The coefficients obtained were compared with those obtained earlier. Tables J, K, L, and M in the Appendix (correspon- ding to Tables 10, G, H, and I respectively) summarize the results. Relationships did not change appreciably. In addition to this correlation analysis, some regres- sion analyses were performed using both non-transformed and transformed variables to determine the extenttxawhich know- ledge of characteristics of household members might permit prediction of index scores. As in the case of the correla- tion analysis, a sub-program of the Statistical Package for the Social Sciences (SPSS) (Nie et a1, 1975) was used for these regression analyses. The multiple regression tech- nique used involved forward (stepwise) inclusion in which the order of inclusion is determined by the respective * Personal communication from David S. Wiley. 76 contribution of each variable to explained variance. Inde- pendent variables are entered only if they meet certain statistical criteria. The SPSS program for forward stepwise inclusion will not enter independent variables that do not achieve satisfactory values of F and tolerance. Few of the regressions were statistically significant. Results are summarized in Tables N, O, P, Q, R, and S. In general, it appears that the predictive capacity of education and urban experience of heads of household and total house- hold income on index scores is minimal. However, it is in- teresting to note that those characteristics studied may serve as better predictors of residential evaluations by male heads of household than by female heads of household. Concerning the dimension of house quality, however, these variables do not significantly contribute to predictions of evaluations for male heads of household, but may have some predictive relations to evaluations by female heads of household. Implications Further Research Although some methodological concerns were raised in this study, there appears to be little support for the general hypothesis that there is a relationship between educa- tion and urban experience of heads of squatter households and total income and household heads' evaluations of residential situation. In future tests of this hypothesis, a better 77 measure of evaluations of residential situation is needed. This study has yielded some information which can be incorporated into the development of such a measure. It has been revealed that aspects of tenure, rent cost, interpersonal relations, toilets, roads and transportation, water, schools, electricity, and clinics appear most frequently in residential evaluations made by heads of squatter households. Measures should deal with these speci- fic concerns in such a way as to determine how household mem- bers rank their importance. It might be possible, for example, to design a study which includes scales on which the perceived seriousness of problem areas is assessed. The survey instrument used to collect the data used in this study did contain several questions which had res— pondents assess the seriousness of such problem areas as water, toilets, refuse, transportation, schooling, and health care. For example: 61. Do you consider getting water in this com- pound is a serious problem, somewhat of a problem, or not a problem at all? These questions were not included inthis study, however, as they did not seem to fit as well with the purposes of the study as did those open-ended questions that were used. Nevertheless, responses to such questions could be correlated with characteristics of household members while controlling for effects of actual residential conditions. Certainly the relationships between characteristics of household members and their evaluations of residential 78 situation reflecting needs is but a part of a larger pro- cess of adjusting housing to housing needs. Further atten- tion needs to be given to other relationships among house- hold characteristics (other than education, urban experience, and income), level of satisfaction with housing conditions, aspirations and plans of household members, factors which constrain or facilitate implementation of plans (e.g. alter- nativesikn:action, resources, information), and actual deci- sions that are made. For example, the differences found in this study between male and female household heads' evaluations of residential situation seem to indicate a need to understand more fully gender roles and how priorities, aspirations, and goals related to housing vary for different household members depen- ding on these roles. In order to understand preferences, decisions, and behaviors related to housing at the level of the household unit; it would also be useful to investigate relationships among roles, aspirations, and goals of different household members and actual decisions and behaviors. Pro- grams might be tailored to those different individuals who have specific influences on behaviors of all household members. On a more conceptual level, more thought might be given to linking those characteristics studied and/or other charac- teristics of household members through the concept of house- hold resources and testing relationships between resources and other components of the housing adjustment model. 79 Whereas resources, both economic and non—economic, are generally viewed as means through which.pe0ple can achieve their goals, specific resources may conceivably be linked to people's aspirations, decisions, and actual behaviors con- cerning their residential situation. It would be useful as far as policy-making and program planning are concerned to understand whether certain resources such as specific infor- mation, skills, or money, are associated with residential decisions and behaviors of household members. It might then be possible to anticipate how residential patterns may change with other social changes such as expansion or decline of the labor force or of educational opportunities. An attempt to control for effects of actual living conditions was not made in this study, thus leaving open the possibility that variations in evaluations are caused by the influence of actual conditions in addition to or instead of the influence of other factors such as the selected charac- teristics of household members. It is important that future studies control for such effects. For example, one might investigate variations in evaluations of individuals living in the same conditions by developing some objective measure of housing or compound quality or by grouping segments of the study population according to particular compounds of residence. Compounds in Lusaka vary according to such features as distance to city center, number of schools, and source of water (National Housing Authority, 1972). 80 Most of the studies on housing adjustment in the con- tinental United States found family life cycle variables such as age of household head and number and age of children to be particularly useful predictors of housing adjustment plans and behaviors. Although there was no specific indica- tion that such variables are relevant to the context of squatting, it could be useful to test relationships between these other characteristics of household members and resi— dential evaluations among a population of squatters. For example, it might be hypothesized that those families with school age children give higher priority to schools than do families with infant children. It is also important to determine whether relationships between characteristics of household members and evaluations of residential situation exist for only certain of the dimensions of residential situation, whereas other needs are so basic and important as to appear in evaluations of all residents. Clean water supply may be an example of this; it is important to everyone regardless of their income or education. Given this possibility it may be useful to give more attention to those aspects of a model of housing adjustment which could incorporate some assessment of whether basic needs are met, i.e. actual conditions of residential situation. Lloyd (1979a) suggested that housing-related behaviors of squatters could be viewed as a rational decision-making processin which household members act in order to attain 81 certain goals and aspirations. However, their choices may be forced upon them in some cases because of limited re- sources or alternatives. This perspective, when considered along with the findings of this study and other studies of squatter housing, suggests that a model to understand housing behaviors in the context of squatting would involve the following components: 0 actual housing conditions 0 characteristics of household members a household members' level of satisfaction with residential situation 0 plans and goals of household members a social, political, and economic constraints on household members 0 actual decisions and behaviors Policy Implications Much of the recent literature on squatter housing, urban development, and housing policy and standards, indicates that squatter housing has proved to be a viable response to acute housing shortages in areas undergoing rapid urbaniza- tion. Squatters are able to meet some of their needs such as low-cost shelter, access to work, and access to friends and relatives through their own initiatives or self-help. However, there are indications that in order to meet pressing needs, squatters need technical, material, and financial support. Policies that reflect these pressing needs and that provide support are needed. Findings of this study indicate that squatters would like clean water supplies, better roads, more schools, 82 improved toilets, and better house structures. However, such changes need to be implemented in a context which does not je0pardize already existing benefits derived from living in squatter settlements. This study indicates that squatters like their economic and tenure situation and their access to jobs and friends. In light of this finding, the construc- tion of site and service schemes which may be more expensive or farther away from jobs seems not to be the answer. Up- grading of existing housing areas, a policy promoted by Martin (1977) and others, seems to be a more viable alter- native. APPENDIX: ADDITIONAL TABLES 83 .wcwmcsou mo mammoon ooH Hmuou uoc mmE muaooummn .zOHONSUOm Hmwuaopwmou mo moxOHmOm no moxwa co mcowumonv ou am>ww mp3 uncommon moo ummmH um nowaz How mpaonmmsoe mo nomads men OH 2 Hmuon OOH O H O O O OH OH ON NH N OOO O OH NO NO ON OOH OOH OHN NOH OO OO6H>HOO OOH O O N OH OO OHHHOOO OOO N O HN OHH OON Oasoaaoo OOH O O O OH NN OO OOO H H OO HOH OON ONO OHHHOOO OOOOO OOOHHOHO OOH O O N O NN OO NO OOOHOOHOH OOO H H OH OO OOH OOO ONO HOOOOOOONOOOH OOH O H N NO OOO O O HO NON OOHOOOOH OOH O H N OH ON OOOHOHOOOO OOO N N HO NOH NOO OHEOOOOO OOOHH Am 2 O z N z N z N z N z N z e z N z N z N z N z OHOHOH. OH O O N O O O O N H O aowumaam>m mucom 30mm on wmaonmwsom mo unmoumm mam Honfisz mo cowmcoawo NOOOHOaH co wouoom mo cowusnwuumwmv cowumsufim Hmwucomfimmm mo moxOHmOn mam woxwd .< canny 84 .wcwmnsou mo owamoon ooa Hmuou uoc ewe munoouom n .moxwamwm no moan Hmwucomwmmn so macauwosv new on ao>Ow mp3 uncommon woo unmoH um £0w£3 Mom mmaonomaos mo Humans one OH 2 HouOHm OOH H O OH ON OO OH OOO HH NO OOH ONN NOO ONH OOOH>HOO OOH O H O OO cam H m Ow mmw OuOHmav mcboasoo OOH O O ON OO OOO H NO OON OOO OOHHOOO mane: OONHHOOO OCH 0 o w No me mcowumaou OOO H O ON OOO OOO HOOOOHOOHOOOH OOH H O no OOO N OO ONO OOHHOOOH ooa n ma mm mcowuwmaoo OOO ON OOH NNN OHaocoom mmvmd. Am 2 N N z N N N z N z N z N z N z N z N z aHOHOH. O O O N H O cowumoam>m muoom comm um mmaozmmoom mo uGoonm mam Monfisz MO COfimCQSHD AmmowmaH co mouoom mo GOOanOHumNQV coOumouOm HmHqumOmom mo moxwamwn mam maxed .mmmom mam: .m mafime .wawpcsou mo omsmoon 00H Hmuou uoc hoe mucmouomn .moxOHmOp Ho moxOH Hmwucomwmmu no mcowumosv one on Go>Ow mp3 uncommon oco ummma um £0O£3 Mom mfiaOfiomson mo Romano mam OH 2 Hmuoem OOH O O O NH ON NO OH OOO H N NO HNH OON OHO NOH OO6H>HOO OOH H O NO >HHHOOO OOO HH OO OOO Ocsoasoo OOH O O ON ON OOO H OO HON. OOO OOHHOOO OOOOO OOOHOOHO CON 0 m mm co mGOOumHmH .5. NOO H OO OON HOO HOOOOHOOHOOOH .8 OOH O N OO NOO N HN OOO cOHHOooH OOH N OH mm mcowuwmaoo NOO OH OOH OOO OHaocoom OOOHH ON 2 N z N z N z N z N z N z N z N z N z N z N z OHOOON. OH O O N O O O O N H O cowuwnam>m ouoom 30mm um mpHosmmDom mo unwoumm mam Hoefinz mo cowmamafin I l|l ll AmoowmcH Go mouoom mo GowusnOHumOov aowumnufim HmOunmmOmmm mo moxOHmOn pom moxOA .mpmom onaom .0 manna 86 Table D. Years of Education of Male and Female Heads of Household Respondent Years of Education 0 1-2 3-4 5-6 7-8 9-10 11.6r Totala more Male head of household 216 42 222 199 180 63 32 954 Female head of household 484 67 183 103 91 13 2 943 aIncludes those households for which information was provided. Table E. Years of Urban Experience of Male and Female Heads of Household Respondent Years of Urban Experience 0-4 5-9 10-14 15-19 20-24 25 or Totala more Male head of household 110 251. 168 '135 126 162 952 Female head of household 241 277 132 103 91 94 938 aIncludes those households for which information was provided. 87 Table F. Total Household Income f Number of Households at Each Level of Income Total (kwacha per montha) 0-20 21-40 41-60 61-80 81-100 over 100 94 213 282 161 90 130 970 al kwacha is valued at approximately US$1.40 88 OO.VO HO OcOOHOHOOHOO NNo.: who. moo. mmo. omo. omo.u oEoo:O a s « oHoemmaon Hmuoe Ooo.u moo. mmo. mHo.u Hoo. mao.n . moon onEom mo moaowuooxo amen: Nqo.n Ono. «no. NNo. mmo. moo.u moon mamaom * a mo cowumoamm Ooo.u owo.: ooo. oHo.u moo.u omo. mom: made mo a moaowumoxm amen: moo. wma. «no. mmo. Hmo. omo.u moo: mama a mo cowumosom NOOxHHV HOONHHOHOO HOOOHHOHOO HOOOHHOHOO NOOOHHO NOOOHHV mGOOumHom huwamso huflamso mcowuwpcou HOGOOHooHoucH mooO>uom pcdooaou mmsom aoOumooq oflaocoom momma maonomsom GOOumDHm>m mo Cowmaoawo mo moOumOuouoouoso Amaowumaouuou uaoaoz uosmoum acmumomv COOumduOm HoOuCopOmom mo moxNHmOo mam moxOA ou mfioocH paosomaom HOHOH pom mpmom maonowsom mo monowuooxm coon: poo cowuoonmm mo mownmcowumaom .0 canoe 89 OO.VO HO HOOOHOHOOHOO mmo.u Omo. moo. NHo. NHo. omo. oEoocO s N maoeowsos Hmuoa mao. Owo.u HHo.u omo.u Hmo.- Hmo. moms mama mo a oucowumoxo coon: ouo.u oma. mno. mmo. wmo. coo.n poo: mama « a a s a mo cowumonom HOOOHHO HOOOHHOHOO NOOOHHOOOO HOOOHHOHOO HOOOHHO NOOOHHO mcowumaom muwamso muOHmDO mcowuwmaoo HmaoOHmououcH wmoO>Hmm panooEou mmnom cowumoog oOEonoom cowumsHm>m mo CowmaoEOn ammo: paommmsom mo mowumwuouomumeo Amaowumaouuoo uaoaoz nosmoum coaumomv Cowumnuwm Hmwucooflmom mo moxOHmOo mam moxOA .mmmom can: ou oaooGH pace nomoom HOHOH mom poem on2 mo ooaoOHonm amen: mom nowumoomm mo newnmcowumaom .m maan 90 mo.vo um unmoOMchOm s oao. HNo. Hoo.u moo. Ooo. Nmo. mEoocO mHosmmnoc Hmuoe omo. oeo. «so. omo. moo. oao. poo: onEom mo moamwuooxo some: Hmo.u smoo. mmo. ammo. coo. mmo. moon onEom mo cowumosmm 39:6 HOOOHHOHOO NOOOHHOHOO HOOOHHOHOO 39:5 HOOOHHHO mcowumaom muwamso muOHmno mcowuwpaou HmaomumonoucH mouw>uom mobooaoo omoom cowumooq oOEocoom memo: mHonomaom COOHOSHO>M mo cowmcoEOo mo mowumwuouomumso AmmoOumHohuoo ucoaoz unboowm acmumomv aoHumDuOm HOOHComNmom mo moxOHmOo pom moxwa .mpmom onEmm ou osoocH paos tempo: HmuoH new one: onEom mo oocownooxm Conn: mam coOumoSpm mo moanmcowumaom .H oHan 91 OO.va HO HOOOHOHOOHO: ‘O‘ wmo.u HNH. Hmo. omo. Ooo. moo. oEoocO pHosomno: s * Hmuou mo woq mmo.n mmo. woo. woo. mmo. «Ho.n moon madame no a « cowumosmo mouosvm OHo. oma. «mo. «no. «No. one.: omen mama mo * # GOOumonmo moumnom Ammquv Amoxwamwov AmoxOHmOov AwoxOHmOov Amoxwuv Amoxflqv mcowumaom huwamso muwamso mcowuwmaoo HmcoOumoHoucH mooO>uom mcfloofioo omsom GOOHOUOA owaoaoom COHUQSHmN/m MO COH wCQEHQ mmmom mHonomsom mo moOumOuouomumen Amcowumaouuoo ucoEoz uosmoum Conumomv coOumnuOm Hmwuaomwmmm mo mcowumSHm>m on Amoqv oBoocH maonomnom Houoe one owner maonomwom we Amumnomv cowumosmm mo mownmcowumaom .5 manmfi. 92 OO.va HO HOOHHOHOOHOO Omo.: NNH. mmo. Ono. moo. NNo. oSooGO maoeomson O # Hmuou mo woq moo.u mmo. ooa. moo. mmo. Noo.n poms cameom mo « aowumoapo poumsom OHo.u mmH. mNo. Ono. «Ho. mmo.u moo: mama mo * * aoOumonmo monmsum NOOxHHO NOOxHHOHOO NOOOHHOHOO NOOxHHOHOO NOONHHO NOONHHO mcowumaom Oanmso Oanmso maoqumnou HMGOOHooHouaH mooO>Hmm mcsooaoo omnom GOOHOUOA odaocoou GOOum3Hm>m mo cowmcmewo modem maoemmsom mo wowumwuouomumsu chowumaouuoo uamfioz moamonm GOOHmomv coaumsuwm Hmwuampfimom mo moxOHmOo mom moxOA ou Awogv ofiooaH maonomsom Hench mam momma mHonomnom mo Aoumsomv HHoOumoHHmm mo moOemaoOumHom .M maan 93 OO.VO HO HOOHHOHOOHOO omo.u «NH. woo. mmo. Nmo. omo. oaoocH mHoeomson * r e s Hmuou mo wOH omo.u OHH. Ooo. mmo. OOo. moo.u one: one mo O a s * coHumopmo ponmnvm NOOOHHV NOOOHHOHOO NOONHHOHOO NOOOHHOHOO NOOxHHO NOOOHHO wCOHumHom huHHmno OHHHOSO wcowuHmcoo HmaomuoououcH mmoH>uom mcbooaou omsom GOHuOUOH ONanoom modem on2 GOHuwaHo>m mo concoEHo mo moHumHHouomHmao AacOHumHounou ucoaoz.uoaooum acmumomv aOHumsuHm HmHucopHmom mo moxHHmHo mam moxHH .mmmom on2 cu Aonv oEoocH mHonomsom HmuoH mam poem on2 mo Awumnvmv GOHumoamm mo moHanOHumHom .H oHan 94 OO.VO HO HOOHHHHOOHO o.» woo.n ammo. HNo.u *ooo. Hmo. Ooo. ofioocH mHosomsoe Hmuou mo moH Nqo.u Hmo. moo. Nmo. HHo. mmo. memo: mHmaom mo e a coHumono moumnvm NOOOHHO NOOOHHOHOO NOOOHHOHOO HOOOHHOHOO NOOOHHO NOOOHHO wGOHumHmm OHHHOSO OHHHODO mCOHuHocou HmcoOHoonucH mooH>uom mcboofioo ammo: GOHHOUOH oHeocoom GOHumnHm>m mo COHmamfiHo modem onaom mo OOHumHnouomumnu oaoocH mHonomsom HmuoH mam mmom onEom mo Aon5omv COHumoomm mo moHanOHumHom AmaOHuoHoHHoo uaoaoz uosooum GOOHmomv coHumDuHm HmHucomHmmm mo moxHHmHo mam moxHH .momom onawm ou AwOHv .z OHHOO 95 .oo>mH£oo uoa mums GOHumsvm Gonmoume m oucH oHanHm> m mo Ounce pom Emuwouo mmmm mew an moHHsvoH mocmHoHou mam h mo moDHm> Ououommeummn .mHHouHuo HOOHumHumum aHmuuoo Home Omen «H OHco monouoo mum moHanum> panacea. noocH .oocmHum> moaHmHQXo ou oHanHm> some mo COHunnHuuaoo o>Huomomou use an mocHEHmump OH COHmsHocH mo Homuo onuHaHEtOcH mmHomHum> ucoocooomaH mo connHocH mmHaooum oumzuomm oqo. Odo. NNo.: mmo.u omo.: Ono. mmo. maOHumHou Hmc0mummuoucH ooo. Heo. OOH. OHH. oNo.u mOH.n OHH. mooH>How NOO. NOO. ONO. OOO. ONO. OOO.- NOO. OHHHOOO mcbooaoo omm. moo. owo. moo.n mno. Noo. omo.u HOuHHmHHHH mmsom mom. moo. oHo. oHo. oeo. oHo.: ,wHo. nOHHOUOH HHo. OHo. _moH. HNo.n n wmo. ewo.r mGOHuHocoo oHEocoom mEooHHH mmom 395m mo omen onme poem on2 «.0 poem onz OHonOsom OHEHHOea SOHO H6 BHHOBOO OOSHHOea OOOHO H6 SHHOQOH mocmo muom COHumsHm>m uHchme mo moHHMHHmN/ uaomcooomaH mo COHmcoEHo A mHthmG< cOHmmmHmwmv SOHumDuHm HmHuaoonom mo mcoHumsHm>m co oaoocH oHosomDom mWHOH mam momma oHoaomaom mHmEmm mom on2 mo mocmHuooxm snow: was noHHMUSpm mo muommmm .z oHan 96 .mm>oH£om uoa oHoS SOHumnuo aOHmmouwmh m oucH oHanHm> m mo mound HoH Emnwouo mmmm use an moHHsva mocmuoHou mam m mo moaHm> auouommmHummn .mHHmuHHo HOUHumHumum GHmunou Home Omen «H hHao monouco who mmHanum> ucomcmo ummcH .ooamHHm> moaHmHaxo ou oHanHm> Home mo SOHunanuaoo o>HuooomoH men an mmcHEHouom OH consHocH mo Hoouo oeu 50H£3 5H moHanum> ucmmaooomcH mo GOHODHocH omH3omum pumshomm ONO. OOO. OOO.- , OOO.- NOO.- OOO. OHO. OOOHHOHOH Hoa0mwmonoucH ooo. mmo. HmH. NHH. mmo.u 00H.n 00H. mm0H>Mom omH. HHo. moo.u moo. mmo. mqo.n Hmo. huHHmDU mcboofioo mmm. woo. woo. woo.: doc. moo. «mo.u NSHHHOO:U mmsom wHo. moo. mmo. NHO. omo. NHo.u woo. nowumUOH NwH. woo. «mo. mHo.u @ Hoo. moo.n mGOHuHmcoo oHaocoom 9505 Honor oHSHmNH mo Humor onEmm poem mHmZ .uHo poem on2 pHoeomsom oucmwuooxm comm: mo coHumonmm oucwHHooxm away: mo cOHumuomm GUCQU muwm COHUNSHmN/m uHMchHm WH moHomHHm> uaomnooopaH mo aOHmfioBHo A mehHmam GOHmmonwmv :OHumnuHm HmHuaomHmom mo mcoHumsHm>m co AwOHV oaoocH p OSMOSOm Houoy mam women oHosomnom onaom mom on2 mo ooCoHuooxm amen: mam Ammo: my COHumosmm mo wuoommm .o oHan 97 m mo Ouuco wow Smuwoua mwmm one he poHHzoou moamHoHou mam m mo mosHm> kuouommmHumm .po>oH£om no: ohm3 SOHumsvo cOHmmonoH m oucH mHanHm> n .mHHouHHo HOUHumHumum aHmunoo Home Omen MO ch0 monouco mum moHpmHHm> Homecoo noch .ooamHum> mochHoxo ou oHQMHHm> Sumo mo cOHuoeHHuaoo o>Huooomou mew so pooHEHouop mH consHoaH wo Homuo osu HUHHB 5H moHanHm> uampaoompaH mo :OHmsHoaH omHzooum ounSHomm ooo. omo. ono. Hwo.n moo. AmoxHHmHov mmoH>Hmm oHo. HHo. who. moo. Nmo. AmoxHHmHov OHHHmav moooofioo OOH. OOO. O ONO.- HOO. HOOOHHOHOO muHHmav mmnom omo. OHo. ooo. Nqo. «No.1 AmoxHHv GOHuHmsou UHaocoum M8805 mNHHodflamo .ERHHodSH pdgfixaom ogfifluoofimefipeu MOHHJUQHHHH comm coHumsHm>m mocmonchHm mm moHomHHm> unomcoooch mo COHmcoaHo OOHHOHHHO HOHHOOOHOOO O6 OOxHHOHO OOO OOOHH AmmHthmo< conmmeomv .mmmmz mHmZ no oEooaH oHonomno: Hmuoy mam mpmom oHonmmsom on2 mo mocoHnooxm amen: mam nOHuwoamm mo muoommm .m oHomH 98 .HHoNrmHHHoo Hoe 983 coHumHHHHo coHOOouwou O oucH OHLOHHOO. O Ho Ounce pom Emuwouo mmmm one an moanva moamHOHou mom m mo OoDHO> muouommOHummb .mHuouHHo HOUHuOHuOuO chuuou Home Oozu HH mHao commune mum OOHDOHHO> unopcoo ummsH .oocmHHO> poaHmmeo ou mHnOHHm> Sumo Ho GOHunnHuucoo o>HuoooOmH one an moaHEHouom OH COHODHocH mo Homuo OSu HUHSB :H OOHHOHHO> ucooaooomaH mo GOHOaHocH OOHsoouO mumahomm ooo. wmo. ooo. Nmo.u ono. AOoxHHOHoV Oo0H>Hom «00. «Ho. moo. a sec. AOoxHHOHpV OuHHmnw ocuooaoo mmm. ooo. mHo. mmo.n moo. AOoxHHOHmV huHHODv omsom Nmo. ooo. omo. Hmo. mno.u AOoxHHv aOHuHmaoo UHEocoom mafia: mmflmwdéamo “Hamiflfiz pHoeomsom OOGOHHooxm omen: Ho cOHuoonom Ouom GoHuOsHm>m moamonchHm m OOHHOHHO> unaccomomcH mo GOHOcQEHo A OHOAHOG< cOHOOmeomv GOHumnuHm HmHucomHOom Ho OoxHHOHo pom OoxHH .Ommom OHO: mo Aonv oEoocH mHOHOanm Hence one Ommom mHOSOOsom OHO: Ho monOHHmoxm coon: mam Amnmsvmv coHumoapm Ho Ouoommm .0 OHQOH 99 .po>OH£oO uo: whoa aOHumavo mOHOOOHwou w oucH OHQOHHO> O Ho Opuoo HOH Emuwouo mmmm one On omuHDvou oocOHmHou mam m we OmnHO> OuouommOHummn .OHHouHuo HOoHuOHumuO chuHmo Home menu HH tho monouao who OOHHOHHO> ucmmaoo nomcH .ooomHHO> moonHoxo ou OHanHO> some mo GOHunnHHucoo o>HuoooOOH onu he mocHeumuom OH coHOoHocH mo Hmpuo one 56H£3 aH OOHQOHHO> HammaooomcH Ho aOHOaHocH OOHaoouO mumsuomm Hmm. ooo. «Ho. Hmo. ooo. AOoxHHOHmv Om0H>Hmm mmm. moo. moo.u oqo. Ono. AOmxHHOHpv OHHHOSU manooeou Hmo. moo. mmo. omo. who. AOoxHHOHmv OHHHmdv mOnom omN. moo. mmo. mmo. Hmo. AOOMHHV cOHuHmaoo oHBoaoom maxen. omflmmHmaaHmo Hammmxaewm oHoeomsom wocwHHooxm comma Ho_:0Huooopm muom COHumsHO>m mocmonHame m OOHHOHHO> unopaooopaH Ho GOHOGQEHQ A OHOhHmo< :OHOOouwomv :OHumouHm HmHucOoHOom Ho OoxHHOHo mam OoxHH .Ommom OHOEOM no oEooGH oHo£OOnom Hmuoy mam Ommom oHonomsom OHmaom mo mocOHHmoxm amen: mom GOHumoSmm mo Ouoommm .m OHQOH 100 .mo>OHnom uo: ouo3 GOHumDvo coHOOmeou m oucH OHHOHHO> a mo OHHCO How Emuwono mmmm one he mOHHsvou oonHOHou mam m Ho OoaHm> AuouomHOHuOmn .OHHmuHHo HOOHHOHumuO aHmuuoo Home Aosu HH tho commune mum OOHQOHHO> Homecoo noch .mocmHHO> pocHOmeo on oHnOHHm> Sumo Ho COHuanHuuaoo o>HuoooOou onu he moCHEHoumo OH aoHOsHocH Ho Homuo onu nngs CH OOHanwm> HomecooopaH Ho cOHOaHoaH OOHzoouO phmsuomm ONN. OOO. OOO. OOO. NOO. HOOOHHOHOO OOOH>Hom ONH. NOO. OOO.- OOO. NNO. NOOOHHOHOO OuHHmno mcdooEou NNO. HHO. NOO. OOO. NNO. NOOOHHOHOO OuHHmsv mOsom NOO. NOO. O NNO. OOO. NOOOHHV QOHuHmaoo UHEocoom mEooHHH poem OHmEom .Ho Humor OHOEom pHoeomnom mocwHuooxm_cmmHD. Ho COHuOosom mumm GOHumnHm>m oocmoHHHame m OOHQOHHO> Homecomoch mo GOHOCQEHQ A OHOOHOc< GOHOOOHwomV GoHumDuHm HOHucmmHOom mo OoxHHOHo paw OoxHH .Ommom OHOEom mo AwOHv oEoocH oHonomoom Hmuoa mam Oowom oHoeomaom OHmEom mo mocOHHooxm amen: one Aouwnomo GOHumoamm Ho Ouoommm .m oHnOH BIBLIOGRAPHY BIBLIOGRAPHY Andrews, P., Christie, M., & Martin, R. 'Squatter' manifesto. Ektistics, 1972, 34, 108-113. Andrews, P., Christie, M., & Martin, R. Squatters and the evolution of a life style. Architectural Design, 1973, 43, 16-25. Bates, R. H. Patterns of uneven development: Causes and consequences in Zambia. Denver: University of Denver, 1974. (Monograph series in world affairs, 1973-1974, ll) Brown, L. & Moore, E. The intraurban migration process: A perspective. In L. S. Bourne (Ed.), The internal structure of the city. New York: Oxford University Press, 1971. Clark, W. A. V. & Cadwallader, M. Locational stress and residential mobility. Environment and Behavior,. 1973, 5, 29-41. Dwyer, D. J. People and housing in third world cities: Per- spectives on the problem of spontaneous settlement. London: Longman, 1975. Gross, I. H., Crandall, E. W., & Knoll, M. M. Mana ement for modern families (4th ed.). Englewood Cliffs: Prentice-Hall, Inc., 1980. Gugler, J. On the theory of rural-urban migration: The case of subsaharan Africa. In J. A. Jackson (Ed.), Migration. Cambridge: Cambridge University Press, 1969. Gugler, J. What is urbanization? An operational approach. African Urban Notes, 1969, 4, 9-25. Herbert, J. D. Urban development in the third world: Policy guidelines. New York: Praeger Publishers, 1979. Hutton, J. (Ed.). Urban challenge in East Africa. Nairobi: East African Publishing House, 1972. Jackman, M. E. Recent population movements in Zambia: Some aspects of the 1969 census. Lusaka: University of 101 102 Zambia, 1973. (Institute for African Studies, Zambian papers, No. 8) Jackson, J. A. (Ed.). Migration. Cambridge: Cambridge University Press, 1969. Lloyd, P. C. Shantytowns in developing nations. Current Anthropology, 1979, 29, 114-117. (a) Lloyd, P. C. Slums of hope? Shantytowns of the third world. New York: Penguin Books,41979. (b) Maasdorp, G. & Humphreys, A. S. B. From shantytown to town- ship. Cape Town: Juta Company Ltd}, 1975. MabOgunje, A. L., Hardoy, J. E., & Misra, R. P. Shelter ' provision in developing countries: The influence of standards and criteria. Chichester: John Wiley and Sons, 1978. Martin, R. Housing Options: Lusaka, Zambia. Ekistics, 1977, 44, 88-95. Martin, R. The ecology of a squatter settlement. Archi- tectural Review, 1969, 145, 213-214. Morris, E. W. & Winter, M. A theory of family housing adjustment. Journal of Marriage and the Family, 1975, 37, 79-88. National Housing Authority. Squatter settlements in Lusaka. Lusaka: National Housing Authority, 1972. Newman, S. J. & Duncan, G. J. Residential problems, dis— satisfaction, and mobility. Journal of the American Planning Association, 1979, 45, 154¥166. Nie, N. H., Hull, D. H., Jenkins, J. G., Steinbrenner, K., & Bent, D. H. SPSS: Statistical package for the social sciences (2nd ed.). New York: McGraw-Hill, 1975. Norwood, H. C. Squatters compared. African Urban Notes, 1975, l (Series B), 119-132. Okraku, I. O. The family life cycle and residential mobility in Puerto Rico. Sociology and Social Research, 1971, 55, 324-340. Payne, G. K. Urban housing in the third world. Boston: Routledge and Kegan Paul, 1977. 103 Peil, M. African squatter settlements: A comparative study. Urban Studies, 1976, l3, 155-166. Rossi, P.H. Why families move: A study of the social psychology of urban residential mobility. Glencoe: Free Press, 1955. Sabagh, G., van Arsdol, M.D. & Butler, E. W. Some deter- minants of intrametr0politan residential mobility: Conceptual considerations. Social Forces, 1969, 49, 88-98. Simmance, A. J. F. Urbanization in Zambia. New York: The Ford Foundation, 1972} Speare, A. Residential satisfaction as an intervening variable in residential mobility. Demography, 1974, ll, 173-188. Stren, R. E. Housing inequality and housing policy in Zambia. Berkeley: University of California, 1975. Taylor, R. C. Migration and motivation: A study of deter- minants and types. In J. A. Jackson (Ed.), Migration. Cambridge: Cambridge University Press, 1969. Tipple, A. G. Self-help housing policies in a Zambian mining town. Urban Studies, 1976, l3, 167-169. Turner, J. F. C. Housing priorities, settlement patterns, and urban development in modernizing countries. American Institute of Planners Journal, 1968, 34, 354-363. Turner, J. F. C. and Fichter, R. (Eds.). Freedom to build. New York: Macmillan, 1972. van Velsen, J. Urban squatters: Problem or solution. In D. Parkin (Ed.), Town and country in centraliand eastern Africa. London: Oxford University Press, 1975. Williams, F. Reasoning with statistics (2nd ed.). New York: Holt, Rinehart and Winston, 1979. Nell. .‘Iif II 3 1‘ III I ‘1 1- {{‘iu CHIGRN STQTE UNIV. LIBRARIE “SI II IIIIII II III II IIII IIIII IIIIIIIIIIIIIII