Research methodology : design and analysis issues

r--ýz - - - ---c RESE iR I E A P, , LOGY: DES3IGN AND MÅLY33S YSSUES by Michael Ear '-eg: Daz '.el Ka'rl ranr Ed-i:-dc, V-'.z June, 184 Wter Supply a i, riB Dvalope-nt D cartiraent Operations P--liv StafF J'e WorlCId ank Jt - reS r0 2 -i- "0 1 1 olyrp t1S S tl c. C iJ. r,.Or -ichael Bpmberger is a sociologist iLh t-e Water Supply and Urban Development 1epari ent of the World Bank, Dais t%'aafmann is an economist with the East Africa Industrial Development and Finance Division of the World Bank. Eduardo Velez riork. with the Instituto SER de Inv,stigaciun in Bogota and was th-a SER Project .irector for thi3 study. orking Paper No. 3 Research Project No. 672-37 Income Formation and Expenditures of Poor Urban Kouseholds in Cartagena, Col..mbis. ABSTRACT This paper is one of a series of working papers which report the principal findings of the study "Income Formation and Expenditure Patterns Among Poor Urban Households in Cartagena, Colombia," (RP0672-57) The present report describes the methodology used in the study and covers the following topics: background and conceptual framework; operationalization of the research; data analysis and interpretation; and the research methodology in retrospect. The questionnaire and a description of the available data are presented in annexes. An important feature of the research was the use of an interdisciplinary approach in which anthropologists, sociologists and econo- mists worked together on design, implementation and analysis. The study was conducted in the Zona Sur Oriental of Cartagena, Colombia in a low-income, largely squatter area in which the World Bank was developing an integrated upgrading project. The project included land reclamation, upgrading of basic services and credit and technical assistance for housing and community improvement projects. The project was intended to affect, directly or indirectly, most of the estimated 100,000 population of the Zona. The research involved the application of a questionnaire covering demographic and economic characteristics of the households and their inter- household transfers of mon ey and goods, to a representative sample of 507 households; and a series of in-depth anthropological case studies with some of these households. The objectives of the research were to demonstrate the importance of interhousehold transfers as survival and economic development strategies for the urban poor and to examine their operational implications for the design and implementation of future urban development programs. Table of Contents Page No. CHAPTER ONE BACKGROUND AND CONCEPTUAL FRAMEWORK 1. Origins of the Study and Key Research Issues....... ..................1 2. The Conceptual Framework..............5 CHAPTER TWO OPERATIONALIZATION OF THE RESEARCH 1. The Interdisciplinary Approach............12 2. The Definition of Key Concepts...........14 3. The Content and Evaluation of the Questionnaire..........38 4. Ethnographic Case Studies............49 5. Sample Design**...........52 CHAPTER THREE DATA ANALYSIS AND INTERPRETATION 1. Analysis of Network Structure and Network Usage....... ..............55 2. The Role of the Case Studies..............58 CHAPTER FOUR THE RESEARCH METHODOLOGY IN RETROSPECT 1. Strengths and Weaknesses of the Methodology...............61 2. Lessons for Future Research.............62 ANNEXES 1. Summary of the Questionnaire..........64 2. Description of the Available Data Sets and Their Documentation............69 3. Questionnaize Spanish).........72 References Cited in the Text.........86 Reports Available on the Research Project.........89 (Continued) (Table of Contents - Cont.) TABLES Page No. Table 1 - Percentage Share of Transfer Income Out of Total Household Income by Sex of Head of Household.............6 Table 2 - Significance of Transfer Income for Low Income-Households.....................7 Table 3 - Information on Network Structure Contained in the Questionnaire. .......... . .....25 Table 4 - The Definition of Network Structure........28 Table 5 - The Analysis of Network Usage.............31 Table 6 - Use and Limitations of Available Data on Effective and Potential Networks.......................47 TabiOe 7 - Stages in thP Sample Design.......54 CHAPTER ONE: BACKGROUND, CONCEPTUAL FRAMEWORK AND KEY ISSUES Section 1: Origins of the Study and Key Research Issues The origin of the present study was an exploratory analysis by two economists (Kaufmann and Lindauer, 1980) of socio-economic data generated as part of an evaluation of the First El Salvador Sites and Services Project (Bamberger, Gonzalez-Polio and Sae-Hua, 1982). The analysis suggested that interhousehold transfers appeared to be a significant, but largely ignored, part of total household income for certain categories of project families. The findings had immediate policy relevance as they showed that female headed households, one of the poorest sectors of the target population, tended to be among the main beneficiaries of transfers and that the inclusion of these transfers in the estimation of total household income could significantly affect estimations of project affordability. A further unexpected finding was an inverse.relationship between number of children and the amount of increase transfer receipts. The conventional economic approach (Becker, 1974; 1981) to the first finding was to suggest tht the tranfers to female headed households were an informal alimony payment. The initial reaction to the second finding was to try and dismiss it as a statistical "quirk" in the data. When this did not prove satisfactory it was suggested that the finding reflected some type of economy of scale in the production and consumption of basic needs, implying that additional children did not raise significantly the amounts of resources required by the household to raise the extra children. A review of the anthropological and sociological literature suggested the need to broaden considerably the conventional economic approaches to this issue. At least four intriguing themes were revealed by - 2 tha review of this literature (CLicussed in more detail in Kaufmann, 1982; and in the companion working papers by Parris, and Bamberger and Parris). First, anthropological studies of primitive societies (Sahlins, 1968; Polanyi, 1957) argued that the provision of assistance to members of the tribe or group was widely generalized survival strategy. Members of the group had a right to this assistance without any direct requirement of reciprocation. Second, sociologists and social psychologists (Homans, 1958; Blau, 1964; Meeker, 1971; and Emerson, 1976) had developed a number of conceptual models for the analysis of a wide range of social exchanges. In contrast to the previously cited anthropological studies, most of the exchange theories assumed that transfers would be reciprocated. Third, a number of studies of low-income urban communities in the U.S. (Stack, 1974) and in developing countries (Lomnitz, 1977; Nelson, 1979) demonstrated that interhousehold transfers formed part of a "survival strategy" for households at the margin of the modern urban sector. These studies also showed that many and perhaps most of the transfers were of goods and services rather than money. A number of studies suggested that interhousehold transfers were of particular importance to female headed households (Nieves, 1979; Stack, 1974; Buria, 1977; Nelson, 1979), largely because they tended to be excluded from formal sector employment and credit institutions. Finally, an extensive African literature (Moock, 1978; Nelson, 1979; Muller, 1975; Abbott, 1976; Weisner, 1976; Mukras and Oucho, 1983) showed that networks usage should not be viewed simply as a survival strategy for the urban poor but also serves a number of economic development functions, particularly with respect to the transfer of human and economic resources between urban and rural areas. -3- When the original El Salvador data was reinterpreted using some of these concepts (Kaufmann, 1982) it was apparent that women in the informal sector were very effective at establishing dynamic social networks. Childcare was one of the important fi:-:ms of network interaction. A re-estimation of the multivariate regressions indicated that, after controlling for income, education and family size, a significant "female-headedness" effect could be identified. Kaufmann argued, based on some of the anthropological evidence, that "female-headedness" represented one of the socially accepted "eligibility criteria" for being considered eligible to receive transfers. Similarly, the negative impact of number of children on transfer receipts was re-interpreted as an example of people flowing towards resources rather than resources flowing towards people. A better off household will often prefer to take a child into their household rather than send money to its parents so that the number of children in higher income households will tend to increase while the number of poorer households will decline. Simply put, the inadequate understanding of the social dynamics of the populations being studied had led to a mis-specification in the estimated regression; more children did not imply less tranfers, but rather the existence of transfers implied access to a social network, which in turn led in many cases to less children at home, as a result of child-care responsibilities undertaken by the network. Before specifying correctly the regression it was even necessary to revise the definition of a household, so as to include those household members (particularly but not exclusively children) who were temporarily living elsewhere. Economists have long been aware of the dangers of dealing with "black boxes" in their statistical work. Correlations and other statistical measures of association can only be understood through the use of other types of data, hence the need to develop a conceptual framework. However, th- problem is not totally resolved when such a framework is either too stylized and abstract - implying that little explanation will be provided for each specific result - or too static - implying that no explanation will be given of the dynamic processes which have occurred. For the economist the problem is further compounded when some of the explanatory factors which would ideally be part of the analytical framework ar not purely economic in nature. As illustrated by the previous example, our topic of enquiry, interhousebold economic interactions, was clearly plagued by "black-box" problems. It became evident that a purely econometric approach to this issue would yield unsatisfactory and probably misleading results. It was apparent that an interdisciplinary approach, akin to what some labor economists have done when integrating rigorous economic tools with institutional analysis, was needed. In this case the integration would be achieved by the formation of a research team comprised of economists, sociologists and anthropologists. In this way it was hoped to combine in-depth anthropogical studies of individual households and their networks; sociological skills in sample surveys and the analysis of networks; and econometric approaches to the construction and testing of models of interhoushold transfer behaviour. The following are some of the key methodological and research issues to be addressed through this interdisciplinary approach (the policy issues are discussed elsewhere): 1. How can a social network be defined and studied? 2. How can non-monetary tranfers such as goods, services, people and possible information be defined and measured? 3. How can the reasons (motivations) for transfer receipt and giving be identified and studied? 4. Should the basic unit of analysis be the individual, the household., the extended family or some other social grouping? 5. What is a household, and how can we define who are its members? The following section describes the conceptual framework used in the design of the study and in the interpretation of the findings. Chapter Two then explains the way in which thr! above concepts were operationalized and Chapter Three explains the basic analytical procedures which were used. Section 2: The Conceptual Framework Previous work with socioeconomic data from El Salvador and the Philippines had pointed out the importance of inter-household income transfers on the economic welfare of the urban poor. A high proportion of urban family income appeared to be derived from income transfers from non-household members. For Santa Ana, El Salvador, the evidence shows that one-third of all families residing in the informal sector received some transfer income during the 1976 survey month. In Tables I and 2 we perceive the importance of transfer income for the lower socioeconomic strata of the population for the Salvadorean and the Philippine urban setting, respectively. Evidence and concepts emerging from the urban anthropological literature were then utilized to complement the quantitative information from El Salvador and the Philippines and to formulate hypotheses concerning the role of inter-household transfers in the survival strategies of the urban poor 1/ in LDCs. Some accepted notions were challenged. The main contentions were:- a) The conventional wisdom of the sociological literature -- that urban households do not rely significantly on other households (or next of 1/ For a full account of the original conceptual framework, see Kaufmann (1982). Table 1 Percentage Share of Transfer Income Out of Total Household Income by Sex and Head of Household Percent of Households Percentage Share Male Female of Household Income* Headed Headed All 0 76.2 40.2 67.4 1 - 20 13.6 14.2 13.8 21 - 40 7.6 19.1 10.4 41 - 60 0.0 12.2 3.0 61 - 80 0.0 6.5 1.6 81 - 99 0.0 0.0 0.0 100 2.6 7.9 3.9 TOTAL 100.0 100?o 100.0 Percent of All Households 75.4 24.6 100.0 *Share transfers/(transfers + all other income) Source: Daniel Kaufmann "Social Interactions as a Strategy of Survival Among the Urban Poor: A Theory and Evidence," page 12, Doctoral Dissertation, Harvard University, 1982. -7- Table 2 Significance of Transfer Income for Low Income Households Santa Ana, El Salvador, 1976 Earnings % Receive Recipients Distribution a/ Transfers Transfer/Total Income First Decile 58 0.66 Second Decile 48 0.25 Third Decile 21 0.05 Fourth Decile 21 0.04 Fifth Decile 28 0.06 Sixth Decile 26 0.06 a/ Since the data set c.imprised only the bottom 65 percent in the income distributions, no figures are provided for the top four deciles. Source: Daniel Kaufmann "Social Interactions as a Strategy of Survival Among the Urban Poor: A Theory and Evidence," page 12, Doctoral Dissertation, Harvard University, 1982. -8- kin) as they did in the rural areas before the process of urbanization set in -- was not an adequate representation of the reality in the urban areas of LDCs. Our information and the notions emerging from the anthropological literature suggested, instead, that dynamic socio-economic interactions between autonomous households do take place. b) The economists" view that inter-household income transfers, if they existed, were neither large enough nor sufficiently stable to be regarded as' operationally relevant, warranted a critical examination. c) Similarly, the notion that such transfers, when they did take place, were flowing from the well-off to the destitute in the urban slums, and hence responded to altruistic motivations from the donor's standpoint, was considered inadequate given the growing evidence that many transfers were coming from households of low socio-economic status. The conceptual framework for the Cartagena study was based on the following set of alternative speculations. 1. Risk averse households. Among low income households in the informal sector, where incomes may not be sufficient for basic needs fulfillment, the degree of risk aversion to the possibility of falling below a certain minimum income is postulated to be rather high. 2. High variability of incomes. Due to the nature of labor market behavior in the developing cities, labor incomes are subject to significant fluctuations. Since many households lack effective access to alternative formal institutions (social security, formal capital and insurance markets are extremely limited), which could reduce these risks the objective risk for much of this population is high indeed. -9- 3. Incentive to diminish risk through informal mechanisms. The household's risk aversion, their high variability of income, and their lack of access to formal insurance mechanisms provides a strong incentive to search for alternative informal "risk aversion" arrangements. An incentive exists to use social networks to reduce income fluctuations. Those households with a social network with similar socioeconomic conditions and who share mutual trust perceive a potential benefit from insuring with each other against adverse contingencies. When household income falls below a certain minimum, transfers may be received from other households in the network whose income at that point is higher. When the economic situation is reversed, income transfers would flow in the opposite direction. 4. Earmarking. Clearly, tranfer income, by increasing total incomes, will affect the expenditure pattern of the household. Additional disposable income will be available for spending on basic needs in general, and on shelter-related expenditures in particular. It is important to ascertain, however, whether certain expenditures (i.e., basic needs) increase disproportionately when transfer incomes (as opposed to other income sources) flow in. In other words, the possibility of the composition of expenditure being affected by the composition of income ought to be considered. The "earmarking" may arise either because the transfer-donor has implicit or explicit expectation as to how the transfer is to be utilized -- "donor-induced earmarking" -- or because the recipient requests the transfer for a specific expenditure -- "receiver-induced earmarking." 2.1 The Framework in Perspective The contractarian/insurance framework outlined above is neither the only plausible way of conceptualizing inter-household interactions, nor can it - 10 - fully explain all of the complex motivations and rights, which affect network exchanges. However, the simplifying abstraction of the "insurance/ contractarian" framework provides the conceptual underpinning necessary to approach the empirical investigation, and to emphasize certain important aspects of interhousehold interactions largely ignored in the economics and sociological literature. The framework emphasizes that economic interactions between house- holds are significant, and are neither random nor intrinsically unstable, but respond to specific economic motivations within which the insurance/ contractarian motive is important. It is recognized, however, that other motivations such as investments, maintenance of social ties and pure altruism can also play a significant role in explaining social network interactions. 1' 2.2 Some Hypotheses about the operation of social networks From our previous discussion it is hypothesized that empirically we are likely to find: a) a significant amount of inter-household transfers. b) a substantial portion of these transfers respond to a specific economic need. c) complexity in the direction of the flows. Transfers do not not only flow from rich to poor; they also flow between households of similar socioeconomic status, whose incomes are subject to temporary fluctuations. 1/ Some degree of complementarity to our insurance framework would be possible in the motives mentioned above (e.g. the maintenance of social ties motive could be seen as a necessary "insurance-premium payment" within the contractarian framework). - 11 - d) complexity in the structure of motivations. Implicit or explicit "insurance-contractarian" schemes are intrisically more complex than, say, transfers that respond to purely altruistic motivations. e) the degree of monetization in the informal urban economy, although significantly higher than in the rural areas, is far from complete. Hence not all the resources that are transferable are in the form of monetary income; in-kind transfers and services (including rent free accomodation) potentially play an important role as well. - 12 - CHAPTER TWO: OPERATIONALIZATIOI OF THE RESEARCH Section 1: The interdisciplinary approach In order to fully incorporate economic, sociological and anthropological literature and methods into the design and analysis of the study; the core research team included economists, sociologists and anthropologists, and a constant interaction was maint-',ned between the three approaches throughout all stages of the study. The following were some of the main functions assigned to each discipline: Anthropology: Exploratory studies to understand the nature of the problem and to help operationalize the main concepts. More intensive case studies were then prepared to illustrate different types of transfer behaviour. The linkages to the quantitative aspect of the survey helped avoid one of the main weaknesses of many ethnographic studies; namely that it is not possible to evaluate the representativity of the cases selected. Sociology: Operationalization of the research concepts and the design and application of the sample survey. Analysis of the structure and usage of networks. Economics: Contribution to the conceptual framework and the definition of research hypotheses. Concepts were defined so they could be used to test hypotheses derived from the economic literature. Application of an econometric model for the analysis of transfer behaviour. The anthropological study was not just a qualitative cross-check on the quantitative survey; but rather a key element in the formulation and operationalization of the research. The ethographic studies provided more - 13 - information on the variety of exchanges made between households, verified their purposes, pinned down the circumstances under which exchanges were given- and requested, and examined how these changed over time. In so doing, it was hoped a better perspective might be gained on how Cartagena household inves.t- ment behaviour could contribute to the objectives of the Bank project by helping mobilize the resources for housing improvements. The interdisciplinary approach also demonstrated how the two major, and usually conflicting, orientations of sociology can be integrated in a single study. The qualitative, phenomenological, "verstehende" approach associated with Schutz, Weber and the ethnomethodologists (Garfinkel and Cicourel) was represented by the ethnographic studies. The positivistic, behavioural approach associated with Durkheim, Parsons, Nagel, exchange theorists and mainstream economics was reflected in the sample survey and the quantitative analysis. The interdisciplinary team worked closely together in the field, in Bogota and in Washington. An iterative approach to developing the question- naire was utilized. A preliminary questionnaire was applied to a number of households, then the questions and answers were discussed by the team, changes in the questionnaire implemented, and the revised questionnaire tested again. Throughout this process particular attention was paid to issues which were either regarded as conceptually thorny and/or whose field-survey operationalization was not considered straightforward. Examples included a conceptually adequate and operationally relevant definition of a household and of a social network; strategies to obtain information on interhousehold transfers, other sources of income from rent and land, etc. - 14 - On the basis of this daily iterative process a more structured questionnaire was drafted, and eventually subjected to the conventional pretest. Section 2: The definition of key concepts 2.1 Defining the Unit of analysis The conventional basic unit of analysis in microeconomics has been the individual, whose interactions with the formal labor and capital markets lie at the core of the study of labor economics and consumer behavior. However, the household-economics strand of the mainstream literature, as represented by members of the Chicago School and by many authors in the development literature, has advocated the use of the household as the basic unit of analysis. It is argued that the household is the relevant economic decision-making unit and reflects the joint "tastes" of the various individuals in the household. For the purpose of studying the strategies of survival among the urban poor in developing cities, this study considers it appropriate to view the household as the basic unit of analysis. The individual takes few important economic decisions independently from the other individuals in the household. On the other hand the extended family is too large a unit for most types of urban inquiry. It is argued that the economic decisions of interest to the present study (labor force participation, most expenditures, etc.) are taken at the household level, - 15 - However appropriata the household might be as a basic unit of analysis in the present context, it is not the only unit of analysis. A larger unit that encompasses a group of basic households will also be relevant, since it describes the context within which social initeractions among households occur. 2.2 The household and household membership Although most socio-economic surveys use the household as the sampling unit and frequently as the unit of analysis, there is considerable ambiguity in the way a household is defined. Most U.S. literature on the household economy has explicitly or implicitly assumed that the household can be equated with the nuclear family. A similar assumption can be observed in much of the sociological literature which argued that the extended family has lost most of its functions as a consequence of urbanization. The sociological and anthropological literature usually defines the household as the social group of all individuals who share the same residence and who possibly also share the same entrance or access to a residential unit. For the purposes of the present study this definition has a number of problems. Studies in Latin America and Africa have shown that there are constant fluctuations T'th respect to the group of people residing in a particular structure. Nelson's work in Kenya (1978) for example showed continual movement between rural and urban areas, particularly of children. She also showed that female headed households frequently offered short or long-term accomodation to other women who had been abandoned by their husbands or who were seeking to establish themselves in the city. Burja's work in the coastal regions of Kenya (1977) showed the importance which women attach to owning their own house so that this can become their economic and social base. Once the house was owned, the woman was free to invite other women to - 16 - come and stay and to develop a complex division of labor between childrearing and economic activities. In Latin america, Nieves (1979), Lomnitz (1977), and Peattie (1969) have all documented the fact that households are "floating coalitions" rather than permanent groups. Stack (1976) and others have documented a similar phenomenon in the United States. The exploratory ethnographic studies in Cartagena demonstrated a similar phenomenon with a large number of members particularly children moving from one family group to another in response to fluctuating economic conditions. Most families had relatives living in or near the community and people, particularly children, would move backwards and forwards between different family groups. It was found that many families who were too poor to be able to offer financial help, would provide accomodation to friends or relatives. The issue also arose therefore of whether to consider 2 brothers, both with their own wives and children, who lived in the same structure as one household or two. Bender (1967), in his theoretical liscussion of the household recognizes that anthropologists have distinguished analytically between the concepts of "family" and "household" referring to the former as a kinship group and the latter as a residence group that carries out domestic functions (Bohannan 1963, Keesing 1958, Solien de Gonzales 1960). It is possible to distinguish three semi-independent variables in the classification of households, namely kinship, residential proximity and domestic functions. Bender echoes Solien de Gonzales in his characterization of "economic cooperation, and socialization of children" as the principal economic functions of the family (1960, p. 106). Bender then elaborates: "The basic feature of 'domestic activities is not that they are necessarily associated either with families or with co-residential groups, but that they are concerned with the day-to-day necessities of living, including the provision and preparation of food and care - 17 - of children. They differ from various other econo,mic and social functions in that they are well suited to be carried out by small groups in a wide range of societal circumstances. For this reason, they are most frequently carried out, at least in part, by co- residential families. They can be, and often also are, carried out by co-residential units that are not families, and in a few instances they are carried out, in part, by families that are not co-resi- dential units. Within the same society, domestic functions are frequently-carried out by more than one kind of small group, and in highly specialized industrial societies they can be to a great extent efficiently taken over by special-purpose groups. And human ingenuity is such that nearly all domestic functions can be carried out by special-purpose groups; witness: the Israeli Kibbutz as described by Spiro" (p. 499). Kemper (1974 and Lomnitz (1977) follow the conceptualizations of Bender by considering households as kin groups that can either live in the same residential unit without engaging necessarily in domestic sharing, or can share domestic functions without necessarily living in the same residence. 1i A different approach to this issue, also in the anthropological literature, is taken by Bossen (1980). After recognizing the existence of ambiguities in the Bender/Lomnitz concept for the purposes of studying strategies of survival among the urban poor, Bossen proposes a different definition, in which a household is a group of people sharing in a collective endeavor to accumulate durable assets. These holdings are primarily in the form of housing and other relative non-liquid assets. The household develops economic security through an exclusive family membership that implicitly (or legally) defines claims on corporate investment. l' For our purposes, this definition does not resolve the conceptual problem discussed previously. Insofar a6 the "corporate investment unit" 1/ Lomnitz describes the broad spectrum of domestic activities: "preparation and consumption of meals, child care, leisure activities, ritual duties, and a wide range of acts of kindness and mutual cooperation" (1977, p. 99). 2/ Bossen (1980, p. 6). - 18 - encompasses various independent economic decision-makers (in terms of labor market and consumption decisions), we will be unable to explain the interactions among the various members of the corporate unit. Thus, this aefinition of household will not satisfy our search for an appropriate basic unit of analysis. The above discussion shows the complexity of the issue as it seems that a group larger than the residential unit will often have a significant influence on decisions relating to labor market behaviour, expenditures and the distribution of people between different residential units. The possibility therefore exists of using some larger unit, for example the extended family or the "social network" as the unit of analysis for studying labor market behaviour and expenditures. As suggested previously, such a characterization may be adequate for some "traditional" rural settings, where labor force participation, "mutual cooperation" , co-residence, and expenditure are all features imbedded within the same relevant decision-maker. However, in urban areas, and specifically in Cartagena, we believe that most major decisions on the utilization of monetary and human resources are made by a smaller, largely though not exclusively residential, unitwhich we shall define as the household. However, it is clear that the decision making behaviour of the household unit is significantly affected by its access to, and interactions with, a much larger, mainly non-residential, group which we shall define as the social network. Labor market and expenditure behaviour will be modelled using the household as the unit of analysis, and including variables describing social network interactions among the explanatory variables. At a later stage of the analysis the possibility may be explored of using the social network as the unit of analysis to study community level outcomes such as different rates of completion of community infrastructure; - 19 - proportion of temporary hosuehold members; levels of participation in community development activities etc. However, Section 2.3 of this chapter will discuss the methodological problems involved in using a social network as the unit of analysis. For the purposes of this study the household will be taken as the "basic unit of decision-making". This basic unit is comprised of a group of individuals who decide jointly on the allocation of labor supply and the allocation of earnings accruing from these labor market interactions. This "unit", depending on the particular labor market outcomes and specific expenditure needs, might or might not choose to interact with other economic units. But this definition of the household is not sensitive to whether this unit does or does not interact with others (neither is it sensitive to the nature and intensity of the interactions). This definition is therefore exogenous to the process to be explained. This definition ought to be operationally useful. If it can be presumed that the household, as defined above, will generally share the same residence, then identification according to co-residence might be a useful starting point. Since, in some cases, more than one relevant decision-maker will live under the same roof (say, two independent nuclear families), care had to be exercised in identifying the various "households" living together. Our own field experience during the preliminary in-depth interviews process revealed the inadequacies of either relying on a 'household head' count or on that unit that shares the same roof as a way of operationalizing a definition of the household. The first suffers from conceptual and empirical ambiguities - since it is often unclear what is meant by the head of the household and how exactly he/she is to be identified, and it is in many instances difficult to distinguish between various (potential) heads within a house in circumstances - 20 - where there are divisions of labor regarding main household responsibilities, including decision-making. As stated previously, the second operationali- zation suffers from the fact that a house cannot be necessarily associated with one household. A specific experience in the field illustrates in the issues. We interviewed a man living in a poorly built shack. He had arrived in Cartagena not long ago, was not interested in creating or using an interactive social network (he was a -network-interaction avoider'), instead he considered as his most precious asset a large machete which be proudly retrieved from under his bed. In that house he lived with his female companion, with his younger brother and his companion, and with their respective children. However, no meaningful socio-economic interaction seemed to exist between the two brothers. Aside from the large machete the only possessions in the house were two old beds and two kitchen stoves, one for each brother and his family. Patterns of food preparation and consumption appeared, therefore, to be better indicators of the boundaries of each household than the housing structure itself. After repeated field experiences it became clear that sharing food from the same olla (pot) is an important factor in the definition of the household group. For poor families, eating is a basic survival strategy and great importance is placed on who pools resources. Households are clear in their own mind about who does and does not use the olla and this seemed to be a good indicator of membership in the primary economic resource pooling group. The definition of household used in the study sought to take into account co-residentiality, resource pooling, joint decision making and acceptance of a single household head. Taking into account these factors the household was defined as that group of people who normally: - 21 - (a) Reside under the same roof. (b) Pool resources (defined as sharing the same "olla") (c) Recognize the same household head The issue of how to classify members who are temporarily absent (or present) is discussed below. Where families live under the same roof but have independent eating arrangements, they would be considered as two separate households (as in the example mentioned earlier of the two brothers). Families understand that once food is shared on a regular basis the second family has become integrated into the first household and has lost its autonomy. The above definition was use as a guideline, and in case of any ambiguity the supervisor was asked to make the decision. The household head is an important concept for at least 3 reasons. First, it is a reference point for studying kinship relations. Second, it is often assumed that the household head is the best person to provide information on many aspects of family life. Third, previous studies (Kaufmann, 1982; Buvonic, 1978) have shown that the behaviour and economic conditions of male and female headed households differ significantly. Kaufmann, for example, showed that even after controlling for income and other household attributes, female headed households are more likely to receive transfer income as "femaleness" is considered as an important eligibility criteria for receiving transfers. A number of different definitions of household head have been proposed, among which: the person defined as head by the household; the main economic contributor and the person with the highest status (based on age, traditional values etc). None of these definitions is completely adequate in all situations. The policy adopted in the present study was the following: - 22 - The household head is normally defined as the person accepted as head by the household members. This definition should however be compared with the following indicators: (a) The main economic contributor (b) The person who is considered to have the maximum authority and to be the principal decision maker (c) The person considered to have the highest status (d) Permanent residence in the house. Where inconsistencies are found, a final decision is made by the supervisor. In practice the application of this concept did not prove to be very problematic. The definition of who is a household member is complicated by the fact that a number of people who would normally be considered as part of the household (children in particular) may be temporarily absent. At the same time there may be one or more people temporarily residing in the structure. The issue is important as previous studies have shown that a considered as part of the household (Murillo and Rothlisberger, 1981). if these temporarily absent members.are considered as part of the household, then the estimated volume of interhousehold transfer flows will decline significantly and the estimated household income will increase. For this reason the following concepts were used: Household members temporarily absent refers to "people who are considered to be permanent members of the household but who are temporarily absent for some special reason. These persons have not established their own household and they plan to return to the respondent household. - 23 - No specific time limnit was placed on when they would return and the respondents definition of temporary was accepted. The survey includes information on how long the person has been absent so that some arbitary cut- off point could be used in subsequent analysis. Temporary household members refers to people living in the household who are defined by the household head as being temporary. Again the definition is that given by the household head. 2.3 A Conceptual Framework for the Study of Social Networks / 2.3.1 Introduction Social Network Research is one of the areas of the social sciences in which some of the most sophisticated analytical techniques have been developed but where, with the exception of small group research, these techniques have had the least impact on the ways in which empirical research is conducted. There are several reasons for this separation between theory and practice. Firstly, even with the use of computers, many of the existing analytical techniques can only be applied with relatively small data sets. Unfortunately, individuals and households frequently have network contacts with very large numbers of people so that the size of the potential matrix or graph can rapidly become unmanageable. For example, Kapferer (1973) found that some of the married couples he studied in East Africa had contacts of one form or another with up to 1,000 people. Similarly, Lee (1969) estimated that the average American family "knows" about 1,000 people. Secondly, many of the analytical techniques require that information be obtained on the interactions between all members of the network and the researcher must usually be content with studying a "personal network" 1/ Taken From Working Paper No. 1. - 24 - consisting only of the people with whom the subject of the study is directly, or indirectly, in contact (i.e., no information is obtained on the interactions among the people who form part of the subject's network). Thirdly, most networks are "multiplex" in that a number of different "commodities" (money, goods, etc.) flow through the network. This makes the analysis much more complicated than in the case of controlled laboratory experiments where only one type of exchange is allowed to take place. Due to these factors, a number of recent reviews (Barnes 1979; Mitchell, 1979) have been quite pessimistic about the advances which have been made in the development of empirically testable network theories and methods of analysis. The researcher tends to find himself faced with an impressive array of mathematical techniques, but with few testable hypotheses. The purpose of the present discussion is to develop a conceptual framework which can be used to describe the structure of social networks and the patterns of network usage in the low-income areas of Cartagena covered by the present research project, and to understand the functions of inter- household transfers as survival and development strategies for low-income households. At the same time the approach takes into account the limitations of the available data. 2.3.2 Information on newtworks available from the study Table 3 summarizes the information on network structure and network usage which is included in the survey questionnaire. In addition, extensive qualitative information on both topics was collected in the anthropological case studies (Working Paper No. 4). The questionnaire included a matrix on which respondents were asked to indicate the geographical location of all their relatives as well as the frequencies of visits. Similar information was also obtained for non-kin with whom there was an exchange of money or goods. - 25 - Table 3 Information on Netmrk Structure and aetwork Usage in the Facuesta sobre Ingresos y Gastos en los Hogares 1. Formation of separate households: 1. Ibnetary transfers sent (for each household receiving transfer from respondent) - parents of head - gift, loan, etc. - parents of spouse - times sent during year - brothers of head - average per month - brothers of spouse - amount last month - children - form of reciprocation - ex-spouse of head - hcw money used - friends from whom help is - sent to household or individual received or to whom given - who initiated transfer 2. For each of above separate 2. Ibnetary transfers received by households: the respondent: (detailes as for 1) - age of head 3. Goods transferred to other - sex of head households: (for each household - relation to respondent receiving transfer) household - frequency during year - where lives --sent last month - economic level compared to - type of transfer (bocks, respondent food, etc.) - frequency of visiting respondent - estimated money value - sent to household or individual - who initiated transfer 4. Goods received by respondent: (details as for 3) 5. Teaporary household menbers: For each temporary menber: - age - sex - reason for living with respondent - ecoaomic situation of home coffpared to respondent - time living with respondent - labor force activity 6. 1mbers teqrporarily absent: - relaxation to household head - sex - age - where living - with whom living - economic situation of respondent conpared to this household - activity - receives mney from respondent - sends money to respondent - 26 - The matrix recorded information for each exchange of money or goods on the value, frequency, kinship relatin between giver and receiver, purpose for which the transfer was made and the form of reciprocation. The survey had the disadvantage that no information was obtained on non-kin with whom interactions other than the exchange of money or goods took place (such as visits or the exchange of information or services), so it is not possible to estimate the total number of people or households forming part of the personal network. 1 Quantitative information is available on two distinct types of network: (a) The EFFECTIVE NETWORK, consisting of all households, both kin and no-kin, with whom there is an exchange of money or goods; and (b) The KINSHIP NETWORK, consisting of all primary kin members, irrespective of whether there is an exchange of money or goods. I/ Exchanges of information and services through "personal" and "social" networks are indeed numerous and important for SEZ households. It was found that vendors, construction workers, housewives and other categories of persons living in the Zone often have regular contacts with friends about which their spouses or childrei, know relatively little, but which are crucial for the household's general welfare. Vendors, for example, help each other de-bulk and re-bulk their wares at the beginning and end of the marketing day, watch each other's stands when they takebreaks, pass along messages to kin or other friends, etc. Construction workers rely on each other for tips on finding work and to speak for each other's character to foremen. Housewives will baby-sit for each other's children, take turns going shopping or getting water, watch each other s houses when they run errands, etc. all are likely to turn to each other as sources of assistance in an emergency "on the job." Small amounts of cash also constantly flow through these personal networks, as advances and returns for purchasing snacks, drinks, cigarettes, lottery tickets, and other small items. The net financial value is high, however, because they ease the work burden for the exchangers. The more frequently these exchanges are made, and the more diverse they becomme, the greater is the likelihood that the "personalized" relationship will become part of the household's social network, as men and women acknowledge the importance of these "exchanges" to their spouses. - 27 - The information from these two sources is combined into what could be called the ESTIMATED LATENT NETWORK. For the reasons given above, this concept is likely to substantially underestimate the LATENT network as non-kin are not included if they have not given or received transfers. Information is also included on the transfer of people between households and data is available on the characteristics of temporary household members and on members who are temporarily absent. Finally, information was obtained on the perceived economic level, relative to the subject's household, of all other households in the networks. This permits analysis of the relative importance of network transfers from higher income to lower income households and of transfers between households of similar economic levels. 2.3.3 Definition of networks and transfer commodities Table 4 shows the way that network structure was operationally defined in the study. For the purposes of the present study, a NETWORK is defined as a system of relations between individuals or households 11 through which one or more COMMODITIES are, or potentially could be, transferred. The present study analyzes the transfer of money, goods and, to some extent, people. It is possible to study either LATENT NETWORKS, which are the systems of social relations through which a COMMODITY actually flows. In the present study the analysis of distance between network households, the frequency of visits and the number of kin could be considered as indicators of a latent network through which the transfer of money or goods could potentially flow. 1/ The points between which commodities flow are often referred to a NODES. - 28 - Table 4 The Definition of Network Structure -[PARENT OF HEAD CHILDREN OF EITHER Whether or SPOUSE not transfercE take place SIBLINGS OF HEAD SIBLINGS FSOS DISTANCE EX-SPOUSE Neighborhood Zona HOUSEHOLD ROSTER Cartagena Demographic and Colombia employment Outside characteristics Colombia of each member FREQUENCY OF VISITING Weekly Monthly More than If transfers . yearly take place OTHER RELATIVES OR Yearly FRIENDS Less Never - 29 - If the aaalysis is restricted to the transfer of a single COMMODITY (for example, money) this is defined as a SIMPLEX NETWORK. If, on the other hand, the analysis covers the exchange of two or more COMMODITIES, then this is a MULTIPLEX NETWORK. The present study permits the analysis of multiplex networks as information is available on the transfer of money, goods and, to some extent, people. 2.3.4 Types of transfer The flow of any COMMODITY between households or individuals is referred to as a TRANSFER. It is useful to distinguish three types of transfer: (a) EXCHANGES, in which there is an immediate or expected future reciprocation. (b) GIFTS, in which there is no direct form of reciprocation. (c) REMITTANCES, in which the transfer takes place between members or households within the same family. In this case, there is no immediate reciprocation in the usually understood sense as the remittance represents a reallocation of resources within the same family (economic) group. In practice, however, it is often not easy to categorize transfers as strictly belonging to one category or another. Sometimes return for an "exchange" takes so long that the original donation becomes a "gift". More importantly, Zona dwellers make distinctions within these categories. Not all "remittances" between Zona households related by kinship are reciprocated in the same amount of time. Those kin-linked households who have great trust (confianza) in each other are generally less concerned about when and if theyll receive "return transfers" than other kin-linked households who are less socially or physically "close" in their relationship. Geographical - 30 - distance plays an important role here: SEZ inhabitants remit income more freely to their parents and "close" or "favorite" siblings than they do to siblings with whom they have little regular contact or to more distant relatives. some "remittances", in other words, may closely resemble "exchanges" or "gifts" 2.3.5 Network usage Table 5 shows how network usage was defined in the study. NETWORK USAGE refers to the patterns, volume and frequency of transfers within a network. The volume of transfers within a given time period may be measured in one of three wyas: (a) GROSS TRANSFER, where receipts and remittances are measured separately. (b) NET TRANSFER, which is the difference between receipts and remittances. (c) TOTAL NETWORK USAGE. This concept, suggested by Butz (1981), is the sum of receipts and remittances. This is a useful way to express the overall importance of transfers in a given community or group and is a better indicator of the volume of network activity than net transfers. Net transfers is often given a misleadingly low figure as receipts and remittances may largely cancel each other out. In fact, in any closed system the sum of net transfers must be zero. Table 5 The Analysis of Network Usage MONEY PARENT OF HEAD Receipt/Remittance Fmequency WhthrArRENT 0F SPOUSE AmutWhether or______________ Use not transfers CHILDREN OF EITHER SPOUSE Reciprocation take place How initiated HOUSEHOLD ROSTER GOODS Demographic and Receipt/Remittance E- employment Frequency characteristics Type of good of each member Value -----""Reciprocation How initiated- PEOPLE Age/sex/kinship Time Economic activity Remitting to or receiving from If transfers respondent take place OTHER RELATIVES OF Comparison of FRIENDS economic level of origin and residence - 32 - 2.3.6 The functions of interhousehold transfers Three broad categories of transfer usage can be distinguished: (a) SURVIVAL or income maintenance function, in which transfers are used to help a household to maintain a certain minimum level of basic needs consumption or to resolve a household emergency such as death or illness. Several households interviewed for the intensive case studies received remittances for survival from their children. One elderly couple was entirely dependent on remittances from their five children in Venezuela, while a female-headed household containing three adopted children received about 80% of its basic needs income from the head's biological daughters in Cartagena and Barranquilla. In a rare instance, an elderly woman living alone had her subsistence needs met by several neighboring households. More commonly, two or more households unrelated by kinship provide each other with cash and services. In one such case, reflective of the "contractarian" exchange posited by Kaufmann (1982), one household helped another meet its basic needs for survival, while the other returned services and occasionally cash for the first household's "development". (b) DEVELOPMENT function, in which the transfer is used to improve future economic or social conditions of the household. Examples would include the use of transfers to start or expand a business, to acquire a new house, or to finance a child's continuing education. SEZ households do send and receive transfers for development purposes. One woman occasionally sent her sister cash to expand here small business, selling fried - 33.- sweets at a foodatand along the Avenida Pedro de Heredia, the commercialized southern boundary of the Zona. Several households ceded small lots to relatives, so that the latter could construct better or less expensive dwellings on them, than if they attempted to purchase land from strangers. A number of other households sent transfers to favorite nephews and grandchildren, "earmarking" these for the youths" education. (c) MAINTENANCE OF SOCIAL TIES. This is a somewhat complex but necessary category which includes many types of gifts to friends and relatives, or to religious, cultural or political groups. In this case, there is no immediate expectation of reciprocation and the purpose is to maintain ties to a network which may prove useful in the future. A good example are the remittances to rural areas of origin by urban migrants. This type of social maintenance has been extensively studied in Africa (Moock, 1978; Mukras and Oucho, 1982) where large amounts of resources are transferred to non-relatives and to community organizations in rural areas of origin with no immediate expectation of reciprocation. Nearly all households interviewed intensively in the Zona maintain social ties of this kind with particular relatives and friends. These ties are kept up for the mutual emotional support the households give each other, for taking care of each other's occasional "wants" rather than "basic needs", and for insuring that the exchange partner will help out in case of an emergency. "Wants" in this context refer to things or services desired by one's relatives or friends, which they can certainly live without, but which serve as tangible - 34 - proxies or remindars that a valued social relationship for support exists between the exchangers. Transfers of this sort range from serving as a compadre or comadre at a friend's or relative's child's religious sacrament, to exchanging toys, cigarettes, and other non-essential goods, to sharing household chores or working together on a small drainage problem, to passing along valuable information concerning jobs, bargains, political support from patrons, etc. Individually, transfers made to maintain social ties have different degrees of financial worth. Taken together, they represent multi-purpose, socially and emotionally significant bonds. As in the case of transfer types, it is not always simple in practice to identify transfer types, it is not always simple in practice to identify transfer functions as belonging to one category or another. Providing occasional, small-scale financial support or labor services to help a relative make small improvements on a house could be considered both a development function and illustrative of the maintenance of an important social tie. Allowing relatives to live rent-free on zonedwellers property, or charging them less to live there than one would charge strangers functions both to insure the relatives' survival or development and maintains the social tie between owning and renting kin. There are a number of operational methods by which the function of a transfer can be identified, however. One way is to ask the giver how he or she expects the transfer to be used. A second way is to ask the receiver why the transfer was received or how it has been used. Both of these approaches are often not very satisfactory as transfers are frequently part of continuing patterns of interaction and respondents may have difficulty in isolating the specific purposes of a particular receipt or - 35 - remittance. Thus, although observation and statistical analysis usually can identify patterns underlying transfer behaviour, respondents often claim that most transfers were gifts which were not given for any particular purpose. To avoid these problems, a final method is to infer the implicit motivation from transfer behaviour by estimating statistically the impact of transfers on expenditure patterns. For example, in El Salvador (Kaufmann, 1982) and the Philippines (Urban and Regional Economics Division, 1980), it was found that access to housing investment opportunities increased the receipt of transfers and it was argued that the transfers were, to some extent, "earmarked" for this purpose. In the latter case, the transfer was often directly used to purchase food or to cover other basis expenditures, but at the same time this freed resources which the recipient family was then able to invest in housing. Earmarking As discussed in the section describing the conceptual framework, the possibility of the household earmarking the incoming transfers for specific expenditures had to be considered. Such earmarking might have been donor or receiver-induced. We included questions in the survey attempting to capture information on the type of expenditure to which the transfer was to be allocated (general expenditures, education, food, housing, others) and on who was the initiator of the transfer. Both the field work and the questionnaire results made it clear how difficult it is to obtain meaningful information in a structured questionnaire form when dealing with issues which are not directly quantifiable, and are of a rather complex and implicit nature. Household income is fungible in terms of possible expenditures; the expectations and responsibilities of the donor and the receiver regarding the uses of the transfer are quite subtle and implicit; the complex social - 36 - interactions among households in the network does not make it necessarily obvious who the initiator of the transfer is or what the main motivator of the remittance was. From the attempt to find out information on earmarking we learned the clear limits of a structured socio-economic questionnaire, notwithstanding the careful iterative quantitative/qualitative approach we had taken in designing it. The in-depth interviews turn out to be much more adequate when dealing with issues that include perceptions, expectations, subtle interactions, implicit understandings, attitudes, etc. The only promising quantitative route for utilizing the survey data to probe into the earmarking issue is to try to explain, through multivariate analysis, the expenditure patterns of the household and see whether the transfer-intensity (proportion of transfers in total income) is a significant determinant of these expenditures. In this case the revealed behavior and quantitative information turned out to be better suited for analysis through a socioeconomic questionnaire and the subsequent data analysis than the inclusion of more qualitative questions. The Contractual Arrangement A varied gamut of contractual arrangements exist behind inter- household transfers. The very fact of the reciprocity imbedded in these transfers suggest that they are expectations, rights and obligations accompanying these flaws. Once again, our field work experience indicated that the notion of contractual arrangements, in the minds of the interviewed households, is more subtle, complex and implicit than a more western-oriented, legalistic view of a contract. Hence, formal categories such as transfer-as- a-gift versus transfer-as-a-loan (as asked in one question in the survey) did not work adequately in the questionnaire. - 37 - It proved equally difficult to obtain information from the transfer receiver household about the way the transfer was to be paid back. Direct questions could not capture the household's own approach to social inter- actions, which in most cases resembles more on implicit-insurance contract than on explicit loan contract. The qualitative aspects of the reciprocity arrangements and the rights and obligations of these transfers were much better captured by the anthropological work. However, the quantitative survey did provide insights by comparing magnitudes and directions of transfers made and received with the socioeconomic characteristics of the household. 2.4 Household Incomes, Expenditures, and Wealth The.iterative and interdisciplinary approach to devising the questionnaire highlighted the need to proceed in a very detailed but flexible fashion regarding questions on incomes and expenditures. However, if these procedures are followed it is possible to obtain a very reliable data set on incomes and expenditures and to at least partially overcome many of the commonly voiced fears and criticisms regarding measurement of incomes (and their underestimation) and of expenditures (and their overestimation). Consistency checks with other questions in the survey as well as with the in- depth interviews indicate that the data has a high reliabilityj! From an analytical standpoint it has proven useful to have gathered data on household wealth. Data on the value of the house, the assets of the household, including economically useful animals were obtained. It appears that the household wealth is important in explaining social network interactions. Furthermore, the data indicates that the correlation between 1/ For the ways in which data on incomes and expenditures was obtained, see Annex 1. - 38 - household wealth and income, whether current or (an approximation of) permanent, is quite low. Since in many circumstances the household's behavioral patterns depend significantly on the composition of its wealth, it may be important not to overlook the collection of information on this variable. In particular for the urban informal sector, where incomes are unstable and fluctuate widely during the family life cycle, an income measure, however adequate, could prove to be a very imperfect proxy for wealth. Section 3: The Content and Evaluation of the Questionnaire li 3.1 Main content areas Annex 1 summarizes the main content areas of the questionnaire (in English) and Annex 3 gives the complete questionnaire in Spanish. In addition to the main sections on social networks and transfer behaviour (see below), information was also obtained on: Tenure: It was expected that security of tenure would be an important determinant of transfers, particularly those affected by investment considerations. Housing quality and methods of finance and construction: Housing quality and value are both indicators of socio-economic status and useful measures of housing improvement (as a consequence of the project). Information on housing finance and construction methods (particularly the use of family labor) is likely to reflect the impact of house improvements on transfer flows of money and labor. 1/ The complete questionnaire, interviewers guide and coding manual are available in Spanish in "Aspectos metodologicos para estudiar patrones de formacion de ingresos y gastos familiares entre familias pobres urbanas." Instituto SER. February 1984. Bogota (Available through the Water and Urban Development Department of the World Bank). - 39 - Demographic characteristics of the household: Age, sex, kiiship, education were obtained for each member. In addition questions were included on whether each member was permanent or temporary. For the latter group information was obtained on the reasons for joining the household, time of residence and a comparison of the economic level of their family of origin with the respondent household. Information was collected on each member so as to permit the creation of aggregate measures such as household size, dependency ratios etc. Economic characteristics of the household: Economic activity, type of organization for whom working, hours, earnings and time of permanance in the job. Information was also obtained on income fluctuations during the past 12 months. The information was collected for each household member, thus permitting the construction of aggregate economic indices such as amount and sources of income, stability of income, number of members economically active etc. Other sources of income: Income from small businesses, cultivation of crops or animals, rental income and transfers. Information was also obtained on the ownership of capital goods so as to permit a comparison of income and wealth. Household expenditures: Information was included on each of the main expenditure categories. 3.1.1 The structure and usage of social networks A. Network structure Information was obtained on the numbers, geographical location, frequency of visits to respondent household and family with whom living, for each of the following: - 40 - a) Parents of head b) Parents of spouse/companion c) Siblings of head d) Siblings of spouse e) Children of spouse or head f) Ex-spouse. Geographical location was classified into the following 5 ordinal categories: 1) Outside Colombia. 2) In Colombia 3) In Cartagena 4) Within the Zona 5) Same barrio These categories combined with frequency of visits permitted the construction of 3 indices of intensity of interaction: i) Weighted index of distance. Scores from 5 (nearest) to 1 (most distant were given to each location.). The location index was then defined as: D = EN N W Where: N1 = No. of network members in the ith geographical zone. W1 = weight of ith geographical zone. ii) Weighted index of frequency of visits. Frequency of visits was coded as follows: (5) Weekly or more frequently (4) Monthly - 41 - (3) Less than monthly but more than once a year (2) Yearly (1) Less than yearly (0) Never The frequency index was then defined as: F E F *N i i Where:- Fi = No. of network households in the ith frequency of visits Ni = Weight of the ith frequency category iii) Combined distance and frequency index: Frequency and distance were combined to produce the following index: FD = E F D *WD ii 1i Where: F. iDi = No. of households in the ith Distance category and the ith Frequency category WFi = Weight for ith frequency category WDi = Weight for ith Distance category B. Types of transfers through the social network Information was obtained on the following types of transfer "commodities" (a) Money (b) In-kind transfers (c) People moving to live with another family (d) Visits to other network households - 42 - C. Monetary transfers The following information was obtained on all monetary transfers remitted from, or received by, the household: (a) Did a monetary transfer take place during the past 12 months (b) Was the transfer a gift, loan etc (c) Frequency of transfer during past 12 months (d) Average amount per month (e) Amount of transfer last month (f) How was the transfer reciprocated (g) What was the money used for (b) Was the money sent to the household or to a particular person (i) Was the transfer requested by the recipient or sent on the initiative of the sender. D. In-kind transfers Information was obtained on the remittance and receipt of the following: (a) Food (b) Clothing (c) School books and materials (d) Household articles For each in-kind transfer similar information was obtained to that for monetary transfers. In addition an estimate was obtained for the monetary value of the transfer. - 43 - E. M4ovement of peole through the social network The following information was obtained on all temporary (as defined by household head) members of the household: (a) Age (b) Sex (c) Kinship relation to head (d) Reason for residing in the household (e) Comparison of the economic level of the household of origin with the respondent household. (f) Time living in the household (g) Economic activity Similar information was obtained on permanent household members (as defined by the household head) who are temporarily absent. In addition the following information was obtained: (a) Where living (b) With whom living (c) Does he/she receive support from respondent household (d) Does he/she send support to respondent household F. Assistance from social network in times of emergency The following were defined as potential economic crisis situations: (a) Serious illness (b) Death (c) Childbirth (d) Difficult economic situation in which unemployed and without money. -44 For each case information was obtained on which of tha following sources of assistance were used: (a) Own resources (savings) (b) Relatives (c) Friends (d) money lender (e) Formal sector (social security, employer etc) 3.2 Evaluation of the questionnaire The interdisciplinary approach, and the preparation of case studies proved extremely useful in the design of the survey. In particular, the constant feedback from the qualitative studies helped the researchers understand and measure more adequately the concepts related to social networks and their usage. Comments are presented on the evaluation of each of the main sections of the questionnaire: Tenure The tenure situation was straightforward to measure and no problems were encountered. Housing quality and methods of finance and construction No problems were encountered in the collection of information on quality, methods of finance or construction. The only possible difficulty related to imputed rental and sales values. However, in some parts of the zona there is a well functioning housing market so that people have a reasonably good idea of values. Given the nature of the survey there were no apparent reasons why respondents would deliberately distort the information. - 45 Demographic characteristics of the household The information proved straightforward to collect. In general the concept of household head was well understood, with the only potential problem relating to households where a woman who had virtually been abandoned by her husband or companion was the de facto head. In these cases it can sometimes be difficult to establish whether the effective head of household is the wife or the (former) husband. Economic characteristics of the household and other sources of income Although it is always difficult to obtain accurate information on household income when a large proportion of the population works in the informal sector and when most sources of income are unstable; the survey data was consistent with the information obtained in the in-depth case studies. We earned income are probably more accurate than in most studies. The estimates of total household income are likely to be more accurate than is normally the case due to the much more careful measurement of transfer income (a source which is either ignored or only covered very superficially by many studies). Expenditures No particular problems were encountered, and the survey findings were generally consistent with the case studies. Network structure and network usage The problems which arose were mainly conceptual. Respondents were quite willing to provide information on their networks and how they were used, but in several cases it proved difficult to form precise questions. The following are some of the difficulties encountered: -46- (a) It was not possible to operationally define the social network in a way which would include all persons with whom the household had some type of interaction. Thus although it was a simple matter to get respondents to list their relatives and where they lived, it was not possible to obtain a complete list of friends and acquaintances in the same way. After much discussion and experimentation, respondents were asked to list friends and other relatives from whom they had received, or to whom they had sent money or goods during the past 12 months. This means that information is not obtained on non-relatives who form part of the social network but with whom no exchange of money or goods took place. This can be seen in Table 6. Thus we can distinguish between a "potential" and an "effective" network with respect to relatives; but for non-relatives we can only examine the effective network. In practical terms it is probably not too serious a problem as the only information we lose refers to the number of people from whom no type of assistance was received. (b) In retrospect it would have been useful to have included more detailed information on distant relatives and non-relatives who form part of the social network. Godparents (compadres), uncles and aunts in particular, play an important role and special categories should probably have been included for them. The reason for not doing this was to avoid the interview becoming excessively long and detailed, but despite this a future survey should include these additional specific categories. (c) It was only possible to measure exchange of money, goods and people. Initially we had hoped to include exchange of services such as care of children, help with house construction or repair and preparation of food. However, after experimentation it was decided that services were too complex to operationalize and they were omitted. For example, if a child is left all - 47 - Table 6 Use and limitations of available data for the study of effective and potential networks 1. ANALYSIS OF EFFECTIVE NETWORKS (i.e observed transfers) Required information available on both relatives and non-relatives Relatives Non-Relatives Information on transfers YES YES 2. ANALYSIS OF POTENTIAL TRANSFER NETWORKS (identification of all members of the social network)' Required information not available on non-relatives with whom there was no network transfer Relatives Non-Relatives Information YES YES on transfers Information on people with whom YES MISSING no transfer took place - 48 - day with another woman how do we measure the value of this s8rvice? The woman would have stayed at home in any case so the cost to her of providing the service is quite low. In this community no-one pays money for baby-sitting so it would have been difficult to estimate the market rate. It is also difficult to measure the number of hours for which the service was provided. Often the child stays in its own home or plays outside with the other woman simply keeping an eye on it. It would not be reasonable to treat this service as being equivalent to placing the child in a kindergarden. This omission is important as the exchange of services plays an important role in a low-income community of this kind, and the computation of the value of the services may affect our estimates of real income. (d) One of the hypotheses we wished to test related to the "ear- marking" of transfers (i.e. transfers which are given on the understanding that they will be used for a particular purpose). In most cases respondents were not able to identify a specific purpose for which a transfer was given, as they considered each transfer to be part of a complex and continuing pattern of exchange. Even though the receipt of a particular transfer might be related to the fact that a child was beginning school, the money might have been used to purchase food or another basic necessity. This in its turn might have freed money from the household budget which was then used to buy school materials. However, it proved to be very difficult for households to trace through this complex series of events. (e) It was not possible to describe the overall characteristics of the social networks as this would have required the collection of information on the patterns of interaction between all households in the network. In the present study information was only obtained on relationships between each network household and the respondent household. In other words if there were - 49 - 10 households referred to by the respondent (A,B,C,...... J) and if the respondent household is classified as X; we have information on network interaction between X and each household (XA, XB etc) but we do not have information on the interactions between other households (AB, AC etc). This means that it is not possible to describe the overall characteristics of the network (density, total volume of exchange etc). This is not too serious a limitation in the present case as we can assume that the samples of networks we have observed are a random sample of all network interactions of households in the± Zona. (f) It is likely that an error may exist in the comparison of the economic level of respondent housholds and of other households in their network. Respondents were asked to indicate for each network household whether their general economic situation was better, about the same, or worse than that of the respondent's household. Within the constraints of the present study it was not possible to visit other network households to obtain an independent check on their economic level. However, the ethnographic studies suggested that families have,a clear idea of how their condition compares with that of other households and they seemed willing to make the comparison so we have no reason to expect a systematic bias. The judgement about economic levels was based more on the perceived permanent economic condition of households rather than on how much they might have ear:ned in a particular month. Section 4: Ethnographic Case Studies As mentioned earlier, the review of the extensive sociological and anthropological literature on social networks and survival strategies demonstrated clearly that it would not be possible to adequately understand - 50 - and analyze interhousehold transfer behaviour purely on the basis of a structured survey. Previous ethnographic studies of social support strategies among poor Latin American households, conducted by Oscar Lewis, Lisa Peattie, Janice Pearlman, Larissa Lomnitz, William Norris and Isabel Nieves, among many others, revealed that the extent of inter-household reciprocity, gift-giving and dependency varied with a complex array of factors, including levels of income, types of employment, closeness of kinship ties, degree of confianza or trust among households, migration histories, etc. Just as these factors changed over time, so did the households' strategies for support and the types of individual transfers given and received. The in-depth study of Zona households followed the quantitative survey's concern for delineating the kinds of, and motivations behind, transfers in cash and kind to and from Cartagena households. It went on to probe into the extent of non-commodity transfers such as exchanges of services, information and affective support, to see how households evaluated the role and importance of all exchanges made vis-a-vis one another. This tactic, it was felt, would help clarify both the general aims of the transactions, that is identify whether and how families aided each other for "survival," for simply maintaining their social ties, or for their "develop- ment" through investment, as well as to specify better the circumstances triggering individual transfers. By conducting a participant-observation study, moreover, investigators could sensitize themselves to the range of rights and obligations attached to all transfers, and thus more fully understand how Cartagena households themselves calibrate the meaning of helping each other on a regular, sporadic or emergency basis. Repeated visiting of households would also facilitate obtaining at least fairly - 51 - reliable "household histories" and genealogies, which would give a perspective on how long families had aided each other or why some households had chosen to focus more strictly on their own welfare. The anthropological study, in short, would not be just a qualitative cross-check on the quantitative survey. It would seek out more information on the variety of exchanges made between households, verify their purposes, pin down the circumstances under which they were offered and requested, and examine how these changed over time. In so doing, it was hoped, a better perspective might be obtained on the potential complementarities between interhousehold transfers and the Bank-supported urban development project. Towards these ends, the anthropological study contacted about 30 households for lengthy interviewing about the history of their households and their network contacts; their patterns of transferring (gift-giving, exchange and remittance); their motivations for doing so, their expectations of their network partners; their views about the project, etc. Six of the families contacted during November-December, 1981, were re-visited throughout April- May, 1982, and provided the core for the analyses made about household survival and development strategies, forms of exchange, and consequences of transfer-giving, transfer-receiving, transfer-seeking, and transfer-avoiding processes. All six were interviewed at least three times and visited informally as well; the investigator spent mornings, afternoons and evenings with three families in particular, representing a wide range of network behavior. Friends, relatives, co-workers and neighbors of the six households were met and information gathered about their perceptions of the relationships bringing them together. Interviews were largely conducted at or near the - 52 - households' dwellings, but also took place at their worksites, recreation areas and contact points with friends and patrons, both inside and outside the Zone. The methodology and findings of the case studies are described in Working Paper No. 4 Section 5: Sample Design The South East Zone (the Zona) in which the study was conducted, consists of a long narrow strip of uncontrolled housing. The Zona is approximately 5,400 meters long from East to West and is bordered on the South side by a lagoon ("Cienaga de la Virgen"), and on the North by the main commercial highway (Avenida Pedro de Heredia). The houses become progressively poorer and the sanitary conditions worse as one moves nearer to the lagoon. This is not surprising as the lagoon is subject to frequent flooding and is also highly unsanitary due to the accumulation of untreated waste material. In contrast many of the houses close to the road are 2 storey brick constructions with expensive doors and windows and other indicators of relative affluence. Differences in the level and quality of physical infrastructure were closely related to the progress of the ICT (Instituto de Credito Territorial) upgrading in each sector. For the purposes of the study 4 parallel strips were defined on the basis of distance from the lagoon. These were crossed by 4 horizontal sectors based on the stage of development of the ICT upgrading project. In this way a total of 16 sectors were identified. The sectors were not used in sample selection but they were used in the analysis. - 53 - According to an ICT Census, conducted in 1981 there were a total of 11,610 structures in the Zona, housing an estimated 90,000 people. The area contained approximately 402 blocks ("manzanas"). A total of 80 blocks were randomly selected, and a listing was made of all str!ctures in each block. Each structure was visited and all households (as defined in Section 2) were listed. It was estimated that a sample of approximately 500 households would be sufficient to permit estimations of the proportions for the most important variables with a confidence interval of + 5% at the 95% confidence level. A sample of households was then selected in each block with the number of households selected being proportional to the total number of households in the block. The number of households selected on each block varied between 4 and 9. By using these procedures a 2 stage self-weighted stratified sample of households was drawn. For most of the analysis the sample was treated as if it were a simple random sample. Theoretically the efficiency of the estimates could have been increased by using more complex weighting procedures. However, this was not done, partly to avoid additional complexities in the analysis, but mainly because most of the analysis was based upon multiple regression and non-linear estimations where weighted samples should be avoided. Table 7 Stages in the Sample Design CHARACTER- SECTORIZ- SELECTION ISTICS OF ATION OF OF THE ZONA THE ZONA BLOCKS SELECTION OF HOUSEHOLDS 90,000 Random Listing and Estimated that Self-weighted sample people selection mapping of sample size of in each block with 11,600 of 4 all 500 households sample size proportional structures blocks structures sufficient to to number of households. 402 blocks from each and permit estimation Minimum sample = 4 and section households lof key variables maximum = 9. in each "~-with 5% confidence blocks selected interval and 95% Precision could be block confidence level increased by stratifying into the 16 sectors. 4 strips parallel to the lagoon 4 vertical strips based on progress of ICT infra- structure. Total of 16 sectors SECTORS NOT USED IN SAXPLE SELECTION - 55 - CHAPTER THREE: DATA ANALYSIS AND INTERPRETATION Section 1: Analysis of network structure and network usage 1.1 Network structure 1.1.1 Description of network structure The characteristics of the network structure were defined in terms of: (a) The SIZE OF THE ESTIMATED LATENT NETWORK: The sum of all kinship related households in the network (irrespective of whether there was any reported transfer interaction) plus non-kin network households with whom interaction took place. As mentioned earlier, it was not possible to include non-kin households with whom no transfers took place. Consequently the estimated latent network is smaller than the total latent network would be if all non-kin could be included. (b) The LOCATION OF NETWORK HOUSEHOLDS: The number of households living in the immediate neighborhood, in the Zona, in Cartagena, in other parts of Colombia and in other countries. A summary index of distance was created by assigning weights to each category. (c) The FREQUENCY OF INTERACTION: The number of households in each of a set of frequency categories and summary indexes were created by assigning weights to each category. (d) The RELATIVE ECONOMIC LEVEL OF THE RESPONDENT HOUSEHOLD: Comparison of the number of network households who were perceived by the respondent as having a higher, similar and lower economic levels. (e) KINSHIP RELATIONSHIPS: Summary of the kinship relationships of all network households to the respondent household. - 56 - 1.1.2 The determinants of network structure 6 indicators of network structure were defined: (a) No. of people in the network (b) No. of households in the network (c) Frequency of interaction (d) Combination of interaction and distance (e) Network households with a higher economic level (f) Network households with a better or equal economic level. The values of each variable were first compared for male and female headed households. Each variable was then regressed on a set of indicators of the socio-economic characteristics of the household. 1.2 The volume of network usage 3 indicators of network usage were used: (a) Transfer receipts by Zona hcaseholds. (b) Transfer remittances by Zona households. (c) Total network usage. Defined as the sum of receipts and remittances. 1.3 Patterns of network usage 1.3.1 Descriptive analysis The following indicators and analyses were used to describe network usage: (a) Numbers of households involved in network transactions. (b) Comparison of the characteristics of households who give and receive transfers. Households were classified into transfer GIVERS (who gave but did not receive transfers); RECEiVERS (who received but did not give); NON-USERS (who neither received nor gave); and USERS (who both received and gave). The households were first compared in terms of non-transfer household - 57 - income. A Discriminant Analysis was then conducted in which the socio- economic characteristics of all 4 groups were compared. The analysis worked well and a number of significant differences were found. (c) Descriptive analysis of monetary transfers sent from Zona households. Transfers were analyzed in terms of the relative ages of sender and receiver, the kinship relation, geographical location of recipient and the relative economic levels of sender and recipient. The purpose of the transfer and the reciprocal obligations entailed were also examined (d) Descriptive analysis of monetary transfers received by Zona households. A similar analysis was conducted as for (c) 1.3.2 The determinants of network usage The following indicators of network usage were defined: (a) Value of transfers given (b) Value of transfers received (c) Proportion of network households to whom transfers remitted. (d) Proportion of network households from whom transfers received. (e) The proportion of transfers which take place within the Zona (as opposed to transfers to or from areas outside the Zona). Each of these was regressed on a set of socio-economic characteristics of the household. All of the regressions identified statistically significant household attributes although the R squares were in general relatively low (in most cases not exceeding .35). 1.3.3 Use of networks in times of emergency An analysis was made of the types of network support obtained in times of household crises (births, deaths, sickness and unemployment). - 58 - 1.4 Transfer of people through social networks An analysis was made of the characteristics of household members who were temporarily absent and of temporary household members. The groups were compared in terms of age, sex, activity and their economic contribution to the respondent household. The information was then used to estimate the net economic impact of absent and temporary members on the economy of the household. Section 2: The role of the case studies Chapter 2 stressed the interdisciplinary approach used in the study and the close intergration of the ethnographic case studies and the quantitative analysis. The case studies were used in the early stages of the study to help understand the social and economic functions of transfers and to assist in the formulation of hypthoses and the design of the research instruments. The case studies also play an important role in the analysis and interpretation of the results of the study. First, the case studies are presented in their own right as one of the major components of the research methodology (see Working Paper No. 4). They show the important role which social networks play in all aspects of the li - of Zona families. The unstable economic and social conditions mean that very few households are self-sufficient and that most are constantly seeking help of one sort or another from their friends, neighbors and relatives. Households are interlinked by a complex set of expectations and social debts, and most actively seek to cultivate or maintain their network contacts. The case studies illustrate the complex nature of this web of interactions and exchanges and demonstrate why respondents had such difficulties in replying to - 59 - questions about the obligations created by a particular transfer or the way in which it was intended to be used. The detailed ethnographic descriptions provide the background against which the quantitative information relating to a sample of particular transfer events can be understood and evaluated. The second function of the case studies is to test the validity and practical usefulness of the patterns of transfer behaviour identified in the statistical analysis. Working Paper No. 1 identified the following patterns of transfer behaviour, and demonstrated the existence of statistically significant differences between them: (a) Transfer receivers (b) Transfer givers (c) Transfer users who both receive and give (d) Non-users who deliberately avoid participation in a transfer network (e) Non-users who do not have access to a transfer network. The Working Paper also included case studies of households belonging to each of these categories. One of the functions of these cases is to evaluate the explanatory usefulness of the statistical categories. They test the utility of the categories by seeking to understand whether they have any meaning to Zona households. Are households aware that they are transfer avoiders or transfer receivers, and is this part of a deliberate strategy on their parts? The cases show that families are clearly aware of the importance of social networks and that most of them are able to articulate their network strategies. A store-keeper, for example, understands that she must provide certain types of services to her neighbors to be considered as socially acceptable and hence to be able to cultivate and maintain customers. A second respondent was very deliberately develoging a social network, whereas another - 60 - family had a clear natwork avoidance stratagy. These illustrations help to validate the statistical categories and to understand the significance of the statistical relationships. A third, and closely related function of the cases is to help us to understand the meaning of the statistical findings. Why, for example, are certain categories of households "transfer avoiders", and why is the average income of "transfer users" (households who both give and receive transfers) higher than the income of families who only receive transfers? These three functions demonstrate that the case studies are not simply an appendix to the (more important) quantitative analysis, but have an equal status in the design and interpretation of the study. Although it is true th,at the utility of the case studies would be considerably reduced if they were not complemented by the statistical analysis, it is equally true that the statistical findings would be of very limited usefulness without the qualitative studies to help us understand what the statistics "really mean". - 61 - CHAPTER FOUR: THE RESEARCH METHODOLOGY IN RETROSPECT 1. Strengths and weaknesses of the methodology In general the research methodology and the conceptual framework worked well, although there are a number of ways in which the concepts and measurement of networks and network usage could be refined in a future study. The interdisciplinary approach worked extremely well and the quality of the questionnaire and our understanding of the dynamics and meaning of transfers were considerably enhanced by the case studies. It proved particularly effective to have a core team of economists, sociologists and anthropologists working together throughout the project so as to maintain a constant interaction between the disciplines. A number of ways were found in which the definition and measurement of networks and their usage could have been refined. First, more attention should have been paid to the study of interactions between non-kin network members, and between more distant relatives. Aunts, uncles and godparents in particular play an important part in the network. Second, it was clear that the exchange of services is very important and should have been given more attention. This was not done due to the conceptual difficulties of measuring the quantity and value of many types of services, but clearly this is an issue which should be studied in the future. Third, the study was not able to adequately describe the concepts of reciprocity and the types of obligations incurred by different types of transfers. This is a topic where the case studies can make an important contribution. Finally, the issues of childcare and the transfer of people (particularly children) through the network, should be given greater attention. Other studies, particularly in Africa, have shown - 62 - that children are often sent to live with relatives in order to maximize their educational opportunities. It would be interesting to explore whether a similar phenomenon occurs in Colombia. An important methodological issue relates to the moment in time at which the study was conducted. Due to unanticipated delays in project implementation, most families had not yet received the full benefits from the ICT upgrading program. Consequently it was not possible to evaluate what would be the final impact on transfers once the full benefits from the project were perceived. This problem was probably unavoidable when the study was designed, but it clearly limits the way in which the results can be interpreted. 2. Lessons for future research The following recommendations are proposed for future studies of this topic: 1. The interdisciplinary approach should be used in much the same way as in the present study. It is helpful for all members of the research team to be involved in the preparation of case studies to more fully understand the dynamics of the communities being studied. 2. A more intensive use of anthropological case studies should be included so as to: (i) Understand the strategies for developing and maintaining social networks. (ii) Study more fully the differences between Network Users, Receivers, Givers and Avoiders. (iii) Understand the reciprocal obligations generated by transfers - 63 - 3. The issue of the exchange of services should be studied more intensively, an attempt should be made to define the main types of services, to quantify them an to place some value on them. One issue of importance is to determine the extent to which poor familites use services to reciprocate receipts of money. 4.- Child care and the transfer of children between households should be studied. The research should investigate the reasons for the transfers and their functions as both survival strategies and investments in human capital. 5. In a future study it would be useful to conduct interviews at two or more points in time to observe the changes in transfer behaviour at different stages in the evolution of the project. One option would be to repeat the survey with the same sample with perhaps an interval of a year between the two applications. A second option would be to select a small panel of (say) 25-50 households and reinterview them periodically to observe their patterns of transfers. One successful example of this approach was the study by Mila Reforma in Manila in which 100 families kept continual records of income and expenditures (including transfers) over a period of several years (Urban and Regional Economies Division, 1980). - 64 - ANNEX 1 SUMMARY OF THE CONTENT OF THE QUESTIONNAIRE Section Contents 1. Identification Location of the dwelling 2. Characteristics of Region of origin head and spouse Time-in the community Education 3. Description of Number of rooms dwelling Construction.materials 4. Tenure and method Who owns the house of construction How built and financed Impact of ICT project on quality of house Comparison of economic level of house owner and tenants. Imputed rent and sales value 5. Social networks Assistance in times of Who was assistance obtained from during periods of emergency sickness, death, childbearth or other emergencies. A. Register of house For each member: hold members Age Sex Permanent or temporary member Economic activity of members over 10 years of age. B. Employment For each household member: Self employed o- who working for Months working during past 12 months Average days worked during past month Hours worked per day Earnings during past month in main employment Alternative sources of earnings and how much earned Highest and lowest earnings during past 12 months Affiliated to Social Security C. Other family groups Kinship relation of other household heads in the structure No. of members of other households Comparative economic level of other households Does respondent help other households Does respondent receive help from other households - 65 - Annex 1 Page 2 of 5 Section Contents D. Household members who are temporarily absent For each member: Relation to household head Sex Age Where living With whom living Economic level of residence compared to respondents household Economic activity Does he/she send money Does he/she receive money E. Register of all members of the social network Information was requested on all of.the following categories of people who do not form part of the respondents household: a, Parents of the household head b. Parents of the spouse or companion c. Siblings of household head d. Siblings of spouse/companion e. Children of head or spouse f. Ex-Companion of the head In addition respondent was asked whether there are any other friends, relatives of god parents who either give or receive monetary or other types of assistance. The following information was obtained on each member of the network: a. Age of household head where living b. Sex of household head c. Relationship of household head to respondents household d. Where living e. Comparison of economic level of network household with that of respondent f. Frequency of visit, to respondent household - 66 - Annex 1 Page 3 of 5 Section Contents E-1 Monetary transfers For each network household from respondent the following information was obtained: household a. Was money sent during the past year b. Was it sent as lpan, gift, etc. c. How many times sent during the year d. Average amount sent per month e. How much sent last month f. How was the transfer reciprocated? With money, labor etc. g. What was the money used for h. Was the money sent to the household in general or to a particular person i. How was the transfer initiated? (by request from recipient etc). E-2 Monetary transfers The same information as for E-1 was obtained received by for each receipt of money respondent household E-3 In-Kind transfers Similar inforiation was obtained except that: from respondent a. Respondents were asked to estimate the household value of the transfers. b. Details were obtained on what was sent. E-4 In-Kind transfers Similar information as for E-3 received by respondent household 6. Non labor market Information was obtained on the amount of money income received during the previous month from: a. Severance pay b. Pensions c. Scholarships d. Transport or other types of subsidy e. Rent from property f. Sale of animals g. Transfers from friends or relatives h. Pawning of household goods i. Sale of vegetables or other produce from garden j. Lottery etc k. Charitable assistance - 67 - Annex 1 Page 4 of 5 Section Contents 7. Household business a. Possession of land for agricultural production b. Are any crops currently sowed c. Does the household own any other business and if so where is it located 8. Ownership of a. Number of pigs animals b. Number of fowl c. Number of horses or cows 9. Household The household was asked to indicate how much expenditures was spent on each of the following items. In each it was indicated whether the figure was daily, weekly or monthly: a. Cooking fuel b. Water c. Electricity d. Household items e. Lotteries and gambling f. Clothing g. Education h. Medicine i. Rent j. Materials for repairing or improving the house k. Food 1. Cigarettes, rum and cinema m. Financial help to parents n. Debt payments 0. Transport p. Other expenses Ownership of the following consumer durables: a. Refrigerator b. TV c. Loud speakers d. Cassette radio e. Sewing machine f. Liquidizer g. Electric fan h. Living room furniture How much is saved each month Respondents were also asked which of the following expenditures would they make if their regular income was increased by $80 per month: - 68 - Annex 1 Page 5 of 5 Section Contents a, Debt payment b. Housing improvements c. Set up a business e. Electrical goods f. Education g. Clothing h. Food i. Install services (light etc) Respondents were also asked how they would spend a lump sum of $ 800 which was received only once. - 69 - ANNEX 2 DESCRIPTION OF THE AVAILABLE DATA SETS AND THEIR DOCUMENTATION 1. Characteristics of the Data The data refers to a survey of 505 households conducted in the Zona Sur Este of Cartagena. The original code contained 1198 variables. Subsequently additional coding was carried out to create a further 80 variables (relating mainly to the characteristics of temporary household members). The codebooks, with both the original and the new variables, are available in Spanish in both Washington and Bogota. The information is available in the form of raw data and SPSS and SAS files. Most of the analysis, and the creation of new variables was done on the SAS file. Both' the raw data and the SAS files were created on IBM computers and it is much more difficult, although not impossible, to use them on other computers. The sample design (see Chapter 2, Section 5, of this report) permits stratification by geographical sector. However, for most purposes it is much simpler and more economical to treat the data as a simple random sample, particularly in those cases where regression analysis will be used. The data has been thoroughly cleaned and in the opinion of the research team is internally consistent. Chapter 2, Section 3, describes some of the validity tests which were conducted on the data. 2. Data and Documentation Available in Washington Any enquiries relating to the data should be addressed to Michael Bamberger, Water Supply and Urban Development Department, The World Bank. The following 3 tapes are available: Tape 1 Tape 1 is the original tape created by the Instituto SER in Bogota. It contains three files: 1. REW 181 Tape Dump Ingresos Datos A Tape WT, This file contains the Cartagena study Raw Data. Printout available. - 70 - Annex 2 Page 2 of 3 2. REW 181 Tape FSFl Tape Dump Ingresos Banco A Tape Dump # Dire Banco A Tape WTM This file contains a SAS data set created by the Instituto SER. Printout available of the file definition. 3. REW 181 Tape FSF 2 Tape Dump Ingresos SPSS A Tape WTM This file contains an SPSS data set also created by the Instituto SER. Printout available of the file definition. The data recording was done by an IBM computer. The tape has no labels. To read the tape use the LOAD procedure. Density = 1,600 BPI. Tape 13 This tape contains three large and final data sets. The first SAS data set is CAR 5 which contains the original data set from Colombia, checked for internal consistencies and other tests. It also contains the new variables created for the analysis. The file contains the most clean, coherent and reliable information. The second SAS data set file is CARTAG4 which contains the additional variables which were coded and incorporated into the file. The "Las ojas de codificacion" prepared by Instituto SER are attached. This file contains variables which are not included in the directory of variables; their names are V1201-V1278. A copy is available of the file creation procedures. A printout of the first observation of each variable is also available and so is a list of all the variables. Tape 21 This tape contains some of the SAS files used to read the data. a. CARTAG1 and CARTAG2 were the initial files used to read the raw data. b. CAR5 is the SAS file used to create CAR5 SAS data set. c. CARTA4 is the SAS file used to create CARTAG4 SAS data set, d. CAR3451A, CAR3451B, CAR3451C, CAR3451D, CAR3451E and CAR3451F are the SAS files created to run the final regressions. - 71 - Annex 2 Page 3 of 3 3. Data Available in Colombia All enquiries relating to data available in Colombia should be addressed to Eduardo Velez, Instituto SER de Investigacion, Apartado Aereo 1978, Bogota, Colombia. The raw data file and the original SAS and SPSS files are available in Bogota together with their corresponding documentation. - 72 - ANNEX 3 QUESTIONNAIRE (Spanish) Note: The questionnaire was designed so that the information on the names and relationships of the members of the social network (Page E), can overlap with pages El (remittance of monetary transfers); E2 (receipt of monetary transfers); E3 (remittance of in-kind transfers); and E4 (receipt of in-kind transfers). This was achieved by making pages El through E4 smaller so that the information on the social network could be read at the left of each of these pages. CONI *DENIAL ENCUESTA SOBRE TTT OI~1 CONFIDENCIAL INGRESOS Y GASTOS Solo para uso EN LOS HOGARES estad Istico CARTAGENA - ABRIL/MAYO 1982 CAPITULO ¡ IDENTIFICACION DE LA VIVIENDA Y DEL HOGAR No. L . .LLL. Barrio LJ Zona 1 L . 1 1 Sector L_ No. de hogares Li Cercana a canal ( ) 1 Lejana de canal ( ) 2 REGISTRO DE VISITAS . Completa ()Incompleta (Rechazo 2a. Completa (ocompleta (Rechazo 3a. Completa (Incompleta Rechazo CAPITULO 11 IDENTIFICACION JEFE Y COMPAÑERA(O) 10. Y hasta qué curso hizo la compañera? Nombre Jefe: - Ninguno ( 1 - Elemental incompleto 1 1 2 - Elemental completo 1 ) 3 1. En cuál municipio nació el Jefe? - Bachillerato incompleto 1 1 4 - Bachillerato completo ( 1 5 - Universidad ( 1 6 1 19 2. En cuál municipio nació la Compañera? NA( )8 NA CAPITULO III NA -CARACTERISTICAS DE LA VIVIENDA 3. Hace cuantos años vive el jefe en Cartagena? 11. Cuántos cuartos tiene la casa, incluyendo la sala? Años Toda la vida ( NA No. de cuartos )99 4. Y hace cuántos años vive la compañera en Cartagena? 12, Material del techo de la casa Años ______ (OBSERVE) Toda la vida ( ) NA - Cartón o Suplex 1 1 - Zinc ( 12 5. Cuántos años hace que vive el Jefe en la zona? - Eternit 1 1 3 - Tejas 1 4 - Concreto 1 ) 5 Años |_. - Otro, cuál? ( 1 Toda la vida ( 9 6. Y cuántos años hace que vive la compañera en la zona? 13. Material del piso del cuarto principal (OBSERVE) Años L..L|J- Tierra ( ) 1 Toda la vida ( 1 NA ( ) - Cemento ( >2 - Baldosa - Baldosín 1 1 3 7. Hace cuántos años que vive el Jefe en esta ( )9 vivienda? Años L L 14. Tiene la casa conexión con el acueducto? Toda la vida 1 - No 1 )1 8. Y cuántos años hace que vive la compañera - sí ( ) 2 en esta vivienda? 1 >9 Años Toda la vida ( ) NA ( 15. Qué tipo de servicio sanitario tienen en la vivienda? 9. Hasta qué curso hizo el Jefe en la escuela? - Ninguno -1 )>1 - Hueco u hoyo seco 1 )2 - Ninguno ( 11 - Taza SIN tanque de agua - Elemental incompleto ( ) 2 conectada a pozo séptico ( ) 3 - Elemental completo ( ) 3 - Taza CON tanque de agua - Bachillerato incompleto ( ) 4 conectada a pozo séptico ( 1 4 -Bachillerato completo ( )5 - Taza CON tanque de agua - Universidad 1 1 6 conectada al alcantarillado ( ) 5 ( >9 1>9 CAPITULO IV PROPIEDAD DE LA VIVIENDA 24. Y para rellenar el lote o para construir la casa o para ha- cerle mejoras, de donde sacaron los materiales o el dine- 16. Esta casa es de Uds.? ro? - NO ( 1 -AI8 - SI 1 , 2 -sa (MARQUE HASTA DOS EN ORDEN DE IMPORTANCIA) - Préstamos del gobierno, como del ICl, 17. Y es: BCH, etc,? Li 1 - Del Jefe del hogar? 1 - Ahorros o ingresos propios? LJ 2 - Del compañero (a) ? ( 2 - Préstamos empresa donde trabaja? L_ 3 a 2 - De ambos? 3 - Regalo empresa donde trabaja? L_J 4 - De Uds, y otra familia que vive en la - Regalo de familiares o amigos? LJ 5 a.26 misma case? 4 - Préstamo de familiares o amigos Li 6 siga - De Uds, y de un familiar que vive NA ( )8 en otra casa? ( ) 5 Marque 1 en el que considere - De Uds. y de otra persona? 6 Pase el más importante. Y 2 en la NA8- segunde categor(a en Importancia NA )8 21 A NO PROPIETARIOS 25. Y a esos familiares o amigos tuvieron que: - Pagarles todo pero sin intereses? ( ) 1 18. Entonces la casa es de: O les pagaron con intereses? ( >2 - Un familiar que vive en la casa? ( )1 - O solo les pagaron algo? ( ) 3 - Un familiar que vive en la zona? ( ) 2 - O les pagaron con trabajo o ayudas? ( ) 4 - Un familiar que vive fuera de la zona? ( ) 3 - No les pagaron nada? ( ) 5 - Otra persona que vive en la casa? ( )4 NA 1 18 - Otra persona que vive en la zona? ( 1 5 - Otra persona que vive fuera de la zona? ( ) 6 26. Y esos familiares o amigos, en general, tenían una situa- - Otro, quién? ción econ6mica: NA ( ) 8 - Peor? ( )1 - Igual?d ( )2 19. Y cuánto paga por alquiler al mes? - Mejor que la de Uds.? 1 ) 3 - Nada NA ( >8 - S/mes _ LJLLLJ NA ( )8 20. En general el dueño de la casa tiene una situación econó. 27. Desde que empezaron a construir o arreglar o a mejorar mica: la casa con qué personas lo han hecho? - Peor? ( )1 No 1 Si 2 - Igual? ó 1 2 - Solo con miembros del hogar? 1 ( 1 2 Li - Mejor que la de Uds.? ( ) 3 - Con obreros contratados? ( ) 1 ( ) 2 LJ NA 1 )8 - Con amigos sin pagarles? ( >1 )2 L.iJ - Con familiares sin pagarles? ( 1 ( ) 2 Li_j - Con amigos pagándoles? 1 ( )2 LJ - Con familiares pagándoles 1 ( ) 2-.L 20a, De los cuartos de la casa cuántos tienen Uds. en arriendo? NA ( >8 No. de cuartos Pase NA ( )8 a A PROPIETARIOS 28. Desde que el ICT comenzó el proyecto de relleno de ca- ¡les y de la zona han tenido Uds. que hacer una mejora o cambio a la casa? 21. Cómo consiguieron la casa: - Consiguieron el solar y la construyeron? ( )1 siga - NO ( )1 - 30 - La compraron ya construida? ( )2 - Si ( )>2 -29 - Compraron las mejoras? ( >3 a 24 NA ( >8 - La heredaron ¿ se la dieron? ( 14 NA( 8 29. Por qué tuvieron que hacer esos cambios o mejoras? 22. Cómo consigueron el solar: - Se hundió la casa? ( ) 1 - Lo compraron? ( 1 1 - Tenían planeado hacerlo? ( ) 2 - Se lo regalaron? 1 ) 2 - Para aprovechar y mejorar la casa? ( ) 3 - Lo invadieron? 1 ) 3 - Otro, cuál? ( ) NA 1 >8 NA ( )8 23. Rellenó el solar antes de construir la casa? 30. Y desde que empezó el proyecto del ICT y están rellenan- do y haciendo calles, su casa: - NO ( 41 - Ha mejorado? 1 )1 - Si ( )2 - Sigue igual? ( 2 e NA 1 )8 - O está peor? ( )3 -síga NA ) 8 40. Cuánto reciben al mes por el alquiler? 31. Por qué dice Ud. que está peor? (MARQUE HASTA DOS) -Nada ( - Barro . Inundación ¡ 1 NA { )8 - Casa hundida Li 2 - Canales obstruyeron vía LJ 3 - Perdió espacio an la casa L_i 4 41. Y en grneral la condición económica de esos parientes y - Otro, qué? L-1 amigos es: L-L_l - peor? ( )1 NA ( )8 - igual? )2 - o mejor que la de Uds.? ( )3 NA ( )8 32. Uds. todavía están pagando la plata que consiguieron pa- CAPITULO V ra la casa, o ya la pagaron toda? RECURSOS DEL HOGAR EN SITUACIONES EXTREMAS - Todavía la están pagando ( ) 1 - Ya la pagaron toda ( ) 2 NA ( ) 8 42. En los últimos cinco (5) años Ud. o algún miembro de su hogar ha estado gravemente enfermo? 33. Cuál papel de propiedad de la casa tienen Uds.? - NO ( )1 A44 - Ninguno ( 1i )2 SIGA - Carta o papel o recibo? ( ) 2 - Promesa (minuta) de compra.venta ( ) 3 - Documento del ICT 4 43. ué hizo Ud. o a quién recurrió para pagar lo que nece- - Escritura Notaría < 5 sitó? - Otro, cuál? ._(_) NA ( )8 34. Aproximadamente cuánto cree Ud. que vale el alquiler mensual de esta casa hcy en dia? NA ( )8 S 1 í L.L.i L NA ( 8 44. En los últimos cinco (5) años murió algún miembro de su hogar? - NO )1 -A46 35. Y cuánto cree que puede valer esta casa hoy en día? - si )2 -SIG 45. Qué hizo Ud. o a quién recurrió par^a pagar lo que le co- .. 1 ibraron por el entierro? NA ( >8 36. Alquilan Uds. a otra persona o familia algún cuarto o lo- NA ( )8 cal de esta casa? - NO ( ) 1 A 38 46. En los últimos cinco (5) años ha nacido algún niño en es- - Si ( )2 -SIGA te hogar? - NO ( 11 -A48 37. Y por cuánto al mes? - Si 2 -SIG S _ I i ~J NA ( ) 8 47. Qué hizo Ud, o a quién recurrió para pagar los gastos del parto? 38. Los miembros de su hogar, que viven aquí, son dueños de otra u otras casas? - NO ( )1 AA4 - Si ( -)- Queda(n) solo en la zona? ( 2 NA 1 >8 Queda(n) solo fuera de la zona? ( 3 SIGA Quedan en ambos? ( )4 NA( 8 48. Cuando Uds. han estado en una situación económica di- 39. En esa(s) casa(s) viven: fícil, sin trabajo ni dinero, cómo han hecho para solucio- - Solo familiares? 1 narla o a quién han acudido para que les ayude? - Solo amigos? 1 ) 2 - Familiares y amigos? ( ) 3 - Otras personas? ( > 4 NA ( )8 A REGISTRO DE TODOS LOS MIEMBROS DEL HOGAR Comenzando con el JEFE, Digame Sexo E,d¿d Dígame si TEMPORALES MAYORES DE 10 AÑOS registre el nombre de todas cuál es el !Ps personas del hogar que parenteco vive aqui; durante e¡ 0ítimo mes han con el -H 1 Y cuál es la Y el hogar En ls umitos Cual ha sido la actividad vivido en esta vivienda JEFE: - 2 temporal? 1 razón más de, 5 alnos cuan. más irroortante que importante está en una tos años na ha hecho en el último mes: Despuês dol JEFE, registre o perma- para qué. situación vIvIdo aca? e' nombre de la esposa nente? 2 viva acá? - Mejor? 1 Trabajar 1 o compañera; luego los - Igual ? 2 - años 1 Trab. y Estud. 2 hijos a hijas por orden de - Peor? 3 -3 a 4 2 Hogar y Trab. 3 edad; continúe con los -1 a 2 3 Buscar Trabajo 4 parientes y, por último non Que la -Menos Hogar 5 los no parentes de Uds. que 1 4 Estudio 6 Jubilado 7 Otro, qué? 8 1 2 3 4 5 6 7 8 9 01 T ) 1 siga Si código 1, 2, 3 6 4 pase 01 ________ _L_J L-JJ P ( ) 2 ag L-J Li al cuadro 8 esta persona. 02 T )1 siga Si código 1, 2, 3 6 4 pase 02 1..t_J P ) 2a LJ LJ al cuadro B esta persona, T ( 1 siga Si código 1, 2, 3 6 4 pase 03 L_J L_LJ P ( ) 2a9 Li L__J l- al cuadro B esta persona, 04 T >1 siga Si código 1, 2, 3 6 4 pase _ L-L_J P ) 2 a9 al cuadro B esta persona. 0sT ( 1 siga Si código 1, 2, 3 6 4 pase LiL 1- P ( 1 2 a9 L Li Li acuadro B esta persona. 06 T 1 siga Si código 1, 2, 3 6 4 pase L_ L_L_ P 1 ) 2 a9 Li Li L al cuadro B esta persona. T 1 siga Si código 1, 2, 3 6 4 pase 07 L_ i ) 2 3 9 - L_j L_i al cuadro B esta persona. 08 T 1 siga Si código 1, 2, 3 6 4 pase L L-1J P 1 2 a9 -i Li L al cuadro B esta persona, 09 TI 1 siga Si código 1, 2, 3 6 4 pase 0L_J _JJ P 2 a9 L-J Li Li alcuadroBestapersona, ¡ T ) 1 siga Si código 1, 2, 3 6 4 pase 10 ._] L.J.J P 1 2a9 L Li L- al cuadro B esta persona, T 1 1 1 siga Si código 1, 2, 3 6 4 pase L L PLi1) 2 ag L_ Li Li al cuadro B esta persona. 12T 1 siga Si código 1, 2, 3 6 4 pase Li L-.-L P ( 1 2 a9 Li Li L al cuadro B esta persona. T 1 siga Si código 1, 2, 3 6 4 pase 3L LJ P ( 2 a9 Li L L- al cuadro a esta persona. 14 T 1 siga Si código 1, 2, 3 6 4 pase L _ - P ( 1 2 a9 L. L_ L i cuadro B esta persona. T 1 siga Si código 1, 2, 3 6 4 pase 15 L. .L JJ P ( 1 2 ag Li L_ Li al cuadro B esta persona, 16T 1 siga Si código 1, 2, 3 6 4 pase Li L_L.J P ( ) 2 a9 Li Li Li a cuadro B esta persona, 17 TI 1 siga Si código 1, 2, 3 6 4 pase L_J t_ 1i P( 2 a9 L Li LJ al cuadro B esta persona. 18 T 1 siga Si código 1, 2, 3 6 4 pase Li L_1_J. P ) 2 s9 LJ LJ Li al cuadro B esta persona. 19 T ( ) 1 siga Si código 1, 2, 3 6 4 pase Li '-.J-.J P 1 ) 2a9 Li L L- al cuadro B esta persona. TI ) 1 siga Si código 1, 2, 3 6 4 pase 20 L_ Lt_J PI( ) 2 a9 Lj LJ Li al cuadro B esta persona. T ) 1 siga Si código 1, 2, 3 6 4 pase 2-_J1 l-.__ P 12 a9 [_ L l cuadro B esta persona. T ) 1 siga Si código 1, 2, 3 6 4 pese Li L-.L_J P ( 1 2 a Li Li Li al cuadro B esta persona, TI ) 1 siga Si código 1, 2, 3 6 4 pase L _ J_ J P 1 1 2 a Li Li Li al cuadro B esta persona, TRABAJADORES DEL HOGAR Registre el nombre y el número En el trabajo esta persona es: Durante los En promedio Cuántas horas Cuánto ganó el Ganó algo por Cuánto ganó........... el Está afliado al ISS de orden de todas las personas rabaaor-obrero? 1 ultimos cuántos di¡as al día trabajó mes pasado........ en trabajos distintos al mejor mes y el peor o a Cajas? .-ayudante familiar? 2 12 meses a la semana . . . . - . . su trabajo principal? principal? mes en los que tienen codigo 1, 2, 3 6 4 en - empleado? 3 cuántos trabajó el en promedio? mes -s lo1 la columna 9 del cuadro A. -independiente? 4 meses último mes? ultmos 12 meses? S1 -dueño? 6 trabajó? -NO ( )2 -b.usca trabajo? 6 1 2 3 4 5 6 78 9 $ si 1L. Peor L LLL... si 1 NO 0 ) O Mejor LJ.L±LJ NO [] 2 $ si IJ.L.. Peor L JiL si [ 1 NO ( ) 0 Mejor L J J NO [12 L _ s i L . L . Peo r L L L . . . i s i L 1 L_J--L $ L.' ILLI.JI NO > 0 Mejor L 1 1 1 i NO []2 Si ¡jjjj Peor L I L L si E 1 NO ( ) 0 Mejor L 1 1i 1 1 NO [12 C OTROS JEFES DE HOGAR QUE HABITAN EN LA MISMA VIVIENDA Dígame el nombre de todos Cuál es el parentesco entre el Cuántas personas del hogar de La situación económica de Uds. los ayudan: Y ellos les ayudan a Uds. los JEFES JEFE de este hogar con el ........................ viven en esta hogar de........................ - Con dinero? 1 - Con dinero? 1 de los otros hogares hogar de ........................? vivienda? en relación con la de Uds. es. - Frecuentemente - Frecuentemente con que habitan en - Mejor? 1 con tareas domésticas? 2 tareas domésticas? 2 esta misma vivienda. - Igual? 2 - Cuidando niños? 3 - Cuidando niños? 3 - o Peor? 3 - Comida y otros? 4 - Comida y otros? 4 1 2 3 4 5 6 No. LJI ] Li Li Li No. L. LJ Li Li No. LJ L_J L_ D _MIEMBROS DEL HOGAR TEMPORALMENTE AUSENTES Dígame el nombre de todos los Cuál es la relación Cuántos Dónde está Con quién vive? Y en el hogar Y que está haciendo allá? .... -.......... les Y nandarán dinero miembros de este hogar que de ...................... años viviendo? Con parientes 1-~ donde vive tiene 1 mandó dinero el mes PASADO? estén temporalmente por con el Jefe de este H - 1 tiene? -Zona? 1 Con amigos 2-- una situación -Labores? 3 fuera y que no hayan formado hogar? M = 2 -Cartagena? 2 Solo 3 - Mejur? 1 -Vive allá? 4 -NO -NO otro hogar distinto a éste -Otra ciudad? 3 Sitio de trabajo 4 A8 - Igual? 2 -Buscando trabajo? 5 --sli3 -Venezuela? 4 Ejército 5 - O peor? 3 -Otro? 6 ' . -Otro 7 5 O tro 6 Que ja de Uds ? Cuánto? Cuanto? 1 2 3 4 5 6 7 8 9 - 10 L_J iI _- J L-i L_ NO ( NO ( ) si ( --$SI ( )_$ L_J w L L_ L-J L_ NO() NO I L_Ji LiLiL NOII NO i s ·쉔 7 E TRANSFERENCIAS MONETARIAS DESDE ESTE HOGAR ý- 17 ýý w -- -~- - 1 En el Utirno Lo mandaron En lo que va En promedio Y el mes Para corresponder Para que la gaste Ese dinero lo C6mo empezó arlo Uds. la han corno: del año, cuán- Cuánto la rnan pasado cuanto por esta ayuda, en qué cosas le mandan a ... esta ayudat mandado dinero Préstamo tas veces le dan al mes? le enviaron? ....... 1 ............ los manda a Uds., esa ........... 1 ........ a, .................. 7 pagadero han mandado devuelve : plat3? 0 al hogar en -Uc!s. ,crnarcn la cuando se dinero? - El dinero? 1 -Ayuda gral, 1 generai? -iueda ? 2 iniciativa? 1 -Sí 1 réstamo - Les trabaja? 2 -Educ3ci<Sn 2 -Ellos p.dieron? 2 -No 2 con plazo - Ambas cosas73 -Alimento 3 -Individuo 1 -Otro, qui? _fijo? 3 -Vivienda 4 ~Hogar 2 Otro qué? Otro, qué? -Otr- qué? - 2 3 4 6 8 9 Si ( ) 1 - NADA 0 NO 1 2 S Li NADA )0 NO 2 S L-L-I-L-J 1 Li Si i NADA )0 NO 1 27 S ! 1 1 1 Li si 1 - NADA 0 NO 2,p- Li S L-i L-i si 1 - NADA 0 NO 27- S LJ-L 1 1 SI 1 - NADA )0 NO 2 L-1 Ll 1 1 1 1 S Li SI NADA 0 NO 2 1 S 1 1 1 1 Si J NADA 0 NO( )2 J S 1 1 1 1 Si NADA 0 NO( 2 S 1 1 1 Li si NADA 0 NO 2 S 1 1 si 1 NADA )0 NO 2 Li 1 S 1 1 1 i Li L-i si 1 NADA 0 NO )2---- S 1 1 Li Li NADA )0 NO 1 2 S 1 si NADA )0 NO 2 S 1 1 Si i NADA 0 NO 2 S L-L-L-L-J Li L-i SI 1 NADA )0 NO í 2 L-J L-i L-1 L-J SI J NADA 0 NO )27- IL-J J S 1 1 L-i L-1 si 1 - NADA )0 NO 2.p- S 1 1 1 1 si i - NADA )0 NO( )27, S 1, 1 ! si 1 1 - NADA ) 0 NO )27,-- 1 J S i 1 1 1 si 1 NADA )0 NO t 1 2 Li L-1 Li si NADA )0 NO 1 12 J S 1 ' 1 ; i si ( 11 1 NADA )0 NO( )2 L-J S 1 1 1 si NADA )0 NO 2 -n;» S 1 i 1 1 si NADA )0 NO 1 L_J IADA 0 J Li L-i LJ Li -2 TRANSFERENCIAS MONETARIAS HACIA ESTE HOGAR En el último El (ella) se los En lo que va En promedio, Y el mes Para corresponder Para que la gaten Ese dinero el (ll Cómo empezó esta al3o Uds. han mandan coro: del año cuán to les pasado cuánto por esta ayuda, en qué cosas les lomanda a alguien ayuda: -Recalo? 1 recibido dinero -Preitamo cuantas veces m3nda al les mandó Uds: m3nda él (ella) en especial o al de ellos? png::dero ha rand do mes? a Uds.? la Plata? hogar en general? ouedon 2 dinero? -Dan dinero? 1 -Ayuda gral. 1 -Préstamo -Le trabajan? 2 -Educación 2 . 1 con plazo -Dinero -t-trabajo?3 -Alimentos 3 -IndividUo 1 -Ellos propusteron2 -No 2 fio7 3 -Otro, qué? -Vivienda 4 -Hogar 2 -Otro, qué? 3 -Otro, qué? -Otro, qué? 1 2 3 4 5 6 7 8 9 SI 1 - NADA ( )0 NO) 2 Li _ LJJJJ L_ L_J si ) 11- NADA ( )0 NO 12r L- LiLiLu L-J L2 1 _ _i si ()1- NADA ()0 NO 127 Lii iLI.J L Li L J si ( 1 -- NADA ( >0 NO ( 27 Li > I > L__J L_J L_ L Si (11- NADA> 0 NO( )27 Li Li L__J Si(1 - NADA ( >0 NO (27.i L0_.J ___i L i Li L_J Li Li Si 1- NADA( 0 NO( )2 , Li LLLIJ L Li L_ Li si ) 1 - NADA ( )0 NO ( )2 7 Li LLLLLJ Li Li Li sI ( 1- NADA ( )0 NOI 2- Li |_| _ _ L > L_ Li L_ L_ Si 1 11- NADA ( ) 0 NOI_)2-, L_ _ JLj i Li Li Li L_ si ) 11- NADA ( >0 NO ( 2)72 Li L __ L Li_1 si í 11- NADA ( )0 NO ( 2·-:;:----_ _JLL_ Li LJ Si 1 11- NADA ( >0 NO )2-2- iL Li L si ()1- NADA ( >0 NO (27 Li Li si 1 - NADA 1 )0 NO __2 _ _L_J sl 1 -. NADA 1 )0 NO( )27 Li Li Jtj L_J si 1 - NADA ( >0 NO l2 L Li L L L_J SI 111- NADA ( >0 NO 1 1 27 Li LLJLi 13 Li L si 1 -- NADA (0 NO )2 - Li Li2 .J L_J Li L_ SI 1 1- NADA 1 )0 NO ( ) 2 Li Li Li L-J síí( )1 --NADA> 0 NO( ) 2 L Li Li L Li si ( 1 1- NADA( 0 NO (2 - [L ._J LJ L-J LiJ si 1 NADA (>0 si 1 NADA ) 0 NOI ) 2 L- Li Li LiJ Si NADA (0 NOI 27 iL iL E-3 TRANSFERENCIAS EN ESPECIE DESDE ESTE HOGAR Cada cuánto tiempoEsta ayuda........ Como se empezó En el último a Y el mes Y qué fue lo En total se la mandan a al- esta ayuda? alo Uds. le manden Uds, pasado Uds. que mandaron? cuánto vale quien en especial, enviaon a - Menos que -Us oa enviaron ......... - mensual 1 enviaron? -Comida 1 aproximadamente a al hogar en ge- ron la ni- articuos como - Mensual 2 -Ropa 2 'o que enviaron? neral? ciativa 1 comida, ropa, - S 1 -Libros libros escolares, - 5 meses 3 -No 2 escolares 3 -Individuo 1 -Eilos pro- ollas, etc.? 6 año -Artícuios -Hogar 2 pusieron 2 -S( 1 1 a hogar 4 2 psron 2 -No 2 - Más de -Otro, qué? 1 ano 5- 1 2 3 4 5 6 7 si 1 1 - Jsi1 11- D 0n S NO ) 2 _J NO{ 2 a6 1 2 3 4 LLL,LJ L-J L_J si < 1-- jsi) 1- DDE E NO( 2 L_ NOI 1 2 36 1 2 3 4 L_J Li Si 1 - si( 11_ ELDE s NO( 127 i NOI 12a6 1 2 3 4 L L.J Sil 1 - si i 1- DE D S NO( 2 7 Li NO) 2a6 1 2 3 4 L L Li sií 11- síí)1- DDDD NO 12;" L_J NOI 1 2 a6 1 2 3 4 Li Li si < - si 1 NO NO 2 a6 1 2 3 4 L_ L sI ) 11- sI 1 11- DDDD S NO )l2 -. L_J NO ( 2a6 1 2 3 4 Li Li si 11- si 11- DDDD S NO( 12-- L_j NOI 2a6 1 2 3 4 Li Li SI t 11- si t- DDE LIjs NO 1 2 L_J NO( l2 aQ 1 2 3 4L Li SI 1 11- si 1 - E L D s NO 2 NO i2a6 1 2 3 4 L_LJ- si ( ) 1 --si ( )1 - s NO 2_ O 1 2a6 1 2 3 4 Li L Si 1 Si NO( ) 2-rL_ N 0 : 2 a6 1 2 3 4L-L_ si 1 11- Li SI 11- DDE 1 NO( )27¿_j NO 12 a6 1 2 3 4 sil 11- si >1_- E E El í NO 2 NO t 12 a6 1 2 3 4 Li L Si ( 1 --SI ( )1 - NO )27- L_J NO 2 a6 1 2 3 4 L si 1 1 - si 11- DEED s NO( i2 L_J NOI 2 a6 1 2 3 4 WLL L LJ Li si 1 - si 1-1- ED ED 0 s NO 127 L_i NO( 2a6 1 2 3 4 W,Li Li Li Sil 11-- si í 1- DE E sEl NO( 27'_ NO( 12a6 1 2 3 4 LL.L.J LJ Si 1 1-- Li i 1 DEDE s _ _ NO 12 LJ NOI )2a6 1 2 3 4 LLL,..L,i Li Li Sl 1 ) 1 -...LIDE- E E] s NOI 12L_ NO ( 12a6 1 2 3 4 i.LLJ LJ Li sil 11- L i ( 11- E LID L L NO 12- [ NO( )2a6 1 2 3 4 WLLL,,L_ _ si - si 11- ELIE 11 LL NOI )2- L_ NOI 12a6 1 2 3 4 [ si ( 1- si 11- D L LI s NOl2- Li _ NO 12a6 1 2 3 4 L i,,,j Li Li si )1- si l 1- E LI] n S NO( 12- Li NO( 12a6 1 2 3 4 LL,L,J_J Li Li si 1 11- si < )1- 1 DE s NOI í2-,-_ NOI )2a6 1 2 3 4 L-L-i_L-1 Li Li si< 11- si 1 1- E l DE S NO( )2-- Li NO( 12a6 1 2 3 4 L Li sI11 ______ E-4 TRANSFERENCIAS EN ESPECIE HACIA EL HOGAR En el (ltimo Cada cuanto tiem- y el año Y que fue lo que Y cuánto vale Esta ayuda la Cómo empezó es- enviado a Uds cosas para Ud. ? pasado les enviaron? aproximada- da.................. a taayuda? * -Menos que enviaron -Comida 1 mente lo que alguien en artículos com rrensual 1 -Ropa 2 enviaron ? especial o e- -Uds. propu- ropa, libros, -VensuDi 2 -Libros sieron? 1 comida, cllas, -2 a 5 rreses 3 3 hogaren etc.? -6 meses a -S ( í e l general? -Ellos p unaio 4 -No ( )2 -Artículos dieron? 2 -S( 1 -Más de 1 año 6 hogar 4 -Individuo 1 -No 2 -Otro, qué? -Hogar 2 -Otro, qué? 1 2 3 4 5 6 7 si ( t- Si 1 -- []EEE S_ NO( 2 Li NO) 2 a6 1 2 3 4 {__[ Li Si 1 -- EEEE n- NO) 2. ;L 1 2 3 4 L LJ si( 11- si 1 - EEEE S NO 127 L_ NO ) 2 a6 1 2 3 4 L_L Li si 1 -- sI ( 1 - E S EEl NO( )27 Li_ NO( 2 a6 1 2 3 4 _LJ L-i SiC 1-- si( 1 -- ED E n $ NO )27 Li NO( 2 a6 1 2 3 4 __L L_ Si 1 1- SI 1 -- E n3 E S NO( 27 Li NO) 2 a6 1 2 3 4 Li __ si 1 - si )1 - - E E S NO )2 . L_J NO( 2 a6 1 2 3 4 Li Li si( 1- SI 1 - E E n n NO( 2 - Li NO) 2a6 1 2 3 4Li Si 1 -- si (1- EEEL] S_ NO )2 7 Li_ NO( 2 a6 1 2 3 4 L Li sil 1-- si 1 -- EEEE S NO( )2--- NO ) 2a6 1 2 3 4 _J Li si 1 - si 1 -- LIEEE n _ NO) 2.. Li NO 2 a6 1 2 3 4 Li Li si 1 - si ( - oE E $ NO( )2 - Li NO )2 a6 1 2 3 4 LJ Li SI ( 1- Si 1 - EEEE S NO( 2 -w Li NO )2 a6 1 2 3 4 L_L iL L_ Li si 1 - sí 1 i1- E ] S 0_El NO( (2)s2 Li NO) 2 a6 1 2 3 4 LLJUJ Li Li Sil 11- si 1 -- EEEE __S NO) 2 Li NO) 2 a6 1 2 3 4 LLLUJ Li Li sil 1 Si E-.- E E E S - NO( )27 _J NO ) 2 a6 1 2 3 4 LLL! J L Li si ( 1- si 1 11- [ E]E S - NO ( r Li NO ) 2 a6 1 2 34 LLLJ Li Li SI ( 1 -- SI1 -- EE E S NO )27- L_J NOI 2a6 1 2 3 4 L.LLii Li Li si ( )- si 1 - j [ ]E S NO )27 Li NO) 2 a6 1 2 3 4 L..-L Li SI í 5 i1 it - E E E: ES NO( 12 , Li NO( 12a6 1 2 3 4 Li Li si ) )1 si< 1- o n E:] S - Sl 1 11- SilO1 2a6 ___ NOI i2.- Li NOI )2a6 1 2 3 4 Li Li Si (t- sil 11- L]E0 SD NOI l2- L_J NOI )236 1 2 3 4 1Li Li Si1- s 11+1- E E E n S- NOI )2-7. Li NO( )2a6 1 2 3 4 Li si í It - si 1 11- E E E S NO' l2.., Li NOl 12a6 1 2 3 4 L..) __j SI l1- si 1 11- EE E S NO 1i2 NOI )2a6 1 2 3 4 Li si 1- si 1- E E S NOI l2-- L_ NO( 12 a6 1 2 3 4 I Li Li CAPITULO VI 1 21 1 1 CAPITULO ViI OTROS INGRESOS DISTINTOS A LOS DE TRABAJO NEGOCIOS DEL HOGAR Registre en este Capítulo los ingresos que recibió el 12. Este hogar tiene un terreno, rosa o parcela para la pro- hogar o las personas de este hogar diferentes a los duccion agrícola? que recibieron por el trabajo que realizaron durante - No ( 1 A 14 el último mes. - sí 2 SIGA Durante el mes pasado algún miembro de este hogar recibió dinero por: 13. En la actualidad tiene sembrado algún cultivo? 1. Cesantías? -No ) 1 -No ) - Sí 1 )2 -Sí 1 -- Cuánto? S L 1 L 1 i ( )8 2. Pensiones o jubilaciones? - No 1 ) 14. Este hogar tiene algún negocio de su propiedad? -Sí ( 1 -- Cuánto? S i i L 1 1 - No 1 -A 16 3. Becas, Medias becas o ayudas para el estudio? - Sí ( ) Dónde queda? - No 1 )0 -Si ( ) -- Cuánto? SIJi 1 l J J - En la casa 2 - En lazona ( )3 - En Cartagena 1 ) 4 4. Concepto de Subsidio Familiar, de transporte o de otros - Otro sitio ( 1 5 Subsidios? 1 )8 - No ( )O -Sí 1 ) -+- Cuánto? S I L 1 1 l 5. Arriendo de piezas, casas, tierra, etc.? 15, Oué hacen en ese negocio? - No )0 _ -Sí 1 1 -+- Cuánto SI 1 6. Venta de animales como cerdos, gallinas, etc.? - No 1 1 C $CAPITULO ViI -Sí- Cuánto? S PROPIEDAD DE ANIMALES 7. Dineros enviados por parientes? o amigos? 16. Tienen Uds. cerdos en la vivienda? - No 1 )0 -Sí 1 ) -+- Cuánto? Si I 1 1 1 J - No ( 1 - Sí ( ) Cuántos? LL_J 8. Empeño de algún bien del hogar? - No ( )o 17. Tienen pavos, gallinas o gallos? - si* ) - Cuánto? S 1 1 1 1 [--l - No 1 ) - Sí 1 ) Cuántos? L . I.L 9. Venta de artículos producidos en una parcela o rosa del hogar, como yuca, ñame o arroz? - No ( )0 18, Tienen caballos o vacas? -Sí 1 1 --- Cuánto? S 1 1 1 1 l 1 - No ( ) - Sí 1 ) Cuántos L.L'LJ 10. Concepto de loterías, chance, rifas? - No ( )0 SIGA EN EL CAPITULO VIII -Sí ( ) -- Cuánto? S L_J1 1 L 1L i 11. Concepto de ayudas de instituciones del gobierno, de empresas privadas o entidades benéficas? - No ( )0 - Sí ( Qué? -- Cuánto? S 1 1 1 1 L -1 CAPITULO IX GASTOS DEL HOGAR 131 L I En el último mes cuanto gastaron Uds. en: Diario Semanal Mensual Para co acón unicamente 01 Combustible para cocinar? S S - S J 02 - Agua (Acueducto compra)? S s S L 1LLI. 03 - Servicio de luz? S S - SL- 04 - Cosas para la casa (Ollas, adornos, etc.)? $ $_ sL 05 - Lotería • Cnance? SS L 06 - Ropa? _ S - S L J 07 - Educación, (Mensualidad, libros, etc.)? S S S 08 - Medicina S S - S _____ L LL 09 - Alquiler s_ _ S - s _ L 1 1 10 - Materiales o arreglos para la vivienda? S Li 1 1A- Comida (Gastos de mercado - tiendas) S s S [ 118- Comida (por fuera de la casa) sS - S ÍA 1 1 12 - Cigarrillo - ron - cine? s S S 13 - Dio ayuda a parientes o amigos? 5 S $L 14 - Ayuda a parientes o amigos? S S - S 15 - Pagó deuda (préstamo, compra, etc.) S S - S _ L LiLUI 16 - Transporte? S S - S Li L 17 - Gasto en alguna otra cosa? SS_ Li _J Dígame para cada uno de los siguientes artículos, si lo tienen 27 - Ud. pudo ahorrar este mes? y cómo lo consiguieron: - No ( ) Cómo lo -Sí ( ) Cuánto? LLLLI ATRTICULO Consiado: 1 c idro ara: 28 - En promedio cuanto ahorra Ud. al mes? Regalo fam. 2 Propio? 1 Nada Regalo amigo3ý Regalo? 2 Prestado 4 Préstamo? 3 18 - Nevera? 4o L 2_- -7 29. Si de ahora en adelante Uds. se ganaran S4.000.oo más de lo que normalmente se ganan, en qué se lo gastarían? 19 - T.V.? NO Li LJ2 (REGISTRE LAS DOS MAS IMPORTANTES) 20 - Equipo de sonido? NIO 1 - i 2J 01 - Pagaría deudas L 2" - Radio Cassette NO( Li _L 02 - Mejoraría vivienda Li 03 - Pondría negocio Li 22 - Máquina de Coser 7 L U-- 04 - Ahorraría Li 23 - Licuadora? o L J L_J 05 - Electrodomésticos Li 06 - Educación Li 24 - Abanico? o Li LJ7 07 - Ropa Li 25 - Muebles de s -08-Comida Li comedor? NO! 2 LJ LJ-7 09 - Pondría servicios (luz) Li 10 -Otro, cuál L_J 26 - Otra cosa s0 A NO 1 2 27 Registre 1 en el más importante y 2 enel siguiente. cuál? 30. Y supongamos que el mes que viene (y solo el mes ese, no los meses siguientes) tuviera S40,000.oo, en qué se Nombre Supervisor los gastaría? (Registre solo las dos respuestas más importantes) 01 Pagaría deudas L__ 02 Mejoraría vivienda 03 Pondría negocio L 04 Ahorraría L-_ 05 Electrodomésticos L_ Observaciones Supervisor 06 Educación Li 07 Ropa L_J 08 Comida L 09 Pondría servicios (luz) Li 10 Otro, cuál? Li Registre 1 en la más importante y 2 en la siguiente. Nombre Encuestador Nombre Codificador Fecha _j_j Da Mes Duración de la entrevista Minutos Observaciones del Encuestador Observaciones Codificador - 86 - REFERENCES CITED IN THE TEXT ABBOTT, Susan (1976), "Full time Farmers and Week-end Wives: An Analysis of Altering Conjugal Roles" Journal of Marriage and the Family. 165-174., BAMBERGER, Michael, GONZALEZ POLIO, Edgardo and SAE-HAU, Umnuay (1981) "Evaluation of the First El Salvador Sites and Services Project." Urban and Regional Report No. 80-12. Urban and Regional Economics Division. The World Bank. BARNES, J.A. (1979), "Network Analysis: Orienting Notion, Regorous Technique or Substantive Field of Study?" in Paul Holland and Samuel Leinhardt (Eds) "Perspectives on Social Network Research" New York. Academic Press. BECKER, Gary (1974), "A Theory of Social Interactions," Journal of Political Economy," November-December. BECKER, Gary (1981), "A Treatise on the Family," Harvard University Press. BENDER, D. R. (1967), "A Refinement of the Concept of Household: Families, Coresidence and Domestic Functions." American Anthropologist, 69. BLA.U, Peter (1964), "Exchange and Power in Social Life" New York. Wiley. BOSSEN, L. (1980), "The Household as Economic Agent," Mimeo. University of Pittsburg. BOHANNAN, P. (1969), "Social Anthropology," Holt, Rinehart and Winston, New York. BURJA, Janet (1977), "Production, Property, Prostitution and Sexual Politics in Atu" Cahiers d'Etudes Africaines 65:13-39 BUTZ, William and STAN, Peter (1981), "The Microstructure of Malaysian Interhousehold Exchange Networks" UCLA Labor Economics Workshop. April 14. BUVONIC, Mayra, YOUSSEF, Nadia and VON ELM. Barbara (1978) "Women Headed Households: The Ignored Factor in Development Planning" International Center for the Study of Women, Paper prepared for the Agency for International Development. EMERSON, Richard (1976), "Social Exchange Theory" Annual Review of Sociology Vol. 2 335-362 HOMANS, George (1958), "Social Behaviour as Exchange" American Journal of Sociology. 62: 597-606 - 87 - KAPFERER, B, (1973), "So-ial Network and Conjugal Role in Urban Zambia: Towards a Reformulation of the Bott Hypothesis." in J. Boissevain and J. C. Mitchell (Eds) "Network Analysis: Studies in Human Interaction." The Hague. Mouton. KAUFMANN, Dani and LINDAUER, David (1980), "Basic Needs, Interhousehold Transfers and the Extended Family" Urban and Regional Report No. 80-15. Urban and Regional Economics Division. The World Bank. KAUFMANN, Dani (1981), "Social Interactions as a Strategy of Survival Among the Urban Poor: A Theory and Some Evidence." Doctoral dissertation, Harvard University. KEESING, F. (1958), Cultural Anthropology, Rinehart and Co., New York. KEMPER, R. (1974), "Family and Household Organization Among Tzintzumtazan Migrants in Mexico City," in W. Cornelius and F. Trueblood (Eds.), Latin American Urban Research. LEE (1969), "The search for an abortionist" Chicago University Press. LOMNITZ, Larissa (1977), "Networks and Marginality: Life in a Mexican Shantytown" Academic Press. New York. MEEKER, B. F. (1971), "Decisions and Exchange". American Sociological Review. 36:489-95. MITCHELL (1979), "Networks Algorthms and Analysis" in "Perspectives on Social Networks." Academic Press. New York. MOOCK, Joyce (1978), "The Content and Maintenance of Social Ties between Urban Migrants and their Home-Based Support Groups: The Maragoli Case." African Urban Studies. Number 3. Winter 1978-79. Michigan State University. MUKRAS and OUCHO (1982), "Migration Remittances and Rural Development in Kenya". Unpublished. Nairobi. MULLER, Maria (1975), "Actions and Interactions: Social Relationships in a Low-Income Housing Estate in Kitale, Kenya. Afrika-Studicentrum. Stationsplein. Leiden. Holland. MURILLO, Gabriel (1983), "Colombian International Labor Migration: Trends and Perspectives." UNESCO, Population Division. NELSON, Nici (1978), "Female-centered Families: Changing Patterns of Marriage and Family Among Buzaa Brewers of Mathare Valley." African Urban Studies. Number 3. Winter 1978-79. Michigan State University. NELSON, Nici (1979), "Women Must Help Each Other: The Operation of Personal Networks Among Buzaa Beer Brewers in Mathare Valley, Kenya." In "Women United, Women Divided: Comparative Studies of Ten Contemporary Cultures" Indiana University Press. 77-98. - 88 - NIEVES, Isobel (1979), "Household Arrangements and Multiple Jobs in San Salvador." Signs: Journal of Women in Culture and Society. Autumn 1978. Vol 5 No. 1 PEATTIE, Lisa (1969), "The View from the Barrio" Ann Arbor, University of Michigan Press. PEATTIE, Lisa (1981), "Marginal Settlements in Developing Countries" Annual Review of Sociology. 157-175. POLANYI, K (1957), "The Economy as an Instituted Process" In Trade and Market in the Early Empires. Edited by K. Polanyi and C. M. Arensburg and H. W. Pearson. New York. Free Press. REFORMA, Mila and OBUSAN, Ricci (1981), "Household networks and survival strategies among the urban poor: Monetary and non-monetary transfers among selected families in Tondo." Urban and Regional Report No. 81-22. Urban and Regional Economics Division. The World Bank. SAHLINS, Marchall (1968), "On the Sociology of Primitive Exchange." In The Relevance of Models for Social Anthropology. Edited by M. Banton. Monographs No. 1. London. Tavistock. SOLIEN DE GONZALEZ, N. L. (1960), "Household and Family in the Caribbean," Social and Economic Studies, 9. STACK, Carol (1976), "All Our Kin: Strategies For Survival in a Black Community" Harper and Row. New York. WEISNER, Thomas (1976), "The Structure of Sociability: Urban Migration and Urban Ties in Kenya." Urban Anthropology. 5: 199-223 URBAN AND REGIONAL ECONOMICS DIVISION (1980), "A Study of Income and Expenditure Patterns of Households in the Tondo Foreshore Area." Urban and Regional Economics Division. The World Bank. (Summary of a Report Prepared by the Research and Analysis Division. National Housing authority. Philippines. - 89 - REPORTS AVAILABLE ON THE RESEARCH PROJECT Working Papers 1. The Structure of Social Networks in the Zona Sur-Oriental of Cartagena - Michael Bamberger and Scott Parris, Report No. UDD-50. 2. Income Transfers and Urban Projects: Research Findings and Policy Issues - Daniel Kaufmann, Report No. UDD-56. 3. Survival Strategies and Support Networks: An Anthropological Perspective - Scott Parris, Report No. UDD-58. 4. Patterns of Income Formation and Expenditures Among the Urban Poor of Cartagena (Final Report - Project No. 672-57) - Michael Bamberger and Daniel Kaufmann, Report No. UDD-63. "Aspectos Metodologicos para Estudiar Patrones de Formacion de Ingresos y Gastos entre Familias Pobres Urbanas". Instituto SER de Investigacion, Bogota. (Detailed explanation of the research methodology, including the questionnaire, interview guide and code book.)