ASPECTS OF INTRACOMMUNITY RESIDENTIAL MOBILITY AMONG MEXICAN AMERICANS IN MICHIGAN Thesis for the Degree of M. A. MICHIGAN STATE UNIVERSITY MARGARET JOHNSON ADAMS 1970 - .g ,4. o ...’ —¢--_-_$qv~ofl.o”OO—aw ‘ a , .‘o PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE I JUN o 7w ] I I. 3:27 I MSU Is An Affirm-rive ActIorVEquel Opportunity Institution ABSTRACT ASPECTS OF INTRACOMMUNITY RESIDENTIAL MOBILITY AMONG MEXICAN AMERICANS IN MICHIGAN BY Margaret Johnson Adams The central problem of this thesis is the explana- tion of the process of residential mobility which is de- fined as intracommunity mobility. The problem is approached from the theoretical and empirical standpoints. Residen- tial mobility is conceptualized as the process through which a household attempts to balance its housing with its housing needs. Housing needs are defined as spatial require- ments. A paradigm of residential mobility is presented in which the following elements appear: family life cycle, housing requirement, dwelling space, housing supply and access to the housing market. The latter concept is three dimensional and refers to the position of a household in terms of financial resources both money and credit, informa- tion on the housing market and discrimination, social or demographic, which may limit or enhance its access. Margaret Johnson Adams Hypotheses are derived dealing with family life cycle, housing requirement, dwelling space, financial access and inertial influences on residential mobility which concern tenure status, duration of residence in com- munity and time in present house. Residential mobility is conceptualized in two parts: past mobility and Planned mobility. Past mobility is measured by the ratio of in- town moves to duration of residence. Planned mobility, the primary dependent variable, is operationalized as mover or stayer dependent upon whether the household anticipates a future move. I With the additional interest in extending the study of residential mobility to minority groups, these hypothe- ses are tested with survey data obtained from a sample drawn from the Mexican American population in Michigan out- side the Detroit SMSA. The resulting cross-tabluations are analyzed with the aid of a statistical measure of associa- tion, gamma. The Mexican American sample can be described as primarily urban and as having large households and moder- ate incomes. It is also predominantly home-owning rather than renting. The land contract proved to be an important method of home financing enabling low income households to achieve home ownership. Forty percent of the household ex- pected to move. The household expecting to move tends to be young, renting, in a small dwelling and to have a medium Margaret Johnson Adams income and high past mobility. The household expecting to stay tends to be older, homeowning, in a large dwelling and to have low past mobility and either low or high income. In general family life cycle as measured by age of head of household is consistently a strong independent var- iable relating to planned mobility. The relationship is influenced by household size; it is stronger in small house- holds and weaker in medium and large ones. Low past mobil- ity, owner tenure status, long duration of residence in community and long time in present house indicate, in des- cending order of importance, inertial influences on planned mobility. Neither household size alone nor household size controlling for dwelling space, representing housing adjustment were found to be related significantly to expec- ted mobility. Dwelling space alone has some influence. The higher the income of a household, the greater the fi- nancial access to the housing market through home owner- ship, financing with less risk and obtaining more dwelling space. ASPECTS OF INTRACOMMUNITY RESIDENTIAL MOBILITY AMONG MEXICAN AMERICANS IN MICHIGAN BY Margaret Johnson Adams A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Sociology 1970 ACKNOWLEDGMENTS I wish to express my appreciation to Professor Grafton D. Trout, Jr., for providing access to the data for this thesis and for his guidance and encouragement while pursuing the research. My thanks are also extended to Professors J. Allan Beegle and Denton Morrison for their helpful cOmments. Finally, I would like to thank my husband, Tom, my daughter, Sandra, and my son, Tom, for their patience, understanding and support during my graduate studies. ii TABLE OF CONTENTS ACKNOWLEDGEMENTS. . . . . . . . . LIST OF Chapter I. II. III. IV. V. TABLES AND FIGURES. . . . INTRODUCTION 0 O C O O O O THEORETICAL FRAMEWORK . . Conceptualization . . . A Paradigm of Residential Hypotheses. . . . . . . METHODS . . . . . . . . . Operationalization. . . Data Collection . . . . Method of Data Analysis ANALYSIS OF DATA. . . . . SoCial Characteristics of Testing of Hypotheses . Summary . . . . . . . . v-» CONCLUSIONS . . . . . . . BIBLIOGMPHY O O O O O O O O O O 0 iii 13 13 21 23 24 24 28 31 34 34 37 54 61 68 Table 1. 2. 12. 13. 14. 15. 16. 17. 18. LIST OF TABLES AND FIGURES A summary of Operationalization. . Planned mobility by past mobility in town of present residence . . . Planned mobility by past mobility by tenure Planned mobility by past mobility by age . Planned mobility by past mobility by house- hold size. . . . . . . . . Planned mobility by past mobility by income . Planned mobility by age of head of household. . . . . . . . . Planned mobility Planned mobility Planned mobility Planned mobility dwelling Planned mobility Tenure status by by age by household size. by age by income. by household size by household size space . . . . . , by dwelling space income. . . . . . Method of financing by income. . . Dwelling space by income . . . . . Planned mobility by tenure status. Planned mobility by duration of residence. Planned mobility by time in house. iv 39 40 41 41 43 45 45 47 48 49 50 51 51 53 53 55 Table 19. Planned mobility by time in house by tenure . . . . . . . . . . . Figure 1. Model of family life cycle . . . . . CHAPTER I INTRODUCTION American life is urban, and its population is mobile socially and geographically. The concern of this paper is with residential mobility, more specifically that occurring within communities. Intercommunity moves are primarily job oriented and dominated by the positive "pulls" of the new location whereas intracommunity moves are primarily housing oriented.l The process of intracommunity residential mobility has consequences in the urban social structure, the demo- graphic composition of the city, and for both the mobile and non-mobile households in terms of comfort, standard of liv- ing, access to urban facilities and opportunity for partici- pation in urban activities to mention just a few. As Rossi has indicated, "Basic research into residential mobility is of importance because mobility is one of the most important forces underlying changes in urban areas."2 1Nelson Foote, et al., Housing Choices and Constraints (New York: McGraw Hill Book Co., Inc., I960), P. 153. 2Peter H. Rossi, Wh Families Move (Glencoe, Illi- nois: The Free Press, 1955;, p. 2. In addition to the theoretical interests of sociolo- gists and other social scientists in this process of resi- dential mobility, there is practical justification for an investigation of this aspect of American life. There is a need to know the manner in which the process operates in and interacts with the situation of an increasing shortage of housing and a changing composition of housing supply. In the concluding pages of Why Families Move, Peter Rossi presents some guides to future research.3 Generally, these guides involve an extension of his methods to other populations, more intensive research on the relationships uncovered and the measurement of pertinent variables and an extension of research in residential mobility in new direc- tions such as spatial patterning and housing values. This thesis represents an attempt to follow the first named guide wherein Rossi suggests an extension of research not only to more representative populations but also to special types. such as minority groups.4 The focus will be on intracommun- ity residential mobility among Mexican Americans in Michigan. Therefore, the present study is not a replication but an extension of Rossi's work. An investigation of residential mobility among Mexican Americans is important for several reasons. Many members of this minority group were once migrants. In the 3Ibid., pp. 181-184. 41bid., p. 182. process of resettlement in Michigan, and.even afterward, housing is the most critical problem this group faces. In addition, the housing moves represent a series of adjust- ments which are perhaps more crucial.for this group than for more affluent populations having wider-choices. As a general introduction to the subject of intra- community residential mobility, a review of previous stud- ies is presented. Intracommunity mobility is commonly de- fined as intracounty mobility. The extent of residential mobility may be gauged by census data which indicate that during each of the past 15 years approximately 20 percent of the population has moved at least once during the pre- ceding year. This total residential mobility is composed one-third of intercounty, interstate and interregion moves and two-thirds of intracounty moves.5 .The annual proportion of local residential mobility varies from 12 to 14 percent of the total population and from 66 to 70 percent of all movers.6 Analyses of reasons for intracommunity moves indi- cate that approximately 60 percent of local moves are related to housing dissatisfaction and approximately 40 percent are 5Foote, Housing Choices, p. 159. 6Henry S. Shryock, Population Mobilit Within the United States (Chicago: Community and Famin Study Center, UniversIEy of Chicago, 1964), pp.-66-67. due to forced moves or are a "by—product of some other decision."7 A major study of residential mobility was made by Peter Rossi in the early 1950's.8 A.primary aim of this research.was to demonstrate the utility and rewards of the survey method for this subject area. Rossi approached the study of the process of residential mobility from three standpoints: area mobility, household mobility and factors in the decision to move. Rather than focusing on past mo- bility, this study defined mobility in terms of desires and plans for moving. Mobility is interpreted as the mech- anism by which housing is brought into adjustment with hous- ing requirements.9 Rossi's findings regarding area mobility indicate that while mobile areas contained more single households and childless families than stable areas it could not be interpreted that such household types are more mobile than others. Actually "all household types within mobile areas had higher mobility than their counterparts in more stable environments. . ." and the mobility rates of each area were were interpreted as a ". . .characteristic of the area 7Foote, Housing Choices, p. 154, and Rossi, Why Families Move, p. 135. 8Rossi, Why Families Move 91bid., p. 180. itself rather than a function of the household types which are to be found there."10 Full familes were found to be the most potentially mobile household type in all areas. Mobile households tended to be larger, younger and renters rather than owners.ll A Mobility Potential Index, based on age, household size and tenure, and a Complaints Index both correlated well with mobility inclination but not too well with each other. When combined, the two indexes accurately predicted 75 percent of the mobility inclined households. Dissatisfaction with dwelling space emerged as the most important complaint, and size was the most important criterion in evaluating prospec- tive residences. In conclusion Rossi interprets the meaning of resi- dential mobility and the reasons for extensive urban resi- dential shifting in the following manner: The findings of this study indicate the major function of mobility to be the process by which families adjust their housing to-the hous- ing needs that are generated by the shifts in family composition that accompany life cycle changes. Janet Abuelughod studied residential mobility among center-city residents of high quality dwellings in New York, loIbid., p. 44. llIbid., p. 71. lzIbid., p. 9. Philadelphia and Chicago. The aims of this research were to identify the characteristics of such inner city dwellers, to determine if and how this type of housing selected these residents and to indicate the housing values and mobility rates of this population.13 A knowledge of the sample characteristics is impor- tant in order to appreciate the findings of this study. If a husband, wife and minor children are considered a "typical" American family, the large majority of these center-city residents are atypical. Only 15 percent were families with minor children, and of these 65 percent had no children in school. The three most frequent household types were couples over 45 with no minor children at home, older individuals with no attachments and younger individuals with no attach- ments.14 All residential mobility studies have found current tenure status, i.e., renting or owning, to be a good predic- tor of mobility, but the relationship between these two var- iables was stronger than usual in the Abu-lughod sample. A mobility scale was constructed along three dimensions: length of time in previous dwelling (regardless of tenure held), length of time in current dwelling, and intentions of moving within two years.15 Because of the nature of the 13Janet Abu-lughod, "Appendix" in Foote, Housing Choices, pp. 387, 390. 14Ibid., pp. 391-392. 15Ibid., p. 418. mobility scale used, Abu-lughod claims "evidence of a two- way causal relationship between tenure and mobility. . . owners are less mobile because 'owning a house' makes them so. . ." and ". . .ownership itself selects out the least mobile elements of the population (mostly the immobile buy homes)."16 A clear relationship of stage of family life cycle and mobility was found with mobility decreasing through the progressive stages. The data indicated a strong relation- ship between age and mobility and between building type and 17 mobility. Unlike Rossi's study, no significant relation— ship between household size and mobility was found.18 Wendell Bell associated life style with movement to the suburbs.19 Among the three life styles presented, emphasizing familism, career or consumership, the familism orientation is considered to have an important influence in the choice of the suburbs as a place of residence. He concludes from his study of Chicago suburbs that a career emphasis has only small influence on residential mobility. 161bid., p. 419. 17Ibid., pp. 419-421. 18Ibid., p. 419. It was pointed out that there were few large families in the sample. 19Wendell Bell, "Social Choice, Life Style, and Suburban Residence," in The Suburban Sommunityed. by William M. Dobriner (New York?* G.P. Putnam's Sons, 1958), pp. 225-247. Gerald Leslie and Arthur Richardson examine resi- dential mobility with a combination of life cycle and career pattern variables.20 As did Rossi, they defined mobility as stated intention to move and used Rossi'erobility Poten- tial Index (age, household size and tenure status) to repre- sent the family life cycle influence. Influence of career pattern is measured by education, perceived class differ- ences, social mobility expectations, attitude toward present dwelling and toward present neighborhood. In delineating a path of greatest increments to correlations "of stated mobility intentions to succeeding independent variables," four variables.emerged as sufficient 'to account for the .76 correlation of all eight variables to mobility intentions.21 These were, in descending order of importance, social mobility expectations, perceived class differences, attitude toward house and education. The pre- dictive equation based on these four variables proved to be highly accurate without the inclusion of any life cycle variables.. The strong association of upward.social mobility with residential mobility indicated in this study is in contrast to Bell's findings. The authors believe this may 20Gerald R. Leslie and Arthur H. Richardson, "Life Cycle, Career Pattern and the Decision to Move," American zSociological Review, 26 (1961), 894-902. 21 Ibid., p. 897. be due to the different population involved, that of the suburbs by Bell and a residential subdivision by their own study. They find, however, no necessary inconsistency with Rossi's findings because the latter's sample was more heter- ogeneous than that drawn from the subdivision where the range in household type, age and tenure status was narrow.22 Leslie and Richardson conclude that both sets of factors are necessary to an explication of voluntary resi- dential mobility. A paradigm is offered which "assumes that both the need for more living space as the family increases in size and the need to adjust housing to changes in social status are potent forces inducing families to move. The push toward residential mobility would be greatest when the two forces act in concert and least when neither is operative."23 .Leslie and Richardson also present a general model of the process whereby the life cycle and career pattern variables enter into the decision to move-.-24 Complaints about present dwelling are posited as an intervening vari- able between the independent variables, stage of family life cycle and career pattern, and the dependent variable, resi- dential mobility. They point out that complaints are not 221bid., pp. 899-900. 231bid., p. 900. 24Ibid., p. 901. 10 merely a function of objective dwelling deficiencies but also of the perceived opportunities to ameliorate these con- ditions by moving. In the course of acknowledging that no claim of adequate empirical foundation for their model is made, the authors indicate that "the relevant studies to date have differed sufficiently in general purposes, in populations studied, and in methodological detail to make it possible that the differences in their findings are arti- facts thereof."25 A longitudinal approach to the study of moving is provided by Albert Chevan in the desire to specify the roles of "precedents and antecedents of a residential change. . . more fully than in the cross-sectional approach" particularly 26 Each in terms of its association with family life cycle. household was questioned on its residential history, work history, marital status history and family history. After obtaining the histories from his large sample of households with both husband and wife in their first marriage in the Philadelphia-Trenton SMSA, Cheven constructed a rate of moving which included both frequency of moves and time in 251bid., p. 902. 26Albert Chevan, "A Longitudinal Approach to the Study of Moving," Paper presented at the annual meeting of the Population Association of America, Boston, Mass., April 18-20, 1968, p. l. 11 27 Each family history was sub- in the residential area. divided into three-year exposure periods with each period indicating the number of moves made, time spent in the area and number of children born during the period. The findings from this study indicate a decline in rate of moving as the duration of marriage increases.28 Higher rates of moving occur when children are born during a period, but this rate declines "from one period to the next, indicating that the birth of a child in one period does not have the same effect as if it had occurred in a 29 The effect of the birth of children in previous period." one period is influenced by whether a move was made in the previous period. Chevan also finds a family life cycle effect on moving which is independent of the birth and growth of chil- dren. The housing adjustments of always.childless couples tended to be concentrated in the early years of marriage with the decline in rate of moving as marriage progressed being more gradual than the decline of couples with children.30 271bid., pp. 4-5. 281bid., p. 6. 291bid., p. 7. 3°Ibid., pp. 9-10. 12 Another article on residential mobility was that by Sabagh, Van Arsdol and Butler.31 In this case the approach was theoretical rather than empirical.. Voluntary residen- tial mobility is seen as having both structural and social psychological components which must be taken into consider- ation. The authors work from the assumption that "mobility determinants can be analyzed in terms of push-pull factors pertaining to origin and destination and frictional factors 32 A series of interrelated impeding or facilitating moves." propositions is presented which pertains to the four basic dimensions by which residential mobility is conceptualized. These are: (1) family life cycle and familism; (2) social mobility and social mobility aspirations; (3) residential environment and residential needs and aspirations; and (4) social and locality participation. The intervening fric- tional factors are availability of desired housing, infor- mation about residential opportunities and financial resources . 3 3 31George Sabagh, Maurice Van Arsdol, Jr., and Edgar W. Butler, "Some Determinants of Intrametropolitan Residen- tial Mobility: Conceptual Considerations," Social Forces 46 (September, 1969) 88-98. 321bid., p. 94. 33Ibid., passim. CHAPTER II THEORETICAL FRAMEWORK Conceptualization As indicated in the introduction, residential mobil- ity here refers to change of residence not involving change in community. This is not to say that intercommunity migra- tion, which is also part of the general mobility of American society, does not also contribute, as does residential mo- bility, to change in urban areas with ramifications in such spheres as the urban social structure, institutional social control, the viability of local organizations, the nature of interpersonal relationships, the demographic composition and the physical structure of the city. However, inter- community migration requires a different conceptualization and research design which in combination with that of resi- dential mobility would prove to be unwieldy if feasible. Furthermore, as indicated previously, residential mobility constitutes the majority of total population mobility. The topic will be approached from the standpoint of the movement of households rather than from an areal or spatial stand- point. Therefore, the unit of analysis will be the household. 13 14 The residential mobility process will be conceptual- ized in two parts: past mobility in the town of present residence and planned mobility. Past mobility is defined as the amount of residential shifting-which has taken place since arrival in the present community of residence. Planned mobility is defined as potential mobility in terms of plan- ning a future move. Therefore, planned mobility status will be either stable or mobile depending upon whether or not a move is anticipated. This conceptualization of current sta- tus has been used in several studies following the check on its validity by Rossi.34 Planned residential mobility is the primary dependent variable under consideration; past mobility in town of present residence is to be used as an independent variable. The effect of family life cycle on residential mo- bility will be investigated primarily.~ Essentially the idea of a family life cycle is that there is a succession of changes through which a typical family progresses from its inception to its dissolution. The various stages of this hypothetical family may be summarized as characteriz- ing an expanding family in the early years, a stable one in the middle years, and a contracting one in the later years 35 of its "life." The changing demands and resources of a 34See comments in Operationalization section of Chapter III. - 35See Foote, Housing Choices, pp. 97-118, for a more detailed description of the family life cycle stages. 15 family progressing through the cycle should be reflected in changing housing needs, housing adjustments and expectation of moving. Evidence of such a relationship has been pro- vided by several studies as indicated in the introduction. Before elaborating on the conceptualization, it is necessary to state how it is to be related to the Mexican- American population in Michigan. The approach to be used is that of developing a paradigm for the general case of residential mobility in American society. Hypotheses will be derived which apply in general also. The data analysis will contain an evaluation of whether and in what manner this ethnic group conforms or diverges from this concep- tualization. The notion of adjustment is central to the paradigm \ of planned residential mobility to be developed. The ele- _ - -.._ rum-..“ ments to be considered in this adjustment process are the stage in family life cycle, housing requirements, the pres- ent dwelling space, the access of the household to housing I and the situation in terms of housing supply. The stage of family life cycle reflects the composi- , tion of the household. Once marriage occurs and the house- I hold is established, the size is constant until the child- bearing stage is reached during which Size increases. The next stage, child-rearing, is represented by a plateau in terms of size. Then there begins a decrease in household size as children leave and become established outside the 16 family. After this period size is constant again until the family is dissolved by the death of one of the partners. Each successive stage also represents the increasing age of the head of the household. The cycle assumes the absence of three generational households. If these typical stages were represented along axes of size and time, the result would be a general increase in size through time up to a maximum during the child-rearing stage and a general decrease from there onward. The changes in size occurring over the course of the family life cycle are a major consideration in the adjust- ment process with which this thesis is concerned. The second element, housing requirement, refers to the spatial needs of the household.36 These requirements are determined by the size of the household and are at a maximum level during the child-rearing stage. The third element involved in the process of adjust- ment is the present dwelling space the household occupies. This refers to the physical space available within the resi- dential unit. The adjustment process is essentially an 36The author is, of course, aware that there are other needs beyond those of space which may influence a household's requirement such as desirability of neighbor- hood, location with reference to facilities and place of work, life style, social mobility, etc. However, the latter concerns will not be dealt with in this paper. 17 an attempt by the household to balance its present dwelling space with its housing requirements. The fourth element is the access of the household to housing. The concept of access is multidimensional. Three dimensions may be discerned concerning the amount of financial resources, information and discrimination. The access of a household to housing may be more free or more restricted according to its position on each of these three dimensions. Financial considerations, both with respect to income available and ability to obtain credit, affect not only the size and quality of housing obtained but also the rental or ownership basis of its possession by the house- hold. Access to the housing market also depends upon know- ledge of that market, and, therefore, on the information the household has at its disposal. The third dimension. refers to discrimination which may affect access. This includes any social or demographic characteristics of the household, e.g., racial, ethnic, size, etc., which would tend to limit or enhance its position in.terms of access. The final consideration in the adjustment process is the availability of housing. This is the physical supply of housing of all types. The four elements mentioned above Operate within this situation. Residential mobility is the process through which the household attempts to balance its housing with its hous- ing needs. Housing requirements reflect the composition of 18 the household and change as this composition changes. When a change in requirements is evident, the household attempts to adjust its housing to meet them. This involves an adjust- ment to larger or smaller dwelling space according to the new household size. The ability of the household to effect this balance is mediated by its access to housing, i.e., its financial and information resources and any impairment due to dis- crimination. These housing requirements and the ability of the household to enter the housing market operate within the context of the supply of housing extant. The effect of family life cycle on residential mobility is expected to be a reduction in the amount of mobility from early to late stages. The complete residen- tial history of a household would be expected to show that adjustments are more likely to take place in the earlier part of the history than during the later part although readjustments may occur toward the end also. This is due to the higher probability of household size changes during the early stages, the likelihood of increasing financial resources, a decline in experimental housing adjustments and the inertia associated with home ownership which is more likely to occur as time passes. The psychological and financial ties to an owned dwelling plus the greater ability of an owner as opposed to a renter to make structural spatial adjustments in the house contribute to this inertia. This 19 could also be considered as a component of the "axiom of inertia" proposed in intercommunity migration.37 The assump- tion is "that an individual's propensityrto move is a func- tion of, among other things, his length of residence in the community."38 A schematic representation of the trends in house- hold size, dwelling space and income which are to be expec- ted during the course of the family life cycle is found in Figure 1. The abscissa in the figure is the time dimension. The time periods (Tl through T6) represent stages in the family life cycle as follows: T : the initial 25person household formed upon marriage 1 and continuing until a child is born. T2: the child-bearing stage. T3: the child-rearing stage. T4: the child-launching stage. T5: the later 2-person household after the children are established elsewhere. T : the household is dissolved with the death of one of the partners. As is shown the dwelling space follows household size in T1' lags behind in T2 and the early part of T3. Once the maximum size is reached in T3, dwelling space is brought into adjustment. At T4 the dwelling space outruns 37Kenneth C. Land, "DUration of Residence and Pros- pective Migration: Further Evidence," Demography 6 (May, 1969) 133-140. 38 Ibid., p. 133. '20 _ mane... on me . .9 me me .e _ _ 7% I 32°.nt 29.3.6... >>On_ mEOOm “.0 mum—232 Eco—Eocze .. nwa onoxmmnoz _ _ Io... 21 the decreasing household size as adjustments are usually made only after the last child has left the household. Dur- ing T5 there is typically another delay in adjusting housing to the 2-person household again. At the-final stage, T6, another delay may also occur in adjusting housing. A Paradigm 9: Residential Mobility On the basis of the foregoing conceptualization, the following theoretical presentation is made. Postulates: I. A household evolves through a family life cycle in which the household size increases in the early stages and decreases in the late. II. Requirements for housing change as the stage of family life cycle changes. Definition: The housing requirement is determined by the size of the household. III. Over time, the household attempts to balance the housing space it occupies with its housing require- ment. Corollary: The longer the time span-the more-likely the equilibrium is reached. Assumptions: Ar-l. The adjustment process takes place within the avail- able supply of housing. It is assumed to be constant. 22 A-2. Income tends to increase with the age of the head of household. A-3. All households have equal information on the housing market. A-4. There is no blockage of access by discrimination.39 Propositions: P-l. Low past mobility is associated with the expectation of moving and vice versa. P-2. Expected mobility becomes more likely from early to middle family life cycle stages and then becomes less likely. P-3. Households with high housing requirements and low dwelling space expect to be mobile. P-4. The adjustment process is affected by the access of the household to the housing market. P-5. Time and home-ownership have an inertial influence on household residential mobility. 39This assumption for a Mexican American sample is not as unreasonable as might be expected. The survey sched- ule contained the question, "Have you ever had trouble find- ing the kind of house where you wanted to live at a price you could afford to pay, just because you were Mexican American?" Only 14 percent of the households responded yes to this question. However, many mentioned repeatedly diffi- culties in obtaining housing they could afford but attri- buted this to the general shortage of low-cost housing, not to ethnic discrimination. It is possible that the distri- bution of low-cost housing minimizes the likelihood of Mexi- can Americans encountering discrimination because such hous- ing is found in areas of transition and areas which are racially mixed. 23 Hypotheses For the purposes of this investigation, the follow- ing hypotheses are derived: H:1 H:6 Mar" ,The greater the mobility in the past theflesg likely it is that a move is expected. Expected mobility has a curvilinear relationship with age of household: the likelihood increases from younger to middle ages then decreases. The larger the household size in comparison to the present dwelling space (number of rooms) it occu- pies, the more likely a move is anticipated. The higher the income, the higher the probability that a household owns its dwelling. The higher the income, the more likely house financ- ing is by mortgage. The higher the income the larger the present dwell- ing space. Households with owner tenure status are less likely to expect a move. The longer the duration of residence in the community, the less likely a move is planned. The longer the duration of residence in present house, the less likely a move is planned. CHAPTER I I I METHODS Operationalization Family life cycle is measured by the age of the head of household and the household size. The time axis in Figure 1 can also be interpreted as age. Size of house- hold is measured merely by the number of persons in the household. Therefore, typing of individual families in cate- gories such as childless couples, couples with grown chil- dren, couples with school age children, etc., will not be feasible. In all but a few cases the household includes only members of the nuclear family and only occasionally three generations. Households with female heads and house- holds with unmarried male heads in the sample will be in- cluded in the data analysis. These two categories repre— sent 10 percent and 2 percent, respectively, of the sample. Since housing requirement is interpreted as the need for space, it is also measured by the number of persons in the household. The particular composition-of members can- not be taken into consideration. Dwelling space is measured by the number of rooms. Although a commonly used measure of dwelling space, it 24 25 ignores such considerations as size of rooms, function of rooms, the size of a household and the age, sex and rela- tionship of its members. Access to the housing market for this-research proj- ect is concerned only with the financial dimension. The financial resources of the household are measured by income which is household income not just that of the head of house- hold and the method of home financing utilized. The latter variable will provide an indication of the credit available to the household. Given a minority population known to con- tain large families and expected to have lower than average incomes, knowledge of the means by which a household may achieve home ownership is important. A discussion of the two types of financing, mortgage and land contracts, is useful. The opportunity of financing by land contract in Michigan may be an important avenue to ownership for such a group as the Mexican-Americans. A mortgage is held by a bank which imposes restrictions on the type, location and age of dwellings it will finance and also has more stringent credit criteria which a client must meet. Land contracts, on the other hand, are usually held by indi- viduals whose restrictions, standards and credit investiga- tions are often more lenient. A land contract also offers the advantage of a lower down payment, no closing costs and may extend the payment period beyond the current 25 year limit imposed on a mortgage. In the case of a land contract 26 the deed to the property is retained by the lender until payment is completed whereas with a mortgage the purchaser is given the deed immediately. A distinct disadvantage to the land contract is that repossession may follow immediately upon failure to remit a single house payment promptly. In contrast repossession procedures on mortgaged houses run a matter of months and provide opportunities for the mortgage holder to be re-instated and to protect his investment. The less stringent credit requirements and the possibility of rapid repossession with the land contract make it possible for unscrupulous creditors to take advantage of lower income households. The inertial influences on planned mobility are through the variables of past mobility in the town of present residence, duration of residence in community, time in pres- ent house, and tenure. Past mobility is measured by the number of moves by the household in present community of residence in conjunc- tion with the duration of residence there. A move is defined as any change in residence regardless of the time a dwelling is occupied, and no distinction is made between voluntary and forced moves. Duration is measured~from the age of 16 onward for those respondents who lived in the community dur- ing childhood. This variable is measured in tenths of a year. The histories of the households are examined for data on these two variables, and an index of mobility is constructed for each according to the following equation: 27 total number‘Of‘inetown moves duration of residence in town Index of Mobility = In this manner, both the number of moves and the exposure to 40 The variable time in moving are taken into consideration. present house is simply the number of years, measured in tenths. In previous studies of residential mobility, tenure status has been measured simply as owning or renting. In addition to the forementioned dichotomy, ownership in this study will be expanded to specify the method of financing used to purchase a house and to specify those owners who have completed the purchase. Therefore, tenure status will be subdivided into the following categories: renter, land contract owner, mortgage owner and fully paid owner. Planned mobility, the basic dependent variable, is measured by the response to the question: "Are you planning to move from this house eventually, or do you think you will remain here permanently from now on?" Unfortunately the question places no restriction on the period of time during which a move is anticipated, for example, during the next year or two years. Rossi examined the relationship between 40A similar approach to the measurement of a house- hold's moving experience was used by Chevan. He identified a move as only those changes of residence which involved a minimum of four months occupancy and constructed a variable which indicated moves per hundred years of regional exposure: total number of moves X 1200 R = total months of exposure Chevan, "Longitudinal Approach," p. 5. 28 stated mobility inclinations and actual mobility in his project. His sample was querried on their desire to move within the next 10 months. A follow-up on actual mobility only 8 months later revealed that 96.percent of those house- holds planning to stay actually did so and‘80 percent of those expecting to move had actually moved indicating that stated intentions are a good indicator of actual behavior.41 Further support is provided by Leslie and Richardson who made a similar check 10 months after the intention to move in 12 months was stated. In this case, 97 percent of the stayers stayed and 85 percent of the movers moved.42 A summary of the Operationalization of the concepts involved in the present study is given in Table 1. Data Collection This thesis is a secondary analysis of data collec- ted during November and December, 1967, and January, 1968, for ”Mexican Americans in Transition, Migration and Employ- ment in Michigan Cities," by Harvey M. Choldin and Grafton D. Trout, Jr.43 41Rossi, Why Families Move, p. 107. 42Leslie and Richardson, "Decision to Move," p. 897. 43Sponsored by the U. S. Dept. of Labor, Office Manpower Research, Contract No. 81246632. 29 Table 1. A summary of operationalization . . Level of Concepts/Variables Indicators Measurement 1. Family Life Cycle Age of head of household ordinal 2. Housing requirement Household size ordinal 3. Dwelling size (size of present unit) Number of rooms ordinal 4. Access to housing market Income ordinal Method of financing ordinal* 5. Inertial influences Past mobility in town of present residence Index of mobility ordinal Duration of residence Years in community ordinal Time in present house Years in house ordinal Tenure Tenure status ordinal* Method of financing ordinal* 6. Housing supply Assumed constant -- 7. Planned mobility Move plans. . ordina1* *The level of measurement is given as used in this thesis. Although the indicators accompanied by asterisks are often considered nominal and in some instances are dichot- omous, it is reasonable to consider them as ordinal because in their conceptualization there is an underlying dimension that is at least ordinal in nature. For example, the dichot- omy of planned mobility (movers-stayers) can be represented as the split of a continuous variable. The expectation of moving cannot realistically be considered as "all or nothing" categories; it would vary by degrees of expectation of stay- ing. This expectation of moving is slight and movement across the range would indicate that the expectation of staying de- creases and the expectation of moving increases. At the op- posite end, the expectation of moving would predominate over that of staying. Method of financing is another case. The Statuses of renter, land contract owner, mortgage owner and paid owner represent increasing degrees of commitment to hous- ing. Tenure status, renter-owner, is the dichotomous split of this variable as low and high commitment. 30 This research project used "controlled selection sampling" to provide a probability sample of the Mexican- 44 Counties American population outside the Detroit SMSA. were selected from the 17 counties in Michigan which were estimated to contain at least 100 Mexican American families. These 17 counties which are located in the southern half of the Lower Peninsula of Michigan contain virtually all the Mexican American population in Michigan outside the Detroit SMSA. The four counties which had more than an estimated 800 families were deliberately selected for the sample with the remaining 13 counties being classified into three strata with approximately the same estimated number of settled Mexican Americans and on the basis of the presence or absence of a large city in the county. In addition the geographical distribution of the counties was taken into account. One county was selected from each stratum with probability pro- portionate to the estimated number of settled Mexican Ameri- can families in the county. For one of the counties selec- ted, the estimated number of Mexican American families was so low that, in order to obtain an overall sampling rate consistent with the other counties, it was necessary either to double-weight a subsample of them or to select a second 44Roe Goodman and Leslie Kish, "Controlled Selection-- A Technique in Probability Sampling," Journal gf the American Statistical Association, 45 (1950) 350-372. 31 sample county from the same stratum and to represent that stratum by a probability selection of two counties rather than one. The latter choice was made with the second county being selected on the basis of the personal judgment of the research directors rather than at random. After the counties were selected, the Mexican Ameri- can households within them were identified and listed by contacting informed persons in various organizations and through the use of city directories and telephone books from which persons having Spanish surnames were listed. A pilot study indicated that the two main errors in the lists so obtained were non-locatable addresses and names of non- Mexican Americans. Names were then selected from these lists for interview. The interviewing was performed by bilingual and biliterate Mexican American residents of the counties under study after receiving interview training. Method 2: Data Analysis The data for this thesis were analyzed with the use of the Control Data Corporation 3600 and 6500 computers using available statistical programs. The primary method_of statistical analysis used is a measure of association, gamma. This measure was chosen for several reasons. It is apprOpriate for measuring associa- 1Lions between ordinal variables, and all of the variables 32 used in this thesis can be interpreted as having an ordinal level of measurement as indicated in Table l. Gamma also has the advantage of being directly inter- pretable. Costner has indicated that gamma can be inter- 45 "The numer- preted as the proportional reduction of error. ical value of" . . .gamma. . ., "disregarding sign, gives the percentage of guessing errors eliminated by using knowledge of a second variable to predict order."46 Therefore, gamma has both a value and a sign (+ or -) with a range of -1.0 to +1.0. A gamma of -l.0 indicates that predicting a reverse order between variables would allow perfect predictability, and for +1.0 the converse is true, i.e., perfect predictabil- ity is the result of predicting the same order on both vari- ables. When the number of pairs with reverse orders on the two variables is equal to the number with same orders, a gamma of 0.0 is found. "Values intermediate between 0 and 1.0 indi- cate the degree to which guessing errors may be reduced by utilizing knowledge of order on a second variable."47 45Herbert L. Costner, "Criteria for Measures of Association," American_Sociological Review, 30 (June, 1965) pp. 341-353. 46John H. Mueller, Karl F. Schuessler and Herbert L. Costner, Statistical Reasonin in Sociology (Boston: Houghton Mifflin Co.,‘l970). P. 887 47 Ibid., p. 289. 33 All computations for gamma are based on the untied pairs for the particular variables in question. It is a symmetrical measure of association, and, therefore, the prediction of order can be from either variable on the other on the basis of one calculation. CHAPTER IV ANALYSIS OF DATA Social Characteristics 2; the Sample HEEEE' The approximately 8,000,000 Michigan popula- tion includes approximately 100,000 Mexican Americans. The Mexican American population in Michigan is primarily urban, and this is reflected in the sample. Of those Mexican Ameri- cans outside Detroit SMSA, 75 percent live in cities with populations of 50,000 to 250,000, 12 percent live in rural places and another 12 percent in small towns of 2,500 to 25,000. 5E2.22§.§E§' The households surveyed had 90 percent male heads and 10 percent female heads. The ages of male heads of household in the sample were distributed approxi- mately in quarters in the following categories: those 29 and under, 30-39, 40-49 and 50 and over years. The 69 fe- male heads of household tended to be older than the 626 male heads with median ages of 43 and 40 respectively. Of the 3,776 persons represented in the 695 house- holds, 67 percent were under the age of 25. The youth of the Mexican American group may be indicated in contrast with 1960 census data showing that only 46 percent of total Michi- gan population appeared in this category. 34 35 Marital Status. Marital status among male heads of household was distributed as 95 percent married, 2 percent widowers, 2 percent single and 1 percent divorced or sepa- rated. Similar categories for the female heads were 34 per- cent widows, 25 percent separated, 23 percent divorced, 7 percent single and 10 percent married but providing the main source of income for the household. Household Size. The sizes of household tended to be large ranging upward to 15. This distribution is presented below. It should be noted that better than three-quarters of the households contain three to eight persons. The mean household size is 5.5 persons. Number of Persons Percent l 2.2“ , 2 .8.5. 3 14.5, . 4 - 16.3 5 14.8 6 12.8 7 9.6 8 8.8 9 5.6‘ 10-15- 6.9 Total 100.0 (695) Income. Income levels of the households range from less than $1500 to more than $15,000 per annum. The modal group (32.1%) have incomes of $7,000-9,999. Another quarter have incomes of $5,000-6,999. The remainder of the sample can be divided as those with less than $3,000 annual income (8.5%), those with $3,000-4,999 (15.7%) and those with more 36 than $10,000 (15.3%). An income level of $7,000 divides the sample households approximately in half. Annual Income 'Percent Under $1500 1.7‘ $1500-1,999 1.6 $2,000-2,999 W5.2 . $3,000-3,999 6.6 $4,000-4,999 9.1 ‘ $5,000-6,999 26.5__. $7,000-9,999 ,32.l _ ,,_,. $10,000-14,999 ”‘13.4 ‘"2 $15,000 and over 2.3 Total 100.0* (695)* *Includes 1.6% no information Education. The level of education of the heads of household in the sample was low; 52 percent had completed 6 or fewer years of schooling. Years of Schooling Completed Percent No schooling 12 1-4 years‘ 23 5-8 years 33 9-11 years 17 12 or more years 15 Total 100% Occupation. The occupations of the male heads of household in the sample are predominantly in the operative category (63.2%). The next largest occupational category is that of craftsmen, foremen (14.2%) followed by laborer (6.5%). The remaining 16.1 percent are scattered among ser- vice, managerial, farm laborer and clerical occupations. 37 Tenure Status. The large majority (69.5%) of the Mexican American households in the sample own or are buying their homes. Almost half of the home owners finance their purchases through land contracts. Tenure Percent Rentl 30.5 (212) Own 69.5 (482) Land Contract 48.0 (229) Mortgage 28.0 (133)2 Fully paid 24.0 (114) 1 Includes 3.6% who do not pay rent but are pro- vided with housing through their job. 2This total does not add up to 482 as 6 cases had no information. Testing 9: Hypotheses Planned Mobility. Planned mobility refers to the status of moving plans for a household. Those anticipating no move are classified as "Stayers" and those anticipating a change in residence as "Movers." In the sample of 695 households, 281 (40.4%) planned to move and 349 (50.2%) planned to remain. The remaining 65 households (9.4%) fell into the category of "don't know" or "no information" and were dr0pped from the main analysis of planned mobility. In comparison to national figures presented in the Introduction, this represents a high degree of mobility in the Mexican American population. 38 An analysis of the "don't knows" revealed that they tend to be like stayers. They tend to be distributed in pro- portion to the total number and have approximately that pro- portion across the independent variables related to expected mobility. There is a somewhat higher proportion at the $5,000 or less income stratum and also in renters rather than owners, but when looking at the "don't knows" themselves, they tend to be low income and renters. Past Mobility ig_Town 2: Present Residence. The Past Mobility Index used to measure past mobility is the ratio of total number of in-town moves to the duration of residence in town. The data obtained from this index are dichotomized at the median to represent low and high past mobility. The re- sults of relating this index to planned mobility are pre- sented in Table 2. It can readily be seen that low past mobility is as- sociated more strongly with stable than mobile households. There is a tendency for households with low past mobility to remain less mobile and a less marked tendency for high mobile households to continue also. These results indicate a moderate association between these variables but in a direc- tion opposite to H:l in which an inverse relationship was expected. In an effort to clarify this relationship a series of control variables were applied. First, the relationship between planned mobility and past mobility was controlled 39 Table 2. Planned mobility by past mobility in town of pres- ent residence. Past Mobility Expected Mobility Low ~ High Movers 33.7 55.4 Stayers 66.3 44.6 Total % 100.0 100.0 Total (309) (307) 2 x = 29.4 d.f. = 1 p > .001 Gamma = -.42 for tenure (Table 3). The result is a reduction in the level of significance and also in gamma for renters and paid owners. Among buyers, which includes both land con- tract and mortgage owners, and paid owners the direction of the original relationship is maintained. However, among renters the direction disappears indicating that renters are more likely to be movers regardless of amount of past mobility. The table also reveals that renters regardless of planned mobility are very likely to have high past mo- bility, buyers are approximately evenly divided between low and high past mobility while paid owners are character- ized by low past mobility. The chi square and degrees of freedom are summed over the partial tables as suggested by Blalock in order to give an indication of overall level of significance. 46Hubert M. Blalock, Social Statistics (New York: McGraw-Hill Book Co., Inc., 1960), pp. 238-239. 40 Table 3. Planned mobility by past mobility by tenure. ‘ Tenure Renters Buyers Paid Owners “PIahhed Past Mobility Past Mobility Past Mobility Mobility- Low High Low High Low High Movers 60.4 72.3 29.2 40.3 27.7 50.0 Stayers 39.6 27.7 70.8 59.7 72.3 50.0 Total % 100.0 100.0 100.0 100.0 100.0 100.0 Total (48) (137) (178) (154) (83) (14) Renters: x3 = 2.3 d.f. = 1 p > .25 < .1 Gamma = -.24 Buyers: x2 = 4.46 d.f. = l p > .05 < .025 Gamma = -.44 Paid Owners: x = 2.7- d.f. = l p > .1 < .05 Gamma = -.26 Total: x2 :‘9.47 d.f. = 3 p > .025 <.01 Table 4 presents the results of the introduction of the second control variable, age of head of household. In this case the measures of association are substantially re- duced as are significance levels. The direction is main- tained except in the youngest group where direction changes. The young age group is predominantly movers regardless of past mobility. The 29 or less group is much more likely to have high past mobility, the 50 or more group much more likely to have low past mobility with the intermediate group more evenly divided. The third control variable used is household size (Table 5). The large household sizes represented in this Mexican American sample could be expected to restrain 41 Table 4. Planned mobility-by past mobility by age of head of household. Age of Head of Household 29 or less 30-49 50 & over Planned Past Mobility Past Mobility Past Mobility Mobility Low High Low High Low High Movers 75.0 69.2 40.5 51.2 20.0 30.0 Stayers 25.0 30.8 59.5 48.8 80.0 70.0 Total % 100.0 100.0 100.0 100.0 100.0 100.0 Total (16) (107) (163) (170) (130) (30) 29 or less: x§ = .22 d.f. = 1 p > .75 < .5 Gamma = .14 30-49 : x2 = 3.8 d.f. = l p > .05 < .025 Gamma = -.21 50 & over : x = 1.4 d.f. = l p > .25 < .1 Gamma = -.26 Table 5. Planned mobility by past mobility by household size. Household Size. 1-4 persons. 5-7 persons 8-15 persons Planned Past Mobility Past Mobility Past Mobility Mobility Low High Low High Low High Movers 29.8 59.5 40.4 52.1 29.7 53.3 Stayers 70.2 40.5 59.5 47.9 70.3 46.7 Total % 100.0 100.0 100.0 100.0 100.0 100.0 Total (124) (126) (111) (121) (74) (60) 1-4 persons: x = 22.27 d.f. = l p > .001 Gamma = -.55 5-7 persons: x = 3.09 d.f. e l p > .05 < .025 Gamma = -.22 8-15 persons: x = 7.67 d.f. = l p > .01 < .005 Gamma = -.46 42 mobility. The direction of the zero order table is main- tained in all partials and significance levels remain high. A change is noted in the gamma: it is increased in 1-4 per- son households, decreased in 5-7 person households and re- mains approximately the same in the largest households. The percentage of stayers in each household size strata are: 1-4 persons, 55.2%; 5-7 persons, 53.4%; 8-15 persons, 59.7%. Although the largest households are more likely to be stay- ers, the differences in percentages are not large. The mid- dle group may reflect adjustments being made during the child- bearing stage while at the same time representing not overly difficult spatial demands to be met by new housing. Income is the last control introduced in the rela- tionship between planned mobility and past mobility in town of present residence (Table 6). Here again the direction of relationship and high levels of significance are retained. The strength of association is increased in the lowest and highest income groups and reduced in the two middle groups. The latter group may be at an income level where a move is more feasible. If one examines the percentage of stayers in each stratum the following information is obtained. Income '% of Stayers less than $5,000 61.2 $5,000-6,999 54.4 $7,000-9,999 50.0 $10,000 or more 61.3 Table 6 . 43 Planned mobility by past mobility by income. INCOME (in thousands) $4.9 or less $5.- $6.9 Planned Past Mobility Past Mobility Mobility Low High . ,.Low.... High Movers 24.6 52.9 38.5 51.8 Stayers 75.4 47.1 61.5 48.2 Total % 100.0 100.0 100.0 100.0 Total (69) (70) (65) (84) A” $7 - $9.9 $10 or more Past Mobility Past Mobility Low High Low High Movers 41.9 58.1 22.0 53.8 Stayers 58.1 41.9 78.0 46.2 Total % 100.0 100.0 100.0 100.0 Total (93) (93) (41) (39) $4.9 or less: x: = 1.6 d.f. = 1 p > .001 Gamma = -.54 $5-6.9 : x2 = 2.39 d.f. = l p > .25 < .1 Gamma = -.25 $10 or more : x = 8.6 d.f. = l p > .005 < .001 Gamma = -.61 Total: x = 27.39, d.f. = 4 P > .001 44 It can be surmised that the high proportion of stayers in the lowest income category may reflect the lack of financial resources to gain access to the housing market. The equally high proportion in the highest income category may reflect resources sufficient to have made a satisfactory adjustment in housing since“this group has the greatest ability in terms of income to accomplish the adjustment. Age gquead 9: Household. The results of the test- ing of H:2 are given in Table 7. These results indicating astrongy highly significant association between age and planned mobility echo the findings of all research on resi- dential mobility. A monotonic inverse relationship appears. This is not in accord with H:2 which specified a curvilinear one. A further division of the youngest age group might possibly have helped. However, it is possible that the early age of marriage among Mexican Americans and the tendency for child-bearing to begin soon after marriage may account for ‘the greater proportion of movers in the youngest age group. ' In line with this reasoning, household size was in- troduced as a control variable (Table 8). In the partial dealing with the largest household size (8-15) the associa- tion between age and planned mobility is substantially re- duced as is the level of significance. This particular par- tial is marred, however, by the low cell frequencies in the youngest age group. The association is increased in 1-4 per- son households and remains approximately the same in the 45 Table 7. Planned mobility by age of head of household. Planned Age Mobility_ 29 8 under 30-39 40-49 50 & over Movers 70.4 54.1 36.8 21.8 Stayers 29.6 45.9 63.2 78.2 Total % 100.0 100.0 100.0 100.0 Total (125) (185) (155) (165) £7= 78.86 d.f. = 3 p > .001 Gamma = .51 Table 8 . Planned mobility by age of head of household by household size. Household Size 1-4 persons 5-7 persons Age I Age Planned 29 50 I 29 50 Mobility or less 30-49 or over.i or less 30-49 or over T Movers 71.1 47.6 18.5 I 68.4 48.7 20.5 Stayers 28.9 52.4 81.5 31.6 51.3 79.5 Total % 100.0 100.0 100.0 100.0 100.0 100.0 Total. (83) (84) (92) (38) (152) (44) In” 8-15 persons Age Planned 29 50 MObilitY or less 30-49 or over Movers 75.0 41.3 34.5 Stayers 25.0 58.7 65.5 Total % 100.0 100.0 100.0 Total (4) (104) (29) 1-4 persons: xg= 49.23 d.f. = 2 p > .001 Gamma = .65 5-7 persons: x = 19.6 d.f. = 2 p > .001 Gamma = .53 8-15 persons: x = 2.42 d.f. = 2 p >..5 < .25 Gamma = .23 46 5-7 households. Therefore, the association between age and planned mobility loses strength as household size increases. The monotonic relationship is maintained in each case. It is risky to compare the youngest groups across household size because of the low cell frequency mentioned above. For the oldest group the likelihood of moving plans increases as household size increases. This could indicate the anticipa- tion of a decrease in household size or further attempts to adjust housing to large household size. The proportion of stayers in each partial increases as household size increases. Another control, income, was applied to this rela- tionship (Table 9). In this instance, significance levels remained high, the monotonic relationship was sustained and the strength of the measure of association was maintained in the $5-6.9 income stratum while being reduced in the $7-9.9 stratum. In the other two cases the measure was strengthened. Household Size and Dwelling Space. A surprising result is found in relating household size to planned mobil- ity. A weak association is evident in Table 10. A strong association between large size and inclination to move has been found by Rossi and most researchers. However, when the range of household sizes is examined for previous stud- ies of residential mobility, differences appear. The larg- est household size found in Rossi's study is 5 persons.48 48Rossi, Why Families Move, pp. 73ff. 47 Table 9. Planned mobility by age of household head by income. Income (in thousands) $4.9 or less $5 - 6.9 Age Age Planned. 29 50 29 50 Mobility or less 30-49 or over or less 30-49 or over Movers 61.3 48.1 17.5 65.9 46.6 22.5 Stayers 38.7 51.9 82.5 34.1 53.4 77.5 Total % 100.0 100.0 100.0 100.0 100.0 100.0 Total. (31) (54) (57) (41) (73) (40) $7 - 9.9 $10 or more . Age Age P1anned-p 29 50 29 50 Mobility or less 30-49 or over or less 30-49 or over Movers 80.0 44.4 41.7 71.4 39.3 13.3 Stayers 20.0 55.6 58.3 28.6 60.7 86.7 Total % 100.0 100.0 100.0 100.0 100.0 100.0 Total (35) (117) (36) (7) (61) (15) $4.9: x3 = 19.4 d.f. = 2 p > .001 Gamma = .57 $6.9: x2 = 15.41 d.f. = 2 p > .001 Gamma = .51 $9.9: x2 = 15.02 d.f. = 2 p > .001 Gamma = .41 $10+: x = 7.27 d.f. = 2 p > .05 (.025 Gamma = .63 Total: x2 = 57.10 d.f. = 8 p > .001 48 Table 10. Planned mobility by household size. Number of Persons in Household Planned Mobility 1-2 3-4 5-6 7-8 9-15 Movers 35.5 47.7 45.0 47.9 38.6 Stayers 64.5 52.3 55.0 52.1 61.4 Total % 100.0 100.0 100.0 100.0 100.0 Total (62) (197) (177) (117) (83) .2. x = 4.6 d.f. = 4 p > .5 < .25 Gamma = .03 It is interesting to note that Abu-lughod found no signifi- cant relationship between these two variables but indicates 49 (She that there were few large families in her sample. does not state specific numbers.) The sample under investi- gation here, of course, provides a very wide range in house- hold size. Table 11 gives the results with which to test H:3 that the larger the household size in comparison to the pres- ent dwelling space (number of rooms) it occupies, the more likely a move is planned. As H:3 is stated an increase in the proportion of movers across household size is to be ex- pected in each dwelling space category. Table 11 does not indicate this. However, if each household is examined sep- arately, it can be seen that the proportion of movers in each household size group is greater in the smaller dwelling space than the larger. The greatest change occurs in the 49Abu-lughod, Housing Choices, p. 419. 49 Table 11.‘ Planned mobility by household size by dwelling space. Number of Rooms l-5 6-8 Planned Number of Persons Number of Persons Mobility 1-4 5-7 8-15 1-4 5-7 8-15 Movers 46.8 56.1 47.7 41.6 39.7 37.6 Stayers 53.2 43.9 52.3 58.4 60.3 62.4 Total % 100.0 100.0 100.0 100.0 100.0 100.0 Total (158) (98) 3” (44) (101) (136) (93) 1-5 rooms: x: = 2.19 d.f. = 2 p >.5 <.25 Gamma = -.08 6-8 rooms: x = .31 d.f. = 2 p >.9 <.75 Gamma = .05 middle size households. The partial representing smaller dwelling space is equally divided between movers and stayers; the one representing larger dwelling space is proportioned as 40 percent movers and 60 percent stayers. The strength of association for this table is quite low as are the sig- nificance levels. Using dwelling space as a control does not change the original relationship between household size and expected mobility. In examining dwelling space alone with planned mobil- ity, the relationship is stronger and the significance level is higher (Table 12). Once again a monotonic relationship is observed. 50 Table 12. Planned Mobility by dwelling space. -. MESIIEEY 1_4 Nugber of Room: 7-8 Movers 52.3 47.7 41.8 37.8 Stayers 47.7 52.3 58.2 62.2 Total % 100.0 100.0 100.0 100.0 Total (151) (149) (158) (172) 2 x = 7.9 d.f. = 3 p > .05 < .025 Gamma = .21 Income. Three hypotheses offered are related to income. In H:4 it is expected that the higher the income, the greater is the probability that a household owns its dwelling. In Table 13 it can be seen that H:4 is clearly supported with an association of moderate strength and high significance. It is something of a surprise to note how high the proportions of owners are in the lower income categories. The results of relating method of financing to in- come are found in Table 14. In this case H:5 stating that the higher the income, the more likely house financing is by mortgage is supported. The association is moderate and at a high level of significance. In addition to indicating sup- port for H:5, Table 14 contains some very interesting find- ings. As would be expected, proportions of renters decrease with increasing income. In comparing land contract with mortgage financing, support is found for the reasoning that 51 Table 13. Tenure status by income. M Income (in thousands) Tenure $2.9 $10 Status or less $3-4.9 $5-6.9 $7-9.9 or more Own 49.2 53.7 64.1 77.2 88.9 Rent 50.8 46.3 35.9 22.8 11.1 Total % 100.0 100.0 100.0 100.0 100.0 Total (59) (108) (184) (224) (108) x2 = 52.11 d.f. = 4 p > .001 Gamma = -.42 Table 14. Method of financing by income. ._._— Income (in thousands) Method of $4.9 $10 Finance or less $5-6.9 $7-9.9 or more 'Renters 46.4 38.0 23.7 12.0 Land Contract 22.3 33.2 38.8 38.0 Mortgage- 9.5 14.7 21.9 37.0 Paid 21.8 14.1 15.6 13.0 Total % 100.0 100.0 100.0 100.0 Total (179) (184) (224) (108) 2 x = 75.87 d.f. = 9 p > .001 Gamma = .20 52 land contract financing might provide an important avenue toward home ownership for lower income groups; the propor- tion of land contract holders are considerably greater than mortgage holders in lower income strata. The paid owner stratum also is interesting in that the lower the income, the more likely a house is paid for. In fact paid owners represent 55 percent of total owners in the $2,999 or less strata. It seems reasonable to assume this may reflect older heads of household on retirement incomes. The third hypothesis relating to income is H:6 stat- ing that the higher the income, the larger the dwelling space. The data in Table 15 tend to support H:6 with a highly significant although not strong association. Inertial Influences. Although tenure has been exam- ined in relation to income, it remains to indicate its rela- tion to moving plans of the household. In H:7 it is expected that owners are less likely to be movers than renters. In Table 16, H:7 is supported with a strong association at a high level significance. It is interesting to note that mortgage owners are slightly more likely to be movers than land contract owners. In addition the difference between buying owners and paid owners is not as great as might be expected. The last two hypotheses refer to duration of resi- dence in the community and in the present house. Both are expected to be associated with lower probability of planned Table 15. 53 Dwelling space by income. Income (in thousands) Dwelling $2.9 $10 Space or less $3-4.9 $5-6.9 $7-9.9 or more 1-4 rooms 45.8 30.3 25.0 18.3 14.8 5 rooms 18.6 26.6 27.2 25.4 20.4 6 rooms 20.3 24.8 23.4 24.6 26.9 7-8 rooms 15.3 18.3 24.5 31.7 28.0 Total % 100.0 100.0 100.0 100.0 100.0 Total (59) (109) (184) (224) (108) x2 = 35.59 d.f. = 12 p > .23 Table 16. Planned mobility by tenure status Owners Planned . Land Mobility Renters Contract Mortgage Paid Movers 68.9 24.0 36.2 30.6 Stayers 31.1 66.0 63.8 69.4 Total % 100.0 100.0 100.0 100.0 Total (190) (212) (127) (98) 2 x = 66.69 d.f. = 3 p > .001 Gamma = .40 54 mobility, H:8 and H:9 respectively. Duration of residence in community for this sample ranged from .1 to 47.5 years. This data was divided into quartiles and related to planned mobility. The results are presented in Table 17. These data support H:8 as once again a monotonic relationship is in evidence. Although the association is highly significant, it is not very strong. Although there will be some influ- ence of age in long duration it is noted that the associa- tion is not as strong (Gamma = .29) as that of age with planned mobility (Gamma = .47). Table 18 gives the association between time in pres- ent house and expected mobility. In this case the relation- ship is not so clean cut. The likelihood of planning a move is the same whether residence is for less than 3 years or whether for 3 to 10 years. The association is weaker than that of duration of residence in community with planned mobility. When controlling for tenure (Table 19), the asso- ciation reverses for buyers and renters and remains the same for paid owners but significance levels are low. Paid own- ers are more likely to plan a move when in residence for 3- 10 years than in the shorter or longer time periods. Summary The results relating past mobility in town of pres- ent residence to planned mobility (Hzl) can be summarized as follows. Stayers are more likely to have had low past 55 Table 1?. Planned mobility by duration of residence in community. Planned Duration (1n years) Mobility .l-5.l 5.2-12.6 12.7-19.4 19.5-47.5 Movers 56.6 50.0 38.3 31.6 Stayers 43.4 50.0 61.7 68.4 Total % 100.0 100.0 100.0 100.0 Total (159) (154) (154) (155) x2 = 24.14 d.f. = 3 p > .001 Gamma = .29 Table 18. Planned mobility by time in house. Planned Time in House (in years) Mobility 3 or less 3.1-10 10 or more* Movers 45.9 45.8 32.8 Stayers 54.1 54.2 67.2 Total % 100.0 100.0 100.0 Total (183) (212) (116) x2 = 6.25 d.f. = 2 p > .05 < .025 Gamma = .14 56 Table 19. Planned mobility by time in house by tenure. Tenure Buyers Paid Owners Time in house Time in house Planned . 3 , 10 3 ~~‘~ 10 Mobility or less 3.1-10 or more or less 3.1-10 or more Movers 34.0 38.6 31.8 18.2 42.9 26.3 Stayers 66.0 61.4 68.2 81.8 57.1 72.7 Total % 100.0 100.0 100.0 100.0 100.0 100.0 Total (106 (145) (44) (ll) (28) (57) Renters Time in house Planned 3 10 Mobility or less 3.1-10 or more Movers 68.1 74.4 64.3 Stayers 31.3 25.6 35.7 Total % 100.0 100.0 100.0 Total (64) (39) (14) Buyers: x2 = .95 d.f. = 2 p > .975 < .950 Gamma = -.15 Paid Owners: x2 = 3.28 d.f. = 2 p > .25 < .1 Gamma = .12 Renters: x2 = .61 d.f. = 2 p > .975 < .950 Gamma-= -.03 57 mobility and movers, high past mobility, and, therefore, no support is given to H:l. In the attempt to clarify this re- lationship it was found that it virtually disappered when controlling for age thus indicating that age tends to explain the relationship between past and planned mobility. In the other control cases of tenure, household size and income, it was found that the original relationship depended upon the level of each variable. It holds for buyers but is reduced for renters and paid owners. It is increased for small house- holds, reduced for medium size households and maintained for large ones. In the case of income, the original relation- ship is increased in the lowest and highest income strata and reduced in the middle income strata. In each case the change in gamma indicates the change in the proportion by which error in estimating the order of pairs can be reduced from a random estimation. The relationship of family life cycle and planned mobility was the concern in H:2. The younger the head of household the more likely it was that he would plan to move. A monotonic relationship which is strong and highly signifi- cant exists between increasing age and increasing proportion planning to stay. In investigating the effects of household size on this relationship, it was found that in general the strength of association between age of head of household and planned mobility decreased as household size increased. It it notable that in the oldest age group, which had the 58 greatest likelihood of being stayers in the original rela- tionship, the likelihood of being movers increases as house- hold size increases. Income also specifies conditions under which the relationship is increased, maintained or reduced although the differences are not as large as in the house- hold size partials. It was expected that the better the housing adjust- ment in terms of space which the household had achieved the more likely it anticipated staying. The results indicated virtually no association in terms of gamma between the house- hold size-dwelling space adjustment and planned mobility. However, the proportion of movers in each household size group is greater when in smaller dwelling space than the larger. In examining each independent variable separately with relation to planned mobility, it was found that dwell- ing space has some association and household size has none. Access to the housing market through financial re- sources was the concern of H:4, H:5 and H:6. In the case of tenure (H:4), it can be clearly stated that as income increases the proportion of owners increase also. In addi- tion, only at the lowest income level ($2,999 or less) is the proportion of owners less than half. Support is also provided for H:5 dealing with method of financing. Financ- ing through mortgage is more likely as income increases. While the same general description can be madeOf land con- tracts, the latter method is more likely than the former at 59 lower income levels. While households in the lowest income stratum are most likely to be renters, they are also more likely to be paid owners than households in any higher stra- tum. Income is also directly related tO dwelling size (H:6). Inertial influences on planned mobility are antici- pated for tenure, duration of residence in community and time in present house. Tenure does have such an influence on planned mobility (H:7). Renters are the most mobile, but the least mobile are land contract owners. This perhaps reflects the greater likelihood of land contract than mort- gage holders among lower income groups. Duration of resi- dence has a monotonic relationship with planned mobility: the longer the duration the more likely a household is stay- ing (H:8). Time in present house has a weaker association with planned mobility (H:9). At the 3 years or less level and the 3.1 to 10 year level, the likelihood Of staying is the same but at greater than 10 years it increases. Tenure was used as a control on time in house but the significance levels obtained were very low. The greatest inertial influ- ence is exhibited by tenure, followed by duration in commun- ity and time in present house. In general family life cycle as measured by age of head of household is consistently a strong independent vari- able relating to planned residential mobility with the expec- tation of staying more likely with increasing age. This re- lationship is influenced by household size; it is stronger 60 in small households and weaker in medium and large house- holds. Low past mobility, owner tenure status, long dura- tion of residence in community and long time in present house indicate, in descending order of importance, inertial influences on planned mobility, i.e., the tendency for households to remain in presently occupied dwellings. Nei- ther household size alone nor household size controlling for dwelling space, representing housing adjustment were found to be related significantly to planned mobility. Dwelling space alone has some influence. The higher the income of a household the greater the financial access to the housing market through home ownership, financing with less risk and obtaining more dwelling space. CHAPTER V CONCLUSIONS The Mexican American population from which this sample is drawn can be described as primarily urban, young and having large households. The latter trait is predom- inantly due to large numbers of children rather than to three generation households. The population is charac- terized not by very low incomes but by moderate incomes. It can also be described as predominantly a home owning population. Forty percent of the sample planned to move. The household planning a move in this Mexican American sample can best be described as young and with renter status. In addition it tends to have high past mobility, medium income level and to occupy a small dwell- ‘ing space. The household planning to stay can best be described as Older with owner status. Its mobility has been low in the past, and it tends to occupy a large dwelling space. The income level of the stayer household is either low or high. On the basis Of this knowledge, the Mexican Ameri- can population can be compared with the model presented in Figure 1. Given the early age of marriage and tendency 61 62 for child-bearing to begin soon after marriage, T1 would cover a shorter time or age span. As a consequence of larger numbers of Offspring, T2 would be longer although this would depend upon the spacing of children born to the household. In the model, T is equal to T4 on the 2 supposition that it would take approximately as long to "launch" a particular number and spacing of children as to bear them. This could be expected to hold for the Mexican Americans also. The longer that T becomes, the 2 shorter T with stable household size would be. In the 3 case of very large households T3 may be greatly attenuated or even eclipsed entirely since the oldest child could leave the household soon after or even before the youngest child is born. The expectation that T2 and T4 would be longer would also result in a shorter T This is borne 5. out by the data showing large households with heads over 50 years of age. The results in some cases did not support the con- ceptualization presented in Chapter II. This is true in particular for the relationship of past mobility in town of present residence to planned mobility. The general premise is that the household adjusts housing to its re- quirements over time and that each successive adjustment would represent a more satisfactory one. On this basis the more adjustments made in the past the less likely a future move is planned. Assuming that the conceptualiza- tion is valid, the lack of support could be due to the 63 indicators used or to variables which were not included. Age of head of household did reduce the relationship. The index of past mobility is considered a strong indi- cator of past mobility in town of present residence whereas the interview question which is the source of planned mobility data is weak. This is a hazard of sec- ondary analysis, and a recommendation for future research would be to construct a question which has a time limit on expectation of residential mobility and also has or- dered response categories such as "very likely will move," ”somewhat likely will move," etc. The analysis also failed to indicate that the housing adjustment achieved by the household was a factor in planning a move. Although there are other require- ments beyond those of space which influence residential mobility, spatial requirements are considered to be an important element. The suggestion is made that it is probably the method of measuring the housing adjustment which is causing difficulty. As was mentioned in Chapter III, the number of rooms is a crude measure of dwelling space, and controlling household size may not result in the best approximation of the housing adjustment Of the household. The paradigm is still considered to be viable. Refinements are especially indicated for the two areas just mentioned.. The search for valid indicators which 64 can be used to evaluate the theoretical framework-must-be 'cdfitinued in this case astinmmany-Others in sociology. A comparison of present findings with those Of previous studies of residential mobility is hampered by the variety in methodology and in populations studied. Certain congruencies are evident, however. Tenure status was a good predictor of moving plans in every case in 'which it was included except one. The exception was the Leslie and Richardson sample in which only 9 percent of the households were renters. In the Mexican—American sam- ple in which owner status is-refined, the results showed renters to be the most mobile followed by mortgage owners, land contract owners and paid owners. Family life cycle stage was consistently highly related to moving plans with again the Leslie and Richard- son data being an exception.’ The latter study used age as and indicator while the other studies, except for the pres- ent one, used household composition. The Leslie and-Rich- -ardson sample was more homogeneous than the others and with- out a wide age range- This may account for the difference in findings. When using household size as an independent vari- able the results are more varied. Rossi with a large, het- erogeneous sample found that moving plans were more likely among large households. Abu-lughod with a center city sam- ple found no relationship as did Leslie and Richardson and 65 the present study. The longitudinal study by Chevan may provide an explanation in this area. When controlling for the duration of marriage, he found no relationship between moving plans and the number of children present. However, when examining the households in successive three year periods, the addition of a child does affect moving plans for that period. Moreover, what has occurred in previous periods in terms of moves and whether other children were born also has an influence on moving plans in the current period. The housing adjustments occur- ring earlier may retard future moves as additional chil- dren are merely accommodated into existing space. Chevan states that the presence of children of itself is not a good predictor of_future moves. This suggests that a cross-sectional approach may not be able to reveal ade- quately the relationship between household size and resi- dential mobility. This may also Offer an explanation for the Mexican American sample, that is, that larger numbers of children are accommodated into present housing. This may be espe- cially likely given the financial resources available. Al- though income levels were not as low as might be expected for the Mexican American sample, when examined in the light of the sizes Of the households, the ability to make housing adjustments is restricted. The addition of children may 66 cause a strain on the budget that is more salient than the strain on dwelling space. In view of many Of the characteristics of this pop- ulation, it is remarkable that 70 percent own homes. The education level, occupation, income and size of these house- holds aside from their ethnicity would seem to indicate sub- stantial barriers to home ownership. The land contract method of financing available in Michigan is considered to be an important factor in this level of ownership. It clearly represents an important means toward that end among lower income households. The background of many Mexican American households would not meet the usual criteria for obtaining a mortgage. Rental units suitable for large households are typically in short supply at any price level. This may provide an added impetus toward ownership for this group. The ability Of an owner to make structural modifi- cations would be more important to a large than a small household. The fact that 40 percent of the sample plan to move indicates that a large portion of the current housing adjust- ments are unsatisfactory. There must also be other house- holds with equally unsatisfactory adjustments but which see moving plans as unrealistic because of financial resources. The difference in planned mobility between land contract owners and mortgage owners in which mortgage owners are 67 more likely to be mobile probably reflects the different income spread between the two groups. A final conclusion concerns the implications for housing policy which this study provides. The need for reasonably priced, large size housing is clearly indi- cated by the Mexican American population. Yet it is pre- cisely in this category that little in the way of new housing is being provided either as rental or owned units. The study also suggests that it would ease the housing situation for lower income groups if, given a supply of suitable housing, a method of financing with features comparable to the land contract were available at a na- tional level. SELECTED BIBLIOGRAPHY Abu-lughod, Janet, "Appendix" in Housing Choices and Constraints, edited by Nelson Foote pp 51. New York: McGraw-Hill Book Co., Inc., I960. Bell, Wendell, "Social Choice, Life Style, and Suburban Residence," in The Suburban Community edited by William M. Dobriner, New York: G. P. Putnam's Sons, 1958. Blalock, Hubert M., Social Statistics, New York: McGraw- Hill Book Co., Inc., 1960. Chevan, Albert, "A Longitudinal Approach to the Study of Moving," paper presented at the annual meet- ing of the Population Association of America, Boston, Mass., April 18-20, 1968. Costner, Herbert L., "Criteria for Measures of Associa- tion," American Sociological Review, 30 (June, 1965), 341-353. Dobriner, William (editor), The Suburban Community, New York: G. P. Putnam's Sons, 1958. Foote, Nelson, Janet Abu-lughod, Mary Mix Foley and Louis Winnick, Houging Shoicgp and Constraints, New York: McGraw-Hill Book Co., Inc., 1960. Goodman, Roe and Leslie Kish, "Controlled Selection--A Technique of Probability Sampling," Journal of the American Statistical Association, 45 (1955), 350-372. Land, Kenneth C., "Duration of Residence and Prospective Migration: Further Evidence," Demggraphy, 6 (May 1969): PP. 133-140. Leslie, Gerald R., and Arthur H. Richardson, "Life Cycle, Career Pattern and the Decision to Move," Ameri- can Sociological Review, 26 (1961), pp. 894-902. 68 69 Mueller, John H., Karl Schuessler, and Herbert L. Costner, Statistical Reasoning ip Sociology, Boston: Hough- ton Mifflin Co., 1 70. Rossi, Peter H., Why Families Move, Glencoe, Illinois: The Free Press, 1955. Sabagh, George, Maurice Van Arsdol, Jr., and Edgar W. But- ler, "Some Determinants Of Intrametropolitan Resi- ential Mobility: Conceptual Considerations," Social Forces, 48 (September, 1969): PP. 88-98. Shryock, Henry 8., Population Mobility Within the United States, Chicago: Community and Family Study Cen- ter, University of Chicago, 1964. . . .(Irfnpllllllfil) . ‘IIIIIIIIIIIIIIIII