.- -~ - < -- 0‘ ~v—v-wwvvqwhcnnnu‘woql THE PROCESS OF “RETIRED” MiGRAflON _ ' TO ST. PETERSBURG, FLORIDA Dissertation for the! Degree 6f FILE. - _ . . , JOHN CHARLES emu *1 , , -_ x; :l‘ri1973" ' ” -7V>,'U;_;aa~.:e ................ ‘ L’""ARV " Michigan State University This is to certify that the thesis entitled 1 THE PROCESS OF "RETIRED" MIGRATION TO ST. PETERSBURG, FLORIDA presented by John Charles Catau has been accepted towards fulfillment of the requirements for 1213.1). .degree inJepaxhnem: of Geography Date 9/7//77 07639 *' A...- _... 3 BINDING av r — “ME 8 SUNS' 300K 3".an INC. LIBRARY amo: RS SENNGPQ‘ITJIICIIEI; MICH SI Ill/////Il/li///Iill//7ll?ll‘ l ’””ARY / iChigan State University I ill/l :3 In their que 2: are associate 215; scientists izfcrtunate be Erioral princf As an exam with mantel} Alt tough ‘ 3313!; gf the " £1385. severe $1 irxerstat: ABSTRACT THE PROCESS OF "RETIRED" MIGRATION TO ST. PETERSBURG, FLORIDA By John Charles Catau In their quest for an understanding of the patterns and processes that are associated with spatial behavior, geographers, amng other social scientists, have virtually ignored the elderly. This omission is unfortunate because the aged do not always conform to normative behavioral principles. As an example, this study has focused upon one form of elderly mobility, namely, post-retirement migration. Although the participants in this movement constitute a small pro- Portion of the total aged population, their residential flows have Produced several significant spatial patterns. The destinations of aged interstate migrants, for instance, have been concentrated in but a few comunities in the warm weather states of the South, the South- wEst, and the West. The projected increase in peOple over 65, and the ahticipated improvements in retirement programs, including the possi- bility of a younger retirement age, will yield an increase in the rela- tlive dimensions of this retired migrant group. Consequently, in antici- Pfition of these developments, this study sought to examine the salient characteristics of the retired migration process. new 0f the U iE‘ized three ““5 gable in 39Ed mg imam: motivatiol gggvaigrate becau- ;;::es‘.zed that re‘ anpmicular, a 2::Le of distance mg pogulu di 22:21: can be a; 32121, the print 33553185; that the teas. Because the 41:56. through “5rezired migri * MY months We discuss ‘ Erasing 1 fire «on prep it “ "9 the 'z-{E e‘he 2‘ facu refit c John Charles Catau A review of the theoretical foundations of migration analysis suggested three areas in which general migration theories were in- applicable in aged migration. The first discrepancy concerned the role of economic motivations. Whereas most studies have shown that people usually migrate because of job or business considerations, it was hypothesized that retirees are in a position to emphasize other elements, and in particular, amenities. 0 The second discrepancy, which concerned the role of distance in retired migration, led to the postulate that the very popular distance-decay proposition will require revision before it can be applied successfully in amenity-related movements. Finally, the principles of stage migration were examined, and it was suggested that they were of little consequence in the retired migration process. Because the questions under consideration could not be adequately answered through aggregate statistics, a personal interview survey of 245 retired migrants was conducted in St. Petersburg, Florida, during the early months of 1972. The discussion of the salient findings was organized into four menjor sections. First, several of the personal characteristics of the respondents were examined and summarized. With this background, the remaining sections emphasized: (l) the migrant origins; the migration preparations; and (3) the actual migration process. Among the more conspicuous personal attributes of the respondents Were the facts that: (1) they were all white; (2) they came from n“Odest financial backgrounds (50 percent of those responding had pre- re‘tirement incomes of $10,000 or less, and only 3 percent earned at £323,000); and (3 silty. Also, two eeiapsed time beti She migrant orig stat of the count aim this north: 3136:2151)ng cont: '1'. 35 amenities W5 :' a gust-retiremen 112‘: Once the GE 2izictioh of di! is the migran tilts to migrat ELPezersbuzg ei1 3;;2‘. facilities stress to live. Ni. "35w ‘ f iVe ‘: nth Nearl- 3.f;v . .3 the}: move 3:! \EL- John Charles Catau least $20,000); and (3) they had a past history of residential stability. Also, two age variables were employed to show that the mean elapsed time between retirement and migration was just two years. The migrant origins were concentrated in SMSA's in the Northeast quadrant of the country. Whereas 87 percent of the national sample came from this northern region, the areas inmediately adjacent to m; Petersburg contributed very few migrants. This suggested that the mfll of amenities was a more important consideration in the selection ofeapost-retirement home than the friction created by the intervening space. Once the decision to move to an amenity area was made, however, the friction of distance was reasserted. As the migrant origins inferred, amenities supplied the major stimulus to migration. Nine out of every ten retirees were drawn to St. Petersburg either by its climate, its recreational facilities, health facilities, or because they perceived it as,a nice place for retirees to live. Economic motivations were conspicuous in their absence. Eighty-five percent of the respondents had a prior knowledge of the City. Nearly all of these had made at least one personal visit before they moved there; and 96 percent indicated that these trips were their most important source of information. This suggested a linka“between St. Petersburg's dual roles as a tourist attraction and a retirement center. TVO direct indications of pre-migration planning failed to supply eVid . . ence of extensive preparations. there was little ev gazes directly to St. lit“: the vast majori siestined to end in Men the lack of Ecalyses were freq fizz-anal research wa John Charles Catau There was little evidence of stage migration. Eighty-six percent migrated directly to St. Petersburg from their pre-retirement home. And for the vast majority of the respondents, the migration process was destined to end in St. Petersburg. Given the lack of prior investigations in many of these areas, the analyses were frequently as heuristic as they were definitive. Additional research was suggested. THE in THE PROCESS OF "RETIRED" MIGRATION TO ST. PETERSBURG, FLORIDA BY John Charles Catau A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Geography 1973 To Janice . . . the one person who made it all possible ii i; effort such as apart of many peoplt 2er retired peo; 5%. Without thei :52; have never bee ‘= director of the negation. He was as the use of t The author is (1‘ run As the l sue advisor dur ‘e. imnn provided is forEbearance, 'TLtten "rough“ d The other we ACKNOWLEDGMENTS An effort such as this is founded on the advice, assistance, and support of many people. Of paramount importance was the cooperation of the many retired people who consented to serve as respondents in the survey. Without their assistance, the major portion of the research could have never been completed. One of these people, Mr. Claghorn, the director of the Senior Citizens Center, Inc. , deserves special recognition. He was helpful in many ways, but in particular, he volun- teered the use of the center as a base of operations in St. Petersburg. The author is also appreciative of his association with Dr. Stanley [L Brunn. As the principal advisor in this dissertation, and in fact, as the advisor during the author's tenure at Michigan State University: Dr- Brunn provided both timely guidance and intellectual stimulation. His forebearance, especially with the missed deadlines and the hand- written "rough" drafts, is sincerely appreciated. The other members of the dissertation and guidance committee, 0" Lawrence Sommers, Dr. Robert Thomas, and Dr. Allan Beegle, were als° very helpful. Their contributions are sincerely appreciated. Particular gratitude is extended to Mr. Sherman Hollander who skillfully performed the cartographic work; and Mrs. Joyce Ingalls, § 8 . “per tEnlist, who eXpertly prepared the final manuscript. iii he author also w: Insanity Computer Ce: L“; axial Science Res pixie through suppc Eximon. finally, the aut L‘s. Janice, whose c 2 :ention . The author also wishes to acknowledge the Michigan State University Computer Center, and especially, the Computer Institute for Social Science Research. The use of their facilities was made possible through support, in part, from the National Science Foundation. Finally, the author's greatest debt is owed to his long-suffering vufe, Janice, whose contributions and sacrifices have been too numerous to mention . iv M Tables . . . "J: of Figures REVIEH Stud TABLE OF CONTENTS Page list of Tables . . . . . . . . . . . . . . . . . . . . . . . . . viii list of Figures . . . . . . . . . . . . . . . . . . . . . . . . x GHWTER I THE EMERGENCE OF POST-RETIREMENT MIGRATION. . . . . . . 1 The "Elderly Explosion" . . . . . . . . . . . . . . . 3 The Emergence of Retirement Systems . . . . . . . . . 14 A Growing Interest in Retirement . . . . . . . . . . 18 The Growing Importance of Post-Retirement Migration . . . . . . . . . . . . . . . . . . . . . 20 The Spatial Components of Post-Retirement Migration: An Introduction . . . . . . . . . . . 23 II PROBLEM STATEMENT IN MIGRATION THEORY . . . . . . . . . 29 The Theoretical Foundations of Migration Analysis . . 33 Are Aged Migrations Economically Motivated? . . . . 35 What is the Role of Distance in the Process of Aged Migration? . . . . . . . . . . . . . . . . . 42 Do the Principles of Stage or Stepwise Migration Apply in Retired Migration? . . . . . . . . . . . 47 Prospectus . . . . . . . . . . . . . . . . . . . . . 48 III REVIEW OF THE LITERATURE ON AGED MIGRATION . . . . . . 50 Studies of Aged Migration Based upon Indirect Methods of Measurement . . . . . . . . . . . . . . 51 or :02st (Cont Studies of amicnher Studies of Analysis 3 THE ST. PETE The Ratio: St. Peter The Spa Fact< Fact Fact Fac‘ 7 mmuu: Operati SamPliJ The TABLE OF CONTENTS (Cont'd.) Chapter Page Studies of Aged Migration Based upon the Census and Other Governmental Statistics . . . . . . . . . 57 Studies of Aged Migration Based upon Survey Mlysis O O O O O O O O O O O O O O O O O O O O O 67 IV THE ST. PETERSBURG STUDY AREA . . . . . . . . . . . . . 81 The Rationale for the Selection of St. Petersburg . . 82 St. Petersburg: General Characteristics . . . . . . 92 The Spatial Structure of the Population . . . . . . 96 Factor 1 Youthful Fringe . . . . . . . . . . . . 98 Factor 2 Socioeconomic Status . . . . . . . . . 102 Factor 3 Mobility . . . . . . . . . . . . . . . 103 Factor 4 Old Age-~Inner City . . . . . . . . . . 105 V ATTRIBUTES OF THE SURVEY DESIGN 111 Operational Definitions . . . . . . . . . . . . . . . 111 Sampling Procedures . . . . . . . . . . . . . . . . . 114 The Sampling Frame . . . . . . . . . . . . . . . . 114 Type of Sample . . . . . . . . . . . . . . . . . . 117 Sample Size . . . . . . . . . . . . . . . . . . . . 121 The Questionnaire . . . . . . . . . . . . . . . . . . 125 VI CHARACTERISTICS OF THE RETIRED MIGRATION PROCESS . . . 126 Background Characteristics . . . . . . . . . . . . . 126 Race . . . . . . . . . . . . . . . . . . . . . . . 127 Sex and Marital Status . . . . . . . . . . . . . . 130 Age . . . . . . . . . . . . . . . . . . . . . . . . 132 Income Before and After Retirement . . . . . . . . 136 Pre-Retirement Mobility . . . . . . . . . . . . . . 140 vi TTJFOTSTNTS (Cont in: Migrant Ori The Dist: The Size lfigrant Pr Migrant The Deg with Migrati Character The Pr The Pr ProSpe 3: SUMMARY pg Backgro Race Sex 5 Age InCo Pre- Migrar The Si21 Migra M19 The TABLE OF CONTENTS (Cont'd.) Chapter Migrant Origins . . . . . . . . . . . . . . . . . . . The Distribution of Origins . . . . . . . . . . . . The Size of the Community . . . . . . . . . . . . . Migrant Preparation . . . . . . . . . . . . . . . . . Migrant Motivations . . . . . . . . . . . . . . . . The Degree and Source of Pre-Migration Familiarity with St. Petersburg . . . . . . . . . . . . . . . Migration Planning . . . . . . . . . . . . . . . . Characteristics of the Migration Act . . . . . . . . The Presence of Stage Migration . . . . . . . . . . The Presence of Seasonal Migration . . . . . . . . Prospects for Future Mobility . . . . . . . . . . . VII SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . Background Characteristics . . . . . . . . . . . . . Race . . . . . . . . . . . . . . . . . . . . . . . Sex and Marital Status . . . . . . . . . . . . . . Age . . . . . . . . . . . . . . . . . . . . . . . . Income Before and After Retirement . . . . . . . . Pre-Retirement Mobility . . . . . . . . . . . . . . Migrant Origins . . . . . . . . . . . . . . . . . . . The Distribution of Origins . . . . . . . . . . . . SizeofComunity................. Migrant Preparations . . . . . . . . . . . . . . . . Migrant Motivations . . . . . . . . . . . . . . . . The Degree and Source of Pre-Migration Familiarity with St. Petersburg . . . . . . . . . . . . . . . Page 141 141 150 155 157 167 172 175 176 182 185 187 189 189 190 190 191 191 192 192 193 193 193 194 EETCONm’TS (Cont Pre-Miqra Characteris The Pres Seasonal ProSpec‘ thorgraphy . . . peziix A. Hateria St. Pe‘ 1* n1: ' f.’“‘$x B. Questl TABLE OF CONTENTS (Cont'd.) Pre-Migration Planning . . . . . . Characteristics of the Migration Act The Presence of Stage Migration . Seasonal Migration . . . . . . . . Prospects for Future Mobility . . Bibliography................. Appendix A. Materials Related to the Factorial St. Petersburg, Florida . . . . . Appendix B. Questionnaire . . . . . . . . . . viii Ecology Page 195 195 195 196 197 199 208 212 1 Resident Populat 65+, 1870-1970 . 2 Life Expectancy 3 beading States t of the Total P0} 1 The Top 21 Coun Percent of the 5 Pobility of Per ‘0 March. 1970 5 Source of Aged p"ls'ldence in 1 7 Ralative Impoz Migration to E LIST OF TABLES Table 1. Resident Population Trends--A11 Ages, 1790-1970; and 65+, 1870-1970 . . . . . . . . . . . . . . . . . . . . . . 2 Life Expectancy for Given Sex and Age . . . . . . . . . . 3 leading States of the United States in Terms of the Percent of the Total Population Aged 65 Years and Over . . . . . . . 4 The Top 21 Counties in the United States in Terms of the Percent of the Population Aged 65 Years and Over . . . . . . 5 MObility of Persons Aged 65 Years and Over, March, 1969 to MarCh ' 1970 O O O O O O O O O O O O O O O O I O O O O O O 6 Source of Aged Migrants (65 Years and Over) to South Atlantic and Pacific Regions, 1935-1940, by Region of Residence in 1935 and by Sex . . . . . . . . . . . . . . . . 7 Relative Importance of Different Factors in Motivating Migration to Florida Among 193 Retired Men . . . . . . . . . 8 Three Indications of the Rank Order of Motivations Among 193 Male Florida Retirees . . . . . . . . . . . . . . . . . 9 Residential MObility Plans, Desires, and Behavior, 1961-1962, and the Proportion Successfully Anticipating Mobility Be- haVi or ' by Age 0 O C O O O O O O O O O O O O O C O O O O . . 10 Residential Mobility Plans, Desires, and Behavior, 1961- 1962, and the Fraportion Successfully Anticipating Mobility Behavior by Selected Variables for the Older Population . . 11 Urban Communities (2,500+ Population) in Florida Ranked Via the Percent Aged 65 Years and Older--l970 . . . . . . . 12 Florida SMSA's Ranked Via the Percent Aged 65 Years and 13 Older-197o coml-‘9081tion of the Factors . . . . . . . . . . . . ix 0 9O 99 :3 ;: 12.51.35 (Cont' d 1: gomosition of t1 kites of the R Crigin. . - - - l'; _ . «v. Erited States Si Survey Responds] L7 Figrant Origins 15 Figrant Motivat . . «.1..- Push Factors As Process 0»... n J The Presence 0: Prior to the Mt ,. . . is The Influence Decision to Mo 3:“: The Pelations'r ~"~'?ration Mobi a.” 0»- ine Pelationsi “95% of Ten- 4lSt of Varia IQ «‘aCtors and E LIST OF TABLES (Cont'd.) Table Page 14 Composition of the Sample Clusters . . . . . . . . . . . . 122 15 Indices of the Relative Importance of the States of origin 0 O O O I O O O 0 O O O O O O I O I O O O O O O O O 144 16 United States SMSA'S Ranked As Points of Origin for the Survey Respondents . . . . . . . . . . . . . . . . . . . . 151 I7 Migrant Origins: Size of Community . . . . . . . . . . . 156 M3 Migrant Motivations . . . . . . . . . . . . . . . . . . . 160 19 PuSh Factors As a Component of the Retired Migration Prwess O O O O O O O O O O O O O O O O O O I O O O O O O 166 m) The Presence of Friends and/or Relatives in St. Petersburg, Prior to the Move . . . . . . . . . . . . . . . . . . . . 170 21 The Influence of Resident Friends and/or Relatives on the Decision to Move to St. Petersburg . . . . . . . . . . . . 170 22 The Relationship between Pre-Migration Planning and Post- Migration Mobility . . . . . . . . . . . . . . . . . . . . 181 IS The Relationship Between the Type of Residence and the Length of Tenure . . . . . . . . . . . . . . . . . . . . . 184 A1 List of Variables Used in the Analysis . . . . . . . . . . 208 A2 Factors and Factor Loadings . . . . . . . . . . . . . . . 210 are 1 Suited States Population Age 2 Gains and Lo: 1940-1950 . . 3 Number and Re 1940-1950, By i Absolute Incr in Florida, 1 5 Percentage of 1900-1970 . . E Mia“ Age 0: 7 Florida 5m “2’ Factor 2 SC '- (L) Factor 3 MC 3:: Factor 4 Census Trac Largest . . Distributio .1 TWO Indi CE S Distributio ProportiOna Figure 10 11 12 13 14 15 16 17 LIST OF FIGURES United States Counties with at Least 15 Percent of Their Population Aged 65 Years and Over . . . . . . . . . . . . . Gains and Losses from Net Migration Among the Aged, 1940-1950 e e e e o o e o e e o o o e e o e e e o o e o e 0 Number and Relative Importance of Aged Migrants to Florida, 1940-1950, By county 0 o e e e e e e e e o e o e o e e o 0 Absolute Increase of Persons Aged 65 Years and Over in Florida, 1900-1970 . . . . . . . . . . . . . . . . . . . Percentage of Persons Aged 65 Years and Over in Florida, 1900-1970 e o e e e e e e o e o e e e o e o e e e o I o o 0 Median Age of Florida, 1900-1970 . . . . . . . . . . . . . Florida SMSA's and the Location of St. Petersburg . . . . . Factor 1 Youthful Fringe . . . . . . . . . . . . . . . . . Factor 2 Socioeconomic Status . . . . . . . . . . . . . . Factor 3 Mobility . . . . . . . . . . . . . . . . . . . . Factor 4 Old Age and Inner City . . . . . . . . . . . . . Distribution of the Age Group, 65 Years and Over . . . . . Census Tracts in Which the 65 and Over Group is the Large 8 t I I I I I I I I I I I I I I I I I I I I I I I I I I Distribution of the Sample Clusters . . . . . . . . . . . . Two Indices of the Sample's Age Distribution . . . . . . . Distribution of Migrant's Origins . . . . . . . . . . . . . Proportional Contributions to the Sample, By State . . . . xi Page 10 54 55 84 85 85 94 101 101 104 104 109 109 120 133 143 146 2:13P FIGURES (Cont Finn ‘n‘ Tuber of Migra Size I I O I I LIST OF FIGURES (Cont'd.) Figure 18 Number of Migrants Contributed to the Sample, By SMSA Size I I I I I I I I I I I I I I I I I I I I I I I I xii THE E Adam liv to his c Seth, ar 800 year the time died. Few things es "Ess it can be CHAPTER I THE EMERGENCE OF POST-RETIREMENT MIGRATION Adam lived a hundred and thirty years and begot a son to his own image and likeness, and called his name Seth, and the days of Adam after he begot Seth were 800 years, and he begot sons and daughters, and all the time that Adam lived came to 930 years, and he died. Genesis V: 3-5 Few things escape the influence of aging. And as a ubiquitous process, it can be both desirable and distasteful. Undeniably, while the Years pass, a bottle of wine may become a connoisseur's treasured Possession. A new automobile, on the other hand, will suffer the costly consequences of "aging depreciation" the minute it is removed from the dealer's showrooms These bewildering qualities of aging are Particularly evident when they are viewed from the human perspec- tive. In some societies age is equated with experience, knowledge, and r68pect. In others, the aged person is ignored and even chastised. Whatever the impact, a man's age is a unique attribute which cOwnences at the moment of his birth. It is unequivocally determined by the increasing passage of time. Each man "is subject to the pro- cesses of bodily growth, maturation to adulthood, slow decline in "1901:. and eventual death" (Bogue, 1969, 150). Stated differently, this 13 the process extending from infancy through old age. 0f 1 5:56, every individu a; hhile biological need at an entirelj Quite obviously :arsally accepted run, many def initf :c‘."-. the pathologic. gear at about age 13?. Burgess observ one crit legal. eligibll fits at age 65 « cal cri average Western - 1 '3‘.“ 0L ‘1‘“. I urst attempté isotzological def: lastence (Havig-h' Without ente La fairly acct istrited States “use: ~ *Tmonly ident \ 1 a35: filth: the firm deny are e “Se course, every individual ages at a different rate and in a different way. While biological aging may proceed at one pace, mental aging may proceed at an entirely different pace. And some never reach old age. Quite obviously, it has been nearly impossible to establish a universally accepted set of limits for old age. But in the process of trying, many definitions have appeared. Barker has recognized that "both the pathological and the 'normal‘ symptoms of old age begin to appear at about age 50 and become prevalent after 70" (Barker, 1966, 13). Burgess observed that one criterion of the lower limit of old age would be legal. At what age under the law is a man or woman eligible for an old age pension or for insurance bene- fits administered by the government? . . . Also, the age 65 gets unexpected support according to a biologi- cal criterion of aging. For all males at birth the average life-expectancy for 13 European countries of Western culture is 65 (Burgess, 1960, 5). Havighurst attempted to avoid the use of chronological age by proposing a sociological definition of aging in terms of reduction in social competence (Havighurst, 1960) . Without entering into this methodological controversy, we can gain a fairly accurate indication of the number of elderly people in the United States by utilizing the age groupings employed by the United States Census Bureau.1 On this basis. the elderly P<>Plflati°n is Commonly identified as the group aged 65 years and over. A"I039 the many terms used to identify this older age group the elderly, the aged, senior citizens, and retirees. These "111 be used interchangeably in the text of this study. Also, “tater!!! :‘retired" migration is in reference to movements which occur is r lieutenant. Throughout this dissertation, retired migration used interchangeably with aged migration. are : items The growth of graphic trends 2 could attempt t assolute incre 5:2: the ”elderlg 15:: age groups. is total populat ricator. Table tithe United Sta lit-nation rela1 In 1900 one but“?! later th 237.3%rcent, whe treat. The p: 1““ every 12 E“ C H ' «- Outlnued. FYI-Cent VS. 21 Math“ in 1 THE "ELDERLY EXPLOSION" The growth of this age group is one of the most significant demographic trends to occur in recent United States history. Although we could attempt to assess this growing importance by referring to the absolute increase in senior citizens, a better measure would con- sider the "elderly explosion" in conjunction with the growth of all other age groups. Consequently, the increase in the proportion of the total population that is 65 years and over is a very valuable indicator. Table 1, which supplies a capsular history of the growth of the United States population, both young and old, also includes information relating to this old age index. In 1900 one out of every 25 Americans was 65 or over. A half century later the number of persons aged 65 and over had increased by 297 percent, whereas the population as a whole gained by only 98 Percent. The proportion rose to 8.1 percent, or to the point where one in every 12 had passed his sixty-fifth birthday. And the trend has continued. Between 1960 and 1970 the under 65 population in- creased only about half as fast as did the 65-plus population (12.5 Percent vs. 21.1 percent). As a result, the proportion of the total Population in the 65-plus age group rose from 9.2 percent in 1960 ‘10 9-9 percent in 1970. Today, every tenth American is an Older American. Efforts to explain this changing age structure have usually emFhasized two other demographic trends: 1) increasing life ex- PECtancy and 2) declining fertility. The former relationship is hardly surprising. One would expect that if people live longer, not only Will greater numbers of them reach old age, but also. the tile 1. Resident 1870-197t 225‘s Nmrber far} (000) K i 3.929 ~~ 5,308 3.2.. 9:638 :59 23,192 1*: 31.443 .370 391818 1%: 50.156 1&3 62,948 335‘ 75,995 in 91,972 33 105,711 in 122.755 f“ 131,669 \ Ur \ .333 151' 32‘ :33 179.32 m3 203116! e, Herman ReSult Table 1. Resident Population Trends--All Ages, 1790-1970; and 65+, 1870-1970 All Ages 65+ Percent Percent Change Change Percent From From of Census Number Previous Median Number Previous All (Year) (000) Census Age (000) Census Ages Continental United States Only 1790 3,929 -- -- -- -- -- 1800 5,308 +35.1 -- -- -- -- 1810 7,240 +36.4 -- -- -- -- 1820 9,638 +33.1 16.7 -- -- -- 1830 12,866 +33.5 17.2 -- -- -- 1840 17,069 +32.7 17.8 -- -- -- 1850 23,192 +35.9 18.9 -- -- -- 1860 31,443 +35.6 19.4 -- -- -- 1870 39,818 +26.6 20.2 1,154 -- 2.9 1880 50,156 +26.0 20.9 1,723 +49.4 3.4 1890 62,948 425.5 22.0 2,417 +4o.3 3.8 1900 75,995 +20.7 22.9 3,080 +27.4 4.1 1910 91,972 +21.0 24.1 3,950 +28.2 4.3 1920 105,711 +14.9 25.3 4,933 +24.9 4.7 1930 122,755 +16.1 26.5 6,634 +34.5 5.4 1940 131,669 + 7.2 29.0 9,019 +36.0 6.8 1950 150,697 +14.5 30.2 12,270 +36.0 8.1 \ United States, Including Alaska and Hawaii \ 1950 151,326 -— 30.2 12,295 -- 8.1 1960 179,323 +18.5 29.5 16,560 +34.7 9.2 1970 203,166 +13.3 28.1 20,050 +21.1 9.9 SCNURCE: Herman B. Brotman, "The Older Population Revisited: First Results of the 1970 Census," Facts and Figures gg_01der Americans, U.S. Department of Health, Education, and welfare, Administration on Aging, 1971, p. 5. .orgi J... ,nu K nerv- t"- - 0|:O ,.... an. --.a, 4 "wu- in: --J 1" number of years spent in old age should also increase. And as shown in Table 2, there is no doubt that life expectancies are on the rise. Whereas a male child born in 1900 had a life expectancy of 48 years, one born in 1967 could expect to live 67 years. Although the table stops at this latter date, the trends are so consistent, we can assume that the increases have continued. Of course the primary factors responsible for this trend have been increased utilization of new knowledge from advanced medical and health sciences, and increased te<=hnological "controls" over the environment. But significantly, for the most part older people are not living any longer than the older people of the past lived. A person of 70 in 1900 could expect to live 9Enora years and today a person of 70 can expect to live about 10 [£an years. In other words, a principle factor contributing to the sharp increase in the number and proportion of older people has been the reduction of death rates of infants and young children (Smith, 1951, 30). A sudden decline in the death rate at any age tends to pass on to the higher age classes a larger proportion of the individuals who hQVG already been born. Where death rates are extremely high, only a very small proportion of the individuals succeed in attaining the upper rung of the population pyramid . . . A decline in the death rate at a particular age has the tendency to increase the proportion of the pOpulation at that age and at all later ages (Bogue, 1969, 154).2 If biological research supplies appropriate knowledge of the aging \ no": 2Prior to ending this discussion of life expectancy, we should 1': e that "shorter life expectancy for Negroes produces a smaller ( lgportion of older persons. The 22.7 million Negroes of all ages t1} 70) represent 11.2 percent of the total resident population; to: 1.6 million older Negroes account for only 7.8 percent of the Q1 older population" (Brotman, 1971, 3). frable 2. Life Expectancy for Given Sex and Age Male Female Years 0 20 40 6S 0 20 40 65 11900-1902 47.9 38.4 27.7 11.5 50.7 39.9 29.1 12.2 11929-1931 57.7 44.9 28.7 11.7 61.0 47.1 30.9 12.8 11949-1951 65.5 48.9 30.8 12.7 71.0 53.7 35.1 15.0 31967 67.0 49.6 31.4 13.0 74.2 56.3 37.3 16.4 SOURCE: Philip M. Hauser and Raul Vargas, "Population Structure and Trends," in Ernest Burgess (ed.), Aging ig_Western Societies, Chicago: U. of Chicago Press, 1960, p. 42 and U.S. Department of Commerce, Bureau of the Census. .‘. . ‘e. ‘4 process in human beings, it is possible that the life span itself will ultimately be extended" (Thomlinson, 1965, 121). At our current level of knowledge, however, it is doubtful that a significant ex- pansion of the biological life span will occur any time in the near future. In either case, we can expect larger proportions of the popu- lation to live long enough to approach the physical limit of life. A second school of thought has conjectured that the "elderly explosion" is closely linked to decreasing fertility trends. One Of the leading advocates of this point of view is Albert Hermalin. In his opinion: contrary to popular impression, the increase in the U.S. in the percent of the population over 65 from 4.1 percent in 1900 to 9.2 percent in 1960 was due to the decrease in fertility and not the lower death rates. In this period, mortality improvement had only a slight effect on age composition and that was in the direction of a younger population (Hermalin, 1966, 451-469). The logic behind this argument is that "whenever birth rates fall, in comparison with a previous level, children constitute a smaller proportion of the total population than formerly. As a result, the p°Dulation pyramid has a comparatively smaller base" (Bogue, 1969, 153). From 1880 until 1940 United States fertility rates declined Stea.dily. In addition to contributing to an increase in the elderly 3egnlent of the population, this trend also helped to produce a sig- nificant increase in the median age of the United States (Table 1). a‘QtV-ween 1870 and 1950, the median age climbed ten full years from 20 ~ 2 to 30.2. But the post-World War II "baby boom" reversed both of t 11388 trends. By 1960 the proportion of the population in the child- hQQd ages was back up to very nearly the same level as it had been ‘1“ 4 . .A' II On. n- no- unl- '|‘A “a! . n . .‘t u, I’- em. in . 42:9: 0.. ‘ V 5 I4*& tit" sum ,. 0“.“ ”an KR" I.) o 0‘ (D u I 8’6 e to, e A I... in 1910. Additionally, between 1950 and 1960 the median age dropped by almost a year (0.7). This same decline continued during the next decade, and as a result, in 1970 the median age was down to 28.1 years (Table l) . The impact of these various trends will become especially clear in the future. Official United States Census projections beyond 1980 call for a decrease in the proportion of the total population 55 and over due to the low birth rate of the depression years. This decrease "should continue until about 2010 when people born during the baby boom of the 1940's reach the 65-plus category. After 2010 there is a very rough estimate that as many as one of every five Americans will be 65 or older" (Barker, 1966, 5) . As might be expected, the increase in the elderly segment of the Population has not been spatially constant. In support of this state- ment, we can point to Beale's recent contention that "the occurrence 0f extreme age distributions--young or old--in this country is almost antirely a phenomenon of rural and small town areas" (Beale, 1969, 417) . Unlike the national trend, however, these small scale variations are Primarily the result of age selective migration. In particular, it has long been recognized that migration streams tend to contain a Preponderance of young adults. And consequently, "a continuing prO‘T-QSS of migration tends to produce an excess of persons of ages 20‘44 in growing areas and deficiencies of the same group in losing arses" (Thomas, 1938, 19). A situation similar to this has existed in the United States 31'.an the Depression. At that time, the rural areas started to exP'srience a significant out-migration of their youthful population. I‘ 5‘ A_; 0:” II «he 0 ti} ‘MM '01. . ‘ e ‘ .,_ ‘ M‘ 7‘] . 4 n m: a- -‘- e... Pm. 'I IL. .n V ' In: “-1 This trend has continued up to the present. "From 1950 to 1960, the rural and predominately rural counties of the United States that had net out-migration lost 40 percent of their youth who reached 20 years of age during the decade" (Beale, 1969, 415) . Unless offset by high fertility among the persons remaining, and by comparatively rapid out-migration of middle and old age people, these rates served to raise both the average age of the remaining population, and the Proportion of older people. In Figure 1, the spatial components of this trend emerge strik- in91y clear. Of those counties in the United States with 15 percent 01‘ more of their total population aged 65 years or more (the national average in 1970 was 9.9 percent), the overwhelming majority appear in the agricultural mid-section.3 In fact, a concentrated region of Older counties appears between North Dakota and Minnesota in the north, and Texas in the south. If the states of the United States are ranked on the basis of the proportion of their total population aged 65 and OVer (Table 3), the oldest state is Florida, but the next seven states appear in this north-south .old age belt. Although he operated on the basis of median age, Beale discovered a very similar concen- trmtion in 1960: From the Southern Corn Belt extending southward to the Hill country of Central Texas is a roughly triangular- shaped area of about 300,000 square miles in which the majority of all non-metropolitan counties, including several large contiguous blocks, have a median age \ i 3As a point of information, there are 3,071 counties and parishes 13 the United States. or these, 501 (16.3%) had between 15.0: and -9% of their total population aged 65 years or over in 1970; 98 (3~2%) had between 20.0% and 24.9% in this age group: and only 9 ‘33) had 25.0% or more. , e \ t .34 ~_ d)nU a» 21‘ “A: 4‘ six? EFF «I- ..inVl‘ ZED. n (.JaniAuvf- QB. .lILr {AU 7 . \ N , \ .PZU .V. .1 r. val\ u — >1! .s..>) 44.1%..Zszivww smut-51L: .U Huh-z.- \ , s —V\-.~v« s Ohm. .20....143uOmOL KO mawzwu m5 “UUKDOm “Numufig co. 3 cm a. 5872.» no ezuuxus 10 mm>o oz< 92m» 3 30¢ 295.53.. Eur» to Fzmomma a. 5%.. 5 rt: 3:238 mu._.<._.m 3.22: ._ magi .7. 2'5 .6..- ‘An n Table 3 . 11 Leading States of the United States in Terms of the Percent of the Total Population Aged 65 Years and Over 1. 2. 10. :SCWIRCE: Florida (14.6%) Arkansas (12.4%) Iowa (12.4%) Nebraska (12.4%) South Dakota (12.1%) Missouri (12.0%) Kansas (11.8%) Oklahoma (11.7%) Maine (11.6%) Massachusetts (11.2%) U.S. Department of Commerce, Bureau of the Census, 1970. 1970 Census of Population and Housing. ‘C o'. 1‘32 a} p f 12 above 35.0 (U.S. = 29.5). The proportion of elderly people is anywhere from 2/3 to 100 percent higher than the national average. Young adults and young children are couparatively few in number. This is the end result of prolonged age-selective out-migration on a population of low to moderate fertility, here and there abetted by in-movement of older people for retirement (Beale, 1969, 417). The predominance of Florida, and in particular the counties of Central Florida, is especially significant. Unlike most of the other areas, Central Florida is not characterized by youthful out-migration. In fact, young adults are actually moving into this area. The primary reason why the proportion of elderly people has continued to increase is that there has also been an even greater in-migration of older People. The size of this group has been sufficient enough to not only counterbalance the impact of the young migrants, but also to cause six of the seven leading counties in the United States in 1970 to come from Central Florida (Table 4). Except for Texas, the other States noted as retirement centers do not emerge as clearly. And Texas is a relatively unique state since its predominance can be Jdirflted with both youthful out-migration and elderly in-migration. The situation in California and Arizona is complicated by heavy in‘migrations of all age groups: consequently, the elderly concen- trations are not as outstanding as might be expected. T. Lynn Smith has recognized the complexity of the internal varj-ations in the proportion of people over 65: The situation in each state represents a specific balancing of forces. If one merely considers the migration factor, one state . . . may now have a high proportion of the aged because of a boom at the opening of the century which attracted to it tens of thousands of young men men and women: whereas another must attribute the high proportion of old people in its population largely to the immigrants who settled there just before the outbreak Tablxe 4. 13 Percent of the Population Aged 65 Years and Over 1. Charlotte, Florida 2. Pasco, Florida 3- .Manatee, Florida 4. Pinellas, Florida 5- Sarasota, Florida 5- Hamilton, Texas 7- Citrus, Florida 8. Elk, Kansas 9. Mills, Texas Llano, Texas SCWHMCE: (35.1%) (31.6%) (30.2%) (29.5%) (28.6%) (26.5%) (26.0%) (25.7%) (25.5%) (24.9%) 11. 12. 13. 16. 17. Sierra, New Mexico Throckmorton, Texas Kerr, Texas Coleman, Texas Bosque, Texas Eastland, Texas Osborne, Kansas WOodson, Kansas Chautauqua, Kansas Furnas, Nebraska St. Clair, Missouri 1970 Census of Population and Housing The Top 21 Counties in the United States in Terms of the (24.7%) (24.3%) (23.8%) (23.8%) (23.8%) (23.7%) (23.2%) (23.2%) (23.2%) (23.2%) (23.2%) 14 of the first Wbrld war . . . [anotheg] may be experi- encing a rapid aging . . . primarily because the men and women who pioneered in its development during their youth are now reaching the advanced ages . . . another should attribute the increasing percentage of the aged . . . largely to the fact that considerable numbers of its youth have been migrating to other areas . . . finally, it may resulg] from having been receiving large numbers of elderly migrants from other sections of the country (Smith, 1955, 8-9). Although many very important trends emerge from the preceding discussion, one supercedes all others. Succinctly stated, the United Sturtes is experiencing a marked increase in its elderly population, and this growth is displaying significant spatial patterns. The inqplications of this demographic process are as widespread as they are pro found . THE EMERGENCE OF RETIREMENT SYSTEMS One of the most important implications is closely related to the emergence of retirement as a universally acceptable consequence Of old age. When the development of stable retirement policies is coupled with increasing life expectancies and the general expansion of the elderly population, a situation arises in which a greater trunnber of persons spend an increasing number of years in an economically ‘Inuproductive role. As one might expect, the repercussions of this QQVelopment are very significant. In order to appreciate them, one Bl\cauld begin by recognizing that retirement is a relatively new phenomenon in our society. Up until recent years, there was no widespread retirement of c)Zl-der people until they were compelled to leave the labor force by I”lysical incapacity, illness, infirmity, or senility. Consequently, a.8 recently as 1900, "68 percent of all men 65 and over were in the 15 labor force" (Burns, 1954, 385) . By 1930 this proportion had declined to 54 percent and by 1966 to 30 percent. In 1970, only 13.5 percent of all older families had heads who were year-round full-time workers. This sharp drop in elderly employees is primarily the result of tech- nological innovations and the extensive application of mechanical Power. Together, these forces have created a tremendous increase in worker productivity. With fewer people capable of performing the tasks that had formerly required the services of many, it became feasible to develop a series of stable retirement policies. Suddenly people began to retire for reasons other than health. Some desired the leisure which a<=<=ompanied their new position: others were literally forced into retirement by generational competition between the young and the old. Until approximately 1935, the retiree's means of support was an entirely personal matter. Ironically, the Depression caused many older Americans to become aware of their common goals and problems. As a consequence, "Dr. Francis Townsend initiated a plan for a guar- anteed monthly income) this innovation posed a threat to the estab- lished government pattern of dealing with the elderly, namely no direct relations" (Smith and Marshall, 1970, 1). However, in 1935 the federal government did respond to Townsend's initiative with the first Social Security Act. In part, this program represented "an a“'ltempt to partially alleviate a glutted labor market by enticing out old workers through minimal economic incentive, while at the same titne establishing a clear principle upon which mandatory retirement leicies could be instituted" (Hill and Marshall, 1968, 26-27) . The Social Security Act institutionalized the concept of retirement 16 as a part of public policy, while also establishing retirement on a pensionable basis as a means of partial protection against a major in- dustrial risk, namely old age. Because the benefits were available to only those people who had retired, it laid the basis for compulsory re- tirement as part of our national labor policy. Initially, "the pensionable protection against old age under the Social Security Act was, from the economic point of view, both inadequate auui inflexible" (Burns, 1954, 384). Fortunately, this stimulated eco- nomic and social innovations, including the introduction of employer- iJ1i1:iated private pension systems to supplement social security payments. Stxill, since many of the first pension arrangements were voluntarily Created by management, their coverage was very restricted. In 1954, BuI‘ns observed that "at present, only about 500,000 or less than 4 per- °ent of older people 65 and over are receiving these Li—ndustriafl pen- Sions" (Burns, 1954, 388) . The majority of retirees were eligible to reoeive nothing more than very inadequate social security benefits. Al- though some were able to supplement their income through savings, most ret:irees were faced with a barely subsistent level of living. Ob- ‘"i<>usly, the prospects of retired life were hardly appealing, and as a reSult, many workers resisted compulsory retirement practices. The evolution of labor unions, and the subsequent development of extensive, and in some cases, very lucrative, pension systems has changed this. Those workers who are not represented by a labor ‘1rtion are very rare. As Ullman noted in 1954, "industrial unions have been obtaining retirement provisions so generally that the I“-‘lll'tber of retired workers will increase enormously in the future" (Ullman, 1954, 124) . In addition, people are retiring at even TM 17 earlier ages. In fact, one of the benefits mentioned most frequently today is a "30 and out" program. Obviously, if concessions of this nature are instituted, people will be retiring after only 30 years of service; and conceivably, retirees may soon be leaving the labor force at the age of 50. Recently, in addition to the fuller development of pension systems, there have been developments in medical programs (e.g., Medicare), community health projects, subsidized housing for the elderly and food stamps. Periodic increases in social security bene- fits have also been instituted in an effort to at least keep pace With the rising cost of living. But most elderly people are still far from wealthy. In 1970, hillf of the almost 7.2 million families with 65-p1us heads had money incomes of less than $5,100 per year. (In contrast, younger families had a median of $10,500.) At the bottom of the income distribution, about 150,000 families (2 percent) had annual incomes of less than $1.000 or about $20 per week. A total of some 750,000, or 10 percent of the older families, had incomes of less than $2,000; and a total °f more than 1.7 million, or almost one-fourth of all older families, had less than $3,000 (Brotman, 1971b, 2). Perhaps a more accurate indication of the relative deprivation of retirees can be obtained fli‘om the federal government's statistics on poverty. In 1969, 1 0 250,000 older families were below the poverty level. This repre- 8erated 17.6 percent of the total. Due to an increase in social Se(:urity payments, which over-balanced consumer price increases, born of these figures declined in 1970. In that year, poverty fami- lies decreased by 80,000 to 1,170,000, or 16.3 percent of all the 18 families with heads 65 years or older. This improvement will probably be temporary, however, since the cost of living has an uncanny ability to overtake benefit increases. And even with the rather extensive decline in poverty levels, a sizeable segment of poor elderly families Clearly remains. These statistics are certainly disturbing, but they can not hide the fact that significant advances have undoubtably occurred. The most important trend to emerge from the preceding discussion is that increasing numbers of senior citizens are spending an increasing r“.Ill'ober of years in retirement. And, in general, the programs designed to aid these retirees represent great advances over those that existed at the turn of the century. Further, they will probably continue to improve . A GROWING INTEREST IN RETIREMENT Retirement is much more than the simple act of leaving the active labor force. It also heralds far-reaching changes in a person's role as a functioning member of society, carrying with it profound impli— cations for interrelated changes in one's social status, social activities, health conditions, financial capability, and self percep- tion (Eteng and Marshall, 1970, l) . As might be expected, with retirement emerging as a stable social pattern, it has also come to It is interesting to examine a timely quote from 1954. Recog- nizing the insufficient governmental support, Tibbits declared that "even today the major obstacle to making adequate provision for this new element (retirees) in American society would seem to be the outlays essential for military defense" (Tibbits, 1954, 305). 19 occupy a place of central concern in contemporary Western society (Hill and Marshall, 1968, 3). One of the best indications of this growing interest in retire- ment, aging, and old age is available in the literature of the many academic disciplines that consider these topics. Without attempting a complete review of these materials, it is possible to make several Salient observations. First, two separate disciplines, geriatrics and gerontology, have emerged to consider the various aspects of aging and Old age. As a branch of medical science, geriatrics is primarily cOncerned with the physiological and pathological consequences of aging. Gerontology tends to be much broader. In addition to scien- tific studies of physiology and pathology, a gerontologist might also cOnsider the psychological processes of aging, or perhaps the many SOcial and economic aspects of retirement and old age. Of course, given these latter topics, it is not surprising to learn that several of the social and behavioral sciences have also cOntributed to our understanding of old age-related phenomena. The Overwhelming majority of the work has come from sociology, economics, and psychology.5 In particular, the topics of traditional concern have been the problems and the adjustments associated with retirement and aging. As an example, Tibbits has identified several of the Principal problems facing the elderly individual: the maintenance of income to meet the requirements of active and healthful living; discovery of new occupa- tions or social roles; finding the opportunity for social contacts, companionship, and affection; Other disciplines, the most notable of which is geography, have contributed very little. 20 maintenance of health; and procurement of suitable living arrangements (Tibbits, 1954, 303-304). HiJL].and Marshall have summarized the many studies of personal ad- justment by noting that: very few of these studies have focused on the changes which occur in the individual's life after he retires and the relationship of these changes to the individual's personal adjustment to retirement. Rather, most studies have focused on the present life situation of the retiree, that is, his present level of income, health, etc., and their relationship to his present level of adjustment . . . (Hill and Marshall, 1968, 1). THE GROWING IMPORTANCE OF POST-RETIREMENT MIGRATION One of the many topics that has received only limited attention is post-retirement migration. This is especially surprising when one realizes that in recent years, it has become clear that many elderly people do move upon retirement. As an example, between 1958 and 1959 about 10 percent of the 11.8 million households con- taining one or more persons 65 years and over moved (Barker, 1966, 7). According to the Current Population Reports of the United States Bureau of the Census, of the 19,700,000 people over 65 in 1969, 8.6 per- cent" or 1,700,000 lived in a different house in the United States in 1970 (United States Bureau of the Census, 1971b). Although these mdbility rates are well below the level of younger individuals, it is obvious that with over 1.5 million senior citizens moving each year, retired migration is a very significant demographic trend.6 According to estimates of the Current Population Survey, of the 200 million persons one year old and over living in the United States in March, 1970, 36.5 million, or 18.4 percent, had been living at a different address in March, 1969. 21 There is reason to believe that a significant portion of this mobility may be closely linked to the fact that retirement "constitutes an important break in the individual's life cycle" (Honnen, Eteng, and Marshall, 1969, 39). In fact, one of the most consistent findings in migration literature is the principle that "the heightened pro- pensity to migrate at certain stages of the life cycle is important in the selection of migrants" (Lee, 1966, 297). For a relatively long period, age, which is a crude indicator of family life cycle stages, has been found to be one of the crucial determinants of residential mobility of all types, including internal and interna- tional migration (Bogue, 1959; Shryock, 1964; Thomas, 1938; Thomas, et al., 1957). In the process, many people have associated the high mobility rates of young adults with new marriages, families expanding with the birth of children, and moves related to the husband's employ- ment. Traditionally, once this period is completed, mobility has tended to decline in a very consistent manner. Recently, however, it has become apparent that a small but significant increase may occur after retirement. As Barker has observed: Upon retirement the working man automatically becomes a senior citizen. Such a change in occupational activ- ity requires the elderly male to face potential embar- rassment and loss of status if he remains in the same immediate community, because: 1) he usually has a re- duced scale of living (less income); 2) he may desire not to be like the other '1ost souls' who haunt the shop or campus without purpose or function; and 3) he may experience mental depression brought on by living in the presence of former duties, responsibilities, and active associates. Consequently, retirement by its very nature in American society, creates built-in pressures on the elderly man to change not only his type of housing, but its location as well (Barker, 1966, 4). Stated more 51 5:. also 9‘93 1554, 361). mditio 5m identifi amen grO“ Parental famj {C mntain. 53.2115 PM“ solischeider . :10: desil’e ‘ fies their n! (Goldscheide Lia: the dev mst-retiren 22 Stuated more succinctly, "retirement not only cuts the ties to the job, hunt also greatly loosens those to the community of residence" (Hoyt, 1954, 361). Additionally, retirement usually coincides with what sociologists have identified as the "empty—nest" phase of the family cycle. As children grow up and leave home to establish their own nucleated, parental families, large houses may become too big and too costly to maintain. Consequently, desires for smaller and cheaper units should produce sizeable residential shifts. In the opinion of Goldscheider, however, "a significant proportion of the older popula- tion desire to move but do not since appropriate housing which satis- fies their needs and which they can afford is not readily available" (Goldscheider, 1966b, 84). In a contrasting view, Barker indicates that the developers of retirement communities are reacting to the post-retirement housing market: the developers . . . are filling the current housing needs of the same families for whom they built tract houses 15 to 20 years ago. The families who bought '3 bedrooms and 2 baths' in 1950 have now raised their children and are looking for smaller units designed to meet the needs of their new position in the life cycle (Barker, 1966, 30). Undeniably, many variations in living arrangements are sought, and, in fact, the majority of elderly citizens may desire to remain in their pre-retirement residence, but the key element in this development is the potential mObility that is created by retirement and smaller families. Closely related to this increased potential is the emergence of a nearly universal system of financial support for the retiree. Earlier comments have shown that although still inadequate, the :3me designed regresent signif i firm the perspect aggear to serve 7: increasing propm homer words, ' iajority of olde retirement no 10 1355, 617). The 5‘5- Rate a matte flied ‘10 remail is live with hi one COW-hunky, 3353030115 r01e :3: than ever t shtially Stanc .10 have the ‘ Ms and move My cg“ ”finial Envi \6 Shaw % Co It is a; 23 programs designed to support senior citizens are relatively new, and represent significant advances over past programs. When viewed from.the perspective of increasing mobility potential, these programs appear to serve as the main catalysts responsible for transforming increasing proportions of thlS potential into operative human action. In other words, "the economic necessity which compelled the great majority of older people to remain in their home community after retirement no longer exerts its full force" (Burgess, Hoyt and Manley, 1955, 617). The post-retirement place of residence is becoming more and more a matter of choice. Whereas, traditionally, the older person tended to remain in the home of his middle years or, if he moved, 'u>live with his children or be limited to changes within his immediate home community, the potential for the maintenance of independent and autonomous roles for the elderly is more prevalent and widespread now than ever before. With relative financial security, and with spatially standardized payment procedures, the elderly can now choose "to leave the communities in which they have resided for 30 or 40 Years and move a thousand miles or more into a strange but reputedly congenial environment" (Harlan, 1954, 333). 222.82atial Compgnents g£_Post-Retirement Migration: ‘§§_Introduction It is apparent that post-retirement migration may occur at a variety of geographical scales. As an example, consider the informa- tion presented in Table 5. Consistent with the patterns displayed by migrants of all ages, the majority (69 percent) of all elderly People that changed their residence between 1969 and 1970 remained within the same county. Since intra-county shifts can incorporate 9M. Table 5. 24 Mobility of Persons Aged 65 Years and Over, March, 1969 to March, 1970 Number I. Total Population Aged 65 Years and Over (1969) 19,713,000 A. Non-Movers 1. Residing in Same House in Both 1969 and 1970 18,008,000 B. Movers 1. Residing in a Different House in 1970 1,690,000 2. County Level a. Residing in the Same County 1,173,000 b. Residing in a Different County in 1970 517,000 3. State Level a. Residing in a Different County in the Same State 319,000 b. Residing in a Different County in a Different State 198,000 c. Residing in a Different County in a Different But Contiguous State 78,000 d. Residing in a Different County in a Different, Non-Contiguous State 120,000 SOURCE: Percent of Total 100.0 91.3 8.6 2.6 1.6 1.0 0.4 0.6 U.S. Bureau of the Census, "Mobility of the Population of .: March, 1969 to March, 1970," Current Population Reports, Series P-20, No. 210, Jan. 15, 1971b. the U.S 25 many types of movements, this is hardly a surprising statistic. Not cxfly do nearly all intraurban and intrametropolitan shifts remain within a single county, but many of the rural to urban shifts also fail to cross county boundaries. Additionally, the dominance of short distance movements is undoubtably related to the fact that while spending a number of years in a particular area, most people develop both social contacts and an understanding and knowledge of the various attributes of the region. In the Opinion of Barker, this . . . is a vested interest which one is increasingly reluctant to give up in the later years of life. To render this familiarity obsolete by a long distance move is an option the elderly, the least adaptable segment of the population, seldom choose willingly (Barker, 1966, 63). In light of this strong inertia, it is surprising that, in a Single year, nearly 200,000 retirees elected to surrender their "familiarity investments" and migrated to other states.7 Although this number is but a small portion of the total retired migrants (12 percent), in several respects, it is the most significant por- tion. In addition to overcoming strong forces of inertia, long distance elderly migrants have tended to concentrate in relatively few destinations. In particular, the interstate flow of retirees has been spatially concentrated in the warm.weather states of the South, the Southwest, and the West. Florida, Arizona, and California have predominated.8 7Of course, prior to the development of retirement systems and the associated improvements in the financial security of retirees, this inertia was undoubtably even stronger. 8While Figure 1 displays the importance of Florida very clearly, a large scale in-migration of younger people has concealed the imr portance of California and Arizona. Althoug' the movement, 1338-1940 dec in the growti that 'Califo: L940 than we heigration second in th leaflet in a 11-3 Fewem the State a- with festinatiOn decade dis; net mi's'ratj Flt)Iida, a til‘e lEade Parent) , net Eigrdt weesixth 26 Although it is difficult to ascertain the actual magnitude of the movement, rough estimates have been developed. As early as the 1930-1940 decade, California and Florida were the overwhelming leaders in the growth of their elderly population. T. Lynn Smith has estimated that "California alone had 57,000 more persons age 65 and over in 1940 than would have been the case had the growth not been swelled tw'migration from other states" (Smith, 1951, 23). Florida ranked second in the absolute growth of aged people (+33,000) and was the leader in the relative growth with a 25.1 percent increase. In 1940, 11.3 percent of the persons aged 65 and over in Florida had moved to the state during the preceding five years (Smith, 1951, 20). With the exception of the emergence of Arizona as a popular destination for aged migrants, the statistics for the 1940-1950 decade display the same general tendencies. California had the greatest net migration gains with an increase of 130,000 senior citizens. Florida, a distant second in absolute growth (+66,000), was the rela- tive leader with a 38.8 percent gain, followed by Arizona (25.3 percent), and California (17.0 percent). In other words, "because of net migration gains, the actual aged population in 1950 exceeded the expected by two-fifths in Florida, by one-fourth in Arizona, and by one-sixth in California" (Hitt, 1954, 196). Perhaps an even better indication of the importance of aged migration is represented by the fact that in Florida, 62.5 percent of the total increase in oldsters in this decade was due to net migration. Arizona and California ranked second and third in this respect with increases of 44 percent and 38.3 percent, respectively (Hitt, 1954, 197). Since i‘fiidblYI t absolute it :ese of Flt tier. incre. then left is at leas P’Jp‘zlation this latte as; coditions 5‘ tCtdl p¢ ‘ 9035 ‘1 to y. n . s‘nui 27 Since 1950 the relative situation has changed very little. Un- deniably, the most impressive development has been the tremendous absolute increases that have occurred. As an example, consider the case of Florida. Between 1940 and 1950, this state added a net migra- tion increase of 66,000 retirees. In the next 10 years, over a quarter of a million (260,000) more persons 65 years and older moved to Florida than left the state. This same statistic for the 1960-1970 period is at least 350,000. During this same period, Florida's over 65 population increased by 78 percent (+435,000) to 968,000.9 Although this latter figure is a composite of both migration additions and additions from the normal processes of aging, it can be compared to a total population gain of "only" 37 percent. The increase in the aged segment of our national population was 21.1 percent. Obviously, the states of Florida, Arizona, and California are capturing more than their share of the "elderly explosion." And significantly, none of these statistics include the unknown number of "temporary migrants" who travel to the South and West during the winter months and return to their permanent homes for the rest of the year. Even if this latter group is ignored, however, it is readily apparent that aged migration is a very important feature of contemporary living in the United States. Given the projected increase in people over 65, the anticipated improvements in retirement programs, including the possibility of a younger retirement age, and the easy access to amenity states, it seems inevitable that aged migration will continue to develop and expand. 9A3 a result, 14.6 percent of the state's total population is now composed of older persons (Table 3). In light of caplete understa: in: patterns that after. involves a tune: social 5 Eities involved uemmplev COnSj :5 retired migra: has introduced a Wit-eon, In ‘ 28 In light of this development, it is essential that we seek a complete understanding of both the processes that are involved, and the patterns that are produced. For the migrant, the decision to move often involves a change from.one social system at the point of origin to a new social system at the point of destination. For the com? munities involved, the decision to move has obvious importance. As an example, consider the implications of the spatial concentration of retired migrants. Furthermore, the phenomenon of aged migration has introduced a new set of considerations into migration literature and theory. In the next chapter, several of these considerations will be identified and examined. Thus of aging ha “ mstem v4 “gratis“ firtual i CHAPTER II PROBLEM STATEMENT IN MIGRATION THEORY "If a family has an old person in it, it possesses a jewel" (Chinese Proverb). Thus far, "it would seem that those interested in the problems of aging have known little of migration study; and those competent in migration study have been little interested in the problems of the aged" (Smith, 1951, 16). This relative lack of interest in aged migration merits additional attention. And logically, we might begin by examining the salient factors that have combined to produce this virtual ignorance. In conjecture, several appear to predominate. One fairly obvious element is the relatively recent emergence of the importance of long-distance retired migration. Prior to this development, students of the aged had already cultivated an extensive interest in a variety of other topics. Of course, the same statement can be made for those involved in migration studies, however in that instance, the situation was complicated by a clear, and very consistent pattern of migration selectivity. In study after study, the rates of migration were found to be low for children and adolescents, to rise sharply to a peak in early adulthood, and then to fall off with increasing age (McInnis, 1971, 196). Whereas persons aged 25-29 29 we discovered tc age identified a! manning majO me migratory YC Perhaps an cf aged mgratiOJ éetailed statist is relatively 1 is United Stat! tion analysis m tatulate. and n to 1940, the S) zenial census based upon son ibase populat “1" m death rat‘ 30 were discovered to be the most mobile population segment, the elderly were identified as one of the most stable. For obvious reasons, the overwhelming majority of migration studies have concentrated on the more migratory youth. Perhaps an even more fundamental explanation for the scarcity of aged migration analyses is the historic lack of comprehensive and detailed statistics. In fact, this same element can be used to explain the relatively recent development of almost all migration studies in the United States. Clearly, those interested in any phase of popula- tion analysis must rely heavily upon official agencies to collect, tabulate, and make available, the basic data. Unfortunately, prior to 1940, the subject of migration figured only slightly in each de- cennial census of population. Studies of large scale migration were based upon some rather risky estimation procedures. Starting with a base papulation, early investigators would frequently use birth and death rates to calculate an expected population at a later date. In this manner, discrepancies between the expected size and the actual size could be attributed to net migration (D. Thomas, 1938; webber, 1956; Hitt, 1952, 1954: Smith, 1954a, 1954b, 1955). The 1940 Census schedule included the first questions about migration per se. Establishing a precedent that has carried over to the 1970 Census, the respondents were asked to disclose their place of residence, on the same date, five years previously, i.e., April 1, 1935. In the process of employing these data, researchers were able "to determine the net movement of persons 65 years of age and over into or out of each of the states during the five years, 1935-1940" (Smith, 1951, 18). Unfortunately, the statistics were not sufficient mash to deter: given state wen for county unit; zigrant exchang census regions: fiest. One of ti. of the componer occurred after meEd censu: and rePrOducti- 3‘-‘?ion. and St is notah tam"? to mg "ere basically Che ”alas PlaCe win-try) I and in the Umiter-1 .L a! A r ~5~1ons wer. reFor.r . ‘3 which 31 enough to determine the extent to which the elderly migrants to any given state were drawn from any other given states; and tabulations for county units were impossible. The best available indication of migrant exchanges concerned interchanges between the four principal census regions: the Northeast, the North Central, the South, and the West . One of the first research undertakings to perform an analysis of the components of population growth with units smaller than states, occurred after the 1950 Census (Bogue, 1957). In taking advantage of improved census tabulations, Bogue made estimates of net migration and reproductive change for each state economic area, economic sub- region, and standard metropolitan area. As notable census improvements have continued, information per- taining to migration has also improved. In the 1970 Census, there were basically two questions in the schedule dealing with migration. One was place of birth (state in the United States or named foreign country), and the other was place of residence in 1965 (county or city in the United States or named foreign country). Although these questions were exactly the same as those used in 1960, "the published reports which have been planned will contain significant new tabula- tions" (Long, 1971, 21). Additionally, there will also be greater opportunities for making special studies based on one's own tabulations from.data in the 1 in 100 sample available on computer tape.10 0 Even with these advancements, however, in many respects, the census materials are still relatively inadequate. A more de— tailed discussion of these weaknesses will be presented later. As might be unrest in mg“ 1243. migratm“ ' of disciplines h attributions ha escronics. and 9 In general aajor componentE Ltd, have placc Lizinq that niq population, soc E’s-her of perso tion, and race. iron the factor avariety of ec lent and income fixation anal} Patial proces i«Thence of d It Would 1° any 0f thes T, A‘- . ‘13“‘pllnary C is , an Curre v.15» 32 As might be expected, with the expansion of statistics, the interest in migration has also increased. In the short period since 1940, migration analyses have expanded immensely. Of course, a number of disciplines have been involved, but perhaps the most outstanding contributions have come from the fields of demography, sociology, economics, and geography. In general, demographers have stressed migration as one of the major components of population change. Sociologists, on the other hand, have placed a heavy emphasis on migration differentials. Recog- nizing that migrants do not constitute a random sample of the total population, sociologists have tried to evaluate the importance of a number of personal characteristics, e.g., age, sex, occupation, educa- tion, and race. The economist's interest in migration has also focused upon the factors that influence the propensity to move. Naturally, a variety of economic indicators have been identified such as employ- ment and income differentials. Finally, the geographer's role in migration analysis is very clear. By definition, migration is a spatial process producing spatial patterns. And in particular, the influence of distance has received widespread attention. It would be a serious mistake to attribute a degree of superiority to any of these approaches. Additionally, it would be just as foolish to assume that the scientists in each discipline operate in a pro- fessional vacuum. In fact, to the contrary, a great deal of inter- disciplinary contact has frequently occurred. The concept of migration as an occurrence affecting nearly every person has been a common and unifying interest. One of the been the emerge: cried of sever. tendency to ign words, in many thin the fra THE Any meal £33110!) ana firing as E 3331M and ! :ePtiVe Prin- m“ by Stag iifferencES fefiles alt-Io} Ltd 7) the 33 One of the most notable consequences of this cooperation has been the emergence of a significant theoretical structure. A cursory review of several of the aspects of this structure will show that the tendency to ignore aged migration is a very serious mistake. In other words, in many respects, retired migration does not fit accurately within the framework of these theories. THE THEORETICAL FOUNDATIONS OF MIGRATION ANALYSIS Any meaningful consideration of the theoretical foundations of migration analysis must begin with the work of E. G. Ravenstein. Laboring as early as the 1880's, he employed migration data for England and several other nations to develop a series of very per- ceptive principles relating to: 1) migration and distance; 2) migra- tion by stages; 3) streams and counterstreams: 4) rural-urban differences in the propensity to migrate; 5) the predominance of females among short distance migrants; 6) technology and migration; and 7) the dominance of the economic motive (Ravenstein: 1885, 1889). Perhaps the most outstanding feature of these generalizations is their durability. In the 80-plus years since their initial publi- cation, Ravenstein has been frequently quoted and occasionally chal- lenged. And even though there have been literally thousands of migration studies in the meantime, very few additional generalizations or "laws“ have been advanced. In the opinion of Everett Lee, “Ravenstein's papers have stood the test of time and remain the start- ing point for work in migration theory" (Lee, 1966, 47). Significantly, Lee is responsible for one of the few general theories of migration to appear in the last half-century. His straightforward cc restatement of the missions regain means and count the process, Lee associated with 1 of destination; Every act ’35 each of the: flat influences L) the schema s lPYOdUCt of t} origin, and th BQCdUSe 1% also addec‘ almlus of p' mogniZed th enough to We exists. and ‘ 34 straightforward conceptualization, which is essentially an updated restatement of the Ravenstein model, represents an effort to derive conclusions regarding the volume of migration, the development of streams and counterstreams, and the characteristics of migrants. In the process, Lee identified four sets of pertinent factors: those associated.with the area of origin; those associated with the point of destination; intervening obstacles; and personal factors. Every act of migration involves an origin and a destination, and each of these points is characterized by a set of conditions that influences both the migration decision and the migration process. In the schema set forth by Lee, the simplest form of migration is a product of the interplay between the pushes or repulsions at the origin, and the pulls or attractions at the destination. Because the real situation is usually much more complex, however, Lee also added the two other sets of forces. Instead of a simple calculus of pushes and pulls deciding the act of migration, he recognized that the balance in favor of a move must be sufficient enough to overcome both the natural inertia which almost always exists, and the intervening obstacles that lie between the individual and the potential alternative sites. Personal factors are important because the set of advantages and disadvantages at both the origin and the destination, and the intervening obstacles, will be defined differently for every individual. Consequently, a person's attributes will influence how he perceives a Particular set of circumstances; and thus, they will also help to exPlain his behavioral reaction. The Rave descriptive. easily accommc nously, the a serious discr tion in the 1' will consider 53:38 migrate: 2) the dist: It the Proce: ligation th :iqration. DECGEQ readi ‘- We has that evfiry I need that t] 5e. ice, or a MESiremie 35 The Ravenstein and Lee models are general, and basically very descriptive. In most respects, the migration of retired people is easily accommodated within their framework. But as was hinted pre- viously, the agreement is not complete. In fact, there are several serious discrepancies. Three of these will come under closer examina- tion in the remaining portion of this chapter. Specifically, we will consider: 1) the push-pull mechanisms that are thought to pro- duce migratory responses, and in particular, the economic motives; 2) the distance factor; and 3) the phenomenon of stepwise migration. In the process, it should become indisputably Clear that general migration theories are inadequate in our attempts to understand aged migration. The need for additional empirical analyses will also become readily apparent. 533 5333 Migrations Economically Motivated? We have noted that migration research begins with the premise that every migratory movement "is either a response to some impelling need that the person believes he cannot satisfy in his present resi- dence, or a flight from a situation that for some reason has become undesirable, unpleasant, or intolerable" (Bogue, 1969, 753). Because each migration act has it own unique set of pushes and pulls, it would be foolish to attempt to catalogue the infinite combinations that might appear. But one set of considerations seems to constantly re-emerge in the migration literature. As Olsson has noted, most theories include the idea that individuals migrate in order to raise their level of income and that migrations are a means of achieving a state of spatial income equilibrium (Olsson, 1965, 23). The ind be seeking ma the migration glaces or reg “can increase is predicted In its Simple i"... there ~11 is lDi tion to the u Wat can be fl use MVe. I meat 0n ma .1”. hoth inCc anionic vie mtility 0f Althol mtivesr Q“ are much 1: E:- digraph er the sh . 36 The individual migrant or groups of migrants are assumed to be seeking maximwm economic advantage. Thus, from this perspective, the migration decision is essentially an economic choice between places or regions of work. Where, through migration, the migrants "can increase their earnings by more than the costs of migration, it is predicted that they will undertake to move" (McInnis, 1971, 197). In its simplest form, a model of this relationship would show that: M. . Eu 3 f‘EIC) i where gal-is the flow of migration from region i to region j in rela- i tion to the number of people in region i; E is the gain in earnings that can be obtained through migration; and C is the cost of making the move. In other words, people migrate because as sellers of labor, intent on maximizing their return, they perceive regional differences in both income levels and economic opportunities. Clearly, this economic view of migration is also a theory of the geographical mobility of labor. Although Ravenstein recognized the significance of economic motives, the first attempts to empirically verify this perspective came much later. During the 1930's Carter Goodrich edited a research monograph entitled Migration and Economic Qpportunity. Based upon the thesis that the flow of internal migration is highly responsive to economic opportunities, with large outflows from areas of economic hardship and submarginal economic income, this book initiated a whole series of efforts designed to explore the relationships between migra- tion and various economic indicators (Goodrich, 1936). In c were consi mdergoing from those Blanco (lS S‘I;;ort fc that "labc areas when A St CIOSely r« 0f their : (1962) pr. are imPar SPatial i 37 In one area of investigation, Shryock (1951) found that there were consistent flows toward those sections of the country that were undergoing economic expansion and metropolitan growth, and away from those sections where the economy was "mature" or declining. Blanco (1963), Lowry (1966), and Muth (1971) have compiled subsequent support for this contention; and recently, Pursell (1972) concluded that "labor seems to flow from areas with limited opportunities toward areas where superior economic opportunities exist" (Pursell, 1972, 261) . A second set of empirical investigations has considered the closely related topic of spatial earnings differentials. On the basis of their research in the United States, sjaastad (1960) and Raimon (1962) presented persuasive evidence that state income differentials are important factors in accounting for interstate migration. The spatial income equilibrium implied in this discovery has been con- firmed by D. Thomas and her associates (1957, 1960, 1964), and by Lianos (1970). Studies of economic migration have also linked fluctuations in the level of migration to business cycles, and changes in the level of economic activity (D. Thomas gt 31,, 1957, 1960, 1964; Bogue, 1969). And still others have shown that "not all individuals move from regions of low to regions of high average earnings. Indeed, for every flow of people in that direction there is a significant, but typically smaller flow in the opposite direction" (McInnis, 1971, 199). The list could continue, but in view of the current objectives, it would be both excessive and irrelevant. Our primary interest in the principles of economic migration is to evaluate their 38 applicability in terms of retired migration. Instead of an intensive analysis of each individual research finding, we need to ask if the basic tenets of the economic migration theories are of value in ex- plaining the migratory behavior of senior citizens. Do elderly people migrate in response to spatial differentials in both earnings and economic opportunity? Can we accept Porter's contention that migration only takes place in connection with changes of employment (Porter, 1956)? Actually, regardless of age considerations, any theory of migra- tion that is based solely upon economic elements is inadequate. As Bogue has noted, "at no point should it be inferred that economic factors and the spatial adjustments being made by the economy are the only forces at work to determine the direction, size, and composition of migration streams" (Bogue, 1969, 793). The so-called rational, economic man can exist only rarely in the real world. It might even be said that retired migrants are very nearly non-economic. Because they are retired, these individuals can ignore considerations in their search for a new home. Once they are financially dependent upon a pension and social security benefits, they no longer need to reside near a specific job site. And, since these methods of support are primarily spatially constant, the importance of earnings differentials is also neutralized. In composite, retirees are relatively free to emphasize other considerations. For many, this has meant a switch to non-economic factors such as amenities and services. In this respect, retirees have come to represent a new, and emerging form of migration that is gaining importance for people of all ages. Although it is nearly impossible to indications t spport of th theme-1930 based on laws JHnn‘m Sever a1 Year: in the Pacif; One of intereSt in 1954' he dec 39 impossible to set the precise origins of this movement, there are indications that it may have started as early as the 1920's. In support of this conjecture, M. Bright and D. Thomas recognized that the pre-l930 migrations to California far exceeded the expectations based on laws of migration. They surmised that an important part of the migration to California has been of a hedonistic rather than a primarily economic character and has been motivated more by climate and legend than by superior job opportunities (Bright and Thomas, 1941, 778). Several years later, in 1952, McKinley discovered a similar pattern in the Pacific Northwest. He observed that although, historically, migration within the United States was always associated with improved economic op- portunities, the permanency of war migration into this region was apparently strongly influenced by the psy- chological factor of taste for the region and its climate and other preference imponderables" (McKinley, 1952, 9). One of the first investigators to express more than a passing interest in amenities was a geographer, Edward Ullman. Writing in 1954, he declared that for the first time in the world's history, pleasant living conditions--amenities--instead of more narrowly defined economic advantages, are becoming the sparks that generate significant population increase . . . the new 'frontier' of America is a frontier of comfort (Ullman, 1954, 119). And there can be little doubt that for most people, the truly essen- tial component of this comfort is climate. In the opinion of Ullman, ”climate is probably the most important regional amenity" (Ullman, 1954, 123). Those with a choice consistently favor the pleasant climates. Although we might logically conclude that individual per- ceptions will produce a multitude of favorite locations, in reality, most people seem to agree that warm climates are the most desirable. This prefer the United received di As an exam; firizona hoI This fact that I Spread, am are fairly Hence, thi in their m 40 This preference has regional implications because very few areas of the United States can satisfy this requirement. Those that do, have received disproportionate numbers of amenity-oriented migrants. As an example, we have noted the appeal that Florida, California, and Arizona hold for retirees. This regional disparity, based on climate, is reinforced by the fact that climate can be easily combined with other, more evenly spread, amenities. Culture, education, sanitation, and other comforts are fairly consistently spread within the continental United States. Hence, this frees more people to concentrate on climatic discrepancies in their migration decision-making processes. The freedom to stress amenities is even stronger among retirees because their social security benefits are spatially constant. In other words, a retiree residing in Michigan can expect to receive the same basic financial support if he moves to a warmer climate in Florida.11 To his credit, Ullman also recognized the importance of other amenities. They "do exert a pull; mountains and beaches, hunting, fishing and other sports, beautiful New England towns, all come to mind" (Ullman, 1954, 123). In support of this viewpoint, we might examine the conclusions derived from Rikkinen and Alanen's study of 11Before we consider other significant amenities, we should note that climate-related migrations are not entirely free of economic considerations. Indeed, economic forces may differentiate between those who move to a more pleasant climate, and those who remain in their pre-retirement residence. Additionally, warmer climates may even furnish minor economic incentives. As Greenwood and Gormely have observed: "not only is sunshine itself an attractive force, but also the cost of living is lower (lower cost of fuel, housing, etc.) the more temperate the climate" (Greenwood and Gormely, 1971, 150). 41 intra-Minnesota migration among the elderly (Rikkinen and Alanen, 1970). Through the use of cohort analysis, these geographers were able to identify a series of age-related migration flows. The cohort immediately preceding retirement exhibited the same general patterns as the total population. But after retirement, this agreement began to weaken: Whereas metrOpolitan counties with the greatest suburban population continued to grow, several counties further to the northwest of the Twin Cities also experienced migration increases. This latter area (The Lake District or the Big Moraine Region) is noted for its multitude of lakes, rolling and wooded hills, and general vacation- oriented characteristics . . . it would appear . . . that certain conditions, such as the existence of high amenity features, must play a significant role in at- tracting recently retired persons to some counties (Rikkinen and Alanen, 1970, 14-16). Our contention that retirement produces a decline in the work-home bond, and that it may foster a subsequent increase in an amenity- home relationship, is supported. Additionally, this example also shows that for some people there are other amenities that rank even higher than climate. In fact, although the oldest cohorts migrated only rarely, it became clear that they tended to form a rather unique conglomeration around nursing homes and boarding-care facilities (Rikkinen and Alanen, 1970, 17). Finally, there have been several other references to the non- economic impulses associated with migration. Unfortunately, most have been either indirect or perfunctory. Meier (1968) and Gould (1967) have attributed the attraction of many migrants to cities in Florida, Texas, California, and Arizona to a search for amenities. Similarly, Brunn (1972) has identified comparable influences in the emerging patterns of "new urbanization" in the South and the Southwest; and aelbock (1 of United The sonic pri1 early ret increased AlthOLigh Put apps \— 42 Helbock (1968) found analogous factors in the emergence and growth of United States' "new towns." The point is clear. Migration theories based solely upon eco- nomic principles are inaccurate. A more standardized income structure, early retirements coupled with benefits, more leisure time, paid vaca- tions, a longer life span, and in general a more affluent society have increased the importance of other considerations (Brunn, 1972, 4). Although all population groups have been affected, the greatest imr pact appears to have occurred among the elderly. Aged migrants are increasingly free of economic restrictions. It behooves us, therefore, to seek a better understanding of their migratory activities. Over the long run, this should help us to revise several of the theoretical foundations in the migration literature. What is the Role of Distance in the Process g§_Aged Migration? Although the motives of migration have always been a very popular topic of analysis, concern for the impact of space (distance) as a component of all migration systems has also received considerable attention. The importance of distance has emerged with the realization that migration is more than a simple calculus of the comparative ad- vantages and disadvantages of the place of origin and the potential places of destination. Additional considerations are produced by inter- vening obstacles. And undeniably, the most salient obstacle is distance. If migration is to occur, the deterrent effect created by the inter- vening space must be surmounted. Hence, distance comes to represent the costs of migration. On the one hand, distance is a proxy for the actual transportation costs. On the other, it represents the psychic costs of are those hisfami] case, the Th: the natui Pas'enste: D. Thoma: 43 costs of moving. If the former are the monetary expenses, the latter are those which result from the reluctance of an individual to leave his family and friends and move to unfamiliar surroundings. In either case, the magnitude is related to the intervening distance. Through the years, there have been many attempts to identify the nature of these relationships. We have noted the efforts of Ravenstein (1885) and Lee (1966). Other contributions have come from D. Thomas (1938) and Olsson (1965). D. Thomas observed that long distance migrants may be expected to be differen- tiated from short distance migrants, and to be dif- ferentiated in varying degrees from the non-migrating population of the community of origin and the settled population of the community of destination (D. Thomas, 1938, 7). In response to his own excellent review of the distance factor in migration theory, Olsson proposed that "not only the number, but also the characteristics and perhaps also the motivations of migrants, very well could be a function of distance" (Olsson, 1965, 27). Although general statements of this nature possess obvious significance, perhaps the most extensive consideration of the distance component has come through a series of mathematical formulations. Classified as interaction models, these endeavors have gone beyond migration to encompass all forms of human interaction. Most of the early efforts were patterned after the principles of Newtonian physics. In general, the gravity concept postulates that "an attracting force of interaction between two areas of human activity is created by the population masses of the two areas, and a friction 12Gerald Carrothers has furnished a somewhat outdated, but very good historical review of these developments (Carrothers, 1956,94-102). against i 1956, 94) of intere size of 1 tance be Endel ha concentr Pro-duct Pcrtiona Thus, 25 44 against interaction is caused by the intervening space" (Carrothers, 1956, 94). More precisely, this viewpoint holds that the amount of interaction varies directly with some function of the population size of the two centers and inversely with some function of the dis- tance between them. Perhaps the most renowned statement of this model has come from Zipf: "the 'force' of interaction between two concentrations of population . . . is directly proportional to the product of the populations of the two centers and inversely pro- portional to the square of the distance between them" (Zipf, 1946). Thus, Zipf's hypothesis assumed the form: where: Po is the population at origin; P is the population at d destination; and D is the distance separating o and d. A second popular model of interaction is Stouffer's model of intervening opportunities (Stouffer, 1940). According to this hypo- thesis, the flow of migrants between two places is inversely related to the number of opportunities for the migrants to satisfy their needs that intervene between them. In other words, the number of per- sons going a given distance is directly proportional to the number of intervening opportunities. Obviously, in contrast to the gravity models, this paragon attributes a different role to the spatial com- ponent. Although based upon the premise that migration is costly and that the mobile person will cease moving as soon as he encounters an appropriate opportunity, Stouffer's hypothesis also implies that the friction of distance is not necessarily continuous. In support 45 of this belief, Rose has demonstrated that "intervening Opportunities during migration mean different things to upper, middle, and lower class people, to Negroes and whites, and to eastern urbanites and Western rural people" (Rose, 1958, 423). Hence, in this model, a constant distance factor is impossible. This apparent disagreement has introduced considerable controversy into the migration literature. Whereas most people agree that dis- tance performs an important function in explaining the spatial alloca- tion of migrants, a significant debate exists concerning the precise nature of this function. Geographers have questioned the value of simply utilizing intervening physical space. Instead, they have shown that the costs of traversing this space may be much more meaning- ful. In accordance with these beliefs, physical distance has been modified by both time and cost factors to produce concepts such as "effective" and "economic" distance (Isard and Freutel, 1954). Others have advocated the use of psychological or perceived distance. Con- comitantly, many have also come to realize that an extra unit added to a long movement may be of less importance than an extra unit added to a short movement (Carrothers, 1956, 97). The simple inverse re- lationship expressed in most interaction models has been tested empirically and various exponents have been substituted, ranging from one-half to over three. Although these diverse discoveries have uncovered an added complexity, one fact remains relatively unchanged: distance does introduce friction into the migration process. Obviously, as the intervening space increases, the propensity to migrate decreases at some undefz' essential, are the applicabii migration. I: bound retiree as one moves should become is hardly the 30m from the OVEICOming ti] “5 in partic 'I'Orking Years habitat. Cor In gen. rddted Mgr influences a 21" e , a. a dlSta W est (Arizon {if EShold he if each “Ea ObviOu P0: E traditi 46 at some undefined rate. Efforts to identify this rate, while undeniably essential, are beyond our current concern. Instead, we must question the applicability of even the simple distance-decay proposition in aged migration. If the principle is correct, the major origins of Florida- bound retirees should be from near and adjacent states. Conversely, as one moves to the north and to the west of Florida, the contributions should become proportionally smaller and smaller. In reality, this is hardly the case. In fact,the overwhelming majority of migrants come from the Northeast and the Midwest. Apparently, the costs of overcoming the intervening space are secondary to the pull of amenities, and in particular, to the pull of climate. After spending their working years in a colder climate, many retirees favor a warmer habitat. Consequently because of the small area of subtropical climate within the country, California and Florida largely escape come petition. Their amenity pulling power is reinforced by the relative uniqueness of their environment, which enables them to exert a pull even across half a con- tinent (Ullman, 1954, 130). The recent emergence of Arizona further substantiates this observation. In general, interaction models fail to accomodate amenity- related migration streams. But this is not to say that all spatial influences are inoperative. To the contrary, there are indications that a distance threshold exists between the South (Florida) and the west (Arizona and California). Although hardly impenetrable, this threshold helps to differentiate between the primary source regions of each area (Friedsam, 1951; Smith and Marshall, 1970, 8). Obviously, although the retired migration process resists the more traditional spatial explanations, we can not assume that it is aspatial. logical th fl'ienced h quate exis the spatie 47 aspatial. In addition to the apparent distance threshold, it seems logical that the decision to either move or stay may be heavily in- fluenced by the amount of intervening space. Undeniably, the inade- quate existing models need to be revised if they are to help determine the spatial role in amenity-related movements. Qg_the Principles 2; Stage 2£_Step-Wise Migration Apply i§_Retired Migration? As a final example of the shortcomings of migration theory, we can examine the literature pertaining to stage or step-wise migration. Succinctly stated, the basic concept in this theory is that rural inhabitants move individually toward large cities via gradual moves through smaller centers. Theoretically, as each migrant steps up the urban hierarchy, he is replaced by another migrant who is also ad- vancing through the system. As might be expected, the concept of step-wise migration has been implicit in many migration studies. In a recent review, Olsson (1965) identified several of the outstanding contributors. One of the most intriguing aspects of this work has been the discovery that the theory performs best when it is applied in a sit- uation with an obvious primate city. As an example, R. Thomas (1968: 1972) has uncovered very clear step-wise movements among the migrants in Guatemala. Unfortunately, the United States does not satisfy this desirable precondition. In fact, in addition to lacking a primate city, it is apparent that the American migration streams are much more complex than the original step-wise theory would seem to warrant. Hierarchical flows may exist in the rural-urban movements, but on national segment of Even that people Of stages, configurat. can lll-af Coasiderat Seeking re judging fr iPpkent t 48 a national scale, this form of mobility is but a relatively small segment of the total system. Even if we softened the original hypothesis to indicate simply that people tend to move to their ultimate destination via a series of stages, it seems logical that few retirees would fit within this configuration. At such an advanced stage in the life cycle, oldsters can ill-afferd a series of residential shifts. Both time and money considerations preclude this. Instead, we might expect the amenity- seeking retiree to move at once to his ultimate destination. And judging from the size of many of the retirement communities, it is apparent that hierarchical considerations are at most a minor factor. PROSPECTUS The primary objective in the preceding sections of this chapter was to pose questions. Concomitantly, we have noted many of the weaknesses in the existing theoretical structure of migration. Perhaps the most serious inadequacies occur in the literature relating to aged migration. In many respects, the traditional models are unable to produce satisfactory results when they are applied to this special form of mobility. The questions pertaining to aged migration are both relevant and important. They deserve our attention. In this spirit, the remainder of this dissertation will concentrate on an empirical analysis of retired migration to St. Petersburg, Florida. First,we will review the scant literature on aged mobility. Then, after es- tablishing both the study area and the research design, we will con- sider the product of a survey analysis. While analyzing this informa- tion, several objectives will predominate: 1) to characterize the :igrant er and the b: the spati into the 49 migrant and his selection process; 2) to analyze both the structural and the behavioral components of this process, and thereby identify the spatial attributes; and 3) where possible, to fit this information into the existing theoretical structure. CHAPTER III REVIEW OF THE LITERATURE ON AGED MIGRATION Florida was becoming a winter resort for northern city people. One could travel by sleeping car from Boston to Jacksonville or by steamship from New York. More than 40,000 persons had come thus the previous winter. Thousands were moving to Florida to settle, enjoy the subtropical climate, grow oranges, and get away from city pressures (F. Ratzel, 1876). Although we have cited many references in the course of our discussions thus far, most of the literature dealing specifically with aged migration has either been treated very lightly or not at all. The main purpose of this chapter is to identify the contributions of those relatively few items that fall into this category. In this review, the articles will be grouped according to their methods of data collection. Three main categories will be utilized: the indirect; the direct-—census and other governmental statistics; and the direct--survey analysis. This method of presentation will permit us to judge the strengths and weaknesses of each approach. Thus, in addition to a simple review of the literature on aged migra- tion, we will also develop an important, although admittedly super- ficial, acquaintance with several of the primary methods of migration analysis. 50 $331158 OF AGE In our p1 statistics in t1 were forced to initial remark applicable in either inadequ investigators or more indire 11‘35 been the 5 following logj 31qu in tha1 SPECified am, rat migration rat e f°r a co 51 STUDIES OF AGED MIGRATION BASED UPON INDIRECT METHODS OF MEASUREMENT In our previous discussion of the development of migration statisticsixithe United States Census, we noted that many early analyses were forced to employ indirect measures of migration. Although our initial remarks were deliberately general, this same principle is applicable in the specific case of aged migration. When faced with either inadequate census materials or no materials at all, many investigators accomplished age migration studies by resorting to one or more indirect indices. Perhaps the most popular indirect approach has been the survival ratio method.13 This procedure involves the following logic: given a population cohort at time A, the total change in that cohort by time B, which is equivalent to A plus a specified amount of elapsed time, is the result of two factors: net migration and mortality. Hence, if the mortality or survival rate for a cohort can be estimated, then it is possible to determine 13Bogue (1969) has identified and discussed three indirect measures of internal migration: the vital statistics method; the survival ratio method; and the place of birth method. The popularity of the survival ratio method is easily explained. First, the vital statistics indicator obtains a measure of net migration by subtracting reproductive population change from total change. Both births and deaths are considered. Obviously, this procedure is most accurate when the subject population includes new-born children. Since it is impossible for older cohorts to have reproductive additions, this technique is of questionable value in judging aged migration. The place of birth method is even more suspect. By comparing a person's birthplace with his current residence, we can account for a maximum of one move in each person's lifetime. All intervening moves are ignored. Unfortunately, as a person's age increases, we can expect this technique's accuracy to decline. And thus, it is certainly not a desirable method of estimating aged migration. (A more detailed discussion of the problems associated with this latter approach will come in a later portion of this chapter.) the estima formula: Estimated We can ill V6 know t1 211960 (( lid a sun five year Period by 4.900) an, 5% Thi 1965 EXce lite this The the nuthe the Vario 502-346“on Cansus sr not Vary Rmc a“'all'sis‘ n = 19.4). a I: 1951' dmade. has. . QSlZE 52 the estimated net migration for the cohort by utilizing the following formula: Estimated Net Migration = Cohort B - [ZCohort A)(Estimated Survival Raté] We can illustrate this procedure with a hypothetical example. If we know that a particular county had: 1) 5,000 residents aged 60-65 in 1960 (cohort A); 2) 4,950 residents aged 65-70 in 1965 (Cohort B); and a survival rate of 98 percent for the same cohort for the same five year period, then we can estimate the net migration for that period by multiplying Cohort A by the survival rate (5,000 x .98 8 4,900) and then subtracting the result from Cohort B (4,950 - 4,900 = 50). This latter figure shows that the actual number of cohorts in 1965 exceeded the expected amount by 50. And logically, we can attrib- ute this difference to net migration. The key element in the formula is the survival rate. Ideally, the number of deaths for a given cohort should be known for each of the various subsections in the study area. Unfortunately, this in- formation is rarely available. And as a result, in addition to using census survival rates, investigators must assume that this rate does not vary spatially when age-sex cohorts are being considered. Among the first individuals to employ this technique in the analysis of aged migration were T. Lynn Smith (1951), Homer Hitt (1954), and Irving webber (1956). Although Smith's work was published in 1951, it actually concerned national trends during the 1930-1940 decade. In particular, two aspects of elderly migration were emphasized: l) the volume of the net movement into or out of each state; and 2) the identification of the specific counties that were on the rec aged 65 ye ragritude In the pro and Easter Persons mi nized the It a When, afte during the the rOle 1 rates of ; Mformed trying to migrant “5 1950 Eisratio, Fired the 1940. 01 results . 53 on the receiving end of a significant amount of migration among persons aged 65 years and over. Smith hoped to measure both the absolute magnitude and the relative importance of elderly population movements. In the process he identified Southern California, peninsular Florida, and Eastern Texas, as "the three principal areas to which elderly persons migrate" (Smith, 1951, 27). To his credit, Smith also recog- nized the conspicuous absence of aged migration analyses. It may be that Homer Hitt (1954) was responding to this situation when, after noting the uneven growth of the aged in the 48 states during the preceding inter-censual periods, he vowed to determine the role that internal migration had in producing these differential rates of increase (Hitt, 1954, 195). To achieve this goal, Hitt performed two major operations. First, he continued Smith's work by trying to estimate the volume and relative importance of the net movement of elderly persons into and out of each state between 1940 and 1950 (Figure 2). Then, with the intent of ascertaining whether migration among the aged was increasing in importance, he also com- pared the movement between 1940 and 1950 with that between 1930 and 1940. Of course, this latter process produced strong affirmative results. Whereas Smith and Hitt employed the survival ratio method on a national scale, webber (1956) concentrated his efforts in Florida. The earlier studies had discovered spatial growth differentials among the aged population in the various sections of the state. Webber hoped to measure this phenomenon by analyzing elderly movements into the state and its 67 counties during the decade ending in 1950 (Figure 3). He concluded that certain counties "were favored by ZO.F(KQ:Z FNZ .20! OOG.I 0V9. .0N0< NI... 0203:“ he mwmmnuJ 02¢ 02:90 ..k & Uhk-nh 54 bowel St Ste. Onm_IO¢m. .omcd NI». @2032 29.53.92 hmz 20¢“. wwmeJ 024 «2.46 .N manor... ¢m>o 02¢ n0 owa< mh2(¢o.! no ¢w0232 ‘ “ME 3. : NUMBER AM i mews n no _‘ o . RI: El AI 1 55 FIGURE 3. NUMBER AND RELATIVE IMPORTANCE OF AGED MIGRANTS TO FLORIDA. I940-I950, BY COUNTY ...... RATIO OF ACTUAL TO EXPECTED POPULATION AGED 65 YEARS AND OVER g '00 g I40 I80 220 NUMBER AGED 65 YEARS AND OVER IN I950 WHO MIGRATED WRING THE PRECEEDING DECADE Mi. IOO Afler Webber {/9567 retired peo housing, an of the stat certain the to realize Mebber, 1' Altho was conduc e-e'loyed t 1Eigration and Alanen 05 this re mL‘fl’fition 99nera1 HL‘ tiSiOn to fact that PIO‘Jide d mach less Alanen, 1 Evidence tours a1 labor f0: 50th th‘ 56 retired people because of mild climate, accessibility to coasts, housing, and recreational facilities, and the institutionalization of the status of the retired person" (webber, 1956, 327). Webber was certain that areas possessing these characteristics would "continue to realize larger increases in the aged part of their population" (Webber, 1956, 327). Although these early analyses were surprisingly spatial, each was conducted by a sociologist. More recently, two geographers employed the same indirect approach in a study of the intrastate migration of persons over 50 years of age in Minnesota (Rikkinen and Alanen, 1970). In the preceding chapter we noted the hypothesis of this research: "that there is a threshold age after which the migration patterns of older aged cohorts begin to differ from the more general migration trends" (Rikkinen and Alanen, 1970, 5). The de- cision to utilize the survival ratio method was predicated by the fact that "official census publications of the United States do not provide direct total migration data for counties and smaller areas-- much less for individual components of the population" (Rikkinen and Alanen, 1970, 6). Data from the 1950-1960 census interim produced evidence that the hypothesized threshold was a reality, and that it occurs at a time when the majority of people retire from the active labor force. Among their concluding remarks, Rikkinen and Alanen noted that: while the migration of older people to and within California and other warm weather states is a commonly recognized occurrence, intrastate migration, even within a state outside one of the more 'popular' retire- ment areas of the United States, is no less complex, varied, or multifaceted (Rikkinen and Alanen, 1970, 17). 57 Although these studies of aged migration are obviously signifi- cant, several serious limitations emerge in their indirect methods of migration estimation. We have noted that the accuracy of these pro- cedures is threatened by a number of very risky assumptions. Also, the statistics that are generated are aggregate estimations, and thus, they furnish virtually no insight into the behavioral aspects of migration. A geographer's interest in the spatial manifestations of aged migration can not be fulfilled, since even the very basic information pertaining to actual origins and destinations is unavail- able. Further problems arise when the oldest cohorts are considered. In particular, since they are so greatly affected by mortality, it is impossible to use the survival rate to derive any very detailed conclusions about their mobility (Rikkinen and Alanen, 1970, 13). STUDIES OF AGED MIGRATION BASED UPON THE CENSUS AND OTHER GOVERNMENTAL STATISTICS In lieu of the preceding indirect data manipulations, many investigators have turned to governmental agencies for their migration statistics. Of course the most popular source of this information has been the united States Bureau of the Census. The publications emerging from the decennial census of population and housing are inadequate in many respects, but they do offer a convenient indication of migratory activities. Since 1940, respondents have been asked, in one form or another, to supply both their place of birth, and their place of residence at an earlier date.14 Recently, in an effort to 14Under the method of organization adopted in this chapter, the place of birth statistic is an enigma. Whereas Bogue (1969) identifies it as an indirect indication of migration, it is also 58 measure residential mobility, the Census Bureau added a question con- cerning the length of residence in the current housing unit.15 With such a variety of statistics available from a national survey, it is hardly surprising that studies of aged migration have also assumed a variety of forms. As an example, consider an early effort by Friedsam (1951). After recognizing that few if any attempts had been made to analyze the migration of the aged to the Pacific coast states and to Florida, Friedsam turned to census tabulations of the division of residence in 1940, by the division of residence in 1935, to analyze the gross interregional movements of persons aged 65 years or more. He discovered that the census materials were far from ideal. In fact, the only way to evaluate the movements into Florida was to analyze the movements into the eight states comprising the South Atlantic region. Even with these restrictions, however, Friedsam.was still able to uncover some crude, but interesting pat- terns. He was one of the first to hint at the existence of a thresh- old between the "trade areas" of the Pacific region and Florida: "the South Atlantic region draws only about 13 percent of its total in-migration of aged persons from the four regions west of the Mississippi, while the Pacific region draws approximately 40 percent primarily derived from Census publications. The decision to examine an article based upon place of birth information (Dyer, 1952) in this section is admittedly arbitrary. It could also have been placed in the preceding section. 15In 1970 the 15 percent sample was asked: "when did [Ehe head of the householé] move into this house (or apartment)?" The alterna- tive answers were: 1968 to 1970; 1965 to 1967; 1960 to 1964; 1950 to 1959; 1949 or earlier; and, always lived in this house or apartment. 59 of its aged in-migrants from the five regions east of the Mississippi" (Firedsam, 1951, 238) (Table 6). He also identified differences between the patterns of elderly migrants and the patterns of migrants of all ages. In particular, the tremendous attraction of Florida for the aged migrant emerged strikingly clear. Another of the early studies based upon census tabulations was Donald Dyer's indirect analysis of the origin of Florida's population (Dyer, 1952) . Unfortunately, Dyer, a geographer, made no attempt to restrict his analysis to any specific cohort. Consequently, in the context of aged migration, his effort is only mildly significant. If one aspect is worth noting, however, it is his technique of analysis. Simply stated, Dyer used place of birth statistics to infer popula- ti°n origins. This procedure can produce meaningful results only if we can assume that, on an interstate level, Floridians migrated no more than once. Thus, if a person is both born in Florida and resides in F1Orida, he is considered to be a non-migrant. All others must be migrants because their place of birth differs from their place of residerice. Under these circumstances, the process of determining the c>3:‘:i.gins of Florida's population is no more involved than the task of ~ 8:"u‘Ply counting the number of people born in each state. \ 1 fact 6The apparent difference in these figures was attributed to "the par that: there is a general trend of migration of all ages to the wQSt while migration into the South Atlantic is much more selec- tiv I. e (Friedsam, 1951, 238). 60 .. .IDIIlifil ,I .l.’ I. .mmu .m .Smmc o .NmoHBcouoo. mo Hmcuson =.moumum copes: on» as wood on» no cowusumaz HmcofimmquoucH= .Bvauwum .n .m "mumoom ¢.m H.H m.N m.NH I m.m b.mN o.mm b.0H maufiom w.m H.H m.N H.NH I m.m b.5N H.mm 0.0H 0H6: H.m H.H m.N m.mH I m.m 0.0N H.mm c.0H H0908 sowmwm owucmaum nusom I m.mH ¢.m m.H m.m m.Hm m.NN m.oa v.m oasfiom I m.mH v.HH m.H N.m a.mm h.ma m.m m.N was: I m.hH m.0H h.H m.m v.Nm v.HN H.0H m.N Hmuoa GOamom oamwoum ofiuaoum camucsoz Hmuucmo Hmuucmu naucmHum Hmnucoo Hmuusmu owucmHuc womamcm xom can cowumcwumoo susom susom susom suuoz cuuoz oats“: 3oz ummz uwmm puss umum mousom on mcavuooom mucsumwz mo cofiusnauumwa ommucoouom . 5m 3 Ba 23 5 moamgmmm no game 3 8.732 .233“ condos 2. crass: 53m 8 :25 can as.» me muses was to ass 6 33 61 Unfortunately, the initial assumption is only rarely accurate. As an example, consider the common case of a person who is born in one state (e.g., Michigan), only to move to another state (e.g., Ohio), before ultimately moving to Florida. According to Dyer, this person's origin would be recorded as Michigan and not Ohio. The repercussions in a migration analysis are apparent. And of course, as the number of intervening residences increases, the degree of inaccuracy also increases. These same problems exist when a native-born Floridian returns to the state after residing elsewhere. Is he really a non- migrant? Realities dictate that the place of birth statistic can Offer little more than a very crude indication of migration origins. We must re-emphasize the fact that Dyer's objective was to commant upon the origin of Florida's total population. His analysis did not stress elderly cohorts, and therefore, we should not be overly d1 8a13°P<31Lnted by the incidental contributions that he made to the literature on aged migration. In fact, in this respect, Dyer's work is representative of most geographic undertakings because, with few exceptions, the geographer's attention has only rarely centered upon e“I'deil‘l‘ly migrations.17 Fortunately, this has not always prevented him fr°m making some rather significant discoveries. As an example: ““sider the work of Edward Ullman (1954) . His primary concern was \ 17To the contrary, geographers have virtually ignored all aspects prie e aged. Among the few direct analyses in existence are G. E. S 8'ti's examination of urban housing for the elderly (1970); and Golant' 3 analysis of Th__e_ Residential Location a_n__d Satial_ Be- W~ of the Elderly (1972). Indirect references, although signifi- of more numerous, are still very restricted. A large proportion ”thine!!! have emerged from studies relating to the population age ecolotures of urban areas. In particular, nearly all factorial gies have yielded an age factor, and with few exceptions, this 62 not elderly migration. And yet, we have noted the importance of his arguments concerning the role of amenities in regional migration. His observations are an essential component in the complete under- standing of elderly population movements. Subsequent geographic efforts by Kariel (1963), Helbock (1968), and Brunn (1972) have added support to Ullman's work. Two other geographic studies deserve our attention. First, there is P. Simkins' discussion of the "Distribution of the Aged in Pennsylvania" (1964) . In his attempt to explain regional variations in the growth of the aged, Simkins placed a heavy reliance upon age- selectivity in total migration. Unfortunately, the migration of the elderly was virtually ignored. Instead, Simkins characterized the loca3|. differences in the distribution of the elderly as a residual °°mPOnent of the migration trends of the younger, more numerous age groups. In this manner, he reasoned that the proportion of the aged in a given population will increase where the younger adults have Out~migrated in large numbers. And conversely, where there have been large numbers of young in-migrants, the proportion of the elderly residents among the total population will decrease (Simkins, 1964, 183) . In a subsequent consideration of the aged population in Pennsyl- Vania. Schnell (1970) argued for a more direct investigation of the Spatial trends among the elderly. From this perspective, he discovered \ di suggzicm has shown the urban core to be demographically old, and the appro 3 to be proportionally young. Even a completely different theseaeh by Coulson (1968) produced nearly the same pattern. Because aged various efforts are, at best, incidental to our understanding of They migration, they are beyond the scope of our current considerations. 111 not be included within our discussions. 63 that "almost 83 percent of the regional variation in the growth of the aged can be explained by the size and trends of the 55-65 cohort" (Schnell, 1970, 228). To his credit, Schnell also recognized the importance of migration among the older inhabitants. In fact, he identified the next important area of research as "the determination of the demographic and non-demographic factors which influence the migrational behavior of the aged" (Schnell, 1970, 230) . In his opinion, this should include a rather broadly based study of the social and economic characteristics of the elderly, as well as a careful examination of the nature of the communities to which aged migrants are attracted or re- pelled, and those in which they remain after retire- ment (Schnell, 1970, 230). One of the few attempts to actually assess the importance of amallities in elderly migration has come from the work of two economists, 5- Barsby and D. Cox (1970) . In a somewhat unique approach, they 1911(”had frequently cited attractions such as climate, water resources, and Clean air, and concentrated upon the more neglected appeal of p “bl-1C goods. Their stimulus came from C. Tiebout's arguments that: 1) geographic mobility could serve as a substitute for a market of public goods; and 2) that families are mobile and responsive to variations in local public sectors (Tiebout, 1956) . The decision to employ statistics concerning the elderly was based upon two considera- tions . First, it is plausible that the public sector preferences of elderly persons are more homogeneous . . . 3 an second, older persons have less attachment to the private sector of the economy, and so can give more weight to the public sector in making locational choices (Barsby and Cox, 1970, 3). 64 In a multiple correlation-regression analysis of Arizona retirees, nine public sector variables were utilized. These included indices concerning eligibility requirements and size of old age assistance, state income tax exemptions and credits, property tax, and expendi- tures on education, health, hospitals, and public welfare. In some- what of a surprise, Barsby and Cox were forced to conclude that the data on the migration of elderly persons offer scant support for the alleged relationship between aged mobility and the public sector (Barsby and Cox, 1970, 10). Up to this point our discussion has centered predominately upon interstate migrations. The empirical investigation that follows this chapter will continue this emphasis. It would be incorrect to assume from this, however, that elderly people do not also move at a much smaller scale. In fact, intraurban shifts probably far out- number all other forms of aged mobility. Appropriately, C. Gold— scheider, a sociologist, has considered the redistribution of the older population in the Los Angeles metropolitan area between 1940 and 1960 (Goldscheider, 1966b). In addition to examining the older citizens' participation in the suburbanization process, Goldscheider also hoped to comment upon changes in the relative concentration of the older and younger populations. Concomitantly, he divided the metropolitan region into both settlement areas and distance zones. He found that although all age segments were moving away from areas of older settlement and areas nearer the metropolitan center, the older population was doing so at a slower rate (Goldscheider, 1966b, 83). In the course of three census periods this relative stability among the elderly produced a discernible change in the age structure 65 of the metropolis. Between 1940 and 1960 the differences between age groupings accentuated, and in particular, the older areas, and the areas closer to the center became increasingly populated by senior citizens. Subsequent factorial ecologies have shown that this trend is a common feature in most of the metropolitan areas of the United States. At this juncture we might ask if the Census and other govern- I mental statistics are suitable for spatial analyses of aged migration. I Obviously, they possess some merit. we can employ them to establish gross migration streams. we can learn the magnitude of these movements, and we can even identify several of the socio-economic correlates that characterize them. But is this enough? Can we comment upon the specific reasons why individual retirees migrate? Can we say anything about the process of aged migration? Unfortunately, we can not. The governmental data suffer from the same deficiencies that characterize all aggregate statistics. This is a particularly crucial weakness when the subject is migration; because migration is an individual (or family) act. The characteristics of both the origin and the destination are defined differently for every migrant. And of course, different people are affected in different ways by the same set of obstacles. In the opinion of Everett Lee, "the decision to migrate . . . is never completely rational, and for some persons the rational component is much less than the irrational component. We must expect, therefore, to find many exceptions to our generalizations" (Lee, 1966, 292). 66 Aggregate statistics tend to conceal too many of the features that are essential to a complete understanding of the retired migration process. In the specific case of the Census, the aggregate difficulties are compounded by other considerations. We have identified many of the problems that are associated with the place of birth statistic. Unfortunately, the other set of migration statistics, those based upon the place of residence five years previously, are also inadequate. As an example, this latter procedure can supply no special information concerning those people who move more than once in the designated five year period. Nor can it identify those who migrate but return to their place of origin. Investigations based upon the census are also restricted to government-established time periods; thus, measure- ments are available only once in each decade. Among geographers, it is acknowledged that the census publica- tions are also hindered by the size of their reporting units. The 1970 Census, for instance, has the capacity to show in-migrants and out- migrants over the five year period 1965-1970, for any type of areas defined in terms of major cities or whole counties (Ferris and Long, 1971, 121). But specific county exchanges are impossible to determine, and it is not proper to assume that all parts of a county or city lose or gain migrants at the same rate. In other words, large units may conceal very significant spatial patterns. 18In its Current Population Surveys, the Bureau of the Census does conduct an annual measure of migration with a one year interval. Although potentially valuable, this survey is actually inferior to the regular census. In particular, the yearly procedures supply very little information. Their primary function is to furnish an indication of large-scale mobility trends. As a source of materials for migration analyses, they have little value. 67 Given these problems using aggregate data, we can appreciate the role of a more behavioral approach to the study of aged migration. By soliciting individual responses to survey questions, it is possible to overcome many of the difficulties. The investigator can accumulate information that is more specific to his purposes. He can establish his own intervals, and he can also question the retired migrants about their individual motives and adjustment procedures. Since the specific origins and destinations can be identified, the actual spatial implica- tions can also be noted. In the next subsection, we will focus upon several of the more prominent attempts to identify and assess the behavioral elements that are salient in aged migration. STUDIES OF AGED MIGRATION BASED UPON SURVEY ANALYSIS Those who have employed survey methods to collect data concerning aged migration have done so for a variety of reasons. As early as 1952, Blanchard used a mail questionnaire to compile a "comprehensive guide" of 111121532113 Hg! 32 Retire (Blanchard, 1952) . In essence, this effort amounted to little more than an unsophisticated discussion of 75 potential retirement places. The schedules, which were sent to elderly alumni from 13 eastern colleges and universities, were used as the primary source of information about the subject communities. Although Blanchard's objectives precluded many of the anticipated benefits of survey analysis, we can rejoice from the fact that studies of this nature have been relatively rare. The vast majority of survey work has focused upon the behavioral components of aged migration. For instance, relatively voluminous contributions have been based upon the realization that "migration necessitates personal readjustments \' d- «.L--’ 68 on the part of the older people who move, and social readjustments on the part of the communities in which they take up residence" (webber, 1956, 323). As might be expected, sociologists have tended to dominate this area of inquiry. In his dissertation, one of the early leaders, Walter McKain (1947), examined the social participation of old people in a California retirement community. He focused upon memberships in organizations and attendance at both secular meetings and religious services to corroborate the common sense observation that social participation declines with advancing age. Several years later, webber (1954) reached a similar conclusion after performing a nearly identical analysis in two Florida communities, west Palm Beach and Orlando. Prior to the latter study, Webber had also conducted a survey for the Florida State Improvement Commission (Webber, 1951). This earlier effort was in response to the agency's desire to identify the outstanding social implications of aged migration. The survey was conducted among the retired white population of St. Petersburg. And in several respects, it was a pace-setting investigation. For one of the first times, an effort was made to identify the impact of retired migration on a community and its facilities. Also, webber's work seemed to stimulate several additional analyses on related topics. Three years later, Hoyt (1954) conducted a survey among the retired residents of a mobile home park in Bradenton, Florida. In addition to the characteristics of the park and its inhabitants, he also focused upon the residents' social participation, and their attitudes toward mobile home living. During the same year, Harlan (1954) conducted another survey in St. Petersburg. Following webber's n... * I "p. an 69 initiative, he recognized that the intrusion of a "new" population group may institute both individual and collective adaptations. The functional relationships among the individuals and institutions of a community may be altered. And as a community adapts, both ecological and social changes occur (Harlan, 1954, 332). Of course, St. Petersburg, where the proportion of the aged was increasing through migration, represented an excellent opportunity to define the nature and extent of these collective responses. Harlan found that "the presence of these persons is clearly reflected in demographic data, in the occupations of the younger population, and in the types of local business and professional activities" (Harlan, 1954, 332).19 Several years later, webber (1961) used the same general area (Pinellas County) to analyze the relationship between the health needs of the aged, and the health resources of the community. Whereas these latter efforts have tended to emphasize several of the collective responses to aged migration, it should also be noted that a concomitant development has focused upon personal adjustments. The aforementioned studies of social participation can be used to illustrate this trend. Additionally, in 1952, Granick undertook a survey of the senior population in two Florida communities, St. Cloud and Winter Park. 19As an example, Harlan discovered that "medical, dental, and related personnel are conspicuous. The community supports 14 Chiropo- dists, 19 chiropractors, 15 optometrists, 33 osteopaths, 28 practi- tioners, 61 dentists, and 147 physicians and surgeons. There are also proportionally numerous nurses, technicians, laboratory assistants, and dieticians. Among the physicians and surgeons, approximately one- sixth indicate that their practices are primarily devoted to the treatment of conditions associated with aging, including diseases of the heart, skin, and digestive tract" (Harlan, 1954, 339). 2 '<._.\-. I.» 70 avowed purpose was "to describe various aspects of [ihe retireeéfl eral adjustment" (Granick, 1952, 419). More recently, the Department of Rural Sociology at the University Wisconsin (Madison) has published a series of studies under the eral topic of "Retirement and Migration in the North Central States." hough each of these studies has assumed a slightly different per- ctive, the overriding objective has been to "provide needed and ningful information concerning problems of adjustment among re- ed persons" (Eteng and Marshall, 1970, 17). The act of retirement consistently viewed as a process which produces changes in the e situation of individuals. The articles that are relevant to current interests also consider the additional impact of post- irement migration. In either case, survey methods are employed gauge the social-psychological responses of the affected population. is assumed that personal adjustment levels will be reflected the levels of retirement and life satisfaction. The first analysis to incorporate the responses of retired rants appeared in 1968 (Honnen, Eteng, and Marshall, 1968). Based a data drawn from three communities in Wisconsin and two communities Plorida, this study amounted to a comparative socioeconomic analysis. mg the topics considered were: 1) the various background charac- Lstics of the retirees (e.g., age, marital status, health situation, incial status, and education) ; 2) some of the factors which in- enced the survey respondents in their decision-making process iing to retirement; and 3) why some retirees decided to move to ride while others chose to stay in Wisconsin. t...— 71 Subsequent reports also have utilized this comparative approach. example, in 1970, Eteng and Marshall performed a series of uare tests on data from communities in Wisconsin, Florida, and a. As might be expected, comparisons were made across the three 3, and between the Wisconsin non-migrants and the Florida and (a migrants. The sociological analysis centered around five major .95: 1) background characteristics; 2) retirement decision- I, including anticipatory socialization; 3) retirement transition; Incomitant socialization: and 5) the level of satisfaction with :spective retirement communities (Eteng and Marshall, 1970). .atter item reappeared in an ensuing study which focused upon the .ty retirement centers in Florida and Arizona (Smith and Marshall, In addition to analyzing and comparing the responses of the ant populations, Smith and Marshall also briefly considered Lstorical background and functions of the privately planned ement community. These Wisconsin studies are typical of the many efforts that been made to evaluate personal and community adaptions to ement and migration. The emphasis is upon concepts such as ipatory and concomitant socialization, financial and health cations, life satisfaction and morale. Although the significance is work is self-evident, there are many other aspects of aged tion that also need to be examined. From a spatial perspective, ‘virtual disregard for the place of origin and the actual process gration is regretable. The Wisconsin surveys were designed so only people who had resided in the North Central region of the ad States prior to retirement were included in the samples. [‘3 72 [sequently, no information is furnished concerning migrants from er areas of the country. And even the precise origins of the quali- irespondents were never really analyzed. Items such as reasons moving, and reasons for selecting a particular community, have ' rarely been considered by those with a primary interest in the stment characteristics of retired migrants. Fortunately, those with different interests have also conducted 2y research. Still within sociology, we can note the efforts of :55, Hoyt, and Manley. In 1954, Manley utilized a rather e data source to compare various characteristics of migrants can-migrants among a group of retirees (Manley, 1954) . The nation came from the personnel records of a large metropolitan ment store. In his efforts to identify the attributes of those e more likely to migrate, Manley recognized the significance factors: economics and mobility. Then, in the following year . he joined forces with Burgess and Hoyt to produce a study of Instruction of Scales for the Measurement of Migration After went." In essence, this study was an attempt to discover the for migration. Over 180 retired residents of a mobile home :y in Florida were interviewed according to seven criteria of .on: climate, health, economics, nobility, activities, relatives, nds. Responses were arranged on a five-point scale for the resence of a criterion as a motivational force, and also on sity scale for the weight of the criterion as a motivating The research findings, which are depicted in Tables 7 and 8, that climate and economic factors were the most important ‘- ‘u. u~-~TF 73 .e 7. Relative Importance of Different Factors in Motivating Migration to Florida Among 193 Retired Men Importance of the Factor Factor First Second Third None Total Score Its 96 73 ll 13 445 h 50 12 10 121 184 mic (Living Costs) 12 16 20 145 88 Lty (Travel) l7 l4 5 155 84 .ties 4 22 27 140 83 s 4 15 36 90* 78 ves 5 4 23 148“ 46 Iy (Farm or Rest) 5 7 4 177 33 .des 48 cases where the person had friends in Florida. des 13 cases where the person had relatives in Florida. E. W. Burgess, G. C. Hoyt, and C. Manley, " The Construction of Scales for the Measurement of Migration after Retirement," Sociometrx, 1955, p. 621. ‘5 ‘1 ‘ M-“_‘. I no 74 Table 8. Three Indications of the Rank Order of Motivations Among 193 Male Florida Retirees Rank Order of Motivations Reported Motivations Presence Intensity Importance Climate l l 1 Economics 2 2 3 Mobility 3 3 4 Friends 4 6 6 Activities 5 4 5 Health 6 5 2 Relatives 7 7 7 SOURCE: E. W. Burgess, G. C. Hoyt, and C. Manley, "The Construction of Scales for the Measure- ment of Migration After Retirement," Sociometry, 1955, p. 622. 7S Iences in migration.20 Friends, health, and relatives rated lowest Iess, Hoyt, and Manley, 1955) . Up to this point, nearly all of the studies have centered upon ees in either Florida or Arizona. It would be improper to assume this, however, that the California retiree has been overlooked. ct, at least two notable surveys can be identified. In the first, n Goldscheider (1966a) reported on a portion of a longitudinal y of residential mobility in the Los Angeles SMSA. He was ing to the lack of knowledge concerning "the differential pro- :y to move among various social and economic categories within Lder population" (Goldscheider, 1966a, 103). And in the process, so considered the degree to which sub-groupings within the older .tion realize their nobility plans and desires. Actually, 'heider did not limit his analysis to the traditional 65 and over oup. Instead, he lowered the age limit to 50. Also, most of bility that is discussed is intrametrOpolitan. This obviously the applicability of the conclusions, especially from the point I»: of the interstate migration of retirees, but it does not I: Goldscheider from making some very interesting observations. Is an example, in addition to reaffirming the widely known Itive stability of the older population, he also discovered Ie people aged 50 years and over are less likely to plan or 0Although the significance of economic forces may appear to it our earlier discussions of the non-economic nature of aged on, we must remember that the specific economic consideration migration studies is employment. In this instance, however, nomic reference is the lower costs of living in Florida which, se, is closely related to the Florida climate. There is really radiction. 76 to move, and are less successful than the non-older population icipating their mobility behavior (Table 9). 0n the basis of 10, he concluded that within the older population, the lower conomdc status groupings, the non-married, one-person house- renters, Protestants, and those who were more mobile in the decade, were both more mobility-prone and more successful in pating their mobility behavior (Goldscheider, 1966a, 107). The second noteworthy California study was published as a hesis for the Department of City and Regional Planning at the sity of California (Barker, 1966). The general topic was the ment housing market, but more specifically, Barker concentrated alifornia retirement communities. Given the various aspects 3 special form of housing, he elected to emphasize both the ‘e of these communities, and the market which they serve" (Barker, ix). As part of this latter pursuit, Barker found that the market for retirement communities is metropolitan, not state (ional. The elderly are prepared to exchange obsolete housing, but not their vested interest in the metropolitan area. Roughly 90 percent of the dwellers in successful retire- ment communities came from the metropolitan area in which the retirement community is located (Barker, 1966, 81). the empirical support for these statements is restricted to two Lities in the San Francisco Bay area, it is unfortunate that r speaks in such a general manner. It may be true that a person .ng in that region may be reluctant to leave, but can we make Ime generalization for new retirees in Michigan and New York? 1 another vein, can we use Barker's observations to say anything non-metropolitan retirees? 77 .voH .a .lommao flu .amoaoudouoo.mm Hdduson :.coHuoHomom nooao on» no auwaaooz Hoaucooemom Howucouomuaoe .uooeonomoaoo .u ”mumoom o.e~ H.vH «.mm H.H~ mva uo>o one mo n.5m 0.0H «.mm m.vm emu voIom ”mom H.0N «.ma o.Hm ¢.m~ mom Ano>o one omc oceusasmom nooao m.ae m.e~ e.Hm ~.vm Hem doeudaddod nooHOIcoz m.>m m.m~ m.me H.om mum cOwuoHomom deuce noe>moom NomHIHomH mmmHIHomH nomHIHomH momeo .oz huwafiooz .uuo>oz .o>oz 0» .m>oz 0» oceuemeoeuc¢ unwound masseuse anaconda maaommmooosw unmouom usoouom ucoouom sOquOmOum one was an .HOabonom hufidanoz mcwumm .momHIHomH .uow>mcwm can .mouemoo “cause haasmmmooosm smcmnfih xu4442354 454Jae”'\-’II I I II 78 o.mH v.m m.m~ m.ma HHH ocoz m.H~ H.HH H.vm 0.5a mos moo m.vn h.v~ v.mn m.om ova one: no N m.mv o.Hn m.~v m.mu no one: no n oomHIommH .mo>oz mo umosoz H.mm m.om m.Hv m.mm oma noncom o.oa m.m m.mm m.mH new sense undo meeeflmza m.mH m.m ~.em m.- HHH who: no mouse m.om v.vH m.mm m.om mma common 039 o.am a.~m o.¢m m.h~ voa caused «so scauwmomsoo oaosomaom m.m~ n.mH v.mm m.m~ mma ooauuschoz >.vm «.ma m.on m.am mwm oceans: usueum Havens: m.- o.oa m.m~ m.mn on season m.H~ m.ca v.am H.5N on oaaosuco v.n~ H.0a n.Hn m.mH Ham acounououm addendum a.m m.m h.av «.mm em ouwcsIcoz m.>~ h.ma o.om o.m~ mvm ovens doom uoa>osom momHIHoma momHIHomH mmmanamma momma .oz muaoewue> mundane: odeudmeoeudd .uuo>oz .m>oz on .o>oz on aaasmmmouosm ucoouom unwound mcwuwmoo anaconda ucoouom useouom 79 .voa .m .loomao HN .muoaoudouoo .MN Accusoo :.co«usaomom HmoHO o3» MO xuwflwnoz Hmwucmoammm HmwucmHOMMflo: .uoofiocomoaoo.o "momDOm o.mm m.m~ v.om n.m~ we unaaoo doom zoo m.ma m.HH H.on m.- mm ueHHoo deem amen o.o~ o.eH m.en o.m~ mo udHHoo opens sou m.m~ m.oa o.m~ o.o~ sod uoHHoo does: some sofiuomsooo o.m~ «.me o.~m e.o~ mmo ooo.mm due» ammo o.- «.aa m.m~ m.ma «we ooo.mm seen and: Assesses odoodH n.5m H.ma m.nn e.o~ mos nosoo on» menus cum m.m~ 5.6H o.m~ o.- mod dunno sumaneum ~.H~ m.m m.~m m.ma mm nm>o ecu «voodoo con—8:3 noe>unmm Nomaaaome Nomauaoma I done . assessor odeunmeoeucd .muo>oz .m>oz on «MMMQMon 0 oz mmanueun> %HHSMNW000§W UCUUHUA unwouflm OCflHHmmfl OGHGSQHQ unwouom ocuauuu 80 If we can focus upon a single impression that emerges from this law of the literature on aged migration, it would be the serious lciency of spatial contributions. Quite obviously, the geographic Its have been disappointingly infrequent. In the survey approach Iged migration research, they have been completely non-existent. I we recognize that the entire process of aged migration is spatial, that as such, it produces spatial implications, we can appreciate need for remedial action. Very simply, it should be our task nalyze these e1ements,and consequently, to contribute to the total rstanding of the aged migration phenomenon. In this spirit, the owing chapters will focus upon a survey analysis of retired ants to St. Petersburg, Florida. Once the methods and background rmation have been established, the attention will shift to the ey results. Through this process, we can expect the necessary adent to emerge. CHAPTER IV THE ST. PETERSBURG STUDY AREA Some people judge Florida by Miami and will have none of it; others think of Florida in terms of old men waiting to die on St. Petersburg beaches . . . (Blanchard, 1952, 166). If you don't enjoy the old people who sit on green benches on the sidewalks, play shuffleboard, and keep happy on a very small income, you had better go somewhere else. But if you feel congenial with those old people, if their life, their spirit and their friendliness strike a responsive note within you, St. Petersburg may be your place (Blan- chard, 1952, 182). Some people call St. Petersburg 'The City of the Three C's': the canes, the crutches, and the cripples (A Resident). There are several prerequisites to the proper analysis of survey information. As an example, it is essential for us to know why and how the study area has been selected and defined; and if only a portion 0f the subject population was included in the survey, we must also examine the sampling procedures. These considerations will help to datermine the appropriate methods of analysis. They will also aid in establishing the degree of generality that can be attributed to the Survey results. In other words, unless we examine the methods of field inVEStigation, we can not be sure if our conclusions are restricted to only the survey respondents. It may be that they are applicable to t he entire pOpulation. And quite Obviously, this is a crucial considera- tion. 81 82 To this end, the next two chapters will consist of a discussion of the field procedures. First, various aspects of the study area will be identified. Then, in the next chapter, the characteristics of the survey design will be described and explained. THE RATIONALE FOR THE SELECTION OF ST. PETERSBURG In Chapter I, we noted that if elderly movements are classified according to scale, they closely resemble the patterns established by the total population. Short moves predominate. In fact, a large majority (69 percent) of all the elderly people in the United States who changed their residence between 1969 and 1970 remained in the same county. And only 12 percent, or approximately 200,000, moved to a different county in a different state. This latter figure is very deceptive because it belies the true importance of the aged inter- state migrant. The residential shifts at this level have been far from random, and consequently, the spatial implications are particularly significant. For many years, California and Florida were the most favored destinations. More recently, Arizona has also become highly attractive. This relative concentration lends broad support to the conten- tion that amenities are a primary concern of retired migrants. It also holds special significance for the person interested in conducting survey research among aged interstate migrants. In particular, while searching for an appropriate study area, this knowledge should be most influential. Immediately, a large majority of the states can be eliminated. Those that remain can be examined in greater detail, and in the process, a precise site can be designated. 83 If we consider the three leading states on a county level, an interesting situation develops. Whereas California and Arizona fail to display the anticipated concentration of senior citizens, Florida's patterns are very impressive (Figure 1). On the basis of the percent of the total population aged 65 years and over, the central Gulf Coast possesses the largest relative concentration of elderly residents in the entire nation. Of the 67 counties in Florida, 19 (28 percent) have at least 15 percent of their residents aged 65 years and over. Six of the nation's seven leading counties are located along the Gulf Coast, and in each instance, at least one-fourth of their population has reached the retirement age. In comparison, California has only five counties (9 percent) with over 15 percent retired; and Arizona has only one county (7 percent) at this level. Over the long run, the growth of Florida's elderly population has been truly outstanding. Between 1900 and 1940, the number of senior citizens increased from 14,000 to 131,000. The magnitude of this in- crease is striking, but it has been dwarfed by the post-werld War II developments. During the 30 year period ending in 1970, the elderly group soared to nearly a million people (Figure 4). Of course the entire state population has also mushroomed, but significantly, the under 65 age group has been unable to keep pace with the expansion of the aged.21 In support of this observation, Figure 5 indicates that 21In the last decade alone, nearly every area of the state ex- perienced a significant change in its age composition. Due in part to changing birth rates and in part to highly age-selective migration, the number of children under five declined, and all other age groups increased. The largest growth (+79 percent) was registered by the 65 and over age-cohorts. Number of residents aged 65 and‘over (in thousands) 84 _____ _fl ._‘__ _--- FIGURE 4 mac I— V T 7 ABSOLUTE INCREASE OF 989.366 I PERSONS AGED 65 YEARS AND OVER IN FLORIDA. l900-I970 553,474 :1221‘ ' ' 550 if“ ‘ ' r ....... ..... ....... ..... ...... ----- ........ ....... ........ ....... ..... ...... ...... ........... ............ ........... ............. ........... ............. 250 237914 .......... .......... ZOO IOO IOO .;.;.; ;_ Source: “in... U.S. Census of PapquII'on I960 I970 Isoo IsIo 85 FIGURE 5. PERCENTAGE OF PERSONS AGED 65 YEARS AND OVER FLORIDA, ISOO-ISTO ..... Percent aged 65 and over I970 FIGURE 6. MEDIAN AGE OF FLORIDA, l900-I970 "’HMTW” T ”I” ”"7 Age 25 20 I900 I9IO Saurcs: U.S Census of POpuIaII'on the pt I stea 1'11 I97c conconi t 12.3 years When natal si mmmit WI) nc of th 0f sma is unk 05 sale is a di and the tiVe C01] promisir‘ K... 86 the percent of the total population aged 65 years and over has shown a steady rise from a level of 2.5 percent in 1900 to the point where, in 1970, Florida led the nation with a figure of 14.6 percent.22 A concomitant increase in the median age has led to a national high of 32.3 years in 1970 (Figure 6). When the statistics for individual communities are examined, the spatial significance of these trends is clearly established. Certain communities have enjoyed immense popularity among the elderly. By 1970 no fewer than 37 urban places in Florida had at least 30 percent of their population aged 65 years and over (Table 11). The number of smaller settlements with a similar proportion of elderly residents is unknown. Information of this nature is particularly valuable in the process of selecting a study area. If we can assume that, in general, there is a direct relationship between the proportion of elderly residents and the proportion of aged migrants, then the knowledge of where rela- tive concentrations exist can be helpful in pinpointing the more promising locales. Of course, the final selection must be based upon a number of other criteria. In all cases, the research objectives must be kept in perspective. A quick inspection of Table 11 will show that the communities on the list vary both in governmental status and in size. Of the 37 22The national average was 9.9 percent, and surprisingly, neither California (9.0 percent) nor Arizona (9.1 percent) could match this figure. The large scale migration of senior citizens into these states has been counter-balanced by a combination of relatively high rates of natural increase and substantial additions of non-elderly migrants. Table 87 Table 11. Urban Communities (2,500+ Population) in Florida Ranked Via the Percent Aged 65 Years and Older--1970* Percent Aged County or 65 Years Name Neighboring City Population and Older 1. DeBary (U) Volusia 3,154 58.6 Sanford 2. Buena Vista (U) Dade 3,407 52.2 Miami . Beacon Squier (U) Pasco 2,927 49.0 4. Tammrac Broward 5,078 48.8 Ft. Lauderdale 5. Miami Beach Dade 87,072 48.7 Miami 6. Cedar Hammock- Manatee Bradenton South (U) Bradenton 10,820 45.3 . Zephyrhills Pasco 3,369 44.6 . Newport Richey Pasco 6,098 44.4 . Ormond-by-the-Sea VOlusia 6,002 43.8 Datona Beach 10. Port Charlotte (U) Charlotte 10,769 43.8 11. St. Cloud Osceola 5,041 43.1 12. Lauderdale-by-the-Sea Broward 2,879 42.9 Ft. Lauderdale 13. Bay Harbor Islands Dade 4,619 42.2 Miami 14. Venice Sarasota 6,648 41.7 Sarasota 15. Englewood (U) Sarasota 5,182 41.6 16. Lehigh Acres (U) Lee 4,394 41.6 Fort Myers 17. Venice South Sarasota 4,680 41.4 Sarasota 18. Deltona (U) volusia 4,868 40.2 Sanford l9. Surfside Dade 3,614 40.0 Miami 20. Longboat Key Manatee 2,850 39.1 Sarasota-Bradenton 21. Dunedin Pinellas 17,639 38.7 St. Petersburg Table 22. Ken 23. Jasmi 24. Gulf 15. Palm W N. N. a. 88 Table 11. (Cont'd.) Percent Aged County or 65 Years Name Neighboring City Population and Older 22. Kenneth City Pinellas 3,862 38.6 St. Petersburg 23. Jasmine Estates (U) Pasco 2,967 37.4 24. Gulf Gate Estates (U) Sarasota 5,874 37.3 Sarasota 25. Palm Beach Palm Beach 9,086 35.9 26. Gulfport Pinellas 9,730 35.7 St. Petersburg 27. Holmes Beach Manatee 2,699 35.3 Bradenton 28. Bayshore Gardens (U) Manatee 9,255 35.3 Bradenton 29. Lake WOrth Palm Beach 23,714 34.9 Palm Beach 30. St. Petersburg Beach Pinellas 8,024 34.3 St. Petersburg 31. Collier Manor- Broward Cresthaven (U) Pompano Beach 7,202 33.7 32. Belleair Pinellas 2,962 33.3 Clearwater 33. Tavares Lake 3,261 32.9 34. Kensington Park (U) Sarasota 3,138 31.5 Sarasota 35. Fort Myers Beach (U) Lee 4,305 31.0 Fort Myers 36. St. Petersburg Pinellas 216,232 30.6 37. North Bay Dade 4,831 30.0 Miami *Only those places with at least 30 percent of their total population aged 65 years and over are included. (U) = Unincorporated SOURCE: U.S. Bureau of the Census, Census of Population: Population Characteristics (Florida). 1970, General 89 places, 14 are unincorporated. Unfortunately, in the detailed reports of the Bureau of the Census and other government agencies, these places are virtually ignored. Thus, if the survey results are to be supple- mented by aggregate statistics, an unincorporated location is unaccept- able. For similar reasons, the smaller incorporated places are also at a disadvantage. From this point of view, standard metropolitan statistical areas (SMSA's) are most desirable. In addition to serving as a fundamental reporting unit for most government departments, these administrative regions are also the only areas for which information is published on a census tract basis. In Florida, nine SMSA's were delimited in 1970. Only four have a proportion of elderly residents which is in excess of the national average (Table 12). Three of these (Fort Lauderdale- Hollywood, West Palm Beach, and Miami) are located along the South Atlantic Coast, and the fourth (Tampa-St. Petersburg) is situated in the midst of the aforementioned Gulf Coast concentration. If we subdivide this latter region on a county basis, a striking dichotomy emerges. Hillsborough County, with only 10.4 percent elderly residents, barely exceeds the national average. Pinellas County, on the other hand, has the fourth highest concentration of senior citizens in the nation with 29.5 percent. But even this figure is somewhat deceiving. Of the 23 incorporated places listed in Table 11, six are situated in Pinellas County.23 23These include: Dunedin (38.7 percent); Kenneth City (38.6 per- cent); Gulfport (35.7 percent); St. Petersburg Beach (34.3 percent); Belleair (33.3 percent); and St. Petersburg (30.6 percent). 90 Table 12. Florida SMSA's Ranked Via the Percent Aged 65 Years and Older- 1970 Percent Aged SMSA Name Constituent Counties 65 Yrs+ l. Tampa-St. Petersburg Hillsborough 20.3 Pinellas 2. Ft. Lauderdale-Hollywood Broward 17.9 3. west Palm Beach Palm Beach 17.3 4. Miami Dade 13.6 5. Orlando Orange 9.6 Seminole 6. Jacksonville Duval 7.5 7. Pensacola Escambia 6.4 Santa Rosa 8. Gainesville Alachua 6.3 9. Tallahassee Leon 5.5 SOURCE: U.S. Bureau of the Census, Census of Population: 1970, General Population Characteristics (Florida). 91 One of these, the city of St. Petersburg, was selected as the study area for this dissertation. With "only" 30.6 percent of its population aged 65 years and over, St. Petersburg ranks near the bottom of Table 11. Yet if we consider the size of each community, the situa- tion changes drastically. Just 4 of the 37 communities exceed 15,000 residents, and St. Petersburg is the unchallenged leader. It is nearly 2.5 times the size of the second largest city (Miami Beach), and over 9 times the size of the third largest (Lake Worth). There are more elderly residents in St. Petersburg than there are total residents in every urban place on the list except Miami Beach. Among cities of equal or greater size, St. Petersburg's statis- tics become even more impressive. Within the state there are three cities larger than St. Petersburg. The closest any comes to the 30 percent level of elderly residents is Miami with just 14.4 percent. On a national scale, the closest challenger is Portland, Oregon (14.8 percent). For a number of reasons, therefore, St. Petersburg was deemed a logical selection as a study area. To briefly recapitulate, it is situated in the midst of the nation's most outstanding concentration of senior citizens. And operating under the assumption that a significant portion of this concentration is the result of migrant additions, it is apparent that the area is especially appropriate for a study with the current objectives. Also, St. Petersburg is part of a unique SMSA. As a metropolitan unit it is the national leader in the level of elderly residents. When subdivided into its component parts, the SMSA also yields the greatest relative concentration of senior citizens among all central cities of 200,000 or more. 92 St. Petersburg is a logical choice in still two other respects. First, as a community with a diverse population, it avoids many of the potential research problems that can characterize planned retirement communities. These may be part of the community regulations, or they may even be implied. As an example, income is a particularly effective discriminator. When the housing options are expensive enough to elimi- nate certain middle and low income groups, the prospects for drawing a representative sample can be seriously threatened, if not entirely eliminated. From a second perspective, St. Petersburg profits from a compara- tively long tradition as a residential center for senior citizens. The quotations introducing this chapter offer supportive testimony to the city's rather unique reputation. The planned retirement community, on the other hand, is a relatively new feature on the United States housing landscape. Most developments have appeared since 1960 (Barker, 1966). If temporal developments are an important component of analysis, there- fore, a city similar to St. Petersburg is required. In the current context, the changing character of migration systems made St. Petersburg a logical study area. Its environment will allow us to compare the migration systems of an earlier period with those of the present. ST. PETERSBURG: GENERAL CHARACTERISTICS Once the decision to focus on St. Petersburg had been made, and before beginning the actual field investigation, it became necessary to develop at least a cursory understanding of the city. For obvious reasons, this familiarity is an invaluable prerequisite to planning a successful survey design. 93 From a number of general sources, it was determined that the city of St. Petersburg occupies approximately 56 square miles of land and freshwater lakes on the southernmost tip of Pinellas County. Located midway on Florida's west coast, St. Petersburg is surrounded on three sides by Tampa Bay, Boca Ciega Bay, and the Gulf of Mexico (Figure 7). This peninsular location has played an important role in the evolution of the city. Initially, the isolated situation posed a serious barrier to growth. Incorporation did not occur until 1892, and as recently as 1900 there were only 1,600 permanent residents. Only after 1920, when bridges and highways made the area easily accessible, did a notable growth trend emerge. Almost immediately, St. Petersburg became a preferred retirement center. The census of 1920 showed a population of 14,237. During the boom years that followed . . . persons settling in the community were relatively old. By about 1930 . . . both the tourist and resident populations contained large numr bers of aged and retired persons. The resident population in 1930 numbered 40,389 . . . and 3,008 (7.5 percent) were 65 years of age and over (Harlan, 1954, 333). During the next decade the community came to be characterized by a number of attributes that have since persisted. In particular, a seasonal fluctuation of population became very evident. Through the winter months, a large contingent of temporary residents would swell the city's population. And in a continuation of the preceding trends, these winter visitors had a high average age (Harlan, 1954, 333). The Second WOrld war introduced a short hiatus in St. Petersburg's rapid growth. But almost immediately after its conclusion, the expan- sion resumed at an even higher level. In fact, during the 1950's the city nearly doubled. As the decade began there were 97,000 full- time residents: by the 1960 Census, the figure had climbed to 181,000. 94 § 3:233 «mam mm. .wS mu \ \0003>.._JOI . -3353: 50... V .653 34a emu; V ss oxammmmema pm-zomxoaax\.\ wumwdz440: 3353.3 :8... . ‘ y . immovaw \\zomwz..Eo mmzz. 024 uo< 040 w 105.04.... ... ucaoi Odgo-I log ..... - a .......... 0: ammxmhwa Hm >._._..:mo—z n «9.04“. .o. mane... 105 is obvious that the inner city has lost its appeal. Apparently, many more migrants are moving directly into the St. Petersburg periphery. This is typical of the suburban expansion in nearly all major United States cities. Additionally, since most of these fringe tracts were shown to be areas of relatively high status in Factor 2, it would appear that mobility is linked to social status. One section of the older and more central portion of St. Petersburg does display moderate to high mObility patterns. The reference is to tracts 13, 14, 15, 35, and 36. These tracts comprise the center of St. Petersburg's senior citizen population.27 It is apparent that St. Petersburg is still a very attractive location for many retired migrants. Unfortunately, the Census Bureau's statistics prohibit meaningful statements regarding the origins of these people. Hopefully, the empirical section of this thesis will help to rectify this omission. Factor 4 - Old Age--Inner City. The significance of the age structure of St. Petersburg's population was first encountered in Factor 1; it is re-emphasized in Factor 4. In fact, these two factors alone account for 50.5 percent of the total variance in the original raw data matrix. Whereas the first factor indicated age related dif- ferences in family structure, the current dimension is more indicative of differences in life style. The positive loading variables show that areas with positive factor scores are characterized by: high dependency ratios (numerous retired people), large numbers of widowed 27Tract 13 had 50 percent of its population over 65 in 1960: Tract 14 had 65 percent; Tract 15 had 74 percent; Tract 35 had 55 percent; and Tract 36 had 53 percent. 106 people, low incomes, many renters (particularly in structures with three or more apartments where the rent is low), and numerous families with no automobiles. The variables median age of males and population over 65 load so highly on Factor 1, they do not link-up with this di- mension. It is obvious, however, that old age serves as the common feature of these positive scores. A significant exception to this general condition is presented by the positive loading of the variable, percent nonwhite. In Factor 1, this measure was associated with the younger population groups. Similar youthful family structures caused this apparent linkage. In the current case, however, nonwhite populations are more closely aligned with the elderly age group and the obvious explanation points to similar life styles. Very simply, the black ghetto is also an area of high de- pendency ratios (numerous children under 15), low incomes, few auto- mdbiles, and many multi-unit, low rent apartments. The intriguing' feature of this relationship is displayed in Figure 11. The black ghetto, comprising tracts 9, 10, 11, and 16, is immediately adjacent to the "senior citizens ghetto" (Tracts 13, 14, 15, 35, and 36).28 The final variable to load positively on this dimension (DHCENCIT-- different house same central city) is a mobility index. Unlike those of the preceding factor, however, this variable depicts intra-urban movements. In the case of the nonwhite population, Berry and Horton 28Although they are beyond the scope of this analysis, a number of interesting questions do arise. In particular, what is the relation- ship between the black and old age communities? It would be especially interesting to examine the situation in Tract 12 where 32 percent of the population is over 65 and 27 percent is nonwhite. 107 have recognized that "pressure for housing for Negroes is intense, prices for equivalent accomodations are higher, and Negro families move frequently, probably in a ceaseless search for a decent home" (Berry and Horton, 1970, 348). Economics, discrimination, and in some in— stances, choice have served to restrict this search to the central city. The processes involved in the high mobility level among the elderly are less clear but it seems likely that they are related to the high proportion of renters and the low levels of car ownership. This is to say that although many senior citizens do not experience the residential inertia that accompanies home ownership, they also can not afford to move out of the central city because they lack the flexibility of automobile transportation. The negative loadings appear to be typical of growing, middle- class suburbs. The people are married, employed, have middle incomes, and live in recently built, single-family houses. Once again it is apparent that migrants from other areas of the United States favor these suburban locations. Factor 4 presents a very distinct spatial pattern (Figure 11). There is a clear division based upon age, race, and life style. The inner city-suburban, and old age-youth dichotomies can hardly be de- picted with greater clarity. In summary, it is obvious that generational differences serve a very significant role in determining the spatial structure of St. Petersburg's population. Factors 1 and 4, which together account for over 50 percent of the variance in the original data matrix, are in- dicative of age-related distinctions in both family patterns and life 108 styles. In many respects, St. Petersburg may be viewed as a spatial dichotomy composed of the "old core" and the "young periphery." Although these observations are based upon the 1960 Census, they are no less significant.29 Undoubtedly, as a result of the dynamic nature of cities, a similar analysis, based upon the 1970 data, would produce a slightly different set of results. Given the evolutionary nature of the urban spatial structure, it is less likely that drastic changes would occur in just ten years.30 For some investigations even a minimal shift is significant. But since the primary objective in this case has been to establish a preliminary background for a subse- quent survey analysis, the minor changes are much less significant. Our main concern has been to develop an understanding of the "nature" of St. Petersburg. From this perspective, the significance of age differentials in the 1960 patterns is extremely clear-cut. A quick inspection of Figures 12 and 13 will show that the situation has changed very little. The elderly continue to comprise a major segment of the total population.> In fact, if we were to divide the population of each census tract into five age groups (Children under 5; Youth, 29For obvious reasons, it was necessary to conduct the factor analysis prior to the actual field survey. Unfortunately, the results of the 1970 Census were still unpublished at that time. It was not until the field experience had been completed that the more recent information became available. This set of circumstances made it imr perative to use 1960 statistics. 30A recent study by K. Haynes furnishes strong support for this observation. After comparing the patterns produced by a 1961 analysis ‘with those emerging from.a 1951 analysis, he concluded that "although the internal consistency of factors is not as great as might be de- picted, the consistency is reasonably reliable. The spatial distribu- tion of factors created by this analysis is extremely stable through time" (Haynes, 1971, 334). 109 Ohm. .cozoSaoa *0 3980 .mj 683m 7 o.~ nod ........ ..... ...... ....... .0.“ '1 a 0H0Hnum 0 a ”n ...” o a. a H................. x .. mm. a ,.n........a...». 0 0 0 0HHHOHOH0 ~— nu...u...uuun. "nun“ O. 0 .n.n.nun.n.n.n.n.n. 0000000 00000000H0H0H0H0H0Ha v0H0H0H0H00 0 ...... ”xv. ”an”. : .. .HHHHHHHDN. “Nun“ "nu”nuuunnu“ A u.“ .H.H.”.u.....u.u. 23 .o. . . . . “RR.“ ”H.” HOUOH0”. n.— ”0”.” H. n. H 0F. “a 0000 0 0H0H0H0H00H0H0nunun v0”. “NO". n“ 0”“0 0 000000000000 000000 0.0.0 0 000uc0a 0 H0000 HHHHHHH”. _N 0n0u ””0H0H0HHHH ”0H mw:mwwmmmwnuummwu mequq.wmwn . . :4:eaa.a ”Wm"...mwm m». "we... u” “we“? mapmnww ”.....m. A . . . nuns “mun. mun?” rude ”nun” N nun .....umaaamma 3.3%.. . 5.9 .nwmmmuwmmfiwm_m....~mwww enema” a .nmnmnm.§i£_ . 000000000000000 "W000” v0 0 0 0 0 0 0 00HHHHHH0 0000000000ufl§0000 Nara" ”HHHHH ....... . . . o. . . . ........... ”mun” . .nuuu "mum“ ..F.» .03.. . . ..... U... H . _ H .M Wmnmumnmum. hp. ‘ pupppvpbp a J 35.3.“. .......... ~0c°¢ ”NH” m‘uuu N. ................ ...OCIOIOOUA M O .3 .nuunununnn“ . H . H . H . ....... . .....wwwmummmmmmm x; meme”was.new .. ....... ................. ............................. ............................ ............................. . ............................ " ............................ ................... .................. .................. .................. ........................... ........................... .......................... ........................... .......... .......... .9: pzuucua Swami are m. aw .525 «m6 ozq no. we» whinesm” mean» no .amwnm mu“ :91; z. «.63: mamzmo a: to 6.595% .n. UKDOE N. MKDGI _omammmmtma, .5 110 5-19; Young Adults, 20-34; Mature Adults, 35-64; and the Elderly, 65 and over), we would discover that the elderly comprise the largest group in over half (54 percent) of the cases. Spatially, the "old core" and "young periphery" are still apparent. The overall evidence is convincing. St. Petersburg does represent an excellent choice as a study area for a survey analysis of aged migra- tion. But with over 66,000 residents aged 65 years and over, and with an unknown number of younger retirees, it is obvious that only a portion of the potential respondents can actually be interviewed. We can eliminate some of the original group by establishing a precise defini- tion of the subject population. The final selection of respondents, however, must be based upon a logical set of sampling procedures. CHAPTER V ATTRIBUTES OF THE SURVEY DESIGN OPERATIONAL DEFINITIONS The survey population was identified as those households that migrated to St. Petersburg, Florida, EEEEE.the retirement of the major family wage-earner. Notice that the definition does not include an age criterion. This omission is linked with the recognition that a single standard retirement age is nonexistent. Undeniably, the most popular approach is to associate retirement with age 65 for men, and 62 for women. But there is nothing sacred about these limits and people can and do retire both before and after they achieve these levels. If an age re- quirement were utilized, therefore, "early" retirees would be virtually eliminated from consideration. This could be a very serious omission if the age at retirement is a major influence in the migration process. As implied in the preceding chapters, the key element in the definition is the stipulation that the major wage-earner must have been living elsewhere when he or she retired. It makes no difference if his previous residence was either in another Florida community or in another state. As long as the family members moved to St. Petersburg after the retirement, the household is a part of the eligible population. In an effort to avoid any unnecessary inaccuracies, the decision as 111 112 to the identity of the major wage-earner was left up to the respon- dents.31 Occasionally it was difficult to determine if the major wage— earner was actually retired. To illustrate, consider the case of a person who retired and migrated to St. Petersburg but who then took either a full or part-time job in the city. Although it is possible to argue that the individual is no longer retired, and hence that he is no longer an eligible subject, the approach in this analysis was to consider the person's status as he migrated to St. Petersburg. This decision was based upon our interest in the retired migration process. Obviously, from this perspective, the key characteristic is the status that existed during the move. Once the retirees had reached St. Petersburg, their employment activities were of little consequence. The decision to employ households instead of only males or only females is in response to the high mortality rates of this age group. A preliminary census analysis had indicated that the number of single person households in St. Petersburg was quite high, and as might be expected, the widowed group was especially large.32 If we recognize 31In this manner, a number of arbitrary judgments were avoided. As an example, if the husband was retired but the wife had continued to work, the respondent was asked to indicate which of the two was the major wage-earner when they both were employed. If it was the husband, the household was eligible for further consideration, but if the wife was always the major wage-earner, then the household failed to meet the requirements and it was eliminated. 32Among persons aged 14 years and over in 1970, 23 percent (23,000) of the females were classified as widows and 6.6 percent (5,000) of the males were classified as widowers. From an already low sex ratio of 82 for the entire city, the ratio falls to 22.3 when only ‘widowed persons are considered. The difference in these figures is 131981? explained by the sex differential in mortality. the c to as in t1 to St earns popu] membe were a wid circr Vanta the 1 life, the f is tr the s thEn it we Sid 113 the close relationship between age and mortality, it is only logical to assume that the largest proportion of these individuals is found in the highest age cohorts. Quite a sizeable number probably migrated to St. Petersburg when their spouses were still alive. Whenever this action followed the retirement of the major wage- earner, the household was obviously a legitimate part of the defined population. It would be a serious mistake to ignore the remaining member on the basis of his or her sex. To illustrate, if only males were eligible as respondents, all the qualified households with only a widow remaining would be eliminated from consideration. Under those circumstances, the more recent migrants would receive an unfair ad- vantage during the selection of a sample. Very simply, since most of the retirees arrive in St. Petersburg in the later stages of their life, the longer their residence, the greater the chances that one of the family members will pass away. Consequently, unless the household is the unit of interest, the remaining individual could be ignored in the sample. Because some retirees spend the winter months in Florida, and then return to their "permanent" homes for the remainder of the year, it was necessary to make still another decision concerning the nature of the population. Specifically, should these seasonal migrants be accepted as valid subjects? As a compromise solution, it was decided that a minimum residence requirement of six months would be observed. Hence, in order to be eligible, the retiree must average at least six months of each year in St. Petersburg. Those staying less than this amount were not con- sidered residents. For them, the trip to St. Petersburg was viewed as more the p also to St two I it wa ignor 1:: lm Sim:- 114 more of an extended vacation than an actual move. Although some of the people that average at least six months in the city are undoubtedly also vacationers, most are simply part-time residents. They have moved to St. Petersburg, but not entirely. Instead they prefer to maintain two residences. As abonafide segment of the post-retirement population, it was determined that these seasonal migrants were too significant to ignore. SAMPLING PROCEDURES The Samplinngrame Once the target population was defined, and before a sample could be drawn, it was necessary to locate an appropriate sampling frame. Theoretically, this frame should be an adequate, complete, accurate, up-to-date, and convenient list of the qualified population units (Moser, 1958, 121-122). Whenever feasible, duplications must be avoided. In social science research these stringent requirements are very rarely satisfied. With a target population comprised of aged migrants, the criteria may even be impossible to fulfill. Migration is a con- tinuous process, and as a result, it is very difficult to maintain an up-to-date listing of the participants. Even if a cut-off date is employed, the frame can include several units that have since changed their residence. If these people become part of a sample, it will be necessary to relocate them before they can be interviewed. Among the elderly, another disruptive component is high mortality. Sampling frames which include non-existent units are inaccurate. Clear] chance other analy: POPUIJ all t] ment . and p. if th would Opera The f ”Oulc in Si bECa: finch Cate den at \o OVQ til Car 115 Clearly, because the aged possess such high levels of mortality, the chances of obtaining an inaccurate frame are also quite high. In addition to these general problems, there were also several other difficulties which were specific to the characteristics of this analysis. A re-examination of the exact attributes of our desired population will show that the ideal sampling frame would have included all the households that migrated to St. Petersburg after the retire- ment of the major family wage-earner. If a household made regular and periodic movements to another area, the members could only qualify if they averaged at least six months a year in St. Petersburg. A list of people over 65, or perhaps one of retired residents, would have closely approximated our desired frame, but in view of the operational definitions, neither would have been entirely adequate. The former roster would fail to include young retirees, and the latter would incorrectly include those who retired after already residing in St. Petersburg. Seasonal migrants created another serious problem because it was impossible to gain any indication of their magnitude, much less their identity. Given this situation, a number of options were available. The easiest choice would have been to forego a sampling frame and simply resort to a nonprobability approach.33 Unfortunately, the lack of h; L. 33It should be noted that the original efforts fell into this category. Without the benefit of an adequate sampling frame, it was decided that an appropriate alternative would be to interview retirees at the Senior Citizen Center, Inc. This facility, which is located in a census tract with 80 percent of its residents aged 65 years and over, is an activity center for the elderly.' Members pay nominal dues each year and this entitles them to take part in the center's various activities including dancing, bingo, card parties, shuffleboard, checkers, state societies, and bus objec It be a SEE cons whic and H119: alth the 116 objectivity inherent in this procedure was too crucial to be ignored. It became evident that a sampling frame was necessary, and as a result, a search for the "best available" frame was initiated. Emerging from this probe was the decision to utilize the §£:_ Petersburg City Directory (Polk, 1971). With certain reservations, this source afforded a sufficient list of retirees. The directory consists of a partial enumeration of the St. Petersburg population in which the head of each household is listed in two ways: alphabetically, and sequentially by street address. Among the scant information published for each of the units is a designation of the residents' employment status. On this basis, therefore, it was possible to iden- tify retirees. Unfortunately, it was impossible to determine if they migrated to St. Petersburg before or after their retirement. Also, although the relatively recent publication date (1971) was an advantage, the emphasis upon permanent residents was a disadvantage. Specifically, trips. Seasonal memberships are available for visitors (the total number of members was estimated as 1850), and nonmembers can also take part if they pay a fee either for the day, or for the specific event. Although the daily attendance was difficult to measure, for some func- tions, in particular dances, it was known to exceed several hundred people. On this basis, it seemed both logical and efficient to conduct all the interviews at the center. Permission was granted by the direc- tor, Mr. Claghorn, and work began. Before long, it became obvious that the situation was less than ideal. Most of the retirees came to the center to engage in some sort of an activity. This preoccupation seriously restricted the opportunities for a personal interview. Additional problems arose because a large proportion of the people that were contacted were only temporary visitors in the city. To illustrate, 30 people were approached as potential subjects and half were ineligible seasonal residents. Of the remaining 15 contacts, 12 consented to the interview. In view of the fact that it took more than a full week to accomplish this disappointing total, and that the ap- proach was obviously subjective, a decision was made to change to a new technique. an unknown list. Whe accepted a a Exes 2 In a function 0 to be avoi then randc Of these It verse, the Consider-at It 1' can °nly t accurate 6 City diI‘EC Elements. St. Peters 117 an unknown number of seasonal migrants were undoubtably absent from the list. When all aspects were considered, however, the directory was accepted as an appropriate sampling frame. Thg'gyp§_g£'8ample In any survey the type of sample design that is utilized is a function of a number of considerations. If bias in the selection is to be avoided, and if the precision of the results is to be calculable, then random methods of selection must be employed. The precise form of these methods will be influenced by the characteristics of the uni- verse, the objectives of the analysis, and the significance of practical considerations such as time, labor, and money costs. It is also apparent that regardless of the design, the sample can only be as precise and representative as the sampling frame is accurate and complete. In the current context, we have noted that the city directory could not provide a perfect roster of the population elements. Although it did yield a list of the retirees residing in St. Petersburg, that list could only have been as accurate as the direc- tory was itself. Even if we assume complete precision from this per- spective, the frame was still less than perfect. Some of the units on the list were actually ineligible because they retired aftg£_they migrated to the city. Other units, such as seasonal migrants, were eligible but not included. As a proportion of the total frame, these elements were relatively insignificant. But as a source of error that existed even before the sample design was selected, they were too important to ignore. With t represent t) the best prc jective and design was s Cluste EOSt import; located in 1 Rated. om Consideratic a 1°uld have been a very difficult task. The negative correlation coefficient cent black problems 0: of both bl 9 comprise high prepc The geneity o as it may The propc is actua] recall 1:} tion bec< disparitj apparent t° the c Particip Th Wbeer O] 128 coefficient (r = -.49, significant at the .001 level) between the per- cent black and the percent aged 65 years and over, is indicative of the problems one would have in identifying an area with a high concentration of both blacks and retirees. Of the 56 tracts in the city (1970), 9 comprise the black ghetto, and of these, just three show a relatively high proportion of senior citizens.37 The second circumstance which contributed to the racial homo- geneity of the sample is implied in these statistics. As surprising as it may seem, there are relatively few black retirees in the city. The proportion of senior citizens in the black population (6.6 percent) is actually below the national average (7.0 percent). And when we recall the level of white retirees in St. Petersburg, this low propor- tion becomes especially striking. Although a full explanation for this disparity is beyond the current objectives, it is hard to ignore the apparent importance of a racial differential bathe movement of retirees to the city. Either retired blacks favor other communities, or they (participate in post-retirement migration in much lower proportions. The available evidence supports the latter interpretation. As Ian example, in a 1956 discussion of migration streams in Florida, I. ‘Nebber observed that: 37Tract 216 presented the best combination with 59 percent black, and 34 percent aged 65 years and over. The other possibilities were Tract 201.01 with 17 percent black and 30 percent elderly, and Tract 218 with 24 percent black and 31 percent elderly. Unfortunately, e‘rdeen these combinations do not guarantee the presence of a large num- be): of black retirees. As an example, although 24 percent of the popu- 1ar1:ion in Tract 218 is black, and 31 percent is aged 65 years and over, Only 7 percent (36) of the 538 aged residents are black. The 5 Years, but this relati forces. Tl “We. rec blacks in 1 terms of do'~1btedly Poor as e1 for whites A se QEneral C11 in the 801; (United St of migrant vey teem, Herthafist. a similar Unf< certainly I [although] large numbers of white persons have migrated to Florida subsequent to retirement, it is usually assumed . . . that Negroes have not shared in this movement in appreciable numbers. This notion is based in part on the existence in the state of the traditional majority-minority social structure of the South, which is believed to make it unattractive to members of nonwhite races, and it is reinforced by the absence of noticeable facilities for non- white retired persons (Webber, 1956, 326). The social environment has probably improved in the subsequent years, but black aged migration remains relatively low. Apparently, this relative stability is a reflection of a more fundamental set of forces. The United States Senate Special Committee on Aging, for in stance, recently outlined several pertinent "hazards" facing aged blacks in the United States (Special Committee on Aging, 1971). In terms of the propensity to migrate, the most important hazard is un- doubtedly poverty: "old Negroes are more than twice as likely to be poor as elderly whites--50 percent in poverty compared to 23 percent for whites" (Special Committee on Aging, 1971, vii). A second deterrent to migration may be something as basic as the general distribution of the black population. Over half still reside in the South, and among the elderly, this same proportion is 61 percent (United States Census of Population, 1970). The forthcoming discussion «of migrant origins will show that an overwhelming majority of the sur- “ney respondents came to St. Petersburg from the heavily populated Northeast. If we assume that the system of black aged migration displays ll similar spatial bias, the relative lack of black retirees in St. Petersburg is less surprising. Unfortunately, the racial composition of the sample prevents any additional comments on black migration. It is a subject which is certainly worthy of more intensive empirical analyses. Sex __d_ _Ma_r1 It is females. I birth, defi 105. As th causes the in 1970 tqu lation of T continued I females Va over 65 ag 75 Years When hummug ratio Was natiOne]. é residents‘ Um fact ( 111.5. T tionndire designed it was Po usually r diameter 130 _S_e_:_:_ a_n<_i_ Marital Status It is a well-established fact that there are more males born than females. In the last three decades, as an example, the sex ratio at birth, defined as the number of males per 100 females, has been a steady 105. As the population ages, the longer life expectancy among females causes the ratio to decrease. This trend is reflected in the fact that in 1970 there were 94.5 males for every 100 females in the total popu- lation of the United States. Below age 18 the early male predominance continued but at a decreasing level. The numerical superiority of females was most apparent among the elderly. More specifically, the over 65 age cohorts had a ratio of 61.5, and among the people aged 75 years and over, there were 43.8 males for every 100 females. When this knowledge is combined with the age structure of St. Petersburg, it is not surprising to observe that the city's overall ratio was 78.2. Although this figure is obviously well below the national average, it should be noted that the sex ratio of the elderly residents, which was 62.2, was much closer to the mean. In view of the fact that the sample used in this analysis was drawn from the latter age group, we might expect a similarly low ratio among the survey reapondents. This was not the case. Male respondents were over- represented; and in fact, the sex ratio of the sample was a very high -111n5. This anomaly is partially explained by the nature of the ques- tionnaire. To recapitulate, the questions in the schedule were dGaigned to gain information about the major family wage-earner. If it was possible, whenever a married couple was contacted, the husband ustually replied. This meant, of course, that the various background Characteristics, including sex, reflect the husband's attributes. The f recognizes large major the sex rat sex structu married hou cedure was t0 87.9. direction 0 PrEdominan majority. Notak average. . “Plain, bi CEding dis if t1'Iere i the elderl requisit‘e 1970 CenSL (74 PErCe In fact, The of prOpor my 27 p for the w (Mesh-On Singing 131 The full impact of this procedure becomes evident when one recognizes that there were 141 married couples in the sample. If a large majority of these households were represented by the male member, the sex ratio would obviously supply a poor indication of the true sex structure of the whole sample. As a remedy, one could assign each married household a value of two in the calculations. When this pro- cedure was adopted in the current analysis, the ratio fell from 111.5 to 87.9. In other words, this simple revision completely changed the direction of the relationship. From a situation indicative of male predominance, the ratio shifted to a level where it reflected a female majority. Notably, the sample ratio was still well above the St. Petersburg average. This shortage of women in the sample is difficult to fully explain, but several pertinent factors are discernible. From the pre- ceding discussion, it is apparent that a low sex ratio can only occur if there is a large number of single person (female) households. Among the elderly, the longer life expectancy of females satisfies this pre- requisite by producing a large, proportion of widows. According to the 1970 Census, the typical elderly man in the United States is married (74 percent), but the typical elderly woman is a widow (54 percent). In fact, only 36 percent of all aged females are married. The high ratio in the sample is indicative of an irregular set 0f proportions. Sixty-seven percent of the females were married and only 27 percent were widows. This may partially reflect a tendency fol? the widow population to be less cooperative in responding to the cI‘v1¢"38tionnaire, or it may also be an unfortunate implication of the 8‘milling techniques. The key question, however, is one which can not be answered migration? with those after migra sequent nov intensive a £95 TherI the major I sane persow As F bemeen th 65 is not 62 a femal at 65 the The their Sixt theme“ . 132 be answered here: Does marital status have an impact upon post-retirement migration? Specifically, how do the migration rates of widows compare with those of married couples? How frequently do women become widows after migration, and does this change in marital status initiate sub- sequent movements? These, and other related questions, deserve a more intensive analysis. are There are two pertinent measures of age available in the survey: the major family wage-earner's age at retirement; and the age of this same person when he or she moved to St. Petersburg. As Figure 15 discloses, two-thirds of the respondents retired between the ages of 60 and 69. The obvious popularity of ages 62 and 65 is not unexpected since these are the "legal" retirement ages. At 62 a female worker may retire with full social security benefits, and at 65 the same option is open to men. The fact that one-fourth of the major wage-earners retired before their sixtieth birthday is emblematic of the importance of early re- tirement in the United States. For some of these people, the decision 38As a prelude to the interpretation of these data, it should be anoted.that social scientists have discovered that some people misstate ‘thedr age. In the view of Thomlinson, "statements regarding the age of very old persons are of highly doubtful accuracy, for after they reach age 60 or so, many people exaggerate their antiquity as a point of Pride" (Thomlinson, 1965, 121) . On a more general scale, Bogue has noted that if people are asked to report their ages directly, "those With less education and those who wish to hide their age tend to give a biased answer or to round their age to the nearest digit ending in either zero or five" (Bogue, 1969, 148) . Occasionally the misstate- ment may even be unintentional. Among the elderly, for example, a faulty memory may lead to inaccurate estimates. In any case, the inter- pretation of age data should always be approached with caution. fit FIGURE I5. TWO IND W LII! i wiirl KIWI -, XOCODOOLK Frequency 133 FIGURE I5. TWO INDICES OF THE SAMPLE'S AGE DISTRIBUTION Age at retIrement (2:62 yrs.) Age when moved to St. Petersburg (Y: 63.8 yrs.) 60 & O I -.-. -.____q to lea sidera favor; are a that trend lengt ular 2.9. will so;s Smal CO ul it i The anal 933‘ Of 1 sYnc SizE 134 to leave the working force early was unavoidable because of health con- siderations; for others it was a voluntary choice which reflected a favorable set of pension arrangements. If recent labor negotiations are a true indication of future developments, it is quite conceivable that young retirees will expand immensely. To be specific, a current trend in union contract demands is a policy of retirement after a fixed length of employment. In these plans, instead of adhering to a partic- ular age, eligibility for retirement will be based upon work experience, e.g., 30 years of service. If and when these demands are accepted, it will be possible for people to retire in their late 40's and early 50's.39 At the other extreme of the age distribution the proportions are smaller. Just 19 percent of the migrants retired after age 65, and only 4 percent retired after age 70. Although this shortage of "old retirees" could be a normal situation produced largely by compulsory retirement, it is also suggestive of the importance of age in migration planning. The implication is that there is an inverse relationship between the age at retirement and the propensity to engage in long distance migration. In support of this observation, Honnen, Eteng, and Marshall have 39Given the degree to which young retirees are prevalent in this analysis, and in anticipation of these future developments, it is essential to reassert the necessity for defining retirees on the basis Of their employment status and not their age. Retirement is not synonymous with age 65 and to operate as if it is, is to ignore a sizeable, and growing proportion of the total retired population. discove had ret Florid. They c be the Eteng, indie; Proce: exactj Peter: a com; DEarL three The m films who ‘ it s. freq that abrt thes 135 discovered that whereas 27 percent of a sample of Wisconsin retirees had retired at or above age 70, a similar sample of retired migrants in Florida showed that just 7 percent were still working after age 69. They concluded that "the older the retired person, the stronger will be the tendency for him to remain in his state of origin" (Honnen, Eteng, and Marshall, 1969, 22-24). By adding the second distribution in Figure 15, we can gain an indication of the elapsed time between the act of retirement and the process of migration. Although the two distributions do not coincide exactly, they are very similar. Evidently, most retirees move to St. Petersburg either immediately after they leave the labor force or after a comparatively short hiatus. A specific tabulation discloses that nearly half of the retirees (48 percent) moved within a year, and within three years, over three-fourths (77 percent) had changed their residence. The mean elapsed time for the overall sample, including those who moved almost immediately, was just over two years. If we consider only those who waited at least one year, the same statistic rises to four years. With most of the retirees waiting only a short period to move, it seems probable that some form of pre-retirement planning was a frequent occurrence. This is a logical conclusion since it is unlikely that a retiree would decide to leave his home, friends, and relatives abruptly. Later in this chapter we will consider several aspects of these migration preparations. It should also be noted that the elapsed time statistic is based upon the age of the subject when he or she moved to St. Petersburg. Hence, if the migrant took a little longer than average to reach the city, it ‘1 volved in Income B_e£ Cons characteri the reSpo ditions, ' dEfinitiv a more he toPits the Firs Came to S1 50 Percen' my 3 pe acteristi ti“ by I Wore than has attra 136 city, it is conceivable that a number of "steps" or "stages" were in- volved in the process. Income Before and After Retirement Considerable caution is required in the analysis of the income characteristics of the sample because nearly half (48.7 percent) of the respondents refused to divulge the information. Under these con- ditions, it is presumptuous to consider the more salient generalizations definitive. Instead, until additional information becomes available, a more heuristic approach is suggested. Fortunately, there are two topics that hold particular promise. First, it is clear from the survey that most of the migrants came to St. Petersburg from modest financial backgrounds. Whereas 50 percent had a pre-retirement family income of less than $10,000, only 3 percent managed to earn at least $20,000.40 Although this char- acteristic is somewhat surprising, it concurs with an earlier observa- tion by I. Webber. Specifically, on the basis of a survey conducted more than two decades ago, Mr. Webber indicated that "St. Petersburg has attracted people from all walks of life, but . . . not, for the 40Because the respondents retired at different times, even these statistics can be deceiving. Quite simply, the value of the dollar fluctuates. Therefore, if two individuals with the same pre-retirement incomes retired at a 20 year interval, it would be improper to equate their incomes. According to statistics from the United States Depart- ment of Labor, for instance, the purchasing power of a dollar (1967 = $1.00) was $1.38 in 1950, and only $.86 in 1970. In effect, a $10,000 income in 1970 would have purchased over $15,000 worth of goods at the 1950 prices. Obviously, if absolute comparisons are to be meaningful, adjustments for this situation are essential. When relative classifi- cations are involved, the adjustments are less crucial but still imr ;portant. Temporal inequities need to be considered in all cases. most part, haven for n Alth only wealtk not be int tion diffe is quite c °f POSt-re is dead reflected Wong non ha eatlie Etenga an Of a Samp than $S,c 137 most part, people of great wealth . . . [it] has evidently become a haven for retired people with modest resources" (Webber, 1951, 94). Although this finding is in contrast to the common belief that only wealthy retirees can afford to migrate to Florida, it should not be interpreted as an indication that income is an unimportant migra- tion differential. To the contrary, the literature on this subject is quite consistent: on an interstate scale, income is a good predictor of post-retirement mobility.41 The decision to migrate to another state is clearly influenced by the retiree's financial status. This is reflected in a recent comparative analysis of the pre-retirement income among non-migrants in Wisconsin, and migrants in Arizona and Florida: While about two-thirds of the Wisconsin retirees had a [pre-retirement] annual income of less than $7,000, only one-third in Florida and Arizona were within this income range. In other words, there were more low income retirees in Wisconsin than there were present in the other two samples (Eteng and Marshall, 1970, 37). An earlier study had produced similar results. Specifically, Honnen, Eteng, and Marshall discovered that while more than half (53 percent) of a sample of Wisconsin retirees had pre-retirement incomes of less than $5,000, in the case of Florida retirees, only 18 percent of the full-time residents and 9 percent of the seasonal residents fell into this same income category (Honnen, Eteng, and Marshall, 1969, 31). Given the fact that many of the retirees in St. Petersburg came from modest financial backgrounds, and that there are other communities 41The interstate stipulation is necessary because in a study of intrametropolitan residential mobility, Goldscheider discovered that "not only do members of higher economic groupings among the older popur lation have lower rates of residential mobility (behavior and attitudes) but, of those planning or desiring to move, a smaller proportion successfully anticipated their mobility behavior when compared to mem- bers of lower socioeconomic groupings" (Goldscheider, 1966a, 105). where peol Miami Bea have spat favor a d migrants“: in effec1 tain fine "Open" or available °bleCtive future c, The the diffe P0! most acccmpan: Survey 0 that We of 70 Pe or “me fiQUre a to °°nc1 curtailm “new ( \ 42 whereas necks Specific percent Harshall 138 where people from wealthy backgrounds predominate (e.g., Palm Beach, Miami Beach), is it possible that the retirees' financial standings have spatial manifestations? That is, could it be that wealthy retirees favor a different type, or class, of community than the "poorer" migrants? Are the admission policies of some retirement communities, in effect, discriminatory because they only attract people with a cer- tain financial standing? And if so, does not this contrast with an “open" community, like St. Petersburg, where all levels of housing are available?42 Although questions of this nature are beyond our current objectives, they are fundamentally important, and hence, they deserve future consideration. The second salient generalization related to incomes concerns the difference between the pre-retirement and post-retirement incomes. For most people, the process of withdrawing from the labor force is accompanied by a drastic reduction in earning capacity. In their survey of Sun City (Florida) residents, Smith and Marshall discovered that over 70 percent of the respondents reported an income discontinuity of 70 percent or more. Whereas 84 percent had incomes of $10,000 or more before they retired, only 26 percent were still above that figure after they retired. This revelation prompted Smith and Marshall to conclude that "not only does the retiree suffer from economic role curtailment, he is also faced with changes in consumption and life style" (Smith and Marshall, 1970, 13-14). 42As tentative support for this hypothesis, we can note that whereas over half of the St. Petersburg respondents had pre-retirement incomes of less than $10,000, an earlier survey had shown that in a specific planned retirement community (Sun City, Florida), more than 80 (percent of the respondents had incomes exceeding $10,000 (Smith and (Marshall, 1970, 13). The retirement percent ha about thei experience $5,000 had $10,000 ha pared to the unpre 139 The St. Petersburg respondents followed a similar pattern. Before retirement, 8 percent had annual incomes of less than $5,000, and 50 percent had incomes below $10,000. Later, when they were questioned about their post-retirement income, it became clear that many had experienced sizeable reductions. The proportion earning less than $5,000 had increased to 61 percent, and the segment with incomes below $10,000 had expanded to 92 percent. When this latter figure is com- pared to the national standard (80 percent with incomes below $10,000), the unpretentious nature of the sample is reaffirmed. At this point, a word of caution in the interpretation of the above is approriate. Low post-retirement incomes do not necessarily mean low standards of living. Frequently, the elderly supplement their incomes by using savings or other previously accumulated assets. They also have smaller households to support, and because many own their homes outright, they make no rental or mortgage payments. In the process of trying to develop a realistic poverty level for the elderly the fed- eral government has taken many similar elements into consideration. Still the fact remains that every fourth senior citizen is poor.43 While it is virtually impossible to determine the number of "officially poor" respondents, the preceding statistics present a clear indication that many are either below the poverty level or very near to it. Also, the pre-retirement income figures seem to indicate that very few subjects were capable of supplementing their incomes from extensive 43For a more comprehensive discussion of l) the problems involved in measuring elderly poverty; and 2) some of the pertinent statistics on this topic, consult Measuring Adequacy of Income (Brotman, 1971b). savings. DOt res1 are-re P udgrati Elect 1 single mobili Circum P€rson tie hj to mic. t0 lCnl abili. time I virtu. will 1 l‘etirQ these a Stro reaSOn. resPOnc of n°n~n 140 savings. The obvious conclusion is that aged migration to Florida is not restricted to the wealthy. Pre-Retirement Mobility Past research has substantiated the importance of chronicity in migration. Defined as a tendency for observed mobility rates "to re- flect repeated and frequent movement by the same individuals rather than single moves by a larger number of persons," this concept implies that mobility decisions are linked to past experience as well as current circumstances (Morrison, 1971, 178). Most proponents argue that "a person who has once migrated and who has once broken the bonds which tie him to the place in which he has spent his childhood is more likely to migrate again” (Lee, 1966, 294). Each succeeding move is thought to lower the subsequent inertia even more. And conversely, "the prob- ability of remaining in the same place of residence increases as the time of residing there increases" (Olsson, 1965, 33). For the elderly, this concept has some intriguing aspects. By virtue of their age alone, it is conceivable that many senior citizens will have developed a very strong resistance to moving by the time they retire. Those with a history of infrequent migration, and in particular those who have remained stable for many years, can be expected to possess a strong attraction for their "home" and friends. If we carry this reasoning one step further, it is logical to also expect the survey respondents to come from relatively mobile backgrounds. The ideal test of this hypothesis would involve a comparison of the pre-retirement mobility of the sample with that of a similar group 0f non-migrants. However, since the structure of the survey prevented 141 this approach, an alternative method was necessary. Specifically, the respondents were asked to indicate the number of times they had moved in the 30 years preceding their retirement. Their answers fail to support the chronicity concept. Sixteen percent of those who could remember had never moved, and 72 percent had moved three times or less. Only 14 percent had changed their residence at least six times. To- gether these statistics make the fact that the respondents elected to move to St. Petersburg even more conspicuous. Quite simply, many of the migrants were forced to overcome immense inertia barriers. MIGRANT ORIGINS Having identified the major personal attributes of the sample, we can turn our attention to several aspects of the aged migration process. This section will be devoted to an examination of the migrant origins. In response to a question which was posed in Chapter II, the discussion will focus primarily upon the role of distance hoaged migration. Subsequent sections will consider both the preparations for migration, and the characteristics of the actual migration procedure. The Distribution 22 Origins According to the evidence supplied in Chapter II, most theories of migration recognize the importance of the friction of distance. In essence, they acknowledge an inverse relationship between distance and migration; as intervening space increases, the financial and psychic costs of overcoming that distance also increase and consequently, the propensity to migrate decreases at an undefined rate. Obviously, if this simple distance-decay proposition is applicable in aged migration, the major origins of Florida-bound retirees should 142 be concentrated in the near and adjacent states. Also, as one moves to the north and west of Florida, the contributions should become propor- tionally smaller and smaller. The validity of these expectations can be tested in this analysis by examining the distribution of the respondents' origins (Figure 16). This procedure produces a striking set of results. An overwhelming majority of the subjects came to St. Petersburg from the Northeastern quadrant of the country. In fact, if we rank the states on the basis of their contribution to the total migration stream (Table 15), the seven leading states, which were responsible for 79 percent of the na- tional migrants, comprise a nearly continuous horizontal belt stretching from Massachusetts to Illinois.44 Of the nine census divisions, just three (New England, Middle Atlantic, and East North Central) accounted for 87 percent of the national total. It could be that the pre-eminence of the Northeast quadrant is a result of a predominance in the number of peOple aged 65 years or more. If this were true, instead of the regional patterns in Figure 16 representing a distance or directional bias in elderly migration, they would simply reflect the distribution of elderly residents in the United States. At first glance, this argument appears to be quite creditable. Although Figure 1 displays a relative concentration in the Midwest, of the five leading states in terms of the absolute number of residents aged 65 years or more, four (New York, Ohio, Pennsylvania, 44 . . Fifteen subjects, representing 6 percent of the total sample (N a 245), came to the city from outside of the United States: four- teen were Canadian citizens; one originated in the Panama Canal Zone. Unless otherwise noted, the proportions referred to in this part of the text are based upon only the national migrants (N = 230). 143 l-|||J WFZ<¢O_! k0 KUOIDZ . . ® .4 / \ a mz_o_mo .mhz<¢0.2 no 20.._.Dm_x._.m_o .m. wmaoi 144 Table 15. Indices of the Relative Importance of the States of Origin* Absolute Percent of Index of Number of the Total Relative Respondents (N = 245) Contribution New York 39 15.9 173 Ohio 34 13.9 296 Michigan 30 12.2 342 Pennsylvania 27 11.0 186 Illinois 25 10.2 198 New Jersey 15 6.1 186 Massachusetts 12 4.9 163 Ontario 11 4.5 - Indiana 6 2.5 104 Wisconsin 6 2.5 108 Connecticut 4 1.6 121 Maryland 4 1.6 113 Missouri 4 1.6 61 Florida 3 1.2 27 Minnesota 3 1.2 65 Quebec 3 1.2 - California 2 .8 10 Delaware 2 .8 450 New Hampshire 2 .8 225 Virginia 2 .8 50 District of Columbia 1 .4 100 Georgia 1 .4 22 Iowa 1 .4 24 Kansas 1 .4 31 Kentucky 1 .4 24 Louisiana 1 .4 27 Maine 1 .4 67 Nevada 1 .4 200 Tennessee 1 .4 21 West Virginia 1 .4 4O 1 .4 Panama Canal Zone *The following states were not represented in the sample: Vermont, Rhode Island; North Carolina; South Carolina: Alabama: Mississippi: Arkansas; Texas: Oklahoma; Nebraska; North Dakota: South Dakota; Montana; wyoming; Colorado; New Mexico: Idaho: Oregon; washington; Utah: Arizona: Hawaii; Alaska. 145 and Illinois) also rank high as contributors to the sample. The lone exception is California. In an effort to neutralize this condition, and thereby gain a clearer image of the migration patterns, an index was calculated for each state. This involved two components: the percent of the total sample which originated in that state and the percent of the total aged pOpulation which resided in that state. A value of 100 indicates that the number of migrants sent to St. Petersburg was in exact proportion to the number of senior citizens in the population. The figures ex- ceeding 100 occur whenever a state is over-represented in the migration stream. Conversely, an index of less than 100 indicates that the state sent fewer migrants than would be expected from the magnitude of its aged segment. The distribution of this index is portrayed in Figure 17. The patterns in this map are such that most of the preceding comments still apply. The Northeast and North Central regions continue to dominate as origins: and several of the states in these areas were truly outstanding contributors. Perhaps the most notable example is Michigan, where for every 100 migrants that could be expected, 342 actually originated. The other leaders include: Ohio (296); Illinois (198); Pennsylvania (186): New Jersey (186): New York (173): and Massa- chusetts (163).45 When these listings are compared with the 45The index value for each contributing state is presented in Table 15. The states that failed to send any migrants to St. Petersburg have a value of zero. Among the others, three are very misleading. The high indices for Delaware (450), New Hampshire (225), and Nevada (200) are more a result of their relatively low levels of aged residents than their high levels as migrant origins. Together they only accounted for five migrants. It would be a mistake, therefore, to include them in a list of the leading origins. 146 m2; 5 .3353 m7: op 39.595on .Ezoizoaoma N. wane: aw. ms 147 absolute rankings in Table 15, only the order of the major origins changes. Up to this point the discussion has centered exclusively upon where the migrants were from. In order to comment upon the theoretical implications of this movement, however, it is equally important to identify where they Eg£g_ngt_from. As Figure 17 indicates, with the exceptions of Nevada and Maryland, every state south of the Mason- Dixon Line, and west of the Mississippi River, was under-represented in the sample. This paradoxical discovery deserves additional analysis. To begin, according to the friction of distance concept, as one moves north of Florida, the number of migrant origins should become proportionally smaller. In reality, the opposite occurs. The antici- pated inverse relationship between distance and the number of migrants is, in fact, positive. Instead of the near and adjacent states contrib- uting the largest share of retirees, the states farther to the north hold that distinction. This can be illustrated by focusing on the area east of the Mississippi River. The 14 states comprising the northern part of this region contributed 87 percent of the national migrants. Their combined index of contribution was 195. The 12 states in the southern portion, on the other hand, contributed just 7 percent of the national total, and their combined index was only 33. If the only factor which determined the number of migrants was the relative size of the aged component, both areas would have an index of 100. Because the South supplied only about one-third of its pro- portional share, however, and because the North supplied nearly twice as many migrants as expected, it is clear that other determinants were also involved. In order for the friction of intervening space to be an 148 appropriate consideration, the preceding values would have to be switched. As it stands, there were other, more important, elements, one undoubtedly being the search for amenities. The precise motivations of the respondents will not be examined until the next section of this analysis, but it is apparent that for many of the retirees the costs of overcoming the intervening space were secondary to the pull of amenities, and in particular, to the pull of the St. Petersburg climate. Given a national system of social security payments, retirees can also ignore that variable in favor of climatic comparisons. As a logical extension, this latter factor may be partly responsible for the under-representation of the southern states. Specifically, as Lee has postulated, "the volume of migration within a given territory varies with the degree of diversity of the areas included in that territory . . . A high degree of diversity . . . should result in high levels of migration" (Lee, 1966, 292) . In this case, the climatic differential between the northern states and Florida is obviously suffi- cient enough to stimulate a substantial flow of retirees. But in the southern states, where the differences are smaller, the flow is reduced. It is important to recognize that for each individual, the crucial element in the evaluation of the disparity is a perceptual judgement. If he is to migrate in response to climatic amenities, the retiree must perceive a significant difference between the conditions in his pre- retirement environment and the climate in St. Petersburg. This evalua- tion appears to emphasize the winter differential. Thus, whereas northern migrants can discern a clear disparity, the residents farther south appear to have greater difficulty in perceiving a significant 149 difference. In some instances a difference may be evident, but since it is exceeded by the forces of inertia (e.g., the friction of distance), the retiree is likely to be more reluctant to migrate to St. Petersburg. This latter observation hints at a potentially serious problem. Quite simply, it is naive to think that climatic amenities are the only pertinent explanation for the patterns in Figure 17. They are undeniably important in aged migration, and in fact, because they frequently take precedence over spatial considerations (viz., distance), it is essential that the interaction models be revised to include an amenity component. As the forthcoming discussion of migrant motivations will show, however, other elements also play significant roles in the process of retired migration. It is particularly important to realize that not all the movements are aspatial, that is, they do not all completely ignore distance. Whereas the patterns east of the Mississippi River supply clear evidence that an inverse relationship does not always exist between distance and the number of migrants, the national patterns display a definite distance influence. With the exception of Nevada, every state in the area west of the Mississippi River was under-represented in the St. Petersburg sample. Could it be that a distance threshold exists between the amenity areas in the South (Florida) and those in the West (Arizona and California)? On the basis of the origins they discovered in two retirement communities, one in Florida and one in Arizona, Smith and Marshall hint that the answer to this question may be yes: the more industrialized North Central States (with the exception of Illinois) sent a greater proportion of mi- grants to Florida. The more agricultural and more western states predominated in Sun City, Arizona. This effect may be interpreted as regional since the respondents from the 150 older midwestern, industrial states are closer to Florida, and the agricultural states are closer to Arizona (Smith and Marshall, 1970, 8). Unfortunately, the two sociologists did not pursue this subject in any greater detail. And without information concerning the specific origins of the retirees residing in Arizona and California, it would be pre- sumptuous for us to expand on the preceding ideas. This is another area which would certainly benefit from additional field surveys. gfig_§i§g.2£_the Community One final characteristic of the migrant origins comes from a reexamination of Figure 16. Just as most of the origins are concen- trated in but a few states in the northeastern section of the country, so are they concentrated in but a few areas within those states. Specifically, the respondent origins are dominated by large metropolitan centers. As the patterns indicate, three SMSA's (Chicago, New York, and Detroit) are truly outstanding. Together they accounted for nearly one-fourth of the total sample. In a similar manner, if all the metro- politan areas which contributed retirees to the migration stream are ranked according to the size of their contribution, the top 11 SMSA's account for 50 percent of the national sample and 47 percent of the total sample (Table 16). Unfortunately, the factors responsible for this predominance are difficult to identify. The obvious temptation is to emphasize the importance of population size. New York is the largest SMSA in the United States, and Chicago and Detroit rank three and five, respectively. The second largest SMSA, Los Angeles-Long Beach, only added one migrant, but in view of the preceding discussion, this is to 151 Table 16. United States SMSA's Ranked As Points of Origin for the Survey Respondents Number of Rank Among Name of SMSA Migrants U.S. SMSA's Chicago, Ill. 21 3 New York, N.Y. 20 1 Detroit, Mich. l9 5 Pittsburgh, Pa. 10 9 Cleveland, Ohio 9 12 Boston, Mass. 8 8 Philadelphia, Pa. 7 4 Buffalo, N.Y. 6 24 Newark, N.J. 5 l4 Akron, Ohio 5 48 Flint, Mich. 5 67 Allentown-Bethlehem-Easton, Pa. 4 58 Washington, D.C.--Md.-Va. 3 7 Milwaukee, Wis. 3 l9 Paterson-Clifton-Passaic, N.J. 3 22 Toledo, Ohio-Mich. 3 46 Syracuse, N.Y. 3 51 Canton, Ohio 3 80 South Bend, Ind. 3 113 Racine, Wis. 3 162 Baltimore, Md. 2 11 Cincinnati, Ohio-Ky.-Ind. 2 21 Miami, Fla. 2 25 Columbus, Ohio 2 35 Dayton, Ohio 2 39 Hartford, Conn. 2 49 Grand Rapids, Mich. 2 61 Youngstown-Warren, Ohio 2 62 Binghampton, N.Y.-Pa. 2 100 Atlantic City, N.J. 2 159 Los Angeles-Long Beach, Calif. l 2 St. Louis, Mo.-Ill. l 10 Minneapolis-St. Paul, Minn. 1 15 Kansas City, Kansas-Mo. 1 26 New Orleans, La. 1 31 Tampa-St. Petersburg 1 32 Louisville, Ky.-Ind. 1 40 Sacramento, Calif. 1 41 Albany-Schenectady-Troy, N.Y. 1 45 Norfolk-Portsmouth, va. 1 47 Jersey City, N.J. l 55 Springfield-Chicopee-Holyoke, Mass. 1 63 Bridgeport, Conn. l 76 Wilkes Barre-Hazelton, Pa. 1 87 Utica-Rome, N.Y. 1 89 152 Table 16. (Cont'd.) Number of Rank Among Name of SMSA Migrants U.S. SMSA's York, Pa. 1 90 Lancaster, Pa. 1 94 Reading, Pa. 1 102 Fort Wayne, Ind. 1 112 Rockford, Ill. 1 116 Lorain-Elyria, Ohio 1 122 Columbus, Ga.-Ala. l 128 Scranton, Pa. 1 130 Lawrence-Haverhill, Mass.-N.H. 1 133 Lowell, Mass. 1 140 Whelling, W. Va.-Ohio l 154 Springfield, Ohio 1 172 Portland, Me. __1_ 183 TOTAL = 191 (78 percent of total sample) 153 be expected. In other words, the large SMSA's in the West and the South (e.g., San Francisco-Oakland; Houstonyanuieven Atlanta) would be ex- pected to follow the same patterns of under-representation which char- acterized the states in those areas. The best section to test a size postulate, therefore, is in the Northeast region. With this in mind, Figure 18 is a scatter diagram of the northeastern SMSA's depicting both their population size (ex- pressed in logarithms) and their contributions to the migration sys~ tem. The curvilinear relationship is roughly composed of two parts. First, the metropolitan centers with at least 1,000,000 people show a positive trend. Philadelphia, with nearly 5 million inhabitants but only seven migrants, and New York, 11,500,000 residents and just 20 migrants, are major deviations. The areas with less than a million people comprise the second portion of the graph. And for these centers, a relationship is almost totally absent. In other words, for this category, city size can not be used to explain the predominance of some metropolitan centers and the relative insignificance of others. We can illustrate the irregularity of this distribution by re- ferring to several specific examples. For instance, eight SMSA's contributed three migrants each, and yet their respective United States population ranks ranged from 7 to 162 (Table 16). Racine, Wis- consin, with a metropolitan population of 171,000 contributed as many migrants as Milwaukee, Wisconsin (population: 1,400,000). Also, al- though Indianapolis, Indiana (1,110,000) failed to send a migrant, Flint, Michigan (500,000) furnished five. The obvious conclusion is that factors other than population size are involved. At this point it is conjecture, but it seems SMSA Population in thousands (expressed in Logs) FIGURE NUMBER OF MIGRANTS CONTRIBUTED TO THE no,ooo 9,000 epoo 7.000 spoo 5,000 4,000 3,000 2,000 r,ooo 900 800 700 ‘ 600 154 SAMPLE. BY SMSA SIZE ,.__., -- --—*—» -—--A — —— -—-- —-~ — - -.., —- -—-— —— - --— —- :——-——-4 O 0 .-—————————————-~ ———M- -....- ~— ,1 ~— —-----—.- .....___, O O __ ...—..m- ‘- . _ _. _*--‘__—__ ..l_.__. O O O O O Q 0 O h—"“‘— "T“ . ’ “ “* ‘ O O 0 . ___ -_ __ O . O _ --...e... l. .. 1 1. _ .__ - _ _ (p (D _ H___ __ ____ 7 __H" . ---__-_ I- _, ... ,, . L D : 0 1, . 8 I 1 1 . . O 1 O __“1-_...1___ r.._l____.l_.. 1-4...-“1. .. __1 _. 1.. .--n_ --.L..-._A._t.__1._..._L-.r-_.1-_._._.,| _ L 1 4 i L I 2 3 4 5 6 7 8 9 l0 ll I2 l3 l4 l5 l6 l7 l8 IO 20 2| 22 Number of migrants 155 possible that, particularly in the case of the "smaller" centers, the information flow between friends and relatives may have been an im- portant consideration. Once the initial retired household is situated, they may, in turn convince other retirees from their "old hometown" to join them in the Florida sun. And if this second wave of retirees should locate in the same neighborhood, their chances of being in- cluded within a single cluster would naturally be increased. The validity of this idea will be examined in greater detail in the next section. One final observation is necessary. In short, although 78 percent of the total sample came from metropolitan areas, a smaller proportion actually resided in larger cities (Table 17). As an example, over one-third (37 percent) originated in communities of less than 50,000 people. Given the fact that county units are used to delineate SMSA's, however, this differential is not surprising. Clearly, many people had resided within the metropolitan counties but outside of the central city limits. The fact remains, however, that the sample is undeniably urban. (wily nine respondents (3.7 percent) came from communities with less than 2,500 inhabitants. MIGRANT PREPARATIONS The decision to migrate is very rarely impulsive. To the contrary, the migration process usually commences as soon as an individual, or family unit, entertains the first thought of living elsewhere. It does not terminate until the household is "permanently" settled in a different residential site. In between, the migrants frequently engage in a number of crucial preparations. They begin by compiling a mental, if not a physical list, of the desirable characteristics. Then, on the 156 Table 17. Migrant Origins: Size of Community No. of Percent of the Size Category Migrants Total Sample Under 2,500 9 3.7 2,500 - 9,999 22 9.0 10,000 - 49,999 60 24.5 50,000 - 99,999 21 8.6 100,000 - 249,999 33 13.5 250,000+ 95 38.8 Information not available 5 2.0 245 100.0 157 basis of these criteria, they identify potentially acceptable destina- tions. These are evaluated, and the "best" site is selected. If this latter location is sufficiently desirable, in other words, if it compares favorably with the point of origin, the migrants may begin to plan their actual move. The objective in this section is to briefly consider three aspects of retired migrant preparations: (1) the factors which led to the decision to move to St. Petersburg, i.e., the migrant motivations; (2) the degree and source of familiarity with St. Petersburg (prior to the move): and (3) the extent to which the retirees planned their move S . Migrant Motivations The very fact that the retired migrants decided to move to St. Petersburg is indicative that strong motivations were involved. Research efforts have consistently shown that migration usually will not occur unless the motivations are sufficient enough to overcome the barriers of inertia. And it is clear that these barriers were formidable for many of the survey respondents. In addition to a natural inertia, which almost always exists, the retirees were also faced with the restrictions created by a drastic reduction in their earning capacity. Very few individuals were wealthy before they left the labor force, and thus, the subsequent losses were large enough to force several family incomes below the federally recognized poverty levels. It is only logical to assume that this reduction became an important consideration in the decision-making process. 158 Another source of resistance was identified in the earlier dis- cussion of the respondents' pre-retirement mobility. In short, since the probability of remaining in the same place increases as the time residing there increases, the relative stability of the sample should have continued after retirement. Given the low levelscnfmigration which characterize the United States elderly population, it is clear that similar forces are very effective in persuading large numbers of retirees to disregard migra— tion as a realistic retirement Option. But in view of the fact that the survey respondents were able to overcome not only these forces, but also those that were created by the substantial intervening space, it behooves us to identify and examine the powerful motivations that were clearly involved. Consistent with migration theory, these motivations can be classi- fied into two general categories: (1) push factors, which reflect dissatisfaction with the point of origin; and (2) pull factors, which reflect the attractions perceived at the point of destination. Tradi- tionally, the emphasis has fallen upon economic elements, and in partic- ular, upon employment considerations. Evidence to this effect is presented in Chapter II where a relevantly crucial question was posed for the first time: are aged migrations economically motivated? The suggested answer, which has reappeared several times in the literature and in this study, and especially in conjunction with the migrant origins, was that aged migrations are frequently motivated by non- economic considerations. The predominant element varies but in the context of the current sample, it has been suggested that amenities play a crucial role. 159 In an effort to test this contention, the subjects were asked to indicate the two most important reasons for their move to St. Petersburg. These responses are tabulated in Table 18. Unfortunately there are two weaknesses in this approach. Although it is conceivable that more reasons could have been involved, the respondents were restricted to giving only two motivations. Fortunately, the severity of this complaint is reduced when we realize that 77 percent of the migrants chose to list only one item. On the other hand, the second problem is more serious. By requesting only the most important reasons, it is impos- sible to evaluate the relative significance of any other factors. If, as in this case, it is desirable to also judge how unimportant various considerations are, the preceding approach is of no assistance. With this dilemma in mind, therefore, the retirees were also asked to evaluate the specific importance of nine motivations. These items were derived from the literature and they are presented in Appendix B as part of the interview schedule.46 The insignificance of the traditional economic motivations is clearly portrayed. In response to the open questions concerning 46The schedule also shows that the intended procedure was to have each migrant evaluate each item on the basis of a five point scale. Unfortunately, before too many interviews were completed, it was evi- dent that the design was somewhat faulty. Very few respondents chose to differentiate between the five degrees of importance. Instead, it was quite common for a person to view the decision as a dichotomy: the item was either important or unimportant. Hence, the five point scale was abandoned and a dichotomous choice was adopted in its place. 'The responses of those few people who had acted on the basis of the «original scale were adjusted to reflect this revision. 160 00.00m 00¢ 00.00H mom 00.00H mom qmaoe mn.0n 00H mh.os 00H 00.0 0 um3mco oz mm.m h NN.H m mo.a v msoocmHHoomaz n0.m m v0.m m m0.a o mcw>fla Hmowfiocoom em.n 0H Hm. m mm.0 0H mooufiuou mom oomHm 00Hs\ufi ooqu vm.n 0H vv.m 0 00.v NH mm>flumaou u0\oce mocofium 00.0 0 00.0 0 00.0 0 mGOwumuooncoo mmocwmom no 900 pm.m m 00.0 0 nm.m m muouomm £05m 00.0 HH mm.m b mo.H v mofiufiaaomm HMGOaumouomm mm.nm no vo.m m Hm.mm m0 mcoflumuwoflmcoo spammm mm.oo 00H m0.HH mm 0H.mm mmH oumEHHo no nonumoz meEMm Hobos .oz meEmm Hmuoa .oz mamfimm Hmuoe .oz commom mo ucmouom mo unmouom mo unmouom Hmuoe usmuuomEH umoz usmuuomEH umoz osooow meannessuoz saunas: .ma «Home 161 their most important influences,not surprisingly, no migrants listed job or business considerations. Also, when they were subsequently asked to indicate the specific importances of these same items, 99 percent replied that they were of no importance. The only motivation which is even remotely economic, is also only incidentally related to the economic forces mentioned in the literature. Specifically, nine respondents elected to move to St. Petersburg because they perceived it as a more economical place to live. Although a part of the savings was attributed to lower taxes, the consistently moderate climate was also favored because of the associated savings in heating and clothing COStS . 47 Indirectly, therefore, the economic influences furnish a degree of credence to the suspected pull of natural amenities. As Table 18 indicates, this support is hardly necessary. The most attractive amenity, climate, was mentioned as the primary reason for moving by over half of the retirees (55 percent). When those who listed this 47This latter observation requires two explanatory comments. First, tax benefits were an important consideration for more than nine retirees. A total of 55 respondents (23 percent) admitted that taxes were significant in their own decision to migrate. The SMSA origins suggest that a sizeable portion of this group was fleeing from the high taxes that characterize the urban Northeast. A supplemental attraction was undoubtably the $5,000 homestead exemption that all retired Florida homeowners are eligible to receive. The second comment concerns the nature of the St. Petersburg climate. A recent municipal publication describes the climate in this manner: "the surrounding waters . . . tend to temper the cold of winter and the heat of summer, resulting in a very mild year-round climate, with enough difference between summer and winter to afford variety and interest" (Land Use Plan, 1970, 4). The same description continues by noting that the average yearly temperature for the 48-year period between 1921 and 1968 was 73.9 degrees. The January and July means were 63.3 degrees and 82.7 degrees respectively. 162 factor as the second most important reason are included, the proportion rises to a full two-thirds. But even these impressive statistics understate the total impact of St. Petersburg's climate. To illustrate, the direct question concerning this element produced an affirmative response from 94 percent of the total sample. Thus, with near unanimous agreement, the members of this migration stream disavowed the importance of job and business considerations, and emphasized the significance of a pleasurable climate. On the basis of the existing theoretical framework, this attitude is an anomaly. Among the elderly in St. Petersburg, however, it is simply representative of the powerful attractions created by an entire set of amenities. Although not nearly as evident as the climatic fac- tor, several people also claimed that the recreational facilities in the city constituted a major appeal. Fishing and other water sports were mentioned, and one individual even indicated that he came to St. Petersburg because of the keen shuffleboard competition. Another group of 16 retirees recalled that their primary moti- vation was the overall appeal of the city. According to their percep- tions, St. Petersburg was simply a "nice place" for retirees to live. Also, notwithstanding the fact that no one specifically mentioned them as primary considerations, it is interesting that 23 percent of the sample considered the large number of retirees, and the accompanying facilities as important elements in their decision-making process. The facilities constitute an amenity, but can the same be said for Foamaraderie"? Finally, even those who stressed the significance of health motivations can be interpreted as having reacted to the pull of 163 amenities. In particular, the 62 respondents in this category, which represent 25 percent of the total sample, were usually attracted by the warm St. Petersburg climate. As Blanchard has suggested: "northerners suffering from [Er susceptible té] arteriosclerosis, diabetes, chronic nephritis, heart troubles, and a number of other chronic ail- ments, as many old people are, would benefit from permanent or seasonal migration southward" (Blanchard, 1956, 56). The American Medical Association, in 1885, was even more specific when it designated the St. Petersburg area as the world's healthiest place to live. Thus, a total of 217 migrants were primarily drawn to St. Peters- burg by either the climate (125), the recreational facilities (4), health considerations (62), or just because they perceived it as a nice place to live (16). This represents 88.6 percent of the total sample and it is a clear indication of the salient impact of amenity motivations in aged migration. The remaining motivations are also important to note. They in- clude the group which perceived St. Petersburg as an economical place to live (4); those who were attracted to the city by friends or rela- tives (12); those who were primarily "pushed" to St. Petersburg (8); and those who migrated in response to miscellaneous considerations (4).48 The friends and relatives group is surprisingly small. According to Ullman, "as more people settle in pleasant areas, they themselves will exert an agglomerative pull, bringing in still more newcomers" (Ullman, 1954, 128). Either Ullman's hypothesis is erroneous or the 48This latter group includes one person who came to St. Petersburg looking for a wife; another who was attracted by the city's reputation; and a third who appreciated the flat terrain. The other person ended up in St. Petersburg because he made a wrong turn while driving to Tampa. 164 current under-representation was a product of still another factor. For instance, it seems quite likely that the nature of the current topic may be responsible for a deceptive set of responses. Because we are concerned only with the single most important influence, it is conceivable that friends and/or relatives may have simply been less important than some other motivation, i.e., climate. The fact that this latter suspicion is supported by a second set of responses is significant. To illustrate, 21.6 percent agreed that the persuasion of friends was an important component in their decision to move to St. Petersburg; 28.6 percent made a similar declaration concerning the impact of relatives. Also, in recognition of the fact that the persuasive persons may actually reside elsewhere, the respon- dents were asked if they migrated to the city in order to be near friends and/or relatives. The proportions fell to 9.4 percent for friends and 15.9 percent for relatives. But the existence of an agglomerative pull is still very apparent. This camaraderie concept will receive addi- tional attention in a subsequent portion of this chapter. Up to this point, the various motivations have been essentially the same. Each has represented an attractive feature of the St. Peters- burg environment. The final category is in contrast to this situation because it includes those migrants who were reacting to the unfavorable conditions at their points of origin. Even though there are only eight households in this division, it would be improper to conclude that push factors were unimportant in the current migration stream. A retiree who migrated for climatic reasons could have just as easily been responding to an undesirable climate at the point of origin as to the attractive climate in St. Petersburg. And in fact, most 165 migrants probably considered the conditions at both sites before they decided to move. With this understanding, an effort was made to evaluate the role of push factors in retired migration. The retirees were asked if there was anything about their previous residence that encouraged them to move, and 51 percent answered affirmatively.49 When asked to specify these push factors, the migrants listed several considerations (Table 19). Two of these are especially notable. Almost two-thirds of those supplying answers indicated that they were encouraged to move by the climate at their points of origin. Usually the specific impact of either low temperatures, snow, or a damp environment were mentioned. Another sizeable group of retirees (25 percent) fled from the problems associated with central city deterioration. In their opinion, the pre-retirement residence was rendered undesirable by high crime rates, pollution, congestion, and the expansion of black neighborhoods. An intriguing question emerges from this discussion of migrant motivations. In view of the fact that many of the desired qualities were also available in other communities, and in fact, in other states, why did the migrants select St. Petersburg? The survey responses show that just over 40 percent of the sample did consider at least one other destination. And of the 102 people in this category, 75 thought of settling elsewhere in Florida. The remaining options range from other 9Conversely, half of the sample was ngt_pushed away from the points of origin. Because these people still migrated, however, it is apparent that the St. Petersburg attractions were very strong. With nothing to encourage the retiree to leave, the barriers of inertia would naturally be very immense. Migration could only occur, therefore, if the inertia were neutralized by a sufficient level of attraction. 166 Table 19. Push Factors as a Component of the Retired Migration Process Absolute % of Those % of Total Factor Frequency Responding Sample Weather-climate 79 64.8 32.2 Central city deterioration 30 24.6 12.2 Too expensive 8 6.6 3.3 No place for retirees to live 1 .8 .4 Lonely, too far from friends 2 1.6 .8 Forced to move 2 1.6 .8 Push factors were not important (no response) 123 -- 50.2 167 popular retirement areas such as California (15), Arizona (14), and Texas (7) to places like the Bahamas (1), Colorado (1), Washington (2), and New Mexico (1). Although the survey information will not permit us to test their importance, several plausible explanations are available. For instance, given the distribution of migrant origins, it is clear that a Florida destination may have been selected because it was closer than the other popular retirement states. This possibility was discussed in a previous section of this chapter. When faced with a choice of sites within Florida, the traditional reputation of St. Petersburg may have been a significant variable. Also, in view of the fact that most people favor known circumstances, it is conceivable that St. Petersburg was selected because the migrants were familiar with the city's numerous attributes. The significance and source of this pre-migration famil- iarity form the essential framework of the subsequent discussion. The Degree and Source 9£_Pre-Migration Familiarity_with St. Petersburg Nearly 85 percent of the retirees indicated that they were familiar with St. Petersburg before they migrated there. Unfortunately, because this statistic is based upon personal judgments, it is very difficult to interpret. The standards were left up to each individual's own discretion, and hence, the avowed levels of familiarity undoubtedly ranged from a passing interest, to an intense and intimate understanding of the city. In a similar manner, it seems likely that this perceived knowledge had an irregular impact in the decision-making process. Whereas one migrant may have strongly emphasized his acquaintance during the selection of St. Petersburg, another may have totally ignored 168 his own familiarity in favor of some other set of circumstances. In this respect, the latter group is similar to the 15 percent who claimed no prior knowledge of the city. Together these two segments offer clear proof that familiarity is not an essential prerequisite in aged migration. As a logical consequence, one might also assume that familiarity is completely unimportant. But such a conclusion would be false. We can be reasonably sure that a large majority of the migrants were influenced in some way by their prior knowledge of the city. Thus, as a component of the aged migration process, this phenomenon warrants additional analysis. We can begin by focusing on the primary sources of the migrants' information. By their own admission, most of the retirees' information was acquired through visits to the St. Petersburg area. Ninety-seven percent of those with prior knowledge indicated that they had made at least one trip to the city before they decided to move there. And of these, 96 percent indicated that the visits were their most important sources of information. Although notable in themselves, these statistics also have several interesting implications. As an example, the tradi- tional reputation of St. Petersburg as a focal point for retirees may be connected to its role as a tourist center. Once the city had developed both functions, the retirement attractions were continually displayed to future retirees. This was partly because a combination of the spreading practice of paid vacations, and the increased mobility of the American population, meant that more and more Americans were able to visit the amenable regions of the country (Ullman, 1954). As they travelled in Florida, many vacationers 10‘) were attracted to St. Petersburg by its tourist facilities. Once in the city, these people were introduced to the pervading "retirement environment." In view of the significance of personal visits as a source of familiarity, and also of the apparent importance of familiarity in the selection of a retirement home, it seems safe to assume that an unknown portion of these vacationers later returned to the city as retired residents. On this basis, therefore, we can project a continua- tion of the St. Petersburg retirement function. Additional support for this latter prediction can be obtained from the fact that the second most important source of information was contacts with friends and relatives. Only 6 percent of those with prior knowledge viewed these contacts as their most important source, but this is largely due to the overwhelming importance of personal visits. The actual contribution from friends and relatives was influ- ential enough to warrant additional examination. As Table 20 indicates, most of the respondents had a friend and/ or relative residing in St. Petersburg before they moved there. Where- as only 27 percent arrived without any resident acquaintances, exactly half of the migrants had at least one friend and just over one-third had at least one relative. These figures are interesting, but from our current objectives they are also only minimally important. It is more pertinent to ask if these residents were influential in the re- spondents' decision to select St. Petersburg as a post-retirement home. And thus, it is with this understanding that Table 21 was compiled for analysis. Of the 178 migrants who had a friend and/or relative living in the city, 95 (53.4 percent) admitted that they were influenced by these 170 Table 20. The Presence of Friends and/or Relatives in St. Petersburg, Prior to the Move Friends Not Present Present Total R Present 22 5 12 2 34 7 e ° ( 55) ' ( 30) ° ( 85) l a Not : present 27.3( 67) 38.0( 93) 65'3(16O) v e 3 Total 49.8(122) 50.2(123) 100.0(245) Table 21. The Influence of Resident Friends and/or Relatives on the Decision to Move to St. Petersburg Friends Not No Resident Influential Influential Friends Total : Influential 5.3(13) .8( 2) 14.7( 36) 20.8( 51) : Not t influential 4.5(11) .8( 2) 7.8( 19) 13.1( 32) i v No resident : relatives 21.6(53) 17.2(42) 27.3( 67) 66'1(162) Total 31.4(77) 18'8(46) 49.8(122) 100.0(245) 171 acquaintances. Because this represents 39 percent of the total sample, it is clear that personal contacts were much more important than the preceding discussion has indicated. The precise nature of the influence is uncertain but it seems logical that it ranged from a simple exchange of information to the point where some subjects were actively encouraged to migrate to the city. Partly because of the established importance of personal visits, the impact of resident friends and relatives may even be greater. In particular, just as we reasoned that several retirees were attracted to St. Petersburg as tourists, so can we conclude that others were attracted to St. Petersburg by their personal acquaintances. They may fail to perceive the relationship, but several retirees were undoubtably introduced to the city during a social visit. Indirectly, therefore, the presence of friends and relatives was a source of information. And this is especially notable because it offers further support for the earlier projection concerning the continued growth of St. Peters- burg's retirement function. As long as friends and/or relatives have both a direct and indirect influence, aged migrants will continue to be attracted to the city. Before concluding this discussion of the migrants' important sources of information, we should also emphasize those sources that were unimportant. Specifically, radio, television, and all forms of printed materials (including magazines, newspapers, and mailed ad- vertisements) were very rarely mentioned. In fact, just 16 people spoke of one or the other, and of these, 10 noted the importance of newspapers. The available literature indicates that a similar pattern of response has appeared in several previous studies (Eteng and Marshall, 172 1970; Smith and Marshall, 1970). For instance, after examining the replies of their Florida and Arizona samples, Eteng and Marshall con- cluded that: little use was made of magazine and newspaper articles, promotional literature supposedly put out by developers and realtors in the retirement community, books, and other reading materials from state agencies and chambers of commerce in the retirement community. At least 85 percent of all the retirees in both samples failed to obtain retirement information from any of these sources (Eteng and Marshall, 1970, 29). Needless to say, realtors, promoters, and chambers of commerce should be intrigued by these discoveries. Migration Planning If we assume that a retiree has reached the point where, on the basis of his personal movtivations, and perhaps his familiarity, he has decided to migrate to St. Petersburg, an obvious question arises concerning the extent to which he plans his actual move. Of course, it is entirely possible that the retiree could have developed some sort of preliminary plans before he even selected a precise destination. As an example, many people anticipate their later years by trying to accumulate a "retirement nest-egg." Activities of this nature are notable, but they are also beyond our current concern. Instead, we are interested in those plans which specifically prepared the retiree for his eventual move to St. Petersburg. To a large degree, the extent of these plans should be closely related to the amount of elapsed time. A retiree who moves almost immediately after his decision can not engage in a great deal of plan- ning. But one who waits a substantial period is in a favorable position to consider the various alternatives, and hence,to develop a functional 173 migration plan. Unfortunately, the survey responses shed very little light on this subject. This is largely because there is no specific information available concerning the elapsed time between the selection of St. Petersburg and the subsequent move there. The only indication as to when the decision to move was made is with respect to the act of retirement. The subjects were asked if the decision came more than five years before they retired (14 percent answered yes); less than five years before they retired (23 percent); less than five years EEEEE they retired (54 percent); or more than five years a££§£_ they retired (9 percent). The fact that a sizeable majority made their decision after they had left the labor force is particularly notable if we recall an earlier observation. Specifically, in the discussion of Figure 15, we noted that most of the retirees had moved to St. Petersburg either immediately after they retired or after a comparatively short hiatus. The precise statistics showed that the mean elapsed time was approximately two years. Forty-eight percent moved within a year, and 77 percent moved with three years. In com- bination, these figures show that although nearly half of the respon- dents moved within a year of retirement, nearly two-thirds of the sample did not even decide to move until after they had already withdrawn from the labor force. It seems clear, therefore, that the time available for planning was very short for most of the retirees. This fact is reflected in the responses to the two specific ques- tions concerning the extent of pre-migration planning. First, the retirees were asked if they owned any property in St. Petersburg before they actually moved there. Just 34 (14 percent) answered affirmatively, and of these, 22 owned their property for less than a year. It may 174 be that this low occurrence is related to the nature of the com- munity. As a large city with a moderately long history, St. Petersburg differs from many of the more recent attempts to create planned retire- ment communities. The latter usually begin with a single company and a large section of undeveloped real estate. The promoter sells plots of land, usually on an installment plan, and frequently there is very little pressure to assume immediate occupancy. Some communities will even differentiate between the sections that are scheduled to be de- veloped at once, and those that are to function as areas for future expansion. Hence, a person who is planning for his imminent role as a retiree can purchase a portion of land and simply hold on to it until he is ready to put it into use. In the meantime, his costs are minimal. Until the land is completely paid for, the taxes are usually assumed by the developer. Even those who have finished buying their land have fairly low taxes because of the undeveloped nature of the land. The situation in St. Petersburg is very different. Overall, a person hoping to buy for the future is severely restricted. Vacant land is rare; and in view of the complex forces involved, it is dif- ficult to anticipate the future situation. Potentially promising areas are subject to the competitive forces of a market system and this can make the initial costs prohibitive. If the purchase is completed, competition can add to the cost of simply holding the land in reserve. Also, the urban taxes will frequently be much higher. Under these con- ditions, it is usually perceived as a wise decision to wait until just before moving to purchase property. At that time the relevant variables can be evaluated instead of anticipated. This may be part of the reason 175 why so many of those who bought property before they migrated, did so only a year in advance. And in a similar vein, it may also be a reason why only 14 percent even bothered to make a preparatory purchase. The second attempt to measure pre-migration planning concerned the extent to which the retirees had made previous arrangements for a place to live. Obviously, this indicator is related to its predecessor. The retirees who owned property in the city before they moved there, undoubtedly knew where they were going to reside. The current measure is more encompassing, however, because it also recognizes that not everyone wanted to move directly into a privately owned home. Those who wished to begin in rented facilities, and who also made arrangements before moving, also qualify as planners in this category. Since only one-fourth (27 percent) of the sample had prior know- ledge, the previous contention that only a few of the retirees actually engaged in an extensive pre-planning process is further supported. Ad- ditionally, it is interesting that 86 percent of those who had made arrangements did so during a personal visit. This reinforces the previously established significance of those visits. Not only did the migrants receive most of their information in this manner, but several also used them as vehicles for accomplishing their pre—migration plan- ning. If we are to gain a complete understanding of the aged migration process, these preliminary visits need to be analyzed in much greater detail. CHARACTERISTICS OF THE MIGRATION ACT Once the retiree has completed his preparations, which we have found to be varied both in their nature and their extent, he is ready to 176 commence the actual move. This final section will examine several of the retirees' movement characteristics. Particular attention will be focused upon a question which was originally posed in Chapter II: Do the principles of stage or step—wise migration apply in retirement migration? Another topic of major concern will be those retirees who engage in seasonal movements. They deserve special attention because their semiannual flows represent the epitome in amenity-induced migra- tion. The Presence g£_Stage Migration In the previous discussion of stage migration, we noted that the basic concept states that rural inhabitants move individually toward large cities via gradual moves through smaller centers. Empirical in- vestigations have shown that the theory performs best when it is applied in a situation with an obvious national primate city. Under a strict interpretation, therefore, it is unrealistic to expect the current sam- ple to support this hypothesis. Not only are United States' migration streams much too complex, but there is also no clear-cut national primate city, and as we have seen, very few migrants originated from a strict rural setting. As a simple revision, the original tenets can be relaxed to indi- cate that long distance migrants tend to move to their ultimate des- tinations via a series of intermediate steps. In our earlier discussion we indicated that even this thesis is probably inappropriate in retired migration. At such an advanced stage in the life cycle, senior citizens can ill—afford a series of residential shifts. Instead, we hypothesized 177 that amenity—seeking retirees will generally move directly to their ultimate destinations. The empirical evidence supports that proposition. Eighty-six percent of the migrants came to St. Petersburg straight from their pre- retirement home. Of the 34 retirees (14 percent) who made at least one intermediate stop, 16 first went to another community in Florida. This suggests that these migrants were still attracted to the Florida ameni- ties, but it also raises several questions concerning the decision to switch to St. Petersburg. With job and business considerations elim- inated, and with climatic considerations virtually neutralized, what other factors led to the eventual intrastate shift? Unfortunately, this small sample is somewhat biased because it only includes current residents of St. Petersburg. There must also be other migrants who after moving to the city, decided to leave for another residence in another Florida community. This hinders any attempt to estimate the magnitude of these intrastate shifts, and in view of the information available from the survey, we also can not come ment on either the migrant motivations or any of the other relevant factors involved in this subsequent migration process. More attention is clearly needed in this area. Seven of the migrants made their initial move to a different location within their "home" state, and only 11 came to St. Petersburg via an intermediate stop in another state. This latter group, which constitutes just 4.5 percent of the total sample, deserves further inspection because if any of the migrants followed a stage process, they probably fall into this category. With this understanding, it is important to re-emphasize the group's small dimensions. Also, 178 many of the combinations are difficult to explain. One retiree, for instance, traveled from Delaware to Florida via Michigan. Another reached Florida after first traveling from New Jersey to New York. Four of the combinations are intriguing because they involve intervening stops in states which also have reputations as retirement centers. Two retirees, one from Illinois and one from Minnesota, traveled through Texas; a second person from Minnesota went to Arizona; and a New York migrant traversed the nation when he moved to California before switching to Florida. In many respects, these people are similar to those who moved to St. Petersburg from another Florida community. The attraction of amenities would have been virtually neutralized, and yet the motivations were significant enough to permit migrations which covered several hundred, if not more than a thousand miles. Unfortunately, the survey responses shed little light on this problem. Climatic considerations were the most important motivation in two cases, and health and recreational facilities were crucial in the re- maining instances. As unlikely as it seems, could it be that the mi- grants perceived and reacted to differences in amenities? Although questions of this nature deserve future attention, we should not lose sight of the fact that the overwhelming majority of the migrants did 225_move to St. Petersburg via stages. Most assuredly, one major reason was age. The retirees were simply too old to waste their time reaching the St. Petersburg area. Direct movements imme- diately after retirement are temporally most efficient, and as we have seen, they were also the most popular approach. Another preventive agent may have been the financial implications of step-wise migration. Coming primarily from modest backgrounds, few of the migrants could 179 afford the unnecessary luxury of intermediate stops. And indeed, this thought gives rise to what may be the most important explanation of all. Specifically, most of the retirees were familiar with St. Peters- burg, and in fact, many had made personal visits. This meant that the city was a known entity with known attributes. And since the retirees were migrating in response to those characteristics, why should they have stOpped along the way in some other, less attractive place? Before we conclude this discussion of stage migration we should check to see if a similar phenomenon occurs during the migrants' search for a permanent residence in the city. The preceding evidence suggests that it does. Specifically, major support can be derived from the fact that only 66 migrants, representing just 27 percent of the total sample, arrived in the city with a knowledge of where they were going to reside. Conversely, 73 percent had no idea. It seems logical, therefore, to assume that this latter group commenced a residential search immediately upon their arrival. Those with good fortune undoubtably found a suitable home in a very short time. But those with less luck were forced to select temporary quarters while their search continued. During this stage, the central part of the city, with its numerous retirement hotels, apartments, boarding houses, etc., must have been an attractive location. As permanent sites were secured, the migrants advanced to a different stage in the residential process. Theoretically, subsequent shifts could also be placed in this context if they represented additional . . . 50 "stages" in the retirees' search behaVior. 0 . . . . . This Situation rarely arose in reality because only 12 percent of the total sample moved more than once. 180 If the previous reasoning is correct, it should be reflected in the survey responses. There is an expected relationship between the level of pre-migration planning, as represented by prior knowledge of a place to live, and the degree to which the migrants have changed their residence since arriving in the city. This hypothesis can be tested with the information presented in Table 22. Given the previous argu- ments concerning the use of temporary quarters, we would expect that a large majority of those who changed their residence would be migrants who came to the city with no prior residential plans. And indeed, this position is supported by the data. Additionally, we would expect that those with prior arrangements would have little need for temporary lodging, and hence, that they would display low levels of intraurban mobility. This position is also supported. Perhaps the strongest indication that the hypothesis may be in- valid, however, is the fact that the largest group of retirees, totaling 104 people, came with no prior knowledge, and yet they also have never changed their residence. Either these people were fortunate enough to find a permanent residence immediately, or their search period was so short that they never really perceived their intermediate loca- tion as a residence. In either case, a stage migration process was never adopted. The suSpicions raised by this discovery are confirmed when the chi-square statistic is computed. There is no relationship between pre-migration planning and post-migration mobility. Although several retirees relied upon temporary lodging, in the context of the entire sample they are too insignificant to support the postulated existence of a widespread stage process. 181 Table 22. The Relationship between Pre-Migration Planning and Post- Migration Mobility "Before you moved to St. Pete, did you know where you were going to live within the city?" Yes No Total Have Y°u Yes 18 74 92 ever Changed No 48 104 152 your residence in 9" *Insufficient information was available from one respondent. Chi square = 2.44 with l d.f. Result: Not significant at .05 level. 182 The Presence 2£_Seasonal Migration For one-fifth of the sample, migration is a seasonal process. Although the precise proportion differs for each individual, almost with- out exception this group spends part of the year, and especially the winter months, in St. Petersburg, and the remaining portion, including the summer months, in another community frequently located in their "home" state. Clearly, a major stimulus for these movements is seasonal changes in the weather. The participants are able to take advantage of the best and avoid the worst of two different climatic systems. For this reason, therefore, these semiannual flows represent the epitome in amenity-induced migration. During an earlier survey in Florida, Honnen, Eteng, and Marshall recognized the need for a careful study of this special category of retired migrants. Their related comments included an intriguing series of questions. Do they migrate every winter to warm climates? Do they choose the same community every time they migrate and why? Do they maintain two or more residences in view of the seasonal nature of their migrating characteristics, and also in view of their generally higher standard of living? To what extent are they socioeconomically un- productive vis-a-vis other migrant and non-migrant re- tirees? In what other ways are they distinct from their retired counterparts? (Honnen, Eteng, and Marshall, 1969, 52). These and other related questions concerning the seasonal migrants' attributes, and the various components of their migration process, deserve serious attention. Yet in view of (1) the current objectives; (2) the available information; and most of all, (3) the magnitude and significance of this task, it would be a serious mistake to try to supply the answers in this analysis. 183 The current group of seasonal migrants is most assuredly not representative. In our discussion of the operational definitions we noted that retirees were not included in the sampling frame unless they satisfied a six months per year residence requirement. This criterion alone precludes an unbiased sample because, as the preceding questions imply, seasonal migrants may limit their stay to as little as two to three months. Additional bias can be traced to the use of the city directory as a sampling frame. Those with a short tenure probably would not invest in a "permanent" residence in St. Petersburg. Financially, it would be very inefficient to purchase and maintain quarters if they were only going to be utilized a maximum of four to five months each year. Also, a "permanent" dwelling would tend to prevent the migrants from diversifying their winter destinations. The logical option, therefore, is to stay in the temporary facilities offered by apartments, retirement hotels, boarding houses, and motels. Unfortunately, the transient residents in these areas do not appear in the city directory. With this knowledge, one might wonder how 49 seasonal retirees managed to be included in the current sample. Table 23 furnishes a partial explanation. The part-time residents display a unique and very significant preference for mobile homes. Because they live in the city at least six months each year, this group of migrants is willing to make a firmer residential commitment than most of the short tenure migrants; many are still either reluctant, or unable, to go so far as to purchase a single family home, however, and mobile homes are an attractive alternative. They require a relatively small investment, but offer permanent lodging. 184 Table 23. The Relationship Between the Type of Residence and the Length of Tenure Length of Tenure Type of Seasonal Permanent Residence Migrants Residents Total Mobile Home 39 44 83 Other 10 152 162 Total 49 196 245 Chi square = 56.34 with l d.f. Result: Significant at .001 level. 185 The need for a more intensive analysis of seasonal migrants is reaffirmed by their unique residential status. If retirement communities are to anticipate and plan for this special group of retirees, they must identify not only their housing preferences, but also the other kinds of facilities they require. Usually these demands will magnify those of the permanent residents, but on occasion, they may also require special services. Their specific impact can be established only if the communities also determine (1) the magnitude of their seasonal influx, and (2) the characteristic features of the retirees and their periodic migration processes. Subsequent research is suggested. Prospects for Future Mobility Technically, the migration process does not terminate until the household is "permanently" settled in a different residential site. Is it possible, therefore, 1x) measure the end of retired migration? In other words, can we ever definitely state than an individual retiree has completed his or her migration process? Until that retiree dies, probably not.51 Longitudinal studies offer a temporary solution to this dilemma; but if they are terminated while the retiree is still alive, the pos- sibility will still exist that the migration process could be resumed. In another approach, we can comment upon the prospects for future mobility by relying upon each retiree's self-professed plans. This method is weakened somewhat by the fact that a person's plans are not 51And even then it can be argued that some retirees have a final dimension added to their migration process when their bodies are trans- ported elsewhere for burial. It would be interesting to examine the spatial manifestations of this "post-mortem migration." 186 always synonymous with his subsequent behavior (Goldscheider, 1966a). But as an indication of the possibilities that exist at a single point in time, it is a valuable procedure. For instance, when asked if they were thinking of moving again, 86 percent of the current sample replied that they were not. Among the seasonal migrants this response represented an intention to maintain the same periodic destinations. For the rest of the migrants the planned stability was really an ex- tension of their St. Petersburg residential behavior. Only 12 percent of the migrants had moved more than once since reaching the city, and a similarly small proportion were even contemplating any future moves. In this latter category, just 15 respondents were thinking of leaving St. Petersburg. The indication is clear that for most of the retirees, the migration process was destined to end in St. Petersburg. CHAPTER VII SUMMARY AND CONCLUSIONS In their quest for an understanding of the patterns and processes that are associated with spatial behavior, geographers among other social scientists have virtually ignored the elderly. This omission is unfortunate because the elderly do not always conform to normative behavioral principles. Generally, they are more susceptible to more ailments, which when contracted, tend to be more serious, and to last longer. When these health problems produce either temporary or permanent disability, or when they cause bed confinement, the aged person's normal movement patterns are clearly affected. Other senior citizens experience de- clines in their mobility because they lose or surrender their right and/or ability to drive. Widespread poverty is another limiting factor because, among other considerations, it frequently restricts the older person's residential mobility. A retiree may want to move to another part of the city, but in view of his financial status, he may no longer possess the ability to join in the normal market processes that char- acterize urban residential movements. The end result is a concentration 187 ( lltl-IIII... 188 of the elderly in the older, and frequently poorer, sections of most large United States Cities. To a large extent these financial restrictions are the product of the drastic income reductions which characteristically follow re- tirement. And indeed, many of the other changes that accompany re- tirement also have spatial implications. The very decision to with- draw from the active labor force is, in effect, a decision to also withdraw from one of the most significant regular movements of people in the United States today--the journey to work. In its place the retiree frequently substitutes irregular social trips, mid-day shopping excursions, visits with the doctor, and perhaps even an occasional trip to the senior citizen center. Hence, his new behavior patterns have both spatial and temporal implications. This study has focused upon several aspects of still another retirement-related phenomenon, namely, retired migration. In contrast to the preceding developments, the participants in this movement con- stitute a small but increasing proportion of the total aged population. This numerical minority is somewhat deceptive, however, because it fails to portray the true significance of post-retirement migration. Consider, for instance, the fact that the major destinations are concentrated in just three or four states. One of these, Florida, experienced a 78 percent increase in its elderly population between 1960 and 1970. Although part of this net gain of 435,000 elderly people was produced by normal aging processes, the major portion clearly came to Florida 2In one of the rare and very recent examples of geographic re- search in this area, S. Golant (1972) has examined the residential loca- tion and spatial behavior of the elderly in Toronto, Canada. 189 through interstate migration. Such a rapid increase has obvious im- plications for the state, and especially for the individual communities. Also, additional significance arises because there are strong indica- tions that the retired movements may be representative of a new form of general migration (Brunn, 1972). This suggests that aged migration differs from that of the general population. And on the basis of that hypothesis, this study was initi- ated to (1) characterize various aspects of the aged migration process; and (2) identify and subsequently suggest the necessary revisions for, any theoretical anomalies which may be associated with this new topic in migration research. Given the lack of prior geographic investigations in this area, the research was frequently as heuristic as it was definitive. Therefore, as we review the more salient findings, we will also identify some of the more intriguing research possibilities. Most of the information was obtained through a personal interview survey of 245 retired migrants living in St. Petersburg during the early months of 1972. BACKGROUND CHARACTERISTICS Race Although 7.8 percent of the total aged population in the United States is black, all of the survey respondents were white. This un- expected homogeneity is undoubtably at least a partial product of the combined effects of residential segregation in St. Petersburg and clus- ter sampling. But it also suggests that either retired blacks favor other communities, or they participate in post-retirement migration in 190 much lower proportions. Before the impact of these components can be evaluated, however, a study of black retirees must be completed. Sex and Marital Status There are 78 males for every 100 females in St. Petersburg. Due to the longer life expectancy of women, the sex ratio falls to 62 for the elderly. Among the survey reSpondents, however, the ratio was 111. This large discrepancy was partly the result of males dominating as the respondents for most family units. But when both members of a married couple were included, the ratio still only fell to 88. Clearly, there were other factors responsible for the under- representation of women. One of the first suspicions is that marital status has an impact upon retired migration. Perhaps the propensity for married couples to migrate to St. Petersburg is greater than that of widows and other single females. And since many of these women become widows after they reach the city, a question arises concerning their subsequent mobility rates. egg Two-thirds of the respondents retired between the ages of 60 and 69; and another one-fourth retired before they were 60. If pension systems and other retirement programs continue to improve, the latter group of "early" retirees should increase, and concomitantly, this should lead to an increase in the number of people free to move to retirement communities. The lack of respondents at the other end of the age continuum suggests an inverse relationship between the age at retirement and the propensity to migrate. However, to test this hypo- thesis, the characteristics of non-migrants must be identified. 191 The mean elapsed time between retirement and migration was just two years: in fact, 48 percent of those responding left for St. Peters- burg within a year. This suggests that many of the subjects had at least considered their move before they retired. Income Before and After Retirement Post-retirement migration is not restricted to the wealthy. Nearly half of the sample refused to divulge this information, but those who did frequently came from rather modest backgrounds. Fifty percent had pre-retirement incomes of $10,000 or less, and only 3 percent earned at least $20,000. It may be that the retirees' financial standings have spatial manisfestations. Wealthy migrants may favor a different type or class of community than the more modest migrants. Also, some communities may indirectly, and in some instances, even purposefully, discriminate against a certain class of retirees by establishing restrictive ad- mission requirements. These hypotheses need to be tested. All the migrants had experienced a drastic reduction in their earning capacities since they retired. Pre-Retirement Mobility The concept of chronicity in migration was not supported. In fact, in the 30 years before they retired, most of the migrants were comparatively stable; 16 percent had never moved, and another 56 percent had moved three times or less. This meant that most of the sample was forced to overcome immense inertia barriers. The motivations that led to the decision to move must have been extremely powerful. 192 MIGRANT ORIGINS The Distribution 2£_Origins The first indication that retired migration is theoretically anomalous came from the distribution of the migrant origins. Instead of an inverse relationship between the migrant flows and the distance from St. Petersburg, the patterns display a direct association. The areas immediately adjacent to St. Petersburg contributed very few migrants; those further to the north, and especially those in the Northeast quadrant of the country, supplied 87 percent of the national sample. Evidently the pull of amenities was a more important considera- tion in the selection of a post-retirement home than the friction created by the intervening space. The more distant states in the North pre- dominated as origins because the perceived amenity differential was greater. Once the decision to move to an amenity area was made, the friction of distance appears to have reasserted itself. This is evidenced by the fact that the area west of the Mississippi River contributed virtually no migrants. The logical conclusion is that when the retirees in this area decide to move to amenities, they choose those areas which are closest, namely, California and Arizona instead of Florida. To test this conclusion, and the related hypothesis that there is a directional bias in the interstate flow of Northeastern retirees, the origins of those in Arizona and California retirement communities need to be examined. 193 Size 9£_Community Seventy-eight percent of the total sample came from metropolitan backgrounds. In fact, three of the five largest SMSA's in the United States (Chicago, New York, and Detroit) contributed nearly one-fourth of the total sample. This discovery led to a question concerning the relationship between the population of a SMSA and the size of that community's contribution to the St. Petersburg migration stream. For metropolitan centers in the Northeast with a population of one million or more, these variables displayed a positive trend. Among smaller SMSA's, however, there was no discernible relationship. It may be that information flows among friends and relatives are a crucial explanation for the excessive contributions which originated in some of the smaller centers. And if this is true, it may also be that there are concentra- tions of friends and/or relatives in the same neighborhoods within St. Petersburg. These latter postulates could be important components in the retired migration process, and therefore, they deserve addi- tional consideration. MIGRANT PREPARATIONS Migrant Motivations Most of the retirees required strong motivations to overcome the forces of inertia created by the intervening space, their advanced age, their past residential stability, and their modest financial status. As the distribution of the migrant origins suggested, amenities supplied the major stimulus. Nine out of every ten retirees were drawn to St. Petersburg either by its climate, its recreational facilities, health considerations, or because they perceived it as a nice place for 194 retirees to live. The strongest pull was exerted by the climate. Ninety-four percent of the respondents agreed that climate was an im- portant consideration, and 55 percent identified it as the single most important motivation. In contrast, virtually all theories of migration place a similar emphasis upon economic elements. Instead of amenities, they argue that most migrants respond to job or business changes. This basic disagreement is extremely important because it identifies another inadequacy of the traditional theoretical structure. As other segments of the migrant population begin to place greater emphasis on amenities and other non-economic considerations, the need for theoretical re- visions will become even more imperative. Frequently the appeal of St. Petersburg was supplemented by a repulsion from the point of destination. Half of the sample admitted that push factors were important in their move, and among these, the more crucial considerations were an unfavorable climate and urban deterioration. The Degree and Source 2£.Pre-Migration Familiarity with St. Petersburg Since 15 percent of the sample were unfamiliar with St. Petersburg before they arrived as new residents, one can conclude that familiarity is not an essential prerequisite in retired migration. The most important source of information for the rest of the respondents was personal visits. Ninety-seven percent of those with a prior knowledge of the city had made at least one visit, and of these, 96 percent indicated that these trips were their most important source of information. This suggests a linkage between St. Petersburg's dual roles as a tourist attraction and a retirement center. Many 195 people arrived initially as tourists and returned later as retired residents. Consequently, as long as St. Petersburg remains a major tourist area, it should also continue to be a center of retirement, and hence, a major destination for aged migrants. Pre-Migration Planning The previous discovery that the mean elapsed time between retire- ment and migration was just two years, and that 48 percent of those responding had left for St. Petersburg within a year, suggested that many of the subjects engaged in some form of pre-migration planning. Surprisingly, therefore, 64 percent of the sample did not even decide to move to St. Petersburg until after they had retired. This combina- tion of statistics indicated that many migrants had very little time for planning. And in support of that view, the two direct indications of pre-migration planning in the survey failed to supply evidence of extensive preparations. Specifically, those retirees who had antici- pated their mobility by either purchasing property or making previous arrangements for a place to live were in a very small minority. It may be, however, that migrants to St. Petersburg are not representative in this respect. As an "open" city, St. Petersburg differs from the many planned retirement communities where parcels of land are intentionally sold for future occupancy. The extent and precise nature of these differences are unclear. CHARACTERISTICS OF THE MIGRATION ACT The Presence 2£_Stage Migration There was very little evidence of stage migration in the survey. Eighty-six percent of the retirees migrated directly to St. Petersburg 196 from their pre-retirement homes. Because half of those few who did "step" to the city came from other Florida communities, however, a question exists concerning their motivations. Specifically, with the same general amenities, and with job and business considerations eliminated, why would a retiree leave one Florida community for another? Could it be that they perceive and react to relatively small amenity differentials? Seasonal Migration An unknown proportion of the St. Petersburg population is comprised of seasonal migrants. They represent the epitome in amenity-induced migration because, by regularly shifting their residence, they are able to take advantage of the best, and avoid the worst of two different climatic regimes. The precise decision varies with each individual, but usually the cooler months are spent in St. Petersburg, and the warmer months are spent somewhere farther to the north. Since the survey respondents were required to reside in St. Petersburg for a minimum of six months each year, those seasonal migrants who restricted their stays to only the cooler months were not included. Even with this limitation, however, temporary residents still constituted one-fifth of the sample. Of course, the significance of this group of retirees extends beyond their mere presence. The many fluctuations created by their periodic movements have profound implications for both the origins and the destinations. If these communities are to successfully accomo- date these changes, many aspects need to be examined. In this vein, several promising research topics were identified. 197 Prospects for Future Mobility For the vast majority of the respondents, the migration process was destined to end in St. Petersburg. At the time of the survey, only 6 percent were considering a permanent move out of the city. As post-retirement migration continues to increase, the desir- ability, and eventually, the necessity of being able to identify its salient characteristics will become more and more apparent. If the favored destinations are to anticipate the various demands of these new residents, and hence, if they are to make proper plans to accomodate their arrival, the communities will require extensive information. They will need to know the personal characteristics of the retirees, their origins, and their migration processes. In this respect, there- fore, the research described in this thesis should be beneficial. But there are other areas of inquiry, some of which have already been identified, that will also require intensive analysis. New communi- ties, for instance, could benefit from the experiences of some of the more traditional destinations such as St. Petersburg. They could identify and examine the concomitant community adaptations. Specif- ically, what impact did the influx of large numbers of retirees have on St. Petersburg's tax base? Did these new residents require special facilities? How did they influence the various municipal services? The subsequent success of each community will require it to keep abreast of the activities and preferences of its elderly residents. As an example, given the heavy dependence of the elderly upon public transportation systems, urban planners will need information per- taining to the spatial behavior patterns of the aged. 198 Geographers, and other social scientists, have the potential to help supply most of this information. To date, however, their capaci- ties have been left virtually untapped. The significance of aged migra- tion in itself, and as a representation of changing trends in the migra- tion patterns of the general population, is such that these past atti- tudes and actions must be revised. The challenge exists. More research is imperative. BIBLIOGRAPHY BIBLIOGRAPHY Aldridge, Gordon J. "The Role of Older People in a Florida Retirement Community." Unpublished Ph.D. dissertation, University of Michigan, 1955. Barker, Michael B. California Retirement Communities. Berkeley: University of California Press (Center for Real Estate and Urban Economics), 1966. Barsby, Steve L. and Dennis Cox. "Public Sector Influence on Interstate Migration of the Elderly." Tuscon: University of Arizona Press, 1970 (Mimeographed). Beale, Calvin L. "Demographic and Social Considerations for United States Rural Economic Policy," American Journal g£_Agricultural Economics, 51 (May, 1969), 410-427. . "Natural Decrease of Population: The Current and Pro- spective Status of an Emergent American Phenomenon," Demography, 6 (May, 1969), 91-99. Berry, Brian J. L. and Frank E. Horton (eds.). Geographic Perspective 23 Urban Systems. Englewood Cliffs: Prentice-Hall, Inc., 1970. Blanchard, Fessenden 8. Where £9_Retire and How: A_Comprehensive Guide. New York: Dodd, Mead and Co., 1952. Blanco, Cicely. "The Determinants of Interstate Population Movements," Journal 2f Regional Science, 5 (1963), 77-84. Bogue, Donald. "Internal Migration," The Study 2£_Population, eds. P. M. Hauser and O. D. Duncan. Chicago: University of Chicago Press, 1959. . Principles g£_Demography. New York: John Wiley and Sons, Inc., 1969. Bowles, Gladys and James D. Tarver. Net Migration gf the Population, 1950-1960, by_Age, Sex, and Color. Vol. I, States, CountiesL Economic Areas, and Metropolitan Areas. Washington, D.C.: U.S. Government Printing Office, 1965. 199 200 Bright, M. L. and D. s. Thomas. "Interstate Migration and Intervening Opportunities," American Sociological Review, 6 (1941), 773-783. Brotman, Herman B. Income and Poverty in 1970: Advance Report. U.S. Department of Health, Education, and Welfare: Facts and Figgres 22_Older Americans, 1971a. . Measuring Adequacy 2E Income. U.S. Department of Health, Education, and Welfare: Facts and Figures 92_Older Americans, 1971b. . The Older Population Revisited: First Results 2f the 1970 Census. U.S. Department of Health, Education, and Welfare: Facts and Figures 22_Older Americans, 1971c. Brunn, Stanley D. "The New Urbanization: Emerging Patterns in the South and Southwest." East Lansing: Michigan State University, 1972 (Mimeographed). Bultena, Gordon L. and Vivian wood. "The American Retirement Community: Bane or Blessing," Journal g£_Gerontology, 24, 2 (1969). Burgess, Ernest W. (ed.) Aging i2_Western Societies. Chicago: University of Chicago Press, 1960. Burgess, Ernest W., G. C. Hoyt, and C. R. Manley. "The Construction of Scales for the Measurement of Migration after Retirement," Sociometry, l8, 4 (1955), 616-623. Burns, Robert K. "Economic Aspects of Aging and Retirement," American Journal 2£_Sociology, 59, 4 (1954), 384-390. Carrothers, Gerald A. P. "A Historical Review of the Gravity and Poten- tial Concepts of Human Interaction," Journal 2£_the American Institute 2: Planners, 22, 2 (1956), 94-102. Coulson, Michael R. C. "The Distribution of Population Age Structures in Kansas City," Annals RE the Association gf American Geog- raphers, 58 (1968), 155-176. Cox, Dennis and Steven Barsby. "Arizona and Retirement Migration," Arizona Review, 17, 11 (1968), ll-14. Dyer, Donald. "The Place of Origin of Florida's Population," Annals 2;. the Association 23 American Geographers, 42 (1952), 283-294. Eteng, William I. A. and Douglas G. Marshall. Retirement and Migration $2 the North Central States: A_Comparative Analysis: Wisconsin, Florida, Arizona. Population Series, No. 20. Madison: Univer- sity of Wisconsin Press, 1970. 201 Fisher, Jacob. "Trends in Interstate Migration Among the Aged," Social Security Bulletin, 11 (1948), 8-12. Friedsam, H. J. "Inter-Regional Migration of the Aged in the United States," Journal 2f Gerontology, 6 (1957), 237-242. Gilbert, E. Reid and Douglas G. Marshall. Retirement and Migration ig_ the North Central States: Plannipg for Retirement. Population Series,No. 21. Madison: University of Wisconsin Press, 1970. Golant, Stephen M. The Residential Location and Spatial Behavior 2f. the Elderly. The University of Chicago Department of Geography Research Papers, No. 143. Chicago: University of Chicago Press, 1972. Goldscheider, Calvin. "Differential Residential Mobility of the Older Population," Journal 9£_Gerontology, 21 (1966a), 103-108. . "Intrametropolitan Redistribution of the Older Popula- tion," Pacific Sociological Review, 9 (1966b), 79-84. Goodrich, Carter g£_§l, Migration and Economic Opportunity. Philadel- phia: University of Pennsylvania Press, 1936. ‘SLGould,P. R. "Structuring Information on Spacio-Temporal Preferences," Journal 2£_Regional Science, 7 (1967), 259-274. Granick, Samuel. "Adjustment of Older People in Two Florida Communi- ties," Journal g£_Gerontology, 7 (1952), 419-425. Greenwood, M. J. "An Analysis of the Determinants of Geographic Labor Mobility in the United States," Review 2£_Economics and Statistics, 51 (1969), 189-194. Greenwood, M. J. and P. J. Gormely. "A Comparison of the Determinants of White and NonWhite Interstate Migration," Demography, 8 (1971), 141-155. Harlan, William H. "Community Adaptation of the Presence of Aged Per- sons: St. Petersburg, Florida," American Journal 2£_Sociology, S9 (1954), 332-339. Havighurst, Robert J. "Problems of Sampling and Interviewing in Studies of Old People," Journal pf Gerontology, 5 (1950), 158-167. Haynes, Kingsley E. "Spatial Change in Urban Structure: Alternative Approaches to Ecological Dynamics," Comparative Factorial Ecology, Brian J. L. Berry (ed.), Economic Geography, 47 (1971-Supplement), 324-334. Helbock, Richard. "New Towns in the United States," The Professional Geographer, 20 (1968), 242-246. 202 Hermalin, Albert I. "The Effect of Changes in Mortality Rates on Population Growth and Age Distribution in the United States," Milbank Memorial Fund Quarterly, 44, 4 (1966), 451-469. Hill, Richard and Douglas G. Marshall. Retirement and Migration ip;E§g_ North Central States: Personal Adjustment 32 the Retirement Transition i2_Three Wisconsin Communities. Population Series, No. 15. Madison: University of Wisconsin Press, 1968. Hitt, Homer L. "America's Aged at Mid-Century: Number, Distribution, and Pattern of Change," in Living ig_the Later Years, T. Lynn Smith (ed.), Gainesville: University of Florida Press, 1952, 9-29. . "Migration Trends Among Older People." Paper read at the Sixth Annual Meeting of the Gerontology Society, Inc., 1953. . "The Role of Migration in Population Change Among the Aged," American Sociological Review, 19 (1954), 194-200. Honnen, James S., W. I. A.Eteng, and D. G. Marshall. Retirement and Migration i2_the North Central States: Comparative Socioeconomic Analysis--Wisconsin and Florida. Population Series, No. 19. Madison: University of Wisconsin Press, 1969. Hoyt, G. C. "The Life of the Retired in a Trailer Park," American Journal 2£_Sociology, 59 (1954), 361-370. Isard, Walter and Guy Freutel. "Regional and National Product Pro- jections and Their Interrelations," in Long Range Economic Pro- jection. Vol. 16, Studies i§_Income and Wealth. Princeton: Princeton University Press, 1954. Kaplan, R. L. "Factors Causing Migration Among the Elderly," Bulletin 2£_the Institute 2£_Gerontology, 17 (1970), 3-6. Kariel, Herbert. "Selected Factors Areally Associated with Population Growth Due to Net Migration," Annals g£_the Association 2£_Ameri- can Geographers, 53 (1963), 210-223. Lee, Everett S. "A Theory of Migration," Demography, 3 (1966), 47-57. Leslie, G. R. and A. H. Richardson. "Life Cycle, Career Pattern, and the Decision to Move," American Sociological Review 26 (1961), 894-902. Lianos, Theodore. "A Stocks and Flows Approach to Migration," American Journal g£_Agricultural Economics, 52 (1970), 442-443. Lloyd, R. G. "Social and Economic Circumstances of Retirement," Sociological Research, 40 (1956), 179-182. 203 Long, Larry H. "Migration Research and the 1970 Census," in Research and the 1970 Census, A. L. Ferris (ed.). Oak Ridge, Tenn.: Southern Regional Demographic Group, 1971, 121-130. Lowry, Ira 8. Migration and Metropolitan Growth. San Francisco: Chandler Publishing Company, 1966. McInnis, Marvin. "Age, Education, and Occupation Differentials in Interregional Migration: Some Evidence for Canada," Demography, 8 (1971), 195-204. McKain, Walter C. "The Social Participation of Old People in a California Retirement Community." Unpublished Ph.D. dissertation, Harvard University, 1947. Manley, Charles R, "The Migration of Older People," American Journal pf Sociology, 59 (1954), 324-331. Meier, R. L. "The Metropolis As a Transaction-Maximizing System," Daeadlus, Fall (1968), 1291-1313. Michelson, L. C. "The New Leisure Class," American Journal 2f Sociology, 59 (1954), 371-378. ’ Miller, David L. "Adaptations of Aged Persons to Relocation." Unpub- lished Ph.D. dissertation, University of Chicago, 1964. Monahan, William and D. G. Marshall. Retirement and Migration ig_£§g. North Central States: Retired Persons i2_Three Wisconsin Com- munities. Population Series, No. 14. Madison: University of Wisconsin Press, 1968. Morrison, Peter. "Chronic Movers and the Future Redistribution of Population: A Longitudinal Analysis," Demography, 8, 2 (1971), 171-184. ,\Moser, C. A. Survey_Methods pf Investigation. London: Heinemann Educational Books, Ltd., 1958. Muth, Richard F. "Migration: Chicken or Egg?" Southern Economic Journal, 37 (1971), 295-306. Okun, B. "Interstate POpulation Migration and State Income Inequality: A Simultaneous Equation Approach," Economic Development and Cultural Change, 16 (1968), 297-313. Olsson, Gunnar. Distance and Human Interaction. Regional Science Research Institute Bibliography Series, No. 2. Philadelphia: Regional Science Association, 1965. 204 O'Rourke, Desmond. "A Stocks and Flows Approach to a Theory of Human Migration with Examples from Past Irish Migration," Demography, 9 (1972), 263-274. Pan, J. 8. "Social Characteristics of Respondents and Non-Respondents in a Questionnaire Study of the Aged," American Sociological Review, 15 (1950), 780-781. Pastalan, L. A. and D. H. Carson (eds.). Spatial Behavior 2f Older People. Ann Arbor: Institute of Gerontology, 1970. Polk, R. L. and Co. Polk's St. Petersburg City Directory, 1971. Richmond: R. L. Polk and Co., 1971. lel Porter, R. "Approach to Migration through Its Mechanism," Geografiska Annaler, 38 (1956), 317-343. on vex Priest, G. E. "An Investigation of the Elderly in the Urban Environment with Special Reference to Their Housing." Unpublished M.A. thesis, Simon Fraser University, 1970. Pursell, Donald E. "Determinants of Male Labor Mobility," Demography, 9 (1972), 257-261. Raimon, Robert L. "Interstate Migration and Wage Theory," Review gf_ Economics and Statistics, 44 (1962), 428-438. Ratzel, F. Stade-Und Culturbilder Aus Nordamenka. Leipzig: Brockhaus 1876. Ravenstein, E. G. "The Laws of Migration," 2 vols. Journal g£_the Royal Statistical Society, 48 and 52 (1885, 1889), 241-301; 167-227. Rikkinen, Kalevi and Arnold Alanen. "Migration Trends of Older Aged Persons in Minnesota," Acta Geographica, 21 (1970), 5-18. Rose, Arnold M. "Distance of Migration and Socio-Economic Status of Migrants," American Sociological Review, 23 (1958), 420-423. IRimmel, Rudolph. "Understanding Factor Analysis," Journal 2£_Conflict and Resolution, XI (1967), 440-480. Eiauer, Carl 0. "The Formative Years of Ratzel in the United States," Annals 2£_the Association 9£_American Geographers, 61 (1971), 245-254. Scflrnell, George A. "Factors Affecting Recent Trends of the Aged Popu- lation: A Case Study," Aging and Human Development, 1 (1970), 219-230. 205 Schnell, George A. "Geographic Association Between Migration and Selected Aspects of Age Structure," Proceedings 2f the Pennsylvania Academy 2£_Science, 43 (1969), 153-156. Sheldon, Henry D. "Future Trends in Our Older Population," The New Frontiers 2£_Agi g. Wilma Donahue and Clark Tibbitts (eds.). Ann Arbor: The University of Michigan Press, 1957, 65-75. Shevky, E. and W. Bell. Social Area Analysis: Theory, Illustrative Application, and Computational Procedures. Stanford: Stanford University Press, 1955. Shryock, Henry S. Population Mobility Within the United States. Chicago: Community and Family Study Center, 1964. "Redistribution of Population, 1940 to 1950," Journal 2f the American Statistical Association, 46 (1951), 417-437. Simkins, P. D. "Distribution of the Aged in Pennsylvania," Proceedings g£_the Pennsylvania Academy 9£_Science, 38 (1964), 183-188. sjaastad, L. A. "The Costs and Returns of Human Migration," Journal 2: Political Economy, 70 (1962), 80-93. . "The Relationship between Migration and Income in the United States," Papers and Proceedings 2£_the Regional Science Association, 6 (1960), 37-64. Smith, John and Douglas C. Marshall. Retirement and Migration i§.£§§_ North Central States: Two Planned Retirement Communities. Population Series, No. 23. Madison: University of Wisconsin Press, 1970. Smith, T. Lynn. "The Changing Number and Distribution of the Aged Negro Population in the United States," Phylon, 18 (1958), 339-354. . "The Changing Number and Distribution of the Aged Popu- lation," Journal 2£_the American Geriatrics Society, 3 (1955), 1-14. . "The Distribution and Movements of the Aged Population," Journal g£_Business (1954), 108-118. . "The Migration of the Aged," Growing with the Years. New York Joint Legislative Committee on Problems of the Aging, 1954, 69-80. . "The Migration of the Aged," Problems 2£_America's Aging Population. T. Lynn Smith (ed.). Gainesville: University of Florida Press, 1951. Stouffer, Samuel A. "Intervening Opportunities," American Sociological Review, 5 (1940), 845-867. 206 St. Petersburg Planning Commission. Land Use Plan. St. Petersburg: St. Petersburg Planning Commission, 1970. Thomas, Dorothy 3. "Age and Economic Differentials in Interstate Migra- tion," Population Index, 24 (1958), 313-325. . Research Memorandum gn_Migration Differentials. New York: Social Science Research Council, 1938. Thomas, Dorothy S., gE_§13 Population Redistribution and Economic Growth in_the United States. 3 vols. Philadelphia: The American Philosophical Society, 1957, 1960, 1964. Thomas, Robert N. "Internal Migration to Guatemala City, Guatemala, C. A." Unpublished Ph.D. dissertation, Pennsylvania State Uni- versity, 1968. "The Migration System of Guatemala City: Spatial Inputs," Professional Geographer, 24 (1972), 105-112. Thomlinson, Ralph. Population Qynamics. New York: Random House, 1965. Thornthwaite, C. W. Internal Migration i2_the United States. Phila- delphia: University of Pennsylvania Press, 1934. Tibbitts, Clark. "Retirement Problems in American Society," The Ameri- can Journal 2§_Sociology, 59 (1954), 301-308. Tiebout, Charles M. "A Pure Theory of Local Expenditures," Journal 2: Political Economy (1956), 416-424. Ullman, Edward L. "Amenities as a Factor in Regional Growth," Geograph- ical Review, 44 (1954), 119-132. U.S. Bureau of the Census. Census g£_Population and Housingrp 1970 Census Tracts Final Report: Tampa-St. Petersburg SMSA. Washington: U.S. Government Printing Office, 1970. . 1970 Census pf Population and Housingprlorida: General Demographic Trends for Metropolitan Areas, 1960-1970. washington: U.S. Government Printing Office, 1971. . Components QEDPqulation Change, 1950-1960, for Countiegi SMSA's, State Economic Areas, and Economic Subregions. washington: U.S. Government Printing Office, 1962. . "Mobility of the Population of the United States: March, 1969-March, 1970," Current Population Reports. Series P-20, No. 210, 1971. U.S. Congress. Senate. Special Committee on Aging. "The Multiple Hazards of Age and Race: The Situation of Aged Blacks in the United States," A Working Paper, 1971. 207 Webber, Irving. "Adjustment to Retirement Made by Aged Persons," Proceedings 2f the Louisiana Conference 92_Aging. Baton Rouge: Louisiana State University Press, 1958. . "The Effect of Migration on the Number and Distribution of the Aged in Florida," Journal 2£_Gerontology, 11 (1956), 323-327. . "The Organized Social Life of the Retired: Two Florida Communities," American Journal 9£_Sociology, 59 (1954), 340-346. . "The Retired Population of a Florida Community," Problems of America's Aging Population. T. Lynn Smith (ed.). GainesvilIEt University of Florida Press, 1957, 87-101. . The Retired Population 9£_St. Petersburg, Its Character- istics and Social Situation. Tallahassee: Florida State Im- provement Commission, 1950. . "Significant Socioeconomic Factors in Measuring Health Zipf, Status and Health Needs," Aging: ‘A_Regiona1 Appraisal, Carter C. Osterbind (ed.). Gainesville: University of Florida Press, 1961, 82-93. G. "The P1P2/D Hypothesis in the Intercity Movement of Persons," American Sociological Review, 11 (1946), 677-686. APPENDICES APPENDIX A. MATERIALS RELATED TO THE FACTORIAL ECOLOGY OF ST. PETERSBURG, FLORIDA APPENDIX A. MATERIALS RELATED TO THE FACTORIAL ECOLOGY OF ST. PETERSBURG, FLORIDA Table A1. List of Variables Used in the Analysis Xl - (SEXRATIO) Sex Ratio X2 - (POPUNDlS) % of the Population Under 15 Years Old X3 - (POPOVR65) % of the Population Over 65 Years Old X4 - (DEPRATIO) Dependency Ratio X5 - (FERRATIO) Fertility Ratio X6 - (POPMARRD) % of the Population Married X7 - (POPWIDOW) % of the Population Widowed X8 - (POPNONWH) % of the Population Nonwhite X9 - (POPFBORN) % of the Population Foreign Born X - (PUND8EDU) % of Persons 25 Years Old and Over with Under 10 . 8 Years of Education X11 - (P4COLLG+) % of Persons 25 Years Old and Over with 4 Years or More of College X12 - (SAMHOUSE) % of Persons 5 Years Old and Older Residing in the Same House in 1960 as in 1955 X13 - (DIFHOUSE) % of Persons 5 Years Old and Older Residing in a Different House in the U.S. in 1955 (as compared to 1960) X14 - (DHCENCIT) % of Persons 5 Years Old and Older Residing in a Different House in the Central City of This SMSA in 1955 X15 - (DHNORWST) % of Persons 5 Years Old and Older Residing in a Different House in the North and West in 1955 X16 - (DHSOUTH) % of Persons 5 Years Old and Older Residing in a Different House in the South in 1955 X17 - (INCUND3.) % of All Families With Incomes Under $3,000 in 1959 X18 - (INC3.-6.) % of All Families with Incomes of $3,000 to $6,000 in 1959 208 209 Table A1. (Cont'd.) X19 (INC6.-10) % of A11 Families With Incomes of $6,000 to $10,000 in 1959 X20 (INC10-14) % of All Families With Incomes of $10,000 to $14,000 in 1959 X21 (INCOVRlS) % of All Families With Incomes $15,000 and Over in 1959 X22 (MPROFTCH) % of Males Employed As Professional, Technical, and Kindred Workers X23 (MMNGROFF) % of Males Employed As Managers, Offs., and Propr's, Including Farm X24 (MLABORER) % of Males Employed As Laborers, Except Mine X25 (LBFRATIO) Labor Force Ratio X26 (MLSINLBF) % of Males (14 Years and Over) in Labor Force x27 (FLSINLBF) % of Females (14 Years and Over) in Labor Force X28 (MUNEMPLO) % of Males (14 Years and Over) Unemployed X29 (RENTHOUS) % of A11 Housing Units Renter Occupied X3o (VACTHOUS) % of All Housing Units Available Vacant (For Sale or Rent) X3l (HOUSSOND) % of All Housing Units Sound X32 (HOUSlUNIT) % of Housing Structures with One Unit X33 (HOUSBUT+) % of Housing Structures with 3 or More Units X34 (BLT50-60) % of Housing Units Built 1950 to March, 1960 X3S (P/POWNHS) Median Numbers of Persons Per A11 Owned Occupied Units X36 (P/PRNTHS) Median Number of Persons Per all Renter Occupied Units X37 (TOTPOP60) Total Population in 1960 X38 (POP/HHLD) Population Per Household X39 (MDNAGEML) Median Age of Males X (MVIN5860) % of A11 Housing Units Moved into 1958 to 40 March, 1960 X (RNTLT60.) % of All Renter Occupied Units With Gross Rent 41 Less Than $60.00 X42 (RT60-100) % of All Renter Occupied Units With Gross Rent of $60 to $100 X43 (RICO—150) % of A11 Renter Occupied Units with Gross Rent of $100 to $150 X (UTNDAUTO) % of All Occupied Units With No Automobile Available 44 210 mmw. II II II 0mm. II II II OHBdmmmA mN vmm. II II II II II II 0mm. mmmomdqz VN mNm. II II II II II o¢b.I II mmomuzzz MN mmm. II II II II II mam.I II moemommz mm Hmm. II II II II II oam.l II mHm>Oqu AN mmm. II II II II II mmm.I II VHIOHUZH 0N vmn. II II II mom.l II II II OHI.oUZH ma mom. vam. II II II II mam. II .oI.mozH ma omm. II II II Hmm. II 0N0. II .mQZDUZH ha cow. II II II II www.l II mmv. mBDOme 0H Nvm. II II II hmm.I mmo.l II II Bmzmozma ma own. II II II Nam. II II II BHuzmqu vH mum. II II II II mmm.I II II mmDOmmHD ma New. II II II II boo. II II Mm30m24m NH Hmm. II II II II II Hmm.I II +oqqoovm Ha NNm. II II II II II vmm. mam. DomeZDm OH mom. II II II II II II 0mm.l zmommmom m chm. II II II omv. II II won. mzzozmom m Hhm. II II II 6mm. II II wmo.l ZOQH3mOm h mmm. II II II 0mm.I II II mmv. ommmzmom o Nmm. II II II II II I vmm. OHBdmmmm m omm. II II II mom. II II mm@.l OHemmmmo v mom. II II II II II II Nmm.I mmm>0mom m mum. II II II II II II Nmm. mHQZDmOm N mvm. II II II II II II mmo. OHBdmxmm H N3 5 m m v m N H mmHQMHum> "xufiamcseeoo uouomm uouowm uouomm uouomm nouomm uouomm nouomm «mocflpmoq Houomm pcm muouomm .N< manna .3320 53 «>2 met M Eu M me... mocmu mgu ca mcfixa mmcflomoa .manmu on» no cowumuwumumucw oumuaaflumm 09 .cofiuwmom xmswum> Hashes u on muouomu mwxm mamflocflum mo cofiuwuou scum ocauHSmmu coausaom onsuosuum mameflm any scum mum mocwomoa Ham. 211 oomm. «vow. ommm. moon. mono. omvo. mom. mucmaum> mo cowuuom Iona o>wumassao ommo. mmmo. Hovo. «mam. mmma. ommfl. mmom. mocmHum> mo cowuuomoum mom. II II II mvm. II II II 0924029: vv mam. II II mom. ovv.I II II II omHIooam mv oom. II II omo.I II II II II ooanoam we coo. II II II Hmo. II Nov. II .ooeqazm aw vow. II II II II oam.I II II ooomzH>z ov mom. II II II II II II mom.I gamodzoz mm mom. II II II II II II 0mm. oqmm\mom on non. II II II II II II Hon. oomomaoe hm mam. II II II II II II omm. oxezmm\m on How. II II II II II II Ham. mmzzom\m mm mmm. II II II ooo.I II II II ooIomeAm vm Hmm. II II II oom. II II II +Bommoom mm omm. II II II nmm.I II II II BHZDHoDom mm own. II II II II II hom.I II 020mmoo: Hm vow. II II II II mom.I II II maomeo¢> on How. II II II moo. II II II moomezmm on man. II moo. II II II II II oqmszDz om moo. II II II II II mvv. ovo. mququm mm mam. II II II II II II mam. mququz on mg n o m v m N H moanmwuu> "wuHHmcsEEoo Houomm uouomm uouomm nouomm uouomm uouomm nouomm A.U.ucouv .Nm wanna APPENDIX B. QUESTIONNAIRE “‘*‘ -:I “__a“ .-r ... g .. APPENDIX B. QUESTIONNAIRE Identification Address Census Tract Name (optional) Race (CHECK ONE) Caucasian ____ Negro _____Other ( ) Sex Male ____Female Were you born in the United States? ____ Yes (Go to #5) ____ No IF NO, in what country were you born? Marital Status Single _____Married ____ Divorced ____ Separated Widow _____Widower How old was the major family wage-earner when he or she retired? Years How old was the major family wage-earner when he or she moved to St. Pete? Years How old is this same person (the major family wage-earner) today? Years the person in reference is deceased IF THE RESPONDENT AND THE EX-MAJOR FAMILY WAGE-EARNER ARE DIFFERENT PEOPLE, what is your age? Years 212 10. ll. 12. l3. 14. 15. 16. 17. 213 What was the occupation of the major family wage-earner when he or she retired? Into which of these categories did your family's income fall just prior to retirement? INDICATE WITH THE NUMBER ONE _____Under $5,000 .____ $15,000-19,999 ____ $30,000-39,999 _____$5,000-9,999 _____$20,000-24,999 _____$40,000-49,999 $10,000-14,999 _____$25,000-29,999 ____ $50,000 and over &_; Into which of the above categories does your current income fall? i INDICATE WITH THE NUMBER TWO 1 Present Employment Status V a A. Is anyone in this family currently employed full-time (40 hours/ week)? Yes No B. Is anyone in this family currently employed part-time (less than 40 hours/week)? Yes No When the major family wage-earner retired, where was he or she living? State County City, Town, or Metropolitan Area (if applicable) Was this home located: On a farm ____ in the suburbs _____in the central city in the countryside (but not on a farm) How many times have you moved since retiring? Once (Go to #19) ____ Twice ____ 3 times ____ 4 times more than 4 times IF YOU MOVED MORE THAN ONCE BETWEEN RETIREMENT AND YOUR MOVE TO ST. PETE, list the location of each of the places you lived between your home at the time of retirement and your first home in St. Pete. (List these in the order of their occupancy. In other words, begin by giving the location of your first home after retirement and end 18. 19. 20. 21. 22. 23. 214 with your last address prior to St. Pete. STATE COUNTY CITY, TOWN OR METRO AREA (if applicable) Was your last home prior to St. Pete located: On a farm in the suburbs in the central city in the countryside (but not on a farm) How long have you lived in St. Petersburg? Less than 12 months (less than 1 year) 36-47 months "“' (3-4 years) 12-23 months (1-2 Years) 48-59 months (4-5 years) 24-35 months (2-3 years) 5-10 years 10-15 years more than 15 years Have you ever changed your residence in St. Pete? Yes No (GO TO #24) IF YES, how many times have you moved? Once Twice 3 times more than 3 times IF YES TO #20, prior to your current location, what was your last address in St. Pete? (Street and house number) IF YES T0 #20, when did you move from this address to your current address? Less than 6 months ago 7-12 months ago 13-18 months ago 19-24 months ago more than 24 months ago 24. 25. 26. 27. 29. 30. 31. 215 As close as you can remember, how many times did your family move in the 30 years preceding retirement? None 1-3 times 4-5 times 6-7 times 8-9 times 10 or more times What was your most important reason for moving to St. Pete? What was your second most important reason for moving to St. Pete? Here is a list of reasons why different people move. How important was each for you? (CHECK ONE ON EACH LINE) of no very importance great great somewhat slight at all the persuasion of friends the persuasion of relatives weather or climatic considerations large number of retirees in St. Pete tax benefits in Florida the advice of a doctor to be near friends to be near relatives job considerations other reasons Was there anything about your previous residence that encouraged you to move away to a new location? Yes No (GO TO #30) IF YES TO #28, what were these factors? When you were thinking about moving to St. Petersburg, did you consider any other p1ace(s) as a possible new home? Yes No (GO TO #32) IF YES, where were these other places located? (In what state(s)?) Florida California Arizona Other ( ) 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 216 When did you decide to move to St. Petersburg? More than 5 years before retiring Less than 5 years before retiring Less than 5 years after retiring More than 5 years after retiring Before moving to St. Pete, did you have any friends living in the city? Yes No (GO TO #35) IF YES, did these friends influence your decision to move to St. Pete? Yes NO Before moving to St. Pete, did you have any relatives living in the city? Yes (GO TO #37) No IF NO, before moving to St. Pete, did you have any relatives in Florida? Yes No (GO TO #40) IF YES TO EITHER #36 or #37, did these relatives influence your decision to move to St. Petersburg? Yes NO IF YES TO EITHER #36 or #37, did these relatives have any doubts about your moving to St. Pete? Yes No (GO TO #40) IF YES, what were these doubts? Other than relatives in Florida, did anyone else advise you against moving to St. Pete? Yes No (GO TO #43) IF YES, who were these people? relatives friends doctor other ( ) 217 42. IF YES TO #40, what were these doubts? 43. Did you own property in St. Pete before you moved into the area? Yes No (GO TO #45) 44. IF YES, how long did you own that property before moving to St. Pete? less than 1 year l-5 years 6-10 years ll-lS years more than 15 years 45. Were you familiar with St. Petersburg before you moved here? Yes No (GO TO #47) 46. IF YES, what was your most important source of information? 47. Here is a list of several sources of information about St. Pete. How much influence did each one have on your level of information before you moved here? (CHECK ONE ON EACH LINE) very no very great great some little influence influence influence influence influence at all a. personal visits to St. Pete b. friends c. relatives d. television e. radio f. magazines 9. mailed advertisements h. newspapers 1. other ( ) 48. How many times did you visit St. Pete before deciding to move here? Never (GO TO #52) Once Twice 3 times More than 3 times 49. During your visit(s) what impressed you most about the city? 50. 51. 52. 53. 54. 55. 56. 57. 218 During your visit(s) what impressed you least about the city? How long was it between your last visit and your actual move to St. Pete? less than one year 1-2 years 2-3 years 3 years or more Before you moved to St. Pete, did you know where you were going to live within the city? Yes No (GO TO #54) IF YES, how did you make these arrangements? During a visit by mail by telephone through a realty agency through friends or relatives other ( ) Since you moved to St. Pete, how frequently, on the average, have you left the area for a period of a week or longer? Never less than once a year once a year twice a year 3 times a year more than 3 times a year Do you consider St. Pete to be your permanent home? Yes (GO TO #57) No IF NO, where is your permanent home? State County City, Town, or Metropolitan Area (if applicable) The last time you left St. Petersburg for more than a week: a. Did you travel out of state? ____ Yes No b. Did you visit friends? ____ Yes ____ No c. Did you visit relatives? _____Yes No d. Did you return to your "old home"? ____ Yes ____ No e. Did you go on business? _____Yes No f. Did you go on a vacation or sight-seeing trip? Yes No 58. 59. 60. 61. 62. 63. 64. 65. 219 Would you encourage other retirees to: a. Live in the South? _____Yes No b. Live in Florida? _____Yes No c. Live in St. Petersburg? ____ Yes _____No d. Live in your neighborhood? _____Yes No e. IF THE RESPONDENT IS LIVING IN A HOTEL OR APARTMENT COMPLEX, live in your apartment complex? Yes No IF YES TO ANY OF THESE QUESTIONS, why? (Why is St. Pete a good home for retirees?) good facilities good climate tax benefits lots of retirees to make friends with other ( Are you satisfied with St. Pete as a retirement home? Yes No Is there anything about living in St. Pete that you don't like? Yes No (GO TO #63) IF YES, what don't you like about living in St. Petersburg? Are you thinking of moving again? Yes No (GO TO #65) IF YES, are you thinking of moving: a. Back "home"? Yes No b. Out of St. Pete? Yes No c. Out of Florida? Yes No If you could afford it, would you like to move back to your "old home"? THANK YOU SO VERY MUCH!!!!