Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Transition areas : a study of location factors affecting low-income housing Policzer, Irene 1983

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1981_A8 P64.pdf [ 9.35MB ]
Metadata
JSON: 831-1.0095458.json
JSON-LD: 831-1.0095458-ld.json
RDF/XML (Pretty): 831-1.0095458-rdf.xml
RDF/JSON: 831-1.0095458-rdf.json
Turtle: 831-1.0095458-turtle.txt
N-Triples: 831-1.0095458-rdf-ntriples.txt
Original Record: 831-1.0095458-source.json
Full Text
831-1.0095458-fulltext.txt
Citation
831-1.0095458.ris

Full Text

'.TRANSITION AREAS: A STUDY OF LOCATION FACTORS AFFECTING LOW-INCOME HOUSING by IRENE POLICZER B. Arch. Universidad de Chile, 1966. A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS i n THE FACULTY OF GRADUATE STUDIES School of Community and Regional Planning We accept t h i s study as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April ; 1981 © Irene P o l i c z e r f 1981 I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s . f o r an advanced degree a t t h e U n i v e r s i t y o f B r i t i s h C o l u m b i a , I a g r e e t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s may be g r a n t e d by t h e head o f my department o r by h i s o r h e r r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n o t be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . Department o f School of Community and Regional Planning The U n i v e r s i t y o f B r i t i s h C o l u m b i a 2075 Wesbrook P l a c e V ancouver, Canada V6T 1W5 Date Apr 1 1* 1981. 1Q\ ABSTRACT i i Transition areas located at the fringes of Central Business D i s t r i c t s are, i n most c i t i e s , one of the important r e s i d e n t i a l location options f o r the lowest income groups. The dynamics of c i t y growth result i n a process of abandonment of those areas by the high income groups and occupation by the poor; most neighborhoods i n those areas have a low l e v e l of housing maintenance and low rental values. Some housing programs, such as NIP, RRAP, attempt to improve the housing conditions of the poor by upgrading the housing stock i n those areas. I t i s f e l t that, by subsidizing housing repairs and neighbohood improvement programs, two objectives can be achieved: better housing f o r the poor and neigh-borhood s t a b i l i t y . At the same time, there i s evidence i n some North American c i t i e s of a reversal of the suburbanization process: some medium-to-high income groups which t r a d i t i o n a l l y tend to locate i n suburban areas, now are locating i n old-central neighborhoods. The houses are extensively renovated, and some of these areas are gradually becoming new middle-to-high class r e s i d e n t i a l d i s t r i c t s . This trend raises some concern with respect to the effects of th i s process on low-income r e s i d e n t i a l options. Although there i s some evidence that the g e n t r i f i c a t i o n process may produce dislocation problems for the poor, there seems to be l i t t l e agreement as to the . significance of th i s problem and the type of housing p o l i c i e s that would be more appropriate to ensure adequate housing for the poor i n areas undergoing g e n t r i f i c a t i o n . This research has four major objectives: 1) To i d e n t i f y the role of t r a n s i t i o n areas on low-income r e s i d e n t i a l loca-t i o n . 2) To i d e n t i f y those variables that can explain the g e n t r i f i c a t i o n process i n central neighborhoods. i i i 3) To assess the effects of g e n t r i f i c a t i o n , p a r t i c u l a r l y on low-income r e s i -dential location options. 4) To assess the effects of housing and neighborhood improvement subsidies on low-income location i n gentrifying areas. The method chosen was that of theoretical research. A review of d i f -ferent bodies of location theory was used to derive a conceptual location model which combines economic, socio-ecologic and dynamic components of r e s i d e n t i a l location. The model, i n turn, was applied to analyze the four research areas l i s t e d i n the objectives. As a general conclusion drawn from the analysis, i t i s suggested that the g e n t r i f i c a t i o n process defines a planning situation characterized by co n f l i c t i n g goals and long-term uncertainty. The analysis provided some insight as to the type of uncertainty involved, the nature of the goals con-. f l i c t , and some indicators that can be useful for housing policy i n gentrifying areas. Since the g e n t r i f i c a t i o n process appears to be very recent i n Canada, most of the evidence presented- i n t h i s research i s based on US l i t e r a t u r e . However, the approach taken has attempted to focus on those variables that would appear to be more applicable to the Canadian scene. The model presented i n t h i s research can be used for a number of planning purposes., one of which i s measuring and understanding '.the, occurrence and significance of gentrif i c a -t i o n i n Canadian c i t i e s . i v TABLE OF CONTENTS Page Chapter 1, Background 1 1.1. Middle-Class Resettlement 1 1.2. Research Outline 8 Chapter 2, Economic Factors of Location 15 2.1. Economic Theory, Basic Concepts 15 2.2. Economic Factors and Low-income Location 21 2.3. Transition Areas: Economic Definition 23 Chapter 3, Socio-ecological Factors of Location , ........ 26 3.1. Urban Ecology and City Growth 27 3.2. Social Factors of Segregation 31 Chapter 4» Dynamic Factors of Location 43 4.1. The F i l t e r i n g Process 44 4.2. The Neighborhood Change Process 66 Chapter 5, The Arbitrage Model 75 5.1. Arbitrage: Main Concepts 76 5.2. Arbitrage: Discussion , ... 83 5.3. Arbitrage i n Transition Areas 90 Chapter 6, The Gen t r i f i c a t i o n Process 102 6.1. A Multidimensional Location Model 103 - 6.2. The Nature of Gentrif i c a t i o n 1.1.5 Chapter 7, Policy Implications 1.40 7.1. Summary of Findings 1-41-7.2. Policy Implications • 153 7.3. Future Uses of the Location Model , -.. x6>0 Bibliography 166 V LIST OF DIAGRAMS Page Diagram 2.1 20 Diagram 2.2 20 Diagram 2.3 24 Diagram 2.4 25 Diagram 4.1 48 Diagram 4-2 49 Diagram 4-3 50 Diagram 4.4 51 Diagram 4.5 52 Diagram 4.6 52 Diagram 4«7 • •• 53 Diagram 4.8 54 Diagram 4-9 55 Diagram 4-10 57 Diagram 4-11 58 Diagram 4-12 59 Diagram 4.13 59 Diagram 4.14 • • • 60 Diagram 4.15 60 Diagram 4.16 60 Diagram 4.17 61 Diagram 4.18 61 Page Diagram 4.19 62 Diagram 4.20 62 Diagram 4.21 .................. 62 Diagram 4-22 63 Diagram 4-23 63 Diagram 5.1 ........... 79 Diagram 5.2 84 Diagram 5.3 ................ 85 Diagram 5.4 ••• 92 Diagram 5.5 93 Diagram 5.6 .... 94-Diagram 5.7 96 Diagram 6.1 . 108 Diagram 6.2 I I4 Diagram 6.3 123 Diagram 6.-4. ....... 130' Diagram 6.5 130" Diagram 6.6 131> Diagram 6.7 132 Diagram 6.8 133 Diagram 6.9 J34. 1 .CHAPTER 1: BACKGROUND 1.1. MIDDLE-CLASS RESETTLEMENT IN OLDER URBAN NEIGHBORHOODS: An overview. In the past twenty-five years, few issues have presented, as great a challenge to urban policymakers as the decline of central . c i t i e s . During t h i s period, large numbers of middle and upper-income, families, chose to l i v e i n suburban communities. As families migrate from the c i t y to the suburban ring, they often leave behind a decaying core increasingly populated by low-income and minority families. Despite evidence that these trends are continuing, there are signs that considerable neighborhood r e v i t a l i z a t i o n i s occuring i n c i t i e s across US and Canada. Though a few of these areas were site s of Urban Renewal, pro-grams, most have undergone renovation and restoration through private invest-ment a c t i v i t y . The significance of th i s trend l i e s i n i t s stark contrast to the urban-to-suburban migration patterns which have t r a d i t i o n a l l y predominated i n metropolitan;areas. This movement, which has been termed " g e n t r i f i c a t i o n " , "Neighborhood resettlement", "market (private) r e h a b i l i t a t i o n " , etc., i s not unprecedented. Theorists such as Firey (194-7), Hoover and Vernon (1959), and Birch (1971) have i d e n t i f i e d i s o l a t e d urban locations where deviations from t r a d i t i o n a l norms have occurred. But growing evidence suggests that, at least i n the US, the incidence of middle-class immigration to older central areas has increased rather substantially. (Dennis Gale, 1979. p. 293) In a survey of public o f f i c i a l s and r e a l estate o f f i c i a l s i n 143 c i t i e s i n the US, Black found that 48% of communities over 50,000 population had i n 1975 some degree of private market, non-subsidized housing renovation underway i n older r e s i d e n t i a l neighborhoods. Another survey, of public o f f i -c i a l s and l o c a l c i t i z e n organizations i n the t h i r t y largest US c i t i e s , d i s -covered that resettlement was occurring i n almost a l l of them. (Gale, 1979, p. 293). In a study of forty-four c i t i e s done by the National Urban Coalition 2 i n 1978, substantial market r e h a b i l i t a t i o n was,identified i n almost 75% of the cases. The vast majority of those neighborhoods (nearly 90%) reported the onset of r e h a b i l i t a t i o n a c t i v i t y only within the past eight years. (National Urban Coalition, 1978, p. 4.). One important key to understanding the reasons for t h i s apparent departure from c l a s s i c a l precepts of r e s i d e n t i a l location theory i s the develop-ment of broad-based data from opinion surveys of resettlement households. Unfortunately, such a comprehensive study does not yet exist. Nevertheless, a number of i n d i v i d u a l , separately conducted surveys have been performed recently i n American resettlement neighborhoods and t h e i r results provide some indica-tions as to the i d e n t i t y of the r e s e t t l e r s , t h e i r geographic origins, the reasons for t h e i r choice, and some effects on low-income r e s i d e n t i a l displace-ment. The summary of evidence presented i n t h i s section has been drawn mainly from three sources: Dennis Gale (Gale, 1979) summarizes data related to demographic characteristics of the new c i t y dwellers, t h e i r geographic origins and the i r reasons for relocation. Howard Sumka (Sumka, 1979) centers on the extent of the r e v i t a l i z a t i o n movement and some effects on low-income disloca-t i o n . Further evidence on the displacement problem i s drawn from the study carried out by the National Urban Coal i t i o n (NUC, 1978). P r o f i l e of the new inner-city dwellers Data sources permit observations on s i x demographic characteristics of r e s e t t l e r households: household size, r a c i a l composition, annual income, age, education and occupation. There i s also evidence of the geographic origins of the r e s e t t l e r s . C o l l e c t i v e l y , these data lend considerable weight to popular characterizations of the r e s e t t l e r s . "The most t y p i c a l such household i s childless and composed of one or two white adults i n t h e i r l a t e twenties or t h i r t i e s . College educated, often possessing graduate education, the household head i s most l i k e l y a professional 3 or (less l i k e l y ) a manager. The annual household income varies among metropo-l i t a n areas but i s l i k e l y to range between $15,000 and $30,000 with several re s e t t l e r s earning more than $4-0,000. Doubtless, many of those earning higher incomes are composed of two workers. For the most part, the above evidence seems to be supported by more descriptive accounts of resettlement i n several American c i t i e s . " (Gale, p. 295). Evidence i n terms of geographic origins of the r e s e t t l e r s , on the other hand, does not support the "Back-to-the-City-Movement" idea. Though many observers have assumed, with l i t t l e or no evidence, that r e s e t t l e r s are mostly d i s s a t i s f i e d , former suburbanites, more recent research shows, that. "... a r e l a t i v e l y small minority of households moved... from the c i t y ' s en-c i r c l i n g ' suburbs... Less than 20% of res e t t l e r s surveyed i n Atlanta, Boston, Cambridge, and Washington said that they had done so." "In fact, more appear to have located i n some c i t i e s from outside the metropolitan area altogether... More than one-half (and i n some cases as many as 90%)... moved to the renovating area from somewhere within the municipal boundaries... About one-half of the r e s e t t l e r s had moved from an apartment and two-thirds had been renters i n t h e i r previous location." " I f these figures are even roughly representative of r e s e t t l e r s i n other c i t i e s , they suggest that most are first-home buyers. I t i s l i k e l y that they migrated to the c i t y to attend college or graduate, school or to take employment there. After working a few years, they accumulated enough ca p i t a l to make a down-payment on a house and were encouraged to do so by .their r i s i n g incomes and favorable federal tax p o l i c i e s . " (Gale, 1979, .p. 296). Causes of the g e n t r i f i c a t i o n process As already discussed, the existing evidence supports the notion that the new inner-city s e t t l e r s are not the t y p i c a l suburban family trying to move back to the c i t y ; rather, the r e s e t t l e r s are mostly new households who prefer the innercity to the suburbs. The question.arises, then, as to why these new urban households prefer the inner c i t y . Several reasons have been suggested by different researchers: A) Rising suburban prices for new homes due to i n f l a t i o n i n labor, materials, financing costs, and to r e s t r i c t i v e growth controls,. have "forced buyers with l i m i t e d income to stay i n c i t i e s " . (Gale p. 301). The evidence shows, though, that the re s e t t l e r s have r e l a t i v e l y high incomes, so suburb-an locations are i n fact quite affordable fortthem. B) Household transportation costs are considerably higher when two household members are employed, so i t becomes economically ra t i o n a l to.minimize these costs. The counter-argument i s that, with two bread-winners, income i s also higher, and t h i s increased income would overcome the higher commuting costs. C) Household size and absence of children. Since the r e s e t t l e r s are mostly childless couples, they do not need spacious yards, schools and other f a c i -l i t i e s associated with suburban areas. From th i s point.of view, the. loca-tio n choice appears to be economically r a t i o n a l . D) Different values ( l i f e s t y l e changes), associated with the baby-boom genera-t i o n . The new s e t t l e r s would place a high value on the c u l t u r a l , h i s t o r i c a l , or architectural character of an area. E) Investment potential perceived by the new s e t t l e r s . Increases i n property values are indeed s i g n i f i c a n t i n the newly rehabilitated areas.. Some neighborhoods (Chicago's Wicker Park, Houston's Heights Area, San Francisco's Fillmore, Capitol H i l l i n Washington) have experienced home value increases of 500% to 1,000% i n the l a s t f i v e to eight years ("Housing", June 1980, p. 16). The evidence shown by surveys i n the US supports some of these ex-planations moretthan others. For instance, a survey carried out-in the Mount Pleasant and Capital H i l l sections of Washington (Gale 1976, 1977), showed that 5 the most highly-rated reasons for choosing to locate i n those areas were: Investment potential of the house purchased Relatively affordable price A c c e s s i b i l i t y to place of employment Ar c h i t e c t u r a l / h i s t o r i c a l value of the house/neighborhood. . Of lesser importance were the soc i a l and c u l t u r a l attractions of c i t y l i v i n g . The stage theory Most researchers seem to support the notion that the resettlement process tends to have a similar pattern i n different areas. This pattern i s bas i c a l l y a process of change i n family composition and preferences as the 'resettlement' area consolidates over time. Gale's study of the Mount. Pleasant and Capital H i l l areas, for instance, showed that the more recently renovated neighborhood (Mount Pleasant) had a higher proportion of young, male adults, while the older area (Capitol H i l l ) had higher proportion of women and families. The respondents of the two areas also showed different, degree of committment to t h e i r house/neighborhood (as shown i n 'intentions to move' and 'reactions to the dwelling 1 type of questions); those i n the. newer area showed a s l i g h t l y less favorable attitude and higher predisposition, to move. . This and other similar evidence suggest that the r e s e t t l e r s can be c l a s s i f i e d according to the stage of the r e v i t a l i z a t i o n at which they entered the neighborhood. "Early entrants tend to be 1 risk-oblivious 1;.they are followed by the 1 r i s k - t a k e r s 1 and, f i n a l l y , by the 'risk-averse' (Pattison 1977, as quoted by Sumka, p. 4-82). The following quote summarizes how the stage process i s perceived: "Early i n the r e v i t a l i z a t i o n of a neighborhood, come the pioneers. Advance characterizes them as 'middle-class people with lower-than middle-class incomes: a r t i s t s , writers, musicians, etc.' Others may be 'households unsure of t h e i r welcome' i n a t y p i c a l middle-class neighborhood, such as gays and i n t e r r a c i a l couples. Also, 'persons with a special f e e l for. the potential of a neglected grand home' - such as architects and decorators - w i l l move during the early stage of r e v i t a l i z a t i o n . " 6 "Later comes - as one Brooklyn pioneer says - 'the doctor and lawyer crowd. And then the suburban crowd with big bucks from the sale of th e i r homes." (From a r t i c l e i n 'Housing' - June 1980, about the report "US Housing Markets" published by Advance Mortgage Corp., USA). Neighborhood displacement The effects of private investment i n rehab on displacing low-income residents have attracted the attention of a number, of researchers; there i s also a growing movement ( i . e . i n c i t i e s with s i g n i f i c a n t rehab a c t i v i t y , such, as San Francisco) of community groups protesting against the displacement, problem occurring i n rehab neighborhoods. Yet there seems to be l i t t l e agree-ment as to the extent of the displacement problem, i t s nature, and i t s i m p l i -cations i n terms of government policy. Those who argue that displacement i n an increasingly serious problem compare the effects of rehab with those of Urban Renewal: "...There i s a great deal of information about the fate of displacees from urban renewal, highway, and other public works displacement experience, and there is.no reason to assume that the fate of present day displacees i s any d i f f e r e n t . . . These e a r l i e r studies found that a large proportion of the families displaced i n the 1950's and 60's -i n some cases as many as 86% of the study population - relocated to substandard housing... Those who claim that t h i s gloomy picture of . the past does not have relevance for today's displacement must be charged with the obligation of proving otherwise. U n t i l that time, i t i s important to act under the assumption that these displacement:; effects are just as true today as they were i n the 1930's, the 940's^' the 1950's and 1960's." (Hartman, _197.95Sp. 488) Displacement i s often associated to the role of r e a l estate speculators who begin acquiring properties i n areas they expect w i l l be a t t r a c t i v e to middle- to upper-income households. As speculators move into an area, the f i r s t to be affected are resident centers. Owner-occupants appear to be affected at a l a t e r stage, and t h e i r loss i s considered to be mostly the,-loss of a po-t e n t i a l l y higher s e l l i n g price, since they move out with the idea that the neighborhood i s s t i l l i n decline. The view that displacement i s not a serious problem i s based mostly on the following arguments: A) The r e s e t t l e r s are a small group, and occupy r e l a t i v e l y small areas; though 7 some neighborhoods may be.seriously affected, the nation-wide incidence of the problem i s not s i g n i f i c a n t . B) The phenomenon i s mostly of a temporary nature, produced by the preferences of a very special segment of the population (The baby-boom generation). As th i s group becomes older and th e i r values change, the g e n t r i f i c a t i o n trend w i l l tend to disappear. C) Empirical research to date has focussed primarily on the r e v i t a l i z a t i o n process i t s e l f , rather than i t s secondary effects; the analysis of d i s -placement raises d i f f i c u l t conceptual and measurement problems which no. one has addressed systematically. Therefore, the existing evidence i s not s u f f i c i e n t l y r e l i a b l e to assess the effects ofprivate rehab on displacement. Empirical research, so f a r , does not provide a clear answer to the two debating positions. On one hand, the main findings of the study done by the National Urban Coalition suggest that: The incomes of households moving into rehab neighborhoods are higher than those of the previous residents - but not s i g n i f i c a n t l y higher.. Professional and white-collar xrorkers tend to displace blue-collar workers and the unemployed. The elderly are most often displaced. The study done by the. National Urban Coalition centers on the characteristics of residents who move out, but say l i t t l e about.the new location of those groups. On the other hand, Pattison's study of the Bay Vil l a g e i n Boston suggests that many homeowners realized s u f f i c i e n t c a p i t a l gains from the sale of t h e i r properties to f u l f i l l a long-standing goal of moving to the suburbs. Others were able to retain t h e i r properties and convert them to rentals that generated..income s u f f i c i e n t to allow them to move to the suburbs. (Sumka, 1979, p. 4-85-4-86) The conclusions from these findings are often contradictory and most of them contain important sources of bias, either underestimating or overestimating the incidence of displacement. There seems to be . s t i l l a lack of a clear theoretical framework and a carefully.designed operational counter-part (Typical problems: how to. define displacee; how to i s o l a t e rehab from other reasons for non-voluntary moves). Furthermore, the existing research seems to center on the actual displacee; there i s l i t t l e ( i f any) evidence on the effect of g e n t r i f i c a t i o n i n terms of depriving new low-income households from a potential housing option. 1.2 RESEARCH OUTLINE The purpose of th i s research i s to provide some theoretical back-ground that can be helpful for the understanding of the nature of the g e n t r i f i cation process, i t s causes and i t s effects on low-income residents. The research centers on the following issues: 1) I d e n t i f i c a t i o n of the role of t r a n s i t i o n areas i n low-income r e s i d e n t i a l location. 2) I d e n t i f i c a t i o n of those variables that can explain the g e n t r i f i c a t i o n process in, central areas. 3) I d e n t i f i c a t i o n of some eventual effects of the g e n t r i f i c a t i o n process, p a r t i c u l a r l y on low-income location options. A) Assessment on the effects of housing and neighborhood improvement subsidie on low-income location options i n gentrifying areas. These four major research objectives, i n turn, are used to derive some policy implications on low-income housing programs i n gentrifying neigh-borhoods. The research consists, i n a review of different approaches to r e s i -dential location theory. The selected l i t e r a t u r e has been c l a s s i f i e d into two major categories: A) Equilibrium variables of location, and B) Dynamic factors of location. Both categories include l i t e r a t u r e drawn from the f i e l d s of economics and human ecology. 9 These two categories are somewhat a r t i f i c i a l , since a) the e q u i l i -brium models contain some dynamic elements and-b) there i s some overlap between economic and ecological approaches; rather than.restrictive categories, they are used as a general reference to order the discussion. Because of the broadness of the topic, the l i t e r a t u r e review i s necessarily very general; rather than a deep and precise analysis of each loca-t i o n factor, the intention has been that of presenting a broad spectrum..of variables and t h e i r interrelationships. Within the itime arid resource l i m i t a -tions, the research attempts to capture some of the multiple dimensions, i n which the location game takes place and how these multiple dimensions interact and change i n the process. Diagram 1.1 shows an outline of the research. The following i s a b r i e f summary of contents by chapters;, showing the major concepts developed i n each one. Chapters 2 and 3 present an overview of equilibrium factors of loca-t i o n from the point of view of economics and socio-ecology. ECONOMIC FACTORS OF LOCATION are presented i n Chapter .2. This chapter includes a summary of Alonso's r e s i d e n t i a l location theory and an application of t h i s theory for the d e f i n i t i o n of t r a n s i t i o n areas. Section 2.1; Economic Theory, i s a summary of Alonso's Central Place theory. This part contains the main assumptions and concepts of economic theory. Residential location i n t h i s approach i s seen as the result of.a bidding process i n x^hich each individual and firm reaches an equilibrium by choosing an optimum price and quantity of three consumption goods: land, distance to CBD and other (composite) goods. Section 2.2: Location and Income, discusses some implications for low-income 10 CHAPTER I I "1 ECONOMIC THEORY Location* function of: income; land quantity; accessi-b i l i t y  r CO o S0CI0-EC0L0GY Location, function of: s o c i a l segre- | gation; growth -J CHAPTER I I I TRANSITION AREAS Defi n i t i o n ..J CHAPTER IV HOUSING DYNAMICS The f i l t e r i n g process NEIGHBORHOOD DY- NAMICS Neighborhood l i f e -cycle; invation succession r CHAPTER V The Arbitrage Model L I CHAPTER VI TRANSITION AREAS Dynamics of change pro-cess; effects on low-income housing " 1 SYNTHESIS: A MULTIDIMEN- SIONAL MODEL Equilibrium: a c c e s s i b i l i ty matrix Dynamics: Arbitrage modell I "~1 GENTRIFICATION Causes (from accessibi-l i t y matrix); Effects (from Arbitrage model) CHAPTER VII "~I 1 POLICY IMPLICATIONS Role of t r a n s i t i o n areas; causes of gentrif ication 1; effects of g e n t r i f i c a -t i o n ; effects of rehab; uses of model J Diagram 1.1: Research outline 11 r e s i d e n t i a l location. Location for.those groups.is suggested to be a function of the r e l a t i v e d e s i r a b i l i t y of land as compared to a c c e s s i b i l i t y for the high-income groups. Section 2.3: Transition Areas: d e f i n i t i o n , applies the main concepts of econo-mic theory to define the location and role of t r a n s i t i o n areas. Transition areas are defined as those areas i n which developed urban land changes to different uses as a resul t of population growth. SOCIO-ECOLOGICAL FACTORS OF LOCATION are discussed i n Chapter Three. This chapter contains two sections: Section 3.1: Socio-ecology and.City Growth, describes r e s i d e n t i a l location as related to the growth process; the theory presented i n t h i s section provides additional background for the d e f i n i t i o n of t r a n s i t i o n areas discussed i n section 2.3. Section 3.2: The Segregation Process, discusses r e s i d e n t i a l location, as the result of a segregation process between different s o c i a l groups. Groups are c l a s s i f i e d i n the three dimensions of s o c i a l rank, race-ethnicity.and-family status. A summary of personalistic and non-personalistic factors affecting the segregation process i n each of these dimensions is.presented,. as well as some of the implications of th i s process for the centralization of some groups i n areas around CBD. Chapters 4- and 5 center on the dynamics of the location process. Location dynamics i s approached i n terms of housing turnover and neighborhood change theories. THE FIRST APPROACH TO LOCATION DYNAMICS, i n Chapter 4, contains the following parts: 12 Section J+.l, The F i l t e r i n g Process, which. centers on the dynamics of housing turnover. F i l t e r i n g i s considered to be a mechanism whereby low-income groups have access to housing under free market conditions. The discussion centers on definitions of f i l t e r i n g and analysis of some factors related to the role of f i l t e r i n g i n providing housing for the poor. Section 4-2: The Neighborhood Change Process, which describe physical and so c i a l components of neighborhood change. The two theories presented (neigh-borhood l i f e - c y c l e and invasion-succession) highlight some of the factors affecting the change process i n t r a n s i t i o n neighborhoods. THE SECOND APPROACH TO LOCATION DYNAMICS i s the content of Chapter 5. The dis-cussion i n t h i s chapter centers on the Arbitrage Model, which integrates housing and neighborhood dynamics into one comprehensive theoretical framework. This chapter has three sections: Section 5.1: The Arbitrage Model, presents the main concepts and assumptions of the model. Section 5.2: Discussion., suggests some of the positive and negative aspects of the model from the point of view of the preceeding body of theory and suggests some additional elements that could be introduced to i t . Section 5.3' Arbitrage i n Transition Areas, applies the main concepts of the Arbitrage Model for the analysis of the change process i n Transition areas. The analysis i n t h i s section centers on the location factors affect-ing r e s i d e n t i a l options for the poor and includes an assessment of the role of rehab i n t r a n s i t i o n areas. SYNTHESIS: THE GENTRIFICATION PROCESS, i s the content of Chapter 6. This chapter has two parts: 13 Section 6.1: Summary of Location Factors: a Multidimensional Model, which suggests a framework of location theory that encompasses the location factors reviewed i n the preceeding chapters. The r e s i d e n t i a l location game i s .pre-sented as a multidimensional decision process i n which different groups interact within the c i t y . Social groups are defined i n three dimensions: s o c i a l rank, ..family status and ethnicity; i n each of these dimensions, they are assumed to. make a choice related to f i v e dimensions of the c i t y : geometry, geography and the three dimensions of the s o c i a l groups at play. The interrelationships.between these variables are summarized i n an a c c e s s i b i l i t y matrix; the i n i t i a l output of the matrix i s a sp a t i a l d i s t r i b u t i o n of population and housing (neighbor-hood) quality. Housing and neighborhood quality, as well as population d i s t r i b u t i o n , are assumed to behave as intervening variables i n the a c c e s s i b i l i t y matrix: an i n i t i a l (equilibrium) situation yields a certain neighborhood and housing quality pattern. This pattern would affect successive location changes derived from population, growth. The changes would, i n turn, affect the existing housing and neighborhood quality and continue the change.process. Section 6.2, The Gentrification Process, applies the.location model summarized above to derive some possible factors which could explain the g e n t r i f i c a t i o n process. The causal factors of g e n t r i f i c a t i o n are derived from the accessibi-l i t y matrix and described i n terms of changes affecting the different dimen-sions of-the location game. Thus, causal factors are. assumed to be those factors which, by affecting the c i t y geometry, geography, and/or segregation 14 attitudes, would resu l t i n different a c c e s s i b i l i t y values i n the multidimen-sional matrix. Possible effects of the g e n t r i f i c a t i o n process are derived from the Arbitrage model. Assuming that g e n t r i f i c a t i o n i s triggered by one or.more of the suggested causal factors, the model i s used to detect resulting changes i n low-income housing options i n the affected neighborhoods, i n surrounding areas and i n overall housing supply. POLICY IMPLICATIONS are discussed i n Chapter 7. The implications include: A summary d e f i n i t i o n of the role of t r a n s i t i o n areas on low-income location. A summary of causal factors of g e n t r i f i c a t i o n . A summary of possible effects of g e n t r i f i c a t i o n , p a r t i c u l a r l y on low-income location options. A summary of possible effects of rehab i n gentrifying areas. . - A discussion of policy implications for the design of low-income housing programs i n gentrifying neighborhoods. - A discussion of some future uses of the location model. CHAPTER TWO:'ECONOMIC;FACTORS OF LOCATION 15 The choice of 'where to l i v e ' i s one of the most c r u c i a l decisions made by urban households - cr u c i a l not only for them but to the urban area as a whole - for the sum and synthesis of th i s r e s i d e n t i a l decision by households i s one of the prime determinants of the structure and character of our urban areas. The explanation of why households l i v e where they do, however, i s a complex and d i f f i c u l t task; the concerns involved are numerous and diverse, ranging from the price of land and i t s determinants to the so c i a l characteris-t i c s of neighborhoods. Theoretical approaches to re s i d e n t i a l location emcom-pass areas of economics, sociology, geography; each of these, approaches high-l i g h t s some of the multiple dimensions i n which the location game takes place. This chapter summarizes the main concepts of economic theory, which approaches location as a trade-off process between a c c e s s i b i l i t y and quantity of land. The concepts drawn from economic theory are used to .present an i n i t i a l explanation of low-income centralization around the Central Business D i s t r i c t . 2.1 ECONOMIC THEORY; BASIC CONCEPTS The basic notion of location as a function of distance from other l o -cations of interest has existed i n economics for a long time. The.earliest l i t e r a t u r e that dealt' with t h i s relationship centered on a g r i c u l t u r a l land. While Ricardo attributed differences i n land-rents primarily to f e r t i l i t y d i f -f e r e n t i a l s (Alonso, 1964., p. 3) , Von Thunen was the f i r s t to show that.the rents of land around a marketplace would also be determined by their proximity to that marketplace. The process of competitive bidding between potential users of various parcels of land would simultaneously determine both the l e v e l and rate of decrease of rents for land progressively farther from the central marketplace. (Alonso, 1964, p. 3,-4-). The r e l a t i o n between location, land price and transportation cost.: 16 wasapplied to urban land by R.M. Hurd (1903): "Since value depends on economic rent, and rent on location, and l o -cation on convenience, and convenience on nearness, we may eliminate the intermediate steps and say that value depends on nearness." (Alonso, 1964., p. 6). In 1926, R.M. Haig applied Von Thunen's basic hypothesis to urban r e s i d e n t i a l land. Location, i n Haig's model, was presented as a trade-off bet-ween a c c e s s i b i l i t y to the Central Business D i s t r i c t and. housing, rents: "An economic a c t i v i t y i n seeking a location finds, that, as i t approa-ches the centre, site-rents increase and transportation costs decline. As i t retreats from the centre, site-rents decline and transportation costs increase... The the o r e t i c a l l y perfect s i t e i s that which furnishes the desired degree of a c c e s s i b i l i t y at the lowest cost of f r i c t i o n . " (Haig, 1926, p. 4.22-4.23) "In choosing a residence purely as a consumption proposition one buys a c c e s s i b i l i t y precisely as one buys clothes or food. He consi-ders how much he wants the contacts furnished by the central loca-t i o n , weighing the 'costs of f r i c t i o n 1 involved - the various possible combinations of s i t e rent, time value, and transportation costs; he compares this want with other desires and his resources, and he f i t s into his scale of consumption and buys." (Alonso, 1964-, p. 8) These concepts were formally stated by Alonso (1964) and Wingo ( l96l) , and form the basis of what i s now called the 'economics of re s i d e n t i a l loca-t i o n ' . By considering the supply and demand for r e s i d e n t i a l , i n d u s t r i a l and ag r i c u l t u r a l land, Alonso formulated a general theory of the .determination of urban land uses and land rents, based on the p r i n c i p l e of a c c e s s i b i l i t y . This theory provides a basic, common theoretical framework for all.recent economic models of r e s i d e n t i a l location and trip-to-work. A summary of i t s main con-cepts (related to re s i d e n t i a l location) i s given here. Basic assumptions of the model are: A) The featureless p l a i n : The c i t y i s assumed to s i t on a pl a i n which i s uni-form and contains no h i l l s , views or any other, feature. Each location i s equal i n terms of so c i a l status, land and geographical characteristics. Municipal services and tax rates are uniform throughout the ,city. B) The economic man: I t i s assumed that i n d i v i d u a l households w i l l t r y to 17 maximize' s a t i s f a c t i o n by consuming the goods they, l i k e and avoiding those they d i s l i k e . Each individual w i l l spend whatever income i s available i n maximizing t h i s economic s a t i s f a c t i o n . C) There i s no housing stock: No fixed structures exist on the ground. Deci-sions of individuals and firms consider location only i n terms of amount of land and distance to employment; housing attributes are considered non-existent. D) Market equilibrium: There i s a fixed number of households and firms i n the market. Each in d i v i d u a l knows- the price of land at every location. Land i s bought and sold by free contract, without i n s t i t u t i o n a l or any other r e s t r a i n t s . E) The si m p l i f i e d c i t y : A l l employment and a l l goods and services are av a i l a -ble only at the center of the c i t y . The c i t y has unlimited room to expand, and this expansion i s r a d i a l around the Central Business D i s t r i c t . F) The decreasing price of land: From the point of view of the i n d i v i d u a l , the price of land i s fixed. This price decreases as transportation costs i n -crease. Land has a maximum price at CBD, where transportation costs are zero, and a minimum price at the farthest location. . . Within Alonso's framework, each household i s assumed .to derive u t i l i t y from the r e s i d e n t i a l s i t e (in terms of size of the lot)., and..from a composite commodity representing a l l other goods and services, while d i s u t i l i t y arises from commuting a c t i v i t y . In order to maximize u t i l i t y subject., to budgetary outlays, the household may trade-off a centralized location.against larger l o t sizes and/or the consumption, of other goods and services. The representative household's u t i l i t y function i s expressed as: u =? U('z,q,d) where g_ represents the quantity of land consumed; z represents the consumption of a l l other goods and services and d i s the distance between the centre of 18 the c i t y and the household's.residence. The household faces a budget cons-t r a i n t of the form: Y = P.z + R(d).q + (d) where Y represents the income of the household, P represents the price per unit of the composite commodity (z), R i s the price per unit of housing ( l ) and T i s the transportation cost of commuting, which are both assumed to be. a function of distance d from the CBD. I t i s assumed that these expenditures exhaust income, i . e . , there are no household savings. (Gera, Kuhn, 1977, p. 19) " . . . U t i l i t y i s maximized when the household allocates i t s income i n such a way as to equalize the ratios of the marginal u t i l i t i e s of the goods to t h e i r respective prices. For achievement of location-a l equilibrium, a process of substitution takes place between mar-ginal increases i n commuting costs... and marginal savings i n the housing expenditures... as the household considers sit e s at i n -creasing distances from the CBD. For the household's u t i l i t y to be at a maximum, i t must locate at a distance from the center of the c i t y where the marginal rent-cost i s equal to the marginal journey-to-work cost." (Gera, Kuhn, 1977, p. 21) Individual equilibrium i s described i n terms of bid-price curves. A bid-price curve for a resident i s defined as "the set of prices f o r land the indi v i d u a l could pay at various distances while deriving a..constant l e v e l of s a t i s f a c t i o n ; that i s to say, i f price of land were to vary with, distance i n the manner described by the bid-price curves, the individual.would be i n d i f f e r -ent among locations" (Alonso, 1964, p. 59). Alonso emphasizes three points i n th i s respect: l ) a bid-price curve refers to a given individual;. 2) a bid-price curve refers to a given l e v e l of s a t i s f a c t i o n ; since there i s an i n f i n i t e number of levels of s a t i s f a c t i o n for an i n d i v i d u a l , there would be for each ind i v i d u a l a family of bid-price curves corresponding to different levels of (l) . v....v. To avoid thorny d e f i n i t i o n a l problems and confusions between different terms, the term 'price' i s used i n i t s generic sense and i n -cludes market expressions such as contract rent, sales price and cost of ownership. Price i s defined as "the amount of money the occupant pays or would pay to the landlord for the right to the use of a unit of land (for example, a square foot or an acre) for a given period of time. Thus, the price of land times the quantity of land i n a s i t e represents the payment for the use of the s i t e " . (Alonso, 196/+, p. 16) 19 s a t i s f a c t i o n ; and 3) a bid price bears no necessary.relation to the actual market price of the land at each location. "A bid-price i s hypothetical, merely saying that, i f the price of land were such, the ind i v i d u a l would be s a t i s f i e d to a given degree". (Alonso, 1964, p. 59) The most important characteristic of a bid-price curve i s that i t has a negative slope; i t s value decreases with distance from the Central Busi-ness D i s t r i c t . Two factors account for t h i s negative slope:, increasing commu-ting costs and the increasing d i s u t i l i t y with distance: "... Because outward movement produces d i s u t i l i t y . . . (the marginal cost of movement).... must i n effect be a saving i n order that the bid-price curve maintain a given l e v e l of s a t i s f a c t i o n . This saving must occur i n land since commuting costs are increasing. The mar-ginal cost of movement i s equal to the quantity of land times the change of the price of land plus the increase i n the cost of commu-tin g . Bid price, then, has been defined so that, the income effect of cheaper land w i l l counteract the depressing effect of commuting costs on income, and w i l l permit the consumer to maintain.a given l e v e l of s a t i s f a c t i o n by substituting land and the composite good for a c c e s s i b i l i t y as distance from the center increases." (Alonso, 1964, p. 71) The steepness of each bid price curve can be viewed.as 'an indica-t i o n of how much an indiv i d u a l cares to be near the center i n order to avoid commuting': "The slope, and thus the shape, of bid price curves are ultimately determined by the tastes of the i n d i v i d u a l . . . The slope of a bid price curve may be equated roughly to the added (marginal) expense and bother of commuting as distance from the center increases. Thus, an indiv i d u a l who does not mind walking may.be expected to have a gently sloped curve up to the l i m i t of what he considers walking distance; i f he d i s l i k e s other means of .transportation, beyond th i s l i m i t his curve w i l l be f a i r l y steep... On the other hand, an indiv i d u a l who detests walking and i n s i s t s on driving everywhere w i l l have decreasing increments of nuisance and expense as distance increases. His bid price curve has decreasing steep-ness with distance." (Alonso, 1964, p. 90) In general, Alonso argues that individuals with steeper bid-price curves w i l l locate closer to the center of the c i t y . This i s shown i n diagram 2.1, which i l l u s t r a t e s one bid-price curve for each of three individuals. Diagram 2.1 "... The indiv i d u a l or firm with the least steep bid-price curve w i l l s ettle farthest from the center... The next ind i v i d u a l towards the center w i l l be the one with the next least steep bid price curves... Thus, the chain i s formed, l i n k by l i n k , u n t i l the price paid by the user with the steepests bid price curve at the center of the c i t y i s established." (Alonso, 1964, p. 89) The process of adjusting the values of different bid-price curves i n order to f i t actual market prices ( i . e . a situation of market equilibrium) i s explained i n terms of supply and demand factors. The supply element i s the t o t a l amount of land at any given distance from the center; the demand element i s the t o t a l number of households for any given shape of the bid-price curves. Since each curve represents a set of prices and locations with which each i n d i -vidual achieves a given l e v e l of s a t i s f a c t i o n , the process of adjusting to supply and demand factors can be viewed as a mechanism whereby each individual chooses a l e v e l of sat i s f a c t i o n such that he finds an optimum location within a given set of market prices. — - C B D . '- . D I S T A M C E : Diagram 2.2 21 Diagram 2.2 shows one example of- how this adjustment takes place. An i n i t i a l set of bid-price curves was derived for a certain l e v e l of s a t i s -f a c t i o n . I f t h i s i n i t i a l set (showed i n dotted l i n e ) yields less locations than bidders i n the market, each household would have to adjust by moving to.a next l e v e l and paying higher prices. This may i n turn result i n more land, available than bidders, i n which case the set of curves adjust downwards.. After a certain number of i t e r a t i o n s , a correct solution should be found i n which the t o t a l land equates the t o t a l number of bidders and the envelope formed by the set of bid-price curves coincides with actual market prices. 2.2 LOCATION AND INCOME The preceeding section presented a very general overview of the main concepts of Alonso's theory. The next step i n the discussion w i l l be the ana-l y s i s of how t h i s theory explains the r e s i d e n t i a l location of low-income groups. The discussion should provide some elements for a future analysis of how the main concepts of economic theory can be applied to explain the role of t r a n s i -t i o n areas on low-income r e s i d e n t i a l location and, i n l a t e r chapters., to derive some possible causes of the g e n t r i f i c a t i o n process. The. following quotes summarize Alonso's view of low-income r e s i d e n t i a l location: "In large American c i t i e s the higher-income residents locate farther away from the center of the c i t y than those of lower incomes. The p r i n c i p a l explanation offered has been that, as the c i t y grows, the wealthier b u i l d t h e i r houses on the vacant land at the periphery, leaving t h e i r old houses near the center to the poorer. This expla-nation i s based on population growth. However, i t can be shown that i n a c i t y which does not grow (that i s , one corresponding to our s t a t i c model), the same pattern of peripheral r i c h and central poor obtains, given certain tastes. .(Alonso, 1964, p. 106 - Emphasis added) "... The peripheral location of the wealthy i n American c i t i e s may be explained without referring to the growth of c i t i e s , and w i l l not be l i m i t e d to rapidly expanding c i t i e s . . . Given a strong appetite for land, so that the holdings of land vary greatly.with income, the wealthier are affected r e l a t i v e l y less by the costs of commuting because they spread these costs over larger holdings. Consequently, the r i c h are price-oriented whereas the poor.are location-oriented. Less a c c e s s i b i l i t y being bought .with increasing income, accessibi-l i t y behaves as an i n f e r i o r .good. This.is a possible explanation of the paradox encountered i n American c i t i e s , , of.the poor l i v i n g on expensive central land and the r i c h on cheaper peripheral land..." (Alonso, 1964., p. 109) The explanation of the central location for the poor.is based on the argument that high incomes lead to more gently-sloped curves, therefore out-bidding i n the market for locations farther away from CBD.. The argument goes as follows: Given a strong appetite for land, the wealthier w i l l tend to have larger holdings. When holdings are large, .a small rate of exchange i n price i s necessary to offset the increased transportation costs. For instance, a man holding one acre needs only a.one-dollar.decrease i n price to offset a one-dollar increase i n commuting costs. A man holding one-tenth of an acre needs a $10 decrease i n land price to offset the;same one-dol l a r increase i n transportation costs. This factor points to a gentler slope for higher incomes. On the other hand, as land holdings increase, a c c e s s i b i l i t y becomes more scarce, therefore more valuable; i t could be expected, then,, that i n d i v i -duals with large holdings would be w i l l i n g to s a c r i f i c e more.land i n order to gain a small change i n a c c e s s i b i l i t y ; t h i s factor would suggest a steeper slope as income increases. The actual steepness w i l l depend, i n the end, on the r e l a t i v e overall value given to land quantity as compared to a c c e s s i b i l i t y . Again, if.we assume that the appetite for land i s strong, i t can be expected that the.increased desire for a c c e s s i b i l i t y w i l l not offset the overall desire ...for large quantities of land. The net effect between these two opposite effects, then, w i l l point to a gentler slope for high income groups, as long as the dominant tastes are for large holdings. In other words: given a strong appetite f o r land, the wealthy, who can afford high commuting costs, w i l l prefer locations where they can purchase 23 large holdings. The poor, who cannot afford high commuting costs, w i l l outbid at central locations; they w i l l be able to do t h i s by s a c r i f i c i n g land quantity, i . e . they w i l l acquire very small holdings at a very high unit price. I f the dominant tastes were reversed, i . e . a c c e s s i b i l i t y were con-sidered as more desirable than land quantity (which i n Alonson's.view would be the case of the pre-industrial town or some Latin American c i t i e s ) , the loca-t i o n pattern would reverse accordingly: the high-income groups would locate i n central areas and the poor would locate i n the periphery: "The American case, where the., peripheral, suburban locations are occu-pied by the well-to-do, corresponds to a slowly decreasing marginal rate of substitution. In the pre-industrial town, where there were sl i g h t d i f f e r e n t i a t i o n i n the size of the l o t s , "the leading trades-men l i v e i n the centre of town, the common people i n the periphery" Similar conditions obtain i n other c i t i e s of the world today i n countries such as India and those of Latin America..." "The centrally located luxury apartments correspond to a minority of the wealthy, who strongly object to commuting and do not want a p a r t i c u l a r l y spacious s i t e . This corresponds to a rapidly decreasing marginal rate of substitution, and consequently steep bid-price curves." (Alonso, 1964, p. 108-109) V Thus, i n Alonso's view, assuming that location i s a trade-off between a c c e s s i b i l i t y and quantity of land, the s p a t i a l d i s t r i b u t i o n by income w i l l be a function of overall preferences between these two consumption goods; the high-income groups would tend to locate where the most preferred of those goods i s maximized, and the lowest income groups w i l l tend to locate where the least preferred of those two goods can be optimized. This notion w i l l be referred to i n future ! chapters, when other factors of location are discussed. 2.3 TRANSITION AREAS: ECONOMIC FACTORS Transition areas located at the fringes of the Central Business Dis-t r i c t are t r a d i t i o n a l l y considered one of the important location options for the poor. A review of the main economic concepts of res i d e n t i a l . l o c a t i o n theory has provided some explanation of the location of low-income groups i n those areas. This explanation assumed.an 'equilibrium' sit u a t i o n , that i s , a 21 situation i n which the t o t a l size of the population.is fixed. However, the location and role of t r a n s i t i o n areas may be also con-sidered- i n a growth situation. The dynamics of• c i t y growth result i n a process of change of use i n those areas, originated by the expansion of the Central Business D i s t r i c t . The economic theory of r e s i d e n t i a l location,. although i t assumes a s t a t i c model, can be used to make some inferences i n terms of popu-l a t i o n growth. This p o s s i b i l i t y has been stated by Alonso: " I t should be noted that, a l l other things equal, with the poor i n the center and the r i c h i n the periphery, sheer population growth w i l l push back the boundary between the r i c h and the poor, i n c l u -ding some of the houses that were r i c h into the poor area. This phenomenon i s quite ancient, as noted i n a 1503 act of Parliament: •Great mischiefs d a i l y grow and increase by reason of. pestering the houses with diverse fami l i e s , harboring of inmates, and converting houses into several tenements, and exacting the erecting of new buildings i n London and Westminster'. However, this, process should not be confused with suburbanization, since overall density is-increasing rather than decreasing." (Alonso, 1964-, p. 110 - Quote from L. Mumford, 'The Culture of C i t i e s ' , 1938, p. 123). The following discussion introduces some growth factors into the equilibrium model, i n order to define some characteristics of t r a n s i t i o n areas. Further background i n terms of the role of growth on r e s i d e n t i a l loca-t i o n w i l l be presented i n the next chapter. In an equilibrium sit u a t i o n , different users w i l l .locate according to the r e l a t i v e steepness of th e i r bid-price curves. In a simple ca.se where two groups are competing, a bid-price system can be depicted as follows: _PIS\AWCE: Diagram 2.3 The firms, with a steeper curve, w i l l locate between 0 and r, and w i l l pay prices ranging from p_^  to p_. Residents w i l l locate between r and t ^ and w i l l pay land prices ranging from p_ to 0. Point p_, i n t h i s case, i s a t r a n s i t i o n point, i n which either of the bidders w i l l locate and both w i l l pay the same price f o r the land. Let us assume now that there i s an overall expan-sion i n the. c i t y : population has grown, and more land i s needed.for the expan-sion of firms. Pressures from growth w i l l result i n an up and outwards s h i f t of the two bid-price curves: CBD ^ fV )p 7 DISTANCE Diagram 2. J+ This s h i f t results i n an overall increase i n land prices. I t also results i n an expansion of the urban land (from to t ^ ' ) . At the same time, the commercial area has grown from 0-r to 0-r'. The new t r a n s i t i o n point, r', i s now farther away from CBD. The area r - r ' can be defined as a t r a n s i t i o n area. 'Transition' i s defined as a process of change over time, from one highest, and best use to another. I t should be noted that what defines a t r a n s i t i o n .area i s not the actual change of use, but the expectation of change. After the change has taken place, the area consolidates with the new use and a new t r a n s i t i o n area appears i n the next outer ring. CHAPTER THREE: SOCIO-ECOLOGICAL FACTORS OF. LOCATION 26 The f i r s t part of the analysis of Transition Areas was a general overview of economic factors of location at a c i t y l e v e l . Starting from the Central Place Theory, a d e f i n i t i o n of Transition Areas was derived, and two types of variables related to the location of the poor i n those areas were described: 1) In an equilibrium situation, location of the poor i n central areas results from a bidding process whereby low-income groups s a c r i f i c e quantity of land i n order to gain a c c e s s i b i l i t y to CBD. 2) In a growth situation, location of the poor i n Transition Areas results from the r e l a t i v e l y lower rental values derived from the expected expansion of CBD and the resulting change i n land use i n Transition Areas. The f i r s t type of variables stems from economic theory. The second part, as w i l l be seen i n this chapter, has been suggested by human ecologists who focus on urban growth as the major variable determining r e s i d e n t i a l location. This f i r s t d e f i n i t i o n of Transition Areas focussed on those factors that relate the location of p a r t i c u l a r individuals to certain physical areas of the c i t y . A l l the'-variables considered (distance to CBD, quantity of land, income...) relate an indiv i d u a l to a location, but say l i t t l e about the so c i a l dimensions of the location game. The purpose of t h i s chapter i s to explore some of the socio-ecological variables related to the role of t r a n s i t i o n areas i n r e s i d e n t i a l location. The discussion w i l l s h i f t from the concept of 'transition areas' to that of 'transition neighborhoods' by focussing on the characteristics and type of change process of the so c i a l groups that tend to locate i n tr a n s i t i o n areas. This chapter contains a discussion of the following issues: 1) A general overview of those theories i n Human Ecology which relate 27 r e s i d e n t i a l location to the urban growth process. 2) A general overview of some of the Vsocio-ecological approaches to resident-i a l location which relate location to soci a l segregation between different groups. The socio-ecological variables introduced i n t h i s chapter add new dimensions to the location game. At the same time, they provide a background for the analysis of neighborhood change that w i l l be discussed i n the follow-ing chapter. 3.1. URBAN ECOLOGY AND CITY GROWTH. The term 'Urban Ecology' was introduced by Park and Burgess i n 1921 and represents an attempt to systematically apply the basic theoretical scheme of plant and animal ecology to the study of urban communities. The f i r s t studies i n urban ecology emerged i n France and England more than a.century ago; those early studies centered on the d i s t r i b u t i o n of crime, suicide and other soc i a l phenomena, and i n i t i a t e d what today i s an important aspect of human ecology: the study of the sp a t i a l d i s t r i b u t i o n of inte r - r e l a t e d s o c i a l variables. Robert Ezra Park (Human Ecology, 1936) argues that a basic process i n human relationships i s competition, largely involving a struggle for space. This competition involves an automatic.and unplanned degree of co-operation,, forming what i s called competitive co-operation. The struggle for existence, based on competitive co-operation, determines the s p a t i a l d i s t r i b u t i o n of persons i n urban areas. The organization of the ecological community i s based upon the dominance.of the Central Business D i s t r i c t . Dominance, competition and human mobility are regarded as resulting i n the process of centralization of services, concentration, segregation, invasion and succession of popula-tions. Burgess ('The Growth of the City: an Introduction to a Research P r o j e c t 1 , 28 1925) developed the famous concept of the f i v e concentric zones of growth: "The t y p i c a l process of the expansion of the c i t y can best be i l l u s -trated... by a series of concentric c i r c l e s , which may be numbered to designate both the successive zones of urban extension and the types of areas differentiated i n the process of expansion." "(A c i t y tends to expand r a d i a l l y ) . . . from i t s CBD ( i ) , to an area of tr a n s i t i o n , which i s being invaded by business and l i g h t manufacture ( I I ) . A t h i r d area ( i l l ) i s inhabited by the workers i n industries who have escaped from the area of deterioration ( l i ) but who desire to l i v e within easy access of th e i r work. Beyond th i s zone, i s the res i d e n t i a l area (IV) of high-class apartment buildings or of ex-clusive ' r e s t r i c t e d 1 d i s t r i c t s of single family dwellings. S t i l l farther, out beyond the c i t y l i m i t s , i s the commuters' zone -suburban areas, or s a t e l l i t e c i t i e s . . . " "This chart brings out clea r l y the main fact of expansion, namely, the tendency of each inner zone to extend i t s area by the invasion of the next outer zone. This aspect of expansion may be called succession, a process which has been studied i n d e t a i l i n plant ecology." Burgess';model i s based on the observation of the s p a t i a l d i s t r i b u -t i o n i n Chicago i n 1920 and has, therefore, a li m i t e d generality. However, his description of the growth and expansion of the c i t y i n terms of expansion, succession and invasion, has had a considerable influence. As w i l l be seen l a t e r , Burgess' notion of location i n terms of con-centric rings successively moving outwards as the c i t y grows and expands, constitutes the theoretical foundation upon which most of the neighborhood change studies are based. Amos H.. Hawley ('Ecology and Human Ecology', 194-4-5 'Human Ecology', 1950) suggests that community structure i s based primarily upon, di v i s i o n of labour. The/organization of sustenance a c t i v i t i e s - the way a community organizes i t s e l f for survival i n a par t i c u l a r habitat - results i n a sp a t i a l d i s t r i b u t i o n . Those a c t i v i t i e s least able to withstand the time and cost of more distant locations and with a maximum need of a c c e s s i b i l i t y w i l l tend toward a central location. Hawley summarizes a theory of land values i n rel a t i o n to re s i d e n t i a l location": "The r e s i d e n t i a l property on high priced land i s usually i n a de-teriorated condition, for since i t i s close to business and i n -d u s t r i a l areas i t i s being held speculatively i n anticipation of 29 i t s acquisition by more intensive and therefore more remunerative land use. In view of that probability owners of such property are not disposed to spend heavily for maintenance or to engage i n new r e s i d e n t i a l construction. Hence the property can command a r e l a t i v e l y low rent for family use Conversely, new resid-e n t i a l structures appear on low-value lands, lands that.have few i f any alternative uses The tendency to high rental valuation i s minimized somewhat by the lowered general a c c e s s i b i l i t y to places of employment and specialized service that greater distance involves. Thus while land values, i n the main, grade downwards with distance from concentrations of associational.units, rental values f o r r e s i d e n t i a l property tend to vary inversely with land values." (Alonso, p. 9-10) The paradox of low income families l i v i n g on high priced land while wealthier families l i v e on cheaper land i s explained i n Hawley's theory i n terms of a process over time i n a growing c i t y . Growth or i t s expectation i s i m p l i c i t i n the behavior attributed to the speculators and i n the expanding area which results from the continuous addition of new houses to the p e r i -phery. The concentric zones models described above explain neighborhood change process i n terms of invasion, succession, extension, and concentration, i n which land use and population type i n one zone i n i t i a l l y invade and then succeed that of the adjacent zone. Change i s envisioned as being much l i k e the dropping of a pebble i n a lake: a series of ripples radiate from the point of impact pushing into each other i n th e i r outward flow. Much of the l i t e -rature on neighborhood change stems d i r e c t l y from the concentric zone .model. Homer Hoyt's sector model (Hoyt; "The Structure and Growth of Residential Neighborhoods i n the United States", 1939) presents the most pro-minent alternative to the concentric rings theory. In a study of 14.2 Ameri-can c i t i e s , Hoyt argued that the d i s t r i b u t i o n of population i s organized more i n terms of homogeneous, pie-shaped sectors that run from the c i t y ' s CBD toward the periphery. Thus, the high-income areas tended to be found i n one or two major sectors rather than i n the concentric zones described by Burgess. The soc i a l status of an area would not be related to distance from CBD; indeed, the types of neighborhoods that one would pass through i n 30 driving from the center of the c i t y would ba s i c a l l y depend on the sector i n which one started. The sector theory explains growth i n terms of a x i a l development along major transportation arteries and l i n e s of least resistance. The key element i n change i s that of high-status residence. Its location and expansion sets the parameters for the development of the other areas. Some'df the basic pro-positions describing sectoral change are (Schwirian, 1977, p..185-186): - High-rent r e s i d e n t i a l growth tends to proceed from the given point of o r i g i n , either along established l i n e s of travel or toward another existing nucleus of buildings or trade areas. - The zone of high-rent tends to occupy high ground that i s free from r i s k of floods and to spread along lake, bay, r i v e r , and ocean ports where the water-front i s not used by industry. - High-rent r e s i d e n t i a l d i s t r i c t s tend to grow toward the section of the c i t y that has free open country beyond i t s edges and away from 'dead end' sections that are prevented from expanding by natural or a r t i f i c i a l barriers. - The high-priced r e s i d e n t i a l neighborhood tends to grow towards the home of community leaders. - Sometimes trends i n the construction of. o f f i c e buildings, banks., and stores p u l l the high-priced r e s i d e n t a i l neighborhoods i n the same, general.direction. - High-grade r e s i d e n t i a l areas tend to develop along the fastest existing \ transportation l i n e s . - Deluxe apartment areas tend to develop near the business centers i n old es-tablished r e s i d e n t i a l areas. - The growth of high-rent neighborhoods continues i n the same direction for a long period of time. - High-rent neighborhoods do not skip about at random i n the process of move-ment. They follow a d e f i n i t e path i n one or more sectors of the c i t y . - I t i s possible, under certain conditions, for high-rent areas to 'double-31 back', or return toward the center of the c i t y . - High-rent areas tend to be adjoined by medium-rent areas; sharp disjunctions i n rental areas are not frequent. Harris and Ullman ("The Nature of C i t i e s " , 194-5) suggest s t i l l , another alternative pattern of c i t y geometry, the multiple nuclei model. Change i n the multiple-nuclei model i s treated in. terms of the .factors that account for the emergence of separate nuclei within the c i t y . The four factors that are involved i n the development of the nuclear pattern are.: 1.) Certain a c t i v i t i e s require specialized f a c i l i t i e s located i n only a few.sections of the metropolis. 2) Certain l i k e a c t i v i t i e s p r o f i t from adjacent congregation. 3) Certain unlike a c t i v i t i e s are antagonistic or detrimental to each, other. 4) Certain a c t i v i t i e s are unable to afford the costs of the most desirable loca-t i o n a l sites.. The obvious question at th i s point i s : Which of the three different models best f i t s the actual geometry of most c i t i e s ? Although there i s no clear answer to th i s question and researchers have different.opinions on this matter, some additional elements related to the shape of urban, areas can be drawn from the analysis of the ^segregation process between .diff.erent s o c i a l groups. This w i l l be discussed i n the following section. 3.2 SOCIAL FACTORS OF SEGREGATION. From the early work to the most recent formulations, Human Ecology has seen r e s i d e n t i a l location as a process of segregation between different groups: "By i t s very nature, urbanization has .led to.the r e s i d e n t i a l segregation of persons and groups within c i t i e s . .Given large concentrations of population, given the heterogeneity.of i n d u s t r i a l society,, and given differences i n l i f e s t y l e and culture, i t i s almost inevitable that persons w i l l sort themselves from other .persons on. the,basis of sub-j e c t i v e l y important c r i t e r i a . I t would indeed be amazing to f i n d urban concentrations where a l l neighborhoods or blocks or c i t i e s had exactly the same types of persons." (Avery Guest, p. 269). 32 A starting point for the analysis of r e s i d e n t i a l segregation - i t s degree, s p a t i a l patterns, and i t s causes - i s tlie.Burgess theory of urban growth. Burgess' ring concept seemed to be primarily a h e u r i s t i c device to show how a c t i v i t i e s and groups become segregated, concentrated, and centralized i n the c i t y . The zones were meant to indicate general tendencies of location, rather than hard and fast areas. I t follows from his theory than lower status persons, single and unattached persons (including homeless men and women), and ethnic and r a c i a l minorities should generally be located near the center of the c i t y (centralized), while higher status persons, third-generation Americans and families with children should generally be located on the urban outskirts (decentralized). Within Burgess' model, groups become segregated i n the c i t y , not because they have a particular l i k e or d i s l i k e for other gorups, but because changes i n the c i t y ' s structural or morphological features lead them to l i v e i n certain areas of the c i t y . Within the context of possible causes of segregation, the theory emphasizes the non-personal causes-Following Burgess's framework, Zorbaugh developed the concept of 'natural areas', which can be considered a f i r s t attempt to explain the nature of neighborhoods. A 'natural area' i s a geographical area characterized by a physical i n d i v i d u a l i t y and by the. c u l t u r a l characteristics of the people who l i v e i n i t : r "A natural area i s a geographical area characterized by a physical i n d i v i d u a l i t y and by the c u l t u r a l characteristics of the people who l i v e i n i t . . . Just as there i s a plant ecology whereby, i n the struggle for existence, l i k e geographical regions become associated with l i k e 'communities' of plants, mutually adapted, and adapted to the area, so there i s a human ecology whereby, i n the competition of the c i t y and according to definable processes, the population of the c i t y i s segregated over natural areas into natural groups. And these natural areas and natural groups are the 'atoms' of c i t y growth, the units we t r y to control i n administering and planning the c i t y . " (Zorbaugh, 'The natural areas of the.city', 1926, i n Theodorson, 'Studies i n Human Ecology, 196l, p. 4-7). Zorbaugh sees these natural areas as human communities with similar t r a i t s which, i n the process of struggle for survival, become adapted to a p a r t i c u l a r geographical area. Again, the adaptation process relates to the geographical characteristics of each area, rather than to the interaction between members of these communities. A more formal framework which provides methodological tools to iden-t i f y the s o c i a l variables that define different s o c i a l groups i n urban areas has been more recently developed by Schevsky and B e l l (Schevsky and Bell,. 'The Social Areas of Los Angeles', 194-9). Social Area. Analysis i s a method for the study of s o c i a l differenciation and s t r a t i f i c a t i o n of groups. The o r i g i n a l formulation of t h i s method was presented i n a study of urban phenomena i n Los Angeles. The objective of the study was to understand urban aggregations, not as isolated, self-contained units, but as parts of a wider, system or re-lationships; i t was f e l t that the factors that define different groups i n urban areas were a r e f l e c t i o n of the more si g n i f i c a n t characteristics of modern society. Starting from three main postulates concerning i n d u s t r i a l society, Schevsky and B e l l derived a set of indexes, obtained from census tract data, which relate to each of the postulates. The three main constructs are: 1) Social Rank (economic status), which would r e f l e c t the changing d i s t r i b u t i o r of s k i l l s as a s i g n i f i c a n t d i f f e r e n t i a t i n g factor among individuals i n modern society, and would be measured by education, employment status, value and quality of home. 2) Urbanization (family status), which would be a consequence of the increas-ing d i f f e r e n t i a t i o n of function and resulting changes i n the ways of l i v i n g (women entering the labor force, alternative family patterns). Indicators of t h i s index are: age and sex, persons per household, owner-ship status, house structure. 3) Segregation (ethnic status), derived from the increase i n the complexity of organization: (increasing d i v e r s i t y , segregation and i s o l a t i o n of. groups) and measured by race, country of b i r t h , citizenship. I t should be noted that the term 'segregation' here i s used i n a r e s t r i c t e d meaning, since i t describes d i f f e r e n t i a t i o n by ethnicity only. Two hypotheses i m p l i c i t i n the formulation of this typology were tested i n the Los Angeles study: 1) that the three types are necessary to account for the observed s o c i a l d i f f e r e n t i a t i o n between urban sub-populations; and 2) that the indexes constructed to measure the three factors are unidi-mensional measuring instruments. The results of the analysis supported these two hypothesis and, i n addition, provided some evidence that the three factors are adequate, as well as necessary, to account for most of the observed variation between sub-populations that can be detected by census data. Van Ardsol, C a m i l l e r i , and Schmid have carried out an extensive study i n which soci a l area analysis was applied to ten large American c i t i e s . The results of t h e i r study led them to conclude that t h i s approach has high generality and i s highly applicable to the c i t i e s studied. In eight of the ten c i t i e s studied the data supported the construction of the variables of so c i a l rank, urbanization and segregation on the basis of the s p e c i f i c indexes u t i l i z e d by Shevsky and B e l l . A study of certain areas of San Francisco, done by B e l l , provides further evidence on how s o c i a l area analysis, provides a framework for the i d e n t i f i c a t i o n of the: main d i f f e r e n t i a t i o n factors i n urban groups. (Theodorson, 'Studies i n Human Ecology 1, 1966, p. 132). Social Area Analysis provides a useful set of indicators.to measure the characteristics that define sub-urban groups i n urban areas. These sub-groups, or s o c i a l areas, are groups with similar scores on three s o c i a l dimensions:, s o c i a l rank, family status and ethnicity. Although these i n d i -cators do not describe the type of interaction that takes place between different individuals or groups, i t can be hypothesized that people.living i n a p a r t i c u l a r type of s o c i a l area would systematically d i f f e r with respect to attitudes and behavior from persons l i v i n g i n other types of s o c i a l areas. Schevsky 1s; and B e l l ' s typology defines .segregation to refer only to 35 one type of group d i f f e r e n t i a t i o n , which i s ethnicity; but i n fact groups separate themselves according to other factors as well. Avery Guest gives a broader meaning to the concept: (Avery Guest, 'Residential Segregation i n Urban Areas'. "Contemporary Topics i n Urban Sociology, 1977, p. 269-335). "Segregation means d i s s i m i l a r i t y i n r e s i d e n t i a l d i s t r i b u t i o n -that i s , the tendency of one type of person to l i v e separately from another type of person i n one small area i n the c i t y . . . . The opposite to segregation would be integration, and a process i n which segregation i s abolished would be called desegregation". "Segregation... may also take the form of concentration. I f mem-bers of a group are segregated so that a l l l i v e i n one region- or large area of the c i t y , they are concentrated. In the case of black-white segregation, the existence of large r a c i a l ghettoes could be considered an example of concentration. But a group may be segregated without being concentrated, as would be indicated by a pattern on c i t y blocks being completely white or completely black, but these blocks being randomly distributed around the c i t y . . . " (Guest, 1977, p. 273-274). Guest differentiates three dimensions of segregation: by so c i a l status, by family composition and by race/ethnicity. These three dimensions are more or less coincident with Schevsky's and Be l l ' s typology; the only difference being the d e f i n i t i o n of segregation as a general phenomena, and ethnicity as only one dimension of i t . In terms of the reasons for segregation, Guest distinguishes two types of forces: 1) Non-personalistic forces, which produce an ind i r e c t type of segregation, result of physical features and characteristics of the c i t y , rather than the attraction of rejection between different s o c i a l groups. Non-personal segregation i s considered as a product of two types of causes: Sentiment and symbolism attached to certain areas and elements of the c i t y . This factor relates to elements such as h i s t o r i c a l character of a neighborhood, traditions associated with a pa r t i c u l a r area, etc. - Morphological or structural features of neighborhoods, such as size and age of the housing stock, provision of open spaces. 2) Personalistic forces, which are conscious acts of individuals or groups 36 to l i v e close or far away from each other. Personal segregation would be due to: - The actions of individuals i n terms of choosing to l i v e with groups having similar values and l i f e s t y l e s . - The actions of i n s t i t u t i o n s or specialized agencies within the communi-ty; for instance, i t i s frequently allegued-that r a c i a l segregation i n US'.-cities, while sought by some individuals, i s ac t i v e l y created by the actions of realtors who may refuse to show certain neighborhoods to blacks. Guest argues that, although l i t t l e i s known about the causes of re s i d e n t i a l segregation, the evidence suggests that family segregation and segregation by soci a l rank are primarily a result of non-personal factors, while ethnic-racial segregation results primarily from the desire to l i v e close or far away from other groups. A summary of the argument presented by Guest i s given below. A) Family-l i f e cycle segregation: In Guest's view, segregation by stage i n l i f e - c y c l e i s the direct result of the part i c u l a r housing needs of each group i n terms of size and age of the housing. These housing needs can be summarized by the following categories: Households of married couples with children present: this group should be located i n spacious housing, and r e l a t i v e l y new housing since they can be expected to move into houses when they start the child-bearing period and remain i n them for a long period of time. Since most, of the new housing i s added at the periphery of the..urban areas, families i n the c h i l d -bearing stage would locate i n those suburban single-family dwellings. Households of married couples i n the age period over L5, generally without children present: people i n th i s group would locate i n spacious older housing, " i f only as a r e l i c of past housing needs"; they would age along with the neighborhood. Because of the older type of housing, they would be 37 located i n more central areas than the young families. - Households i n which no or very few married couples or children are present, generally containing the young, single population, the-jolder, widower popu-l a t i o n , and tfeodivorced-separated population. This group would need less spacious dwelling and should tend to locate i n higher density areas; the old age component may also l i v e i n old, spacious units i n central neigh-borhoods. In both cases, th'eslocation Is close to the core of the c i t y . Since t h i s group represents a r e l a t i v e l y shorter stage i n the l i f e - c y c l e , one can predict that apartment house areas of the c i t y would be more transient i n character (higher mobility rate). In t h i s view, segregation by family status i s primarily non-personal, and results from structural or morphological features of neighbor-hoods rather than from sentimental features. Several studies done i n recent years support t h i s notion. For instance, Peter Rossi (1955), i n a study of r e s i d e n t i a l mobility i n Philadel-phia, found that most of the families moving or planning to move were doing so to adjust housing needs with l i f e - c y c l e changes. The space, quality, and structural soundness of housing was the overwhelming concern i n t h e i r housing choice. (Guest, 1977, p. 286). A study of r e s i d e n t i a l segregation by l i f e -cycle stage, i n Cleveland, Ohio (i960) showed that differences by family status did not explain segregation; most of the segregation could be explained•by. differences i n housing types (Guest, 1977, p. 287). S i m i l a r l y , Guest's study of segregation by family types i n 16 other metropolitan areas (1970), showed similar findings. Guest also found that the centralization-decentra-l i z a t i o n pattern was related to family status rather than s o c i a l rank or ethnicity: "In fact, the centralization-decentralization pattern was much more characteristic of families than of s o c i a l status groups, white ethnic groups or r a c i a l groups, suggesting that family composition i s the most clear-cut characteristic distinguishing the center from the outskirts of the American metropolis. On the whole, the study of the seventeen metropolitan areas showed •38 l i t t l e tendency for family segregation to vary i n clearcut concentra-tion patterns other than centralization-decentralization." B) Social class segregation: The issue of segregation by s o c i a l rank has provoked a wide-ranging debate on whether i t results from personalistic or non-personalistic causes. More has been written about segregation by s o c i a l rank than by any other dimension, but less i s probably known d e f i n i t i v e l y . There has also been a long-standing debate on whether s o c i a l class groups are segregated on the basis of centralization-decentralization, or whether i t follows other patterns, such as Hoyt's wedge-shaped areas emanating from the CBD. Exponents of the non-personal, structural factors i n determining soci a l class segregation argue that persons of higher s o c i a l status have certain tastes and desires f o r certain types of neighborhoods, and use t h e i r income to purchase location according to those tastes; those preferred neigh-; • borhood attributes would be primarily: l i t t l e congestion, l i t t l e business and industry, generous amount of internal and open space. Though not always ex-p l i c i t , i t follows from this argument that the lower status persons would locate not according to t h e i r tastes but i n those areas not chosen by the higher status groups. The arguments presented by Guest i n support of.this theory are based on a study done by Otis and Beverly Duncan of occupational.segregation i n the Chicago Metropolitan D i s t r i c t i n 1950. This study was replicated i n .various other metropolitan areas. A l l those studies showed that segregation indexes by occupation were moderate (greater than those for family groups, and similar or smaller than those for ethnic groups). Segregation was found to be generally correlated to income and occupational prestige. Another study done by Guest on 17 SMSA (Guest, 1971) shows that the proportion of white-collar workers i n each area (as opposed to blue-collar workers) was highly related to the types of housing, p a r t i c u l a r l y number of rooms and structural conditions of the house. 39 How t h i s evidence supports the non-personalistic theory i s not clear i n Guest's discussion. The study by Duncan and Duncan appears to show that there i s segregation by s o c i a l status, but says l i t t l e about i t s causes. Guest's study, on the other hand, relates location to differences i n housing types by occupational class; the correlation between housing type and occupa-t i o n a l class does not explain whether housing quality i s a cause, an effect or a concomitant factor to location preferences. Another type of non-personalistic location factors commonly associa-ted with s o c i a l rank i s that of sentiment and symbolism attached to certain neighborhoods. One leading proponent, Walter Firey ('Sentiment and Symbolism as Ecological Variables', 194-5), maintains that space has a symbolic value, not only cost-imposing q u a l i t i e s . Space takes meaning for man through c u l t u r a l d e f i n i t i o n , and c u l t u r a l values intervene between the physical environment and the human community. Firey suggests that the maintenance of Beacon H i l l , i n central Boston as an upper-class r e s i d e n t i a l d i s t r i c t can only be explained i n terms of symbolic values. In Guest's view, however, the Beacon H i l l case would be a residual, rather than a dominant phenomenon, since the overall pattern i n Boston, with only that exception, i s s t i l l that of -centralization-decentralization-by 'social rank. S t i l l another c r i t i c i s m of the ecological view sustains that actions of groups may result i n class segregation. Higher status neighborhoods would defend the i r t e r r i t o r y and prevent the entrance of others by c o l l e c t i v e behav-i o r , or by i n d i r e c t means such as r e s t r i c t i v e zoning and land use controls. In terms of geographical location by s o c i a l rank, current evidence seems to suggest that income-class variations with distance from CBD.are not very s i g n i f i c a n t . Rather, segregation would follow patterns related to the location of transportation corridors, i n d u s t r i a l areas and other geographical features. 40 In Guest's words: "As I have shown recently (Guest 1971, 1972b), the actual tendency of s o c i a l status to vary with each mile of distance from CBD i s not great, and thus we are r e a l l y trying to explain s l i g h t tendencies, i f present at a l l , of status to increase with distance from CBD." (Guest, 1971, p. 299). Kent P. Scwirian (p. 187-193) shows that, i n a study of US c i t i e s i n I960 by Schnore and Winsborough (1972), while 90 c i t i e s showed a centralization pattern (higher status i n outer rings), 24 communities showed a reverse . gradient (higher status i n central areas), and 70 urbanized areas showed .a pattern i n which both the lowest and highest classes were centralized while middle-class was decentralized. The same study, though, showed that changes i n time pointed to upper-class groups becoming less centralized and lower groups becoming more centralized. The following quote from Scwirian £p-—192-)- gives an interesting insight on how the existing evidence might be suggesting a de-velopmental theory of land use: "Schnore speculates that there might well be a sequential pattern of ecological change as societies modernize. In the early pattern of c i t y development the 'inverse gradient' characterizes the c i t y i n which the affluent population l i v e s near the c i t y ' s center and the lower groups l i v e towards the periphery. With growth, modernization, and aging the central portions of the c i t y become less attractive r e s i d e n t i a l l y and the upper classes migrate towards the periphery. The vacated areas i n the inner portions of the c i t y become targets for those urbanites who cannot survive the s t i f f e r competition for fringe housing." (Schwirian, 1977, p. 192) In other words, the.*location d i s t r i b u t i o n of highest and lowest groups, or the r e l a t i v e centralization of the:poor, would be an indicator of the degree of urbanization of an area. C) Ethnic segregation: U n t i l recently, most sociologists treated.ethnicity as.an aspect of s o c i a l rank. Ethnic i d e n t i f i c a t i o n was seen as a temporary manifestation of the migrant's low economic status i n American society. And i t was presumed that d i s t i n c t i v e ethnic t r a i t s would disappear as the migrant was assimilated economically into the society. -41 ' This perspective, i n seeing status and ethnicity as very s i m i l a r , seems to follow the view that r e s i d e n t i a l segregation arises primarily from nonpersonal causes. Ethnic groups would become segregated primarily because th e i r income status relegate them to certain types of housing and neighborhoods i n the c i t y . However, there may be certain personalistic components of segregation .because the ethnic group, due to i t s low status, seeks out others of the same ethnicity i n order to protect and defend i t s interests. More recent research seems to suggest, however, that segregation by ethnicity cannot be explained by income alone. For instance, Nathan Kantrowitz (1973). (Guest, 1977, p. 304.) suggests, using the New York metropoli-tan area as an example, that a high degree of re s i d e n t i a l segregation remains among the foreign stock population of New York. This view i s supported i n two studies of r e s i d e n t i a l segregation i n the Toronto metropolitan area: In one study, A. Gordon Darroch and W.G. Martson (1971) found that segrega-tion among ethnic groups i n Toronto could be explained only slightly, by the soc i a l class differences of the groups. In another study, Darroch and Nartson (1969) argued for a multidimensional concept of ethnic .residential segregation, showing that ethnic segregation could be explained by a varie-ty of factors: period of immigration to Canada, r e l i g i o n , mother tongue, and country of b i r t h . They showed that ethnic segregation could not be explained by period of immigration to Canada alone, indicating .that i t was not simply an a r t i f a c t of having new, low status immigrants. (Guest, 1977, p. 305). - The actual location patterns a r i s i n g from ethnic segregation are not well understood. There i s no clear evidence to prove that ethnicity produces either a centr a l i z a t i o n pattern, a sectoral pattern, or some other type of clustering. Some studies suggest that ethnic groups s p a t i a l l y concentrate on non-personal symbolic bases. For instance, Christen Jonassen (1945), (Guest, 1977, p. 307) argues that Norwegians i n New York City have located L2 close to the sea and semi-rural areas, because they remind Norwegians of their native land. Similarly, a study by Kosa ('Hungarian Immigrants in North America: their Residential Mobility, and Ecology', 1956) suggested that location and movement of Hungarian groups could be explained in terms of their cultural values. A good summary of what seems to be the current view in terms of residential segregation in the three dimensions suggested by the Social Area Analysis typology i s K.P. Schwirian 'club sandwich' theory: -(-Schw-i-rian-,—1-9-7-7T .p..~-1987"' "While the social area analysis perspective developed f a i r l y inde-pendently of the classic models there has been some interest i n investigating the patterns of urban social areas in terms of con-centric zones, sectors, and multiple nuclei. In studies of cities in the United States and Canada i t was found that, overall, status tended to be distributed more in terms of sectors than distance gradient. Familism was distributed more by distance gradient than by sector. The main reason for this i s that the familism's index major component i s the relative concentration .of single-dwellings . as opposed to multiple-dwelling units. New housing.at the fringe tends to be single-family while other neighborhoods in the urban . core are being converted from single-family to multiple-dwelling units. Ethnicity was distributed neither by sector nor by distance but tended to appear more in terms of highly clustered nuclei... Thus i t seems that the city i s much like a multilayered club, sand-wich. Over the basic grid of streets we have a layer of social status sectors; then a layer of housing types distributed primarily by density gradient; a l l topped off by a layer of nucleated racial and ethnic minorities. Social Area Analysis attempts to separate these layers and examine their relative contribution to the city and then recombine them to observe their total configuration." (Schwirian, 1977, p. 198) CHAPTER FOUR: DYNAMIC FACTORS OF LOCATION 43 The review of location theory presented i n the preceeding chapters has centered on economic and socia l factors related to the centralization of low-income groups around the Central Business D i s t r i c t . From the point of view of economic theory, t r a n s i t i o n areas were defined as those areas i n which a change i n land use process i s underway, this, change being produced by the dynamics of growth and the resu l t i n g expansion of the Central Business D i s t r i c t . Growth, and the expectations of land use change derived from i t , were suggested to result i n a deterioration of mainte-nance standards i n t r a n s i t i o n areas; t h i s i n turn would result i n r e s i d e n t i a l location of low-income groups i n those areas. On the other hand, the overview of some of the socia l . components of re s i d e n t i a l location suggested that centralization of tha.-poor • i n areas around the Central Business D i s t r i c t i s one of the so c i a l dimensions i n which the location game takes place. Other types of so c i a l segregation (by family status and ethnicity) are l i k e l y to'result i n the centralization of pa r t i c u l a r family types and ethnic groups. Since the main element defining t r a n s i t i o n areas i s the change pro-cess taking place i n them, a key element to understand t h e i r role i n terms of low-income residential, location i s the analysis of th i s change process,, i t s causes and mechanics, and the effects of the process on r e s i d e n t i a l segrega-t i o n . Studies of location dynamics stem primarily from the same body of so c i a l ecology presented i n Chapter Three (section 3.1). The basic notion underlying a l l the studies of location dynamics i s that, as the c i t y grows and expands, new housing i s added i n peripheral areas. Housing additions, com-bined with the expansion of the Central Business D i s t r i c t , produces an outward flow of the urban residents. This outward flow has been described either i n terms of successive u concentric rings around CBD (as described by Burgess), or i n terms of a x i a l movements along different sectors of the c i t y (as described by Hoyt's sector model). In both cases, they assume that different groups move to successively outward areas as the c i t y expands. This chapter contains a discussion of some of the dynamic factors that influence the .role of t r a n s i t i o n areas i n terms of low-income r e s i d e n t i a l location. The discussion centers on two approaches to r e s i d e n t i a l mobility: 1) The F i l t e r i n g process, which explains mobility from the point of view of housing quality. The f i l t e r i n g process i s considered to be a market mechanism whereby low-income residents have access to housing. 2) The Neighborhood Change process, which includes those theories that des-cribe physical and s o c i a l components of neighborhood change. 4.1 THE FILTERING PROCESS. One of the concepts underlying p o l i c i e s of maintenance subsidies f o r deteriorated housing i s the f i l t e r i n g down concept. The basic.idea of f i l t e r -ing i s that additions to the housing stock w i l l start a chain of moves i n which each household, of the chain moves to a better housing; successive users of each building belong to lower income groups. At the end, of the process, the units are ' f i l t e r e d ' to the lowest income group. This process i s consi-dered to be the mechanism whereby the poor acquire housing under free market conditions. In t h i s scope, housing maintenance p o l i c i e s assume that f i l t e r i n g i s an e f f i c i e n t mechanism to provide housing for the poor. Subsidies for maintenance have the objective of improving the quality of the housing thus provided, and at the same time these p o l i c i e s are supposed to protect the physical and s o c i a l f a b r i c of neighborhoods by preserving theold, deteriorated buildings. The analysis of f i l t e r i n g i n the following section w i l l center on 45 the issue of how f i l t e r i n g provides housing for the low-income groups. This section contains: - A review of different definitions of f i l t e r i n g ; An analysis of the mechanics of f i l t e r i n g and i t s effects on low-income housing. 4.1.1 DEFINITIONS OF FILTERING. The concept of f i l t e r i n g can be traced back to Hoyt's sector theory. Hoyt's essential notion was that, as a c i t y grows, the fashionable, r e s i d e n t i a l d i s t r i c t moves outwards from the center, always i n the same direction. Obso-lete houses l e f t behind by the well-to-do become occupied by the poor, p a r t i -c ularly the new immigrants, so that the wedge-shaped path taken by the high-income group contains both extremes i n the income spectrum. Middle-income families group themselves about the well-to-do, hoping that some of the status of the fashionable areas w i l l rub off to them (Wallace Smith, ' F i l t e r i n g and Neighborhood Change; 1970, p. 65). William Grisby (The .Filtering Process, 1963) provides a good review of alternative definitions of f i l t e r i n g : A) The t r a d i t i o n a l d e f i n i t i o n (Radcliff, Urban Land Economics, 1949) relates changes i n housing price to changes i n occupancy: "This process ( f i l t e r i n g ) i s described most simply as the change i n occupancy as the housing that i s occupied by one income group becomes available to the next lower income group as a result of decline i n market price, i . e . i n sales price or rents value." ,(Grisby,1963, p.96) This d e f i n i t i o n assumes that rent and occupancy changes occur simul-taneously and are simply two sides of the same coin. Of these, the key ele-ment i s considered to be occupancy change, the value change being the permis-sive factor. B) A reformulation was.provided by Fisher and Winnick (Ernest Fisher and Louis Winnick,. 'A Reformulation of the F i l t e r i n g Concept', 1951). This d e f i n i t i o n recognizes that change i n price and change i n occupancy do not 46 always occur simultaneously: " F i l t e r i n g i s defined as a change over time i n the position of a given unit or group of dwelling units within the d i s t r i b u t i o n of ' housing prices and rents i n the community as a whole." (Grisby, 1963, p. 96) Thus, f i l t e r i n g i s measured i n terms of change i n r e l a t i v e value rather than i n terms of changes i n occupants. C) Grisby suggests s t i l l another concept - that of f i l t e r i n g as an improvement i n r e a l housing standards among low-income fami l i e s . This welfare concept suggests that f i l t e r i n g occurs only i f low-income housing standards have been raised: "Such a d e f i n i t i o n would hold that f i l t e r i n g (change i n housing prices and rents) must be measured while holding income, quality and space per person constant, or i n more relaxed form, that f i l -t ering occurs only when value declines more rapidly than quality so that families can obtain either higher quality and more space at the same price, or the same quality and space at a lower price than formerly." (Grisby, 1963, p. 97). Grisby's d e f i n i t i o n also excludes the change i n occupancy element. The only difference between th i s and Fisher's concept seems to be the welfare notion that f i l t e r i n g occurs only when a rea l improvement i n housing for the poor occurs. As such, i t does not consider the concept of ' f i l t e r i n g up* (or change to a lower quality housing) a.s.part of the f i l t e r i n g concept. D) Wallace Smith ( F i l t e r i n g and Neighborhood Change, 1970) argues that the key element i n the d e f i n i t i o n of f i l t e r i n g i s the change i n occupancy, rather than the quality change. He approaches the concept of f i l t e r i n g i n terms of market behavior: "Optimization i n the use of a durable good requires s h i f t i n g i t about among different classes of users as i t s r e l a t i v e usefulness declines or r i s e s . The.shifting about, i n i t s e l f , affects welfare only as i t i s an e f f i c i e n t or an i n e f f i c i e n t reallocation... F i l t e r i n g i n res-ponse to deterioration of the housing stock could well leave aggre-gate and indiv i d u a l housing welfare unchanged." "I f f i l t e r i n g i s to be understood as a basic pattern of market be-havior, then i t must be defined as a response to any change i n condi-tions of supply or demand - i n the number or types of households and their incomes, i n the physical quality of the stock or any portion of i t , or i n construction of pa r t i c u l a r types of new units. The res-ponse must be measured i n terms of occupancy of pa r t i c u l a r houses or neighborhoods by some different class of households." 47 " F i l t e r i n g . . . i s simply a change i n the allo c a t i o n of housing units among households of different types." (W. Smith, . 1 9 7 0 , p. 74--75) In summary, the concepts of f i l t e r i n g discussed above seem to des-cribe two d i s t i n c t types of processes which, although interelated, do not always occur simultaneously. The f i r s t approach, or market approach, defines f i l t e r i n g as a process of change i n the allo c a t i o n of the housing good to d i f -ferent submarkets; t h i s approach centers on the housing unit i t s e l f and the way i t changes i n occupancy within a market system. The second d e f i n i t i o n , or welfare approach, centers on the relationship between housing quality, neigh-borhood characteristics and overall housing standards. In the analysis that follows, the two concepts w i l l be used at. d i f -ferent stages. . The f i r s t part of the analysis (mechanics of f i l t e r i n g ) w i l l be based on the market concept of f i l t e r i n g . The welfare approach w i l l be discussed at a future stage, as part of the. factors determining neighborhood change. 4.1.2 MECHANICS OF FILTERING. Wallace Smith defines f i l t e r i n g as a basic pattern of market behav-i o r . As such, he argues that f i l t e r i n g i s the market response to changes i n conditions of supply and demand: changes i n the number or types of households, changes i n income., changes i n the physical quality of the housing stock or portions of i t , construction of pa r t i c u l a r types of new units. Each of these external changes would have an effect i n terms of housing turnover. The analysis that follows centers on the effect of each of these external factors on. the process of housing turnover. The analysis w i l l attempt to show how the f i l t e r i n g process works i n terms of providing housing for the poor, and how t h i s process i s affected by subsidies to housing main-tenance. General assumptions for the analysis are: A) An hypothetical community consisting of three families, d i f f e r i n g as to income but hot otherwise, and three houses d i f f e r i n g as to quality. B) An i n i t i a l a l l o c a t i o n i n which the lowest-income family occupies the lowest quality house (house 1 i n diagram 4.1) and the highest-income family occupies the highest quality house (house 3 i n diagram). ( l ) O O O Diagram 4.1 C) Housing quality, refers to the building characteristics only. There are no location differences between the dwelling units. The analysis i n t h i s sense would be similar for the all o c a t i o n of any other durable goods (cars for instance) that can be resold i n the.market.. D) High-income household i s defined as a household which has access to new housing; low-income household i s a household which cannot afford new housing; middle-income household i s defined as a household.which cannot purchase new housing but can purchase a high-quality accommodation. E) A l l the households are home-owners. A discussion on some eventual effects on tenants w i l l be included at a l a t e r stage. The f i r s t part of the analysis centers on some l i k e l y effects on housing moves of the following variables: housing deterioration; income changes; population changes, and additions to the housing stock. The second part of the analysis will.introduce rehab as an additional variable i n the process. (•i)-'-< *,»..*A discussion of why t h i s i s an optimum market d i s t r i b u t i o n can be found i n Wallace Smith, 1970, p. 77-78.," In a no-rehab situation, then, the following situations can be expected: Case I : Housing deterioration. (Population and income remain constant.) I f the change i s just, a time change, the^overall housing stock can be expected to. deteriorate due to age alone. Housing deterioration w i l l produce, d i s s a t i s -faction with the existing a l l o c a t i o n and pressure to add a new housing unit i n the market. The effect of deterioration can be depicted as follows: i z 3 W E U J A S A N D O M E D HOUSE H O U S E 'Diagram L.2 The new unit added w i l l be purchased by the highest-income house-hold. This w i l l start a chain of moves whereby each household w i l l move to a higher quality housing and the l a s t house i s eventually abandoned. In t h i s case, there has been a f i l t e r i n g process but no r e a l improvement of the overall housing sit u a t i o n : the effect of f i l t e r i n g i s that of recuperating the housing standard existing before aging. I f no units are added to the market, aging, with no investment i n rehab results indeed i n an overall deterioration of housing standards. Case 2: Income increases. (Time and population are held constant.) Overall income increases can be expected to resu l t i n dissatisfaction.with the exist-, ing housing a l l o c a t i o n , since each household can afford higher housing quality. To the extent that t h i s pressure produces the construction.of a new dwelling unit, the effect can be depicted as i n the following diagram: 50 o o IB 1 2 3 UEAJ HOUSE MOUSafc Diagram 4.3 I t should be noted that i n t h i s case, though income i s the ^ triggering element, the direct cause of the improved housing situation, i s i n fact the f i l t e r i n g process; to the extent that income increases are spent on other goods, there i s a housing improvement i n terms of a f f o r d a b i l i t y , but no r e a l improvement i n housing quali t y . Thus, income increases have a positive effect i n terms of housing improvement through f i l t e r i n g . I t i s worth pointing out that t h i s improvement occurs regardless of the d i s t r i b u t i o n of the increased income.... Taking the extreme case: only the high-income household has a higher income, i t can be expected that t h i s household w i l l be w i l l i n g to s e l l his former.unit at a r e l a t i v e l y lower price i n order to purchase the new unit i n the market; the lower price w i l l make t h i s unit more affordable f o r the next...household, which i n turn can lower his price by a f r a c t i o n of his gain; thus, the effect of higher income can be expected to percolate to at least some extent a l l the way down to the bottom group. On the other hand, i f only the bottom household has an income im-provement, t h i s household w i l l be w i l l i n g to pay a premium for the next housing i n the chain; the middle-income household w i l l f i n d that he can s e l l . at gain and spend part of th i s gain i n offering a premium price, f o r the highest quality house; t h i s i n turn w i l l improve the high-income household's p o s s i b i l i -ty to purchase a new unit. 51 Case 3: Population increase, income and age remain constant. The effect of population growth varies, depending of the income group of the new households. Let us consider,them case by case: 3a) New household, high income: Population growth creates pressure for addi-tions i n the housing stock. I f the new unit i s added, the effects arenas follows: o o HOUSE.HO House Diagram 4«4 The new household w i l l compete with the existing top-level household to purchase the new unit; whoever gets the new unit, the ' f i l t e r i n g ' , i f any, i s blocked at the top l e v e l , and the chain w i l l not continue.. At the same time, the bidding for the new house w i l l raise the price of t h i s ..unit. The owner of the f i r s t 'existing' house w i l l t r y to pay at least., part. of the higher price by r a i s i n g , i n turn, the s e l l i n g price of his unit., To the extent that this process continues a l l the way down, i t can be argued that prices-are 'pulled up' (raised by p u l l i n g from the top). The effect of a new high-income household,, then,:is neutral i n terms of f i l t e r i n g but negative for a l l groups i n terms of prices.. There i s i n t h i s case both an increase i n prices and a deterioration of housing conditions (less a f f o r d a b i l i t y ) . 3b) New household, medium income:. The diagram on.the next, page shows the effect of a new unit i n t h i s case, which can be summarized as follows: 52 Diagram 4.5 The newcomer w i l l bid for the house available for the old resident of the same income l e v e l ; either of them w i l l move to i t , and the other w i l l block the f i l t e r i n g a t . t h i s l e v e l . Again, there w i l l be a pulled-up increase of prices for the lower, levels of the chain. At the same time, the bidding raises the.price of the top-level units, which w i l l then experience a 'pushed-up' price. As opposed to the.bottom half, t h i s pushed-up price can be considered ben e f i c i a l f o r those groups, since i t improves their, p o s s i b i l i t y of moving up the chain. Thus, a medium-income newcomer w i l l improve the housing situation of the top groups and deteriorate that of the bottom groups. For the existing medium-income household, the new situation may be either better, or worse, de-pending on the selling-price/purchase-price d i f f e r e n t i a l he can achieve i n the bidding. I t i s worth considering the effect of a newcomer i n a case where no units are added t.o_the market: . . . . - ....... vJ_ HOUSEHOLD . Diagram 4-6 As the diagram shows, the effect would be neutral f o r the top group and an additional deterioration for the lower group, which would experience a reverse f i l t e r i n g and crowding at the bottom of the housing supply. 53 3c) New household, low income; The low-income newcomer can be expected to bid for the house occupied by the existing low-income household. Diagram 4 . 7 The effect i s a 'pushed up' price a l l the way up the chain. For the bottom household, again, the, new situation depends on the relative price he gets for the house he sells and the one he purchases. If the pressure created by the;new household i s not enough to result in a new dwelling unit, the effect can be expected to be, again, crowding at the bottom level and no f i l t e r i n g process. In general, a l l the cases of f i l t e r i n g produced by population growth can be expected to have some effects in common: A) Population growth results in some degree of f i l t e r i n g down only when i t i s accompanied by housing additions. Population growth without housing addi-tions produces in a l l cases crowding at the bottom level.; depending on the income level of the new household, there may be also a 'fi l t e r i n g up' process which results in a deterioration of housing quality for. the medium-income household as well. B) In a l l cases, the last (bottom) dwelling unit remains occupied, either by the same household or by a low-income newcomer. Additions to the housing stock produce different degrees of f i l t e r i n g , but the process never reaches the bottom household. The effect, of population, growth i s , thus, that of lengthening the f i l t e r i n g chain and blocking the f i l t e r i n g pro-cess at the level.of the newcomer income group. C) The effect of population growth on the f i l t e r i n g process i s regressive: 5A the lower the income.group, the higher the probability of having a deteri-oration of housing quality resulting from growth. In t h i s respect, the lowest income groups are indeed paying the highest cost originated by population growth, and the eventual gains f o r other groups. D) The effect of f i l t e r i n g on improvement of low-income housing has a negative - correlation with population growth: the more growth and the higher the • income of the newcomer, the lower the probability of improved housing for the poor. Case A: Income decreases: (Other factors are held constant.) Lower purchasing power produced by factors such as i n f l a t i o n , or economic recession can be ex-pected to deteriorate the a f f o r d a b i l i t y aspect of housing. To the...extent that higher costs (such as taxation) are s i g n i f i c a n t , someitype of reverse f i l t e r i n g process can be expected: Diagram 4-»8 In the theoretical situation i n which no investments i n rehab are considered, the effect can be expected to be vacancy for the top l e v e l dwelling unit and crowding at the bottom l e v e l . The vacancy i s l i k e l y to occur at the l e v e l i n which the income deterioration i s most severe. For instance, i f only the medium-income household suffers an income deterioration, t h i s i s l i k e l y to result i n vacancy at the medium l e v e l and no change for the high-income household. In a l l cases, though, crowding can be expected for the bottom households. Thus, the r e l a t i v e rank of existing vacancy i n an urban area could be an indicator of decreasing purchasing power for the correspond-ing income group.. Case 5: Population decrease: The effect of population reduction Is more or less similar to that of income increases: a vacancy w i l l occur at the l e v e l of the out-moving household, and th i s vacancy can be expected to trigger a chain of moves whereby the remaining households down the f i l t e r i n g chain may move to better housing. Q HOUSEHOLD our o o '. KBAWOOWED H O U S E Diagram 4.9 The diagram depicts a situation i n which a medium-income household moves out. In general, i t can be expected that the effect of population decrease i s favorable for the.lowest income groups, regardless of the income l e v e l of the out-moving household. 56 L.1.3 The Effects of.Rehabilitation.. Some housing programs i n Canada, such as the Residential Rehabilitation Assistance Program (RRAP) and the Neighborhood Improvement Program (NIP), attempt to improve the housing conditions of the poor by subsidizing investment i n housing and neighborhood improvement. For instance, the main objective of RRAP has been stated as follows: "... To improve the housing conditions of low and moderate income people through assi s t i n g i n the repair and conservation of existing r e s i d e n t i a l buildings." (CMHC, RRAP delivery handbook.,- 1978, p..2) This type of program makes grants and loans available.to households that meet certain conditions i n terms of income, type of occupancy and existing housing qual i t y . The loans and grants are used to undertake repairs and/or improvements required.to meet minimum standards of maintenance. This section attempts to apply the main, concepts of the f i l t e r i n g -process to assess some effects of government subsidies for housing improvement.. Two types of subsidies w i l l be considered: A) Subsidies provided for different groups with a decreasing amount as income increases. This type of subsidy would l e v e l - o f f the capacity by group to invest i n housing maintenance and upgrading; t h i s type of .subsidy i s des-cribed i n the following analysis as a s i t u a t i o n - i n which . a l l income groups invest i n r e h a b i l i t a t i o n . . . . B) Subsidies with an income c e i l i n g that makes them available f o r the lowest income group only. This subsidy i s described as a situation i n which only the bottom group Invests i n r e h a b i l i t a t i o n . . The analysis that follows centers on the same variables,discussed i n the preceeding section. To each of the external factors of f i l t e r i n g , the option of investing on housing improvement (rehabilitation) i s introduced. Case 1: Rehabilitation i n a situation of housing deterioration-Case 1 i n the preceeding analysis referred to the effects ;of housing deterioration i n terms of f i l t e r i n g . I f r e h a b i l i t a t i o n i s introduced as an additional variable i n t h i s case, the following effects can,be expected: - Rehabilitation in.general can be seen as a factor reducing.the pressure f o r adding new units i n the market. To the extent that r e h a b i l i t a t i o n i s i n fact preferred over new construction, the effect would be that of recoverin the housing quality l o s t by aging, with the advantage that the so c i a l costs of moving are avoided. - Rehabilitation f o r the bottom group i s l i k e l y to result i n the l e v e l l i n g -off of the two lower groups, with eventually some swaps between them. The groups that do not invest i n r e h a b i l i t a t i o n w i l l create pressure to add a new unit i n the market: ABAWDOMED NEUJ House OMIT Diagram 4.10 In t h i s case, group 3 w i l l move to the new unit, and group 2 w i l l move to unit 3, but group 1 w i l l not move. I f we assume that the cost of moving to the next unit i s at least partly financed by the .selling, price of the unit currently occupied by each group, then group 2 i s i n fact worse-off i n t h i s case, since the unit he leaves w i l l have no buyer and therefore his cost of moving w i l l be higher. So the benefit obtained by r e h a b i l i t a t i o n i n the bottom group w i l l be paid by the next group i n the chain. I t can be argued that t h i s cost w i l l extend to some degree a l l the way up the chain and produce an overall "pulled-down" decrease i n demand/price; t h i s effect, w i l l most l i k e l y be higher f o r the lower groups (with the exception of the very bottom one) and diffuse upwards. This assumes that the effect of r e h a b i l i t a t i o n i s . that of l e v e l l i n g off the two lower groups. I f the. investment i n r e h a b i l i t a t i o n i s n o t . s u f f i -cient to upgrade unit 1 to a 2 l e v e l , then unit 1 w i l l move to position 2 58 anyways. The effect of r e h a b i l i t a t i o n will.be i n t h i s case n i l , since nobody w i l l use the improved house. On the other hand, i f the r e h a b i l i t a t i o n investment upgrades unit 1 to a l e v e l higher than unit 2 (say, between 2 and 3), the effect to be expected would be a re-ranking of the households i n order to match the new quality rank: s > o o 7^  ABANDONED HOUSE. UWlT Diagram 4.11 Household 2 w i l l want to move to unit 1 which, though not as good as 3, i s probably more affordable. Household 1 may want to move to unit 2 which, again, w i l l l e t him end up with a better house and s t i l l have some money, l e f t i n his pocket. This relocation reaches an equilibrium when the new quality rank coincides with the income rank. t The result i s a net gain for groups 2 and 1 (quality improvement at a lower cost) and a net loss for group 3,. which now has no buyer for the house he wants to abandon. In terms of overall social cost, though, addition of a new unit results i n the abandonment of house 3, the best one of the old units« The following policy implications can be derived from the preceeding analysis: - A subsidy for r e h a b i l i t a t i o n should be substantial i n order to have any effect at a l l ; i f :.tbe'subsidy i s not s u f f i c i e n t , the bottom group w i l l abandon the r e h a b i l i t a t i o n unit anyways, or w i l l not be interested i n the subsidy. - A subsidy for r e h a b i l i t a t i o n should not be t i e d to permanence .clauses. In the cases described above, such a permanence clause would not allow the re-arrangement of the ranks allocated to each household, thus impeding the 59 optimum al l o c a t i o n to occur. Case 2: Income, increases. The effect of r e h a b i l i t a t i o n i n t h i s case i s similar to situation 1, depicted i n diagrams 4.10 and 4.11. The only difference i s that 'benefits' i n t h i s case are actual gains i n quality, while i n case 1 they are only the recovery of quality l o s t with aging and deterioration. Case 3a): Population increase, high income. A) Rehabilitation for a l l groups: This case would be similar to a no-rehabi-l i t a t i o n case i n which r e h a b i l i t a t i o n does not off-set the pressure for adding a new unit. Either with or without r e h a b i l i t a t i o n , i f there arev_no additions to the housing stock, the effect to be expected i s a reverse f i l t -ering with crowding at the bottom of the f i l t e r . Q O O O 7F -HOUSEHOLD Diagram 4.12 Addition of a new unit results i n a one-step f i l t e r i n g at the top and quality improvement at a l l l e v e l s : o o Diagram 4.13 o T— ..*L2I i t (Ota) HOUSEHOLD B) Rehabilitation.for the bottom group only: Rehabilitation at the bottom level'without housing additions produces again crowding at the bottom,, with a l i k e l y higher pressure from upper groups to occupy the improved dwelling: 60 Q O O O -.2 NE.UJ HOUSEHOLD Diagram 4-. 14-Rehabilitation combined with housing additions results i n a variant of case 1, i n which there i s now a potential purchaser for the top existing unit: o o o 1'HOO: ME.LJ 'NEW UMIT USEHOLD Diagram 4-. 15 Depending on the l e v e l reached by unit 1 after the upgrading, a swap of occupants may occur between households 2 and 1; the newcomer will.purchase either the existing top unit or the new unit added to the market. . I f the im-provement i s not si g n i f i c a n t , household 2 w i l l be r e l a t i v e l y worse-off, since the opportunity to upgrade his housing standard i s blocked. o o o /NEI HOU W E . U l/MELJ UfJlT j SEHOLD Diagram 4-.16 3b) Population increase, medium income, age and income constant. A) Rehabilitation for a l l groups w i l l have an effect similar to that of a high income newcomer. The only difference i s that the f i l t e r i n g chain goes a step lower. In other words, there i s crowding at the bottom i f no units are added, and overall improvement when.new units are introduced. 0 Rehabilitation for the bottom group, with no new units added increases the pressure f o r crowding at the bottom of the chain: Q O O O N E . U J HOUSEHOLD Diagram 4.17 With addition of new units, the situation i s as follows: o O o Z/f* s 3-/ W E . l O H O U S E H O L D NE.U) OMIT Diagram 4.18 The newcomer offsets the decrease i n price imposed upon house 2 by the lack of interest from household 1. Or the newcomer may outbid, him .for the purchase of unit 3, i n which case the medium income household ends, up at the bottom of the quality scale. Again, a swap between 2 and 1 may occur ( i f the improvement i s big enough to raise unit 1 above a 'medium' l e v e l ) . In. terms of policy implications, t h i s i s probably the optimum case for a RRAP-like type of program. Note that the permanence clause, though not a problem, does not improve the case either; furthermore, i f a re-ranking results with the upgrading, the permanence policy can prevent the. two lower groups from re-ordering t h e i r positions to a quality rank equivalent to t h e i r income positions. 3c) Population increase, low-income. A) Rehabilitation for a l l groups: i f no units are added to the housing 62 market, there i s no f i l t e r i n g process; the.new household w i l l t r y to occupy the lowest quality house; the.result to be expected i s either crowding or displacement at the-lowest-income l e v e l . Q o o o N E U H O U S E H O L D Diagram L.19 If. additions to the housing stock occur as well as r e h a b i l i t a t i o n , a f i l t e r i n g process a l l the way up the chain can be expected. o l/!s -WEI HOO o o W E U HOUSEHOLD o j N E W U M l T Diagram i.20 There i s an overall improvement i n th i s case. The only, problem i s that, i f r e h a b i l i t a t i o n i s subsidized, the cost of the subsidy i s , i n fact, paid by the newcomer, who w i l l not have to purchase an improved, more expensive house. B) Rehabilitation for the bottom group, no housing additions: again, the . effect to be expected i n t h i s case i s an additional pressure to displace the occupant of the improved unit: Q O o I HOUSEHOLD Diagram A.21 Additions to the housing stock produce, again, a l i k e l y l e v e l l i n g -off of the two lowest groups, and a higher, housing cost for.the newcomer. The newcomer may occupy either unit 2 or unit 1, but i n both cases w i l l encounter an a f f o r d a b i l i t y problem. •<t-HOUSEHOLD Diagram 4.22 ^ > o ^ ' o o. El NEW Case. 4- Income decreases. The-likely effect of r e h a b i l i t a t i o n for the lowest income household i n a situation of decreasing incomes can be depicted as follows: ' 1^ " o • o o Diagram 4.23 Households 3 and 2 can be expected to compete i n t h i s case for the purchase of the renewed unit; t h i s w i l l push-up the price of t h i s unit, and household 1 w i l l l i k e l y move. Since his income i s also assumed to be de-creasing, though, i t i s not clear where t h i s household w i l l move. In.a land-lord-tenant situation the result would be outright displacement. In an owner's situation,- the most l i k e l y result seems to be higher aff o r d a b i l i t y problems for a l l groups. I t should be noted that most of the preceeding analysis refers to homeowners. In case where the low-income occupants are tenants, a 'displace-ment effect can be expected i n a number of cases. For instance, i n cases of population growth and no housing additions (see diagrams 4.1V, 4.19, 4.21) the l i k e l y result i s a high proportion of evictions and higher rental values f o r 64 tenant-occupied dwellings. 4.1.!4 SUMMARY OF FINDINGS The preceeding analysis showed how, i n an hypothetical community consisting of three households (high, medium and low-income), the f i l t e r i n g process takes place. The analysis centered on the effects on .housing improve-ment through f i l t e r i n g of the following variables: population changes, income changes, housing deterioration, additions to the housing stock and r e h a b i l i t a -t i o n p o l i c i e s . The findings can be summarized as follows: A) In a l l cases, the factor that triggers the f i l t e r i n g down process i s the addition of a new dwelling unit i n the market. This addition may be either the construction of a new dwelling, or a vacancy l e f t by out-migration. I f no new houses are available, the f i l t e r i n g down process does not occur and i n some cases (such a population increase) a reverse process.takes.place, i n which the bottom households experience crowding and/or displacement. B) The effects of income increase are generally positive i n terms of housing improvement through f i l t e r i n g . - The improvement occurs regardless of the di s t r i b u t i o n of the increased income. C) The effect of pupulation growth i s negative, p a r t i c u l a r l y f o r the poor. Growth produces a longer f i l t e r i n g chain and different degrees of improve-ment for the medium-to-high groups. But the effect of growth i s also that of blocking the f i l t e r i n g chain, thus leaving the lowest group unaffected by the eventual benefits of the f i l t e r i n g process. In a growth situation, housing .additions provided by the private market should, thus, be combined with some type of housing pol i c y d i r e c t l y addressed to the poor. D) Rehabilitation results i n maximum benefits for a l l groups only when, additions to the housing stock are produced at the same time. Rehabilita-t i o n without housing additions has a net benefit only when the cause of 65 f i l t e r i n g i s income increase. In a growth s i t u a t i o n , . r e h a b i l i t a t i o n with no housing additions produce housing quality improvements but does not change the major problem of growth: crowding and displacement of the poor. Rehabilitation results i n benefits for at least some groups i n all.cases, but i t also results i n costs for some groups. The most common case i s that i n which a low-income newcomer encounters higher housing costs, after' re- . h a b i l i t a t i o n takes place; another eventual loser from r e h a b i l i t a t i o n i s the medium-income household (see case 3a, diagram 4-.16.). The effect of r e h a b i l i t a t i o n subsidy i s n i l unless the subsidy i s s i g n i f i -cant i n terms of housing improvement. I f the s i t u a t i o n i s that of income deterioration,, maintenance subsidies for the lowest quality housing may have an additional problem f o r the poor, since the.renovated homes would be more desirable for higher-income groups. Although the owner of the renewed house may eventually gain. in. the process, i n case of tenants the displacement problem becomes worse. 1.2 THE NEIGHBORHOOD .CHANGE PROCESS. The preceeding section was a discussion of r e s i d e n t i a l mobility from the housing quality point of view. The discussion suggested some of the external factors that may be related to the rate of housing turnover and i t s effects on the housing options of different income groups. One major assump-ti o n i n the analysis was that the choice of housing was based on housing characteristics alone, that i s , no neighborhood or location attributes were associated with different types of housing. An alternative approach to r e s i d e n t i a l mobility i s based on the l o -cation components of housing choice. Residential mobility within t h i s approach would be seen as the result of changes i n the physical and s o c i a l components of different neighborhoods. The question would be, thus, how d i f -ferent groups are affected by external change factors, and how.these changes affect the location/patterns i n different areas of the c i t y . The dynamics of location can be approached from two perspectives:. A) The focus can be on each pa r t i c u l a r area of the c i t y ; the analysis i n this case would center on the process of occupancy changes and changes i n the physical aspect of each neighborhood. B) The focus can be on each of the s o c i a l groups and how they move through-out the c i t y . These two approaches do not necessarily r e f l e c t different theoretical positions i n terms of r e s i d e n t i a l mobility. They both explain the same phe-nomenon, only from a different point of view. In fact, most of the research on r e s i d e n t i a l location.dynamics take to some extent the two approaches, with different degrees of emphasis on one or the other. Since the main objective of t h i s research, though, i s the.study of t r a n s i t i o n areas, approach a) seems to be more appropriate i n t h i s case. Therefore, the emphasis here w i l l be on how the segregation process,, or the. dynamic components of i t , affect the physical and soci a l aspects of those 67 areas affected by change. Some elements of perspective b) r e s i d e n t i a l mobili-ty, w i l l be included only to the extent that they help to support different theories related to the change process i n Transition Areas. Again, a starting point for the discussion of neighborhood change has been provided by Burgess 1 theory of urban growth. VJithin Burgess 1 theory, groups become segregated within .the c i t y as a resul t of_morphological changes produced by c i t y growth. In.terms of neighborhood change, the implications of Burgess' theory of growth can be summarized as follows: "Early i n i t s history, a neighborhood contained.a variety of popu-l a t i o n groups, although higher status persons, native urban dwellers, and families with children would predominate. As the CBD expanded, the previously peripheral neighborhood would become de-sired for other uses., setting off a process of invasion and driving out higher status persons, native Americans, and families with c h i l d -ren. Eventually, a complete process of succession would occur so that the neighborhood population would disproportionately consist of lower status residents, the foreign born ethnic population, and homeless men and women, or at best, the single and childle s s popu-l a t i o n . " (Guest, 1977, p. 281). Thus, external .causes of central c i t y i n s t a b i l i t y were seen as occa-sioned by growth - both of the commercial heart and of new immigrants to the c i t y . While these forces caused decay of neighborhoods, they, also created an economic rationale for the continuing usage of outlying areas. With the decline of immigration and.the decline of the CBD, the inner zones (the pro-, cessing places for new immigrants to the ci t y ) lose a portion .of t h e i r j u s t i -f i c a t i o n for being. With need diminished, the ultimate fate of the worst structures i s abandonment... The Burgess theory l a i d the basic concepts f o r two main.approaches to.neighborhood change.in.transition areas: the 'neighborhood l i f e - c y c l e ' and the 'invasion-succession', theories. These are complementary,, rather than opposite views of the neighborhood chance process: while the f i r s t one ex-plains change i n terms.of.income segregation and focusses on the physical as-pects of neighborhood change (deterioration, density changes), the second one. centers on so c i a l aspects of neighborhood change and introduces segregation 68 by race/ethnicity as a major variable. L.2.1 THE NEIGHBORHOOD LIFE-CYCLE. Hoover and Vernon (Anatomy of a Metropolis, 1959) offered a f i v e -stage model of neighborhood evolution i n t h e i r c l a s s i c study of the New.York Metropolitan Region; .these-five stages were hypothesized as follows:.. Stage 1: Residential development:, transforms undeveloped r u r a l land t o . r e s i -dential use. The building boom of the 1920's defined t h i s stage for many Northeastern North American c i t i e s . Most urban neighborhoods, at th i s stage . . develop as single-family dwellings. There i s a high population growth, high vacancies of short.duration because the construction rate may run ahead of po-pulation growth. Stage 2: Transition stage.: comprises a time of aprtment construction. Many of the undeveloped sites are either patches of open space bypassed i n the f i r s t . building wave, or obtained through the demolition of the oldest single-family dwellings. This stage i s most evident i n the inner rings of metropolitan areas, and i s also characterized by both substantial construction and popula-tion growth. Sometimes the prospects of a neighborhood moving from i t s i n i t i a l , (single family) stage of development to the t r a n s i t i o n stage can be quite alrming for some.residents, who see this.change prospect as destructive to the essential character of the .area. Eventually the fears of the residents.can be galvanized into c o l l e c t i v e s o c i a l action aimed at prohibiting the; dncursion of multiple dwellings. Stage 3: Downgrading: can occur several years l a t e r . In t h i s period the neighborhood can lose a number of dwelling units through demolition, accident-, a l destruction, abandonment, or conversion to other uses. Population and den-s i t i e s increase through the crowding of existing structures by the newest immigrants to the region. The group of young families generates additional 69 strains on an aging infrastructure. This downgrading stage i s often associated with the 'slum invasions' by segregated ethnic and minority groups, since the physical conditions of the. dwellings become crowded or dilapidated. Many slum owners enjoy a f i n a n c i a l advantage for maintaining deteriorated housing through tax write-offs, low taxes on the deteriorated structures,, almost no improvement expenditures, and high income return from the large number of renters i n the crowded structures. . . ... Stage L: Thinning-out: occurs after the immigrants have settled, down. Both population density and .dwelling occupancy decline i n t h i s period. This i s mainly a procesS;;of household size shrinkage as children and boarders .move.out. Higher mortality rate characterizes t h i s somewhat older population; and fo r -merly subdivided dwellings .begin to merge as vacancy rates begin to. r i s e . Building abandonment spreads and dangerous structures are demolished. Stage 5: Renewal: i s the f i n a l stage, in. which the obsolete areas are replaced by new multiple-family units. Most often t h i s has been either subsidized moderate or low-income housing, or luxury apartments. In almost every case, th i s stage i s i n i t i a t e d through public intervention.. The neighborhood l i f e - c y c l e model suggests that density, dwellings, and the l o c a l population.of urban areas regularly change through time, as areas age and move from stage to stage.. Not a l l neighborhoods pass through a l l the stages, and the rates of change vary considerably. Schwirian (1977) suggests several factors influencing the l i f e - c y c l e process: Housing supply: pressures for neighborhood change i n urban areas, are affected by. the r e l a t i v e rate of growth of new houses (at the periphery) to population gain. Access to employment and mode of transportation: areas gaining a c c e s s i b i l i -ty w i l l tend to undergo more fi e r c e competition i n the urban housing land market. Social characteristics of residents, and t h e i r c apability to mobilize res-70 ponses i n order to maintain or.to order neighborhood change. The l i f e - c y c l e model was formulated i n 1959 and, as presented here, does not consider more recent trends to preserve and r e v i t a l i z e central neigh-borhoods. Nevertheless, i t can be considered a useful starting point for the study of the dynamics of neighborhood change. 4.2.2. THE INVASION PROCESS. . The invasion-succession theory evolved from studies of r a c i a l dis-crimination i n USA. Stemming from the general framework provided by Burgess 1 •' theory of growth, i t intended, i n i t s o r i g i n a l form, to describe .the patterns of location of Negro population i n northern US c i t i e s . More recently, i t has been applied i n a more general way to describe a process of occupancy change i n neighborhoods. Duncan and Duncan define <• succession- as--follows:<•• "An area undergoes 'succession' when one type of land use replaces another. The term .'residential succession' means, more s p e c i f i c a l l y , the replacement of one population group in. an area by another. The i n i t i a l population and i t s successor may d i f f e r with respect to eco-nomic function, s o c i a l status, ethnic or national background, race, or other s o c i a l l y s i g n i f i c a n t characteristics or a combination of characteristics." "Accompanying r e s i d e n t i a l succession may be changes i n the density of population, the composition of households, the ways i n which re-s i d e n t i a l structures are used, the character of l o c a l i n s t i t u t i o n s , etc. Such changes, however, are not considered to be invariable concomitants of succession." (Duncan & Duncan," 1957, p. 108) In Duncan and Duncan's view, succession i s considered to have four main stages: penetration, invasion, consolidation and piling-up, recognizing that there i s no sharp l i n e dividing these stages and the sequence i s not necessarily complete in. a l l cases: Penetration i s the i n i t i a l stage, i n which some Negroes enter an area occupied by whites. Invasion occurs when penetration i s followed by the movement into an area of substantial numbers of Negroes. 71 Consolidation refers to the continued increase in.number and proportion of Negroes i n an area, after invasion has been accomplished. P i l i n g up i s the stage.when, after complete occupancy by.Negroes, Negroes continue to increase i n numbers. P i l i n g up represents an increase i n Negro population without a corresponding increase i n . l i v i n g space. The invasion-succession, process whereby Negro population.would occupy i n i t i a l l y white neighborhoods was i n i t i a l l y seen as a ' s e l f - f u l f i l l i n g prophecy' produced to a great, extent by the r a c i a l attitudes of whites: . "When a few Negro families do come into a white neighborhood, some more white families move away. Other Negroes hasten to take the i r places, because the. existing Negro neighborhoods are overcrowded due to segregation. This constant movement of Negroes into white neighborhoods makes the bulk of the residents f e e l that t h e i r neigh-, borhood i s doomed to be predominantly Negro, and they move out -with t h e i r attitudes.against the Negroes reinforced. Yet i f there were no segregation,, t h i s wholesale invasion would not. have occurred. But because i t does occur.,, segregational attitudes are increased, and the v i g i l a n t pressure to s t a l l the Negroes at the borderline i s kept up." (Taeuber & Taeuber, 1965»<p..',21- quoted fromrGlinnar. Myrdal, An American Dilemma, 194-4-) • This view assumes that the invasion i s i r r e v e r s i b l e . I r r e v e r s i b i l i t y i s i m p l i c i t In the widely accepted concept of a 'tip. point', defined as "the percentage Negro i n an area.which exceeds the l i m i t s of neighborhood tolerance for i n t e r r a c i a l l i v i n g " (Taeuber & Taeuber, 1965,"'p, 100-)'.. -Once- -the -percentage Negro passes the t i p point,,, i t i s assumed that whites w i l l leave the area at an accelerated rate and be replaced by Negroes u n t i l the neighborhood becomes ent i r e l y Negro. The p a r t i c u l a r characteristics of the succession.proeess i n terms of r a c i a l attitudes between white and black populations i n North American c i t i e s are not relevant here. Nevertheless, some aspects of i t are worth.discussing, since they can be applied to, broader d e f i n i t i o n s of the succession process. For instance, Taeuber & Taeuber showed.that the magnitude and the rate of the process could be related, to factors, such.as population growth and housing construction. Their findings can be summarized as follows: - The largest succession areas tended to be i n those c i t i e s i n which the r e l -72 a t i v e growth rate of black population, (as. compared to that, of the white population) showed the highest growth d i f f e r e n t i a l . (For instance, a c i t y 'i with a 3% black population growth and a 1% white population growth would have l a r g e r succession areas than another c i t y where the growth rates were 3% and 2% f o r the same groups..) The areas where succession.proceeded, at a f a s t e r rate were also those with higher growth d i f f e r e n t i a l s . . - This pattern was affected by the addition of newr.houses at the periphery and t h e i r a v a i l a b i l i t y to,, black population: "In the large Northern.cities, white population growth, .on net balance, tends to be absorbed in. the suburbs,, while increases-in Negro popu-l a t i o n are absorbed.... i n . central c i t i e s . A high rate of Negro i n -migration (with)......restrictions .on ava i l a b l e housing, outside e x i s t i n g Negro areas will.:necessarily,result i n the replacement.of white po-pulation by Negro population i f . . . . these areas.are already built-up. ... In the South.... due to ,the l e s s e r volume-of. Negro immigration and the a l t e r n a t i v e offered by new re s i d e n t i a l , building, the pressure, f o r Negroes to occupy dwellings formerly occupied- by whites i s l e s s . . . " . "Much of the sociologists'.knowledge of the nature of r a c i a l change i n urban neighborhoods has come from the study of Northern c i t i e s during decades of extremely high Negro in-migration coupled with net losses i n white..population..... The. data, presented.... demonstrate that the fortunes, of residential.neighborhoods .in a c i t y are to a large extent t i e d .to broader changes, occurring i n the metropolitan area and the economy as a whole. An important implication.of the findings here i s that as the source of Negro population growth i n c i t i e s s h i f t s from a predominance of ..migration to a -predominance of natural increase,, and.as the-urban, growth context'is a l t e r e d , modifi-cations of the patterns of r a c i a l . change i n r e s i d e n t i a l neighborhoods may be expected." (Taeuber and Taeuber, 1965, p. 125). 4..2.3. NEIGHBORHOOD CHANGE .AND .RESIDENTIAL MOBILITY. A basic notion underlying neighborhood change.theories i s . t h a t , as the c i t y grows and expands,, households tend to move to adapt t h e i r needs, to the changing environment. Since most-of the new housing, i s b u i l t in. peripher-a l areas, i t i s assumed, that adaptation to housing needs tends to.produce an outward flow of some g r o u p s w h i l e some other groups in-migrate to the older housing i n central, areas. Empirical research on. r e s i d e n t i a l mobility supports t h i s notion i n d i f f e r e n t ways. Some of the findings of those studies can be summarized as follows: A) Residential mobility, i s related to housing needs.: the degree of housing turnover has been suggested to be highly related to the degree of s a t i s -faction with housing. For instance, an extensive study carried out by Abu-Lughod and Foley ("The consumer votes by Moving, I960), on r e s i d e n t i a l mobility i n USA suggested, that d i s s a t i s f a c t i o n with current housing under-l i e s : ;the majority of the decisions to move; the major sources of discon-tent would be, according to. t h i s study: space within the dwelling unit, the s o c i a l and physical characteristics of the surrounding neighborhood and the cost of housing. Dissatisfaction with housing and neighborhood quality would be the result of changes i n housing needs as well as changes i n the'[quality of the dwelling i t s e l f or i t s environment.. B) Residential mobility i s related to family status: most of the d i s s a t i s -factions with housing and neighborhood characteristics derive from changes i n family status. Changes i n family size, increase or decrease., i n .number of children, aging, were • reported by Abu-Lughod in...major changes resulting i n housing d i s s a t i s f a c t i o n . C) Residential mobility i s - r e l a t e d to ethnic status: the same study suggested that, while white households tend to move mostly from the c i t y to the suburbs, with a minority, of whites moving from suburbs to central c i t i e s , the non-white groups tend. ,to move primarily within the c i t y proper.> D) The mobility pattern that.created the suburban - s h i f t i s not.a simple move, but a complex of different moves. The phrase " f l i g h t to the. suburbs", according to Abu-Lughod,.,.is- f a r from accurate: an important, part of the-suburban population has .been drawn, not from the c i t y i t s e l f , .but.from small towns and farms or., suburban areas i n other. regions, and represents an actual in-migration;.into the area. Those who do move from c i t y to suburbs rarely do so i n one move: "The suburbs are not reached by one 7i f l y i n g leap, but by a series of short hops, each one a . . l i t t l e farther out from the center of town". (Abu-Lughod, I960, .'p. 190). E) The relationship between mobility and income i s not clear: on. one hand, as discussed i n page 4-01 income differences result i n different, types of centralization-decentralization patterns, although changes in,time tend to lead to higher c e n t r a l i z a t i o n for low-income groups and decentralization for higher-income groups... ,0n the.other hand, Abu-Lughod suggests .that low-income groups tend to be more mobile than higher-income groups, although income increase expectations would tend to result i n higher mobility. CHAPTER . FIVE: , THE ARBITRAGE MODEL 75 The preceeding discussion consisted i n an overview.of. different as-pects of location dynamics. From.the point of. view of housing quality,, the analysis suggested some variables affecting residential, mobility and the. effects of t h i s mobility on low-income housing options. From the point,of view of neighborhood dynamics, a general overview was presented-of ...some . variables affecting the process of neighborhood .change. In. general.,,, housing turnover and neighborhood change was related to the following variables,-: -A) Housing characteristics and th e i r r e l a t i o n with housing needs and .afford-a b i l i t y for different groups. B) Neighborhood characteristics, i n terms of physical and s o c i a l components of neighborhoods. C) External factors, such as population growth, changes i n income, additions to the.housing stock. .. • D) Social attitudes, i n terms of segregation between different .groups. These-four sets of factors were detected to.be inter r e l a t e d : . i n a dynamic situ a t i o n , housing and neighborhood characteristics were .found to be affected by external growth factors; the effect of external., factors, was found to be related to segregation attitudes; and the changes i n housing-;and neigh-borhood characteristics would, i n turn, affect at least some.external factors, such as housing additions and moves to different types of.housing. The f i r s t part of the discussion (the f i l t e r i n g process).centered on changes i n housing occupancy while holding neighborhood characteristics-constant. The second part (neighborhood change) reviewed some approaches to neighborhood change. The purpose of t h i s chapter i s to combine housing and..-neighborhood variables and to.discuss how these variables interact i n terms of low-income r e s i d e n t i a l location i n . t r a n s i t i o n areas.. The theoretical framework .chosen for t h i s analysis i s the Arbitrage Model as presented by Leven et a l , 76 ("Neighborhood Change: Lessons i n the Dynamics of Urban Decay", 1976). This chapter contains: 1. A presentation of the main concepts of the Arbitrage. Model..-2. A discussion of some of the advantages and problems of this.model i n terms of the body of location theory previously discussed. 3. An application of the Arbitrage Model for the assessment, of the role, of Transition Areas i n terms of low-income r e s i d e n t i a l location. 5.1. THE ARBITRAGE MODEL: MAIN CONCEPTS. The Arbitrage Model as developed by.Leven. et. a l represents .an, attempt to capture the complex set of variables related to .neighborhood change. I t combines previous research on f i l t e r i n g and neighborhood change- theories i n one comprehensive formulation which includes a l l . the previously, mentioned variables that intervene i n the process: housing quality, neighborhood ..attri-butes, external factors and so c i a l attitudes of segregation. •. -. . The model i s partly based on the concept of f i l t e r i n g ; down.and partly based on theories of market r a c i a l t r a n s i t i o n . By•redefining the. con-cept of f i l t e r i n g , and defining housing i n terms of 'housing bundle' .(that i s , a commodity which includes both the dwelling unit and the neighborhood.sur-rounding i t ) , Leven et a l extend previous- theories to a broader view of the housing decision-making process. As opposed to the c l a s s i c economic theory which approaches location, i n a s t a t i c situation as the result of an equilibrium, between;supply, and demand factors, the Arbitrage Model i s a model of market disequilibrium,: and describes the type of forces that come into play when mismatches between supply and' demand occur. This mismatch i s produced by. external .changes, (such as changes i n population and changes i n the housing stock) and,, by.producing s h i f t s i n the market, results i n changes i n the all o c a t i o n of housing to different groups. A starting element for the formulation.of .the Arbitrage model was provided by Martin.J. Bailey (Bailey, 'Notes on the Economics of,Residential Zoning and Urban Renewal', 1959). Bailey described a market.model of.resi-.. dential clustering that resulted from preferences f o r . l i v i n g i n •high-income, rather than low-income neighborhoods and for l i v i n g i n , neighborhoods, with're-la t i v e l y few blacks. . I f one group preferred to l i v e among others of the.same, group and l i v e d i n the boundary between groups only at a discount, and., i f . mem-bers of the other group paid a premium over. what, they would pay ..to live., i n housing surrounded by t h e i r own group, there i s a price incentive,, for .changing land use,from those groups paying the discount to those paying, the premium. This process of s h i f t i n g users continues u n t i l prices i n the. boundary:-area reach an equilibrium value; prices w i l l be highest where the-high-Income groups l i v e together and lowest where the lowest-income, groups are.clustered. Leven et a l extend t h i s basic reasoning and introduce to i t other parameters.. Basic d e f i n i t i o n s of the model are: .. . . . A) "Housing" i s defined as a bundle of services extending, beyond, the dwelling unit. The housing bundle includes.neighborhood attributes and expected future.status of those attributes: . . . . . The amount a household i s w i l l i n g to pay for a p a r t i c u l a r dwelling, unit depends on multiple considerations: structural q u a l i -t i e s such as size and soundness; a c c e s s i b i l i t y ; , scope, and .quality of public.services, and the several characteristics, that define neighborhood quality. C o l l e c t i v e l y , we c a l l a l l these characteris-t i c s of the property the housing bundle." "The estimated value depends not solely of the characteristics asso-ciated with the property today, but also, on the expected future status of those characteristics. To the household, then, value i s the present value of the future, stream of services expected from the property." (Leven et a l , 1976, p. 35). B) " F i l t e r i n g " i s defined i n a broad meaning that includes both changes in. housing quality and changes i n occupancy: " F i l t e r i n g takes place when a household,. without change in. i t s income or tastes, experiences a change in.-its housing bundle to a different rank on i t s scale of preferences. " F i l t e r i n g up" - .or upranking -occurs when the change i s to a more preferred bundle; ' f i l t e r i n g 78 down' - or downranking - when i t i s to a less preferred bundle." "... F i l t e r i n g can occur even i f the household does not experience a locati o n a l change i n i t s dwelling unit. The d e f i n i t i o n e l i c i t s two si g n i f i c a n t corollary d e f i n i t i o n s : for. active and passive f i l t e r i n g . . Active f i l t e r i n g occurs when a household experiences a change i n the ranking of i t s housing bundle by moving to a different.unit. Pas-sive f i l t e r i n g occurs when the household does not move but experi-ences a change i n the ranking of i t s housing nevertheless." (Leven et a l , 1976, p. 96) This d e f i n i t i o n of f i l t e r i n g i s similar to what was previously d i s-. cussed as the welfare approach to f i l t e r i n g , but i t has at the same time a sig n i f i c a n t difference: by focussing on the change experienced by the.house-hold rather than the change experienced by the dwelling unit, i t also contains the mobility element; t h i s i s reflected i n the two corollary d e f i n i t i o n s of passive and active f i l t e r i n g . C) Households are assumed to change residences volun t a r i l y for two reasons: l ) when family circumstances change, such as a change i n size, income., or place of employment, and 2) when the household perceives i t can obtain the desired quantity of services at a more favorable price at another l o -c a t i o n This second factor i s associated with the passive f i l t e r i n g down of the neighborhood and constitutes the central element of the arbitrage process. . , D) Arbitrage i s the change i n location produced by mismatches i n supply and demand: "The Arbitrage process i s the mechanism whereby, supply s h i f t s from one submarket to another; sometimes th i s occurs by households moving from one neighborhood to another, sometimes by a change i n the:;housing bundle so that the neighborhood can serve a new c l i -entele." (Leven et a l , 1976, p. 37) The Arbitrage process describes.. "a moving boundary between high-and low-income or status areas, where a protracted mismatch between.supply and demand makes:lt profitable to convert from high-income to successively lower-income occupancy. In end stage i s reached when values have declined so far that i t i s no longer profitable to maintain, the units f o r occupancy. They are, then abandoned, vandalized, and eventually demolished." (Leven et al, 1976, p. 38) 79 Eow does the arbitrage process come into play? Figure L.2A depicts a simple situation i n which only two groups are at play. These two groups are: - Group one, high-income households, defined as those households which can afford new housing. - Group two, low-income households, which are those who cannot afford new housing. These groups could also be defined i n other soci a l dimensions. For instance, black groups i n USA, who do not have access to houses i n the suburbs, could be defined as group two as well. The diagram depicts the i n i t i a l position of these two groups. This i n i t i a l position implies two major assumptions: - The r i c h prefer to l i v e away from the poor. The poor prefer to live- close to the r i c h . These location preferences introduce soc i a l segregation attitudes into the model and result i n the following location pattern: A-• • ! ~~ D I S C O U N T • • D O • • o n • • D O rll<5H~IN'COMET &OLMCAEY A E E A A E E A PEEMHM o o o o lOU-INCOME AEEA • o. THlSH-INCOME H O U S E H O L D L O W - I N C O M E .HOUSEHOLD (SOOecE f~ LEVEN ET AL.,.: I<370~, P 4o) i Diagram 5.1 Part of the highest-status groups locates i n an 'exclusively r i c h ' area. This group has the most desirable location and pays the highest housing values. At the other extreme i s the group who l i v e s i n an 'exclusively poor' area and pays the lowest rents. In between, there i s a boundary area i n which 80 part of the residents are r i c h and part of- them are poor. Those high-status households w i l l locate i n the boundary area only at a discount price: the negative effect of l i v i n g close to the poor results i n lower housing values . At the same time, the low-income households are paying a premium for l i v i n g closer to the r i c h . Thus, the price of housing i n the boundary area w i l l be somewhere between the high-status and the low-status areas. As long as the supply remains a reasonable match to the demand of each of these submarkets, the market.may be said to be i n equilibrium, a con-d i t i o n under which arbitrage remains inoperative. This equilibrium, though, can be disturbed i n a number of ways: On the demand side, population increases on the high-income s i d e . w i l l inspire new construction: new houses are added i n a new high-income area, and the rest of the neighborhoods remain unchanged. Population increases on the lower-income side, though,, w i l l bring about a different process: since there i s no resource to new construction, families must double-up; increased demand pushes prices upward within the low-income neighborhood and the premium paid for boundary units increases.. This widening gap i n boundary prices makes i t profitable to s h i f t occupancy from high-income to low-income occupancy i n the boundary area, and at th i s point arbitrage comes into process. On the supply side, demolition of low-income housing (growth of CBD, cons-truction of new highways) w i l l generate a shortage of housing for low-income families. Again, i t would become profitable to s h i f t housing from high-income to low-income occupancy i n boundary areas. This movement, i n turn, would reduce the stock for high-income housing and result i n pressures for new additions i n the new high-income neighborhood. On the other hand, the process can also be triggered by additions of houses at the periphery due to income increases, rather than population growth. New houses w i l l i n i t i a t e a chain of moves i n which a high-income 81 household w i l l move to the new unit and leave a vacancy which in.turn w i l l be occupied by a household moving out of the boundary area. This new vacancy, i n turn, gives an opportunity for a low income household to move into the bounda-ry area. The net result i n th i s case i s the abandonment of a unit i n the ex-clu s i v e l y low-income sector. The role of expectation: the tipping point In the i n i t i a l s i t u a t i o n , the three described areas had a r e l a t i v e l y stable composition: two areas were exclusively occupied by one income group, and the boundary area had an income mix i n which the s t a b i l i t y was due to the price-location trade-off i n which each of the two component groups made an 'equilibrium' type of decision. The arbitrage process a l t e r s t h i s equilibrium: although the two extreme areas continue to have only one type of residents,, the boundary area has altered i t s r e l a t i v e proportion of low-to-high-income population. I t has undergone a 'passive f i l t e r i n g down' i n the perception of the remaining r e s i -dents of t h i s area. I t i s here where the 'expectation' factor comes into play: the increase i n low-income percentage has altered the 'tipping, point', or the maximum percentage of low-income households that the higher group i s w i l l i n g to tolerate at the price they paid. The area i s perceived as 'going poor' or 'going black'. At t h i s point, lenders w i l l be less w i l l i n g to finance housing improvements and maintenance, and the remaining high-income households i n that area w i l l exert pressure to move out to the exclusive high-income group, with the resulting new additions to the high-income housing stock i n new high-status neighborhoods. Expectation, then, behaves as a ' s e l f - f u l f i l l i n g prophecy', and results i n a continuing of the arbitrage process, even after the mismatch between supply and demand for low income groups ceased to exist. To the extent that the pressure results indeed i n new housing additions, low-income groups 82 w i l l continue to move into boundary areas. The end stage of th i s process, unless halted by public action, i s the abandonment of the i n i t i a l low-income area, the f i l t e r i n g of the boundary area to a low-class neighborhood, and the f i l t e r i n g of the i n i t i a l high-income area into a boundary neighborhood. In summary, the arbitrage model suggests that: Neither age of housing stock or r a c i a l (income) character of a neighborhood does, i n an by i t s e l f , have any causal relationship with the succession process. The reranking of housing values i s a function and a consequence of changes i n the neighborhood socio-economic levels (socio-economic composition). The movement i s most often a downgrading following a drop i n average income (by increasing the proportion of low-income households), but i t could be also an upgrading due to increase i n income. A neighborhood i n immediate proximity to one which i s experiencing or has experienced the downranking sequence w i l l , unless external constraints are imposed, experience the same succession sequence. "We see the housing stock as an inert commodity, the values of which are determined by the housing preferences of the householders. These preferences express concern not solely with the physical cha-r a c t e r i s t i c s of the housing (as i n the t r a d i t i o n a l approach), but with a whole constellation of attributes (those c o l l e c t i v e l y desig-nated 'neighborhood characteristics') that have greater impact i n setting housing values than such physical t r a i t s as age or modern-i t y , once regarded as decisive i n determining market values and i n i t i a t i n g the cycle of succession". (Leven, l&tf-'SU.. p,»/A8-JL9) Neighborhood change, then, as described i n the arbitrage model, i s i n i t i a l l y the result of external forces (population growth, housing additions ...) The i n i t i a l effect i s accelerated or continued by the so c i a l perception of the neighborhood status. Physical deterioration i s seen, at best,, as a concomitant element brought about by the perception of change (expectation), by lenders and investors who are not w i l l i n g to spend i n maintenance because 83 of the 'unstable 1 character of the neighborhood. 5.2. THE ARBITRAGE MODEL: DISCUSSION. The model presented by Leven et a l . and outlined i n the preceeding section i s an attempt to capture the main variables affecting the dynamics of neighborhood change. The authors complement the presentation with empirical evidence drawn from St. Louis; the succession process, as detected i n different St. Louis neighborhoods, appears to support the main concepts of the arbitrage model. Although there seems to'be some evidence on the r e l i a b i l i t y of the model i n other case studies, (Meadows, AIP Journal Vol. 44 #3, July 78) i t i s probably too early at t h i s point to assess i t s value i n terms of accuracy to describe actual phenomena. Therefore, the discussion here does not attempt to compare the model with actual empirical evidence. Rather, the .focus w i l l be on the in t e r n a l consistency of the model and i t s relationship with the body of theory discussed i n previous chapters. In reviewing the advantages and problems of the model, an attempt w i l l be made to bring together the set of dynamic factors affecting neighbor-hood change. This, i n turn, w i l l be the starting point for the next section, i n which the Arbitrage model w i l l be applied to assess the .role of t r a n s i t i o n areas i n terms of low-income r e s i d e n t i a l location. On the positive side 1). The model s h i f t s the emphasis from a descriptive approach that charac-terizes the neighborhood change theories reviewed before (neighborhood l i f e - c y c l e , invasion-succession) to an explanatory one. I t i s i n fact a theory of why people move. The causal relationships relate neighborhood change to two types of variables: 84 - External factors, such as population growth, additions to the housing stock, a v a i l a b i l i t y of housing to different groups. - Social factors or segregation attitudes between groups. These two sets of independent factors produce a combined effect which i s defined as 'expectation' and behaves as an intervening variable i n the process. The effect i s twofold: household moves and physical deterioration of housing, with abandonment as the f i n a l stage. Diagram 5.2 shows the re l a t i o n -ship between these variables. POPULATION" CHAW6E HO0SIW6 STOCK CHAMSE T\ / / Diagram 5.2. The important innovation introduced i s what i s shown as a loop i n the diagram: the m u l t i p l i e r effect produced by the 'expectation' variable, and the relationship between th i s 'expectation' effect and so c i a l (segregation) attitudes. 2) The reasoning can be extended a b i t further by introducing the concept of 'tipping point' and re l a t i n g i t to the 'expectation' variable. I f 'tipping point' i s defined as the percentage of one group relative, to another i n an equilibrium situation and at a certain housing price, i t would be possible to measure the value of the 'expectation' effect as re-lated to the two causal factors (changes i n demand-supply and segregation attitudes): In cases i n which there i s a strong discrimination against a group, a small change i n the tipping point would be s u f f i c i e n t to trigger the chain of moves. Examples of thi s case would be the white-to-black succession pro-cesses i n some North-American c i t i e s . I t could be said i n th i s case that expectation i s very i n e l a s t i c to household moves. o Conversely, i f discrimination i s low, the expectation effect would not come into play u n t i l the tipping point has changed substantially. If there i s no discrimination at a l l (which could be the case when the two groups belong to different family status), the expectation, element would not come into play. The following diagram depicts the relationship among the three variables: EXPECTATION EFFECT F O P U ' L A J I O M CHANGE (%) -Diagram 5.3. I t should be pointed out that 'expectation' does'not necessarily mean ' f l i g h t out'. I f we assume that a certain mobility.rate i s normal i n any community, the same expectation effect i s produced by a not-moving-in type of reaction. This set of variables provides at the same time one indicator of neighborhood s t a b i l i t y : i t can be expected that a 'stable' community w i l l have no more moves than those produced by s t r i c t demand-supply changes, while an 'unstable' neighborhood would have a high compounded effect produced by the expectation variable. 3) The model focusses on income and r a c i a l changes i n a neighborhood, but i t can have a more general application i f the 'expectation' variable i s approached as discussed above. I f we start with the c l a s s i f i c a t i o n of groups i n terms of soci a l rank, family status and race-ethnicity, i t should be possible to i d e n t i f y the type or types of s o c i a l d i f f e r e n t i a t i o n at play for each p a r t i c u l a r succession case; an expectation value.could then be found for the aggregate effect of the r e l a t i v e discrimination-predict-able for each type. For instance, succession by a different ethnic group of similar s o c i a l rank should produce a lower expectation.value than would be the case i f there i s also a s o c i a l rank difference between the two groups. 4) The 'attitude' element discussed so f a r refers only to the degree of d i s -crimination of the 'high' group towards the 'low' one, but the model also assumes that the lower groups want to l i v e close to the high.ones. I t could be argued that t h i s attitude i s not always true i n r e a l l i f e (a number of examples could be shown i n which ethnic or low-income minorities do not want proximity to other groups). Nevertheless, within, the .logic of the arbitrage model, t h i s assumption i s not necessarily the result of a so c i a l attitude: since 'low groups' are defined, as those who cannot afford new housing, the assumption must necessarily hold true by d e f i n i t i o n , since the existing 'high income' housing are the only housing alternatives (other than crowding) f o r increasing demand i n the low income group.. This apparent redundancy of assumptions can prove to.have useful implications for public policy, i n terms of provision of new housing for the poor: To the extent that the assumption r e f l e c t s a true attitude, location of low 87 income housing must consider the effect of 'stigma' ('separation from the high income groups). - On the other hand, a v a i l a b i l i t y of new housing for the poor through public programs may indeed reduce the pressure for succession to take place, to the extent that the advantages of new housing overrides the perceived bene-f i t s of l i v i n g close to the r i c h . On the negative side 1) The model i s unidi r e c t i o n a l , because changes are seen as invasion of 'high' areas.by ;Mow' groups, but does not e x p l i c i t l y include the reverse process, although t h i s p o s s i b i l i t y i s accepted i n the definitions of f i l -t ering. The fact i s that, within the model, high income groups w i l l always move to new housing and new housing i s assumed to be i n peripheral areas. The p o s s i b i l i t y that high groups prefer old houses (or. central locations) could eventually be included, but the model does not provide any causal variable to explain how t h i s reverse process could, occur. 2) The model i s also somewhat deterministic, since i t considers abandonment as the only end state. Although abandonment i s a very topical, problem i n USA, similar cases i n Canadian c i t i e s are r e l a t i v e l y rare.. This i s not necessarily a d i f f i c i e n c y of the model i t s e l f ; rather, the focus on aban-donment derives from the particular situation i n which the model was i n i t i a l l y used. In f a c t , by using the same l i n e of reasoning developed i n the model, i t may be possible to detect some explanations as to why US and Canadian c i t i e s are different i n t h i s respect: One possible explanation relates to the exogenous variables (demand and supply changes). I t can be argued that, other things equal, abandonment results only when additions to the housing stock exceed increases i n popu-l a t i o n . I f Canada has indeed a lower r a t i o of new-constructions to new-88 households, i t would be reasonable to predict a lower incidence of abandon-ment. - The second explanation may be the difference i n racial.segregation: Canada does not have the black discrimination problem t y p i c a l of US c i t i e s ; within the model, t h i s would r e f l e c t i n lower 'expectation' values and, therefore, lower pressure for completing the succession process. In other words, abandonment i s not necessarily the end stage of the process; i t s r e l a t i v e occurrence depends on the r e l a t i v e effect of the expecta-t i o n factor i n terms of compounding the process, which i n turn depends on the degree of discrimination. To the extent that discrimination i s low and housing additions more or less match population increases, there w i l l ..always be a c l i e n t f o r the lowest income house. 3) The model rejects the notion that housing quality (age, soundness...) has an effect on the succession process. Quality i s seen, at best, as a con-comitant to succession, resulting from the causal variables that determine succession. A case could be made, though, for tfre.notion that quality, even i f i t i s a result of soc i a l and demographic forces at play, does have a causal relationship with intentions to move. In diagram 5.2. (p. 84) showing the relationship among different variables, deterioration should be considered as an intervening variable, similar to 'expectation'. The dotted l i n e i n the diagram would show the m u l t i p l i e r effect of deteriora-t i o n , which should be added to the 'expectation' effect. 4) There i s no consideration to CBD expansion as a variable related to popu-l a t i o n growth. This i s somewhat contradictory with the more t r a d i t i o n a l theories which explain succession as resulting primarily, from CBD expansion. One way of introducing CBD expansion i n the model may be the assumption that every increase i n population produces additional demand for CBD space. The effect of population growth, then, would be twofold: increase i n demand (new unit needed) and decrease i n supply (a unit changes from 89 r e s i d e n t i a l to non-residential.use). This argument w i l l be developed i n the next section dealing with implications of t h i s model for housing poli c y i n t r a n s i t i o n areas. 5) Variables related to land a v a i l a b i l i t y are not included i n the model and they would be d i f f i c u l t to introduce. Implicit i n the model are two assumptions:. 1) there i s a fixed stock of housing i n each area, and 2) there i s unlimited land to expand. To the extent that, these assumptions can be manipulated by policy ( i . e . upzoning to higher densities, urban growth boundaries), the succession process would be dif f e r e n t , and the result of t h i s change i s not clear. 6) Although one of the major advantages, of t h i s model as compared to.other theoretical formulations i s the introduction of some socia l variables (se-gregation attitudes), there are s t i l l a number of other s o c i a l forces at play which affect the process of neighborhood change. For instance, the l e v e l of cohesiveness and organization existing i n each neighborhood, the capability and interest to take active p a r t i c i p a t i o n , either to. encourage the process or to stop i t by impeding the penetration of other groups, are without any doubt important factors to consider. These factors are hardly measurable and i t i s almost impossible to generalize them i n a way that permits t h e i r inclusion i n a model. Perhaps a more reasonable approach would be to consider them on t h e i r own merits i n each case, and leave to the good judgement of the researcher to assess the i r i n c i -dence i n terms of the r e l i a b i l i t y of the model i n the l i g h t of those s p e c i f i c situations. 7) In terms of geographic and geometric factors of location, the model explains the dynamics of the change process but does not consider the s p a t i a l d i s t r i b u t i o n of the two moving groups. I t i s i m p l i c i t that the movement occurs outward from CBD, since new houses are i m p l i c i t l y assumed to be added i n peripheral areas. Nevertheless, while t r a d i t i o n a l theory r e l a t e s t h i s outward movement to income increases, t h i s .mobility i s rela t e d i n the model to the passive f i l t e r i n g down of houses and neighborhoods. It could be argued that, to the extent to which the optimum distance to CBD i s indeed re l a t e d to income, t h i s outward move, without income increases may r e s u l t i n a new type of non-equilibrium i n which the f i l t e r i n g down of neighborhoods results i n locations f a r t h e r away from CBD than the optimum distance f o r a given income. I f that i s the case, at some point of the process a 'come-back-to-the-city' movement would be reasonable to pr e d i c t . .This f a c t o r , again, does not a l t e r the model, but makes i t necessary to use i t i n conjunc-t i o n with other l o c a t i o n f a c t o r s . \ t ..u-int.'..,. -In summary, the Arbitrage model i s probably the most comprehensive t h e o r e t i c a l approach to date to neighborhood dynamics. I t has clear advantages over former approaches i n terms of higher explanatory value. Although there are a number of variables not taken into consideration, most of them can be.introduced without changing the basic structure of the model. S t i l l , some other f a c t o r s of l o c a t i o n dynamics not taken into account cannot be e a s i l y introduced; per-haps the way to consider these remaining variables, i s on a case-by-case basis. 5.3. THE ARBITRAGE MODEL AND TRANSITION AREAS. The discussion i n the preceeding sections centered on how the A r b i t -rage model provides a general framework to understand the dynamics of neigh-borhood change; i n t h i s section, an attempt w i l l be made to apply these general concepts to the s p e c i f i c change process which takes place i n areas around CBD, and the e f f e c t s of t h i s process on low-income r e s i d e n t i a l l o c a t i o n . The analysis w i l l include an assessment of the e f f e c t s of rehab programs i n terms of neighborhood preservation and housing improvement f o r low-income groups. 91 A f i r s t assessment cof the role of rehab programs was .presented i n Chapter At- the f i l t e r i n g process; i n that chapter, rehab was approached i n terms of housing quality alone, with no consideration of neighborhood change i n the process. The discussion that follows introduces neighborhood characteris-t i c s , s o c i a l and location variables i n the analysis, thus complementing the conclusions related to the effects of rehab i n t r a n s i t i o n areas. At the same time, the following discussion i s intended to provide a framework for the analysis of the dynamics of the g e n t r i f i c a t i o n process that w i l l be presented i n Chapter 6. These three dimensions of the role of rehab (in terms of housing turnover; i n terms of neighborhood change dynamics, and i n terms of dynamics of the g e n t r i f i c a t i o n process), are.summarized i n Chapter 7, Policy Implications, and constitute the central part of the analysis of the role of rehab i n terms of r e s i d e n t i a l location for low-income groups i n t r a n s i t i o n areas. In order to introduce some of the variables discussed i n the previous section, which are not contained i n the i n i t i a l formulation of the model, some assumptions w i l l be added to the i n i t i a l set. Starting with t h i s enlarged set of assumptions, the analysis w i l l show how.some external factors, (.population changes, additions to the housing stock, rehab subsidies) affect the dynamics of neighborhood change. The basic assumptions of the analysis can be described as follows: A) Two groups of households are at play: high-income households, who can afford new housing, and low-income households, who cannot afford new housing. B) The high-income groups want to l i v e away from the poor, while the poor prefer to l i v e close to the r i c h . C) Low-income groups l i v e closer to CBD and high-income groups l i v e towards the periphery of the c i t y . This s p a t i a l d i s t r i b u t i o n 'makes economic sense' i n economic terms. D) Each addition i n population has a twofold effect: Effect.on demand (a new unit i s needed) - Effect on supply (a unit changes from r e s i d e n t i a l to non-residential use) (1) E) For sim p l i c i t y , i t i s assumed that each area has i n i t i a l l y the same number of units. F) The price of the dwelling units increases with distance from 'the CBDr..; ; This re f l e c t s both segregation preferences and economic preferences i n terms of distance from the CBD. G) There i s unlimited land to expand at the periphery of the c i t y . Diagram 5.4 below-depicts the i n i t i a l position of the two groups i n three different areas: Area '1;, the exclusively poor area i s adjacent to CBD and has the lowest prices. Area ,2. , the boundary area, has an 'equilibrium' proportion of high-and low-income households. Area '3 , the exclusively r i c h area i s farthest from CBD and has the highest values. (CBD) (D L O U ) . © . 1 W C O M E B O O M D A E Y I N C O M E • • . Q _ • O Q _ • o_ • O O .. • _CL o O Diagram 5.4 The analysis that follows introduces the same external factors a l -ready discussed i n the f i l t e r i n g process; a s l i g h t l y different sequence has been chosen t h i s time in.order to explain the mechanics of the arbitrage process. ( l ) . . . . . . . This assumption r e f l e c t s the effect of population growth on CBD expansion. The actual rate of population/commercial growth i s not r e a l l y l / l ( in Vancouver i t i s probably about 5/1 - the l / l rate has been chosen only f o r s i m p l i c i t y . 93 Case 1; Aging of the housing.stock (population and income are held constant) The Arbitrage model posits that aging of the housing stock alone does not have an effect on turnover. As long as the social characteristics of the neighborhoods don't change, each group w i l l remain, i n t h e i r locations. The effect of aging of the housing stock i s suggested to be different degrees of maintenance (depending on the a f f o r d a b i l i t y of the different groups), but no . moves. I t should be noted that, other things being equal, aging of the housing stock affects the lowest group more than the high ones: since those groups w i l l l i k e l y spend less i n maintenance, aging w i l l result i n increasing differences i n quality between the different types of housing. income, i t can be argued that aging does have an effect on turnover as well. The c r i t i c a l variable i n t h i s case i s the boundary area; i f the i n i t i a l house-holds d i s t r i b u t i o n i n th i s area r e f l e c t s s o c i a l segregation attitudes, i t can be expected that, as low-income dwelling units deteriorate i n the boundary area-, this physical deterioration can be perceived by the, other groups i n the boundary area as a downgrading of the neighborhood. These high-income house-holds w i l l f i n d that i t i s not worthwhile to spend i n maintenance i n a decaying neighborhood and w i l l l i k e l y create pressure to move to the exclusively r i c h neighborhood. Diagram 5.5 The turnover i n th i s case i s started i n the boundary area and results i n the addition of a new unit i n the periphery. At the same time, On the other hand, to the extent that maintenance l e v e l i s related to the unit l e f t vacant i n the boundary area creates an increase.in supply, the effect of which on prices i s l i k e l y to result i n a household moving from the exclusively poor to the boundary area. Thus, the arbitrage process continues as described i n the preceeding section. Case 2: Population increase, high income (other factors held constant) As stated i n the i n i t i a l assumptions, the effect of population increase i s twofold: Demand effect, which w i l l result i n one unit added at the periphery, with no arbitrage effect. Supply-effect, which w i l l create pressure to convert one unit i n the fringe area to CBD uses. This new pressure w i l l reduce the stock available for the low-income groups; prices i n t h i s area w i l l increase, and i t may be worthwhile f o r some houses i n the boundary area to s h i f t from high to low-income occupancy. The succession .process i s , thus an arbitrage process, and can be depicted as follows: chain and a change of expectation i n two areas: the boundary area i s perceived as 'going poor', and the low-income area i s perceived as 'going commercial'. The i n i t i a l move, before the expectation effect, produced a r e l a t i v e improvement of the homeowners who moved, since the demand for CBD space Diagram 5.6 Thus, the high-income newcomer i s producing an upward move i n the 95 allowed them to s e l l at a price such that they could afford a better quality hous£.- For tenants, the effect i s l i k e l y to be simply a higher rent and eventual displacement from the low-income area. After the expectation effect, though, the situation i s somewhat diff e r e n t : Expectation 1 (area going poor) i s l i k e l y to result i n a drop i n prices i n the boundary area, so that homeowners who i n i t i a l l y ,didn^t:move,; became worse-off. The longer they stay i n the neighborhood, the higher the price d i f f e r e n t i a l between thei r current location and the next one available to them. Expectation 2'(area going commercial) pushes up prices i n the low-Income neighborhood; t h i s increase i s an increase i n land values and is. accompanied by a decreasing interest i n the dwelling units, since the expected use i s not r e s i d e n t i a l . Thus, the net result i n both cases i s neighborhood deterioration, produced by a change i n expectation which i s not matched by an. equivalent increase i n demand. Deterioration w i l l increase the expectation effect and behave as an additional m u l t i p l i e r i n the process. Case 3: Population increase, low income (other factors constant) Diagram 5.7 depicts the effect of an increase, i n the number of low-income households. As i n the preceeding case, there i s . a twofold effect: reduction i n supply produced by CBD expansion, and increase i n demand produced by the direct housing needs of the newcomer. The effect i s similar to case 2, but the new low-income housing needed results i n an additional succession chain. © © © • • O • o O • • O • o - ->o- - * o -• • O N E W r > - € 3 - - o - - o 1 N Q J 1 HOUSEHOLD -J i - . -. Diagram 5.7 Again, there are two types of expectation effects: area ( l ) i s expected to go commercial and area (2) i s expected to go poor. More households and less dwelling units available i n the low-income area produce a price increase i n that area. Owners i n the intermediate area may f i n d i t profitable to s e l l t h e i r houses to low-income households and the arbitrage process, as described before, i s triggered by a double impact on demand increase and supply decrease i n the fringe area. The low-income newcomer has to face a situation of increased prices for his f i r s t dwelling unit, and has no previous equity gain to pay th i s increased cost. So i n th i s case, the price of the process i s paid by the new, low income population. Similar situations are encountered by tenants i n fringe areas, who now have to pay higher rents due to the increase i n price of t h e i r dwellings. In t h i s case, there i s an increase i n price for both the land .and the dwelling units i n the fringe areas; the r e l a t i v e incidence of each factor, i n determining the f i n a l price depends on the r e l a t i v e interest (or bidding price) offered by the two competing users. In t h i s case, again, and more so, there i s an improved situation f o r the high-income households who move to the 'better' area,, and a deteriorated price f o r those who prefer to stay i n th e i r neighborhoods. I t should be noted that households i n area 1 and 2 are making a choic< based on housing price alone, t h i s price assumed to be deteriorating due to the changes i n so c i a l composition of each area. To the extent that other location factors are important for them ( i . e . '.character1 of each neighborhood i n terms of physical features, h i s t o r i c value, type of so c i a l services, etc.), every household i s indeed facing a different type of trade-off: I f they remain i n th e i r areas, they lose by price deterioration of th e i r dwelling units. I f they move, they lose the eventual benefits of existing neighborhood cha-r a c t e r i s t i c s . Case 4: Income increases (other factors constant) Increases i n income may have several alternative effects, depending on which group receives the :.income increase and how important are the segrega-tion factors of location as compared to the income/distance to CBD relationship. To the extent that location i s primarily determined by segregation attitudes, income increases are l i k e l y to have l i t t l e effect in,terms of ar-bitrage. Overall income improvements can be expected to.result i n higher maintenance and upgrading of a l l the neighborhoods, with a negligible effect on turnover. Income increases for the bottom group may produce a chain, of moves as described i n the discussion on the f i l t e r i n g process (see ^Chapter 4) since the moves are accompanied by income increases, they do not affect the perceived character of the affected neighborhoods and therefore, there i s no expectation effect. Income increases for the r i c h group, on the other hand, may induce some households i n the boundary area to move to the exclusively.rich neighborhood and trigger an arbitrage process as described before. To the extent that location i s related to the relationship between income and distance to CBD, income increases will.tend to produce an.outwards flow of those groups 'receiving the income improvement. Again, t h i s move w i l l 98 have l i t t l e expectation effect i f the i n i t i a l move i s produced by an improve-ment for the bottom group households, but may result i n an .arbitrage started at the boundary area i f the high-income groups are the ones who receive the increased income. In t h i s case, the l i k e l y result would be. a r e l a t i v e l y higher deterioration of the boundary area, since the potential c l i e n t s for the vacated dwellings would move only at a loss i n terms of a c c e s s i b i l i t y , and they would try to compensate th i s loss only by paying a discount price for the new location. In f a c t , these two types of effects suggest a more general considera-t i o n that should be introduced into the analysis. I f the starting position of the two groups represents an equilibrium i n two dimensions, (distance to the:CBD..and proximity to desired s o c i a l groups) each moving decision w i l l be a trade-off between gains i n one of these dimensions and losses i n the other one. For instance, a move of a low-income group produced by population growth may be worthwhile i n terms of s o c i a l gains, but at the same time represents a loss i n terms of a c c e s s i b i l i t y ito ;the :GBB.-,..-This*-l®ss:-:m^  may 1 result i n just a higher a f f o r d a b i l i t y problem for that group.. Thus, i t can be argued that, to the extent that each location repre-sents an equilibrium between different dimensions of the location decision, the household moves produced by external factors a l t e r t h i s equilibrium .in d i f f e r -ent ways and produce different types of mismatches between demand and supply. The analysis tools provided by the arbitrage model are not s u f f i c i e n t to assess these effects, and c a l l s for some additional theoretical framework to be introduced. This l i n e of argument w i l l be pursued i n Chapter 6, i n which some elements of a multidimensional model w i l l be suggested. The remaining part of t h i s chapter w i l l center on some possible effects of rehab p o l i c i e s on the neighborhood change process. 99 The role of housing improvement.subsidies . (rehabilitation) As discussed i n Chapter 4 (see page 56 ), housing r e h a b i l i t a t i o n p o l i c i e s have two main objectives: neighborhood preservation and low-income housing improvement. These two objectives are expected to be met by subsidies i n housing maintenance and neighborhood improvement.projects. , From the analysis above, i t seems reasonable to expect the following effects of r e h a b i l i t a t i o n : 1) Housing improvement subsidies f o r low-income groups are l i k e l y to produce . net benefits when aging i s the major reason for housing turnover and arbitrage. The subsidies would i n this case compensate the differences i n maintenance capability between groups; by improving the physical appearance, they would reduce the number of moves produced by the perceived physical deterioration of the low-income households. 2) In a l l cases, housing improvement subsidies improve the housing situation of those groups who decide not to move, since better maintenance can be expected to counteract at least to some extent the drop i n prices produced by the succession process as described, p a r t i c u l a r l y i n the boundary areas. 3) By subsidizing housing and neighborhood improvement, the direct effect of the programs i s to counteract the physical deterioration effect accompa-nying succession. 4) However, the preceeding analysis has shown that physical quality of housing and neighborhoods are only a concomitant factor i n the process, the real, causal factors being external factors such as population increase, additions to the housing stock, segregation attitudes. Thus, the effects of rehabi-. l i t a t i o n i n terms of neighborhood preservation w i l l be at best weak, since they are aiming at the effects rather than the causes of neighborhood change. 5) In terms of low-income housing, housing repair subsidies improve the situa-t i o n of those groups that don't move. However, the group that i s .paying the highest cost i n the process i s the new, low-income household; another 100 loser i s the low-income tenant. Tenants of subsidized .rehabilitation units may be considered protected to the extent that subsidies are t i e d to rent c e i l i n g clauses, but the effect of growth affects tenants of other units as w e l l . For these tenants, and for the newcomers, the effect of r e h a b i l i t a -t i o n can be considered negative, since r e h a b i l i t a t i o n increases the proba-b i l i t y of overall price increases i n r e h a b i l i t a t i o n areas. In terms of low-income households e l i g i b l e f o r r e h a b i l i t a t i o n , there i s a positive effect i n the a f f o r d a b i l i t y aspect of the problem; to the extent that low income groups cannot afford new housing, i t can be expected that, t h e i r capability to invest i n adequate maintenance w i l l also be lim i t e d ; i n th i s sense, r e h a b i l i t a t i o n represents an actual improvement, and behaves very much l i k e any other type of income subsidy. At present, one of the conditions to gain r e h a b i l i t a t i o n subsidy under RRAP program i s the requirement;:to stay i n the r e h a b i l i t a t i o n dwelling for.a period of f i v e years. The l i k e l y effect of t h i s p a r t i c u l a r clause i s worth some comments. Looking again at one of the 'arbitrage' diagrams (for example, population increase, low-income, diagram 5.7, two l i k e l y effects appear reasonable to expect: a) The housing options for each low-income newcomer i s related to.the mobility rate of the low-income households. For the newcomer, the net effect of the permanence clause i s similar to that of reducing supply: each r e h a b i l i t a t i o n unit i n the low-income area i s , i n fact, moving out of the market u n t i l the permanence time expires. So again, the benefit of the subsidy i s paid i n terms of higher pressure over the remaining units, increased prices and crowding, b) . In terms of the CBD expansion, each houseowner e l i g i b l e for r e h a b i l i -tation faces a choice: either they s e l l to a commercial bidder, or they, gain the r e h a b i l i t a t i o n subsidy. Their interest i n r e h a b i l i t a t i o n w i l l depend, therefore, on the benefit d i f f e r e n t i a l between these two options. In other words, the amount of.the subsidy must be substantial indeed to compensate for the actual or expected, gains derived from CBD expansion. 8) I t would be worth comparing r e h a b i l i t a t i o n with at least one alternative oriented to act upon some 'triggering factors': additions to the low-income housing stock i n central areas. The effect of t h i s type of policy ® " ~ 1 O O O O o : o .HOUSE.HC2L.D.I I | Diagram 5.8 Direct increase i n the low-income housing supply makes new housing available for low-income groups. The mismatch i n supply and demand needs not,, i n t h i s case, resolve by f i l t e r i n g s but i t i s solved by the new stock. The succession process does not take place at a l l . The low-income groups (both existing and newcomers) have an improved housing, situation, since the increase, i n supply, and t h e i r effect on price, improves t h e i r housing a f f o r d a b i l i t y . Neighborhood s o c i a l balance i s also preserved by maintaining the 'tipping point' unchanged. This type of solution can be achieved i n a number of ways, for instance: - By extending the use of r e h a b i l i t a t i o n for conversions of old housing into two or more dwellings. By encouraging i n f i l l and innovative housing proj ects i n central areas i n which higher densities can be achieved without s i g n i f i c a n t l y changing the character of the area. can be depicted as follows: C&D © • • • o • 1 1 M E U 1 1 UNITS o • on o CHAPTER SIX:.THE'GENTRIFICATION.PROCESS 102 In preceeding chapters, a general overview was presented of some of the current approaches to .residential location. The review has included eco-nomic and socio-ecological theories of location; from both the economic and the s o c i a l point of view, the analysis has shifted from s t a t i c (equilibrium) loca-. t i o n models to theories of location dynamics. The analysis of different bodies of theory has been used to define t r a n s i t i o n areas, to explain the role, of these areas i n terms of r e s i d e n t i a l location f o r the poor, and to derive some implications related to the effects of rehab subsidies on low-income location options. Throughout the analysis of different variables affecting r e s i d e n t i a l location i n t r a n s i t i o n areas, a certain type of location preferences has been more or less i m p l i c i t l y assumed which resulted i n peripheral locations for the high-income groups. These preferences relate, among other variables,, to the assumptions that new housing i s b u i l t i n peripheral areas and preferred by the r i c h , and that income increases correlate to larger quantities of land and increasing distance to the .CBD. On the other hand, the evidence presented i n the. f i r s t chapter sug-gested that a new type of location pattern i s emerging i n some North American c i t i e s . This new phenomenon, which has been called g e n t r i f i c a t i o n , middle-class resettlement or private rehab, i s characterized by a change.in location preferences whereby middle-to-high status groups which t r a d i t i o n a l l y locate i n suburban d i s t r i c t s would now prefer old central neighborhoods. The purpose of th i s chapter i s to apply the main concepts contained i n the review of location theory for the analysis of the g e n t r i f i c a t i o n process. The analysis should provide elements that could be useful for the following objectives: I d e n t i f i c a t i o n of those variables that can be used to predict the nature and extent of the g e n t r i f i c a t i o n process i n different areas. I d e n t i f i c a t i o n of some effects o f . g e n t r i f i c a t i o n on low-income location options, and the factors related to those effects. I d e n t i f i c a t i o n of some effects of rehab programs i n gentrifying areas, i n terms of low-income location options. In other words, the analysis should provide some tools that a policy-maker could use to improve low-income housing programs i n gentrifying neigh-borhoods. The f i r s t part of the chapter i s a synthesis of the main location factors derived from the body of r e s i d e n t i a l location theory.. The synthesis i s presented i n the form of a multidimensional location model.which contains both the equilibrium and the dynamic variables related to r e s i d e n t i a l location. The second part of the chapter applies t h i s multidimensional model for the assessment of possible causes and effects of the g e n t r i f i c a t i o n process, as derived from the location factors suggested i n the model. 6.1 FROM ALONSO TO ARBITRAGE: A MULTIDIMENSIONAL LOCATION MODEL. The analysis starts with Alonso's view of the c i t y . The variables considered i n the Central Place Theory are consistent with, the following set of assumptions, as described i n Chapter 2,.(p. 18:? The c i t y i s a featureless p l a i n Man i s an economic man - There i s no existing housing stock There i s a fixed population - There i s one (central) CBD - There i s unlimited land to expand Given these assumptions, the location game relates an economic dimension of individuals (income) to a geometric dimension of the c i t y (distance to CBD, size of the l o t ) . The result of th i s two-dimensional r e l a -104. tionship i s a location for each individual and a price, for each piece of land. Let us f i r s t review the main elements of th i s two-dimensional r e l a -tionship. Then, by releasing each of the i n i t i a l assumptions, we w i l l examine how the new dimensions introduced affect the location game. In a two-dimensional.game, location results.from the equilibrium between a u t i l i t y curve and a budget l i n e . Both curves depict the assumptions that:, each individual wants to l i v e close to CBD, and each individual, wants large quantities of land. The trade-offs between a c c e s s i b i l i t y and quantity,of land res u l t i n a r a d i a l d i s t r i b u t i o n around the CBD i n which the gradient.is a function of income. The slope of the income gradient i s determined, by, the re l a t i v e steepness of the bid-price curves for each income group,.which i n turn i s determined by overall preferences for quantities of land as compared to a c c e s s i b i l i t y . From the discussion i n Chapter 1 (see pages 29 to 37) it.should be remembered that the income gradient does not necessarily yigld.increasing income with distance to the CBD. This i s the case when overall,preferences are f o r large quantities of land rather than a c c e s s i b i l i t y . In eases of high transportation costs, generally low incomes (as i n some South-American c i t i e s ) or, more generally, stronger desire for a c c e s s i b i l i t y over land quantity, a c c e s s i b i l i t y i s considered a scarcer commodity and high-income groups w i l l tend to have steeper bid-price curves; the lowest-income groups.in those cases w i l l outbid only at more distant locations. Two important notions are impor-tant to remember from the previous discussion: - The income gradient i s a function of the r e l a t i v e value of a c c e s s i b i l i t y to the CBD as compared to other location factors (such as quantity of land). The location of low-income groups i s a function of the.high-income group's, u t i l i t y curve. Low-income groups w i l l tend to locate where the factors least preferred by the high-income groups are optimized. 105 F i r s t as sumption, relaxed: the ci t y , i s not ...featureless. By relaxing the f i r s t assumption, the dimension added i s c i t y geography. For si m p l i c i t y , two types of features w i l l be considered as part of the same dimension or geography: natural features (topography, shorelines, sun exposure...), and man-made features (cemeteries, transportation corridors...). The same reasoning behind the central place theory can be applied to th i s dimension. I f we assume that individuals want to locate close to certain features and f a r away from others, location can be expressed again i n terms of acc e s s i b i l i t y . . Individuals now face a trade-off game i n two dimensions: - A c c e s s i b i l i t y to the CBD vs. quantity of land. A c c e s s i b i l i t y to positive features vs. distance from negative ones. Two sets of u t i l i t y curves could be derived, one for each of these two dimensions. The resulting u t i l i t y function would, again, have to meet a budget constraint.. Thus, an optimum location could be found for each income group. The income gradient around the CBD becomes a more complex income shape i n which the high-income groups are l i k e l y to outbid i n those areas with positive geographic features, and the low-income groups are l i k e l y to locate i n those areas where the negative features are found. I f transportation corridors are considered negative features of the c i t y , f o r instance, the emerging pattern at th i s point starts to .resemble Hoyt's wedge.-shaped income d i s t r i b u t i o n : low-income groups w i l l locate closest to the transportation corridors and each sector w i l l have a- p a r t i c u l a r income-gradient, depending on the geographical characteristics of each of them. Second assumption relaxed: Man i s not only an economic man. From the con t r i -butions of Social Area Analysis i t was shown that soc i a l typology can be.des-. cribed by the three categories of s o c i a l rank, families and ethnicity. These categories can be interpreted as three dimensions of each in d i v i d u a l . The standing of each individual i n each of these three dimensions w i l l relate to the individual location preferences, both i n terms of the c i t y dimensions (geometry and geography) and the corresponding dimensions of the other.groups at play. In other words, segregation attitudes can be defined.as a soc i a l dimension of a c c e s s i b i l i t y ; preference to l i v e close or f a r away from certain s o c i a l groups adds new terms of trade-off to the location game. What emerges at t h i s point i s an a c c e s s i b i l i t y matrix, i n which.individuals w i l l t r y to balance-off a l l the dimensions of t h e i r a c c e s s i b i l i t y preferences... Diagram 6.1 depicts one of the possible ways i n which t h i s accessi-b i l i t y matrix can be visualized. Some of the trade-offs involved i n terms of s o c i a l segregation are summarized below. A) A c c e s s i b i l i t y to the c i t y (geometry and geography). In terms of c i t y d i -mensions, i t was suggested i n Chapter 3 that different s o c i a l types have different a c c e s s i b i l i t y requirements. For instance, the soc i a l rank d i -mension behaves more or less l i k e the economic man described above: the trade-offs i n terms of s o c i a l rank relate to both distance to. CBD and geo-graphic features. The location pattern by s o c i a l rank w i l l thus .be similar to the sector-radial gradient combination already described.. Family status, as suggested-by numerous empirical research studies (see pages 36-5.8) tends to have a clear r a d i a l gradient: family size increases with distance from rfehe- CBD.- -It CQ,uld,:i thus.,-be.. concluded- tha.-t-, family; status relates mostly to.the geometry of the c i t y . I f family size i s substituted for income i n Alonso's u t i l i t y curves, a family-status location function.could be derived. In terms of ethnicity, the a c c e s s i b i l i t y trade-offs would appear to be mostly related to the geographic dimension of the c i t y (remember the example of Norwegians i n New York, p. who look for features resembling those of. t h e i r native land). The ethnicity dimension, on the other hand, i s a changing 107 one: as the groups become integrated into the.new socity, the ethnicity dimen-sion tends to become less important and other factors tend to prev a i l i n the location decision. B) A c c e s s i b i l i t y to other s o c i a l groups i s one way of describing what was previously defined as segregation attitudes. Although t h i s aspect.of loca-t i o n i s probably the one i n which least i s d e f i n i t e l y known, there seems to be at least some evidence to suggest that: In terms of soci a l status, high-income groups would prefer to l i v e f a r away from low-status ones, but the preferences of the low-status persons towards the high-status ones are not clear. Familism seems to be neutral i n terms of segregation. There is,hardly any evidence suggesting either attraction or rejection attitudes between groups having different family status only. - Ethnicity would appear to result i n the desire of at least some groups to l i v e close to t h e i r own group, mostly i n the i n i t i a l period of settlement.. Long-time residents would have different types of closeness-distance preferences to new immigrants, depending on the particular ethnic/racial o r i g i n of the immigrants and the ethnic composition of the old residents.. Thus, the positive/negative components of the ethnicity dimension of th e , a c c e s s i b i l i t y matrix cannot be generalized and should be derived from the s p e c i f i c analysis of the groups at play ..in each case. The.location factors summarized up to t h i s point suggest at least f i v e dimensions for the a c c e s s i b i l i t y matrix. Two of them refer to the c i t y (geometry and geography), and three of them, refer to the individuals (social rank, family status and e t h n i c i t y ) . Each of these f i v e .dimensions would y i e l d a s p e c i f i c type of a c c e s s i b i l i t y requirements and trade-offs for each of the three dimensions of the s o c i a l groups at play. This multi-dimensional set of interactions i s depicted i n Diagram 6.1. 1.08, The diagram shows the relationships suggested f o r the a c c e s s i b i l i t y matrix which appear to be consistent with the evidence presented i n Chapters 2 and 3 (Economic and Socio-Ecologic Factors of Location). I t should be noted that, although these values are reasonably consistent with the preceeding evidence, they represent only a f i r s t and very general approximation which should be developed by future research Into a more accurate and precise formulation. \ .LOCATION XpiMENSlONS ciTY D I M E N S I O N S S O C I A L G E O f A E T K Y 6EO&KAPHY SOCIAL RfkNK. FPvNULlSIA ETHMICITY " S O C I A L . \ DISTANCE TO t B D L O T S I Z E POSITIVE FEATURES NEGATIVE HUSH L O U WITH CHILHtEM WITHOUT CHILDREN NAT WE B O K M F o e e i d w < HI6H o +- + + ' < LOU) + O + 4- -2 £ WITH CHILDREN o +• 2 • * q o u. WITHOUT :CHILDECN . + o 2 a ? o o >-y-0 .NATIVE -o + + : — + H 2 tr FOKEKSW BOE.M + o + H ;+H + H + (-) — .Positive correlation -• .Negative correlation 2 No clear evidence Q No correlation Note: Symbols represent r e l a t i v e rather than absolute preferences. For example, the 0 value i n Children/CBD box indicates that house-holds with children are less interested i n CBD a c c e s s i b i l i t y than childless households. Diagram 6.1 1:C'9J At t h i s point, the i n i t i a l l y r a d i a l income gradient has been modified by the whole set of trade-offs i d e n t i f i e d .in.the a c c e s s i b i l i t y .matrix. .The emerging location pattern i s l i k e l y to be very similar to the 'club sandwich' described by Schwirian (see quote on page 4-2), i n which each of.the soc i a l d i -mensions adds a different layer to the o v e r a l l r e s i d e n t i a l configuration. The role of income: In terms of soci a l class segregation, income was considered as an indicator of s o c i a l rank; income:.in t h i s respect would be regarded as one of the parameters of the a c c e s s i b i l i t y matrix and, as such, i t "'Is one of .the variables related to the u t i l i t y function. In other words, the role of income would be described i n t h i s respect as one of the dimensions of the preferences of each i n d i v i d u a l . At the.same time, income has an additional effect. I f we apply again the basic notion of economic theory, i . e . that the location function represents the equilibrium between a u t i l i t y function and a budget constraint, the new role of income can be explained i n terms of aff o r d a b i l i t y : ... •. In the. multidimensional model, the location decision.is i n fact des-cribed as the choice of optimum amounts of a larger set of consumption goods. To the initial.consumption of land and distance to CBD (geometric.choice), four other 'consumption goods' have been added: .distance to•geographic features and distance to the three s o c i a l dimensions of the different groups.. The optimum amount of a given set of consumption goods an individual . w i l l purchase i s that i n which the preferences of each individual - are.at. equi-librium with the quantities he can purchase with his income within a .given set of prices. Thus, to the extent that these added dimensions of location.have. indeed an effect on prices, location i n t h i s multidimensional model w i l l be. again a function of income; optimum location w i l l be the location .in which each ind i v i d u a l can purchase the optimum quantities of these multidimensional loca-tion factors given a certain set of preferences and a budget constraint. ' 119 Again, i t would be reasonable to predict that higher income groups are l i k e l y to outbid i n those areas where the amount of .the most desirable l o -cation factors i s maximized (better geographical features, higher-status neigh-borhoods and so on) and lower income groups w i l l tend to locate where the least preferred or i n f e r i o r goods are maximized. Thus,. location.of the high income groups i s primarily related to t h e i r own preferences i n terms of family, ethnic and s o c i a l characteristics while location of the lower income groups i s , again, related primarily to the preferences of the higher groups. Third assumption relaxed; housing stock i s added. The next step of the analysis i s the introduction of housing quality as an additional dimension. . Let us assume that, given a fixed population, an optimum location was found for .each indiv i d u a l i n terms of the a c c e s s i b i l i t y matrix previously described. The s i -tuation at t h i s point i s not unlike that of a group of s e t t l e r s having., different s o c i a l status, family and ethnic characteristics, who start a new. town.around a production/commercial centre i n a new land. After some.deliberations, the group can be expected to d i s t r i b u t e s p a t i a l l y around t h i s centre and .occupy a certain amount of land. Each household w i l l acquire a piece of land and build a dwelling unit appropriate to t h e i r rank, family and ethnic preferences; neigh-borhood services can be also expected to be b u i l t at this.point.. Thus, the housing bundle i s suggested to be i n i t i a l l y . t h e result of the ' f i r s t round' of the location game. Location and housing quality are at equilibrium: each household has b u i l t the best possible.house" at'the best pos-sib l e location. Housing quality within t h i s framework i s not an independent variable, but the result (or concomittant factor) of the f i r s t f i v e dimensions of location. Fourth assumption relaxed; population, grows. The introduction of population growth has the effect of a l t e r i n g the i n i t i a l equilibrium of location factors. I l l Either natural growth or immigration w i l l bring.new households to the hypothe-t i c a l equilibrium town. These new households w i l l have .varying characteristics i n terms of s o c i a l , family and ethnic status. I t can be expected that new-comers w i l l have similar preferences to those of the groups already l i v i n g i n the c i t y . At-:-the same time, since the old residents have already, occupied the land around the centre, the only place where housing can be added i s at the periphery of•the c i t y . At t h i s point, housing quality i s more than, the result of t h e . i n i t i a l location round and behaves as an additional dimension of the location game. To the extent,that newcomers have housing preferences similar, to those of .the old residents, t h e i r location decision w i l l be influenced by the housing quality component. Several points are worth noting here: A) Population growth i s related to CBD expansion. The immigrant comes attracted by the growth i n the ci t y ' s economic base, and the a r r i v a l , i n turn, results i n additional demand for commercial and other services. Whatever the effect of t h i s economic m u l t i p l i e r , the sp a t i a l counterpart of i t i s the change i n use of some units to non-residential use.. ;.So the newcomer brings both an additional demand and a reduced supply of the housing stock. B) The supply-reduction effect has been described i n Chapter.5 i n terms of succession process i n which residents of the inner.ring move to successive-, l y outlying areas. This i n i t i a l move produces an expectation ..effect which behaves as a m u l t i p l i e r and accelerates the succession process. C) In addition to the supply-reduction effect, there i s the demand eff e c t . produced by the housing needs of the newcomer. To.the extent that the location needs of the newcomer matches the distance/land quantity position, i n peripheral areas, population growth proceeds without.altering the i n i t i a l equilibrium i n th i s respect. I f , on.the contrary,, the newcomers have different types of needs, a different process would occur. This pro-112 cess was described i n the arbitrage model i n terms of . succession by two different income groups i n one l i n e a r direction, starting from CBD and yielding a new.house i n the periphery. D) To the extent that location follows the multi-dimensional matrix, though, the succession process i s not necessarily l i n e a l . For instance, low-income groups and. childless households may prefer central locations; . people of p a r t i c u l a r ethnic groups w i l l want to l i v e close to.those.of.the same ethnic o r i g i n : high-income families may want to avoid lowlands .event-, ua l l y located i n the periphery, and so on. So the- location f i n a l l y , .chosen by the newcomer w i l l trigger a multidimensional chain of moves, i n which each affected.household w i l l re-assess the equilibrium factors.of the i n i t i a l position against the cost-benefits of several alternative next locations. At t h i s point, the .'housing market i s no longer i n equilibrium: the i n i t i a l mismatch i n supply and demand has' produced several multipliers which accelerate and a l t e r the succession process i n a number of ways... Some.-.of these disequilibrium factors have already been described i n . Chapter 5 .•(i.e..: .CBD. . expectation effect,, 'area going poor' expectation e f f e c t ) . Some b.thers are s t i l l worth pointing out: • People move but the housing bundle i s fixed. This housing, bundle.includes not only a housing quality component; i t also represents- an,optimum accessi-b i l i t y / l a n d quantity position. The next position may represent an Improve-ment over the 'going poor' neighborhood, but i s at the same time a. deterio-ration of the. a c c e s s i b i l i t y optimum.- . . ... ... The number of units added at the periphery i s a function of population, growth. I f the growth rate i s s i g n i f i c a n t l y high, the total.amount of houses needed at each stage may be such that exceeds the land available.at any given distance f o r the equilibrium l o t / s i z e - a c c e s s i b i l i t y . combination. Whoever moves to the new units w i l l be gaining i n terms of housing quality, 113 but the geometry of the new location will.be farther away from the optimum. In short, the resu l t of population growth., i s to transform an i n i t i a l equilibrium situation i n one of continuum disequilibrium:, s o c i a l attitudes w i l l accelerate the process beyond-the actual demand-supply .needs. . Each new posi t i o n w i l l r e s u l t i n an.improvement for one of the elements.of. theiaecessi-, b i l i t y matrix, but a deterioration of other elements.. This, continuum of,.dise-,.. quilibrium positions relates.to the number of variables i n the accessibility... matrix: the more dimensions we consider, the more succession effects we .can reasonably.predict. F i f t h assumption: there i s one central CBD. The assumption of one central, em-ployment centre w i l l not be relaxed i n . t h i s analysis. The purpose of the. whole exercise has been to highlight some variables affecting the role of t r a n s i t i o n areas around.the CBD. For this purpose, the introduction .of a.multi-nuclei type of model of urban structure would only serve to. add complexity without s i g n i f i c a n t l y a l t e r i n g the major concepts highlighted i n the analysis. Sixth assumption:, there i s unlimited land to expand. This l a s t assumption i s i n fact part of the geometric dimension of the c i t y ; i n this-:sense,, i t s relaxa-t i o n would not introduce a new dimension to the location game;, rather,.-it would a l t e r the distance-to-CBD/quantity-of-land relationship i n the u t i l i t y , f u n c t i o n . Restrictions to.land expansion reduce the t o t a l amount of land- for, urban devel-opment. Land being generally more scarce, overall prices and.densities .can be expected to increase; thus, each in d i v i d u a l can purchase less land ..at; a higher price at any given distance than would be the case i f no r e s t r i c t i o n s to land a v a i l a b i l i t y are present. Two types of r e s t r i c t i o n s to land a v a i l a b i l i t y can be visualized: 1) natural barriers to, expansion., (such as lakes,, mountains, floodplains) and . man-made barriers, such as urban growth boundaries or freeze of. .agricultural J-'.In-land. One.aspect, of land.restrictions, worth pointing out i s the r e l a t i o n bet-ween land a v a i l a b i l i t y and the central location, of ithe-I.GBDv..-:.;,lf.v for instance, do not allow a perfectly r a d i a l expansion but l i m i t growth" to only some sectors around.the CBD, as shown i n the two cases-depicted i n Diagram 6.2, the effect i s overall larger-commuting distances for .any .given .population:•• the lack of land i n some, sectors must be compensated;by.additional expansion, of the remaining sectors where there i s land available. To the extent that, commuting distances have a threshold determined by each.individual 1s income. and time available for commuting, this threshold i s reached at an- e a r l i e r stage of growth i n c i t i e s with natural barriers to r a d i a l expansion. (A)_ Unrestricted expansion (B) Restricted^expansion Diagram 6.2 In summary,, the multidimensional model outlined above, i s suggested as a theoretical framework, that encompasses.the most .important, location, factors .re- -viewed in.the preceeding chapters. Social, groups are defined in, three dimen-.. sions: s o c i a l rank, familism and ethnicity; i n each of these ..dimensions, they as assumed to make a choice related to f i v e dimensions of the c i t y : geometry, geography and the three dimensions of the other groups, at play. The.interre- • lationships between these variables are.synthesized,in an accessibility/matrix. One output of th e - a c c e s s i b i l i t y matrix i s an .aggregate set.of u t i l i t y functions which .represent the preferences of each individual.,for; the, five'.di-mensions. The optimum location for each individual i s suggested to be that i n , 11-5 which t h i s aggregate u t i l i t y function.meets a budget constraint. For the higher-income groups, the actual location i s suggested to be a function .of t h i s group's preferences i n terms of family status, s o c i a l rank and ethnicity, while for the lowest-income groups location i t i s suggested to be primarily a func-t i o n of the high-income groups' preferences. Housing and neighborhood quality are suggested to behave as i n t e r -vening variables i n the a c c e s s i b i l i t y matrix: an i n i t i a l (equilibrium) location yields a certain housing and neighborhood quality pattern. This pattern would affect successive location decisions and location changes derived- from popula-ti o n growth. The changes would, i n turn, affect the existing pattern and con-tinue the process. 6.2 THE NATURE OF THE GENTRIFICATION PROCESS The body of theory outlined i n preceeding chapters has attempted to explain some of the characteristics of the re s i d e n t i a l components of Transition Areas, as.they appear i n most North American c i t i e s . These characteristics can be summarized as follows: In terms of population composition, Transition Areas tend.to contain dispro-portionate amounts of low-income and lower status persons, f i r s t generation ethnic and r a c i a l groups, and small households without children. A . r e l a t i -vely s i g n i f i c a n t proportion of this population i s transient i n nature. In terms of physical characteristics, T.A. tend to be composed mostly of old, deteriorated dwelling units and declining neighborhoods. There i s a r e l a t i v e l y higher proportion of rental accomodations over owner-occupied dwellings. In terms of neighborhood change process, T.A. show a process of change from re s i d e n t i a l to CBD type of uses. This change, and the expectations derived from i t , tend to accelerate the decline process and extend t h i s process to adjacent neighborhoods. 116 The theoretical framework provided some insight as to the major r e s i -dential location factors related to these characteristics of T.A. However, as shown i n the introductory part of th i s chapter, a new trend appears to be emerging i n some North American c i t i e s : t r a d i t i o n a l l y , old and deteriorated housing i n central areas seem to be attractive f o r some, middle-to-high income groups; t h i s trend would result i n changes i n population composition, physical characteristics and change process i n Transition Areas. The purpose of th i s chapter i s to provide some theoretical background that can be helpful i n explaining the causes, characteristics and effects of th i s new location trend, p a r t i c u l a r l y i n terms of low-income groups. By b r i e f l y reviewing the location factors derived from the body of theory, an attempt w i l l be made to i d e n t i f y those factors that may be related to t h i s change i n trend and the effects of i t on the overall, location game. This section contains a discussion of the following issues: 1) de-f i n i t i o n of g e n t r i f i c a t i o n , 2) causes of g e n t r i f i c a t i o n , 3) effects of gentri-f i c a t i o n and 4) effects of rehab subsidies. The f i r s t part i s based on d e f i -nitions of related concepts, such as f i l t e r i n g . The second part i s based on the a c c e s s i b i l i t y matrix described i n the preceeding section. The t h i r d and fourth parts apply the Arbitrage model of neighborhood change. 6.2.1 DEFINITION OF GENTRIFICATION The evidence presented i n the introductory chapter describes a change i n location trends which has been defined as 'gentrification.', 'private r e h a b i l i t a t i o n ' or 'resettlement' by different researchers. This phenomenon i s characterized by both a change i n location preferences of some soc i a l groups and a change i n housing quality i n some central areas; these areas t y p i c a l l y undergo, a process of privately financed restoration of old buildings, which are then occupied by some middle-to-high-income .households. A d e f i n i t i o n of th i s process could be approached from two points of view: A) The focus can be on the household; g e n t r i f i c a t i o n i n th i s case would be defined as a change of location preferences. B) The focus can be on the dwelling unit, i n which case g e n t r i f i c a t i o n would be defined as a change i n housing quality or occupancy pattern. The two approaches refer to the same process, only with.different points of view; for the purpose of th i s research, the f i r s t approach appears to have at least two advantages: - A d e f i n i t i o n based on housing quality changes would be d i f f i c u l t to. d i f f e r -entiate from the concept of f i l t e r i n g , , which i s characterized by housing, quality changes as wel l . Furthermore, a d e f i n i t i o n based on preference changes makes i t possible to derive the causes -of the process from the l o -cation preferences described i n terms of the a c c e s s i b i l i t y matrix. On the other hand, the process :of resettlement described by the evidence presented before appears to be suggesting two d i s t i n c t types of phenomena: i n some cases, resettlement occurs around well defined areas, while i n . others, the trend r e f l e c t s a more general change i n the population composi-t i o n i n central areas with no clear location focus. Although the f i r s t case i s -more apparent i n terms of perceived phy-s i c a l changes in-the affected areas, the second one i s probably more frequent. Furthermore, the f i r s t case i s l i k e l y to be strongly related to sp e c i f i c cha-r a c t e r i s t i c s of certain areas (such as c u l t u r a l or geographic t r a i t s ) , while the second case i s l i k e l y to be more cl e a r l y related to more general location trends. Given.these considerations, i t would be reasonable to define the gentrification. process as a change i n the socio-economic composition of some central neighborhoods,, from a predominance of low-income households to a higher proportion of r e l a t i v e l y higher-income residents; - th i s location trend i s l i k e l y to result i n the upgrading of the physical quality of the affected neighborhoods. 6.2.2 CAUSES OF GENTRIFICATION From the theory research, i t was suggested that location i s a process whereby s o c i a l groups interact with the c i t y . The interaction was suggested to take place i n several dimensions: the s o c i a l groups were categorized .in- . terms of. s o c i a l rank, familism and ethnicity. In each of these dimensions, groups would make a choice related to f i v e dimensions of the c i t y : geometry,, . geography, and the three dimensions of the other s o c i a l groups at play. Housing and neighborhood quality was suggested to be an intervening variable:, an i n i t i a l (equilibrium) situation would result i n a certain housing quality, d i s t r i b u t i o n ; t h i s d i s t r i b u t i o n would affect successive location changes derived from population growth. The changes would, in. turn, affect housing and neighborhood quality and continue the change process.. On the other hand, g e n t r i f i c a t i o n has been defined as a change i n • location preferences of a certain s o c i a l group. The t y p i c a l component of t h i s group was described as a young, childless household with high education and medium-to-high income; most of the r e s e t t l e r s are white,, but other components of the ethnicity dimension are not clear from the existing- evidence. This s o c i a l group i s t y p i c a l of the f i r s t stage of the g e n t r i f i c a t i o n process; as resettlement consolidates over time, the s o c i a l composition.appears to show a higher proportion of families with children and lower education/income status. Our task, then, consists i n the i d e n t i f i c a t i o n of the attitudes, of t h i s group i n at least two dimensions (familism, s o c i a l rank) as they relate to the f i v e dimensions of location i n the a c c e s s i b i l i t y matrix. In each of ,119 these dimensions, the analysis.should show some possible factors that may lead to,the observed change i n attitude. The dimension of housing quality should be treated as an intervening variable related to both the causes and the effects of the g e n t r i f i c a t i o n process. The analysis starts with the two-dimensional game which relates the economic dimension of the s o c i a l groups with the geometry of the c i t y , and continues with the successive introduction of the remaining dimensions. Gentr i f i c a t i o n and c i t y geometry. Assuming that location relates to i n d i v i - . duals preferences i n terms of l o t size and distance to the GBD, the f i r s t part of the analysis should detect those changes with which high-income groups' pre-ference for central'locations would make economic sense. Within.Alonso's model, the users of land with steeper bid-price curves locate nearer the center of the c i t y . The wealthy w i l l locate i n the periphery as long as the steepness of the bid-price curves tend to decrease with income increases. Decreasing steepness with income occurs, when land i s generally valued more than a c c e s s i b i l i t y . In economic terms, this i s the , same as saying that a strong appetite f o r land yields higher rates, of substi-tution for land (positive rate) than the (negative) substitution rate for distance. This has been summarized by Alonso as follows: "Given a strong appetite for land, so that the holdings of land vary greatly with income, the wealthier.are affected r e l a t i v e l y less by the costs of commuting because they spread those costs over larger . s i t e s . Consequently, the r i c h are price-oriented whereas the poor are location-oriented. Less a c c e s s i b i l i t y bought with increasing income, a c c e s s i b i l i t y behaves as an i n f e r i o r good.-" (Alonso,. 1964, p. 109) Thus,, the change i n steepness of the bid-price curves.with income increase i s a function of the r e l a t i v e u t i l i t y of land as compared with the 120 d i s u t i l i t y of transportation. I t can.be argued, then,, that a reverse situation may occur (that i s , steeper slopes with increasing income) i n cases when the extra u t i l i t y of land i s generally considered less valuable than the extra d i s u t i l i t y of distance. In Alonso's words, again: "The American, case, where the peripheral, suburban locations are occu-pied by the well-to-do, corresponds to the case of a slowly decreasing, marginal rate of substitution. In the pre-industrial town, where there was s l i g h t d i f f e r e n t i a t i o n i n the size of the l o t s , 'the leading tradesmen l i v e i n the centre of town, the common, people i n the p e r i -phery* . Similar conditions obtain i n other c i t i e s of the world today i n countries such as India and those of Latin America." (Alonso, 1964 p. 108-109. Quote from G. Paulson, The Study of Gities,. 1959). At the same time, i t was submitted i n Chapter 2 that location of low-income groups would tend to be i n places such that they consume less of the good considered 'superior' and more of the good considered i n f e r i o r , because t h e i r - a f f o r d a b i l i t y c e i l i n g would not l e t them outbid the r i c h i n other loca-tions. Given these considerations, some cases in.which a. reverse income/ . steepness situation (steeper curves with increasing incomes) would occur may be: A) Relatively low appetite f o r land: i n cases where large parcels are not. greatly desired, i t can be.expected that the size of the parcels would not change greatly with distance. . The marginal u t i l i t y of land would be less than the marginal d i s u t i l i t y of distance; therefore, high income groups, would tend to show steeper bid-price curves. This case w i l l , be examined l a t e r on, when population growth factors are introduced. B) Higher transportation cost: t h i s would produce generally steeper land/ price gradients; prices w i l l be extremely high i n central locations and reach close to 0 'value at r e l a t i v e l y short distances. The effect of t h i s .in changing the income gradient can be shown In two ways: I f the price of land in.central location becomes so high that the low-income group cannot buy even a minimum parcel, then these groups . w i l l have to choose those.places where land price approaches 0 value. By s a c r i f i c i n g 121 distance (paying the high commuting costs) they can get a r e l a t i v e l y cheap piece of land.. . Transportation being a scarce commodity, the marginal d i s u t i l i t y of trans-portation w i l l tend to be higher than the marginal u t i l i t y of land. As income increases, it.can be expected that the optimum location would tend • to give more economies i n transportation (shorther distances). Thus, to the extent that f u e l becomes expensive and scarce, it..can be expected that people w i l l t r y to maximize transportation savings.over savings i n land. In terms of a market equilibrium model, t h i s yields more central location for high income groups. C) Population Growth:, sheer population growth may a l t e r the' a'cc'eilsibili*: ty/land quantity relationship. The pattern of increasing income with distance assumes that land i s generally preferred over a c c e s s i b i l i t y ; however, a c i t y expanding r a d i a l l y with no r e s t r i c t i o n s to land supply yields increasing quan-t i t i e s of land with increasing distances. Even i n an i n i t i a l case where land i s considered more desirable, i t i s reasonable to expect that,, i f population grows beyond a certain threshold (which may be defined, for instance, as a maximum commuting time within a given transportation technology), the r e l a t i v e d e s i r a b i l i t y of more central location would become more important than the desired for land quantity. D) " Zoning: j _ n general, i t can be argued that zoning increases the d e s i r a b i l i t y of land, to the extent to which i t r e s t r i c t s the overall supply; i f land i s generally scarcer, the bid-price curves would tend to be gentler; i n th i s sense, zoning would lead to peripheral location of the.rich. However, the effect of zoning depends on the actual regulations i n each case. For instance, i f central areas are zoned for very low densities, purchasers would be forced to buy more land than they would under free market conditions; i f that i s the case, site, prices i n central areas can be expected to become higher than what the lowest-income groups can afford. This effect can be increased i f , at the 122, same time, l o t sizes i n peripheral areas are smaller than the optimum land-quantity/ distance-to-CBD relationship. Thus, the effects of zoning cannot be generalized. I t would be rea-sonable to expect that, i f minimum l o t sizes i n peripheral areas are s u f f i -c i e n t l y small and l o t sizes i n central areas are r e l a t i v e l y high, the r i c h would tend to outbid i n central locations, where they would obtain both land quantity and a c c e s s i b i l i t y to CBD. 'E) Urban.Growth boundaries: The effect of urban growth boundaries must be consi-dered i n conjunction with population growth. To the extent that the already developed area of a c i t y i s close to saturation under current zoning regula-tions, increases i n population must result i n additions of new housing i n peripheral areas. Under free market conditions, the urban area w i l l expand u n t i l a l l the new households achieve the optimum l o t - s i z e and distance combi-nation. I f a l i m i t i s fixed to expansion, however, the size of the new sites has to be such that the t o t a l number of sites matches the number of new house-holds within the boundary. In rapidly growing areas, t h i s may result i n s i t e sizes smaller than the optimum. I f such i s the case, high income groups w i l l be w i l l i n g to pay a premium for more central locations. This effect may be compounded over the years. I f we v i s u a l i z e urban growth as successive rings added to the periphery of the c i t y , each new ring w i l l have a l o t size optimum for the population, income l e v e l and available land at the corresponding time of; growth. To the extent that population growth rate increases, and r e s t r i c t i o n s to urban expansion also increase over the years, t h i s may well result i n l o t sizes decreasing with distance. Lots being actually smaller i n peripheral areas, the trade-off i s no longer d i s -tance versus quantity: centrally located l o t s may represent a net gain i n both quantity and distance. Thus, population growth i s l i k e l y to produce, again, a change i n 123 location trends at some point. I f population.grows without r e s t r i c t i o n s to expansion, the threshold would be determined by a maximum commuting time after which a c c e s s i b i l i t y becomes a more desirable good than land, quantity. I f , on the other hand, r e s t r i c t i o n s to expansion are present, the threshold i s reached when l o t sizes become less than optimum i n peripheral areas. The effect of population growth may be multiplied i f ..growth boundaries (such as mountains, water bodies or other natural barriers) obstruct a perfect r a d i a l expansion around CBD. I f growth can occur only towards some directions, expansion must be constrained i n smaller available land at any given distance; or higher commuting time for any given amount of land. Thus, both the distance and the quantity of land thresholds are reached at an e a r l i e r stage of growth. F)Income r e s t r i c t i o n s : Deterioration of overall purchasing power due to i n f l a -t i o n or other reasons are l i k e l y to produce a preference for generally l i v i n g closer to the c i t y centre. This i s consistent with the i n i t i a l assumption that the steepness of the bid-price curves decreases as income increases. The effect of decreasing purchasing power i s depicted i n diagram 6.3. Diagram 6.3 In t h i s case, the overall preference f o r more central location would push down the land prices i n peripheral areas, which would sta r t to show the low-maintenance, deterioration effect characteristic of t r a n s i t i o n areas. Again, this factor points at central locations for high-income groups and pe-124 r i p h e r a l locations for the poor. Ge n t r i f i c a t i o n and c i t y geography. Changes i n physical characteristics of a c i t y may result i n changes i n location preferences. I f we include within geo-graphic features both natural and man-made physical characteristics, the question i s : does a general suburbanization process produce physical changes such that location preferences may be altered? . . The causes, characteristics and effects of the suburbanization pro-cess belong to a f i e l d of location theory which has not been explored i n th i s research. However, at least one factor i s worth some speculations: Traditionally, central areas are considered to be both commercial and em-ployment centers. As such, they have a location 'plus' (savings i n trans-portation costs), but also what could be described as a 'geographic minus', i . e . the negative e x t e r n a l i t i e s usually associated with commercial and i n d u s t r i a l uses (noise, t r a f f i c congestion, etc.). To the extent that suburbanization results i n a reduction of both commercial and i n d u s t r i a l a c t i v i t i e s i n central areas, i t can be argued that at least somei of these negative e x t e r n a l i t i e s are also reduced. At the same time, the savings i n transportation tend to occur towards peripheral locations, at least for some groups, such as i n d u s t r i a l workers. I t could be expected, then, that the a c c e s s i b i l i t y function i n both i t s geometric and geographic dimensions would tend to show a reverse pattern: peripheral locations for the groups which value a c c e s s i b i l i t y to i n d u s t r i a l employment, and central locations for groups who are less related to i n d u s t r i a l or commercial jobs and less affected by the remaining e x t e r n a l i t i e s of central locations. Social dimensions o f . g e n t r i f i c a t i o n . Social groups were described i n terms of three dimensions: s o c i a l rank, familism and ethnicity. In each of them, each group would have location preferences with respect to geometry and geography 1-2 5' of the c i t y , and the three dimensions of the remaining groups. We have ex-plored some possible factors.related to the income/geometry and geography d i -mensions. The remaining task i s , therefore, the analysis of at least one additional dimension (familism) with respect to both the physical and so c i a l dimensions of location, and income with respect of the socia l dimensions. Familism and physical t r a i t s . As discussed previously, family status i s . strongly related to distance from the CBD.:' central. locations are. t r a d i t i o n a l l y preferred by singles and childless couples, while peripheral locations are generally preferred by families with children. From th i s perspective, then, i the g e n t r i f i c a t i o n trend does not represent, at least i n i t s i n i t i a l stage, a change i n location preferences; rather, i t would be just, the effect of a higher proportion of young, childless households. I f familism i s considered alone, the l o g i c a l implication would be that the g e n t r i f i c a t i o n trend would tend to dissapear as the baby boom generation gets older and other groups become dominant. Familism and income. This prediction should be scrutinized more carefully, though. To the extent that the new groups have both an income and a familism dimension, a preference o r i g i n a l l y derived from family status may have an income effect: increasing proportions of high status persons change the socia l character of a neighborhood. I f i t i s true that high status households tend to locate close to one. another, and there i s no segregation between .different family status groups, i t could be argued that the new high status groups may produce an expectation effect and p u l l other high status groups to the up-graded central neighborhoods. This expectation effect seems to be supported by the stage theory of g e n t r i f i c a t i o n discussed i n Chapter 1: i t would appear, from the existing evidence, that the young, high income households are followed, as the g e n t r i f i c a t i o n process consolidates over time, by other groups t r a d i -t i o n a l l y associated with suburban locations. Segregation.attitudes. I t has been argued that, although the primary reason for location of the poor may be non-personalistic factors (affordable housing, a c c e s s i b i l i t y to employment), an area perceived as 'going poor' drives out other, higher status groups, thus f a c i l i t a t i n g theMmove-in of additional low-income groups. Some si g n i f i c a n t changes with respect to th i s t r a d i t i o n a l view have occurred i n more recent years: Urbanization brings about increasing d i v e r s i t y of soci a l types. In terms of family status, for instance, the unmarried, childless status of young house-holds has evolved from a transient stage i n the l i f e - c y c l e to a more perma-nent choice of l i f e s t y l e for larger sectors of the population. This i n -creasing d i v e r s i t y brings increasing variety of housing and location needs. - Related to increasing d i v e r s i t y i s a change i n attitudes among different groups: as a wider range of types appear, the differences between groups become less apparent, and interaction increases. I t i s significant,., f o r instance, that those households surveyed by the N.U.C. who had moved to rehabilitated housing generally perceived s o c i a l mix as a positive factor, and many of them cited exposure to a variety of l i f e s t y l e s as one of the reasons for th e i r location choice. The irony of th i s situation i s that, according to the same survey, private rehab resulted i n dislocation of previous residents who could not afford the higher prices, maintenance re-quirements and assessment values produced by large-scale rehab. Thus, the rehab process at the end defeated one of the very purposes that started i t . Housing quality. One of the main explanations of the decentralization of some groups i s related to housing needs and preferences. The assumption i s that a) young families with children prefer large, new housing, b) high-income 127. groups can afford and prefer new housing, c) new housing i s mostly provided i n peripheral areas, d) therefore, families with children and high status persons tend to locate i n a decentralized pattern. Thus, decentralization would relate to the extent to which high income groups prefer and can afford new housing. On the other hand,, existing evidence on private rehab i n central areas suggests that; a) the income status of res e t t l e r s i s higher, but not always s i g n i f i c a n t l y higher, than that of old residents; b) resettlers mention the a r c h i t e c t u r a l / h i s t o r i c character of old houses as a positive factor, c) resettlement, often relates to h i s t o r i c designation of buildings and neighbor-hoods, d) high cost of suburban housing i s often mentioned as a reason f o r private rehab i n central areas. Thus, there seems to be a change i n both a f f o r d a b i l i t y and prefer-ences attributed to some .income groups. These two changes are probably i n t e r -related, and i t i s hard to say which one i s more important. I t could be ex-pected, however, that i n areas where suburban housing prices are escalating and/ or there i s evidence of interest i n the h i s t o r i c preservation of buildings and neighborhoods, some degree of private r e h a b i l i t a t i o n i s also l i k e l y to be present. The problem i s , again, the effect of t h i s trend on housing prices. To the extent that middle-to-high groups i n i t i a t i n g the resettlement process produce an increase i n housing prices i n the target neighborhoods, the low-income groups are out-bid by other middle-incomes, thus continuing the process and displacing the older residents. 6.2.3. EFFECTS OF GENTRIFICATION The location factors that have been suggested as possible causes of the g e n t r i f i c a t i o n process appear to have different long term effects on low-income location options. Causes such as income r e s t r i c t i o n s , increase i n 128 -transportation costs would produce a reversal of the income gradient, with the r i c h locating i n central areas and the poor i n the periphery. The geographic dimension of suburbanization would contribute to the central location of the r i c h with the poor locating close to the i n d u s t r i a l areas' i n the suburbs. The long term effects of s o c i a l factors of g e n t r i f i c a t i o n are not very clear. To the extent that g e n t r i f i c a t i o n i s merely the result of an i n -'crease i n the number of small, childless households, i t can be argued that the phenomenon would become less s i g n i f i c a n t as th i s group moves to the next stage i n the family l i f e cycle. On the other hand, the neighborhood upgrading pro-duced i n the process may be s u f f i c i e n t l y s i g n i f i c a n t to attract other middle-, to high-income groups; the expectation effect would, thus, continue the up-grading process. In either case, at least some displacement of the poor can be expected but the location options for t h i s group are not clear. F i n a l l y , causal factors such as zoning and urban growth r e s t r i c t i o n s may produce varying effects on low-income location, depending on the actual shape of the growth boundaries and the l o t - s i z e / a c c e s s i b i l i t y relationship resulting i n each case. Thus, the long term effects of th i s type of causal factors are hard to generalize and should be assessed on a case-by-case basis. In terms of the short-term effects of the process, the analysis of causal factors has not given s u f f i c i e n t insight as to the manner i n which the. process develops over time. Even i f peripheral location, of the poor i s to be expected i n the long term, t h i s eventual pattern may either be reached in.one single move or develop gradually by successive moves to outlying areas. The analysis that follows attempts to provide some insight on the dynamics of the g e n t r i f i c a t i o n process as related to the causal factors pre-viously discussed. This analysis should be useful to i d e n t i f y the groups affected during the process and th e i r short-term options i n terms of residenial location. The theoretical framework for t h i s analysis i s the Arbitrage Model 12.9^  as presented i n Chapter 5 and applied to discuss the dynamics of change i n Transition Areas. To the set of . assumptions used as a starting point i n the analysis of Transition areas (see,p. .9-10 »> -"the following variables are i n t r o -duced: A) A new dimension i n the s o c i a l groups i n which high-income households have two types of family status: families with children and childless households. B) Location preferences for the new groups are: Families prefer peripheral areas, while childless households prefer central locations. - No special preferences of different family status households to l i v e either close or f a r away from one another. The analysis of the dynamics of t r a n s i t i o n areas contained the assumption that the CBD. .expands ^wi-t-h-every increase in-population. .This assumption w i l l not be maintained here, since i t would complicate the analysis and does not bring additional insight to the effects of CBD expansion previously dis-cussed. This set of assumptions, plus the general assumptions of the Arbitrage model, describe the starting position of the different groups. T,bothis i n i t i a l p osition, the l a s t assumption introduced i s that a l l the new households are small, chil d l e s s households with middle- to high-income. Diagram 6.4 depicts the i n i t i a l position. An exclusively r i c h neigh-borhood composed of families with children i s located i n the periphery: a boundary area contains an equilibrium mix of high and low-income households. Close to the-.CBD,jare -located, .the,;low-income- households,? -plus'-one high-income, small-family household. 130-O • HISH- INCOME CHILDLESS HOUSEHOLD HIGH-IWCOMe UirH CHILDREN LOW-INCOME . HOUSEHOLD (SD) © © • O o • • o • O o • • o • O o Diagram 6.; I f a new household of high-income and small family status i s added to the scheme, t h i s addition w i l l start a chain of moves i n which the new household locates, close to CBD. The stages of th i s sequency of moves i s de-picted below. The f i r s t move, depicted i n diagram.6.5, i s the direct location of the new household. A low-income, household has/ineentlvectd move^  to the next possible option i n the boundary area. A household i n t h i s area moves to the exclusively r i c h neighborhood i n which, i n turn, a household results located i n a new dwelling unit i n more peripheral locations. (CBD) • • • 1 o • D - f O • O ® o o x> o o NELU HOUSEHOLD Diagram 6.5 This f i r s t move has resulted i n better housing for a l l the house-holds within the chain of moves; the price paid by the newcomer for his dwelling has permitted a low-income household to afford a unit i n a better i 3 r ' area; t h i s i n turn allowed a high-income household to finance'a new dwelling unit. The arbitrage has, "thus favored these households. I t should be noted that t h i s ' f i r s t round' effect i s not unlike the effect produced by the addition of a low-income household to the existing po-pulation: the newcomer i s i n fact financing the upgrading of a l l the households moving up the f i l t e r i n g chain (see diagram 6.6 below). The.only loser i n th i s f i r s t round i s the eventual newcomer, of low income, which would be encounter-ing a more affluent competitor for the central unit and therefore, a higher bidding price; the low-income newcomer has indeed, i n th i s case, his housing option closed by the high-income household's preference for old, central housing. LOUJ E C ! HIGH o MEU3" OUIT :HCOSEHOLDS -O Diagram 6.6 From the neighborhood point of view, though, the effect i s somewhat diff e r e n t . The central.area i s upgrading, but the boundary area i s downgrading, while the high-income area remains unchanged. The low-income households re-maining i n the central area now have a 'better' neighborhood surrounding them, but t h i s quality improvement may also bring higher pressures for better maint-enance standards, higher taxes and similar a f f o r d a b i l i t y problems. The house-holds remaining i n the boundary area, on the other hand, experience a net de-ter i o r a t i o n of housing prices, since t h i s area i s starting to be perceived as 'going poor'. As shown i n the diagram, t h i s f i r s t round effect may trigger a con-tinuation of the arbitrage process: the upgrading of the central, neighborhood may attract additional households of high-income status. 132 The second move, depicted i n Diagram 6.7, shows the effect of t h i s type of ex-pectation: i n th i s case, a household i n the exclusively r i c h area has chosen to move to the upgraded central neighborhood. IPBD O MEU UMlf Diagram 6.7 This move brings about, again, improvement for those who participate i n the chain of movesi The price for t h e i r housing improvement has been paid t h i s time by the high-income household i s l i k e l y w i l l i n g to pay a previum for the house he purchases. At the same time, the neighborhood change process started by the new household (upgrading of the central area, downgrading of the boundary area) i s continued by th i s second round of moves. The t h i r d move, depicted i n diagram 6.8, shows a l i k e l y move started i n the boundary area. The high-income household remaining i n this area has experi-enced a passive f i l t e r i n g down of his neighborhood below the q.uanlity he expected at the price he i n i t i a l l y paid for his location. This household i s , thus, l i k e l y to s e l l at a discount and t r y to move to the exclusively r i c h area. Depending on the preferences of the high-income households, two types of moves can occur i n the exclusively r i c h area: either a household builds a new unit i n peripheral areas, or a household decides to move back to the up-graded central neighborhood: 133,-(So) -O • • • • • O o o o •<>-K>-o "*UMIT 6. • 6 • • Q • • O • O • 1 - O o Diagram 6.8 Households of low income s t i l l remaining i n the central area, on the other hand, are experiencing several changes: thei r neighborhood i s becoming more expensive, which may result i n higher property taxes, higher pressure to improve maintenance standards. cAt the same time, the adjacent neighborhood i s downgrading to a point i n which they can expect to purchase a dwelling unit at a lower price. Depending on the impact of the a f f o r d a b i l i t y problem, they may f i n d i t worthwhile to move to the downgraded neighborhood. This pressure to move out w i l l be higher i f the upgrading of the central neighborhood i s attracting families' previously l i v i n g i n peripheral areas. At t h i s point, the decision to move for the low-income household i s no longer produced by house improvement incentives, but by price and afforda-b i l i t y incentives. Thus, the effect of the g e n t r i f i c a t i o n process at th i s stage can be considered a displacement effect. Low-income households are moving to cheaper, more deteriorated areas i n which they lose not-only i n terms of neighborhood quality, but also i n terms of a c c e s s i b i l i t y to the.CBD. The preferences of the high-income households have also another type of effect: to the extent that these households s t i l l prefer .new houses i n peripheral areas, the process continues more or less as i n the t r a d i t i o n a l Arbitrage model discussed i n Chapter 5 (see, f o r instance, case 3), and houses i n central areas are either abandoned or purchased at a low price by new, low-income households. On the other hand, i f the upgrading expectations i n the gentrifying areas are si g n i f i c a n t enough to attract other high-income house-holds, there i s no further additions to the housing stock. Diagram 6.9 shows these two alternatives. 1 * • D • 6 • • O • • o 'ABANDON o • • o p o A J E U J . _ U M I T : _ 6 • 6 " O - -o-• • o D O • o I • p 0 t . . . __ _ _ _ B - • Diagram 6.'9 In a situation of r e l a t i v e l y f i x e d population, t h i s effect can be viewed as a more e f f i c i e n t u t i l i z a t i o n of existing resources, since old houses are recycled and land i s saved i n peripheral areas. In a case of r e l a t i v e l y high population growth, though, the price of the g e n t r i f i c a t i o n process i s paid, again, by the new, low-income household, who encounters a reduced housing stock and a s t i f f e r competition for an eventual location i n central areas. Summary of short-term effects. The analysis has suggested, an eventual sequence of moves triggered by the addition of one high-income household to a predomi-nantly low-income area. To the extent that the g e n t r i f i c a t i o n process more or. less follows the described pattern, i t s effect can be summarized as follows: l ) Two types of effects can be expected: - The direct effect produced by the allo c a t i o n of the high-income house-hold ( ' f i r s t round' i n the sequence). This f i r s t move improves i n the short term the housing quality of the households that move up the chain, and results i n some degree of so c i a l changes i n each of the affected neighborhoods. 1 3 5 - The in d i r e c t or longer term effect produced by the change expectation, as the central area i s expected to 'go r i c h ' and the adjacent areas are expected to 'go poor'. In the boundary area, t h i s effect i s similar to that produced by the t r a d i t i o n a l arbitrage process: the i n i t i a l down-grading triggers additional moves and continues the process. In the gentrifying area, the upgrading expectation may trigger addi-t i o n a l moves of other high-income households. This, i n turn, accelerates the change process i n the adjacent neighborhoods. 2) Throughout the process, those households that move f i r s t have at least some i n i t i a l housing improvement; i n l a t e r stages, t h i s improvement deteriorates as the downgrading of the boundary area consolidates over time. . At the end, the low-income households locate i n a predominantly low-income area and lose i n terms of a c c e s s i b i l i t y to the >CBD. 3 ) The households that prefer to stay i n th e i r neighborhoods experience the neighborhood deterioration and i t s effect on prices; when they move, they do so at a discount for the dwellings they s e l l and a premium for those they purchase i n order to obtain the desired neighborhood quality. L) The low-income newcomers suffer the effect of increased prices.and less housing options as the t r a d i t i o n a l l y low-income central areas change i n character;' these newcomers are probably paying the highest price of the process. 5 ) For the old residents of low-income, t h e i r loss i s that of a f f o r d a b i l i t y , as pressures for better maintenance-or taxesv.pu-shes .them out,.of the up---• -grading, neighborhoods. ^Ifv-they., sell-;'-'they--trade-off a better. aff o r d a b i l i t y for..a deterioration of th e i r location. Factors related to the occurrence of these .effects may be:. A) Size of the chi l d l e s s , high-income household: To the extent that these households are not s i g n i f i c a n t . i n numbers, t h e i r location preferences are 1-36 not l i k e l y to produce any special expectation effect and the process would proceed with no difference from the t r a d i t i o n a l arbitrage process. I f th i s population i s s i g n i f i c a n t , on the other hand, t h e i r location i s l i k e l y to trigger additional moves and accelerate the process. B) Preferences of the high-income groups: These groups are choosing between the t r a d i t i o n a l large sites/low a c c e s s i b i l i t y of peripheral areas and an eventual s t a t u s / a c c e s s i b i l i t y gain i n central locations. To the extent that the perceived assets of the central neighborhoods are not important for them, the expectation effect i n central neighborhoods i s not l i k e l y to occur. C) Population growth rate: The housing improvement effects eventually produced i n the process are higher for the low-income groups i n a situation of.fixed population. A high population growth situation diminishes, the low-income improvement options. The extreme case occurs when, population, growth i s mostly low-income, since the new low-income groups are the most negatively affected i n the process. 6.2.4. EFFECTS OF REHABILITATION In the discussion of the Arbitrage model i t was argued that physical quality of housing and neighborhoods should be viewed- as an intervening varia-ble i n the location game; as such, a certain quality would result from an i n i t i a l location decision determined by the multiple dimensions of the aceessi-. b i l i t y matrix. This i n i t i a l quality would, i n turn, add another.factor to the location decision as new groups enter the game and households change location throughout the c i t y . The occurence of g e n t r i f i c a t i o n , on the other hand, was suggested to. be related primarily to external factors that would change the location varia-bles described i n the a c c e s s i b i l i t y matrix (such as changes i n .transportation costs, income, a v a i l a b i l i t y of land, etc). Given at least some of these ex-137 ternal causes, the g e n t r i f i c a t i o n process would affect. both the physical qua-l i t y and the s o c i a l configuration of the affected. neighborhoods. Within t h i s scope, the effect of r e h a b i l i t a t i o n subsidies for housing and neighborhood improvement would be that of changing part of the effects of the g e n t r i f i c a t i o n process. The impact of r e h a b i l i t a t i o n would, depend on the re l a t i v e importance of the s o c i a l factors of neighborhood change as compared to th e i r counterpart i n terms of physical quality of housing and neighborhoods. . In other words: to the extent that the physical.appearance of a neighborhood i s perceived as an important element of the 'quality' or 'status' of a neighborhood,, p o l i c i e s oriented to improve t h i s physical appearance would affect the., decisions to move and, therefore, the neighborhood change, process. I f , on the other hand, the decisions to move are primarily based on the.social configuration of the neighborhoods (who l i v e s where), the moving decisions would not be s i g n i f i c a n t l y affected by r e h a b i l i t a t i o n subsidies. The effects of r e h a b i l i t a t i o n suggested .below assume, that both phy-s i c a l quality and so c i a l factors play a role i n the decisions to move.. Thus, r e h a b i l i t a t i o n subsidies would have at least some'effects on the. dynamics of neighborhood change triggered by g e n t r i f i c a t i o n . These effects depend on which areas are selected for:.'improvement: as was previously discussed,., two neighbor-hoods are affected throughout the process: the central, gentrifying area (or exclusively poor area) and the adjacent neighborhoods (or boundary,areas). In the boundary area, g e n t r i f i c a t i o n would have an effect .similar to that.of the t r a d i t i o n a l succession process: deterioration and increasing .proportions of low-income, households. In the central areas, on the other hand.,, the effect, i s that of physical quality improvement and higher proportions of high status groups. Both effects produce a change i n expectations, which contribute to accelerate'the process. 138, Rehabilitation subsidies in.central areas: In the exclusively poor areas, the eff.ects of g e n t r i f i c a t i o n were suggested-to.. relate to the. occurrence of the expectation effect: to the extent that these areas are perceived to be 'going r i c h ' , they would attr a c t other high-income households, previously locating i n peripheral areas. To the extent that physical quality of neighborhoods affect the process., i t can be expected that the upgrading produced by these additional moves would.be accelerated by r e h a b i l i t a t i o n and neighborhood improvement sub-si d i e s . . In terms of low-income housing options, i t was suggested that low-income residents in. gentrifying areas encounter primarily an a f f o r d a b i l i t y problem derived from taxation and eventual pressure to improve maintenance . standards. Rehabilitation subsidies would act upon one aspect of th i s problem by financing higher maintenance, but does not affect the taxation effect.. . Thus, households e l i g i b l e for r e h a b i l i t a t i o n programs would face a.more complex choice: either they use the program and have part of.the..affordability problem reduced, or they get the f i n a n c i a l benefit of s e l l i n g and moving to.the. adja-cent area. Their decision w i l l probably depend on the relative.'pressure for s e l l i n g as compared to the advantages of not-moving, and th i s i n turn would depend on the magnitude of the a f f o r d a b i l i t y problem. For those households not e l i g i b l e for the program (say, those just above the maximum income e l i g i b i l i t y c e i l i n g ) , the situation i s somewhat, d i f -ferent: to the extent 'that r e h a b i l i t a t i o n subsidies are successful throughout th e i r neighborhoods, t h e i r a f f o r d a b i l i t y problem would increase, since both the r e h a b i l i t a t i o n of surrounding houses and th e i r effect i n attracting new. high-income households would add up to increasing pressure to move out. F i n a l l y , low-income newcomers who would eventually locate, i n the de-teriorated, low-income neighborhoods would encounter, after r e h a b i l i t a t i o n , an. addi t i o n a l factor against t h e i r location options in. those areas. Their loss has two causes: the increased demand for central housing, and the decreased 139 rate of new housing construction originated by the change i n preferences of the high-income households. Rehabilitation subsidies i n adjacent areas. The effect of g e n t r i f i c a t i o n , as the t r a d i t i o n a l succession process, i s that of changing the boundary area from •a mixed to an 'exclusively poor' status; t h i s i s accompanied by a physical de-te r i o r a t i o n produced by the income li m i t a t i o n s of the in-moving groups. Deterio-ration, both i n physical and so c i a l terms, would accelerate the succession process. The effect of r e h a b i l i t a t i o n subsidies i n th i s area would, therefore, be similar to those suggested previously (Chapter 5) as general effects of re h a b i l i t a t i o n on t r a n s i t i o n areas: some degree of slowing down of the succes-sion process, better a f f o r d a b i l i t y for the low-income groups. Again, t h i s effect,; though generally positive, i s r e l a t i v e l y weak, since i t acts upon an intervening variable, rather than on the actual causal factors of neighborhood change. As i n the previous analysis of r e h a b i l i t a t i o n i n t r a n s i t i o n areas, the same effect, with additional advantages for a l l groups, would be achieved by direct housing construction i n either the central or the boundary area: the increase i n low-income housing supply represents for the poor an opportunity to remain i n areas more desirable i n terms of a c c e s s i b i l i t y and neighborhood quality, at the same time preserving the so c i a l composition of both the central and the boundary areas. CHAPTER SEVEN:.. POLICY IMPLICATIONS The preceeding research i s an overview of location factors, affecting the location options for low-income groups i n t r a n s i t i o n areas around the Central Business D i s t r i c t . The general overview of location factors has been used for the analysis of..one par t i c u l a r type of neighborhood change process: g e n t r i f i c a t i o n or middle-income resettlement i n central neighborhoods. The research has.been, undertaken with the following purposes:. To i d e n t i f y the role of t r a n s i t i o n areas i n terms of low-income r e s i d e n t i a l location. To i d e n t i f y those variables that can explain the g e n t r i f i c a t i o n process i n central neighborhoods. - To id e n t i f y the effects of the g e n t r i f i c a t i o n process, p a r t i c u l a r l y i n terms of low-income r e s i d e n t i a l location. To a.ssess the effects of housing and neighborhood improvement subsidies on low-income location options i n gentrifying areas. The research started with a review of different bodies.of location theory which provided some insight as to the major economic, social, and physical forces affecting urban r e s i d e n t i a l location, both from a s t a t i c and .a.dynamic point of view. From the general overview, a multidimensional location.rmodel was derived, which attempts to synthesize the major variables at play ,and the way they interact in. the location game. The model, has an 'equilibrium' component (summarized'in .an accessibi-l i t y matrix) and a 'dynamic' component (summarized i n the.Arbitrage model). The a c c e s s i b i l i t y matrix was used to derive some causal factors that. can. explain the ..occurrence, of the g e n t r i f i c a t i o n process and. l i k e l y long-term effects on low-income location.options. The Arbitrage model was used to assess, the dynamics of low-income location i n t r a n s i t i o n areas, the effects of gentri-f i c a t i o n in..this process, and some l i k e l y effects of r e h a b i l i t a t i o n and neigh-borhood improvement p o l i c i e s on low-income r e s i d e n t i a l location options. 1418 This chapter contains: l ) a summary.of the main, findings derived from the theoretical analysis i n terms of ..the .four main objectives l i s t e d above, 2) a discussion of some policy implications derived, from..these findings and 3) a br i e f discussion of how the suggested location model could be used for public policy. 7.1 SUMMARY OF FINDINGS 7.1.1. TRANSITION AREAS AND LOW-INCOME LOCATION The location model presented i n Chapter 6 suggested that, location, preferences.of each in d i v i d u a l could be synthesized i n an a c c e s s i b i l i t y matrix i n 5 dimensions. Two of these dimensions (geometry and. geography) would define the c i t y , while the remaining three (social, rank,, familism, ethnicity) .would. . define the s o c i a l groups at play. An additional dimension (housing and neigh-borhood quality) would result from the interaction between .the s o c i a l groups and the c i t y and.behave as.an intervening variable i n the. process. In an equilibrium situation, the a c c e s s i b i l i t y matrix ..defines the preferences of each i n d i v i d u a l . The actual location choice .would..be-at the point where, the aggregate preferences meet a budget constraint. .  Thus.,, whatever the preferences of each group i n each of i t s s o c i a l dimensions,., the ..actual loca-t i o n i s always related to income. Location of.high-income groups was suggested to coincide., with those areas i n which the most desirable location .factors ..are optimized, while location of the lowest-income groups would.be at those points where the least preferred location factors prevail.. . . Transition .areas, on the other hand, were defined.in Chapter 2.as those areas i n which.a change in. land use i s underway. This change i s i n most cases from r e s i d e n t i a l to CBD type of uses i n areas surrounding the employment,. centre, but i t can. also be. a change from one re s i d e n t i a l use to.another, a change from commercial :to r e s i d e n t i a l use, etc. ' 14*2). The role of t r a n s i t i o n areas on low-income r e s i d e n t i a l location i s related to three major types of location factors: A) In an equilibrium situation, low-income location i n t r a n s i t i o n areas relates to the geometric dimension of location: low-income groups would tend to concentrate around 'the - CBD l o j i g r a s ^ a ^ desired (or more affordable), than large land holdings. B) In a growth situation, low-income location i n those areas i s related to CBD expansion: low-income groups would tend to concentrate i n those areas where growth expectation results i n low maintenance standards. In this sense., the s t a b i l i t y of low-income neighborhoods i n central areas i s related to factors such as population growth, income changes and other factors affec-ting the rate.of CBD expansion. C) In both the equilibrium and the growth situations, low-income location i n central areas relates to s o c i a l dimensions of location: an i n i t i a l location of the poor i n central areas may drive away other, higher status groups, thus increasing the centralization of the poor; t h i s effect may be stronger i f ethnic minorities and other seggregated groups also locate i n central areas. Thus, ^segregation attitudes and existence of minority groups are factors affecting low-income location i n t r a n s i t i o n areas. In terms of neighborhood change dynamics, t r a n s i t i o n has been defined as a process of change over time from one highest and best use to another. In t h i s sense, what defines t r a n s i t i o n areas.around CBD i s not the actual change, but the change expectation. After the change of use has taken place, the area consolidates with the new use and a new t r a n s i t i o n area appears i n the next outer ring. Thus, the change process can be described as a continuous sequence of moves to successively outer areas. 7.1.2 CAUSES OF GENTRIFICATION Gentri f i c a t i o n has been defined as a change i n location preferences ; whereby some medium to high-income households, which during the 1950's and 60's, preferred new dwelling units i n peripheral areas, tend to locate i n old, de-teriorated housing i n central areas. This process of private r e h a b i l i t a t i o n results in. the upgrading of the physical and social status of the affected neighborhoods. The possible causal factors that could explain the occurrence of the ge n t r i f i c a t i o n trend have been derived from the f i v e dimensions of the location preferences synthesized i n the a c c e s s i b i l i t y matrix, and can be summarized as follows: 1) In terms of c i t y geometry, g e n t r i f i c a t i o n would be related to factors that may affect overall preferrences for land as compared to a c c e s s i b i l i t y to CBD. In general, t h i s type of factors would tend to reverse the income gradient and y i e l d , i n the long term, central location for high-income, groups and peripheral locations f o r the poor. Some factors related to the accessi b i l i t y / q u a n t i t y of land relationship are: Higher transportation costs and/or r e s t r i c t i o n s i n f u e l consuption.. - Zoning regulations resulting i n r e l a t i v e l y large l o t s i n central areas and r e l a t i v e l y small l o t s i n peripheral areas. - Urban growth boundaries, either f i x e d by policy or resulting from geo-graphic barriers. Population growth.and c i t y expansion beyond a certain 'threshold' commuting time. Income r e s t r i c t i o n s . 2) In terms of c i t y geography, g e n t r i f i c a t i o n may be related to changes i n the re l a t i v e attractiveness of both central and peripheral areas resulting from suburbanization. By reducing the negative externalities t r a d i t i o n a l l y associated to central areas, suburbanization of i n d u s t r i a l areas may increase the attractiveness of central neighborhoods for r e s i d e n t i a l uses. 3) In terms of s o c i a l dimensions of location, g e n t r i f i c a t i o n would be related to those factors that affect the choices of individuals i n t h e i r family and ethnicity dimensions as related to the physical components of the c i t y , and factors modifying the preferences of different groups to l i v e close or apart from one another. The long-term effects of these changes on low-income location options are not clear and tend to vary depending on the causal factor i n each case. Some changes affecting the soci a l dimensions of location may:be: Higher proportion of childless households which are less interested i n large sites and more interested i n the a c c e s s i b i l i t y to employment and amenities associated to central locations. Increasing d i v e r s i t y of so c i a l groups and th e i r housing and location needs. As new types of so c i a l groups develop (such as the change of the childless status from a transient to a more permanent choice of l i f e s t y l e ) , at least some of these new groups can be expected to have l i f e s t y l e s associated with central c i t y rather than suburban li v i n g . . Changes i n "segregation attitudes between different groups resulting i n higher interaction and increasing attractiveness of areas presenting a div e r s i t y of groups and l i f e s t y l e s . The i n i t i a l stage of the g e n t r i f i c a t i o n process i t s e l f , which, by a l t e r -ing the so c i a l status expectations of central neighborhoods, may tend to a t t r a c t other groups t r a d i t i o n a l l y l i v i n g i n high-status peripheral areas. L) In terms of housing quality and neighborhood characteristics, g e n t r i f i c a -t i o n may be related to the increasing costs of housing in. peripheral areas and changes i n preferences from new to old houses with h i s t o r i c or archi-tectural value. Since housing quality i s suggested to be an intervening variable i n the location game, i t can also be expected that the housing quality improvement produced by private r e h a b i l i t a t i o n would behave as an additional attraction for high-income groups i n central areas. 7.1.3 EFFECTS OF GENTRIFICATION The dynamics of neighborhood change triggered by an i n i t i a l location of some high-income groups i n central areas produce a number of changes i n terms of low-income location options. The location of the high-income house-hold starts a chain of moves i n which other high-income groups, at the other end of the chain, start a new unit. To the extent to which the. dynamics, of change can be accurately described by the Arbitrage model, four major types of effects can be expected: l ) Effects on r e h a b i l i t a t i o n neighborhoods. The new s e t t l e r s are l i k e l y to. produce an improvement of the physical appearance of the affected neighbor-hoods and a corresponding increase i n housing prices. Some of the effects of the process, as described i n Chapter 6, can be summarized as follows: Price increases may force out tenants, who are l i k e l y to be the f i r s t ones displaced. Homeowners may be better-off i f they get the benefit of the increased price by s e l l i n g at the right time and obtain a better dwelling unit i n a more desired location. However, low-income owners who do not want to move face two problems: a f f o r d a b i l i t y of housing, when tax increases and eventual.pressure. for better maintenance.standards, result, from private r e h a b i l i t a t i o n : changing character of the neighborhood; involuntary moves due to the a f f o r d a b i l i t y problem. The housing stock may be reduced either i n absolute or.in r e l a t i v e terms the change from rental to owner-occupied homes reduces the number of rental units i n the market; . . . the r e h a b i l i t a t i o n of old, converted rooming housing to single-family use may produce an absolute reduction of the housing stock. On the other hand, private, r e h a b i l i t a t i o n i n some North American c i t i e s has resulted i n the recovery of old,.abandoned housing and the r e v i t a -l i z a t i o n of whole areas. Thus, the effects of r e h a b i l i t a t i o n i n reducing the housing stock cannot be generalized. However, some factors are worth consider-ing when private resettlement appears to be taking place: A) Type of housing occupancy: since the f i r s t and most l i k e l y to be displaced are tenants, a high proportion of rental accomodations would c a l l f o r s p e c i f i c actions i n t h i s respect. B) A v a i l a b i l i t y of equivalent housing i n surrounding areas. C) Relative income levels of in-coming and out-coming groups. The lower the. income l e v e l of old residents, and the higher that of r e s e t t l e r s , the.most l i k e l y an a f f o r d a b i l i t y problem due to increase i n housing prices on the neighborhood. D) Segregation attitudes, related to income, family status or ethnic differences between the two groups. Although d i v e r s i t y of l i f e s t y l e s i s often a t t r a c t -ive for the new s e t t l e r s , this attitude may not be shared by the old r e s i -dents, and tensions may occur. For instance, i f the old neighborhood has .a high proportion of ethnic minorities, resettlement may be. perceived.in a racial-ethnic dimension. Tensions may be even higher i n the opposite case: low-income, majorities displaced by high-income minorities. E) Existing s o c i a l f a b r i c and community services i n the neighborhood: r e s e t t l e -ment may be highly disruptive i n areas with a high degree of cohesiveness and personal interaction of p a r t i c u l a r groups. For instance,, senior c i t i -zens may be very sensitive to privacy, quietness; new s e t t l e r s with higher mobility, presence of youth or children, may affect t h e i r neighborhood i n a negative way.. I f displaced, they may lose access to special services or friends. In the case of some ethnic groups, clustering may be desired as a way to preserve the i r language and c u l t u r a l t r a i t s , 2) Effects i n surrounding areas. Displacement resulting from private rehabi-l i t a t i o n produces an increased demand for low-income housing i n surrounding areas. The process, as discussed i n Chapter 6 using the Arbitrage model, i s not unlike the chain of moves produced i n other cases of invasion-succession: neighboring areas f i l t e r down, middle-income groups.move out, suburban housing i s b u i l t i n peripheral areas. This process i s also com-parable with the effects of urban renewal: "Thus, the urban renewal projects i n the c i t y caused large numbers of poor and black families to f i n d new housing somewhere. The Arbitrage model f a i r l y well describes how t h i s was accomplished. These renewal projects, which were meant to bring industry and„,middle-income fami-l i e s back to the c i t y , may have caused more middle-income families to leave through the mechanism of Arbitrage than were accomodated i n the subsidized middle-income housing actually r e b u i l t within the renewed areas, even when such housing was actually b u i l t . " (Leven et al,l977, p. 168-169) Some effects are worth pointing out: Newcomers from central areas encounter higher prices. For tenants, t h i s i s a net loss i n terms of a f f o r d a b i l i t y , For homeowners, the cost/benefit depends ( i n terms of housing quality alone) upon the price d i f f e r e n t i a l between the old property and the newly purchased one. Since the primary reason for the low-income homeowner to move to adjacent areas was assumed to be a f f o r d a b i l i t y , i t can be expected that a) he would purchase the lowest quality (cheapest) house i n the new neighborhood, b) he would undertake minimum, i f any, repair work. The neighborhood f i l t e r s down as low income population increases.. Although the i n i t i a l increase i n demand may push the prices up, as the area i s . per-ceived to be 'going poor', decreased interest from other middle-income households may.result i n a long-term price decrease. Homeowner residents face two choices: a) s e l l the property at an early stage, thus benefiting from increasing demand, and move to a more suburban location. Or b) remain i n the neighborhood for a longer period, thus having, i n the end, a net.loss i n housing price. The choice w i l l depend on the desirabi-l i t y of suburban locations, the degree of attachment to the neighborhood, and the household's perception of the change i n the neighborhood. Tenants would face a s t i f f e r competition and also a deterioration of t h e i r housing quality since landlords may choose to save i n maintenance and rent the units to the increasing low-income population. 3) Effects on overall housing supply. We have seen that private r e h a b i l i t a -t i o n may produce a reduction of low-income housing stock i n the r e h a b i l i t a -t e d t^x-o'n neighborhoods. In addition to t h i s , some overall effects should be noted: Decrease i n housing options for new, low-income households. New, low-income population that would normally locate, either temporarily or more or less permanently i n rundown central housing encounter t h i s choice closed by the middle-income r e s e t t l e r . Effect on turnover rate: change from rental to owner-occupied accomoda-tion generally relates to longer permanence. On the other hand, central areas are t r a d i t i o n a l l y preferred by some transient groups who need the services and opportunities of central locations. These groups w i l l increase the bulk of rental accomodations i n adjacent areas; thus,, these groups lose i n terms of a c c e s s i b i l i t y to CBD. This effect cannot be .ge-neralized, since i t depends on the sp e c i f i c location of r e h a b i l i t a t i o n areas, which, may not affect those central locations at a l l ; but i t points out the need to assess the r e l a t i v e importance, of a c c e s s i b i l i t y f to CBD for/those groups either displaced or not allowed to move i n after / private r e h a b i l i t a t i o n takes place. On the other hand, i t was suggested that private r e h a b i l i t a t i o n , and the chain of moves triggered by i t , i s assumed to produce an outwards move process whereby the l a s t household i n the chain w i l l start a new unit. I f this, i s the case, at least some of the stock reduction produced by private rehabi-l i t a t i o n may, be balanced by normal market mechanisms. I t should be noted that i f the new, middle-income household.had chosen a new house rather than an old one i n central areas, the chain.of moves would have not taken place. Thus, private r e h a b i l i t a t i o n , to the extent that 1*93 i t produces new units through the arbitrage process, has a positive effect on housing improvement through f i l t e r i n g for some groups. Some qu a l i f i c a t i o n s to t h i s conclusion should be pointed out, however: A) The occurrence and volume of new housing construction i s p o s i t i v e l y related to the segregation attitudes between different groups: to the extent that changing neighborhoods are perceived as 'bad' by some groups, the succes-sion process proceeds and new units w i l l be produced by market demand. B) The occurrence and volume of new construction i s p o s i t i v e l y related to the d e s i r a b i l i t y of suburban l i v i n g and w i l l proceed as long as the high-income groups prefer new housing over old units i n central areas. As the expectation effect produced by the i n i t i a l upgrading of gentrifying.areas. increase, other high-income groups would move to these areas, thus reducing the pressure for new housing construction. C) The occurrence and volume of new construction i s negatively correlated to the degree of attachment of individuals to the i r neighborhoods. I f house-holds prefer to maintain t h e i r locations, because of neighborhood attachr-ment or lack of interest f o r suburban l i v i n g , the pressure for new cons-' truction w i l l be reduced; overall price structure i n developed areas w i l l r i s e , and bottom groups w i l l have increased crowding.. The relationship between new additions and preferences, f o r suburban . type of housing i s p a r t i c u l a r l y important. So f a r , most of the conclusions are based on the assumption that at least some households (high-income,, large families) tend to move to peripheral areas. However, th i s assumption i s not very c l e a r l y supported by the existing evidence on private r e h a b i l i t a t i o n ; on the contrary, i t would appear that, as the resettlement process consolidates over time, i t attracts increasingly larger proportions of medium-income fam i l i e s : "Early i n the r e v i t a l i z a t i o n of a neighborhood, come the pioneers:... 150 7 a r t i s t s , writers... gays, i n t e r r a c i a l couples. Also, persons with a special f e e l for the neglected potential of a grand home... "Later comes... the doctor and lawyer crowd. And then the suburban crowd with big bucks from the sale of t h e i r homes." (Hoti-sing;^June''1980,, p.l°) "... S l i g h t l y over 10% of the neighborhoods which indicated the pre-sence of families with children indicated that t h e i r numbers had increased after r e h a b i l i t a t i o n . In many cases, the rehabilitated areas were composed of a mixture of singles.childless couples and families with children." "Neighborhoods attracting families with children include St. Louis' Lafayette Square, Indianapolis' Lockerbie Square... Philadelphis Cedar Park..." (NUC report,. -1978V p. 7) I) Effects on o v e r a l l location trends (long-term effects).. The effects of g e n t r i f i c a t i o n pointed out so f a r are those produced by a p a r t i c u l a r action (location i n central areas) undertaken by a p a r t i c u l a r s o c i a l group (small, middle-income households), on low-income r e s i d e n t i a l location. This action however, may be just a f i r s t indication of a more profound change i n loca-t i o n trends; i f that i s the case, some more structural, long-term effects can also be expected. An assessment of the long-term effects of the process must be based on a more clear knowledge of the r e a l forces at play which are producing t h i s change i n preferences. I f private resettlement i s caused primarily by more or less temporary factors, such as preferences of some groups for older housing, stage i n the l i f e - c y c l e of the baby-boom generation, etc., the long-term effect w i l l be r e l a t i v e l y less s i g n i f i c a n t as t h i s generation moves to the next stage i n l i f e - c y c l e . However, there i s some evidence of more structural factors at play, a was discussed i n Chapter 6. In terms of low-income r e s i d e n t i a l location, the effects w i l l depend on which are the dominant underlying forces. For instance: A) G e n t r i f i c a t i o n due to economic forces (transportation r e s t r i c t i o n s , income r e s t r i c t i o n s , etc): t h i s set of location factors would tend to.produce a reversal of the income gradient, to a future equilibrium situation with the highest income groups i n central areas and the poor towards the p e r i -phery. .1513 B) Morphological changes i n the c i t y (suburbanization of i n d u s t r i a l areas, l i m i t s to urban expansion, l o t size configuration by area...): the effects of these factors w i l l vary, depending on the p a r t i c u l a r characteristics i n each case. C) Social changes (less segregation between groups, more d i v e r s i t y of so c i a l types, decreasing attractiveness of suburban , : l i f e-styl.e;; etc.').,. may ', result i n either a change of the shape of the income gradient (such as cen-t r a l i z a t i o n of low-income groups i n intermediate areas, with 'urban' high status groups i n central areas and 'suburban' high status groups i n p e r i -pheral areas), or different degrees of integration between different groups at each location. Most l i k e l y , the g e n t r i f i c a t i o n trend i s being caused by a combina-t i o n of temporary and structural changes occurring i n urban areas. This multi-p l i c i t y of factors at play creates a situation of uncertainty with respect to future location options for low-income groups; as the process, develops over time, these groups are l i k e l y to be pushed out from successive locations and encounter increasing d i f f i c u l t i e s to enter the location market altogether. This situation i s a constraint for planning strategies available because of t h i s long-term uncertainly i n terms of optimum location f o r low-income groups. In.cases where a g e n t r i f i c a t i o n process appears to be present, planning strategies should be s u f f i c i e n t l y f l e x i b l e to adapt to changing c i r -cumstances and include a continuous process of monitoring i n order to detect those indicators that eventually suggest new changes. 7.1.4 EFFECTS OF SUBSIDIES TO HOUSING AND NEIGHBORHOOD IMPROVEMENT The occurrence of g e n t r i f i c a t i o n was suggested to be related primari-l y to external factors that would change the location variables as described i n the a c c e s s i b i l i t y matrix. Housing and neighborhood quality changes would re s u l t from these i n i t i a l changes i n location trends and, i n turn, contribute a new factor i n the overall location game. In t h i s scope, the effect that can be expected from subsidies to housing and neighborhood improvement would be that of alt e r i n g some of the effects of the g e n t r i f i c a t i o n process; at the same time, although.housing im-provement per se i s not a direct causal factor, at least some modification, i n the dynamics of the process could be expected i n terms of modifying part of the expectation effects associated with the process. This effect on expectation would be related toi the r e l a t i v e importance given to housing quality as compared with s o c i a l configuration as determinant components of a neighborhood's character. Rehabilitation subsidies i n gentrifying areas.can be expected to contribute to the upgrading'of these areas originated by the ge n t r i f i c a t i o n process. In th i s sense, r e h a b i l i t a t i o n would increase the upgrading expectation of these areas, thus contributing to accelerate the process. Rehabilitation subsidies, at the same time, would improve at least, p a r t i a l l y the a f f o r d a b i l i t y problem encountered by the low-income residents. However, for those low-income households not e l i g i b l e for the program, the effect would be just the opposite: by producing an overall upgrading, the re-sulting price increases and taxation, along with the higher maintenance pres-sures, would worsen the a f f o r d a b i l i t y problem for those groups. For eventual newcomers of low-income, r e h a b i l i t a t i o n subsidies on existing neighborhoods would also have the negative effect of higher prices arid higher demand from other groups in. the areas where these low-income newcomers would locate. Rehabilitation subsidies i n adjacent areas, on the other hand, can be expected to counteract to.at least some extent the downgrading of these neighborhoods. This physical improvement should slow down to some extent the succession process produced by the 'going poor' expectation. 153) However, t h i s effect can be considered weak, since i t does not touch the major determinants of the succession process: increased demand for low-income housing i n intermediate areas, provision of new houses i n peripheral areas only, and s o c i a l ..segregation attitude between the in-coming and the out-coming groups. Households e l i g i b l e f o r r e h a b i l i t a t i o n , on the other hand, have a net gain from r e h a b i l i t a t i o n subsidies i n terms of a f f o r d a b i l i t y ; i n t h i s sense, r e h a b i l i t a t i o n subsidies are similar to any other income subsidy for low-income groups. Alternative p o l i c i e s may meet the main objectives of r e h a b i l i t a t i o n programs (better low-income housing, neighborhood preservation) i n a more e f f i c i e n t way. For instance, a policy of higher densities and direct construction of low-, income housing i n central areas both increases the housing options for the poor (either old.residents or newcomers) and, by reducing the displacement effect produced by g e n t r i f i c a t i o n , contribute to a better preservation of ad-jacent neighborhoods. To the extent that g e n t r i f i c a t i o n i s caused by the desire to l i v e close to a d i v e r s i t y of groups, t h i s type of higher density/new low-income housing a.lso contributes to achieve t h i s d i v e r s i t y goal. Programs such as i n -centives to i n - f i l l housing, innovative housing design, may represent indeed.an optimum balance between the need of new housing and the preservation of both physical and s o c i a l components of neighborhood character. 7.2 POLICY IMPLICATIONS 7.2.1 GENERAL IMPLICATIONS The preceeding analysis has attempted to i d e n t i f y some effects of the g e n t r i f i c a t i o n process and r e h a b i l i t a t i o n subsidies i n gentrifying areas on 154 location options for different residents, p a r t i c u l a r l y low-income groups. In terms of public policy, the analysis suggests a planning situation defined by two major characteristics: A) Gen t r i f i c a t i o n i s characterized by a situation of long-term uncertainly. Although the process appears to be i n i t i a l l y triggered by the sp e c i f i c preferences of a r e l a t i v e l y transient group (a certain stage i n l i f e - c y c l e of the baby-boom generation), there i s some evidence at the same time of more profound changes i n location trends. To the extent that those trends are i n fact appearing, the long-term effect on low-income location options i s not clear, and w i l l vary depending on which are the major forces under-ly i n g the process. Some factors point at peripheral location of the poor, some others point at low-income location i n intermediate areas (between two peaks of high-income groups i n both central and peripheral areas), and some factors suggest a continuous succession of moves and increasing d i f f i -c u l t i e s to enter the location market altogether. B) Gentr i f i c a t i o n i s characterized by a situation of c o n f l i c t i n g goals. Both the process i t s e l f and eventual r e h a b i l i t a t i o n subsidy.programs i n gentri-fying areas tend to achieve some planning goals to the detriment of others. For instance,, the following c o n f l i c t s appear to be present: 1) G e n t r i f i c a t i o n brings improvement of the physical appearance of neigh-borhoods. This i s consistent with objectives related to h i s t o r i c pre-servation and the l i k e . However, i f displacement occurs, the improve-ment of some areas w i l l bring about the deterioration of other, adjacent areas. From th i s point of view, r e h a b i l i t a t i o n subsidies i n both resettlement and adjacent areas would amelliorate the problem; subsidies would be probably more e f f i c i e n t i n non-gentrifying neighborhoods. 2) Middle-class resettlement i s consistent with overall objectives of so c i a l mix. However, i t s effects on housing prices and maintenance 155 pressures may defeat this.purpose by producing a f f o r d a b i l i t y problems and eventual.displacement of some low-income groups. Rehabilitation subsidies for homeowners have a lim i t e d effect, here: maintenance may be subsidized, but the very improvement i t produces, multiplies, the price-increase type of a f f o r d a b i l i t y problem (i.e.. increase i n taxes).. This type of c o n f l i c t c a l l s for r e h a b i l i t a t i o n subsidies t i e d with p o l i c i e s oriented to increase the housing stock. 3) Resettlement reduces housing options for -in-coming' low-income groups, i n gentrifying areas. Rehabilitation subsidies, on the other hand,, while protecting the rest of the residents, further reduces the a v a i l a b i l i t y of housing f o r new, low-income households. Thus, r e h a b i l i t a t i o n .subsidies should be t i e d to new housing construction i n central areas. 4) G e n t r i f i c a t i o n may be r e f l e c t i n g profound changes i n location trends; in-the long term, there i s uncertainty with respect to a) which locations w i l l be open to low-income groups, b) how these locations w i l l meet location and housing needs for low-income groups. Rehabilitation programs, on the other, hand, aim.to immediate r e l i e f . Furthermore, two clauses often t i e d to r e h a b i l i t a t i o n (permanence.in the housing f o r a period i n order to gain forgiveness, and r e s t r i c t i o n of subsidies to pre-defined areas),, may c o n f l i c t with long-term options or needs.,; This .. uncertainty c a l l s f o r f l e x i b l e p o l i c i e s , which can adapt to changing .needs and options. 5) Private r e h a b i l i t a t i o n i s consistent with goals of preservation of resources, by making use of previous investment. However, middle-income resettlement, unless i t triggers new cons-truction through f i l t e r i n g and arbitrage, reduces the t o t a l supply of housing. Thus the advantages i n terms of resource preservation.are more apparent than r e a l , since the need for new construction, i s not reduced., and i n some cases i t may be increased. 7.2.2 FACTORS AFFECTING THE EFFECTS OF GENTRIFICATION'AND' REHABILITATION The implications suggested above refer to general effects that g e n t r i f i c a t i o n and r e h a b i l i t a t i o n subsidies may produce.in gentrifying.areas. The incidence of these effects w i l l vary i n each p a r t i c u l a r case,, depending on the nature of the long-term forces producing the g e n t r i f i c a t i o n trend. At the same time, l o c a l circumstances may suggest s p e c i f i c implications for policy,-depending on the factors present i n each, area undergoing gentrification*. Some factors that may affect the choice of policy i n gentrifying areas are suggested below: A) At the neighborhood l e v e l : Occupancy l e v e l (vacancy rates.): private r e h a b i l i t a t i o n targeted, to abandoned buildings w i l l maximize e f f i c i e n t u t i l i z a t i o n of .resources and produce a minimum displacement effect. Even i f r e h a b i l i t a t i o n , goes to other units, a high vacancy rate would indicate that the d i s -placement problem i s l i k e l y to be less s i g n i f i c a n t , as long as:vacant units have r e l a t i v e l y similar prices as the rest of the units.. . Type of housing occupancy: a high percentage of tenants would c a l l for some controls over r e h a b i l i t a t i o n of multiple-dwelling, f o r single- • ;• family use.. Since tenants are l i k e l y to be the f i r s t ones displaced., r e h a b i l i t a t i o n subsidies should, include tenant protection clauses,, either rent : ceilings or direct subsidy for tenants, and tenant p a r t i -cipation i n the r e h a b i l i t a t i o n decision. - A f f o r d a b i l i t y l e v e l of residents: residents who can afford tax increases produced by ove r a l l neighborhood upgrading are i n fact getting a benefit i n terms of appreciation of thei r housing value. Homeowners below this a f f o r d a b i l i t y l e v e l may be pushed out (forced to s e l l ) and would need a special type of subsidy. Social differences (income, age, ethnicity...) between.the in-coming groups and the old residents: big income differences may result i n an improvement of housing beyond what old residents can afford, thus creating tensions between groups. H i s t o r i c designation, of neighborhoods are l i k e l y to be rejected by very low-income residents. Rehabilitation subsidies and other programs should aim at minimizing the differences i n housing quality that each group can achieve. Ethnic, age differences, though they can eventually be an asset to the upgrading neighborhood, may produce additional tensions. Neighborhood p o l i c i e s should contem-plate the active p a r t i c i p a t i o n of a l l groups, oriented to detect tensions and f i n d a way to minimize them. Social f a b r i c , type of community services: a closely-knit neighborhood may suffer more severe disruption by massive resettlement. The location .and -type of the" improvement" works. should.-;;in: this-_case.be controlled,-, -in-order to minimize t h i s effect. Rehabilitation of buildings providing services to the community (grocery store, recreation centers) should be scrutinized. Relative importance of location for old residents: preservation p o l i c i e s should give p r i o r i t y to those households for which the i r location i s highly related to a c c e s s i b i l i t y to work, friendship and other neighbor-hood t i e s . Those groups less sensitive to location may. i n some cases be better-off since the sale of t h e i r houses i n upgraded areas may allow them to move to more desirable locations. - Land a v a i l a b i l i t y , existing densities: since at least two groups are competing for the same area, and a t h i r d group (new, low-income house-158 holds) i s also affected, available land for. development,, or potential for higher densities would present a good potential in-order to accomo-date the increased demand.. In any case, r e h a b i l i t a t i o n combined with housing stock increases should be encouraged. At the c i t y .level: Population growth rate and characteristics: since one of the maj.or problems of the g e n t r i f i c a t i o n process i s the.deterioration of.housing, options f o r new, low-income population,, the rate of low-income- immigra-tion i s an important element i n the assessment of the costs and .benefits of the process. At the same time, the rate, of growth of the type.of. households that trigger the process i s also important: a low proportion of these types of households would not produce a si g n i f i c a n t change i n trends; t h i s would happen only i f these groups grow at a rate such, that they a l t e r the 'tipping point' and change the expectations i n the affected neighborhoods. Land a v a i l a b i l i t y i n peripheral areas: t h i s factor relates d i r e c t l y to the occurrence of g e n t r i f i c a t i o n and the magnitude of the expectation' effect, which w i l l be higher i n cases.where r e s t r i c t i o n s to urban, ex-pansion reduce the average size of the l o t s in.peripheral areas.. - Rate of new building construction: the proportion of new ,construction as related to the rate of population growth may be an indicator, of the effects of segregation attitudes i n multiplying the succession process./ This would be true regardless of the occurrence of g e n t r i f i c a t i o n . Building construction beyond the direct demand produced by growth would-indicate, other things equal, at least some degree of i n s t a b i l i t y i n , some neighborhoods produced by the al t e r a t i o n of the s o c i a l mix con-sidered acceptable by some groups. On the other.hand, • a r e l a t i v e l y low rate of new building construction may be indicating the occurrence of g e n t r i f i c a t i o n , since i t may be resulting from changes of preferences of some groups for old rather than new housing. I f th i s i s the case, other factors should be analyzed to assess the impact of th i s p a r t i c u l a r indicator. Transportation changes, such as new t r a n s i t systems, new transportation corridors, may affect the a c c e s s i b i l i t y component of the location variables. Improved a c c e s s i b i l i t y can be argued to play against the occurrence.of g e n t r i f i c a t i o n to the extent that g e n t r i f i c a t i o n i s indeed related to the r e l a t i v e importance of a c c e s s i b i l i t y to CBD as compared to other location factors. Improved a c c e s s i b i l i t y , i n any case, reduces some of the problems affecting low-income groups i n gentrifying areas, to the extent that these improvements reduce the time and monetary costs of commuting i n areas where the poor may be displaced i n the g e n t r i f i c a t i o n process. In summary, the g e n t r i f i c a t i o n process appears to represent a planning situa-t i o n characterized by c o n f l i c t i n g goals and long-term uncertainty. Rehabilita-tio n subsidies i n t h i s situation contribute to achieve some goals but defeat others. The analysis submitted i n t h i s research suggested some of the groups affected (either p o s i t i v e l y or negatively) by both g e n t r i f i c a t i o n and rehabi-l i t a t i o n , some indicators that can be useful for the assessment of the incidence of these effects i n sp e c i f i c cases, and some planning tools that could be used either i n conjunction or as an alternative to r e h a b i l i t a t i o n i n order to improve the achievement of pa r t i c u l a r goals. The conclusions and implications outlined above are primarily based on the conceptual location model that was derived from the review of different bodies of location theory. Although the major variables presented i n the model appear to be consistent with most of the approaches to location theory that could be reviewed i n the research, the v a l i d i t y of the conclusions and i m p l i -1.60 cations i s s t i l l subject to the accuracy of the model i n describing r e a l l i f e s ituations. In th i s sense, both a more complete development of the model and i t s testing by empirical research may affect the set of conclusions presented here. This uncertainly with respect to the r e l i a b i l i t y of the model adds to the uncertainty of long-term effects as predicted from the model i t s e l f and suggest a more general implication for planning policy. Whatever the p r i o r i t i e s determined by the policymaker i n terms of different planning goals, the actual planning decisions should be such that permit the achievement of these goals within this framework of uncertainty. Thus, rather than one pa r t i c u l a r program (such as r e h a b i l i t a t i o n or neighborhood improvement subsidies), this situation c a l l s for a more comprehensive planning strategy which includes a range of different p o l i c i e s to apply as different factors change i n each case and at each point i n time; at the same time the strategy should include a monitoring system capable of detecting changes over time of the long-term trends. The research has suggested some of these possible planning tools, and some indicators that could be helpful formonitoring. The actual choice of plans and programs depends on the p r i o r i t i e s determined by the policymaker. 7.3 FUTURE USES OF THE LOCATION MODEL The research could be considered finished at th i s point since, the conclusions and implications already outlined have met, within the time and resource l i m i t a t i o n s , the main objectives i n i t i a l l y stated. However, the l o -cation model developed throughout the research i s an additional output that can serve a number of uses. Some comments on the uses of the model were, therefore, considered worth pointing out. The location model as presented i n the research is. conceptual i n nature. I t provides a very general framework to synthesize a set of variables and t h e i r interrelationships. I t i s at a very i n i t i a l stage of development 161' and, as such, needs further study oriented to define more accurately each of the variables, to design an operational counterpart, and to devise a more precise measuring of the interrelationships among the main variables. Given these considerations, some comments can be made on how the model works and what type of uses could be expected from i t . Description of the model. The model i s i n fact a set of two interrelated sub-models: an a c c e s s i b i l i t y matrix, which summarizes location preferences i n f i v e dimensions, and a dynamic model of household moves. The dynamic model des- . cribes the occurrence and volume of household moves and the a c c e s s i b i l i t y matrix describes the s p a t i a l d i s t r i b u t i o n of residents, either with a f i x e d population or the d i s t r i b u t i o n of moves as detected i n the dynamic model. In general, the model permits the transformation of a set of data obtained from census i n f o r -mation and land surveys into a s p a t i a l d i s t r i b u t i o n of residents and household moves. In doing t h i s , i t allows a description of different areas i n terms of physical and s o c i a l characteristics, as well as a prediction of possible changes. One way of describing the mechanics of the model, i t s inputs and out-puts, may be as follows: 1) The s p a t i a l characteristics.of a c i t y can be plotted i n a map, and areas can be i d e n t i f i e d with similar a c c e s s i b i l i t y to CBD, l o t size configuration and geographic t r a i t s . Each of these areas i s given a f i r s t , two-dimen-sional a c c e s s i b i l i t y coefficient (geometry and geography dimensions). 2) The s p a t i a l d i s t r i b u t i o n of the existing population, defined by s o c i a l status, familism and ethnicity, can be also plotted. This d i s t r i b u t i o n would affect the i n i t i a l boundaries of homogeneous areas. An a c c e s s i b i l i t y c o e f f i c i e n t can then be derived for the resulting areas; t h i s new c o e f f i -cient would include, both the two i n i t i a l dimensions plus the s o c i a l segre-gation coefficients derived from the existing s p a t i a l d i s t r i b u t i o n c 162 3) Housing quality by area can be predicted '/and checked against the actual housing qu a l i t y i n different areas, and a value can be found f o r the housing (neighborhood) quality dimension;): At t h i s point, the a c c e s s i b i l i t y matrix i n a l l the dimensions has been b u i l t , and r e l a t i v e aggregate a c c e s s i b i l i t y coefficients have been found for each of the homogeneous areas. L) Given an a c c e s s i b i l i t y matrix which describes both the physical and the s o c i a l components of location, the dynamic part of the model can be used.. The next input i s , thus, population forecasts by s o c i a l type and forecasts of housing additions. 5) The next output i s a forecast of moves. By assigning the new population to areas having a c c e s s i b i l i t y coefficients similar to those chosen by the f i r s t residents, the model should be able to detect where the moves would st a r t , i n which direction they are l i k e l y to proceed, and how the process would affect the s o c i a l and physical characteristics of both the affected and the surrounding neighborhood. 6) To the extent that the a c c e s s i b i l i t y matrix accurately describes segrega-t i o n attitudes, the model should also show i n which cases t h i s i n i t i a l change i n location affects the s t a b i l i t y of a neighborhood. " I n s t a b i l i t y " i n t h i s sense would be measured as the difference between the i n i t i a l s o c i a l composition and the one resulting after the f i r s t i t e r a t i o n of moves; th i s i n s t a b i l i t y i s assumed to be d i r e c t l y correlated to the seggregation a t t i -tude between the two groups at play. Thus, an i n i t i a l set of data (land surveys, distance to CBD, l o t sizes, population - existing and forecasted - by rank, familism and ethnicity, housing construction) -can been used to predict a pattern of r e s i d e n t i a l d i s t r i -bution, a pattern of location moves and a pattern of physical and soc i a l changes i n different areas of the c i t y . Uses of the model. To the extent that these general data are available, some possible uses of the model are suggested as follows: 1) The model can be used for direct forecasting: the f i r s t and more obvious use would be that of predicting the population composition and i t s counter-part i n terms of housing and neighborhood characteristics i n different areas, and the changes that can be predicted from the population forecasting. 2) The model can be used to detect eventual problem areas: t h i s f i r s t fore-casting should be able to detect cases i n which growth produces some types of problems, such as household moves beyond what could be considered the l i m i t i n terms of neighborhood s t a b i l i t y insufficiency of land or housing, stock i n s p e c i f i c areas, displacement of some groups to less .desirable locations, types of s o c i a l mix incompatible with some types of segregation attitudes, etc. 3) The model can be used to assess advantages and problems of planning p o l i -cies: a number of p o l i c i e s can be introduced to the model by assuming a spec i f i c effect of these p o l i c i e s on the values of the a c c e s s i b i l i t y matrix. For instance: Zoning regulations affect the l o t sizes i n the affected areas: there-fore, they modify the geometric dimension of the a c c e s s i b i l i t y c o e f f i -cient. Urban growth boundaries can be introduced i n a similar way. Rehabilitation subsidies modify the housing q u a l i t y dimension and the resulting coefficient i n each of the affected areas. - Transportation programs modify the CBD a c c e s s i b i l i t y or the geometric coeffic i e n t i n those areas where transportation i s improved. Immigration p o l i c i e s affect the s o c i a l c o e f f i c i e n t , and so on. These coefficients modified by policy would affect the overall s p a t i a l d i s t r i b u t i o n and the pattern of moves. The model can be used to predict those changes and detect eventual gains or losses for different groups 16*£ i n terms of location changes. 4.) The model can be used for planning analysis: some external forces not con-t r o l l e d by the policymaker (such as changes i n soci a l attitudes, economic expansion or recession) could also be introduced by assuming an effect on the coefficients of the a c c e s s i b i l i t y matrix. The incidence of such ex-ternal factors could be derived by comparing outputs of the model before and after the. .occurrence of these external phenomena. Another way of using the model f o r analysis would be reversing the mechanics of the model: for instance, changes i n location trends could be de- . tected by comparing a c c e s s i b i l i t y matrices derived from an area at two d i f f e r -ent points i n time; differences i n s p a t i a l d i s t r i b u t i o n of residents would result i n different values of some of the a c c e s s i b i l i t y c o e f f i c i e n t s . This type of analysis may be helpful to explain the causes of the changes i n location trends. This l a s t use i s essentially the method that was chosen for the analysis of the g e n t r i f i c a t i o n process, i n which some causal factors of gentri-f i c a t i o n were related to factors that would change the values of the accessi- . b i l i t y matrix. The analysis was carried out i n a very conceptual l e v e l , since the purpose of the research was to point out general rather than s p e c i f i c effect of the g e n t r i f i c a t i o n process. A more s p e c i f i c application i s also possible, however,, i f s u f f i c i e n t data are available. For instance, an assessment on the probability of gentri-f i c a t i o n to occur i n Vancouver could be based on essentially the same reasoning. Let/us assume that i n Vancouver the following variables are present: High rate of population growth (high immigration rate) - Limited land to expand (Agricultural Land Reserve, urban growth boundaries, natural geographic.barriers). Saturation of population capacity under the current zoning. - Relatively low densities i n central areas. 165 - Growing importance of h i s t o r i c preservation movements. Given these assumptions, i t would be reasonable to predict that at least some degree of private r e h a b i l i t a t i o n should be underway or should occur i n the near future. By deriving the corresponding a c c e s s i b i l i t y matrix, before and after the occurrence of these eventual causal factors, and. introducing population growth forecasting by soc i a l type, i t should be possible to predict, where this g e n t r i f i c a t i o n process i s l i k e l y to s t a r t , which neighborhoods would be affected and which groups would gain and lose i n the process. At the same time, i f g e n t r i f i c a t i o n indeed occurs i n Vancouver, planners would be faced to the choice of p o l i c i e s which would present c o n f l i c t -ing planning goals: the housing needs of the newcomers are i n c o n f l i c t with the rights of the residents to have thei r housing choice protected. Control of urban expansion c o n f l i c t s with the need of new housing. Goals of di v e r s i t y of l i f e s t y l e s would be both looked for and defeated by high-income resettlement. The model presented here may prove a useful tool to assess i n more spe c i f i c ways the incidence of these c o n f l i c t s and the effects on location options of p o l i c i e s oriented to meet each of these goals. 166 BIBLIOGRAPHY Abu-Lughod, Janet. "The Consumer Votes by Moving", N. Foote, J . Abu-Lughod, Mary M. Foley & Louis Winnick, Housing Choices and Housing Constraints. McGraw-Hill Book Co., I960. Alonso, William. Location and Land Use, a General Theory of Land Rent. Harvard University Press, 1964. Bailey, Martin. "Notes on the Economics of Residential Zoning and Urban Renewal". Urban Land Economics, Vol. 33, #5, August 1959, p. 288-292. Barret, Frank.A. "Residential Search Behavior:. A Study of Intra-Urban Reloca-ti o n i n Toronto." Geographical Monographs #1, York University -Atkinson College, Toronto, 1973. Burgess, Ernest W. The Growth of the City; An Introduction to a Research Project. George A. Theodorson, editor. Studies i n Human Ecology, The Pensylvania State University, 1961. CMHC - U.S. Department of Housing & Urban Development. R e v i t a l i z i n g North- American Neighborhoods: A Comparison of Canada and U.S. Programs for  Neighborhood Preservation and Housing Rehabilitation, C.M.H.C., Ottawa, 1979. - Neighborhood Improvement Program Operators Handbook. CMHC, Ottawa, 1978. - CMHC & the National Housing Act. CMHC, Ottawa, 1977. - Federal Housing Programs: A Quick Review for Ready Reference. CMHC, Ottawa, 1978. - Residential Rehabilitation Assistance Program: Standards for the Rehabi- l i t a t i o n of Residential Buildings. CMHC, Ottawa, 1978. CMHC - Residential Rehabilitation Program: Delivery Handbook. CMHC, Ottawa, 1978. Downs, Anthony. "Alternative Futures f o r the.American Ghetto". Michael " Stegman, editor, 'Housing & Economics-; The American Dilemna' . The MIT Press, Mass., 1970. Duncan, Otis Dudley; Duncan, Beverly. "The Negro Population i n Chicago: A Study of Residential Succession." The University of Chicago Press, 1957.. Firey, Walter. Sentiment and Symbolism as Ecological Variables. ,;G. ,-iTheodorson, editor, ;Studies .in"Human"Ecology,--.Pensylvani-a State University, 1961. Gale, Dennis E. "Middle Class Resettlement i n Older Neighborhoods: The Evidence and the Implications". A.P.A. Journal, Vol. 45, #9, July 1979, p. 293-304. Gera, Surendra; Kuhn, Peter. "Residential and Job Location and the Journey-to-Work: A Review and Theoretical Perspective". Urban Paper No. 1, Economic Council of Canada, Ottawa, 1977. CMHC CMHC CMHC CMHC • 167 Grisby, William G. "The F i l t e r i n g Process"... Housing Markets and Public Policy, University of Pennsylvania Press, 1963. Guest, Avery M. "Residential Segregation i n Urban Areas". Kent P. Schwirian, editor, Contemporary Topics i n Urban Sociology, The Ohio State University. New Jersey, 1977. Gutman, Emil. Housing: A Problem i n Skid Row Rehabilitation. M.A. Thesis -School of Community Planning - U.B.C., 1972. Hartman, Chester. "Comment on Neighborhood Re v i t a l i z a t i o n and Displacement: A Review of Evidence". Journal of the American Planning Association, Vol. L5, U, October 1979, p. LSS-L9L. Hawley, Amos H. "Ecology & Human Ecology". G. Theodorson, editor, Studies  i n Human Ecology, Pennsylvania State University, 1961. Hoover, Edgard M.; Vernon, Raymond. Anatomy of a Metropolis. Harvard University Press, 1959. Hoyt, Homer. The Structure, and Growth of Residential Neighborhoods i n the United States. Federal Housing Administration, N.Y. 1939. Jonassen, Christen T. "Cultural Variables i n the Ecology of an Ethnic Group". G. Theodorson, editor, Studies i n Human Ecology, Pennsylvania State University, 1961. Kosa, John. "Hungarian Immigrants i n North America: Their Residential Mobility and Ecology". G. Theodorson, editor, Studies i n Human Ecology, Pennsylvania State University, 196l. Kuper, Leo; Watts, Hilstan; Davies, Ronald. Durban: A Study i n Racial Ecology. Golumbia University Press,.London, 1958. Lansing, John B.; C l i f t o n , Charles W.; Morgan> James N. New -Homes and Poor  People: A Study of Chains of Moves. Ins t i t u t e . f o r Social Research (I.S.R.) The University of Michigan, 1969. Lessinger, Jack. The Case for Scatteration. JAIP, Vol. 28, §3, August 1962, p. 159-169. Leven, Charles L..; L i t t l e , James T.; Nourse, Hugh 0.; Read, R.B. Neighborhood  Change - Lessons i n the Dynamics of Urban Decay. Praeger Publishers Inc., N.Y., 1976. L i t t l e , James T. Residential Preferences, Neighborhood F i l t e r i n g , and Neigh- borhood Change. U.S. Department of Housing and Urban Development. Institute for Urban and Regional Studies, Washington University, St. Louis, 1974. L i t t l e , James T.; Nourse, Hugh 0.; Phares, Donald. The .Neighborhood Succession  Process: A Summary. Institute for Urban & Regional Studies, Washington University, St. Louis, 1975. Meadows, George Richard; C a l l , Steven T. Combining Housing Market Trends and Resident Attitudes i n Planning Urban: R e v i t a l i z a t i o n . AIP Journal, Vol. LL, #3, July 1978, p. 297-305. 168 The National Urban Coalition. Displacement:.. City Nieghborhoods i n Transition. N.Y.: N.U.C., 1979. Park, Robert Ezra. "Human Ecology", G. Theodorson, editor, Studies i n Human  Ecology, The Pennsylvania State University, 1 9 6 l . Sands, Gary. "Housing Turnover: Assessing i t s Relevance to Public Policy". AIP Journal, Vol. 42, #4, October 1976. Schwirian, Kent P. "Internal Structure of the Metropolis". K. Schwirian, editor, Contemporary Topics i n Urban Sociology, General Learning Press, New Jersey, 1979. Seeman, Albert L. "Communities i n the Salt Lake Basin". GV Theodorson, editor, Studies i n Human Ecology, The Pennsylvania State University, 196l. Shevky, Eshref; B e l l , Wendell. Social Area Analysis, G. Theodorson, editor, Studies i n Human Ecology, The Pennsylvania State University, 1961. Smith, Wallace F. Housing, the Social and Economic Elements. University of Ca l i f o r n i a Press, 1971. Smith, Wallace F. F i l t e r i n g & Neighborhood Change. Michael Stegman, editor, Housing & Economics: The American Dilemna. The MIT Press, 1970.-S t e r n l i e l s , George; Burchell, Robert; Hughes, James; James, Franklin. "Housing Abandonment i n the Urban Core". AIP Journal, Vol. 4-0, §5, September 1974., p. 321-324.. Sumka, Howard J. "Neighborhood R e v i t a l i z a t i o n and Displacement: A Review of Evidence". Journal of the American Planning Assoc.,, Vol. 4-5, #4-, October 1979, p. 480-4-87. Taeuber, Karl E.; Taeuber, Alma F. Negroes, i n C i t i e s : Residential Segregation  and Neighborhood Change. Aldine Publishing Co., Chicago, 1965. Theodorson, George A. editor,. Studies in.Human Ecology. The Pennsylvania State University, 196l. Watson, Chris S. "Vacancy Chains, F i l t e r i n g , and the Public Sector". JAIP, Vol. 40, #5, September 1974, p. 346-352. Western Research Corporation. Ltd. Study of the Economics of Conversion i n the  Kit s i l a n o Area, City of Vancouver Planning Department, Vancouver, 1975. Wingo, London J r . Transportation and Urban Land. Resources of the Future, Washington, D.C, 1961. Zorbaugh, Harvey W. "The Natural Areas of the City". G. Theodorson, editor, Studies i n Human Ecology, The Pennsylvania State University, 196l. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0095458/manifest

Comment

Related Items