the social spatial segregation in the cities of latin america

8/20/2019 The Social Spatial Segregation in the Cities of Latin America

1/44

The Social Spatial Segregation in theCities of Latin America

Francisco Sabatini

Inter-American Development BankSustainable Development Department

Social Programs Division


2/44

Dr. Francesco Sabatini is Profesor at the Institute of Urban Studies of the Catholic University of Chile.He holds a Ph.D in Urban Planning from the University of California, Los Angeles, a degree in Sociologyand a Master in Urban Development from the Catholic University of Chile


3/44

Presentation

The cities offer Latin America and the Caribbean their best opportunity for economic and social devel-

opment. Aside from concentrating on more than two thirds of the population, it is estimated that urbanactivities will generate more than 75% of the expected growth of the Gross Domestic Product in the nexttwo decades. Therefore, in order to improve the competitiveness of economic activities in national andglobal markets, it is necessary to not only maintain healthy economic policies and eliminate commerce

barriers, but also to improve the cities’ abilities to provide an efficient platform to support the establish-ment and development of many types of companies. Adequate provision of infrastructure and good livingconditions, factors attracting skilled labor and industrialists to the cities, are crucial initiatives of localeconomic development, yet they are not sufficient. Access to well-paying jobs and good urban servicesare critical in order to increase the populations’ opportunities to live according to their desires and values.Yet, true social development will not occur unless concrete measures are taken to remove other barriers,including spatial segregation of the poorest households and ethnic or cultural discrimination. Social in-clusion and economic development are equally important in reducing violence and other antisocial behav-

iors. In summary, a more inclusive city is a more productive city, encouraging growing markets for local products and services, thus contributing to the acceleration of economic growth.

This document analyzes the first of the barriers mentioned above, spatial segregation of the pooresthouseholds. It discusses the characteristics and trends of residential segregation in the cities of LatinAmerica, its causes and consequences, the state of research in this field and the policies that could controlspatial segregation. The document emphasizes the fact that segregation is a complex phenomenon withsome positive dimensions from the perspective of the social policies, as it could help improve their target-ing and efficiency. The negative dimensions of the phenomenon are also identified, such as social stig-matization of the low-income or minority-occupied neighborhoods. These considerations are important inthe implementation of one of the central proposals of the Bank’s Social Development Strategy, whichadvocates the coordination of policies and programs in the territory. 1

I hope the publication of this study helps disseminate the available knowledge of this phenomenon to public policy managers and Bank staff, contributing to improve the design and execution of territoriallycentered social development policies and programs.

Eduardo RojasPrincipal Urban Development SpecialistSocial Development DivisionSustainable Development Department

1 I.A.D.B., “Social Development. Strategy Document,” Washington, DC, 2003.


4/44

Table of Contents

Introduction 1

Characteristics and Trends of Segregation 3

Definitions and Precisions of Approach 7

Popular Explanations of Segregation 12

Impacts 18

Policy to Control Segregation and its Relation to theTerritorial Strategy of Social Program Coordination 23

AnnexThe Latin American Segregation Paradigm 33

References 37


5/44

Introduction

This report discusses the characteristics and trends

of residential segregation in the cities of LatinAmerica, the causes and consequences, the statusof the research within this field and the policiesthat would control segregation. As the report con-cludes, there is a discussion of the possible impli-cations of the Inter-American DevelopmentBank’s (I.A.D.B.) continual efforts in targetingand improving the efficiency of the urban social

policies.

One of the three goals of the Bank’s new SocialDevelopment Strategy is the territorially integra-

tion of the provision of social services. One mustconsider the fact that since 1994, the Bank allo-cates 50 percent of its resources to the poor. Forthis reason, the Strategy has as much importancefor the Bank as for the countries of the Region.

Questioning the relevancy of segregation is par-ticularly adequate considering how segregatedLatin American cities are. Nevertheless, linkingthe strategy of targeting and social policy effec-tiveness with the reality of our segregated cities

provokes yet another question that we should not

ignore: Are we implying that the spatial segrega-tion of the poor urban communities, a fact unani-mously condemned, could help a strategy orientedat improving social policies? The segregated loca-tions of the poor would favor the strategy to terri-torially integrate the provision of social services.Wouldn’t this be a contradiction?

Most of the Latin American researchers and urbanspecialists would reject the idea that Governmentaims “to take advantage of” segregation of the

poor people in order to make the social programsmore targeted and efficient. This is not only con-cerning moral repairs. Likewise, critics would callattention to the politics of a government’s implicitlegitimization of segregating the poor through thistype of action. This will amount to an acceptanceof the noticeable social inequalities, a fundamen-tal cause of the existing spatial segregation in thecities.

However, the conclusion of the present document

is different. Segregation is a complex phenome-non with a clearly positive dimension: the spatialconcentration of the studied social group, whichcould help to improve the targeting and efficiencyof the social policies. Complementarily, the mostnegative dimensions of the phenomenon, specifi-cally the configuration of neighborhoods and so-cially homogenous areas and their social stigmas,are the least helpful in the territorial targetingstrategy of social policies. On the other hand, thisreport concludes by emphasizing the general con-tribution that the social policies can make by con-

trolling segregation, beyond specific strategies ofterritorial coordination.

The first chapter of this report focuses on thecharacterization of the phenomenon of urban resi-dential segregation in Latin America, based onavailable studies and in our own research efforts.We will describe the traditional pattern of segre-gation in Latin American cities and its most recenttrends. We will emphasize familiar aspects, forexample; the strong spatial concentration of theupper and growing middle classes and the concen-

tration of the poor as well as other generallyavoided socially diverse groups present in the af-fluent areas of our cities. Amidst these trends, wewill emphasize the modification of the traditional

pattern of segregation recognized in most recentdecades.

The second chapter contains a composed defini-tion of segregation and various precisions con-cerning approach and methodologies. The defini-tion highlights different objective aspects of seg-regation, which are important in deciphering theirdistinct practical implications. In addition, wewill emphasize the relevance of the subjectivecharacteristics of segregation. The formation ofterritorial stigmas, as much by society as by the

poor people themselves, has significantly negativeimplications. Finally, among the precisions ofmethod and approach, we will discuss the natureof the segregation process, considering the posi-tive as well as the negative consequences, and the

1


6/44

importance of the geographic scale of the locationand the method of measurement.

In the third chapter, we critically analyze the pre-dominant approaches to the urban residential seg-regation in Latin America. We will review, with

particular concern, a series of statements and per-spectives of the causes of the phenomenon that weconsider incorrect. Specifically, we will attempt toreject the virtual “demonization” that has beenassociated with segregation, creating a more bal-anced and practical vision, favorable to policydesign. This balance, as well as the pragmatic ap-

proach, is strengthened when emphasizing thatthis is a phenomenon rather than a problem, and

by stressing its changing character as a process. Inone annex, we include a more detailed critique ofsegregation through an analysis of the most popu-

lar approach in Latin America. We consider itsignificantly detrimental to the advancement of

public policy in issues of urban residential segre-gation.

The fourth chapter focuses on the impacts on thequality of life and the perpetuation of poverty.We will emphasize the effects of social disintegra-tion caused by spatial segregation, in a contextdefined by economic liberalization and otherchanges associated with the “globalization” of oureconomies. We will demonstrate that in past dec-

ades, spatial segregation of the poor, having bothnegative and positive effects, has been deprived ofthe latter and faces a worsening of its negativeconsequences.

In the fifth chapter, we propose policy measuresdesigned to neutralize the worst effects of segre-gation, while at the same time aiming to take ad-vantage of the positive impacts of the territorialcoordination of social services, which can emergefrom distinct situations and geographic scales ofsegregation. Segregation is not bad per se, as we

would have argued in previous chapters. The poli-cies and measures proposed guide the process ofsegregation towards solutions with positive im-

pacts on the poor and the social policies benefiting

them. We will emphasize the importance of adopt-ing a mixed strategy of spatial interventions, suchas the reduction of the scale of segregation and, inspecial cases, the dispersion of the poor, mixedwith other social policies. Additionally, we willfocus on those strategies helping to improve theaccessibility and mobility of the poor within the

boundaries of each city, and other strategies thatincite greater possibilities of interaction within thedifferent social groups.

The character of this report is strongly influenced by the limited empirical research on residentialsegregation in Latin American cities. On onehand, the studies that put the hypothesis to the testare scarce, a basic requirement of scientific re-search. On the other hand, currently there are nostatistical series or comparable measurements be-

tween cities, except maybe the most recent case inBrazil.

Nevertheless, the research of the subject of resi-dential segregation through statistical studies hasintrinsic limitations, which justifies different ap-

proaches, such as the present, in which the em- piric-qualitative analysis and the conceptual interpretation reinforce themselves. The statistics usedin developing countries, and especially in theUnited States, which has the greatest tradition andsophistication in segregation measurement, show

serious methodological and theoretical limitations.For example, it is worth noting that the dissimilar-ity index, vastly utilized internationally due to itssimplicity, measures a dimension of segregationthat proves to be the most positive of the phe-nomenon. Larger dissimilarity indexes do notnecessarily represent a negative factor.

Along with these, we will argue that the we cannot solve the needed advancement required in thesegregation analysis in Latin America thoughadded data and applications of quantitative meth-

ods, but mainly through empirical research thatallows us to surpass the overly simplistic visions

by which we have interpreted the reality of ourcities.

2


7/44

Characteristics and Trends of Segregation

Throughout most of the 20 th century, the cities ofLatin America exhibited a pattern of residentialsegregation similar to the European “compactcity” model. The groups higher on the social scalewere concentrated in the central area, containingalso the best quality of construction and architec-ture.

The cities decay, socially and physically, towardsthe periphery, with the exception of the geo-graphic area in which a sort of cone of “modern”city formed during the 20 th century. Examples ofsuch geographically influenced construction areresidential and commercial neighborhoods occu-

pied by high-income and rapidly ascendinggroups. In this aspect, the Latin American citydoes not differ much from the Continental-European Mediterranean city. Paris, like manycities of the old continent, has a concentrated coneof families with higher income that have settled ina defined geographic area - in this case, towardsthe West. Perhaps the most notable difference be-tween the Latin American upper class and theirEuropean counterparts is the faster pace at whichthey have deserted the center of the cities.

An alternative model of a capitalistic city is thatof the Anglo-American suburbs. There, the upperclass, inspired by an anti-urban ideology of prot-estant origins, occupied the periphery of the cities,while the lower class individuals populated thecentral areas.

The suburbanization of the urban upper class wasa process initiated as early as the middle of the20th century. In London, a pioneering city of thisurban revolution, suburbs emerged, even when thetechnology of transportation (carriages) was rather

precarious. The evolution of the train, the street-car and, much later, the automobile, would facili-tate the development of the suburb, having a sig-nificant impact on the United States.

The model of the compact city does not alwaysappear in pure form in Latin America. There are atleast three explanations:

• The degree and historical period in which theupper class have left the center of each city,similar to the degree of concentration of thesegroups in a particular area of urban growth,varies from city to city. For example, in Bo-gotá the degree of spatial concentration of theupper class is greater than in Mexico City; andin Lima, the upper class left the central areasmany decades before their counterparts inSanto Domingo.

• The influence of the cultural pattern of thesuburban city has affected our continent, as weexplain later in the presentation, much more inthe adoption of architectural fashions and ur-

ban styles than in the identity formation andconsolidation of the social group present in thesuburban movement in the countries of origin.

• Due to the importance of the European coloni-zation in the formation of the urban systems,most of the cities of Latin America were con-structed along the coast or a river, introducingrandom geographic factors having an influenceon the urban form and moving away the au-thentic model cities.

TRADITIONAL PATTERN OFSEGREGATION

We can summarize the “traditional” pattern ofLatin American segregation as having the charac-teristics listed below. After this enumeration, wewill see that the economic reform and politicalchanges occurring around the beginning of the1980s encouraged altercations to this pattern. Thecharacteristics are:• The significant spatial concentration of upper

class groups and the ascending middle classesto an extreme in only one zone of the city, withits apex in the historical center, and a clear di-rection of expansion towards the periphery (re-ferred to as the “high-income neighborhood”).

• The conformation of ample housing areas forthe poor, mainly in distant and poorly serviced

peripheral areas, but also in deteriorated sec-tors close to the city-center.

3


8/44

• The significant social diversity of the “high-income neighborhoods”, where, along with vir-tually all of the upper class, certain middle andlow class groups live, with the important ex-ception of “laborers,” “informal dwellers” or“marginalized groups,” as they have domi-nated the poorest groups in different periods.

The first two characteristics are amply recognizedin the specialized literature, even amongst authorsoutside of the Region (an example of a recent textis by Meyer and Bähr, 2001). However, this isnot true concerning the third characteristic, whichis generally ignored.

Concerning the final characteristic, the social di-versity of the “high-rent neighborhoods,” it is im-

portant to contrast the situation of segregation in

Latin American cities with cities in the UnitedStates. In the US, suburbs tend to be more ho-mogenous in social terms. In fact, a communitygenerally consists of a neighbors’ association that,working openly or implicitly with the local mu-nicipality, resort to a series of legal and formalmeasures to exclude the lower class (with restric-tions on building height, minimum lot size, speci-fication of architectural typology etc.). The homo-geneity is clear in racial terms: on average awhite resident of a metropolitan area of the UnitedStates lives in census tracks where 83 percent of

the population is white, whereas the typical blackresident lives in census tracks where only 54 per-cent population is black (Briggs, 2001 based onthe census data through the year 2000). The ex-treme social diversity in areas of discriminatedgroups, especially Afro-Americans, is expressedin the fact that the denomination of “ ghettos” isoften applied to segregated areas where the Afro-American population represents only 40 percentof the population (Jargowsky, 1997).

The Latin American cities demonstrate an inverse

situation: the areas inhabited by the very poor aremuch more socially homogenous than the residen-tial areas of the upper class. For example, in 1990in Mexico City, the upper class (7.5 percent of the

population) represented only one third of the population of the 23 delegations and richest mu-nicipalities of the city (out of 183). However, the

poorest social layer (18 percents of the popula-tion) represented 79.4 percent of the occupants of

the 35 delegations and poorest municipalities ofthe city (Rubalcava and Schteingart, 1999). Re-cent studies in Rio de Janeiro produce similar re-sults, confirming the social diversity of even themost affluent areas of Latin America (Preteceilleand Ribeiro, 1999; Ribeiro, 2000).

THE LARGE SCALE OF SEGREGATION

We can integrate the first two characteristics ofthe traditional Latin American pattern of segrega-tion noted with the idea of large-scale residentialsegregation. In fact, many specialists mention thischaracteristic of Latin American cities in literatureand debates. Nevertheless, it needs two explana-tions:• It is also a characteristic of cities in other re-

gions of the world, such as in the United

States, where rich and poor, Anglo-Saxon andminorities, appear to be clearly segregatedwhen observing the overall territory of the cit-ies; and

• A methodological question in this judgment israised, relative to the inclusive viewpoint bywhich it is based. To the contrary, if we fo-cused on the smaller geographic scales, wewould have to conclude that the “Latin Ameri-can high-income neighborhoods” are not verysegregated, because of the social diversity re-cently discussed.

How can we make the condition of diversity in the“high-income neighborhoods” compatible withthe obvious characteristics of large-scale segrega-tion shown by the Latin American cities? The key

point here is that certain neighborhoods excludethe poor, groups that make up more than 50 per-cent of the population of each city. Hypotheti-cally interpreted, one could attribute this fact totwo historical factors:• The prevalence of the European urban culture

in the Latin American upper class, more so

than the Anglo-American, leads them to repro-duce in the zones where they concentrated inthe 20 th Century, the social and city structureof industrial European cities. All but the “in-formal” poor are admitted to the project 2.

2 We argue later that Latin American urban upper classhave undertaken somewhat of a historical project inconstructing the suburbanized areas like pieces of “cit-

4


9/44

Those groups publicize the condition of being poor countries, thus societies exclude themfrom the “high-income neighborhoods,” caus-ing substantial agglomerations of “informal”

poverty in the peripheries of our cities.• The intrinsically speculative operation of the

land markets result in the proprietors settingland prices for higher income social groupsthat overflow into the area. This mechanism,which we will discuss in detail later, wouldhave functioned as a method of expulsion ofthe poor families from the “high-incomeneighborhoods” within a very short period af-ter their origination.

OTHER TYPES OF RESIDENTIALSEGREGATION

The Latin American city shows other forms ofresidential segregation in addition to separationaccording to socioeconomic status. There are ra-cial, ethnic and age differences in the urban popu-lation that have some manifestation in specialterms. Unfortunately, researchers have practicallyexcluded these types of segregation from socialresearch (some exceptions are the works of Telles,1992a and 1992b; Germain and Polèse, 1996; andHiernaux, 2000) 3. The efforts have been concen-trated in the socioeconomic aspects, and eventhese efforts have been minor and deficient, as we

will argue in Chapter 3.

ies within developed countries” having a significantcontinental European influence.3 Cities of the United States with the majority of the

population of Hispanic origin represent an area of eth-nic segregation cases that Latin American researchesshould study in the future (Mike Davis, 2000, in astudy of this phenomenon).

In any case, the few existing empirical studies prove that segregation of racially or ethnicallydiscriminated groups tends to correspond withsegregation of low socioeconomic groups (in thecase of Brazil, see Telles, 1992a). Coincidently, astudy relating to Spanish immigration in BuenosAires during 1850-1930 dismisses the notion thatethnic enclaves had formed in the city (Moya,1998).

In comparison with the United States, where eth-nic and racial residential segregation is verymarked, Latin American cities present a different

panorama. According to the only apparent comparable quantitative study to those of the UnitedStates, the ethnic segregation of the African popu-lation in Brazilian cities is significantly smallerthan in the Americans (Telles, 1992b). Although

the study consists of outdated numbers, the subse-quent table replicates the data in this study. We

base this on values of the index of dissimilarities,the most used index in the international study ofsegregation. The value indicates the percentage ofthe population of African origin having to changeresidential area (census district) in order toachieve a homogenous distribution within the city.

TABLE 1Index of Residential Dissimilarity by Race in the 10 Greater Brazilian Metropolitan Areasand in select Metropolitan Areas of the United States: 1980.

BRAZILLIAN CITIES INDEX US CITIES INDEXSao Paulo 37 New York 73Río de Janeiro 37 Los Angeles 86Belo Horizonte 41 Chicago 76Porto Alegre 37 Detroit 87Recife 38 Philadelphia 77Salvador 48 Washington 69Fortaleza 40Curitiba 39Brasília 39

Belén 37

5


10/44

With respect to differences in age groups, theanalysis of the complete census information dem-onstrates a greater presence of children living inthe periphery and the elderly living in the centralareas 4. This is predictable in cities growing atrelatively high rates, with a significant immigra-tion, especially when immigration has “special-ized” in the poor, within more extended families;similar to what has happened in the last decades inLatin America.

CHANGES TO THE TRADITIONALPATTERN OF SEGREGATION

The traditional pattern of segregation, which be-gan settling in the 20 th century, has recently dem-onstrated that it not absolute. Since 1980, the pat-tern has undergone important changes because of

the following new dynamics:• The introduction of alternatives to residential

development for the high or middle-incomegroups outside the “high-income districts,” thatis outside of the traditional areas where thesesocial groups are concentrated, in many inci-dences, in the center of low-income settle-ments (Bearings, 1990; Sabatini, 1997; Cal-deira, 2000).

• The emergence of shopping, office and ser-vices subcenters, outside the CDB and the“high-income districts,” usually in the crossing

of radial and ring roads where you can accesswide market opportunities (Gorelik, 1999;Frúgoli, 2000).

• The generalization of the increases in land prices throughout the urban areas, making itinevitable to locate new housing developmentsfor the low-income groups outside the cities inthe surrounding region.

4 We have seen segregation maps according to agegroups in Mexican (Germain and Polèse, 1996) andArgentinean (Towers, 1999) cities.

• The appearance of discontinuous residentialgrowth patterns favoring smaller urban centersand the use of rustic houses in patterns, oscil-lating between vacation and permanent resi-dences, fueling the “sprawl” as the dominantgrowth typology (Hack, 2000).

• The urban renovation of deteriorated centralareas, focused as much on the restoration ofold houses for residential or other uses, as inthe construction of high-rise residential build-ings for the middle class (Hardoy and Gutt-man, 1992; Red, 1999).

The two preceding changes are numerically lessimportant than the first three, and are confined tocities of countries or regions with higher levels ofeconomic development. Sao Paulo, a city exhibit-ing most clearly the five changes and conforming

a true “urban region,” mimics changes in the urban skyline maintained by the principal cities ofthe developed world – a phenomenon also entitled“diffused city ” (Dematteis, 1997).

We will discuss the causes of the deliberatedchanges when we discern the factors behind the

process of social spatial segregation (Chapter 3).The discussion concerning whether or not thesechanges have become more evident in an increas-ing number of cities and represent a disruption inthe traditional pattern of segregation is secondary.

It is more important to examine the possible urbanand social implications. Chapter 4 focuses on theimpacts of segregation and its recent changes.

6


11/44

Definitions and Precisions of Approach

The social segregation of urban space, also knownas residential segregation, is a spatial phenomenonwith complex connections to social differencesand inequalities, a complexity usually leading toconfusion. For this reason, it is important to pointout what we mean by segregation and the distin-guishable dimensions of the phenomenon.

GENERAL DEFINITION

In broad terms, residential segregation corre-sponds to the spatial agglomeration of families of

a similar social condition, regardless of how wedefine the social differences. Segregation is de-termined according to conditions of ethnicity, mi-gratory origin, age or socioeconomic status, alongwith other conditions. In Latin America, we focusour attention on socioeconomic segregation, as thefew completed empirical studies focus on thistype, ignoring other forms of social separation ofurban space. This is understandable consideringthat the distinct social inequalities, of income andrank or social class, represent the most salientcharacteristics of social structure in Latin Ameri-

can countries - more than poverty, in any case.

COMPOUND DEFINITION

Segregation requires a composed definition toaccount for its differences, which have distinctimplications, as much in terms of its social andurban impacts as with respect to the public policy.Here, we will differentiate three dimensions ofsegregation:• The degree of spatial concentration of the so-

cial groups;• The social homogeneity presented by different

inner city areas and• The prestige or social stature (or lack of stat-

ure) of different areas or neighborhoods ineach city.

The first two are objective dimensions of segrega-tion, registered in thematic maps of the city, aswell as through statistical indexes such as the in-

dex of dissimilarity. We mentioned these dimen-sions earlier as the most prevalent internationalmeasurement tool of segregation. In thematicmaps, they color areas to mark the location of dif-ferent groups; and the indexes measure the degree

by which the numerical significance of each socialclass within the communities create differencesamong neighborhoods and inner city areas.

The third dimension, relative to the prestige of theneighborhood, is of a subjective nature. This re-fers to the images, perceptions, reputation and

territorial stigmas designated by the city’s popula-tion to some of its neighborhoods. At one ex-treme, the social prestige of the neighborhoodsforms the basis for real estate value for promotersand in the realization of capital gains (land rents)for its residents; and at the other extreme of thesocial scale, the stigmatization of the neighbor-hoods contributes to various forms of social disin-tegration.

“NEW POVERTY”: A CRUCIAL TREND

The subjective dimension of segregation is centralin some key processes taking place in contempo-rary cities, including those of Latin America:• On one hand, in the growth of the urban real

estate sector, we explain an international phe-nomenon by the liberalization of the tradition-ally regulated urban markets and the signifi-cant flow of capital that it attracts. Segregationis a significant condition in many residentialreal estate businesses.

• On the other hand, the “new poverty” that isincreasing in the cities (according to the Euro-

pean denomination), is clearly associated withthe surge or reinforcement of territorial stig-mas: neighborhoods full of drug dealers, delin-quency, school desertion and adolescent preg-nancy, among other forms of social disintegra-tion. These neighborhoods are similar to theareas known as black “ghettos” in NorthAmerican cities, which are being replicated inother regions:

7


12/44

- In Europe, in the form of immigrantneighborhoods; and

- In Latin America, through the conversionof old precarious workers’ neighborhoodsinto concentrations of unemployed or un-deremployed populations politically mar-ginalized, where the social problems dis-cussed are increasing.

Although there are significant differences betweenvarious neighborhoods of “new poverty,” it is im-

portant to recognize its proliferation and associa-tion with territorial stigmas. These are alimented

by the population in general, and even by the au-thorities, and by the residents of the discriminatedneighborhoods. However, the stigmas have objec-tive foundations, among which changes in “objec-tive” segregation are prominent. The reduction in

the degree of social heterogeneity in the discrimi-nated and poorest residential areas of the cities (aneffect of the second dimension of segregation)appears to be a universal phenomenon of the era,in which social inequalities continue to grow.

In the United States, while the segregation of the black population has declined in recent decades(notable in the decrease of the dissimilarity in-dex), the social problems in the ghettos have be-come more severe. Whether through the impover-ishment of the middle class black families resid-

ing in these neighborhoods, through their emigra-tion towards the suburbs, or due to both, suchghettos are considered more homogenously poorthat before (Massey and Denton, 1993). The pre-carious settlements of Latin American cities havethe same problem. Greater rates of unemploy-ment and more flexible labor regimes have cre-ated more socially homogenous neighborhoods interms of poverty. The manner by which the in-crease of objective segregation (dimension 2) andother contextual factors explain the surge of terri-torial stigmas will be the focus of Chapter 4.

CONCEPTUAL APPROACH

We can summarize the approach adopted to ana-lyze social segregation, its causes, effects and im-

plications for urban policies through the four as-sertions discussed below. We will evaluate thefourth concept, concerning the nature of the seg-

regation process, more thoroughly, as it summa-rizes the first three assertions:• Residential segregation is a phenomenon, not a

problemThe effects of residential segregation can be

positive as well as negative. As many studieshave concluded and major cities show, theformation of ethnic enclaves is positive, for the

preservation of the minority class cultures andthe enrichment of the cities, which eventually

become more cosmopolitan. Qadeer (2001)discusses the case of Toronto, recognized asone of the more cosmopolitan and multicul-tural cities of our times.

We can attribute the most recent outbreak ofviolence in immigrant neighborhoods – thecase of several British cities during 2001, a cri-

sis that gave rise to a governmental commis-sion and special report on the problem (seeCantle, 2001) - more to the spatial accumula-tion of poor and other discriminated familiesthan to ethnic residential segregation. In fact,the European ethnic enclaves, unlike the ghet-tos of African-Americans, depict significantethnic diversity, including the representation ofa substantial percentage of the families of thehegemonic nationality in each city (French,Swedish or British families, according to each

particular case).

On the other hand, with respect to the positiveeffects of segregation, it is important to clarifythat spatial segregation of social classes is usu-ally a function of “normal” or “comprehensi-

ble” social practice, specifically in the searchfor social identities or the eagerness of thecommunity to attain a better quality of life.The fact that not all families in a particular citylive within favorable conditions, which allowsthem to validate their situation through a col-lective identity and encourages them to im-

prove their quality of life, does not necessarilyendorse the situation as socially condemnable.

• Residential segregation is a constituent ele-ment of social realitySociety does not exist outside the space, as isoften assumed in the traditional disciplinaryculture of social sciences. The space itselftakes on social meaning and plays various rollsin social practices. Specifically, segregation is

8


13/44

a part of particular social practices of thegreatest importance:- Segregation is important in the formation

of social identities. In dynamic societieswith powerful processes of social mobility,such as those of capitalist societies, spatialsegregation is a common resource in con-solidating the identities of the ascendinggroups, or used in defending the identitiesthreatened by these changes. Accordingly,we consider contemporary western citiesmore spatially segregated than pre-industrialist cities.

- Spatial segregation is part of the mecha-nism offering a socially differentiated ac-cess to public goods or collective con-sumption, such as landscape, environment,neighborhood safety and, in general, qual-

ity of life.- Segregation is instrumental in building

more communitarian styles of social life.Confidence is an important social capitalin the construction of a social life with astrong communitarian and less individual-istic content. Confidence, as demonstratedthrough social research, requires that the

parts involved in the relation exhibit simi-lar lifestyles, a condition known as socialhomogeneity. The Lomnitz (1977) study,in the “colonies” of Mexico City, is per-

haps the most widely recognized in LatinAmerica, which examines these issues; al-ternatively, Durston (2001) emphasizes therole of confidence in the development of“social capital” in Latin America.

• The geographic scale under which segregationtakes place is significant in its effects.The negative impacts of segregation are asso-ciated with a less significant interaction be-tween social classes. Spatial isolation of the

poor or discriminated groups, and their percep-tion of this condition, perpetuates social disin-

tegration. Therefore, if segregation occurswithin a reduced geographic scale, like a smallcity or through the conformation of smaller,socially homogenous neighborhoods, the nega-tive effects of segregation can be less signifi-cant or non-existent. However, when segrega-tion intensifies in broader scales, exceedingmargins of “the walking scale” and limiting theoptions of physical interaction between social

classes, spatial segregation can become coun-terproductive, especially for the poor.

• Residential segregation is a process, not asituationThe term itself, borrowed from Biology, has aclearly dynamic and incessantly shifting con-notation. Nevertheless, in the tradition of urbanstudies, particularly in Latin America, this im-

plication has lost validity. It is possible that thestatic bias established by the general conceptof residential segregation is due to the pre-dominance of architects and geographers in thefield of urban planning.

A process approach to segregation must start byexamining it as a socially collective fact that weshould come to understand. Segregation is notsolely a problem; it is also a phenomenon, with its

own right to exist and possibly, a more or less predictable phase of evolution. What are the explanations and motives behind segregation? Dothird parties practice this type of auto-segregationor spatial exclusion? Wouldn’t this behavior re-flect comprehensible or reasonable aspects?When unfolding the process dimension, we openan intellectual avenue for these considerations anda minor objectivity in the analysis.

There are several interpretations along this line.Perhaps, the most recognized at an international

level is the “assimilation” approach. This de-scribes the stages of the assimilation process of acity’s ethnic minority group transitioning from

preliminary spatial segregation to a permanentdispersion within the city. Fundamentally, theconcept is based on the dynamic situation of the

North American cities during the 20th century.Students of the School of Chicago, headed byRobert Park, elaborated this concept early on. Inany case, in the discussion of the possible policiesto control segregation in Chapter 5, we debate theneed to analyze this approach further.

In Latin America, urban planners and demogra- phers have pondered and debated the existence ofevolutionary systems of spatial segregation of

poor immigrants within the cities. Perhaps the best-known interpretation is that which occurswhen families settle in central areas of the city ina first stage of integration. After integratingthemselves in employment, social and eventually

9


14/44

political networks, these families move to the pe-riphery, whether through illegal land occupancy,lot purchases through business deals of question-able legality or through government sponsoredhousing solutions. Families exchange room space,legal and material security (the most common pe-riphery homes in Latin American cities are builtmainly on private property) for accessibility and

proximity to occasional jobs. 5

According to the composed definition of segrega-tion, there is a transition between a situation offirst stage residential segregation (spatial concen-tration of the group and urban space shared withother groups) to a second stage (social homogene-ity of the space). In time, poor families tend toisolate themselves from other social groups. It is

possible to counteract the effects through other

methods of social integration, in particular,through employment or economic integration and

political participation. Labor flexibilization andthe political marginalization of the poor, compo-nents of the present context, remove this layer andadd “negativity” to spatial segregation, as we willargue later.

However, in “normal” conditions, the systemreaches a positive stage: corresponding physically,to the urban assimilation of the settlement to thecity; and socially, to an acceptable level of social

heterogeneity. The effects of time and fortuitousexperiences of each stage create greater socialdiversity amongst the families. Motivation of thesettlers is of essence in the self-building of theirhomes or in neighborhood improvement efforts.Within each city, there are a series of goals con-tributing to the collective imagination, such asthose traditionally adopted by the Chilean settlers.Their hope is to move beyond the stage of “camp-ing” (illegal invasion) to becoming a “settlement”recognized by the authorities (who urbanize andlegalize the settlement), until finally forming a

“villa.” They associate the “villa” with “formal”residential construction of various degrees of con-

solidation: consistent architectural style, homescomplying with normal urban standards of con-struction or the quality of the building materials.

5 In 1959, the relators of an important seminar organ-ized by CEPAL, regarding Urbanization Problems inLatin America, argued the necessity of carrying out“re-localization programs” designed to provide housingto the migrants “which were crowded together inslums” (Hauser, 1962: 72).

This explanatory model does not have a general-ized spectrum since, among other explanations,

poverty has become a “product” generated more by the cities themselves than by the migrations,and also because in much of Latin America themigrants opt to live with relatives already living inthe periphery (Hauser, 1962: 317; Peattie, 1987).Yet, it exposes certain clues to help understandthe evolution of residential segregation of the poorand middle-class up through the present time. In

particular, under aggravated conditions of eco-nomic and employment uncertainty, similar toconditions today, there is a tendency to blend so-cial classes within the space in order to increase

their sense of security. According to the tradi-tional explanation, this security is what the innercity slums offer to immigrants. The crisis thusincites a backward progression of spatial segrega-tion of poor families in Latin American cities to-wards the poor homogenous periphery. Thisaborts the possibility of realizing the final stage of

progression, a stage in which the old poor periph-eral neighborhood is eventually assimilated withthe rest of the city, in which an adequate level ofsocial diversity is reached and finally in which theindividual no longer identifies himself as an “im-

migrant,” rather as a respected resident of the city.

This trend of social integration under crises takeson two forms: the impoverishment or “populariza-tion” of central areas and the arrival of middleclass families affected by the crisis in the poorneighborhoods of the periphery. This occurred inthe 80s in Sao Paulo, Brazil. The economic crisis

produced significant alterations in the segregation pattern of the city, creating a backward progres-sion in the segregation progress. One study (byRolnik et. al. 1990), which we will examine more

thoroughly in a subsequent chapter, describesthese changes, concluding that the crisis reflecteda “perverse setback in segregation” due to the im-

poverishment of middle class groups and furtherintensification of social inequalities.

During the most recent decades, changes in the process of segregation are apparent in LatinAmerican cities. Meffert describes the movement

10


15/44

of poor families from the periphery to central ar-eas as a form of pacific occupation of the city bythe poor in Mexico City (1990 – cited by Mires,1993: 101). Earlier, Matos Mar discussed similartrends in Lima (1988); and Mires suggested thatthe modern Latin American city “has been in-vaded from its own interior,” emphasizing self-regulation theories as possible explanations forthese changes (1993).

Currently, the mounting economic and social cri-sis within various Latin American countries may

be stimulating social spatial mixing in order torevert feelings of insecurity.

Accordingly, we will examine an Argentineancase in Chapter 4, Section C. In this case, the ex-treme social insecurity instigated by the crisis in-fluenced a broader social mixing within the space,supporting the notion that social isolation is themost detrimental aspect of residential segregation.

Finally, among other “process related” aspects ofresidential segregation fitting into our approach,we will discuss the connection between segrega-tion and social differentiation, and the develop-ment or defense of social identities. As we willargue, residential segregation is associated morewith social differentiation than social differences.

11


16/44


17/44

pointed out earlier. The idea or theory of the mir-ror, as an interpretation of the origin of segrega-tion, is inconsistent with this characteristic. Theanalysts frequently ignore this inconsistency, andwhen they do discuss it, they do not offer a suffi-cient explanation.

Alternatively, the second explanation, implicatingthe operation of the real estate markets, is contra-dictory to the specific urban development proc-esses taking place in certain cities, some older andothers more recent:• The densification of higher-income neighbor-

hoods, old and new, through high-rise housingconstruction for families of lower than averageincome for the area.

• These projects allow promoters to significantly profit and their indirect impact is to reduce

residential segregation. Considering the sig-nificant concentration of good quality facilitiesand services in these neighborhoods, there isalways a strong demand to live in high-rise

buildings. This introduces two relevant pointsto our discussion: the “verticalization” of theseneighborhoods and its effect on the reductionof segregation are important processes; also,there is no significant social (and cultural) re-sistance to social spatial mixing, a similar factof the suburbs of North American cities.

• The dispersion of closed condominiums for

middle and high-income families around theurban periphery, many in areas already popu-lated by the poor.

• They established this practice in the eightieswith unequal intensity in a variety of LatinAmerican cities, from Santiago, Chile to SanJose, Costa Rica; from Puebla, Mexico to Riode Janeiro, Brazil. Consequently, Villaça dis-cusses the relative spatial proximity that is tak-ing place between different social groups in themajor Brazilian cities (1998). Rolnik, et.al. af-firm that in San Paulo during the eighties, an

“increase in relative proximity among the dif-ferent social groups in the space” took place(1990). Portes found that, in the case of Bo-gotá, certain high-income residential develop-ments were being established in poor areas(1990). Towers (2001) asserts the same in thecase of Buenos Aires.

• Earlier, Alexander Portes predicted this phe-nomenon in some Latin American capitals

(1990), connecting it to the regressive trends inincome distribution, that in turn caused an in-crease in poverty in the cities. Later, thechange would be interpreted as a widespreademerging trend in cities of Latin America,caused mainly by changes in the private realestate sector (Sabatini, 1997) – which we willanalyze later in the document.

• At the same time, this dispersion of condo-miniums represents an increase in residentialsegregation, as well as a reduction of its geo-graphic scale. The guards and fences emerge atthe same time that the physical distance be-tween the rich and the poor diminish.

Finally, the third explanation, implying that LatinAmerican countries imitate the cultural patterns ofdeveloped countries, does not distinguish between

urban spatial forms and architectural styles, nordoes the explanation distinguish between thosereligious cultures continuing, to this day, to makea difference in the models of urban structure incapitalistic societies. Catholic societies have mostclearly preserved the continuity of the pre-industrialist city, along with some protestant so-cieties, such as the Lutherans. These societies donot undertake the urban Anglo-American revolu-tion, which prompts segregation of the upper classin the suburbs.

Finally, the social diversity of the “Latin Ameri-can high-income neighborhoods” contradicts thesuburban model, which has social homogeneity asone of its essential characteristics. The adoption ofEuropean architectural styles is confused with theadoption of the patterns of urban segregation ofnations that have been the leaders of capitalistdevelopment in the last century: first England, andlater the United States.

REINTERPRETATION OF THE CAUSESOF SEGREGATION

Now, we offer a secondary interpretation of thesethree popular explanations of the origin of urbansegregation in Latin America, striving to makethem consistent with the facts:

With respect to the explanation attributing segre-gation to social inequalities, it appears to us thatresidential segregation reflects processes of social

13


18/44

differentiation more than social differences. Theconnection between social differences and spatialsegregation is inverse rather than direct, as as-sumed habitually. Certain groups newly develop-ing their identity, as well as existing groups fear-ing that their identity is being threatened, oftenresort to spatial segregation in order to stabilizetheir identity. It is common, when feeling threat-ened, for emerging middle class groups in dy-namic economies, or ethnic minorities, to createneighborhoods of ethnic enclaves.

In this manner, when differences and inequalitiesare threatened, or when they are minor or indis-tinct, groups increasingly resort to spatial segrega-tion. Conversely, when social differences areclear and profound, due to prevailing distinctionsof status (as apparent in the European pre-

industrialist societies and in the majority of exist-ing Latin American societies), the groups can ef-fectively share the urban space. The European

pre-industrialist city showed a significant socialmix in the space, and the Latin American citiesexhibit considerable diversity in their “high-income neighborhoods.”

This interpretation stems from a dynamic approach to segregation, incorporating “the spatialcomponent” within social structures and not out-side them in the manner of a “reflection.” It also

highlights the personal motivation in the modifi-cation of segregation patterns (in Fishman, 1987and Sennett, 1970 similar interpretations are madeconcerning segregation in cities of England andthe United States, respectively).

In regards to the explanation implicating the ac-tivities of real estate developers, it is important toemphasize that land rents, which these agents in-tend to capitalize, do not depend on the socioeco-nomic level of the residents, rather on their abilityto pay per square meter of constructed or qualified

land. Sometimes the social exclusion (forced spa-tial segregation of undesirable social groups oractivities) is a method of increasing the profits,however at other times; high-rise constructionencourages this increase by expanding the volumeof floor space available for sale. Often, these

business deals thrive when they succeed in attract-ing foreign groups of a lower income class incomparison with the residents of the respective

areas. In summary, developers relate pragmati-cally with segregation, profiting whether broaden-ing or decreasing its effects.

Nevertheless, real estate markets and segregationhave other more significant relations, of which thefollowing two seem the most pertinent to this

presentation: the inversion of the relationship be-tween land use and land price due to the uniquecharacteristics of markets and the dependency ofdevelopers on the structure of land prices in thecity.

In the first relation, we must focus on the intrinsi-cally speculative nature of urban land markets andthe relevant impact they have on the segregation

pattern. Given the peculiarities of urban land asinfinitely differentiated, only partially inter-

changeable and non-reproducible, among othercharacteristics, the causal relationship betweenland use and land price is usually reverted. Land

prices, determined by uses, eventually turn into afactor excluding uses. Owners set their price

based on the expected use. When a neighborhood begins to attract higher income residents, ownersspeculatively raise their prices anticipating “en-hanced” uses, creating an obstacle for lower in-come groups wanting to settle there – somewhatof a self-fulfilling prophecy. We have given thismechanism, particular to the imperfect land mar-

kets, the definition “spatial propagation of landspeculation” (Sabatini and Arenas, 2001; Sabatiniand Smolka, 2001). This evokes a transition from“type 1” segregation (spatial concentration of agroup), which is not entirely bad, to a “type 2”segregation (social homogeneity of the neighbor-hood), which can produce negative effects, as de-

picted in Chapter 4.

The second correlation between land markets and patterns of segregation refers to the subordinationof location decisions of the principal Latin Ameri-

can developers in the city land pricing structure.This subordination was important in the consoli-dation of the traditional pattern of segregationthroughout the 20 th century. The illegal land set-tlers chose the cheapest land because of the de-creased risk of legal repression and eviction; andthe State housing programs began implementingtheir projects on less expensive land, thus lower-ing their overall costs. Illegal land settlement and

14


19/44

State housing programs have forced poor peopleto live in places already stricken with poverty –the low income of the residents explains the lowerland prices (Smolka, 2002). Private real estatecompanies, partially due to their traditional atomi-zation, executed smaller projects designated forareas where the social condition corresponded tothe demand fulfilled by the projects 7. Surpris-ingly, these three distinct agents abided by thesame spatial logic, having to do with the landmarkets, gathering support in favor of the tradi-tional pattern of segregation.

The explanation concerning the imitation of cul-tural patterns is vague, as we implied earlier.

Nonetheless, we believe the explanation pointsout an undeniable fact. The Latin American upperclass has been culturally “dependent.” They have

arduously tried to recreate the reality of developednations within their own. In their efforts, theyhave counted on State assistance and often thecomplacency of other social groups.

This enduring and sustained eagerness invokes aweak social identity, constantly evolving. As we

pointed our earlier, when social identities are frag-ile, groups resort to spatial segregation. It is rea-sonable to interpret the construction of “high-income neighborhoods” in Latin American citiesas a resource to the spatial form, with the goal of

developing the less chauvinistic identity of a de-veloped country.

As indicated earlier in this presentation, all groupsof the industrial European social structure fit inthese “high-income neighborhoods,” which areunderstood as genuine pieces of city of developedcountries. This is consistent with the catholicethos of Latin America. Consequently, the “high-income neighborhoods” are exceptionally diverse,excluding only those who do not conform to thisdefinition or objective: the “informal” or “margin-

alized,” groups in a class below the formal labor-ers. Their presence deteriorates the identity of the“city of a developed country” that they have beentrying to build. 8

7 We discussed the changes to the traditional pattern inChapter 1.8 These aspirations of European identity have had anexpression in urban planning philosophy and practice

amongst Latin American leaders for quite some time.In Santiago in 1872, one Chilean leader, Major Ben-

jamín Vicuña Mackenna, proposed the creation of acentral road to exclude the poor “informal” groupsfrom their “own city” and “culture.” In addition, theViennese urban planner Karl Brunner proposed a planto transform Santiago, having an explicit image of the“European” social structure (Vicuña Mackenna, 1872;Brunner, 1932).

The scarce development of the economies forcedthe concentration of all private and public effortsin the construction of this affluent and “modern-istic” exception to a single area of the city, in themiddle of the under-developed space. Conse-quently, it is important to point out the notableconcentration of the upper class in city spaces(segregation in the first dimension).

Accordingly, the coincidence between strong so-cial inequalities and noticeable spatial segregationof the cities is quite apparent as well as deceptive.

REINTERPRETATION OF THE PRESENTTRENDS EXHIBITED BY SEGREGATION

With respect to the present trends of segregationin our cities, the most common interpretation is

the theory that segregation is growing because ofincreasing social inequalities, while refusing orignoring the facts that indicate something differ-ent 9. Furthermore, the assertion that segregation isescalating is not qualified regarding the scale orthe dimensions in which it occurs.We will analyze these issues in the Annex, em-

phasizing that the empirical studies carried out inLatin America are not usually designed to testtheories, rather to “demonstrate” that segregationis increasing, a fact considered obvious.

Hence, there are flaws in many of the studies car-ried out to interpret the changes of the traditional

pattern of segregation. There is an apparent ideo-logical resistance to the idea that spatial segrega-tion could be reducing; at least in some sectors ofthe city periphery. Regardless, this resistance iscoherent with the perception of segregation as amirror image of social inequalities, given that re-cently, they have been increasing.

9 See “Changes in the traditional pattern of segrega-tion” in Chapter 1.

15


20/44

CONCLUSION

In conclusion, the main causes of social spatialsegregation in Latin American cities are the fol-lowing, differentiating between causes operatingat the level of the motivations of the different ac-tors and aggregated or “systematic” causes. In-cluded in this summary are the major factors thathave influenced the changes effecting segregationin most recent years.

They stress the following motivations:• The first is the enthusiasm to construct a mul-

tiple class identity of city (and society) similarto that of developed countries. In recent dec-ades, the generation of profits no longer re-quires the spatial concentration of private and

public resources, as it was required in earlier

eras, explaining the relative dispersion of the“modern neighborhoods” within the space ofeach city. This change represents new oppor-tunities to decrease distances and to increaseinteractions between different social groups.

• The increase in value of real estate property,whether as business (for developers and otherinvestors who profit from land appreciation)or as family assets, is a contributing factor

promoting the social homogeneity of thespace over the social motivations (of socialidentity) of the high and middle-income

groups that tend to self-segregate. Manywealthy families do not like the idea of poorer

people living close, because, according to popular belief, it could affect the increase inthe value of their property. Accordingly, thisconviction operates as a specific form of theself-fulfilled prophecy we mentioned earlier.There could be, at the same time, more cul-tural space for social mixing, or for the reduc-tion of physical distances between socialgroups, than what the traditional pattern ofsegregation demonstrates. The decrease of

high and upper middle class group concentra-tion in “high-income neighborhoods,” due tothe construction of gated communities in otherareas of the city, reinforces this concept. 10

The middle class Latin American suburb ismore of a physical reality, imitating traditionsof Anglo-American architectural and urban

planning elements, than a replication of theorganized neighborhoods of the North Ameri-can suburbs, which exclude certain socialgroups while developing social class identi-ties. Although elements of real estate valori-zation and development of group identities areevident in both societies, they have a differentimpact, influencing policy design. Both thesecular instability of Latin American econo-mies and the most segregated character of so-cial structure have more influence in the seg-regation of real estate valorization than in theformation of social class identities.

10 Coincidently, the Report of Human Development inChile 2002 of the PNUD discovered, through a nationalsurvey, that 63.3% of Chileans claimed to feel no dis-advantage in living near impoverished families.

• Additional motivations of segregation includethose relative to quality of life. Groups that

have the option to select certain areas of thecity with sufficient access to public and col-lective goods (goods that are hard to obtainindividually) tend to concentrate in one areaof the city. Accordingly, segregation actuallyimproves the possibilities for families to enjoythe landscape, nature, environment and resi-dential security.

• Secondary in importance is the incidence ofaffirmation and defense of social group identi-ties in the self-segregation of upper incomeand other emerging groups. Culturally, be-

cause our societies harbor significant socialdifferences, lower levels of social mobilityand hierarchal relations between groups andclasses, they are more open to a greater socialmix in the city space. The reduction in thescale and intensity in segregating the poorestgroups, indisputably the most significant chal-lenge our cities face, is more feasible in thisrespect than it initially appears.

• The following “systemic” factors impactingsegregation and its evolution should be noted:

• The accumulation of the poor and the forma-tion of extensive homogenous areas where the

poor reside are the result of the aggregate impact of the forces and motivations just de-scribed. The upper and middle-incomegroups build cities according to their interests,marginalizing the underprivileged from theenjoyment of material and symbolic benefits

16


21/44

afforded by the neighborhoods they con-structed.

• As we have discussed earlier, the managementof the developers’ land pricing structure inmany cities was crucial in stabilizing whatwould become the traditional pattern of urbansegregation throughout or continent. In par-ticular, we must emphasize the role of illegalland settlers and state housing programs in theexisting concentration of poverty in the cities’

periphery.• Since the eighties, the liberalization of the

land markets, the concentration of real estatecapital, the adoption of the gated communitymodel and the construction of significant re-gional urban infrastructures, particularly inroad and transportation, are amongst the sig-nificant factors that have affected the modifi-

cation of the traditional pattern of segregation.Due to the considerable size of the real estate

projects and a favorable institutional and eco-nomic context, real estate developers have be-gun to spread out their commercial and resi-dential land investments, no longer restrictingthem to “high-income neighborhoods” andcentral areas, as they did before. In particular,the location of middle and high-income resi-dential developments in lower-income dis-tricts allows them to capture significant landrents. The relatively large size of the projects

allows the investors to restructure residentialsegregation on a reduced spatial scale.

• Essentially, these projects promote a changein the segregation scale, with more intensesegregation on a smaller geographic scale.The social homogeneity of these new commu-nities is very high, although the physical dis-tance to the residential areas of poorer fami-lies diminishes.

• The spatial dissemination of land speculationfrom downtown and “high-income neighbor-hoods” to other areas of the city is one of thesystematic factors affecting modification ofsegregation in most recent decades. The dis-

persion of residential developments, busi-nesses and “modern” services induces the tra-ditional speculative vortex of the land marketsto spread to the whole city. This fact, alongwith the increasing formalization of the landmarkets promoted by land tenure regulariza-

tion, policies and programs, is a contributingfactor in the expulsion of new poor familiesoutside the cities. In addition, new highwaysand an increase in automobile ownership helpto re-create, on a larger scale, the traditionalsegregated urban models where the poor peo-

ple live in the periphery, yet with the impor-tant caveat that the previously settled poorfamilies benefit from the reduction in thescale of the segregation.

17


22/44

Impacts

We pointed out earlier that segregation has posi-tive as well as negative effects. Before we reviewthe known impacts of segregation in Latin Ameri-can cities, which are few given the scarcity ofstudies, we will discuss the effects of segregationfrom a broader international perspective.

POSITIVE AND NEGATIVESEGREGATION; VOLUNTARY AND

FORCED

The two objective dimensions of segregation sig-nificantly differ with respect to their effects. Thespatial concentration of a social group (first di-mension), whether ethnic or socioeconomic, usu-ally has a positive impact, as much for the groupas for the city and the community. The preserva-tion of the customs and identities of the ethnicgroups, known as “minorities” in cities of devel-oped countries, and the social and political “em-

powerment” of the Latin American urban poor areexamples of the beneficial effects of spatial con-centration.

However, when the spatial social homogeneity isstrong (second dimension), the effects tend to benegative, particularly in the social disintegrationof the poor, an effect that we will demonstratelater in this document. It is important to note thatwe relate the most negative effects of segregationto their involuntary character. The poor and dis-criminated groups are excluded from certainneighborhoods and areas of the city. Then, theyare forced to concentrate in the worst areas of thecity by land markets, social housing policies andforced eradications.

The distinction between these two objective di-

mensions of segregation relates to the differences between voluntary and forced segregation. Thespatial concentration of a group (first dimension)does not exclude new families of different socialconditions from moving to the area. There arefamilies of inferior social condition, or from ra-cially or ethnically discriminated groups, that pre-fer to live in neighborhoods where groups ofhigher social status predominate. Conversely,

other families prefer living with those of equalstatus in segregated neighborhoods, where theyfeel comfortable and can rely on the security ofthe social networks of mutual aid, which are usu-ally quite strong in these communities.

It is important to point out that spatial concentra-tion of social groups is a form of segregation that,on the extreme, could be a result of the citizensexerting their free will. We could classify thisvoluntary form of segregation as “natural,” as it isassociated with the affirmation of social identities,the respect for well-defined values, or the searchfor a better quality of life. Besides, it originates inthe personal choice of dwelling location of theindividuals and their families, which is a value initself.

It is also true that discriminated or impoverishedgroups appear spatially concentrated, and that thiscircumstance is far from being the result of theirown preferences. Nevertheless, the localization ofthese groups in our cities shows a relative disper-sion to areas of the urban periphery, includingdeteriorated central areas, rather than a concentra-tion in one internal zone, as often occurs with

higher-income groups. Conversely, these residen-tial areas show a noticeable absence of families ofdifferent social conditions. Social spatial homoge-neity is the most outstanding characteristic of thesegregation of lower income groups in LatinAmerican cities, and this homogeneity is clearlynot the result of the collective free location deci-sions of their members.

In effect, social spatial homogeneity, the secondobjective dimension of segregation, can hardlyoccur without mediating the use of the power. It is

extremely improbable that highly homogenousdistricts will conform as an effect of the aggrega-tion of individual decisions. The State, exerting amonopoly of legitimate use of force within thesociety, is involved, either directly or indirectly, inevery form of compulsory segregation (Marcuse,2001).

18


23/44

Nevertheless, the apparently spontaneous segrega-tion resulting from the free operation of land mar-kets has an involuntary component. Not every-thing is voluntary and free in the residential seg-regation of higher-income groups. High densityof land uses make it possible for poorer familiesto settle in practically any area of the city, even inareas of high land costs. In fact, it is usually asmart business decision for real estate investors to

build high-density houses for lower income fami-lies in prestigious neighborhoods.

In cities of developed, as well as developing,countries, it is common for lower income families,and even racially discriminated groups, to expressa desire to share neighborhoods with other socialgroups, even the groups that discriminate them(see Squires et.al, 2001 for the case of the Afri-can-American population in the United States).

If the desire exists for the poor to move to areaswhere higher income families concentrate, and ifthis is a good real estate investment decision, howdo we explain the emergence of socially homoge-nous areas? In reality, the State always plays a

part in the emergence of this form of segregation.In affluent neighborhoods, the legal system andspecifically the construction and urban norms,

play a key role in excluding the unwanted familiesfrom the area. There is a recurrent requirement

for minimum lot sizes with little technical justifi-cation. This takes place in neighborhoods and ar-eas well endowed with infrastructural material andservices that could support much higher densities.The social exclusion objectives, and not the tech-nical needs, are behind the motivations definingminimum size for lots. Poorer families cannot buysuch large lots, even if the unitary land price islow.

They apply land use and building norms to protectneighborhoods from the intrusion of other settlers,

as well as other legal arrangements, with the North American suburbs representing the mostdeveloped prototype. The communities of theUnites States are organizations that specify physi-cal and architectural criteria for the neighborhood,to protect the neighborhood from unwantedgroups. Conversely, real estate agents and the

banks issuing mortgages usually resort to veiledforms of social discrimination with similar pur-

poses when they coincide with their interests.Thus, the segregation of the suburbs would be,largely, a planned process and not only spontane-ous, but a forced reality for individuals excludedfrom these places, exercised by those who merelyseem to practice their freedom of choice.

In the Latin American city, as we have already pointed out, this form of segregation is strongwithin the poorer communities and less significantin affluent areas. The forms of coercion, similarto those that explain the relative absence of poorfamilies in the wealthy neighborhoods, cause theagglomeration of poor households or an emer-gence of socially homogenous settlements. With-out the possibilities of attaining land in goodneighborhoods, due to the high land prices as wellas the zoning standards preserving the “exclusiv-

ity” of those neighborhoods, poor families areforced to buy or illegally settle on land in segre-gated locations. The State also develops socialhousing projects in a spatially segregate manner,mainly for the same reasons.

The terms enclave and ghetto differentiate bothforms of segregation (Peach, 2001; Boal, 2001).The African-American ghettos represent types ofcoercive segregation causing negative effects, notonly for those living within the ghettos, but alsofor the entire city. According to Massey and

Denton, a ghetto is defined as a set of neighbor-hoods exclusively inhabited by members of onesocial group, in which virtually all the members ofthat group live (1993: 18-19).

We should also discuss the wealthy ghettos, emphasizing their overall negative effects, includingthose affecting the residents themselves. Boredomis one of these negative consequences, and isseemingly to blame for the gentrification processin many cities. Often, groups of young people,children of the suburban residents, seek out the

energetic ambience of the older, deteriorated cen-tral neighborhoods. For decades, researchers haveimplicated a lack of social diversity as the expla-nation behind this prominent characteristic of thesuburbs in many North American cities (Sennett,1970). Alternatively, ethnic or socioeconomic en-claves stand up as a contrast. These are voluntaryforms of segregation, whose outcome is more

19


24/44

positive than negative. Social homogeneity is nottheir trademark.

As these involuntary forms of segregation are re-lated to the action of the State, and because theyconcentrate the negative effects, the policies tocontrol segregation, in principle, have ample roomfor action. By modifying what the State acts ordoes not act upon, without reducing or restrictingthe localization decisions freely adopted by thecitizens, it would be possible to make progress incontrolling the detrimental effects of residentialsegregation.

INFLUENCE OF THE SEGREGATIONSCALE 11

When considering smaller neighborhoods, par-

ticularly in medium size cities, segregation couldhave less negative effects, or simply not have anyat all, even in the areas marked by social homoge-neity (second dimension). However, when low-income and homogenous neighborhoods are sur-rounded by other low-income neighborhoods, seg-regation tends to reach a more vast geographicscale, generating negative consequences that didnot exist without this agglomeration of poverty.

Once again, we return to the example of the Afri-can-American ghettos in the United States, whose

negative effects have been broadly documented.As the Massey and Denton definition indicates,the ghetto consists more of an agglomeration ofneighborhoods than single neighborhoods.Massey and Denton also indicate these groups asthe most isolated and geographically encapsulated

people of the United States, while also emphasiz-ing that “they live in large and contiguous settle-ments of densely occupied neighborhoods that are‘packed’ around the urban centers” (1993: 77).

The scale is significant because it worsens the

effects of physical, employment and social isola-tion of these groups, considered the most relevantaspect of spatial segregation in what matters forsocial integration. Alternatively, a critical concern

in the materialization of the effects of social disin-tegration is the phenomenon known as subjectivesegregation. The physical isolation of the othersocial groups causes an intensification of this sen-sation, as well as the conviction of “being a sur-

plus.”

11 The empirical data provided in this section and thefollowing two sections of this chapter, in support of theargument, comes from research carried out in recentyears in the main Chilean cities.

This isolation not only heightens the feelings ofnonconformity of the residents themselves, yetalso influences the image that residents of otherneighborhoods have of these homogenous ag-glomerations of poor and discriminated groups.The broad scale of segregation and the intensity ofits dimension 2 represent the principal spatial fac-tor contributing to the emergence of territorialstigmas. The disdain and repulse of other groupsis an important precursory factor of social disinte-gration of these objectively and subjectively seg-

regated groups.

Studies carried in Chilean cities validate the nega-tive social effects of the geographic scale of thesegregation of poor groups (Sabatini et.al. 2001aand 2001b). The statistical correlations betweenthe degree of social spatial homogeneity and so-cial problems are greater when the analysis fo-cuses on contiguous census districts of low-income groups than when analyzed separately.The qualitative study of pairs of poor segregatedurban neighborhoods, with different geographic

scales within each city, confirms those results.According to those studies, the social problemsthat increase due to the agglomeration of poorcensus districts include low scholastic perform-ance, unemployment, adolescent pregnancy andthe indolence of the youth (homes of young peo-

ple who neither study nor work, a group that typi-cally falls into drug addiction problems and delin-quency).

The geographic scale of segregation shows con-tradictory trends during most recent decades in

Latin American cities. In one direction, the scaleseems to be contracting, while in another direc-tion, broadening.

We connect the first positive trend to the evolu-tion of real estate and land markets. Previously,we pointed out the relative dispersion, apparent inmany Latin American cities, of enclosed residen-tial condominiums, commercial spaces and “mod-

20


25/44

ern” office centers outside the traditional areaswhere high-income groups and activities concen-trate. We have also attributed this change withinthe cities to the liberalization of the land markets,the expansion and concentration of the real estatecapital and the improvement of urban infrastruc-ture, particularly the roads.

From the point of view of public policy, weshould not ignore the market trends that influencea decrease in the physical distances between so-cial groups, as restricted as this phenomenon is tocertain parts of the urban periphery and as limitedas the effect is on certain poor groups. In any case,it is another empirical revelation that the poor arenot segregated inevitably by the markets, and itoffers a historic opportunity to reverse segregationto some degree. It is important to reinforce these

market processes within the policies to controlsegregation.

The second trend corresponds to the increase ofthe segregation scale, particularly for newly im-

poverished households. Earlier, when we dis-cussed the relationship between land markets andsegregation, we emphasized the phenomenon thatwe have labeled “propagation of land specula-tion.” In the context of the liberalization and re-duction of regularization policies of land marketsimplemented in recent decades in Latin America,

this phenomenon affects the whole of the urbanspace, stimulating spatially generalized increasesin land prices. In this manner, the increase in thesegregation of the poor would reach a regionalscale (Sabatini and Smolka, 2001). At a maxi-mum, poor families who access a house should doso in areas far from the urban fringe, whether inopen spaces or less populated central areas.

INCREASE IN THE NEGATIVE EFFECTSOF SEGREGATION

The segregation of the poor groups in the LatinAmerican cities has both urban and social im-

pacts. Among the first type of impacts, it is worthmentioning the accessibility difficulties and thedeficiencies of services and urban facilities in cer-tain residential areas. Among the second, the

problems of social disintegration explained earlierare worth discussing further. They represent forms

of impoverishment or social degradation linked tothe disadvantages of physical isolation.

The urban impacts are well known. Spatial segre-gation causes the poor people of our cities to be-come even poorer. However, the social disintegra-tion effects, and their vast social impacts, are rela-tively recent. We could say that segregation isshowing a “negativeness” which was not previ-ously noticeable with such intensity and scope. InChile, the empirical studies illustrate that, beforethe 80s, segregation of the poor did not have thesame socially disintegrating effects that are appar-ent today. In addition, some social variables, suchas scholastic performance, employment and indo-lence of the youth, showed less negative valueswithin the most segregated, poor census tracks(socially homogenous) in comparison with poor

census tracks of greater social diversity (Sabatiniet.al., 2001b). According to Touraine’s famouslocution, this segregation, in a political context of“centrality of the marginalized groups,” favoredthe organization and the social and political “em-

powerment” of the poor people. Clearly, the cir-cumstances in the case of Santiago, in which wedetect this positive effect, may constitute an ex-ception, while demonstrating at least one possibil-ity that has apparently disappeared.

Thus, we conclude that while in the past, segrega-

tion of the poor had both negative (urban) and positive (education, employment, and family as-sociated with the political and social strengtheningof its territorial base) effects, the most complexeffects of social decomposition are now worsen-ing. The explanations are spatially “objective,”associated with the relentless increase in the in-tensity and scale of segregation of most of the

poor, and spatially subjective, associated with thesurge of territorial stigmas and the reinforcementof existing stigmas. In addition, there are non-spatial explanations, related to the “flexibility” of

the employment markets and the advancement ofa political system “based on marketing research”that limits the political participation of the poor invoting.

During acute social crises, under uncertain andinsecure conditions of “economic globalization”and increased flexibility in labor relations, thenegative consequences and disadvantages of spa-

21


26/44


27/44

Policy to Control Segregation and its Relation to the TerritorialStrategy of Social Program Coordination

In this panorama of precarious information andempirical studies, in which even the theoreticaldialogue is biased, it is difficult and rather daringto make policy proposals. For most, the panoramaof urban segregation in Latin America has impor-tant specificities beyond the shared characteristicsamong cities of the region, and the policies shouldaddress these differences. In addition, as we willemphasize, the policies to control segregationshould be “calibrated” to the cultural characteris-tics and circumstances of each situation in a me-thodical trial and error type of process. Countrieswith a tradition of such policies, particularly theUnited States, understand the necessity to care-fully design and periodically modify the policiesin order to obtain the desired effects and preventthem from generating an outcome which is com-

pletely opposite of what they set out to accom- plish.

Nevertheless, there are sufficient hints as to thereality of residential segregation in the Region,

particularly concerning its effects and trends,helping to provide the general foundation of a pol-icy to control segregation. The subsequent pagescontribute to this task.

GENERAL OBJECTIVES OF A POLICY TOCONTROL SEGREGATION

The general objectives of a policy to control resi-dential segregation should be to stimulate socialintegration. In order to do so, the policy must dif-ferentiate the explainable and positive compo-nents of segregation from the negative ones, espe-cially those affecting the poor.

Simultaneously, this policy should balance thecombination of two possible methods of socialintegration: the traditional method of social mobil-ity, and the integration from diversity, predomi-nantly emphasized today.

In fact, the trendy concepts of urban social inte-gration encourage the efforts to surpass the old

“assimilation” models of social integration, re- placing them with models that increase integrationfrom the diversity. Specifically, the classic inter-

pretation of the sociologists of the School of Chi-cago, from the early 20 th century, pointed out anassimilation of minorities into the predominantsocial and physical structures of the city, loosingtheir initial spatial segregated status. The critiqueof the school of Chicago and the new concepts ofintegration seem to suit the cities in which integra-tion, and similarly residential segregation, prob-lems include clear ethnic and racial components,as in the United States, Canada and Europe.

Nevertheless, how can we apply these new em- phases to our cities, where the most expl

the social spatial segregation in the cities of latin america

Documents