Central-city residential migration

Material Information

Central-city residential migration integration and the changing distribution of income
Axelrad, Michael J
Publication Date:
Physical Description:
113 leaves : charts, maps ; 28 cm


Subjects / Keywords:
Minorities -- Housing -- Colorado -- Denver ( lcsh )
Minorities -- Housing ( fast )
Colorado -- Denver ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 105-109).
General Note:
Cover title.
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Planning and Community Development, College of Design and Planning.
Statement of Responsibility:
by Michael J. Axelrad.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
13756410 ( OCLC )
LD1190.A78 1986 .A96 ( lcc )

Full Text
U1 fl 7 0
5 b 3 7 3 7

CENTRAL-CITY RESIDENTIAL MIGRATION: Integration and the Changing Distribution of Income
Michael J.
A Thesis Submitted to the Graduate Faculty in Partial Fulfillment of the Requirements for the Degree of Master of Planning/Community Development
University of Colorado at Denver Denver, Colorado
May, 1986

I. SUMMARY...........................................................1
II. INTRODUCTION......................................................3
III. DENVER HOUSING MARKET .......................................... 18
IV. MINORITY HOUSING MARKETS.........................................42
VI. CONCLUSIONS......................................................92
APPENDIX: THE SPSS PROGRAM......................................Ill

1. The Denver-Boulder SMSA: Population Growth, 1970-80............ 22
2. The Denver-Boulder SMSA: Household Incomes, 1969-79............ 24
3. City of Denver: 1980 Median Household Incomes by Ethnic
Group, 1980................................................ 43
4. City of Denver: Distribution of Census Tracts by Percent
Minority, 1970-1980........................................ 49
5. Comparative Variable Responses in Minority Census Tracts. ... 51
6. Compared Multiple Regression Coefficients ....................... 57
7. Market Characteristics in Census Tracts Experiencing
Black Integration...........................................65
8. Market Characteristics in Census Tracts with Potential
for Black Integration.......................................69
9. Southeast Denver Census Tracts by Potential for Black
10. Market Characteristics in Census Tracts with Potential
for Hispanic Integration .................................. 76
11. Southwest Denver Census Tracts by Potential for Hispanic
12. Denver-Boulder SMSA: Concentration of the Black Population,
1970-80.................................................... 84
13. Denver-Boulder SMSA: Household Income Distribution in
Predominantly Black Neighborhoods, 1970-80 ................ 86
14. Denver-Boulder SMSA: Concentration of the Hispanic
Population, 1980 .......................................... 87
15. Denver-Boulder SMSA: Household Income Distribution in
Predominantly Hispanic Neighborhoods, 1970-80........... 90
16. City of Denver: Indices of Neighborhood Integration...............91

1. The City and County of Denver, 1970 Census Tracts................20
2. The Denver-Boulder SMSA..........................................21
3. City of Denver: Location of Oldest Housing Units, 1980 .... 31
4. City of Denver: Location of Densest Census Tracts, 1980. ... 31
5. City of Denver: Upper Income Residential Sectors ............... 33
6. City of Denver: Location of Black Neighborhoods, 1970....... 44
7. City of Denver: Location of Black Neighborhoods, 1980....... 45
8. City of Denver: Location of Hispanic Neighborhoods, 1970 ... 46
9. City of Denver: Location of Hispanic Neighborhoods, 1980 ... 47
10. City of Denver: Census Tracts Undergoing Initial
Integration, 1970 to 1980................................. 64
11. Southeast Denver Census Tracts by Potential for Black
12. Southwest Denver Census Tracts by Potential for Hispanic

The Relationship Between Age and Income, 1970 and 1980. .
The Relationship Between Density and Income, 1980 ...........
The Relationship Between INDENS and Income, 1980.............
The Relationship Between Owner Occupancy and Income, 1980 .
The Relationship Between Average Sales Price and Income, 1980
The Relationship Between Percent Black and Income, 1980 .
The Relationship Between Percent Hispanic and Income, 1980. .
The Relationship Between IHISP and Income, 1980 .............
Southeast Denver Census Tracts by Percent Owner-Occupied and Median Household Income, 1980................................
Southeast Denver Census Tracts by Percent Sold and Percent
with Multiple Sales, 1980-84 ...........................
Southwest Denver Census Tracts by Percent Owner-Occupied and Median Household Income, 1980................................
Southwest Denver Census Tracts by Percent Sold and Percent
with Multiple Sales, 1980-84 ...........................

Studies of minority residential patterns have generally focused upon central city-to-suburb patterns. Clark and Frey have identified the advantages of suburban location for whites, as suburbs have greater exclusionary powers and an advantageous position to capture population and employment growth. Although the disparities between central cities and suburbs remain an important social and policy issue, a wider framework of residential migration patterns is needed, one which distinguishes neighborhoods by market characteristics as well as municipal location. Within the central city, a diversity of housing types exist. In addition to a greater definition of neighborhood types, migration studies will benefit by an delineation of the differences in migration patterns for minority groups, which will be examined in this study for the central city of Denver, Colorado.
When minority patterns are compared with the patterns for the entire market, the black housing market shows patterns similar to the city-wide market, although it has been hindered by lower levels of income. This affirms the contention of researchers that discrimination in the past has prevented black households from obtaining equal quality housing.
The recent patterns of limited integration in small concentrations into previously white markets indicates a loosening level of discrimination. Blacks who have obtained middle-class status have become more acceptable to

the white middle class. Thus, new patterns of discrimination are based upon income level and class status rather than race.
The Hispanic market, in contrast, shows some unique patterns. Unlike the black market, which is slowly integrating in small concentrations into white neighborhoods, Hispanic households have tended to invade neighborhoods as the population has grown and to locate in greater concentrations, effectively "tipping" the neighborhood from white to Hispanic. Generally, blacks and Hispanics have located in separate areas of the city and high concentrations of both minority groups within a single neighborhood is rare. Market patterns show that higher concentrations of blacks in Hispanic neighborhoods will significantly lower the income level in the neighborhood. The Hispanic population, therefore, tends to be a more cohesive community, showing higher concentrations of Hispanic households regardless of race and less likely to locate in white or black neighborhoods.
The lessening discrimination felt by the growing minority middle class has not extended to a lessening discrimination felt by the entire minority population. The largest share of minority households are still located in a small number of neighborhoods which have the highest concentrations of non-white persons. Thus, while integration has become possible for some households, discrimination has limited this possibility for the majority of minority households, who are still trapped in older residential areas, with limited opportunities for quality housing. Future policy should therefore focus on upgrading these neighborhoods and providing educational and occupational opportunities for these areas.

Residential segregation occupies a key position in patterns of race relations in the urban United States. It not only inhibits the development of informal, neighborly relations between whites and Negroes, but ensures the segregation of a variety of public and private facilities.^
Since the mid-1960s, when the previous quote was written, a number of social trends and public programs have combined to lessen the existent segregation in American society: "the civil rights movement and the Great Society programs of the 1960's as well as the general strength of the national economy have led to a significant narrowing of the black-white income gap since 1960.From the late 1960s until the early 1980s, the Federal government used a combination of legislative action and judicial response to promote integration and the dispersal of low-income housing.^
The enactment of key pieces of legislation including the Fair Housing Act of 1968 (FHA) and the Housing and Community Development Act of 1974 (HCDA) provided some relief from segregation. Federal initiatives, however, have been severely limited for a number of reasons, but primarily "because of the absence of enforcement powers, [and] the failure of the Supreme Court to set an unequivocal standard for measuring discriminatory behavior."^ The result has been that only a small portion of the minority population has been able to integrate into white society, where a level of discriminatory behavior remains.
Regrettably, some twenty years after Martin Luther King's ennobling "I have a Dream" speech and almost a decade and a half after

enactment of the nation's first reasonably comprehensive civil rights laws, vast numbers of black, brown, and yellow Americans remain unable to fully participate in or benefit from the American way of life.^
Minority households continue to be segregated into specific housing markets, and these generally are of lower quality. Because of their lower income levels, minority households have had last choice for quality housing units, which has often meant ghetto-type slum dwellings.
This discrimination, however, has lessened. Evidence shows that more blacks are now able to afford housing in suburban areas.^ With the limited effect of Federal action, the levels of integration occurring in American society are attributable to gradual social change: acceptance of minorities in the housing market. Although a level of racial discrimination still exists, for a growing number of Americans the idea of a black or Hispanic household living on the same street does not invoke the same hysteria that it would have 30 years ago, or in some neighborhoods as recently as 10 years ago. As more minority households have assimilated into white society, the remaining levels of discrimination may be more strongly rooted in economics than in race: households "are more sensitive to income than to racial composition.
Patterns of Residential Relocation: A Survey of Existing Literature
Recent research has focused upon the differing patterns of minority and non-minority migration. Most frequently, this has meant comparing black and white patterns; little research has been done on hispanic patterns, which are so important in the Denver area. Nonetheless, this paper assumes that many of the Hispanic patterns will generally resemble black patterns.
These patterns are based upon classical theories of urban growth, primarily the concentric theory of Burgess. The basic pattern of metropol-

itan growth occurs as the concentric rings expand outward. New homes constructed on the edges of the urbanized area are consumed by upper and middle-income residents who are capable of paying the higher cost of housing at these locations. The older housing units, as they age and deteriorate, become affordable to lower-income residents. Across time, the individual neighborhood passes into consecutively lower-income housing markets, following "the ecological model of invasion, transition, and succession.
This process is often referred to as the "trickle-down" or "filtering" process.
Towards the city's center, the old, deteriorated units have become
slums, housing the lowest-income classes.
According to the principal theory of urban growth, there are gradients in residential land use, with persons of lower socioeconomic status living closer to the central business district.
Just as metropolitan Negroes tend to live in the central city, so might it be anticipated that within the central city, Negro residential areas would tend to be centrally located.
Downs notes that the "dynamics are essentially the same [for] poorer
households, or minority-group members, or both.''^
Generally, the outward movement of the population occurred because of
a large number of factors but one factor in post-World War II expansion was
"the substantial increase in black migration from the rural south.
Originally, the new housing units were within the boundaries of the central
city, but the
duality of city and suburbs was ensured as central annexation gave way to the fragmentation of the suburban periphery, creating numerous competing municipalities, each empowered to manage externalities and serve as its own gate keeper.^
This fragmentation provided further restrictions on minority mobility,

as suburban communities enacted ordinances which were economically exclusionary:
Many suburban communities, supported by a longstanding judicial bias in favor of the validity of municipal land-use regulations, have adopted no-growth, limited growth, and biased growth land-use controls. These controls influence not only the availability of land for residential development, but also the price of land and therefore the cost of residential development. ^
This has resulted in a self-reinforcing pattern of exclusion: "Lower
suburban population densities result from the evolving capacity of suburbs
generally to attract higher-income households opposed to higher-density
residential development."^
During the 1960s and 1970s, Frey has posited, "it is not likely that recent white out-movement from large central cities is heavily influenced by interracial housing dynamics as had been the case in the 1950s.
But the on-going suburbanization of the white population, commonly referred to as "white flight," certainly has a number of causes. In northern cities, judicial enforcement of busing plans to achieve public school integration has commonly been identified as a factor. Yet, Frey argues that the continuing suburbanization has occurred primarily in response to other factors:
Significant racial desegregation in central-city schools and the occurrence of racial disturbances during the period contributed little to the explanation of city-to-suburb "white flight," whereas ecological features of the SMSA and city-suburb fiscal disparities proved to be important determinants. One racial factor-- the percentage of the central city population which is black did influence white out-movement.
The result of "white flight" has been a severe erosion of the tax base of the central city as suburbanization "served to redistribute a good deal of the metropolitan area's tax base out of the central core."^ According to the traditional theory, the lowest-priced residential market continues to

expand, occupying an increasing portion of the core city. "As a consequence, a more disadvantaged population has been left behind which contributes less to, and demands more from, city financial resources."1* Closely tied to the spread of the low-income residential population is the increasing minority population which has been stranded by white suburbanization.
Because the suburbanization phenomenon has been confined almost exclusively to the white population, the black share of central city populations has increased noticeably over the past thirty years. In short, black movement both into and within the metropolitan area is restricted to the central city, while white movement occurs on a metropolitan-wide basis. As a result, the high levels of poverty and unemployment that have come to be concentrated among urban blacks have also come to be concentrated within large central cities.
The housing opportunities for minority households are linked to the
expansion of the metropolitan market, however.
Poor households must enter the center of the area and gradually spread outward, successively displacing higher-income households.
Thus the trickle down process requires, in effect, a constant outward movement of all income groups.^1
The minority population's housing quality improves, therefore, as the "white flight" continues, leaving a supply of higher-quality units into which the more capable households can move. A second process occurs when the expanding black population "invades" a white neighborhood. Following the initial entry of a few minority households,
[ijrreversibility is implicit in the widely accepted concept of a "tip point," Once the percentage Negro passes the tip point, it is assumed that whites will leave the area at an accelerated rate and be replaced by Negroes until the neighborhood has become entirely Negro. ^
Although researchers have not found "a single, universally applicable tipping point which can explain and predict the point at which neighborhoods will irreversibly change from white to nonwhite,"23
as late as the 1970s,

research showed that whites "tended to be reluctant to stay in neighborhoods
into which blacks were moving or to buy homes in interracial areas.
One dimension of housing discrimination is revealed by the fact that "the
black 'invaders' were often higher in status than the resident whites.
Thus, minority households generally must earn more than non-minority
households in order to occupy housing of equal quality: "there is evidence
of discriminatory behavior against blacks that takes the form of exclusion
and price discrimination on the part of housing suppliers."^
Recent Trends Affecting Migration Patterns
During the 1960s and 1970s a number of sociological and economic
changes occurred, resulting in some loosening of the discriminatory patterns
created by the filtration process. One important change has been the growth
of a black middle class, as documented by Wilson:
As race declined in importance in the economic sector, the Negro class structure became more differentiated and black life chances became increasingly a consequence of class affiliation.^
Connected to this, and resulting from it, is a second trend, greater black
It involves a market increase in the level of black movement to the suburbs and suggests that the black population is, at last, permitted to participate in the suburbanization process.
A third trend "involves a changing preference on the part of metropolitan
whites for a central-city residence.These three trends have created
significant new opportunities for a portion of the minority population, yet
overall have separated the minority into a middle-class of "haves" and a
lower-class and under-class of "have-nots," who are left with practically
no opportunity to enhance their life chances.

Analysts have agreed that the growth of a minority middle-class has
affected only a small portion of the minority population.
The civil rights movement of the sixties made a fundamental difference in the lives of younger, educated blacks but achieved only marginal gains for older, less-educated blacks. The movement's goals were generally oriented more toward status and dignity for the black middle class than toward the bread-and-butter issues for the black working class.^0
The growing class distinctions within the black community has resulted from
educational and occupational opportunities:
A fairly elaborate black class structure has developed, especially since 1940, with three clearly distinguishable groups-- an underclass population of largely unskilled laborers, domestic workers, and welfare recipients; a stable working-class population of largely semi-skilled blue-collar workers; and a middle-class population of largely semi-skilled blue- and white-collar workers. On the basis of this classification, roughly 35 percent of the black population today is in the underclass, about 30 percent in the working class, and approximately 35 percent in the middle class.
According to Pettigrew's assumption, the 35 percent who belong to the black middle class are primarily younger blacks who were able to take advantage of the educational and social benefits which resulted from the civil rights movement in the 1960s.^2
Concurrently with "white flight," a growing share of employment located in suburban areas. Sadly, the jobs most suited for the urban poor and the concentrations of the minority population were precisely the types of jobs which were relocating in suburban areas: "Dispersion is more marked among blue collar jobs in manufacturing, wholesaling, and distribution, least pronounced in white collar jobs and in service industries."^3
Most employment growth will be in lower-paying jobs. A national trend "has been the pronounced decline of most of the industrial sectors."34 The American economy has seen "the permanent loss of high quality employment

when manufacturing jobs are not replaced or are replaced by low-paid, low-quality services jobs."33
These trends combined have isolated the low-skilled minority worker
from employment opportunities as the blue-collar jobs in manufacturing,
wholesaling, and distribution "are moving to places that are scarcely
accessible by public transportation."36
Today, the economic and technological revolution of advanced capitalist society threatens to solidify the position of the black underclass. With the lack of job expansion in the manufacturing sector, and the fact that the better-paying jobs in the service industries require training and education, the poor urban blacks face a situation in which the higher paying jobs that can be obtained without higher education and/or special skills are decreasing in central cities' not only in relative terms, but often in absolute numbers.3?
In contrast, "the black middle class has expanded greatly in recent years; consequently, most middle-class black families are first-generation occupants of this status."3
As blacks acquire middle-class status, many now find single-family
suburban housing available to them. Early movements of blacks to the
suburbs actually reinforced patterns of segregation:
As increasing numbers of blacks gained access to suburban communities during the 1960s, the concentration in a few suburbs with higher than average existing black populations reinforced racial segregation in the metropolitan ring."39
During the 1970s, "more prosperous black households seeking single family
residences have probably secured the greatest freedom of suburban movement,
though there is evidence this movement is significantly restricted.The
loosening restrictions on black suburbanization is the result of a number of
social trends:
The greater suburban movement of blacks can be linked to rising black social status, some elimination of past discriminatory practices and improvements in race relations, which appear to be

reducing the constraints that have prevented blacks from fulfilling their preferences for suburban locations.
In spite of the increased number of skilled minority workers filling
professional job positions, the combined problems of access and training
"are largely responsible for the creation of a semipermanent underclass in
the ghettoes."^ And this underclass is facing the loss of the better
residential units in inner-city neighborhoods by the intrusion of "gentrifi-
cation." A definition of gentrification which focuses upon the effects upon
the existing population has been developed by Clark:
gentrification is defined to be the upgrading of resident incomes in older, lower income neighborhoods through:
(1) the displacement or replacement of lower income by higher income households on individual parcels, through the reuse of existing structures or residence-to-resid-ence conversions;
(2) the elimination of lower income units through conversion to nonresidential use or vacant status; and
(3) the addition of higher income units on vacant or nonresidential parcels.^
This process has been viewed favorably by public officials; "the return of upper income whites is looked upon as a means toward slowing the pace of tax base erosion, if not as a stimulus toward central city revival.In addition, many urban policy makers "would prefer central-city revitalization in the interest of the poor, though the poor seem to be at a distinct disadvantage in capturing the benefits of core-area redevelopment."^
In actuality, the process of gentrification, which has been dubbed "trendi-fication" in Australia,offers benefits only to the upper-class residents who are expected to move into the revitalized neighborhood. "In the process, many downtowns are being converted into bourgeois playgrounds."^ The dislocation of the existing residents has forced the poor, predominantly minority market outward, "invading" adjacent neighborhoods, often

at a cost which the poor household is unable to bear. Gentrification, however, may have slowed the rate of white suburbanization. Simultaneously, the growth of the black middle class has allowed blacks to capture a greater share of the supply of suburban housing, although some level of discrimination still exists:
Increasing black concentration is more likely to occur in communities with a weaker property tax base and in suburbs with higher tax rates [indicating] that the dual housing market effectively steers blacks to disadvantaged communities.2*^
Hispanic Residential Patterns
Hispanics have been subject to as much discrimination as the black population, and in some respects have fared Worse. Unlike the blacks, Hispanics have not had any popular leaders such as a Malcolm X or Martin Luther King who have been able to spearhead a movement for civil rights for urban-dwelling Hispanics.
Few studies have been conducted on the migration patterns of Hispanics. However, research has revealed some differences between Hispanic and black patterns. First, the term "Hispanic" includes a number of diverse ethnic groups: "They include Cubans, Mexican-Americans, Puerto Ricans, and small groups from other Latin American nations, such as Colombia and Guatemala."2*^ These separate groups show a diversity of ethnic characteristics, primarily skin color. "Those with light-colored skin integrate easily into Anglo neighborhoods, unlike light-colored blacks. As a result, Hispanics are less residentially segregated than blacks."^
At the same time, the Hispanic population as a whole tends to remain in the same neighborhood regardless of class:
as Hispanics improve their economic circumstances they tend to remain in the same locations rather than, as many blacks do, move to "better" neighborhoods. They add rooms or purchase more homes

nearby. Therefore Hispanic neighborhoods typically contain a broader spectrum of incomes than black and Anglo neighborhoods.^
Hence, one would find some stratification of the Hispanic population based
upon skin color. Among Hispanics who are unable to pass for whites, the
tendency is to remain within the established community. Since the Hispanic
population is growing at a faster rate than the total population, the
Hispanic community is occupying a growing number of neighborhoods. Thus,
the classical pattern of transition and succession from white to non-white
is the theoretical framework explaining Hispanic migration.
Lifecourse Migration Patterns
Recent migration patterns among blacks and whites, Frey has argued, have become more similar, as suburban relocation has become more available to black households. Even with the greater propensity to stay within the city shown by whites, the general trend is still to move to the suburb.
Frey suggests that the the series of moves a family makes across time follows a set "lifecourse (age-related) migration patterns.
In the past, differences between black and white patterns were based upon "societal influences which affect destination preferences (in the case of whites) and constraints that have prevented the fulfillment of preferences (in the case of blacks).^3 Overall, a family's needs will be similar regardless of race, and these needs will change across the lifecourse of a family.
The act of migration is a strongly age-related phenomenon which is far more likely to take place among young adults below age 35 than among the older segment of the population.^
The "normal" lifecourse process of white households involved a series of moves associated with its needs at that point in time:

the family's first several moves were actually directed to a central city destination. However, these city-destined moves were made in the early ("pre-child" and "child-bearing") stages of the life cycle and each of these was of fairly short duration. Later moves to suburban destinations were associated with the more sedentary ("child rearing," "child launching," "post child," and "widowhood") stages of the life cycle and were generally associated with long durations of stay.-^
The different migration patterns for blacks during the post-war period-- the
low levels of suburbanization-- were "due to a lack of access rather than
preference.While white levels of suburbanization were determined by age
categories, suburbanization rates for "the black population, in the
aggregate, did not vary appreciably across age categories.
By analyzing recent suburbanization changes, Frey has found that among
blacks, migration patterns "are significantly more suburban directed than
those of previous decades."^ Among whites, however, patterns have not
changed. "Their 1975-1980 destination propensity rates provide for a
greater suburbanization of non-blacks.
The population within a neighborhood, therefore, is comprised of a
number of sub-groups. First, there are families and older couples, who
settled in the neighborhood when they were raising a family. During that
time, whether the neighborhood was in the outer portion of the central city
or in a suburb, the neighborhood was a prime middle-class market, affordable
yet far enough away from areas in which it would be considered "dangerous"
to raise a family. At the same time, the neighborhood could have been a
temporary location for a young family which moved there yet planned to move
farther away from the central city.
As the neighborhood ages, more and more development occurs beyond it
and the growth of the lower-class population may "invade" areas closer to
the neighborhood. At this point, a larger share of the homes in the

neighborhood would be occupied by younger families which have not yet begun to raise children. These families would eventually move outward towards a suburban location. According to recent trends, a growing share of these families would be non-white.
Purpose and Scope of This Study
Thus, the migration patterns within a housing market are complex, dependent upon demographic and supply characteristics. The delayed accessibility of single-family markets to minority households has resulted in a pattern in which outer neighborhoods within the central city and suburban neighborhoods had been primarily white, except for those areas which had been "invaded" and "tipped" to minority markets.
Generally, the research on migration patterns has been concerned with central city/suburban discrimination and migration. The data for differences within the various central city neighborhoods has not been studied, except very generally in studies which have discussed interdependent market processes, such as Clark's study of gentrification,^ or processes of minority "invasion.In this study, however, the focus will be upon the market characteristics of central city neighborhoods, which frequently resemble suburban neighborhoods.^ Thus, the overall theoretical framework which describes suburbanization of minorities is assumed to hold true for newer central city neighborhoods, as well.
This study will describe the extent of change within the central city neighborhoods of Denver, Colorado across the period from 1970 to 1980. The key emphasis will be upon the extent to which market variables govern the distribution of income within these neighborhoods. The changing distribu-

tion will indicate the changing position of these neighborhoods within the central city market and the entire metropolitan market.
First, the variable relationships for all census tracts will be analyzed to describe the patterns of the entire Denver market. Following this, the patterns within census tracts with high percentages of black or Hispanic residents will be compared with the entire market to discern the unique patterns of change within the minority markets. Patterns of integration and the extent of integration as they are revealed by the changing distribution of income will then be described.
Based upon the theoretical framework cited above, it is expected that a certain amount of integration has occurred across the period of study.
Within the Denver neighborhoods which had been completely or overwhelmingly non-minority prior to 1970, a certain extent of integration will have taken place by 1980 for both the black and Hispanic populations.
Although little research on Hispanic migration was available, the analysis of market data is expected to indicate the differences between the patterns for the two minority groups. Overall, the comparatively larger size of the Hispanic population, combined with its stronger community ties, is expected to result in the more classical pattern of invasion and replacement of the non-minority population, with less division of the population by income level. The black population, on the other hand, may not have sufficient size or buying power to have caused "tipping" in previously nonwhite neighborhoods. Instead, the middle-class black population has integrated into a number of white communities, with a wider geographical range of activity, but at lower concentrations.

The extent of the on-going integration is allowed partially by the growing acceptance of the minority middle-class into society. As Wilson and Pettigrew indicate, educational and occupational opportunities have only reached a limited portion of the minority population. This middle-class minority population has become acclimated to white society, sharing its norms. The remaining level of discrimination is governed more by economic position than race.
This, however, means that the majority of the black and Hispanic populations will still be trapped in the lower-class housing markets, aggravating the gap between income levels in the lower- and middle- income areas. Reflected in the income distribution, there is expected to be greater segregation by income level within the Denver market.

Method of Analysis
Using the theoretical framework created by other researchers, this thesis will describe the actual performance of the Denver housing market. This analysis utilizes an extensive data base containing demographic information from the U.S. Census and the Denver Regional Council of Governments (DRCOG) collected at the census tract level.^ The third source of data was the annual Housing Detail Report^ produced by the Denver Planning Office. This latter source also contains census tract level information and provides an extensive array of market information. The SPSS program used for the data base is reprinted in the appendix, including a complete list of variables.
The analysis will consists of a cross-sectional study of a variety of demographic and market variables, which will indicate the changing nature of the relationship between these variables and each's spatial relationship throughout the city. This study assumes that the "trickle-down" process is applicable to a significant share of Denver's neighborhoods, so that a cross-sectional study will "rank" the census tracts, indicating their positions within the entire market.
The Housing Detail Reports track data for 117 census tracts, although the number increased between 1972 and 1983, as the census tract boundaries were changed for the 1980 census. To make all analyses comparable, the 1970

tracts were used throughout. Of the 117 tracts in the 1970 census, this analysis includes 115 (Map 1), excluding only 41.05 and 44.02. Tract 41.05 consists primarily Stapleton International Airport and contains only a small amount of residences. Tract 44.02 contains Lowry Air Force Training Center, which contains only military-provided housing. These units are unaffected by market activity.
The City and County of Denver: The Changing Role of the Central City
The Denver-Boulder SMSA contained six counties in the 1980 census: Adams, Arapahoe, Boulder, Denver, Douglas, and Jefferson (Map 2). The major urban area in Colorado, it is also extremely fragmented, containing 37 incorporated municipalities. This political fragmentation, expectedly, has arisen out of the desire of a number of metropolitan residents to form communities with economically exclusive growth ordinances. The result of this fragmentation has been the concentration of the lower-income and minority populations within the central city.
In response to Denver's annexations during the 1970s, the suburban communities lobbied for the Poundstone Amendment to the state constitution, a law which virtually prohibits Denver from annexing any more land. In its land-locked position, Denver has seen population and employment growth occur in the SMSA, but has been unable to capture a significant share.
Like most central cities, Denver saw a net loss of population from 1970 to 1980, as a growing share of the population sought suburban housing.
Across that decade, the population of the central city fell from 514,678 to 492,365, while the entire SMSA grew from 1,235,936 to 1,618,461 (Table 1). Denver's population, as a share of the entire region's, declined from 41.6%
in 1970 to 30.4% in 1980.

Map 1
The City and County of Denver, 1970 Census Tracts


.£ 2
J >
o c


Map 2
The Denver-Boulder SMSA

Table 1
The Denver-Boulder SMSA: Population Growth, 1970-80
Total Population (% of SMSA)
1970 1980
City of City of
Denver SMSA Denver SMSA
514,678 1,235,936 492,365 1,618,461
41.6% 30.4%
Black Population (% of SMSA)
(% of Total Populati
47,011 50,188
) 9.1% 4.1%
59,252 77,786
12.0% 4.8%
Hispanic Population (% of SMSA)
92,348 173,687
(% of Total Population)
18.8% 10.7%
Source: U.S. Bureau of the Census, 1970 Census of Population and Housing, Census Tracts, PHC70-2-138, Denver SMSA and
U.S. Bureau of the Census, 1980 Census of Population and Housing, Census Tracts, PHC80-2-138, Denver-Boulder SMSA.

In 1970, Denver contained 47,011 black residents, 93.7% of the metropolitan total; the suburban counties combined contained only 3,177. By 1980, some integration of the suburbs had occurred; in this year Denver only contained 76.2% of the total metropolitan black population. Within Denver, the black population increased from 9.1% of the population in 1970 to 12.0% in 1980.
The central city also contains a significant share of the regional Hispanic population, though the concentration is not of the same magnitude as the black population. The percent of the population of Spanish origin was not tabulated in the 1970 census, so no cross-sectional comparison of the Hispanic population is possible. In 1980, however, Denver contained 92,348 persons of Spanish origin, 53.2% of the metropolitan Hispanic population. Within Denver, Hispanics accounted for 18.8% of the population, compared with a share of 10.7% for the entire SMSA.
The central city is also the location of most low-income areas as well (Table 2). From 1970 to 1980, the entire SMSA saw increases in the share of households with incomes between $30,000 and $74,999 (1970 incomes are in 1980 dollars to correct for inflation) and a slight decrease in the share with incomes below $15,000. At the same time, in the central city the share of households with incomes below $30,000 increased.
Across the decade, the median household income in the SMSA (corrected for inflation) stayed relatively constant, increasing by 1.1% from $19,769 to $19,982. The median household income in the central city declined from $16,420 to $15,507, illustrating the suburbanization which occurred during the 1970s and the increasing concentration of lower-income households in the

Table 2
The Denver-Boulder SMSA: Household Incomes, 1969-1979
A. 1970 Households by 1969 Incomes
Income Households % of Total Households % of Total
$0-$14,999 86,424 46.6% 146,010 36.9%
$15,000-$29,999 65,854 35.6% 162,056 41.1%
$30,000-$49,999 25,073 13.5% 67,853 17.2%
$50,000-$74,999 3,810 2.1% 10,406 2.6%
$75,000 or more 4,160 2.2% 8,753 2.2%
TOTAL 185,321 100.0% 395,078 100.0%
Median: $16,420 $19,769
B. 1980 Households by 1979 Incomes
Denver SMSA
Income Households % of Total Households % of Total
$0--$14,999 102,637 48.5% 219,207 36.0%
$15,000-$29,999 72,065 34.0% 232,803 38.3%
$30,000-$49,999 27,578 13.0% 118,935 19.5%
$50,000-$74,999 6,460 3.0% 26,030 4.3%
$75,000 or more 3,264 1.5% 11,711 1.9%
TOTAL 212,004 100.0% 608,686 100.0%
Source: Denver Regional Council of Governments, Profiles of 1970-80 Socio-Economic Change by County and Census Tract (Denver: DRCOG, 1983)

central city. In 1980, the median income in Denver was 22.4% lower than the metropolitan median.
Physical Variables
The variables analyzed in this chapter fall into two groups. The first group describes the actual physical condition of the housing supply, while the remainder all reflect market characteristics, which include a mixture of supply, demand, and price characteristics. The two physical variables used in this analysis are the age of the housing supply and the density. According to theories of neighborhood change, the most fundamental determinant of the income level in the neighborhood is the age of the housing stock. In actual performance, census tracts in Denver indicate that such a relationship exists, shown in the following equations:
In 1970, INC = 25,684 - 199.8 (AGE); R2 = .453
In 1980, INC = 25,511 - 195.3 (AGE); R2 = .441
In both equations, INC represents the median income for all households within a census tract (in 1980 dollars to correct for inflation), and AGE equals the average age of all single-family units in the census tract.
The similarity between the intercept and slope values in the two equations reinforces the conclusion that aging has a constant effect on .income which has not changed over time. It is a fixed characteristic that older housing is occupied by lower income households. Generally, the relationship states that for each additional year the median income declines by $195 to $200. The average ages for 1970 may be skewed slightly, as these were equated solely by subtracting 10 years from the average ages in 1980. The average age may be overestimated in newer census tracts which saw construction

during the 1970s and in older tracts where some older units have been demolished or converted to non-residential use.
The correlation coefficients (R^ values) indicate that the relationship is fairly strong, but there is a large amount of variation from the regression equation (Graphs la and lb). Not all of the high-income tracts had low average ages; of the tracts with median incomes exceeding $25,000 in 1970, the average age ranged upward to 25. In 1980, three out of 14 tracts with median incomes exceeding $25,000 had average ages above 30. For higher income areas, in which site rent is higher, aging of the housing stock does not result in deterioration at the same rate that it occurs in lower income areas where site rent is lower.
In both 1970 and 1980, a number of census tracts had incomes significantly lower than expected. This could indicate that age has a greater effect on the income level that is represented in the regression equation; the incomes may decline more sharply and then level out at higher age levels as the theory predicts. None of the submarkets alone, low-, medium-, or high-income, shows a strong relationship between age and income, only the combined interrelationships between age and income including all submarkets established the regression lines in Graphs la and lb. Filtering out of various sub-markets does not result in stronger correlation coefficients.
Density is closely tied to both age and income level. Post-World War II housing has been built at lower densities, a trend which has continued even as land costs have skyrocketed.^ Within Denver, where the older neighborhoods were built at lower densities than the cities of the midwest and northeast, the density gradient should not represent a sharp decline

Graph 1
The Relationship Between
Age and Income,
1970 and 1980
A. 1970
M733 S3997
B. 1980
SPSS satcn svsrrr
o uir ?* icr

away from the center of the city, but a gradient does occur, as the newer housing stock has been built at low densities.
The initial attempt was to establish this relationship as linear (Graph 2). Even though the correlation coefficient equals .409, the leveling out of densities for lower income markets creates a curved shape rather than a linear relationship. Even using a log transformation, the relationship is not strengthened. The relationship between income and the resultant variable EDENS, e(10*H(DENS/33,333)), still creates a curved shape and the value decreases to .415. The strongest relationship between density and income is established using the inverse of density, INDENS (Graph 3). The equation correlation coefficient improves to .474.
The relationship between the age of the homes and density is also fairly well established, with a coefficient of .494.
Age and density are the key variables describing the physical characteristics of the housing in a particular sub-market. The two variables exhibit a linear relationship, and both are related to the income level, represented by the median income. Generally, the oldest housing stock is located near the Central Business District, as the concentric circle theory suggests. Although the circles are uneven and often discontinuous, the oldest homes are located roughly in a circle around the Central Business District (Map 3). Density also does not cluster into neat concentric rings. The densest residential areas, nonetheless, are also the oldest areas, distributed about the Central Business District (Map 4).
These basic structural characteristics generally define the housing market. The low income housing market is concentrated in the central part of the city, in the older, denser residential areas. Moving outward

Graph 2
The Relationship Between
Density and Income, 1980
3044 lO 31220 20 27896 30 24372 40 21248 30 17924 60 14600 70 11276 BO 7932 90
4629 00
4130 37 lC1** 48 17300 60 23993 72 30610 83 37223 93 43041 07 30436 IB 37071.30 63606 41 70301 33
37868 00 34344 10 31220 20 27896 30 24372 40 21248 30 17924 60 14600 70 11276 BO 7932 90 4629 OO
SPSS BATCH SYSTf'H 20 22 04 04/21/86 PAGE 3

STD ERR OF ES7 - 33S7 4cq- INTERCEPT (A) 23308 91372 SLOPE (B) -0 36994

Graph 3
The Relationship Between
INDENS and Income, 1980
*77 31 BOO? 17 11139 or 13310 Bt! 136t>2 74 1B214 60 20366 46 22918 32
34 3-4 10
27896 30 *
24372 40
21248 30
1 7924. 60
14600 70
11276 80
7932 90 *


4629. OO * 2
373 66 2927.32 327 38 7631.24 9983.09 12334.93 14686.81 17038.67 19390 33 21742.39 24094.23
37868 00 34344 10 31220 20 27896 30 24372 40 21248 30 17924 60 14600 70 11276 80 7932 90 4629 OO
SPSS BATCH SVSTCM 20:22.04 04/21/06 PACE 6
COPPELATION (R > - 0 6=030 R SQUARED 0 47436 SIGNIFICANCE 0 ooooo
STD ERF OF EST - 313t 12069 INTERCEPT < A > 7010 18313 SLOPE (B) 1. 31881

Map 3
City of Denver,
Location of Oldest Housing Units, 1980
City of Denver,
Location of Densest Census Tracts, 1980

from the center, income level increases while density declines. The newest housing on the edges of the city will be occupied by the highest-level income groups. This simplified overview neglects varying market conditions, but for a large portion of the census tracts in Denver, these basic physical characteristics interact according to this paradigm.
The most easily identifiable exception to this pattern is the distribution of higher-income census tracts within the older part of the city. Two upper-income "sectors" are well established in older Denver; one extends southeast from the Country Club through Belcaro into Virginia Village, and the second encompasses the Hilltop neighborhood in east Denver (Map 5). Upper-income areas are more stable, for many economic and social reasons.
The higher unit cost restricts the demand for these areas. Built to higher standards, these homes also are more structurally sound, tending to keep prices high. The limited demand for upper-income units also dictates that most new units will be priced for the middle-income market. Finally, having more economic power, the higher-income households form an exclusive group--a group which more readily can control any potential changes to their neighborhood. High-income neighborhoods, therefore, are in a position to maintain a wealthier character, even as they age. The two existing upper-income sectors demonstrate this.
Market Variables
Among the market characteristics, the variable with strongest relationship with income level is the percentage of units which are owner occupied (Graph 4). The relationship is established by this regression line:
INC = 6326.2 + 195.4(00),

Map 5
City of Denver,
Upper Income Residential Sectors


c c
k_ o
> c
- >
w Q3


> X)
O 3

Graph 4
The Relationship Between
Owner Occupancy and Income, 1980
r ile SUErIL E CENSU- STATTCKGRAH 0~ i not.-,- r ic . CATE o: INT8C 1 4 70 1 4/06) 24 30 CENSUS TRACT 33 90 43 DATA 30 20 22 04 (ACROSS) OOSO 33 lO 62 70 72 04/21/86 30 81 PAGE 11 90 91 30
3TB6B 00 I I I I I I 37868 00
I I 1 I I I I
34344 10 i I I * 34344 10
1 1 I I I I I
31220 20 I I I I I I I I I I 31220 20
27896 30 I I 27896 30
I 1 I * I
I I * I I
24372 40 * I 1 24372. 40
21248 30 I I I I I I 21248 30
I < # I
I # I
1 I * I 2 * I
I * * I
1 79Z4 60 I 17924 60
I *
1 1 * #1 # 1

i * I * I I
14600 70 * 1 * I 2# 14600 70
i 2 1 # I I
i * * 1 I I
i * I #* * I I
* I * * I I
1 1276 80 I * I 1 1276 80
i * ** I * I I
i I * # I I
i * I * ** I I
i I I I
^32 90 I I 7932 90
i I
i I I I
i * * 2 I I I
4629 00 * * I I 4629. OO
0 30 9 90 19 30 29 10 38 70 48 30 37. 90 67 30 77 10 86 70 96 30
SPSS BATCH SYSTEM 20 22 04 04/21/06 PAGE 12
CORRELATION STD ERR OF EST - 4012 0195j INTERCEPT (A) - 6326 20071 SLOPE (B) - 193 42774

Where INC = 1980 median household income and
00 = the percentage of all units which were owner occupied
in 1980.
This relationship has a correlation coefficient (R^) of .538, signifying a strong relationship.
The relationship was expected to be one of the strongest; home ownership is a solid value in American society, and this is proven by this variable correlation. In fact, a large number of low and moderate income census tracts show higher percentages than predicted by the regression equation. This indicates that a number of households of low and moderate incomes will spend a relatively larger portion of their income on home ownership rather than renting. Rental units are perceived as having poorer maintenance, inferior locations, higher densities-- all the conditions which are associated with a declining neighborhood. Households who choose to rent are generally perceived as those who are not yet earning a sufficient income to afford home ownership.
A similar relationship exists between income and sales price, which shows this relationship:
INC = 5193.8 + .176(PRICE)
for the year 1980. This equation has an R^ value of .445, which is fairly strong. In actuality, this relationship is not uniformly linear for all income groups (Graph 5). In the lower and lower-middle income ranges, home ownership is available only for households who spend a disproportionately high percentage of income on housing. At the higher end of the scale, household spend a smaller portion of their income. Thus, rather than demonstrating a linear relationship, the actual connection between these variables is a discontinuous curve, with a slope which gradually diminishes

Graph 5
The Relationship Between
Average Sales Price and Income, 1980
04 04/21/be- PAGE
' 2LE
SUBFILE CEN=-J'-. SCATTEPCFAr or r'1......
i*7-* c.
3^666 00
3*34 10
31 220 ro
27696 30
24572 40
21246 50
17924 60
14600. 70
11276 60
7952 90
629 00 *
12600 00
2728'; OO 419^0 G( 56640 00 71320 OO 86000 OO 100680 00 119360. OU 130040 00 144720 00 139400 00
37868 00
34344 lO
31220 20
27896 30
24372 40
21248 30
17924 60
14600 70
11276 80
7932 90
4629 OO
ST b ERR OF EE* - 5144 455J7 INTERCEPT (A)
20 22:04 04/21/86 PAGE 9
3193 78116 SLOPE 0 17628

as price increases.
The relationship between income and the racial characteristics of the population show a weaker relationship. The relationship between income and the percentage of the population which is black has a correlation coefficient (R) of .053, skewed by the thirteen tracts with percentages exceeding 30%. But even when these thirteen census tracts are removed and the regression line is recalculated, the value only increases to .106.
The distribution of the census tracts on this graph does indicate a relationship between the black population and income. Out of 115 census tracts, 97 fall below the dashed line in Graph 6. Thus, for most census tracts, a low median income does not necessarily indicate a high percentage of the population is black. However, higher percentages which are black are related to lower median incomes, so that the general relationship can be expressed by this equation:
INC < 43138 2569(%BLACK)
This relationship, therefore, indicates that an amount of discrimination exists in the housing market. Although a primarily white population does not necessarily indicate a higher income level, the inverse is true: a larger black population in a census tract indicates lower income levels.
Similarly, the percent of the population of Hispanic origin does not show a strong linear relationship; the R^ coefficient is only .296 (Graph 7). A stronger relationship exists using the inverse of the percent Hispanic (Graph 8):
INC =11875.7 + 250.5(1/HISPANIC)
Using this equation, the correlation coefficient increases to .488. Thus

Graph 6
The Relationship Between
Percent Black and Income, 1980
3706B 00
34044 lO
31220 20
270*6 30
24072 40
21248 OO
17*24 60
14600 70
11276 0O
7*02 *0
462* OO
SPSS BATCH SYSTEM 19 08 00 04/21/86 PAOE 3
O 10601 18060 97620
0 00042 -423 70609

Graph 7
The Relationship Between
Percent Hispanic and Income, 1980
rn. * <. t'f : LN r A i 14-P; CLNf:!? TRACI I>A T A
jle cens j:
4 67 12 3 19 63 27 12 34 60 42 08 49 37 37 03 64 33 72 02

3706S 00 I I I I I 1 I I I I
34344 10 I I I l I I I I I 1 I 1 I I I I I
31220 20 I I I I I I I I I 1 I I I 1
I I # I 1 I I
27596 30 * I I
I * I I I
1 r I 1 I
24372 40 * I I *
I * I 1
I 1 2 I Z I I X 1
21248 30 i * I I 1 I I
I 2 ** 2 I I I
I #** * I I I
2 * * I I
17924 60 I I
Z * * I # I
I # I* 2 I I
I ** * I I I
14600 70 ** * * I 2
I ** I I I
I 2* I I I
I 2 * I # I
I #* I * I
11276 80 * I I
I * I * * I
I * 1 I # * I
I * r I I
z * i I
7932 90 * i 1
I * z I I
1 i 4 I
4629 00 i I

o 43 8 41 13 e* 23 38 30 B6 3B 34 43 82 33 31 60 79 6B 27 73 76
SPSS BATCH SYSTEM 20 02 39 04/21/86 PACE I 1
STD ERR OF E ST - 3945 07061 INTERCEPT (A) 19947 30924 SLOPE (B) -184 24693

Graph 8
The Relationship Between
IHISP and Income, 1980
f 1 L i
sup' he cfnsl
00 *
L r- j(j' L'A i L ~ 14/6o) CENSUS TRACI DATA
P 39 O 49
O 60 0 71
O 9? 103
37068 OO
34?44 10
34344 10
31220 20
2786 30
27096 30
24372 40
* 24372 40
21240 30
21240 30
17924 60
17924 60
14600 70 22 *
1 1276 80 *
11276 0O
7932 90
7932 90
4629 OO -*-*2
O 33 O 44 O 33 O 63
4629 OO
O 76 O 87 O 97 1 08
20 02 39 04/21/86 PACE 13
3068 31143 113
0 40830 11873 66896 0
0 OOOOO 23040 74330 0

INC=Median Household Income in 1979 AGE=Average Age of Single Family Homes in 1980 00=Percent of All Single Family Homes Owner Occupied in 1980 IHISP=1/Percent of Population of Hispanic Origin PRICE=Average Sales Price of Single Family Homes in 1980
All the variables discussed above except for density and percent black
contributed to this equation, emphasizing the complexity of market choice in
the city of Denver. The excluded variables resulted from the strong
interrelationships between the set of market variables. The colinearity of
these variables filtered out the excluded variables.
Variations on this same equation were attempted; the percent of Spanish origin and the density were fed into the equation instead of their reciprocal values and other variables were added to the equation. None of these improved the equation significantly; the correlation coefficient (adjusted R^) remained near .9 and the standard error of the equation could not be reduced below 2,400. The size of the standard error, however, could indicate the need to include a wider range of variables such as proximity to other land uses, structural characteristics, or other factors to fully explain the distribution of income. It could also indicate an intangible quality in market choice; beyond all the market characteristics there may exist an unexplainable quality in a neighborhood which makes it more desirable. In any case, the multiple regression equation indicates that the distribution of the various sub-markets within Denver is a function of a number of structural and demographic characteristics.

Location of Minority Housing Markets
For blacks, the classic pattern of neighborhood residence in American cities is a ghetto located in older neighborhoods close to the central district.
The spatial distribution of minority groups within the housing market has resulted from economic and social factors both, but an overwhelming factor has been discrimination. The income of minority workers historically have been lower than white incomes. Even with the progress across the last two decades, this has continued to be an economic burden. In the 1980 census in Denver, black median household income was 24 percent lower and the Hispanic median 22 percent lower than the white median (Table 3).
Within Denver, both the black community and the Hispanic community have lived in distinct sectors (Maps 6-9), the black population concentrated northeast of the CBD and the Hispanic population northwest and west of the CBD. Although the definition of "Hispanic" used by the U.S. Census has changed, making comparisons between two points in time difficult, Map 7 reproduced the map used by James et al. Across that decade, it is clear that the minority populations have spread into larger numbers of neighborhoods, but the expansion of these populations has predominantly occurred within the same sectors. Blacks spread eastward along Colfax Avenue and Montview Boulevard into the suburban community of Aurora. Although the black population dispersed between 1970 and 1980, the

Table 3
City of Denver,
1980 Median Household Incomes by Ethnic Group
1980 Median Difference wit
Ethnic Group Household Income All Households
All Households $15,506
Black Households $12,344 -20.4%
Hispanic Households $12,606 -18.7%
Source: U.S. Bureau of the Census, 1980 Census of Population and Housing, Census Tracts, PHC80-2-138, Denver-Boulder SMSA.

Map 6
City of Denver,
Location of Black Neighborhoods, 1970
OVER 75%
50 to 74.9% 25 to 49.9% 10 to 24.9% 5 to 9.9%

Map 7
City of Denver,
Location of Black Neighborhoods, 1980
over 75%
50 to 74.9% 25 to 49.9% 10 to 24.4%
5 to 9.9%

Map 8
City of Denver,
Location of Hispanic Neighborhoods, 1970
over 75%
50 to 74.9% 25 to 4Q.QV-
Minorities in the Sunbelt, p.

Map 9
City of Denver,
Location of Hispanic Neighborhoods, 1980
>* > o
CO ** >
a c 00 CD o
w "O W
TJ CO O c J O > o' > n
w o c O o D
u. CD Q 3 O I o
over 75%
50 to 74. 9%
25 to 49. 9%
10 to 24. 9%
5 to 9.9%

overall residential patterns were clearly similar. One main difference was the spreading of blacks through a number of the older post-war Aurora neighborhoods. By 1980, those areas were a solid and direct extension of the black residential concentrations within the city of Denver.^
In 1970, most of the black population in Denver was concentrated in a small number of census tracts (Table 4); 97 tracts had a black population which was 5 percent or less of the total population. By 1980, some deconcentration of the black population had occurred; only 73 tracts had a 5 percent or less black residents. The number of tracts with 5 to 10 percent black increased from 4 to 20 and the number with 10 to 25 percent increased from 4 to 7. The number of tracts with high concentrations of blacks, 25 percent or more, increased from 12 to 15, with the net growth occurring in tracts with 25 to 50 percent black.
The Hispanic residential pattern is marked by greater diversity. Only 39 tracts in 1980 had 5 percent or fewer Hispanics. The categories from 5 percent up to 75 percent all had significant shares of the total distribution, while only one tract had more than 75 percent Hispanic. The Hispanic population also saw a spreading into a larger geographic area, but the core of the sector saw a greater concentration of the hispanic population. Simultaneously, Hispanics "moved into suburban neighborhoods to the north, south, and southwest of the city of Denver."^
Although the extensions of the minority sectors indicates that the housing market in Denver and the adjoining suburbs is demonstrating some integration of housing markets, the large concentration of minority populations in the center of these sectors indicates that a large portion of the minority populations are still living in inferior housing markets,
whether for social or economic reasons.

Table 4
Distribution of Denver Census By Percent Minority, 1970-1980 Tracts
1970 1980
PERCENT BLACK Number of Tracts % of Total Number of Tracts % of Total
Less than 5.0% 97 84.3 73 63.5
5.1-10.0% 4 3.5 20 17.4
10.l%-25.0% 2 1.7 7 6.1
25.l%-50.0% 3 2.6 6 5.2
50.l%-75.0% 3 2.6 3 2.6
More than 75.0% 6 5.2 6 5.2
Total 115 100.0 115 100.0
less than 5.0% 39 33.9
5.1-10.0% 19 16.5
10.1-25.0% 22 19.1
25.1-50.0% 17 14.8
50.l%-75.0% 17 14.8
More than 75.0% 1 0.9
Total 115 100.0
Source: U.S. Bureau of the Census, 1970 Census of Population and
Housing, Census Tracts, PHC70-2-138, Denver SMSA and
U.S. Bureau of the Census, 1980 Census of Population and Housing, Census Tracts, PHC80-2-138, Denver-Boulder SMSA.

Variable Relationships in Minority Neighborhoods
Generally, a review of the Denver minority housing markets reveals that a great deal of progress has been made. But a number of factors indicates that these are still tenuous first steps towards full social integration. Although the minority population is spreading into a widening area, there are still large concentrations. Minority households still have to deal with educational and social differences which have resulted in lower job expectations and lower income. The lower income will remain a strong factor preventing fuller levels of integration. In this section, minority market performance is compared, showing the combined effects of lower income and higher concentrations of minority populations.
Detailed differences between the housing market in predominantly white areas and minority neighborhoods has been investigated by creating the same variable interrelationships described in the previous chapter, but filtering out the census tracts according to the percent black or the percent Hispanic (Table 5). The census tracts were filtered first of all for those which had a minority content which exceeded the city-wide median-- 2.1% for percent black and 10% for percent of Spanish origin. The second filtering includes the tracts in the top quartile for each variable-- 7.5% for percent black and 30% for percent of Spanish origin. Although the differences between black and Hispanic concentrations may seem large-- the percentages used for the Hispanic market are four times as large as for the black neighborhoods-- the black market actually may feel harsher discriminatory effects as their distinctive color may often be more noticeable than for Hispanics, who often can blend into white society. The median value for percentage black, however, is only 2.1%. This value seems too small to have

Table 5
Comparative Variable Responses In Minority Census Tracts
Dependent Variable: 1980 Median Household Income
Response for census tracts in which: Percent Black Percent Hispanic
Independent Variable Tracts >7.5% >10% >30%
Intercept Slope 25,511.0 -195.3 20,604.6 -147.0 21,538.2 -165.3 16,991.4 -104.2
Density Intercept Slope 24,000.0 -.414 20,443.2 -.327 19,427.2 -.314 16,486.2 -.249
Percent Owner-Occupied Intercept Slope 7,043.5 184.3 5,477.1 168.2 4,874.8 169.0 4,418.9 152.9
Price Intercept Slope 6,261.3 .162 7,301.0 .084 3,560.6 .202 1,484.5 .240
Percent Black Intercept Slope 19,253.7 -712.0 9,947.3 41.0 15,963.5 -996.6 13,589.5 -692.3
Percent Hispanic Intercept Slope 19,947.0 -184.2 14,452.4 -153.1 16,676.4 -111.1 16,375.9 -107.3

a significant effect upon the housing market, so the only filtering for black neighborhoods in this study uses the upper quartile value, 7.5%.
When the relationship between age and income is filtered for higher percentages of minority population, the intercept values decrease and the slope levels out. This suggests that the income in housing of similar age is lower in minority markets. Aging also has a lesser effect in minority households, indicating that the sensitivity to aging (and deteriorating conditions) is stronger in higher-income (white) households. Minority households either are less sensitive to aging or are less able to react to aging conditions by relocating. This last point is only common sense; earlier the relationship between the percent black and hispanic and income has been established, and these minority households are unable to bear the cost of moving to better (newer) housing units.
Because the slope for all census tracts is steeper, notch points exist where the income level for all census tracts will fall below the income levels for census tracts with higher concentrations of minority households. This occurs in older housing markets; for neighborhoods with higher concentrations of blacks, the notch point occurs at 100 years and for neighborhoods which are 30% or more Hispanic at 94 years. Thus, this does not define any market niche where affluent minority households are likely to aggregate, it simply illustrates how concentration of minority population has an adverse effect on income, resulting in a low-income housing market in which minority households are concentrated. The ability of the predominantly white middle class to move to newer housing units on the urban periphery has been discussed at length, but the white housing market generally has

controlled older housing as well, relinquishing control only on the oldest, inferior areas.
White households which live in older housing are either affluent enough to maintain or upgrade their properties (the former account for the inner edge of the "upper-income sector" and the latter are tagged "gentrifiers"), with the white market maintaining a higher level of income even where the housing ages up to almost 90 years. In the lowest income areas, the white households are less affluent households who are forced to live in these areas. Although these households are not necessarily poorer than the minority households in the same area, the low incomes prohibit them from finding better housing. This confirms studies which have indicated "that whites pay less for a house, ceteris paribus, in the black ghetto and areas of racial transition than elsewhere.
A similar set of relationships exist with density. Census tracts with higher percentages of minority population have lower intercept values and lesser slope gradients. Because of the relationship between location, age, and density, the parallel reaction of the market to age and density was expected. As with age, there are notch points along the regression lines where income is actually predicted to be higher in census tracts with higher percentages of minority residents. Again, the explanation lies in the innermost tracts with the lowest income white households which are not affluent enough to afford housing in lower-density, newer areas.
In 1980, the effects are greater upon the Hispanic neighborhoods than the black areas. In the relationship with age, the intercept values for the Hispanic neighborhoods are lower, especially for the highest concentration of Hispanic population. While all other intercepts exceed $20,000-- though

these still lag behind the $25,500 value for the entire Denver market-- the intercept for census tracts which contain 30% or more of Spanish origin falls to less than $17,000. In the density equations, the intercept values are lower for both Hispanic equations, while the equation for all census tracts with greater concentrations of black population has an intercept which is only $850 less than for the entire market. Additionally, the slopes are more level in the equations with the higher concentration of hispanic population.
The relationship between the share of housing units which are owner occupied and income reveals that in minority markets even a high percentage of owner-occupied units does not result in incomes as high as in the total market. All the market segments have similar slopes, but this does not indicate that home ownership is readily available for minority households. Already, the relationships between income and the percent owner-occupied has been discussed. In addition, a higher concentration of minority population in a census tract has been shown to cause a lower median income. Therefore, the minority markets will have lower percentages of owner-occupied units. Thus, the similar slopes in Figure 25 are misleading; the maximum percent owner-occupied in the minority markets is lower than in white markets. Although these equations indicate that ownership has the same beneficial effects for minority neighborhoods (stability, higher values, higher income, etc.) as it has for white neighborhoods, economic reality in 1980 states that higher owner occupancy is not yet attainable in minority markets.
The combined effects of lower income and higher minority concentration upon price provides a basic relationship which starts to quantify the actual cost of discrimination in the housing market. This set of equations is

deceptive; it indicates that in the Hispanic markets, income will be higher than in the city as a whole when the average sales price exceeds $60,000. The Hispanic and black markets, however, are lower-priced markets, so generally similar-priced markets result in lower incomes for minority areas. In simpler terms, households in minority markets are paying a larger share of their incomes for housing. Only in the lowest-priced markets (where the average 1980 sales price would be less than $14,000) would the income in the black neighborhoods exceed the expected income for the entire market.
When the percentage of the population which is black or Hispanic is filtered, the relative exclusivity of each minority market is highlighted. This indicates a primary difficulty in the process of integrating the housing market. Each minority group originally congregated in a specific area for social and cultural well-being. Across time, the segregated nature of each community may diminish the value of that market in the eyes of other minority groups, as well as the white middle-class. This possibly explains why the growth of each minority "sector" has occurred radially outward and to the south as a way of providing more distance from the other minority market. Perhaps the former occupancy by a minority community has "tainted" that market. Certainly, the most attractive homes are newer ones and these are the middle-class, predominantly white markets.
Within the hispanic markets, small increases in black population result in significant decreases in income. With an increase of only one in the percent black, income declines by practically $1,000 in the larger Hispanic market. This represents a sharper decline than in the entire Denver market. Two possible conclusions emerge. First, a combination of a higher

concentration of Hispanics and blacks effectively creates a significantly larger total concentration of the minority population, and this higher minority percentage accounts for the lower income levels. But a second, more disturbing conclusion must also be considered: perhaps black households experience more discrimination from Hispanics than from whites.
The inverse does not seem to be true. The effect on income of the percent hispanic has a similar slope in the total Denver market and in neighborhoods where concentrations of black population are highest. A one percent increase in the concentration of hispanics has resulted in a decline of less than $200 in the median household income.
Overall, the market dynamics of the different variables is summed in the multiple regression equation (Table 6). The equation constructed in the previous chapter was reconstructed, again filtering out the lower percentages of hispanic and black population. To create the strongest equations, more individual cases need to be included. Thus, the hispanic equation includes all tracts in which the hispanic population equalled or exceeded 10% of the total population, the median value. The black equation includes only the cases where the percent black exceeded the upper quartile value, 7.5%.
The comparative regression equations indicate that for the most part the response of income to the independent variables in the minority census tracts is similar to the response for the entire market. Neither minority market included the variable IHISP, which could affect an accurate comparison of the equations. The minority neighborhoods tend to show a sharper response to the percent owner occupied, highlighting the lower values for this variable in minority neighborhoods: if these neighborhoods had values

Table 6
Compared Multiple Regression Coefficients
Dependent Variable: 1980 Median Household Income
Independent Variable
Multiple Regression Coefficients for:
All Percent Black Percent Hispanic
Tracts >7.5% >10%
Percent Owner-Occupied 117.7
Price .105
Age -80.0
IHISP (1/Percent Hispanic) 63.7
Constant 6055.8
Standard Error 2626.6
Adjusted . 842
158.2 124.2
.106 .121
-33.7 -55.1
Not in Equation Not in Equation
2007.6 4052.5
1221.0 1817.7
.947 .935

similar to white neighborhoods, the result would be comparatively higher incomes for minority residents. Hispanic neighborhoods, but not black, show a significantly flatter relationship with age and price. This indicates that Hispanic households are less sensitive, perhaps, to the value of their housing units, affirming that Hispanics tend to remain in the same locations regardless of income level. Black neighborhoods, on the other hand, have an identical relationship between price and income with the entire set of census tracts. Both black and Hispanic neighborhoods are less sensitive to age than white neighborhoods, which again ties higher concentrations of minority population to lower income, lower percentages of owner occupancy, and lower prices. Generally, the lesser response to age results from lower income, which restricts the housing choice of minority households .
Overall, the comparative regression coefficients highlight some of the same conclusions indicated earlier. As expected, the presence of higher concentrations of minority population in a census tract has been connected with older, denser housing, lower levels of owner-occupancy, and lower housing prices, for which households must pay a higher percentage of their incomes. These relationships indicate a level of discrimination is present in the market, yet this is still a difficult, spurious relationship. The discrimination may be economically rooted, with the difference in income level accounting for the level of segregation. Yet a history of educational and social discrimination based upon ethnic status certainly has a role in the lower economic class in which a large share of the minority populations
still find themselves.

These equations indicate that some integration of middle-income neighborhoods is occurring. Both the black and Hispanic populations have gradually spread into a greater geographic area. Suburbanization opened up a considerable supply of homes and subsequently set the preconditions for expansion of the minority markets southward. Black patterns have tended to echo white market patterns, showing similar relationships between income level and other market variables, indicating that the key constraint for blacks has been exclusion from the majority of non-minority neighborhoods.
In black neighborhoods, the higher incomes are found in the neighborhoods which have higher percentages black, indicating that the poorer black residences are located in transitional areas. The black population seems, therefore, to be segregated by income levels, even within the areas of highest concentrations of the black population. More affluent black households have accepted a "separate but equal" succession of neighborhoods throughout the northeast part of the city.
Hispanics, on the other hand, show some different patterns, especially in the lessened response of income to the proportion of residents which are Hispanic. Unlike blacks, the Hispanic pattern has followed a pattern of invasion and concentration in a widening number of neighborhoods. The lessened slope in the equation relationg percent Hispanic to income (Table 5) indicates that the Hispanic population has not segregated into neighborhoods based upon income level.

Neighborhood Transition
The term "tipping" entails a transition from a white neighborhood to significant minority concentrations. It was noted earlier that researchers are in general agreement that "tipping" is more a concept than an actual process. Instead, the patterns of integration should show limited intrusion into previously white markets, although the Hispanic pattern of concentration in neighborhoods has been noted.
Because of the changing definition of "Hispanic" used by the U.S. Census, comparison of the larger minority groups in Denver is difficult. In this chapter, a certain amount of Hispanic migration will be inferred, but the actual transition of a market is best illustrated through a longitudinal analysis of the influx of black households into previously white neighborhoods. Within Denver, the lower percentages of blacks in most census tracts has already been noted. There are few tracts which gained significant percentages of blacks from the 1970 to 1980 census. For this section, the process of integration will be studied in census tracts which contained less than 5% blacks in 1970 and in which the percentage increased by more than 5% between 1970 and 1980.
Research has indicated that residents are more sensitive to changes in surrounding neighborhoods than in their own, as they are watchful for the encroachment of minority households.

Households reveal themselves to be sensitive to economic status of a neighborhood's residents as measured by mean or median income.
Households are willing to pay a premium for sites at a distance from concentration of nonwhite population and inferentially prefer neighborhoods with a low probability of socioeconomic change.
Households reveal themselves to be somewhat more sensitive to the racial composition of adjoining neighborhoods than the racial composition of the neighborhood itself. Furthermore, they are more sensitive to income than to racial composition.
There are substantial premiums placed on units in 'desirable' locations as opposed to low-income, low public service neighborhoods.^
The white residents, to protect themselves from the intrusion of minority households and in fear of falling property values, start to leave the neighborhood.
As property becomes vacant through normal turnover, its owner have no alternative to selling or renting to blacks, and the neighborhood gradually becomes more and more black in occupancy.^2
Theoretically, neighborhoods which have higher percentages of owner-occupancy are more affluent and have higher housing prices. Housing costs are higher and, ceteris paribus, will be less likely to undergo a rapid transition. (This stability is fundamental to the structure of the Denver market: two higher-income sectors stretching from the older part of the city have maintained stability even as surrounding neighborhoods have undergone series of changes.) Thus, the percent owner-occupied would be a key variable explaining where integration will occur.
Other studies indicate this same conclusion, but various researchers disagree on the relationship with other variables, especially income.
Cronin indicates that

Compared to nonminority households, minority households search in and move to neighborhoods with: a lower mean household income; a higher proportion of households below the poverty line; a higher proportion of black households; [and] a lower proportion of occupied units that are owner-occupied.^
McDonald, in contrast, cites studies which have shown, "The probability of
tipping was increased by increase in the median income in the tract and the
percentage of houses of sound condition.Although this seems to defy
normal market behavior, it is essential to remember that in the black
housing market the higher median income is linked to higher percentages of
rental occupancy. The expansion of the black housing market
did tend to be into rental housing. This is probably the expected result because higher income white renters are probably very mobile Furthermore, the blacks who were the first to move into the white areas were likely to be relatively mobile and high-income persons.^
This latter characteristic is important; the minority market contains a variety of income groups and a large share of the relocating households belong to the higher groups, whereas the lower-income minority households aggregate in greater concentrations in the older centers of the minority "sectors." Downs has indicated the same conclusion, that the initial black households moving into a previously nonminority neighborhood are the more affluent households of the black population. Once the transition from white to black has begun,
the initially depressed prices of housing tend to rise. Not only are the incomes of most blacks moving in higher than the incomes of most white residents, but white property owners can charge blacks more than they could have charged white renters and buyers.'
Black Integration Patterns in Denver Neighborhoods
From 1970 to 1980, 15 census tracts in Denver saw the initial stages of integration of black residents (using the definition described at the

beginning of this chapter). Most of these tracts were adjacent to the existing black sector, located along East Colfax (Map 10). The black sector also spread to the west, as the North Capitol Hill and Capitol Hill neighborhoods began to integrate. In addition to the eastward movement along major arterials, some southward movement of the black market is also evident, using Holly St. and Quebec St. as conduits. Census tract 50, located between these two north-south arterials, increased from 1% black in 1970 to 6% in 1980. The southeast market within Denver also experienced an influx of black households. Tracts 68.01 and 68.02 also increased in percent black, resulting in 6% black in 68.01 and 7% in 68.02. While these are still relatively small percentages, less than the city-wide averages, these actions represent the first steps of a major southward migration of the black population.
Low levels of owner occupancy has been associated with the transitional markets. Within the city of Denver, the average percent owner occupied in 1980 was 50.6%. All tracts except one had percents owner occupied lower than this average (Table 7). The average within these census tracts is 24.7%. Low percentages of owner occupancy are generally the rule for transitional neighborhoods, but within the established black sector, owner occupancy exceeds 80% in a number of tracts.
McDonald's research notwithstanding, the median income in the selected census tracts declined. Throughout Denver, the average value for the median income declined from $18,410 to $16,375 (as could have been predicted from the relationship between age and income: for each additional year in the average age, income will decline, theoretically, by roughly $200). Within these census tracts, the median income declined by more than $3,000

Map 10
City of Denver,
Census Tracts Undergoing Initial Integration, 1970 to 1980
% Black 1980> 10 Integrating Tracts

Table 7
Market Characteristics in Neighborhoods Experiencing Black Integration
% Owner- Median Sales , 1976-80 Sales , 1980-84
Census Tract Occupied, 1980 Income, 1980 % Sold % with Multi-Sales % Sold % with Multi-Sales
26.01 0% $6,061 46.2% 0.0% 27.2% 0.0%
26.02 6% $6,758 46.1% 14.4% 31.0% 8.6%
27.01 3% $8,666 58.0% 25.0% 61.6% 15.4%
27.02 12% $13,242 45.4 % 17.8% 37.2% 6.2%
27.03 8% $9,803 42.0% 19.7% 35.3% 8.6%
32.01 16% $11,901 46.0% 13.7% 30.9% 4.1%
37.01 25% $11,779 34.3% 8.9% 29.2% 5.6%
37.02 20% $12,557 40.6% 12.9% 26.4% 5.5%
37.03 29% $13,051 36.7% 12.3% 27.4 % 4.3%
43.01 28% $13,896 30.6% 11.1% 23.9% 4.7%
44.01 40% $14,255 40.0% 14.5% 24.2% 4.4%
50.00 51% $19,544 28.1% 6.5% 16.4% 1.7 %
54.01 30% $10,795 30.4% 8.7% 29.6% 3.7 %
68.01 53% $24,006 52.2% 20.6% 21.0% 4.3%
68.02 39% $23,502 43.1% 9.1% 18.6% 2.6%
City-wide Avg. 51% $16,375 33.9% 9.9% 24.4 % 6.1%

except in 68.02, which only declined by $1,497. The decline in income in these tracts averaged $4,954, nearly 150% more than the city-wide average.
Generally, the integrated markets had all been predominantly white prior to 1970. Only the Capitol Hill (27.01, 27.02, and 27.03), North Capitol Hill (26.01 and 26.02), and South Platte (54.01) neighborhoods have large Hispanic populations. The first two are the western-most end of the black residential sector and represent an area of interaction between the two minority populations. In addition, the North Capitol Hill neighborhood sits on the edge of the encroaching Central Business District and has lost many of its housing units. This neighborhood has lost significant population, although efforts are currently underway to reverse this trend.
The South Platte neighborhood is an anomaly in the Denver market, as it has established a black market totally separated from the northeast and east part of Denver. Its large minority population may have arisen from the large amount of industrial and railroad-related land uses in the vicinity, which have made the area an inferior location for residential land.
This neighborhood, also, has encroached upon the hispanic sector.
The introduction of minority households generally is considered to disrupt the stability of the neighborhood; once a few minority households move in, panic ensues, property values decline, and the existing households seek new housing. In Denver, the census tracts which started undergoing a process of integration certainly showed comparatively high amounts of property turnover. During the period from 1976 to 1980, an average of 41.3% of all single family homes in integrating census tracts were sold at least once, compared with 33.3% for the entire city. A higher percentage of properties also experienced repeat sales during the same period. In the

integrating tracts, 13.0% of the single family homes were sold two or more times, compared with a city-wide average of 9.9%. Two factors should be considered in concert with these high percentages of sales. First, this was a period of great social upheaval in the City of Denver, with the introduction of a mandatory busing program which, as other researchers have pointed out, may not be the primary cause of suburbanization, but it a contributing factor. Second, the high inflation experienced in the late 1970s led to a higher number of rapid home sales. The entire market experienced instability during this period. Following 1980, interest rates climbed rapidly, causing a significant decline in the number of home sales.
The concept of "tipping" would indicate that the instability in the neighborhood would continue unabated until a transition to a minority market had been completed. In the selected tracts in Denver, however, the instability of the 1970s has not continued. During the 1980-1984 period, the number of homes which were sold in these tracts averaged 29.3%, compared with 24.4% for the entire market. However, five of the 15 tracts actually had lower percentages than the city-wide figure. Possibly, as the neighborhood stabilizes at a new level of integration, income will stabilize as well, removing the stigma of "declining property values" and lower quality of life which had been thought to accompany the influx of minority households .
Additionally, the tipped tracts had comparatively few repeat sales. On average, 5.3% of the homes in these tracts were sold more than once from 1980 to 1984, compared with 6.1% for the entire market. The contention that integration introduces a period of instability does not seem to be true in Denver. Only four tracts had higher percentages of repeat sales than the

city-wide averages, and all four of these are in North Capitol Hill or Capitol Hill, areas which are undergoing a number of different transitional processes, including pressure from the CBD to convert to non-residential uses.
Potential for Further Black Integration
As previous research predicted, lower levels of owner occupancy and adjacency were key indicators of areas in which integration to higher percentages black occupancy would most likely occur. Within Denver, these were the most reliable indicators, especially in the census tracts along East Colfax. Some leap-frogging of minority households to the southeast portion of the city defied the adjacency requirement, but these neighborhoods have relatively high amounts of rental units. They are also situated along major arterial streets, Hampden Avenue and Havana Street, and the southward migration of the black population along Holly and Quebec Streets was indicated earlier. The presence of these other arterial streets make these areas more accessible to the black population.
In addition, an initial period of increased home sales may indicate a transitional period in a census tract which is connected with integration of the market. The percentage level of sales should be less than during the 1976 to 1980 period, but a higher than average percentage of sales or multiple sales from 1980 to 1984 could indicate that a neighborhood is ripe for integration.
Census tracts adjacent to the established black sector and throughout southeast Denver (Table 8) were analyzed for integration potential. In the initial step, tracts were graphed according to percent owner-occupied and median income in 1980 (Graph 9). The census tracts in which blacks had

Table 8
Market Characteristics in Census Tracts With Potential for Black Integration
% Owner- Median Sales , 1976 i-80 Sales , 1980 i-84
Census Occupied, Income, % with % with
Tract 1980 1980 % Sold Multi- Sales % Sold Multi- Sales
28.01 28% $13,789 32.8% 11.0% 28.8% 5.5%
28.02 16% $10,405 40.8% 12.0% 33.3% 6.8%
28.03 16% $13,244 30.1% 6.7% 29.4 % 5.7%
29.01 42% $12,901 36.1% 12.6% 26.5% 6.7%
29.02 65% $17,422 28.6% 8.9% 25.5% 4.6 %
30.01 62% $14,440 31.1% 9.4% 25.9% 5.1%
30.02 59% $16,215 29.2% 9.2% 29.4 % 3.9%
30.03 43% $12,810 35.6% 11.3% 23.3% 3.8%
30.04 38% $12,506 32.3% 9.1% 24.1% 4.1%
30.05 94% $26,020 24.1% 6.2% 16.2% 3.0%
32.02 15% $15,568 40.8% 13.6% 26.5% 6.0%
32.03 71% $26,397 29.0% 5.9% 24.5% 4.7%
33.00 77% $23,222 29.7 % 8.1% 24.5 % 3.8%
34.00 77% $20,640 29.3% 8.1% 22.8% 3.9%
35.00 60% $12,578 30.9% 9.0% 19.5% 4.0%
38.00 37% $14,009 27.5% 4.3% 24.1% 4.1%
39.01 79% $28,492 25.3% 5.6% 18.9% 3.1%
39.02 87% $18,725 27.5% 5.3% 24.3 % 3.8%
40.01 47% $17,618 17.2% 7.8% 21.1% 4.4 %
40.02 92% $34,442 25.6% 5.1% 16.2 % 1.9%
40.03 85% $21,364 37.8% 12.2% 24.8% 4.8%
40.04 87% $37,868 21.5% 5.1% 16.4% 2.2%
42.01 73% $20,690 34.6 % 9.6% 23.2% 4.0%
43.02 71% $20,774 30.2% 9.8% 25.0% 2.6%
43.0£ 93% $35,873 22.4 % 4.8% 17.2% 3.1%
43.04 68% $19,866 29.5% 9.3% 23.1% 4.9%
43.05 57% $26,868 24.3 % 5.6% 16.8% 2.1%
49.00 34% $15,294 31.6% 7.6 % 20.5% 3.3%
51.01 13% $14,875 29.5% 7.8% 19.8% 3.4%
51.02 81% $23,348 32.0% 8.5% 18.1% 2.9%
52.00 87% $27,965 33.3% 7.4% 19.4% 3.3%
53.00 61% $20,278 28.2% 7.2% 16.1% 2.8%
68.03 84% $27,100 35.1% 7.5% 18.9% 3.2%
68.04 65% $33,605 43.0% 9.3% 29.6% 5.8%
69.01 30% $17,897 24.6 % 3.7% 16.4% 3.0%
69.02 41% $26,113 26.3 % 6.7% 17.6% 2.4%
City-wide Avg. 51% $16,375 33.3% 9.9% 24.4 % 6.1%

Graph 9
Southeast Denver Census Tracts
by Percent Owner-Occupied and Median Household Income, 1980
% Owner
City-Wide Mean: $16,375

started to integrate were included on this graph to show their relationship with the other tracts in southeast Denver and the tracts were then grouped loosely according to their position on the graph. Subsequently, the tracts were graphed according to the percentage of units sold between 1980 and 1984 and the percent with multiple sales (Graph 10). Surprisingly, none of the tracts exceeded the city-wide average for multiple sales and only four exceeded the average for percent sold.
Although no clear relationships emerge between groups and either variable on Graph 10, theoretically those tracts with the higher values for either variable are more likely to integrate during the 1980s. All the tracts exceeding the city-wide average for the percent of homes which were sold between 1980 and 1984 were considered to have more potential and were advanced to the next higher group, resulting in the groups shown in Table 9. When mapped (Map 11), they indicate that southeast Denver will most likely continue to integrate. The greatest potential exists in the older homes in south central Denver, bridging the gap between the existing black area and the South Platte neighborhood, and in tracts along Colorado Blvd.
The areas which show the least potential are, predictably, higher income areas, especially Belcaro, Virginia Village, and Hilltop. All of these neighborhoods are within upper-income "sectors." Even as surrounding neighborhoods see further integration, these upper-income areas have the economic power to continue to exclude minority households. The integration that is occurring, however, is part of an on-going process. As black households have slowly closed the income disparity with white households, they have been able to integrate with more ease. Eventually, black households

Graph 10
Southeast Denver Census Tracts
by Percent Sold and Percent with Multi-Sales, 1980-1984
Percent Sold
City-Wide Mean: 6.1%
( ): Estimated Level of Potential from Graph 9.

Table 9
Southeast Denver Census Tracts
by Potential for Black Integration
High Potential Medium Potential
28.01 33.00
28.02 35.00
28.03 39.02
29.01 40.01
29.02 40.03
30.01 43.04
30.02 43.05
30.03 53.00
30.04 69.01
32.02 69.02
40.02 40.04
51.02 52.00

Map 11
Southeast Denver Census Tracts
by Potential for Black Integration

could close the income gap completely, allowing them the economic power to locate in all neighborhoods.
Contiguity has also been identified as a key variable resulting in the introduction of minority households into a neighborhood. Census tract 42.01 has been estimated to have little potential for integration. Its location, however, adjacent to a large area with high concentrations of blacks, may increase the potential. The high income and owner occupancy may prevent this.
Contiguity to integrated areas also should lead to more integration in the census tracts in central Denver, located between Colfax and Alameda.
The southeast corner of Denver, in which some integration has occurred, could lead to more location of black households in these neighborhoods and in the neighborhoods north of Hampden.
Potential for Hispanic Integration
The Hispanic population, being larger than the black population, had spread over a wider geographic area. All neighborhoods in southwest Denver (Table 10) had accommodated a population of at least 5% Hispanic in 1980 and all tracts but three had exceeded 10%. Theoretically, higher values would need to be present to qualify as a "tipped" neighborhood.
Without comparable figures for 1970, the actual increases in Hispanic integration cannot be traced. Within the southwest portion of Denver, however, the census tracts with percentages exceeding 19%-- the city-wide average-- are considered in this report to have already undergone significant integration and those with values lower than the city-wide average show potential for integration. In southeast Denver, the relationship between

Table 10
Market Characteristics in Census Tracts
With Potential for Hispanic Integration
% Owner- Median Percent Sales , 1976 i-80 Sales , 1980 '-84
Census Occupied, Income, Hispanic, % with % with
Tract 1980 1980 1980 % Sold Multi- Sales % Sold Multi- Sales
13.01 75% $17,261 30.4% 28.2% 9.3 % 18.2% 3.9%
13.02 72% $16,718 33.1% 31.3% 9.7% 19.2% 2.8%
14.01 60% $16,262 28.1% 28.3% 7.9% 17.8% 3.1%
14.02 69% $16,610 33.3% 39.3% 14.5% 21.1% 6.0%
14.03 55% $12,260 25.1% 33.1% 12.5 % 21.4% 4.9%
45.01 54% $14,545 29.4% 36.8% 13.8% 22.6% 5.0%
45.02 50% $11,193 53.0 % 35.0% 14.7% 25.5% 6.6%
46.01 88% $18,409 18.7% 33.0% 12.6% 19.8% 4.0%
46.02 57% $16,276 27.0 % 34.4 % 10.4% 21.0% 5.4%
46.03 84% $19,295 15.8% 37.7% 14.8% 24.3 % 5.2%
47.00 70% $20,694 13.5% 24.2 % 6.1% 15.3% 2.0%
48.01 96% $26,875 14.0% 22.6% 5.6% 14.1% 1.8%
48.02 65% $19,314 15.2% 29.0% 9.3% 16.4% 2.1%
54.01 30% $10,795 58.7% 30.4% 8.7% 29.6% 3.7 %
54.02 38% $12,068 35.4% 33.8% 9.2% 20.6% 4.5%
55.02 33% $26,838 6.9% 29.3% 3.6% 17.7% 1.8%
55.03 87% $26,692 9.3% 42.8 % 13.7% 19.9% 3.0%
119.02 51% $24,666 7.4% 30.9% 7.4% 19.9% 4.4 %
119.03 75% $28,393 11.0% 33.6% 7.3 % 16.9% 2.8%
120.01 89% $27,022 4.7% 30.3% 8.8% 17.1% 1.9 %
City-wide Avg.51% $16,375 19.4% 33.3% 9.9% 24.4% 6.1%

income and percent sold did not easily reveal patterns of integration. The southwest tracts, however, reveal some clearer patterns.
All tracts in which the percent Hispanic exceeded 19%, the median incomes fell below $17,500, while the tracts with lower percentages of Hispanics all had incomes exceeding $18,000. Owner occupancy was not strongly related to income; all tracts in the southwest area except four had percentages of owner occupied which exceeded the city-wide mean of 51%. Of the four which fell below the mean, one tract, 55.02, had a median income of $26,838.
Unlike the southeast area, southwest Denver's tracts did not undergo a wild flurry of sales and resales during the late 1970s. In the southwest tracts an average of 31.6% of all homes were sold between 1976 and 1980, compared with 41.3% in southeast Denver and 33.3% for the entire city. Two conclusions can be derived from this. First, the area probably contained a fairly high percentage of Hispanics prior to this time, so "white flight" may not have had the same effect it did in the predominantly white southeast Denver area. Second, the higher percentage of Hispanics had occurred because this area is generally older than southeast Denver, so the market attraction would not be as high. The newer census tracts in the area,
55.02, 119.02, 119.03, and 120.01, have not yet felt the intrusion of the minority population; these are still a higher income area, with an average income of $26,729 in 1980, compared with an average of $16,880 for all other southwest tracts.
Like the southeast area, home sales have declined during the 1980s. On average, 19.8% of the homes in the area were sold between 1980 and 1984, compared with a city-wide figure of 24.4%. Only 3.7% of the homes had

repeat sales, while throughout the city the average was 6.1%. Only three southwest tracts had more than 24.4% sales and none exceeded the city figure for repeat sales.
To determine the future pattern of Hispanic location, the analysis performed for black population in southeast Denver was repeated. First, the percent owner-occupied and median incomes were graphed (Graph 11) and the results were used to group the census tracts. Tracts with incomes lower than $17,500 and owner occupancy below 60% were considered to have the highest potential for additional Hispanic location. Tracts with one of these two characteristics were listed as medium potential and those with neither as low potential. A second graphing, based upon percent sold and percent with repeat sales (Graph 12), was used to adjust the assessed potential of each tract. Those tracts grouped to the right of the line had a higher percent of sales and/or higher percent of multiple sales. These tracts were moved upward from their previous grouping, resulting in the final grouping of tracts according to potential for further Hispanic integration (Table 11 and Map 12).
Unlike the analysis for black expansion, the analysis indicates that most further expansion of Hispanic households will occur in tracts which already contain higher percentages of Hispanic population. Among the tracts with lower than average percent Hispanic, only 119.02, which has a relatively low median income, and 55.02, which has a higher percentage of rental units, will see significant increases of Hispanic population, confirming Downs' conclusion that Hispanic households tend to remain in the same area even as their incomes rise.^

Graph 11
Southwest Denver Census Tracts
by Percent Owner-Occupied and Median Household Income, 1980
% Owner Occupied

Graph 12
Southwest Denver Census Tracts by Percent Sold and Percent with Multi-Sales, 1980-1984
with Multi-Sales
City-Wide Mean: 6.1%
( ):
Estimated Level of Potential from Graph 11.

Table 11
Southwest Denver Census Tracts
by Potential for Hispanic Integration
High Potential Medium Potential Low Potential
Census Tract Percent Hispanic, 1980 Census Tract Percent Hispanic, 1980 Census Tract Percent Hispanic, 1980
13.01 30.4% 13.02 33.1% 47.00 13.5%
14.01 28.1% 46.01 18.7% 48.01 14.0%
14.02 33.3% 46.03 15.8% 48.02 15.2%
14.03 25.1% 55.02 6.9% 55.03 9.3%
45.01 29.4% 119.03 11.0%
46.02 27.0% 120.01 4.7%
54.01 58.7 %
54.02 35.4%
119.02 7.4%

Map 12
Southwest Denver Census Tracts
by Potential for Hispanic Integration
___6th Ave.
Established Hispanic Area High Potential Medium Potential
Low Potential

The Extent of Integration
The dual processes of suburbanization and integration have brought higher-quality housing into reach for a larger number of minority households, yet the caveat attached to this by Wilson, Pettigrew, Frey, et al. is that the minority middle-class, which is able to take advantage of the increased opportunities for quality housing only accounts for a small portion of the total minority population. The loss of the minority middle- class in central city neighborhoods also results in the greater segregation of the housing market, minority and non-minority, by income class. This latter effect is expected to be stronger in the black market than for Hispanics; one unique feature of the Hispanic residential pattern is concentration with less regard to income level.
The black population started from a greater level of concentration in 1970 (Table 12), when 89.9 percent of the black population was concentrated within census tracts within which 10 percent or more of the total population was black. By 1980, this share had lessened to 79.5 percent, which, although a lower share, indicates that integration has been accessible to a relatively small share of the black population. The established black sector in the northeast part of the central city in 1980, while declining in total population, saw an increase in the number of black residents. This area in 1980 contained 45,901 black residents, 59 percent of the total black population in the entire SMSA.
Those census tracts which had undergone initial integration during the 1970s contained 4,676 black residents in 1980, 8.5 percent of the total population. This accounted for only 8.1 percent of the blacks in the city, and 6 percent of the metropolitan total. Thus, integration only occurred

Table 12
Denver-Boulder SMSA, Concentration of the Black Population,
A. 1970
Total Black Percent
Population Population Black
City of Denver Tracts: % Black >10 64,542 41,775 64.7%
Percent of City Total 12.5% 88.8%
Percent of SMSA Total 5.2% 83.2%
% Black <10 and >5 19,459 1,682 8.6%
Percent of City Total 3.8% 3.6%
Percent of SMSA Total 1.6% 3.4%
City of Denver Total 514,678 47,011 9.1%
SMSA Total 1,235,936 50,188 4.1%
B. 1980
Total Black Percent
Population Population Black
City of Denver Tracts: % Black>10 83,238 45,901 55.1%
Percent of City Total 17.7% 79.5 %
Percent of SMSA Total 5.1% 59.0%
% Black<10 and >5 55,259 4,676 8.5%
Percent of City Total 11.2% 7.9%
Percent of SMSA Total 3.4 % 6.0%
City of Denver Total 492,365 59,252 12.0%
SMSA Total 1,618,461 77,786 4.8%
Source: DRCOG, Profiles of 1970-80 Socio- Economic Change

for a limited number of blacks during the 1970s. Only 31,885 black residents in the SMSA lived outside of the established black sector in 1980, less than half of the total black population.
In 1970 and 1980 both, census tracts in which the percent black equalled or exceeded 10, the majority of households had incomes below $15,000 (Table 13). Although the share of households in this range declined, from 63.5 to 59.9 percent, the decline was slight. The total number of households in tracts with higher concentrations of blacks increased by almost 16,000 and 55 percent of these had incomes below $15,000. The City of Denver as a whole experienced net losses in most income ranges. The number of households increased by 26,512 from 1970 to 1980 and 16,674 of these, 63 percent, were in the lowest income bracket. The majority of this net growth, 8,755 households, occurred in tracts with black populations of 10 percent or more.
The Hispanic population, spread over a greater geographic area, has not had the extreme concentration found in the black community, but integration of the Hispanic community, like the black, has occurred slowly. In 1980, Denver contained 57,067 Hispanic persons and 64.1 percent of these were concentrated in census tracts where the percent Hispanic exceeded 30% (Table 14). An additional 24.6 percent of the Hispanic population resided in tracts where Hispanics accounted for 10 to 30 percent of the population. Almost half of the metropolitan Hispanic population resided outside of the central city, but tracts in Denver where Hispanics account for 10 percent or more of the population still contain almost half of the total Hispanic population in the SMSA.

Table 13
Denver-Boulder SMSA,
Household Income Distribution in Predominantly Black Neighborhoods, 1970-80
1970 Households by 1969
_______City of Denver
% Black>10
5<% Black<10
Household Income Number % of Total Number % of Total Number % of Total Number % of Total
Less than $15,000 13,864 63.5% 5,649 48.2% 85,963 46.5% 146,010 37.0%
$15,000-$29,999 6,198 28.4% 3,635 31.0% 66,100 35.7% 162,056 41.0%
$30,000-$49,999 1,513 6.9% 1,664 14.2% 25,120 13.6% 67,853 17.2%
$50,000-$74,999 144 0.7% 369 3.2% 3,794 2.1% 10,406 2.6%
More than $75,000 114 0.5% 391 3.3% 4,077 2.2% 8,573 2.2%
Total 21,833 100.0% 11,708 100.0% 185,054 100.0% 395,078 100.0%
B. 1980 Households by 1969 Incomes
City of Denver SMSA
% Black>10 5<% Black<10 Total
% of % of % of % of
Household Income Number Total Number Total Number Total Number Total
Less than $15,000 22,619 59.9% 20,893 48.0% 102,637 48.5% 219,207 36.0%
$15,000-$29,999 11,513 30.5% 13,522 31.0% 72,065 34.0% 232,803 38.2%
$30,000-$49,999 3,148 8.3% 6,244 14.3% 27,578 13.0% 118,935 19.5%
$50,000-$74,999 364 1.0% 1,897 4.4% 6,460 3.0% 26,030 4.3%
More than $75,000 136 0.4% 1,001 2.3% 3,264 1.6% 11,711 1.6%
Total 37,780 100.0% 43,557 100.0% 211,566 100.0% 608,686 100.0%
Source: DRCOG, Profiles of 1970-80 Socio-Economic Change

Table 14
Denver-Boulder SMSA,
Concentration of the Hispanic Population, 1980
Total Population Hispanic Population Percent Hispanic
City of Denver Tracts:
% Hispanic >30 111,587 57,067 51.1%
Percent of City Total Percent of SMSA Total 23.8% 6.9% 64.1% 32.9%
% Hispanic <30 and >10 120,215 21,953 18.3%
Percent of City Total Percent of SMSA Total 25.6% 7.4% 24.6% 12.6%
City of Denver Total 469,385 89,078 19.0%
SMSA Total 1,618,461 173,687 10.7%
Source: DRCOG, Profiles of 1970-80 Socio-Economic Change

The tracts in which Hispanics accounted for 30 percent or more of the total population saw a greater concentration of low-income households, compared with the black census tracts (Table 15). Within these Hispanic areas, 62.7% of the total households had incomes below $15,000, significantly higher than the city-wide share of 48.5%.
Thus, the black and Hispanic populations, while gaining some access to other central-city and suburban neighborhoods, are still concentrated in the established "sectors" in the City of Denver; the majority of the black population in the SMSA resides in northeast Denver and almost half of the Hispanic population in the metropolitan area is located in northwest and west Denver, with a large share in lower-income neighborhoods surrounding the CBD.
Quantitative Measures of Integration
Two indices which quantify the extent of integration in the central city are the dissimilarity index'' and Schnare's segregation index. ^ The former index measures the difference between the proportion of two groups in the subareas (census tracts) of the city. Higher scores indicate higher levels of dissimilarity. The segregation index, which measures exposure of two subgroups of the population to each other, also indicates higher scores for higher levels of segregation.
When the 1970 and 1980 indices for white/black integration are measured (Table 16), some integration of the city is evident. The dissimilarity index declined from .850 in 1970 to .707 in 1980, indicating a small amount of equalization of the proportion black throughout the city in 1980. The segregation index fell from .619 to .473, indicating greater levels of exposure of the white population to the black.

The larger Hispanic population, however, indicated lower levels of segregation in 1980. The dissimilarity index for Hispanics was .537 and the segregation index was only .264. Without comparable measures for 1970, a level of desegregation to compare with the black population is not possible. The stronger community ties in the Hispanic community and the prevalent pattern of succession from white to Hispanic which has been demonstrated in west Denver indicate that less desegregation has occurred in this minority population. James indicated that
exposure rates of Hispanics shifted very little during the decade. Little change in the overall segregation of Hispanics occurred during the 1970s in Denver.^
Earlier in this paper, some discrimination between the Hispanic and black populations was indicated by the statistical correlations in minority neighborhoods. The indices bear out this conclusion. In 1980, the dissimilarity index was higher between the black and Hispanic populations, .726, than between either minority group and the white population. The segregation index showed a higher score between the black and Hispanic populations, .307, than between whites and Hispanic but was lower than the white/black
index in 1980.

Table 15
DenverBoulder SMSA,
Household Income Distribution in Predominantly Hispanic Neighborhoods, 1980
City of Denver % Hispanic > 30
% of
Household Income Number Total
Less than $15,000 24,733 62.7%
$15,000-$29,999 11,760 29.8%
$30,000-$49,999 2,573 6.5%
$50,000-$74,999 306 0.8%
More than $75,000 64 0.2%
Total 39,436 100.0%
City of Denver Total SMSA
% of % of
Number Total Number Total
102,637 48.5% 219,207 36.0%
72,065 34.0% 232,803 38.2%
27,578 13.0% 118,935 19.5%
6,460 3.0% 26,030 4.3%
3,264 1.5% 11.711 1.9 %
211,566 100.0% 606,686 100.0%
Source: DRCOG, Profiles of 1970-80 Socio-Economic Change

Table 16
Indices of Neighborhood Integration
White/Black Integration
1970 .850
1980 .707
White/Hispanic Integration
1980 .537 .264
Black/Hispanic Integration
1980 .726 .307
Logan, p. 879 and James, pp. 206-207.

The minority populations of Denver have accomplished an amount of integration, in Denver's suburbs and in the predominantly white neighborhoods in the central city. The massive Hispanic population has moved into practically all census tracts in west Denver and the black population has integrated into neighborhoods in east and southeast Denver. This integration has generally not "tipped" any Denver neighborhoods; integration has generally not led to greater instability. Black integration patterns have been marked by migration in small amounts into a number of census tracts while Hispanic integration patterns have been characterized by gradual succession into a neighborhood.
Further growth of the minority markets is expected, based upon market patterns in Denver. The black population will spread into south and southeast Denver, most likely following the established pattern of gradual movement into a census tract in small concentrations. Generally, the low concentrations will prevent any market "panic" so that black "tipping" is not likely in southeast Denver.
Hispanics, on the other hand, will continue to increase their concentrations in neighborhoods in southwest Denver without adding significant Hispanic population into the predominantly white neighborhoods south of Evans Avenue. The exposure of Hispanics to whites and blacks, however, has

changed little since 1970. The expansion of Hispanic markets will likely continue to follow a pattern of succession into neighborhoods.
This integration, however, involves only a small portion of the minority population. In 1980, the overwhelming majority of the black population in Denver still resided in census tracts in which blacks accounted for 10% or more of the population. The census tracts which had between 5 and 10 percent black in 1980 only contained 8.1% of the city's black population. The Hispanic population, as well, remained concentrated in tracts in which 30 percent or more of the population was Hispanic. Practically half of the Hispanic population in the entire SMSA still resided in Denver in neighborhoods in which the Hispanic population exceed 10 percent.
The separation of the minority populations into the middle-class, which has gained educational and occupational opportunities, and subsequently better housing, and the lower- and under-classes which remain trapped in lower- quality housing has also led to greater segregation of housing markets by income level. The concentration of households with incomes below $15,000 was not significantly abated between 1970 and 1980. Hispanic neighborhoods saw a similar concentration in 1980.
Implications for Further Study
The majority of previous studies dealing with minority migration patterns and migration have focused on central-city to suburb moves. This study has added an extra dimension to the literature on migration by adding intra-city migration to the overall migration paradigm. Central cities are not monolithic in their housing supply; a model of migration should account for the different types of housing available within the central city. An overall model would include a variety of different types of relocations:

inner-city to outer-city, inner-city to suburb, outer-city to suburb, etc. Each of these general locations could then be delineated by housing types and other market variables.
This study focused only upon the overall changes revealed by census tract level data. The housing choices made by individual households can be deduced, but details of housing search and relocation can not be finely distinguished using these data. The next step in this research should focus on these individual moves as revealed by the U.S. Census' Annual Housing Survey and such local sources as the Polk City Directory.
An overall model of housing market migration patterns could also benefit from a longitudinal study of the moves within a neighborhood:
Frey's research has pointed out the different types of households which inhabit a single neighborhood.^ These include the more mature households which relocated into the neighborhood when it existed on the urban periphery and the younger households who move into in at a later time, as an intermediate move prior to locating farther from the urban center.
This study could establish a model for the stages in a neighborhood's development so that the "trickle down" process in the neighborhood could be easily defined by the pattern of moves into and out of the neighborhood. Certainly, the research introduced in this paper could benefit by an examination of these patterns. The overall patterns of integration identified in this study could be more finely detailed through an investigation of individual household moves.
Policy Responses to Recent Integration Patterns
Debate has focused on three kinds of policy that are distinguished by differences in the way they apporach the fact of racial segregation. .

The first has sometimes been called "ghetto economic development." This policy would attempt to relieve poverty by taking advantage of possibilities for direct action within the racial ghetto itself. The second policy, a sort of polar opposite to the first, would attempt to relieve ghetto poverty by helping and encouraging the population of central city ghettos to disperse into the largely white metropolitan suburbs. The third policy, standing logically between the other two, would use the conventional tools of antipoverty policy to integrate urban racial minorities into the central city's economy.^2
The second alternative, dispersal of the minority population, has been
attempted extensively in the Denver SMSA. This paper has emphasized the
segmentation of the minority, especially black, population into different
classes. Suburbanization and integration into central-city neighborhoods
have occurred, mainly in spite of public policy. James has noted that
"Colorado has long been an aggressive state in enacting and enforcing bans
against discrimination."^ In 1959, the state enacted the Colorado Fair
Housing Law, which is "tougher in some ways than is the principal federal
statute, Title VIII of the Civil Rights Act of 1968."^ Like the federal
act, however, the Colorado Law has weakness concerning "provisions for the
recovery of damages by complainants."^^
As well, Denver "has made aggressive use of federal housing subsidy
programs,"^ with a large share of the units being located in the suburbs.
The result, however, has not been wide-scale decentralization of minority or
low-income households:
Subsidized units built in Denver's Anglo suburban neighborhoods were occupied by Anglos. National evaluations have concluded that dispersing subsidized housing did not foster either socioeconomic integration or racial and ethnic integration. The main reason is that units built in suburban areas were inhabited largely by residents of the same area. ^
The conservativism of the 1980s has led to cutbacks in federal support for housing programs: ""The most intense and sustained cutbacks in recent