~w——v— - —"— —-_—' URBAN RENEWAL: PATTERNS OF POPULATION AND HOUSING CHANGE Thesis for flu Degree of pII. D. MICHIGAN STATE UNIVERSITY Carole Ellis Wolff 1966 9’ L I ‘iT 7‘. 4. P. V THESIS 3 1293 7048 ‘ M7- T? I” . . ILJ: {a '.'. \. .‘IILY «re This is to certifg that the thesis entitled URBAN RENEWAL: PATTERNS OF POPULATION AND HOUSING CHANGE presented In] Carole Ellis Wolff has been accepted towards fuIIiIIment of the requirements for Ph ° D ° degree in _S_.O_C i__O logy Date I ABSTRACT URBAN RENEWAL: PATTERNS OF POPULATION AND HOUSING CHANGE by Carole Ellis Wolff I The research reported in this dissertation attempted‘ to measure the effects of a government program--urban renewal—~on patterns of change in housing supply and the character and distribution of the population in two major cities. The patterns of ecological change observed in the two renewal cities (Washington, D. C. and St. Louis, Missouri) were compared with those changes which occurred in a third city (Cincinnati, Ohio) which did not enter the renewal program until late in 1959. The study was designed as a "natural" experiment. The independent variable (urban renewal) occurred in two cities -(the experimental cities) but did not occur in a third city (the control). Comparable data on selected population and housing variables were taken from the 1950 and 1960 Censuses of Population and Housing. Before and after comparisons were made on these variables for both the experimental and control cities. - . In addition to the routine use of medians and per- , dam; .“Qfipfltages, a measure of change in the dependent variables was tht ' 3' t, ., Carole Ellis WOlff introduced. This measure was composed of changes in the value of a variable from 1950 to 1960 for each census tract. These were summed and a mean difference and standard devi- ' ation of the difference computed for each city. Differences in the direction and magnitude of change on these variables - could then be compared for the three cities. The basic hypothesis of the research was one of "no difference" between the experimental cities and the control in changes in the condition and size of the housing supply and in the character and distribution of the population. The two renewal cities (St. Louis and Washington, D. C») experienced decreases in the proportion of substandard and overcrowded housing, while in the control city (Cincinnati) ”T, substandard and overcrowded housing increased. While the proportion of non—white occupied housing increased in-all three cities, home ownership among Negroes was more wideeg} spread in the experimental cities than in the control. The reason for this is related to a difference in population change between the renewal and non-renewal cities. The proportionate increase in Negro population was roughly the. same for all three cities, but the two experimental cities lost from a fourth to a third of their 1950 white population -while Cincinnati's declined only seven percent. There was raise a general improvement in the quality of the dwelling é’ e units occupied by Negroés?’ The proportion of substandard '1‘. -, Carole Ellis Wolff and overcrowded units among non-white occupied housing de-4g clined substantially in the renewal cities and increased substantially in the control city. /x In the renewal cities, the income levels of the popula-~/’. ‘fltion did not increase as much as in the control city, and one renewal city actually experienced a major decline in K I /skilled manual and white collar workers. These and related ‘fkfdata suggest that renewal contributes to the exodus of middle Lclass population from the central city. This accelerated exodus of white, middle class residents from the renewal cities appears to be the result of a redis- tribution of population within the experimental cities. Many tracts in all three cities experienced large increases in populations. But in the control city, the new arrivals in these tracts more often came from outside the central city. In the renewal cities, they came more often from elsewhere in the central city. Population redistribution within the city appeared to take the following form: 1) Large numbers of lower status Negroes displaced ~. ;. . ».fi '115 c-Jncopt II. ‘. a”! Magu‘p‘vdtu. .3 2.16? col Knee. SIC '1 ‘ ' ifigg- .:.:2 1._2.£3-‘ . .~-. "5 “111.111.11.155... «warm Ema! 3109:“: D. c.: Nvernmnnc Printingo ‘ 999¥9‘ i.n "Metrcpolitan 4.. CHAPTER I INTRODUCTION TO THE PROBLEM Introduction The 20th century in the United States has been charac— terized by the phenomenal growth of metropolitan areas—- large concentrations of population around one or more cities of over 50,000 population.1 In 1960, 65% of the U. S. popu- lation lived in one of the 212 Standard Metropolitan Statis- tical Areas.2 And slightly less than half of that 63% lived in the metropolitan rings, outside the central city.3 Thus, 1The term, Standard Metropolitan Area, was introduced by the Bureau of the Census to refer to a county containing a city of 50,000 or more, plus contiguous counties which, according to certain criteria, are essentially metropolitan in character and are socially and economically integrated with the central city. In 1960, this concept was renamed, Standard Metropolitan Statistical Area. See U. S. Census of Population: 1960. General Social and Economic Character- istics, United States Summary. Final Report PC(1)-C. (Washington, D. C.: Government Printing Office, 1962), p. . XXXI. Leo F. Schnore, in "Metropolitan Growth and Decentrale ization." American Journal of Sociology, 65 (September 1957), .1pp. 171-180, gives an historical perspective on metropolitan ‘ growth and decentralization. 2U. 8. Bureau of the Census, U. S. Census of Population: 1960, op. cit., table 101, p. 257. Since 1900 there has been a steady increase in the number of SMSA's and the proportion of the population living in them. There were 52 in 1900 with 31% of the population and 162 in 1950 with 56.8%. See Murray Gendell and Hans L. Zetterberg, A Sociological"A1manac fer) the United States (New York: Charles Scribner's Sons, 1964, 2nd edition), Table 9.14, p. 95. aGendell and Zetterberg, ibid. over the past 60 years, the nation has become more and more metropolitan and, within the metropolitan areas, increasingly suburban--if suburban is defined as anything within the SMSA, but outside of the central city.4 This progressive decon- centration of home, work and shopping center was underscored by the discovery that almost all of the 25 largest central cities lost population during the 1950-1960 decade. Schnore suggests that this may be due to the failure of large cities to annex adjacent territory and population and to the fact that new migrants to the large metropolitan areas are coming,¢/ directly to the metropolitan rings rather than, as in the past, migrating first to the central city and then diffusing outward. The rings, consequently, are being fed by streams of migrants from outside the area and by ex-residents of the central city. The literature on urban research since the early 1950's has reflected the decentralization of population, housing, +Schnore, op. cit., p. 172. In 1950 over half of the population growth in the U. S. took place in the metropolitan rings. 5Schnore, op. cit. In this article, Schnore cites Donald J. Bogue and Emerson Seim, "Components of Population Change in Suburban and Central City Populations of Standard Metropolitan Areas, 1940 to 1950," Rural Sociology, 21 (September—December 1956), pp. 265-275, to the effect that dur- ing the 1940-1950 decade, city and suburban populations reproduced at or above replacement levels, leading one to the conclusion that the loss of city population in the subsequent decade was more likely due to out—migration. ‘- and industry. The Census developed the Standard Metropolitan Area concept for the 1950 Census and Bogue, Schnore, Taeuber and others have been researching and documenting the overall changes taking place in metropolitan America.6 The use of high speed computers has made it possible to compare the changes in all cities of a certain size with changes in all other cities in different size categories, so that generali- zations about patterns of structure and change in cities can / now be made in a way that was not possible only a short time ago. Most of the studies of suburban life are post-World War II 6Schnore, op. cit.; see also some of Schnore's other articles: "The Growth of Metropolitan Suburbs," American Sociological Review, 22 (April 1957), pp. 165-75; "The Separa- tion of Home and Work: A Problem for Human Ecology,“ Social Forces, 52 (May 1954), pp. 556—45; "Components of Population Change in Large Metropolitan Suburbs," American Sociological Review, 25 (October 1958), pp. 570-75. Some of Donald Bogue's contributions to the study of metropolitan decentralization include: Mptropolitan Decentralization: A Study of Differgp- tial Growth (Oxford, Ohio: Scripps Foundation for Research in Population Problems, 1950); Components of Population Change, 1940-1950 (Miami, Florida: Scripps Foundation for Research in Population Problems, 1957); Populationpgrowth in Standard Mptropolitan Areas, 1900—1950 (Washington, D. C.: Government Printing Office, 1955). Also relevant are: Amos H. Hawley, The Changing"Shape‘6f'Metropolitan America: Deconcentration Since 1920 (Glencoe: Free Press, 1956); Henry S. Shrycock, Jr., "Population Redistribution within Metropolitan Areas: Evalu- ation of Research,“ Social Forces, 55 (December 1956), pp. 154—59; Bernard Lazerwitz, "Metropolitan Community Residential Belts, 1950 and 1956," American Sociological Review, 25 (April 1960), pp. 245-52. Many of these relatively recent studies compare central city and suburban populations as in the follow- ing article by Karl and Alma Taeuber, "White Migration and Socio-Economic Differences Between Cities and Suburbs," Ameri» can Sociological Review, 29 (October 1964), pp- 718-29- ' and more books with "suburb" in the title come off the press each year.7 The Problem in the Context of Recent Urban Research In the 1950's and '60's, the central city was studied largely in comparison with the rest of the metropolitan area. Much has been learned about the changes in the central city through these comparisons. We know that, in general, the large central cities are losing population, particularly the white, middle class population, and are gaining Negro resi- 8 dents. We know, too, that overcrowding, which was 7Some of the books reflecting this post—war interest in our growing suburbs are: J. R. Seeley, R. A. Sim, E. W. Loosley, Crestwood Heights (New York: John Wiley and Sons, 1956); Benjamin Chinitz, ed., City and Suburb: The Economics of Metropolitan Growth (Englewood Cliffs, New Jersey: Prentice Hall, 1964); Robert Wood, Suburbia: Its ngple and Their Politics (New York: Houghton, 1959); Wilfred Owen, Egg Mgtropolitan Transportation Problgm (Washington, D. C.: Brookings 1956); William Dobriner, Class in Suburbia (Engle- wood Cliffs, Prentice Hall, 1965). 8The Taeubers, op. cit., suggest that migrants are of higher status than non-migrants, that large cities contribute to their own suburbs and to other metropolitan areas more high- status persons than they receive and suburbs receive more of such migrants than they lose; the end result is that the socio— economic level of the central cities is diminishing and that of the suburbs increasing. In another article, "The Changing Character of Negro Migration,“ American Journal of §pciology (January 1965), pp. 429-41, the same authors report that non— white migrants to a number of large cities are also of higher status than Negro non-migrants and of equal or higher educa- tional attainment than the non-migrant white population. v particularly bad in 1950, has been alleviated due to the massive building activities in the suburbs and that vacancy rates, which were close to crisis levels in 1950, have in— creased to a comfortable 7.7% across the country.9 The ecologists would argue that these changes are the result of “natural" ecological processes——of supply and demand in hous- ing, of space and economical operation for industry in the *7 suburbs, and of increasing dependence on the family car, better freeways, and the declining efficiency of forms of public transportation. What happens when a government program sets out to “interfere" with these natural ecological processes is the \/ subject of this dissertation. What happens when a conscious and "artificial" attempt is made to interfere with the market in housing and urban land values in an effort to prevent the exodus of the white middle class to the suburbs and the deterioration of residential property values in the central city?10 In contrast to much of the recent urban research, 9See Ben J. Wattenberg, and R. M. Scammon, This USA (New York: Doubleday, 1965), pp. 520. Also, for a study of changes in housing in six metropolitan areas and two central cities, see Beverly Duncan and Phillip Hauser, Housing a Metropolis--Chicago (Glencoe, Illinois: The Free Press, 1960). l°The "write-down" is a device, incorporated into the Housing Act of 1949, which allows local urban renewal agencies to purchase slum areas in the central core where land is ex- pensive because of its location and potential use, clear and prepare these areas for development, and then sell them at a reduced price to redevelopers who agree to build in accordance with the plans for that area. The government reimburses the local agency for 2/5 of the loss incurred. / A the focus of this study is on the central city and a measure- ment of the effects of the Urban Renewal Program on changes in the population and housing within it. The Urban Renewal Program was part of the Housing Act of 1949, legislation which was intended to encourage large scale construction of housing to alleviate the shortages caused by 15 years of depression and war. After the war, the men re- \/ turned to marry and establish families. The GI Bill, low interest FHA and VA loans, and the post-war prosperity made it possible for these new families to seek their own housing-- an important change from the pre-war pattern of doubling up with parents and other relatives. In addition, unattached individuals, such as students, young professionals, and middle aged spinsters, sought independent living quarters from their families of orientation. The Census reports that the number of primary unrelated individuals (single heads of households living alone or with some unrelated person) in- creased 78% from 1950 to 1965--which represents a demand for some 5.7 million housing units.11 Not only was there not enough housing, but that which Egg available immediately after the war was in very bad condition. At the time of the 1940 Census, 40% of all the housing units in the U. S. had no bathtub or shower, a third had no flush toilet, almost half cooked on wood, coal, or oil stoves, and 78% used wood or coal as heating fuel. By contrast, in 1960, only 12% 11Wattenberg and Scammon, op. cit., p. 58. of all housing units had no bath or shower, 10% no flush toilet, 5% continued to use wood, coal or oil stoves for cooking and 16% still heated with wood or coal.)12 Only 51% of all housing units were classified as standard in quality by the Census in 1940; in 1950 the percent had risen to 65%, in 1960 to 81%. Between the last two censuses, the percentage of dilapidated housing decreased from 9.8% to 5.2% in 1960. Duncan and Hauser, in their study of six metropolitan areas, found that the improvement in housing quality was due mainly to new construction in the suburbs, rather than to demolition or rehabilitation of older dwellings.13 This corresponds with Wattenberg and Scammon's observation that by 1966, half of all housing units in the U. S. will be of post—war con- struction (roughly 50 million housing units).14 The Urban Renewal Administration was only one of several .g/ agencies formed by the government to repair the cities and J rehouse the population. In terms of financing, acres of land affected, or number of housing units built, it was not even a major program. But, while other parts of the housing program (i.e., the low interest loans) primarily affected the metropolitan rings (because this was where new housing could be mass produced on large plots of vacant, relatively 12Wattenberg and Scammon, op. cit., p. 246. 13See Duncan and Hauser, op. cit. 14Wattenberg and Scammon, op. cit., p. 245. _—Im~ - ... cheap land), urban renewal occurred almost entirely within the central cities. The first Housing Act of 1957 provided some money for mortgages, but was primarily a public housing program. As housing became more plentiful, the desirability of public housing among potential occupants decreased and real estate and construction interests could lobby more effectively against it. The Housing Act of 1949, while containing some provision for additional low-cost public housing, was a much \// broader program designed to facilitate private home construc- tion and to eliminate slums and blighted areas. The declara— tion of national housing policy in the Housing Act reads: The Congress hereby declares that the general wel- fare and security of the Nation and the health and . living standards of its people require housing produc— tion and related community development sufficient to remedy the serious housing shortage, the elimination of sub-standard and other inadequate housing through the clearance of slums and blighted areas, and the realiza- tion as soon as feasible of the goal of a decent home and a suitable living environment for every American family, thus contributing to the development and rede- velopment of communities and the advancement of the growth, wealth, and security of the Nation. The Urban Renewal Program, established by this second Housing Act, was originally designed as a program of “spot" clearance of sub-standard and inadequate housing and of new residential construction on cleared slum sites. Over half of the acreage had to be devoted to residential use. Because of the exist- ing housing shortage in the early 1950's, the slum clearance . program, with its subsequent dislocation of low income families, drew severe criticism. The supply of low income housing was inadequate for the demand. Weak housing codes and the impossibility of enforcing them contributed to the fact that blight appeared to be increasing faster than re- development. In the 1954 amendments to the Housing Act, \/ better provisions were made for the relocation of families and the notion of rehabilitating or conserving existing struc- tures was introduced. But despite the fact that local urban renewal agencies were made formally responsible for the re- location of families from the renewal areas, follow—up studies consistently find that only a small proportion of families come to the agency for relocation help and, of those, many are relocated into other substandard housing and most are forced to pay higher rents for their new location. In response to criticism, more emphasis was also placed on over- all planning. By this time, however, local governing bodies and business groups had become more familiar with the re— newal program and its possibilities. The 1954 Act also in— cluded some support for non-residential uses in renewal areas (up to 10% of the grants—in-aid). By 1961, downtown businessmen were pressing for more help from the renewal program and central city mayors were worried about their declining tax bases. The 1961 amendments made the Urban Renewal Program more of a downtown development// program than a housing program. Up to 50% of the grants—in— aid could be used for non-residential purposes, and housing, , particularly for the low—income population, was all but — )- 10 forgotten. As Kaplan, Greer, and many others have observed, renewal areas are no longer selected because they represent the most blighted residential sections of the city; the basis for choice is the potential, “higher" use of the area.15 In summary, from the first Housing Act in 1957 to the present, there has been a shifting emphasis: from a stress 3 on slum clearance and the provision of housing, particularly 1 for the low-income population, to an increasing emphasis on non—residential construction and overall planning for the re—‘ development of the central city. In the written documents and the public speeches of official spokesmen for the Urban Renewal Administration today, several aims of a renewal program in the city core are specified or implied: 1) to improve the downtown area so as 4/ to attract new business and old shoppers to the central city, thereby increasing the tax base of the financially strapped cities; 2) to provide attractive and convenient “in-town” housing for middle and upper class constituents (which also means increased revenues for the city in the long run); and 5) to relieve the city of at least part of its burden of services to a lower income, and largely Negro, resident 15Scott Greer, Urban Renewal and American Cities: The Dilemma of ngpcratic Intervention (New York: Bobbs- Merrill Company, Inc., 1965). 11 population.16 In terms of the ecological changes documented by Schnore, Bogue, the Taeubers, and others, the aims of re— newal are to reverse some of the trends observed between 1940 and 1956: to decrease the concentration of Negroes in the central city, to attract back part of the higher status (and \ problem—free) white population, and to prevent a complete de- centralization of places to work and shop. If we are to understand the possibilities and potentialities of planned change in an increasingly complex society, it is important to know, as precisely as possible, what the actual effects of an experiment in this type of change have been. Several sociologists have been suggesting the type of research reported on in this dissertation. Donald Bogue, editor of a volume entitled, Needed Urban and Metropolitan Research, called for fewer community case studies and more comparative urban research.l7 With programs like renewal in 16The appelation, "Negro Removal," has been given to urban renewal by the civil rights movement. While there is no question but what the Negro has been disproportionately affected by renewal programs, this author does not believe that the primary motivation for relocating large numbers of Negro families is one of simple prejudice. :Because the urban lower class today is predominantly Negro, Negroes are very much involved with all of the major social problems facing the city--crime, prostitution, education, welfare. It is the problems--rather than the people-—that the city fathers are trying to evict through renewal. The term, "Negro removal," however, does have a factual basis. About two-thirds of all persons relocated from renewal project areas have been Negro. 17Donald J. Bogue, ed., Needed Urban and Metropolitan Research (Miami, Ohio: Scripps Foundation Studies in Popu- lation Distribution, 1955), No. 7. y 12 mind, he emphasized that social planning must rest on a knowl- edge of the facts. Joseph Lohman wrote the chapter on the need for research into the redevelopment and control of slums and blighted areas. Relative to renewal, he suggested that there is a need to measure all of the effects, favorable and unfavorable, which could be attributed to the renewal pro- gram.18 Specifically, he suggested taking a sample of re- developed areas and 1) comparing the population residing in those areas before and after redevelopment and 2) comparing the resident population with those in adjacent areas which weren't redeveloped. Lohman made another proposal which was less directly related to renewal. He suggested a test of several competing hypotheses which attempt to explain R. D. McKenzie's observ— ation that population growth rates were inversely related to distance from the Central Business District and that the V/ cores of large cities were actually losing population. Warren Thompson confirmed this observation by computing growth rates for census tracts grouped by one or two miles intervals from the center of the city outward. Is the loss of central city population due to a strong outpouring of white population in response to a non-white invasion? Is it due to a “a dis— placement of regidential land uses by more intensive land uses, able to pay a higher economic rent on the land?' ‘ laBogue, ibid., p. 56. 15 Or could it be that, in cities built before the automobile to accommodate higher densities of population, the exodus merely represents the effort of population to reach a new density equilibrium appropriate to the present modes of trans— portation?19 Lohman proposed that all three competing hypotheses be preferably contained in a single study so that the relative contributions of each might be measured. He felt that this research should also try to discover what types of people left the inner city tracts and what kinds of people remained. Finally, Leo Schnore, in his article on "Metropolitan Growth and Decentralization, calls for more research into the demographic composition of different parts of the metro- politan area and the migrant streams that flow between them—- into the redistribution of residential population. Only out of an understanding of the processes of change can a meaning— ful theory of urban growth be developed.2° Statement of the Problem The present study derives some of its basic hypotheses from the work of Schnore, Bogue, Taeuber, and others of the ecological school. It has also been influenced by the literature on urban renewal, which will be reviewed in a 19Bogue, ibid., p. 20. 2°Schnore, op. cit., p. 177. 14 later chapter. The focus is not on the renewal process itself, but on the effects of this program on the distribution and character of the central city population and on the nature of the housing supply in the central city. The study differs from most others in the area of urban sociology that the author is acquainted with in that: 1) it is quasi—experimental in design; that is, it is a natural experiment with an independent variable (urban re- newal) "naturally” occurring in two of three cities (the ex- perimental cities) and not occurring in a third (the control city). Before and after comparisons are made for both the experimental and the control cities. 2) it uses a comparative approach in which the cities are matched, not only on size, but on the type of population, the nature of the labor force, the relative age and condition of the housing, and the general location of the cities. 5) it uses, as the major analytical tool, a measure of change in the dependent variables with the census tract as the unit of analysis.21 Three large cities, all border cities between South and North and older cities in the eastern half of the United States, were chosen. They are of approximately the same size, and all have relatively large Negro populations. Two of the cities were major participants in the urban renewal program during the 1950's (the experimental cities--St. Louis, Missouri and Washington, D. C.); the third city (the control city-- \ Cincinnati, Ohio) did not begin to participate in the program until late in the decade and, at the time of the 1960 Census, . 21Both the use of an experimental design and the use of means and standard deviations of the differences on selected population and housing variables are exploratory in this study. . ' ( 15 had only one small project in the execution phase. Two of the cities, one experimental (St. Louis) and one control (Cincinnati), were quite similar in the age and condition of their housing, the occupational make—up of their population, and the educational and income level of the residents. In terms of Hadden and Borgatta's classification of American cities based on a factor analysis of some 65 variables, the profiles of the three cities are roughly similar and St. Louis and Cincinnati are particularly alike (see Table I).22 All three are relatively high in density, percent non-white, and high on the deprivation index; all three were quite low in terms of population increase, and the percent living in single dwelling units and relatively low in average income; they were about average in the median age of the population. Where differences in the profile exist, St. Louis and Cincinnati usually differ from the District: the District has more foreign born, is residentially more mobile (with fewer persons living in the same house, 1955—1960), and has many more migrants. In view of the differences between the two experimental cities, it is not expected that the effects of renewal, if any, will be reflected to the same extent; but if both cities show similar patterns of change--and this 22See Jeffrey K. Hadden and Edgar F. Borgatta, American Cities: Their Social Characpgristiqg (Chicago, IllinOis: Rand McNally & Company, 1965). The profile includes 11 varia- bles on which all U. S. cities over 25,000 are ranked and assigned decile scores (0-9). The variables and the three , cities' rankings on them are included in the accompanying table. € n A. ’ ‘ m -X- Profile of Cities Table I. Issues uorqeonpg squexfirw queoxaa 09-996? esnoH ewes queoxea aseexouI uorq —eIndoa queolea 95v uerpew uxog ubralog eqqu—uon quaoxed xepuI UOIQEAIIdaQ smoouI uerpew fiqrsuea sqrun burIIema exfiurs (3.000) uorqundoa quom SSTQID 16 O 0502 Cincinnati 0 0750 . Louis St 0765 1 C. D. Wash., Their Social Character pp. 80-95. 1965). Rand McNally & Company, Jeffrey Hadden and Edgar F. Borgatta, American Cities Illinois * istics (Chicago, 17 pattern differs from that in Cincinnati-—the conclusion that this pattern of change is due to renewal will be strengthened. In a study with an experimental design, the basic hy- pothesis is always one of "no difference." In the present study the null hypothesis will be applied to changes in two sets of variables at three stages of analysis. Specifically, it is hypothesized that there will be no differences observed between the experimental and control cities with respect to the types of change: 1) in the condition and size of the housing supply, and 2) in the character of the population (race, median income, median education) when (a) the data are summed for each city as a whole, (b) the mean changes (on a census tract by census tract basis) are computed for the city as a whole, and (c) changes in the renewal tracts alone are compared with changes in a selected group of "non-renewal" tracts which experienced major changes in 4 or more population and housing variables. One additional hypothesis and different type of analysis was attempted. One of the most important effects which many authors familiar with urban renewal claim has occurred is that the program has simply moved the "slums“ from one area 23 of the city to another. Martin Anderson claims that the program has aggravated the housing shortage for lower income groups by destroying more low—rent housing than it created.24 23“Slums,“ for the purpose of the research here, will be defined as an area of overcrowded or substandard housing occupied by persons of low education and income. Z‘Martin Anderson, The Federal Bulldozer (Cambridge, Massachusetts: Massachusetts Institute of Technology Press, 1964). ‘1 18 His basis for this claim: in 1950, there were 1,850,000 dilapidated dwelling units in the U. S. (by Census definition); the Bureau of the Census found that 90% of the occupants of this housing paid less than $40 per month rent and that 77.6% of the families and individuals in nonfarm dilapidated house ing earned less than $1,000 a year. Renewal has occurred chiefly in neighborhoods where the housing is deteriorated, if not dilapidated, so that many dislocated families would be in this deprived segment of the population. In contrast, the rent distribution of private residential construction (which does not include public housing) in urban renewal areas (8292 units completed in 1962) pgggp at over $100 a month, with a median of $195 in 1962.25 Alvin Schorr echoes the same idea.26 He says that renewal has created a problem in the low income housing market in two ways: 1) housing has not been available for relocatees at the time they are diSplaced so they crowd "temporarily“ into another neighborhood, converting it into a "slum"; and 2) redevelopment has replaced the slums with more middle and high income housing than public housing units, further depleting the housing stock available to the poor. Few displaced families have ever returned to their former neighborhood. Robert Weaver, former head of the Urban Renewal 251bid. ‘aeAlvin Schorr, Slums and Social Insecurity (Washington, D. C.: Social Security Administration, Report No. 1). Mn. -1 19 Administration, also recognized that relocation frequently "created additional slums and brought blight into new areas."27 Speaking of the District, he acknowledged that “. . . some of the older residential areas were adversely affected. Southwest Washington was cleared only at the price of creating the need for additional clearance in parts of Northwest Washington and the spread of blight in a segment of the Northeast."28 Renewal was not intended to be a welfare program, al— though many liberals supported it in the hopes that it would provide better housing for the poor and, as if by magic, decrease crime, delinquency and prostitution. Slums are more ‘ than old housing—-they represent a location for certain il— legitimate occupations, a place to live for persons who, because of racial discrimination, have limited choice of residences, and a place to live for people who choose ppp to Spend what money they do have on housing. As long as there is a demand for inexpensive housing and as long as the demand exceeds the supply, physical slums cannot be eliminated. Redevelopment has been predominantly concerned with the up- grading of land use and may continue to pursue this objective, clearing first one area and then another until the slums, its people and its problems are beyond the city limits. 27Robert C. Weaver, The Urban Complex: Human Values in ' Urban Life (New York: Doubleday, 1964), p. 55. i 281bid., p. 54. A5 5 I 20 It is important to know, therefore, in evaluating the effects of the urban renewal program, whether the lower income groups and the physical conditions in which they usually live are simply being shifted from one area of theR\// city to another, or whether some improvement in the housing available to this group is occurring in the process. This 3 hypothesis can neither be simply stated nor easily tested; however, a compromise is better than no test at all. When clearance for renewal occurs, large numbers of people are displaced and large scale changes occur in terms of the number of housing units in a census tract, their condition and the number and type of residents. If those displaced are relocated throughout the city and not moved en masse to certain parts of the city, then no changes of equal magnitude, but opposing direction, would occur in the areas of the city not directly affected by renewal. If, on i the other hand, those displaced tend to move to certain t/// ; sections of the city in relatively large numbers, then we 1 might expect to find several census tracts undergoing large scale changes in several housing and population variables and on some of these variables the changes should be in an opposing direction from those observed in the renewal tracts. It is hypothesized, then, that there will be no census tracts in the renewal cities experiencing changes equal in magni- tude but opposite in direction to those changes observed in the renewal tracts. i ._- 2 21 Census data from the 1950 and 1960 Census will be used to test the hypotheses. Comparable information on selected population and housing variables was collected for each census tract in each of the three central cities. The census tract was the basic unit of analysis. Changes in these census tracts over the decade will be compared for the experimental cities, which had several renewal projects enter the executive phase early in the decade, and the control city, which did not have a project enter the execution stage until late in 1959.29 Outline of the Dissertation The remaining chapters of this dissertation will discuss: 1) the methodology of the study, including a descrip- tion of the three cities, the characteristics of the housing and the population which will serve as the dependent vari— ables, and the statistics used to analyze the data (Chapter II); 2) other literature which relates to the basic hypothe— ses of the study (Chapter III); 5) the basic findings of the study (Chapters IV and V); 4) an interpretation and discussion of these findings as they relate to the basic hypotheses (Chapter VI); and 29This information was communicated to the author in letters from the Urban Renewal Agencies in each of the three : cities. 22 ,5) an overall summary of the dissertation, including a critical discussion of the different approach and methods used in this study and some suggestions for other research in this general area (Chapter VII). hurry)” . ‘- .::i .J‘xxfu tam -, - . . n .vrl 51"} . ‘ .. - “r- n -ov :1 a na)‘: 'Cttiixp~ . , .~.: .. ..-. the “-3.0%; and the M”. ;- . I "l: mu onto: the tx-cu-v 0‘07: pk... :1“ ”.--- -. . - _ ~ . ‘ ’O‘lhfl “3:1 he 3"."1hufel‘ t3 Danifil ht! cm .0 Daniel Ka'z meager: Hgghgls L11 L13 - ,4. '\ York: Dr‘zden. 195.!) by Abraham min: 4+. ~- - 1 8w ItanClICOI Chandler Publishing] : , -; " paid:- the can “use“: import-lint ‘* , -- got." :0 Wang. fl » -3.— "4' (...;- '-. _ q €54 _ p) T" __ 7 $1- >-- v f‘uo 5".) :1 as: 'd ‘ ..._ ‘\?1b"yi-- *.¥ 7 -_.,. CHAPTER II METHODOLOGY The design of this study is that of the natural experi— ment.30 Before and after measurements on selected population and housing variables were taken from the 1950 and 1960 Census.31 Two cities were called the experimental cities because a considerable amount of clearance and redevelopment took place within them between the two Censuses. A third city was labeled the control city because it was not a major participant in the urban renewal program in the 1950's and the little renewal that took place did not enter the execu— tion phase until late in 1959. 30The term is attributed to Daniel Katz (Leon Festinger and Daniel Katz, Research Methods in the Behavioral Sciences, New York: Dryden, 1955) by Abraham Kaplan (The Conduct of Inguiry, San Francisco: Chandler Publishing Co., 1964). Kaplan prefers the term "social experiment.“ Whatever the word, the concept refers to an experiment in which changes are pro- duced by an independent variable introduced, not by the scientist, but by the policy maker or practitioner. Kaplan gives as exe- amples the relocation of the Japanese in California and the de- segregation of schools in the southern U.S. The resulting changes, he notes, are often more clear and drastic than could ever be produced in the laboratory. On the other hand, the problem of controls is more serious as well. See Kaplan, p. 164. 31U. S. Bureau of the Census, U.S. Census of P0 ulation and Housing: 1960. Census Tracts. Final Report PHC (1) - 27, 151, 166. (Washington, D.C.: U.S. Government Printing Office, 1962), and U. S. Bureau of the Census, U.S. Census of Population, 1950. Vol. III Census Tract Statistics. Final Report PHC(1)-11, 47, 59. (Washington, D.C.: urs. Government Printing Office, 1952.) 25 III I ’ 24 Description of the Sample The sample consists of all census tractsa2 within three cities. Two of the cities, Washington, D. C. and St. Louis, Missouri, were classified as "leading urban renewal cities" between 1949 and 1960 by HHFA on the basis of the amount of Federal funds spent or reserved for renewal activity in those cities up to 1960. Washington ranked 4th in total Federal funds Spent (16.1 million), 5th in Federal funds reserved but unspent (40.4 million) and 5rd in funds spent per capita ($20.90). St. Louis was 8th in total Federal Funds spent (8.6 million), 6th in Federal funds reserved but unspent (57.0 million), and 7th in funds spent per capita ($11.50).33 These two cities, therefore, have been called the "experimental" cities because they have been exposed to the experimental condition——urban renewal. The third city, Cincinnati, Ohio, differed in that it did not undergo major renewal during the 1950-60 decade. It has been called the “control“ city because f it did not undergo the experimental condition. 32With the exception of those tracts which only contain a public institution. Where the tract is a mixture of public and private uses, it has been included because in the census, ‘ group quarters are not counted as occupied housing units and are not included in the housing inventory. Where population characteristics are being compared, however, some caution will be called for because the population in group quarters '.' ii included in the population figures. ~ 33The figures on which this discussion is based are from Harold Kaplan, Urban Renewal Politics: Slum Clearance in A‘j '. Newark (New York: Columbia University Press, 1965), p. 5. (if ‘I ‘h ‘lIlPF""""""""""""""""""""-"""'------------ 25 There were several reasons for the selection of these particular cities. Washington, D. C., was the first city decided upon because the author had been involved in research on renewal in that city and the observations made in the course of the research led to the formation of the basic hypotheses of this thesis. Since the District is one of the more atypical cities in the U. S.——because of the large white collar labor force, the emphasis on public administra- tion and service industries, the largest proportion of Negroes of any U. S. city, the existence in 1950 of large areas of underdevelOped land within the central city, and the lack of home rule——it was felt that a second city of similar size and equally broad participation in the urban renewal program should be chosen. And since the urban renewal program has been intimately involved with the Negro population in the cities, it was felt that the second city should be relatively similar to Washington in the proportion non—white. Three variables alone——size, participation in urban re- newal, and the proportion non—white——narrowed the field of choice considerably. Table II provides information on the extent of the renewal programs in the execution stage as of December 51, 1959 for all cities of 500,000-999,000 and a few smaller ones in the eastern half of the U. 8.34 34The cities in the western U. S. were either not large enough, had too few non—whites, or were not sufficiently dilapidated to require much help from the Urban Renewal Administration. — .— nun-n—‘n— h'IT D .U .D .eonmsa mmB .noaumuu Inflaflfipd Hm3wsmm swan: .mocmmd mosmcam mEom new msflmsom .mmma .Hm nmnfiwomn .mhouomnfln uoohoum Ammmmwwlmmmmm was CH cm>flm mH ammum soaumamfioo Ho .QOHusowxw .msflssmHm wzu Umfiommn was uowmoum HmHsofluumm m Hmsumsz .mmma .Hm nonsmomn ou HOHHQ ommUw coausowxw msu commune Us: soan3 huwo cw>flm m sH muomflonm Ham How muflcs msflaamsp tam .mumoo .mmmwuom mo soaumafism m “communes sump was .0 .0 .mm coumcHSmmk .soflumuuchHEpd Hm3msom swans .mosmmd monmcflm «50m tam mcflmsom .ommd 26 .om wasp .wofiumflumuomumno pomflmnm HMEMQMMImmmMW Eoum coxmu mm3 magma mflnu CH coaumfiuowew one mam ohmm ooo.NmN.om «.mwm m.mm >m>.a>w dam.dmm mucmHu< smm meow ooo.Now.m> m.mmw o.ma amm.dmm.a mm>.wmw mflsoa .um mama mmmw ooo.w¢m.mma o.m¢m «.mm mmo.¢m¢.d m>a.mom soumGHSmmz smm mam ooo.m¢o.¢a m.>mfi m.wa Ham.mw¢.fi mom.¢fim UGMH0>0HU Shea NNNN ooo.mmn.am m.nma w.mm mem.wmm.d 00>.mwm muoEHuHmm saw was ooo.mmm.fia >.mwa a.¢fi smm.¢aw med.wmm huflo mmmcmx mwa emm ooo.oo«.m N.>ma m.mm www.mmm smo.mmm EarmGHEHHm mm men ooo.mmm.m $.sma m.m Hmm.mmm mam.mom opmaoa amm swam ooo.m¢a.mm «.mm m.md mmm.mam.m wow.m>m. smusnmuuflm mm Nome ooo.mom.m >.sm m.ma No¢.¢0m.y www.mom Humc:flosflo mom mooa ooo.o>m.mw m.m¢ N.>a www.mmw1 www.mmw xum3mz muses muses mumou uomuu ommfi ommd ommd huflo mcHHHmzp mcHHHOBU pomfloum Hm3ms wuflo Hmuuswu soap soap .pum .puwnsm mmouo now as mufls3lcoc Inasmom Imasmom waco muomuu HMBmQMMIQW mmuud unmoumm dwzm huHU mmHuHU Honucoo can Hmucmfiflummxm Hmflusmuom mo consummfiou < .HH manna 27 Detroit, Chicago, and New York were too big. Atlanta, with a large acreage involved in renewal, had a fairly modest gross project cost in comparison with St. Louis and Washington. It is also considerably smaller than either of these cities, both in terms of central city population and in terms of the SMSA as a whole. Atlanta has a high proportion of Negroes, but it was felt that its position as a definitely southern city, albeit a progressive one in race matters, might affect the results of renewal in a way that would not be true for a more northern city. The two cities chosen as the experimental cities were of almost identical size, with large Negro populations and urban renewal projects begun in the early 1950's. They constituted two of the three cities with the greatest acreage affected and the largest budgets for renewal. If the same patterns of change occurred in these two cities, despite some important differences between them, renewal could be the cause. On the other hand, changes occurring in both renewal cities could simply represent the "normal" patterns of change in large American cities during the 1950's. So a control city was chosen—-one which was similar to the two renewal cities in as many ways as possible, but which differed principally in that it was not a major participant in the urban renewal program during the 1950's. Newark and Toledo were eliminated as potential control cities because of their 28 small size. Toledo also had a relatively small Negro pOpu- lation and Newark, it was felt, might be unpredictably affected by its proximity to New York City and its central position in the megalopolis of the northeastern U. S. Pittsburgh is below average (for the 25 largest cities) in proportion non-white, and while fairly low in acreage af- fected by renewal, it is about average in gross project costs for cities in its size category. The control city chosen was second from the bottom in terms of acreage affected and 4th lowest in gross project costs. Furthermore, Cincinnati and one of the experimental cities, St. Louis, were remarkably similar on many other variables which, conceivably, could affect changes in population and housing characteristics. In the classical experiment, the subjects in the experi- mental and control group should be as much alike as possible. In a natural experiment such as this, we could only match the subjects--the three cities—-on as many crucial variables as possible. If the renewal cities do not differ in any system- atic way from the control city on variables which might be related to population and housing changes, then a possible relationship between different patterns of change and the ex- perimental condition can be entertained. If the renewal program has had a unique impact on housing and population changes in the city, it would be expected that the patterns of change in the District of Columbia and St. Louis would be similar to each other and that both would differ from the patterns observed in Cincinnati. 29 The three cities, first of all, are roughly similar in size, are above average in the proportion of the population which is Negro, and are all "border" cities between North and South.35 In 1950, Washington, D. C. had a population of 802,000, St. Louis, 857,000, and Cincinnati, 504,000. During the 1940-1950 decade, all three central cities gained in population: St. Louis by 5.0%, Cincinnati by 10.6%, and D. C. by 21.0%. In the first two cities the change was com- pletely due to natural increase as both experienced a net loss in migration (-5.6% in St. Louis, —1.5% in Cincinnati). Washington's net migration rate was +4.7%.36 Migration has always played a bigger part in Washington's growth because it was the nation's capitol. In the metropolitan area, in the 1940's, migration contributed more to overall growth than did reproductive changes; but in the central city, natural increase accounted for three times as much of the population increase as migration did. The reverse was true for St. Louis and Cincinnati: in both the metropolitan area as a whole and in the central city, natural increase was more important than migration in the area's growth.37 35During the 1940's, the non-white net migration rate to all three cities was almost identical: Cincinnati, 26.2; St. Louis, 28.5; and Washington, D. C. 28.8. See Bogue, Components of Population Change, 1940-1950, op. cit., Appendix Table III. 36Bogue, Components of Population Change, 1940—1950, 0p. cit., Table II. 37Ibid. 50 The three cities are similar in that they have a greater percentage of Negroes than the average for the 25 largest cities. In 1950, the District of Columbia was 55.4% non-white, St. Louis 18%, and Cincinnati, 15.6%; the 25 city average was 15.9%. In 1960, Washington was 54.8% Negro, St. Louis, 28.8%, and Cincinnati, 21.8%; and the 25 city average was 20.7%. Similarly, the three cities were below average in the growth rate of their Negro population. While non-whites in the 25 largest cities increased 55.5% on the average, D. C's non-white population increased 47.5%, St. Louis', 59.9%, and Cincinnati's, 59.4%. Changes in the population and housing characteristics of a city could be related to many different factors, among them the age and condition of the existing housing structures in the city, the occupational make-up of the city, and the educational and income level of its residents. In 1950 the cities were not-identical on these characteristics; but two of them (St. Louis and Cincinnati) were quite similar. This means that one experimental and one control city are not sub- stantially different on these variables. The second experi- mental city differs from both of them, which would only strengthen the conclusion that renewal is an important independent variable if both St. Louis and Washington, D. C. evince similar patterns of change during the decade despite their differences in occupational distribution and housing characteristics. 51 One factor on which Cincinnati is more similar to St. Louis is in the age of its housing supply: 81.8% of Cincinnati's housing was built prior to 1950; 89.4% of St. Louis' housing was built before that time; but only 61% of D. C.'s housing was that old. In fact, 22.5% of the District's housing was built in the 1940's compared with 5.8% in St. Louis and 8.5% in Cincinnati. This was apparently due to the great demand for living quarters for the enlarged government work force during World War II. On the basis of age alone, the housing supply in Washington was less in need of renewal than in either Cincinnati or St. Louis. A second factor which could influence housing changes during the decade is the condition of the existing housing supply in 1950. Again, the control city is more similar to one of the experimental cities (St. Louis) than the two re- newal cities are to each other. In 1950, 28.47% of the hous- ing units in St. Louis and 50.48% of those in Cincinnati were classified as substandard by the Census; by comparison only 12.40% of the units in the District were so classified. Population and housing changes in a city could also be influenced by the occupational distribution of the labor force. While the two renewal cities differ from each other in occupational structure (see Table XXI), there are no systematic differences in occupational distribution between the two renewal cities and the control. Washington differs most strongly from St. Louis and Cincinnati in its high 52 Table III. Age of Housing Structures in Three Cities, 1950 and 1960 (In percents) City 1950 1960 St. Louis Number Reporting 251,915 222,758 Year Built: 1940 3.8% 1959-March '60 .43% 1930-39 6.8% 55-58 1.84% 1920-29 18.9% 50-54 5.10% 1919 & earlier 70.5% 40-49 4.65% "' ~ 59-before 88.00% D. c. Number Reporting 220,325 247,840 Year Built: 1940 22.3% 1959-60 1.6% 1930-39 17.8% 55—58 4.0% 1920-29 21.1% 50-54 10.3% 1919 & earlier 39.9% 40-49 20.9% 59-before 65.2% Cincinnati Number Reporting 157,580 147,106 Year Built: 1940 8.5% 1959-60 1.1% 1930-39 9.6% 55-58 3.7% 1920-29 16.1% 50-54 6.7% 1919 & earlier 65.7% 40-49 9.2% 59-before 79.5% Source: U. S. Census of Population and Housing, op. cit., Table 5, 1950 and Table H-1, 1960. 55 percentage of clerical and professional personnel and in its low percentage of operatives and kindred workers: 51.1% of the city's male work force is in professional and clerical occupations; 47.5% in white collar jobs. St. Louis and Cincinnati are primarily industrial cities with the greatest concentration of the labor force in skilled and semi- skilled occupations (craftsmen, foremen, and operatives): 35.5% of the male work force in St. Louis and 39.6% in Cincinnati are in white collar jobs. Conversely, 42.5% and 55.1% of the labor force in Cincinnati and St. Louis are en- gaged in manual occupations (excluding service workers and private household employees) as compared with only 58.5% in the District in such occupations. The figures for women in the labor force reflect the same differences. In St. Louis and Cincinnati 52.1% of the women work in white collar jobs whereas in D. C. 67.1% are so employed. Conversely, 26.6% and 22.5% of the women in the former cities are in manual occupations as compared with only 7.5% in Washington, D. C.38 As for changes in the occupational distribution during the decade, all three cities increased in the percentage of the labor force engaged in professional and technical occu- pations and all three declined in the percent in clerical and sales occupations. There were no radical changes in the 380. S. Census of Population and Housing, op. cit., Table 2, 1950 and Table P-5, 1960. 54 distribution of the labor force in any of the three cities. The 1960 Census provides information on the percent of the labor force employed in different types of industry. Assuming no radical shifts between 1950 and 1960, Table IV provides a further description of Washington as primarily an administrative center and of St. Louis and Cincinnati as predominantly engaged in manufacturing. Data The basic sources of data are the Census publications from 1950 and 1960.39 The PHC(1) series was used and com- parable information was taken from the 1950 and 1960 publications. There were no changes in census tract boundar- ies for these cities during the decade. However, in Washington, D. C., 1950 census tracts 25, 52, 54, 57, 75, 74, 76, 77, 78, 88 and 95 were subdivided at the time of the 1960 Census. Institutional tract 25.2 was omitted and the subdivided tracts were renumbered from 1 to 125; consequently, while there were 95 numbered in 1950, there were 125 numbered tracts in 1960. Where the data could be summed, 1950-1960 comparisons were made between the original 1950 tract and the sum of the 1960 tracts into which the original tract was subdivided (see Appendix A). Where the data could not be summed (as in median rent or education), the value for the 39See Footnote no. 51. 55 .mIm manna ..DHU «we "msflmdom pom COHumHsmom m0 momsmo .m .D “muusom .sa sore mama mmuooeose III** .mmu4 HMUHumHumum cmuflaomouumz Unmocmum momma .m m.m m.m accosmmmmoum Rogue 6.6 0.4 m.m m.s m.m s.m coeumooom e.m m.m m.m m.m m.m s.m psormmsom muo>eum e.ee o.me o.oe m.ee m.m m.oe moons genome m.m m.m s.m m.m o.fi m.m moons mammoaoss o.m s.m m.m N.m m.m s.m ooeomoeoememss Hofio m.fi m.m m.m s.m III III omouaflmm e.om o.mm m.em o.mm N.o m.s ooeuouommooms o.m s.m m.m o.e o.m m.o soeuoouumooo o.e o.e --- m. III *4--- mouse: soeo omen spec swam mono *emzm announce HumocflUCHO mflsoq .um .U .Q ..nmmz mo Oahu CH osmonfim ucmonmm come .moeoso moose so seomeoeH as oeossoaosm .sH oases 56 original 1950 tract was compared with each of the tracts which resulted from its subdivision. In Cincinnati, there were 110 tracts in 1950; in 1960, one tract--number 86—-was subdivided into two and one new tract was annexed, making a total of 112 tracts in 1960. In St. Louis, no changes were made in the tracts; one institutional tract was omitted. A few changes were made in the classification of infor- mation in the Census itself. In the following cases, changes were made so that the 1950 and 1960 information were ppp comparable: 1) Residential mobility was measured for a five year period preceding the 1960 census, whereas it was only obtained for a one year period preceding the 1950 census. Respondents were asked at the time of the 1960 census whether they had lived in the same house in 1955, or, in the case of the 1950 census, in 1949. Using this measure, we would expect more mobility in 1960 than 1950 because of the greater time span included in the question asked by the census. ‘ 2) A second measure of residential mobility appeared for the first time in the 1960 census. This measure refers to households rather than persons. The question was asked whether the household had moved into the unit between 1958-59, between 1954-57, or earlier. 5) In the 1950 census, the item, "number of housing units reported on," differs slightly from the item, "all housing units in tract." In the 1960 census, the two items are the same . 57 4) Rent comparisons, by race, presented a problem be- cause a) no gross rent was available in the 1950 census and b) nothing but gross rent for non-whites was available in the 1960 census.40 Only differences in contract rent for the total population in a census tract may be compared for the two years, and contract rental differences between the races may be compared for 1950, but no direct racial comparisons can be made for 1960 nor can differences in rents within each racial group between 1950 and 1960 be established. In the following case, a change in the classification of information was made, but, according to the Census, a combination of items in the 1960 census are the equivalent of a single item in 1950: 5) The 1950 item, "housing units with no private bath or dilapidated" is the equivalent of the 1960 items, "dilapidated," "sound, lacking other plumbing facilities," and "deteriorating, lacking other plumbing facilities." The population characteristics used in the study are described in the Code Book for Deck 5 and Deck 1 (columns 4-55) (see Appendix B). The housing characteristics used are described in the Code Book for Deck 2 and Deck 1 4OContract rent is the price agreed upon by landlord and lessee, regardless of furnishings, utilities, or services. Gross rent is the contract rent plus the average monthly cost of utilities and fuels if these are in addition to the contract rent. 58 (columns 55-58). Most of the analysis makes use of seven population variables and five housing variables. The seven population variables are: 1) mediaglschool years completed (total population)-- 5:6-8 2) median school years completed (non-whites)--5:9-11 5) median family income (non-whites)--5:12-16 4) non-white population--5:19-25 5) total population--5:24-28 6) median family income (total population)--5:25-29 7) percent non—white--5:50-55 These variables were selected, out of all the information available in the Census, as those which would best describe the population in socio-economic and racial terms. For the sake of simplicity, the median was used rather than working with the proportion of people in given income and education categories. The five housing variables are: 1) median contract rent (total population)--5:29-51 2) percent of housing substandard--1:59-61 5) percent non-white occupied housing units--2:45-45 4) percent overcrowded housing units--2:46-48 5) percent non-white occupied overcrowded housing units--2:49-51 These were chosen because they reflected best several im- portant characteristics of the housing supply which, conceiv- ably, could be affected by urban renewal. sFor instance, the condition of the housing (whether substandard or not) and the degree of overcrowding were factors which the Urban Re- newal Program specifically intended to alter. Occupancy by race and income level were also important in the aims and results of the renewal program. h 41Deck 5, columns 6-8. 59 Census data have been used by the demographers for some time. Generally, however, the data are used to compare large areas: concentric zones, central city vs. suburban rings, cities of various size, urban vs. rural, etc. It is relatively infrequent that the data are manipulated on a census tract basis, as in the construction of ecological correlations.42 Leslie Kish43 did so when he studied dif- ferentiation within the concentric zones of metropolitan areas. He observed that zonal gradients obscure a great deal of information and found that there was more differentiation between tracts in the inner zones than in those farther out. The same criticism can be made of census tracts which are heterogeneous in character; but the tracts cover a much smaller area and population, and are originally drawn to be as homogeneous as possible. Consequently, they are not quite as arbitrary a unit as the concentric zone. Aside from the difficulty of within-tract heterogeneity, there are three other qualifications about the use of census 42Donald L. Foley, "Census Tracts and Urban Research," Journal of American Statistical Association (December 1955), pp. 755-42, documents five different ways in which census data can and have been used: 1) descriptive use: incidence of a single factor; 2) descriptive use: two or more separate in- cidence patterns overlapping each other; 5) time-series use: in which changes, by tracts, are reported for stated periods of years; 4) analysis of relationships: using ecological correlation; and 5) statistical index form: as in the indices of social rank and familism constructed by Eshref Shevky and Wendell Bell, Social Area Analysis (Stanford: Stanford Unis versity Press, 1955). 43"Differentiation in Metropolitan Areas," American Sociological Review, 19 (August 1954), pp. 588-598. 40 data in this study. First, there is great variability around the mean population per census tract, particularly in the experimental cities, and, in those cities, particularly in 1950. This has implications for the use of medians in the analysis. In order to minimize the error in copying data by hand from the census reports and in order to keep the task of this thesis within manageable proportions, census tract medians for family income, education, and rent were used in- stead of the computation of proportions in various income, education, and rental categories. There is-a loss of precision incurred in this practice which is most serious when a "median of medians" is computed for all census tracts in a given city without weighting the medians for differences in population size. Where changes in the median value of one of these variables are being measured for each census tract, the author feels that the error is not as serious, since we can treat the median in this instance as a "score" on each variable. Second, while the Census has sharpened and refined its categories of housing quality, there has been no system— atic validation of the "dilapidated" and "deteriorated" concepts.44 And third, in the Census measurement of resi- dential mobility (the number of persons living in a different house within the central city), there is no way to distinguish inter-tract from intra—tract mobility. If this had been ‘ 44The author is indebted to Donald L. Foley, op. cit., for bringing this to her attention. 41 available, a more precise test of some of the hypotheses would have been possible. Concepts and Definitions Several words are used throughout this dissertation which have a precise operational definition. The words and their definitions are: 1) 2) 5) substandard housing unit: housing units in the 1950 Census that were classified as dilapidated or having no bath and those in the 1960 Census that were classi- fied as dilapidated, or sound--lacking other plumbing facilities--, or deteriorating--lacking other plumb- ing facilities. overcrowded housing unit: housing units that had over 1.01 persons per room. a significant or major change: since tests of statis- tical significance are rarely made in this disserta- tion, these two words are used interchangeably to refer to a 1950-60 change in a given variable for a particular census tract which exceeds 1 1 or 2 standard deviations from the mean change (or mean difference) on that variable for the city. Where statistical significance is meant, it will be so stated. 4) renewal tracts-—those tracts, any part of which was involved in an official urban renewal project which entered the execution phase between January 1, 1950 and December 51, 1959. Methods of Analysis The data are analyzed in three stages. In the first stage, the information on selected variables are summed for the city as a whole and expressed in the form of medians and percentages. The three cities can then be compared in 1950 42 and in 1960 or changes in these summary measures can be com- pared for each city. In the second stage, changes in the value of selected variables for each census tract are summed and a mean dif- ference and standard deviation of the difference computed. The formulas used for the mean and standard deviation of the differences are: X x (2d)2 — __ 2( 1.1950 - i1960) 2d 2 _ Eda _ T where xi1950 refers to the value of a variable for a particu- Xi1960 refers to the value of lar census tract in 1950 and the same variable for the same census tract in 1960. N is the number of comparisons, or, in other words, the number of paired census tracts for which values of the variable are known in both 1950 and 1960. Both the mean and the standard deviation of the difference are used descriptively in the analysis. That is, the mean expresses the average change of a particular variable which occurs in the census tracts of one of the three cities. T-tests can be used to test the significance of the difference between the mean changes in the renewal and control cities. The standard deviation tells something about the distribution of these changes--whether the tracts are relatively uniform in the magnitude of change on this variable or whether there are wide differences among the tracts in the way in which they change on this variable. Furthermore, it would be of substantive interest if the two t] in Y? T.)— v.. .aL. fiv 45 renewal cities were observed to have much larger standard deviations on some of the variables which might plausibly be affected by renewal and the control city was observed to have a much smaller standard deviation on these variables. F tests can be used to test the significance of the differ- ence between variables in the renewal and control cities. One other use is made of the standard deviation in the analysis. If a variable is normally distributed, two-thirds of the cases should fall within.i 1 standard deviation of the mean and 95% of the cases should fall within i 2 standard deviations of the mean. Therefore, those tracts which fall outside the first standard deviation are relatively unusual in the magnitude of their change on a particular variable; those which fall outside the second standard deviation are very unusual in the size of the change on that variable. It is assumed that the renewal tracts--and perhaps those tracts serving as "receiving" areas for those displaced by renewal--will fall in these categories of major or significant change. Furthermore, if the change in a variable is ppp normally distributed, this will be recognizable because more of the tracts will fall in the extreme categories on either the plus or minus side of the distribution. In the third stage of analysis, the changes in the values of selected variables are averaged for the renewal tracts alone so that the effects of urban renewal on those areas which are formally a part of the program can be directly measured. 44 One additional step in the analysis is taken to corres- pond with the added, more substantive hypothesis discussed in the preceding chapter. Those tracts which experienced major changes on four or more of the twelve selected popula— tion and housing variables--excepting those classified as renewal tracts-—were grouped together, labeled non-renewal tracts, and the changes observed therein compared with the changes observed in the renewal tracts. CHAPTER III REVIEW OF THE LITERATURE The demographic literature from which the hypotheses vuzre derived was discussed in Chapter I. In this chapter, Literature directly related to renewal and its effects will 1x3 surveyed in an attempt to predict which null hypotheses vwill be rejected and which sustained. Those who have studied the urban renewal program have made many observations-—\/ :xnne substantiated by research and others by experience-- 0 Vdiich might suggest the direction of differences in the types of changes in the renewal and control cities if urban renewal luas, in fact, had an impact on the development of the central cityu This chapter, therefore, will be organized around the tMm>'basic null hypotheses and the findings at each level of annalysis that the applicable literature leads us to expect. 1. There will be no differences between the experi- mental cities and the control with respect to the types of change in the condition and size of the housing supply. The first level of analysis is concerned with overall Changes in each city.‘ According to the Bureau of the Census, the number of substandard dwelling units, nationwide, de- Creased.between 1950 and 1960. Duncan and Hauser, in the 45 46 latter part of the decade, reported on changes in the hous- ing inventory between 1950 and 1956 in six selected metro- politan areas.45 For two of the six Standard Metropolitan Areas-—Chicago and Philadelphia--data are also given separatelywfl for the central city. The housing situation in the six SMSA's improved, quantitatively and qualitatively, during the seven year period, despite the concern that "the demolition entailed in urban renewal will exacerbate the depression and war— "46 The size induced housing shortage in metropolitan areas. of the housing inventories increased anywhere from 10% to 40%, with the growth rate between 17-26% for most of the SMSA's. As a result, vacancy rates in five of the six areas increased. According to the authors the improvement in the housing stock was due almost entirely to new construction (from 10 to 40 new units per 100 1950 dwellings) rather than demolition (one to three per 100 1950 dwellings). The number of new housing units gained by the conversion of formerly single- family units was offset by the number of units lost through merging several units within a single structure (as frequently happens in rehabilitation). However, most of the new con- struction was in the suburban rings; the growth rate for hous— ing in the two central cities was closer to 5%. 45Beverly Duncan and Philip Hauser, Housing a Metropolis—- Chicago (Glencoe, Illinois: The Free Press, 1969). The six were: Boston, Chicago, Detroif,‘Los Angeles, New York, and Philadelphia. ’ 461bid., p. 5. 47 There was an improvement in housing quality, too. The proportion of units classified as substandard drOpped sub- stantially in each SMSA-—from a third to a half of them and by 1/12 in New York City. This, it should be noted, may be \// entirely due to the volume of new construction, rather than to any substantial demolition of substandard units. There was an overall decrease in room crowding for both whites and non-whites though the rate in 1956 for non-whites was still four times that of whites. The improvement in overcrowding in non-white occupied units may be due to the general loosen- ing of the housing market and to the fact that non-whites \V have been able to "spread out" more in the central city as a result of the huge out-migration of whites to the suburbs. According to Robert Weaver, the white exodus and the non- v/ white in—migration to the central city has resulted in a larger, more diversified and better quality housing stock being turned over to non-whites.47 Finally, home ownership became more prevalent in each SMSA; in four of the cities N/ rental housing decreased both numerically and proportionately. Not everyone is in agreement with Robert Weaver that the changing racial composition of the cities necessarily results in improved housing for Negroes, particularly where urban ‘/ renewal has led to the large scale dislocation of many Negro families. Relocation efforts connected with renewal were 47Weaver, op. cit. 48 criticized for failing to accurately appraise the existing housing supply in terms of the family size, composition and budget of the population to be moved; the restrictions which race and ethnicity place on a group's potential housing market were often ignored, it was charged. The Housing and Home Finance Agency, as early as 1955, warned communities that slum clearance "could result in a worsening instead of the desired improvement, of the housing conditions of Negro and other racial minority families if the administration of these programs resulted in decreasing the living space presently available in any community to such groups.“48 Only a limited number of predictions can be based on Duncan and Hauser's study since they were mainly concerned with changes in the metropolitan areas rather than the central cities. Since the central cities in the present study are somewhat smaller than those studied by Duncan and Hauser, we might expect renewal to have a slightly more noticeable effect and the growth rates might be slightly higher, particu- larly in the renewal cities. In line with the trend in the other six SMSA's, a general improvement in the quality of housing might also be expected in the cities in this study: substandard housing should decrease, at least propprtionately, and overcrowded housing should decline. If Weaver is correct, there should also be general improvement in the quality of 48"Living Space Available to Racial Minority Families," Housing and Home Finance Agency, Local Public Agency Letter No. 16, February 2, 1955. 49 \ non-white occupied housing units. On the other hand, if those more critical of renewal are correct, the depletion of low income housing available to Negroes through renewal and the construction of luxury housing to replace it should result in increased overcrowding among non-whites. Essentially, the two opposing arguments can be stated this way: 1) if renewal is "successful," it will result in fewer whites leav- ing the city and, as a consequence, in limited improvements, if any, in Negro-occupied housing; 2) if it is not "successful," the relocation process itself will hasten the white exodus from the central city, thus Opening more of the better quality housing stock to Negro occupancy. Morton Grodzins, in fact, argues for the latter; he feels that urban renewal programs, by displacing large numbersIof Negro families, have had the \/// effect of hastening the succession of adjacent residential areas to all-Negro occupancy.49 The second and third levels of analysis emphasize changes which occur in the census tracts within a city. Those tracts cleared for renewal during the decade might be expected to be very different in 1960 from what they were in 1950; other tracts which served as receiving areas for those displacedefiip there was a tendency for many people to re- settle in the same area--should also show some important differences over the decade. The literature on relocation 49Morton Grodzins, The Metropolitan Area as a Racial Problem (Pittsburgh: Pittsburgh Press, 1965). 50 and urban renewal suggest some findings which may be expected at this stage of the analysis. Several authors have pointed out that those displaced by renewal are frequently relocated in other substandard , / housing, or if the housing is better, the price is generally higher.50 A private study in Philadelphia in 1957 found 70% of the relocated families living in housing where the standards did not meet those of the housing code; the comparable fig- ures given by the Urban Renewal Administration varied from 8% to 55%.51 A study done by the University of Southern California between 1955 and 1958 found that of 41 cities en- gaged in renewal, 26 did not provide any assistance for families trying to relocate. In those, of 25,000 families diSplaced, 70% entered substandard housing in the same or adjacent neighborhoods. In the 15 cities where help was provided, 17,000 families were moved, a third ending up in 2 Gans, too, was critical of the relo— substandard housing.5 cation aspects of renewal.53 He pointed out that the West Enders in Boston had to bear the financial burdens of higher 50For instance, see P. Marris, "A Report on Urban Re- newal in the United States" in Leonard Duhl, The Urban Condition: People and Policy in the Metropolis (New York and London: Basic Books, Inc., 1965), pp. 115-154; Herbert Gans, The Urban Villagers (Glencoe, Illinois: The Free Press, 1962). 51Alvin Schorr, 0p. cit. 521bid., p. 64. 53Herbert Gans, "The Failure of Urban Renewal: A Cri- tique and Some Proposals," Commentary, 59 (April 1965), pp. 50-56. 51 rentals for new apartments that were often no better than the ones they left behind.54 Frequently, those they left behind were not always in the worst possible shape. Gans found the Same thing in Boston that Kaplan had noted in Newark--clearance areas were .chosen not because they had the worst slums, but because . 5 they offered the best sites for luxury housing.55 The orig- \ inal goal of redevelopment officials in Newark had been "midele income housing on cleared slum sites."56 Neither of the two goals sought--slum clearance or the provision of middle income housing within the city--proved attainable. Developers and the Federal Housing Authority were unwilling to build or invest in middle income housing on a small area of cleared land in the midst of a Negro ghetto because they felt that it was a bad economic risk. The Federal Division of Slum Clearance, on the other hand, was unwilling to finance a clearance of the entire ghetto and a building of a "city within a city." So the renewal authority was forced to compromise. In finding sites that were acceptable to FHA and the developers they had to bypass the worst blighted 54A study of the West Enders by Chester Hartman in the November 1964 issue of the Journal of the American Institute of Planners found that median rents rose from $41 to $71 per month after a move. Cited in Gans, ibid. 55Forty percent of the West Endeis had lived in good housing. 56Harold Kaplan, Urban Renewal Politics: Slum Clearance in Newark (New York: Columbia University Press, 1965), p. 15. 52 areas and clear instead neighborhoods which were adjacent to relatively middle class areas and which were, themselves, in only a moderately rundown condition. There is one other observation about the relocation process that is relevant to the first hypothesis. Alvin Schorr suggested that people prefer to live in the same neighborhood. He cited a study done by Chapin 25 years ago where 90% of a group of slum residents in Minneapolis relo- cated within a radius of one mile. A very recent study of mid-century urban renewal found the same thing among those who relocated without official help.57 On the basis of these studies, we would expect that the renewal tracts in Washington, D. C. and St. Louis were not the worst tracts in terms of overcrowded and substandard housing in 1950; that if luxury housing did in fact replace the slums, the increase in rents in renewal tracts would be well above the city average; that non-renewal tracts with a large proportion of substandard housing in 1950 would experience sizeable increases in population due to those relocated from the redevelopment areas; and that tracts ad- jacent to the renewal areas would also experience large increases in population over the decade. 2. There will be no differences between the experi- mentalfiand control cities with respect to the types of change in the character of the population. S7Schorr, op. cit., p. 64. 55 Both the demographic and the renewal literature attest to the increasing non-white population in our central cities and the continual movement of middle income whites to the suburbs; many cities appear to be moving toward a pattern of / segregation between the central city and the metropolitan ‘\/ rings.58 Morton Grodzins reports that the increase in non- whites in the 14 largest central cities between 1940 and 1950 averaged 67.8% as opposed to 5.7% for whites.59 In the succeeding decade the pattern was much the same: the average increase in non-whites for the 25 largest cities was 55.5% as opposed to a 4.5% loss in the white population. At the same time that cities were becoming increasingly Negro, non-whites were also becoming increasingly segregatedwfl from whites within the city. McEntire found that the resi- -dential segregation of non—whites increased between 1940 and 31950 in both St. Louis and Washington, D. C., as well as in 60 He suggested that this several other major U S. cities. may have been due to the housing shortage following the war; if so, then a reversal of this trend should be observed in 58According to Donald Bogue, Components of Population Change, 0p. cit., p. 54, the net migration gain of the sub- urban rings in all SMSA's was only 4.8% non-white. However, from the standpoint of patterns of Negro migrants to the metropolitan areas, it is significant that 21.6% of all non— white migration to these areas during the 1940's went to the suburban rings. 59Grodzins, Op. Cit. 60McEntire. opI cit. M 54 the 1960 data for the major cities.61 The question is: did renewal retard or facilitate the mass movement of whites from city to suburb and did it have any appreciable effect on the segregation of Negroes within the city? Both Weaver and Grodzins imply that renewal has indirectly led to increased residential segregation within \V/’ the city.62 The third stage of analysis focuses on the changes in the renewal tracts themselves. Almost all of the authors writing on renewal have remarked on the disproportionate numbers of non-whites affected by the renewal process. Duncan and Hauser reported that "a standard dwelling with non-white occupants is somewhat more likely to be demolished than is a standard dwelling occupied by whites. A sub- standard dwelling occupied by non-whites is somewhat more likely to be demolished than is a substandard dwelling with II 83 white occupants. Whether owner or renter occupied, non- white housing in the cities is more likely to be dilapidated 81According to Karl E. Taeuber, southern cities con- tinued to increase in segregation while northern cities declined somewhat between 1950 and 1960. See his article, "Negro Residential Segregation, 1940-1960: Changing Trends in the Large Cities of the U. S.," paper read at the Annual Meetings of the American Sociological Association, 1962. 82The segregation of Negroes in the city from whites in the suburbs is not the same as segregation of Negroes and whites within the city. Even though non-white population constitutes an increasingly large proportion of the city population, it is theoretically possible for them to be evenly distributed throughout the city. 63Duncan and Hauser, op. cit., p. 85. 55 or lacking certain plumbing facilities. Davis McEntire cites official statistics on 115 renewal projects in 77 cities: 45,000 families had been relocated by the end of 1957 of whom more than 50,000 were non-white. Public housing within urban renewal projects up to 1957 were tenanted almost entirely (97%) by non—whites.64 If the renewal tracts in St. Louis and Washington are "typical," we would expect that the majority of persons in the affected tracts would be Negro and that, if any of the projects are completed by the 1960 Census, any large low income population in the renewal tracts would probably indi- cate occupants of public housing and these we would expect to be non-white. Summary of the Expected Effects of Renewal on Population and Housing Change Expectations concerning the short—run effects of renewal would differ, depending upon the perspective one takes for predictions. Central city mayors, their city councils, and officials of local renewal agencies would have looked for an \f’ improvement in the quality of housing in the city, as older and dilapidated dwellings are demolished and replaced by new apartments and town houses. The relocation agencies could assure the mayor that those diSplaced persons coming to the agency for relocation were placed in housing that was standard 64McEntire, op. cit., p. 558. 56 in quality, though frequently higher in price than the home in the slums. As the new in—town apartments.were finished and new occupants settled in, the above—mentioned officials would probably predict that their city would lose fewer of its white middle class population than cities which did not make the effort to renew. As for the Negro population, it may be that the city officials thought that eliminating the slums would discourage the in-migration of non-whites to the central city, that some of those displaced would leave the city altogether, and that many of those remaining would accept relocation in public housing, thereby improving the qualitycxftheir housing and assuring their resettlement in already Negro neighborhoods. The predictions of sociologists and others who studied various aspects of the renewal process would be somewhat different. Most would probably agree with the city officials that, proportionately, housing in the central city would improve as a result of renewal, if only because of the elimi— nation of many substandard dwellings. Even without much demolition, the proportion of substandard dwellings would probably decrease due to the large apartment complexes which were built in most big cities during the 1950. Overcrowding—- so typical of slum areas--would be expected to decline over- all, but this would not be a consequence of renewal alone. The opening up of tracts of new homes in the suburbs would take much of the pressure off of central city housing; this 57 could be called a ”pull" factor. At the other end, the dis- placement of large numbers of Negro families would almost certainly mean the invasion by Negroes of formerly white neighborhoods, which would serve as a "push" factor in white migration to the metropolitan rings. This would open up more homes and neighborhoods to Negro occupancy which, alleviating the pressure for housing among Negroes, would lead to a decline in overcrowding in non-white occupied housing. The difference,here, between the renewal and non- renewal cities would be one of degree rather than direction of change. At the same time, this author would expect that overcrowding would increase in other neighborhoods in the city, at least in the short run, because most displaced families do not go to the relocation agencies for help, cannot or choose not to Spend much of their income on housing, and therefore seek out areas where cheap housing is avail- able. Whenever the demand for inexpensive housing is high, landlords will appear who are willing to subdivide houses and reap the profits that overcrowding provides. Renewal destroys a good part of the market in cheap housing, but does nothing to eliminate the demand for the commodity. For the 1960 Census, it could also be predicted that the number of housing units would not grow as rapidly in the renewal cities because, while many older units had been de- molished, many of the projected new units would not have been completed. /, 58 AS for the loss of white population from the central cities, most scholars would probably not be as hopeful as the city officials. There are too many factors attracting people to the suburbs--good schools, lower taxes, cheaper and newer homes, more space——and these perquisites of sub—“" urban life have generally been open only to whites. The question is: dpes rgpewal contribute to or retard this out- YE:E_E£EEEE£222”'MOSt of the sociological literature would define renewal as a "push" factor in accelerating the loss of white population, particularly in the Short run. The displace— ment of 20-40,000 people, approximately two-thirds of them non-white, is bound to have an impact on a community. That many people will seriously overcrowd any existing Negro neighborhood. It seems reasonable to assume that those resi- dents of the receiving areas who can afford to do so will move on to less crowded territory. If the number of Negroes in the city remains the same, or increases Slightly, they will be forced to seek housing in formerly white neighbor- hoods. An invasion of middle class white areas by middle class Negroes fleeing a lower class invasion of 20,000 in their own neighborhoods is likely to be on such a large Scale that whites throughout the city would react with concern for the values of their property. This would give an additional "push" to the many other factors pulling whites toward the suburbs. CHAPTER IV TESTS OF THE NULL HYPOTHESIS In this chapter the basic findings relating to the hypo- theses of the study will be described. The chapter is - structured as follows: The first two stages of analysis will be discussed in order. In the first stage, changes in the population and housing variables for the city as a whole are compared for the renewal cities and the control. In the second stage, the average changes on the population and housing variables for the census tracts in each city are com- pared. Within each stage, the data relevant to the two basic null hypotheses will be described, first that data relating to the null hypothesis concerning changes in the housing variables, and second, that relating to the null hy- pothesis concerning changes in the population variables. Stage 1: City—wide Comparisons Housing variables Three of the housing variables--the percent of substand- ard and the percent of overcrowded housing units and the percent of non-white occupied substandard units--showed the predicted differences between the renewal and control cities. 59 60 Both Washington, D. C. and St. Louis, Missouri, had higher median rates of substandard and overcrowded housing in 1950 than did the control city; and both the renewal cities showed a reduction in these median rates between 1950 and 1960, while Cincinnati's rates on both variables increased over the decade (see Table V). While the increase in over- crowding for Cincinnati is a very Slight one, the trend is upheld by the computations of the mean percents on these variables. Differences can also be observed in the changes in the standard deviations around the means of substandard housing: in the two renewal cities, the standard deviations decrease Slightly, while in Cincinnati the increase in the standard deviation is significant beyond the .01 level of confidence.6 The standard deviation of the mean percent of non-white occupied substandard units decreased in both renewal cities (significantly in the District: F < .05) and increased sig- nificantly in the non-renewal city (F < .01). The range of scores around the means for overcrowding were remarkably stable. The housing variables which did not Show a different pattern of change in the experimental and control cities 65A value of F was computed with 100 degrees of freedom in both the numerator and the denominator. The table of the F Distribution to which the value was referred is found in Helen M. Walker and Joseph Lev, Statistical Inference (New York: Holt, Rinehart, and Winston, 1955), Table X, pp. 466-469. 5 .mHOSB o no mpflo one now been bumuusoo coflomE one 00 coaumEonummm cm MHmEflm ma pH “momaomfi mo smaoma m we mflcu poop muoz** .mfia .oomfi uo: .omme "oaro .eumcsaucao .mma .omma "mma .ommfi ".02 .mHSOA .um .mmfi .Omma umm .omma ".U .Q .coumcflcmmz "mum owmmn mum mUHumHumum mop SUHQB so m.z may .omumum mnaBumnuo mmmHSD .SOHDMHSQOQ muHQBIsoc ucmfloflmmsm m m>mn muomuu Ham uoc mmsmumn o0>Ho>sfl kHHMSms mum muumuu uo3mm .mmuHQBIsoc mcH>Ho>sH mmHQMHum> co “ommmn ma spasms on» SUHSB :0 2 may on momma mammSucmnmm SH mnmnfisz* 61 Ammv Ammv moans www.ma mmm.am omo.m mmm.am o.m 6.0m oumosmumnsm mUH£3 Isos unmoumm Amadv **soHumHsmom I I I I nee mmw Hmuou .usmu uomuucoo swaps: Aomv Ammv mafia: pmo3ououm>o mam.ma amo.mfi sso.ma man.om m.om m.sm omflmsooo muas3 Iooc mo usmuumm muflss msflmson mes.m ems.m www.mfi mom.¢a m.aa m.mfi pmo3ououm>o mo unwonmm Ammo *iooo meets msaooos oom.mm mae.fim www.mw mm¢.mm m.¢w m.mm omHQSUoo muflnz Icon mo usmunmm mafia: msflmsoo «mm.¢a mmw.¢a mom.oa emo.mfi m.w ¢.m photomumnsm mo unmouwm ommfi ommd omma omma ommfi ommfi mCOHu5H>mo mommz menace: pumpsmum .U .Q .COumsHsmMB "mmHQMHum> moansom omuomamm so .Ommd paw Omma .muumue Had Mom modumflumpm wumEEsm .m> dance 62 .m> OHQMB .A**V ODOCUOOM 00m ** .m> manna .A*v muocuoom 00w * Ammo Ammo when: aoo.om omm.mm mm¢.ad oem.¢fi m.om H.sm photomquSm muHSB loos unwoumm Aefiav AONHV **:oflumasmoe I I I I mmw mmw Houou .ucmu uomuucoo CMHpmz Awmv Ammv muses omUBOHUHm>o www.ma oms.ma mam.m moo.m m.sm m.sm omflmsouo mufl£3 loo: mo unwound mafia: mafimsos omm.m mes.oa mam.ma mefi.om e.¢fi a.Sfi omo3ououm>o mo unmoumm Amnv *Ammv muflss mswmson www.mm www.mm ecs.ma mmm.afi 0.0s H.H¢ poemsuoo mua£3 Icon mo unwound muss: mcflmso: was.fim mam.sm www.mfi mmm.sm m.aa m.mfi osmosmquSm wo unmoumm owmd omma ommd omma owma ommfi mSOHuMH>mo memos mosses: Unmocmum mcfimsom omuomamm so .mefi pom Omma Husowmflz .mHDOA .um "mmHQMHum> .muomuB Had How moaumHuMum humEEdm .Q> OHQMB 65 .m> OHQMB .A**V muOG#OOm mmm** .m> manna .A*V muocuoom mmm* Ammv Ammv muss: mes.¢m www.0a mam.aa mem.m N.mm a.mfi photomquSm muH£3 Icon unmoumm AOOHV Amoav **s0HumHsmom I I I I mmw mma Hmuou .uswn pomuusoo soaps: Amnv Ammv mafia: U®o3ououm>o oem.ma msm.ma owm.m mmm.m m.wm N.om pmflmsooo muH£3 Ico: mo unmoumm moans moanson amfi.aa Mmm.fia Nmm.wfi www.mfi m.mfi m.ma owoBououm>o mo unmoumm Ammo *xsmv meats moamoos omo.am mma.mm N¢N.Sa wmm.m «.wm ¢.mm pmflmsuoo muH£3 Icon mo ucmonom muflss mcflwson www.mm mma.m amm.mfi mmm.m m.m N.N photomquSm mo unmoumm ommd omma omma ommfi Coma omma mc0fluMH>mo mammz monsomz photomum Caro .HuMGSHUSHU unmanneum> moamsom omuumamm so .Ommd paw omma .muomue Had How mUHumHuMHm mumafism .o> mHQMB 64 were: the percent of non-white occupied housing units, the percent of non-white occupied overcrowded units, and the median contract rent. In all three cities the percent of non-white occupied housing units, whether computed in medians or means, increased and so did the standard devi- ations around the means. The difference between the absolute values of the medians and their matching means is due to the degree of segregation of the Negro, particularly in St. Louis and Cincinnati. In roughly three-fourths of the tracts in these two cities there are no non-white occupied 5 The median and the means on this variable in the units.6 District are much closer together because a little over 2/5 of the tracts in both 1950 and 1960 had non-white occupied housing units. In all three cities, the median percent of non-white occupied overcrowded units went down. However, the pggp percent and the standard deviation went down in the experi- mental cities, but went up slightly in the control city. In Cincinnati the difference between the 1950 and 1960 means on this variable is somewhat greater than the difference between the two means on overcrowding in general, which sug— gests that the reason for the increase in overcrowding in Cincinnati is due to pressure in non—white occupied units. 66When the means were computed, all the tracts with zero values on this variable were averaged into the mean; for means computed only on those tracts with sufficient non-white occupants, see Table IX. 65 This is confirmed by the information in Table V, which shows a much higher rate of overcrowding in non-white occupied tracts. Finally, in all three cities the median rent went up and no differences at this stage can be determined between the renewal and non-renewal cities. In summary, the null hypothesis is clearly rejected at the first level of analysis for three of the housing vari- ables (the percent substandard, the percent overcrowded, and the percent of non-white occupied substandard units) and tentatively rejected for a fourth (the percent of non-white occupied overcrowded units). It clearly cannot be rejected for two of the variables: the percent of non-white occupied units and median contract rent. Other information about the general housing situation in the three cities will be helpful in the later discussion and interpretation of these findings. Proportionately more new building took place in the District than in either St. Louis or Cincinnati: of those units reporting, 15.9% had been built during the 1950‘s in the former city, 7.4% and 11.5%, respectively, in St. Louis and Cincinnati. Concomitantly, Washington‘s total housing supply increased 14.5% over the decade. This can be compared with an increase of only 5% in Chicago and Philadelphia, as reported by Duncan and Hauser.67 ist. Louis' supply actually declined by less 67Duncan and Hauser, Op. cit. 66 than 1%; and Cincinnati‘s increased 4.47%. While the actual number of new units were very close in Cincinnati and St. Louis, Cincinnati had a smaller number of housing units to begin with and St. Louis demolished a large number of its housing stock during renewal. There is further support for the contention that renewal restricts the housing supply through demolition and inade— quate reconstruction: the vacancy rates in 1960 were lower in the two renewal cities (2.99% in D. C. and 5.92% in St. Louis) and slightly higher in the non-renewal city (4.20% in Cincinnati). There is also some support for Robert Weaver's assertion that renewal and the exodus of whites from the city opened up a greater variety of housing to Negroes. Table VI gives the percent of owner-occupied units that are occupied by non— whites. In all three cities home ownership among non-whites about doubled. It is more likely that this is the result of a general increase in the number of Negroes inhabiting the central cities, and of other social forces at work during the 1950's. First, the potential for Negro home ownership was widened by increasing educational and job opportunities and an improvement in family income for non-whites. Second, the . FHA and VA programs initiated after the war opened up housing “J opportunities in the suburbs for whites confined to the over- crowded cities during the depression and war. AS vacancy 67 oa.om mm.oe oe.ma sa.om ao.me ma.mm ooaosooouuousou mUflSBIoos unmoumm ao.oe ea.m so.oe om.m oo.ae ma.om ooaeoooo uoszo OHH£3ICOG #COUHOQ oa.mm ma.em om.oe mm.am mo.mm mm.mm ooaooooo Iumc3o ucmuumm oooe omoe oooa omoe come omme mason .um sumsoHUCHU .o .Q .coumsHSmmz Ommfi pom ommd .mmHuHO mouse CH muHCD omHQSUUOIHmucmm pom Hma3o CH momsmno .H> mamas 68 rates increased following the housing boom, better opportuni- ties for Negroes as well as whites appeared in the urban housing market. Finally, during the 1950s, there was increas- ing utilization of government-underwritten mortgages by non- whites. Weaver noted that the proportion of all FHA and VA loans going to non—whites had increased from 2.5 to 2.2, respectively, in 1950 to 2.5 and 5.7 in 1960. Whereas 18% of all non-white mortgage-holders in 1950 had VA and FHA loans, 29% had such loans in 1960. Also, over the decade, there was an increase of 254,000 non—white home owners holding con- ventional mortgages.68 The Executive Order for Equal Oppor- tunity in Housing, signed by Kennedy in 1962, helped to strengthen open-occupancy policies which had been formally adopted in many cities and states and Specifically banned any discrimination in housing covered by FHA and VA mortgages. Urban renewal, however, may be responsible for the dif— \¢\ ference in the ratio of owner-to-renter occupied units among non-whites. In a city which was only 22% non—white in 1960 (Cincinnati), 75% of all rental housing was occupied by non- whites and only 11% of the owner-occupied units were owned by Negroes. In D. C., a greater percentage of owner-occupied units than of rental units were occupied by Negroes and in St. Louis the proportion in rental units was not quite double that in owner-occupied units. The reason for this may lie in 68Weaver, op. cit., pp. 265-264. 69 the fact that Cincinnati did not experience the great out- migration of white population that the two renewal cities did over the decade (see Stage 1: Population). Population variables Differences between the renewal and control cities were observed on three of the population variables-—median family income for the total pOpulation, non-white median family in— come, and the total population per tract. The medians of total population per tract were much higher in the renewal cities in both 1950 and 1960. Both the medians and the means suggest a greater average decrease in the per tract popula- tion in the renewal cities and there are significant decreases in the standard deviations (F < .01) for St. Louis and D. C., suggesting that the population was more normally distributed among the census tracts in 1960 and much more skewed (toward the over-populated end) in 1950.69 In the control city, the difference in average tract population is very small and there is very little change in the standard deviation over the decade. Since the difference in tract Size existed in 1950, that cannot be attributed to renewal; it may be that Cincinnati, as a city, was tracted by the Census later than 69Another indication of skewness and its direction is the difference between the median and the mean in each census year; in 1950, the mean is further from the median and con— siderably higher than it, indicating large extreme values, and in 1960, the mean and the median are very close together, indicating a more balanced distribution. 70 St. Louis and Washington and that, therefore, the average tract population is closer to the Size recommended by the Bureau of the Census. It is the change between 1950 and 1960 in the experimental cities and the lack of change in the control city which may be attributed to the renewal process. The difference in family income, both for the total population and for the non-white population, is one of degree rather than direction between the renewal and control cities. The median "median family income" for the total population increased 49% in the District, 51% in St. Louis, but 97% in Cincinnati; similarly, the median "median family income" for non-whites increased 79% in the District, 106% in St. Louis, but 162% in Cincinnati. A possible explanation for this difference in degree is that renewal aggravated the exodus of middle class families and converted a greater number of the city's tracts to lower income residences. Cincinnati, on the other hand, did not lose as much of its middle class population and consequently reflects more accurately the in- creased affluence of both whites and non-whites in the 1960's. This interpretation is somewhat strengthened by the fact that 15.8% of all families in Cincinnati had incomes over $10,000 while only 10.8% of those in St. Louis did. However, the differences discussed here are partly a matter of the distribution of income groups rather than a description of 71 .meooq .um ea emm. .ommmn mum moflumeMum Hmnuo Ham LUHQB so m.z may mom m m>mo muomuu Ham uoc mnemomn om>Ho>aH maamsms mum muomuu Hmzmm .HpMSSflUSHO ca Rmo. paw .uofluumfln mnu SH &wa.ma mm3 owmd CH muHSBIsos ucmoumm smaomfi m£B*** .m> magma ou muoauoom mnu mom .SOHumHsmom muH£3laoc ucmflUAMMSm .mmuHS3Isos mcH>Ho> IGH mmHQMHHm>.oo “comma ma amends ens SUHSB so 2 msu ou Home“ monocusmumm SH nnmnfisz** .coaumHsmom Hmuou mnu now new mmuafl3lsoc Mom soaumosom Aflzw OEOUSH mHHEMM smwoma won ou mGOflumEonummm mamaflm mum menu “womaome mo mamaomfi mum mmmnu umnu muoz* oe.am me.mm ma.om mo.mm oa.mm ma.mm ***ouas3Iooo usoonmm I I I I masse momma *soaemeoooo stoop .mEoosH maflfimm swaps: «mam mmom, mmeo mono Noam oaoo eoaumeoooo Hopes Ammv Aamv muHSBIaos .mEousfl I I I I moose omame *seasmu spasms Ammo Ammo *immuaszIeoo I I I I m.m m.m sowumospm endows I I I I m.ee e.ma soaooeoooo stoop *.:0Humodom swaps: oooa omoa oome omoa come omoe mCOHuMH>0p memo: nonfipmz ouooeoom coaumHsmom omuomamm so .Ommfi Ugo Ommd .U .Q .coumcflsmmz "moanmaum> .muomne Had mom mOHumHumum wumEEsm .MHH> OHQMB 72 .MHH> OHQMB 00m *** 6G6 ** .* amm.em ama.om meo.em emm.me oe. om. *IrooaszIooo unwound *COfiumHsmom immao image stoop .msoose I I I I moose scene season spasms moon memo ammo mega ammo aeoo soaumeoooe spoon Awmv AONV muflnzlcos .mEOUGH I I I I «some songs *sHasou coasts Ammv ANNV **mufl£3Icos I I I I m.m fi.m cowumosom soaps: I I I I m.o a.o soaumeoooo Houou *.c0flumosom scape: oooa omoa come omoe oooe omme wCOHUMH>me mammz mGwHUOS photomum coaumHsmom omuumamm so .Ommfi pom Omma HHSOmmHE .mHsoq .um "mmHQMAHm> .muumua HH¢ How moflumflumum humfifism .QHH> mHQMB 75 .MHH> OHQMB 00m *** Ucm ** 4* ooo.em www.mm omo.oa eae.oe N.oe ma. 44*ouas3Isoo usmoumm *COflumHsmom Aaaav Hmuou .mEoosH I I I I seams eoome season coasts ooem oamm area Nome seas ease soaumeoooe Hmuos Aomv Aaav mpHoBIco: .mEoosfl I I I I osoee emmee *seasou spasms ions ioeo *roueszIsos I I I I m.m w.m coaumusom CMHCOE I I I I e.m e.m soaumgoooo Houou *.s0Humosom GMHUOS oooe omoe oooe omoe oooe omme mCOHDMH>mU mammz menace: cementum soflumHsmom omuomamm so .omma pom ommfi OHQO .Humccflosflu "mmHQMHHm> .muomna Had Mom mUHumHumum mumfifism .UHH> magma 74 the income level of the cities.70 It is further strengthened by shifts in the occupational composition of the city popu— lations as opposed to the relative similarity of the labor force for the SMSA's (see Appendix C). St. Louis declined in the proportion of skilled craftsmen and foremen, while Cincinnati increased considerably in this fairly well paid segment of the blue collar group; the proportion of service workers and semi-Skilled operatives increased in St. Louis and decreased in Cincinnati. While the white collar portion of the labor force (including professional and technical personnel, managers, officials and proprietors, and clerical and sales) went from 50.9% to 29.8% in Cincinnati, it de- creased from 27.8% to 24.6% in St. Louis. No differences were observed between the experimental and control cities in the percent non-white and the median education of the total population and of the non-white popu- lation. When the true medians for the three cities are examined, however, there is a difference: Washington's population decreased in average education from 12.0 in 1950 to 11.7 in 1960; St. Louis' population increased only slightly from 8.7 to 8.8; but Cincinnati's average education went up from 9.0 to 9.7, a difference which fits in with the findings discussed in the preceding paragraph. 70The true median family income for 1950 and 1960 in the three cities was: Washington: 1950, $2975; 1960, $5995; St. Louis: 1950, $2718; 1960, $5555; Cincinnati: 1950, $2644; 1960, $5701. 75 On the surface there appears to be no real difference between the experimental and control cities in terms of changes in the percent non-white: the median percent over all census tracts increased in all three cities. The inter- esting difference is between St. Louis and Cincinnati, and Washington, D. C. Because Washington has had a much higher proportion of Negroes, more census tracts in that city have had a high non-white population. St. Louis and Cincinnati, though relatively high in the proportion non-white among major cities, follow more the pattern of southern cities where Negro families are scattered in little pockets throughout the city; in both these cities there are many tracts with .1% or less non-white, whereas in D. C., there are sizeable Negro populations in almost all tracts (see Appendix D). Consequently, the median non-white population per tract in St. Louis and Cincinnati is yppy low. The discrepancy between the means and medians suggest how skewed the distribution is. It may be that Cincinnati is becoming more like the northern cities in the distribution of its non-white population. It appears (from the closeness of the mean and median in 1960) that it at least became less segregated in 1960 than it was in 1950. This is confirmed by the following table (Table IX). By comparing Table VIII with Table IX, it can be seen that there are some very real differences between the renewal and non-renewal cities and the patterns of change in the white and non-white population. First, all three cities were 6 7 Rooa mm. ao.m sm.m mm.m mm.mm mm.mm Coma Rooa ms. mm.s ma.> a¢.m ma.aa om.¢w omma aumscaocao Rooa am. ma.N NO.N mm.m Os.ma mm.mh omma Kooa mm. mm.m aw.¢ ¢¢.m ¢S.ma am.¢m omma mason .um Rooa aa. mm. ow.m om.m om.ma m>.o> owma Rooa ma. mm.m mm.w em.mm mo.NN mw.m¢ omma .U .Q .soumsanmmz muons» Had Ra mIa smIoa memN eulom +05 m £ua3 muomue mamcmo ca moa>aq coauMHSQOm mua£3lsoz amuoe mo coauuomoum muHSBIsoz mo unmoumm sm>a0 .uao .mo .wnauoouw . NH OHQMB "mousom m.¢I m.mm s.om m.ma mmauau ummmHMa mm o.SI a.mm m.am o.ma aumssaoqao o.¢NI m.mm m.mm o.ma mason .um m.mmI m.s¢ m.¢m ¢.mm .O .Q .soumsanmmg coaumHsmom mua£3 soauMHSQOQ ouanBIsoc Omma omma mmauao ca mmcmoo unmonmm ca mmcmno unmonmm mua£3Icoc unwound ooaanmma .eoaoeo mouse on ooeezIeoz mo eoflouoeoue are ca momcmgo .HHH> magma 77 'below the 25-city average in the growth rate of their non- white population. St. Louis and Cincinnati were practically identical in the percentage change of the non-white popula- tion. Where the three cities differed was in the percentage change in the white population: while all three lost white population during the decade, the Size of the loss in the two renewal cities was much greater than that in the control. Cincinnati was much closer to the 25-city average in the per- centage loss of its white population; D. C. and St. Louis lost a third and a fourth respectively of their white popu- lation. This would appear to be fairly strong confirmation that renewal accelerates the exodus of whites from the central city. However, other possible causes of the same phenomenon must be considered, and, if warranted, eliminated. The loss of white population is extreme in comparison with the average for the 25 largest cities. Something must differentiate St. Louis and Washington from other cities of the same size. One possibility, which might affect the move— ment of population from the central city, is density. Table X Shows that the two experimental cities were much more dense than the control and, concomitantly, the metr0politan ring of the control city was more densely settled than the corres- ponding rings of the experimental cities. Perhaps the vacuum created by low density suburbs and the pressure of extreme density in the central cities led to the shift of population. 78 For these three cities only, there is a direct relationship between the 1950 density and the loss of population during the 1950-1960 decade, and an inverse relationship between the density of the suburban rings and the loss of central city population. However, the physical location of these_ three cities must be considered in the comparisons. Washington and St. Louis are both situated so that their sub- urban hinterlands are "expandable"; that is, there is open land around these metropolitan areas into which population can flow freely. Cincinnati, on the other hand, is hemmed in by the hills of Kentucky to the south and the Dayton SMSA on the north. This may explain why the Cincinnati SMSA was more dense than either of the others in both 1950 and 1960. This direct relationship between density and population change is not upheld when all of the 25 largest cities are examined. In Table XI there is a rough correlation between the direction of population change (loss or gain) and density. This relationship, however, is explained by the fourth column in the table which indicates the amount of land annexed to the central cities between 1950 and 1960. All of the cities which gained in population during the 1950's, with the ex— ception of New Orleans, experienced annexation over the de- cade. All of the cities which lost population, with four exceptions, experienced no annexation during the decade. In three of the four exceptions, the additional parcels of land were quite small. Within the two groups of cities--those which lost population and those which gained-~there is no 79 one on poops mm3 huosou somnmmmmb .Nmma pom Nmma .omma ca émzm mason .um ..U .Q .soumcanmmz .moammo msa Iusaum usmEcum>oo .m .D .AusmEmammsm uumuumfld aouaumaumum dv Nmma poo Nmma .xoom mumn muao pom mucsoo .msmomo 05» mo smmusm .m .5 was some corms mum mason were ea memo weer o.mo mom was.oma.a sos.a swam mo umpaamSwu m.eI www.ma www.mos am mace soap s.om mam.a amm.aoo.m mme.a swam soumsasmms o.mm mas aao.oam.a oma.m «mam no umoaamsmu m.NaI www.ma omo.oms as auto soap m.ma moo moa.ooo.m ama.m «mam mason .um a.me asm sso.mom mmm emzm 00 possesses m.oI aom.o omm.mom as auto soap m.ma moa.a emo.aao.a one emsm aumssaosao coma a.aaa was www.mmm nae.a swam mo possessou o.am ama.ma maa.mom as auto soap m.am eoo moo.eoa.a ooe.a «mam soumsasumz m.mm mmm www.4mm mm¢.m swam no umocaosmu o.m oeo.ea ooa.omm as auto seep e.aa sow , amm.amo.a omm.m swam mason .um m.om aam. do¢.oo¢ mmm dwim wo Hmpsamfimn o.oa aaa.o mom.mom a.ma Sago soap m.ea mmm.a moa.eom one swam Homeeaosao omma womuwo msaomumnm wuamcma seaboasmom mmaae dmzm nm>o coaumHsmom ca meadow mummuusa wmmucwonwm ca mend $I *mea flaw omma .mmmud amoaumauMum cmuaaomonumz pumpcmum 009:8 How huamcmfl Dam mwu¢ Ucmq .N manB 80 o I m.a o.oa+ mammauo 302 o a.mma+ a.mm ¢.sm+ soumsom o m.sma+ w.mw ¢.mm+ moaamo s.mma m.m+ «.ma m.mm+ mmaomS¢ moq s.moa m.mm+ m.mI m.a+ spas manner 0 o.am+ s.me m.mw+ 0ac0u:< com m.mm m.¢+ «.ma m.ma+ um>cma o s.sa+ m.ma a.ma+ mauummm s.sm N.m+ m.SI n.0I aunsdaucao o o.ma+ 0.5 m.aa+ maaommsmaosa m.m¢m >.N+ >.mI ¢.>I maaommmscaz m.mma I w.maI a.aI muoEauamm m.sma m.m+ m.maI m.¢I onwam>mao «.mm I a.maI a.oaI smuonuuuae m.mm a.aa+ m.m m.ma+ mmxsmBHaS o.mam I m.mmI m.sI .0 .n .coumsasmms a.me¢ I m.mmI >.mI uaouuma m.mmm I o.¢NI m.NaI mason .um w.oma I m.maI N.mI 0ammmsm N.oaom I m.maI m.mI meseeoooease m.mo I a.saI o.maI coupon m.mse I ¢.waI m.¢I oumaocmum smm a.aoa s.oa m.NaI o.aI ommoaso m.me I m.mmI m.SI xHMBmz a.mma I s.mI a.aI spas snow 302 am3mc0u mmaaa mumsvm ca ommanmma mucwoumm ca omma ca .Ommanmma .mmsmno coaumH .OmIomma mmcmsu .muamsmo mo Hmouo mmmmuod .pcma omxmcsd lemon muaoz coauMasmom msaosmommo ca mmauao mmauao ummmumq mm 0:» ca wmcmsu coauMasmom one wuamcmm .HN magma 81 apparent relationship between density of the central city and the size of the gain or loss. Some Of the more densely settled cities lost a smaller prOportion of their white popu- lation than did either St. Louis or Washington. It is perhaps significant that most Of the cities which gained in population and which were able to annex adjoining areas during the 1950's were in the southern and western parts of the U. S. In comparison with the older cities of the east, these are newer cities in sections of the country which have been industrializing and developing major urban centers largely since World War II. Houston and Dallas are partially products Of the space age, moving from 14th and 22nd among U. S. cities in 1950 to 7th and 14th by 1960. Some of these cities--unlike Newark, Boston, Philadelphia, and Cincinnati, for example-—have fairly Open hinterlands in which to expand; and the Texas cities are particularly favored by easy annex— ation laws within the state. Because of their relative youth, it is not surprising that these rapidly growing cities were less involved with the renewal program than the Older and more densely settled cities of the east. In the case of the three cities studied, the loss Of population was related, not only to the density Of the central city, but also to the lower density of the surrounding metro— politan rings. In order to test the possibility that it is the difference in density between central city and suburban ring which leads to a change in population, the 1950 and 1960 82 densities for city and ring were computed for those eastern cities which were considered for inclusion in the present study. In seven Of the eleven cities the density Of the SMSA increased; in four, where new counties were included in the SMSA in 1960, density decreased. In all eleven central cities, density decreased, and in all eleven metropolitan rings, density increased.71 There is no apparent connection between the density of the ring in 1950 and the size of the decrease in density in the central city over the decade. A second factor which could be responsible for the ex- treme loss of white population in the two experimental cities is the proportion Of Negroes in the city or a dramatic change in their proportion of the population. Housing may prove to be the most difficult area for integration in the U. S. and it is possible that a significant increase in Negro popu- latiOn may send a large number Of whites scurrying for the suburbs. From an examination Of the three cities studied neither the proportion Of Negroes nor the change in this pro- portion appear tO explain the difference in the loss of white population between the experimental cities and the control. Cincinnati and St. Louis are roughly Similar in the percent non-white in 1950, yet Cincinnati lost significantly fewer whites over the decade. St. Louis and the District were very 711n all but three Of the 25 largest cities, central city density decreased between 1950 and 1960. The exceptions were Los Angeles, Denver, and New Orleans. 85 different in the percent non-white, yet they both lost large proportions of their white population. Cincinnati and St. Louis were almost identical in the proportionate change in non-white population over the decade, yet they differed sig- nificantly in the loss of white population. This same lack of relation was upheld by a comparison Of the percent non- white and the extent Of population change in the eleven cities considered for inclusion in this study. The third possible factor affecting the differences in the loss of white population between the experimental and control cities is the independent variable in this study. There is some additional support for concluding that the ex- tent Of renewal (as measured by the amount of acreage affected by renewal projects in execution as Of December, 1959) may be a contributing factor in the loss of white population. It is possible that the actual number of Negroes in a city, or an increase in the proportion Of Negroes, may escape the notice of white residents as long as the minority population and any growth in it are contained in separate neighborhoods. There may be some invasion Of white neighborhoods adjoining Negro enclaves, but this would probably not affect whites in other neighborhoods further away. However, the large scale movement Of Negroes around the city--such as occurred in Washington and St. Louis following the relocation Of 20,000 to 50,000 people--would attract the attention Of white property owners and perhaps set Off, in reaction, a large scale flight to the suburbs. Also, such massive dislocation of a visible minority 84 would (and did in the experimental cities) mean invasion on a grand scale and Of more and distant neighborhoods than would have occurred as a result of simple population growth in the minority community. If this interpretation is correct, one would expect that<\fi>z the larger the area affected by renewal, the greater the (1‘7 number Of families displaced; and if a sizeable proportion of these families are Negro, it would be expected that a correspondingly large displacement Of white population would follow. Table XI indicates that the two experimental cities dif- fer from the control and from most of the other 25 largest cities in the amount Of acreage involved in renewal in these cities. St. Louis and Washington rank third and fourth in acreage in renewal, outranked only by two much larger cities, Chicago and Philadelphia. For the eleven eastern cities which are roughly similar in size and proportion non-white, there appears to be a general relationship between the amount Of acreage involved in renewal and the direction and magnitude Of population change (see Table XII). When one considers the number of complicating factors, the differences in Size and density and the fact that the renewal projects in the several cities are in different stages of execution and completion, it is somewhat surprising to find so direct a relationship. In summary, neither density nor the percent non-white appear to account for the loss Of white population from the 85 Table XII. The Extent Of Renewal and Population Change Cities in Ascending Population White population order of renewal change 1950-60, change, acreage, 1950 in percents 1950-1960 Newark -7.6 -26.8 Cincinnati* -0.5 -7.6 Pittsburgh +0.8 -15.4 Toledo +7.5 -0.2 Birmingham +4.6 +4.9 Kansas City +1.5 -2.5 Baltimore -1.1 -15.6 Cleveland* -4.2 -18.6 Washington D. C. -4.8 -55.5 St. Louis —12.5 -24.0 Atlanta* +47.1 +45.2 *Cincinnati and Cleveland annexed 2.2 and 6.2 square miles respectively over the decade; Atlanta annexed 91.5 square miles. 86 central cities. Both high density and extensive participa- tion in the urban renewal program are characteristic of the Older eastern cities and would thus seem to be functions Of the age Of a city. When factors such as size, age, and percent non-white are roughly controlled for, as in the com- parison of the eleven cities considered for inclusion in this study or in the comparison of the three actually chosen, a relationship between the extent of renewal and the loss of central city population appears. Not only did the renewal cities lose more of their white \// population; the non-white population that was left behind seemed to be more segregated from the whites within the cen- tral city--another prediction from the literature that appears to be upheld by the data. In Table IX it can be seen that in Washington and St. Louis, the percent Of the Negro popula- tion living in tracts of over 75% non—white occupancy increas— ed; in Cincinnati, this proportion decreased. While the Negro population in all three cities declined in tracts with a lesser percent non-white, Cincinnati had a greater proportion of its Negro population living in such tracts in 1960. While all the growth in the non-white population in the renewal cities was in the most heavily non-white tracts, the growth in Cincinnati was in the tracts with 50-74% non-white. In summary, the null hypothesis regarding differences on selected population variables at the first level of analysis can be rejected for changes in the median family income of 87 both the total population and the non-white population and for changes in the average per tract population. Using the city-wide medians based on census tracts, the null hypothesis could not be rejected for changes in the median education of both the total population and the non-white population or for changes in the percent non-white. When the true medians for education (based on individuals rather than census tracts) are compared, the hypothesis of no difference can be rejected. Regardless of the measures used, the null hypothesis could not be rejected for changes in the percent non-white; however, on two other population variables, important differences between the renewal and the control cities were found. The renewal cities lost considerably more Of their white population and the remaining Negro population became more segregated within the central city. Population and housing variables in non-white occupied tracts Since renewal disproportionately affects the non—white population, one place to test for differences between the experimental and control cities would be in those tracts with sizeable non-white populations. Table XIII gives the same summary statistics for those tracts. Differences between the renewal and control cities appear on five variables: the per— cent of substandard units, Of overcrowded units, of non-white occupied substandard and overcrowded units, and the percent non-white. In the renewal cities, the percent of substandard Table XIIIa. Tracts, 1950 and 1960, and Housing Variables: Summary Statistics for Non-White Occupied on Selected Population Washington, D. C. Standard Means deviations Percent of 1950 1960 1950 1960 Substandard housing units 20.066 12.027 14.482 14.490 Non-white occupied housing units 40.592 61.865 50.156 51.772 Overcrowded housing units 18.507 16.651 8.416 8.951 Non-white over- crowded units 27.968 21.972 9.828 9.929 Non-white occupied substandard units 29.769 12.542 19.905 12.857 . Non-white 57.255 66.849 26.070 29.254 Total population 9,574 6,606 6,719 2,628 Table XIIIb. Tracts, 1950 and 1960, Summary Statistics for Non-white Occupied on Selected Population and Housing Variables: Louis, Missouri Standard Means deviations Percent Of 1950 1960 1950 1960 Substandard housing units 50.072 28.205 28.569 21.721 Non-white occupied housing units 46.154 65.015 54.875 28.686 Overcrowded housing units 50.587 22.860 8.914 8.048 Non-white overcrowded units 54.924 27.675 7.549 7.955 Non-white occupied substandard units 57.565 58.578 27.655 29.899 Non-white 68.895 70.792 25.859 25.720 Total population 10,142 8,158 2,865 5,176 Table XIIIc. Summary Statistics for Non-white Occupied Tracts, 1950 and 1960, on Selected Population and Housing Variables: Cincinnati, Ohio Standard Means deviations Percent Of 1950 1960 1950 1960 Substandard housing units 15.460 54.225 15.978 28.770 Non-white occupied housing units 57.988 55.177 55.056 51.591 Overcrowded housing units 25.492 24.494 11.185 11.178 Non-white over— crowded units 26.618 27.594 10.478 10.291 Non-white occupied substandard units 16.088 55.118 15.755 51.472 Non-white 74.720 65.454 21.244 26.476 Total population 7,609 5,018 2,594 1,544 91 housing units in non—white occupied tracts declined consider- ably, both for the general population and for non-whites, while the reverse was true Of Cincinnati. Similarly, the proportion of overcrowded units--among the general population and among non-whites--went down in the experimental cities and went up in Cincinnati. While the average percent non- white increased in the non-white occupied tracts Of the experimental cities, the average percent non-white in such tracts decreased in the control city. This corresponds to the findings in Table IX Of increased segregation in the renewal cities. NO differences between the renewal and non-renewal cities were found on two variables: changes in the average tract population and the percent of non-white occupied units. In all three cities, the average tract population, in non-white occupied units, declined, suggesting a spreading out of the Negro population within the city. And, in all three cities, in tracts with sizeable non—white populations, the percent Of non-white occupied housing units went up. Stage 2: Census Tract Analypis Housing variables The second level of analysis focuses on the mean changes, iby'census tract, of selected population and housing variables J'J’I‘the three cities (see Tables XIV and XV). Significant dif- felrences in the direction and size of changes in three 92 variables were found between the renewal and non-renewal cities. Tracts in Washington and St. Louis exhibited an average decrease in the percent Of substandard housing units, in the percent of overcrowded housing units and in the per- cent Of overcrowding in non-white occupied housing units; tracts in Cincinnati exhibited average increases on all three variables.72 NO differences were observed between the experimental and control cities in terms Of mean changes in the proportion of non-white occupied housing and in rent.73 Since several tracts in each city lacked rent information in one Of the two years, the mean changes are artificially deflated and the standard deviations artificially inflated in Table XIV; for a more accurate appraisal Of the average change in this varia- ble, by census tract, see Table XVI. In that table the average change is still about the same in the three cities; what is different is the size Of the standard deviation. In the renewal cities, the average deviation around the mean increase in rent was significantly greater (F < .01) than in the control city. This suggests that there were no really 72T-tests of differences between means were significant beyond the .01 level. 73Generally,'maans based on a sample Of medians would not be statistically sensible. However, if we treat the median Contract rent in a particular census tract as a "score" on iflmat variable and are interested in the average change in that SCKDre from 1950 to 1960, then the mean difference Of pairs of mediians makes substantive sense. 95 major shifts in median rents in the census tracts Of Cincinnati that could not be accounted for by inflation; from Figure 1, it can be seen that the tracts are more nor- mally distributed around the mean in the control city and more skewed in the direction of large increases in rent in the experimental cities. In summary, the null hypothesis concerning mean changes in selected housing variables can be rejected for three varia— bles: 1) the percent Of substandard housing, and 2) the percent Of overcrowded housing, both among the general popu- lation, and 5) among non-whites;it cannot be rejected for two variables: median contract rent and the percent of non— white occupied housing units. However, a null hypothesis predicting no differences between the standard deviations around the mean changes in the renewal and control cities can be rejected for one variable: median contract rent. Population variables There are significant differences between the mean changes on four population variables in the renewal and non- renewal cities. The census tracts in the renewal cities lost Significantly more population on the average, gained significantly more non-whites (Table XV) and failed to gain as much in terms Of median family income for both the general jPOpulation and for non-whites only.74 74In all cases, the values of T are significant beyond the . 01 level. 94 In only one case is the difference between the standard deviations of the mean change significant. The tracts in the two renewal cities have much larger deviations around the mean change in number Of non-white residents than is true of Cincinnati. Since Cincinnati and St. Louis experi- enced the same percentage increase in Negro population over the decade, this difference in the variances must be due to the large scale dislocation of Negro population as a result of urban renewal. NO differences were Observed on three variables at this stage Of the analysis: median education, for the total popu- lation and for non-whites only, and the percent non-white (see Table XV)- In summary, the null hypothesis for the second stage Of analysis can be rejected for three housing variables, all three of which were clearly or tentatively rejected at the first stage: the percent of substandard housing, and the percent Of overcrowded housing, both among the general popu- lation and among non-whites.7S The null hypothesis at the second stage can also be rejected for four population varia- bles (total population, non-white population, and median 75The mean and standard deviation of the differences for :non-white occupied substandard housing was computed for tracts with non-white occupants in at least one Of the two jyears; there was an average loss Of Negro-occupied substandard units in the renewal cities and a large average gain in such fuausing in the control city. The null hypothesis for this \Niriable was rejected at stage 1 and, on the basis Of the <fi1mputation just described, can be rejected at stage 2. 95 .mcomaummfioo mma mo amuou m msaMmE .omma ca msoama>a095m mua mo some now 05am> was nua3 Ownmmfioo mmB uomuu omma may mom moam> map .OOEESm on no: Oasou sump mnmnB “mm msomanmmfioo owmanmma mo Hones: may mcame .uass m mm omummnu mnoB muomnu omma amum>mm mo soama>aonsm may mo uso omummuo muomnu Coma mmonu .omfifisw on Oasoo sumo map mnmfiz amo.om www.ma omo.aa aom.m ¢a>.ON ma©.>a moa.ml MOm.ONI mma.NI ema.I >ma.ml mam.mal mwa.ma www.ma mn¢.ma mm>.m www.mm www.ma mma 5mm.m m>>.ONI mms. mmN.¢ mom.ml mmm.s Nom.ma mNO.mN www.ma mw¢.m m>m.¢m mdm.aa *mm «mm.m www.mal aam.m am¢.a mom.>aI mmm.m fl 2 muass photomum lash muasz Icon ucmonwm coauMHsmom amuou .usmu pumuusoo swaps: moans OOO3OHUH0>O teamsooo oua£3 Icon mo unmonmm muass msamsos UOO3OHO In®>o mo unwound muass msamson omamsooo OuaSB Icon mo.usmoumm muass msamsos onwosmquSm mo unwound Gm N Om X Om N aumssaocao madoq -um .U.Q .ooumsanmms mmanmaum> omma pom omma consumm moanmaum> msamsom Omuumamm co mmocmnmmwan m0 mcoauma>wa OHMOGMDm can memo: .>HN manna 96 .mcomaummEoo mma wo anuou m mcaxmfi .owma ca msoama>aUQSm mua mo some now Ozam> mnu £ua3 omnmmfiou mm3 uumnu omma mop How 09am> mnu .OOEESm on Dog Oasoo sumo OHOQB “mm mcomanmmeoo omIomma mo Hogans mzu msaxmfi .uans m on Omummuu mumB muomnu omma amum>mm mo COama>aOQSm map mo use Omummnu muons» omma mmosu .OOEESm on Oasoo sumo mnu mnm£3* u‘ aaa oma *mo u z amo.oa mme.oI mom.am aaa.oI aaa.mm oao.oa ouaszIooe ueoouoe COHHMHDQOQ oea.moma ooo.mmamI moo.mmma oos.aomaI ooo.maaa moo.momaI stoop .osoooa maaEmm cmaomz www.mmma www.mm amo.oama ame.aoa Ooo.mmmm oao.ame eoeuoaseoo ethos omo.mmma mao.maNI moo.aaam www.masI moo.oeem aeo.0smaI coaumaoeoe mua£3lsoc amuoa moo.moaa mom.omoI oma.mmaa moo.mmoI ome.smem www.momaI ouasaIeoeuosoosa haafimw cmaomz aao.m omo.aI mae.m amm.aI os4.m eoe.aI muaszIsoo COHuMUDGO GMHUOZ moo.a mmN.I oas.a mom.I omm.a mam. coaumeoeoo Honor .soaumusow swaps: no x on x on x aumasaocao I masoq .um .U .Q .ooumcanmmz mmanmanm> omma pom omma cmm3umm wmanmanm> coaOMasmom omuomdwm so moocmnmwwaa a0 mSOauMa>mQ onmocmum Ono mommz .>x dance 97 .muau pom manmanm> comm suaB moanm> 2 was ca coaumEnomsa on naaB muomnu Omuamm mnmSB moosmnmmmao some so comma ma waQMu mane .ooumoasaeo soon «>6: oooa uo omoa possum so sea oma u z mam.oaaa omo.oaomI moo.aem mam.eemaI ooa.mao aeo.moeaI ooaumeseou stoop .mEousa waafimw smaomz so ooa maa u z asm.m www.mmI Nma.ma eoo.eNI mme.aa oma.mmI coaumeoeoe Hmuou .HCOH UUMHUCOU GMHCOZ oa ea as u z aeo.omo ooo.amoaI aom.oma mmo.smmI ooa.ooo ecs.memaI mouaszIeoo .mEoosa >aaEMw smaomz ea om mo u 2 mo. omo.I so. aa.I ooa.a moa.I ouassIsoe -- coaumosom swaps: moa mma ama u z moo. amo.I omm. emo.I maa.a eaa. soauoasooe Hmuou soaumosow cmaomz em mm No I z www.ma aem.m eeo.a mae.e mso.m asa.e moans mnemoos pmo3ououw>o mo ucmunmm am on no x on x aumssausau masoq .um .U .Q .sOumcanmmS mmanmanm> *oooa out omoa ommsuom TOHQMHHMNV Umuuwamm CO mOOCOHOMMHQ .HO mCOHuma>OQ UHMUGmum Ucm mflmwz .H>X OHQMB 98 family income for both the total population and non-whites only); on three of these (non-white population is the ex- ception) the null hypothesis was also rejected at Stage 1. The exception is Significant, for while there were no im- portant differences between the added Negro populations in the renewal and control cities as a whole, there were im- portant differences in the shifting around of the Negro popu- lation within each central city. Skewed distributions One other way Of looking at the changes which occurred in the population and housing variables of three cities is to study the distribution of census tracts around the mean change in the value Of a variable (see Figures 1 and 2). Examination of both tables suggests that in the renewal cities there are more extreme changes in pppp directions from the mean, while in the control city the great majority Of the extreme changes are in one direction—-that of increasing the proportion Of substandard, non-white occupied, and overcrowded housing units.76 The same trend appears among the population 76The trend is strongest on the housing variables. 40-45% of the major changes, by census tract, in the renewal cities were in a positive direction; that is, almost as many tracts made dramatic changes in one direction as in the other. In Cincinnati, on the other hand, only 27% of the tracts with significant changes over the decade changed in a positive direction. Number of changes exceeding 1.1 or 2 Standard Deviations of 99 variables, though to a lesser extent. These trends would seem to support the contention that urban renewal is, to some extent, a moving of the slums from one part of the city to another. Major improvements in the quality of housing in one area are accompanied by major deterioration in the qual- ity Of housing elsehwere in the city. The distribution Of tracts around the mean percent Of sub—standard housing units is quite different in the renewal and control cities. In the renewal cities, those tracts which exceeded the let and 2nd standard deviations were more often at the plus end Of the distribution, meaning that more tracts experienced Significant losses in the proportion Of substandard units. On the other hand, those tracts in Cincinnati which exceeded the lst and 2nd standard deviations were all at the minus end of the distribution, meaning that those tracts experienced major increases in substandard housing. Similarly, in the proportion of non-white occupied units, the tracts in all three cities were Skewed to the negative end Of the distribution, but a few tracts in the renewal cities experienced major losses in the proportion Of such Five Housing Variables for Three Cities.* +1 or 2 SD's -1 or 2 SD's St. Louis 29 56 65 Washington, D. C. 54 85 159 Cincinnati 16 57 15 * c The same tract may be counted more than once if it experienced major changes on more than one variable. 100 units (three of the ten in Washington were renewal tracts and three of the five in St. Louis) and only one in Cincinnati did—-the tract which began renewal at the end Of 1959. The two renewal cities, on the average, decreased in the proportion of overcrowded housing, whereas the control city increased slightly. While the standard deviations on this variable were very similar in all three cities, many more tracts in the renewal cities were at the extreme ends Of the distribution (beyond.i 1 or 2 standard deviations). In St. Louis 15 tracts experienced major decreases in overcrowding; 12 major increases (21% of all tracts); in D. C., 17 experi— enced major decreases and 25 major increases (42% of all tracts); but in Cincinnati, eight tracts showed major de— creases and 12 major increases (18% Of all tracts). There are differences in the distribution Of tracts around the mean on several of the population variables as well. The reader may recall that the average increase in median family income is much greater in Cincinnati than in the experimental cities, although the standard deviations are not greatly different. This suggests that the increases in income in the two renewal cities are reduced-~in the averages--because of a number of tracts which changed in the Opposite direction. Data in Figure 2 confirm this. The distribution around the mean in Cincinnati is skewed in the negative direction (toward increases in median family in- come), whereas in Washington and St. Louis the tracts were more normally distributed. 101 .OPQ .soumsanmmz "manmanm> msamson omuomamm so mocmnmwmao some mop Ossonm muomnu msmcmo mo coausnauumao .ma mnsmam muass msamsos, means msa soaumHsmom omozouonm>o moans mcamson Imson Omamsooo muass mcamsoo amuou .ucmn mua£3lsos OOOBOHOHO>O muao3lsos OumosmquSm nomnucoo smaomz mo unwound mo usmonmm mo usmuumm mo unmoumm . I e O FIIL N IN mi IuV 3 Im 1 o T0 1.. _ -oa m w Na n .3 s n ma m Ima 1 S ION Nw aI a+ N+ NI aI a+ N+ NI aI a+ N+ NI aI a+ N+ NI aI a+ N+ msoauma>mn Unmoomwm 102 .ausommaz .masoq .um "moanmaum> msamsos omuomamm so mocmnmmmao coma may ocsoum muomuu mumsmo mo coausnanumaa .Qa onsmam muass msamson muass mSa soaumasmom omo3ononm>o moans mcamson Insos OOHQSUUO muass msamson amuou .usmu mua£BIcoc omozouunm>o mua£3Isoc Unmoomumnsm uomnusoo swaps: mo unwound mo unwound mo unmonmm mo unwonmm N Is Im Im .oa INa .ea Ima Ima .ON NI aI a+ N+ NI aI a+ N+ NI aI a+ N+ NI aI a+ N+ NI aI a+ N+ msOauma>mQ pnmosmum SJSEIL snsueo go JeqmnN 105 oano .aumcoausav "moanmanm> msamson pouOOHOm so mocmnwmmao some on» bosons muumnu momcmo mo coausnanuman .oa mnsmam nuans msamson muasd msamsoz SOaumHsmom OOOBOHOHO>O muass mcamson teamsuoo muass msamsos amuOu .usmn mua£3lsos omo3ononm>o mua£3Isos photomumnsm uomnusoo cmaomz mo ucmonmm mo unwound mo usmonmm mo unmonmm L _ o MJ’ILIII— 1N m .uv m. Im O .m J 3 .0a a w .Na n 3% s n .ma 3 9 .ON NI aI a+ N+ NI aI a+ N+ NI aI a+ N+ NI aI a+ N+ NI aI a+ N+ msoauma>ma Unmocmum 104 Standard Deviations +2 +1 —1 ~2- +2 +1 —1 —2 +2 +1.-1 —2 +2 +1 -1 -2 20. 18. 16. 4‘3 14. L) :3 Ed 12- :9, 10. m 5 8. LL— (J “5' 5‘ a 4. g 2 O l— I E Median edu- Median edu- Median Median family cation,total cation, non- family in- income,total population whites come, non- population whites Standard vaiations +2 +1 -1 -2 +2 +1 -1 —2 +2 +1 -1 -2 ZOI 18~ 16~ ..‘3 (J 14‘ rd 6:. mi 23 10- 2:” on L3 8 <3 5 as 4. .r .Q g 2‘ =2 0 E‘igure 2a. Non-white population Total popu- Percent non- tion white Distribution Of census tracts around the mean difference on selected population variables: Washington, D.C. 105 Standard Deviations +2 +1 -1 -2 +2 +1 -1 —2 +2 +1 -1 —2 +2 +1 -1 —2 20- 18* 16' 14‘ 12‘ 10‘ I 6‘1 4- 2* T‘s fl Median edu- Median edu- Median family Median family cation,total cation, non- income, non- income,total population whites whites population Number of Census Tracts Standard Deviations +2 +1 —1 —2 +2 +1 -1 -2 +2 +1 -1 -2 20- 18‘ 16. 14. 12. .... 10. 8 .. 6. 41 Number 01‘ Census Tracts 2. 0 Non-white Total popu- - Percent non- population lation white Figure 2b. Distribution Of census tracts around the mean difference on selected population variables: St. Louis, Missouri . Number of Census Tracts Number Of Census Tracts Figure 2c. 106 Standard Deviations +2 +1 —1 -2 +2 +1 -1 -2 +2 +1 —1 —2 +2 +1 -1 —2 20. 18* 164 14‘ 12' 10‘ 84 5. 4. 2. 0 m —1 Median edu— Median edu- Median family Median family cation,total cation,non- income,non- income,total population whites whites population Standard Deviations 20.+2 +1 -1 -2 +2 +1 -1 -2 +2 +1 -1 -2 18< 16I 141 12- 10‘ 8. 6I 4. 2. 0 Non-white population lation Total popu- Percent non- white Distribution Of census tracts around the mean difference on selected population variables: Cincinnati, Ohio. 107 Increases in non—white population was typical Of all three cities. The average increase was unusually large for the District and larger in the two renewal cities than in the control city. The standard deviation for changes in the non-white population was quite large for D. C., and much larger in both D. C. and St. Louis than in Cincinnati. The renewal tracts in the experimental cities were prominent among those tracts which lost large numbers of Negro resi- dents: in St. Louis five Of the six tracts with changes greater than -1 standard deviation were renewal tracts; in D. C. two of the five were. More of the tracts experienced extreme gains in non-white population in the renewal cities; fewer of those in Cincinnati did (15% in St. Louis, 20% in the.DiStrict, 10% in Cincinnati). At the same time that the average tract in the three cities was gaining in non-white population, the average tract was losing in terms of overall population. Again, the re— newal cities were more similar to each other than they were to the control city: both St. Louis and Washington lost Iconsiderably more population per tract than did Cincinnati. IVaShington, again, experienced a wider variety Of and more extreme changes, as indicated by the larger standard devia- ‘tion. The two renewal cities differed from the control city in that the distributions of the tracts in the former Vfixre much more skewed toward losses Of population, whereas time Cincinnati tracts were about evenly Split; all Of the 108 renewal tracts in the experimental cities, save one, were included in those tracts which suffered major losses of population. CHAPTER V THE RENEWAL TRACTS AND THE PATTERN OF RELOCATION While the inferences concerning the effects of urban re- newal on changes in the city as a whole are somewhat indirect, those changes in the census tracts affected can be Observed much more directly. This chapter will, first, describe the major changes which occurred in those tracts identified by each city's urban renewal agency as undergoing some renewal effort during the decade, and second, attempt to document the pattern of invasion and succession which occurred in city neighborhoods as a result Of the massive dislocation Of families in the renewal areas.77 Stage 5: The Renewal Tracts §pmmation of changes in all renewal tractg In both Washington and St. Louis, the renewal tracts are clustered together on the fringes of the central business Iiistrict in the central part of each city. Table XVII shows tile overall changes which occurred in that part of each city; 77These traCts were identified in letters to the author frcnn staff members in the Urban Renewal agencies in the three Cities. In Washington, D..C., the tracts were: numbers 64, 65. 66, and 67; in St. Louis: numbers 94, 102, 105, 106, 107, 115. 114, and 120; and in Cincinnati: number 2. 109 110 the data in the table are based on a summation Of information from all the tracts involved in renewal. In both cities, the proportion Of substandard housing in the renewal tracts de- clined, although the relative decrease was not great in St. Louis: two-thirds Of the 11,800 remaining housing units in St. Louis' renewal tracts were still substandard. Similarly, the proportion Of overcrowded housing units decreased. In the renewal tracts Of both cities, the proportion Of non-white occupied housing units decreased at a time when the city-wide trend was in the opposite direction. Overcrowding remained fairly high and even increased slightly among non-white occupied units in the District. And the vacancy rate in- creased considerably in the renewal tracts in both cities. With respect to changes in the characteristics of the population of the renewal tracts--in both renewal cities, the size Of the population in the renewal tracts declined, 85% in D. C. and 55% in St. Louis. In D. C. the proportion non- white was exactly the same, before and after renewal; in St. Louis, the prOportion declined from 54% to 42% in 1960. The average education Of whites and non-whites in these tracts .increased—-by a great deal in the District and more modestly in.St. Louis. This, tOO, was in opposition to the overall 'tremd in Washington where the average education for the city decflined. Rents, on the average, doubled in St. Louis and almost tripled in D. C. Gains in median family income were Iflflnarkably slight; only in St. Louis did the increase in the 111 median income for the total population in the renewal tracts approximate that for the city as a whole. Description of changes in each renewal tract Each of the renewal tracts differed somewhat in the types of changes which occurred over the decade since in some, pub- lic housing supplanted the Older structures, in others luxury or middle income housing was rebuilt in their place, and in some no rebuilding had yet occurred at the time of the Census. In the following paragraphs, the detailed changes in each re- newal tract will be described. The reader will recall that a "major" change is defined as one which is greater than 1.1 or 2 standard deviations from the mean difference Of a given variable for a particular city. For example, all four renewal tracts in Washington, D. C. experienced a major decrease in the total population of each tract. Census tracts 64 and 65 were over two standard deviations from the average change in size of population, on a census tract basis, for the District; census tracts 66 and 67 were over one standard deviation from this average change.78 In the District, tract 64 experienced major changes on popu- lation and housing variables; major Shifts occurred on only 78This was a decline, on the average, of 428 persons per 1:ract. The mean difference is positive because the 1960 figure is subtracted from the 1950 figure. Census tract 64 Lost 7860 people, number 65 declined by 5976, number 66 lost all bmt 99 Of a 1950 population of 4440 and number 67 went from 5455 to 2560. 112 Table XVIIa. Summation Of Changes in Washington, D. C. * Renewal Tracts: Variables 1950 1960 Percent Of substandard 2949 _ 125 = housing units 6073 ‘ 49% 1216 10% Percent of non—white 5524 _ 557 = occupied units 5878 _ 60% 1105 49% Percent Of overcrowded 1796 _ 267 = units 5878 ' 31% 1103 24% Percent Of non-white occupied 1287 = 57% 252 = 43% overcrowded units 5524 557 Percent Of all occupied 5878 = 1105 = 0 housing units 6073 96'79% 1216 90°717 Total population 25,490 4218 Negro population 16,264 2662 "Median Of medians" Education 8.1 10.8 Income non-white 2071 2948 Education non-white 7.0 8.9 Income 2169 2869 Rent 54.5 92 *- Tracts 64-67. 115 Table XVIIb. Summation Of Changes in Renewal Tracts:* St. Louis, Missouri Variables 1950 1960 Percent of substandard 15264 = 78% 7854 = 67% housing units 19575 11800 Percent of non-white 9725 _ 5869 = occupied units 19065 _ 51% 10444 37% Percent Of overcrowded 7210 _ 2919 _ units 19063 ‘ 58% 10444 ‘ 28% Percent Of non-white occupied 5802 =,39% 1185 = 51% overcrowded units 9725 5869 Percent Of all occupied 19065 = 10444 = housing units 19575 97°59% 11800 88°51% Total population 69,807 51,427 Negro population 57,592 15,084 "Median Of medians" Education 7.6 8.0 Income non-white 1568 1582 Education non-white 7.1 7.4 Income 1592 2502 Rent 15.0 54.0 ‘*Traots 94, 102, 105, 106, 107, 113, 114, 120. 114 three or four variables in tract 65-67. Washington, D. C. Census tract 64 lost 82% of its 1950 population. In 1960, the tract was 62% Negro, as Opposed to 95% non-white in 1950, and the socioeconomic status Of the area had risen considerably: the median family income had increased (over one standard deviation from the average change) from $1859 to $4817; the average education went up (over two standard deviation from the average change) from 7.1 to 12.1; and the median contract rent changed from $27 tO $127 a month. The increased socioeconomic level Of the 1960 population is apparently due to the new white residents, most Of whom moved in between 1958-60. There are several reasons for this in— ference: 1) the non-white median family income only changed from $1779 to $2764; 2) the non-white median education, while Showing a substantial increase (up one standard deviation), went only from 6.9 to 9.1, nowhere near the median for the tract as a whole; and 5) the white residents are "new" be- cause none Of them over the age Of five lived in the same house in 1955; 44% moved there from another house within the city and the other 56% moved there from outside the city. On “the other hand, 76% Of the non-white population in the tract Inoved, since 1955, from a different house in the central city; cnily 6% Of the non-white population over five in the tract lived in the same house in 1955. For those who moved from annother house in the District, there is no way of knowing vfluether they moved from another house in the renewal tract or frtun elsewhere in the District. 115 In terms of housing, the number of units in tract number 64 decreased from 2572 to 622. There was a significant de- crease in substandard housing (over two standard deviations from the mean); 60.8% Of the units in 1950 were substandard; in 1960, only 10.5% were. There was a significant decline in the percent Of non-white occupied housing units (beyond the second deviation) and in the percent of overcrowded housing units (beyond the first standard deviation), but there was a major increase in the percent of non-white overcrowded units (beyond the second standard deviation). While Negroes con- stituted 62% of the 1960 population, they lived in only 35% cof the occupied housing units and 85% Of the overcrowded units \Mere occupied by non-whites. But, the vacancy rate was a (zomfortable 17% in 1960, suggesting that most Of the housing aavailable was Open to whites in the upper income brackets. Census tract number 65 lost all but 1% of its population cover the decade. In 1950 it had been a predominantly white Eirea (62%); in 1960, 42 of the 60 people living in the tract Imere Negro. Because Of the small number of residents in 1£360, no data were available on income, education, or rent. Tfliere were only ten housing units in the tract in 1960, a{bparently "left overs" from before redevelopment since five CNE them were classified as substandard. Tract number 65, 'tfierefore, was between the demolition and construction stages (Di? the renewal process‘at the time of the census SO that no c3C>mparisons can be made between Old and new residents. All of 116 the Old residents were dislocated. The Old residents were Of lower education (the median years Of school for the tract was 8.1, for non-whites, 7.0), and low income (the median family income for the total population was $2414, for non- whites, $2185), and their average rent was relatively low ($57 a month). Half of the housing units in 1950 were sub- standard, 50% overcrowded. Tract number 65, then, was a_ deteriorated neighborhood Of lower income whites and non- whites, fairly typical of those frequently chosen for renewal. Tract number 66 lost almost all Of its 1950 population. Like number 65, it was a predominantly white (61%), lower socio-economic status neighborhood where the average education ‘was 8.5 years (7.1 for Negroes), the average family income, $2490 ($2026 for Negroes), and the average rent, $59 a month. In 1950, 28% Of the housing units were overcrowded (40% Of those occupied by non-whites), and 56% were substandard. .Also like number 65, number 66 was in the "in-between" stages (of renewal at the time of the census. The few remaining :families had moved to the area from elsewhere in the central <:ity sometime between 1954 and 1960 (81.8%); a fifth had Slived there prior to 1954. ' Tract 67 experienced a major decrease in total population Eindsubstandard housing (from 59% in 1950 to 8% in 1960); it lxsst 50% of its housing units and 57% Of its people. Over- crrowding decreased both among whites and non-whites. The IENDpulation in 1950 was low in education (8.0 average for the 117 population, 6.8 for Negroes), low income ($1924 for the popu— lation, $2156 for Negroes), and low in median rent ($52 a- month). The population was 58% Negro in 1950, 64% in 1960. In both 1950 and 1960 the Negro population had a higher family income than the whites. The increase in family income, however, was below the average for all census tracts. The increase in the level of education (among white and non-white), on the other hand, was significant (over one standard devi- ation from the mean change). The 1960 population does not appear to be radically different from the 1950 population. But there has been considerable shifting Since four-fifths <>f the occupied units were taken over by new tenants within ‘the two years preceding the census, 52% of the residents Inoved in from another house in the central city, 9% had been living in the same house since 1955, and the remaining 58% (came from outside the city. St. Louis, Missouri. In St. Louis, six tracts (94, 102, 1105, 106, 107, and 120) began to participate in the renewal Exrogram well in advance of the census; two others (115, 114), Emitered the execution stage in December 1959. The only Huajor change which occurred in these tracts was that they lczst a large part of their population. The tract 115 popu- laition declined from a 1950 total Of 18,492 to 12,749 in 1960, ‘a- 51% drop; tract 114 lost over half (55%) of its 1950 popu- lation Of 6, 195. 118 Both tracts had a white population (115: 99%; 114: 85%) that was low in education (115: 8.1, 114: 7.6), median family income (115: $2571; 114: $1982), and rent ($15 and $11 a month, respectively). A very high proportion of the housing units in both areas were classified as substandard (113: 80.9%; 114: 89.8%) and a little over a third in each tract were overcrowded. In both tracts, the population was more "mobile" than the average (median) for the city: a greater than average percent of the population over one year Old had lived elsewhere in the city a year preceding the census (10.1% in 115, 11.9 in 114) and a smaller than average percent had lived in the same house in 1949 (84.9% in 115, 82.0% in 114). In Washington, D. C., the reverse was true. .All of the renewal tracts were above the median in terms of the percent of the population living in the same house at the time of the census and one year preceding the census. Tract 94 was made up Of a non-white population (98% in 1950) and remained 90% nondwhite in 1960. It declined from Ia population of 4475 in 1950 to only 259 in 1960. This was El “left over” population, not newcomers to a rebuilt area, 53ince 60% had been living there at least five years at the trime Of the census. The population was similar in character- jmstics to the other renewal tracts—-very low in education ('7.0), income ($1510) and rent ($16). The area declined sig— rLificantly in the proportion of substandard (from 77% to 41%) ialud overcrowded (from 40.4% to 25.5%) housing units. There ‘Vfare 112 housing units remaining in 1960, a fourth Of them 119 vacant. This suggests that the relocation and demolition process was still going on at the time Of the census. Tract 102 was also a largely Negro tract (95%) with a very large population in 1950 (11,648) which was more than halved by 1960 (4,926).79 The population was a poor one ($1582 in 1950; $1515 in 1960) and was Significant in that it actually declined in median income over the decade (over one standard deviation from the mean change). This decline was apparently due to some poor whites living in the tract, since the median income for non-whites increased Slightly (to $1425). The 1960 residents were very similar to those in 1950 in terms of their lack Of education, their low salaries and the cheapness Of their housing. Overcrowding declined consider- ably Over the decade (from 58% to 25%), but the percent Of substandard housing remained high, declining from 81% to 70%. In 1960, there were 1758 housing units in the tract, a de- crease from the 2964 units in 1950. There is the possibility 'that some Of these may be new low income units because 40% (of the units in the tract were occupied between 1958-60-- ‘Nhich is well above the median for St. Louis. Tract 105 was a predominantly Negro area (87% in 1950) Vflhich lost about 70% of its 1950 population (over two standard deviations from the mean change both in its total population and in the number Of Negro residents). The educational level ‘ 79The decrease in total population and in Negro popu- liation were both beyond the second standard deviation from the Infiaan. 120 Of the residents was low in 1950 (7.2), even lower in 1960 (6.7). The median family income went up slightly for the area over the decade (from $1157 to $1525),80 rents more than doubled (from $16 to $57); but the median family incomes Of non-whites went down to $958, the lowest, by far, Of any tract in St. Louis. In 1950, 100% of the housing in this tract was classified by the census as substandard. There was a major decrease in such housing over the decade (over one standard deviation from the mean change), but in 1960 77% of the units were still substandard. The standard housing must therefore either be 116W housing or renovated Older structures; and since renova- ‘tion is expensive and the population did not change in char- aacter over the decade, the suggestion is that new low income llousing was built to replace the Older structures. There was El significant decline in the proportion of the housing units (occupied by non-whites (from 92.6% in 1950 to 65.7% in 1960). ffhere was also a major decline in the proportion Of over— czrowded units--from 57% in 1950 to 20% in 1960. It appears JErchthe data that the new or renovated housing went chiefly tc> white occupants Since the proportion Of substandard units among non-white occupied units was 90.7% in 1960. In both .1s350 and 1960, tract 105 received more Of its residents from OLl'tside the city than was typical of census tracts in the city. 80A significantly small change when the average in- CErWease for the census tracts was an increase of $1808 in median family income. 121 This suggests that the neighborhood may be a "receiving“ area for new migrants from rural areas; such an interpre- tation is strengthened by the extremely low educational and income level of the population in both 1950 and 1960. Tract 106 had a 1950 population Of 12,186 which was halved by 1960. Sixty-three percent of the earlier popula- tion was Negro, 75% of the later.81 The people in this tract were not at the bottom though they were relatively low in education (7.8), income ($1815) and rent ($17). The 1960 population differed very little, on the average, from the 1950 population. The average income went up about $700, the average education only .2 Of a point. Three-fourths of the housing in 106 was substandard in 1950, a figure which declined significantly by 1960 tO 48%. Overcrowding was higher in this tract (59.1%) than in all but six other tracts in St. Louis. In 1960, it graduated to the second most overcrowded tract (55.1). People were still Inoving into housing in the tract in the few years preceding ‘the census (40% of the units were occupied by new residents lbetween 1958-60), though a third of the units were occupied lsy'people who had lived there prior to 1954. Tract 107 was also a predominantly Negro tract (91% in 3.950) which had lost all but 19% Of its 1950 population by 81The size of the tract's loss Of population and of its Ifieegro population Specifically was greater than two and one Sftandard deviations, respectively. 122 1960. The educational level of the tract was low and decreas- ed slightly Over the decade. The average family income was low ($1465) and hardly changed at all over the decade ($1550 in 1960). The income of Negroes in the tract in 1960 was higher than that of whites ($1740). Rent was lower in this tract in 1960 than in any other tract in St. Louis. Practically all of the housing in the tract was substandard, both in 1950 and 1960 (94% and 90% respectively). Overcrowd- ing was higher than 106 in 1950 and was still way above the - median in 1960.82 Since 19% of the remaining units were vacant, this tract could still have been in the relocation and demolition phases of renewal at the time of the census. Zilmost half Of the housing units in the tract were occupied 13y the residents prior to 1954. Tract 120 had a larger white population (64%) in 1950 ‘than in 1960 (55%), and a relatively small base population (5082), Of which it lost about two-thirds by 1960. The edu- <2ational level of this tract was higher than any of the. (Ither renewal tracts in St. Louis (8.5 in 1950, 9.6 in 1960) arui increased to a 1960 level that was above the median for aflLl tracts in the city. The median income went up, but was Stiill considerably below the median for the city. The median Ireint also increased significantly (over one standard deviation fifcmlthe mean change in median rent) to $56 a month, much 82Though the decline in overcrowding was above the first S t andard deviation . 125 higher than the 1960 rents in any Of the other renewal tracts. In 1950, 89% of the housing was substandard, in 1960, 62%. A third of the units in 1950 were overcrowded but only 9% were in 1960, a major decrease (over two standard devia- tions) in overcrowding, both among non-whites and the popula- tion as a whole. However, in 1960 21% of non-white occupied housing was still overcrowded.83 This tract also experienced a major decline in the percent Of housing units occupied by non-whites (from 66% to 25%). The tract was unusual in its mobility prior to the 1950 census. Nineteen percent of the residents had lived elsewhere in the city in 1949 (well above the median for St. Louis census tracts) and only 58% were living in the same house (well below the median for the city). This meant that 25% of the 1950 population had come to the area from outside the city. The same thing was true Of the tract in 1960--a higher than average proportion had moved to their 1960 residence from another house within the central city in the last five years preceding the census and only 25% (compared with a median of 47.6%) had lived in the same house in 1955. The trend for residential mobility with- in the city was accentuated among the non-white residents of the tract. On the other hand, the same proportion or more 83This is a consequence Of the fact that while the number of housing units increased the number Of nOn-white occupied unéts decreased, yet the non-white population declined only 55 . 124 of the non-white residents lived in the same house in 1949 or 1955. The difference, therefore, is that more Of the white residents Of the tract apparently moved to the area from outside the central city. Cincinnati, Ohio. Cincinnati's tract 2 was the only tract in that city affected by urban renewal during the 1950's. Like many of the other renewal tracts it was predominantly Negro (80% in 1950), declined significantly in Negro popu- lation (over two standard deviations), in the proportion non- white (58% in 1960), and in total population (from 7955 to 496). Like the other tracts, its population was low on edu— cation (7.5), income ($1568) and median rent ($16 a month). The tract was above average in overcrowding (55.8% in 1950; 52.1% in 1960) and, percentage-wise, changed little on this characteristic over the decade. A 21% vacancy rate sug- gests that clearance was still in progress. This is further supported by the fact that all Of the remaining housing was classified as substandard (whereas only 15% was in 1950).84 Summary of changes in renewal tracts. Major changes in three Of the variables were consistently associated with the tracts affected by renewal; a decrease in overcrowding, loss of population, and loss of non-white population. Three Of the four renewal tracts in Washington, D. C., and five of the eight in St. Louis, all experienced a major decrease in g 84There were 12 other tracts with a higher percent of Substandard housing, only five others with as large a percent of non-white. 125 the percent of overcrowded housing units in the tracts.85 All of the tracts in the District, seven Of the eight in St. Louis and the one in Cincinnati sustained a major de- crease in the size of the tract's population.86 In St. Louis, particularly (five out of eight tracts), in D. C. to a lesser extent (two out of four tracts) and in Cincinnati, there was a major decrease in the non-white population of the renewal tracts.87 In all three cities the renewal tracts did not necessarily represent the most dilapidated tracts in the city. In two tracts in the District, in the one in Cincinnati, and in only three of the eight in St. Louis was there a sizeable decline in the percent of substandard housing. Patterns Of Relocation What has happened to all Of the people diSplaced by re— newal? The renewal tracts in the District lost 21,272 persons, those in St. Louis, 58,580. If others who have studied renewal are right, most Of these people did not leave the metropolitan area, probably most did not even leave the central city. If Schorr is right, they were within a mile Of their former home. In this study it will be impossible 85In five of the eight tracts, the change was over two standard deviations from the mean. 86In nine of the 12, the change was over two standard + 0.0m 0.0m 0.0m m.mN N.0m Icos mo soaumafism muomuu amzwomu a.0a a.ma 0.0m n.0a m.ma m.mm N.¢m 0.0+ +.> mo coaumEEsm coaumHsmom ODHSBICOG ca maamm HOnmE £ua3 m.0a 0.0a N.>a >.Nm 0.0m m.m+ 0.am 0.mm 0.0a muumnu mo coaumfifism soauMasmom asuou ca mcamm Momma nuaB +.0a a.m s.+N 0.0m «.wa 0.0m 0.am m.s¢ 0.0m muomuu mo coaumafidm muomuu mamsmo aam How 0.ma m.ma 0.+N N.mm m.mm 0.0m N.s¢ 0.s+ N.mm smaome OUHBImuaU aumo masoq coudca aums masoq COuwsa and: masoq coumsa Icaocao .um Iomm3 Iswuoau .um. Inmmz Isausao .um Ifimmz muao amnucwu on“ INUaO amnucmo .mmsom 00amu50 Eoum usmunmm ucmnmmman ca unmoumm mmsom 05mm ca ucmunmm Ommalmmma .noauao amnocoo mouse caches muaaanoz Hmaucmcemmm .HHH>N magma 152 the city as a whole, a much smaller proportion of new resi- dents from elsewhere in the city than was typical, and a pggp greater proportion of new residents from outside the city. It would appear that the tracts which grew most rapidly in Cincinnati did so as a result Of migration to the city from outside whereas the comparable tracts in St. Louis grew more as a result of internal migration within the city. The diSplaced residents of the renewal areas are still, in a sense, "at large" in the city. One other method was used in an attempt to uncover them. Earlier in this chapter, it was pointed out that the renewal tracts experienced major changes on a number Of variables: in addition to Significant losses of population, the proportion Of overcrowded and sub- standard housing decreased, and the character of the pOpu- lation often changed. Since the people who were relocated would still be poor, they would require housing which was inexpensive and marginal in quality, and overcrowding might be necessary in order to Obtain sufficient housing at a real- istic price. Consequently, the “receiving" areas would prob- ably change in many ways in response to the new population. In an effort to pick out these potential "receiving" areas, the author selected all Of the tracts which experienced "major" changes in four or more variables over the decade.90 90In the District, three Of the four renewal tracts were included in this list and in St. Louis five Of the eight were. 155 The renewal tracts which had experienced four major changes or more were grouped together with the rest Of the renewal tracts (see Table XVII) and the remaining non-renewal tracts were grouped together. Then some Of the basic data for the tracts within each group were summed. The summations for the non-renewal tracts can be seen in Table XIX. This third group Of tracts differs from the first two: 1) in having a much higher proportion Of substandard housing in 1950 and 1960; 2) (with the exception of St. Louis in 1950) in having a higher proportion of non—white occupied housing units in 1950 and 1960; 5) in having a lower median education in both 1950 and 1960 (in St. Louis all three sets of tracts had the same median education in 1950); and 4) in having a lower median income in both 1950 and 1960. These "non-renewal" tracts also had a higher than average proportion of overcrowded housing, particularly among non- whites--roughly a third Of non—white occupied housing in these tracts in all three cities was overcrowded. In terms Of the character Of the housing and the people, these tracts are more likely to be the recipients of the re- locatees than either Of the preceding groups. The data in Table XVIII provide further support for this interpretation. A smaller than average proportion in the non-renewal tracts lived in the same house in 1955 and 1960; and a larger than average proportion moved there from another house in the central city. In St. Louis fewer Old residents remained in 154 Table XIXa. Summation of Changes in the "Non-renewal" Tracts:* Washington, D. C. Variables 1950 1960 Percent Of substandard 9529 _ 6259 _ housing units 59015 — 16°1% 65576 _ 9'8% Percent of non-white 18557 _ 55805 = occupied units 57418 ‘ 52°5% 61138 55'5% Percent of overcrowded 9996 _ 10055 _ units 57418 ‘ 17'4% 61138 ‘ 16°4% Percent of non-white occupied 5865 _ 8248 = overcrowded units 18557 ” 51'6% 33803 24°4% Percent of all occupied 57418 = 61158 = housing units 59015 97°29% 63576 95'16% Total population 242161 200740 Negro population 84071 129006 "Median of medians" Education 12.1 10.6 Income non-white 2400 5890 Education non—white 8.8 9.9 Income 5011 4567 Rent 55 71 *Tracts 1,2,25,24,50,56,47,48,55,55,57,59-65,68,72,75-77,81 84,85,87,92,95,95,96,98,104,105,112,122,125. .The "non- renewal" tracts are those which experienced major changes on four or more variables. 155 Table XIXb. Summation of Changes in the "Non-renewal" Tracts:* St. Louis, Missouri Variables 1950 1960 Percent of substandard 8625 _ 5456 = housing units 31207 ‘ 27'6% 31411 17'4% Percent of non-white 5844 _ 20274 _ occupied units 50657 _ 19°1% 29149 _ 69'6% Percent of overcrowded 6209 _ 7408 = units 30657 ‘ 20'5% 29149 25'4% Percent of non-white occupied 2076 = 6255 = overcrowded units 5844 55°5% 20274 30°8% Percent of all occupied 50657 = 29149 = housing units 51207 98°25% 51411 92'8% Total population 102402 105258 Negro population 21919 78875 "Median of medians" Education 8.9 8.9 Income non-white 1720 4140 Education non-white 8.8 9.5 Income 5058 4025 Rent 55 59 *Tracts 27-31,33,35,56,57,98,104. variables. The "non-renewal" tracts are those which experienced major changes on four or more 156 Table XIXc. Summation Of Changes in the "Non—renewal" Tracts:* Cincinnati, Ohio Variables 1950 1960 Percent Of substandard 9525 _ 7485 = housing units 26491 ‘ 55'2% 26378 28'4% Percent of non-white 5401 _ 15150 = occupied units 25775 ‘ 21°O% 24825 52'9% Percent of overcrowded 5970 _ 5218 _ units 25775 ‘ 25'2% 24823 ‘ 21'O% Percent Of non-white occupied 1798 = 5265 = overcrowded units 5401 53°5% 15150 24'9% Percent Of all occupied 25775 = 24825 = housing units 26491 97'29% 26578 94°1O% Total population 86120 74959 Negro population 19209 48228 "Median Of medians" Education 9.5 9.2 Income non-white 1454 4752 Education non-white 8.4 9.5 Income 2870 4195 Rent 51 54 *Tracts 5,6-9,14,58-41,66,67,69,77,80,87. The "non-renewal“ tracts are those which experienced major changes on four or more variables. 157 1960 and more new residents moved there from elsewhere in the city than was true for any other group Of tracts. More can be learned about these tracts which experienced four or more major changes over the decade by comparing them with the renewal tracts in Washington and St. Louis. The re- newal and non—renewal tracts were different in 1950, under— went different kinds of changes over the decade and were not as different from each other in 1960 as they had been in 1950. The most Obvious differences between the renewal and-non- renewal tracts in 1950 were: 1) considerably more of the housing in the renewal tracts was substandard; 2) many more Of the housing units in the renewal tract were occupied by Negroes; 5) more Of the units in the renewal tracts were over— crowded; 4) Over half of the population in the renewal tracts in 1950 was Negro whereas a third or less of the pOpula- tiOn in the non-renewal‘tracts was non-white; 5) the population in the renewal tracts in 1950 was of substantially lower education, and income, and they paid, on the average, much lower rent. The types of changes which occurred in the renewal and non- renewal tracts were quite different. Some of the important differences in the types of changes which occurred between 1950 and 1960 were: 1) the number of dwelling units in the renewal tracts decreased by 55-80%, while those in the non-renewal tracts increased slightly; 2) the percent of non-white occupied units decreased in the renewal tracts, and increased considerably in the non-renewal tracts; 5) 4) 5) 6) 158 the percent of overcrowded units decreased in the re- newal tracts and increased or remained about the same in the non-renewal tracts; the population decreased drastically in the renewal tracts and either increased or remained the same in the non-renewal tracts; the non-white population decreased both numerically and proportionately in the renewal tracts and increased numerically and proportionately in the non-renewal tracts. In the District, the 1960 non-white popula- tiOn (If the non-renewal tracts were half again as large as the 1950 population; in St. Louis, the 1960 Negro population was over two and one-half times greater than the 1950 non-white population; with the substantial increase in non-white population in the non-renewal tracts (and the substantial loss of a similar population from the renewal tracts) it is not surprising to find that the population of the former tracts decreased (or remained the same) in terms of their average education, whereas the educa- tional level in the latter tracts went up overall. Furthermore, while the median rent in the renewal tracts doubled and almost tripled, the median rents in the non-renewal tracts went up by a third to three-fourths. The tracts in Cincinnati which experienced four or more major changes between 1950 and 1960 and which were not involved in renewal activities during the decade resembled the non- renewal tracts in St. Louis and Washington in several ways: 1) while a relatively large proportion of the housing 2) 5) 4) units in Cincinnati's non-renewal tracts were sub- standard the proportion was much smaller than in the renewal tracts of the two experimental cities; the proportion of non-white occupied housing units increased considerably over the decade; the number of overcrowded units occupied by non- whites increased in the non-renewal tracts in all three cities while the number of such units in the renewal tracts declined; the number of non-white residents more than doubled in the Cincinnati tracts and the proportion non-white increased to roughly 64%. 159 Where Cincinnati‘s non-renewal tracts differed from the ex- perimental cities (particularly St. Louis) was in the type of migration patterns within and to the city. More of the pOpu— lation remained in the same house in Cincinnati than in St. Louis; fewer moved to the non-renewal tracts from elsewhere in the central city than was true for St. Louis; and more migrated to these tracts from outside the city than was true for St. Louis. When the three groups of tracts--those experiencing major gains or losses in total or non-white population and those experiencing major changes in more than four variables--are plotted on census tract maps of the three cities, some patterns begin to emerge. Many of these patterns are typical of all three cities, and many Of them are typical of St. Louis and Cincinnati, apart from Washington, suggesting, first Of all, a fairly circumscribed impact for programs such as renewal, and secondly, remarkable similarity in patterns Of change for large American cities. The first and more Obvious pattern is that Of major popu- lation loss in the cores of the three cities and major popu- lation growth at the peripheries. This substantiates ,McKenzie's and Thompson‘s Observations, though it does not provide an answer tO the reasons behind this phenomenon. In all three cities, some Of the core tracts which lost popu- lation were the renewal'tracts; but, also in all three, many were not. SO renewal is not the answer. For the cities .0 ____~p_*~““““_fifl 140 mo ¢ A\Av muumnu eam30swnIsoc= soaumHsmom wua£3lsoc :a AIL mommOH pen A+v meamm Menus coaumHsmom uumuu amuou :a any mumm0a tam va meamm Momma "hex an: .0 .Q .ZOEUZHmm55 vm TWP. (”J 120 ’ 4 ’1 I“ A" so. 7: 3:1: I 4 see page 140 ‘ \‘ a .o I 0 I . L! H a. \ .0. I n c a I V. . .. a u 2.. .: s e I . u 2 u . 5 I A 3 4 _ 3 I K ' o . 2 I . = . no.3“ V i - ... a’ c u . p l c z = a c ‘0 T a e w = ., m a s M m 2 1 .fl m ......S- N c u . I . : S a 22:... .5 ....m .3. n m A “a | ~ u a.“ m J (I ...:u. ..a 1.7.: ...-...: .... nuspm . I ' ... :— l 1:.- I . . s 94% I a . I ® I w w ...: I I I” 7 u :- l I D . ... C n u.- .ovoo: J. n u 4...: .9- :— I. . I A ' \ B nu . e.) . I. s ....... .n...n.fl.:. m. nos-”I... u 6 B v. H A I .. s u . . o 7 I "I . I I r: ..u. ....x 9 . I ... 3 u I u 4 o n.» n m. ..».H-.: J 2 3 m A I "I . m. a m l e :1. ...e... mu . ...I 4 A: u i .. A E. I I .. ... ... c _. . sir. . s . . . . u I»: s u I p u ‘ a 3 =- . v *- uv I .. "was: .... 5. I .... .. ...:- c . I a a u... D "Her-.- 0.» 5 I Eu: & 1 “ D 3 A a 3 \v . I" u o a l.- u o. . e . I on 4 5: A N3 I ... :54...- .5 II n 4 In "I . s c 0 . “ ...-U.S.; M H E.- .l .u huh! I! ‘ I. n 3 A. s flllln .. I . o I- A I. . I\IaI e . a. . n O . I I I I w 1! I .. _ . .. m V — I _ 4.... . 7 n n ... 2 6 «I .IL 5 ’ 4 w I I :3 3 .II D .1... m 5 . l o I 9 142 . It. ~.. 0 .. - ~: 04 - D I c-s rm. '5 I a, lumen- I'IIYIIII IIYCIII$0I (Illllll " Inn-'8 nu: V ll" lou see page 140 Map Key: s. o . H I. i" r A -. ' :0" IV ’ . k a raw .. I :0! 0' €- " U. s a, z ' - ' U I o 0‘_ O \ a o ‘. :2 o. :12 g k v '- '( g ‘ V‘IJ - ‘ A .0 o p I 2 79 o - l I. B7 VlfiII i:- ‘c ‘ - 1- 145 -\ snmcnno rwr (m ‘ f I ' c-flzo : SYCAMOE M ("J "O , c- a: IOIIIIIII ' n m- m. :_ 53 = : 3 : uu _ g 5‘ MI : . - - I. ....” f : ~ ~ 5 3 ~‘ .. . ‘ '0 _ I 'A ’ 47 ‘ l \ a \ a t 5 . ’5 ~ 45 ’ I o .1 I '11 ”VIII /\ s ' ’ V =34 : @“m-m 43 I an no — nu \ CINCINNATI CITY AND ADJACENT AREA 144 involved in this study, one other potential answer can be eliminated--the loss of core population was not due to a non-white invasion of formerly white areas. Most of these areas had sizeable non—white populations in 1950 which they lost over the decade; none of them increased substantially in non-white population; in some of these tracts the propor- tion of whites actually increased. A second hypothesis sug- gested that the loss of population was due to a displacement of residents by commercial and other more intensive land uses. There is support for this in two of the three cities in this study; in Cincinnati, the number of dwelling units declined in four of the five tracts which lost population (excluding the renewal tract); and in Washington, in all four tracts (excluding the renewal tracts) where it was possible to compare the number of dwelling units in 1950 and 1960.91 The third hypothesis suggested by Lohman (see Chapter I)-- that the loss of population in the central city represents an effort on the part of the population to seek a new density appropriate to contemporary methods of transportation--cannot be rejected since none of the data, as analyzed in the present study, is applicable to it. Another pattern is that in both Cincinnati and St. Louis some predominantly white tracts experienced major increases in population and some tracts experienced major increases in 91The other tracts which lost population were halves of tracts which were split after the 1950 Census, so an accurate comparison is not possible. 145 the number of non-white; but no tracts experienced both a major increase in the size of the population and in the number of non-white. This may be due to the fact that the District has such a high percentage of non-whites and that some of the fringe areas of the District that were formerly undeveloped and that were built up during the 1950‘s are heavily Negro areas. Third, in neither St. Louis nor Cincinnati was a major increase in population combined with several other major changes; in other words, the areas which gained population just attracted more of the same type of population already resident there. Areas of increasing non-white population, however, frequently experience other changes in the character of the housing and population. One such change was an in- 2 In the census tracts of all three crease in overcrowding.9 cities, great gains in the non-white population were always accompanied by major increases in overcrowding. Great losses of population were likewise accompanied by decreases in over— crowding (in tracts with minuses and check marks). The areas of all three cities which just showed gains in population appeared to attract new people to the neighborhood who were similar to those already residing there. There were few major changes in these tracts in either education, income or rent. Over half of these tracts in St. Louis and Cincin- nati were all white; a little over a fourth were in D.C. 92This occurred in all of the tracts with a plus and a check mark in St. Louis, all in Cincinnati, and nine of the 12 in D. C. 146 Those tracts which both lost population and experienced other major changes were considerably below the city medians in residential stability for the five years preceding the Census; this was particularly true of the renewal cities. Concomitantly, in the renewal cities these tracts were above average in the proportion of 1960 residents who had moved there from elsewhere in the city; Cincinnati was below average in this regard. This suggests that these areas of declining population, which are also characterized by low income, edu- cation, and rent, may be serving as temporary way stations for those displaced from renewal areas. This interpretation is suggested by: 1) the fact that the 1960 population in these areas matched the 1950 population of the renewal tracts; 2) relatively few people were long term residents of the area (31% in St. Louis and 56% in D.C.); 5) the fact that the area is adjacent to the renewal areas; and 4) the unusual number of people who moved there from elsewhere in the city (51% in St. Louis, 59% in Washington). CHAPTER VI INTERPRETATION OF THE FINDINGS § In this chapter an effort will be made to summarize and interpret all the findings in relationship to the major hypotheses and the relevant literature. Following the pattern of earlier chapters, changes in the housing variables will be discussed first, then those observed in the population varia- bles. A third section will discuss the redistribution of population within the cities. The final section will point out some of the similarities, or regularities in the patterns of change, which seemed to be true for all three cities or for the two which were very closely matched on a number of variables. The Housing Variables It was hypothesized that there would be np difference between the renewal and control cities in terms of changes in the size and condition of the housing supply. Duncan and Hauser had found a relatively modest increase (about 5%) in housing supplies in Chicago and Philadelphia between 1950 and 1956. The rate of growth in Washington was almost three times as high, due primarily to an increase of 25,000 rental units over the decade. Rental units constitute a greater 147 148 portion of the total housing stock in the District because of the "temporary" nature of many of the jobs in the labor force there; another factor could also be the relative youth of much of the labor force. The number and proportion of rental units decreased in both St. Louis and Cincinnati as increased prosperity enabled the more stable work force to purchase their own homes. In this, the latter two cities followed the trend in the six SMSA's studied by Duncan and Hauser. However, St. Louis differed from Cincinnati in the overall growth in its housing supply; while Cincinnati's in- creased 4.5%, St. Louis' housing supply actually decreased by about 10,000 units. About 8,000 of these units were demolished in the renewal tracts. Washington lost a little over half that amount in its renewal areas and engaged in more building activity. Thus, it appears that in one renewal city, extensive demolition resulted in a restriction of the housing supply which was only partially offset by the loss of population over the decade. Both renewal cities had fewer vacant housing units than did the control city in 1960; but the vacancy rate in none of the three approached that "comfortable 7%" which Wattenberg and Scammon described for the U.S. as a whole.93 93However, many of the vacant housing units in the U.S. are in the rural areas and others of general population de- cline such as the mining towns in Kentucky. 149 Two consistent differences between the renewal and the control cities were that in both of the renewal cities the proportion of substandard and of overcrowded housing declined and in the control city substandard and overcrowded housing increased. This difference held true at both stages of the analysis no matter what descriptive statistic was used. This corresponds with Duncan and Hauser‘s findings in Chicago and Philadelphia, two cities which were involved with the renewal program early in the 1950's. In the control city, substandard housing--particularly and overcrowded housing to a lesser extent-~did not just increase in certain restricted areas of the city, but became more general, increasing in many tracts throughout the city. Conversely, in the renewal cities, substandard and overcrowding did not just decline in the re— newal areas, but generally throughout housing in the city. It is possible that the improvement of a fairly large area under public auspices encourages and makes economically feasi— ble private development on the fringes of the renewal area. This may be why many of the areas where four or more major changes have occurred or where there has been a serious loss of population are in tracts which are adjacent to the official renewal areas (see Figures 5a,b,c).94 Consequently, while 94This interpretation is confirmed for Washington, D.C. Helen Mason, in a Master‘s Thesis for Catholic University, reported that the Northwest section of the District lost more of its 1950 population than any of the other three sections of the city except the Southwest--which was the renewal area. In Figure 5 all the tracts which showed major losses in popu- lation (except those in the renewal area) were in the 150 renewal itself may not be directly responsible for many of the changes taking place in our cities, the federally sub- sidized program can be of sufficient size and impact that it makes it possible for private capital to "carry on." Kaplan's work on the politics of renewal strengthens this interpretation; private developers were unwilling to invest in a relatively small cleared area that was threatened by sur- rounding blight. Newark had to compromise by selecting for 2” renewal an area which was less deteriorated but more promising, in terms of its surroundings, for long term investment. This was not true for Washington and St. Louis--the areas chosen were badly deteriorated, overcrowded and overpopulated-- chiefly because the renewal tracts covered an extensive terri- tory which, in both cases, was adjacent to the central business district and, in Washington, the Capitol and other federal office buildings. With luxury apartments and tree-filled malls dotting the landscape, many businessmen were less re— luctant to invest in the upgrading of their property. In addition to these overall changes in the housing in— ventory of the cities which took part in the renewal program, there were also many changes occurring in the renewal tracts themselves. In the renewal areas of both Washington and Northwest. The reason for the population decline, according to Miss Mason, was a shift from residential to commercial land use in the neighborhood. See Helen S. Mason, "A Sociological Analysis of Selected Aspects of Urban Development in Washing- ton, D.C., Master's Thesis, Catholic University, Washington, D.C., May 1962. 151 St. Louis the proportion of substandard and overcrowded hous- ing declined as they did in each city as a whole. In St. Louis the proportion of substandard housing remained high in the renewal area, even though some 8,000 substandard units had been removed from the inventory. This may be because demolition was incomplete at the time of the census. The high vacancy rate (12%) in the area, when compared with a City-wide rate of about 5%, suggests that many homes were still being evacuated. Other changes in the renewal areas also diverged from the predominant pattern in the experimental cities. While, generally, the proportion of non-white occupied housing went up in all three cities, the proportion of non-white occupancy declined in the renewal areas. In Washington this appears to have been due to a slightly higher proportion of white than Negro residents who were not displaced during renewal and to a much higher rate of white in-migration to the area from out- side the city. Since some of the luxury apartments in Washington were ready for occupancy in 1959, they were un- doubtedly the destination of the white in-migrants. In St. Louis, more whites than non-whites were displaced, but, like Washington, the area was fed by a sizeable in-migration of whites from outside the central city; some of these came to luxury housing in one of the tracts, others to "integrated" public units in another tract. 152 Finally, median rents went up considerably in the re- newal tracts, doubling or almost tripling over the decade. This was in opposition to a standard average increase in all three cities of $22 to $24. Because of the extreme increases in rents in some of the tracts in the renewal cities, the~ standard deviations around the mean changes in the two cities were about three times higher than that around the mean change in rents in Cincinnati. Another area of difference between the renewal and con- trol cities was in the condition of housing available to Negroes in the central city. Robert Weaver had argued that renewal would result in a general improvement in the quality of such housing and he was right--at least for the cities in the present study. Independently of renewal, more Negroes during the 1950s were educationally, occupationally, and financially in a better position to purchase homes than previously; and more took advantage of the low-interest FHA and VA loans. Thus, home ownership among Negroes doubled in all three cities in the present study over the 1950-1960 decade. However, in the renewal cities, the ratio of owner- to-renter occupied units among non-whites improved whereas in the control city, it deteriorated.95 In Washington, a greater proportion of owner- than of renter-occupied units ‘95Miss Mason, op. cit., reports that, according to the Census of 1955, Negroes were buying homes in the District at a faster rate than the white people. 153 were occupied by Negroes; in St. Louis, the proportion in rental units was not quite double that in owner-occupied units; but in Cincinnati, the proportion of non-whites in rental units was seven times higher than their proportion in owner—occupied units. The reason for this difference must be the fact that Cincinnati did 29; experience the great out-migration of white population that the two renewal cities did over the decade: Washington lost a third and St. Louis a fourth of their white population between 1950 and 1960; Cincinnati's white population only declined 7.6%. Along with the improvement in home ownership among non- whites, in the renewal cities, there was a general improve- ment in the quality of the dwelling units: the proportion of substandard and overcrowded units among those occupied by Negroes declined substantially in both of the renewal cities and increased substantially in the control.96 The Population Variables It was also hypothesized that there would be no difference between the renewal and control cities in terms of changes in the character and distribution of the_population reSiding in them. In at least one way, the three cities were remarkably 96This pattern for overcrowding was observed in the means for 1950 and 1960, the 1950 and 1960 means in the non-white occupied tracts and in the mean differences; the only exception was in the comparison of medians where overcrowding declined in all three cities. This is probably explained by the general in- crease in all three cities of non-white occupied housing. 154 similar. All three experienced major losses of population in the inner part of the city and major increases in popula- tion at the peripheries. This substantiates McKenzie and Thompson, who found that some of the larger cities were losing\ I population at the center and were growing fastest in the outgnkfI‘ rings. In the three cities in this study, this loss of popu— lation was not due, as some have suggested, to a non-white invasion of formerly white areas. In some of these inner city tracts there was a substantial loss of Negro population and in some the proportion of whites actually increased. Many of these tracts bordered on the renewal areas and on the down- town sections of the cities and in many of them the number of dwelling units decreased considerably--suggesting a possible change in land use patterns from residential to commercial. These border areas would be, in the Chicago school‘s termin- ology, the zone of transition; it may be that as the metro- politan area grows outward, the city core increases in size. One of the main hopes of the renewal program was that it would retard the out-migration of middle class and predomi— nantly white residents to the suburbs. The data in the present study provide strong evidence to the contrary. The two renewal cities lost prOportionately more of their white population than did the control city, even though they ex- perienced roughly similar in-migrations of Negro population. This is what Grodzins predicted--that renewal would accelerate the succession of many city tracts to non-white occupancy. 155 That the renewal cities also lost more of their middle class is supported by several kinds of evidence. First, Cincinnati's median family income (for both the general population and for non-whites only) increased significantly more than the median incomes in either renewal city.97 Second, a much greater proportion of Cincinnati families made over $10,000 a year than was true for its matching city, St. Louis. Third, the average education declined in Washington, increased only Slightly in St. Louis, but went up in Cincinnati. And finally, the labor force statistics indicate that the proportion of skilled blue collar workers increased in Cincinnati, while decreasing in St. Louis; service workers and semi-skilled operatives increased in St. Louis while decreasing in Cincin- nati; and the proportion of white collar personnel decreased more over the decade in St. Louis than it did in Cincinnati. Another overall difference between the renewal and con- trol cities was in the general distribution of population within each city. The average tract population was higher in the renewal cities and higher still in the renewal tracts themselves. But renewal apparently lowered the extreme con- centrations of population; and, judging from the much smaller standard deviations in 1960 in both St. Louis and Washington. spread out the population more evenly throughout the city. 97This is based on the computations of mean differences. The census tracts in the renewal cities were also normally distributed around the average change in family income, while those in Cincinnati were skewed in the direction of large in— creases. 156 In addition to the overall changes in population in the cities, there were also many changes which occurred in the renewal tracts themselves. The renewal tracts were among those which lost major portions of their population between 1950 and 1960. The average education of residents in the renewal areas in 1960 was up over what it was in 1950, the median family income increased slightly. There is something puzzling in this. One of the housing changes in the renewal tracts was an unusual increase in average rent. This means that, while the 1960 resident population is--on the average-- of only slightly higher status, they are paying--on the average--much higher rents. Since the averages involved are medians, it is difficult to conclude anything other than that many families living in renewal areas after renewal are pay— ing out a greater proportion of their income for rent than the families who previously lived there. Several of the re- location studies have found that this is what happens to families displaced from renewal areas: they move to better quarters but pay higher rents. In both St. Louis and Washing- ton, some public housing was built within the renewed areas and people similar in many ways to the former residents moved into them. It is possibly the case that public housing-- while a good bargain-—is nevertheless considerably more expen- sive than a two-room unit in a crowded tenement. Negroes, according to Duncan and Hauser and many other students of renewal, have been disproportionately affected 157 by renewal programs. This was true in the present study-- roughly 60% of those in the areas chosen for renewal were non—white; the relocation of many of that 60% led, in turn, to a sizeable exodus of whites from the experimental cities. With Negroes thus able to "Spread out" a bit within the city, it is not surprising that the average tract increase in non- whites was significantly greater in both renewal cities than in the control. The standard deviations on this variable were also significantly greater than in Cincinnati, which could be interpreted to mean that there were larger dislocations of Negro population in the renewal cities. Despite the ability to "spread out" in the renewal city, it nevertheless appears that Negroes became more segregated, not only within the ..— . --.. fl,”— . . metrope4ttan‘areasT‘but within the central Cities as well. h...- A greater proportion of the total non-white ESEEIEEISE‘IIGed in tracts of maximum Negro concentration in 1960 than in 1950 in Washington and St. Louis; the reverse was true for Cincin— nati. The trend in Cincinnati is that predicted by Davis McEntire of the basis of his findings of increased segrega- tion in major American cities between 1940 and 1950. At the time, he felt that the reason for this increased segregation was the huge migration of non-whites to the cities during the war and the housing shortage which accompanied both the war and the depression. With all the building activity in the 1950's, it was reasonable to expect a loosening of the housing market for both. But this same activity took place in 158 Washington and St. Louis. Why should increased segregation be associated with renewal? There are several possible answers, all of which are tentative guesses. One-—it may be that Cincinnati is more of a “northern" city and that both Washington and St. Louis are more identified with Southern traditions. Southern cities, the Tauebers found, increased in segregation between 1940 and 1960 while Northern cities declined somewhat on this variable during the 50kg: This still leaves the question-~what is there about "Southern traditions" that is leading to increased racial segregation in the cities; traditionally, Negroes were mggg scattered throughout the city in the south. Another tentative interpretation might be that as other forms of racial separation break down (in education, voting, and jobs) in the South, residential segregation may become more im- portant. A second possible answer to the question of why increased segregation in the renewal cities may be more closely related to the renewal process itself. The Housing and Home Finance ‘)”i Agency Warned that slum clearance could result in a worsening of the housing conditions of minority groups through a re- striction of the living space available in any community to such groups. In the renewal cities, we know that the number and proportion of non-white occupied units increased, that overcrowding in these units decreased and that the quality of these units was much improved. But it could still be true 159 that Negroes were informally restricted to certain parts of the city. It may also be true that Negroes at different in- come levels were differently affected by renewal and the consequent opening up of housing throughout the city. Since most of the demolished housing was substandard in qual- ity and most of the new housing is for a middle class market, it may be that the lower class Negro is more restricted than ever in his choice of housing in the central city. The data in Table XX suggest that Negroes of all income levels were very circumscribed in their choice of housing in the late 40's and early 50's and that the freer housing market benefited primarily the upper income Negroes. Relocation Renewal programs in most cities have involved the displace- ment of many families from the condemned homes to new quarters elsewhere in the city. Quite a few studies of renewal have focused on aspects of relocation--the quality of housing moved into, the amount of rent paid in the new housing, the propor- tion of those displaced who received formal help in reloca- ing, etc. In this study, the focus of interest was on the redistribution of population within the city--not just the relocatees, but those who they, in turn, may have displaced. This intra-city migration is difficult to trace and may only be done indirectly with Census data. Therefore, the results reported on in this section are very tentative and exploratory. Table XX. Substandard and Overcrowded Housing in Non-white 160 Occupied Tracts, by Income Quartiles Median % Median % substandard overcrowded 1950 1960 1950 .1960 Washington, D.C.: lst Quartile* 45.5 27.9 54.2 25.6 2nd Quartile 51.9 17.0 26.6 24.1 5rd Quartile 50.1 5.4 50.0 24.2 4th Quartile 10.5 1.6 25.4 11.4 N= (51) (89) (51) (89) Cincinnati: ist Quartile 55.1 55.5 42.9 55.4 2nd Quartile 24.7 66.8 55.8 28.8 5rd Quartile 10.2 18.2 54.8 22.2 4th Quartile 8.1 8.5 19.1 17.7 N= (11 (50) (11) (30) St. Louis: lst Quartile 74.5 70.8 58.1 51.1 2nd Quartile 76.5 46.0 58.7 51.5 5rd Quartile 45.8 16.5 56.6 24.5 4th Quartile 21.9 4.5 26.1 21.6 N= (22) (51 (22) (51) * The first quartile is the lowest one. 161 On the other hand, certain patterns in the data are so clear that the author feels they warrant serious attention. In attempting to trace the movement of population within the city, the place to start is with the population of the re— newal tracts themselves. In 1950, the renewal tracts in Wash- ington and St. Louis were heavily Negro areas of rundown, sub- standard and overcrowded housing; rents were low and the edu— cational and income level of the population was low. Following renewal these same tracts in 1960 had lost a lot of the popu- lation, and a good portion of their dwelling units, and had proportionately fewer Negroes, overcrowded housing units, and, in some cases, substandard units. The object of search, then, was a sizeable low income Negro population that would need to settle in somewhat rundown areas of relatively low rent. As a first possibility, those tracts which experienced a major increase in population over the decade were investi- gated. The thought was that perhaps these 20-40,000 people would show up en masse in several other tracts in the city. But the population in this group of tracts were well above average in income and education, the housing was in excellent condition and many of the tracts were predominantly white. The proportion of Negro occupancy in these tracts changed very little from 1950 to 1960 and where Negroes did live, they were higher in education than the white residents. Many people in these tracts had moved there during the five years preceding the Census from elsewhere in the central city. 162 These mobile people may have represented the forefront of the lower class Negro invasion: the middle class whites would be moving out of neighborhoods into which the more average middle class Negroes were moving and the above average Negroes in these tracts might be doing the same thing. It is also possible that many of these "upper" class Negroes are moving to the area from outside the SMSA. Bernard Lazerwitz98 found that Negro in-migrants to northern and border cities were similar in education to the resident white non-migrants and that migrants in general (both white and non-white) exceeded the white non-migrant population in both educational and occupational status. This indicates a change in the type of Negroes coming to such cities. A second possibility was to investigate those areas which had experienced a major increase in non-white population over the decade. But most of the housing in these neighborhoods had been in good condition in 1950 and improved, percentage- wise, over the decade. The residents were of average income and education and well above average when compared with the non-white averages for the cities. The main change in these neighborhoods was from a white to a predominantly non-white population. A good proportion of the new residents in these neighborhoods had also moved from elsewhere in the central city. It is possible that many of the white families who 98Bernard Lazerwitz, op. cit. 165 moved from this second set of tracts moved into the set of tracts discussed in the preceding paragraph; many, too, were undoubtedly part of the 24-55% who left the city. But the evidence fpr the first half of that statement is particularly strong. The census tracts in St. Louis in that first set of tracts which had experienced major increases in total popula- tion had a much greater proportion of new residents from other parts of the city than was typical for tracts in that city and a very small prOportion of new residents from outside the city limits. The evfdence that this invasion-succession pattern is related to renewal is seen in the fact that in the matching control city, the same set of tracts (which had experienced major gains in population) had more old residents, a much smaller proportion of new residents from other parts of the central city, and a mggh greater proportion of new residents from outside the city--a completely different migra— tory pattern. There was one other possibility. The renewal tracts had changed in many different ways between 1950 and 1960. Perhaps the receiving tracts for the relocatees would also change in many ways in the opposite direction. A group of tracts was selected which had experienced four or more major changes on the selected population and housing variables, and the 1950 and 1960 populations in these tracts were com- pared with those in the renewal areas. The housing in these ”non-renewal“ tracts was in worse condition in both 1950 and 164 1960 (in terms of substandard and overcrowded dwellings) than either of the other two groups of tracts and the population was below average in education and income. In St. Louis this group of tracts gained more new residents from elsewhere in the central city than any other group of tracts. The number of dwelling units in these tracts increased slightly (possibly due to the conversion of single family houses into multi-family units), overcrowding increased or remained the same, the size of the population increased or remained the same and the proportion of non-whitesvwent up substantially. The average education of the resident population went down and rents, relatively, increased very little. One other argument buttresses the possibility that the "non-renewal" tracts may be the major recipients of families displaced by renewal--in all three cities many of these tracts are close to the renewal tracts. Marris and Schorrs indicated that displaced families often relocate within a mile of their former home. Changes Independent of Renewal Several patterns of change in the population and housing variables were observed in all three cities or in the two cities (one experimental and one control) which were closely matched. These changes, then, were independent of urban re— newal. 165 One of the most general changes, which is also true of most other large American cities studied in the last two decades, is an increase in the non-white population (abso- lutely and proportionately) and in the proportion of non-white occupied housing units-—a consequence of the continuing migration of Negroes to our urban areas, of the general ex- clusion of non—whites from the suburbs, and of the migration of whites outward from the central city and directly to the suburban fringe from outside the SMSA. Concomitantly, over— crowding in non-white occupied units went down, on the average, in all three cities although there was still considerably more overcrowding among non-white occupied units than among whites. All of this coincides with Duncan and Hauser‘s findings for Chicago and Philadelphia. While overcrowding was slightly higher in the renewal cities than in the control, the differen- tiation of "scores" on overcrowding was very similar in all three cities and changed little over the decade. Along with the decrease in overcrowding among non-whites went a general decrease in the average tract population in non-white occupied tracts. In the renewal cities, there was a general thinning out of population throughout the city; but this thinning out of the non-white population occurred in all three cities. St. Louis and Cincinnati shared two other similar patterns: in neither city was a major increase in the population of a census tract associated with major changes on several (three) other variables except for one instance in each city where that major population increase was non-white. What this 166 apparently means is that the tracts which grew most during the 1950's were those where new building occurred and new residents moved in who were very similar to those already there. All of these tracts were on the fringes of the city where open land is most apt to be available. Secondly, census tracts which experienced major increases in non-white popylation always experienced an increase in overcrowding. In summary, urban renewal in Washington, D.C. and St. Louis has had several discernible effects on the development of those cities during the 1950's. As hoped, a general im- provement in the housing stock occurred in both cities and this was accompanied by a general improvement in the quality of housing available to members of minority groups. Renewal also resulted in a considerable redistribution of population within the central city and appeared to accelerate the exodus of middle class whites from the central city and the segre- gation of Negroes in the metropolitan area as a whole and within the city itself. CHAPTER V I I SUMMARY, CRITICISMS, AND SUGGESTED FUTURE RESEARCH Summary of the Thesis The research reported on in this dissertation attempted to measure the effects of a government program--urban re- newal--on changes in the housing supply and the character and distribution of the population in two major cities. Urban renewal was, in effect, an attempt to interfere with the basic pattern of ecological change in larger American cities. Most such cities were losing population to their suburbs, growing faster in the metropolitan rings than in the central city, gaining steadily in Negro population, and becom— ing increasingly the home of the poor, the aged, and the minority groups. It was the hope of the Urban Renewal Adminis- tration and of many who supported the program tHat renewal would succeed in attracting the white middle class, their business and their taxes back to the city. In this study, the patterns of ecological change in two cities which took part in the renewal program from the early 1950's on were compared with the changes which occurred in a third city which did not enter the program until late in 1959. The study'was quasi-experimental in design. That is, it was a natural experiment with an independent variable 167 168 (urban renewal) "naturally" occurring in two of the three cities (the experimental cities) and not occurring in a third city (the control city). Before and after comparisons on selected population and housing variables were made for both the experimental and the control cities. The sample included all of the census tracts in three central cities, each a border city between South and North in the eastern half of the United States. Two of the cities, one experimental (SEL/Lguisififlissouri) and one control \/ (CincinnatiL‘Ohio), were very similar in the age and condition of their housing, the occupational make-up of their population, and the educational and income level of the residents. All three (Washington, D.C. was the third) were approximately the same size and had relatively large Negro populations. The 1950 and 1960 Censuses of Population and Housing were the chief source of data. Comparable information was taken from the 1950 and 1960 publications. The data was analyzed in three stages. In the first stage, the information on selected population and housing variables was summed for the city as a whole and expressed in the form of medians and percentages. The three cities could be compared in 1950 or 1960 or changes in these measures could be compared for each city. In the second stage, changes in the value of selected variables for each census tract were summed and a mean dif- ference and standard deviation of the difference computed for 169 each city. Differences in the direction and magnitude of change on these variables could then be compared for the three cities. In the third stage of the analysis, the changes in the values of selected variables were averaged for the renewal tracts alone so that the effects of the urban renewal program on those areas which were formally a part of the program could be directly measured. The basic hypothesis of the research was one of "no dif— ference" between the experimental and the control cities. Specifically, it was hypothesized that there would be no dif- ferences observed between the experimental and control cities with respect to the types of change: 1) in the condition and size of the housing supply, and 2) in the character and distribution of the population When: (a) the data are summed for each city as a whole (b) the mean changes (on a census tract by census tract basis) are computed for each city as a whole, and (c) the mean changes for the renewal tracts alone are compared. On the housing variables, the two renewal cities experi— enced decreases in the proportion of substandard and over- crowded housing, while in the control city substandard and overcrowded housing increased. In the one experimental city which was matched with the control, the housing supply de- clined by about 10,000 units due to the extensive demolition connected with renewal while the control city‘s housing stock increased 4%. The actual amount of new building in the two 170 cities was practically identical. Both renewal cities had fewer vacancies than the control city in 1960, even though vacancy rates varied from 8% to 14% in the renewal tracts. While the proportion of non—white occupied housing in- creased in all three cities, home ownership among Negroes was much more widespread in Washington and St. Louis than in the control city. The reason behind this is related to a dif- ference in population change between the experimental and control cities. While the proportionate increase in Negro population was roughly the same for all three cities (and almost exactly the same for the matched cities), the two re- newal cities lost from a fourth to a third of their 1950 white population while Cincinnati's declined only 7.6%. In addition to the improvement in home ownership among non—whites, there was also a general improvement in the quality of the dwelling units available to minority groups: the proportion of substandard and overcrowded units among non-white occupied housing declined substantially in the renewal cities and increased substantially in the control. Finally, many changes occurred in the renewal tracts themselves. There was a general improvement in the quality of housing in these areas, the proportion of non-white occu- pancy decreased, and median rents doubled or almost tripled over the decade. On the population variables, the three cities experi— enced major losses of population in the central core and major 171 increases in population at the peripheries. This loss of population was not due, as some have suggested, to a non- white invasion of formerly white areas; in some of the inner city tracts there was a substantial loss of Negro population and in some the proportion of whites actually increased. The renewal cities, as opposed to the control, lost proportion- ately more of their white population, even though they experi- enced roughly similar in—migrations of Negro population. They also lost more of their middle class population as indi- cated by their relatively modest increases in median family incomes, as compared with substantial increases on this vari- able for Cincinnati's white and non-white population; by the decline or stability of the average education of their popu- lation, as compared with a definite increase in the level of education in Cincinnati; and, in St. Louis only, by a greater loss of skilled blue collar and white collar residents and a greater gain in service workers and semi—skilled operatives. This accelerated exodus of white, middle class residents from the renewal cities appears to be the result of a con- siderable redistribution of population within the experimental cities. Many tracts in all three cities experienced large increases in the size of their populations. In Cincinnati, the new arrivals were more often from outside the central city. In Washington and St. Louis, they were more often from elsewhere in the central city. The people in these tracts were predominately white, well above average in income and 172 education and the proportion of Negro occupancy in these tracts had changed very little from 1950 to 1960. The Negroes in these tracts were higher in average education than the whites. It is possible that the new arrivals in this group of tracts were running from a Negro invasion of their old neighborhoods. Many tracts in the renewal cities had, in fact, experi- enced very large increases in non-white population. These were neighborhoods where the residents were about average in income and education, where the housing was in good condition in 1950 and in even better, proportionately, in 1960, and where the residents were predominantly white in 1950, but predominantly Negro in 1960. A complete cycle of invasion and succession had occurred in these tracts--from practically zero non-white occupancy in 1950 to 70-80% non-white in 1960. So it is even more conceivable that people moved from this group of tracts to the first group. Apparently, then, these middle class Negroes were themselves running before an invasion of lower class Negroes who were being moved out of the renewal areas-—or possibly just taking advantage of in- creasing housing opportunities to get themselves out of the slums. The former argument is favored because it provides an initial "motive" for the whites to move in the first place. In fact, there was a third group of tracts, from which the more average Negroes could have fled and to which the 175 displaced lower class Negroes could have gone. These tracts were close to the renewal areas, had a sizeable non-white population in 1950 which increased substantially in 1960, and had housing that was in worse condition in both 1950 and 1960 than either of the other two groups of tracts. The popu- lation in these tracts was below average in its level of education and income and the level of education decreased over the decade. A greater proportion of new residents in this group of tracts had moved there from another house in the city than for any of the three groups of tracts. All of this adds up to fairly strong support for the idea that these are the "receiving" areas to which those displaced by renewal moved, setting off a chain reaction which culminated in a considerable redistribution of population within the city. All of this movement of population resulted in another major difference between the renewal and control cities. In the two renewal cities, the segregation of non-white popu— lation within the central city increased between 1950 and 1960, while decreasing in Cincinnati. A greater proportion of the total non-white population lived in tracts of maximum Negro concentration in 1960 than in 1950 in Washington and St. Louis, while the reverse was true for Cincinnati. Limitations of the Study There are several precautions which the author should emphasize. First, while many of the findings were supported 174 by Duncan and Hauser's research in Chicago and Philadelphia, it would not be good to generalize the findings here to other large American cities. The two renewal cities studied follow- ed many of the same patterns of change, but also diverged at important points. Differences in racial composition, age of city, nature of the labor force, the relative importance of migration and natural increase in changes in city size——all of these factors could appreciably affect the impact of an urban renewal program. Second, this research was restricted to the central city. Urban renewal, so far, is essentially a central city program. Its effects are probably more discernable in the city, rather than outside of it. However, there have undoubtedly been re— percussions throughout the metropolitan area. Too many factors would have to be controlled if the effects were to be weighed in such a limited number of SMSA's. A fully ade- quate study of the effects of renewal on changes in the total metropolitan area would require a large number of SMSA‘s-- preferably all of a given size in the U.S.-fwhich could be grouped according to the extent of their participation in the renewal program. ~Another difficulty with using the whole metropolitan area in such a small sample is that of defining where the area ends. If the Census definition were used, it is conceivable that two of the SMSA‘s might be densely settled in all of the counties adjacent to the major city while one of the SMSA's had one or more largely undeveloped 175 counties which still were "essentially metrOpolitan in char- acter and socially and economically integrated with the central city." This could seriously bias the results. If all SMSA's were included, these local differences would prob— ably balance each other. The city, on the other hand, has arbitrary, but very operational, politically meaningful, and well understood boundaries. A third limitation was set by the nature of the data. Because of the interest in the redistribution of population in this research, it is unfortunate that greater detail about intra-city mobility is not provided in the Census. While the inferences about such mobility are fairly strongly sup- ported by the data, the migration of large numbers of people from one particular area of the city to another is still inferential and not as concretely documented as the author would like. A final limitation was set by choice. Some of the find- ings could have been documented in greater depth by studying, as Duncan and Hauser did, the "components of change" in the I housing inventory--whether fluctuations in the size of the housing supply in various parts of the city were due to demo- lition, new construction, conversion (of single family homes to multi-family units), or merger (of multi-family units into single family homes). 176 Suggestions for Further Research The limitations of phi; study lead quite naturally to new ideas for research in this and related areas. One possi- bility was referred to above—-doing the same thing on a much larger scale, controlling chiefly for size of city and extent of involvement in the urban renewal program. A second would include an investigation of the differences between renewal and non-renewal cities in terms of the components of change in the housing inventory in terms of changes in the property values, assessments, and land uses in a) the renewal tracts, b) those tracts adjacent to the renewal area, and c) the city as a whole. A third possible project would be to find out who the people are and where they are coming from who are moving into the renewal areas. Are those who come from outside the city moving from that city's own suburbs or migrating from other metropolitan areas? What age groups are being attréated to the central city and what is their occupational and class background? Will these new residents be relatively stable or will there be a high turnover of population in the inner city tracts? Will renewal, in the long run, still succeed in bringing middle class whites back to the city? Another question that merits investigation is: What is responsible for the apparent association between increased segregation and renewal? Perhaps the massive dislocation of population creates the kind of pressure on the housing market 177 which realtors can use to "block bust" and informally encour- age racial homogeneity in city neighborhoods. Perhaps the changing racial balance in cities makes integration more difficult. Finally, there is a great deal to be learned about pat— terns of residential mobility, both within cities and within metropolitan areas. Are there pathways which particular fami- lies follow as their social position changes? Why do some neighborhoods "turn" when Negroes begin to move in, while others "integrate"? Can we predict what effects open occu- pancy and consequently "open" suburbs will have on housing in the central city? Many new programs and many new laws may bring about radical changes in the city of the future. B IBLIOGRAPHY 178 BIBLIOGRAPHY A. BOOKS Abrams, Charles. Man's Struggle for Shelter in an Urbaniz- ing World. Cambridge, Massachusetts: Massachusetts Institute of Technology Press, 1964. Anderson, Martin. The Federal Bulldozer. Cambridge, Massachusetts: Massachusetts Institute of Technology Press, 1964. Back, Kurt..Slums,pProjects, and People: Social Psycho+ logical Problems of Relocation in Puerto Rico. Durham, North Carolina: Duke University Press, 1962. Bogue, Donald J. Components of Population Change, 1940-1950. Miami, Florida: Scrips Foundation for Research in Population Problems, 1957. Bogue,qDonald J. Metropolitan Decentralization: A Study of Differential Growth. Oxford, Ohio: Scrips Foundation for Research in Population Problems, 1950. Bogue, Donald J. Needed Urban and Metropolitan Research. Miami, Florida: Scripps Foundation Studies in Popula- tion Distribution, 1955, No. 7. Bogue, Donald J. Population Growth in Standard Metropolitan Areas 1900-1950. Washington, D.C.: Housing and Home Finance Agency, December, 1955. Chapin, F. Stuart. Urban Growth Dynamics. New York: Wiley & Sons, 1962. Chinitz, Benjamin, ed. City and Suburb: The Economics of Metropolitan Growth. Englewood Cliffs, New Jersey: Prentice Hall, 1964. Conant, James B. Urban Slums and the Suburban Community. New York: McGraw Hill, 1962. Dobriner, William. Class in Suburbia. Englewood Cliffs, New Jersey: Prentice Hall, 1965. 179 180 Duhl, Leonard, ed. The Urban Condition. New York: Basic Books, 1965. Duncan, Beverly and Philip Hauser. Housing a Metropolis-- Chicago. Glencoe, Illinois: The Free Press, 1960. Gans, Herbert. The Urban Villagers. Glencoe, Illinois: The Free Press, 1962. Gendell, Murray and Hans L. Zetterberg. A Sociological Almanac for the United States, New York: Charles Scribner‘s Sons, 1964, 2nd edition. Glazer, Nathan. Studies in Housingpand Minority Groups. Berkeley: University of California Press, 1960. Gordon, Mitchell. Sick Cities. New York: The Macmillan 1C Company, 1965. ~/ Greer, Scott. The Emerging Cipy: Myth and Reality. Glencoe, Illinois: The Free Press, 1962. Grier, Eunice and George. Privately Developed Interracial Housing: An Analysis of Experience. Berkeley, Uni- versity of California Press, 1960. Grodzins, Morton. The Metropolitan Area as a Racial Problem. Pittsburgh: Pittsburgh Press, 1965. Hadden, Jeffrey K., and Edgar F. Borgatta. American Cities: Their Social Characteristics. Chicago, Illinois: Rand McNally & Company, 1965. Hawley, Amos H. The Changing Shape of Metropolitan America: Deconcentration Since 1920. Glencoe, Illinois: The Free Press, 1956. Higbee, Edward. Squeeze: Cities Without Space. New York: Morrow, 1960. Hoyt, Homer. The Structure and Grgwth of Residential Neigh- borhoods in American Cities. Washington,>D.C.: Govern- ment Printing Office, 1959. Jacobs, Jane. Tbepgeath and Life of Great American Cities. New York: Random House, 1961. Kaplan, Harold. Urban Renewal Politics: Slum Clearance in Newark. New York: Columbia, University Press, 1965. Klein, Woody. Let in the Sun. New York: Macmillan, 1964. 181'— Laurenti, Luigi. Property7Values and Race: Studies in Seven Cities. Berkeley: University of California Press, 1960. McEntire, Davis. Residence and Race. Berkeley: University of California Press, 1960. McKelvey, Blake. The Urbanization of America, 1860-1915. New Brunswick, New Jersey: Rutgers University Press, 1965. Millspaugh, Martin, and Gurney Breckenfeld. The Human Side of Urban Renewal. Baltimore: Fight-Blight, 1958. Owen, Wilfred. The Metropolitan Transportation Problem. Washington, D.C.: Brookings, 1956. Perloff, Harvey. Planningpand the Urban Community. Pittsburgh: Carnegie Institute of Technology, 1961. Rapkin, Chester, and William Grigsby. Residential Renewal in the Urban Core. Philadelphia: University of Pennsylvania Press, 1960. Rodwin, Lloyd. Housing and Economic Progress: A Study of the Housing Experiences of Boston‘s Middle-Income Families. Boston: Harvard University Press, 1961. Rossi, Peter, and Robert Dentler. The Politics of Urban Renewal:4 The Chicago Findings. Glencoe, Illinois: The Free Press, 1962. Schorr, Alvin L. Slums and Social Insecurity. Washington, D.C.: Social Security Administration, Report No. 1. Seeley, J. R., Sim, R. A., Loosley, E. W. Crestwood Heights. New York: John Wiley and Sons, 1956. Shevky, Eshref and Wendell Bell. Social Area Analysis. Stanford: Stanford University Press, 1955. U.S. Bureau of the Census, U.S. Census of Population, 1950, Vol. III, Census Tract Statistics. Final Report PHC(1)- 11, 47, 59. Washington, D.C.: U. S. Government Print- ing Office, 1952. U.S. Bureau of the Census. U.S. Census of Population: 1960. General Social and Economic Characteristics, United States Summary. Final Report PC(1)-1C. Washington, D.C.: U.S. Government Printing Office, 1962. 182 U.S. Bureau of the Census. U.S. Census of Population and Housing: 1960. Census Tracts. Final Report PHC(1TF 27, 151, 166. Washington, D.C.: U.S. Government Print- ing Office, 1962. Van der Post, Laurens. Patterns of Renewal. Pendle Hill Pamphlets. Walker, Helen and Joseph Lev. gpatistical Inferpnce. New York: Holt, Rinehart, and Winston, 1955. Wattenberg, Ben J. and R. M. Scammon. This USA. New York: Doubleday, 1965. Weaver, Robert C. The Urban Complex: Human Values in Urban Life. New York: Doubleday, 1964. Wood, Robert. Suburbia: Its People and Their Politics. New York, Houghton, 1959. Woodbury, Coleman, ed. Urban Redevelopment: Problems and Practices. Chicago: University of Chicago Press, 1955. B. ARTICLES Abrams, Charles. "Slums, Ghettos, and the G.O.P.‘s 'Remedy'," Reporter, May 11, 1954, pp. 27-50. Bogue, Donald J., and Emerson Seim, "Components of Population Change in Suburban and Central City Populations of Standard Metropolitan Areas, 1940 to 1950," Rural Soci- ology, 21 (September-December 1956), pp. 265-275. Foley, Donald L. "Census Tracts and Urban Research," Journal of American Statistical Association, (December 1955), pp. 755-42. Gans, Herbert. "The Failure of Urban Renewal: A Critique and Some Proposals," Commentary, Vol. 59, No. 4, April 1965. Hawley, Amos. "Community Power and Urban Renewal Success," American Journal of Sociology, 68 (January 1965), pp. 422-451. Housing and Home Finance Agency. "Living Space Available to Racial Minority Families." Local Public Agency Letter No. 16, February 2, 1955. 185 Kish, Leslie. "Differentiation in MetrOpolitan Areas,“ American Sociological Review, 19 (August 1954). PP. 388-598. Lazerwitz, Bernard. “Metropolitan Community Residential Belts, 1950 and 19563" American Sociological Review, 25 (April 1960). pp. 245—252. Marris, Peter. "A Report on Urban Renewal in the United States" in Leonard Duhl, The Urban Condition: People and Policy in the Metropolis. New York and London: Basic Books, Inc. (1965), pp. 115-154. Schnore, Leo F. "The Separation of Home and Work: A Problem for Human Ecology," Social Forces, 52 (May 1954), pp. 556-545. Schnore, Leo F. "The Growth of Metropolitan Suburbs," American Sociological Review, 22 (April 1957). PP. 165-175. Schnore, Leo F. "Metropolitan Growth and Decentralization," American Journal of Sociology, 65 (September 1957), pp. 171-180. Schnore, Leo F. "Components of Population Change in Large Metropolitan Suburbs," American Sociological Review, 25 (October 1958), PP- 570-575. Seligman, Daniel. The Enduring Slums," Fortune, December V/ 1957, pp. 144ff. Shyrock, Jr., Henry S. "Population Redistribution within Metropolitan Areas: Evaluation of Research,“ Social Forces, 55 (December 1956), pp. 154-159. Taeuber, Karl E. "Negro Residential Segregation, 1940-1960: Changing Trends in the Large Cities of the U.S.," paper read at the Annual Meetings of the American Sociological Association, 1962. Taeuber, Karl and Alma. "White Migration and Socio-econpmic Differences Between Cities and Suburbs," American Socio- logical Review, 29 (October 1964), pp. 718-729. Taeuber, Karl and Alma. "The Changing Character of Negro Migration," American Journal of Sociology (January 184 C. UNPUBLISHED MATERIAL Mason, Helen S. "A Sociological Analysis of Selected Aspects of Urban Development in Washington, D.C." Unpublished Master's thesis, Catholic University, Washington, D.C., May 1962. APPENDICES 185 APPENDIX A 186 187 CODE NUMBERS ASSIGNED TO CENSUS TRACTS IN WASHINGTON. D.C., 1960* <— q Census Census Census Tract Assigned Tract Assigned Tract Assigned Number Number Number Number Number Number 1 1 42 42 75.7 85 2 2 45 45 75.8 84 5 5 44 44 74.1 85 4 4 45 45 74.2 86 5 5 46 46 74.5 87 6 6 47 47 75 88 7 7 48 48 76.1 89 8 8 49 49 76.2 90 9 9 50 50 76.5 91 10 10 51 51 77.1 92 11 11 52.1 52 77.2 95 12 12 52.2 55 77.5 94 15 15 55.1 54 77.4 95 14 14 55.2 55 77.5 96 15 15 54.1 56 78.1 97 16 16 54.2 57 78.2 98 17 17 55 58 78.5 99 18 18 56 59 78.4 100 19 19 57.1 60 78.5 101 20 20 57.2 61 78.6 102 21 21 58 62 79 105 22 22 59 65 80 104 25.1 25 60 64 81 105 24 24 61 65 82 106 25 25 62 66 85 107 26 26 65 67 84 108 27 27 64 68 85 109 28 28 65 69 86 110 29 29 66 70 87 111 50 50 67 71 88.1 112 51 51 68 72 88.2 115 52 52 69 75 89 114 55 55 70 74 90 115 54 54 71 75 91 116 55 55 72 76 92 117 56 56 75.1 77 95 118 57 57 75.2 78 94 119 58 58 75.5 79 95.1 120 59 59 75.4 80 95.2 121 40 40 75.5 81 95.5 122 41 41 75.6 82 95.4 125 * 1950 tracts were numbered 1-95 as given. The subdivisions of 1960 were collapsed at the time of the 1950 census. Thus, CT 52 in 1950 is equal to CTs 52 and 55 in 1960. ** Tract 25.2, an institutional tract, was omitted. 188 CODE NUMBERS ASSIGNED TO CENSUS TRACTS IN ST. LOUIS, MISSOURI, 1950 & 1960 Census Census Census Tract Assigned Tract Assigned Tract Assigned Number Number Number Number Number Number 1-a 1 8-b 44 17-a 87 firb 2 8-c 45 17-b 88 1-c 5 8-d 46 17-c 89 1-d 4 8-e 47 17—d 90 1-e 5 8-f 48 18—a 91 1-f 6 9-a 49 18-b 92 1-g 7 9-b 50 18-c 95 2-a 8 9-c 51 18-d 94 2-b 9 9—d 52 18-e 95 2-c 10 9-e 55 19-a 96 2-d 11 10-a 54 19-b 97 2-e 12 10-b 55 19-c 98 5-a 15 10-c 56 20-a 99 5—b 14 10—d 57 20—b 100 5-c 15 10-e 58 20-c 101 5-d 16 11-a 59 21-a 102 5-e 17 11-b 60 21-b 105 5-f 18 11-c 61 21-c 104 5-g 19 11-d 62 21-d 105 4-a 20 11-e 65 22-a 106 4-b 21 12-a 64 22-b 107 4-c 22 12-b 65 22-c 108 4-d 25 12-c 66 22-d 109 5-a 24 12-d 67 25-a 110 5-b 25 15-a 68 25-b 111 5-c 26 15-b 69 25-c 112 5-d 27 15-c 70 25-d 115 5-e 28 15-d 71 25-e 114 6-a 29 14—a 72 24-a 115 6-b 50 14-b 75 24-b 116 6-c 51 14—c 74 24-c 117 6-d 52 15-a 75 24-d 118 6-e 55 15-b 76 24-e 119 6-f 54 15-c 77 25-a 120 6-g 55 15-d 78 25-b 121 7-a 56 15-e 79 25-e 122 7-b 57 15-f 80 25—d 125 7-c 58 15-g 81 26—a 124 7-d 59 16-a 82 26-b 125 7-e 40 16—b 85 26-c 126 7-f 41 16-c 84 26-d 127 7-g 42 16-d 85 26-e 128 8-a 45 16-e 86 189 CODE NUMBERS ASSIGNED TO CENSUS TRACTS IN CINCINNATI, OHIO, 1960* Census Census Census Tract Assigned Tract Assigned Tract Assigned Number Number Number Number Number Number 1 1 59 59 76 76 2 2 40 40 77 77 5 5 41 41 78 78 4 4 42 42 79 79 5 5 45 45 80 80 6 6 44 44 81 81 7 7 45 45 82 82 8 8 46 46 85 85 9 9 47 47 84 84 10 10 48 48 85 85 11 11 49 49 86a 86 12 12 50 50 86b 87 15 15 51 51 87 88 14 14 52 52 88 89 15 15 55 55 89 90 16 16 54 54 90 91 17 17 55 55 91 92 18 18 56 56 92 95 19 19 57 57 95 94 20 20 58 58 94 95 21 21 59 59 95 96 22 22 60 60 96 97 25 25 61 61 97 98 24 24 62 62 98 99 25 25 65 65 99 100 26 26 64 64 100 101 27 27 65 65 101 102 28 28 66 66 102 105 29 29 67 67 105 104 50 50 68 68 104 105 51 51 69 69 105 106 52 52 70 70 106 107 55 55 71 71 107 108 54 54 72 72 108 109 55 55 75 75 109 110 56 56 74 74 110 111 57 57 75 75 111** 112 58 58 * *1950 tracts are numbered 1-110 as given. Tract 111 was a newly annexed tract added in 1960. APPEND IX B 190 Column number 1-5 4-8 9-15 14-18 19-25 24-28 29-55 54-57 58-42 45-45 46-50 51-54 55-58 191 CODE BOOK--DECK 1 Description tract number persons over 5 resident in 1960 93 persons over 1 resident in 1950 number in same house, 1955 & 1960 93 number in same house, 1949 & '50 number in different house, central city, 1955 & 1960 93.number in different house, central city,.1949 and 1950. non-white persons over 5 resident in 1960 93 over 1, 1950. non-white persons in same house 1955 & 1960 93_ in same house 1949-50. non-white persons in different house, central city, 1955 & 1960 QR_non—white persons in different house, central city, 1949 & 1950. housing units with no private bath or dilapidated (1950) 93 housing units (1960) that are: 1) dilapi- dated, or 2) sound, lacking other plumbing facili- ties, or 5) deteriorating, lacking other plumbing facilities. number of housing units reported on (1950 only) all housing units in tract non—white housing units with no bath or dilapidated (1950) QB non-white housing units (1960) that are: 1) dilapidated, or 2) sound, lacking other plumbing facilities, or 5) deteriorating, lacking other plumb- ing facilities. number of non-white housing units reported on (1950 only) Column number 59-61 62-64 65-67 68-70 71-75 74-77 78-80 Description 192 percent = percent = percent = percent = percent = percent = Deck identi no private bath or dilapidated number of housing units reporting (1950) 93. dilapidated, sound or deteriorating lacking other facilities Total housing units (1960) number in different house, central city persons over 5 (1960) or persons over 1 (1950) number in same house persons over 5 (1960) or over 1 (1950) non-white persons in different house non-white persons over 5 (1960) or over 1 (1950) non-whiteppersons in same house non-white persons over 5 (1960) or over 1 (1950) fication: 78. . . Deck no. 1, 2, or 5 79. . . year: 0 . . . 1950 1 . . . 1960 80. . . place: x. . .Washington, D.C. y. . .St. Louis, Missouri z. . .Cincinnati, Ohio Column number 1-5 4-8 9-15 14-18 19-22 25-26 27-50 51-54 55-58 59-42 45-45 46-48 49-51 52-54 55—57 58-60 61-65 64-66 195 CODE BOOK--DECK 2 Description tract number number of all housing units in tract total occupied units number of non-white occupied housing units number of housing units with over 1.01 persons per room number of non-white housing units with over 1.01 persons per room number moved into unit 1958-60 (1960 only) number moved into unit 1954-57 (1960 only) number of non-white moved into unit 1958-60 (1960 only) number of non—white moved into unit 1954-57 (1960 only) non-white occupied housing units percent = total occupied housing units in tract ercent = number of housing units with over 1.01 p personspper room total occupied housing units in tract ercent = number of non-white housing units with p over 1.01 persons per room total non-white occupied housing units ercent = number moved into unit 1958-1950 p all occupied units ercent = numbgppmoved into unit 1954-57 p all occupied units ercent = number non-white moved into pnit 1958-60 p all occupied units ercent = number non-white moved into unit 1954:57 p all occupied units percent = number non-white moved into unit 1958-60 all non-white occupied units 194 number non-white moved into unit 1954-57 all non-white occupied units 67-69 percent total occupied units number of all housing units in tract 70-75 percent 78-80 Deck identification. (See Deck 1.) Column number 9-11 12-16 17-18 19-25 24-28 29-51 52-54 55-59 40-44 45-49 50-55 54-57 58-60 78-80 195 CODE BOOK--DECK 5 Description tract number median persons per unit median school years completed (total population) median school years completed (non-whites) median family income (non-whites) non-white median persons per unit non-white population total population median contract rent, total population (1950 and 1960) non-white median contract rent, 1950 p£_median gross rent, 1960 median family income (total population) total population,.1940 (in 1950 deck only) non-white population, 1940 (in 1950 deck only) non-white population total population percent non-white population, 1940 percent = total population, 1940 median gross rent, total population (1960 only) Deck identification (See Deck 1.) APPENDIX C 196 Table XXIa. 197 Occupational Distribution of the Labor Force, For the City and Standard Metropolitan Statis- tical Area, 1950 and 1960:* Washington, D.C. Male Female Occupation City SMSA City SMSA 1950 1960 1960 1950 1960 1960 Professional & 14.7 15.2 21.4 12.2 15.5 15.6 technical personnel. Managers, officials 10.1 8.1 12.6 5.5 5.4 5.9 & proprietors. Clerical. 16.4 15.0 12.2 46.2 57.9 45.2 Sales workers. 6.1 4.4 6.4 5.4 5.9 6.1 Craftsmen & 16.2 12.2 16.4 1.1 9.7 .9 foreman- Operatives & 15.1 12.5 9.7 5.6 4.4 5.7 kindred workers. Private household -- -- .5 11.0 10.8 8.2 workers- Service workers 12.6 12.8 8.0 15.5 14.4 10.7 except the above. Laborers, except 9.2 8.8 5.9 .6 1.0 1.0 mine. Occupation not 1.4 10.8 7.0 1.5 10.2 7.5 reported. *Source: U.S. Bureau of the Census, U.S. Censuses of Popula- tion and Housing: 1950 and 1960, op. cit. Table XXIb. 198 Occupational Distribution of the Labor Force, For the City and Standard Metropolitan Statis- reported. tical Area, 1950 and 1960:* St. Louis, Missouri Male W Female Occupation City SMSA City SMSA 1950 1960 1960 1950 1960 1960 Professional & 6.9 7.4 10.5 9.5 10.0 11.6 technical personnel. Managers, officials 9.4 6.5 10.9 5.5 2.7 5.1 & proprietors. Clerical. 11.7 10.7 9.1 52.5 50.0 55.5 Sales workers. 7.5 5.6 7.5 7.2 6.6 8.1 Craftsmen & foreman. 19.5 17.4 20.9 2.2 1.4 1.5 Operatives & 25.9 24.5 21.2 25.2 17.0 14.6 kindred workers. Private household .2 -- -- 6.8 7.2 5.8 workers. Service workers, 9.4 9.7 6.2 12.4 14.5 15.4 except the above. ~Laborers, except 9.9 8.9 7.5 1.2 .8 .7 mine. Occupation not 1.8 9.4 6.5 2.0 10.0 7.9 *Source: U. 8. Bureau of the Census, 1950 and 1960, op. cit. tion and Housing: U.S. Censuses of Popula- Table XXIc. 199 Occupational Distribution of the Labor Force, For the City and Standard Metropolitan Statis- tical Area, 1950 and 1960:* Cincinnati, Ohio Male Female Occupation City SMSA City SMSA 1950 1960 1960 1950 1960 1960 Professional & 9.5 11.5 12.0 11.5 12.0 11.8 technical personnel. Managers, officials 12.5 9.6 11.5 5.8 3.5 5.2 & proprietors. Clerical. 9.1 8.7 8.9 28.5 28.9 52.4 Sales workers 8.7 7.2 8.0 7.6 6.8 8.5 Craftsmen & foreman. 12.5 16.6 19.5 2.5 1.4 1.4 Operatives & 21.7 20.2 20.5 19.1 15.0 15.5 kindred workers. Private household .4 1.0 1.0 9.8 8.0 6.5 workers. Service workers, 9.0 8.5 6.2 14.8 14.5 15.0 except the above. Laborers, except 8.5 8.5 6.7 .9 .1 -- mine. Occupation not 1.1 9.0 6.9 1.5 11.5 9.1 reported. *Source: U. S. Bureau of the Census, U.S. Censuses of Popula- tion and Housing: 1950 and 1960, op. cit. APPEND IX D 200 201 Table XXII. Racial Imbalance in the Census Tracts of Three Central Cities Number of census tracts where the Washington, St. Louis Cincinnati percent non-white D.CL ‘ . is: 1950 1960 1950 1960 1950 1960 0 - .09 0 0 46 46 22 51 .1 - 5.0 29 27 51 55 56 55 5.1 - 20.0 15 12 9 11 16 17 20.1 - 50.0 25 18 7 6 7 6 50.1 - 80.0 14 26 6 14 4 11 80.1 - 95.0 8 20 6 10 1 6 95.1 - 99.9 6 20 5 6 4 6 Total number 95 125 128 128 110 112 of tracts APPENDIX E 202 205 «w.mm.mfi mma.mm coaumHsmom AmuOu .ucmu maa .fia.m.m.m .¢6.ma.a mm.mw.oa uomuucou smapwz moa muflcs .mm.amuas cs.mm.mm magmsor amazououm>o mm am .am.mm.om was .em.mm.ma muanzucoc mo unmouwm maa.¢oa .moa.ooa moa.moa .mmumm.m> m>.0>lmm .sm.mm.mm .msuaw.w¢ mm.mm .mm.«m.am mugs: mcamson mma .¢m.am.fi> .mm.¢mlmm .mmnom.mm m¢.ma.m pmp30u0n6>o mo usmonmm mma.mda .sda.mdd .mdalaaa .moaumoa .m>.m>.m> muflcs mm mm.mm.mm mm.mm.am mcamson pmflmsooo mmfi .mm.sm.mm .mm.am.md mmuam.mm .om.mm.a mpan3uaoc mo unmoumm mos mmuom we .ms.ms.mm mm.mm.am No.4m.m4 sm.m¢.ma mugs: magmson mma am.am .am.em.oa .sa.mm.m m«.~m.m pumpcmumnsm no unmouwm .m N. a- m+ a+ mCOHumw>mQ pumpcmum mmHQMHum> mcflmsom .o.o .COHDGHSmMB umOHQMHHm> GCHmdom UOHUOHOW GO mQOHUMfl>OQ @HMGCMUM N .HO d u...- mcapmmoxm momsmno smmz nuHB muomue mamsmo wnu mo HmQEsz coaumoHMHucmpH .MHHHNN OHQMB 204 coaumadmom Hmuou .ucmu mos mm omfi.afi.m ma uumuucou cagpmz mafia: mcflmson pmp30uuuw>o mm mos 40d oma.soa mugnzucoc «0 ucmuumm mma mm Nma .amd.mdd «0a .mm.¢m.mm .omfi.moa .soa.moa muses mcnmsos mmd .omusm .Hm.mm.m .moa.¢m.mw mm.ohlmm pmp3onuum>o mo unmoumm mm muflcs .mm.¢m mm mma mcflmson pmwmsooo mmfi .mmlnm .mm.>m.m¢ oma.¢m moa.moa muw£3lcos mo unmoumm moa.moa muflcs mcflmson mma ww.>m moa.¢m .moa.m¢ pumpcmquSm mo unmouwm .m m: an m+ H+ mcoHDMH>mQ pumpcmum mmHQMHHm> mcflmsom endommflz .mHsoq .um unmanmwum> mcflmsom pmuomamm co mCOHUMH>mQ pumpcmum N no H.H mcflpmmuxm mwmcmnu cmmz nuHB muumue msmcmo may mo Hwnfisz GOHDMUHMHucmpH .QHHHXX OHQMB 205 moa.aoa mm.mm .ma.om.sm coaumasmom Hmuou .ucmu am .ma.m .nm.maas¢ am.mm.om ss.s.m pomuucoo cmauws mafia: mcflmsos pmpzououm>o 4N ss.mm s mugnznco: mo unmoumm mos mmumm mugs: manager Hag .sm.ss .aaumm.am mm.mnm om.m.m.a wmu3ououm>o mo unmoumm hm.>> nuns: .mm.sm mm.mm.¢m manages cmamuooo Hag .aaumm .mm.ma.aa m man31aoc mo unwuumm ea.ma mm.mm.«a mung: manage: Hag .ma.s.m ma-m.m.¢.a oumocmumnsm mo unmoumm m N: .7 m+ a+ mcofluma>ma pumpcmum mmanmwum> mcflmsom oazo .HUMGCHUGAU "mmHQmanm> mcflmsom pmuomamm co mcoquH>mQ pumpcmpm N no d.H mcfipmmoxm mmmcmnu cmmz nuHB muomue mamcmo wnu mo Hmnfisz coauMUHMHucmpH .UHHHNN magma 206 NNH.>Hd .maa.maanmaa .aaa.moa.¢oa mag .mm.am.ms.sm mm.am.oo mm .mm.mmumm .mm.mm.am.ma «m.mm.mm .mmumm.m.a mpflszacoc no ucmuumm oma mma NNH.mHa .bfia.0dfi.mm coaumasmom .mm.ma.ma .mm.mm.am .me.mm.mm.m¢ Hmuou .msoocg oma .ma.ga.m .am.mm.mm 55.4m s¢.m¢.om.m Sagan“ cmgoms smumm so .mmanoma.moausm mm.wm.am .mm.ow.mm mm .mmumm.amuss ma .am.mm.mm .am.m¢ua4 coHumHsmom Hmuoe mmauoma.moa was num.mmnmm .maaumaa mm coauma mm .smumm.am .4oa.mm.mm so .mm.m¢.s4 unmom muanzucoz mm.am.ss maa.mm .ms.mm.ma msoocg sagsmm so aoa.mm.mm .sm.mm.ms an .ma.¢a.om cmgums muassucoz mm.om coaumuspm mm as .sm.am.mm am mm.mm.om canoes muassncoz mousm.mo mm.mm .om.mm.mm .mm.mm.ms conumHsmom Hmpou HNH wm.mm.m .mm.m.m.a «m.om .mm.mm.¢m .coHumosum cmacmz .m m- a- m+ a+ mc0fluMH>wQ pumpcmum mmHQMHHm> coaumHsmom .U.Q .coumcanmmz unmanneum> coaumHsmom pmuomamm co mCOHuMH>mD pumpcmum N no H.H mcflpmwoxm mmmcMBU cmmz nufl3 muomue upmcmu mSu mo quEsz coaumoflwflucmpH .m>Hxx manme 207 mm.mm mm mma .«m.mmnsw .mm.sm.ma mma mugssnaoc mo ucmuumm MNHIaNd.m> boa cowumHsmom .mm.m¢.mm .moaINOd Hmuou .mEOUCH was .4a.ma.m .mm.mm.mm madam“ cmacmz maa ¢N6.¢HH we .mOdlmoa .dfia.mOH mNa woa.a .oa.m.N .Noa.¢m .moa.mmlmm coaumHsmom Hmuoa floa.mm mm .¢m.mmumm .Nm.>m.mm moa.moa coauma mNa .am.mN.mN .Nm.om.>N .NOH.¢m moa.moa Ismom mua£3lsoz mEOUCH aaflsmm 4a mw.mm.mm mOH.NOd swapme muHSBIcoz coaumospm ON 40a mm.mm moa.mm.mm GMAme mufl£3lcoz coaumHsmom Hmuou NNH N $06.0fi.a mm.mN.mN ¢¢.>N .COHumospm cmflpmz .m N: H: N+ H+ msofiuma>mn pumpcmum moanmaum> cowumasmom III! I Iv I Husommflz .mflsoq .um "mmHQMHum> cowumHsmom pwuomamm co mGOHDMH>mQ pumpcmum N no d_fi mafipmmoxm mmmcmno cmmz nufl3 muomua msmcmu map MO qu832 COHDMUHwHucmpH .Q>HNN magma 208 sm.ss mm.am gag .mmusm.a¢-mm .Nm.am.¢a m muflnsucoc mo unmoumm moa.moa coaumgsmom .aoa.ms.as am Hmuou .msooca em mm.ma .mm.om.ma .om.ss.m ma.m magsmm cmgnms mos mos .Nafi.oaa mm.ma gag .ooa.mm.m¢ .mm.ss.mm sm.m.m .afi.mum.a coaumasmom Hmuoe as.mm sm.mm coauma flag usm.mm.mm .aa.oa.mm m.4.m m usmom mugn3ucoz msooaa.sflaawm oa mm m canvas mugs3ucoz COHHMUSUO «a 44.4m sm.m canvas mugn31coz moa.moa .ooa.mm.mm mm.ma.oa aoa .am.mm.ms om .mm.mm.mm aoflumasmom Hmuou mos .mm.mm.m .mm.s¢.m mo-mo.mm .mm.ma.n .:0gumosom cmaomz .m NI a: N+ a+ mCOHDMH>mQ pumpcmum mmHQMHum> cofiumHsmom 4' I Oflno .Humcsflocflu "mmHQMHHm> coaumHsmom pmuomamm co mQOHUMH>mQ pumpcmum N no H.fi mcapmmoxm mmmcmnu cmmz EDHB muumua mSmch may mo umnfidz GOHUMUHMHucmpH .U>HNN mHQMB APPENDIX F 209 H.mm m.¢¢ m.m¢ n.mm OHQO .HHMGCAUCHU m.mm N.Hm $.mm m.mm .02 .mHSOA .um m.am o.Nm N.mm o.mm .U.Q .soumCASmmz omma omma omma omma mcam muflv Hmnpcmo «Sm Amucmoumm :HV omma paw omma .mcflm paw muflu Hmuucmu cwaDmm coaumHsmom mo GOHmH>HQ one .H>NN magma .Ammma .umnfimomm .nonmmmmm coaumasmom now coaumpcsom mmmwuom "OHSO .puomxov 210 .ommfiloomfi 4mmm~4 capaaoaouumz pumpcmum CH £u3ou0 coaudemom .msmom .b tampon "mousom H.m¢ m.om m.m 0.4m 4.0 m.ma mags m.ou m.o« o.H ¢.ma 4.0a o.aa Mano Hmuucmu m.ma m.¢a H.¢ N.ON m.w o.Nd Humasflusflo o.mm m.mm 8.4a m.m¢ m.ma m.os mags m.NHI o.m >.ou m.o m.NH ¢.ma huflo Hmuucmo m.ma w.na m.m m.ma m.ma m.mN masoq .um o.Nm a.eaa m.¢m o.mm m.ea m.wa mcan w.¢u o.aN N.mm m.aa N.Nm m.ma muflo Hmwucmo >.mm m.am 0.66 mesa «.mN w.nd coumcflnmmz omnomma omuowma owuomma omnommd omuoama oanooma mpmomo dzm Amucmoumm Gav m.<2m mmune Ga mwpmomn m>Hm Hm>o mmmmuocH coaumHsmom .>Xx mHQMB 211 Table XXVII. Non—white Population in the Ring and Central City, 1960 (in percents) Proportion Proportion SMA non-white non-white in the SMSA in the ring Washington 24.54 6.09 St. Louis 14.51 6.14 Cincinnati 11.96 5.40 MICHIGAN $an5 UNIV. LIBRARIES I)WIWIll)VINIIIWIIIII“)IHIHIIIHWHHI 31293104807734