Histories of the emergence of federally sanctioned mortgage underwriting in the United States have made racial discrimination a central theme. Yet their emphasis on systematic favoritism for whites over blacks tells only part of the story. This article examines how a broader, waning concept of race that encompassed nationality as well as skin color influenced the practice of locational risk rating at the Federal Housing Administration. Using qualitative and quantitative methods to examine the construction of “neighborhood risk” reveals how, from the introduction of the agency’s underwriting standards, raters depended on a more nuanced population classification scheme in evaluating urban areas. The evidence presented here about the differential treatment of nationality groups points to new directions for studies of housing policy in the United States.
Joint Committee on Appraisal and Mortgage Analysis, Proceedings of the National Appraisal Forum Held Under the Auspices of the Joint Committee on Appraisal and Mortgage Analysis, Washington, D.C., November 19 and 20, 1937 (Washington, DC: The Committee, 1937), 55. This meeting, organized by the Joint Committee on Appraisal and Mortgage Analysis of the U.S. Central Housing Committee, brought together representatives from the real estate industry and federal housing agencies to explore "the scope and practical significance of appraisal data." "National Appraisal Forum [CHC 1024]" in box 2, entry 8 PI-164, folder National Appraisal Forum, record group 207, Records of the Housing and Home Finance Agency, Records of the Central Housing Committee, Records of the Committee on Appraisal and Mortgage Analysis, 1935-1942, National Archives and Records Administration. Note that with the American Institute of Real Estate Appraisal being part of the National Association of Real Estate Boards, Atkinson also served as director of divisional activities for the latter organization.
2.
The literature on underwriting at the Federal Housing Administration (FHA) and the Home Owners Loan Corporation (HOLC) covers a range of topics, from redlining, to restrictive covenants, to steering (where to direct prospective buyers), to the role of agency policies in fostering suburban development. Many of these accounts suggest important links between past and present and how government appraisers left a legacy of segregation and inequality still visible on the American landscape. See, e.g., Kenneth T. Jackson, "Race, Ethnicity, and Real Estate Appraisal: The Home Owners Loan Corporation and the Federal Housing Administration," Journal of Urban History 6, no. 4 (1980): 419-52; Amy Hillier, "Redlining and the Home Owners’ Loan Corporation," Journal of Urban History 29, no. 4 (2003): 394-420; Kristin Crossney and David Bartelt, "Residential Security, Risk, and Race: The Home Owners’ Loan Corporation and Mortgage Access in Two Cities," Urban Geography 26, no. 8 (2005): 707-36; Calvin Bradford, "Financing Home Ownership: The Federal Role In Neighborhood Decline," Urban Affairs Review 14 (1979): 313-35; John Metzger, "Planned Abandonment: The Neighborhood Life Cycle and National Urban Policy," Housing Policy Debate 11, no. 1 (2000): 7-40; Guy Stuart, Discriminating Risk: The US Mortgage Lending Industry in the Twentieth Century (Ithaca, NY: Cornell University Press, 2003); Michael Jones-Correa, "The Origins and Diffusion of Racial Restrictive Covenants," Political Science Quarterly 115, no. 4 (2000-2001): 541-68; David Freund, Colored Property: State Policy and White Racial Politics in Suburban America (Chicago: University of Chicago Press, 2007); Kevin Fox Gotham, Race, Real Estate, and Uneven Development: The Kansas City Experience, 1900-2000 (Albany: State University of New York Press, 2002). Such critical analyses date back much earlier, e.g., Charles Abrams’s classic Forbidden Neighbors: A Study of Prejudice in Housing (New York: Harper, 1955).
3.
On the 1920s and 1930s as a period of transition in Americans’ ideas about racial categories in scientific and popular culture, see Elazar Barkan, Retreat of Scientific Racism: Changing Concepts of Race in Britain and the United States between the World Wars (New York: Cambridge University Press, 1992); Matthew Frye Jacobson, Whiteness of a Different Color: European Immigrants and the Alchemy of Race (Cambridge, MA: Harvard University Press, 1998). For a discussion of the earlier views of hierarchies among population groups and their public policy implications, see Stephen Jay Gould, The Mismeasure of Man (New York: Norton, 1981); Garland Allen, "Eugenics Comes to America," in Russell Jacoby and Naomi Glauberman, eds., The Bell Curve Debate: History, Documents, Opinions (New York: Times Books, 1995); and Diane Paul, Controlling Human Heredity, 1865 to the Present (Amherst, NY: Humanities Press, 1995), esp. chap. 6. Standard assessments of the "eugenic problem" of urbanization in previous decades had pointed to how "the great cities are a sterilizing agency, selecting the best part of the population and cutting down its racial contribution." See "Urban Sterilization: Effect of City Life in Cutting Down Increase of a Superior Population Is Shown by New Figures- Native-Born American Stock Fails to Hold Its Own in Most Parts of the United States," Journal of Heredity 8 (June 1917): 268.
4.
Gould’s account of the history of eugenics emphasizes how eugenicists employed statistical reasoning to press their case. Some of the long history of ethnographic cartography is covered in Jeremy Crampton, "Maps, Race and Foucault: Eugenics and Territorialization following World War I," in Jeremy Crampton and Stuart Elden, eds., Space, Knowledge and Power: Foucault and Geography (Aldershot, UK: Ashgate, 2007), 223-44. While a few scholars have pointed to connections between eugenical ideas and urban development in Europe (e.g., Wolfgang Voigt, "The Garden City as Eugenic Utopia," Planning Perspectives 4, no. 3 [1989]: 295-312), direct influences of these biological scientists in the realm of U.S. urban policy remain open to further investigation.
5.
Herbert Nelson, The Administration of Real Estate Boards (New York: MacMillan, 1925), 191. As Nelson’s reference to sociologists indicates, the real estate industry was by no means the only profession tracking the circulation of nationality groups using maps. From social scientists developing urban theory, to newspaper and telephone companies analyzing present and future subscription patterns, to social workers building on the precedent of Jane Addams and her collaborators, other individuals and institutions were seeking similar information about the distribution of nationality groups in U.S. cities during this period, albeit with a different view of the significance of these distributions. Real estate professionals clearly made efforts to review many of these other sources, as comments by several National Appraisal Forum participants attest. See Joint Committee on Appraisal and Mortgage Analysis, Proceedings of the National Appraisal Forum.
6.
Herbert Nelson , "The Real Estate Code of Ethics," Journal of Land and Public Utility Economics1, no. 3 (1925): 273-74.
7.
Homer Hoyt, One Hundred Years of Land Values in Chicago ( Chicago: University of Chicago Press, 1933), 4.
8.
Ibid., 314.
9.
Ibid., 442.
10.
Consistent with the view that population rank and economic rank were closely intertwined, Hoyt interpreted assimilation in economic as well as sociological terms. As he put it, "Racial and national barriers disappear when the individuals in the foreign nationality groups rise in the economic scale or conform to American standards of living." Ibid., 314.
11.
A number of examples of such maps are found in cartographic collections at the U.S. National Archives. See, e.g., "East Orange, New Jersey. Distribution of Colored Populations & Italian Population in Ward III, 1932," in folder City Maps: New Jersey, General, Bergen County, Camden, East Orange, etc. record group 31, FHA, U.S. National Archives and Records Administration Cartographic Collections. There is evidence of earlier plans by the California Real Estate Board to prepare such a map on a statewide scale. As secretary to the board’s president Harry Allen wrote in a letter to Stanford professor Eliot Mears, "I shall also make a map of California, showing the distribution of foreign races by different colored crayons; as I believe this will be a valuable asset for our files." Letter to Eliot Mears from Serena Bland Preusser, March 5, 1927, box 21 office file 1914-1927, Archives of the Survey of Race Relations, Hoover Institution.
12.
George C. Olcott, Olcott’s Land Values Blue Book of Chicago ( Chicago : George C. Olcott, 1935 ), 24.
13.
On the broader effort by the industry to raise its work to the status of a predictive science, see Jeffrey Hornstein, A Nation of Realtors ( Durham, NC: Duke University Press, 2005).
14.
Scholars have documented close ties between the new agency and the growing real estate profession, not limited to matters of appraisal. See, e.g., Pearl Davies, Real Estate in American History (Washington: Public Affairs Press, 1958); Stuart, Discriminating Risk; Marc Weiss, "Richard T. Ely and the Contribution of Economic Research to National Housing Policy, 1920-1940," Urban Studies 26, no. 1 (1989): 115-26; Hornstein, A Nation of Realtors. Accounts to date of the differential implications of FHA policies for whites versus blacks in the United States, and the agency’s commitment to maintaining homogenous residential communities through steering, restrictive covenants, and lending decision making, are part of this story.
15.
Mortgage risk assessment was the job of FHA staff in local insuring offices. This field organization is described in "The Field Organization of the FHA," Insured Mortgage Portfolio (May 1940): 17, 22; a map and list of insuring office locations was frequently printed in Insured Mortgage Portfolio in this period, as well as in the Underwriting Manual itself. Washington headquarters led underwriting schools for the agency’s army of appraisers to further encourage the homogeneity of approach that the new system prescribed. Like mortgages, students were also rated on a scale with implications for their future advancement within the FHA system as "The Training of FHA Underwriters," Insured Mortgage Portfolio (May 1939) explains. Thanks to the wide circulation of the Underwriting Manual, while the primary audience for the manual was federal employees, its techniques were more widely adopted, "voluntarily incorporated into the lending practice of many banks and other mortgage-lending institutions." "A Plan for Analyzing Mortgage Risk," Insured Mortgage Portfolio (May 1939): 16.
16.
A concise summary is presented in "A Plan for Analyzing Mortgage Risk," 16. For a discussion of how this valuation scheme compared to a range of other approaches that real estate appraisers worked with in the 1920s and 1930s, see Stuart, Discriminating Risk.
17.
Similar grading schemes were used in other financial risk evaluation systems. As FHA head Stewart McDonald noted with pride, "What Dun & Bradstreet and the checking with reserve city correspondents are to the commercial loans of a bank; what Moody’s Manual, Standard Statistics and similar rating agencies are to investment securities, the FHA risk-rating system can and should be to the eligible home-mortgage loans of an institution." Stewart McDonald, "Steps Toward a Sound Mortgage Market," Insured Mortgage Portfolio (June 1937): 4. The notion of grading real estate transactions also had precedent in the industry: the era’s multiple listing services assigned A to D grades to listings as indicators of the projected ease or difficulty expected in selling; the grades judged how well the asking price matched appraisers’ consensus as to the property’s value.
18.
On economic background ratings, see "Schedule C, City Structure Report," attached to "Program for the Study of Sixty-Two Cities," in Records Relating to the Economic Data System, 1936, 1941, 1943 box 1, PI-164 entry 30, record group 207, Records of the Housing and Home Finance Agency, Records of the Division of Research and Statistics of the FHA, "Economic Data System" Records Containing Data on Cities, 1937-1945, National Archives and Records Administration. A copy of the 1935 questionnaire for evaluating individual borrowers is reprinted in Stuart, Discriminating Risk, 93.
19.
The two-part review process is described in Joseph Graham, "FHA Underwriters and Their Functions," Insured Mortgage Portfolio (April 1937): 15-16, 24.
20.
The first edition of the manual was issued in 1934; four revisions were prepared by 1938.
21.
On real estate research methods in general, see sec. 501-99, FHA, Underwriting Manual (Washington, DC: FHA, 1935). On mapping mortgage risk more specifically, see sec. 507 as well as the broader discussion under the heading "Control of Neighborhood Ratings." While previous studies have not treated these maps as tools for data processing, it is clear from FHA materials that agency officials, victims of their own successes promoting the new mortgage insurance program, saw their value in such terms. As in the histories of credit reporting and life insurance described by historians of technology, the need to make sense of an ever-increasing volume of data in this rapidly expanding field prompted the adoption of a variety of "mechanical aids" and "automatic devices" to improve the assessment of economic risk. The functioning of not only risk maps but also the entire risk-rating system in such terms is apparent in Ernest Fisher, "Can We Immunize Mortgages," Insured Mortgage Portfolio 1, no. 1 (1936): 16; as well as in the 1938 revisions to the Underwriting Manual, which included a new chapter on "Compilation and Recordation of Data." The histories of these other financial risk assessment systems are documented in Josh Lauer, "From Rumor to Written Record: Credit Reporting and the Invention of Financial Identity in Nineteenth-Century America," Technology and Culture 49, no. 2 (2008): 301-24; and JoAnne Yates, Structuring the Information Age: Life Insurance and Technology in the Twentieth Century (Baltimore: Johns Hopkins University Press, 2005).
22.
Letter from Homer Hoyt to Dr. Fisher July 22, 1935, "Instructions as to procedure in rating neighborhoods." In carton 18, folder 0-109, "Neighborhood Rating Technique: Miscellaneous," record group 31, entry 6, Research and Statistics, National Archives and Records Administration.
23.
Hoyt held several positions at the federal housing agency during the 1930s, including chief of the Section on City Growth and Structure and later principle housing economist, both in the economics and statistics division. Robert Beauregard, "More than Sector Theory: Homer Hoyt’s Contributions to Planning Knowledge," Journal of Planning History6, no. 3 (2007): 248-71, describes Hoyt’s broad career.
24.
"Plan for Analyzing Mortgage Risk," 28. The several versions of Hoyt’s instructions to raters cited in this article, all of which prompted first the identification of worst neighborhood, second the identification of best neighborhood, and third the identification of intermediate-quality neighborhoods, made clear how grading rather than scoring was his primary concern. In the agency’s view, taking such alternate routes to arriving at a risk score for applicants not only did not interfere with the risk-rating system but also provided valuable additions to the underwriters’ toolkit that would "serve to correct any tendency to treat the features and the system as a fetish, and will tend to orient and control judgment in connection with ratings. The minutiae in the system cannot be significant in the absence of broad judgments, on the other hand, broad general judgments are dangerous in that they may fail to give sufficient consideration to important details. Both approaches are necessary to a correct rating." "Plan for Analyzing Mortgage Risk," 28.
25.
An overview of the entire locational risk-rating procedure for all participating appraisers is found in Homer Hoyt, "Exhibit J: The Preparation of Maps Showing the Dynamic Factors of City Growth: Prepared for the Research Division for the Underwriting Staff," January 8, 1935, appended to "Program for the Study of Sixty-Two Cities," in Records Relating to the Economic Data System, 1936, 1941, 1943, box 1, PI-164, entry 30, record group 207, Records of the Housing and Home Finance Agency, Records of the Division of Research and Statistics of the FHA, "Economic Data System" Records Containing Data on Cities, 1937-1945, National Archives and Records Administration.
26.
Homer Hoyt, "Instructions for Dividing the City into Neighborhoods," undated, carton 18, folder 0-109, "Neighborhood Rating Technique: Miscellaneous," record group 31, entry 6, Research and Statistics, National Archives and Records Administration, offered guidance to local insuring offices. While the Underwriting Manual points to the research department or valuation section in insuring offices as overseeing the preparation of these maps, Charlotte Munchmayer, Preliminary Inventory of the Cartographic Records of the Federal Housing Administration, record group 31 (Washington, DC: National Archives, 1952), reports for some offices the existence of a distinct "neighborhood rating unit."
27.
Hoyt, "Instructions for Dividing the City"
28.
Ibid.
29.
Fisher, "Can We Immunize Mortgages?" 17. Fisher’s description of the biological aspects of urban growth is indicative of the dominance of an ecological understanding of cities in real estate and federal housing circles during this period. The priority that FHA staff ascribed to ecological interpretations of urban patterns is apparent in numerous articles appearing in the agency’s journal during this period. See esp. John Magee, "City Growth and Population Patterns," Insured Mortgage Portfolio (September 1937): 6, 22.
30.
Hoyt, "Exhibit J," 6.
31.
That these practices built on his own earlier work is apparent in Hoyt’s later report detailing location rating around the United States, where he cites One Hundred Years of Land Values in Chicago in several places. See Fha, The Structure and Growth of Residential Neighborhoods in American Cities (Washington, DC: FHA, 1939).
32.
Hoyt, "Exhibit J," 2. While in this document he called for men "engaged in such occupations as real estate, social work, surveying, or medicine that would naturally tend to being them into actual contact with the various neighborhoods to be mapped," elsewhere Hoyt expressed a clear preference for real estate professionals, for example, instructing insuring officers to make contact with the secretary of the local real estate board as the first step in the recruitment process. Consistent with Hoyt’s earlier work (as acknowledgments and footnotes attest, local real estate appraisers and others supplied much of the relevant information), this decision to partner with members of the real estate industry also satisfied the FHA’s desire for close collaborations with these professionals to ensure the success of its new underwriting system and to outsource some of the time-consuming collection of local statistics. The agency’s interest in securing their cooperation in obtaining statistical data is a theme in several articles by the assistant director of the Division of Economics and Statistics, James S. Taylor. See James S. Taylor, "Where Is Mortgage Research Leading?" Insured Mortgage Portfolio (January 1938): 24; and James S. Taylor, "Factual Aids for Mortgage Lenders," Insured Mortgage Portfolio (May 1937): 13-15, 21. To be sure, these men were not from the parallel immigrant real estate industry that Everett Hughes noted for Chicago in his dissertation, published as Everett Hughes, The Growth of an Institution: The Chicago Real Estate Board (Chicago: University of Chicago Press, 1931).
33.
Full instructions to these participants are available in "Maps of Dynamic Factors," contained in the folder with the maps prepared by contributors in Clarksburg, Virginia. Folder City Maps West Virginia Clarksburg PI 45, entry 354 (330/10/31/4-3), National Archives and Records Administration Cartographic Collections.
34.
Hoyt, "Exhibit J," 2.
35.
Ibid., 2.
36.
In light of the scientific pretensions of the underwriting scheme, a point of obvious anxiety for Hoyt was the fact that average rent was the only measure presented on all the maps. Hoyt had used rent as an index to neighborhood quality in his 1933 book but appeared concerned it might seem insufficient as an alternative to the quantitative methodology of the underwriting grid. Seeking to balance the ideal of an objective formula for risk assessment with the reality that comprehensive economic and social data were not available for every city, he repeatedly articulated how average rent was a variable encompassing other variables, suggesting that layering the presence or absence of other desirable and undesirable factors on average rent constituted a sound approach to risk assessment. Homer Hoyt, "City Growth and Mortgage Risk," Insured Mortgage Portfolio (December 1936): 10. Also see James Taylor, "A New Map Aid for Mortgage Lenders: Block Data Maps, One of FHA’s New Factual Aids to Mortgage Lending, Have been Described as ‘Magic Carpets’ for Mortgage Lenders; as the Name Implies, They Permit Instant Visualization of Pertinent Facts Concerning Given Localities" (June 1937): 8. The questions posed for each neighborhood enabled cities with more data on hand, for example, details on foreclosure rates, average age of occupants, distance to parks and recreation, and ease of access to municipal services, to factor these variables into the risk-rating equation. Notably, the emphasis on rent was a change from the 1935 Underwriting Manual; this early FHA statement about the importance of maps in location rating similarly had recognized that a systematic assessment incorporating answers to every one of the manual’s questions might not be feasible and so had laid out age of buildings as the equivalent measurable detail.
37.
Homer Hoyt, "Springfield, Massachusetts," 2, 3, in carton 18, folder 0-109, "Neighborhood Rating Technique: Miscellaneous," record group 31, entry 6, Research and Statistics, National Archives and Records Administration.
38.
Although faculty and graduate students at the University of Chicago sociology department get much of the attention, colleagues in geography and economics, including Homer Hoyt, did important studies as well.
39.
Thomas Philpott, The Slum and the Ghetto: Neighborhood Deterioration and Middle-Class Reform, Chicago, 1880-1930 (New York: Oxford University Press, 1978). An overview discussion of the different experiences of Chicago’s immigrant populations in the early twentieth century can be found in Maurice Halbwachs, "Chicago, expérience ethnique," Annales d’histoire économique et sociale 4 (January 1932): 11-49.
40.
"Red Tape Still Delays Action of New FHA Unit," Chicago Daily Tribune, November 25, 1934, 18. In a testament to the close relations between FHA and the real estate industry, the Chicago district FHA office was headed by Carroll Sudler, principal of the self-titled firm.
41.
John Metzger identified another copy in the Ernest Burgess papers at the University of Chicago Archives. See John Metzger, "Social Capitalism in American Cities" (PhD diss., Columbia University, 1999).
42.
See, e.g., in Hoyt, "City Growth and Mortgage Risk," Insured Mortgage Portfolio (January 1937): 5-7 and his later report, Structure and Growth.
43.
As a result, for these cities (Washington, Chicago, Miami, New York, and Detroit) the present-day dynamic factor maps, gathered from sources similar to those used in the basic maps insuring officers prepared, did not provide a comparable check on accuracy. On the planned uses for Real Property Inventories in risk rating, see "Program for the Study of Sixty-Two Cities," in Records Relating to the Economic Data System, 1936, 1941, 1943, box 1, PI-164, entry 30, record group 207, Records of the Housing and Home Finance Agency, Records of the Division of Research and Statistics of the FHA, "Economic Data System" Records Containing Data on Cities, 1937-1945, National Archives and Records Administration. While organized by the commerce department, the survey work of the inventories was carried out by the Works Projects Administration.
44.
These uses are evident in stamped names of former owners of the volumes this author acquired from used bookstores, which included several banks and real estate firms, as well as in the many pages of advertisements that Olcott contained. Homer Hoyt, who had helped to publicize the value of this source in his 1933 book (albeit before Olcott had included race and nationality information), took out advertisements in later editions for his real estate consulting firm. See Olcott, Olcott’s Land Values, A13. Of special interest to these users was how the blue book provided detailed instructions on how to calculate the value of individual structures within the city and surrounding region. Other prominent users of the blue book included sociologists. At the University of Chicago, Ernest Burgess had turned to Olcott’s work as a source in his efforts to construct a scientific urban theory, suggesting its value for obtaining the "more precise and accurate picture of the anatomy" of this city that FHA staff sought. See Fisher, "Can We Immunize Mortgages?" 17; Ernest Burgess, "Residential Segregation in American Cities," Annals of the American Academy of Political and Social Science 140 (1928): 105-15. Reflecting the wide influence of Burgess’s work, from the 1933 edition Olcott’s classifications also included "ghetto" and "rooming house" areas. Olcott’s book, especially in more recent editions, has been widely used by urban economists. See, e.g., Daniel P. McMillen and John F. McDonald, "Urban Land Value Functions with Endogenous Zoning," Journal of Urban Economics 29, no. 1 (1991): 14-27.
45.
Olcott provided land values by block with the assumption that purchase and rental prices could be calculated from these base rates. Rental value, housing value, and land value were used interchangeably in many studies of the era. For example, Hoyt repeatedly elided rental and land value in his 1933 book; in the 1934 special census of Chicago, maps of the city’s economic areas combined rental and purchase information (see Charles Newcomb and Richard Lang, Census Data of the City of Chicago, 1934 [Chicago: University of Chicago Press, 1934].) Notably, Olcott’s catalog of land use in Chicago made fine-grade distinctions among types of buildings, for example listing twelve varieties of single-family homes, because their differences were relevant to appraisers seeking to compute the value of individual structures. For FHA location ratings, broader brush classes were used. Therefore, after coding each individual building type Olcott reported, the author’s data set collapsed these varieties into categories that matched instructions to risk raters, for example, to outline all areas with single-family homes.
46.
Chicago’s grid orientation was well suited to these geographic divisions. There were a few exceptions to the standard half-mile by half-mile sections in new growth areas on the city’s southern edge, but these were recalculated to conform to the half-mile standard. Omitted areas within city boundaries were those covered by parks, cemeteries, or bodies of water. That Olcott’s divisions were synonymous with how real estate appraisers viewed the city was apparent in the fact that the "Section Map of Chicago" appearing in the era’s weekly issues of Real Estate: The Weekly for Owners, Realtors, Architects, Builders replicated Olcott’s division of the city into half-mile squares.
47.
Magee, " City Growth and Population Patterns," 6. This variable is named after the term used in the Underwriting Manual to describe mixed areas.
48.
Confident that the use of dynamic factor maps succeeded in "reducing to a measurable science these factors that hitherto could not be calculated for lack of data," Hoyt laid out seven basic principles of city growth to speed risk raters’ interpretations-first in a series of articles in Insured Mortgage Portfolio (all titled "City Growth and Mortgage Risk" and published between December 1936 and April 1937) and later in an extended report (FHA, Structure and Growth); they were also anonymously reprinted in the 1938 Underwriting Manual (sec. 909; also see Hoyt, "Exhibit J," 6). His summary comments emphasized two points of special concern. The first was relevant to all cities: that desirable residential areas, which tended to move outwards over time on fixed lines toward the urban periphery, left deteriorating neighborhoods in their wake. The second was relevant to industrial cities (and here he used Chicago as his example): that the least desirable residential neighborhoods tended to move outward together with the expansion of industrial areas. As a result, the greatest predictors of areas of declining value were those at the rear of the march of high-grade residential areas and those at the fore of industrial march. (For nonindustrial cities, raters simply were asked to look at the likely expansion of the lowest grade residential areas.) With Hoyt citing his 1933 book alongside the dynamic factor maps shown in The Structure and Growth of Residential Neighborhoods in American Cities as a key source of information, these variables were calculated based on the original maps of expanding industrial and residential areas that Hoyt first created for his 1933 book. Hoyt relied on Olcott as a staple of his research for One Hundred Years of Land Values in Chicago, meaning these maps, seen in Figures 1 and 2, already were keyed to the half-mile sections that Olcott employed. The following calculations comprise the dynamic factor variables: First, the author converted the information in Hoyt’s maps into two data sets, one for the expansion of industrial areas and the other high grade residential areas-both divided into the four generations that Hoyt had tracked. To pinpoint areas of future risk on account of the movement of industry, "df,ind" required identifying among the most recent generation of industrialized areas outlined by Hoyt the adjacent Olcott half-mile by half-mile squares. Of those squares, removed from consideration were those that overlapped with the past three generations of industrial development given the limited likelihood of further expansion. The squares that remained were scored as areas of future risk. For areas of future risk because of the movement of high grade residential areas, "df,hgr," the author outlined, from among the most recent generation of high grade areas, those which were closest to older high grade areas and therefore at greater risk of decline. To do this required identifying all squares adjacent to the oldest three generations of high grade area expansion. Then, among those squares, those that overlapped with the most recent generation of high grade expansion were scored as areas of future risk.
49.
While mortgage risk maps streamlined much of the location rating process, they did not totally automate local evaluations, even on preliminary review. In Chicago as in other cities, the risk district map also included some unmarked, unclassified areas, mostly territory inside city limits as of yet undeveloped, for which individual location ratings would have to be pursued.
50.
In a few cases Olcott pointed to finer distinctions among the locations of racial and nationality groups within an individual half-mile by half-mile area such as "Pol west of RR, Lit east of RR." Where possible this information was matched to the multiple risk observations within a single half-mile square.
51.
These analyses were regressions run in Stata, the former using ordinary least squares for an analysis on the dependent variable "risk" and the latter a logistic regression for an analysis on the dependent variables "riskfha," "riskfhab," "riskfhac," and "riskfhad."
52.
Based on statements from instructions to risk raters about the importance of immediately adjacent blocks, the spatial autocorrelation control was calculated based on contiguity and Manhattan distance.
53.
This study thus builds on the growing body of scholarship using GIS for urban historical analysis. For a review of some of the literature in this area, see Anne Kelly Knowles and Amy Hillier, Placing History: How Maps, Spatial Data, and GIS Are Changing Historical Scholarship (Redlands, CA: ESRI Press, 2008).
54.
While the regression shows D class to be significant at .05, since the confidence interval crosses zero this finding was discounted.
55.
Surprisingly, single-family homes were more positively correlated with risk than were apartments, perhaps because of the presence of numerous luxury apartment buildings in the city near the lakefront where many of the A class areas lay.
56.
While it did not show up in regressions, all commercial areas were classed as D, consistent with warnings in the Underwriting Manual about the problems of mixing residential and nonresidential uses.
57.
Joint Committee on Appraisal and Mortgage Analysis, Proceedings of the National Appraisal Forum, 71. This assessment appeared to hold even in light of some anti-German prejudice across the United States associated with the rise of National Socialism in Germany.
58.
Hoyt, One Hundred Years , 314.
59.
That the strength of the negative correlation with C areas for Poles was less than for Negroes is likely an artifact of the decision to leave many of the city’s least developed areas unclassified. For looking more closely at where exactly Chicago’s Polish immigrants resided finds one-third living in half-mile by half-mile sections that the FHA left wholly unclassified (63 out of the 187 they inhabited) -leaving these populations out of the calculations of comparative desirability represented by the master map. As the Underwriting Manual explained, undeveloped areas would in many cases have low risk ratings suggesting that Poles might ascend the risk index once these areas were classified.
60.
Several other racial and nationality groups tracked by Olcott-Austrian, Chinese, and Croatian- resided exclusively in D-class areas, but they did not show up in any of the regressions as statistically significant.
61.
Hoyt, One Hundred Years , 442.
62.
Ibid., 316. The omission from this list of other groups tracked by Olcott, specifically his categories of Austrian, Chinese, Croatian, and Holland, is likely explained by their comparatively small numbers in the city.
63.
Magee, " City Growth and Population Patterns," 6, 22.
64.
There are two possible explanations for this development. The first is an actual shift in perception of the two groups by the local real estate community. With Poles and Lithuanians already close to each other in the local hierarchy of races and nationalities, such a change certainly was a possibility. A second explanation is that in fact there was no change in their relative value in the eyes of real estate professionals and that Poles’ apparent slippage was instead an artifact of the decision to leave some of the city’s least developed areas unclassified. See note 59.
65.
E.g., Abrams, Forbidden Neighbors; Wendy Plotkin, "Deeds of Mistrust: Race, Housing and Restrictive Covenants in Chicago, 1900-1953" (PhD diss., University of Illinois, 1999); Stuart, Discriminating Risk.
66.
FHA, Underwriting Manual, (1935 edition), sec. 509, 310; Magee, "City Growth and Population Patterns," 6, 22.
67.
FHA, Underwriting Manual, (1935 edition), sec. 509. Following Hoyt’s discussions on specific population groups, here native white refers to Olcott’s Americans; colored refers to Olcott’s Chinese, Mexican, and Negro; and foreign refers to all other groups Olcott tracked.
68.
"Major Reasons for FHA Rejections," Insured Mortgage Portfolio (February 1939): 15. The other half of applicants rejected on account of location came in more detailed review of areas initially rated A, B, or C. As agency officials explained, "While a given location may appear suitable to the preliminary examiners, actual inspection during the complete processing may reveal disqualifying factors." Insured Mortgage Portfolio (February 1939): 15.
69.
As the data presented here attest, not all areas where Olcott identified Negro residents were classified as D areas on the map. It is possible that these neighborhoods were subsequently rejected for mortgage insurance on secondary consideration by FHA appraisers in the field. But the maps alone do not provide evidence of the total redlining of this population group.
70.
"17 Big Cities’ Growth Shown in FHA Maps: Series Will be Valuable to Real Estate and Mortgage Firms," Washington Post, March 28, 1937, R9. "Of interest to real estate dealers and mortgage lending institutions will be a series of city growth maps dealing with 17 major metropolitan areas, preparation of which has been completed by the Federal Housing Administration." These maps had been "originally undertaken merely to facilitate locational risk rating analyses, which must be made for each property insured."
71.
This was linked to the complete processing of many cities’ Real Property Inventories. On the list of cities with block level data available, see Taylor, "New Map Aid."
72.
As the manual explains, data on race and nationality were kept in multiple research files at local insuring offices. See "Compilation and Recordation of Data," which is sec. 18 in FHA, Underwriting Manual (1938 edition).
73.
No similar finding aid identifies whether or not local insuring office files may be available at the National Archives’ regional research centers. As James Taylor noted in Insured Mortgage Portfolio, although FHA field offices sent on to agency headquarters much of their cartographic data, they also retained copies of these materials.
74.
One remaining example is the map of Jackson, Mississippi. See Jackson, Mississippi, "City of Jackson, Mississippi, Prepared by the Research Division of the Federal Housing Administration, Data from the Real Property Inventory," from folder City Maps: Mississippi, record group 31, FHA, U.S. National Archives and Records Administration Cartographic Collections.
75.
The dual uses that Hoyt planned for these types of maps-supplying data relevant not only to risk rating but also to FHA Economic Background Ratings and Housing Market Analyses-may explain why so many of these records were retained. On the work of FHA staff to publicize the value of the Real Property Inventories as data sets for risk analysis, see Taylor, "Factual Aids," 15; Fisher, "Can We Immunize Mortgages?" 17; and "Real Property Inventory Data Available at Division of Economics and Statistics, Federal Housing Administration, August 1, 1938," cited in Munchmayer, Preliminary Inventory. Some cities published their inventories in book form to facilitate wide distribution. Of those that did so, Los Angeles notably adopted Hoyt’s color coding scheme. See Los Angeles County Housing Authority, Real Property Inventory and Low Income Housing Area Survey of a Portion of Los Angeles County California, 1939-1940 (Los Angeles: Los Angeles County Housing Authority, 1939-1940).
76.
The Structure and Growth of Residential Neighborhoods in American Cities, published by the agency in 1939, is best known for popularizing "sector theory" and "filtering" as an alternative to sociologist Ernest Burgess’s "concentric zones" and "succession." Looking at the dynamic factor maps for various cities reprinted in these publications as well as those available in the National Archives finds participants did not rigidly follow Hoyt’s instructions about the specific time intervals to depict.
77.
The Clarksburg and Fairmont dynamic factor maps are among the few in the FHA cartographic collections that still are tagged with mapmakers’ names and occupations as Hoyt requested in his instructions to participating members of the real estate community. See the entire folder contents in folder City Maps West Virginia Clarksburg PI 45, entry 354 (330/10/31/4-3) as well as PI 45, entry 356, folder Fairmont, WV. (330/10/31/4-4), National Archives and Records Administration Cartographic Collections.
78.
Hoyt, "Springfield, Massachusetts," 2, 3. "Map of Buffalo Showing Areas Densely Populated by the Above People [African, Italian, Polish, German, Hungarian], 1937." This map of Buffalo was included in the folder that the local insuring office sent on to Washington, D.C., suggesting its demographic information was integrated into the risk assessment of that city. From folder City Maps: New York Albany Binghamton Buffalo, record group 31, FHA, U.S. National Archives and Records Administration Cartographic Collections.
79.
Hoyt, One Hundred Years, 5-6. While best known for making this argument in developing his sector theory (see figure 28, p. 77, Structure and Growth of Residential Neighborhoods), Hoyt also makes this point in his 1933 book with respect to land values, in his articles in Insured Mortgage Portfolio with respect to city growth, and in the 1939 FHA report with respect to blight determination.
80.
Hoyt, "Exhibit J," 2.
81.
However, in a later study of the comparative desirability of neighborhoods in Cleveland, Real Property Inventory director Howard Whipple Green followed Hoyt’s instructions nearly verbatim, identifying the presence of native whites and immigrants from the British Empire as indicators of areas of high desirability, while Negroes, along with immigrants from Italy and Russia (Jews), were indicators of low desirability neighborhoods. See Howard Whipple Green, Relative Desirability of Census Tracts in Greater Cleveland. Special 1939 Report (Cleveland, OH: Real Property Inventory of Metropolitan Cleveland, 1939).
82.
As with studies of the FHA, the recognition in histories of the HOLC that for agency staff mapping city neighborhoods the presence of Jews, Italians, and "foreigners" in Philadelphia, St. Louis, and other urban areas was a point of comment and concern has not prompted any systematic study of discrimination by nationality group. However, a few researchers have at least sought to assess the status of the "foreign born" alongside native white and nonwhite occupants, discovering through GIS analyses these populations’ neighborhoods were graded more favorably than neighborhoods with nonwhite occupants but less favorably than communities whose residents were native whites. Jackson, "Race, Ethnicity"; Hillier, "Redlining"; Crossney and Bartelt, "Residential Security." The current dominance of the HOLC maps in the literature on neighborhood risk rating is such that when this scholar went to the National Archives to request these FHA maps, archivists brought out the HOLC security maps instead.
83.
Joint Committee on Appraisal and Mortgage Analysis, Proceedings of the National Appraisal Forum, 80; letter from Hill Ferguson, to Mr. Arthur May, November 23, 1937, in box 8, folder Exhibits- National Appraisal Forum, record group 207, Records of the Committee on Appraisal and Mortgage Analysis 1935-1942, National Archives and Records Administration. The YMCA had a close and ongoing relationship with the real estate community. Beginning in the 1920s with the introduction of a national real estate curriculum, local chapters served as sites for professional education, and the national YMCA was a cosponsor with the National Association of Real Estate Boards of the textbook series used in these courses. Charles Abrams’s Forbidden Neighbors had few kind words for the discriminatory implications of May’s later work.
84.
Their italics. Metropolitan Chicago: Summary of an Economic, Real Estate and Mortgage Finance Survey. Confidential Not for Publication (Washington, DC: FHLBB Division of Research and Statistics, 1940), 3, 12, in box 119, folder Metropolitan Chicago, Illinois. Security Map and Area Descriptions Master. Sec I, record group 195, Records of the Federal Home Loan Bank Board; HOLC, Records Relating to the City Survey File, 1935-1940; Metropolitan Los Angeles: Summary of an Economic, Real Estate and Mortgage Finance Survey. Confidential Not for Publication (Washington, DC: FHLBB Division of Research and Statistics, 1939), 3, in box 101, Missouri, California folder, Metropolitan Los Angeles, California Confidential Reports, record group 195, Records of the Federal Home Loan Bank Board; HOLC, Records Relating to the City Survey File, 1935-1940, National Archives and Records Administration.
85.
Ample evidence indicates the two agencies exchanged ideas. Coming together at one-time meetings such as the National Appraisal Forum, they maintained more regular communications thanks to the U.S. Central Housing Committee, which included representatives from both agencies on its Joint Committee on Appraisal and Mortgage Analysis. FHA records indicate the agency kept the HOLC security maps on file in connection with the construction of its Economic Data System (one of Hoyt’s other projects) and comments from Federal Home Loan Bank Board general counsel Horace Russell on how the FHA "was fortunate in being able to avail itself of much of the raining and experience in appraisal and the development of appraisal data by Home Owners Loan Corporation" underscores the two agencies’ close ties. While the orientation toward long-term future risk in addition to present and short-term future risk exemplified in the use of dynamic factors maps was a key distinction between the location rating work of the FHA versus the HOLC, there were many similarities in the mapmaking strategies they pursued. These included their reliance on an ecological understanding of urban processes as the basis for grading neighborhoods on an A-D scale and their reliance on local appraisers for data collection. For these records, see entry 32 Stack info 130/2/1/5 vols. (not boxes) vol. 27 PI-164 Cross Reference to Security Area Maps; Maps Prepared 1936-1940. From "‘Economic Data System’ Records Containing Data on Cities, 1937-1945," record group 207, Records of the Housing and Home Finance Agency. Records of the Division of Research and Statistics of the FHA; Address by Horace Russell General Counsel FHLBB "Appraisal Data and the Control of Mortgage Lending Policy," 7. In binder Confidential Reports on Economic Conditions in Various Cities; Source: Home Loan Bank Board, vol. 19, PI-164 entry 32, record group 207, Records of the Housing and Home Finance Agency, Division of Research and Statistics of the FHA. "Economic Data System" Records Containing Data on Cities, 1937-1945, National Archives and Records Administration.
86.
Description of Area 42 in box 119 (entry 39, stack area 450/68/2/4/7) folder Metropolitan Chicago, Illinois. Security Map and Area Descriptions Master. Sec I, record group 195, Records of the Federal Home Loan Bank Board, HOLC, Records Relating to the City Survey File, 1935-1940, National Archives and Records Administration.
87.
Hoyt, One Hundred Years, 314; Joint Committee on Appraisal and Mortgage Analysis, Proceedings of the National Appraisal Forum, 55.
88.
Chicago Plan Commission, Chicago Land Use Survey (Chicago: Plan Commission, 1942-1943); includes a reproduction of the survey cards.
89.
See, for example, Olcott’s 1942 edition. For city neighborhoods, these updates were halted in the mid-1940s, but profiles of suburban areas continued to add new information; even in the 1960s some of this information was still being compiled.
90.
Especially suggestive of this line of argument are the many urban histories focused on the experience of individual immigrant or ethnic populations; many of these sources provide occasional comparisons between their subjects and the experiences of other prominent local racial and nationality groups.
