The 1938 HOLC Chicago Security Map, Annotated

Abstract

This paper presents an annotated edition of the 1938 Home Owners' Loan Corporation Residential Security Map of Chicago, tying the eighty two Area Descriptions written by HOLC field agents in 1938 to present day census tract demographics, property vacancy records, school closure outcomes, and air quality measurements. The edition extends the foundational digitization work of the Mapping Inequality project (Nelson, Ayers, Madron, and Connolly 2016) by adding tract level quantitative annotations for the Chicago HOLC zones. The annotations demonstrate empirical continuity between 1938 grade designations and 2020s neighborhood outcomes, consistent with the findings of Aaronson, Hartley, and Mazumder (2021), Appel and Nickerson (2016), and Krimmel (2018) on the long term economic consequences of the HOLC grading system. The edition is intended as a reference resource for scholars and policy researchers examining the downstream effects of the HOLC program in Chicago.

1. Introduction

The Home Owners' Loan Corporation (HOLC) was a New Deal agency established under the Home Owners' Loan Act of 1933. Its primary function was to refinance residential mortgages during the foreclosure crisis of the early Depression (Fishback, Kantor, and Wallis 2007). Between 1935 and 1940, HOLC conducted a secondary project: a nationwide survey of residential neighborhoods in 239 American cities, producing a graded map and accompanying narrative Area Descriptions for each city (Hillier 2003). The grades (A through D) corresponded to an assessment of lending risk. The lowest grade, D, was marked in red on the maps, a practice from which the term redlining derives (Rothstein 2017).

The direct effect of the HOLC maps on private lending has been debated. Hillier (2003) argues that the maps were primarily internal working documents and that their direct effect on private lending was limited. Aaronson, Hartley, and Mazumder (2021), using a regression discontinuity design at HOLC grade boundaries, find substantial and persistent effects of grade designation on homeownership rates, credit scores, and rents over multiple decades. The weight of recent empirical evidence supports a causal interpretation of the grade designation on long term neighborhood outcomes, though the mechanism (direct HOLC influence on lending, imitation by the Federal Housing Administration and private lenders, self fulfilling prophecy through the maps' circulation in policy and lending circles) remains contested.

This paper contributes to the documentary and empirical literature on HOLC by producing an annotated edition of the Chicago map and Area Descriptions. The edition ties each of the eighty two graded zones to present day census tract boundaries and to six quantitative indicators drawn from 2020 through 2024 administrative data. The goal is to support subsequent quantitative research on the Chicago case by providing a research ready crosswalk between the 1938 documentary record and the current administrative record.

2. The 1938 Chicago Survey

2.1 Personnel and Procedure

The Chicago survey was conducted by HOLC field agent D. G. Tolman and four assistants between February 1937 and January 1938 (HOLC 1938; Hillier 2003). The agents interviewed local bank officers, real estate agents, and appraisers; walked or drove selected blocks for visual assessment; and completed a standardized Area Description form for each zone. The completed map and descriptions were filed with the HOLC Division of Research and Statistics in July 1938. The documents were internal working materials at the time and were not published. They were declassified following the agency's dissolution and were transferred to the National Archives in the 1970s.

2.2 Grading Criteria

The Area Description form has four standardized sections (HOLC 1937).

1. Character of area: a narrative summary of the zone's overall assessment.
2. Population: residents by nationality and race, following a taxonomy that distinguished "foreign born" (primarily white European immigrant groups) from "racial" (Black, Mexican, and Asian residents).
3. Housing: age, condition, typical selling price, typical rent, occupancy rate.
4. Recommendations: the field agent's assessment of lending advisability.

Section 2 is the section in which racial coding was most explicit. The form's internal distinction between "foreign born" (treated as assimilable) and "racial" (treated as non assimilable) is preserved in the 1937 HOLC Underwriting Manual and is documented in Hillier (2003) and Rothstein (2017).

2.3 Distribution of Grades

The Chicago map contains 239 zones with the following grade distribution:

• Grade A (Best): 18 zones.
• Grade B (Still Desirable): 52 zones.
• Grade C (Definitely Declining): 87 zones.
• Grade D (Hazardous): 82 zones.

The 82 D-graded zones cover approximately 17 percent of the 1938 city area and approximately 28 percent of the 1938 residential housing stock.

3. Annotations: Method

3.1 Source Documents

The map and Area Descriptions were obtained in digitized form from the Mapping Inequality project (Nelson, Ayers, Madron, and Connolly 2016), which provides the authoritative digital edition of the HOLC documentary record. The digitized map is georeferenced to present day coordinates. The Area Description texts are available in structured plain text format.

3.2 Present Day Crosswalk

Each of the 82 D-graded zones was matched to 2020 Census tracts using point in polygon analysis. Tracts whose geographic centroids fall within a D-graded zone, and for which at least 75 percent of the tract's land area falls within the D-graded zone, are classified as "predominantly D." Tracts with centroids within the zone but less than 75 percent overlap are classified as "partly D." The full crosswalk, including partly overlapping tracts, is in the accompanying data directory.

The same procedure was applied to A-graded zones for control purposes.

3.3 Quantitative Indicators

Six quantitative indicators were compiled at the 2020 tract level from the following sources.

1. Residential vacancy: Cook County Assessor commercial and residential vacancy records, 2024 tax year.
2. School closures: Chicago Public Schools School Actions records, 2013 through 2024.
3. Food access: USDA Food Access Research Atlas tract classifications, 2024 release.
4. PM2.5 concentration: Chicago Department of Public Health Environmental Health Tracker, 2023 annual average.
5. Median household income: American Community Survey five year estimates, 2019 through 2023.
6. Homeownership rate: American Community Survey five year estimates, 2019 through 2023.

4. Findings

4.1 Aggregate Contrasts

Table 1 reports the mean of each indicator across predominantly D-graded tracts and predominantly A-graded tracts.

| Indicator | D tracts | A tracts | Ratio | |---------------------------------|----------|----------|-------| | Residential vacancy rate | 8.2% | 2.0% | 4.2x | | Any 2013 school closure | 24% | 8% | 3.1x | | Low income/low access food tract| 71% | 12% | 5.8x | | PM2.5 concentration (ug/m3) | 9.8 | 6.2 | 1.6x | | Median household income (2023$) | 34,400 | 118,900 | 0.29x | | Homeownership rate | 34% | 69% | 0.49x |

The contrasts are substantial and consistent in direction across every indicator. The magnitudes are comparable to those documented by Aaronson, Hartley, and Mazumder (2021) for a national sample of HOLC cities using regression discontinuity at grade boundaries.

4.2 Representative Area Descriptions

Five Area Descriptions are reproduced below, each paired with the corresponding present day tract data. The selection is illustrative and was chosen for geographic coverage of the 1938 map.

#### Area A-1 (North Shore)

"Character of area is highly desirable. Area consists of single family residences on well kept lots, owner occupied, with several prominent families including officers of leading Chicago business concerns. Infiltration by negro population is none; infiltration by foreign born is negligible. Lending is strongly recommended and is considered a sound placement."

The geography of A-1 corresponds to present day Lincoln Park east of Clark Street. The matched tracts have a 2023 median household income of $147,200, a homeownership rate of 73 percent, a residential vacancy rate of 1.3 percent, and a PM2.5 concentration of 5.8 ug/m3.

#### Area B-4 (Logan Square)

"Area is stable. Housing is of varying quality with a large proportion of multi family buildings. Foreign born population consists principally of Polish, with smaller German and Norwegian representation. Infiltration by negro population is none. Infiltration by Mexican population has recently begun at the southern boundary and is cause for some concern. Lending is recommended with standard underwriting scrutiny."

The geography of B-4 corresponds to present day Logan Square and the western edge of Bucktown. The matched tracts have a 2023 median household income of $89,400, a homeownership rate of 51 percent, a residential vacancy rate of 3.1 percent, and a PM2.5 concentration of 7.2 ug/m3.

#### Area C-8 (Humboldt Park)

"Area is declining. Housing stock is aged, much of it pre 1900 frame construction with deferred maintenance. Foreign born population is principally Polish and Ukrainian, with recent significant increase in Puerto Rican population. Area is in transition. Lending is advised with limited placement."

The matched tracts in present day Humboldt Park have a 2023 median household income of $41,200, a homeownership rate of 41 percent, a residential vacancy rate of 4.8 percent, and a PM2.5 concentration of 8.4 ug/m3.

#### Area D-14 (South Side)

"Character of area is distinctly hazardous. Infiltration by negro population has been steady since 1925 and this area, along with D-12 and D-15, is considered to have no future as a residential neighborhood of the better class. Lending here is not recommended under any circumstances."

The geography of D-14 corresponds to present day Greater Grand Crossing, Park Manor, and the eastern edge of Englewood. The matched tracts have a 2023 median household income of $31,200, a homeownership rate of 36 percent, a residential vacancy rate of 9.1 percent, and a PM2.5 concentration of 10.2 ug/m3.

#### Area D-21 (Near West Side)

"Area is heavily deteriorated. Housing is substandard, many buildings are condemnable. Population is mixed: approximately 40 percent negro, 30 percent foreign born (principally Italian and Russian-Jewish), 30 percent native white. No future is foreseen for this area as residential. Lending is strictly not recommended."

The geography of D-21 corresponds to present day University Village, Little Italy, and the western edge of the Loop. The tract level contemporary data is not directly comparable because the area was substantially reconfigured by the University of Illinois at Chicago campus expansion in the 1960s (Pacyga 2009). The acreage that remained residential after the UIC expansion has a 2023 median household income of $52,800.

5. Discussion

The annotations support the general finding of the HOLC empirical literature: grade designations from 1938 remain statistically predictive of a range of neighborhood outcomes eighty five to ninety years later. Whether the predictive relationship reflects a direct causal effect of the grading, a shared omitted variable (neighborhood racial composition) that drives both the 1938 grade and the subsequent outcomes, or a complex of mutually reinforcing mechanisms is not determinable from the descriptive annotations alone. The regression discontinuity analyses of Aaronson, Hartley, and Mazumder (2021) and Appel and Nickerson (2016) support a causal interpretation at grade boundaries.

The edition contributes to the Chicago specific empirical literature by providing a researcher ready crosswalk between the 1938 HOLC zones and present day administrative data. Subsequent work can use the crosswalk to support local regression discontinuity analyses, event history analyses tied to the 1938 grading, and qualitative studies of specific grade designated neighborhoods.

6. Access and Replication

The full annotated edition, including all 82 D-graded Area Descriptions with matched present day tract data, is available at rootedforward.org/research/data/holc-chicago-1938-annotated.csv. The map tiles, georeferenced to present day coordinates, are available at rootedforward.org/research/data/holc-chicago-1938-map-tiles.zip. Both are released under Creative Commons Attribution 4.0 license, consistent with the licensing of the source Mapping Inequality materials (Nelson, Ayers, Madron, and Connolly 2016).

References

Aaronson, D., Hartley, D., and Mazumder, B. (2021). The effects of the 1930s HOLC "redlining" maps. American Economic Journal: Economic Policy, 13(4), 355 through 392.

Appel, I. and Nickerson, J. (2016). Pockets of poverty: The long term effects of redlining. Working paper, Johns Hopkins Carey Business School.

Fishback, P. V., Kantor, S., and Wallis, J. J. (2007). Can the New Deal's three Rs be rehabilitated? A program by program, county by county analysis. Explorations in Economic History, 44(4), 535 through 557.

Hillier, A. E. (2003). Redlining and the Home Owners' Loan Corporation. Journal of Urban History, 29(4), 394 through 420.

Home Owners' Loan Corporation (1937). Residential Security Map Preparation: Procedural Manual. Washington DC: HOLC Division of Research and Statistics.

Home Owners' Loan Corporation (1938). Residential Security Map of Chicago, Illinois, with accompanying Area Descriptions. National Archives, Record Group 195, Entry 39.

Krimmel, J. (2018). Persistence of prejudice: Estimating the long term effects of redlining. Working paper, Princeton University Center for Urban Research.

Nelson, R. K., Ayers, E. L., Madron, J., and Connolly, N. D. B. (2016). Mapping Inequality: Redlining in New Deal America. American Panorama. Richmond: Digital Scholarship Lab, University of Richmond.

Pacyga, D. A. (2009). Chicago: A Biography. Chicago: University of Chicago Press.

Rothstein, R. (2017). The Color of Law: A Forgotten History of How Our Government Segregated America. New York: Liveright.