75 years ago: mapping Greenburgh's African American home ownership using the tools of the 1940 census

When I researched racist violence perpetrated against African Americans in the Florida Panhandle for my book on the Reconstruction Era, one snide Southerner asked why I bothered traveling so far to dig up stories of racial injustice. The questioner’s intent was diversion, but he did have a point.  While my Northern suburban home of thirteen years is not shamed by a history of Jim Crow oppression and the accompanying violence that plagued the Deep South, we Yankees remain stubbornly oblivious to the legacy of segregation and bigotry that surrounds us. 

I’ve lived in Hartsdale in the Town of Greenburgh for thirteen years. I don’t remember any public discussion about the racial patterns and conditions that exist in the town today. Greenburgh appears to have historically identified African American neighborhoods, yet it's not obvious as to why the Fairview neighborhood  - today, bifurcated by a highway - became the home of the majority of Greenburgh’s black residents well before World War II and continues to be so today. The corollary question to housing patterns is school enrollment: why do 3 of Greenburgh’s 10 school districts have a substantial black enrollment while the rest have few black students?  

Other than relying on anecdotes, how do we research the reasons for these historic and entrenched patterns?  Before starting such research, however, we should skeptically investigate the validity of the insinuated premise: that the black population in Greenburgh was historically subjected to segregated housing restrictions which, facilitated the establishment and preservation of nearly exclusively white neighborhoods and schools. And if such segregation did exist, was it imposed, the result of freely-made housing preferences, or some combination? 

To test the premise of segregation, we can map housing data that tracks race.  What is a good starting point?  African Americans have always lived in Greenburgh: census records from the 18th century record dozens of individuals, both free and enslaved.  The pre-20^th century populations, however, were quite small. The choice is arbitrary, but I find that the 1940 US Census, which recorded information about race, origin, housing, occupation, and other categories, is an excellent resource to test the segregation premise (and even better, it’s accessible for free at www.ancestry.com).   

According to the 1940 census, Greenburgh’s population total was 40,145, with the majority (27,587) living in the incorporated villages.  The unincorporated areas (mostly the eastern half of the town) had just started experiencing increased development following construction of the Bronx River and Saw Mill River parkways (the unincorporated population more than doubled in the 1950s).  The town-wide African American (the Census used the term “Negro”) population was 2,754, or 6.9% of the total. I estimate that almost three-quarters of Greenburgh’s black population lived in the unincorporated area, making up about 16% of the total.

The 1940 Census offers abundant information about this community.  For example, Greenburgh's African American population consisted mainly of migrants to the area, with only one-third born in New York State. The native New Yorkers, logically, were heavily weighted toward children.  Much of Greenburgh’s 1940 adult black population spoke in a Southern accent. Of 1,757 African American adults in Greenburgh (those over twenty years old in 1940), only 15% were born in New York State, with the majority having arrived from Southern states plagued by Jim Crow. The largest state of origin was Virginia with 519 adults, followed by North Carolina (395) and South Carolina (278).  Clearly, the early/mid twentieth century "Great Migration" of black people from the oppressive Jim Crow South had a profound impact on Greenburgh, New York.  

Now, let’s dive into the 1940 census data on housing and the question of segregation. Using the search parameters at ancestry.com, we can cull from the 1940 Census all households led by an African American. Greenburgh had 586 such households which mostly rented but also included 160 home owners.  The residential patterns of black renters and tenants in Greenburgh largely follow the map of black home ownership but not entirely.  Much of the discrepancy is accounted for by the then common phenomenon of black domestic employees (maids, chauffeurs, cooks, butlers, etc.) living in or adjacent to the homes of their white employers.  For example, it is surprising to find more than 800 African Americans living in Scarsdale in 1940 – triple the number that reside there today.  With the exception of a handful of households, mostly an extended family on Saxon Woods Road, Scarsdale’s 1940 black residents were almost entirely domestic employees affiliated with white households. Consequently, in 1940 Westchester, black tenancy outside of neighborhoods of concentrated black ownership does not necessarily indicate any particular acceptance of blacks as neighbors or deviation from segregation.  Instead, in my estimation, black home ownership is the more realistic barometer of racial residential tolerance.     

MAP: African American home ownership in Greenburgh (1940) superimposed over contemporary map. While three times more black households rented than owned homes in 1940, the renters' residential patterns were consistent with the ownership locations mapped above.  In addition to the neighborhoods plotted on this map, several African American adults lived as domestics in white households in area otherwise lacking black residents (Map compiled using www.multiplottr.com)

The map of black owned homes in 1940 Greenburgh shows three distinct clusters and lots of blank space.  In the village of Tarrytown, fifteen blacks owned houses mostly clustered in the downtown area formed by the “T” of Wildey and Mechanics avenues.  In nearby Elmsford, a larger group of thirty homeowners lived along a series of north-south avenues (Goodwin, Evarts, Winthrop, Sears, Cabot) in the village’s north east corner.  (While Fairview was scarred by the construction of Rte 287, the Elmsford black neighborhood later suffered the double indignity of bifurcation by the Cross Westchester Expressway (287) in 1960 and then the extension of the Sprain Brook Parkway along its east border in 1969).

The large majority of black homeowners in 1940 Greenburgh, however, lived in the unincorporated community of Fairview.  The limits of black ownership in Fairview is strictly delineated by Hillside Avenue in the east, Old Tarrytown Road in the North, Knollwood/W. Hartsdale Ave on the west, and Tarrytown/Dobbs Ferry roads on the south.  Fairview contained more than 100 black owned homes as well as several hundred black occupied rental properties.  As of 1940, only six blacks had purchased homes in the Parkway homes neighborhood along South Road and its presidential-named side streets to the immediate north-east of Fairview’s confines.  The semi-circular, hill-side streets of Parkway Gardens, soon to be recognized as a community for middle class black families, had not yet been constructed. Interestingly, in 1940, even Fairview was not an exclusively black neighborhood: many whites still lived there, particularly Italian immigrants and Italian-Americans as both homeowners and renters.  The racial mix of Fairview would change with the advent of the “urban renewal” program in the late 1950s.  In a future essay, we'll examine the notoriously deplorable housing conditions that plagued early/mid 20th century Fairview and the policies such as redlining that exacerbated that situation.  

The integration patterns that existed in Fairview in 1940 and, to a lesser extent, in Tarrytown and Elmsford villages, did not extend to the villages of Irvington, Dobbs Ferry, Hastings or Ardsley, which had no black homeowners.  Nor were there black homeowners in the unincorporated area’s Hartsdale neighborhoods adjacent and to the south of Fairview. Only a few blacks rented in these communities with most claiming residences as domestics employees of white families. Edgemont (or Greenville as it was known then), however, claimed one small black-owned house on Central Avenue and also Captain Joshua Cockburn’s Fort Hill Road tudor-style house, which was the most expensive and impressive of all black owned homes in Greenburgh (stately Villa Lewaro in Irvington had passed out of the hands of the Walker family in 1931).

The map establishes the premise of segregated housing patterns.  But we still have the question of whether the patterns of black home ownership – almost exclusively restricted to three neighborhoods toward the northern edge of the town – are the result of freely-made housing choices or the reflection of community enforced restrictions.  In other words, in pre-World War II Greenburgh, could blacks buy and live whereever they chose and could afford housing but preferred these three areas, or were they constrained to limited confines for home ownership by segregation?   

In Northern communities, there did not tend to be “smoking guns” of legislation or local zoning codes explicitly excluding black ownership from certain areas.  Instead, racial residential patterns arose through the interlocking operation of a series of independent trends that cumulatively created segregation over time.

One of the most powerful factors in creating segregation in the suburbs was restrictive covenants that real estate developers drafted to prevent sales of homes in housing developments to blacks, Jews and other minorities. A study of late 1930s and 1940s housing developments in the suburbs surrounding New York City found race restrictive covenants in one-half of subdivisions of 20 or more homes, and five-sixths of developments with 75 or more homes.  Conversely, few homes in developments of fewer than twenty houses had such covenants, although, the study’s author surmised, “existing social pressures” in such small developments could “be depended upon as an adequate barrier against Negroes.” In all, 56% of all homes surveyed were barred to blacks by such covenants.  This study also found race covenants correlated directly with Federal Housing Authority (FHA) financing.  In other words, federal government housing policy promoted segregation. [John P. Dean, “Only Caucasian: A Study of Race Covenants,” The Journal of Land & Public Utility Economics, vol. 12, No. 4 (Nov. 1947), pp. 428-32].

Courts could be relied on to enforce restrictive covenants until they were prohibited by the US Supreme Court in Shelley v. Kramer (1948).  In the late 1930s, Edgemont was the site of one of the nation’s most notorious proceedings when a white resident neighbor sought to enforce the racially restrictive covenant of the “Edgemont Hills” development and force the eviction of the Cockburn family from the stately tudor they had built on Fort Hill Road.  The Cockburns lost the suit in New York State court but the eviction was not enforced before they lost the house as the result of an unrelated financial matter a decade later.  The Cockburn house, distinguished by an anchor mounted on a rock, stands in Edgemont today and the Cockburn story is told in detail at www.thomas-quirk.com.    

In addition to restrictive covenants, federal loan administrators, bank officials and home insurers, as well as collusion among real estate agents and home owners, would effectively steer potential black home owners from certain communities and direct them to black neighborhoods, particularly in the era after restrictive covenants were barred.  In his autobiography, progressive home builder Alan Carnoy reveals how such steering was practiced.  In the early 1950s, Carnoy became concerned that the northern Hillside Avenue area in Greenburgh was gaining a reputation as a black neighborhood, thereby deterring potential white buyers of the homes Carnoy planned to build. Carnoy adopted a dubious tactic to keep this neighborhood from turning primarily black.  Carnoy instructed his sales staff “to gently explain [to prospective black home buyers] that the Home Owners’ Association would have to pass on every application, as we had put aside two acres for a playground in the development. At that time, the idea of segregation was so deeply implanted in people’s minds, that the Negroes turned away without taking offense, the implication of the playground being completely clear in their minds. After the white people became convinced that the development would be all white, our sales became very easy.”  [Alan Carnoy, Democracia Si! (1962), p. 76.]

 In other situations, the enforcement scheme was even less subtle. A newspaper from 1957 reported an incident in Hartsdale Lawn (now, Poet’s Corner) where a group of white homeowners collected funds to purchase a home to prevent its sale to a black New York City judge.  [Atlanta Daily World, Oct. 17, 1957] Municipal and local zoning for mandatory lot and house sizes were also understood as aimed at maintaining the “character” (i.e., whiteness) of neighborhoods. Insinuations were raised in connection with the Greenburgh Town Council’s 1950 “move directed at protecting the ‘recognized character’ of home development in the Greenville section of the town” by prohibiting “the erection of houses under 20,000 cubic feet in volume in the area and to prevent ‘cracker box’ and ‘overnight’ houses.” [Irvington Gazette, Dec. 1, 1965]. 

While these anecdotes are not dispositive proof that Greenburgh housing patterns resulted from deliberate segregation, the cumulative impact of these examples – the widespread use of restrictive covenants in Westchester, the practice of steering, zoning codes, neighborhood intolerance – highly suggest that the ownership housing patterns illustrated by the above map did not result from freely made housing choices by black home buyers during the first half of the twentieth century.

This conclusion, while patently obvious, relies on logic and not objective proof.  I do not have evidence of widespread restrictive covenants in Greenburgh and only anecdotal examples of housing prejudice on which the theory is constructed. Confirmation would require digging into deeds and real property records.   

SaveSaveSaveSave