Myth #10: Example of Laissez-Faire Planning
By Gergely Baics & Leah Meisterlin
Not until 1916 did New York City acquire its comprehensive zoning resolution, the first in the nation. Through the 19th Century, the city’s land-use restrictions comprised a patchwork of fragmented and locally scaled public and private regulatory interventions, including nuisance laws, fire zones, building codes, and deed restrictions in the form of restrictive covenants. Although with this patchwork of caveats, it is fair to describe this early land-use regime as largely unregulated, at least by our contemporary standards. The Commissioners’ Plan was laid out within this historical context: even as it strictly defines the geometry of blocks and streets — and, consequently, a geometry of intersections — the plan remains silent on land use, save the location of a few public spaces. In fact, the 1811 Plan preceded the conceptualization of distinct categories of use. Such categorization arguably first appeared decades later in William Perris’s landmark 1852-54 Fire Insurance Atlas, which catalogued and systematically organized each building in Manhattan by its function, establishing our still familiar taxonomy of residential, commercial, and industrial uses.
In general, the commissioners treated buildings and the built environment as multifunctional, and therefore had nothing to say about which activities should take place where in the future city. Their silence on the subject contributed to the widely held view that the Manhattan grid was an example of laissez-faire master planning.[1] On one level, this is ahistorical: the plan’s reliance on real estate actors to populate new lots and blocks with buildings as they saw fit not only reflected the commissioners’ faith in capitalist property markets, but also their lack of a conceptual framework for distinct land uses. On another level, regardless of the commissioners’ intention, this is somewhat inaccurate: their plan, we argue in a recent article, inadvertently encoded the city’s first land-use algorithm in its street geometry — one of the Manhattan grid’s many unintended consequences.[2]
From where did this algorithm emerge, and how does it work? Following up on our previous blog post (#9), the answer lies in the grid’s peculiar geometry of elongated, rectangular blocks, corresponding to its street network’s regular and predictable spacing of intersections — more frequent along the north-south avenues than the east-west cross streets. As shown, the rhythm between intersections produced a regularized but differential topology of connectivity and accessibility, wherein avenue locations were rendered more accessible than those on the cross streets. We hypothesized: relative accessibility, in turn, operates as a key land-use incentive for real estate actors seeking profitable uses for their properties. Activities that depend on high levels of access (such as commerce and trade) would be likely to occupy avenue locations despite commanding higher rents for their accessibility. In contrast, residential and especially industrial uses, which are much less reliant upon pedestrian traffic, would likely be satisfied with less accessible cross-street lots. Taken together, the result is a set of geometric parameters creating variation in accessibility, thereby ordering land-use incentives and hence land uses.
The mid-19th century is an ideal period to explore the suggested connections between street geometry and land use. By then, Manhattan had grown substantially into its grid plan, and no comprehensive zoning had yet interfered. Testing our hypothesis, Figure 1 presents a sequence of four maps showing land-use intensities per street segment in 1852-54 for single-use commercial, residential, and industrial buildings, and mixed-use residential buildings with ground-floor retail. Single-use commercial buildings concentrated in Lower Manhattan’s emerging business district and along the port facilities of the southern shore, while single-use industrial clusters appeared along the waterfront farther north. Elsewhere ubiquitous mixing prevailed across the streetscape, but not without relevant patterns. Specifically, ground-floor retail concentrated to produce commercial streets in a bifurcated geography. On the grid, retail intensifies almost exclusively on the north-south avenues as expected, and below the grid it crisscrosses the city. Residential and industrial uses observed the inverse pattern: on the grid, they concentrate mainly on the east-west streets (again, as expected) and below the grid display a generally dispersed geography.
Focusing on the story of retail, there is a clear potential counterargument: shopkeepers were logically drawn to the more trafficked arteries, and on the grid this would mean the north-south avenues — wider than the cross streets and hence more accommodating to traffic — irrespective of their more frequent intersections. The period also allows us to test this counterargument to determine whether street width (as an alternative parameter) alone might influence land use outcomes, and the proof is in the maps. The four wide cross streets, specifically 14th, 23rd, 34th, and 42nd which are each 100-feet wide just like the avenues, display no pattern of intense retail concentration. Today, these are all commercial corridors, but not because of the grid’s embedded land-use algorithm, but because of 20th-century zoning. In other words, street width functions as an auxiliary variable augmenting the attraction of avenues for businesses. Across the street network, the truly consequential variable for land-use incentives is the differential access determined by each location’s proximity to intersections. And avenue locations offer greater proximity than their cross-street counterparts, for the most distant midblock lot on an avenue is only 100 feet away from its nearest intersection, whereas the same measure on a typical cross street is three to four times greater.
Still, questions remain. Specifically, how (and to what extent) does the proximity to intersections drive land-use incentives? Further, if the grid encoded these incentives by ordering street intersections, then it stands to reason that these incentives pre-existed the grid. By that logic, should we not find patterns below the grid similar to those we see on the grid? To demonstrate how the proximity to intersections drives land-use incentives, Figure 2 presents a series of probability curves for buildings in 1852-54, depicting the relationship between a building’s land use and its distance from the nearest intersection. For analytical clarity and to compare patterns, we separated Manhattan’s built environment into two halves — on the grid (north of Houston Street) and below it. The graphs clearly show that buildings closer to intersections were more likely to contain ground-floor retail or commercial uses, whereas those farther away from intersections were more likely to serve residential or industrial functions.
Notably, this pattern holds across both halves of Manhattan, although the distinction between uses appears more pronounced where the grid plan was implemented. This discrepancy directly follows the grid’s elongated rectangular blocks, which created cross streets that are typically much longer than streets in Lower Manhattan. In other words, stretching the blocks’ shape effectively stretched the distance between the near-intersection and midblock land uses, which were separated more markedly on the grid plan than below it. What this tells us is that the Manhattan grid did not institute a new spatial logic for land use. Rather it rendered clear an existing — and perhaps more broadly relevant — algorithm of incentives by block dimension, by limiting the variability of its parameters and exaggerating their differences. Without any consideration of land use on the commissioners’ part, their plan’s unique geometry set in motion locational incentives in particular ways, profoundly shaping Manhattan’s land-use development.
To frame our point differently, imagine if the “planners” had drawn a different street geometry for their grid — one with less elongated rectangular blocks or with longer north-south than east-west street segments or perhaps with square-shaped blocks, a common layout for grid plans. In any of these alternatives, the locational logics would have remained the same, but they would have been operationalized differently, resulting in different land-use geographies. The version we have now feels familiar. Over time, we have embraced some of its patterns like New York’s dynamic mixed-use commercial avenues, just as we have zoned out others such as its interspersed midblock industries. After a century of zoning, we forget how much of our day-by-day land-use geography is less the product of deliberate planning, and more the result of choosing whether to codify patterns embedded within the grid’s street geometry — an algorithm for land use, accidental and unintended.
Gergely Baics is Associate Professor of History and Urban Studies at Barnard College, Columbia University. Leah Meisterlin is Assistant Professor in Urban Planning at Columbia University. They are the authors of "Old Maps, New Tricks: Digital Archaeology in the 19th-Century City" and “The Grid as Algorithm for Land Use: A Reappraisal of the 1811 Manhattan Grid.”
[1] Peter Marcuse “The Grid as City Plan: New York and Laissez-Faire Planning in the Nineteenth Century,” Planning Perspectives 2 (1987): 287-310.
[2] Gergely Baics and Leah Meisterlin, “The Grid as Algorithm for Land Use: A Reappraisal of the 1811 Manhattan Grid,” Planning Perspectives (2017): 1-24. Doi: 10.1080/02665433.2017.1397537