About and methodology

About this report

This report was released in July 2023.

This project was conducted under the leadership of Calvin Gladney, president and CEO of Smart Growth America, and Beth Osborne, vice president of transportation. Abigail Grimminger, communications manager, was the primary author of this report, with additional contributions from Stephen Kenny, policy associate. Atlanta and DC data analyses performed by Megan Wright, economic development associate. Research, editing, and additional writing by Steve Davis, AVP of transportation strategy. Editing and strategic direction from Eric Cova, communications director. Graphic design by KarBel Multimedia.

This report was produced through the generous support of the Barr Foundation.

Data sources and methodology

Socioeconomic data to evaluate the impacts of each highway segment, built or unbuilt, was compiled via the U.S. Census Bureau’s 1960 Census and 2022 American Community Survey Census (ACS) Census Tract data. The land impact of each highway segment was defined as the amount of land that would be unsuitable for any use other than a highway. Each highway segment’s total impact was quantified based on the federal standard for highway lane width (12 feet), and standard shoulder width (six feet). Depending on the intended or constructed number of lanes, the direct land impact of the highway was calculated with an additional 200-foot buffer of the physical highway. Our analyses of unbuilt segments use the planned number of lanes, though those figures are extremely conservative. (E.g, numerous plans for 95/70 in DC suggest eight or and as many as 10 lanes—we used six to stay conservative.) History also suggests that highways could have become larger during planning or certainly later expanded. Our estimates overall are extremely conservative, and if anything they undercount the full likely impacts of these highways.

Census data was pulled at the Tract level given that this was the most localized level of data collected in the 1960 Census. The data analyzed was only from Tracts that were within each city’s limits and bisected by the intended highway path. For example, though I-95 and I-70 continue into Maryland; the segment in this analysis only includes the bisected Census Tracts in the District of Columbia. Atlanta analysis stops at the city limits. For each highway segment, the total feet of impact was used to measure the impact of the segment on land value, displacement, and current business conditions. It is important to note that these values are estimates, and extremely conservative ones at that. It is difficult to fully quantify the impact of highway construction on neighborhoods and surrounding areas, so we were conservative in our calculations.

The number of people estimated to have been displaced by highway construction is also different from the number of people who were or are impacted by highway construction. It is nearly impossible to measure the number of people who were negatively impacted by the construction of highways that bisected established urban areas. Similarly, the estimated number of housing units that were destroyed due to highway construction is different from the number of housing units affected by highway construction. It can be assumed that the radial impact of a highway is much larger than what’s included in the 200-foot buffer, and that homes with any proximity to a highway lost a portion of their value.

Highway routes

Existing/built segments were mapped in GIS using precise shapefiles from the FHWA.

Unbuilt highway segments were drawn in ArcGIS, derived from proposed plans from either city/ state/regional governments or various federal entities. (The USDOT was not created until 1967
and the US Department of Commerce managed the highway program until then.) For these unbuilt routes, it’s impossible to say where a road that was never built would have been with much precision, especially when those highways were canceled or defeated before precise routes were selected and full, final engineering plans completed, or construction begun. There were numerous proposed highway plans from the 1940s through the 70s in each city, but detailed engineering was not always completed nor precise routes finalized. In both cities, we have created the best representation of the routing for unbuilt highway segments, which may differ slightly from one or many versions of planning documents. In each case, they represent a reasonable synthesis of the many options proposed and/or studied. This is one reason that we also use extremely conservative estimates for our assessment of the damage, which likely underestimates both scale and scope.


The Georgia State University Library maintains a terrific resource of highway maps and planning documents from the state, many of which are also geocoded to be overlaid with an existing map. For Interstate 485 from the junction with I-75/85 in the center of the city north to I-85 and SR 400, we used these two maps. (One, two) An Environmental Impact Statement was submitted in 1973 for this segment which studied several alternative routes for that segment. The southern segment of 485 from today’s Carter Center and Presidential Library to the south was less advanced in the planning process before it was canceled. Several GSU maps and other 1969 highway maps from GDOT show an approximate route for this proposed southern segment which were used for this analysis. I-485 as proposed would have connected near where I-675 was later built at a junction with I-285, though this analysis stops at the Atlanta city limits at Moreland Avenue.

Washington, DC

There were several primary sources used to create the routes for DC’s potential highways, all of which were at various stages of planning when ultimately canceled. DC’s plans have their roots in an October 1955 study by De Leuw, Cather & Company which first proposed the “Inner Loop” around the core with roughly a figure eight looping around the core of the city. Plans from 1960-1962, including the Kennedy administration’s 1962 “Recommendations for Transportation in the National Capital Region” included the route analyzed here for the Northern Leg (I-66) along U Street and Florida Ave. Another study was released in 1963-1964 which studied numerous alignments for the North Central Freeway, including the northerly turn from the B&O railroad ROW directly through Takoma Park, Maryland. That route was refined in a 1966 Supplemental Study to more closely follow the railroad and take less land in both DC and Maryland, though that plan was never finalized or approved. We also cited maps included in correspondence between the District Department of Highways and FHWA from 1975-1983 to officially “de-map” these planned highway segments after they were canceled, and have them officially removed from federal interstate maps. Those maps were primarily to serve the purpose of broadly identifying the corridors rather than showing them with great precision, but they do show routes derived from earlier plans.

The North Central Freeway (I-95/70) was the closest to construction when it was finally shelved
in 1973, with right-of-way purchased, at least one obstacle (The Taylor Street bridge in Brookland) torn down, and many homes already seized. This route included separate paths for I-70S and I-95 (known as the Northeast Freeway) after the freeway was to split near Fort Totten in DC. The route analyzed here is a synthesis of the routes presented in the 1962 study, an alternative from the 1963- 64 study, and the 1966 supplementary study.

The Northern Leg (I-66) was less advanced when it was finally defeated. The earliest proposals show the routing we analyzed along Florida Avenue and just south of U Street. From 1963-1965, an alternate route for the Northern Leg of running a complex tunnel under K Street NW to connect with the North Central Freeway at New York Avenue was frequently proposed. But this tunnel route was never finalized, sent through the engineering process, or approved, though it enjoyed more local popularity than the more disruptive Florida Ave/U Street proposal. The idea of a Northern Leg was essentially dead once the Three Sisters Bridge (to carry I-66 over the Potomac River) was halted by the courts in 1969 (upheld by the Supreme Court in 1972.)

One last plan worth mentioning is the District of Columbia Interstate System 1971, prepared for the District Department of Highways and Traffic (and FHWA) by DeLeuw, Cather Associates and Harry Weese & Associates. We referred to this plan but largely treated this as useful historical information. This plan was one of the first to propose a less disruptive path for the North Central Freeway by suggesting a fairly unproven and complex underground interchange and tunnel running along New York Avenue and then hewing closely to the railroad right-of-way further north. This is in contrast to nearly all earlier plans which have I-95/70 continuing further north past New York Avenue to near Rhode Island Ave/Florida Avenue and then following Florida Avenue or R Street east to the northward turn along the railroad ROW.

While some who have studied DC freeway battles have pointed to this 1971 document as the “final” plan for DC highways, considering that 1) the National Capital Planning Commission had already deleted the North Central Freeway from their plans in 1968, 2) the U.S. District Court stopped all freeway construction in late 1969 (upheld by the Supreme Court in 1972), and 3) the freeway was ultimately canceled entirely just two years later in 1973, there’s little evidence that this plan from the 11th hour of DC’s freeway battles carried much weight.

Read the rest of Divided by Design

Explore the report’s full content using the sidebar menu at right (or below if on a mobile device), or jump to one of the three parts with the graphics below.

report cover graphic showing a stylized highway cutting through a city.graphic showing a stylized scene of construction of a highway through a city neighborhoodgraphic showing a stylized scene a few blocks away from a highway running through a city neighborhoodgraphic showing a stylized scene of what a neighborhood could look like after tearing a highway down

Don’t miss supplemental maps, videos, and animations in the DC and Atlanta case studies which are not in the hard copy. Download a PDF version of the report.