Many people walking on a pedestrian alley in Buenos Aires, Argentina.
Technical Summary

Walkable Cities

Project Drawdown defines walkable cities as designing and retrofitting urban environments to encourage walking for commuting or transportation. This solution replaces the conventional practice of driving internal combustion engine (ICE) cars in cities.

Walking is the most neglected mode of transportation: the simplest, most sustainable, and cheapest medium of locomotion. In 2018, humanity walked 1573 billion kilometers, or an average of 200 kilometers (130 miles) per person per year—barely 7 minutes per day. We drive seven times as much as we walk. Walkability is still associated only with leisure and recreation in the majority of urban projects around the world; however, the notion of walking as a competitor in the area of sustainable urban mobility is increasing.

Walkability has commonly been measured according to the “7Ds”: 1) Density of activity, 2) Diversity of land uses, 3) Design of the street network, 4) Destination accessibility by distance or time, 5) Distance to transit, 6) Demand management, and 7) Demographics (which is not part of the built environment, but still important). These Ds influence patterns in travel choices based on the characteristics of users. Because there is insufficient measurement of all variables for large numbers of cities, this analysis focuses on population density as the key indicator for walkability.[1]


This solution measures the impact of urban dwellers increasingly switching from ICE car use to walking.

Total Addressable Market

The total addressable market for walkable cities is defined as the total urban mobility, expressed in passenger-kilometers. Several sources are used to project growth of the market from 2014 to 2050.[2] Current adoption[3] is estimated at 4.9 percent of the market, based on data published by the Institute for Transportation and Development Policy (ITDP) and the University of California–Davis (UCD).

Higher population density, if assumed correlated to density of points of interest (residences, workplaces, shopping, leisure, etc.), can lead to increased walking as destinations become easier for residents to get to on foot. It is assumed that the density threshold for walkability is 3000–4000 people per square kilometer (7.8–10.4 thousand per square mile),[5] and that high-density cities could lead to residents walking for 6.5–7 percent of urban trips. By contrast, residents in non-dense cities of less than 3,000 people per square kilometer typically walk around 2 percent of all urban passenger-kilometers. With no single definitive metric to measure a walkable city, and due to the high variability of cities globally, projected growth in urban population density is used as a proxy for the increased adoption of walkability in general.[5] It is important to note that city densities have been declining by up to 2 percent annually.

Adoption Scenarios[6]

Impacts of increased adoption of walkable cities from 2020 to 2050 were generated based on two growth scenarios. These were assessed in comparison to a  Reference Scenario, in which the total amount of walking follows a baseline projection from the ITDP/UC Davis Study.[7]

  • Scenario 1: Adoption is aligned with high growth projections from the 2015 publication of the ITDP/UCD joint report, “A Global High Shift Cycling Scenario,” with 3.5 percent of urban mobility by walking in 2050 (as opposed to 2.7 percent in the Reference Scenario).
  • Scenario 2: Using the default density threshold and walking mode-share values discussed above, the amount of walking that would happen in 1737 cities worldwide is projected using data from Demographia, representing 57 percent of the global urban population. This estimate is then scaled to 100 percent of the world’s urban population.

Emissions Model

Emission reductions are calculated by replacing fuel consumption and indirect emissions associated with the production of cars with walking, which has zero direct emissions. Indirect food emissions (e.g., from food distribution increases caused by eating more) are excluded.

Financial Model

Costs for ICE cars include fixed operating costs (e.g., insurance), fuel costs, and other variable operating costs such as maintenance and depreciation.[8] While increased walking may result in the need to purchase higher quantities of food because of burning more calories, implementation and operating costs are considered to be zero.


The Scenario 1 estimates that 2589 billion passenger-kilometers of urban travel will be done on foot by 2050, or about twice the current distance per urban resident per year. This results in 1.4 gigatons of carbon dioxide-equivalent greenhouse gas emissions avoided, saving travelers US$1.7 trillion[9] in operating costs of driving a car. The Scenario 2 shows a reduction of 5.5 gigatons of emissions by 2050.


This most basic mode of travel has been neglected for a long time; however, our study shows that it can be a part of a set of global drawdown strategies. It is acknowledged that there are challenges to increasing walking. We ignored implementation costs, as those are likely highly variable and city-dependent. Making cities denser and taking other steps to improve walking, however, takes a lot of time and requires skills that may not be available in many cities. Walking more often requires partly a cultural change in many places; as walking may be seen as the “poor man’s option,” changing attitudes must go hand in hand with changing cityscapes.

At the same time, making walking easier in cities can make other sustainable modes of transport more attractive: cycling, e-biking, and mass transit all stand to gain from improved walkability in cities. Ideally, all of these more efficient modes can be promoted together. There are likely some benefits of combined promotion that we have not included in any of our analyses due to lack of data. In any case, walking ease is a critical part of adoption of the other modes, and cannot be ignored.

[1] Population density is only one of seven variables, but it has the most available global data. These seven variables are highly interrelated and are challenging to model independently.

[2] Sources include: the International Energy Agency (IEA), the International Council on Clean Transportation (ICCT), the Institute for Transportation and Development Policy (ITDP) and the University of California –Davis (UCD).

[3] Current adoption is defined as the amount of functional demand supplied by the solution in 2018. This study uses 2014 as the base year.

[4] This simplification hides the fact that no switch happens suddenly at a density of 3000 persons per square kilometer. We use this as a guide to examine the general phenomenon noted in the literature. This model provides a useful first approximation of one of the key variables that might encourage more walking, and what might be the impact of more walking.

[5] This simplified assumption ignores many other interconnected factors that contribute to growth in this solution that are beyond the scope of this study.

[6] Angel et al. (2011) The dimensions of global urban expansion: Estimates and projections for all countries, 2000–2050

[7] This contrasts with the standard Drawdown Reference Scenario, which assumes a fixed percentage at current levels of market adoption, because the standard Reference Scenario is too optimistic given that sources indicate that walking is on the decline (Buehler and Pucher, 2012).

[8] All costs were harmonized with other transportation solutions, but here ownership is included as a depreciation cost rather than as a first cost as in some other solutions.

[9] All monetary values are presented in 2014 US$.