City Observatory believes in using data to understand problems and fashion solutions. But sometimes the quantitative data that’s available is too limited to enable us to see what’s really going on. And incomplete data can lead us to the wrong conclusions.
Our use of data is subject to what we call the “drunk under the streetlamp” problem: An obviously intoxicated man is on his hands and knees on the sidewalk, under a streetlamp. A passing cop asks him what he’s doing. “Looking for my keys,” the man replies. “Well, where did you drop them?” the cop inquires. “About a block away, but the light’s better here.”
When it comes to transportation, we have copious data about some things, and almost nothing about others. Plus, there’s an evident systematic bias in favor of current modes of transportation and travel patterns. The car-centric data we have about transportation fundamentally warps the field’s decision-making. Unless we’re careful, big data will only perpetuate that problem—if not make it worse.
Sometimes Qualitative Data is More Informative
To understand why qualitative data can sometimes tell us more, let’s look at some documentation about the way one American transportation system performs.
Three recent essays from people walking in Houston make it clear that, there, the infrastructure and land use patterns that facilitate safe walking simply don’t exist. The following excerpts are snapshots from a large body of qualitative evidence showing that, in many U.S. cities, walking is a hellish experience.
Writing in Texas Monthly, in an essay entitled “Where the Sidewalks End,” Sukhada Tektel describes her experiences adapting to Houston after living in Mumbai and Toulouse:
Nothing could have prepared me for the disconnectedness of this oil-and-gas mecca: no clear city center, pitiable public transportation, and, most strikingly, no place to walk…For as far as the eyes can see, there are only cars and not a single person on foot.
David Yearsely wrote a different essay, albeit with a similar title (“Where the Sidewalk Ends”), describing wandering about Houston’s downtown and Third Ward while visiting for an organ music gathering. Even traversing the city’s upscale River Oaks district, he describes long, sidewalk-less stretches outside the walled enclaves of the busy four-lane San Felipe Avenue. In ten miles of walking, he encountered only two other pedestrians, both walking their dogs.
At the Houston Chronicle, David Dorantes wrote, “I want to walk, but Houston won’t let me.” Like many migrants to the Bayou City, he has lived in other places where walking is a normal part of everyday life. But not in Houston:
Nowhere else have I ever experienced such fear when walking in the street. I don’t mean that I’m afraid of the people who I meet on the sidewalk. I mean that walking in Houston is a horrific adventure, a pleasure endangered.
It’s unfair to pick on Houston. Large parts of most American cities, and especially their suburbs, constitute vast swaths of hostile territory to people traveling on foot. Either destinations are too spread out, or there just aren’t sidewalks or crosswalks to support safely walking from point to point. Moreover, walking is so uncommon that drivers have become conditioned to behave as if pedestrians don’t exist, making streets even more foreboding.
From the standpoint of the data-reliant transportation engineer, the problems encountered by Dorantes, Yearsley, and Tektel are invisible—and therefore “nonexistent.” Because we lack the conventional metrics to define and measure, for example, the hardships of walking, we don’t design and enforce solutions or adopt targeted public policies.
But when it comes to car traffic, we have parking standards, traffic counts, speed studies, and “level of service standards.” There is simply no comparable vocabulary or statistics for walking or cycling. Traffic engineers will immediately tell us when a road is substandard, or its pavement has deteriorated, or its level of service has become (or might someday become) degraded. We have not collected a parallel array of metrics to tell us that it isn’t similarly as safe, convenient, or desirable to walk or bicycle to common destinations. The American Society of Civil Engineers’ Infrastructure Report Card grades roads chiefly on vehicle congestion and delay (using dubious data, in our estimation). And, as we’ve pointed out, the U.S. DOT’s proposed performance measures for urban transportation further codify this bias by making vehicle delay the chief indicator of how well roads work. The logical result, as Smart Growth America has argued, is that we will end up with a system that optimizes every street for fast-moving cars, with—predictably—negative effects on walking.
The personal stories of pedestrians in Houston are rich and compelling in their detail, but lack the technocratic throw-weight of quantifiable statistics or industry standards to drive different policies and investments in our current planning system.
Will the move to “smart” cities make this worse?
Last month, the U.S. Department of Transportation announced that Columbus, OH, was the winner of its Smart Cities Challenge, beating out six other cities around the country. Google’s city planning subsidiary, Sidewalk Labs, promised to work with the winning city to deploy a wide array of big data and communication tools in order to better plan and operate transit systems. While the Guardian speculated that Google is securing a central position for its technologies in urban transportation markets, we have a different concern.
Sidewalk Labs has sketched out Flow, a flashy new data system for transportation. According to its own descriptions and press reports, it will help cities optimize traffic and parking. Clearly, Flow is primarily concerned with vehicles (cars and transit vehicles alike). But there’s no indication how it will address the movement of people on foot and on bicycles. It’s ironic that an entity called Sidewalk Labs appears more concerned with cars than with pedestrians.
As the old adage goes: If you don’t count it, it doesn’t count. That premise becomes vastly more important the more we define problems in big-data terms. New technology promises to provide a firehose of data about cars, car travel, car delay, and roadways—but not nearly as much about people. This is a serious omission, and should give us pause about the application of “smart” principles to cities and transportation planning.
It will likely amplify the bias that already favors counting cars, but not people. Consider New York City, perhaps the most pedestrian-oriented place in the nation. New York gathers data on pedestrian activity in a twice-annual survey (which counts pedestrian traffic on two different days in May and September at 100 locations). Contrast that with its system that reports vehicle traffic speeds in real time at more than 300 locations.
This isn’t simply a matter of somehow instrumenting bike riders and pedestrians with GPS and communication devices so they are as tech-enabled as vehicles. An exacting count of existing patterns of activity will only further enshrine a status quo where cars are dominant. For example, perfectly instrumented count of pedestrians, bicycles, cars in Houston would show—correctly—little to no bike or pedestrian activity. And no amount of calculation of vehicle flows will reveal whether a city is providing a high quality of life for its residents, much less meeting their desires for the kinds of places they really want to live in.
The fundamental problem is that we’ve designed our cities for the people moving through them, rather than for the people living, working, and being in them. We’re obsessed with getting there rather than being there.
If we want cities that are truly walkable and bikeable–that can become great places to be rather than easy corridors to travel through–we have to listen to more than big data. We need a framework that considers a wide array of evidence of what we’ve done and what we’ve left undone; of what we are, and what we aspire to be. Simply grafting more technology on to today’s imbalanced system will not accomplish this.
The limits of data-driven approaches to planning
City Observatory believes in using data to understand problems and fashion solutions. But sometimes the quantitative data that’s available is too limited to enable us to see what’s really going on. And incomplete data can lead us to the wrong conclusions.
Our use of data is subject to what we call the “drunk under the streetlamp” problem: An obviously intoxicated man is on his hands and knees on the sidewalk, under a streetlamp. A passing cop asks him what he’s doing. “Looking for my keys,” the man replies. “Well, where did you drop them?” the cop inquires. “About a block away, but the light’s better here.”
When it comes to transportation, we have copious data about some things, and almost nothing about others. Plus, there’s an evident systematic bias in favor of current modes of transportation and travel patterns. The car-centric data we have about transportation fundamentally warps the field’s decision-making. Unless we’re careful, big data will only perpetuate that problem—if not make it worse.
Sometimes Qualitative Data is More Informative
To understand why qualitative data can sometimes tell us more, let’s look at some documentation about the way one American transportation system performs.
Three recent essays from people walking in Houston make it clear that, there, the infrastructure and land use patterns that facilitate safe walking simply don’t exist. The following excerpts are snapshots from a large body of qualitative evidence showing that, in many U.S. cities, walking is a hellish experience.
Writing in Texas Monthly, in an essay entitled “Where the Sidewalks End,” Sukhada Tektel describes her experiences adapting to Houston after living in Mumbai and Toulouse:
David Yearsely wrote a different essay, albeit with a similar title (“Where the Sidewalk Ends”), describing wandering about Houston’s downtown and Third Ward while visiting for an organ music gathering. Even traversing the city’s upscale River Oaks district, he describes long, sidewalk-less stretches outside the walled enclaves of the busy four-lane San Felipe Avenue. In ten miles of walking, he encountered only two other pedestrians, both walking their dogs.
At the Houston Chronicle, David Dorantes wrote, “I want to walk, but Houston won’t let me.” Like many migrants to the Bayou City, he has lived in other places where walking is a normal part of everyday life. But not in Houston:
It’s unfair to pick on Houston. Large parts of most American cities, and especially their suburbs, constitute vast swaths of hostile territory to people traveling on foot. Either destinations are too spread out, or there just aren’t sidewalks or crosswalks to support safely walking from point to point. Moreover, walking is so uncommon that drivers have become conditioned to behave as if pedestrians don’t exist, making streets even more foreboding.
From the standpoint of the data-reliant transportation engineer, the problems encountered by Dorantes, Yearsley, and Tektel are invisible—and therefore “nonexistent.” Because we lack the conventional metrics to define and measure, for example, the hardships of walking, we don’t design and enforce solutions or adopt targeted public policies.
But when it comes to car traffic, we have parking standards, traffic counts, speed studies, and “level of service standards.” There is simply no comparable vocabulary or statistics for walking or cycling. Traffic engineers will immediately tell us when a road is substandard, or its pavement has deteriorated, or its level of service has become (or might someday become) degraded. We have not collected a parallel array of metrics to tell us that it isn’t similarly as safe, convenient, or desirable to walk or bicycle to common destinations. The American Society of Civil Engineers’ Infrastructure Report Card grades roads chiefly on vehicle congestion and delay (using dubious data, in our estimation). And, as we’ve pointed out, the U.S. DOT’s proposed performance measures for urban transportation further codify this bias by making vehicle delay the chief indicator of how well roads work. The logical result, as Smart Growth America has argued, is that we will end up with a system that optimizes every street for fast-moving cars, with—predictably—negative effects on walking.
The personal stories of pedestrians in Houston are rich and compelling in their detail, but lack the technocratic throw-weight of quantifiable statistics or industry standards to drive different policies and investments in our current planning system.
Will the move to “smart” cities make this worse?
Last month, the U.S. Department of Transportation announced that Columbus, OH, was the winner of its Smart Cities Challenge, beating out six other cities around the country. Google’s city planning subsidiary, Sidewalk Labs, promised to work with the winning city to deploy a wide array of big data and communication tools in order to better plan and operate transit systems. While the Guardian speculated that Google is securing a central position for its technologies in urban transportation markets, we have a different concern.
Sidewalk Labs has sketched out Flow, a flashy new data system for transportation. According to its own descriptions and press reports, it will help cities optimize traffic and parking. Clearly, Flow is primarily concerned with vehicles (cars and transit vehicles alike). But there’s no indication how it will address the movement of people on foot and on bicycles. It’s ironic that an entity called Sidewalk Labs appears more concerned with cars than with pedestrians.
As the old adage goes: If you don’t count it, it doesn’t count. That premise becomes vastly more important the more we define problems in big-data terms. New technology promises to provide a firehose of data about cars, car travel, car delay, and roadways—but not nearly as much about people. This is a serious omission, and should give us pause about the application of “smart” principles to cities and transportation planning.
It will likely amplify the bias that already favors counting cars, but not people. Consider New York City, perhaps the most pedestrian-oriented place in the nation. New York gathers data on pedestrian activity in a twice-annual survey (which counts pedestrian traffic on two different days in May and September at 100 locations). Contrast that with its system that reports vehicle traffic speeds in real time at more than 300 locations.
This isn’t simply a matter of somehow instrumenting bike riders and pedestrians with GPS and communication devices so they are as tech-enabled as vehicles. An exacting count of existing patterns of activity will only further enshrine a status quo where cars are dominant. For example, perfectly instrumented count of pedestrians, bicycles, cars in Houston would show—correctly—little to no bike or pedestrian activity. And no amount of calculation of vehicle flows will reveal whether a city is providing a high quality of life for its residents, much less meeting their desires for the kinds of places they really want to live in.
The fundamental problem is that we’ve designed our cities for the people moving through them, rather than for the people living, working, and being in them. We’re obsessed with getting there rather than being there.
If we want cities that are truly walkable and bikeable–that can become great places to be rather than easy corridors to travel through–we have to listen to more than big data. We need a framework that considers a wide array of evidence of what we’ve done and what we’ve left undone; of what we are, and what we aspire to be. Simply grafting more technology on to today’s imbalanced system will not accomplish this.
Related Commentary