This is Junk Science Week at the Antiplanner. Each day, I will present an example of how planners rely on junk science to justify some of their more inane ideas. Today, I will focus on New Urbanism and the sense of community.
First, it is worthwhile asking why planners seem to believe in so much junk science. In previous posts, I’ve presented reasons why planning can’t work: the systems planners want to plan are simply too complicated for anyone to deal with. Because there is no real scientific support for planning, planners instead turn to junk science.
Junk science, or as some prefer to call it, pseuodoscience, refers to the use of apparently scientific concepts or data to give a patina of authority to various claims that are not really supported by the data. One symptom of junk science is that people who question the science are often subjected to ad hominem attacks rather than debates over the accuracy of the information.
Junk scientists often work by finding a set of data — any data. It doesn’t matter if the data were scientifically collected or even if they measure anything that is closely related to what the junk scientists are trying to prove. Once they have the data, they search it to see if they can find any correlations. Once they have found such correlations, they assume that correlation proves causation. They then declare that they have proved their point, whatever that is.
One of the nation’s preeminent junk scientists is Robert Putman, whose book, Bowling Alone purports to prove that Americans’ sense of community is collapsing. Putnam gathered together hundreds of polls, surveys, and other data sets, none of which directly measured community, but which together proved, he claimed, that America’s sense of community was declining. For example, the data he found measured such things as “dwindling trust between adults and teenagers” and “the changing observance of stop signs.”
Only two of Putnam’s data sets compared suburbs with cities. One measured the percentage of people who served as officers or committee members of a local group. The other measured the percentage of people who had attended a public meeting on town or school affairs. Both data sets showed higher participation in the suburbs than in the central cities (p. 206). If these things measure a sense of community, Putnam’s conclusion should have been that people have a greater sense of community in low-density suburbs than in high-density cities. Instead, Putnam made the amazing claim that mobility and sprawl somehow “undermines civic engagement and community-based social capital” (p. 205).
Based on this thinly supported claim, Putnam somehow calculated that “suburbanization, commuting, and sprawl . . . account for perhaps 10 percent” of the decline in community participation (p. 283). He therefore strongly endorsed New Urban planning as a way of restoring our sense of community. “It is surely plausible that design innovations like mixed-use zoning, pedestrian-friendly street grids, and more space for public use should enhance social capital,” he wrote (p. 408). In other words, Putnam proposed to apply to the suburbs the same features that are found in the cities that (according to his measures) have a lower sense of community than the suburbs.
Here is a clear example of someone reading their preconceived notions into data sets that are only remotely related to their central thesis. I am not persuaded that any of Putnam’s data actually measure our sense of community. But to the extent that they do, Putnam simply ignored his numbers and claimed exactly the opposite of what those numbers said.
Of course, the idea that suburbs cause of loss of a sense of community, and that New Urbanism can restore it, long preceded Putnam’s book. The 1991 Ahwahnee Principles, which were the original New Urban manifesto, stated that “Existing patterns of urban and suburban development” contribute to “the loss of a sense of community.”
That sense could be restored, the principles added, by following New Urban designs: compact, walkable communities built around civic and cultural centers. Just two years later, one of the first books on The New Urbanism focused on the claim that it was an “architecture of community.”
The architects who wrote these principles had no evidence for their claims, so they naturally embraced Putnam when his book came out. Yet the fact is that much social science literature in the past has blamed cities, not suburbs, for reducing people’s sense of community. (This literature is reviewed in this article, which argues that cities probably aren’t a big problem either.)
More recently, a study by researchers at the University of California and Dublin University reexamined Putnam’s claim that suburbs lead to a loss in community. The study concluded that suburban residents have more friends, more contact with neighbors, and greater involvement in community groups than residents of dense urban neighborhoods.
Of course, this is the same thing that Putnam found. But, unlike Putnam, the researchers concluded that this showed that suburbs do not contribute to a loss in a sense of community. Based on this, the researchers could find no justification for land-use regulation mandating New Urban design. That doesn’t stop planners today from endlessly promoting New Urban designs as a way of restoring people’s supposedly lost sense of community.
When it comes down to it, “sense of community” is perfect for planners. It is not quantifiable, but it taps into everyone’s loneliness and nostalgia for some mythical past when everybody loved one another. (When was that exactly? The 1930s? The 1890s? You’ve got to be kidding.) New Urban developments attract people from a narrow socioeconomic class, so of course they feel a sense of community with one another.
I, too, have a sense of community, in fact, multiple communities: rail fans; Belgian Tervuren dog lovers; and road cyclists, among others. Yet there are few, if any, people in my neighborhood who belong to any of these communities. Fortunately, I have the telephone, Internet, air travel, and yes, even the automobile to bring me together with others in these communities.
As the late Melvin Webber pointed out more than forty years ago, we now enjoy community without propinquity. People who say we need to reshape urban areas in order to improve our sense of community have simply closed their eyes to the real communities of the past fifty years.
JK: Can someone please explain what is “sense of community†and how does it relate to “sense of placeâ€Â, “livabilityâ€Â, â€Âsustainabilityâ€Â, â€Âwalkability†and gullibility.
Thanks
JK
Bowling Alone isn’t science Randal.
It’s a polemic.
It’s not in the science journals. You want to show where this occurs in the journals in order for this premise to not be fatally flawed.
Other than that huge gaping flaw in your argument, I’d hate to say I agree with you, but BA only looks at a small segment of white community and conflates that with all society. You may find this ‘conflation’ argument familiar, eh, Randal?
And, it’s not very scientific to say the researchers could find no justification for land-use regulation mandating New Urban design , Randal, when they don’t say that at all (click the link, folks. Really.) How can you recognize Junk Science when you see it with this sloppy analysis?
What this particular paper actually says is – I’m not saying I think the paper is great and I have problems with some things in there – but it is one way to look at the issue.
It is, simply, not true that everyone in cities has more interactions than everyone in suburbs. I, myself, have more interactions in rural areas than I do in the city. And I’m a friendly guy.
These authors and their peers tend to find that there is no “either-or”, and when we make plans, we cannot assume that, say, everyone is going to think TND is just wonderful or that a 20% open space set-aside is the greatest thing ever. Their results say “it depends, and folks will sort to places that they like”.
If one looks at those z- and t-scores in the appendices the paper does indeed say ‘it depends’. Not the researchers could find no justification for land-use regulation mandating New Urban design .
Their paper does not explain all the variance, so it is not gospel.
Not everyone wants to live in a dense network. And planners know this. Just because you focus on one segment of planning doesn’t mean all planning is this way (that little conflation problem).
You have to make housing for everybody. That certainly makes it problematic when we are laying out towns and cities, but that is the reality on the ground. And people who do this on the ground know this, despite your flawed efforts to say otherwise.
Anyway, you blatantly confuse advocacy with science and do a disservice to your readers.
DS
â€Âsystems planners want to plan are simply too complicated for anyone to deal withâ€Â
This is still unsupported. You’ll note that I am questioning the accuracy of the statement.
â€ÂBecause there is no real scientific support for planningâ€Â
What kind of scientific support are you looking for? Anything showing a correlation seems to be ruled out (see below).
â€ÂOne symptom of junk science is that people who question the science are often subjected to ad hominem attacksâ€Â
It’s odd that you point that out. You haven’t commited any that I know of, but others sharing your position (who also post on this site) have committed quite a few. I noted quite a few early on. I agree that the use of attacks does make me question the oppposite side more, but it doesn’t necessarily mean that the claims by that side are untrue. Perhaps it is just an angry individual and it would be wrong to judge the entire side by the comments of a few who use attacks instead of logic and reasoning.
â€ÂOnce they have found such correlations, they assume that correlation proves causation.â€Â
Is it possible for there to actually be a correlation? By your statement here, you make it sound as though any piece of data that shows correlation cannot be true. How much of a correlation is needed for it to be assumed that there is causation?
Data can be manipulated quite a bit to back up a claim. Those who rely on junk science will also use the data in any way needed to prove a conclusion. For instance, an individual could choose to show only a portion of the data to support their claim while when the data is taken as a whole actually shows the opposite. I remember seeing a density vs. transit ridership graph shown on a web site (can’t find the web address, I’ll see if I can dig it up) some time ago that claimed that density had no affect on transit ridership (or at least a negligable affect), yet when more data points were taken, especially with higher densities, a clear trend seemed to develop that contradicted the claims of the web site. For some reason the web site didn’t show the entire graph, only the portion that made it easy to back their claim.
pdxf said: â€Âsystems planners want to plan are simply too complicated for anyone to deal withâ€Â
This is still unsupported. You’ll note that I am questioning the accuracy of the statement.
JK: Have you noticed that central planning didn’t work in Russia? Neither did it work in China, E. Germany, Cuba or any other country. That is evidence.
pdxf said: â€ÂOne symptom of junk science is that people who question the science are often subjected to ad hominem attacksâ€Â
JK: Just like challenging Al Gore’s climate crap.
pdxf said: How much of a correlation is needed for it to be assumed that there is causation?
JK: 100% perfect, in every detail, correlation DOES NOT PROVE causation.
There must be a rational link and ability to make predictions that come true. Otherwise it is witchcraft, not science.
pdxf said: I remember seeing a density vs. transit ridership graph shown on a web site (can’t find the web address, I’ll see if I can dig it up) some time ago that claimed that density had no affect on transit ridership (or at least a negligable affect), yet when more data points were taken, especially with higher densities, a clear trend seemed to develop that contradicted the claims of the web site. For some reason the web site didn’t show the entire graph, only the portion that made it easy to back their claim.
JK: See http://www.DebunkingPortland.com/Smart/DensityCongestion.htm
On the left you will see the graph of density vs auto walking and transit trips from the Sierra club web site. One can clearly see the relationships: more density correlates with less driving & more walking + transit. What they left out is they the distorted the graph’s density axis to hide the fact that the density at which these effects became significant is about the density of Los Angeles. They also left out the datapoint dots on the lines to hide the fact that only 3 data points, of the many in the original, show the effect.
An incomplete chart of the data is under their graph. It has the transit numbers which are constantly low until you get above 4500 -6250 Per/Sq Mi, densities that most cities will never reach.
Also note that when you add a TOTAL TRIPS line to that graph, it is clear the density is a direct cause of congestion. Another little oversight.
The original chart is at the lower right of that page.
Thanks
JK
2 letters to the editor in the RMN in response to Jennifer Lang’s poorly-argued hit piece belie Randal’s poorly-reasoned argument:
or maybe
So much for the fear argument spreading beyond a small minority’s ideology and to the masses, I guess.
DS