Reason Magazine “has a long piece attacking New Urbanism”:http://www.reason.com/0502/fe.st.crime.shtml co-written by an architectural liaison officer with the West Yorkshire Police and someone from the Thoreau Institute. It would be tempting to suggest the Onionesque headline:
bq. Libertarians: “World would be better if designed by the police.”
Laurence Aurbach has “a detailed rebuttal on the City Comforts site”:http://citycomfortsblog.typepad.com/cities/2005/02/correcting_the_.html .
{ 15 comments }
mw 02.15.05 at 6:49 pm
Very interesting pair of articles–I would like to see the debate extended further. In particular, there are claims made in the Reason piece that are not addressed in the rebuttal, for example:
1. Mixed-used neighborhoods: From the Reason article:
“Newman found, mixed uses “generate high crime and vandalism rates,†and housing units next to commercial areas “suffer proportionally higher crime rates.†More recent research in Baltimore and Philadelphia by Temple University criminologist Ralph Taylor and several colleagues confirms that mixed uses increase both crime and the cost of policing.”
2. Alleys — the Reason article claims that alleys specifically are a bad idea. This is only addressed only briefly in the rebuttal and only in the context of closing ‘streets AND alleys’.
And I think the proposed Onion-style headline isn’t accurate. The libertarians at Reason, I think, would argue that the world would be better not if designed by police, but if not designed by any single higher authority–that is, it would be better if people are allowed to choose their environments rather than have designs imposed on them by (aesthetically, theoreticlly and/or politically-motivated) planners.
The city I live in (or around 115,000) has a monthly magazine that published a crime map in every issue. It is quite obvious that the older neighborhoods with new-urbanism favored features (grid-structured streets, alleys, neighborhood parks, a mixture of stores, apartments, and single-family homes, etc) have clearly higher crime rates. Month after month, that’s where the highest concentration of blue and green dots are. These are not low-income, run-down neighborhoods by any means, and overall crime rates here are quite low, but there’s no doubt that there are higher rates of crime in the older-style neighborhoods.
Where do we live? Somewhere in between–on a cul-de-sac but and older one with houses close to and facing the street and all quite visible to each other. We’re within easy walking and biking distance of schools, parks and downtown, but the nearest apartment buildings and stores are half a mile away, not next door.
Our neighborhood crime rate is very low–obviously lower than the denser, mixed-use neighborhoods, but not quite so close to zero as the brand-new ‘McMansion’ neighborhoods. I would never trade a marginal increase in safety for the isolation and purely car-based existence of the new developments, but many people clearly are willing to do so, and I don’t see any reason to enact zoning laws that will prevent them from having that choice to make.
Lastly, the apparent denial by New Urbanism partisans that there IS any tradeoff between the richness (but greater risk or crime and vandalism) and the isolation (but greater safety) of suburban subdivisions seems refuted by common sense and common experience.
Thomas 02.15.05 at 7:46 pm
The key part of the rebuttal is this:
The bottom line is how our society addresses threats and the fear of strangers. The crude and invasive gates, barricades, fences and street closures that are so much a feature of post 9-11 landscaping (especially in our nation’s capital) are expressions of fear, not strength, and may in the long run be counterproductive. Crowds of people on commercial streets may experience more crime than those in a patrolled mall. But at the same time it is that freedom, openness and interaction that many believe is the most valuable feature of great towns and cities. Those who cherish cities can appreciate Daniel Patrick Moynihan’s attitude toward threats: “Ours must be openness and fearlessness in the face of those who hide in the darkness. A precaution, yes, sequester, no.”
_____________
Which essentially gives the game away. It may be right–I think it likely is right–but it’s not an argument that the folks at Reason have it wrong, but that even if right, they’re wrong.
Jason McCullough 02.15.05 at 8:29 pm
Did everyone just completely ignore Jane Jacobs’ The Death and Life of Great American Cities when it came out in the 1960s?
mw 02.15.05 at 8:49 pm
Which essentially gives the game away. It may be right—I think it likely is right—but it’s not an argument that the folks at Reason have it wrong, but that even if right, they’re wrong.
Yes it does give it away–it’s a clear indication that this is as much a political as an empirical argument.
What the Reason article is arguing (and I think they’re right) is that the idea that New Urbanism design properties (alleys, through traffic, mixtures of retail, single-family, and multi-family dwellings) are likely to produce greater safety from crime…is politically motivated BS. And, therefore, such designs should not be imposed by zoning laws. People should be able to choose such neighborhoods if they want but should also be able to choose more isolated, homogeneous single-family subdivisions if THAT’s what they want.
And I say this as somebody who has chosen to live in a neighborhood that’s relatively ‘new urbanist’ in flavor. My wife and I like being able bike or walk to restaurants or the movies. I like very much that our kids can get together with their friends and go poke around town. But I understand that our chances of being burgled (though quite low) are somewhat higher than they would be out in sprawl land. It’s a tradeoff–life’s full of ’em.
yoyo 02.15.05 at 8:55 pm
“The city I live in (or around 115,000) has a monthly magazine that published a crime map in every issue. It is quite obvious that the older neighborhoods with new-urbanism favored features (grid-structured streets, alleys, neighborhood parks, a mixture of stores, apartments, and single-family homes, etc) have clearly higher crime rates. Month after month, that’s where the highest concentration of blue and green dots are. These are not low-income, run-down neighborhoods by any means, and overall crime rates here are quite low, but there’s no doubt that there are higher rates of crime in the older-style neighborhoods.”
are you sure you aren’t confusing “crimes per acre” with “crimes per house unit/person”?
urban areas are more dense, and so even if crime is occuring at the same rate in both places, crime will appear denser in compact areas.
catfish 02.15.05 at 9:18 pm
“People should be able to choose such neighborhoods if they want but should also be able to choose more isolated, homogeneous single-family subdivisions if THAT’s what they want”
It seems to me that most current zoning laws are much more likely to lead to single family residential areas. In fact, in many places it is difficult to build New Urbanist-like places even if a market exists. I am positive that it is advocates of single use areas rather than New Urbanists who have been much more sucessful in using zoning boards to impose their preferred urban form on everyone else.
mw 02.15.05 at 9:38 pm
are you sure you aren’t confusing “crimes per acre†with “crimes per house unit/person�
urban areas are more dense, and so even if crime is occuring at the same rate in both places, crime will appear denser in compact areas.
I expect density is a factor, but I’m pretty sure that the population density doesn’t vary nearly as much as the crime dot density–in the outer ‘sprawl’ neighborhoods, there are no crime dots at all most months. For example, in Nov 2004 (the first issue I laid my hands on), there were 36 burglaries. Two thirds occurred within a roughly two-square mile area in the center of the map and one third occurred in the the remaining 50 square miles. So that’s 10 burlaries per month per square mile in the center and 1 burglary for every five square miles in the neighborhoods outside the center, which is a 50:1 difference. There’s no way the population densities vary by anything close to that ratio.
Now I should also point out that the central areas are characterized by a high percentage of unmarried (or at least childless) renters, whereas the areas outside are much more heavily single-family. But that’s not necessarily a problem with the Reason analysis–it’s part of the point. If a you buy a home in a mixed-use area where you may be next door to an apartment house, your chances of being victimized go up.
ogmb 02.15.05 at 11:06 pm
It is quite obvious that the older neighborhoods with new-urbanism favored features (grid-structured streets, alleys, neighborhood parks, a mixture of stores, apartments, and single-family homes, etc) have clearly higher crime rates.
This doesn’t establish causation.
Alex 02.16.05 at 12:58 am
I think there should be a pause for noncrapness before whipping the West Yorkshire Police. After all, one’s views are normally formed by experience. WYP have to police some really horrible estates that are near -perfectly designed for easy crime.
Holme Wood comes to mind: bizarre street plan with back lanes and (shudder!) alleys, absolutely no mixed use and very hard to navigate, hence easy to get away from a break-in and hard to police. It was bad enough trying to collect the rubbish (ad hominem: which I did, /hominem) and return the bins.
Add to that its isolation on top of a hill, separated from everywhere else by a deathcrash ringroad, and you can see why it was known for violence and crime.
Not that this is an argument against the NU, far from it, but only that the WYP have painful experience in the policing problems created by stupid architecture (idiotecture? can we get a meme started?) and perhaps deserve a hearing.
Randolph Fritz 02.16.05 at 7:46 am
What garbage. Low-density suburbs are well known to have more crime than than all but the worst high-density neighborhoods; it’s property crime, largely carried out by professionals who break into houses without any neighbors being the wiser.
In the broader scheme of things, it is worth remembering that low-density suburbs require extensive law and regulation to exist; they are as unlibertarian an a spatial order as can be imagined.
This is entirely without any discussion of New Urbanism, about which I have very mixed feelings.
Murph 02.16.05 at 8:08 pm
From what MW is saying, I suspect he’s from Ann Arbor, MI, where I’m a grad student in urban planning. If I’m right, I think his presentation of the local situation is disingenuous.
1. Those “unmarried renters” are University of Michigan students. The burglaries in the central area are predominantly students leaving their doors unlocked and having their laptops swiped. There’s almost never any “breaking” in the local police blotter–just “entering”.
2. The student population makes this a bad example of what New Urbanism advocates. Students tend to be short-term residents of any given housing unit, meaning that they often don’t know their neighbors and often lack a sense of ownership over their neighborhood (frequently, homeowners will actively attempt to deny students any sense of ownership–students have been called “transients” on the local paper’s opinion page). Without this sense of ownership (“eyes on the street”) that *both* Reason and CNU want to see, identifying and preventing criminals is difficult.
3. The density around here does vary quite a bit. Comparing the dorm areas to MW’s neighborhood probably would get a 50:1 ration pretty easily (I’m guessing MW has about a 1/3 acre lot–pretty typical outside of the central area), but even some of the central neighborhoods characterized by standalone houses have densities of 12-14 units/acre, which is still several times the density outside. Additionally, if you look to the student ghetto areas, housing units have 4-6 people per unit, whereas the student-free neighborhoods average closer to 2-3. Still not 50:1, sure, but I’d guess the ratio of central area to outer area densities tops a factor of 10 easily.
Not to say that MW is *wrong* — if I’m guessing correctly that he’s talking about my town, then it is definitely the case that the traditional neighborhood physical features are definitely correlated with higher crime, even on a per person basis. What I’m saying is that, in Ann Arbor’s case, which fits everything in the description MW gives, it’s not causal. What’s causal of the increased crime is the demographics, not the built form.
Murph 02.16.05 at 8:10 pm
From what MW is saying, I suspect he’s from Ann Arbor, MI, where I’m a grad student in urban planning. If I’m right, I think his presentation of the local situation is disingenuous.
1. Those “unmarried renters” are University of Michigan students. The burglaries in the central area are predominantly students leaving their doors unlocked and having their laptops swiped. There’s almost never any “breaking” in the local police blotter–just “entering”.
2. The student population makes this a bad example of what New Urbanism advocates. Students tend to be short-term residents of any given housing unit, meaning that they often don’t know their neighbors and often lack a sense of ownership over their neighborhood (frequently, homeowners will actively attempt to deny students any sense of ownership–students have been called “transients” on the local paper’s opinion page). Without this sense of ownership (“eyes on the street”) that *both* Reason and CNU want to see, identifying and preventing criminals is difficult.
3. The density around here does vary quite a bit. Comparing the dorm areas to MW’s neighborhood probably would get a 50:1 ration pretty easily (I’m guessing MW has about a 1/3 acre lot–pretty typical outside of the central area), but even some of the central neighborhoods characterized by standalone houses have densities of 12-14 units/acre, which is still several times the density outside. Additionally, if you look to the student ghetto areas, housing units have 4-6 people per unit, whereas the student-free neighborhoods average closer to 2-3. Still not 50:1, sure, but I’d guess the ratio of central area to outer area densities tops a factor of 10 easily.
Not to say that MW is *wrong* — if I’m guessing correctly that he’s talking about my town, then it is definitely the case that the traditional neighborhood physical features are definitely correlated with higher crime, even on a per person basis. What I’m saying is that, in Ann Arbor’s case, which fits everything in the description MW gives, it’s not causal. What’s causal of the increased crime is the demographics, not the built form.
mw 02.16.05 at 8:34 pm
Not to say that MW is wrong — if I’m guessing correctly that he’s talking about my town, then it is definitely the case that the traditional neighborhood physical features are definitely correlated with higher crime, even on a per person basis. What I’m saying is that, in Ann Arbor’s case, which fits everything in the description MW gives, it’s not causal. What’s causal of the increased crime is the demographics, not the built form.
Yes we are talking about Ann Arbor. But the relationship between renters, transience, and higher crime is not purely a student phenomenon even here. There are other mixed areas in town with non-student renters and those also seem to have more transience and more incidences of crime.
There are, of course, demographic factors involved (renters tend to be younger and poorer than single-family homeowners). Hypothetically, if you had apartment buildings full of older, wealthier, long-time residents with children, things would likely be different. So in that sense we’re talking about demographics rather than structure. But demographics and structure are not independent factors. Older, wealthier families with kids just DON’T, as a rule, choose apartment living–that’s not the pattern either here (or anywhere in the U.S. save a few large cities). So by seeking out homogenous neighborhoods of single-family, owner-occupied homes, people do tend to decrease their risk of exposure to crime and social disorder. They are making a choice based on both demographics and environmental structure because the two are interdependent.
Laurence Aurbach 02.19.05 at 12:07 am
The empirical evidence on crime in mixed use areas is, appropriately enough, mixed. Some studies have found higher rates of crime, while others have found that the character of the commercial districts can have a beneficial effect. The presence of lively, bustling shoppers tends to inhibit criminal activity. See the article I referenced, “Architecture as Crime Control” for some citations on this topic. mv is exactly right that it’s about tradeoffs. The point of my rebuttal was that the tradeoffs are not nearly as draconian and frightening as Town and O’Toole would have us believe.
Alleys must be properly designed also. Techniques to ensure natural surveillance are extremely important. Carriage houses (a.k.a. “granny flats”) facing the alley can be a very helpful component, and they can bring in rental income too.
If O’Toole and Town are correct that new urbanism increases crime, where is the evidence from new urbanist communities? To quote Rob Stueteville: “Since nearly 500 sizable new urbanist communities are under construction or built in the US — far more than in Britain or any other country — why couldn’t the authors come up with a single example, let alone enough examples or studies to lend credence to their theories?” Rob’s letter to the editor answers this question in detail.
It’s interesting that studies of gated communities find no difference in crime rates compared to similar non-gated communities. And, speaking of tradeoffs, there are other threats to consider besides crime. Here are a couple: Exurban cul-de-sac developments often have below-standard fire, police and ambulance response times because the street pattern is so inconvenient and development is so dispersed. The excessively wide, curvilinear streets and arterials of suburbia encourage drivers to speed, making collisions with pedestrians more harmful or fatal when they occur.
Laurence Aurbach 02.20.05 at 1:54 am
Many studies have confirmed that crime in multifamily housing is related to demographics, not density. Regional policies to deconcentrate low-income housing are crucial to maintain flourishing neighborhoods. A limit of approximately 10% low-income housing in any given building and/or neighborhood can work well.
mw’s blanket assertion that all multifamily housing has a higher rate of crime is compelling. It fits with the generally fear-based policy discussions that are currently running nonstop on C-Span and other media outlets. However, I have found no rigorous support for that assertion, and the National Multi Housing Council actually labels it an “apartment myth.”
This report cites an Arizona study that found:
One critique of that study is that the Arizona multifamily market is composed largely of retirement communities where one would expect low crime rates. If anyone knows of other research on this topic, I would surely like to learn about it.
Some related facts about multifamily housing:
1) Forty percent of Americans living in an apartment do so by choice and not because of their financial situation. That percentage has increased steadily from 28 percent in 1999. Households earning $50,000 or more are the fastest-growing segment of the apartment market, and now total more than 3.6 million.
2) Multifamily housing — and even public multifamily housing — does not diminish nearby housing values. In some cases, it even increases housing values by creating the population base necessary to support shopping, commercial and civic activities, a civic realm with more people out walking, as well as convenient transit service.
— From NIMBY to Good Neighbors: Recent Studies Reinforce That Apartments Are Good For a Community.
Of course, good building design and competent management are always important in achieving high quality multifamily residences.
Comments on this entry are closed.