By this point, researchers and practitioners from around the country (and beyond!) have laid out their problems with TTI’s congestion reports. Here’s a roundup of some of the best:
“The UMR ignores basic research principles: it includes no current literature review, fails to fully explain assumptions and document sources, does not discuss possible biases, has no sensitivity analysis, and lacks independent peer review.”
Rob Bertini, former head of the USDOT’s transportation statistics branch, criticized several fundamental components of the TTI model, including its “arbitrary” assumptions about using free flow speeds as the baseline for computing congestion costs.
Everything is bigger in Texas—which must be why, for the past 30 years, the Texas Transportation Institute (TTI) has basically cornered the market for telling whoppers about the supposed toll that traffic congestion takes on the nation’s economy. Today, they’re back with a new report, “The Urban Mobility Scorecard,” which purports to measure congestion and its costs in US cities.
The numbers (and from time to time, the methodology) changes, but the story remains the same. Traffic is bad, traffic is costing Americans lots of money, and traffic is getting worse. Here’s the press release: “Traffic Gridlock Sets New Records for Traveler Misery: Action Needed to Reduce Traffic Congestion’s Impact on Drivers, Businesses and Local Economies.”
The trouble with TTI’s work is that, to put it bluntly, it’s simply wrong. For one, their core measure of congestion costs—the “travel time index”—only looks at how fast people can travel, and completely ignores how far they have to go. As a result, it makes sprawling cities with fast roads between far-flung destinations look good, while penalizing more compact cities where people actually spend less time—and money—traveling from place to place. These and other problems, discussed below, mean that the TTI report is not a useful guide to policy.
Moreover, its authors have been consistently indifferent in responding to expert criticism, nor has the report been subjected to peer review. The authors continue to report data for 1982 through 2007, even though TTI’s model for those years doesn’t actually measure congestion: it simply assumes that increased vehicle volumes automatically produce slower speeds, which is not necessarily accurate. The report’s data from 2007 and earlier isn’t comparable the data that comes aftewards, and can’t legitimately be used to make claims about whether traffic is better or worse than in earlier periods. And for decades, TTI used a fuel consumption model to estimate gas savings that was calibrated based on 1970s-era cars, and which assumed that fuel economy improved with higher speeds—forever.
At City Observatory, we’ve spent a lot of time digging through TTI’s work and similar congestion cost reports. A summary of our work is in the City Subjects card deck “Questioning Congestion Costs.” Here’s what we’ve found:
The travel time index used to compute costs treats the inability to drive faster than the speed limit due to congestion as a “cost” to commuters.
The predicted increase in congestion between now and 2030 will likely be trivial: an increase in delay in the average daily commute of about 25 seconds.
Predictions of increases in driving and congestion have repeatedly been proven wrong.
Driving is down: the US experienced “peak car” in 2005, and the average number of miles driven per American has fallen seven percent since then from 27.6 miles to 25.6 miles per day.
The national Inrix travel data shows that time lost to traffic congestion in the United States has fallen 40 percent since 2010.
Time lost to traffic congestion is so small for most travelers that it isn’t noticed and has little economic value.
Building enough capacity to eliminate rush hour congestion would be virtually impossible and cost many times more than the supposed value of time lost to congestion.
As we pointed out in with our “Cappucino Congestion Index” published in April, the very premise of the index is silly: it creates an exaggerated perspective of costs of congestion and totally ignores the costs—and ultimately futility—of adding additional road capacity.
We’re in the process of analyzing the details of the latest TTI report. We’ll post additional information here as our analysis proceeds. This post will be updated.
What does it mean for someone to be displaced by gentrification? And in a just world, what do our cities’ neighborhoods look like?
As reported by Next City, a team of researchers at the University of California-Berkeley has put together a an analysis that probes just those questions. But the stilted answers they come up with get at the heart of one of the greatest contradictions in the debate over where America’s cities are going, and where they ought to be.
Let’s back up for just a moment. The Urban Displacement Project, led by Miriam Zuk, wants to use predictive modeling to peer into the future of neighborhood change in the Bay Area. Drawing on previous theory, Zuk and her colleagues categorized every Census tract in the region into one of eight stages of gentrification, and used previous trends to project how much further gentrification might go.
So far, so good. But digging into the Urban Displacement Project’s methodology reveals a curious choice.
In their course of their analysis, Zuk and her team try to quantify the number of people who have been or might be “displaced” as a result of gentrification. How do you know if someone has been displaced? As the authors point out, previous researchers have generally cast a wide net: displacement might occur as a result of any number of factors, from disinvestment that makes a home unpleasant or dangerous to rising prices that force tenants to look for more affordable housing elsewhere.
But rather than attempt to sort through different types of change, the Urban Displacement Project simply assumes that any reduction in the number of low-income people in a neighborhood is a result of displacement. As Zuk acknowledges, this is rather extreme: it doesn’t leave room for someone to move away simply because, as people do, they decide that they would be happier living somewhere else; and it would define a low-income person who remained exactly in the same place but got a better-paying job as a “displaced person.” It’s unlikely that, in the teeth of the recession, this last number is particularly high—but unless we believe that economic progress in poor neighborhoods is impossible, it’s hard to see how this could be a viable assumption going forward.
This notion of displacement also seems to imply that neighborhoods are, or ought to be, entirely static and that their residents never move. We know that this isn’t true for neighborhoods in general, and poor neighborhoods in particular. As the Urban Institute has shown, one of the main ways that poor families improve their economic situation, get access to better schools and reduce their risks of crime is to move to different neighborhoods. As the Berkeley team writes, it’s probably not fair to describe those sorts of moves as totally “voluntary,” since the movers are reacting to deficits beyond their control in communities they may otherwise want to remain in. But in a context that’s heavily focused on gentrification, rolling this other kind of “displacement” may obscure more than it illuminates.
Neighborhood demographic change, after all, mostly depends on two things: who’s moving in, and who’s moving out. That means a neighborhood’s poverty rate will remain the same only if the poverty rate for each of those two groups is exactly the same. It also means that this balance can be thrown off by a change in either group. If the people moving in become less poor, the neighborhood’s poverty rate will fall even if nothing has changed about the people moving out. Even if, that is, people are not being displaced at any greater speed than they were before.
Much of the research on gentrification has focused on trying to untangle these two threads. In doing so, it has often found that neighborhood change is driven as much or more by a change in the “move-in” group than the “move-out” group. One of the most famous studies, by Columbia’s Lance Freeman, found that low-income residents of gentrifying neighborhoods were only somewhat more likely to move than low-income residents of non-gentrifying neighborhoods: the real difference was who was coming in to replace them. The Urban Displacement Project counts this shift in in-migration as displacement—which doesn’t leave room to interpret, say, a decline in neighborhood stigma that causes middle-income people to stop shunning a poor neighborhood as a positive development.
But the most fundamental problem with this definition of displacement is what it implies about the kinds of neighborhoods we want in our cities. If every reduction in the number of poor people in a neighborhood is “displacement”—and we agree that displacement is something to be avoided—then the only conclusion is that every neighborhood must remain exactly as poor as it is now.
Claiming that every neighborhood ought to stay as it is with respect to poverty—or, for that matter, racial segregation, although Zuk et al are mostly focused on economics—would be one thing in a country where the status quo was relatively egalitarian, and where the neighborhoods that people lived in reflected their own agency in a just, fair, and free society. But obviously wherever that country is, we do not live in it. When the status quo is the result of generations of discrimination and cruelty, taking Zuk’s position is chaining ourselves to a continuing legacy of terrible injustice.
In fact, a full 88 of the 129 tracts that the Urban Displacement Project identify as “gentrifying” are still poorer than the Bay Area as a whole. Obviously that doesn’t mean they’ll stay that way; gentrification and displacement is a process along a spectrum. But it does demonstrate how much room there is in many American neighborhoods to reduce the poverty rate without becoming an exclusionary bastion of privilege. The question is whether, once the process has begun, it can be arrested before housing prices push out all or nearly all of the original community of residents. The answer to that is not obvious everywhere, though our report, Lost in Place, showed that nationally, instances in which this process of improvement managed to reduce the poverty rate in previously high poverty neighborhoods are extremely rare. Despite high-profile examples of neighborhood change in some coastal cities, it’s not at all clear that there’s necessarily a gentrification “tipping point” that leads inexorably to the exclusion of all the poor.
None of this is to minimize the fact that, especially in places like San Francisco, rising home prices oftenconstitute an injustice. Clearly, being forced to leave your home is wrong, even if you live in a neighborhood that has been a target of economic or racial segregation. But we need a framework for understanding housing and neighborhood change that allows us to talk about and address the full range of challenges our cities face. A definition of “displacement” that implicitly rejects any change to the status quo is not up to that task.
Instead, we need to ask what will lead to more open, diverse and integrated neighborhoods in our metropolitan areas. Just as it’s reasonable and necessary to ask how we’ll promote greater housing choices in the nation’s wealthy suburbs (a discussion provoked by the Obama administration’s Affirmatively Furthering Fair Housing rule), it’s necessary to think about what rectifying the pattern of segregation in America’s poor neighborhoods would look like as well.
At FiveThirtyEight, Ben Casselman writes: “Katrina Washed Away New Orleans’s Black Middle Class.” It’s a provocative piece showing the sharp decline in the black population of the city of New Orleans, particularly the city’s black middle class. While the city has rebounded in many ways since Katrina, the city’s black population has recovered more slowly, and middle-income blacks especially so. While the white, non-Hispanic population of the city is still below pre-Katrina levels, it has rebounded faster than the black population.
Casselman alludes to the diaspora of the city’s African American population, which is down by nearly 100,000 from pre-Katrina levels. His analysis shows that the black middle class has recovered far more slowly than other demographic groups, but doesn’t say where they might have moved to. And the analysis conspicuously omits one major factor shaping population trends in New Orleans (and for that matter, other U.S cities): the suburbanization of the black middle class. The word “suburb” doesn’t appear in Casselman’s piece.
Casselman is clear that his analysis covers just the city limits of New Orleans, or Orleans Parish. But there’s much more to metro New Orleans than Orleans Parish. Like most US metros, a majority of the region’s population—and most of its population growth—has been in its suburbs. Suburbanization has accelerated post-Katrina. The city’s population is only about 30 percent of the New Orleans metropolitan area, down from about from 36 percent of the metro total in 2000.
The suburbanization of blacks in New Orleans
Those who follow New Orleans closely know that the area’s black population has grown increasingly suburban. The Data Center, a New Orleans-based independent research organization, has tracked the region’s changes before and after Katrina on its website and in a recent report “Who Lives in New Orleans and Metro Parishes Now?” Their data shows the makeup of the metropolitan area according to its constituent parishes for the years 2000 and 2013. Their findings show a stark gap between demographic trends in the city and surrounding suburbs:
While the city has 97,395 fewer black residents, the metro area as a whole has only 66,752 fewer black residents, revealing that the suburban parishes have gained more than 30,000 blacks. Moreover, the metro area as whole has had a net loss of 75,228 white residents. In short, the metro area as a whole is increasingly diverse with gains in blacks, Hispanics, and Asians and losses of white residents in nearly every parish.
While the black population is increasing in suburban parishes, the reverse is true for the white population, according to The Data Center’s report. The white, non-Hispanic population of the suburban parishes has decreased 11 percent, slightly faster than in Orleans parish, compared to an increase the black population of the suburban parishes of 17 percent. (We’ve reproduced the data from the center’s 2014 report below in the Appendix.)
The movement of black Americans to the suburbs is a widespread trend. According to the Brookings Institution’s Bill Frey, the black population of many central cities is decreasing (including in nine of the ten largest cities), and the black population of the suburbs is increasing almost everywhere, with 96 of the 100 largest metropolitan areas recording increasing in their suburban black population. This movement is propelled by the black middle class; Frey notes that the black movement to the suburbs is led by the young, those with higher education, and married couples with children. As Pete Saunders has written, suburban living is still aspirational for many blacks.
Black middle class growing in New Orleans suburbs
But the growing black population in New Orleans’ suburbs is not a representative sample of the region’s African American residents. Rather, Census data suggest that it largely reflects an increase in middle-income and upper-income black households. Data tabulated by the Census Bureau’s American Community Survey show the relative income levels of black households living in Orleans Parish compared to suburban areas. We’ve extracted data from American Fact Finder for 2005 and 2012. (The 2013 data are available, but reflect a change in metro area boundaries, so we use the earlier data for comparability.) While the Census Bureau reports data in the same income ranges in each year, the dollar figures are not directly comparable between 2005 and 2012 due to inflation over that time period. In metropolitan New Orleans, higher income black households are more likely to live in the suburban parishes than are low income blacks. As of 2012, about 72 percent of blacks with incomes under $15,000 live in Orleans Parish, while about 53 percent of blacks with incomes over $35,000 live in one of the surrounding suburban parishes. Comparing the income distribution data for black households in Orleans Parish with those for suburban parishes in 2005 and 2012 shows that while lower income black households have become more heavily concentrated in the city, middle- and upper-income black households have become more likely to live in the suburbs. Here we’ve divided all black households in metro New Orleans into three roughly equal groups based on household income, and reported the share of the metro total in each income group that resides in Orleans Parish in 2005 and 2012. In 2006, about 63.4 percent of the region’s black middle-income households lived in Orleans Parish. This declined to 54.4 percent in 2012. The share of the region’s poorest black households living in Orleans Parish actually increased from about 68 percent to 72 percent. The location of the highest earning third (those with incomes of $40,000 or more) shifted from a majority in Orleans Parish (54 percent in 2005) to a majority living in the suburban parishes in 2012.
A more integrated New Orleans
Overall, the metropolitan New Orleans region has become more integrated. During the decade of the 1990s, black/white segregation in metropolitan New Orleans was actually increasing—a pattern that ran contrary to the national trend. But between 2000 and 2010, the New Orleans metropolitan area recorded a sharp decrease in segregation as measured by the black/white dissimilarity index. According to William H. Frey at the University of Michigan Population Studies Center, the black-white dissimilarity index for metropolitan New Orleans has fallen from 68.3 in 1990 and 69.2 in 2000 to 63.9 in 2010.
A racial dot map of metropolitan New Orleans, showing the state of segregation in 2010. Credit: Cooper Center Dot Map
One of the keys to addressing the black-white earnings disparity is reducing segregation. As we wrote earlier this year, metropolitan areas with higher levels of segregation have, on average, much higher black-white earnings gaps. Similarly, as the work of Raj Chetty and his colleagues has shown, income and racial segregation is a powerful correlate of impaired economic mobility. The problem is exceptionally acute in New Orleans, which ranks 99th of the 100 largest metropolitan areas on Chetty’s index of intergenerational economic mobility.
As New Orleans rebuilds, it has an opportunity to address the historic patterns of segregation that have aggravated the economic plight of the area’s African-American population. It appears that it is making some progress on this front.
It’s fair, as FiveThirtyEight has done, to acknowledge the significant demographic changes that have taken place in New Orleans. Unquestionably, Katrina has had an enormous impact. But the decline of the black middle class in New Orleans also reflects two well-established trends: the national decline of segregation in housing and the movement of higher income blacks to the nation’s suburbs. Fewer blacks live in New Orleans, and more live in its suburbs. While the white non-Hispanic share of the city’s population has increased—at xx percent it’s still a minority—the white, non-Hispanic population of the area’s suburbs has decreased even faster. In the wake of Katrina, metro New Orleans is gradually becoming a more integrated city.
1. Another tall tale from the Texas Transportation Institute. This week, TTI released another episode of its “Urban Mobility Report,” claiming to measure the cost of congestion and track the continued worsening of traffic in American cities. The problem, as Joe Cortright explains, is that these reports are riddled with problems that TTI has declined to address. From failing to distinguish between mobility (the ability to travel great distances) and access (the ability to actually get to the places you want to be); to comparing data from years that used wildly different and incompatible methodologies; to ignoring evidence that adding roads has failed to reduce congestion even by TTI’s own metrics, it’s either time for the “Urban Mobility Report” to wise up, or for reporters to stop giving it such credulous coverage.
2. Are racial tipping points overblown? Daniel Kay Hertz looks at the popular idea of the “tipping point”: the theory that very mild racial housing preferences can lead to total segregation if just a few people of color move into a previously all-white neighborhood. While the theory is elegant, it turns out that the actual evidence is much more mixed, and there is good reason to believe that tipping points are no longer a major factor in most American neighborhoods. That’s important both for our understanding of why segregation happens, and for giving us reason to believe that more integrated urban communities aren’t an impossible dream.
3. Growing e-commerce means less urban traffic. Joe Cortright examines the potential effects on vehicle travel of increasing reliance on online delivery retailers like Amazon. Because these purchases reduce individual trips to brick-and-mortar stores while increasing the delivery density (and therefore efficiency) of delivery trucks, more e-commerce is probably creating a net negative impact on urban vehicle miles traveled.
4. New Orleans’ missing black middle class. Joe Cortright digs a little deeper into an article by FiveThirtyEight‘s Ben Casselman, who reports that the ranks of New Orleans’ black middle class has shrunk substantially since Hurricane Katrina ten years ago. Cortright finds that this narrative only applies to the city proper, however: New Orleans’ suburbs have seen an increase in middle class black residents, strengthening a trend that existed even prior to the hurricane.
The week’s must reads
1. Transportation for America reports that Phoenix’s voters have approved a sales tax increase that will generate more than $17 billion over 35 years to expand the famously sprawling city’s public transit network. Importantly, much of that money will be used not just to build new infrastructure, but to improve service—including adding frequency on bus lines. About 28% of the funds will go to tripling light rail capacity, and 7% will improve city streets, sidewalks, and bike lanes.
2. Robert Puentes at the Brookings Institution is also skeptical of TTI’s “Urban Mobility Report,” and asks what it would take to move towards a people-centered understanding of transportation priorities. In particular, understanding the distinction between simply covering great distances and actually reaching the jobs, stores, schools, and homes that people need access to is crucial.
3. At CityLab, Richard Florida writes about the dubious efficacy of “broken windows” policing. Rather than “public disorder” leading to more serious crime, a study by Robert Sampson and Daniel O’Brien finds that escalating private conflict is a much stronger producer of violence. Florida concludes that rather than focusing on physical signs of disorder, cities would be better off with a people-focused policing strategy that seeks to defuse tensions between neighborhood residents.
New knowledge
1. David Levinson, Brendan Murphy, and Andrew Owen of the University of Minnesota investigate the “safety in numbers” phenomenon: the idea that the presence of more pedestrians and cyclists reduces the crash rate for those non-vehicular street users. The authors find empirical support for this idea, showing that intersections in Minneapolis with more pedestrian traffic have much lower pedestrian crash rates. Exactly how the effect works is still unclear.
2. More evidence that segregation is one of the chief barriers to racial equality: Jeremy Fiel of the University of Arizona examines school segregation and resource allocation in American schools between 1993 and 2010. Fiel finds that “school segregation…was highest when and where resources such as teachers and expenditures were more unequally distributed across schools and school districts.” Mandatory desegregation programs, on the other hand, were successful in reducing levels of desegregation.
3. Patrick Button of the University of California, Irvine has bad news for economic incentives. Looking at motion picture production incentives, Button finds that these measures can have effects in areas where different cities are highly substitutable—like filming location—but not in other areas, like employment or the establishment of film production firms. Button concludes that the importance of agglomeration effects overrides whatever benefits the incentives provide.
The Week Observed is City Observatory’s weekly newsletter. Every Friday, we give you a quick review of the most important articles, blog posts, and scholarly research on American cities. You can sign up to get it in your inbox by clicking “Subscribe” at the top of the page!
Our goal is to help you keep up with—and participate in—the ongoing debate about how to create prosperous, equitable, and livable cities, without having to wade through the hundreds of thousands of words produced on the subject every week by yourself.
If you have ideas for making The Week Observed better, we’d love to hear them! Let us know at jcortright@cityobservatory.org, dkhertz@cityobservatory.org, or on Twitter at @cityobs.
Urban truck traffic is flat to declining, even as Internet commerce has exploded.
More e-commerce will result in greater efficiency and less urban traffic as delivery density increases
We likely are overbuilt for freight infrastructure in an e-commerce era
Time-series data on urban freight movements suffer from series breaks that make long term trend comparisons unreliable.
Delivering packages and reducing urban traffic congestion! Credit: Jason Lawrence, Flickr
Over at the Brookings Institution, Adie Tomer has performed a significant public service by assembling several decades of US DOT data on vehicle miles traveled. A significant weakness of US DOT’s website is that it mostly presents data a single year at a time, which makes it really difficult to observe and analyze trends in the data.
Tomer’s post plots the US DOT data on urban travel by passenger cars, unit-trucks, and combination trucks. He points to the growth of e-commerce, and the recent entry of Jet.com–which aims to be a challenger to Amazon’s dominance of web-based retailing. Tomer speculates that growing e-commerce will lead to more and more delivery trucks crowding urban streets.
He marshals several decades of data on urban truck VMT to claim that urban truck traffic is up an eye-popping 800 percent since 1966.
Another way to see trucking’s urban trajectory is to view aggregate growth since the 1960s. While urban vehicle miles traveled for both passenger cars and trucks grew steadily between 1966 and 1990—in both cases, far surpassing urban population growth—urban trucking absolutely exploded thereafter, reaching almost 800 percent growth until the Great Recession led to reduced demand. That pattern coincided almost perfectly with therise of e-commerce and the use of digital communications to manage shipping for logistics firms like UPS and FedEx and major private shippers like Walmart.
The post concludes by warning us that we need to provide for additional infrastructure for urban freight movement.
With new companies like Jet and continued growth in stalwarts like Amazon, we should expect e-commerce and urban trucking to keep growing. Those patterns bring some significant implications at all levels of government.
On the transportation side, freight investment will need to betargeted at pinch points and bottlenecks. Those specific sites of congestion deliver disproportionate costs to shippers, which get passed along to consumers, and create supply chain uncertainty
But the problems of doing time series analysis with DOT’s VM-1 (vehicle miles traveled) data is not limited to the largely cosmetic problem of web-site layout. The more serious problem is the significant series breaks that underlie the published data. Over time, US DOT has had to make important changes to the way it defines urban and rural areas (as urban development has occurred) and has had to cope with changing data sources. And, to be sure, DOT has tried to improve the accuracy of its estimates over time. The cumulative result of these changes is that it is very difficult to make statistically valid statements about the change in truck traffic in cities. (We’ve spelled out our concerns about the series break in the freight data in a technical appendix, below).
Urban truck travel actually peaked in 2008, and has mostly been declining, except for the past year.
In our view, we ought to heavily discount the published data, and not make comparisons that assume that the pre-2006 data are comparable to the post 2006 data. If we look only at the post-2006 data, a very different picture emerges. For the past six years–a period for which we have apparently comparable estimates, which appear to be not significantly affected by re-definitions of urban and rural areas–there is little evidence that urban truck traffic is increasing. If anything, the data suggest that it is flat to decreasing.
The alarmist implication of the “800% growth” statistic is that urban traffic will be significantly worsened by growing e-commerce sales. For example, the Brookings data prompted bloggers at SSTI to write “Urban truck traffic has boomed alongside the rise in e-commerce. ” and to fret that “If the rapid growth in urban truck VMT is a result of increasing e-commerce deliveries, we are a long way from peak urban truck traffic.”
In our view, such fears are wildly overblown. If anything they have the relationship between urban traffic patterns and e-commerce exactly backwards. The evidence to date suggests that not only has the growth of e-commerce done nothing to fuel more urban truck trips, but on net, e-commerce coupled with package delivery is actually reducing total urban VMT, as it cuts into the number and length of shopping trips that people take in urban areas.
E-Commerce is increasing rapidly; Urban truck travel is flat
The period since 2007 coincides with the big increase in e-commerce in the U.S. From 2007 to 2013, Amazon‘s North American sales increased by a factor of 5, from $8 billion to$44 billion. Between 20007 and 2013, the total e-commerce revenues United States has doubled, from about $137 billion to about $261 billion according to the U.S. Department of Commerce. But over this same time period, according to the US DOT data as tabulated by Brookings, truck traffic in urban areas actually declined. All the increase in e-commerce appears to have no net effect on urban truck traffic.
Does an increase in package deliveries mean increased urban traffic?
It actually seems like that increased deliveries will reduce urban traffic congestion, for two reasons. First, in many cases, ordering on line substitutes for shopping trips. Customers who get goods delivered at home forego personal car shopping trips. And because the typical UPS delivery truck makes 120 or so deliveries a day, each delivery truck may be responsible for dozens of fewer car-based shopping trips. At least one study suggests that the shift to e-commerce may reduce total VMT and carbon emissions. And transportation scholars have noted a significant decrease in shopping trips and time spent shopping.
But there’s a second reason to welcome–and not fear–an expansion of e-commerce from a transportation perspective. The efficiency of urban trucks is driven by “delivery density”–basically how closely spaced are each of a truck’s stops. One of the industry’s key efficiency metrics is “stops per mile.” The more stops per mile, according to the Institute for Supply Management, the greater the efficiency and the lower the cost of delivery. As delivery volumes increase, delivery becomes progressively more efficient. In the last several years, thanks to increased volumes — coupled with computerized routing algorithms — UPS has increased its number of stops per mile–stops increased by 3.6 percent but miles traveled increased by only about half as much, 1.9 percent. UPS estimates that higher stops per mile saved an estimated 20 million vehicle miles of travel. Or consider the experience of the U.S. Postal Service: since 2008, its increased the number of packages it delivers by 700 million per year (up 21 percent) while its delivery fleet has decreased by 10,000 vehicles (about 5 percent).
As e-commerce and delivery volumes grow, stop density will increase and freight transport will become more efficient. Because Jet.com is a rival internet shopping site to Amazon.com, and not a trucking company, its growth means more packages and greater delivery density for UPS and Fedex, not another rival delivery service putting trucks on the street.
So, far from putative cause of worry about transportation system capacity–and inevitably, a stalking horse for highway expansion projects in urban areas–the growth of e-commerce should be seen as another force that is likely to reduce total vehicle miles of travel, both by households (as they substitute on line shopping for car travel) and as greater delivery density improves the efficiency of urban freight delivery.
As David Levinson reports, data from detailed metropolitan level travel surveys and the national American Time Use Study show that time spent shopping has declined by about a third in the past decade. As Levinson concludes “. . . our 20th century retail infrastructure and supporting transportation system of roads and parking is overbuilt for the 21st century last-mile delivery problems in an era with growing internet shopping.”
So the next time you see one of those white or brown package delivery trucks think about how many car based shopping trips its taking off the road.
Technical Appendix: Urban Truck Data
We’re strongly of the opinion that its not appropriate to treat the pre-2006 and post-2007 truck freight data as a single series that represents the actual year to year growth in urban freight mileage. There are good reasons to treat this as a “series break” look separately at the two series. The technical reasons behind this judgment are two-fold.
Series Break 1: Urbanized area boundaries
Tony Dutzik explored this issue last year in a post for the Frontier Group. Briefly, a number of rural roads were re-classified as urban roads (reflecting changes in development patterns over time). This has the effect of biasing upwards later year estimates of urban VMT when compared to previous years. Some part of the apparent increase in “urban” VMT over the past decade has been a result of reclassifying formerly “rural” traffic as “urban”–not more urban traffic.
Series Break 2: Vehicle classifications
US DOT has used different data and different definitions to classify vehicles pre- and post-2007. Methodologically, what USDOT has done is migrated their vehicle classification system from that used in the now discontinued Vehicle Inventory and Use Survey and instead substituted RL Polk data. As a result of this shift in methodology the number of truck miles on urban roads jumped almost 50 percent in one year, from about 102 billion miles in 2006 to about 150 billion miles in 2007. In 2009, USDOT explained how they’d changed their estimating procedures.
The data now on the website for 2000-2006 were estimated using a methodology developed in the late 1990s. FHWA recently developed a new methodology and used it for this year’s Highway Statistics. This methodology takes advantage of additional and improved information available beginning in 2007 when states were first required to report motorcycle data – before that time, the reporting was not mandatory and the data were missing for a few states. Also, the new methodology does not rely on data from the national vehicle inventory and use survey which provided critical data for the original methodology but was not collected in 2007 as planned.
In April 2011, FHWA recalculated the 2000-2008 data along with the 2009 data to estimate trends. However, after further review and consideration, the agency determined that it is more reliable to retain the original 2000-2006 estimates because the information available for those years does not fully meet the requirements of the new methodology. Thus, the original 2000-2006 estimates are now used, whereas the 2007-2009 data are still based on the new methodology.
The author gratefully acknowledges Adie Tomer’s willingness to share the Excel spreadsheets upon which his analysis was based.
OK, we admit it. We’re data geeks. To us, sometimes — well, often — a single number or data set is compelling proof of an important proposition: bare-naked, and with no verbal embellishment or deeply personal anecdote or cutesy infographic.
Here’s the simple number: since 2000, home prices in city centers have outperformed those in suburbs by 50 percent. In graphic terms, it looks like this:
The data were complied by Fitch–the investment rating agency, in a report released with the announcement that “U.S. Demand Pendulum Swinging Back to City Centers.”
In our view, its the most under-reported story of last week, and if you are an urbanist, maybe the most under-reported story of the year. We blogged about it when the report came out, but it bears repeating: If you care about cities, and you’re looking for definitive evidence of the verdict of the market on urbanism—this is it. But we are also resigned to the fact that we are geeks, and stuff that gets our blood-racing leaves most people cold. So I’m groping for an analogy: the most convenient one is to the stock market.
Image a CNN business reporter saying:
“In the market today, city centers were up strongly to a new high”
Or a Wall Street Journal headline
“A bull market for city centers”
That’s the news here. Just as with private companies this price index is a great indicator of market performance. Imagine for a moment if you were CEO of Widgets, Inc, a publicly traded company. Every day, you’d be getting feedback from the market on how well you were doing, and on investor’s expectations for your company’s future. If your stock price went up, it would be a good indication that you were doing better, and that expectations were rising for future performance. Especially if you had a sustained rise in your stock price, and if your company were regularly outperforming both other companies in the widget industry, and the overall stock market. The reason the investment world is gaga over Warren Buffet is pretty much because he’s been able to do just that with the portfolio of companies he’s assembled under the Berkshire-Hathaway banner.
Wouldn’t it be great if we had the same kind of clear cut financial market style indicator on the health and prospects of our nation’s center cities? Wouldn’t it be useful if we could show in a stark and quantitative way how city centers are performing relative to suburbs? That, in essence, is what the Fitch data shows. Fitch’s analysts looked at 25 years worth of zip code level home price data in 50 of the nation’s largest metropolitan areas to track how well city centers performed compared to surrounding neighborhoods and suburbs. They divided zip codes within metropolitan areas into four groups based on their proximity to the city center.
You can’t literally buy stock in a city, but buying a house is the closest thing imaginable. The price a buyer is willing to pay for a home in a particular city or neighborhood is a reflection both of the current value of that location, and the buyer’s expectations of the future character and performance of the neighborhood and city. Add up all the home values in the city, and you’ve got an indicator of the market for the city as a whole.
This Fitch chart is, in effect, a kind of Dow Jones Index for the performance of the nation’s center cities. It clearly shows that over the course of the last housing cycle—beginning before the big run up in housing prices, but then continuing through the housing bubble, and growing during the bust and recovery, is an ever wider edge of city center housing prices compared to more suburban, out-lying locations. And this isn’t a short term aberration or a recession artifact. The Fitch data show the trend emerging in the late 1990s, and growing steadily over time.
While there’s a growing recognition that cities are back, in some quarters there’s denial. The truly great thing about this measure its it definitively puts the lie to the claims by perennial city nay-sayers like Joel Kotkin that the overall growth or size of suburbs is somehow a manifestation of their revealed economic superiority. In economic terms, bigger doesn’t necessarily mean better. In the economic world, market prices, and particularly changes in relative market prices are the best indicator of what’s hot and what’s not. The new Fitch analysis make it abundantly clear that cities are hot, and suburbs are not.
The reason of course is that housing demand can (and is) changing much faster than supply—which is why prices are rising so much. Rising prices are both a positive indicator of the value consumers place on city center living, and a reminder that as we’ve said many times at City Observatory, we’re experiencing a shortage of cities. And the rising relative prices for city locations are the market’s way of saying “we want more housing in cities” and “we want more cities.” While the strength in the housing sector has been an urban focused boom in new rental apartments, the fact is that supply isn’t growing rapidly enough. We aren’t creating new San Franciscos and new dense, walkable, transit-served neighborhoods in other cities as fast as the demand for urban living is increasing—and that means that prices are continuing to rise.
Over at Belt Magazine, editor Anne Trubek is fed up with the overuse of planning cliches in writing about cities. She’s asking, nay demanding, that everyone stop using ten words:
She’s put her finger on something. These words are used, and over-used and sometimes abused. We share Anne’s pain–especially when it comes to the vague and elastic way in which the term “gentrification” is invoked in all kinds of different contexts. Its natural to be frustrated that a complex, multi-faceted concept can’t be boiled down to a single word that can be universally used with precision to mean the same thing to everyone who encounters it. And it’s inevitable that some words will become, in sequence, fashionable and popular and then trite and shop-worn. To be sure, when we turn dynamic verbs into prosaic nouns (walk to walkability), we almost automatically make ourselves more pedantic and boring.
Even so, we must politely, but firmly disagree that we ought to stop using these words.
Why? In a world where 140 character-at-a-time communication is increasingly the norm, having a simple, evocative way of stating your case is imperative–even if it isn’t as precise and nuanced as it might be, and even if some people will occasionally or often misuse the word.
Words have weight and meaning: We need to use them well and wisely, and back them up with powerful illustrations and compelling stories, and where possible, data.
Let’s take walkability. Its a passive, noun-ified, mouthful. But it captures in 11 precious characters a key notion. Its intuitive and resonates with most people. We can illustrate it with images and stories, and thanks to our friends at Walk Score we can measure it (roughly and imperfectly). By giving at name, illustrating with pictures, and measuring it with data, we can raise its profile the the debates and discussions about cities.
We’re fully aware of the consequences of the sloppy use of terms. Take gentrification: Critics of the process rightly assail instances in which the entire population of a neighborhood is dislocated by the in-migration of the wealthy (like say SoHo, in Manhattan). But then people apply that same term to any time any new development occurs in a previously high poverty neighborhood. As we pointed out, Governing employed a definition of gentrification so sweeping that it classified neighborhoods with increasing poverty rates as “gentrifying” because of small increases in property values and educational attainment rates.
But the problem is not the words: its how they’re defined and used. And banning the words serves no useful purpose, and perhaps sets back the conversation, by eliminating some sign posts (and frankly, bumper stickers) that can succinctly raise attention and start conversations. In many urban policy debates banning these words wouldn’t so much enrich the conversation as cede the rhetorical advantage to those who use other words that are not benign and helpful.
The trouble is urbanists (or as Anne might have us say “advocates for building great cities and interesting diverse neighborhoods where people can easily bike and walk to schools, parks, and shops”) are engaged in extended conversations and frequently adversarial debates with others: road builders, neighborhood groups, developers, mayors, city councils and legislators, and the general public, about hard policy decisions. In many cases, its a war of words. And banning just these ten words would be the equivalent of unilateral rhetorical disarmament.
As we pointed out in our essay on “Rules of Thumb” in transportation planning, there is a conventional wisdom about how we design roads that has subtle and profound biases. Words like “level of service” and “functionally obsolete” have embedded, but largely hidden value judgments that greatly influence policy decisions.
Consider how the media has come to use the term “accidents” rather than “crashes” to describe the death and carnage that cars inflict.
We call subsidized, socialized car storage in the public right of way “free parking,” ignoring or minimizing the substantial costs it imposes on everyone.
Highway planners talk about enhancing “mobility”–by which they mean making cars go faster, which paradoxically, thanks to induced demand and sprawl, often causes a decline in accessibility–the ability to reach destinations–especially by means other that a private automobile.
And occasionally, we cripple our own advocacy by using jargony, non-euphonious terms that obscure our message. For example, we use the ugly, unwieldy term “transit-oriented development” to describe neighborhoods and housing where people can easily walk and take transit to many common destinations. (Please, somebody, come up with a better moniker for this!) MobilityLab tells us of efforts to re-brand the mouthful “transportation demand management” as “transportation options” in hopes of getting greater acceptance and more funding.
So we should continue to use all ten words on Anne’s proposed banned word list. But we should take pains to explain them, illustrate them and where possible measure them with data so that they have a depth of meaning that will add to the discussion. And in addition to quantifying, we can use pictures people what some of these seemingly lifeless, technical terms mean in the real world. Take GranolaShotgun’s compelling photo essay on infill development, showing how denser, but still small scale new residential development in cities can enliven streetscapes pockmarked with vacant lots, parking, and auto oriented commercial uses.
But imperfect as they may be words have value and convey meaning. So rather than ban them, we should nurture and polish them, using them carefully and with due care; there are important conversations ahead.
Why are America’s neighborhoods so segregated? For a lot of people, the answer requires reaching deep into history: explaining the rise of the subsidized mortgage market and redlining; racial violence in towns from Cicero, Illinois to Charleston, South Carolina; restrictive racial covenants; blockbusting; and on and on.
But back in 1971, a professor named Thomas Schelling proposed a much simpler answer. It was called the “tipping point” model, and it suggested that near-total segregation might be the inevitable outcome of even very modest preferences for neighbors who look like you.
The intuition is very simple: imagine an all-white neighborhood where white people have a range of preferences about the racial makeup of their block. One day, a black family moves in. All the white families are okay with having a black family on their block—except for one of them, which moves away and is replaced by another black family. Now all the white families are okay with having two black families on their block, except for one of them, which moves away and is replaced with a black family. And so on, until there are no white families left.
But while tipping makes for elegant theory, it’s less clear that it actually works to describe what happened—or is happening—in American cities. The first study to search for empirical evidence of tipping looked at how the racial composition of neighborhoods changed depending on its starting point—and found something that looked very much like Schelling’s tipping point:
This graph compares the percentage of a neighborhood’s population that was non-white in 1970 with racial change over the next decade. What you want to look at is the vertical line at about 5% minority in 1970. If there were no tipping point, you would expect the trend lines on either side of that vertical line to match up. Instead, they’re extremely far apart, suggesting that something happened when a neighborhood hit about 5% black and Hispanic that caused it to suddenly lose lots and lots of its white residents. In other words, a tipping point.
…and the evidence against tipping
But a follow-up study by William Easterly put some limitations on those findings. Instead of simply asking whether tipping points exist at all, it asked whether their results were as extreme as predicted by Schelling. Remember that in the original model, a small number of new black or Hispanic households could tip a neighborhood to being almost entirely black and Hispanic. That’s important for explaining American residential segregation, because many segregated neighborhoods do, in fact, have virtually no white residents.
But that’s not what the Easterly study found. Instead, while most neighborhoods became less white between 1970 and 2000—which makes sense, since the country as a whole did, too—there was very little evidence of a tipping point that led to all-black and Hispanic communities. While racial change did take place, with about 10% of all neighborhoods switching from majority-white to majority-black or Hispanic over that time period, they didn’t follow the pattern predicted by Schelling.
For one, even nearly all-white neighborhoods—the ones on the far side of the “tipping point” that should be racially stable—became much less white over that time period. Even more surprising, neighborhoods where more than 75% of residents were black or Latino actually became whiter from 1970 to 2000. In a way, that makes sense, since we know that racial segregation is slowly declining; but it’s very confusing from a “tipping point” perspective, since even much smaller proportions of people of color are supposed to send white residents running.
So neighborhoods at the two extremes of segregation didn’t behave as predicted by the Schelling model. What about relatively integrated neighborhoods? According to tipping theory, these should be the least racially stable, quickly becoming dominated by one group or another. Instead, they were among the most stable: the typical neighborhood that was 50% white in 1970 changed less than neighborhoods that were 100% black or 100% white.
Why it matters
Now, there are important caveats to all this. This doesn’t at all contradict the large amounts of evidence that Americans, and especially white Americans, have pronounced racial preferences with regard to their neighbors, or that those preferences have major consequences. (Recall the Ed Glaeser and Jacob Vigdor study that explains lower property values in black neighborhoods, everything else being equal, by the artificial drop in demand that results from whites’ avoidance.) Also, because the available data only begins in 1970, we can’t totally rule out that something like the more extreme version of Schelling’s tipping point model actually did lead to segregation before 1970.
Still, in putting limits on Schelling’s model, Easterly’s study helps shine light on some of the dangers of relying too heavily on “tipping” to explain segregation. First, the tipping point model can be seductive because it tells a story about segregation in which there are no real villains: in a world with very mild racial preferences, it’s amazing “how easily segregation arises”—it’s “math”!
But we know that’s a very selective reading of the evidence. The historical record of racial segregation in American cities is clear, and there are lots of villains. They include the federal, state, and local governments, and the white voters that elected them; the realtors and bankers who enforced discriminatory real estate practices; and the many regular, everyday people who reacted to new black neighbors by throwing bricks or burning crosses, and the even greater number who failed to stand up to them.
But the fact that tipping points aren’t nearly as extreme as they’re often made out is also cause for optimism. Taken literally, the tipping point theory would lead to a pretty fatalistic view about the persistence of segregation: if only very small differences in preferences necessarily lead to widespread segregation, then it may be nearly impossible to make progress. In the wake of the Obama Administration’s announcement that it would be more aggressively pursuing the mandate to “affirmatively further fair housing” in the 1968 Fair Housing Act, many skeptical commenters suggested that any attempt to introduce racial or economic diversity in privileged white communities would turn those neighborhoods into “Detroit”—the country’s poster child for white flight.
That’s a theory that’s more than a little self-interested, since many of the people arguing it live in exactly the kinds of communities that have “benefited” from exclusionary practices for decades. Look, for instance, at the outcry in Westchester County, NY, when HUD suggested that some of the policies that had kept many of its towns so privileged, like a refusal to allow subsidized housing or multi-family buildings, ought to change. Or the people who stood up at a public meeting outside St. Louis to insist that, while they personally didn’t have anything against allowing lower-income black students from a nearby district coming to enjoy the high quality of their schools, other people would probably pick up and leave, and soon the whole district would be poor and black.
But its plausibility comes from the broad acceptance of something like the “tipping point” idea. Easterly’s study shows that the breakdown of one kind of segregation doesn’t automatically lead to segregation at the other extreme—and that’s a thin reed of hope that less divided, and less unequal, cities might be more realistic than we think.
As Robert Caro chronicled in his riveting biography “The Power Broker,” the great builder Robert Moses had a foolproof strategy for getting new highways approved. He’d take a little bit of money and get the project started, driving stakes in the ground and manufacturing expectations about future development opportunities. Then he’d dare the Legislature not to give him the money to finish the project. They invariably did.
The political allure of the building big projects–especially bridges and highways–continues to this day. Call it the “Edifice Complex.” With a bonanza of jobs and contract dollars, there are a wealth of constituents who support more spending and the prospect of a prominent ribbon cutting and a tangible evidence of your efforts to reduce congestion and speed traffic. But as Bengt Flyvbjerg has documented the construction of megaprojects–ranging from Boston’s Big Dig, to Seattle’s Bertha boring machine to San Francisco’s expensive new Bay Bridge–is one of repeated and substantial cost overruns. Megaprojects have their own pathology: they are the product of excessive optimism, over-predicting revenues and benefits, and consistently under-estimating costs and risks. Public officials routinely engage in “strategic misrepresentation”–Flyvbjerg’s polite academic term for “lying.” And cost overruns predictably average 30 percent above budget.
Build now, figure out how to pay later
In many respects, one current project, the replacement of New York’s Tappan Zee Bridge, symbolizes much of what’s wrong about the way we try to tackle infrastructure problems in the US.
The Tappan Zee construction site afforded a great backdrop for a photo op with the President. (WNYC).
The Tappan Zee Bridge spans the Hudson about 25 miles north of New York, connecting suburban Westchester County with Rockland County. The original bridge, built in 1955 carries about 138,000 vehicles per day. A $3.9 billion replacement project is now underway, which would build a new and larger freeway bridge, with eight lanes (and space for more). According to the New York Thruway Authority, the state agency charged with building the bridge the Tappan Zee is the largest infrastructure project underway in the country right now.
But even though construction on the replacement bridge is now 25 percent complete, the agency started construction without a complete financial plan explaining how it will be paid for. The Thruway Authority has blocked requests from local newspapers to release financial data and documents. From what is known, it’s likely that the new bridge, once completed will require a doubling of one-way tolls from the current level of $5 per vehicle to $10 or more. Like many highways around the country, traffic growth on the Tappan Zee bridge is stagnant, having peaked in 2004. The danger is that high tolls could further depress traffic across the bridge and produce a financial death spiral, as higher tolls are needed to make up for lower traffic. Since no financial plans or toll revenue forecasts have been released to the public, it’s impossible to say how likely this is. Nonetheless, the Thruway Authority has issued upwards of a billion dollars in debt, and negotiated a TIFIA loan from the US Department of Transportation. And because the Tappan Zee is the largest source of toll revenue for the Thruway Authority, shortfalls in expected revenue from this project are likely to impact the financial condition of other state toll roads.
Project development processes almost invariably focus on the most expensive build alternatives. Early on, engineering estimates suggested that the existing bridge could be renovated at a cost about 40 percent less than building a new highway bridge (Philip Mark Plotch: Politics Across the Hudson: The Tappan-Zee Megaproject, Rutgers University Press, 2015). Nonetheless the state moved ahead with a more expensive full replacement alternative (albeit after stripping out a long-sought transit component to save costs).
The way we design, select and pay for big highway projects tends automatically to generate bloated projects. The reason: the cost to project beneficiaries of pursuing big projects that benefit them is low or even zero. Big projects funded by state or federal gas taxes (or any statewide or regional funding source) generally don’t require any special contribution from the local community or properties that will benefit. With essentially no cost and big local returns, it’s no surprise that local officials clamor for big highway projects. In short, our method for selecting and funding highway projects encourages communities to pursue the biggest, most expensive project they can get because other people will be paying for it. “I’ll have the biggest solution someone else will pay for. ”
Questionable finances, but great politics?
Despite the project’s shaky finances, demonstrable risks, and questionable utility, it is widely seen as a substantial political achievement for Governor Andrew Cuomo. The New York Times described the Tappan Zee project, along with a proposed renovation of LaGuardia Airport’s terminals as the signature accomplishments of Cuomo’s administration. In the eyes of the political observers and academics the Times quoted, the Governor would get credit for building things. One lauded Cuomo for “taking a page from the Robert Moses playbook” and another said “You get credit for things you build, not things you maintain.”
But is it either good government or smart transportation policy? As WNYC’s Andrea Bernstein pointed out, Governor Cuomo unilaterally decided to exclude transit options from the project, and has been anything but transparent about how the bridge will be financed. The state Thruway Commission blocked release of a federal report on bridge financing in which the US Department of Transportation described the state’s financial plan as “hypothetical, misleading and inaccurate.”
And there are good reasons to question the utility of the proposed project. As Streetsblog’s Stephen Miller has said, Governor Cuomo is building a legacy for the 1950s. A new eight-lane freeway bridge–with no provisions for dedicated transit service, and room for widening is a project from another era.
The allure of megaprojects vs. the tedium of maintenance
The Tappan Zee Bridge is just the biggest example of the edifice complex at work. The temptation to spend money on shiny new projects and to defer or under-fund maintenance has an irresistible political logic. New projects provide showy ribbon cutting opportunities that make great campaign fodder–while the costs of debt service and deferred maintenance are largely invisible and gets passed on to your political successors.
Short-changing maintenance, in turn, becomes fodder for a kind of bait-and-switch tactic in which politicians point to rough roads, sub-standard bridges and maintenance backlogs as a justification for new funding–and then once new revenues are approved, shift the money to pay for new projects. Examples of this kind of politically expedient thinking are abundant: Louisiana’s decision to raid its maintenance funds for $21.6 million per year, for the next 28 years, to fund new highway construction is one such example.
Megaprojects suck the life out of state transportation budgets. In Connecticut, a single megaproject is consuming more than 90 percent of available state highway funds, even as the state faces a significant backlog of maintenance on overburdened highways and rail lines.
In “Overpasses: A Love Story,” Politico’s John Grunwald chronicles Wisconsin DOTs continuing plans to build new and wider highways while deferring maintenance. Despite the fact that 70 percent of the state’s highways are in fair or poor condition, Wisconsin spends more money on new projects than on maintenance. And in turn, the chronic fiscal crisis of highways becomes an excuse to underfund transit spending.
Growing reliance on debt-financing for new capital projects, coupled with stagnant growth in gas tax revenues (largely a product of much absolute declines and slower growth in driving) mean that maintenance budgets are steadily squeezed already. Prior to adopting its latest transportation spending package (funded by a 7 cent per gallon increase in the state gas tax) Washington State was on track to spend nearly 70 percent of its state gas tax revenue retiring debt for earlier projects.
The “Git ‘er done” mentality impresses people in some quarters, particularly when it’s their projects that are advanced. But in an era of scarce resources, diminished driving, and deferred maintenance, the reckless borrowing and optimistic projections that fuel big spending for these megaprojects come with huge opportunity costs. Money spent on over-built and under-used highways and bridges means other needed projects, including transit, don’t get built. And while politicians today are still executing the classic tactics from the Bob Moses playbook for pushing megaprojects, the old master-builder left no notes on winning strategies for paying ongoing maintenance costs. The money spent on a few prominent megaprojects isn’t available to pay for maintenance and the debt service burdens incurred to get new construction today will be a burden on transportation finance for decades to come. That’s frequently the real legacy of the edifice complex.
“The automobile tragedy is one of the most serious…man-made assaults on the human body,” wrote Ralph Nader in 1965. “It is a lag of almost paralytic proportions that these values of safety…have not found their way into legislative policy-making for safer automobiles.”
Those words come from the preface of Unsafe at Any Speed, an expose of just how dangerous many midcentury cars were for their occupants, and helped usher in a revolution in automobile safety. In the post-Unsafe world, car makers now advertise protection against death or injury from accidents as one of the major attractions of their products, from arsenals of airbags to cutting-edge sensors that can avoid collisions to begin with.
But fifty years after that car safety revolution, we’re in desperate need of another one. The threat now isn’t the faulty design of cars: it’s the faulty design of streets.
The scope of the problem
Over the course of the 20th and early 21st centuries, American cities and suburbs transformed most of their streets from public spaces designed for walking, biking, and other kinds of transportation (as well as socializing, putting on public spectacles or demonstrations, or just watching the world go by) into highways that are dangerous for anyone not ensconced in a motor vehicle. The cartoon below, which was passed around the urbanist web a few months ago, may at first seem alarmist to many people—until you remember that on most American streets, stepping off the curb without a signalized intersection really would be courting death or serious injury.
Credit: Claes Tingvall
Nor is this just a problem for postwar sprawlsvilles. Until recently, even New York City’s street design guidelines prioritized motor vehicle traffic over all other users of the public way, in ways that the city now acknowledges were extremely dangerous.
The result is that car crashes are one of the leading causes of unnecessary death and injury in the United States. In 2013, according to the CDC, there were 16,121 homicides nationally—to take an extremely serious problem that has received much attention in the past several years. In In the same year, 33,804 people died in car crashes.
Even more alarmingly, many of those were children. For those under 25, car crashes are the leading cause of death. Over 3,000 Americans age 19 and under died in car crashes in 2013. Many of them weren’t even in cars: almost 500 were simply walking down a sidewalk or crossing the street.
And if you add nonfatal injuries, the numbers are staggering. More than three million Americans were sent to the emergency room as a result of car crashes in 2013—more than half a million of them children and teenagers. It’s hard to imagine another part of our daily lives causing this much damage without an outcry to stop it. In fact, there isn’t much else that does: car crashes are the single most common cause of death for teenagers, and even one of the leading causes of death for children too young to drive.
The culpability of urban planning
True, some of this can be mitigated through changing behavior, in particular drunk driving and failing to wear seatbelts. But we shouldn’t be led to believe that these issues are behind all, or even most, of this public health crisis. According to the CDC, about 4 in 5 car crashes that result in the death of a child don’t involve drinking. And while wearing a seatbelt can reduce the risk of death in a car crash by about 50%, two-thirds of younger children, and nearly half of older children and teens, were wearing a seatbelt or appropriate booster seat at the time of the fatal crash.
The fact is that urban planning bears a major responsibility for the incredibly high numbers of serious injuries and deaths on our streets. We’ve known for years that relatively small increases in the travel speed have huge impacts on the severity of car crashes: A car that strikes a pedestrian at 20 mph has a 5% chance of causing a fatality; the probability of death rises to about 40% at 30 mph and more than 80% at 40 mph.
And yet the primary goal of street design, even in relatively densely-packed urban areas, has been to keep cars moving at high speeds. In cities like New York or Chicago, this has often been 30 or 35 mph—or more when drivers go faster than the speed limit. In many other cities, speed limits on arterial streets are 45 mph or more, virtually guaranteeing that full-speed crashes with pedestrians or bicyclists will be fatal.
Other measures, including narrow sidewalks or no sidewalks at all, and large distances between controlled intersections, also make roads more dangerous. These design choices result in cases like that of Raquel Nelson, who in 2011 was convicted of vehicular homicide because a driver hit one of her three children as they crossed the street. Prosecutors accused Nelson of recklessness because she wasn’t in a crosswalk—but there were no crosswalks anywhere nearby.
Nor is all of this some inevitable byproduct of modern life in a developed country. The United States has one of the highest rates of traffic deaths adjusted for population in the rich world: more than twice that of Spain, three times greater than that of the United Kingdom, and four times more than Sweden.
Even if you adjust for the fact that Americans drive more, the United States’ roads still stand out as some of the most dangerous: 20% worse than Germany, 40% worse than Denmark, and 71% worse than Norway.
As we’ve noted before, this is one of the cases where cities and urban living are the solution. Because people drive less and drive more slowly in cities, traffic death rates are lower in more urbanized places.
Nor are dangerous streets an unchangeable part of national culture. In 1990, the US and UK had almost identical road fatality rates. But since then, the US has made much slower progress—and today, we suffer 71% more deaths for the same amount of driving. The difference is worth 14,000 American lives every year.
The good news is that we have examples to follow, both abroad and in our own country. American cities have begun to adopt the Vision Zero program pioneered in Sweden with a proven track record for improving safety. The program includes shifting infrastructure to prioritize safe mobility, rather than pure speed; using new technologies; and changing transportation management and regulation, including lowering speed limits where appropriate, such as New York City’s recent move from 30 mph to 25 mph on many streets.
We know these measures are effective at saving lives and preventing serious injuries. Given the state of American road safety, the only question is what we’re waiting for.
1. The suburbs: where the rich ride transit. In many cities, transit ridership is dominated by a transit dependent population: people who can afford to own private cars don’t use the transit system. But in some places transit is a mode of choice for higher income commuters. Daniel Kay Hertz mines Census data to identify and map high income neighborhoods where transit use rates exceed the regional average.
2. The Edifice Complex & Our Infrastructure Problems. Joe Cortright looks at the political incentives that lead to the pursuit of mega-projects and their implications for transportation finance. Around the country, leaders are dusting off the Robert Moses playbook for build now, pay later highway projects that shortchange other priorities, including transit and maintenance.
3. A War of Words. Belt magazine editor Anne Trubek is tired of a number of words in the common vocabulary of urbanism, and has called for banning, among others “walkability, “livability” and “placemaking.” Joe Cortright disagrees, and points out these terms have a powerful impact and can be illustrated and measured in ways that helps change the conversation about cities.
4. The Dow of Cities. If there were a financial-market style indicator of the health of cities, it would be something very much like the ratio of city to suburban house prices that’s been constructed by Fitch, the investment rating firm. Joe Cortright examines why the big run-up in city home values compared to the suburbs since 2000 is the most powerful evidence yet that the market is turning decisively to city centers. The high and rising price of city centers clearly signals the “shortage of cities” that needs a new policy response.
The week’s must reads
1. Race Wealth Gap not solved by education–at least not when you’re late to the housing market and there’s an epic bust. Monday’s New York Times describes the results of a new St. Louis Federal Reserve Bank study of wealth disparities among racial and ethnic groups. Black and Hispanic college degree holders saw their wealth decline between 1992 and 2013, according to the study, while white and Asian college graduates saw increases. A key factor seems to be the housing market, which accounts for a bigger share of wealth for Hispanics and blacks; these two groups also saw proportionately larger declines in their housing wealth.
2. Housing Vouchers are the subject of two great posts. At the Brookings Institution, Elizabeth Kneebone and Natalie Holmes look at the neighborhood patterns of housing voucher use. They find that while voucher users tend to live in relatively poor neighborhoods, they are less likely than residents of public housing to live in neighborhoods of concentrated poverty. The Urban Institute has constructed a terrific visualization of the distribution of housing vouchers by income level, and compares it to the distribution of benefits from the home mortgage interest and property tax deductions. Voucher benefits go disproportionately to the lowest 15 percent of the income distribution, and have a measurable, if quite modest effect of ameliorating income inequality.
New knowledge
1. In a study of migration patterns in the UK published in Urban Studies, Lance Freeman and his colleagues find little evidence that gentrification is associated with increased rates of out-migration from poor neighborhoods. In our opinion, in describing this piece CityLab downplayed the import of this research: they emphasized the difficulty of measuring gentrification and said that the new Freeman study provided “mixed” evidence. What this study shows is that–as in other published research–there’s actually no data to support the commonly held belief that increased displacement is a regular occurrence in gentrifying neighborhoods. Freeman and his co-authors conclude “The results presented here are for the most part inconsistent with the notion that gentrification leads to widespread direct displacement that manifests itself in higher mobility rates among residents of gentrifying neighborhoods.”
2. A new study published by the journal Nature shows that an abundance of street trees has strong positive effects on measures of self-reported health and well-being. Survey evidence from Toronto indicates that having 10 more trees in a city block, on average, improves health perception in ways comparable to an increase in annual personal income of $10,000 and moving to a neighborhood with $10,000 higher median income or being 7 years younger.Bonus kudos to Nature for publishing the article under a Creative Commons license.
3. The Census Bureau has released detailed geographic data for 2013 on the location of jobs and workers. Its Local Employment and Housing Dynamics (LEHD) is available through its “On-the-Map” mapping application.
1. City home prices outpacing suburbs by 50 percent. Joe Cortright examines a new study prepared by investment firm Fitch looking at the growing value premium in central cities. Since 2000, home prices have grown 50 percent faster in urban centers than in their surrounding metro areas. For hard headed Wall Street types, the analysts sound surprisingly like new urbanists, citing walkability and a “paradigm shift” in attitudes about cities. This is strong evidence of the growing demand for urban living, which Fitch expects to continue. Their outlook for homeownership and suburban growth is decidedly bearish.
2. The next road safety revolution. Daniel Kay Hertz looks at the mayhem that cars cause in urban areas. While we’ve made significant advances in protecting vehicle occupants from the effects of crashes, the way we’ve designed our cities around car travel has created big risks for people who walk or bicycle. The victims are disproportionately young: car crashes are the leading cause of death for those under 25, and more than 3,000 Americans age 19 and under died in car crashes in 2013. Many of them weren’t even in cars: almost 500 were simply walking down a sidewalk or crossing the street.
3. Between high rises and single family homes. The housing affordability and availability problems that plague many of our cities may be aggravated by the missing middle of housing types–smaller-scale, lower-impact duplexes, triplexes and apartment courts, that have all but disappeared with the advent of stringent single-family zoning. Daniel Kay Hertz reviews metro level data on the dominance of single-family housing, and the general paucity of these smaller, 2-4 unit structures. As we look to accommodate the growing demand for urban living, filling the missing middle void is one way to affordably provide a range of housing options in existing neighborhoods.
4. StrongTowns published Joe Cortright’s critical review of Rosabeth Moss Kanter’s recent book on infrastrucuture “Move.” For those hoping that a Harvard Business School professor would do a rigorous, hard-headed look at the deep-seated business model flaws in our current transportation system, the reader will be greatly disappointed. Cortright argues that “Move” overlooks important trends–notably the declining demand for car travel–and fails to diagnose the chief underlying problem with transportation–we get the prices wrong.
The week’s must reads
1. In a new report for the Century Foundation —The Architecture of Segregation — Paul Jargowsky maps the continuing increase in the number of neighborhoods of concentrated poverty in the nation’s urban areas. His data show that the number of people living in high-poverty ghettos, barrios, and slums has nearly doubled since 2000, rising from 7.2 million to 13.8 million. Blacks, Latinos and children are disproportionately likely to live in these high poverty neighborhoods. These data show that the trends we outlined in our 2014 report Lost in Place have continued, and if anything appear to be accelerating.
2. Its one of the most common refrains in almost every land use controversy everywhere: there isn’t enough parking. But is it true? Writing in the Minneapolis Star Tribune Nathaniel Hood shows that its easy to put that glib assertion to a straightforward quantitative test. His article, “How to respond when someone complains there’s no parking” outlines four simple steps that just about anyone can take to measure parking availability. He shows how you can Google Maps to flag parking structures and surface parking lots, photograph under-utilized parking, and even conduct your own parking availability experiments.
New knowledge
1. New evidence for the power of industry clusters. Alexander Klein and Nicholas Crafts of the University of Kent look at the growth of cities and manufacturing productivity over the period 1880 to 1930. They explore the relative contributions of industry specialization (so-called Marshallian externalities) and industry diversity (so called Jacobs externalities) on the rate of employment growth and producitivity in US cities. They conclude that during this time period, growth was strongly influenced by industry specialization (cities getting better at doing the things they were strong at doing already), and that only larger cities exhibited significant gains from having a variety or diversity of industry sectors.
2. Uber and auto crashes. By giving people who drink a ready alternative to driving their own vehicle, ride sharing services like Uber may reduce the number of alcohol related fatalities. A study by two Temple University researchers looks at the correlation between car crash rates and the growth of Uber use in different cities. The researchers exploit a natural experiment–the gradual roll-out of UberX services in different California communities. They find that the introduction of the lower-cost ride dispatch service is associated with a 3.6 percent to 5.6 percent reduction in “motor vehicle homicides”. The study suggests that if UberX availability extended nationally, it might save as many as 500 lives annually.
The Week Observed is City Observatory’s weekly newsletter. Every Friday, we give you a quick review of the most important articles, blog posts, and scholarly research on American cities. You can sign up to get it in your inbox by clicking “Subscribe” at the top of the page!
Our goal is to help you keep up with – and participate in – the ongoing debate about how to create prosperous, equitable, and livable cities, without having to wade through the hundreds of thousands of words produced on the subject every week by yourself.
If you have ideas for making The Week Observed better, we’d love to hear them! Let us know at jcortright@cityobservatory.org, dkhertz@cityobservatory.org, or on Twitter at @cityobs.
OK, we admit we might be a bit obsessed with this story. But if you can, bear with us one more time.
Here’s the most basic fact: The number of newly-built McMansions—single family homes of 4,000 square feet or larger—is down 43 percent since 2007. By any standard that’s a stunning decline. But because the market for smaller homes has declined even more, a commonly used but in this case misleading statistic—median home size—has floated upwards. That’s lead to some fundamentally flawed claims about the U.S. housing market.
Claims that the McMansion is back
Earlier this year, and again over the last week or two, several journalistic outlets—CityLab, Wonkblog, and the Minneapolis Star Tribune, among others—have written stories with the theme “McMansions are back.” Americans want, and are buying, bigger and bigger houses.
These stories are prompted by the Census Bureau’s publication of its annual report, “Characteristics of New Single-Family Houses Completed,” which dutifully reports the number of new single family homes built, by size, and by region of the country. The report focuses heavily on the median size, in square feet of new single family homes.
These outlets focus on the median home size and the share of newly-built homes that are McMansion-sized, and finding, correctly, that this percentage is increasing. The problem is that there are two ways to look at this, and in this case the median is misleading.
The number of McMansions is down substantially
From our point of view, what’s really important to the “McMansions are back” thesis is the actual number of these extra-large houses being built. And when you look at it that way, it turns out that they’re down 43 percent from the peak in 2006.
Single-family construction is still at a half-century low
The only reason that it looks like McMansions are “growing” is that the market for smaller single family homes remains deeply depressed. It’s important to keep in mind that the US housing market is still in worse shape today than at basically any other time in the previous half-century. The seven years since 2007 represent the seven worst years for housing production in the period since 1959. Excluding recession years, for the past 50 years, the US has built a minimum of about 4,000 new single family homes per million people; for the past seven years we’ve averaged about half that rate of construction. It’s really hard to overstate just how awful the single-family housing market still is in the United States. On a population-adjusted basis, we’re still building fewer single family houses now, several years out of the recession, than at the bottom of the recessions of 1970, 1980-82, 1990, and 2001.
The middle class has been pushed out of the single family home market
Just looking at the median size number misses this much larger point about housing markets. The trouble with this misleading median, as we explained in our March post on this subject, is that it is highly influenced by compositional effects. So that when the bottom drops out of the housing market, and Americans of typical means can no longer afford homes or qualify for mortgages, the construction of smaller, low-cost homes evaporates. Basically, in this very depressed housing market, the only people who can afford new homes (and qualify for mortgages, given much, much tighter underwriting standards) are high-income, high-wealth households. Which is why the market for big homes 4,000 square feet and larger is down a mere 43% from the peak, compared to a 60% decline for homes under 1,800 square feet.
The only way the “growth of McMansions” story holds up is if you believe that the composition of homebuyers in today’s very distressed market is representative of what Americans would buy if incomes and lending standards were at “normal” levels. If we were ever to get back to the 1.5 million annual housing start figure that we took for granted prior to 2007 (which seems an increasingly doubtful proposition), the growth in home buyers would come from those with lower and moderate incomes who would be far less likely to buy McMansions.
From our perspective, the big story about American single family homes is that, from micro-homes to McMansions, they’re all still in dire straits. For one of the country’s major industries to be at half its historic rate of production is a big deal, both for housing and the broader economy.
There’s no question that the median size of new single family homes is larger today than a few years ago. But as we’ve explained here, that actually signifies something very different about the housing market, and American’s demand for housing than conveyed by a quick look at the misleading median.
This isn’t actually a post about transit. It’s about land use. But we’ll get there in a second.
Enrique Peñalosa, the former mayor of Bogotá, Colombia, is responsible for one of the most widely-shared quotes in the urbanist world: “An advanced city,” he said, “is not one where the poor use cars, but rather one where the rich use public transport.” The idea is that the end game of “development” isn’t to maximize private resources, but to create such strong public infrastructure that everyone is happy to use it.
But what about the opposite kind of city—one where the rich use public transit more than the poor? What can we say about that kind of development?
Well-to-do suburbanites on public transit
As it turns out, we don’t have to look very far. Above is a map of the Philadelphia region. Blue areas represent the familiar situation in which transit riders are poorer than drivers; but in orange areas, transit riders are richer than drivers. (I’ve whited out both kinds of places if fewer than 5% of all commuters use transit.)
But most of the rest of the region is orange, turning our expectations on their head twice over: not only is it odd that there are places where transit riders are richer than people who drive, but to the extent those places exist, our intuition (or at least mine) would dictate that they would be in central cities, where transit service is very good—in other words, where we come closest to meeting Peñalosa’s vision. Instead, they’re overwhelmingly in the suburbs.
(And note that the map compares people only to their neighbors—so transit riders in the suburbs don’t appear richer just because the suburbs are wealthier in general. This shows that suburban transit riders are wealthier than suburban drivers.)
Nor is Philadelphia unique. The Minneapolis-St. Paul region, for example, shows a similar pattern.
What’s going on?
So how do we explain this? Well, here’s where we get back to land use. In America, people with higher incomes tend to have certain kinds of jobs: in particular, white-collar office jobs in fields like insurance, law, finance, and so on. In many American cities, those jobs are heavily concentrated in the downtown core. In cities like Philadelphia, which has an extensive commuter rail network, or Seattle and Minneapolis-St. Paul, which have a pretty good network of regional express buses, that makes commuting from the suburbs quite convenient: You can walk from the downtown station to your office, avoiding both the frustrations of driving in rush hour traffic and the expense of downtown parking.
But the situation looks very different for lower-income people. Those jobs are disproportionately likely to be blue-collar manufacturing or service sector, which are much more scattered across the metropolitan area. If you live in the suburbs, the prospect of commuting to another suburb by transit is probably pretty bleak: in most regions, very few suburban jobs are walking distance from a rapid transit station, and local suburban buses are often unreliable and too slow to efficiently travel across the massive distances of American metropolitan areas.
Faced with unreliable, extremely slow commutes by transit, most of those blue-collar and service sector workers will just find a way to buy a car and drive—even if it eats into the money they have for other important expenses. And so you end up with a situation where a lot of wealthy people have an easy transit commute to their jobs, but lower-income people do not.
Partly, this is a story about how different kinds of transit investments can have very different impacts: building another commuter rail line, for example, versus increasing frequency on a suburb-to-suburb bus.
Where are the jobs? Where are the stores?
But as I said at the top, it’s really a story about land use. Given the exact same access to transit infrastructure, the wealthy are able to use it more—simply because their jobs are clustered around transit, and other people’s jobs aren’t.
Not so much to walk to from here. Credit: Paul Sableman, Flickr.
Looked at one way, that’s actually good news: rather than spending years planning and scrounging up money for new transit infrastructure, American metropolitan areas can massively increase the usefulness of the infrastructure they have—and improve their residents’ mobility, both physical and economic—by changing their zoning laws to encourage development within walking distance of transit. A few years ago, a meta-analysis of almost fifty travel behavior studies found that a density of destinations in a central location was one of the most important factors in making transit a useful form of transportation.
That applies to jobs, from manufacturing (where possible) to call centers—but since most trips aren’t actually commutes, it should also apply to all the other things people need access to: supermarkets, grocery stores, and homes. In many cases, bringing people, jobs, and amenities closer to transit is easier, and just as effective, as trying to bring transit closer to them.
* I should note that this pattern—with suburban transit riders being wealthier than suburban drivers—doesn’t hold everywhere. The necessary conditions seem to be 1) a concentration of high-paying jobs in a relatively dense downtown; 2) some sort of high-quality suburb-to-city transit, whether commuter rail or express buses; and 3) poor suburb-to-suburb transit service. Where one or more of these conditions are missing, you’ll get very different results. Here, for example, is Miami:
This isn’t actually a post about transit. It’s about land use. But we’ll get there in a second.
Enrique Peñalosa, the former mayor of Bogotá, Colombia, is responsible for one of the most widely-shared quotes in the urbanist world: “An advanced city,” he said, “is not one where the poor use cars, but rather one where the rich use public transport.” The idea is that the end game of “development” isn’t to maximize private resources, but to create such strong public infrastructure that everyone is happy to use it.
But what about the opposite kind of city—one where the rich use public transit more than the poor? What can we say about that kind of development?
Well-to-do suburbanites on public transit
As it turns out, we don’t have to look very far. Above is a map of the Philadelphia region. Blue areas represent the familiar situation in which transit riders are poorer than drivers; but in orange areas, transit riders are richer than drivers. (I’ve whited out both kinds of places if fewer than 5% of all commuters use transit.)
But most of the rest of the region is orange, turning our expectations on their head twice over: not only is it odd that there are places where transit riders are richer than people who drive, but to the extent those places exist, our intuition (or at least mine) would dictate that they would be in central cities, where transit service is very good—in other words, where we come closest to meeting Peñalosa’s vision. Instead, they’re overwhelmingly in the suburbs.
(And note that the map compares people only to their neighbors—so transit riders in the suburbs don’t appear richer just because the suburbs are wealthier in general. This shows that suburban transit riders are wealthier than suburban drivers.)
Nor is Philadelphia unique. The Minneapolis-St. Paul region, for example, shows a similar pattern.
What’s going on?
So how do we explain this? Well, here’s where we get back to land use. In America, people with higher incomes tend to have certain kinds of jobs: in particular, white-collar office jobs in fields like insurance, law, finance, and so on. In many American cities, those jobs are heavily concentrated in the downtown core. In cities like Philadelphia, which has an extensive commuter rail network, or Seattle and Minneapolis-St. Paul, which have a pretty good network of regional express buses, that makes commuting from the suburbs quite convenient: You can walk from the downtown station to your office, avoiding both the frustrations of driving in rush hour traffic and the expense of downtown parking.
But the situation looks very different for lower-income people. Those jobs are disproportionately likely to be blue-collar manufacturing or service sector, which are much more scattered across the metropolitan area. If you live in the suburbs, the prospect of commuting to another suburb by transit is probably pretty bleak: in most regions, very few suburban jobs are walking distance from a rapid transit station, and local suburban buses are often unreliable and too slow to efficiently travel across the massive distances of American metropolitan areas.
Faced with unreliable, extremely slow commutes by transit, most of those blue-collar and service sector workers will just find a way to buy a car and drive—even if it eats into the money they have for other important expenses. And so you end up with a situation where a lot of wealthy people have an easy transit commute to their jobs, but lower-income people do not.
Partly, this is a story about how different kinds of transit investments can have very different impacts: building another commuter rail line, for example, versus increasing frequency on a suburb-to-suburb bus.
Where are the jobs? Where are the stores?
But as I said at the top, it’s really a story about land use. Given the exact same access to transit infrastructure, the wealthy are able to use it more—simply because their jobs are clustered around transit, and other people’s jobs aren’t.
Not so much to walk to from here. Credit: Paul Sableman, Flickr.
Looked at one way, that’s actually good news: rather than spending years planning and scrounging up money for new transit infrastructure, American metropolitan areas can massively increase the usefulness of the infrastructure they have—and improve their residents’ mobility, both physical and economic—by changing their zoning laws to encourage development within walking distance of transit. A few years ago, a meta-analysis of almost fifty travel behavior studies found that a density of destinations in a central location was one of the most important factors in making transit a useful form of transportation.
That applies to jobs, from manufacturing (where possible) to call centers—but since most trips aren’t actually commutes, it should also apply to all the other things people need access to: supermarkets, grocery stores, and homes. In many cases, bringing people, jobs, and amenities closer to transit is easier, and just as effective, as trying to bring transit closer to them.
* I should note that this pattern—with suburban transit riders being wealthier than suburban drivers—doesn’t hold everywhere. The necessary conditions seem to be 1) a concentration of high-paying jobs in a relatively dense downtown; 2) some sort of high-quality suburb-to-city transit, whether commuter rail or express buses; and 3) poor suburb-to-suburb transit service. Where one or more of these conditions are missing, you’ll get very different results. Here, for example, is Miami:
One of the most controversial recommendations from Seattle’s affordable housing task force, or HALA, was to reform zoning laws that only allow single-family homes in certain neighborhoods. That was always going to be a challenge—as Sonia Hirt argues in her history of American zoning, Zoned in the USA, prioritizing and protecting single-family-home-only neighborhoods from other kinds of uses, including multi-family apartments and condos, is possibly the defining feature of American urban planning. Unsurprisingly, opponents launched a campaign warning their fellow single-family-home dwellers of impending doom: multi-story apartments towering over quiet streets, bringing traffic, shadows, and all sorts of unsavory renter types.
But the HALA suggestion was much more modest than that vision of catastrophe. Rather than building highrises, or even midrises, in previously suburban-looking streetscapes, they wanted to allow a kind of housing that’s become practically extinct in many cities: duplexes and triplexes.
Extinct, anyway, in new construction. This kind of mid-density, low-rise housing—including duplexes, triplexes, townhomes, and other low-density multi-family buildings—has been called the “missing middle”: American cities build lots of single-family homes, and (in a certain places) some larger apartment complexes, both in the form of sprawling suburban “apartment communities” and downtown highrises. What we don’t build are the kind of human-scaled, moderately-dense housing that has historically made up the bulk of America’s urban neighborhoods.
Neighborhoods in Chicago, San Jose, and New Orleans with a traditional mix of single-family homes and lowrise multi-family.
Here, for example, is the breakdown of all new construction in the Seattle metropolitan area over the last ten full years:
And it’s not just a Seattle thing. Cities across the country have the same issue:
Why care about housing’s “missing middle”?
Why does any of this matter? Several reasons. In places where housing prices are an issue, small multi-unit buildings can provide lower-cost housing at market rates—and lower construction costs for nonprofit developers building subsidized housing. These buildings can be placed in single-family districts, many of which are low-poverty “opportunity areas,” without disrupting their lowrise character.
Second, duplexes and triplexes can be a low-visual-impact way to add people to a single-family-home neighborhood. The added density, in turn, can make those communities viable for walkable neighborhood commercial districts, or high-frequency, “show up and go” bus service.
In other words, the “missing middle” is a way to diversify and urbanize low-density neighborhoods without drastically changing the appearance or character of quiet, “suburban”-looking streets that residents of single-family-home areas often value. Charles Marohn of Strong Towns has written about this extensively as a kind of gradual “maturing” of residential neighborhoods, rather than “leapfrog” construction that jumps from mostly single-family homes to midrises or highrises right away.
And missing middle housing can be one key to helping older Americans age-in-place, staying in the same community that they’ve been in, but trading a larger, single family dwelling, for a smaller, but still very much in character unit in a duplex, triplex or other one of these “middle” type housing units. Many aging baby-boomers live in suburban, single family tracts where there are few alternatives to down-size in their current neighborhoods. As we’ve noted, all of the net growth in home-ownership in the next two decades is expected to be among households whose heads are aged 65 and older. Many of these households would no doubt enjoy being owners of smaller scale properties in their own neighborhoods–if they were available.
Of course, denser forms of building are still important. Midrises and highrises allow lots and lots of people to live near major commercial corridors, job centers, and transit hubs, all of which is crucial for giving our cities’ residents economic and social opportunities. High-density buildings also minimize disruption in their own way—a 100-unit highrise downtown might only take up a street corner, but 50 duplexes would take up several whole blocks.
Still, it seems like too many of our cities, including Seattle, have just two settings for urban growth: sprawl and full-on urban density. Figuring out exactly where the roadblocks are to a more moderate kind of density—regulation, financing, and so on—may help keep neighborhoods open to more (and more diverse) people, while respecting the look and feel many residents of relatively low-density communities would like to keep.
Since 2000, home prices have grown 50 percent faster in urban centers than in their surrounding metro areas. If your are an urban data geek, like we are, this is big news. A dramatic shift in city-suburb price differentials strongly signals a deep and enduring market demand for cities.
A new research report from investment rating agency Fitch provides strong confirmation for the case we’ve been making at City Observatory: the demand for urban living is strong and increasing. Fitch calls the trend “striking” and summarizes:
Since the mid- 1990s, demand has skyrocketed in many urban centers, with home price growth in the closest distance tiers growing at significantly higher rates than the MSAs in which they are contained.
Compiling Case-Shiller home price data for 4,600 zip codes in 50 large metropolitan since the 1970s, Fitch shows that home prices in urban centers have significantly outperformed home prices in the balance of their respective metropolitan areas throughout the country. The pattern is widespread: it holds for large and high value coastal markets like San Francisco and Boston, but also for middle sized metros like Nashville, Denver and Portland.
The Dow of Cities
Fitch summarizes its national findings in a single chart, showing how home values in the most central urban neighborhoods performed compared to successively more distant tiers of housing.
If you were to construct a “Dow-Jones” Index for cities–something that would be a comprehensive, consolidated, all-purpose summary measure of the relative economic strength of cities compared to their suburbs–it would look very much like the chart prepared by Fitch. The steady divergence between the typical home values for city centers (shown in blue; the densest and most central one-fifth of census tracts) relative to the rest of their respective metropolitan areas is striking. The chart which shows home values indexed to a 1975 base year clearly illustrates the housing bubble that peaked about 2006, and shows that while city centers also experienced absolute price declines (as did the rest of the market), they fared far better than prices in each of the four more peripheral tiers.
The Fitch analysis of relative city-suburb price trends confirms what we showed was happening in the Portland metropolitan area in a commentary we wrote last fall: in Portland, since 2005, the traditional relationship between city and suburban home values has reversed: a decade ago, city center homes sold at an 9 percent discount to those in the suburbs, now they sell at an 7 percent premium. Fitch shows this same pattern holds nationally.
Behind the price trend
The reasons behind this shift are familiar to City Observatory readers, and are laid out in several of our reports. In Young and Restless, we showed that well educated young adults (25 to 34 year-olds with at least a four year degree) are increasingly choosing to live in close-in urban neighborhoods in the nation’s largest metropolitan areas. In Surging City Center Job Growth, we showed how, for the first time in decades, employment growth in city centers was outpacing that in suburbs, in part at least, to the growing desire of firms to locate their operations closer to the preferred residential locations of young workers.
When it comes to explaining these trends, Fitch’s investment analysts sound like dyed-in-the-organic-merino new urbanists:
Updated urban planning that focuses on walkable cities, improved transportation networks and green space has improved the urban quality of life, drawing in multitudes of residents, who, in prior generations, aspired to stretches of lawn in the less dense suburban and exurban rings.
Its notable that Fitch mentions “walkable cities.” A key factor in the attractiveness of city living is walkability, and various studies have shown that increases in the walk score are associated with higher home values. The Fitch charts of city v. suburban home price disparities are strikingly similar to those generated by Zillow economists showing the relationship between walkability and home price performance. As we highlighted at City Observatory last month, within metropolitan areas, home values in more walkable neighborhoods have dramatically outpaced home prices in car dependent locations.
A big deal for the future of housing — and the suburbs
Fitch’s analysis suggests the move back to the city has important implications for the housing market. Most importantly, they don’t expect a resurgence in the falling homeownership rate:
With a trend toward increasing populations in many cities, where a far higher portion of units are rentals than in suburban areas, a return to the lofty home ownership rates seen before the housing crisis of the 2000s is unlikely . . . .
And Fitch is glum about the future of sprawling suburbs and exurbs: the shift of demand to cities implies “long run risks of declining property values in the urban periphery.”
While the strong–and in Fitch’s view, accelerating–price premium for city center living is a harbinger of robust growth in the urban core for years to come, it also signals what we think is a huge challenge for the nation. Rising prices are a clear sign that we have a shortage of cities. Many of our urban problems–especially those related to housing affordability–are directly tied to American’s growing demand for city living, and the relative paucity of places with great urban character. Higher prices are a market signal that we need more and better cities.
The full report -“U.S. RMBS Sustainable Home Price Report,Second-Quarter 2015 Update, Special Report, August 12, 2015-is available at the Fitch website.
Back in June, we catalogued how riders weren’t really abandoning buses—buses were abandoning their riders, with significant cuts to service in many metropolitan areas that appeared to be driving declines in ridership.
Further analysis of transit data since 2000 suggests that bus riders may have another problem: not only are there fewer buses on their route, but when they do show up, they’re slower.
In this chart, we’ve used data from the National Transit Database to compute the average system-wide bus speed, by dividing the total number of revenue miles traveled by the total number of revenue hours operated. This gives us a single, highly aggregated measure of how fast a “typical” bus is moving in each city in each year.
Since 2000, the median urbanized area of a million people or more has seen their average public bus speed drop from 13.6 mph to 12.7 mph, or about 6.6%. That might not seem like a lot, but it adds up to about 20 minutes a week for someone whose commute used to be half an hour each way. Especially if the wait for the bus is longer, adding another few minutes for each trip, that could be significant.
Interpreting these numbers is difficult, though. There are any number of reasons that average speed might decline. The most obvious possibility, perhaps, is simply that there’s more general traffic, slowing buses along with all other motorized vehicles on the road. In some ways, this makes sense. It’s clear that for the most part, larger cities and those with more congested roads have slower buses: the average bus speed in New York in 2013 was 9.5 mph, while Phoenix’s buses traveled at a much faster 12.2 mph. (Of course, as with private vehicle congestion, slower doesn’t always mean worse accessibility: if New York buses are 25% slower, but your destinations in New York are 50% closer, you’ve got better access to where you’re going in New York.)
On the more nefarious side, cutting service could end up slowing buses down if there are now more passengers per pickup, forcing the bus to wait longer at its stops for everyone to get on and pay. While that may be a problem for some lines and in some metropolitan areas, there doesn’t appear to be any overall correlation between cutting service and slower buses:
It’s also possible that average speeds could decline without any clearly negative impacts on service. Our measure of system-wide bus speeds will pick up changes in the mix of route types that a city runs, so some year-over-year changes may reflect that, rather than changes in the speed of any particular route. Imagine, for example, a transit agency that decides to cut back on suburb-to-city commuter express buses with little ridership and instead focus on local routes in neighborhoods where residents, jobs, and amenities are more densely packed. That would almost certainly reduce the agency’s average bus speed, but might at the same time increase the number of people with access to useful transit service.
Still, though it’s impossible to know with this data, that explanation seems unlikely to be behind the change across the more than forty urbanized areas with over a million people included in this analysis. To know for sure, you’d have to look metro by metro—and probably agency by agency—and look at their service changes over the last thirteen years.
At the very least, though, these numbers should give transit advocates pause. While vehicle speed isn’t the only, or even most important, factor in good bus service, it does matter—and slow speeds are hardly inevitable. Better stop spacing, all-door or proof-of-payment boarding, and bus lanes can all dramatically improve speeds. The current downward trend suggests that these interventions may be increasingly needed.
1. Let’s talk about neighborhood stigma. Daniel Kay Hertz reviews some of the literature on the interplay between a neighborhood’s reputation and its disadvantage—and finds a surprising reversal in the conventional understanding of the issue. Rather than problems like greater crime or vandalism leading to bad reputations, researchers like Harvard’s Robert Sampson have found that communities acquire reputations in large part because of factors like racial makeup. Those reputations, in turn, create stigma that actually create many of the disadvantages that they supposedly reflect.
2. Urban buses are slowing down. A look at the data shows that since 2000, the average bus speed in urban areas of at least a million people has slowed by close to 7%, from 13.6 to 12.7 mph. Daniel Kay Hertz goes through why that matters, and a few of the possible culprits.
3. Revisiting Marietta. Joe Cortright returns to the subject of the Atlanta suburb that’s tearing down “naturally occurring” affordable housing for a private commercial development. Joe argues that it’s wrong to single out Marietta—many other cities have simply been more successful in keeping out low-income housing, and low-income people, to begin with. Those municipalities, and the laws that enable them, also ought to come under criticism.
4. The McMansion mirage reappears. Joe Cortright returns to the question of what it means for McMansions to be “back.” Several media outlets have made that claim, following reports that the median square footage of newly-built single family homes is climbing. But actual levels of McMansion construction are down over 40% since before the housing bust.
1. It’s not a read, but The Problem We All Live With, New York Timesreporter Nikole Hannah-Jones’ hourlong documentary for This American Life, is a must-listen. Hannah-Jones covers the problem of school segregation through the story of a school district outside St. Louis that did something very unusual—it accidentally desegregated. What happened afterwards is both a reason for hope and a damning account of the racial hurdles many Americans still aren’t ready to overcome.
2. In “Where Should a Poor Family Live?”, Thomas Edsall criticizes some affordable housing advocates and developers for building housing mostly in low-income neighborhoods, rather than higher-income areas that might lead to greater economic and racial integration. He even notes that several affordable housing organizations wrote an amicus brief to the Supreme Court asking it to strike down the stronger “disparate impact” interpretation of the Fair Housing Act. (For an opposing view, professor Edward Goetz has some critiques of the move to disperse affordable housing.)
3. We don’t generally cover city finances at City Observatory, but thisGoverning piece on how many public pension funds are struggling should alarm anyone who’s committed to using local resources for maintaining important social services, including housing support, transit, and other community development. Philadelphia, for example, saw a return of just 0.5 percent in Fiscal Year 2015, and in New York City, four of five public pension plans saw returns of under 5 percent—but most pension plans are funded under the assumption they will see an average return of 8 percent.
New knowledge
1. At the Urban Institute, three writers argue that existing data on local businesses, while able to shed some light on neighborhood change, is inadequate to the needs of planners and researchers. Christina Plerhoples Stacy, Brett Theodos, and Carl Hedman write that two rapidly gentrifying neighborhoods in Washington, DC, have seen faster growth in total businesses compared to a non-gentrifying district—but, somewhat counter-intuitively, have actually seen a decline in the number of full-service restaurants, while limited-service restaurants spiked. Existing data, however, can’t distinguish between a fast-food McDonald’s or high-priced cafe, and so make it hard to interpret these numbers in a shifting social and economic context.
2. A new paper from Robert Noland and Stephanie DiPetrillo in the Journal of Transport and Land Use provides even more evidence for the impact of transit-oriented development on the choice of transportation. Shockingly enough, even with a bevy of controls, those who lived in housing close to train stations in New Jersey were much more likely to use transit that those who lived further away.
3. Conventional wisdom is that racial segregation in America, though still unacceptably high, has been declining since about 1970. But a new paper from Daniel Lichter, Domenico Parisi, and Michael Taquino in the American Sociological Review challenges that view. While acknowledging that neighborhood-level segregation has declined, the study suggests that municipality-level segregation has actually increased since 1990. Moreover, because municipalities are the level of both social service delivery and exclusionary policies, the authors argue that this may be a more meaningful geography of analysis, casting doubt on progress many believe American cities have made.
The Week Observed is City Observatory’s weekly newsletter. Every Friday, we give you a quick review of the most important articles, blog posts, and scholarly research on American cities. You can sign up to get it in your inbox by clicking “Subscribe” at the top of the page!
Our goal is to help you keep up with – and participate in – the ongoing debate about how to create prosperous, equitable, and livable cities, without having to wade through the hundreds of thousands of words produced on the subject every week by yourself.
If you have ideas for making The Week Observed better, we’d love to hear them! Let us know at jcortright@cityobservatory.org, dkhertz@cityobservatory.org, or on Twitter at @cityobs.
Last month, we questioned why people weren’t paying more attention to Marietta, the Atlanta suburb that is tearing down 1,300 apartments and permanently displacing their low-income residents. We wondered why this large-scale displacement of poor households—most of whom are black or Latino—didn’t generate the same kind of outcry as much more ambiguous situations in urban settings that are criticized as heartless examples of gentrification.
Our commentary generated press coverage in the local Marietta Daily Journal. Reporters for the Daily Journal put our critique to local Councilman Grif Chalfant. Chalfant defended the plan, saying that by demolishing the Franklin Road apartments, the city was dispersing a high-poverty neighborhood. He compared Marietta’s demolition of those homes to the razing of public housing in Atlanta, Chicago, and elsewhere. Of course, the Franklin Road apartments weren’t public housing—they were private homes purchased by the city. And there was still no indication of whether or how their residents would be given relocation assistance, or if they would simply end up in another low-income neighborhood further away.
The main square in Marietta. Credit: Ken Cook, Flickr.
In our original commentary, we conjectured that there’s been a widespread internalization of the notion that suburbs are white and wealthy, and city centers are for people of color and the poor—and that anything that changes this arrangement is a violation of the natural order of things. While that is certainly true, in a sense, it’s unfair to single out Marietta. Ironically, this city found itself with a concentration of affordable housing—and lower income families—largely because it failed to to be as effective at exclusionary zoning as are most suburbs. When suburbs ban or greatly restrict apartment construction, there’s little chance of becoming another Marietta. The more common and more daunting example of suburban exclusivity may the be places that never have any affordable housing.
Consider Marin County, a supposedly liberal part of Northern California where well-to-do local residents have reacted with great alarm that Star Wars director George Lucas is planning on building 224 units of affordable housing on land he owns. (And keep in mind that in this upscale Bay Area suburb, “affordable” means housing for households with annual incomes between $65,000 and $100,000 per year.) Thanks to exclusionary zoning and other measures, Marin County has very little affordable housing of any kind, and is now enthusiastically opposing the small amount they might get. Something similar happened in Darien, Connecticut, when a local couple tried to develop affordable housing in a wealthy suburb.
Neither Marin County nor Darien are planning to tear down any low-income apartments like Marietta is—but perhaps only because they’ve been so effective at keeping out the poor that there aren’t any low-income apartments to tear down.
It may be that Marietta is the victim of two things: the large-scale Euclidean zoning that separated multi-family housing from other sections of the community, causing it to decline in value while surrounding single family neighborhoods continued to be desirable by more solidly middle-class and upper-middle-class residents. If instead of being concentrated in just one part of the community, these multi-family units had been more dispersed in mixed-use, mixed income-developments, they might have avoided the “concentrated poverty” that prompted the city to acquire and demolish them.
A zoning map of Marietta. Yellow areas are zoned for single-family homes only; brown areas are set aside for apartments. The large brown area in the southeast corner contains the homes to be razed. Pink is commercial. Source: Marietta, GA website
The second problem may be the high elasticity of the housing supply in Atlanta, and the region’s widespread sprawl. (Elasticity is the economist’s term for the ease with which new units can be added to a region’s housing stock. “High elasticity” means you can add more housing very easily.) In another metropolitan area, like San Francisco or Washington, DC, 1960s-era apartments might still command a high price, because of the very constrained housing supply. In Atlanta, where it’s relatively easy to build more housing, older apartments can “filter” down market faster—leading to the kind of rapid increase in poverty seen in Marietta. It may be that displacement is less an issue in the Atlanta area than in other metros because new housing is so readily constructed.
Still left unanswered are the fiscal “beggar thy neighbor” qualities of this policy decision. Marietta officials are quite candid that lowering their city’s costs for schools and public safety expenditures are a key motivation for undertaking the demolitions. And while Marietta may shed these costs, it is certain that at least the cost of educating the children displaced from these apartments will be shifted to other school districts.
But for every Marietta that takes the visible step of demolishing some aging apartments that have moved down market, there are dozens (or perhaps hundreds) of other suburbs that never allowed multi-family housing to be built in significant numbers in the first place, and so precluded this question from ever being raised. And while it may be far less visible and obvious, this use of local land use controls to engineer the social and class structure of a community is no less profound in restricting the opportunities of lower income Americans.
My hometown, Chicago, is having a fight over words: in particular, “Chiraq.” That’s a portmanteau of “Chicago” and “Iraq,” which is meant to analogize the city not to that country’s rich cultural heritage, or extreme weather, but to its war. The name seems to have come from a South Side rapper, but has since been popularized by headlines in decidedly non-South Side outlets, at least threedifferentVICE mini-documentaries, and most recently an in-the-works Spike Lee movie.
Most cities might not have a nickname with such staying power, but they’re familiar with the concept. “Killadelphia”; “Murder Worth”; “Bullet Town”; even “Murder Kroger.” Last year, a website called “Judgmental Maps” made a name for itself by posting annotated maps of American cities; some of the entries for Atlanta, to take a city at random, included “Little Crackistan,” “Dangerous Mexicans,” and “Avoid.” The map for Washington, DC, labeled all of the Anacostia area simply “GUNS & AIDS.” Around the same time, an app called SketchFactor announced that it could help you get where you were going while avoiding “sketchy” neighborhoods.
Judging by the debate in Chicago, many people don’t see a problem with these expressions of local reputation. In a widely-shared essay, one prominent writer declared that “Chiraq” was indicative of real problems, from violent crime to concentrated poverty, and “arguing over what to call” that “shameful reality” was simply a tool of distraction—putting up a Potemkin facade to avoid dealing with the real issues.
Part of Judgmental Maps’ map of Washington, DC.
It’s certainly true that the problems “Chiraq”—and, for that matter, “Killadelphia” or “Bullet Town”—was meant to encapsulate are real. (Even the “Murder Kroger” really did witness a murder.) But the second part of the argument—that the name itself is harmless—is simply not true.
Over the last few years, issues of racial and economic segregation have seen a new burst of attention, covering historic issues like redlining and cutting-edge research on the effect of concentrated poverty on economic mobility. But one of the squishier sides of segregation has received much less coverage: stigma.
Stigma creates the very problems it supposedly reflects
In part, that’s understandable: it’s much harder to measure stigma than it is to measure segregation—or even, as it turns out, to measure intergenerational economic opportunity. But that doesn’t mean it’s not a real, and powerful, force. In fact, strong evidence suggests that stigma canhelp create the very disadvantages it supposedly reflects.
In his book Great American City: Chicago and the Enduring Neighborhood Effect, the Harvard sociologist Robert Sampson recounts an experiment that involved taking extensive data, including resident surveys, from neighborhoods across Chicago over several years, from roughly 1995 to 2002. One question was whether they could use data collected in 1995 to predict what would happen to neighborhoods over time—including what would happen to their poverty rates.
The results are amazing: a neighborhood’s reputation in 1995 was a stronger predictor of poverty in 2000 than almost any other variable, including the neighborhood’s poverty level in 1995. Put simply, if you want to predict what a neighborhood’s poverty rate will be in a few years, knowing whether that neighborhood has a bad reputation is just as useful as knowing its actual poverty rate today.
But don’t neighborhoods get bad reputations because they already have serious issues? To some extent, yes. But Sampson also found that neighborhoods stigmatized as places where, say, public drinking, fighting, or drug dealing were a major problem were not necessarily the neighborhoods with the highest levels of those activities. Instead, the race and immigrant status of a neighborhood’s residents had a “more powerful” effect on reputation than the actual amount of undesirable behavior.
Reputation as a mechanism of inequality
In other words, much of the popular understanding of how neighborhood reputation works is backwards. Communities acquire a reputation for being “sketchy” to some extent independently of whether or not that “sketchiness” is real—and in a way that’s heavily influenced by racism. Once they have a bad reputation, however, the stigma helps create the very problems it warns others away from—in part by causing people to avoid the neighborhood. In Sampson’s words, “stigmatization” becomes a “self-fulfilling prophecy,” and “shared perceptions of disorder…appear to be a mechanism of durable inequality.”
Nor is Sampson’s work the only evidence pointing towards this conclusion. In a paper published in April of this year, researchers from NYU found that online classified ads from stigmatized neighborhoods received many fewer responses than the exact same ads that described the poster as being from a non-stigmatized neighborhood, suggesting that a neighborhood’s reputation may affect its residents’ economic opportunities. And anecdotally, many city leaders say the same thing. Our colleague Carol Coletta at the Knight Foundation once asked then-Newark Mayor Cory Booker what his greatest challenge was. His replied, “Getting people to believe that things can be different.” He saw a realistic hope that things could get better as an important factor in creating and sustaining positive change.
One part of a system of disadvantage
To be clear, the geography of inequality is not simply an invention of our heads. Very real economic and social systems stretching back generations have led to dramatic and devastating differences in quality of life from one neighborhood to the next in nearly every city in America, and pretending that those inequalities don’t exist would ensure that we don’t do anything to rectify them.
But it is now very clear that the reputations we help create through the way we talk about local inequalities—perceived or actual—have a real, and dangerous, power. At the extreme, I suspect we’ve always know this: presumably even the defenders of “Chiraq” would acknowledge that popularizing a community as “Little Crackistan” could not possibly do anything positive for that neighborhood’s trajectory. But stigma doesn’t need to be so crass to work perniciously. The challenge, then, is to talk about our deeply unequal cities in a way that is tightly tied to the actual facts on the ground, and not racial stereotypes; to describe, without dismissing or caricaturing.
That’s a pretty vague charge, it’s true. But maybe we can start simply by acknowledging that words, and the stigmas they can represent and shape, aren’t harmless.
