Tuesday, April 18, 2017

Are The Suburbs Really Overtaking Urban Areas?


American suburbia
Photograph by Andreas Praefcke/CC 3.0
Hello Everyone:

Welcome to a new week.  United Airlines is still plagued by bad publicity, resulting from their anal retentive adherence to the rules.  History challenged White House Press Secretary Sean Spicer re-surfaced at the annual Easter Egg Hunt, reading a story to a group of children.  So far, so good.  On to today's subject: are the suburbs outpacing the cities in population growth?

On December 3, 2016, Laura Kusisto wrote in the Wall Street Journal an eye-grabbing article titled "Suburbs outstrip cities in population growth study finds." (http://www.wsj.com; date accessed Apr. 17, 2017).  Ms. Kusisto wrote,

Big cities may be getting all the attention, but the suburbs ares holding their own in the battle for population and young earners...research shows that suburbs are continuing to outstrip downtowns in overall population growth, diversity, and even younger residents...  (Ibid)

Joe Cortright, in his CityLab article, "Are the Suburbs Really Back?" questions whether Ms. Kunsisto's is accurate.  Mr. Cortright writes, "On its face, the article seems to imply that much of what has been written in recent years about a rebound in cities is either wrong or somehow overstated."  Although the WSJ article was quick to paint the report as urban-suburban race for suburbs, an Urban Land Institute press release was more guarded:

Urban construction
Suburban housing markets across the United States are evolving rapidly and overall well positioned to maintain their relevance the foreseeable future as preferred places to live and work, even as many urban cores and downtown neighborhoods continue to attract new residents and businesses, according to a new publication from ULI... (http://www.uli.org)

The full report was published on the ULI website December 5, 2016.  The report, Housing in the evolving American suburb, was prepared the real estate consulting firm RCLCO, which presents their classification of individual census tracts in the fifty largest metropolitan areas.  Mr. Cortright calls the report "...an ambitious undertaking, classifying even census tract in the 50 largest U.S. metropolitan areas according to a new and quite complicated neighborhood taxonomy."  He admits that he has not had a chance to carefully vet the data, but he does have some initial thoughts.

Grading on the Curve
Joe Cortright writes, "The ULI uses its own custom-crafted definition of cities and suburbs.  And it's quite unlike that used by other researchers."  The researchers examined data for the 50 largest metropolitan and group individual metropolitan areas in six different subject headings including: "gateway, sunbelt" and "legacy."  New York is in a category all to itself.  He continues, "Each category has its own definition of what constitutes 'urban,' as well as five different flavors of suburb, ranging from 'established high end suburbs' to 'greenfield value suburbs.'"  Mr. Cortright asks the reader to think of it as grading on a curve because what defines urban in one metropolitan is considered suburban in another metropolitan area.  The first appendix explains "that each category of metropolitan areas had a set of rules for classifying tracts as urban or suburban, and for suburban tracts, their exact sub-category, based in part on the distribution of data for that category of metropolitan areas, but there's no reporting of the exact cut-off points for each category."  Page 43 of the appendix,

For more information on the absolute cuts for any given MSA category, see the table in this appendix (Ibid), do not look for that table, according to Mr. Cortright "no such table appears in either of the report's two appendices."

Google street view of Chicago suburb
The arbitrary classification and statistical thresholds for placing neighborhoods in each category in different metropolitans left some quite puzzled.  Mr. Cortright reports, "According to the ULI report, as measured by population a higher fraction of neighborhoods in Milwaukee (24 percent) are urban than in Chicago (12 percent), San Jose (40 percent) is more urban that either San Francisco (20 percent) or New York (35 percent), and Houston (10 percent) is slightly more urban than Seattle (9 percent)"  A further look at the report's maps of pattern of urban and suburban classifications suggest that thing appear correct;  "the center of every metro area is urban, its surroundings suburban."  That aside, it is not always apparent that the reader would agree where the report authors delineated the boundary between urban and suburban in a specific metropolitan area.  Case in point, residents in the Chicago suburb (pictured above) would be surprised to find that they live in a "suburb" based on the ULI report.

Phoenix, Arizona
Even if there might be some disagreement with the results in some place, Mr. Cortright credits ULI with trying "something different."  He does note "There are some serious limitations to using municipal boundaries to distinguish between cities and suburbs."  Boundaries are fluid and the common practice to view the largest metropolitan as the "city" and everywhere else as "the suburbs."  Cities such as Phoenix, Austin, Jacksonville have large swaths of low-density density development within the largest city limits.  Another point, is in some metropolitan ares, "the largest city represents only a tiny fraction of the metro area-the cities of Atlanta and Miami are only about 10 percent of their respective metros...."  ULI's methodology is a sober attempt to avoid the issue, by examining density, housing types, and neighborhood characteristics.  Fortunately you can follow their map to see how each census tract has been classified.

It is a difficult and in many respects subject task to delineate urban from suburban, and there is ample room for healthy disagreement.  Mr. Cortright additionally credits the authors for being "...pretty clear about their approach (although it would be great if they had included a table showing the actual definitions used in each category of metropolitan areas)."  However, in CityLab's opinion, "...grading on the curve-using different rules to define what constitutes urban and suburban in different metro areas-make it difficult to interpret their national level results."

Moving beyond a binary classification

The Wall Street Journal article's chief claim is "...this ULI study sheds new light on the relative attractiveness and success of cities compared to suburbs."  One of the issues with this manner of analysis is that it divides the metropolitan world in two black-and-white categories: "city" and "suburb."  However, it would be better to look at the grey area: where people actually live.

"Distribution of Young Adults' Distance from City Center"
UVA Demographic Center
One useful tool is a set of charts created by the University of Virginia's Luke Juday.  Mr. Juday used census data and plotted population and demographic features of the population by using a series of concentric circles around the center of each large American metropolitan area.  Similar to the ULI study, he also focused for the 50 largest American metropolitan areas.  Mr. Juday's analysis charts how things have changed between 1990 and 2012.  The above table charts where young adults resided during the study period.

The above chart tracks the distribution of young adults (22-34 years old) by distance from the city center to their neighborhood.  The brown line symbolizes the 2012 distribution, the orange lis is the distributor in 1990.  Two points to note.  "First, both lines slop down to the right: young people are more likely to live close to the center of the metropolitan area than other Americans.  Second, the line for 2012 is decidedly steeper than the line for 1990.  This shows that young adults have a much stronger preference for central neighborhoods now..."  This a more nuanced and powerful proof of locational preferences of young adults than the summary data in the ULI report.

The housing bubble is over

The expansion of the housing is over, something the ULI study neglects to take into account.  The study treats the period 2000 to 2015 as if it was a single time period.  However, anyone paying attention to the housing markets or the economy would notice that this fifteen year period is broken down into two phases.  Phase I: 2000 to about 2007, corresponding to the expansion of the housing bubble which resulted in construction of a plethora of suburban and exurban single family homes.  Phase II: 2008 to 2015, corresponding to the Great Recession and slow recovery, a period where suburban single family residences languished, most of the housing market action has been focused on multi-family residences, mainly apartments in urban areas.

"Primary City vs. Suburban Population Growth"
Metropolitan Policy Program
It is very apparent that the housing pattern has been quite different prior to and after the recession.  Joe Cortright cites the Brookings Institution analysis, Mid-Decade, big-city growth continues, of the city-suburb population trends assembled by Bill Frey (http://www.brookings.edu).  The Brookings Institute report uses a central definition of city connected to the most populous urban areas in each metropolitan ("primary cities"), which is somewhat limited.  Mr. Cortright reports, "For 2000-2010, suburban population grew faster than city population, 1.38 percent per year to .0.43 percent.  In of the years since 2010, city population has exceeded suburban growth."

Joe Cortright's own analysis, Surging City Center Job Growth (http://www.cityobervatory.org: date accessed Apr. 18, 2017), demonstrate similar trends between 2001-2007 (i.e. suburbs expanded faster) and beyond (when city centers expanded faster).

He writes, "Despite the headlines claims in the report, buried in a single paragraph in the report's Appendix B there's an acknowledgement that were very different after the Great Recession than before:"

Using the RCLCO classification, suburban areas were found to have seen a somewhat lower share of growth since 2010, at only percent of population and 76 percent of household growth.   (http://www.uli.org)

Allow Blogger to remind you that the report's primary conclusion about the prevalence of the suburban areas was "that they accounted for 91 percent of population growth over the 15-year period."  Mathematically speaking, this translates to suburbs made up 95 percent of the population growth between 2000-2010.  This implies that the percentage of population growth-mostly in cities-increased 5 percent between 2000-2010, approximately 20 percent during the 2010-2015 period-a serious increase.  Mr. Cortright observes, "The ULI report could more light on the question of city versus suburban growth.  If it had focused on the period since 2007, rather than simply present results that largely recount the unsustainable growth patterns recorded during the housing bubble."

17th Street Oakland, California
The myth of revealed preference

The implied argument backing the statistically based articles on how more Americans are living in the suburbs than cities, signifies a type of implicit preference for the suburbs.  There is a strong body of evidence that more people would live in urban areas if, here is the catch, there more more affordable housing available.  The strongest indicator of this is the increase of urban home prices relative to suburban home prices.  NIMBYism (or as Richard Florida prefers "urban luddites") and other serious obstacles to building more housing in urban areas which means fewer people can obtain homes that they enjoy living.  In his 2006 book Zoned Out, tried to answer the question: "Are local governments just responding to 'market' demand in ensuring that new development is low-density and auto-oriented?  Or is there really pent-up demand for urban neighborhoods that can't be satisfied because of zoning?"  (http://www.cityobervatory.org)

Decatur, Georgia
Mr. Levine looked for the answer in two very different metropolitan areas: Boston and Atlanta.  (Ibid)  He matched comparisons of consumers and neighborhoods to prove that more Americans would rather live in urban areas.

Joe Cortright reports, "As we've pointed out, the rising relative price of housing in the urban core shows a growing demand for urban living."  Fitch Investment Advisers (Ibid) applied zip code level information on housing prices in America's largest metropolitan areas to chart housing prices in close-in city neighborhoods relative to the remainder of metropolitan areas over a 25-year period.  Their data demonstrates that "the premium that buyers are will to pay to live close to the center has accelerated in recent years.  (Fitch divided up large metro areas into five concentric circles based on their distance of the CBD; prices in the closet tier outperformed all other tiers from 2000 onward; The Fitch finding has been separately corroborated by studies from Zillow and the Federal Housing Finance Agency.)

"Home Price Index Over Time Based on City Center Distances"
Fitch Ratings

In Appendix B of Housing in the evolving American suburb, the co-authors touches on the subject of lagging relative suburban home prices.  The co-authors references the study published by the Federal Housing Finance Agency revealing that home prices appreciated twice as quickly in neighborhoods near city centers as in in neighborhoods ten miles from central business district; in the peripheral areas saw even slower price increases.  Mr. Cortright writes, "They argue that we shouldn't be tempted 'to treat these prices trends as demonstrating a shift in housing preferences.'"  However, the counter evidence is not so much a an impeachment of the rising demand for urban locations relative to urban ones. as a reason for this trend.  We have been told that suburbs are becoming less of a magnetic because of the long commutes, while urban amenities are growing.  Mr. Cortright writes, "They also argue that it's easier to build new housing in the suburbs than cities...what this essentially show is that we have a shortage of cities, and an unrequited demand for urban living."

New York City street scene
About diversity

The ULI report claims that the suburbs are almost as diverse as the cities.  It states:

American suburbs as a whole are racially and ethnically diverse.  Fully 76 percent of the minority population in the 50 largest metro areas live in the suburbs-not much lower than 79 percent of the population in these metro areas as a whole.  (http://www.uli.org)

Joe Cortright critiques this statement, "If all suburbs were alike, this might be an almost convincing point.  But a key part of the ULI report is its observation that there are many different kinds of suburbs."  Sharing this information at a high level of aggregation vastly "obscures the differences within suburbs."  Specifically, "...62 percent of the population population of 'economically challenged suburbs' are racial of ethnic minorities, roughly double the fraction of the population in either 'established high end suburbs' (34 percent) or 'greenfield lifestyle suburbs' (27 percent)."  In an aside, "The report doesn't describe its data source, nor does it cite the exact definition of minorities that it uses..."
"Distribution of Black Residents' Distance from City Center"
UVA Demographics Research Group

Returning to the information from Luke Juday's Changing Shape of the American Metropolis spotlights a more nuanced light on the trend of racial and ethnic diversity across the metropolitan-scape.  The chart on the left shows a fraction of the population that is African American's distance from the central business district, aggregated for the 50 largest American metropolitan areas.  As you move toward the center, that fraction gets smilers.  Mr. Cortright writes, "While it's true that African-Americans are more decentralized now (brown line) than they were in 1990 (orange), it remains the case that blacks are about three times more likely to live within 5 miles of the city center than they are to live 15 or more miles."

Although racial segregation has somewhat mitigated in the United States, the cities remain more diverse than the suburbs; suburbans minorities are disproportionately concentrated in economically distressed suburbs, less likely to reside in more upscale suburban neighborhoods.

The verdict: Not proven

Joe Cortright concludes: "The three principal claims in the Wall Street Journal article, that suburbs are growing faster than cities, that they are 'outstripping' them in the growth of young residents, and they are more diverse, are all incorrect."  Between 2010 and 2015, cities grew faster than suburbs; close-in neighborhoods have become magnets for a disproportionate share of millennials, and cities more diverse.  Essentially, you do not need the WSJ to tell you this, this is obvious if you spend any amount of time in either the suburbs or urban areas.  The unfortunate part is that the media and Urban Land Institute study co-authors have position their report as a contest between the cities and suburbs. This obscures the nuances that exist within each the environments.  It is the nuances that tell the real story.

No comments:

Post a Comment