Has the Great Recession begun to reshape Phoenix residential growth?

From 2002 through 2005, the number of new housing starts in Arizona was greater each year than the previous year – with a peak of over 78,000 new homes beginning construction in 2005.  In 2005, the boom ended, with less than 54,000 housing starts – a drop of almost 21% from the year before, and large drops each year after, until a low of less than 11,000 starts in 2011.  2012 saw the beginning of an upsurge again, with over 16,000 new starts.

Arizona, and particularly its urban center of Phoenix, is well-known for this type of boom-and-bust cycle.  However, behind these numbers are some more complex questions:  In the boom period, where in the Phoenix urban area did growth occur?  Was this pattern driven by the preferences of home-buyers, or by zoning and developers?

Urban theorists have argued for many years that the fringe development and sprawl seen in metro Phoenix is at least in part due to the economics of development rather than where homebuyers would prefer to live.  Surveys and discussion suggest that since the housing and financial collapse of 2006, there has been an increasing preference for mixed urbanism as compared to the homogeneous fringe development that characterizes the Phoenix area. 

Is there evidence that a structural change truly is occurring in Phoenix?  Do housing starts since 2006 show a shift towards greater proximity to urban centers, potential for walkability, and mixed land-use settings?

Kevin Kane, a geography doctoral student, determined to look into these questions, with mentoring and support from Abigail York, a professor in ASU’s School of Human Evolution and Social Change, with whom Kane works as a research associate.  A recently-published article in the journal Urban Geography describes their investigation and findings.

Focusing on the city of Phoenix, and using parcel data to allow for fine-grained analysis, the investigation applied theory from economic geography and used the analytic approach of a logistic regression model.

The objective was to identify factors that most impacted single-family residential development – first, during the 2002-2006 real estate boom, and second, during the 2006-2012 global recession -- and to assess whether the drivers of development changed from the first period to the second.

The analysis considered many characteristics of parcels in Phoenix to determine what increased their likelihood of developing into new housing, including the parcel’s previous land use, the type of land uses within the immediate neighborhood, property value, proximity to the light rail and Phoenix’s “subcenters,” as well as demographic characteristics.

 “Some results were very striking,” commented Kane. 

First, the research supported the notion that in the boom period, lower-cost, lower-density areas far from Phoenix’s central business district and subcenters were most likely to develop.  In addition, the research confirmed that development trajectories changed dramatically during the recession.  The number of parcels converted to residential use decreased drastically – from almost 44,000 during the boom to less than 15,000 during the recession, and the previous use of lands that became residential in each period also changed radically.  During the boom, conversions to single-family homes took place on a blend of agricultural land (36% of parcels), vacant land (58%), and land previously in other uses (5%).  During the recession, the percent of conversions from agricultural land dropped steeply to 5%, and the vast majority of conversions – 88% -- occurred on land that had been vacant in 2006.

Other results were more nuanced, but several trends appeared, with some support for the idea that the recession is the start of a long-term downturn in fringe development and upturn in compact growth.  During the recession, the odds of new single family homes being near commercial and public lands increased substantially.   In addition, developers appeared more willing to develop more expensive parcels with more expensive neighbors – as compared with the lower-cost, lower-density fringe locations popular during the boom.   Together, these results suggested a move towards the urbanist vision of mixed-use, centrally-located neighborhoods.

Another surprising result was that zoning designation is particularly incongruous with actual land use.  For example, single-family residential parcels and single-family residential zoning “matched up” less than 60% of the time.  This correlation statistic is lower for multifamily, commercial, industrial, and agricultural categories. Agricultural zoning and agricultural land use were congruous less than 7% of the time in 2006.  However, with the exception of agriculture, these correlations do not decrease during the bust. Kane and his colleagues took this as a reassuring indication that “properly constructed municipal land-use institutions might be effective mechanisms for more equitable land use even in changing economic conditions.”

Beyond looking at Phoenix specifically, this study aimed to offer an empirical model – a testable method – to evaluate the assertion that urban form both affects economic conditions and is affected by them.  The study did offer a test of this kind, and suggests a path to related research. 

 “While establishing statistical causality in a parcel-level, urban change analysis is particularly tricky, what we’re trying to do here is understand the fine-grained spatiality of the recession: namely, is there a shifting locus to development?" Kane explained.

"A lot of work has been done on how the causes of the recession were by their nature place-specific – things like speculation, mortgage deregulation, and then foreclosures.  Beyond that, though, development in Phoenix is dominated by single-family housing and is very path-dependent - in other words, the changes in patterns we see over this recessionary period will likely be with us for a long time.”


Read the complete paper here:  Kevin Kane, Abigail M. York, Joseph Tuccillo, Lauren E. Gentile & Yun Ouyang (2014): Residential development during the Great Recession: a shifting focus in Phoenix, ArizonaUrban Geography, Volume 35, Issue 4.