Residential Construction Trends in
America's Metropolitan Regions
www.epa.gov/smartgrowth
-------
United States
Environmental Protection
Agency
Residential Construction Trends in America's
Metropolitan Regions
John V. Thomas, Ph.D.
Development, Community, and Environment Division
U.S. Environmental Protection Agency
January, 2009
-------
Acknowledgements
This working paper grew out of the effort to update another EPA publication, Our Built
and Natural Environments (EPA 231 -R-01 -002). It was reviewed by Rolf Pendall, from
Cornell University, Chris Nelson from the University of Utah, and John Carruthers from
the Office of Policy Development and Research at the U.S. Department of Housing and
Urban Development. In EPA's Development, Community, and Environment Division,
Mara D'Angelo, Stephanie Bertaina, and Rachel Friedman each played a role in
compiling the dataset. I would also like to thank David Goldberg from Smart Growth
America for posing the question that motivated this detailed examination of metropolitan
construction trends.
Cover Images
Atlantic Station and Midtown Atlanta, GA - Public domain image downloaded from
http: //en. wikipedia. org/wiki/Image: Midtownatl j pg
Lakelands, Gaithersburg, MD - Courtesy Lee Sobel
Colorado Springs, CO - Public domain image, Courtesy David Shankbone
11
-------
Table of Contents
Introduction 1
Assembling the Data 2
Urban Infill and Smart Growth Not Captured by This Analysis 3
Other Limitations of the Analysis 3
Central City Trends 4
Core Suburban Community Trends 10
Key Regions 12
The National Context 20
Insights and Suggested Research Questions 21
References Cited 23
Appendix A - Detailed Summary Table 24
Appendix B - Core Suburban Community Definitions 25
Figures and Tables
Figure 1 - Central City Number of New Residential Construction Permits 5
Figure 2 - Central City Share of Region's New Residential Construction Permits 5
Figure 3 - Central City Share of Residential Construction (Increase / Significant Share) 6
Figure 4 - Central City Share of Residential Construction (Increase / Smaller Share) 7
Figure 5 - Central City Share of Residential Construction (Minimal Change/ Decrease) 7
Table 1 - Central City Share of Metropolitan Residential Building Permits 8
Table la - Difficult to Distinguish Redevelopment from Construction on Greenfield Sites 9
Table 2 - Core Suburban Communities' Share of Residential Building Permits 10
Table 3 - Central City + Core Suburban Community Share 11
Table 4 - Housing Starts by Unit Type - National Total 2001-2007 20
Table 5 - Share by Unit Type 20
in
-------
Introduction
Across the country, many urban neighborhoods are experiencing dramatic
transformations. Parking lots, underutilized commercial properties, and former industrial
sites are being replaced with condos, apartments, and townhouses. In spite of the many
impressive projects, a central question remains: Do such examples add up to a
fundamental shift in the geography of residential construction?
To answer this question, US Census residential building permit data for the 50 largest
metropolitan regions was examined over an 18 year period (1990 to 2007). Specifically,
the amount of permits issued by central cities and core suburban communities was
compared to the amount issued by suburban and exurban communities. The main goal
was to clarify: 1) if there has been a shift toward redevelopment; and 2) in which regions
the shift has been most significant.
The permit data showed that, in several regions, there has been a dramatic increase in the
share of new construction built in central cities and older suburbs. Specifically, in
roughly half of the metropolitan areas examined, urban core communities dramatically
increased their share of new residential building permits. For example:
• In fifteen regions, the central city more than doubled its share of permits.
o In the early 1990's, New York City issued 15 percent of the residential
building permits in the region. Over the past six years it has averaged
44 percent.
o The City of Chicago saw its share of regional permits rise from 7 to 23
percent over the same period.
o Portland, Oregon went from 9 to 22 percent.
o Atlanta, Georgia went from 4 to 13 percent.
• The increase has been particularly dramatic over the past five years.
• Data from 2007 show the shift inward continuing in the wake of the real estate
market downturn.
This acceleration of residential construction in urban neighborhoods reflects a
fundamental shift in the real estate market. Lower crime rates in central cities and
changing demographics are often cited as forces driving this change. The increased
demand for homes in walkable communities close to high-paying jobs has also been
documented by a number of studies (Leinberger 2007, Nelson 2007, ULI 2006). For
example, the 2007 edition of the annual Emerging Trends in Real Estate report singles
out infill and mixed-use development as "best bets":
"Energy costs add fuel to the fire—people want greater convenience in their time-constrained
lives. Far-flung greenfield homes may cost less, but filling the gas tank burns holes in wallets.
Both empty nesters and their young adult offspring gravitate to live in more exciting and
sophisticated 24-hour places—whether urban or suburban—with pedestrian-accessible retail,
restaurants, parks, supermarkets, and offices. Transit-oriented development at subway or light-rail
stations almost cannot miss." (ULI 2006, p. 14)
-------
However, even with solid economic fundamentals, many large-scale redevelopment
projects still require changes in local regulations or public infrastructure investments to
be successful. For example, transit-oriented development often requires updates to
zoning codes, more flexible parking regulations, assistance with land assembly, or
improvements to upgrade water, sewer and local streets (TCRP 2004). Brownfield
properties often need assistance to evaluate contamination and potentially clean up soil
and groundwater.
The clear trend toward more redevelopment has a couple key implications for smart
growth. First, regions often cited as leaders in promoting growth management and
redevelopment (Portland, Denver, Sacramento and Atlanta) are among the medium sized
cities where the shift inward has been most dramatic. Second, in metropolitan regions
with large and diverse central cities with strong ties to the global economy (New York,
Chicago, Boston, Miami, Los Angeles) the market fundamentals are shifting toward
redevelopment even in the absence of formal policies and programs at the regional level.
The following sections of this report cover the trends in more detail. First, a brief
description is provided to clarify how the data was organized, the types of redevelopment
included, and other significant limitations of the analysis. Next, the trends for central
cities and core suburban communities across the 50 regions are described and
summarized in a set of tables. Tables and charts with sub-regional detail are also
provided for the seven regions with the strongest shift toward redevelopment. Finally,
the trends are placed in a national context and key future research questions are
identified.
Assembling the Data
The first step in answering the basic question of how much residential development
might be shifting inward was to assemble Census Bureau residential building permit data
for the 50 largest U.S. metropolitan regions.1 County-level summary files provided totals
for suburban counties.2 However, since many urban core counties include both
developed and undeveloped land, it was important to reach below the county level.
Therefore, the "permit issuing place" files were organized by region to assemble permit
data for each individual jurisdiction within urban core counties. Two kinds of
jurisdictions were of particular importance: central cities and urban core suburbs.
The latter group is important since many larger metropolitan regions do have suburban
communities that are essentially built out. Therefore, increased construction activity in
these places primarily consists of redevelopment. Two criteria were used to identify such
communities: 1) the land area of the jurisdiction did not significantly increase between
1 Annual summary files for 1990 through 2006 were provided by the Census Manufacturing and
Construction Division covering building permits for new residential units.
http://censtats.census.gov^ldg/bldgprmt.shtml
2 The December 2006 definition of Metropolitan Statistical Areas was used as the basis for deciding which
counties were associated with a particular region.
-------
the 1990 and 2000 Censuses3; and 2) the community was within 5 miles of the central
city or within a clear regional boundary, such as a beltway interstate, separating
expanding suburbs from hemmed-in urban core suburbs.4
Urban Infill and Smart Growth Not Captured by This Definition
In this analysis, urban core places were defined in a way that excludes some types of
redevelopment. Since the Census data are provided at the jurisdiction level, it is not
possible to determine where in a permit-issuing city or county the residential units are
being built. Therefore, communities in which development is taking place on both
undeveloped and previously developed land are grouped into the expanding suburb
category. As a result, regional shares reported in the tables and charts below
underestimate the level of infill-oriented residential construction that is actually taking
place in many regions.
For example, Montgomery County, Maryland, a county with nearly one million people, is
a single building permit issuing jurisdiction in this dataset. However, residential building
permits issued by the county include high-rise apartments and condos near Metrorail
stations, as well as detached single-family homes built on exurban farmland. Since there
was no way to make such distinctions in this dataset, Montgomery County was classified
as an expanding suburban community. In other regions suburban cities such as Pleasant
Hill, California, are also expanding onto vacant land as well as issuing permits for infill
development near major rail transit facilities.
In 13 of the largest metro areas, it is difficult to make any distinctions between
redevelopment and suburban expansion with this dataset. In some cases, the central city
has annexed substantial amounts of undeveloped land. In other cases, the central city is
part of a consolidated city/county government and does not separately report building
permits issued within the core urban area from those issued in rural areas.
Finally, building permits associated with transit-oriented neighborhoods developed on
greenfield5 sites are categorized as construction in expanding suburban areas. Therefore,
although major development projects such as Orenco Station in Hillsboro, Oregon, and
King Farm in Gaithersburg, Maryland, are often considered examples of smart growth,
they are not counted as urban core development in this analysis.
Other Limitations of the Analysis
The geographic distribution of commercial development was also outside the scope of
this analysis. There are reasons to expect that office development in many regions would
3 Significant expansion is an indicator of annexation of undeveloped land.
4 In most cases, an inner "beltway" freeway or a group of key intersecting freeways separated built-out
urban core suburbs from expanding suburbs. See appendix for the boundaries used for each region.
5 The term "greenfield" means land that was previously undeveloped.
-------
be more concentrated than residential development. Retail patterns would be more
complex, but probably follow residential trends. Manufacturing, wholesale, and
distribution center development, on the other hand, will tend to be more dispersed than
residential development in most regions. However, since the Census stopped gathering
commercial building permit data in 1995, such analysis would require an entirely
different data source, such as the zip code business patterns data or employment data
from a private data provider.
Finally, an increase in residential construction in urban core neighborhoods translates
only indirectly into increased density. Invariably, some shares of the permits are simply
replacing old housing units with new units at similar density. This is most likely a small
share of the permits in central cities, but it might be a significant share in some suburban
communities where older single-family homes are torn down and replaced with larger
single-family homes.
Central City Trends
Across the 50 largest metropolitan regions, the increased amount of new residential
development taking place in many central cities is striking. Given the fluctuations in
building activity from year to year, examining total building permits presents only part of
the story. Figure 1 shows the total number of building permits for five major cities with
rail transit systems. Figure 2 shows each city's share of all building permits issued in its
region. While the first chart shows the raw number of permits increasing, the trends are
stronger when viewed in a regional context. For example, even though the number of
permits in Miami, Washington, D.C., and Philadelphia declined after the real estate
market began to cool in 2005, they fell less dramatically than new construction in
suburban and exurban jurisdictions.
-------
Figure 1
Central City Number of New Residential Construction Permits
12,000 -,
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Figure 2
Central City Share of Region's New Residential Construction Permits
25% -,
I 20% -
.a 15% -
10% -
5% -
0%
\
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
-------
Looking at the average share over multiple years also helps to clarify the nature of the
trends. Comparing the early 1990s to the early 2000s is another way to look beyond
some of the variation from year to year. In 26 cities, the share has doubled or tripled
since 2000. In many cases, 2006 also represented the highest annual share over the past
17 years (1990 to 2006). Generally, cities can be grouped into four categories:
• Saw a substantial increase, and account for a significant share of new
construction in the region (Figure 3).
• Saw a substantial increase, but still account for a modest share of new
construction (Figure 4).
• Small changes or declines in the central city share of regional construction
(Figure 5).
• Trend is unclear due to central city expansion or consolidated city / county
government (Table la).
Figure 3
Central City Share of Residential Construction
(Substantial increase and a significant share of regional construciton)
50% -
«
•g
H 40% -
| 30% -
aj
g
4_
o
g 20% -
10% -
0%
Q Average 1990-95
• Average 2002-2007
D2007
./
-------
Figure 4
Central City Share of Residential Construction
(Substantial increase, but less than a fifth of regional permits)
60% -,
a Average 1990-95
• Average 2002-2007
D2007
60% -|
Figure 5
Central City Share of Residential Construction
(Minimal change or a decreased share)
50% -
£ 40% -
| 30% -
1 20% -
10% -
H Average 1990-95
• Average 2002-2007
• 2007
\P
sJT
«^
-------
Table 1 - Central City Share of Metropolitan Residential Building Permits
Average Share
1990-95 2002-07 2007
New York City* 15% 44% 55%
Los Angeles 19% 31% 37%
Los Angeles** 11% 14% 21%
Chicago 7% 23% 40%
Dallas 13% 10% 13%
Ft. Worth 5% 16% 16%
Philadelphia 3% 11% 15%
Miami 2% 16% 22%
Washington D.C. 1% 6% 9%
Atlanta 4% 13% 21%
Boston 2% 10% 9%
Detroit 2% 4% 7%
San Francisco 5% 9% 14%
Oakland 3% 6% 6%
San Jose 11% 12% 11%
Seattle 11% 18% 26%
Minneapolis 2% 8% 9%
St. Paul 1% 4% 3%
San Diego 42% 37% 51%
St. Louis 1% 5% 5%
Baltimore 2% 7% 5%
Denver 5% 19% 25%
Pittsburgh 3% 3% 3%
Portland, OR 9% 22% 33%
Cincinnati 4% 6% 7%
Cleveland 4% 6% 6%
Sacramento, CA 9% 23% 25%
Kansas City, MO 12% 19% 13%
Providence, RI 2% 4% 7%
Milwaukee 6% 16% 25%
Richmond, VA 2% 7% 8%
Hartford, CT 4% 5% 2%
Buffalo, NY 6% 4% 7%
Birmingham, AL 5% 10% 9%
Salt Lake City 6% 7% 10%
Rochester, NY 3% 4% 5%
* Manhattan, Brooklyn, Queens, Bronx Boroughs only - excludes Staten Island
** Share with Riverside and San Bernardino Counties included in regional definition
-------
Table 1 a - Difficult to Distinguish Redevelopment From
Construction on Greenfield Sites
Average Share
1990-95 2002-07 2007
Houston* 20% 23% 24%
Phoenix* 28% 25% 34%
Orlando, FL * 9% 13% 11%
San Antonio* 63% 63% 55%
Columbus, OH* 38% 39% 42%
Austin, TX* 48% 36% 39%
Las Vegas* 40% 18% 12%
Tampa, FL* 8% 14% 28%
Raleigh, NC * 30% 38% 36%
Oklahoma City ** 43% 47% 46%
Nashville, TN *** 26% 31% 38%
Jacksonville, FL *** 58% 54% 57%
Memphis, TN *** 66% 55% 54%
Louisville, KY *** 60% 56% 59%
Indianapolis*** 36% 28% 24%
Charlotte, NC *** 60% 53% 54%
* Land area increased substantially in the 1990s.
** Substantial undeveloped land within city boundaries.
*** Consolidated city/county government.
-------
Core Suburban Community Trends
Urban redevelopment extends beyond the boundaries of major central cities. Many older
suburbs near central cities have been built out for some time, and new residential units
are almost entirely built upon previously developed sites. When these communities are
added to the mix, redevelopment's share changes significantly in a few regions.
Specifically, in eight metropolitan areas, urban core suburbs have significantly increased
their share of regional housing starts.
Table 2 - Core Suburban Communities' Share of Residential Building Permits
Average Share
1990-95 2002-07 2007
New York MSA
Core Subuibs in Bergen County, NJ 4% 6% 4%
Washington, B.C. MSA
Arlington County, VA 2% 4% 9%
Alexandria, VA 1% 2% 2%
Boston MSA
Core Suburbs in Middlesex County,
MA 6% 10% 12%
Miami MSA
Core Suburbs in Broward County, FL 5% 22% 26%
San Francisco / San Jose CMSA
Core Suburbs in Alameda, Contra
Costa, and San Mateo Counties, CA 6% 8% 7%
Core Suburbs in Santa Clara County,
CA 4% 5% 7%
San Biego MSA
Core Suburbs in San Diego County,
CA 2% 3% 6%
Minneapolis MSA
Core Suburbs in Hennepin County,
MN 1% 3% 4%
The method for identifying these communities was described above. However, Figures
6-10 provide a visual illustration of urban core suburbs in three regions. The table in
Appendix B also provides definitions for each region.
10
-------
Table 3 - Central City + Core Suburban Community Share
Average Share
1990-1995 2002-07 2007
New York-Northern New Jersey-Long Island, NY-NJ-PA 18% 50% 59%
San Francisco-Oakland-San Jose, CA** 29% 39% 45%
San Francisco-Oakland-Fremont, CA * 14% 29% 35%
San Jose, CA* 66% 74% 76%
Miami-Fort Lauderdale-Pompano Beach, FL 15% 35% 43%
San Diego-Carlsbad-San Marcos, CA 44% 40% 57%
Dallas-Fort Worth-Arlington, TX 30% 30% 33%
Sacramento-Arden-Arcade-Roseville, CA 9% 23% 25%
Kansas City, MO-KS 14% 22% 18%
Los Angeles-Santa Ana-Riverside-San Bernardino ** 23% 24% 32%
Los Angeles-Long Beach-Santa Ana, CA * 35% 48% 54%
Riverside-San Bernardino-Ontario, CA * 6% 5% 6%
CWcago-Napemlle-Joliet, IL-IN-WI 11% 29% 45%
Boston-Cambridge-Quincy, MA-NH 8% 20% 22%
Seattle-Tacoma-Bellevue, WA 13% 21% 29%
Portland-Vancouver-Beaverton, OR-WA 13% 26% 36%
Washington-Arlington-Alexandria, DC-VA-MD-WV 4% 13% 20%
Minneapolis-St. Paul-Bloomington, MN-WI 7% 16% 18%
Denver-Aurora, CO 5% 19% 25%
Hartford-West Hartford-East Hartford, CT 10% 11% 13%
Milwaukee-Waukesha-West Allis, WI 12% 19% 30%
Birmingham-Hoover, AL 5% 10% 9%
Tampa-St. Petersburg-Clearwater, FL 12% 17% 32%
Atlanta-Sandy Springs-Marietta, GA 4% 14% 21%
Providence-New Bedford-Fall River, RI-MA 9% 10% 11%
Virginia Beach-Norfolk-Newport News, VA-NC 12% 36% 36%
Philadelphia-Camden-Wilmington, PA-NJ-DE-MD 3% 11% 15%
Cleveland-Elyria-Mentor, OH 8% 11% 10%
Cincinnati-Middletown, OH-KY-IN 7% 6% 7%
Buffalo-Niagara Falls, NY 20% 19% 24%
Richmond, VA 2% 7% 8%
Baltimore-Towson, MD 2% 7% 5%
Detroit-Warren-Livonia, MI 4% 5% 9%
Salt Lake City, UT 6% 7% 10%
St. Louis, MO-IL 1% 5% 5%
Rochester, NY 3% 4% 5%
Pittsburgh, PA 3% 3% 3%
* Share with MSA defined as Los Angeles and Orange Counties
** Share with larger CMS A definition. Includes Los Angeles, Orange, San Bernardino, Riverside and
Ventura Counties.
11
-------
Key Regions
New York-Northern New Jersey-Long Island MSA
Average Share
1990-95 2002-07
Core Boroughs (without Staten Island)
Manhattan Borough
Queens Borough
Brooklyn Borough
Bronx Borough
1st Tier Suburban Counties (1)
NJ cities within 5 miles of Manhattan
Urban Fringe Counties (2)
15%
4%
2%
4%
4%
16%
4%
69%
44%
12%
10%
13%
7%
21%
6%
35%
2007
55%
14%
12%
15%
8%
19%
4%
26%
80% -i
70% -
•3 60% -
"o
I 50% -
} 40%
«*•
| 30% H
I
& 20% -
10% -
0%
Share of New Housing Starts by County
(New York MSA)
New York Boroughs except Staten Island
1st Tier Suburban Counties (1)
NJ Cities within 5 Miles of Manhattan
Urban Fringe Counties (2)
Source: New Residential Building Permits, U.S. Census Bureau, Manufacturing and Construction Division
(1) 1st Tier Suburban Counties - Nassau and Richmond Counties, NY; Essex, Union, Bergen, and Hudson
Counties, NJ.
(2) Urban Fringe Counties - Rockland, Westchester, Putnam, and Suffolk Counties, NY; Middlesex,
Monmouth, Ocean, Hunterdon, Morris, Sussex, and Passaic Counties, NJ; Pike County, PA.
12
-------
New Housing Starts by County
(New York MSA)
10,000 -,
9,000 -
8,000 -
7,000 -
6,000 -
5,000 -
4,000 -
3,000 -
2,000 -
1,000 -
0 -
•Manhattan borough ^^~ Queens borough
Brooklyn borough —•— Bronx borough
eft
>
Source: New Residential Building Permits, U.S. Census Bureau, Manufacturing and Construction Division
New Housing Starts by County
(New York MSA)
6,000 -,
£ 5,000
e
o
? 4,000
o 3,000
a 2,000
cN CxV eft ol* cb oi° 0\ eft C?> C?> C\V C\V eft CN^ C\^ pi3 CN\
^ ^ ^ $ $ $ $ $ $ N^V ^ 4* <$ ^ ^ ^ ^
Source: New Residential Building Permits, U.S. Census Bureau, Manufacturing and Construction Division
13
-------
Chicago-Naperville-Joliet MSA
Cook County
Chicago
Core suburban cities
1st Tier Suburban Counties (1)
Urban Fringe Counties (2)
Average Share
1990-95 2002-07
25%
7%
4%
37%
38%
37%
23%
6%
33%
30%
2007
50%
40%
5%
23%
26%
Share of New Housing Starts by County
(Chicago MSA)
80% -,
70% -
•a
"o
I 50% H
I 40% -
«*•
| 30% H
& 20% -
10% -
0%
•Chicago
•1st Tier Suburban Counties
Urban Fringe Counties
Source: New Residential Building Permits, U.S. Census Bureau, Manufacturing and Construction Division
(1) 1st Tier Suburban Counties - DuPage, Kane, Lake, and Will Counties, IL.
(2) Urban Fringe Counties - DeKalb, Grundy, Kendall, and McHenry Counties, IN; Japser, Lake, Newton,
and Porter Counties, WI; Kenosha.
14
-------
Atlanta-Sandy Springs-Marietta MSA
Fulton County
Atlanta
1st Tier Suburban Counties (1)
Urban Fringe Counties (2)
Average Share
1990-95 2002-07
17% 22%
4% 13%
45% 37%
38% 41%
2007
28%
21%
32%
39%
Share of New Housing Starts by County
(Atlanta MSA)
50% -,
45% -
40% -
1 35% -
H
1 30% -
o
'Q
£ 25% -
° 20% -
8
£ 15% -
-
10% -
5% -
•Atlanta
•1st Tier Suburban Counties (1)
Urban Fringe Counties (2)
Source: New Residential Building Permits, U.S. Census Bureau, Manufacturing and Construction Division
(1) 1st Tier Suburban Counties - DeKalb, Cobb, Clayton, Douglas, and Gwinnett Counties, GA.
(2) Urban Fringe Counties - Barrow, Bartow, Butts, Carrol, Cherokee, Coweta, Dawson, Fayette, Forsyth,
Haralson, Heard, Henry, Jasper, Lamar, Merriwether, Newton, Paulding, Pickens, Pike, Rockdale,
Spaulding, and Walton Counties, GA.
15
-------
Washington-Arlington-Alexandria MSA
DC, Arlington, Alexandria
Washington, D.C.
Arlington County
City of Alexandria
1st Tier Suburban Counties (1)
Urban Fringe Counties (2)
Average Share
1990-95 2002-07
2007
50%
45%
40%
_ 35%
^ 30%
a
o
I 25%
1 20%
1 15%
10%
5%
0%
Share of New Housing Starts by County
(Washington, DC MSA)
•DC, Arlington County, City of Alexandria
1st Tier Suburban Counties (1)
Urban Fringe Counties (2)
Source: New Residential Building Permits, U.S. Census Bureau, Manufacturing and Construction Division
(1) 1st Tier Suburban Counties - Montgomery and Prince Georges Counties, MD; Fairfax County, VA.
(2) Urban Fringe Counties - Calvert, Charles, and Frederick Counties, MD; Fauquier, Loudoun, Prince
William, Spotsylvania, and Stafford Counties, VA.
16
-------
Denver-Aurora MSA
City and County of Denver
1st Tier Suburban Counties (1)
Urban Fringe Counties (2)
Average Share
1990-95 2002-07 2007
5% 19% 25%
72% 69% 61%
22% 13% 15%
80% -,
70% -
60% -
I 50% -\
O
*5JD
s
n 40% -
$*-
| 30% -
& 20% -
10% -
0%
Share of New Housing Starts by County
(Denver MSA)
•City of Denver
•1st Tier Suburban Counties
Urban Fringe Counties
cCV
Source: New Residential Building Permits, U.S. Census Bureau, Manufacturing and Construction Division
(1) 1st Tier Suburban Counties - Adams, Arapahoe, Douglas, and Jefferson Counties, CO.
(2) Urban Fringe Counties - Boulder, Broomfield, Clear Creek, Elbert, Gilpin, and Park Counties, CO.
17
-------
Portland-Vancouver-Beaverton MSA
Multnomah County
Portland
Gresham
Suburban Counties (1)
Average Share
1990-95 2002-07
17% 28%
9% 22%
4% 4%
76% 72%
2007
36%
33%
3%
64%
Share of New Housing Starts by County
(Portland MSA)
80% -,
70% -
_ 60% -
1
£ 50% -
c
I 40% -
«*•
o
1 30% -
I
* 20% -
10% -
0% -
•City of Portland
•1st Tier Suburban Counties (1)
Urban Fringe Counties (2)
Source: New Residential Building Permits, U.S. Census Bureau, Manufacturing and Construction Division
(1) 1st Tier Suburban Counties - Clackamas and Washington Counties, OR; Clark County, WA.
(2) Urban Fringe Counties - Columbia and Yamhill Counties, OR; Skamania County, WA.
18
-------
Sacramento-Arden-Arcade-Roseville MSA
Sacramento County
City of Sacramento
Suburban Counties
Average Share
1990-95 2002-07
56% 56%
9% 23%
2007
51%
25%
Share of New Housing Starts by County
(Sacramento MSA)
80% -,
70% -
_ 60% -
I
-a 50% -
|
| 40% -
1 30% -
U
fc 20% -
10% -
0% -
•City of Sacramento
Rest of Sacramento County
Suburban Counties (1)
Source: New Residential Building Permits, U.S. Census Bureau, Manufacturing and Construction Division
(1) Suburban Counties - El Dorado, Placer, and Yolo Counties, CA.
19
-------
The National Context
Examining the national trends over the past few years helps place these regional trends in
context. Between 2001 and 2005, the number of residential units built each year grew
dramatically across all categories and regions. In 2006 and 2007, there was a sharp
decline. However, it has been uneven across the housing market:
• Single family units have declined most rapidly, while the construction of
multifamily units has fallen more modestly.
o The number of new high-density residential units has not declined from
the 200,000 units per year level produced at the height of the real estate
boom.
o Construction of rental units is actually up slightly in 2007, while condos
have declined at a rate similar to single-family detached units.
Table 4 Housing Starts by Unit Type - National Total 2001-2007 (in Thousands)
Year Total Single Family
Multifamily
Detached Attached
2001
2002
2003
2004
2005
2006
2007
1,602
1,705
1,848
1,956
2,068
1,801
1,355
1,133
1,198
1,309
1,397
1,494
1,264
901
140
160
190
213
222
201
145
Total
Multifamily
Units
329
346
349
345
352
336
309
For Sale
Units
71
71
87
120
150
151
115
Rental
Units
258
275
262
225
203
185
194
Units in
Large
Buildings
(20+ units)
178
183
196
192
208
206
205
Year
2001
2002
2003
2004
2005
2006
2007
Detached
Single
Family
71%
70%
71%
71%
72%
70%
66%
Table 5 Share by Unit Type
Townhouses Condos
9%
9%
10%
11%
11%
11%
11%
4%
4%
5%
6%
7%
8%
8%
Rental
Apartments
16%
16%
14%
12%
10%
10%
14%
Large
Multifamily
Buildings
11%
11%
11%
10%
10%
11%
15%
Source: U.S. Census Bureau, Residential Construction Branch, Table Ql "New Privately Owned
Housing Units Started in the United States by Purpose and Design."
20
-------
Residential Housing Starts by Census Region
1,200 -,
1,000 -
800 -
600 -
400 -
200 -
2001
2002
2003
2004
Year
2005
2006
2007
Insights and Suggested Research Questions
While these trends reveal a substantial shift in residential construction patterns, they also
suggest that the change is not yet reshaping the face of urban America as a whole. A
large share of new residential construction still takes place on previously undeveloped
land at the urban fringe. In some regions there has been little change in the share of new
construction taking place in central cities. In other regions, central cities have increased
their relative share of building permits, but still account for a small overall share at the
regional level. Although urban core neighborhoods have doubled or tripled their share of
residential construction since the early 1990s, they still account for less than half of all
new residential units in most regions. The "urban infill" share would be larger if
redevelopment in growing suburbs was also considered, but it would still not likely
represent a majority of new construction in more than a handful of regions.
Additionally, evaluating residential construction based on the Census building permit
data provides less geographic detail than could be achieved by studying a single region.
Previous studies have examined patterns within particular regions (Knaap Song 2004).
With the increased availability of GIS-based parcel data in many regions, it is possible to
evaluate residential construction patterns within jurisdictions and answer more precise
questions, such as:
• What percent of residential units are being built upon previously developed
parcels, and how has that share changed over time?
• How much has average residential density increased in various regions?
• What percent of new housing units are being built in walkable / transit-accessible
places?
21
-------
Although this analysis does not directly address these questions, it does provide a broad
picture of the magnitude and direction of residential construction trends across the
country. The results of this analysis raise a set of important research questions for
subsequent work:
• To what extent are these trends driven by real estate market fundamentals versus
public sector policies?
o Land use regulations, infrastructure provision, and incentive programs.
• In regions where urban core communities' share of new construction has
increased, what kinds of projects are driving the trend?
o Transit-oriented development, high-rise buildings in prime waterfront or
downtown locations, redevelopment of former industrial sites,
redevelopment of strip commercial parcels, or large underutilized parking
lots.
• In regions where urban fringe development is still increasing its share, what is
behind such trends?
o Continued decentralization of employment, a weak overall housing
market, deficiencies in urban core infrastructure.
Resolving these questions will provide a more complete picture of the policy implications
of these trends. First, it could further clarify the approaches that most effectively
increased the overall rate of redevelopment. Second, it could also identify specific
policies and strategies that state and local governments can put in place to capitalize on
these trends.
Finally, continued research will also be needed to shed light on the right mix of policies
as we emerge from the current real estate market turmoil. The data suggest that the shift
toward redevelopment continued in 2007 even as the real estate market weakened.
Although the number of building permits in urban core areas slowed, the declines were
more precipitous in outlying areas. However, redevelopment projects are often capital
intensive and constraints on developer's access to credit and cities access to municipal
bonds financing may begin to substantially reduce the pace of redevelopment.
22
-------
References Cited
Nelson, Arthur C. "Where Will Everybody Live?" Working Paper for EPA Publication
The Business Case for Smart Growth.
Leinberger, Christopher B. 2007. The Option ofUrbanism: Investing in a New American
Dream. Island Press.
Song, Yan and Gerrit J. Knaap. 2004. "Measuring Urban Form: Is Portland Winning the
War on Sprawl?" Journal of American Planning Association, Vol. 70, No. 2: 210 - 225.
Transit Cooperative Research Program. 2004. Transit-Oriented Development in the
United States: Experiences, Challenges, and Prospects. Report 102.
Urban Land Institute / PricewaterhouseCoopers. 2006. "Best Bets: 2007" Emerging
Trends in Real Estate 2007.
U.S. Census Bureau, Residential Construction Branch. "Relationship Between Building
Permits, Housing Starts, and Housing Completions." Available at
http://www.census.gov/const/www/nrcdatarelationships.html.
23
-------
Appendix A -Regional Summary Table
Share of Metro Region's New Residential Building Permits
Average Share
New York-Northern New Jersey-Long Island, NY-NJ-PA
Core Boroughs (w/out Staten Island)
Manhattan borough
Queens borough
Brooklyn borough
Bronx borough
Core Suburban Counties
NJ Cities w/in 5 Miles of Manhattan
Fringe Suburban Counties
Los Angeles-Long Beach-Santa Ana, CA
Los Angeles County
Los Angeles
Core Suburbs
Orange County
Urban Core Suburbs
Chicago-Naperville-Joliet, IL-IN-WI
Cook County
Chicago
Core Suburbs
1st Tier Suburban Counties
Urban Fringe Counties
Dallas-Fort Worth-Arlington, TX
Dallas County
City of Dallas
Core Suburbs
Tarrant County
City of Ft Worth
Core Suburbs
Suburban Counties
Philadelphia-Camden-Wilmington, PA-NJ-DE-MD
City and County of Philadelphia
1st Tier Suburban Counties
Urban Fringe Counties
Miami-Fort Lauderdale-Pompano Beach, FL
Dade County
Miami
Core Suburbs
Broward County
Core Suburbs
Palm Beach County
Core Suburbs
1990-95
15%
4%
2%
4%
4%
16%
4%
69%
1990-95
58%
19%
11%
42%
5%
1990-95
18%
5%
3%
27%
28%
1990-95
39%
13%
6%
22%
5%
5%
33%
1990-95
3%
52%
45%
1990-95
33%
2%
5%
36%
5%
31%
3%
2002-07
44%
12%
10%
13%
7%
21%
6%
35%
2002-07
71%
31%
11%
29%
6%
2002-07
37%
23%
6%
33%
30%
2002-07
23%
10%
2%
26%
16%
2%
35%
2002-07
11%
48%
41%
2002-07
49%
16%
4%
22%
12%
29%
3%
2007
55%
14%
12%
15%
8%
19%
4%
26%
2007
72%
37%
10%
28%
8%
2007
50%
40%
5%
23%
26%
2007
26%
13%
3%
23%
16%
1%
35%
2007
15%
47%
38%
2007
53%
22%
4%
26%
15%
21%
2%
24
-------
Share of Metro Region's New Residential Building Permits
Average Share
Washington-Arlington-Alexandria, DC-VA-MD-WV
DC, Arlington, Alexandria
Washington DC
Arlington County
City of Alexandria
1st Tier Suburban Counties (1)
Urban Fringe Counties (2)
Atlanta-Sandy Springs-Marietta, GA
Fulton County
Atlanta
1st Tier Suburban Counties (1)
Urban Fringe Counties (2)
Boston-Cambridge-Quincy, MA-NH
Suffolk County
Boston
Middlesex County
Core Suburbs
1st Tier Suburban Counties
Urban Fringe Counties
Detroit-Warren-Livonia, Ml
Wayne County
Detroit
Core Suburbs
1st Tier Suburban Counties
Urban Fringe Counties
San Francisco-Oakland-Fremont, CA
San Francisco (City / County)
San Mateo County
Core Suburbs
Alameda County
Oakland
Core Suburbs
Contra Costa County
Core Suburbs
Urban Fringe Counties
Riverside-San Bernardino-Ontario, CA
Riverside County
City of Riverside
San Bernardino County
City of San Bernardino
1990-95
4%
1%
2%
1%
52%
44%
1990-95
17%
4%
45%
38%
1990-95
3%
2%
31%
6%
50%
16%
1990-95
21%
2%
1%
62%
18%
1990-95
6%
7%
4%
21%
3%
1%
29%
2%
37%
1990-95
58%
4%
42%
3%
2002-07
13%
6%
4%
2%
32%
55%
2002-07
22%
13%
37%
41%
2002-07
11%
10%
29%
10%
44%
16%
2002-07
27%
4%
1%
56%
17%
2002-07
11%
6%
4%
24%
8%
3%
29%
2%
29%
2002-07
67%
4%
33%
1%
2007
20%
9%
9%
2%
32%
48%
2007
28%
21%
32%
39%
2007
10%
9%
38%
12%
39%
13%
2007
25%
7%
2%
59%
16%
2007
19%
6%
4%
24%
7%
3%
28%
2%
24%
2007
61%
5%
39%
1%
25
-------
Share of Metro Region's New Residential Building Permits
Average Share
Seattle-Tacoma-Bellevue, WA
King County
Seattle
Core Suburbs
Pierce County
Snohomish County
Minneapolis-St. Paul-Bloomington, MN-WI
Hennepin County
Minneapolis
Core Suburbs
Ramsey County
St. Paul
Core Suburbs
1st Tier Suburban Counties
Urban Fringe Counties
San Diego-Carlsbad-San Marcos, CA
City of San Diego
Core Suburbs
Expanding Suburban Cities
Unincorporated San Diego County
St. Louis, MO-IL
City of St. Louis
1st Tier Suburban Counties
Urban Fringe Counties
Tampa-St. Petersburg-Clearwater, FL
Hillsborough County
Tampa
Pinellas County
St. Petersburg
Suburban Counties
Baltimore-Towson, MD
City of Baltimore
1st Tier Suburban Counties
Urban Fringe Counties
Denver-Aurora, CO
City of Denver
1st Tier Suburban Counties
Urban Fringe Counties
Pittsburgh, PA
Allegheny County
Pittsburgh
Suburban Counties
1990-95
48%
11%
2%
27%
25%
1990-95
21%
2%
1%
7%
1%
3%
64%
8%
1990-95
42%
2%
40%
16%
1990-95
1%
78%
21%
1990-95
45%
8%
29%
3%
26%
1990-95
2%
55%
43%
1990-95
5%
72%
22%
1990-95
39%
3%
61%
2002-07
52%
18%
2%
24%
24%
2002-07
24%
8%
3%
7%
4%
2%
59%
10%
2002-07
37%
3%
44%
15%
2002-07
5%
68%
27%
2002-07
53%
14%
11%
3%
37%
2002-07
7%
46%
48%
2002-07
19%
69%
13%
2002-07
39%
3%
61%
2007
61%
26%
3%
20%
19%
2007
26%
9%
4%
6%
3%
1%
59%
9%
2007
51%
6%
29%
14%
2007
5%
69%
26%
2007
61%
28%
9%
4%
30%
2007
5%
52%
49%
2007
25%
61%
15%
2007
38%
3%
62%
26
-------
Share of Metro Region's New Residential Building Permits
Average Share
Portland-Vancouver-Beaverton, OR-WA
Multnomah County
Portland
Gresham
1st Tier Suburban Counties (1)
Urban Fringe Counties (2)
Cincinnati-Middletown, OH-KY-IN
Hamilton County
Cincinnati
Core Suburban Cities
Suburban Counties
Cleveland-Elyria-Mentor, OH
Cuyahoga County
Cleveland
Core Suburban Cities
Suburban Counties
Sacramento-Arden-Arcade-Roseville, CA
Sacramento County
City of Sacramento
Suburban Counties
Kansas City, MO-KS
Wyandotte County
Kansas City, KS
Jackson County
Kansas City, MO
Suburban Counties
San Jose-Sunnyvale-Santa Clara, CA
Santa Clara County
San Jose
Core Suburban Cities
San Benito County
Virginia Beach-Norfolk-Newport News, VA-NC
Core Cities (Norfolk, Portsmouth, Hampton)
1st Tier Suburban Counties
Urban Fringe Counties
Providence-New Bedford-Fall River, RI-MA
Providence County
City of Providence
Core Central Cities
1st Tier Suburban Counties
Urban Fringe Counties
1990-95
17%
9%
4%
76%
6%
1990-95
25%
4%
3%
75%
1990-95
40%
4%
4%
60%
1990-95
56%
9%
44%
1990-95
2%
1%
35%
12%
64%
1990-95
91%
48%
18%
9%
1990-95
35%
52%
13%
1990-95
24%
2%
6%
54%
22%
2002-07
28%
22%
4%
64%
8%
2002-07
17%
6%
1%
83%
2002-07
29%
6%
4%
71%
2002-07
56%
23%
44%
2002-07
4%
3%
39%
19%
57%
2002-07
98%
51%
23%
2%
2002-07
36%
45%
19%
2002-07
26%
4%
6%
52%
23%
2007
36%
33%
3%
55%
9%
2007
19%
7%
1%
81%
2007
26%
6%
4%
74%
2007
51%
25%
49%
2007
5%
5%
35%
13%
60%
2007
99%
46%
30%
1%
2007
36%
44%
20%
2007
24%
7%
4%
52%
24%
27
-------
Appendix B -Core Suburban Community Definitions
MSA
New York-Northern New Jersey-Long Island, NY-
NJ-PA
Los Angeles-Long Beach-Santa Ana, CA
Riverside San Bernardino
Chicago-Naperville-Joliet, IL-IN-WI
Dallas-Fort Worth-Arlington, TX
Philadelphia-Camden-Wilmington, PA-NJ-DE-MD
Houston-Sugar Land-Baytown, TX
Miami-Fort Lauderdale-Pompano Beach, FL
Washington-Arlington-Alexandria, DC-VA-MD-WV
Atlanta-Sandy Springs-Marietta, GA
Boston-Cambridge-Quincy, MA-NH
Detroit-Warren-Livonia, Ml
San Francisco-Oakland-Fremont, CA
San Jose-Sunnyvale-Santa Clara, CA
Phoenix-Mesa-Scottsdale, AZ,
Seattle-Tacoma-Bellevue, WA
Minneapolis-St. Paul-Bloomington, MN-WI
San Diego-Carlsbad-San Marcos, CA
St. Louis, MO-IL
Tampa-St. Petersburg-Clearwater, FL
Baltimore-Towson, MD
Denver-Aurora, CO
Pittsburgh, PA
Portland-Vancouver-Beaverton, OR-WA
Cincinnati-Middletown, OH-KY-IN
Cleveland-Elyria-Mentor, OH
Sacramento-Arden-Arcade-Roseville, CA
Orlando-Kissimmee, FL
San Antonio, TX
Kansas City, MO-KS
Las Vegas-Paradise, NV
Columbus, OH
Indianapolis-Carmel, IN
Virginia Beach-Norfolk-Newport News, VA-NC
Charlotte-Gastonia-Concord, NC-SC
Providence-New Bedford-Fall River, RI-MA
Austin-Round Rock, TX
Milwaukee-Waukesha-West Allis, Wl
Nashville-Davidson-Murfreesboro-Franklin, TN
Jacksonville, FL
Core Suburban Community Definition
(No Change in Land Area 1990-2000, and...)
Bergen County, NJ within 5 miles of Manhattan
LA County West of I-605, South of 1-210 / Hollywood Hills
Orange County South of Imperial Hwy (SR 91) West of Costa Mesa
Freeway (SR 55)
City of Riverside, City of San Bernardino
Inside I-294
Dallas County Inside I-635
Tarrant County, Between Ft. Worth and Dallas Ft Worth Airport or
wtihinin 5 miles of the Ft. Worth Central Business District
None Identified
None Identified
Dade County West of Palmetto Expressway
Broward County West of Florida's Turnpike
Palm Beach County West of I-95
Arlington County and City of Alexandria
Inside I-285
Inside I-95
Within 5 miles of CBD
Alameda County West of East Bay Hills, North of I-238
Contra Costa County West of East Bay Hills
San Mateo County East of I-280 / Foothill Expressway
Santa Clara County West of I-280
None Identified
West of I-405 Loop
Inside I-494 I-694 loop
South of I-8, West of SR-125, North of South Bay Freeway
None Identified
City of Clearwater
None Identified
None Identified
None Identified
None Identified
Inside I-275 E of Hwy 264
North of I-480 West of 1-271 East of the Cleveland Airport
None Identified
None Identified
None Identified
None Identified
None Identified
Inside I-270 Loop
None Identified
Within 5 Miles of Norfolk CBD
None Identified
Within 5 Miles of CBD
None Identified
Inside I-894 Loop
None Identified
None Identified
28
-------
Memphis, TN-MS-AR
MSA
Louisville/Jefferson County, KY-IN
Richmond, VA
Oklahoma City, OK
Hartford-West Hartford-East Hartford, CT
Buffalo-Niagara Falls, NY
Birmingham-Hoover, AL
Salt Lake City, UT
Raleigh-Cary, NC
Rochester, NY
None Identified
Core Suburban Community Definition
(No Change in Land Area 1990-2000, and...)
None Identified
None Identified
None Identified
Within 5 miles of CBD
Inside I-290
None Identified
None Identified
None Identified
None Identified
29
------- |