Authors Authors
DATE
DEPT NAME
PHILADELPHIAFED.ORG | @PHILADELPHIAFED
Manufactured Housing
Communities in Pennsylvania:
The Basics
Manufactured Housing
Communities in Pennsylvania:
The Basics
JUNE 2023
COMMUNITY DEVELOPMENT AND REGIONAL OUTREACH
PHILADELPHIAFED.ORG | @PHILADELPHIAFED
Eileen Divringi, Community Development Research Specialist*
*The views expressed here are those of the author and do not necessarily relect the views of the Federal Reserve Bank of Philadelphia or the Federal Reserve System.
The author thanks Claudia Aiken, Alaina Barca, Stephen Brown, Theresa Dunne, Lance George, Lauren Lambie-Hanson, Rachel Siegel, and Keith Wardrip for their helpful
feedback. Special thanks to Mary Gaiski of the Pennsylvania Manufactured Housing Association for sharing insights and data that greatly enhanced this analysis.
2FEDERAL RESERVE BANK OF PHILADELPHIA
Introduction
1
This style of housing is referred to in policy documents, news media, and academic literature by a wide variety of names, including manufactured
housing communities, mobile home parks, trailer parks, mobile home courts, and various permutations of those words. This report uses manufactured
housing communities (MHCs) as a generic term for this type of land-lease community.
2
For more information, see www.hud.gov/program_oices/housing/rmra/mhs/faqs.
3
Manufactured homes are also distinct from recreational vehicles (RVs) and park model homes, which are generally classiied as motor vehicles.
Although these may be present in some of the MHCs included in this data set, sites that primarily cater to these users or seasonal campers are excluded
from the analysis.
4
This represents an increase from less than one-quarter in 2020 but is comparable with the average share since 2014 (U.S. Census Bureau, 2021).
Land-lease manufactured housing communities (MHCs)
1
are a unique and understudied housing arrangement, in
which groups of manufactured homes are placed on rented
land. These homes are often, although not always, owner-
occupied, and residents typically pay a monthly lot rent
to the landowner in addition to standard housing costs.
This split-tenure model — in which ownership of the home
is separate from the land beneath it — makes this style
of housing a more attainable homeownership option for
many but also increases residents’ vulnerability to inancial
exploitation and displacement (Sullivan, 2018; Aman and
Yarnal, 2010; Genz, 2001).
Although MHCs are often acknowledged as a key segment of
the unsubsidized, or “naturally occurring,” aordable housing
stock (Durst and Sullivan, 2019), relatively little is known
about these communities because of their lack of inclusion
in commonly used public data sets (Sullivan, Makarewicz,
and Rumbach, 2022). Drawing on a novel, rigorously
compiled data set that captures the locations of MHCs in
Pennsylvania, this report provides foundational information
on the size, spatial distribution, and socioeconomic context
of 2,288 communities, home to an estimated 55,900
households, across the state. Findings indicate that MHCs
are located in a range of rural, exurban, and suburban
communities, as well as some lower-density urban areas.
Demand for this style of housing appears particularly high in
the outskirts of large and midsize metropolitan areas, where
MHCs may oer a relatively aordable housing option.
What Are Manufactured Homes and
Manufactured Housing Communities?
In contrast with conventional site-built homes,
manufactured homes are factory-constructed on a chassis,
then transported for installation on a site. Manufactured
homes are subject to the Manufactured Home Construction
and Safety Standards (HUD Code) enacted in 1976
2
and are
technically distinct from mobile homes, which are factory-
built housing units constructed before the implementation
of the HUD Code. However, for brevity, the remainder of this
report will use manufactured homes as an umbrella term for
these units.
3
As a result of improved production standards,
modern manufactured homes are comparable with site-
built homes in terms of quality and resident satisfaction
(Boehm and Schlottmann, 2004), although many older
or improperly installed homes experience severe repair
and maintenance challenges (Aman and Yarnal, 2010;
Lamb, Shi, and Spicer, 2023). These issues contribute to
the generally high prevalence and costliness of repair
needs among manufactured homes (Divringi, 2023).
Contrary to popular perception, most manufactured homes
are not located in MHCs. Approximately two in ive existing
manufactured homes are in MHCs (Durst and Sullivan,
2019), and less than one-third of new manufactured homes
were placed in these communities in 2021.
4
Still, there are
over 43,000 MHCs nationwide, encompassing an estimated
4.3 million homesites (Manufactured Housing Institute,
2022). Although MHCs are present throughout the United
States, the largest numbers of these communities are in
the southeast, Texas, California, and the Rust Belt states
(George and Yankausas, 2011). Additionally, although MHCs
account for a larger segment of the rural housing stock,
recent examinations of the spatial distribution of MHCs
have highlighted their presence in suburban, and even
moderate-density urban, contexts (Sullivan, Makarewicz,
and Rumbach, 2022; Pierce, Gabbe, and Gonzalez, 2018).
What constitutes a “community” of manufactured homes
varies by state and local legislation, which may in turn dier
3FEDERAL RESERVE BANK OF PHILADELPHIA
from how MHC resident organizations would deine their
constituencies. In Pennsylvania, the Manufactured Home
Community Rights Act deines a community as a group
of three or more manufactured homes on a site (68 P.S. §
398.1).
5
This size threshold was used to compile the data set
of MHCs used in this report.
Manufactured housing is widely considered an important
contributor to the unsubsidized aordable housing stock.
Although previous research suggests manufactured
homeownership is more aordable when the homeowner
also owns the underlying land, MHC homeownership
remains substantially more aordable than site-built
homeownership, even after factoring in lot rent and
accounting for neighborhood characteristics and
geography (Durst and Sullivan, 2019). The aordability
of MHC homeownership is largely a result of the savings
associated with factory-built housing. In 2022, the
average cost per square foot for a manufactured home
($72.21) was roughly half that of a new site-built home
($143.83)
6
(Manufactured Housing Institute, 2022).
MHCs have received scant attention from housing
researchers and community development practitioners.
This has been attributed to several factors, including
misguided beliefs that manufactured homes are inherently
low-quality or obsolete and that MHCs are not meaningfully
present in, or connected to, urban areas (Lamb, Shi, and
Spicer, 2023). Negative stereotypes about MHC residents,
rooted in class-based bias, are another likely contributor
to the lack of policy attention to these communities (Aman
and Yarnal, 2010; Furman, 2015).
Even scholars and practitioners who are motivated to
examine MHCs often ind that these communities are
not captured well in commonly used, publicly available
data sets. The U.S. Census Bureau’s American Community
Survey does not distinguish manufactured homes by land
tenure or presence in an MHC, and the American Housing
5
For more information, see www.palawhelp.org/resource/mobile-home-park-tenant-rights.
6
Both igures exclude land costs.
7
The full data set compiled for this analysis includes 2,288 MHCs; by comparison, the DHS data set contains 1,581 records, including some that were misclas-
siied or that are no longer active.
8
Jewell (2003) inds evidence that some homes in MHCs do appreciate, but appreciation is less likely and smaller in magnitude than that for both site-built
homes and manufactured homes on owned land.
Survey provides proxy variables at only highly aggregated
geographies. The Department of Homeland Security
maintains a national geospatial layer of MHC locations,
which was incorporated into the data set compiled for this
analysis, but I ind that this layer substantially undercounts
the number of these communities in Pennsylvania.
7
This
dearth of information on land-lease communities has
spurred recent eorts to construct new, original data sets
leveraging information from multiple sources to provide a
more complete picture of MHC locations and surrounding
contexts (Sullivan, Makarewicz, and Rumbach, 2022). This
report builds on these local and regional eorts, expanding
the scope of analysis to an entire state, encompassing a
wide range of urban, suburban, and rural contexts.
What Challenges Do Residents
of Manufactured Housing
Communities Face?
While MHCs have the potential to reduce upfront barriers
to homeownership through lower purchase costs, there
are signiicant drawbacks to a land-lease arrangement
for MHC homeowners. The most widely cited is that
such arrangements limit wealth-building potential,
and that potential is a powerful motivating norm in
American homeownership policy (Lamb, Shi, and Spicer,
2023; Furman, 2015). Unlike site-built homeowners and
manufactured homeowners who own their land, MHC
homeowners cannot rely on stable or increasing land
values to oset unit depreciation. Even properly installed,
well-maintained units are subject to wear and tear, making
it less likely that the owner will be able to resell their home
at a similar or a higher purchase price (Jewell, 2003; Boehm
and Schlottmann, 2004).
8
Furthermore, unlike site-built
homeowners, who are able to lock in relatively consistent
monthly housing payments through their mortgages, MHC
homeowners are exposed to market pressures through lot
rents, which can increase rapidly and erode the inancial
4FEDERAL RESERVE BANK OF PHILADELPHIA
beneits associated with lower purchase prices (Jewell,
2003; Durst and Sullivan, 2019). A growing number of
MHCs have been acquired by large real estate investment
companies, heightening concerns about the potential for
extractive rent increases (Associated Press, 2022).
These wealth-building challenges are compounded by the
unusual inancial treatment of manufactured homes located
on leased land. Without land ownership, these homes cannot
be titled as real property, preventing homebuyers from
accessing traditional purchase mortgages. Instead, MHC
homebuyers are limited to the less regulated, higher-cost
personal property, or “chattel,” loan market. Prospective
homebuyers often select lenders based on the options
presented at manufactured home retailers, resulting in
relatively limited comparison shopping (Genz, 2001; Kaul
and Pang, 2022). Like subprime mortgage lenders, these
inancing companies specialize in making higher-interest
loans to borrowers with less-than-perfect credit. Still, just
over half of chattel purchase loan applications are denied
(Russell et al., 2021), and the extent and terms of alternative
inancing arrangements for those unable to qualify for these
loans are not widely documented (Canavan, Roche, and
Siegel, 2022). Furthermore, chattel lending is not covered
by the Real Estate Settlement Procedures Act (RESPA),
potentially exposing borrowers to excessive or unexpected
fees and loan costs at closing.
Chattel borrowers whose homes are located in land-lease
MHCs are at a substantially higher risk of default than
manufactured homebuyers who own their underlying land
(Park, 2022), which may relect the intersecting challenges
of modest borrower incomes, high loan costs, and exposure
to lot rent increases. Chattel borrowers who default on
their loans are not subject to foreclosure but rather face
repossession, a much faster process for which consumer
protections vary widely across states.
9
Furthermore, these
borrowers are often ineligible for programs designed to
assist struggling homeowners. For example, despite being
disproportionately lower-income, MHC homeowners
were not covered by the mortgage forbearance
protections provided in the Coronavirus Aid, Relief, and
Economic Security (CARES) Act (Russell et al., 2021).
9
Some states provide repossession protections speciically for manufactured homeowners. For example, in Pennsylvania, a manufactured homeowner must
receive a 30-day advance notice and can stop the repossession by making up back payments and related fees (12 P.S. § 6262).
While most homeowners who keep up with housing
payments can expect some measure of housing security,
MHC homeowners remain at risk of displacement if
the property owner closes the community. There are
many reasons a property owner may close an MHC,
including an inability to inance needed improvements
to aging infrastructure, the enactment of burdensome or
exclusionary local regulations, and market incentives to
convert to a more proitable land use (Abu-Khalaf, Arabo,
and Swann, 2021; Sullivan, 2018). Despite the persistence of
the term “mobile home,” many manufactured homeowners
do not intend, or could not aord, to relocate their
homes, and homeowners in manufactured units are no
more likely to be transitory than those in site-built units
(Boehm and Schlottmann, 2004). The cost of moving a
manufactured home is estimated at $5,000–$10,000,
but it can vary widely by home size and condition
(Ehrenfeucht, 2016). Many older units are not moveable,
and newer multisection models, which have grown in
popularity in recent years, are more costly to relocate
(Aman and Yarnal, 2010; Sullivan, 2018). In the event of an
MHC closure, homeowners who are unable to move their
homes may abandon their units, forfeiting their already
diminished opportunity for asset building (Sullivan, 2018).
In response, several states and municipalities require
relocation assistance payments for residents displaced
by MHC closures (Ehrenfeucht, 2016). However, even
homeowners who can relocate may have diiculty inding
a suitable alternative site nearby because of widespread
exclusionary zoning practices (Dawkins, et al., 2011).
Unlike site-built homeowners
and manufactured
homeowners who own their
land, MHC homeowners
cannot rely on stable or
increasing land values to
oset unit depreciation.
5FEDERAL RESERVE BANK OF PHILADELPHIA
Last, while this report focuses on several issues speciic to
MHC homeowners, roughly 30 percent of MHC households
are likely to be renters.
10
The risk of community closure
may represent an added layer of residential insecurity for
lower-income renters, who already face acute shortages of
aordable units (JCHS, 2022). Although requirements vary
by state, eligibility for relocation assistance is often limited
to MHC homeowners.
Despite these drawbacks, MHCs continue to house
millions of residents nationwide, providing aordable and
attainable homeownership opportunities in a wide range
of communities. Accordingly, it is important for housing
practitioners and policymakers to better understand this
unique tenure type and the challenges facing low- and
moderate-income MHC residents. The remainder of this
report will provide essential background information on
MHCs in Pennsylvania as a starting point for advancing
solutions-based conversations around preserving
aordability and improving housing security.
10
Author’s calculations using the AHS Table Creator, available at www.census.gov/programs-surveys/ahs/data/interactive/ahstablecreator.html. This estimate
is based on the number of renter-occupied units in manufactured homes in groups of seven or more in 2021.
11
Pennsylvania was selected as the focus of this analysis based on research indicating that it is home to the highest number of MHCs across the three states
of the Third Federal Reserve District (George and Yankausas, 2011). Future work will explore the potential to expand this dataset to New Jersey and Delaware.
12
Available at www.census.gov/geographies/mapping-iles/time-series/geo/tiger-line-ile.2019.html.
13
Available at htaindex.cnt.org/download/.
Manufactured Housing
Communities in Pennsylvania
The following sections describe the spatial distribution,
utilization, and community contexts of MHCs in
Pennsylvania.
11
While some information is available
from public data sets, such as the American Housing
Survey (AHS) and Home Mortgage Disclosure Act
(HMDA) data, the primary source is a novel data set of
MHC locations that I compiled from three sources: tax
assessment data assembled by CoreLogic Solutions
(CoreLogic), Homeland Infrastructure Foundation-Level
Data from the Department of Homeland Security, and
the membership list of the Pennsylvania Manufactured
Housing Association. This new data set is intended to
provide a comprehensive inventory of MHCs in the state,
along with limited information on community size and lot
vacancies. For a detailed description of the construction
and validation of this data set, see Appendix A.
To shed light on the characteristics of communities in
which MHCs are located, I used geographic information
system (GIS) software to spatially join the locations of
MHCs to Census Bureau geographies from the TIGER/Line
Shapeiles,
12
which enabled the data set to be merged with
neighborhood-level demographic and socioeconomic data
from the 2016–2020 American Community Survey and
built environment data from the Center for Neighborhood
Technology Housing + Transportation Aordability Index.
13
Size of the Manufactured Housing Stock
According to the 2021 AHS, there are roughly 169,200
occupied manufactured housing units in Pennsylvania. Of
those, nearly one-third (55,900) are in groups of seven or
Even homeowners who can relocate may have
difficulty finding a suitable alternative site nearby
because of widespread exclusionary zoning practices.
6FEDERAL RESERVE BANK OF PHILADELPHIA
more, which are likely to be in land-lease MHCs.
14
While
this accounts for just over 1 percent of the occupied
housing stock in the state, this igure is comparable
with the number of Pennsylvania households in federally
funded public housing (53,558) and Project-Based Section
8 housing (58,369), both of which receive substantial
attention from housing scholars and policymakers (U.S.
Department of Housing and Urban Development, 2021).
From 2017 to 2021, an average of 1,706 new manufactured
homes were shipped to Pennsylvania each year. While
data on the placement of new manufactured units are not
available at the state level, the majority of units shipped to
the Northeast were placed in land-lease communities during
that period, ranging from 55 to 70 percent, depending on
the year (U.S. Census Bureau, 2021). Despite these new
shipments, the number of occupied manufactured homes
(both inside and outside MHCs) declined by 12.5 percent
statewide from 2011 to 2021,
15
most likely driven by declines
in older units that aged out of the housing stock.
14
The American Housing Survey (AHS) does not directly ask manufactured home residents if they live in land-lease communities or parks. Instead, the AHS
reports the number of manufactured homes located in groups across three size bins: one to six homes, seven to 20, and 21 or more. The 55,900 igure should
be considered a rough estimate, as it excludes MHCs with three to six units but may include groups of manufactured homes where homeowners own the land
beneath their properties.
15
Author’s calculations based on U.S. Census Bureau 2011 and 2021 American Community Survey.
Affordability
Table 1 compares recent home purchase loan
characteristics for site-built and manufactured homebuyers
in Pennsylvania. Manufactured homebuyers were generally
lower-income and obtained substantially smaller loans
than their counterparts purchasing site-built homes.
Manufactured homebuyers who used chattel loans
for properties to be placed on leased land (likely MHC
homebuyers) were the lowest-income and received the
smallest loan amounts by far, which is unsurprising, given
that these loans excluded the underlying land. However,
although the median loan-to-value ratio was much lower,
the typical chattel loan carried more than double the
interest rate of a typical manufactured home mortgage.
Controlling for borrower characteristics, such as income
and credit score, would likely narrow but not completely
close this interest rate gap (Park, 2022; Russell et al., 2021).
As outlined in Table 1, the typical MHC homebuyer would
see a monthly loan payment of $376, substantially lower
than the loan payments for both site-built and manufactured
Manufactured: Chattel Loan,
Leased Land
Manufactured: Mortgage Loan,
Direct Land Ownership
Site-Built: Mortgage Loan
Number 2,002 2,794 412,692
Median Applicant Income $52,000 $56,000 $78,000
Median Interest Rate 7.99% 3.88% 3.25%
Median Loan Amount $45,000 $125,000 $215,000
Median Loan Term (Months) 240 360 360
Median Loan-to-Value Ratio 83.9% 90.0% 95.0%
Est. Monthly Loan Payment $376 $588 $936
Comparison of Originated Loan Characteristics by Land Ownership and Build
Type, Pennsylvania, 2019–2021
Notes
Calculations include only originated, irst-lien purchase loans for owner occupancy. Chattel loans to manufactured homebuyers with direct land ownership,
indirect land ownership, and unpaid leaseholds are excluded. Estimated monthly loan payments are based on median loan amounts, median terms, and median
interest rates reported in the table.
Sources
Author’s calculations using 2019–2021 Home Mortgage Disclosure Act data
TABLE 1
7FEDERAL RESERVE BANK OF PHILADELPHIA
homebuyers with mortgages. However, in addition to loan
payments, chattel borrowers are exposed to the cost of lot
rent. According to the 2021 AHS, median monthly lot rent in
Pennsylvania was $370.
16
Factoring this in, monthly rent plus
loan costs for the median MHC homebuyer would be $746.
While this omits certain housing costs, such as property
taxes,
17
this back-of-the-envelope exercise highlights the
critical importance of lot rents to the overall aordability of
the MHC arrangement.
Geographic Distribution
Figure 1 displays the location and size
of MHCs in Pennsylvania. Notably,
MHCs are present in every county
in the state except for Philadelphia
County. Overall, MHCs are spatially
clustered in the populous west
and southeast regions of the state,
and along highway routes in less
densely populated areas. The two
counties with the largest number
of MHCs are Lancaster (140) and
York (109), which are adjacent,
single-county metropolitan areas
with midsize central cities (see
Appendix B for breakouts for
all Pennsylvania counties).
Just under two-thirds (62.5 percent)
of MHCs in the data set are
categorized as medium-sized, with
11 to 99 homesites. Small MHCs with
three to 10 homesites account for
another quarter (26.2 percent), and
the remaining tenth (11.4 percent)
are large MHCs that have 100 or
more homesites.
18
As illustrated in
16
Includes manufactured homeowners who lease their land but are not in an MHC. Median lot rent in Pennsylvania was not signiicantly dierent (at the 90
percent conidence level) from the national median ($414). Relative to the three states with the largest numbers of MHCs (George and Yankausas, 2011), the me-
dian lot rent in Pennsylvania was signiicantly lower than those in in California ($700) and Florida ($600), but not signiicantly dierent than that in Texas ($315).
17
Manufactured homes that are titled as personal property (such as those in land-lease MHCs) are subject to real property taxes in Pennsylvania, which are
not factored into the igures above, although the portion of property taxes applied to the leased land are likely passed through in lot rent. Manufactured home-
buyers with direct land ownership would be assessed taxes on both their home and land, neither of which is included in these igures.
18
Large MHCs can include hundreds of homesites or more. For example, Pennwood Crossing in Bucks County, Pennsylvania, has over 1,000 homesites (see
www.mhvillage.com/parks/22829).
Figure 1, large MHCs tend to be located along the urban
fringe in midsize and large metropolitan areas, particularly
around Pittsburgh (Allegheny County), Erie, Allentown
(Lehigh County), Scranton (Lackawanna County), York, and
Harrisburg (Dauphin County). Small and medium-sized MHCs
are more dispersed and present in less populated parts of
the state, although many are similarly clustered around the
outskirts of cities and suburbs.
Locations of MHCs in Pennsylvania
FIGURE 1
Notes
Urban/Rural classiications are based on the 2010 census.
Sources
Philadelphia FRB Manufactured Housing Community Dataset, Census TIGER/Line Shapeiles, and OpenStreetMap
Erie
Elk
York
Tioga
Potter
Centre
Berks
Butler
Bradford
Lycomin g
Pike
Bedford
Clinton
Warren
Clearfield
McKean
Blair
Crawford
Indiana
Somerset
Luzerne
Wayne
Fayette
Perry
Bucks
Lancaster
Mercer
Franklin
Chester
Clarion
Schuylkill
Cambria
Monroe
Huntingdon
Greene
Venango
Allegheny
Adams
Washington
Westmoreland
Fulton
Forest
Dauphin
Armstrong
Beaver
Susquehanna
Sullivan
Juniata
Union
Carbon
Columbia
Lehigh
Snyder
Cumberland
Wyoming
Cameron
Lebanon
Montgomery
Lawrence
Delaware
Lackawanna
Northumberland
Northampton
Montour
Philadelphia
Esri, HERE, Garmin, (c) OpenStreetMap contributors, and the GIS user community
MHC Size
Large ( 100 Homesites)
Medium (11-99 Homesites)
Small (3-10 Homesites)
Urban Regions
8FEDERAL RESERVE BANK OF PHILADELPHIA
Notably, although MHCs are considered a predominantly
rural housing type (Aman and Yarnal, 2010), Table 2
indicates that a slight majority of Pennsylvania MHCs
(1,225, or 53.5 percent) are in urban areas, with close
to four in ten in large urban regions.
19
Still, given that
rural areas account for less than one-quarter of the total
housing units in the state,
20
MHCs are disproportionately
located in rural settings.
Table 2 conirms the patterns visible in Figure 1. Large
MHCs are a much greater share of the MHC landscape
in large urban regions (20.2 percent) than in rural areas
(5.3 percent). Small MHCs, which are sometimes found
interspersed among site-built homes in lower-density
neighborhoods, are more common in both rural areas (30.2
percent) and small urban regions (30.4 percent) than in
other settings. Medium-sized MHCs account for a relatively
consistent share across community types but are slightly
more common in rural areas (64.5 percent).
19
The Census Bureau delineates urban and rural areas based on land use, residential density, and road connections. Urban area designations based on the
2010 census were the most recent available at the time of analysis and are used throughout this report. In the 2010 urban area designations, the Census Bureau
divides urban areas into two categories – small urban regions (“Urban Clusters”) and large urban regions (“Urbanized Areas”). Large urban regions typically
consist of an assemblage of central cities and adjacent suburbs. Small urban regions typically comprise small towns that are not near a larger central city. Ar-
eas that do not meet the 2010 census urban criteria are classiied as rural. On December 29, 2022, the Census Bureau inalized a new list of urban areas based
on the 2020 census using revised criteria that, among other changes, increased the minimum population threshold for urban classiication. Using the 2020
census urban areas, 82 MHCs (3.6 percent of the total) classiied as urban in this analysis would instead be classiied as rural.
20
Author’s calculations using 2021 data from the AHS Table Creator.
Count
Row % in Size Category
Small
(3–10 homesites)
Medium
(11–99 homesites)
Large
(≥100 homesites)
Rural 1,063 30.2 64.5 5.3
Small Urban Region 372 30.4 61.0 8.6
Large Urban Region 853 19.3 60.5 20.2
Total 2,288 26.2 62.5 11.4
MHCs by Urban/Rural Location and MHC Size
Notes
Urban/Rural classiications are based on the 2010 census. Figures are tabulated at the MHC level and are not unit-weighted.
Sources
Author’s calculation using the Philadelphia FRB Manufactured Housing Community Dataset and Census TIGER/Line Shapeiles.
TABLE 2
9FEDERAL RESERVE BANK OF PHILADELPHIA
To further illustrate the range of settings in which MHCs can
be found, Figure 2 presents the distribution of residential
densities, in households per acre (HH/acre), of the census
tracts containing MHCs. For context, the average residential
density across Pennsylvania is 0.77 HH/acre, which is less
than the average across tracts containing MHCs (1.17 HH/
acre), but higher than the median (0.68 HH/acre).
Although the majority of MHCs are in low-density areas,
the 75th percentile (1.40 HH/acre) is comparable with the
residential densities associated with counties in midsize
metropolitan areas (e.g., Lackawanna County in the
Scranton MSA and Northampton County in the Allentown
MSA). Tracts containing small and medium-sized MHCs
follow a similar distribution as MHCs overall, while large
MHCs tend to be in tracts with somewhat higher densities
(to which these communities likely contribute), aligning
with previous indings about the disproportionately
suburban location of this subset of MHCs.
Residential Density (HH/Acre)
of Tracts Containing MHCs
Notes
Boxes relect 25th to 75th percentile values; interior white lines represent the
medians; and markers represent the mean value.
Sources
Author’s calculation using the Philadelphia FRB Manufactured Housing
Community Dataset, Census TIGER/Line Shapeiles, and Center for
Neighborhood Technology Housing + Transportation (H+T®) Aordability
Index, based on 2010 census blocks.
FIGURE 2
-
2.50
2.00
1.50
1.00
0.50
10FEDERAL RESERVE BANK OF PHILADELPHIA
Lot Utilization
Figure 3 compares the share of MHCs identiied as having
a high rate of lot vacancy by community size and urban/
rural location. An MHC was considered to have high lot
vacancy if 30 percent or more of its homesites were
vacant (i.e., not occupied by a manufactured home)
in the most recent available aerial image.
21
Although
this represents a snapshot of utilization for a relatively
dynamic housing arrangement, lot vacancy can be viewed
as an indicator of MHC demand.
22
Additionally, since
each unused homesite represents forgone income for
the property owner, MHCs with high lot vacancy may be
at an elevated risk for disinvestment or closure due to
inadequate revenue from lot rents.
Just over one in nine MHCs (11.6 percent) in the data
set met the deinition for high lot vacancy, suggesting
signiicant underutilization, given the high threshold
for this category. The prevalence ranged substantially
across MHC size and location. Overall, high lot vacancy
21
See Appendix A for details.
22
A high vacancy rate may also relect the diiculty of adapting older MHCs to accommodate larger units that have grown in popularity in recent years, par-
ticularly in cases in which these modiications would trigger a zoning review.
was more common in rural areas (13.9 percent) than in
other community types and among medium-sized MHCs
(13.4 percent) when compared with smaller and larger
communities. The level of high lot vacancy among small
MHCs (9.8 percent), which require only a few vacant sites
to meet the threshold, was slightly below average and
relatively consistent across locations, although slightly
elevated in smaller urban regions. Large MHCs (5.4
percent) were the least likely to have high lot vacancy,
driven in part by the concentration of these MHCs in larger
urban regions, where MHCs across size categories were
least likely to have high lot vacancy. However, large MHCs
were much more likely to experience high lot vacancy in
smaller urban regions, suggesting that demand for these
communities is concentrated in the housing markets
surrounding larger and midsize cities.
Demographic and Socioeconomic Context
Table 3 compares the demographic characteristics of the
census tracts surrounding MHCs with those of Pennsylvania
Share of MHCs with High Lot Vacancy (≥30 Percent)
Notes
Urban/Rural classiications are based on the 2010 census. Percentages were calculated at the MHC level and are not unit-weighted.
Sources
Author’s calculation using the Philadelphia FRB Manufactured Housing Community Dataset and Census TIGER/Line Shapeiles.
FIGURE 3
All Rural Small Urban Regions Large Urban Regions
Large (≥100 Homesites)
All
Small (3–10 Homesites)
Medium (11–99 Homesites)
11.6%
9.8%
13.4%
5.4%
13.9%
9.3%
16.6%
7.1%
12.9%
12.4%
12.8%
15.6%
8.1%
9.1%
9.5%
2.9%
11FEDERAL RESERVE BANK OF PHILADELPHIA
overall. It is important to note that these should not be
interpreted as the demographics of MHC residents, but
rather as context for the communities in which MHCs are
located. Demographic characteristics of tracts containing
MHCs are remarkably similar across size categories.
Compared with Pennsylvania overall, these areas have a
much higher share of non-Hispanic White residents and,
as a result, lower shares of residents of color, particularly
non-Hispanic Black residents.
23
The share of residents aged 65 years or older is somewhat
higher in tracts containing MHCs than in Pennsylvania
overall, even though the state has a relatively large share
of older adult residents (Kildu, 2021). In some cases,
the presence of MHCs in these tracts may contribute to
this higher share of retirement-age adults, since a subset
of MHCs is age-restricted communities. Additionally,
some unrestricted MHCs evolve into “naturally occurring
23
By contrast, a recent analysis of MHCs in the Houston, TX, metropolitan area found that these communities were disproportionately located in areas with
larger Hispanic or Latino populations (Sullivan, Makarewicz, and Rumbach, 2022).
retirement communities,” as the unique physical, social,
and inancial characteristics of these living arrangements
often appeal to older adults (Tremoulet, 2010).
Table 4 summarizes the socioeconomic characteristics
of the areas surrounding MHCs, both overall and by
community size. Compared with the state, these areas
have somewhat lower levels of educational attainment,
with larger shares of adults having a high school
diploma or less. This is true across size categories
but is more pronounced for the areas around small
and medium-sized MHCs. A similar pattern holds for
household incomes and home values. However, the
areas surrounding large MHCs have similar, if not
slightly higher, household incomes than Pennsylvania
overall, likely owing to the concentration of these
communities in higher-wage metropolitan job markets.
PA
PA Tracts Containing MHCs
All MHCs
Small
(3–10 homesites)
Medium
(11–99 homesites)
Large
(≥100 homesites)
Race/Ethnicity
Share Black 10.6% 1.9% 1.8% 1.8% 2.2%
Share Hispanic/Latino 7.6% 3.2% 3.0% 3.2% 3.7%
Share White 75.7% 92.0% 92.8% 92.1% 90.3%
Share Other/Multiracial 6.1% 2.9% 2.5% 2.9% 3.7%
Age
Share 65 Years or Older 18.3% 20.7% 20.4% 21.0% 20.1%
Demographic Characteristics of Census Tracts Containing MHCs Relative
to Pennsylvania
Notes
Race/ethnicity categories are mutually exclusive. The Black, White, and other/multiracial categories are non-Hispanic; Hispanic/Latino can be of any race.
Sources
Author’s calculations using the Philadelphia FRB Manufactured Housing Community Dataset and U.S. Census Bureau American Community Survey 2020
5-Year Estimates.
TABLE 3
12FEDERAL RESERVE BANK OF PHILADELPHIA
Despite lower rates of educational attainment and more
modest household incomes and home values, the areas
surrounding MHCs do not appear to be, on average,
particularly distressed. For MHC tracts overall and in each
size category, labor force participation and unemployment
rates are comparable with statewide igures, rates of family
poverty are lower, and homeownership rates are markedly
higher. Taken together, these characteristics suggest that,
relative to the state of Pennsylvania as a whole, MHCs often
provide low-cost housing opportunities in low-poverty,
high-homeownership neighborhoods.
PA
PA Tracts Containing MHCs
All MHCs
Small
(3–10 homesites)
Medium
(11–99 homesites)
Large
(≥100 homesites)
Educational Attainment
High School or Lower 43.2% 50.9% 54.1% 50.5% 46.8%
Bachelor’s Degree or Higher 32.3% 24.2% 21.3% 24.5% 28.4%
Employment
Labor Force Participation Rate 62.8% 61.5% 60.8% 61.5% 62.8%
Unemployment Rate 5.4% 4.3% 4.2% 4.4% 4.0%
Income
Family Poverty Rate 8.1% 5.9% 6.5% 5.8% 5.2%
Median Household Income $68,962 $65,165 $63,133 $64,934 $70,442
Housing
Homeownership Rate 69.0% 79.7% 78.9% 79.8% 80.7%
Median Home Value $204,213 $176,114 $167,711 $175,223 $197,607
Share Housing Cost Burdened 27.2% 22.5% 22.4% 22.4% 23.2%
Socioeconomic Characteristics of Census Tracts Containing MHCs Relative
to Pennsylvania
Notes
Estimates for MHC tracts are weighted based on the universe of the target estimate (i.e., population in category, number of families, or number of households).
Educational attainment measures are calculated for the population 25 years old and over. Employment measures are calculated for the population 16 years
old and over. Medians are calculated as household-weighted averages of tract medians. A household is housing cost–burdened if total housing costs equal or
exceed 30 percent of household income.
Sources
Author’s calculations using the Philadelphia FRB Manufactured Housing Community data set and U.S. Census Bureau American Community Survey 2020
5-Year Estimates.
TABLE 4
13FEDERAL RESERVE BANK OF PHILADELPHIA
Takeaways for Policy and Practice
24
For more information, see rocusa.org/whats-a-roc/what-is-a-roc-how-is-it-dierent/.
A few noteworthy themes emerge in this initial statewide
analysis of MHCs in Pennsylvania. The irst is that MHCs
are present across a wide range of rural to midsize urban
communities and, contrary to common perception, are
present in the housing markets that surround larger urban
areas. Furthermore, the lower likelihood of excessive
lot vacancy in these regions suggests that demand for
MHC-style housing is strongest on the outskirts of larger
urban areas, where they may represent a more attainable
option for lower-income homebuyers.
Although the areas surrounding MHCs are not especially
aluent, indings from this initial analysis suggest that
these communities often provide a source of unsubsidized
aordable housing in low-poverty neighborhoods. By
contrast, formally subsidized housing developments
have been criticized for disproportionately concentrating
households in distressed, economically marginalized
neighborhoods (Newman and Schnare, 1997; McClure,
2008). With high construction costs and a shortage of
low-cost for-purchase homes (Choi and Zinn, 2022),
housing practitioners may consider opportunities to
preserve, or even expand, access to MHCs as part of their
aordable housing toolkit.
Researchers and practitioners focusing on MHCs have
proposed several strategies for leveraging the potential
of these communities as a response to the shortage of
low-cost housing (Sullivan, 2018). For example, facilitating
residents’ cooperative ownership of the land beneath
their homes is often discussed as a means of increasing
residential security, while providing MHC homeowners
with an avenue for asset building (NCLC, 2021; George
and Yankausas, 2011; Ward, French, and Giraud, 2006;
Abu-Khalaf, Arabo, and Swann, 2021).
24
Additionally,
although access to small-dollar home purchase loans
remains a challenge for even site-built homebuyers
(Goldstein and DeMaria, 2022), innovations in community
development inance can yield more consumer-friendly
purchase loan products as alternatives to high-cost chattel
inancing (Thomas, 2019).
While this report provides baseline information to ground
our understanding of the spatial distribution of MHCs in
Pennsylvania, future work will take a closer look at pressing
issues aecting the continued aordability and livability
of these communities. Forthcoming research briefs will
explore MHC residents’ access to employment and public
infrastructure, their exposure to climate-related risks, and
other emerging challenges. These briefs are intended to ill
critical information gaps on this understudied segment of
the low-cost housing stock and can help policymakers and
practitioners better understand and respond to the unique
circumstances facing MHC households.
These communities
often provide a source of
unsubsidized affordable
housing in low-poverty
neighborhoods.
Abu-Khalaf, Ahmad, Flora Arabo, and Stevene Swann. Policy Brief: Preserving the Aordability of Manufactured Homes in Land-Lease
Communities. Washington, D.C.: Enterprise Community Partners, 2021.
Aman, Destiny D., and Brent Yarnal. “Home Sweet Mobile Home? Beneits and Challenges of Mobile Home Ownership in Rural
Pennsylvania.” Applied Geography 30 (2010), pp. 84–95.
Associated Press. “Rents Spike as Large Corporate Investors Buy Mobile Home Parks.” PBS NewsHour, July 25, 2022, www.pbs.org/
newshour/economy/rents-spike-as-large-corporate-investors-buy-mobile-home-parks.
Boehm, Thomas P., and Alan Schlottmann. Is Manufactured Housing a Good Alternative for Low-Income Families? Evidence From the
American Housing Survey. Washington, D.C.: U.S. Department of Housing and Urban Development Oice of Policy and Research, 2004.
Canavan, Ryan, Tara Roche, and Rachel Siegel. Millions of Americans Have Used Risky Financing Arrangements to Buy Homes.
Washington, D.C.: The Pew Charitable Trusts, 2022. Available at www.pewtrusts.org/en/research-and-analysis/issue-briefs/2022/04/
millions-of-americans-have-used-risky-inancing-arrangements-to-buy-homes.
Choi, Jung Hyun, and Amalie Zinn. “Eighty Percent of Homes on the Market Aren’t Aordable for Households Earning Median
Incomes or Less.” Urban Wire (blog), December 7, 2022, www.urban.org/urban-wire/eighty-percent-homes-market-arent-aordable-
households-earning-median-incomes-or-less.
Dawkins, Casey J., C. Theodore Koebel, Marilyn Cavell, et al. Regulatory Barriers to Manufactured Housing Placement in Urban
Communities. Washington, D.C.: U.S. Department of Housing and Urban Development, 2011.
Divringi, Eileen. Research Brief: Updated Estimates of Home Repair Needs and Costs. Philadelphia: Federal Reserve Bank of Philadelphia,
2023. Available at www.philadelphiafed.org/-/media/frbp/assets/community-development/reports/23-02-home-repairs-update.pdf.
Durst, Noah J., and Esther Sullivan. “The Contribution of Manufactured Housing to Aordable Housing in the United States:
Assessing Variation Among Manufactured Housing Tenures and Community Types.” Housing Policy Debate 29:6 (2019), pp. 880–98.
Ehrenfeucht, Renia. Moving Beyond the Mobile Myth: Preserving Manufactured Housing Communities. Oakland, CA: Grounded Solutions
Network, 2016. Available at groundedsolutions.org/sites/default/iles/2018-11/Moving%20Beyond%20the%20Mobile%20Myth.pdf.
Furman, Matthew. “Eradicating Substandard Manufactured Homes: Replacement Programs as a Strategy.” Joint Center for Housing
Studies of Harvard University Working Paper 15-3, 2015.
Genz, Richard. “Why Advocates Need to Rethink Manufactured Housing.” Housing Policy Debate 12:2 (2001), pp. 393–414.
George, Lance, and Jann Yankausas. Preserving Aordable Manufactured Home Communities in Rural America: A Case Study.
Washington, D.C.: Housing Assistance Council, 2011. Available at ruralhome.org/reports/preserving-aordable-manufactured-home-
communities-in-rural-america/.
Goldstein, Emily, and Kyle DeMaria. Small-Dollar Mortgage Lending in Pennsylvania, New Jersey, and Delaware. Philadelphia: Federal
Reserve Bank of Philadelphia, 2022. Available at www.philadelphiafed.org/community-development/credit-and-capital/small-dollar-
mortgage-lending-in-pennsylvania-new-jersey-and-delaware.
Jewell, Kevin. Manufactured Housing Appreciation: Steroetypes and Data. Austin, TX: Consumers Union Southwest Regional Oice, 2003.
Joint Center for Housing Studies (JCHS) of Harvard University. America’s Rental Housing 2022. Cambridge, MA: Joint Center for
Housing Studies of Harvard University, 2022. Available at www.jchs.harvard.edu/sites/default/iles/reports/iles/Harvard_JCHS_
Americas_Rental_Housing_2022.pdf.
References
14
FEDERAL RESERVE BANK OF PHILADELPHIA
References
15FEDERAL RESERVE BANK OF PHILADELPHIA
Kaul, Karan, and Daniel Pang. The Role of Manufactured Housing in Increasing the Supply of Aordable Housing. Washington, D.C.:
Urban Institute, 2022.
Kildu, Lillian. “Which U.S. States Have the Oldest Populations?” Web page, Population Reference Bureau, December 22, 2021, www.
prb.org/resources/which-us-states-are-the-oldest/.
Lamb, Zachary, Linda Shi, and Jason Spicer. “Why Do Planners Overlook Manufactured Housing and Resident-Owned Communities
as Sources of Aordable Housing and Climate Transformation?” Journal of the American Planning Association 89:1 (2023), pp. 72–9.
McClure, Kirk. “Deconcentrating Poverty with Housing Programs.” Journal of the American Planning Association 74:1 (2008), pp. 90–9.
Manufactured Housing Institute. 2022 Manufactured Housing Facts: Industry Overview. Arlington, VA: Manufactured Housing Insutitute,
2022. Available at www.manufacturedhousing.org/wp-content/uploads/2022/04/2022-MHI-Quick-Facts-updated-05-2022-2.pdf.
NCLC. Manufactured Housing Resource Guide: Promoting Resident Ownership of Communities. Boston: National Consumer Law
Center, 2021.
Newman, Sandra J., and Ann B. Schnare. “‘… And a Suitable Living Environment’: The Failure of Housing Programs to Deliver on
Neighborhood Quality.” Housing Policy Debate 8:4 (1997), pp. 703–41.
Park, Kevin A. “Real and Personal: The Eect of Land in Manufactured Housing Loan Default Risk.” Cityscape: A Journal of Policy
Development and Research 24:3 (2022), pp. 339–62.
Pierce, Gregory, C.J. Gabbe, and Silvia R. Gonzalez. “Improperly Zoned, Spatially Marginalized, and Poorly Served? An Analysis of
Mobile Home Parks in Los Angeles County.” Land Use Policy 76 (2018), pp. 178–85.
Russell, Jessica, Nora O’Reilly, Karl Schneider, et al. Manufactured Housing Finance: New Insights from the Home Mortgage
Disclosure Act Data. Washington, D.C.: Consumer Financial Protection Bureau Oice of Research and Mortgage Markets, 2021.
Available at www.consumerinance.gov/data-research/research-reports/manufactured-housing-inance-new-insights-hmda/.
Sullivan, Esther. Manufactured Insecurity: Mobile Home Parks and Americans’ Tenuous Right to Place. Berkeley, CA: University of
California Press, 2018.
Sullivan, Esther, Carrie Makarewicz, and Andrew Rumbach. “Aordable but Marginalized: A Sociospatial and Regulatory Analysis of
Mobile Home Parks in the Houston Metropolitan Area.” Journal of the American Planning Association 88:2 (2022), pp. 232–44.
Thomas, Gail. “CDFIs Help Make Manufactured Housing Aordable for America.” CDFI Fund Impact Blog, July 9, 2019, www.cdifund.
gov/impact/70.
Tremoulet, Andrée. “Manufactured Home Parks: NORCs Awaiting Discovery.” Journal of Housing For the Elderly 24:3–4 (2010), 335–55.
U.S. Census Bureau. “Manufactured Housing Survey.” Annual Tables of New Manufactured Homes: 2014–2021 [data set], 2021.
Available at www.census.gov/data/tables/time-series/econ/mhs/annual-data.html (accessed December 2022).
U.S. Department of Housing and Urban Development. (2021). “Picture of Subsidized Households, 2021.” Oice of Policy Development
and Research [data set], 2021. Available at www.huduser.gov/portal/datasets/assthsg.html (accessed December 2022).
Ward, Sally K., Charles A. French, and Kelly Giraud. Resident Ownership in New Hampshire’s “Mobile Home Parks:” A Report on
Economic Outcomes. Durham, N.H.: Carsey Institute, University of New Hampshire, 2006.
Appendix A. Data and Methods
16
FEDERAL RESERVE BANK OF PHILADELPHIA
Philadelphia FRB Manufactured Housing
Community (MHC) Data Set
I developed the Philadelphia FRB MHC Dataset for
this report and future analyses with the goal of
addressing information gaps in MHC research.
25
This
initial iteration covers only Pennsylvania, although
future improvements may include additional states.
The data set provides the latitude and longitude for
all identiied land-lease MHCs and categorizes entries
as small (three to 10 homesites), medium (11 to 99
homesites), or large (100 or more homesites). MHCs
in which 30 percent or more of the homesites are not
in use are categorized as having high lot vacancy.
I used aerial imagery to visually code communities into size
and lot vacancy categories. I determined the thresholds
for these categories early in the data set development
process based on a review of an initial set of conirmed
MHC locations, with the intent of making qualitative
distinctions across communities. An individual lot was
considered vacant if it appeared to previously be the site
of a manufactured home that had since been moved or
demolished based on the most recent available aerial
image. Existing units that may have been unoccupied did
not count toward this vacancy measure. Vacant lots were
included in size category determinations.
In accordance with Pennsylvania’s Manufactured Home
Community Rights Act of 2012,
26
MHCs are deined as
groupings of at least three manufactured homes that
lease the land on which they are situated. Communities
with manufactured homes in which residents own their
underlying parcel are not included this data set, as
they are not subject to the split-tenure arrangement
25
External researchers may be able to access the data set for research produced in collaboration with the Federal Reserve Bank of Philadelphia. Inquiries
should be sent to Eileen Divringi at [email protected].org.
26
For more information, see www.phfa.org/legislation/act156.aspx.
27
For more information, see www.corelogic.com/wp-content/uploads/sites/4/downloadable-docs/capital-markets-data-sources.pdf.
28
For more information, see pmha.org/.
that characterizes land-lease MHCs. Campsites that
cater primarily to nightly or seasonal RV campers
are also excluded, since these are not intended for
long-term residential use. In the construction of this
data set, I made inclusion determinations based on
available information from public records, community/
campsite websites, and other online sources. I
take responsibility for any errors or omissions.
I used the following sources to construct the data set:
• CoreLogic Solutions Property Records Data:
27
This
data set consists of public property assessment
records, including information on land use, address,
and geographic coordinates. A custom query designed
to capture keywords associated with MHCs was used
to generate a list of potential locations. For most
Pennsylvania counties, records were queried from the
2021 tax year. For Sullivan and Warren counties, 2021
tax year data were not available at time of query; as a
result, tax year 2020 data were substituted.
• Pennsylvania Manufactured Housing Association
(PMHA) Membership List: PMHA is a membership
organization that advocates for the factory-built
housing industry in Pennsylvania.
28
Its membership list
includes records with mailing addresses for individual
MHCs, manufactured home builders, and other
manufactured housing-related stakeholders. After
cleaning and iltering this data set based on inclusion
criteria described previously, retained records were
geocoded using the PolicyMap Data Loader tool.
Appendix A. Data and Methods
17
FEDERAL RESERVE BANK OF PHILADELPHIA
• Department of Homeland Security (DHS) Homeland
Infrastructure Foundation Level Data (HIFLD):
29
The
Mobile Home Parks feature class/shapeile contains
mobile home, residential trailer, and recreational
vehicle (RV) parks in the continental United States and
Alaska. The inal data set includes the relevant features
from the Pennsylvania subset of this ile.
• Google Earth:
30
Google Earth is a desktop-based
mapping application that combines recent and historic
aerial imagery with GIS data, making it possible to
search and review aerial imagery for both addresses
and geographic coordinates. Each MHC record was
veriied and coded into size and vacancy categories
using Google Earth aerial imagery. Depending on the
location, the most recent available imagery ranged
from less than a year old to more than ive years old.
The initial data set was derived from a query of the
CoreLogic data intended to identify MHC parcels or
units within MHCs. To deduplicate MHC records within
the query output, I removed records with identical
geographic coordinates, clustered locations within
500-foot buers, and reviewed contextual ields
in the assessment data to determine if the records
pertained to the same site. Following this initial data
set cleaning, I reviewed each retained property record
on Google Earth to conirm its use as an MHC.
In some cases, it was not possible to infer from aerial
images whether MHCs consisted of multiple adjacent
parcels or if each parcel represented a distinct MHC. In
these cases, I cross-referenced parcels with boundary
maps accessed via the Regrid online mapping application.
31
29
The full data set is available at hild-geoplatform.opendata.arcgis.com/datasets/geoplatform::mobile-home-parks/about.
30
Available at earth.google.com/web/. Historical imagery available in desktop version of Google Earth.
31
Regrid is a property data company that maintains a nationwide parcel boundary mapping application that includes information, such as parcel ownership,
from public records data. Cross-referencing the mapped parcel boundaries with satellite imagery helped identify and distinguish between multiparcel and
adjacent MHCs. The mapping application is accessible at app.regrid.com/.
If adjacent parcels had dierent owners of record, the
parcels were retained as separate MHCs. If adjacent parcels
had the same owner but did not appear to share a street
entrance or interior streets, the parcels were retained as
separate MHCs. Adjacent parcels with the same owner
of record and shared entrances or interior streets were
consolidated into one MHC record.
The PMHA membership list and DHS Mobile Home
Parks data sets were used to supplement the outputs
from the initial CoreLogic query. To prevent introducing
duplicates, I overlaid the geocoded PMHA data set
with the cleaned CoreLogic data set and removed any
points in the PMHA dataset within a 1,000-foot buer of
CoreLogic data set points. I repeated this process for
the DHS data set, using 1,000-foot buers for both the
cleaned CoreLogic data set and deduplicated PMHA
supplement. This generated two new lists of MHC
records, which were then reviewed and coded for size
and vacancy using Google Earth. Veriied records from
each input data set were collated into a combined ile. For
a inal deduplication check, I truncated the geographic
coordinates of every record to two decimal places and
veriied duplicate values using Google Earth and Regrid.
Spatial Joins
To examine the community contexts of MHCs, I used GIS
software to spatially join the coordinates of MHCs to three
sets of geographies:
• To classify MHCs as urban or rural, I joined the
MHC data set to the U.S. Census Bureau TIGER/Line
Shapeile for Urban Areas based on the 2010 census,
which includes dierentiations between small and
Appendix A. Data and Methods
18
FEDERAL RESERVE BANK OF PHILADELPHIA
large urban regions. To account for situations in which
an MHC parcel may be partially included in an urban
area while its associated geographic coordinate falls
outside that area, I added a 500-foot buer around the
urban area shapeile before conducting the join. MHC
coordinates that fell within this buer were classiied
as urban. All MHCs that were not spatially joined to an
urban area were classiied as rural.
• I joined the MHC dataset to the TIGER/Line Shapeile
for 2020 census tracts. I used the 11-digit Federal
Information Processing System (FIPS) codes from this
join to merge in tract-level estimates of demographic
and socioeconomic characteristics from the 2016–
2020 American Community Survey. Each MHC was
retained as a record for the analysis, even if multiple
MHCs were located in the same census tract.
• To incorporate data on residential density, I joined
the MHC locations to the TIGER/Line Shapeile for
2019 census tracts and merged this data set with the
tract-level Housing + Transportation Index created
by the Center for Neighborhood Technology.
32
In this
data set, net residential density is calculated as the
average number of households per residential acre for
census blocks in each tract, weighted by the count of
households in each block.
33
At the time of writing, the
most recent available block-level household counts
were from the 2010 census.
34
Each MHC was retained
as a record for the analysis, even if multiple MHCs were
located in the same census tract.
32
Available at htaindex.cnt.org/. The Center for Neighborhood Technology bears no responsibility for the analyses or interpretations of the data presented here.
33
For more information, see htaindex.cnt.org/about/method-2022.pdf.
34
This analysis was repeated with an alternative measure of residential density (gross household density) from the H+T data set based on 2015–2019 Ameri-
can Community Survey data. The patterns in the results were virtually identical, although actual values were much lower because this measure includes blocks
in which no households are present. I report net residential density because I believe this measure is more intuitive to interpret.
35
For multiparcel MHCs, this is the centroid of the parcel record retained in the inal dataset.
Owing to dierences in the inputs used to construct the
Pennsylvania MHC data set, there are some variations
across records in the location of geographic coordinates
relative to the MHC parcel. Most records were derived
from the CoreLogic dataset (1,881 of the total 2,288), which
provided the coordinates of the parcel centroid.
35
However,
a subset of these records (approximately 300) was missing
coordinates in the CoreLogic dataset. These, as well as
the 262 nonduplicate MHCs incorporated from the PMHA
membership list, were geocoded using the PolicyMap Data
Loader tool, which provided the coordinates associated
with each street address. Similarly, the DHS HIFLD spatial
layer, which accounted for an additional 145 nonduplicate
MHC records, provided coordinates corresponding to
street addresses. Future enhancements to the data set
will standardize these coordinate locations. Since the
vast majority of MHCs were wholly contained within a
single census tract, this variation in coordinate locations
is expected to have minimal impact on the analyses
presented in this report.
Appendix B. Manufactured Housing Communities by County, Size,
and Lot Vacancy Status
19
FEDERAL RESERVE BANK OF PHILADELPHIA
County MHCs
Size Category
Share High
Lot Vacancy
Small
(3-10 Homesites)
Medium
(11-99 Homesites)
Large
(≥100 Homesites)
Adams
19 2 13 4 0%
Allegheny
50 10 32 8 18%
Armstrong
20 8 11 1 5%
Beaver
60 20 36 4 20%
Bedford
33 11 22 0 9%
Berks
66 18 36 12 6%
Blair
43 16 24 3 9%
Bradford
44 12 31 1 20%
Bucks
31 3 17 11 6%
Butler
93 20 61 12 15%
Cambria
31 9 18 4 35%
Cameron
5 2 3 0 40%
Carbon
10 2 6 2 10%
Centre
41 17 21 3 2%
Chester
77 21 46 10 8%
Clarion
29 12 14 3 21%
Clearield
37 14 23 0 22%
Clinton
14 1 11 2 29%
Columbia
48 17 29 2 4%
Crawford
62 23 39 0 19%
Cumberland
64 14 37 13 3%
Dauphin
45 13 26 6 9%
Delaware
5 0 3 2 20%
Elk
6 3 3 0 0%
Erie
82 10 54 18 7%
Fayette
34 8 24 2 12%
Forest
1 0 1 0 0%
Franklin
59 20 31 8 8%
20
FEDERAL RESERVE BANK OF PHILADELPHIA
County MHCs
Size Category
Share High
Lot Vacancy
Small
(3-10 Homesites)
Medium
(11-99 Homesites)
Large
(≥100 Homesites)
Fulton
7 2 5 0 0%
Greene
19 8 9 2 11%
Huntingdon
10 4 6 0 20%
Indiana
26 6 18 2 27%
Jeerson
9 2 7 0 22%
Juniata
12 5 7 0 8%
Lackawanna
34 7 25 2 21%
Lancaster
140 44 82 14 1%
Lawrence
34 2 30 2 18%
Lebanon
37 8 23 6 3%
Lehigh
34 3 23 8 6%
Luzerne
45 7 28 10 22%
Lycoming
51 12 35 4 14%
McKean
15 4 11 0 27%
Mercer
53 10 42 1 30%
Milin
18 6 12 0 11%
Monroe
28 1 26 1 4%
Montgomery
16 1 9 6 13%
Montour
5 1 2 2 0%
Northampton
34 3 25 6 3%
Northumberland
14 2 8 4 14%
Perry
20 6 11 3 5%
Philadelphia
0 0 0 0 n/a
Pike
9 1 8 0 0%
Potter
10 2 8 0 10%
Schuylkill
11 1 8 2 9%
Snyder
9 3 6 0 0%
Somerset
35 9 21 5 3%
Appendix B. Manufactured Housing Communities by County, Size,
and Lot Vacancy Status
21
FEDERAL RESERVE BANK OF PHILADELPHIA
County MHCs
Size Category
Share High
Lot Vacancy
Small
(3-10 Homesites)
Medium
(11-99 Homesites)
Large
(≥100 Homesites)
Sullivan
3 1 2 0 33%
Susquehanna
31 19 12 0 10%
Tioga
52 19 32 1 4%
Union
6 1 4 1 17%
Venango
27 9 16 2 15%
Warren
39 17 22 0 23%
Washington
44 6 33 5 11%
Wayne
21 12 8 1 0%
Westmoreland
97 19 60 18 18%
Wyoming
15 2 13 0 27%
York
109 28 60 21 5%
Appendix B. Manufactured Housing Communities by County, Size,
and Lot Vacancy Status
Notes
An MHC was classiied as high lot vacancy if 30 percent or more of its homesites were vacant in the most recent available aerial image. Percentages were
calculated at the MHC level and are not unit weighted.
Sources
Author’s calculation using the Philadelphia FRB Manufactured Housing Community data set and Census TIGER/Line Shapeiles.
PHILADELPHIAFED.ORG | @PHILADELPHIAFED
