Full text of DOI:10.1177/10780870022184534

Bowman, Pagano / URBAN PUORLBICAINESAFFOFRAIVRASCRAENVTIELWAN/ DMarch 2000
TRANSFORMING
AMERICA’S CITIES
Policies and Conditions of Vacant Land
ANN O’M. BOWMAN
University of South Carolina
MICHAEL A. PAGANO
Miami University
City governments own or regulate vacant land and abandoned structures. In this article, the
authors summarize new vacant-land survey data, examine the conditions and causes of vacant
land, analyze city policy toward vacant land, and explore the possible interconnections among
conditions, causes, and policies. They find that vacant land most often is associated with cities
that have expanded their political boundaries, and the number of abandoned structures is
related to a city’s change in population. Thus vacant land and abandoned structures are not
interchangeable indicators of decay and destruction; rather, they have separate causes and need
different policies.
Land is central to a city. Not only is it the spatial dimension in production,
but it is also a place of recreation and shared community experiences. Its
value is determined by its unique characteristics—namely, it is usually of
fixed supply and cannot be created; it is irreplaceable in terms of its configu-
ration, size, and location; it is immobile; and it has the characteristic of being
permanent in that it is generally indestructible (Kivell 1993; Chisholm and
Kivell 1987). Its value is also determined by its availability, condition,
demand, restricted use (zoning), government subsidy, access, the plans and
visions for it, prestige, symbolism, and social value (Coleman 1982; Kivell
AUTHORS’NOTE: Earlier versions of this article were presented at the Urban Affairs Associa-
tion annual meeting in Ft. Worth, 22-25 April 1998, and at the Who Owns America? II confer-
ence in Madison, Wisconsin, 3-6 June 1998. We thank Rosalind Greenstein for comments on ear-
lier drafts and Richard Forgette for advice on the statistical portion of the study. Thanks are due
to the anonymous UAR reviewers. Research assistance was ably provided by Audrey Harris and
Jennifer Hoffman. We gratefully acknowledge the financial support provided by the Lincoln In-
stitute of Land Policy, Cambridge, Massachusetts.
URBAN AFFAIRS REVIEW, Vol. 35, No. 4, March 2000 559-581
© 2000 Sage Publications, Inc.
559
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

560
URBAN AFFAIRS REVIEW / March 2000
1989; Peirce 1995; Pagano and Bowman 1995; Keuschnigg and Nielsen
1996).
Although a primary assumption in urban land use is that the supply of land
is generally fixed by nature and configured by physical constraints (e.g., land
quality, water availability, etc.) and thereby inelastic, in reality the supply
price can be influenced by government intervention. Publicly owned land
that is not developed—and therefore vacant—may be the result of purposive
public policy for ensuring leisure or recreation space (playgrounds, athletic
fields), protecting and augmenting neighboring land values (parks, gardens),
or directing and guiding economic growth (greenfields). Publicly owned
vacant land may also result from low market prices, undesirability (brown-
fields), or acquisition through take-title seizures (tax delinquencies), among
other reasons (Nicholson 1984; Sinn 1986; Poracsky and Houck 1994; Penn-
sylvania Horticultural Society 1995; Ehrenhalt 1997).
City governments own vacant land and regulate privately held vacant land
through their land-use and zoning powers. Kivell (1993, 4-5) contended that
the public sector is the “second major set of urban land owners”—the first is
private corporations—noting that governmental bodies produce services
(e.g., education, hospitals, roads, housing) and, for equity and efficiency pur-
poses, plan and exert control “over land development and urban expansion.”
Much has been written about the interplay of land, economics, and poli-
tics/policies in the urban setting.1 Despite the centrality of land to a city’s
capital pool, there have been few contemporary and comprehensive analyses
of vacant land (for an exception, see Taylor et al. 1995). To be sure, city-
specific vacant-land inventories are conducted,2 but inventories are not state
mandated and therefore only exist on an ad hoc basis. And although there are
studies of specific land-use topics such as infilling and open space,3 there is
little contemporary empirical evidence to inform our understanding of vacant
land (Manvel 1968; Shomon 1971). In this research, we address the gap in the
knowledge base. In particular, we summarize new vacant-land census data,
examine the conditions and causes of vacant land, analyze city policy toward
vacant land, and explore the possible interconnections among conditions,
causes, and policies.
VACANT LAND: DEFINITIONS AND PERCEPTIONS
The term vacant land is both broad and imprecise. The label is typically
applied to various types of unused or underused land. Northam (1971) identi-
fied five types: (1) remnant parcels that are typically small in size, often
irregular in shape, and that have not been developed in the past; (2) parcels
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

Bowman, Pagano / URBAN POLICIES FOR VACANT LAND 561
with physical limitations, such as steep slope or flood hazard, and thus
unbuildable; (3) corporate reserve parcels held for future expansion or relo-
cation; (4) parcels held for speculation, frequently found in transitional areas;
and (5) institutional reserve parcels set aside by public or quasi-public enti-
ties for future development, given need and funding. Thus, in the literature, it
is not uncommon to find the vacant-land label applied to recently razed land,
perimeter agricultural or uncultivated land, derelict land, land with aban-
doned buildings and structures, brownfields, and greenfields.
Conventionally, the value of vacant land lies in its development potential.
The case of West Palm Beach, Florida is instructive. There, landbanking and
infilling transformed a moribund downtown of empty storefronts into a lively
café-lined Euro-boulevard, complete with a plaza fountain. Vacant land pro-
vided the space and the opportunity for the city government to invest in
design-conscious streetscaping and public improvements, and the commer-
cial and housing markets responded favorably. The success of the initiative
spawned a second effort to convert an additional 75 acres of vacant land in the
downtown area into a $400 million mixed-use development (Flanagan 1997).
The label vacant land frequently carries a negative connotation.4 Associ-
ated images include disinvestment, degradation, blight, and decay (Jakle and
Wilson 1992). For example, Coleman (1982) used phrases such as “dead
space” and “disturbed space” to describe bare derelict land, roughly vege-
tated wasteland, abandoned buildings, and an assortment of various tempo-
rary uses such as materials dumps and construction sites. Perhaps the most
creative label is the acronym TOADS, which refers to temporarily obsolete,
abandoned, or derelict sites (Greenberg, Popper, and West 1990).
Less prevalent are other, more positive images of vacant land. Vacant land
can convey availability, space, opportunity, and informality. For some, vacant
land’s value rests on its nonproductive use, at least as conventionally meas-
ured. That is, its value lies in the indigenous flora and fauna that can be found
in these nonmanicured settings. An urban ecology educational effort spon-
sored by the National Science Foundation in the late 1980s used city parks,
rights of way, and vacant lots as “nature’s classrooms.” Similar thinking is
evident in the Metropolitan Greenspaces Program in Portland, Oregon,
which attempts to shift land-use designations from real estate–generated
labels, such as “vacant” or “undeveloped,” to biologically defined designa-
tions, such as “greenspace” or “greenbelts” (Poracsky and Houck 1994; Ehr-
enhalt 1997). Rather than symbols of urban blight, vacant lots, along with
other open space in a city, can be considered “fortuitous landscapes” (Hough
1994).5 From the perspective of archeologists whose students conducted
fieldwork in 17 vacant lots in Tucson, Arizona, the land represents a distinct
kind of urban space. They argued that vacant lots are not empty in a
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

562
URBAN AFFAIRS REVIEW / March 2000
behavioral sense—that is, they are used for foot travel, refuse disposal, stor-
age, children’s play, and adult activity (Wilk and Schiffer 1979). A similar
theme is echoed by McDonough (1993, 15), who contended that it is through
studies of empty spaces rather than the built environment that one learns
about a city, its culture, and its values. “Whether vacant, reserved, open, or
razed, empty spaces thus play crucial roles in the fabric of the city.”
RESEARCH ON VACANT LAND
National data on urban vacant land have been scant. A 1962 land-use sur-
vey of 48 cities concluded that 23% of all urban land was undeveloped; of the
total, 20.7% was vacant, and 2.3% was under water (Niedercorn and Hearle
1963). By comparing the estimated vacant land in 1962 with the earlier city-
specific studies, it seemed that vacant land declined by nearly 3% and resi-
dential land gained by a nearly equal amount. For the 12 cities in the study
that did not change their boundaries through annexation, the decline in vacant
land was even more dramatic, dropping from 18.4% of total urban land to
12.9% in 1962. The researchers concluded that “vacant urban land in the
larger cities is rapidly disappearing” (Niedercorn and Hearle 1963, 17) and
that cities that had nearly depleted their vacant land had reached the upper
limits for employment and population. A 1968 study found that for cities with
populations greater than 250,000, the amount of undeveloped land was
12.5% (Manvel 1968). The median amount of undeveloped privately held
land in those large cities was only 119 acres.
In the most recent national study, Northam (1971) estimated that among
his 86 study cities, only 75% to 80% of the total amount of vacant land was
“buildable,” meaning that the natural environment (i.e., slope, subsurface
conditions, flood hazard) or societal/institutional concerns (e.g., remnant
parcels, geometric shape, demand for “open space”) effectively prevented
buildings from being erected on those sites.
Subsequent work on vacant land has focused on specific questions
(Greenberg, Popper, and West 1990), has studied non-U.S. cities (Kivell
1993), has taken a critical theoretical perspective (Jakle and Wilson 1992), or
has been anecdotal (Gonzalez 1993). For example, a study by the Pennsylva-
nia Horticultural Society (1995) highlighted the pioneering efforts of a few
cities in the creative reuse of vacant land, especially urban gardening. The
lack of comprehensive and cross-city comparative data has made it impossi-
ble to test Kivell’s (1993, 175) contention that “derelict and vacant land is a
significant part of the overall land use pattern of most cities and amounts to a
major problem in a number of them.”
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

Bowman, Pagano / URBAN POLICIES FOR VACANT LAND 563
THE STUDY QUESTIONS
Our research provides a comprehensive and comparative analysis of urban
vacant land. We estimate and assess the amount of vacant land and aban-
doned structures in U.S. cities, identify and measure the kinds of vacant
land–related policies city governments have in place, and explore possible
causal factors and conditions related to vacant land. As noted earlier, there
have been no comprehensive, systematic studies of urban vacant land in more
than a quarter century. The first two hypotheses that guide our inquiry are
drawn from that now-dated work.
Hypothesis 1: Assuming that the trend identified in Niedercorn and Hearle (1963)
and Manvel (1968) has held, it is expected that the aggregate amount of urban
vacant land has declined over time. The aggregate analysis should show that
the supply of urban vacant land is less now than in the decade of the 1960s.
Hypothesis 2: Northam (1971) found that northeastern cities have less vacant land
than cities in other sections of the country. Therefore, cities located in the so-
called “frostbelt” region of the nation should have proportionately less vacant
land than their sunbelt counterparts.
The last three hypotheses reflect more contemporary trends of population
change, economic health, and policy making in metropolitan America.
Hypothesis 3: An inverse relationship is expected between population change and
the amount of vacant land in a city. As a city’s population grows, its vacant-land
stock should decrease.
Hypothesis 4: An inverse relationship is expected between economic growth and
the amount of vacant land in a city. As Niedercorn and Hearle (1963) found, an
improving local economy should reduce the vacant-land stock.
Hypothesis 5: A positive relationship is expected between the amount of vacant
land and city policy activism. Given that vacant land represents an unrealized
resource, city officials should respond to unused capacity with policies de-
signed to encourage the use and reuse of vacant land.
AN EMPIRICAL ANALYSIS OF URBAN VACANT LAND
Between December 1997 and January 1998, a survey was mailed to city
officials in U.S. cities with 1990 populations of 100,000 or more.6 To mini-
mize the likelihood of definitional disparities, a definition of vacant land was
printed on the questionnaire. It read as follows: “Vacant land includes not
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

564
URBAN AFFAIRS REVIEW / March 2000
only publicly-owned and privately-owned unused or abandoned land or land
that once had structures on it, but also the land that supports structures that
have been abandoned, derelict, boarded up, partially destroyed, or razed.”
Although the questions were structured, respondents were provided ample
opportunity to explain their answers as well as to offer caveats and clarifica-
tions. Planning officials in 99 cities returned the survey for an overall
response rate of 50.25%. The average 1997 population of responding cities
was 346,639, with a median of 167,535. Among the respondents are 6 cities
with populations of one million or more, including New York City. An
examination of the response patterns revealed no discernible over- or under-
representation of particular regions or population sizes.7
THE AMOUNT OF VACANT LAND
On average, the territorial size of the responding cities is approximately
66,500 acres (roughly 103 square miles). The land areas vary greatly—the
median city size is 39,100 acres (61 square miles). In the survey, city officials
were asked to estimate the amount of usable vacant land (thus excluding
undevelopable land such as streets, rights of way, submerged land, wetlands,
etc.) within their corporate boundaries. On average, cities have 12,309 acres
of usable vacant land, although the figures vary considerably. (The standard
deviation of 21,057 suggests the presence of a subset of cities with large
amounts of vacant land. For example, Phoenix reports approximately
128,000 acres of vacant land; Fort Worth, 83,000 acres; and Nashville,
82,000 acres.) The median amount of usable vacant land is a more modest
4,857 acres.
To make meaningful comparisons across cities, the vacant-land figure was
converted to a proportion of the city’s total land area. This variable, the pro-
portion of usable vacant land to total land area, provides a better indicator of
the magnitude of the vacant-land situation in a given city than does the abso-
lute “vacant-land acreage” figure. The average proportion of vacant land to
total land area for the 71 cities that provided data is .181 (or 18.1%), with a
median of .13—proportions that are slightly less than those reported more
than 35 years ago (Niedercorn and Hearle, 1963). (The standard deviation for
the proportion of vacant land to total land area is .163.) Niedercorn and
Hearle found that for the 12 reporting cities that had not expanded their
boundaries since 1962, the average proportion of vacant land to total city land
was 12.9%, which is nearly identical to the .131 average from the 20 cities
that had not expanded their boundaries between 1980 and 1997. Finally, cit-
ies with populations over 250,000 presented data indicating that the
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

Bowman, Pagano / URBAN POLICIES FOR VACANT LAND 565
proportion of vacant land was .154, higher than the .125 reported in Manvel’s
(1968) study.
Niedercorn and Hearle (1963) study of 48 cities
Survey sample: 71 cities
Cities with fixed boundaries (1963), 12 cities
Survey sample: 20 cities with fixed boundaries
Cities > 250,000 population (Manvel 1968), 40 cities
Survey sample: 20 cities > 250,000 population
20.7%
18.1%
12.9%
13.1%
12.5%
15.4%
The first hypothesis—that the amount of vacant land is declining—is gen-
erally borne out for the full sample. For particular subsets of cities within the
sample, those with fixed boundaries and those with populations greater than
250,000, the pattern is different: an increase in the proportionate amount of
vacant land for the large cities and no increase for the fixed-boundary cities.
(The percentage increase for large cities is particularly high: 23.2%. Thus, for
those cities, clearly the hypothesis is disconfirmed.)
The earlierstudies did not focus on abandoned structures asan identifiable
component of the vacant-land situation. The survey instrument allowed us to
do so. For the 61 cities that provided data, the average number of abandoned
structures is 1,312 (with a standard deviation of 4,180) and a median of 150.
(The range is rather vast—from 0 to 28,000 abandoned structures [Philadel-
phia].) Cities with large numbers of abandoned structures were the nation’s
largest cities, including Baltimore with 15,000, Detroit with 10,000, and
Kansas City with 5,000. To facilitate cross-city comparisons, the number of
abandoned structures was standardized by population. The average number
of abandoned structures per 1,000 population is 2.42, with a standard devia-
tion of 4.25.
VACANT-LAND CONDITIONS AND TRENDS
The survey listed a set of six general “conditions” that might describe
vacant land, in addition to an “other” category. Those categories were the fol-
lowing: (1) in oversupply, (2) in undersupply, (3) vacant too long, (4) in the
wrong location, (5) odd-shaped parcels, and (6) not assembled in sufficiently
large parcels. Respondents were asked to choose as many categories as
applied to their vacant-land situation. The data suggest that the size, shape,
and location of vacant land remain the most vexing conditions. Of the 91 cit-
ies that responded to this question, the most universal assessment (51 cities)
was that vacant land was not assembled in sufficiently large parcels. Officials
in 41 cities indicated that vacant land tended to be in odd-shaped parcels. A
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

566
URBAN AFFAIRS REVIEW / March 2000
similar number of respondents (40) noted that their stock of vacant land was
in the wrong location. These three characteristics individually, but especially
in concert, limit the redevelopment potential of vacant land. In 25 cities, the
issue is temporal—that is, land has been vacant too long. In line with earlier
research, notably Niedercorn and Hearle (1963), more city officials (29)
characterized their city as having an “undersupply” of vacant land than an
“oversupply” (23). Officials in 26 cities selected the “other” category and
described their vacant land variously. The most prevalent of these “other”
conditions included the holding of vacant land for speculative purposes
(identified by 5 respondents), undevelopable steep slopes (3), the presence of
brownfields (2), the existence of infrastructure problems (2), and classifica-
tion of vacant land as wetlands (2).
The survey question about conditions focused on the current period. In an
effort to uncover trends, city officials were asked whether the amount of
vacant land in their city had increased or decreased over the past decade. Fur-
thermore, they were asked to select from among a list of possible factors
those to which they would attribute the increase/decrease trend. The catego-
ries and frequencies appear in Figure 1. Increases in the amount of vacant
land were attributed to several plausibly interrelated factors. Two attributes,
disinvestment and suburbanization, were the most frequently cited causes.
They were followed by deindustrialization, contamination of land/structures,
and population out-migration. These attributes present a picture of an erod-
ing city, one in which larger forces are having a very localized effect—an
increase in the amount of vacant land. Substantially fewer cities cited a more
positive cause—annexation—as the reason for an increasing stock of vacant
land. City policies were seldom identified as proximate causes for growth in
the supply of vacant land.
For cities in which the amount of vacant land has decreased, one factor
stands out: a growing local economy (see Figure 2). It is the dominant and
unsurprising attribute of a city that is lowering its vacant-land supply. Also
highly cited as attributes are two seemingly related factors: an in-migration of
population and private development initiatives. Two city policies—land-
reuse policies such as infill and land-use policiesin general—are identified as
important factors in more than half of the cities that reported a decreasing
vacant-land trend.
CITY VACANT-LAND POLICIES
As noted earlier, there have been few empirical cross-city analyses of gov-
ernment policy toward vacant land. Yet city governments design the land-use
policies under which private-sector actors operate. A firm’s decision to invest
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

Bowman, Pagano / URBAN POLICIES FOR VACANT LAND 567
Figure 1: Causes of Increased Vacant Land during the Past Decade
or to disinvest (for our purposes, to reduce the stock of vacant land in a city or
to increase it) is structured, in part, by city policy. Thus, through its policies, a
city government action influences the use of privately held vacant land. Some
cities may adopt an aggressively interventionist posture; others may tend
toward a laissez-faire style. Furthermore, city governments have a direct
impact on local land-use patterns in the manner in which they deal with city-
owned vacant land. Some cities may consider publicly owned vacant land a
nuisance to be disposed of as quickly as possible; others may consider it a key
part of a future cityscape.
Cities’vacant-land and abandoned-structures policies can be classified in
two categories: (1) those that regulate privately held vacant land and (2) those
that affect the use or reuse of city-owned vacant land. In the survey, several
broad policy instruments were included for each of the categories.
Policies for Privately Held Vacant Land
A count of city policy adoptions for the regulation of privately held vacant
land appears in Table 1. Four vacant-land policies tend to be widespread in
American cities: (1) the use of a razing fee, whereby the city recovers the cost
of demolishing an abandoned structure; (2) the adoption of policies for unoc-
cupied residential structures predicated on health and safety concerns; (3) the
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

568
URBAN AFFAIRS REVIEW / March 2000
Figure 2: Causes of Decreased Vacant Land during the Past Decade
adoption of policies for unoccupied commercial or industrial structures
predicated on health and safety concerns; and (4) the creation of special pro-
grams to encourage the use or reuse of vacant land or structures (nearly one-
third of these involve the regulation of illegal dumping on vacant parcels). In
addition, slightly more than half of the cities fine owners of unoccupied resi-
dential or commercial/industrial structures on a periodic basis, such as every
90 days. New York City, for example, can fine owners of abandoned struc-
tures up to $2,000; in Kansas City, Missouri, the fine is capped at $500; and in
San Diego, it is $250. Several cities are authorized to imprison the owners of
abandoned structures for health and safety violations. For example, violators
in Columbus, Ohio, and Charlotte, North Carolina, can be sentenced for up to
30 days.
However, few cities assess other financial penalties. Foreclosures, aban-
donment taxes, and registration fees are used by only a small proportion of
the responding cities. (In nearly half the jurisdictions, according to survey
respondents, counties handle foreclosures.) Louisville, Kentucky, for exam-
ple, charges owners of unoccupied residential and commercial/industrial
structures an abandonment tax that is 1.5 times the standard property tax. Salt
Lake City, Utah, charges owners of abandoned commercial structures a $200
mitigation fee and requires a minimum $385 “registration fee” for owners of
unoccupied commercial or industrial structures.
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

Bowman, Pagano / URBAN POLICIES FOR VACANT LAND 569
TABLE 1: Policies Governing Privately Held Vacant Land and Abandoned
Structures
Number
Policy Type
81
65
City requires owners of unoccupied structures to compensate the city for the cost of
demolishing or razing.
City has other special programs to encourage the use or reuse of privately held
vacant land or abandoned structures.
76
City has a policy governing unoccupied commercial or industrial structures that is
premised on the city’s need to protect the citizens’ “health and safety.”
City fines owners of unoccupied commercial/industrial structures.
City forecloses on commercial/industrial property that is in tax arrears.
City imposes an “abandonment tax” on commercial/industrial structures.
City imposes a “registration fee” on owners of unoccupied commercial or industrial
structures.
50
28
16
4
81
City has a policy governing unoccupied residential structures that is premised on
the city’s need to protect the citizens’ “health and safety.”
City fines owners of unoccupied residential structures.
City forecloses on residential property that is in tax arrears.
City imposes an “abandonment tax” on residential structures.
51
28
14
Some cities will also foreclose on property in tax arrears if it has been
abandoned for a certain period of time. New York will foreclose after only
one year (for commercial/industrial property tax delinquencies). Louisville,
Kentucky, and Nashville, Tennessee, will foreclose on commercial/industrial
and residential property after two years of abandonment; Detroit, Michigan,
after three years; and Atlanta, Georgia, five years.
Policies for City-Owned Vacant Land
Cities own a substantial amount of vacant land and a large number of
buildings, often as a result of tax foreclosures. Of the 99 responding cities, 58
and 54 provided information about their ownership of vacant lots and aban-
doned structures, respectively. The average number of vacant lots owned by
cities is 1,069, but the median is only 50 (the range is 0 through 3,000). The
average number of abandoned structures owned by cities is 137, but with a
median of less than 10—most cities own few abandoned structures. For some
cities, nearly all of the city-owned vacant land and abandoned structures were
acquired through tax foreclosures, although officials in many cities claim that
they never acquire land through foreclosures. The average assessed value of
city-owned land and abandoned structures that were acquired through
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

570
URBAN AFFAIRS REVIEW / March 2000
foreclosures is $187,811 (based on responses from 33 cities). This represents
less than 1% of the assessed value of all city-owned vacant land and aban-
doned structures, which averages $8.3 million per city (according to
responses from 23 cities).
As a consequence of owning vacant land and buildings, cities design poli-
cies for the purpose of encouraging use or reuse of city-owned vacant land
and abandoned structures. Of the 99 city officials responding, 78% indicated
that their city had policies for disposing of vacant land it owned. More than
60% of the city governments have created special programs to encourage
neighborhood or community participation in addressing problems associated
with vacant land and abandoned structures. The treatment of city-owned
vacant land varies, but the standard method for disposing of it is an auction
sale. Respondents in 16 cities have adopted policies allowing the disposition
of city-owned land or buildings below fair-market value; 31 cities offer other
subsidies or inducements as an incentive to private individuals to purchase
city-owned vacant land. The location of vacant land is frequently an issue: 17
cities have an “infill” policy that is different from the policy governing the
disposition of vacant land at the city’s fringe. For instance, Columbus has cre-
ated a land-reuse program that has the authority to sell city-acquired land at a
price that the city believes will induce infill and rehabilitation. Many Califor-
nia cities have created a Redevelopment Agency for the purposes of inducing
growth in the cities’centers rather than at their peripheries. Phoenix sponsors
an infill housing program for single-family houses and provides various fee
waivers and expedited review of plans. The city also offers density and height
bonuses as a tool to redevelop the city’s traditional center.
Eleven cities with policies designed to encourage the reuse of city-owned
vacant land also have developed a separate set of policies specifically
designed for the reuse of abandoned structures. Stockton, California, for
example, offers residential properties to nonprofit housing entities and then
to low- or moderate-income families/individuals as an infill incentive. San
Diego requires owners of abandoned structures to develop a plan to return the
abandoned structure to productive use.
CONDITIONS AND POLICIES
To determine whether two of the vacant-land conditions identified by
respondents—oversupply and undersupply—are associated with particular
city policies, bivariate correlations were calculated.8 Cities experiencing
these conditions may use regulatory policies as a means of encouraging the
reuse of abandoned structures or vacant land. Moreover, these cities might be
characterized by the imposition of financial penalties on owners of vacant
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

Bowman, Pagano / URBAN POLICIES FOR VACANT LAND 571
land. By the same token, the use of those policies may be related to the
vacant-land conditions the city is experiencing. Indeed, the analysis of condi-
tions and policies produced a few significant and interesting relationships.
Among responding cities, an undersupply of vacant land is inversely cor-
related with city foreclosure policies (φ = –.22 for commercial property, φ =
–.27 for residential property, both significant at p < .05). These counterintui-
tive findings suggest that undersupplied cities are not employing a policy for
the purpose of expediting the reuse of vacant land. Cities with an undersupply
of vacant land are also unlikely to impose razing fees on property owners (φ =
–.30, p < .01). Cities with an oversupply of vacant land are more likely to have
adopted razing fees that reimburse the city for the cost of demolition (φ = .22,
p < .05). On balance, however, few of the conditions and policies are corre-
lated, suggesting that other variables are more likely associated with vacant-
land characteristics.
EXPLAINING VACANT LAND AND
POLICY ACTIVITY IN U.S. CITIES
In the survey, city officials offered their assessments of the macrolevel
conditions that may have influenced the supply of vacant land in their cities
(see Figures 1 and 2). However, other microlevel variables may account for
the amount of vacant land and abandoned structures in a city. With regard to
the adoption of specific public policies and possible associations with land
conditions, the correlational analysis identified few statistically significant
relationships. Other variables need to be tested to uncover explanations for
policy behavior in a city, especially regarding the regulation of privately held
vacant land. The next section identifies six structural characteristics that have
theoretical appeal as possible explanations for a city’s vacant-land situation
and its policy behavior. The intent is to design a parsimonious model that
explains the proportionate amount of vacant land (and abandoned structures)
and city policy adoptions.
VARIABLES AND MEASUREMENT
As noted in the hypotheses, the variables of region, population change,
and economic growth are likely to affect the amount of vacant land in a city
and the city’s policy adoptions. In addition, two land-related variables—the
change in the city’s territorial size and the elasticity of city boundaries (Rusk
1995)—may also help to explain both the amount of vacant land and a city’s
policy behavior. The inclusion of these variables is based on the expectation
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

572
URBAN AFFAIRS REVIEW / March 2000
that cities with the statutory authority to extend their boundaries most likely
contain more vacant land than cities without this authority. The city’s land
area, then, grows and incorporates undeveloped land, contributing to a larger
ratio of vacant land to total land area compared with cities having land areas
that remain fixed over time. Also included as a possible explanation is a vari-
able that reflects a city’s fiscal situation. In cities plagued by fiscal problems
officials may take a hard look at vacant land for its revenue-generating capac-
ity. This may translate into policy activity.
Independent and dependent variables were created from published na-
tional sources (e.g., U.S. Bureau of the Census) and from responses to the
survey instrument. The six independent variables provide a structural context
for a city government’s vacant-land situation and are operationalized and
coded in the following manner:
•
•
Region is a dummy variable coded 1 if the city is located in the Frostbelt (north-
east or midwest census regions), 0 if Sunbelt.
Change in a city’s population is its change from 1980 to a mid-1990 estimate
or, if the city did not estimate its 1995 population size on the survey, it is the
change in population from 1980 to 1990, according to census data.
•
•
City land area change is the percentage change from 1980 to 1997. Data on city
land area in 1980 are derived from the U.S. Bureau of the Census, and city land
area in 1997 is taken from the survey.
The elasticity of city boundaries refers to a city’s ability to expand its territory
and is taken from Rusk’s (1995) five-part classification.9 Cities that have not
expanded territorially since 1950, or what Rusk labels “zero”-elastic cities, are
coded 1; those that have expanded the most, or what Rusk labels “hyperelastic”
cities, are coded 5. Cities in between the extremes are coded 2, 3, or 4. At the
ends of the spectrum (zero elasticity and hyperelasticity), regional patterning
appears. Hyperelastic cities are found in the southern and western areas of the
United States. Zero-elastic cities are located primarily in the Northeast and
Midwest. In the groupings coded 2, 3, and 4, there is regional mixing.
The economic growth variable is a standard measure of change in a city’s un-
employment rate, which is defined as the difference in a city’s z-score on un-
employment between 1982 and 1991. The data for calculating the z-score are
drawn from census statistics. The national unemployment rate in 1982 (1991)
is subtracted from the city’s unemployment rate in 1982 (1991) and divided by
the standard deviation in the unemployment rate in 1982 (1991) from the re-
sponding cities. Although this results in an imperfect measure (because the
standard deviation is not a statistic from all of the cities in the nation but only
from the responding cities), it is a good proxy. The average unemployment rate
in 1982 was 9.7% for the nation and 9.44% for the responding cities; the aver-
age unemployment rate in 1991 was 6.8% for the nation and 6.44% for the re-
•
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

Bowman, Pagano / URBAN POLICIES FOR VACANT LAND 573
sponding cities. The responding cities’ unemployment rate, then, is slightly
lower than the national average.
•
A city’s fiscal position is measured as the city government’s budgetary ending
balance expressed as a proportion of expenditures for, in most cases, fiscal year
1996. (When FY 1996 data were not available, ending balances from FY 1995
or 1994 were used or the 1996 average ending balance for the city’s population
category.)10
The vacant-land situation (the dependent variable) is captured in five dif-
ferent ways using survey data. Two are land census indicators—the propor-
tion of vacant-land area to total land area and the number of abandoned build-
ings per 1,000 population—and three are policy measures: a commercial/
industrial policy index, a residential policy index, and a financial penalty pol-
icy. As noted earlier, both the vacant-land and the abandoned-structures
measures are standardized, the former by controlling for the geographical
size of a city and the latter by controlling for city population size. Because the
survey included items that referred specifically to abandoned structures, we
are able to capture the vacant-land situation in this way as well.11 The two pol-
icy index measures are additive measures from questions on the survey: the
sum of a city’s commercial/industrial policies (0-7) and the sum of a city’s
residential policies (0-6).12 In each instance, higher index scores reflect
greater city involvement in regulating vacant land. The financial penalty pol-
icy variable is created from three items on the survey. If a city fines owners of
vacant land, levies an abandonment tax on owners of unoccupied buildings,
or charges a fee for the registration of unoccupied structures, it is imposing an
extra financial penalty on owners. (A razing fee is a financial penalty, but as
noted, it is a fairly common policy in cities and therefore not incorporated
into the financial penalty variable.) Cities that use at least one of these finan-
cial penalties are coded 1; those without any of them are coded 0.
CORRELATES OF VACANT LAND AND CITY POLICY
Correlational analysis was conducted to inform the subsequent multivari-
ate analysis. Table 2 presents the bivariate correlations for the city character-
istics and the vacant-land measures. Several characteristics appear to be asso-
ciated with the proportion of vacant-land area to a city’s total land area. With
a correlation coefficient of .437, the elasticity of a city appears to be espe-
cially noteworthy. Cities with more elasticity are cities with higher propor-
tions of vacant land. Also potentially important, albeit with smaller coeffi-
cients, are the land area change, population change, and region. In other
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

TABLE 2: Bivariate Correlations
Proportion of
Number of
Abandoned Structures
Per 1,000 Population
Commercial
Policy Index
(scale of 0-7)
Residential
Policy Index
(scale of 0-6)
Financial Penalty
Policy
(dummy variable)
Vacant-Land Area
to Total Land Area
Land area change, 1980-1995
Significance level
N
.332
.006
68
–.219
.098
58
–.103
.336
90
–.030
.779
90
.029
.781
95
Elasticity index (Rusk 1995)
Significance level
N
.437
.000
65
–.328
.014
56
–.040
.715
87
.036
.738
87
.046
.666
92
Fiscal position, 1996
Significance level
N
.170
.157
71
–.249
.053
61
–.051
.625
93
–.006
.957
93
–.022
.832
99
Population change, 1980-1995
Significance level
N
.277
.021
69
–.376
.003
59
–.233
.025
92
–.209
.046
92
–.128
.210
97
Region
Significance level
N
–.257
.031
71
.257
.046
61
–.001
.994
93
–.037
.727
93
–.053
.600
99
Economic growth, 1982-1991
Significance level
N
.174
.148
71
–.022
.869
61
.034
.748
93
.056
.593
93
.014
.889
99

Bowman, Pagano / URBAN POLICIES FOR VACANT LAND 575
words, the proportionate amount of vacant land is higher in cities with
increasing land area, in sunbelt cities, and where the population hasgrown.
The number of abandoned structures per 1,000 population in a city
appears to involve some of the same variables—notably, population change
(r = –.376), elasticity (–.328), and region (r = .257). However, the direction of
the relationships is different—elasticity and population change are inversely
correlated with the number of abandoned structures, whereas region corre-
lates positively. Higher numbers of abandoned structures are found in cities
with lesser elasticity, lower population growth, and in the Northeast or Mid-
west. Another variable approaching statistical significance is fiscal posi-
tion—cities with lower ending balances appear to have more abandoned
buildings.
Despite the associations of the six variables with vacant land and aban-
doned structures, those same variables appear less useful in understanding
city policy. In the cases of the commercial and residential policy indices, only
one variable produces a statistically significant correlation. Population
change appears negatively associated with policy adoption; that is, cities with
lower rates of population growth have more vacant-land policies on their
books. None of the variables is associated with the financial penalty index.
MULTIVARIATE ANALYSIS
To examine the interactive influence of these variables on vacant land,
abandoned structures, and city policy, multivariate analysis was conducted.
As noted earlier, there is theoretical justification to include the six independ-
ent variables in the vacant-land and abandoned-structures models.13 Some-
what different models were developed for the policy indices. Ordinary least
squares regression was used to test the vacant land, abandoned structures, and
the commercial/industrial policy and residential policy indices. Because of
the dichotomous nature of the financial penalty policy variable, logistic
regression analysis was employed.
Vacant Land and Abandoned Structures
As Table 3 shows, the model for explaining the proportionate amount of
vacant land in a city possessed low explanatory power (adjusted R2 = .16), and
only one variable, the elasticity index, was significant at a conventional level.
Variables that had initially appeared important in the bivariate analysis (land
area change, region, and population change) become inconsequential in the
multivariate analysis. The hypothesized relationships do not emerge. Region
(hypothesis 2) does not affect the amount of vacant land; neither does
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

576
URBAN AFFAIRS REVIEW / March 2000
TABLE 3: Explaining Vacant Land and Abandoned Structures
Vacant-Land Area
to Total Area
Abandoned Structures
Per 1,000 Population
Unstandardized Coefficients
Standard
Unstandardized Coefficients
Standard
B
Error Significance
B
Error Significance
Land area change, 1980-1995
Elasticity index (Rusk 1995)
Fiscal position, 1996
Population change, 1980-1995 0.0001 0.0006
Region –0.0344 0.0469
Economic growth, 1982-1991 0.0159 0.0266
0.0010 0.0008
0.0389 0.0162
–0.1171 0.1651
0.2196
0.0196
0.4809
0.8894
0.4656
0.5535
0.4509
–0.0034 0.0281
–0.6221 0.5325
–4.6383 3.2542
–0.0355 0.0170
0.0904 1.4205
0.6866 0.9562
6.5024 2.0904
0.2452
0.9032
0.2485
0.1605
0.0423
0.9495
0.4762
0.0031
(Constant)
0.0503 0.0663
0.2399
0.1599
R2
Adjusted R2
0.1508
Standard error of the estimate
0.1434
4.0790
F
2.9988
2.5984
Significance
0.0129
0.0292
df
63
54
population change (hypothesis 3) nor economic growth (hypothesis 4). The
model suggests that the more elastic cities have higher proportions of vacant
land vis-à-vis the total land area. The more elastic the city, the more vacant
land it has.
The same independent variables were used to explain the number of aban-
doned structures per 1,000 population (see Table 3). The model explained
15% of the variance, and population change was the only statistically signifi-
cant variable. As hypothesized (hypothesis 3), the lower a city’s population
growth from 1980 to 1995, the higher its number of abandoned buildings per
1,000 population. No other variable—including the hypothesized region
(hypothesis 2) and economic growth (hypothesis 4)—achieved conventional
levels of statistical significance.
Noteworthy in comparing the two models in Table 3 is that the explanation
for the proportionate amount of vacant land is not the same as that for the
number of abandoned buildings. In the analysis, when the focus shifts from
vacant land to abandoned structures, sign changes occur in four of the six
independent variables. Vacant land and abandoned buildings may be related,
but their causal linkages appear different. Any conclusions from this explora-
tory statistical analysis must be tempered by the low number of respondents
and the model’s relatively modest explanatory power; not all respondents
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

Bowman, Pagano / URBAN POLICIES FOR VACANT LAND 577
were able to report the number of abandoned structures or the amount of
vacant land within their territorial limits. Nevertheless, it appears that vacant
land and abandoned structures, although linked in the literature, are not nec-
essarily caused by the same conditions.
City Policies
Multivariate analysis was conducted for the vacant-land policy variables:
commercial policy index, residential policy index, and financial penalty pol-
icy. The explanation for city policy activity is likely to be a function of city
characteristics and a more immediate cause, the amount of vacant land (as set
out in hypothesis 5). Thus, in the policy models, the variables “proportion of
vacant-land area to total land area” and “abandoned structures per 1,000
population” are used as independent variables. These variables, along with
the six city characteristics used in Table 3, were included in regression mod-
els seeking to explain the adoption of policies.
The analysis showed that the selected city characteristics shed very little
light on a city’s policy behavior. None of the variables produces a statistically
significant coefficient, not even the vacant-land and abandoned-structures
variables, thus disconfirming hypothesis 5. None of the three policy models
even approaches modest explanatory power. Part of the failure of the models
to perform satisfactorily may lie in the assumption of a temporal relationship
that may not be accurate for all cities.14 That is, the model’s specification
assumes that the three policy variables occur subsequent in time to the inde-
pendent variables. It is conceivable that a city’s policy adoptions might have
preceded the city’s unemployment shift or its population change. Moreover,
in the survey, some city officials noted that although a policy had been
approved by the city council, it may not have been implemented because of
perceived negative impacts (e.g., tax foreclosures). Thus a policy may be on
the books but not used by the city. These underlying considerations may con-
found the analysis.
VACANT-LAND POLICIES SHOULD
REFLECT DIFFERING SITUATIONS
It is clear that neither the proportionate amount of vacant land nor the
number of abandoned structures is a simple function of region or economic
condition. Far more important in explaining the proportion of a city’s vacant
land to its total area is the elasticity of a city’s boundaries. As for the number
of abandoned structures in a city, the determinant appears to be population
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

578
URBAN AFFAIRS REVIEW / March 2000
change. In other words, cities with the ability to expand their territories are
likely to have more vacant land, and cities lacking that ability tend to have
less vacant land. Moreover, cities that are losing population are likely to have
more abandoned structures, and cities that are gaining population tend to
have fewer abandoned structures. Vacant land as “raw ground,” as it is often
called, presents a different situation than a parcel on which an abandoned
structure sits.
The inability of other theoretically plausible variables to explain vacant
land and abandoned structures is noteworthy. It suggests that sweeping state-
ments about region and vacant land or economic growth and vacant land
should be made cautiously. For example, in the 1963 vacant-land study,
Niedercorn and Hearle (1963) suggested that a city’s employment and popu-
lation potential are constrained by the availability of vacant land, and should
a city’s supply of vacant land shrink below some unspecified optimal amount,
the city’s economic growth rate might be jeopardized. The data from this sur-
vey show no statistically significant relationship between economic condi-
tions and vacant land or abandoned structures. The implications of these eco-
nomic growth findings in light of the earlier study’s conclusions about an
optimal amount of vacant land are in need of further exploration.
City governments must understand the circumstances that have produced
their vacant-land situation and, in response, craft policy solutions that fit
them. The advice “become elastic!” or “stem the population outflow!”
belongs in the “easier said than done” category. That, of course, is one of the
reasons that the inconsequential findings with regard to policy adoptions are
so vexing. There appears to be no simple explanation as to why, for example,
one city imposes a fine on owners of abandoned structures and another does
not.
Vacant land and abandoned structures reflect different situations. To
expanding cities, vacant land represents a resource—a vast supply of poten-
tially developable open space that can be harnessed to pursue a vision for the
city. To depopulating cities, vacant land represents a red flag—an abundance
of boarded-up buildings that signals a neighborhood in decline and a city in
distress. The successful transformation of both kinds of cities depends on
how effectively they manage vacant land.
NOTES
1. See, for example, the collection of readings in Beauregard (1989) or, more recently,
Jewson and MacGregor (1997) and Lauria (1997).
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

Bowman, Pagano / URBAN POLICIES FOR VACANT LAND 579
2. Examples include Carlson and Duffy (1985), Shreeve (1993), and Taylor et al. (1995).
3. Onthe topicsofinfill and open space, see Brett (1982) and Shomon (1971), respectively.
4. In attempting to paint a vivid mental picture of inner-city decline, former Secretary of
Housing and Urban Development Henry Cisneros (1996, 118) spoke of “boarded-up buildings
and trashed vacant lots.”
5. This reoriented, optimistic conceptualization of vacant land is apparent in the title of the
commentary, “Vacant Urban Land—Hidden Treasure?” (Peirce 1995).
6. The 100,000 population threshold acts as a surrogate for city government capacity. The
underlying assumption is that larger cities are more likely to have geographic information sys-
tems that contain data on vacant land.
7. The median population for all cities over 100,000 population in 1996, according to esti-
mates from the U.S. Bureau of the Census, is 172,000. The proportion of all cities over 100,000
that are in each of the census regions is as follows: 34.6% in the West, 34.1% in the South, 19.8%
in the Midwest, and 11.5% in the Northeast. The regional location of the responding cities
closely mirrors the universe: 38.4% in the West, 32.3% in the South, 18.2% in the Midwest, and
11.1% in the Northeast. Nevertheless, because the requirements of the survey were relatively
information intensive, it is possible that cities with readily available vacant-land information
may be overrepresented in the database.
8. It is important to remember that survey respondents identified the conditions that they
considered descriptive of the vacant land in their cities. Our analysis showed that only two of
these conditions—oversupply and undersupply—were significantly correlated with city poli-
cies to encourage the use and reuse of vacant land and abandoned structures.
9. Rusk’s (1995) elasticity measure is calculated by adding a city’s rank based on both its
population density in 1950 and its change in land area between 1950 and 1990 (trebled). All cit-
ies are then placed in quintiles. This methodology was followed for 12 responding cities to the
survey for which Rusk did not calculate an elasticity score.
10. The data are from Pagano (1997).
11. As noted, the definition of vacant land printed on the survey was inclusive of land with
abandoned structures. In the literature, land with abandoned structures is seen as a particularly
pernicious aspect of the vacant-land situation—a condition often referred to as “blight.”
12. The survey identified seven commercial/industrial policies and six residential policies.
The commercial/industrial policies include a policy based on the health and safety of the resi-
dents, tax foreclosure, fines, an abandonment tax, registration fee, a razing fee, and “special”
programs. The residential policies were the same as the seven commercial/industrial policies,
except that the “registration fee” was not included.
13. To ensure that the models were not plagued by extreme multicollinearity among the inde-
pendent variables, diagnostic tests were conducted. Regressing each of the independent vari-
ables on all of the other independent variables is the test recommended by Lewis-Beck (1980).
These tests suggest the absence of extreme multicollinearity among the independent variables in
the models. The highest adjusted R2 value produced in the testing was with the variable popula-
tion change. However, in this case, the R2 of .368 was sufficiently far enough from 1.0 to mini-
mize concerns of extreme multicollinearity. It should be noted that in this test, change in land
area, the elasticity index, and the economic growth variable were significant at the .05 level. Dif-
ferent model specificationsdidnot change the signs or significance of the independent variables.
14. In the interest of space, the tables containing the results for the policy indices have been
eliminated. Interested readers may contact the authors to receive copies of the relevant tables.
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

580
URBAN AFFAIRS REVIEW / March 2000
REFERENCES
Beauregard, R. A., ed. 1989. Economic restructuring and political response. Newbury Park, CA:
Sage.
Brett, D. 1982. Assessing the feasibility of infill development. Urban Land, April, 3-9.
Carlson, C., and R. Duffy. 1985. Cincinnati takes stock of its vacant land. Planning, November,
2-4.
Chisholm, M., andP. Kivell. 1987. Innercity waste land. London: Institute of Economic Affairs.
Cisneros, H. G. 1996. Urban land and the urban prospect. Cityscape 3:115-26.
Coleman, A. 1982. Dead space in the dying inner city. International Journal of Environmental
Studies 19:103-7.
Ehrenhalt, A. 1997. The great wall of Portland. Governing 10:20-24.
Flanagan, B. 1997. Good design creates another Palm Beach story. New York Times, 12 June,
B1, B8.
Gonzalez, D. 1993. Vacant lots, except for red tape. New York Times, 8 October, B1, B7.
Greenberg, M. R., F. J. Popper, and B. M. West. 1990. The TOADS: A new American urban epi-
demic. Urban Affairs Quarterly 25:435-54.
Hough, M. 1994. Design with city nature: An overview of some issues. In The ecological city,
edited by R. H. Platt, R. A. Rowntree, and P. C. Muick, 40-48. Amherst: Univ. of Massachu-
setts Press.
Jakle, J. A., and D. Wilson. 1992. Derelict landscapes: The wasting of America’s built environ-
ment. Savage, MD: Rowman & Littlefield.
Jewson, N., and S. MacGregor, eds. 1997. Transforming cities: Contested governance and new
spatial divisions. London: Routledge Kegan Paul.
Keuschnigg, C., and S. B. Nielsen. 1996. On the phenomenon of vacant land. Canadian Journal
of Economics, April, S534-S540.
Kivell, P. 1989. Vacant urban land: Intervention or the market? The Planner, August, 8-9.
. 1993. Land and the city: Patterns and processes of urban change. London: Routledge
Kegan Paul.
Lauria, M., ed. 1997. Reconstructing urban regime theory: Regulating urban politics in a global
economy. Thousand Oaks, CA: Sage.
Lewis-Beck, M. 1980. Applied regression. Beverly Hills, CA: Sage.
Manvel, A. D. 1968. Land use in 106 large cities. In Three land research studies. Research
Report No. 12. Washington, DC. Prepared for the consideration of the National Commission
on Urban Problems.
McDonough, G. 1993. The geography of emptiness. In The cultural meaning of urban space,
edited by R. Rotenberg and G. McDonough, 3-15. Westport, CT: Bergin & Garvey.
Nicholson, D. J. 1984. The public ownershipof vacant urban land. The Planner 70 (1): 18-20.
Niedercorn, J. H., and E.F.R. Hearle. 1963. Recent land-use trends in forty-eight large American
cities. Memorandum RM-3664-1-FF. Santa Monica, CA: RAND.
Northam, R. M. 1971. Vacant urban land in the American city. Land Economics 47:345-55.
Pagano, M. A. 1997. City fiscal conditions. Washington, DC: National League of Cities.
Pagano, M. A., and A. O’M. Bowman. 1995. Cityscapes and capital: The politics of urban devel-
opment. Baltimore, MD: Johns Hopkins Univ. Press.
Peirce, N. 1995. Vacant urban land—hidden treasure? National Journal, 9 December, 3053.
Pennsylvania Horticultural Society. 1995. Urban vacant land: Issues and recommendations.
Philadelphia: Pennsylvania Horticultural Society.
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014

Bowman, Pagano / URBAN POLICIES FOR VACANT LAND 581
Poracsky, J., and M. C. Houck. 1994. The metropolitan Portland urban natural resource program.
In The ecological city, edited by R. H. Platt, R. A. Rowntree, and P. C. Muick, 251-67.
Amherst: Univ. of Massachusetts Press.
Rusk, D. 1995. Cities without suburbs. 2nd ed. Baltimore, MD: Woodrow Wilson Center Press
distributed by Johns Hopkins Univ. Press.
Shomon, J. J. 1971. Openlandfor urban America. Baltimore, MD: Johns Hopkins Univ. Press.
Shreeve, E. 1993. Marin City USA. Urban Land, July, 34-37.
Sinn, H.-W. 1986. Vacant land and the role of government intervention. Regional Science and
Urban Economics 16:353-85.
Taylor, R., B. A. Koons, E. M. Kurtz, J. R. Greene, and D. D. Perkins. 1995. Street blocks with
more nonresidential land use have more physical deterioration: Evidence from Baltimore
and Philadelphia. Urban Affairs Review 31:120-36.
Wilk, R., and M. B. Schiffer. 1979. The archaeology of vacant lots in Tucson, Arizona. American
Antiquity 44:530-36.
Ann O’M. Bowman is a professor of political science in the Department of Government
and International Studies at the University of South Carolina. She is coauthor of State
and Local Government (4th ed., Houghton Mifflin, 1999). Her current research interests
include interstate relations, world cities, and municipal and metropolitan strategies
designed to encourage use and reuse of vacant land and abandoned structures.
Michael A. Pagano is a professor of political science at Miami University. He has pub-
lished Cityscapes and Capital (Johns Hopkins University Press, 1995) with Ann O’M.
Bowman. His current research interests include a study of municipal and metropolitan
strategies designed to encourage use and reuse of vacant land and abandoned struc-
tures, an assessment of local government financial capacity, and a prospective analysis
on global federalism issues.
Downloaded from uar.sagepub.com at TEXAS SOUTHERN UNIVERSITY on December 16, 2014