“Properties
and people—that’s
the way I think about the two types of data we have here,” Dr. Dennis
Culhane is saying. “There’s property-specific data, and there’s data about
people.”
Culhane,
an associate professor in the School of Social Work, is co-director of
Penn’s Cartographic Modeling Laboratory (CML), which has created a couple
of powerful new computerized mapping programs known generically as Geographic
Information Systems (GIS). The lab’s Web-based programs—the Neighborhood
Information System (NIS) and the Services Utilization Monitoring System
(SUMS)—make it possible to access and analyze vast amounts of data compiled
by different city agencies.
Let’s
say you want to research an abandoned property in the Mantua section of
West Philadelphia. A quick search in the NIS can show you the location
on a detailed map, the owner’s name, taxes owed, liens against it, gas-
and water-shutoffs, amount of frontage, cracks visible from street—even,
in most cases, a digital photograph. That sort of information used to
require visiting about half a dozen city agencies at half a dozen addresses
and asking for help from each of them. (Of the roughly 560,000 parcels
of land that make up the city of Philadelphia, about 23,000 have abandoned
buildings on them, and another 31,000 are vacant lots. Factor in services
and utilities, and the picture gets complicated fast.) Now, those affiliated
with an approved community organization or city agency can just go to
the CML’s Website for the NIS (http:// apollo.gsfa.upenn.edu/ Projects/NIS.asp),
enter their password and start clicking. (The systems are not, at this
point, accessible to the general public, though Culhane says that Philadelphia’s
City Council is in favor of giving the public access to parts of them.)
“It
used to take people weeks to research a single property,” says Culhane.
“Now you can do a property a minute if you want to. It’s a lot more simply
done, and it’s in a spatial context. Before, you would just get a list
of properties, but you wouldn’t necessarily know what’s next to each other,
and you certainly couldn’t tell about houses behind or on the other side
of the street.”
For
those interested in knowing, say, the truancy levels or reading skills
in a given school-feeder area, or the number of low-birth-weight babies,
SUMS can provide that information with a few mouse-clicks, and analyze
it on a graph as well. The practical possibilities of that are almost
incalculable.
Both
programs grew out of Culhane’s scholarly interest in homelessness [“Three
Degrees of Separation,” February 1997] and abandoned housing, the latter
being a sort of leading economic indicator of the former. As he analyzed
the phenomenon of homelessness—using administrative data from city agencies,
which he describes as a “gold mine of information” that didn’t require
much legwork—he became increasingly interested in the houses in which
the homeless had once lived. When he merged their addresses with utility-termination
records, he found that one out of every four homeless people had their
gas turned off before they ended up in the shelter system, and that 10
percent of those who had their gas turned off ended up in the shelter
system. He soon expanded the analysis to other phenomena, such as tax
delinquency.
Oddly
enough, he says, social scientists had not made much use of this sort
of data. “Getting data from the city was always difficult,” he explains.
“Aside from the fact that there were territorial issues and concerns about
letting data out, just the sheer work demand of it was a pain in the butt.
So we began to envision what would be a more systematic way of getting
data from the city that could be available for research—and would have
the city invested in it.”
It
was the new computerized technology that provided the opportunity, he
says: “At the same time that we get these data from the city for research
purposes, we can create an application that runs on the Web that will
enable city agencies to see each other’s data. It has great power and
utility for community groups and the city.” And the result is a “win-win
situation—the University wins, because it gets data for research purposes;
and the city wins, because they benefit directly and immediately from
the fact that these data are integrated and stored in one place.”
“It
is hard to over-value the potential uses of the work of Dennis Culhane
and his colleagues,” says Dr. Lawrence Sherman, director of the Fels Center
of Government and the Albert M. Greenfield Professor of Human Relations.
“It is on the cutting edge of using evidence to make better policy and
to manage operations.”
Last
summer, M. L. Wernecke, the CLM’s co-director, gave a presentation of
the NIS at the Fels Center to members and staff of the Pennsylvania State
Legislature. “They were extremely impressed with it,” says Sherman. “They
immediately saw how useful the system is for a large number of issues
facing state government, including welfare policy and medical care.”
Sherman,
a strong proponent of evidence-based programs [“A Passion for Evidence,”
March/April 2000], notes that as the systems expand to help produce “trend
data,” they will become an “integral part of performance-management systems
for a wide range of public and non-profit community organizations.
“This
is literally using information to make the city safer, attract economic
development, raise property values and save tax revenues through productivity
gains,” Sherman adds. “It is a clear example of the powerful role an engaged
research university can play in revitalizing American cities, as well
as governance in general.”
“We
have a big affordable-housing problem in the city,” says Culhane, “and
one thing we found that’s almost paradoxical is that the homeless tend
to come from neighborhoods not only where there’s the most abandoned housing
but where there is the most crowded housing. You have empty houses next
to crowded houses. And the reason is because people in the crowded houses
don’t have even the income that will enable them to afford the empty house
next door and still be profitable to the landlord. The landlord can’t
make enough money to pay the taxes, and to repair the property, on the
incomes these people can afford for rent.”
He
zooms in on an abandoned house in the 4400 block of Spruce Street. “This
building’s in tough shape,” he says. “Let’s see what [the Water Department]
says about it.” He clicks his mouse. “No services discontinuance; no service
shutoffs. So [the owner] is still paying the water bill.” He clicks again.
“OHCD [Office of Housing and Community Development], in their foot survey
in 1999, said it was vacant; there were no cracks visible from the street;
no visible roof problems; most openings sealed; no visible fire damage;
no apparent illegal activity; house not posted for sale or rent.”
A
few minutes later, Culhane has exited from the NIS program and has clicked
his way over to SUMS, which describes itself as an “online mapping, reporting
and data-analysis tool” that uses a secure Web site to offer data about
Philadelphia children and families. (Culhane is quick to point out that
SUMS offers only “aggregate data about people at a block level,” with
“no people-identification attached to it.”)
The
data comes from the Philadelphia School District; the Department of Human
Services, which handles child-abuse and neglect cases; and the Department
of Health—“the biggest agencies that deal with people in the city,” in
Culhane’s words.
The
project has created a kind of “forum” in which city agencies can work
together and talk to each other. “Those agencies need to plan,” says Culhane.
“Let’s say you’re the School District, and you’re planning a program around
teen mothers. Well, you need to know where the teen births are occurring.
That’s at the health department.” By integrating all that data, he says,
agencies and other users can “look at the different variables, see maps
of those, and do planning that, by its nature, is inter-agency. Because
these agencies are all bumping into each other out there in the neighborhoods
—but they don’t really coordinate their program development.”
Dr.
Richard Gelles, the Joanne and Raymond Welsh Chair of Child Welfare and
Family who serves as co-director of the Center for the Study of Youth
Policy, used SUMS to do an assessment of the programs and services offered
by the Salvation Army of Greater Philadelphia.
“It
was quite valuable in demonstrating where they were very strong and had
programs matched up with community needs and where there were cooperative
opportunities for collaboration with other social-service [programs],”
he says, “and where, in other parts of the city, there were problems where
they had the expertise to provide services but didn’t as yet have a presence.
“I
think it’s an enormously valuable program,” he adds, since it allows social
scientists to “ensure that certain activities are evidence-based and not
based on anecdotes or a fallacious understanding of what the problems
are in a community.”
Culhane
recalls working with the city on after-school programs. “They wanted to
know, not only where are the schools and where are the kids, but where
are the kids who are getting into trouble, and where’s the delinquency,
or the truancy, happening. And then, if you wanted to do specific programming
that related to family stability, you might want to look at the child-welfare
data.
“Imagine
what it used to take to make a map of that,” he says. “Within one agency
alone, it would have taken a long time, and now we can do it immediately.
And then you can actually zoom in—let’s say the Mantua area—and see all
the variability in school-reading performance. Then you can turn on these
things—libraries, recreation centers, schools and parks. You can begin
to see where are the resources relative to what’s happening in terms of
performance in these schools.
“The
Health and Human Service world is really very excited about the SUMS
application,” he says. “We have several cities around the country who
want us to work with them to replicate this project, because it can be
done relatively quickly.”
City
Council is also intrigued. The SUMS’ “Click-Query Matrix” allows users
to choose the geographic level to work with, which includes councilmanic
districts. “This is very powerful for going to city councilpeople,” Culhane
says. “It gives them statistics that they wouldn’t have, and you can view
data elements or create a raw-data map.” It also allows agencies to view
the data in block groups, census tracts, school-feeder areas, health-center
districts and school clusters.
“For
once, now we have the city with a vested interest in sharing data with
us,” says Culhane, “because they can update and keep these things relevant.
And for them, it’s a pretty big change as well, because they’re going
to be sharing their data at a level of geography that they never could
before,” at an astonishingly fast speed. In doing so, it will enable city
agencies to become considerably more productive.
“We
are working with no fewer than 15 different entities in the city,” he
points out. “So in that alone, it’s been a major educational experience.
We dealt with a lot of skepticism and anti-Penn attitudes, both in city
government and among the community groups, that Penn could deliver something
that had a value to the community. Because there was a perception that
Penn was only interested in the data for its own purposes—for publications,
et cetera. So I think this has been a process of trying to reforge lots
of links that were there many years ago between the University and the
city around public policy and city planning.”
