Breaking the Divestment Cycle: Predicting Abandonment and Fostering Neighborhood Revitalization in Baltimore
Since the 1970’s, a certain urban pessimism has pervaded both academic research and public policy on Baltimore. Nothing, it seemed, could be done to stem the tide of divestment. Policy conversations focused on ‘right-sizing’ and even ‘mothballing’ the urban landscape.
Recently however, there has been a groundswell of enthusiasm for the potential of American cities in general, and Baltimore city in particular. After decades of naysaying, policymakers have begun to realize that urban revitalization is possible. The city has seen a profound shift, away from downtown redevelopment and urban renewal to neighborhood based efforts and infill redevelopment that has shown significant promise.
Despite this optimism, there remain significant gaps in our knowledge base. Due primarily to data limitations, the dynamics of divestment and reinvestment are not well understood. Surprisingly little is known, for example, about when and why particular properties are abandoned and the degree to which abandonment has a contagious effect on adjacent properties.
In Baltimore, these questions take on particular urgency. Not only has decades of decline resulted in 17,000 vacant and abandoned properties, but the nature of Baltimore’s housing stock – primarily brick rowhomes – means that costs of revitalization, stabilization, and demolition are high. By modeling urban divestment and investment, we hope to not only expand our knowledge-base, but to aid the City in effective and efficient program implementation.
To close the gap, we created a unique database
With seed funding from the Institute of Data-Intensive Engineering and Science (IDIES) we created a rich parcel-level longitudinal database of Baltimore city’s housing stock merging administrative data provided by the Housing Department of Baltimore City. This data allows us not only to consider the efficacy of various housing interventions, such as the city’s Vacants to Value initiative, but also to address larger urban questions – examining empirically the mechanisms of property divestment, neighborhood decline, and renewal.
The 21st Century Cities Initiative at the Johns Hopkins University recently provided funding to continue the development of this unique resource and to apply this data to a wide range of urban questions. Researchers from JHU’s Applied Mathematics & Statistics and Sociology departments have come together with the Office of Code Enforcement of Baltimore City to tackle this interdisciplinary project.
In the spirit of Jim Gray’s “20 questions”, a set of key problems were identified early on that set the long term goals and spell out the immediate needs to drive our research directions.
Phase 1 – Predicting Abandonment Investment
Urban researchers have long hypothesized that property abandonment spreads through a city following a contagion logic: the abandonment of one property in a neighborhood increases the likelihood that proximal properties will also be abandoned (over and above the neighborhood’s market position). Modeling this dynamic process, however, is far more difficult and requires longitudinal parcel-level data that until recently has been unavailable.
Our approach builds on the previously developed custom database of the geometries of all parcels and buildings in Baltimore, to which new layers of pertinent information can be joined for exploring vacant housing dynamics. Importantly the use of water usage data has allowed us to create a proxy for building occupancy – a key missing piece in modeling abandonment.
Phase 2 – Evaluating Policy
Baltimore city’s innovative blight remediation program, Vacants to Value (V2V), was designed to facilitate transfer of blighted properties back into productive use via receivership and disposition. V2V appears to have had a positive effect in certain areas, but separating the effects of the initiative from larger population and housing market trends remains a nontrivial problem. Urban researchers often struggle to employ rigorous statistical analyses in the context of neighborhood revitalization in part because the level of spatial aggregation available to them (the census tract or neighborhood) is generally too coarse to make plausible comparisons. Our data, in contrast, have the advantage of being scale-independent and measured consistently through time, allowing us to employ more robust matching techniques to better approximate causal inference.
Phase 3 – Modeling Neighborhood Revitalization
The foundation we are laying will allow our inquiry to go further than these two phases. As the project progresses, the dataset will allow us to model all manner of urban transitions, not just measure the effects of the V2V interventions.
Recently Baltimore suffered a fatality caused by a collapsing vacant house. Our team helped to expedite the emergency inspections prompted by this tragedy.
Building on our existing data solution, we integrated building footprint data to look at the divestment from a completely new angle. Using the footprint data, we divided the city’s blocks of rowhomes into contiguous “sub-blockfaces,” accounting for alleys and previous demolitions which leave spaces between rowhomes on the same blocks. The ends of these sub-blockfaces have the highest likelihood of collapse, and were not flagged in any existing database. We visually checked every high-risk building against the ConnectExplorer aerial photography. Over the first weekend, 80 houses were identified as imminent threats and were subsequently demolished.
Such emergency demolitions are like putting out fires: they have to be done and done fast. Baltimore’s Planning and Housing Departments are also at the forefront of systematically scheduling demos across the city. When earlier this year, Maryland’s Governor Larry Hogan announced Project C.O.R.E., a new program to address blight, Baltimore City’s strategic demolition program schedule received a welcomed boost.
In light of the new funding, Baltimore City began carefully selecting the next round of vacant houses for demolition. We were able to streamline this process by identifying groups of adjacent properties that were entirely abandoned — an ideal demolition scenario that increases cost effectiveness and requires no residential relocation. We identified these properties using a sophisticated database query, which folds in all the required constraints and determines the spatial colocation of row houses.
Today, as city officials carefully discussed each property, considering community feedback, architectural preservation, strategic planning, and cost effectiveness, they projected a map containing property information culled from diverse administrative dataset including our identification of standalone clusters.
Of course, the targets identified by our team are only the beginning. All aspects of urban planning need to be considered before tearing down a property. But one building after another, Baltimore City’s strategic demolition program partnering with the Governor’s Project C.O.R.E. will clear the way for new green space, housing and community development.
Meet the Team
Tamás Budavári is Assistant Professor of Applied Mathematics & Statistics in the Whiting School of Engineering at The Johns Hopkins University, where he focuses on mathematical and computational challenges of big data. He is builder of the Sloan Digital Sky Survey and founding editor of the Astronomy & Computing journal.
Philip Garboden is a doctoral student in Sociology and Applied Mathematics & Statistics at The Johns Hopkins University. He holds a masters degree in Public Policy from the same institution. He works as a researcher at the Poverty and Inequality Research Lab. His work focuses on the intersection of housing and urban poverty.
Michael Braverman is Deputy Commissioner for Permits and Code Enforcement at Baltimore Housing. In that role, he oversees the strategic code enforcement piece of Mayor Rawlings-Blake’s Vacants to Value initiative, leading its innovative receivership and targeted demolition programs. Michael has been asked to share his expertise and passion for well-managed, data-driven government with a variety of cities and with organizations including the Federal Reserve Bank Board of Governors, the Center for Community Progress, and the Clinton Global Initiative. He has a J.D from the City University of New York and a B.A. from the Johns Hopkins University.
John David Evans is Director of Analytics for Permits and Code Enforcement at Baltimore Housing. He develops analytic tools for the evaluation and management of housing programs in Baltimore. John David holds a masters degree in Public Policy from the University of Maryland.
Talented undergraduate students Surya Ram (Applied Mathematics / Computer Science) and Kevin Wells (Sociology) are responsible for much for the code development and data management.
Leave your comment below, or reply to others.
Please note that this comment section is for thoughtful, on-topic discussions. Admin approval is required for all comments. Your comment may be edited if it contains grammatical errors. Low effort, self-promotional, or impolite comments will be deleted.
Read more from MeetingoftheMinds.org
Spotlighting innovations in urban sustainability and connected technology
I see the outcomes of Duke Pond as a representation of the importance of the profession of landscape architecture in today’s world. Once obscured by the glaring light and booming voice long-generated by building architects, landscape architects are steadily emerging as the designers needed to tackle complex 21st century problems. As both leaders and collaborators, their work is addressing the effects of rising sea level on coastal cities, creating multi-modal pedestrian and vehicular transportation systems to reduce carbon emissions, reimagining outdated infrastructure as great urban places, and as with the case of Duke Pond, mitigating the impacts of worsening drought.
AI has enormous potential to improve the lives of billions of people living in cities and facing a multitude of challenges. However, a blind focus on the technological issues is not sufficient. We are already starting to see a moderation of the technocentric view of algorithmic salvation in New York City, which is the first city in the world to appoint a chief algorithm officer.
There are 7 primary forces determining the success of AI, of which technology is just one. Cities must realize that AI is not the quick technological fix that vendors sell. Not everything will be improved by creating more algorithms and technical prowess. We need to develop a more holistic approach to implementing AI in cities in order to harness the immense potential. We need to create a way to consider each of the seven forces when cities plan for the use of AI.
In New Zealand, persistent, concentrated advocacy and legal cases advanced by Māori people are inspiring biocentric policies; that is, those which recognize that people and nature, including living and non-living elements, are part of an interconnected whole. Along the way, tribal leaders and advocates are successfully making the case that nature; whole systems of rivers, lakes, forests, mountains, and more, deserves legal standing to ensure its protection. An early legislative “win” granted personhood status to the Te Urewera forest in 2014, which codified into law these moving lines:
“Te Urewera is ancient and enduring, a fortress of nature, alive with history; its scenery is abundant with mystery, adventure, and remote beauty … Te Urewera has an identity in and of itself, inspiring people to commit to its care.”
The Te Urewera Act of 2014 did more than redefine how a forest would be managed, it pushed forward the practical expression of a new policy paradigm.