New Orleans partnered with tech developers to institute a data-driven blight reduction strategy that decreased the number of decaying houses by more than 15,000 in five years. Its centrepiece was a bespoke tool that converted qualitative property surveys into quantitative scores, advising officials whether a blighted property – a dwelling that has fallen into disrepair – should be demolished or sold. This eliminated a backlog of 1,500 cases. Tackling the city’s decaying housing has been a major priority since 2010, when New Orleans was thought to have more abandoned dwellings than Detroit.
Results & Impact
The scheme reduced the number of blight-affected properties by a third in its first five years. It has also significantly increased the speed of decision-making on abatement measures, clearing a 1,500 property backlog, and reducing the number of property violations (houses that fall short of upkeep criteria) by 6%.
City of New Orleans, Enigma, Code for America, Center for Community Progress, Greater New Orleans Foundation, Enterprise Community Partners, Greater New Orleans Community Data Center, The Reinvestment Fund, The Department of Housing and Urban Development, University of New Orleans, Louisiana Office of Community Development, Propeller, New Orleans Redevelopment Authority, Louisiana Land Trust, Federal Emergency Management Agency
BlightSTAT is a data-driven management process that co-ordinates the efforts of city agencies and public bodies to tackle New Orleans’ blight problem. It does this through regular review meetings to evaluate policy and track progress in meeting the city’s blight reduction goals. Underpinning this process is a management program, Lama, which provides details on the status of each property being investigated. This includes inspection reports, hearing outcomes and abatement actions. A data development company was engaged to create software that could advise officials on whether to sell or demolish a blighted property. This was done by analysing 600 cases and allocating a numerical value to the criteria that underpinned each decision. The data results were analysed to develop a regression model which used property inspectors’ qualitative assessments to calculate, on a 100-point scale, whether a house should be condemned or sold. Residents were also informed of complaints against their property to increase their willingness to improve the standard of their houses.
Cost & Value
By the start of its fourth year, the scheme had helped collect $3.4 million in lien foreclosure payments. The number of hearings that resulted in a definitive judgement increased from 49% in 2011 to 72% in 2013.
Running since 2010
New Orleans has tackled urban blight by creating predictive tool to forecast officials’ decisions on whether to sell or demolish blighted properties, clearing a backlog of over 1500 units.
The Blight Scorecard was part of a broader community engagement strategy to crack down on dilapidated dwellings, launched by Mayor Landrieu on coming to power in 2010. This involved deploying enforcement officers more effectively, making greater use of the state’s power to requisition houses and creating a more coherent management structure.
New Orleans has suffered from urban decay for decades, a trend exacerbated by hurricane Katrina. The flooding that followed affected roughly 80% of city dwellings, a large number of which never recovered. New Orleans has also been afflicted by long term population decline. The city’s population peaked in the 1960s and its steady decline created an abundance of housing stock, much of which has been left empty. In 2010 New Orleans was thought to have more dilapidated houses than Detroit.
Central to tackling the problem was BlightSTAT, a management system that used data to track the progress of the city’s blight crackdown. The concept was to bring together all relevant city agencies and stakeholders in regular review meetings where presentations were made about the successes and limitations of the policies they had been implementing.
“When we started looking at how the city was dealing with the problem we made some rather sobering discoveries,” said Oliver Wise, Director of New Orleans’ Office of Performance and Accountability (OPA). “We had two agencies in charge of blight enforcement efforts. You had one for vacant lots and one for structures and those agencies weren’t co-ordinated… In BlightSTAT, we look at the data to see what’s working, what’s not, what we need to do to improve. Basically, it’s a feedback loop to see if we are attaining our strategic goals.”
Although the management approach changed, the city’s data system initially remained much as it had been. “When we started BlightStat, the software was spit and bubblegum and tape and Microsoft excel spreadsheets,” said Wise.
Two to three years into the initiative, New Orleans brought in Lama, a software management tool that allowed the city to track the activities of all agencies electronically and monitor the position of each property in the enforcement pipeline. This included detailed information on property inspections, hearings and abatement actions, as well as inter-departmental meetings and reports.
Although this data management approach helped greatly speed up the process of identifying and acting against blighted properties, it was New Orleans’ innovative Blight Scorecard tool that was key to increasing the speed of abatement measures.
“We had vastly improved our volume of cases coming through the system, but we hadn’t yet changed how we made that decision to either sell or demolish a property,” said Wise. “And so a formidable backlog, an 18 or 24 month backlog, had accrued on the director’s desk.”
To overcome the 1,500 property backlog, A team from OPA was tasked with finding a way to delegate the decision-making process without affecting the quality of assessments. They did this by analysing more than 600 property resolutions, identifying the central factors that determined whether a house was sold or demolished and allocating them numerical values. The OPA team commissioned Enigma, a data development startup, to conduct tests based on the analysis. This led to the development of a tool that conducted regression analysis to quantify inspection officers’ qualitative assessments: the blight scorecard.
“Basically the scorecard acts as a coin sorter,” said Wise. “So when we have a new guilty judgement on a property, instead of that new case just being tossed into a pile with no organisation, our code sorts them into cases that are likely demolitions and likely foreclosures. And with that coin sorter we go much faster. We have many more properties now that are proceeding down to their fate rather than hanging in blight purgatory.”
The blight scorecard converts inspection officers’ qualitative assessments of properties into numerical scores, between zero and 100, which advise officials on courses of action. A score of zero indicates a house should be demolished while a score of 100 means it should be sold. This has cleared the backlog and, although decisions still have to be formally taken by senior officials, the tool allows lower ranking individuals to make recommendations which are almost always upheld. Although some uncertainty remains because scores can be close to the middle of the spectrum, the tool helps senior officials by showing them which properties can be sorted relatively quickly and which require more time and consideration.
Further efficiency savings have been delivered through a nudge program encouraging homeowners to act if their properties have fallen into disrepair. When civic officials are notified of a blighted property they now send a letter to its owners informing them that neighbours have complained. This strategy has resulted in a 6% decrease in violations in inspected properties, reducing authorities’ workload by the equivalent of one full-time inspector.
The blight reduction program has also benefited from direct engagement with the public. BlightSTAT itself encompasses public meetings. “We had a great deal of public engagement from the get go,” said Wise. “Sometimes north of 100 people attending these meetings at 8 oclock in the morning. I think people were pleasantly surprised to see us taking such a rigorous approach to blight management, but really what they came there for was to find out about that one house next to them.”
To provide that street level transparency, a portal displaying information on the position of dilapidated properties in the enforcement timeline, BlightStatus, was launched in 2012. “It’s basically the Domino Pizza tracker for blight,” said Wise. Users type in the address of the property they are interested in and will receive a breakdown of the house’s position in the enforcement pipeline. This has enabled residents to monitor houses in their local area and provided a scalable answer to the most common request for information authorities received.
BlightSTAT has been nationally recognised for its effectiveness. By the start of its fourth year, in January 2014, it had led to the collection of over $3.4 million in lien foreclosure payments. It was chosen by the Ash Center for Democratic Governance and Innovation at Harvard University as a “Bright Idea in Government” in 2012.
(Picture credit: Flickr/Bart Everson)