New Orleans is a city with a long and storied history that has captured the imagination of many Americans with its rich and vibrant culture. But like many older American cities, it also has faced more than its share of problems – even before the levees broke – stemming from the urban decline of the post-World War II period.
Over the decades, the city has struggled to deal with persistent crime, underperforming schools and unemployment. Though a number of government sponsored and community-based initiatives have been launched over the years (especially after Katrina) to try and make things better, progress has been slow. The blame for many of these program failures has been placed at the feet of city government – which was widely acknowledged as inefficient and unresponsive.
Now the city’s newly elected mayor, Mitch Landrieu, hopes to change all that. Having come into office with a mandate to reform New Orleans’ government, the mayor recently announced plans for the deployment of a new data-driven performance management and accountability system modeled off of Balitmore’s CitiStat program.
Though it sounds like a piece of software or new technology, CitiStat is really a management strategy – one that essentially relies on getting all of the top brass from city government in one room on a regular basis to discuss their performance and develop strategies for making improvements. CitiStat itself is model on CompStat, the system created by the New York Police Department under Police Chief William Bratton which has been credited with helping the department lower crime rates in New York (and has been copied to various degrees in other cities – including New Orleans). CitiStat takes the CompStat model and expands it to the citywide level.
While programs like CitiStat aren’t technologies themselves, they rely on large amounts of data collection and near- or real-time analysis (hence the “Stat” part). Using traditional and geographic information systems, staff can generate a snapshot of the city’s performance in different service areas (such as public safety or sanitation). In cases where the city is performing poorly, municipal managers can be held accountable for developing and implementing a plan to make things right – reporting on their progress at the next meeting. Constant measuring, feedback, and adjustments are the name of the game. The outcome, proponents argue, is more efficient, results-oriented government.
Of course, how much the new mayor’s plans for a data-driven city can improve New Orleans depends on a number of factors. For one, good decision making requires good data, and lots of it. In Baltimore’s case, the CitiStat program went hand in hand with the implementation of a high-quality 311 system that can catalog and track citizen complaints as they’re handled – something New Orleans has struggled to do as of yet. Without upgrades to the city’s IT system (something Mr. Landrieu has already hinted at), the benefits from the new system could be limited.
Another concern will be how open or closed the new system will be to the public. Baltimore’s system was largely closed off, and was later criticized for being susceptible to manipulation for political purposes (though no evidence was found to support the allegations). In contrast, Washington D.C.’s variant – called CapStat – has emphasized sharing municipal data not only with government staff but the public as well (the results can be seen in D.C.’s extensive data catalog). This makes the verification of reported data and analyses possible, and helps create more trust between city leaders, line staff, and the community.
This last point is probably too often overlooked (or simply given lip service) in discussions about data-driven approaches to managing cities. While these technologies do hold promise to help cities run better, government isn’t driven by data alone. If New Orleans manages to turn things around, it’s because people both inside and outside the government were able to work together and create solutions to problems that challenge cities all over the world. In truth, the only thing statistical data tables or maps do is get the conversation started.