About two years ago, amid mass protests directed at police abuses and racial injustice, the city of New Orleans banned the use of facial recognition by law enforcement. And now, amid a spike in crime, the city has reversed the ban.
It isn’t that new evidence has emerged concerning the efficacy of facial recognition technology; rather, the political landscape has changed, and New Orleans may have something valuable to tell us about what role biometric technology can be expected to play in policing cities across America going forward.
When the New Orleans City Council first enacted its ban, violent crime in the city was at a multi-decade low. The ban arrived as part of a larger package aimed at limiting police surveillance, placing restrictions not just on facial recognition but on predictive policing technologies and cell-site simulators, which force mobile devices in a given vicinity to share information with them as though they are cellular towers.
The Council had been working on the municipal ordinance throughout 2020, a year of protest sparked in large part by the killing of George Floyd at the hands of police in May. Amid the Black Lives Matter protests, concerns about racism in policing were surely top of mind for some of the councilors; already at the end of 2019, the National Institute of Standards and Technology (NIST) had confirmed that many facial recognition systems showed disparities in performance across different ethnic groups, which in turn can lead to racialized outcomes in real world deployments.
But the Council’s ban was helped along by duplicity on the part of the New Orleans Police Department. The NOPD had for years denied its use of facial recognition technology, and then admitted to accessing the tech through state and federal partners. The City Council passed the ban a month later.
The Most Murderous City in America
Much has changed since 2020. As awareness has grown concerning the racial disparities in facial recognition systems, some vendors have taken measures to improve their machine learning processes and reduce demographic disparities, often publicizing the results.
There have also been significant changes on the streets of New Orleans. The city’s murder rate has surged in recent years, with the number of homicides jumping from 119 in 2019 to 201 in 2020, and then to 218 last year. Halfway through 2022, New Orleans had logged 145 murders, putting it on track to far exceed last year’s tally and making it the most murderous city in America. Other violent crimes including carjackings and aggravated assaults have also been on the rise.
The situation, and a perception of inaction on the part of elected officials, prompted almost 180 New Orleans businesses and civic organizations to band together as “The Nola Coalition” and call for decisive action from the city’s leadership. Including organizations like the Baptist Community Ministries and the National Association for the Advancement of Colored People, the group wants a mix of solutions that span progressive and conservative talking points. On the one hand, for example, the group wants issues like traffic violations and mental health services to be dealt with by groups other than police; and on the other hand, it argues that police should get better pay incentives tied to performance, and access to surveillance tools like gunshot detection and facial recognition.
A Measured Reversal
On the latter demand, the New Orleans City Council has now delivered. But the ordinance does not give the NOPD unfettered access to facial recognition; a number of safeguards have been implemented.
Police can only use the technology for the investigation of violent crimes, and investigators must first exhaust all other means of identifying an individual before requesting access to facial recognition from a supervisor. If the request is approved, the image to be scanned is sent to the Louisiana State Analytical and Fusion Exchange, where multiple staff members will need to agree before returning a positive match to the police investigators.
Addressing the City Council, one senior police official argued that this process will help to ensure that suspects are identified accurately, and avoid the potential for wrongful convictions.
Will facial recognition actually help New Orleans police to rein in the city’s crime spree? NOPD officials have characterized it as an important investigative tool, and Mayor LaToya Cantrell called its return “a tremendous stride towards greater public safety.” Privacy advocates, meanwhile, remain opposed to the reinstatement of the technology. A representative of the group Eye on Surveillance, for example, characterized the move as a refusal “to deal with root causes” with respect to crime in the city; and one skeptical City Council member, Lesli Harris, decried the NOPD’s lack of data concerning the efficacy of facial recognition in solving crimes.
Crime is a complex, multidimensional issue spanning multiple fields including sociology, economics, politics, health, and even climate, and the question of how best to tackle it is beyond the purview of this column. But with respect to the developments in New Orleans, it’s worth noting that there is some evidence that increasing investment into police resources, particularly in a way that is narrowly focused on solving particular problems, can help to reduce crime; and there is also some evidence that actually solving homicides is an effective way of reducing the overall homicide rate – something that Tom Hargrove of the Murder Accountability Project calls a “homicide bonus”. The NOPD would seem to believe that biometrics can help it to do that.
If facial recognition technology really does prove to be a useful tool to the NOPD in the coming months – and if the safeguards in place prevent any notable abuses – New Orleans could prove to be an example for other cities to follow as they look to technological solutions in the fight against crime, including those jurisdictions that are now rethinking their own restrictions on facial recognition and other law enforcement technologies.
Aug. 10, 2022 – by Alex Perala