Case Study 39-2: Oakland's Surveillance Ordinance — Community Control in Practice

Case Study 39-2: Oakland's Surveillance Ordinance — Community Control in Practice

Background

In 2020, the city of Oakland, California, enacted what became one of the most comprehensive municipal surveillance governance ordinances in the United States. The Oakland Surveillance Ordinance — formally the Surveillance and Community Safety Ordinance — established a democratic framework for all decisions about government surveillance technology in the city, requiring community deliberation, public transparency, and ongoing accountability before any surveillance technology could be acquired or deployed.

Oakland did not enact this ordinance in a vacuum. It had a recent history of surveillance controversy: a proposed "Domain Awareness Center" (DAC) that would have integrated city-wide surveillance data had been defeated by community organizing in 2014; the Oakland Police Department had faced years of federal oversight for civil rights violations; and the city's demographics — majority nonwhite, with significant Black, Latino, and Asian communities with historical experiences of over-policing — made surveillance governance a live political issue.

The ordinance emerged from years of organizing by the Oakland Privacy Working Group, a coalition of civil liberties organizations, community groups, and technology advocates who had been advocating for democratic oversight of police surveillance since before the DAC controversy. It is a case study in what community-designed surveillance governance looks like in practice.

What the Ordinance Requires

The Oakland Surveillance Ordinance establishes a multi-step governance process for any surveillance technology the city considers acquiring or using:

Step 1: Public Disclosure and Impact Assessment

Before acquiring any surveillance technology, the relevant department must prepare a Surveillance Impact Report (SIR) — Oakland's version of a privacy impact assessment. The SIR must disclose: - What the technology does, including its technical capabilities - What data it collects, retains, and shares - How it will be used and by whom - The technology's accuracy, error rate, and any disparate accuracy rates across demographic groups - Any privacy risks identified and how they will be mitigated - An assessment of the technology's civil liberties implications

The SIR must be posted publicly at least 30 days before any council consideration.

Step 2: Community Input

The 30-day public comment period is not a formality. Oakland's Privacy Advisory Commission — a body of community members appointed by the City Council — reviews the SIR and holds public hearings. Any Oakland resident can submit comments. The Privacy Advisory Commission produces a formal recommendation to the City Council.

Step 3: City Council Approval

The City Council must vote to approve the acquisition. It can approve, reject, or approve with conditions. Past councils have added specific use restrictions, data retention limits, prohibited use categories, and oversight requirements as conditions of approval.

Step 4: Use Policy

Every approved surveillance technology must be governed by a detailed Use Policy — specifying who can use it, under what circumstances, what data is retained and for how long, who has access, and what happens in the event of a suspected policy violation. Use policies are also publicly posted and can be challenged by residents.

Step 5: Annual Reporting

The city must publish an annual report on how each approved surveillance technology was actually used — how many times deployed, in what contexts, what data was collected, how many law enforcement requests were received and complied with. This reporting requirement creates ongoing accountability rather than a one-time gate at acquisition.

What Happened in Practice

The Oakland Surveillance Ordinance has been in effect long enough to observe its operation across a range of surveillance technology decisions. Several cases illustrate how the process works:

The license plate reader renewal: Oakland had long used license plate readers (LPRs) on police vehicles and fixed locations around the city. When it sought to expand the LPR program, the SIR process revealed that the department had not been following its existing LPR use policy — retaining data beyond the authorized period and conducting searches beyond authorized purposes. The community deliberation process resulted in the council conditioning renewal on reforms to the department's LPR data practices and enhanced oversight.

The facial recognition question: When Axon (which manufactures police body cameras) announced it was adding facial recognition to its body camera software, Oakland's ordinance required the police department to notify the council before any use — because Axon's cameras were already approved but the facial recognition feature would be a new capability. The council voted to prohibit the use of the facial recognition feature, making Oakland one of the first U.S. cities to restrict body camera facial recognition.

The ShotSpotter debate: Oakland had used ShotSpotter acoustic surveillance since 2006. When the contract came up for renewal under the ordinance's framework, the SIR process provided the first comprehensive public accounting of how the system had been used, its alert rate, and how many alerts had resulted in confirmed gun crime. Community advocates used the data to argue that ShotSpotter's performance did not justify its cost and civil liberties impact. After extensive deliberation, the council voted not to renew the ShotSpotter contract in 2023.

Limitations and Criticisms

The Oakland Surveillance Ordinance has been credited with enabling meaningful community participation in surveillance governance. It has also attracted criticism and revealed limitations:

Speed and burden: The multi-step process takes time and places administrative burden on city departments. Police officials have argued that the process moves too slowly to respond to emerging public safety needs. Some surveillance technology contracts have lapsed or been delayed because the approval process was not completed in time.

Regulatory capture risk: The Privacy Advisory Commission is appointed by the City Council, and its composition reflects council politics. Critics have raised concerns that the Commission could be staffed in ways that make it less independent than the ordinance envisions.

Scope limitations: The ordinance covers city government surveillance. It does not cover surveillance by federal agencies operating in Oakland, surveillance by private actors (Ring neighbors, commercial data brokers), or surveillance by state agencies. City residents are subject to extensive surveillance that the ordinance cannot reach.

Enforcement gaps: When departments have violated use policies, the consequences have been administrative rather than legal. There is no private right of action under the ordinance; residents cannot sue the city for surveillance violations. Enforcement depends on political accountability rather than legal accountability.

Despite these limitations, Oakland's ordinance has served as a model. Versions of the CCOPS framework have been adopted in more than 20 cities. The National League of Cities has published guidance on municipal surveillance governance. The Oakland experience demonstrates both the potential and the limitations of democratic governance as a surveillance accountability mechanism.

What Oakland Teaches

The Oakland case teaches several things that are directly relevant to Chapter 39's analysis:

1. Community deliberation changes outcomes. The ShotSpotter non-renewal is the clearest example: without the ordinance's requirement of a Surveillance Impact Report and public deliberation, the contract would likely have been renewed as a routine administrative matter. The deliberation process produced a different outcome.

2. Transparency enables accountability. The revelation during the license plate reader renewal that the department had not been following its own use policy was a consequence of the SIR process — the department was required to document its actual practices, which revealed departures from its authorized practices. This would not have been visible without the transparency requirement.

3. Governance is ongoing, not one-time. The annual reporting requirement creates continuous accountability. It means that the political costs of surveillance violations accumulate over time, rather than dissipating once initial approval is granted.

4. Local governance has limits. The federal, state, and private surveillance that the Oakland ordinance cannot reach limits its effectiveness as a comprehensive response to the surveillance landscape. Local democratic governance is necessary but not sufficient; it requires federal and state frameworks to address the surveillance actors it cannot reach.

Discussion Questions

The Oakland Surveillance Ordinance requires democratic deliberation before surveillance technology acquisition. Is democratic deliberation an adequate governance mechanism for surveillance decisions, or do some surveillance decisions require technical expertise that community deliberation cannot adequately evaluate?
The ShotSpotter non-renewal resulted from the deliberation process producing data about the system's actual performance that was used by community advocates in the political process. Is this the kind of accountability the ordinance was designed to provide? What else would adequate accountability look like?
The ordinance's scope is limited to city government surveillance. How significant a limitation is this in practice? What would need to be in place at other levels to address the surveillance the ordinance cannot reach?
Police officials have argued that the ordinance's deliberation process is too slow to respond to emerging public safety needs. Evaluate this argument. Is speed a genuine requirement that the ordinance should accommodate? Or is the requirement for deliberation itself a feature, not a bug?
If you were designing a surveillance governance ordinance for your own municipality or campus, what elements would you include from Oakland's model, and what would you modify?