A study by the Center for Security and Citizenship Studies evaluated the impact of the Sampa Sampa program, monitoring and surveillance through cameras in the streets of the city of São Paulo. According to the study, the program did not significantly impacted public safety rates. 
THE study evaluated three crime indicators (theft, theft and homicides) and two of police productivity (arrests by flagrant and warrant).
According to the analysis, there was no change in the indices in relation to other cities of the state. The study concludes that the program is expensive and ineffective.
Although constantly released by the mayor of Sao Paulo, Ricardo Nunes, the program, which integrates City Hall and private cameras in a network with tens of thousands of observation points, monitored in a third party surveillance center, did not represent perceptible improvement by the calculations in any of the rates.
Evaluated in two methods, different-in-differences and Event-Study, as a way to avoid failures and dubious analysis, the study compared criminal statistics in police stations in regions where there is adherence to the program with those of regions in which this does not occur, used as a control population.
The portrayed rates already had their own, fall or increase dynamics, which were kept in relation to other locations, according to the Study Center.
“The absence of measurable impact, even after a year of operation and with massive use of public resources and security agents, suggests that facial recognition has been more effective as an instrument of political propaganda than as a public policy based on evidence,” the study points out.
According to the survey, there were also cases of documented identification errors, adding the lack of specific regulation and the history of racial bias associated with this type of technology, makes up a picture that imposes serious concerns about the social cost of the program.
“In a budget shortage scenario and precarious public services, it is urgent to reassess priorities: public safety must be thought of from effective policies, based on evidence and respect for fundamental rights, not technological solutions that promote massive surveillance without concrete results”, criticizes Cesec.
City Hall defends the program
Sought by the report, the city manifested through a note, in which it classified the comparisons made in the study as inappropriate and linked to a failure methodology, with incorrect conclusions.
According to the city, the program deals with the largest monitoring system in Latin America, with more than 31 thousand cameras, and its results prove its effectiveness, such as the arrest of almost 3,000 people in the act, the capture of over 1,550 fugitives and the location of about 80 people missing.
“Another failure of the study method is to compare the capital with cities in the interior of the state, which have different social, economic and criminal reality from the capital São Paulo, a metropolis that corresponds to more than 25% of the state’s population and has a GDP higher than that of countries. Scholars also fail by presenting a loaded introduction of politicized statements, without any support in the data analyzed in the study itself or in statistics and technical evidence, “says the note.
The study, available in website From CESEC, it dedicates two attachments to technical issues.
The city also explained that the São Paulo system validates the alerts with evaluation of human agents, and that there is no record of incorrect arrests in situations of use of the system.
The system also advocates the work of the Civil Police, collaborating with 275 occurrences between November 2024 and May this year.
