Public CCTV to increase security

Video surveillance cameras to reduce crime: Do they serve or just make us believe they do?

For a few years, the installation of CCTV surveillance cameras by municipalities has become fashionable. But we seldom wonder if so much investment really serves to reduce crime. In many areas of big Latin American cities, it is common to run into CCTV security cameras. Installed at the top of poles or buildings, these devices are the eyes of the Police. They are attributed a deterrent power. No criminal wants to be seen in full action.

In a recent El Comercio note, it is noted that between 2017 and 2018 the number of video surveillance cameras installed by the municipalities increased by 44%. Peru for example in Lima, only Groove, for example, has 551 of these cameras. Ventanilla, a district with much less resources, has 350, even more than Miraflores (304).

But this is not a Lima or Peruvian fashion. Outside of Peru, it is an extended practice in countries such as the United Kingdom and the United States, for example. That was what motivated several investigators to assess the effect reported by 41 rigorous studies on the impact of security cameras on crime.

What they found was positive. Surveillance cameras reduced by 51% the crimes committed in massive car parks (very common in both countries). However, its range of impact is limited. They did not reduce home theft or public transport theft.

Should similar results be found in Peru? There is only one study that has answered a similar question. He focused on the video surveillance cameras of the Cercado de Lima. This study, conducted by Noam López, found much more modest results. For every 30 crimes committed in a month, 3 stopped committing.

Possible problems at big scale CCTV

CCTV systems are an instrument within a major production line, which is to prevent crime. They are not the panacea. On the contrary, being part of a production line, they need the rest of the parts to work perfectly. For this, at least three aspects are needed.

First, they need support. It is useless to have monitoring centers with hundreds of cameras if we depend on the eyes (and attention) of people who monitor what happens in a mass of information from hundreds of cameras. In other countries artificial intelligence has replaced those eyes. It is used to analyze, process, systematize information and even to identify suspicious objects and people. Once done, human intervention becomes easier in the following steps (intervene, stop, etc.).

Second, they require timely police response. And in that, the algorithms will always be defeated by people. The notification of a crime can reach the police automatically and in real time. However, you must have enough patrolmen to quickly go to each emergency. Response time is key and there we have a patrol gap in the country to cover.

Third, it matters that the cameras are strategically located. Several years ago, a citizen security manager told me that the mayors installed what he called "political cameras." They were cameras that were installed just because the neighbors asked for them and not because they were necessary. In other cases, the cameras are installed in sites of little use (dark sites, low crime, etc.).

Fourth, they must be interconnected. This is the big problem of Lima. To put it in some way, the signals of the cameras of each district are not compatible with each other, which prevents having a single video surveillance center for the entire Metropolitan Lima. Fortunately, last May the Municipality of this commune issued an ordinance to generate such interconnection. But it will take a while until all this works as an integral system.

As with many ideas on citizen security, what matters is not only the idea itself, but all the resources around it that make this idea work. Therefore, not every imported recipe will work here. The institutional context of the “importer” country matters and the effectiveness of video surveillance cameras does not escape this