A man leaves an SUV on Times Square in New York City. It begins to smoke, and an alert vendor and a nearby police officer bring the NYPD bomb squad to the scene. The man who left the vehicle is captured on video changing shirts, checking to see if anyone is watching him make his getaway.

This is one of the few instances in which police surveillance cameras will earn their keep.

When it comes to deterring crime and terrorism, police on the beat are still the sharpest tool we have. The Times Square plot was foiled by an alert person and a prompt police response, not a camera.

This outcome is not unprecedented. An alert border patrol agent stopped the would-be Millenium Bomber at the U.S.-Canada border, sensing something “hinky” about him. His plan was to drive a car full of explosives to the Los Angeles International Airport in a manner parallel to the one stopped this weekend. Her intuition saved lives in an instance where a camera would have failed. A camera would have caught him on tape at the border, and other cameras could have filmed his final approach to the airport terminal, but cameras don’t sense “hinky” like people do.

This fact is often omitted from pushes for broader surveillance systems. Big city mayors have sold greater surveillance as a deterrent to both crime and terrorism.

Since September 11, 2001, major cities have greatly expanded the number of surveillance cameras on public streets. New York has over eight thousand, installed at a cost of $25 million, and Washington, D.C. has over five thousand. London has over a million surveillance cameras with a bill of $400 million to show for it.

Unfortunately, this has not worked as advertised with regard to terrorism. The cameras did nothing to deter this weekend’s attempt in New York. London, the most camera-laden city of the three — probably in the world — suffered commuter train and bus bombings on July 7, 2005 that killed 52 and wounded around 700. The ever-present surveillance cameras got plenty of footage but deterrence was not among their virtues. The recent suicide bombing on Russian subways highlights the fact that if terrorists are willing to die to deliver their deadly payloads, cameras do nothing to stop them. Hopefully, Washington will not soon face a similar threat.

The cameras have likewise proven little deterrent to run-of-the-mill criminal activity. Roughly 80% of the crime in London goes unsolved. The boroughs with the highest clearance rates (still a measly 25%) have fewer-than-average cameras on their streets. One in a thousand of the crimes solved was a surveillance camera success story. Crime in D.C. remains high. The NYPD solved nearly 60% of its 2009 murder cases, down from the year before and a bit below the national average. Chalk success and failure in this field up to police work, not technology. A camera in the right place and viewing at the right angle is helpful, but don’t depend on them in lieu of investigative prowess.

The New York subway system has over 4,000 cameras alone, but only half of them are operational. A recent double homicide highlighted this fact — the killer made his escape at the Christopher Street station, where the cameras that should have given clues to his identity were inoperable. This is not to say that they are useless, as security cameras helped capture the assailant in a savage midtown beating barely a month before.

Terrorism and crime require both deterrence and response. People provide effective deterrence, but cameras aid in the response, in piecing together the plot and tracking down those responsible. Both options cost money, and arriving at the right mix of the two is a tough decision with shrinking municipal budgets. New Yorkers, and everyone else who is told that more cameras will equal more security, should know the facts when deciding on the right balance between cops on the beat and eyes in sky.