Danger, Will Robinson: Can Robots Protect Us from Corruption?

Technology is a frequent recourse for anticorruption advocates, be it in the form of crowdsourced reporting, tree tracking, or drug verification.  To that list, one can now add one of the Democratic Republic of the Congo’s recent initiatives: robots.

We may not yet be at the point where, like something out of a summer blockbuster, robots can chase down offenders or take the lead in corruption investigations. Nevertheless, building upon their earlier efforts in Kinshasha, a group of engineers has recently been hired to install a traffic robot in Lubumbashi. The robots, eight feet tall and looking like something out of the 1960s, have traffic lights embedded in their torsos and are equipped with cameras which allow them to record traffic violations. The theory behind these cybermen? Robots can’t be bribed, thereby circumventing the notoriously corrupt (human) police force of the DRC, whose officers could either baselessly stop drivers and demand money or be bought off by a driver who truly has committed an infraction.

There are many good things about this initiative. Encouraging Congolese startups (and women in business and science–the engineering team that developed the robots is all-female) seem like worthwhile goals. And if people are somehow intimidated into being better drivers, as some Congolese have claimed is occurring, then the DRC’s horrific traffic accident rate may drop. However, are these robots really effective as corruption-fighting tools?

In theory, the robots are a novel spin on the classic technological approach to fighting corruption, because they mix three traits which, though admittedly found in other anticorruption techniques, are not usually combined:

  • Inability to be bribed: By attempting to remove the human element from traffic interactions, the robot’s creators have developed another way of sidestepping systemic government corruption. (However, humans aren’t completely taken out of the equation. The footage of traffic violations is sent to a police center where it is analyzed for violations.)
  • Monitoring ability: Since the robots record the initial offense (or lack thereof), they provide evidence as to whether or not a police officer is justified if the officer later contacts a driver and imposes a fine. The robots may not mimic the ability of police body cameras to document corruption in any subsequent public-police interactions, but this record of driver behavior could allow a driver to prove that her or she is being unjustly asked for a bribe.
  • Expressive function: The engineers could have merely created a regular stoplight with a camera. Instead, they chose to install their technology in a hulking, semi-personified figure. Beyond the novelty factor, human-like robots could be effective at inducing behavioral change. Think of them like an extension of the classic “eyes in the lunch room” studies, which found that pictures of eyes posted in lunch rooms resulted in, for example, more people depositing money into a lunch room “good faith” payment box. Robots could serve as a middle ground between actual humans, whom drivers feel they can’t trust, and completely un-human traffic light cameras.

Yet there’s reason to remain skeptical. The continued necessity of human involvement should not be underplayed. Since human enforcement of penalties is ultimately necessary, the potential for bribes still remains. Furthermore, actual data about the robots’ effectiveness, beyond the same few anecdotal quotations passed around the Internet, remains nonexistent. True, the Congolese government is buying more of them, but one could just as easily see that as the government investing in a flashy technology created by a group with a compelling story behind it, instead of more substantive anticorruption changes or infrastructure.

Despite my skepticism, I still think the the project is worth keeping an eye on, for two reasons:

  • First, it may be most helpful in decreasing bad police behavior, in contrast to affecting bad driver behavior. A driver who truly has committed an infraction could still attempt to bribe his or her way out of a larger fine at various levels (for example, at an administrative office, or at the office that receives the recorded footage).  On the other hand, with fewer police officers on the traffic beat, there are fewer opportunities for officers to demand bribes.
  • Second, the robots function as an interesting psychological experiment: By both documenting the infraction (so there is still some sense of future consequence for bad behavior, despite the lack of an immediate enforcement agent) and distancing the moment of potential bribery (from immediately after one speeds or runs a red light to much later, when one receives notice of the fine and must go somewhere to pay it), do the robots make it less likely that people will perceive of bribery as a way to escape punishment? It’s not clear that the answer to this question will be yes, but if it is, it could have broader anticorruption implications:  reformers might be able to spend less money closing up every little loophole and work-around, and instead focus on just throwing somewhat crude, less expensive obstacles in the path of people who could be tempted into corruption.  In fact, perhaps it’s best for those obstacles stay primitive—after all, if movies are to be believed, intelligent, powerful robots might find us corruptible humans to be our own biggest problem.

3 thoughts on “Danger, Will Robinson: Can Robots Protect Us from Corruption?

  1. Pingback: 23 AUGUST 2015 | WAR OF THE DRONES

  2. Katie this is really interesting I am particularly intrigued by the idea that the expressive robots could function like the eyes on the wall in the payment experiments. If the cameras are sometimes monitored, they could also function like a panopticon.

    I do have a few purely practical question about the potential to deploy robots in corrupt areas of the developing world (which don’t detract from the intellectual interest of their capacity to thwart corruption if it was feasible to place them)
    1. Electricity flows. In many of these areas, power frequently goes out. Robots would either need to be connected to a power source or would require frequent battery changes. Would there be a chance that law enforcement would go down whenever the power does? If they use batteries, what are the logistical challenges of distributing them?
    2. Cost: Would purchasing, repairing and monitoring footage from these robots significantly increase law enforcement costs over human police officers? If so, how can very poor countries afford them?
    3. Potential theft. Robots would seem valuable. In societies where metal bits are frequently stolen for their scrap value, would there be an incentive to steal the robots that are supposed to enforce the law?

    • Thanks for a great post Katie! As Sarah said, I think the personification component is the most interesting part of this idea. And, it is a component that most other technology based tools with potential for fighting corruption–I would add block chain registries and mobile banking to the list you mention at the beginning of the post–seem to lack. Given psychology research about obedience like the Milgram studies, person-like enforcement tools likely have a lot of potential.

      I would echo Sarah’s concern about theft. In India, at least in Mumbai, I never once saw a public trashcan on the side of a street. Trash is a huge problem, and one that Prime Minister Modi has made a focus through his Clean India campaign. Apparently there were efforts to provide public trashcans, but theft was impossible to overcome. Any trashcan that was mobile enough to be emptied was also vulnerable to theft for scrap metal. Of course, these robots are much bigger, so perhaps theft is less of a concern logistically. But damage is still an issue. What stops someone caught by the robot from damaging it with anonymity to destroy the evidence? Damaged robots across the city could create bigger broken-window-theory style problems related to public perception.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.