In this blog post, Program Manager Thijs Turel discusses the alternatives. He suggests that we should introduce a degree of randomness to surveillance systems.

Complete enforcement is becoming affordable

Gradually our cities are installing sensor networks and camera monitoring systems. These infrastructures make it easier and cheaper for government to register violations, both major and minor. And with costs going down and technological possibilities ever growing, the monitoring systems are increasingly able to register more and more offences. Examples are the traffic speed camera’s (trajectory speed control) at the A2 highway and the camera systems being used to enforce ‘milieu zone’ rules; meant to keep polluting vehicles out of the city. At some point in the near future we might have a system that is able to spot and fine virtually all people who break the rules, even for minor offences.

“With the rise of new technologies, also new possibilities come about. But this does not mean we should use every possibility the technology has to offer.”

Thijs Turèl

Program Manager Urban Data & Intelligence

Tada principles

In the recent coalition agreement, the Amsterdam Municipal Council embraced the so called TADA principles, meant to guarantee responsible digitization. One of these principles is ‘maintaining the human scale’. But to what extent is a government that automatically fines 100% of the smaller offences still in line with the human scale? In Shenzhen, China, a system is used that, based on facial recognition, fines people who violate road rules. Ask a random person in the Netherlands and they will probably feel pretty uncomfortable with this. Will this also become our destiny when we blindly accept the automated fine machine with 100% chance of being caught?

Two challenges

To my mind, there seem to be two challenges we face when this 100% accurate fine system would become operational. The first challenge is that many of the rules and laws we have in place where never meant for complete enforcement. An example: think of someone crossing a zebra crossing with a red traffic light. This is currently prohibited under traffic regulations. But if we would enforce this rule every time, we would unnecessarily restrict people's freedom: There are many situations in which people do not harm anyone by breaking the rule: For example, when there is no other traffic in sight. The purpose of having this traffic regulation that it enables government officials such as police officers to fine people in situations where violation of the rule actually causes problems.

The second challenge is that the new possibility of affordable enforcement regimes is stimulating the creation of new rules. It is doubtful the ‘milieu zone’ rules would ever have come into existence, without the camera system being technologically and economically feasible. And now it has been installed, cities are already thinking about more use-cases to use the enforcement system that has been built.

Now, even when all new rules have been formulated to protect legitimate public interests, this can still lead to an undesirable situation. Collectively, they will gradually restrict people's autonomy. Now it's red light crossing, in a year it will be cycling on the sidewalk, throwing garbage on the street, etc. A frightening prospect, in which we are allowed to do less year after year.

Four possible solutions

But how do we do this when we are drawing up more rules and it is becoming easier to enforce them? I see four possibilities, less rules, more lenient rules, lower fines and the use of ‘artificial dice’.

  1. Less rules. If we were to have less rules, logically the problem described here would not arise. Yet this solution largely seems to be an appeal to be cautious about formulating new rules. It is difficult to imagine that we will start abolishing existing rules (i.e. scrapping the regulation that states that crossing a red light is an offence).
  2. More lenient rules. We can, of course, nuance rules by including exceptions and making them context-dependent. For example: If there is no other traffic within 60 metres crossing is allowed; except during fog/in situations where visibility is obstructed. Yet rules that have many exceptions, or depend on the situation, are complicated. It is not fair to expect people to understand or be aware of all the exceptions.
  3. Lower fines. We can start working with micro-fines. For example: We are no longer fined 65 euros once, and remain undetected for the 99 other times that we walk through the red light. Instead, we are fined, say 65 cents, for each of a total of 100 violations. Advantage: there is no longer a penalty that was intended for incidental application, which suddenly becomes draconian with structural application. Disadvantage: there is a risk that the fine will now rather be interpreted as an economic choice - as a right that you can buy. This seems to be at odds with the moral appeal that should go with fines.

All disadvantages.

  1. So what do we do? Here’s my perspective on solution number four:4. What would happen if we translate the current way of working into the digital world? We artificially reduce the chance of getting caught, by making the surveillance system play dice, introducing a degree of randomness. We set a compliance target. We accept that 100 people daily ignore a red light while crossing the road, and set the chance of being caught in such a way that we get 100 violations. If too many people violate the rule, the chance of being caught increases and vice versa.

Advantages: the rules remain the same, so it is clear what the rules are, and what kind of behaviour we expect from people. But in doing so, we aim for a minimum amount of fines. Disadvantages: is by definition arbitrary whether you will be punished in case of violation. But that's no different than it is now.

Would it work? I’m not sure. But we have to figure out a way to deal with this. This is the only way in which we keep a human scale in our time and age of increasing digitization.

This is a personal blog, opinions are his own.