A Society Can Be Watched Closely and Still Be Free, on Three Conditions

The standard picture puts surveillance and liberty at opposite ends of a slider. More cameras, less freedom. I think the picture is wrong, and that getting it wrong is dangerous, because the cameras are coming regardless and we will need to know what actually protects us once they are here.

Surveillance Is Not the Variable

Start with an observation that should be uncomfortable for everyone. Some of the most closely watched communities in human history were also among the freest, and some of the least watched were tyrannies.

A village of two hundred people has total surveillance. Everyone knows where everyone was last night, who is sleeping with whom, who drinks, who owes money. There is no privacy in any modern sense. And yet villages have ranged from open, tolerant and self-governing to suffocating and cruel. The watching was constant in both. Something else determined the outcome.

Conversely, a modern state can be nearly blind to its citizens and still crush them. Knowing nothing about people is perfectly compatible with imprisoning them at random.

So surveillance does not, by itself, set the level of liberty. It changes what is possible, in both directions. What determines which direction is a set of conditions about who watches, who is watched, and who writes the rules.

Condition One: The Watching Runs Both Ways

The tyrannical village and the free village differ in one respect immediately: in the free one, the headman is watched as closely as the drunk.

A surveillance arrangement is dangerous in proportion to its asymmetry. When the state can see the citizen and the citizen cannot see the state, every increase in capability is an increase in domination, because knowledge is leverage and the leverage all points one way. When the citizen can see the state just as clearly, the same capability becomes accountability.

Concretely: if the police record the public, the public records the police, and the recordings are equally admissible and equally accessible. If the state keeps a file on every citizen, every citizen can read the file the state keeps on any official, including what they earn, whom they meet and what they decided. If a camera is on my street, the feed from the ministry's corridor is available on the same terms.

This is not a hypothetical remedy. Body cameras, freedom of information laws, published registers of interests and mandatory disclosure of official meetings are all partial versions of it, and everywhere they exist they work partially. The failure is that they are partial. Symmetry has to be the principle, not the exception the state grants when pressed.

Condition Two: The Watchers Are Watched

Symmetry between state and citizen is not enough if the people who run the system sit outside it.

Every surveillance apparatus has operators: the people who can query the database, redirect the camera, flag the record. In every actual case I know of, this group has been the least supervised part of the arrangement, on the theory that supervising them would compromise the operation. That theory is exactly backwards. The operators are the point of maximum danger, because they hold the leverage on everyone, and they must be the most supervised people in the society, not the least.

That means every query logged and the log public. Every access to a record visible to the person in the record. No exception for security, because security is precisely the excuse under which the abuses happen. If a state cannot run its intelligence service under those terms, it should have a smaller intelligence service.

Condition Three: The Rules Are Set by the Watched

The first two conditions are about information. The third is about power, and it is the one I care about most.

Whether a surveillance society is free depends in the end on who decides what the watching is for, what may be done with what is seen, and what is forbidden. If that question is answered by the same small group that operates the system and benefits from it, the first two conditions will erode, because the people setting the rules have every incentive to weaken them.

The rules for watching everyone must be set by everyone, directly, and revisited often, because the technology moves and the rules must move with it. This is not a general plea for direct democracy dressed up as a surveillance policy. It is that surveillance is the specific domain where a rule written by the few for the many is most obviously self-serving and most obviously dangerous, and where the many have the clearest interest and the clearest competence to decide. Nobody needs expertise to know whether they want their medical records queried without notice.

Why This Matters Now

The cameras, the sensors, the location logs, the recognition systems are not going away. Arguing against their existence is arguing against the tide, and it has the further disadvantage of leaving the field to the people who will build them anyway and write the rules in private.

The useful argument is about the three conditions. Does the watching run both ways? Are the operators the most exposed people in the system? Do the watched set the rules? Where the answer to all three is yes, I would accept a great deal of surveillance and expect to remain free. Where the answer to any one is no, I would not accept a single camera, and I would expect the ones already installed to be used against me eventually.

The slider was always the wrong picture. Liberty under surveillance is not a quantity of watching. It is a shape.