Security and liberty: a trade-off rather than a choice

A society without security does not allow liberties to be exercised, since fear restricts movement, expression and plans. A society without liberties no longer protects anyone against the one who provides security. The two values support each other as much as they limit each other, which forbids treating one as the price of the other.

That is why the useful question is never which of the two you prefer, but where you place the point of balance, on which subject, and under what guarantees. Put that way, it admits of precise and debatable answers.

Everyday vocabulary complicates matters by associating each value with a camp. Yet real positions are distributed according to subject: the same person may call for more control in one area and less in another, without contradicting themselves.

The first is that of proportionality: is the restriction commensurate with the risk, and is there a less restrictive means of obtaining the same result? This reasoning is at the heart of judicial review of policing measures in most democracies.

The second is that of reversibility: is the measure limited in time, with a review provided for, or does it settle in? Arrangements adopted in exceptional periods have a documented tendency to become permanent, which gives the temporary character particular importance.

The third is that of oversight: who checks, with what means, and what happens in case of abuse? The same measure may be acceptable under independent oversight and unacceptable without, which explains apparently contradictory positions on a single arrangement.

Security measures present a known political asymmetry: adopting them is cheap, withdrawing them is risky, since any incident occurring after the withdrawal will be imputed to whoever decided it. The result is a one-way movement, independent of the evolution of the risk.

This asymmetry is a serious argument in favour of automatic review clauses: they shift the burden of decision, by requiring justification of maintenance rather than of withdrawal. It is one of the few mechanisms that corrects a structural bias without depending on the virtue of decision-makers.

The same person may accept a strong restriction on a risk they judge serious and probable, and refuse it on a risk they judge rare. The trade-off depends on the estimate of the risk as much as on the value granted to the liberty concerned.

This explains combinations often described as incoherent, for example great vigilance about the surveillance of communications combined with acceptance of checks in public space. These are not two opposite positions, but two different estimates of the risk and of the guarantee available.

Faced with a new measure, three questions are enough to situate one's own position: what happens if it is misused, who will be able to notice, and how is it brought to an end? They do not say whether the measure is good, but they reveal what one requires in order to accept it.

It is also the best way of discussing with someone who does not share your estimate of the risk: disagreement about the figure is hard to settle, disagreement about the guarantees can be negotiated.