an indefinitely renewable period, because of their supposed dangerousness, i.e. a "very high probability of recidivism" – a subjective notion if ever there was one. Far from having disappeared from our legislation, despite the commitments made in this respect in 2012, this notion is spawning numerous narratives aimed at assuaging (with minimal effort) the general public – who have due reason to be worried. We have heard talk of "dangerousness", of "individuals to separate from society", of the "precautionary principle applied to justice" to justify locking up individuals on the S-list. We will have been spared nothing all through 2016, during which time a number of established lines have shifted, when an entry flagged on the S-list should only be for internal use by the police services, the contents of which has not always or still not been checked and which, in all cases, has never been validated, either by an adversarial procedure or by a ruling. Is this not ultimately about locking up for as long as possible all individuals who are considered "deviant", the criminal, the "madman", quite ignoring the fact that this person will one day get out, and that it would be in society's best interest for this spell of deprivation of liberty to serve a constructive purpose? 2016 not only saw a significant erosion of fundamental rights in the legislation, but also in the day-to-day reality of the institutions that the CGLPL visited all through the year. Prison overcrowding is only getting worse. The CGLPL has constantly denounced this issue as a breach of people's dignity and a form of inhuman and degrading treatment in the meaning of Article 3 of the ECHR. On 1 December 2016, the prison occupancy rate reached 118%, while the rate in remand prisons hit 141%. In a report published on 20 September 2016, "En finir avec la surpopulation carcérale," the Ministry of Justice produced a spot-on analysis of the phenomenon and stressed the need for balance between the creation of new places and finding alternatives to imprisonment. But, paradoxically, it earmarked almost all budget appropriations to the former. And yet, a satisfactory answer to the problem of prison overcrowding will never be found solely by creating new prison places. Indeed, despite some 30,000 new places having been created over the past 25 years, this very problem has never been so acute: the average rate of 141% in remand prisons hides peaks of 200% in Ile-de-France (the Parisian region) and in Overseas France. The number of remand prisoners (i.e. presumed innocent) meanwhile exceeded the symbolic threshold of 20,000 in 2016 – up by 14% from 2015 – now accounting for a third of all prisoners when they only made up a quarter back in 2015. This finding invalidates the statements made at regular intervals concerning a supposedly "lax" justice. On the other hand, non-custodial alternatives to prison are still wholly inadequate, despite the 15 August 2014 Act which has not gone as far as was hoped: just 2,300 non-custodial sentences have been passed in two years instead of the 8,000 to 20,000 per year that the impact study for the Act predicted. During the visits it carried out in 2016, the CGLPL was able to observe just how nervous magistrates are, in the current context, about imposing sentence adjustments. Prison sentences should only be imposed as a last resort, and yet time and again the CGLPL's teams came across situations in which such sentences smacked as illogical: very short sentences – which contribute in no small part to desocialisation and financial insecurity, and make no impact in terms of rehabilitation because prison rehabilitation and probation services are already stretched to breaking point; sentences served by people whose old age or physical or mental health do not appear to be compatible with being kept in prison, but who remain there anyway for lack of any alternative. What are we waiting for, then, before thought is given to the point of very short sentences and to the continuing imprisonment of people whose health is failing? In some jurisdictions, there is evidence that constructive dialogue between the judicial authority and the prison managers is enabling individual situations to be handled on the margins, by putting forward a sentence adjustment or end of sentence, or postponing imprisonment, which effectively limits prison overcrowding. Such praiseworthy and discreet initiatives do not have any financial impact and their 11