Comment: This table shows the trend reversing from 2000. During the last three decades of the 20 century, the growth in the number of prisoners serving long sentences was constant and marked. The proactive policy of developing the adjustment of short sentences (firstly less than one year and then less than two years) follows fresh growth in short sentencing demonstrated by the statistics on sentencing, whilst long sentences have stabilised at a high level. The reconciliation between counting movements and those in stock shows that the average duration of imprisonment doubled between 1970 and 2008 (2009 CGLPL Report, Page 251, note 2 in the French version). Indicators then continued to increase to 10.4 months in 2013. This increase is confirmed for the average duration of detention within its strict meaning, which increased from 8.6 months in 2006 to 11.5 months in 2013, and stabilised in 2015 and 2016 (10 and 9.7 months respectively) (DAP-PMJ5, 2014-2017). th Additional reference: "L’aménagement des peines : compter autrement ? Perspectives de long terme" (Adjustment of sentences: another way of counting? Long-term outlook), Criminocorpus, 2013, (online: http://criminocorpus.revues.org/2477). 1.7 Incarceration densities and over-occupation of penal institutions Statistical data used by the Prison Administration, total number of prisoners at any given time and operational capacity of institutions, enable it to calculate an "incarceration density" defined as the comparison between these two indicators (numbers present per 100 operational places). The density for all institutions – 117.8 on 1 December 2017 – has no great significance as the indicator varies a great deal according to type of institution: 90 for detention centres and detention centre wings, 75.1 for long-stay prisons and long-stay prison wings, 67.5 for institutions for young offenders, whilst for remand prisons and remand prison wings the average density was 142.1. In addition, the average by type of institution includes variations within each category: - of the 95 sentencing institutions, only 12 had a density higher than 100, including 4 detention centre wings in Overseas territories and 5 open prisons (4) and centres for adjusted sentences (1) in Ile-de-France. This over-occupation concerned 815 detainees in Mainland France and 364 in Overseas France. - of the 130 remand prisons and remand prison wings, 16 had a density lower than or equal to 100 and 114 had a density greater than 100, of which 42 had a density higher than 150. Three remand prisons and remand prison wings exceeded 200, i.e. a population of prisoners more than double the number of operational places (all three in mainland France). Over-occupation of prison institutions is therefore limited to remand prisons by application of numerus clausus to sentencing institutions which are a little below declared operating capacity. For remand prisons, the increase in operational capacity (+2,008 places between 1 January 2005 and 1 January 2015) was less than that of the number of prisoners (+3,742) and density is therefore higher in 2015 than in 2005. Over-occupation of an institution has consequences for all prisoners in it, even if some cells have normal occupation (new arrivals wing, solitary confinement wing, etc.). It is therefore relevant to note the proportion of prisoners based on the percentage of occupation of the remand prison where they are. At 1 January 2017, the vast majority were affected by this situation of over-occupation (91%); over a third (38%) of detainees in remand prisons or remand prison wings were in institutions where the density was greater than or equal to 38. Reference: "Statistiques pénitentiaires et parc carcéral, entre désencombrement et suroccupation (1996-2012)" ("Prison statistics and total incarceration, between clearance and over occupation (1996-2012)"), Criminocorpus, 2014 (online: http://criminocorpus.revues.org/2734). 156