Note: the annual reports of the CICI from 2003 to 2015 allow the first five columns of the table to be presented. The column for accompanying minors was not present before 2011. The last two columns relating to the result of placing and holding in administrative detention centres do not come from the same source. A report by the Senate Finance Committee published on 3 July 2009 and following up on a mission carried out by the Court of Auditors, provided numbers for 2006-2008 with regard to detainees who were finally sent back, excluding voluntary returns. The proportion compared with number of committals can therefore be calculated (last column). The 7th CICI report, published in March 2011, provided the proportion for 2009 (page 77). The following report gave a rate of 42% for CRAs possessing interservice deportation centres (pôle interservices éloignement) and 37% for the rest, but no overall rate. The items set out in the last column of the table for 2010- 2013 are from an informational report from the Senate on CRAs (no. 775 of 23 July 2014). This report also sets out the number of placements in 2013. These figures nevertheless remain linked to sporadic assessments of detention, and have unfortunately not been updated for 2014 and 2015. The number of placements in 2009 has been corrected here compared with previous editions: the new statement of 30,270 placements given initially as the total for France and its Overseas territories (CICI reports for 2009, 2010 and 2011) became in later editions (2011 and 2012) that for Mainland France, whilst the previous edition (27,699 placements) became that for French Overseas départements. Comment: The CICI annual reports do not show how the average rate of occupation is defined and assessed. By applying this rate to capacity, an estimate of the average numbers of persons present in CRAs should be obtained. However this estimate is unreliable as the capacity may have been given for a fixed date (it would not then be the average capacity for the year). Another estimate of numbers would be possible from this table as placements correspond to entries and average duration of stays has been supplied. A lower estimate is arrived at. For 2015, calculating the occupancy rate gives an average total number of 840 prisoners, and a calculation by average stay in detention gives a total number of 835 prisoners. These two methods of calculation show an increase in these prisoner numbers from 2003 (496 or 432 dependent upon the method of estimating) to 2007 (1,285/1,014) and then a drop to 2011 (811/585). The same calculation showed an uncertain result for 2013 (754/795), the first indicating a fall and the second a rise, but the 2015 data showed an increase whatever the calculation method chosen. Relatively little use continues to be made of house arrest, an alternative to detention introduced in 2011: 668 measures in 2012 and 1,258 in 2013 (source: French National Assembly, impact study of the bill dated 23 July 2014). 165