Special Report to the Storting Document 4:3 (2018–2019) 34 These include situations in which inmates choose to be excluded without having signed the appropriate declaration, and situations in which daily association with other inmates is very limited due to building or staffing conditions of a non-urgent nature. Hence, they are not included in the figures. Because of the lack of reliable figures, previously published statistics on decisions on complete or partial exclusion have been left out of this report. Excepted is information about particularly prolonged ­exclusions for the period 2015–2018 and the overview of ­administrative decisions on complete exclusion during the period September–December 2018. These figures are based on separate reporting schemes.104 5.2 Sources of error in day surveys The Correctional Service has conducted nationwide day surveys since 2012 for the purpose of obtaining a more accurate picture of how often and for how long inmates are locked up in their cells. During the period 2012–2018, 19 surveys were carried out on randomly selected weekdays at four-month intervals. In 2018, one day survey was also carried out at the weekend. The surveys are conducted by the Correctional Service Region South, and the regional offices are tasked with obtaining and assuring the quality of the figures obtained from the prisons in their respective regions. The surveys entail reporting by each individual prison of the number of inmates who have no association with other inmates on that particular day, and the number of inmates who have less than two hours’ association with other inmates. As from 2015, the number of inmates with between two and eight hours’ association with other inmates is also reported. 35 There are a number of sources of error associated with the Directorate of Correctional Service’s day surveys. Weaknesses related to the method, such as the low number of survey points (three per year), leave room for random variations, and also prevent the figures from giving a representative picture of the extent of isolation and exclusions in Norwegian prisons. There have been no surveys conducted systematically on weekends, and the category ‘2–8 hours out of cell’ provides no information about the distribution within this category. It is substantially more intrusive and harmful to have 2–3 hours out of cell per day than 7-8 hours. In our review of the day surveys, we have found that there are no adequate procedures for quality assurance of the figures. Among other things, we have found that reported data from some prisons is lacking. Several prisons have also reported figures for the number of inmates with restrictions on association with other inmates that are clearly lower than what we, based on our visits, have reason to believe is actually the case. 5.3 Uncertain figures on the use of security cells Concerning the use of security cells, we have reviewed figures for the period 2008–2018 and information about the duration of administrative decisions for the period 2013–2018. In the process, we found incorrect records of such use, and the figures also did not tally with our findings on the use of security cells made during visits to individual prisons. This makes it reasonable to assume that the number of administrative decisions on the use of security cells is also uncertain, and that the actual figure is probably higher. Signs for marking cell doors. Main findings Our collection and review of figures show that there is no reliable overview of the extent of solitary ­confinement in Norwegian prisons. There are significant sources of error associated with many of the available figures. This means that the responsible authorities do not have access to information necessary to enable consideration of measures to reduce the use of solitary confinement and limit its harmful effects. 104 Prolonged exclusions are reported in accordance with statutory deadlines; see Section 37 fifth paragraph of the Execution of Sentences Act. The figures showing the number of administrative decisions pursuant to Section 37 during the period September–December 2018 are based on the manual reporting scheme. It is highly censurable that, more than 20 years after the Standing Committee on Justice requested such figures, the governmental authorities still lack a reliable overview of the extent of isolation in Norwegian prisons. The figures used in this report must therefore be understood to be minimum estimates throughout.