84 Sections 37 and 38 of the Execution of Sentences Act can both be seen as containing implicit requirements for prevention. Pursuant to Section 37, the measure must be ‘necessary’. Among other things, this means that an assessment shall be made of whether the ­purpose can be achieved by other, less intrusive measures than exclusion from the company of other ­inmates.266 Pursuant to Section 38, such measures must be ‘strictly necessary’, and it is an express ­requirement in the Act that ‘less invasive measures have been unsuccessfully attempted or will clearly be inadequate’. The existing regulatory framework does not, however, confer on the Correctional Service any overall duty to systematically prevent the use of coercive measures such as solitary confinement.267 The Correctional Service ‘shall make suitable ­arrangements for remedying the negative effects of isolation’, and to ‘prevent or remedy the harmful effects of exclusion’. This is different from preventing the use of solitary confinement. The current legislation falls short of meeting Norway’s commitments under international law to prevent situations that entail a risk of torture or other cruel, inhuman or degrading treatment or ­punishment. Special Report to the Storting Document 4:3 (2018–2019) 12.1 The responsibility of governing authorities The main task of the governing authority in preventing the use of solitary confinement is to provide each prison with a framework and conditions that make it possible to use solitary confinement in exceptional cases only (as a last resort) and for as short a time as possible.268 Effective prevention is conditional on knowing where and how measures should be implemented. Such knowledge is dependent on accurate information. ­Chapter 5 showed that there are major and serious gaps in the information about the number of ­administrative decisions on exclusion, and that many are in fact excluded without an administrative decision. This applies to inmates who are placed in solitary ­confinement for financial or practical reasons ­(see ­section 6.1 De facto solitary confinement), and self-isolation for which no administrative decision is required in the opinion of the Directorate (see section 6.3 Solitary confinement by choice). In Chapter 6 Extensive use of solitary confinement and restrictions on association with other inmates, ­reference is made to the large number of inmates who are locked up in their cells for more than 16 hours a day in units defined as communal sections. The ­establishment of minimum standards for association with other inmates, in line with our Nordic neighbours, would presumably go a long way towards preventing such confinement. 85 There are likewise no central guidelines on what the individual prison should to prevent self-isolation or ­isolation as a result of suicide risk (see section 9.3 Use of security cells in connection with mental crises). The Correctional Service’s guidelines for the prevention and handling of self-harm, suicide attempts and ­suicides in prisons provide little guidance on what specific measures staff should take to prevent the solitary confinement of inmates who are considered to be suicidal. The same applies to possible ways of following them up to provide human support and care. The Correctional Service also lacks reliable figures relating to the number of suicide attempts in Norwegian prisons.269 Such figures can constitute important controlling information for obtaining an overall picture of the situation in a prison and of the inmates’ ­general state of health. This is in turn important for the ­prevention of solitary confinement. Our review of decisions on complete or partial ­exclusion and the use of security cells during the period 2014–2018 shows that the grounds seldom include any documentation of measures to prevent exclusion and solitary confinement. The grounds are typically stated in standard phrases such as ‘less invasive measures have been tried/ are clearly impossible’, without any reference to specific measures. This suggest that there is a need for a more systematic approach on the part of the governing authority, among other things to the question of when the conditions in Section 37 of the Execution of Sentences Act are met and what grounds should be given for administrative decisions. The prisons have a duty to report prolonged exclusions and stays in security cells to the regional level (after 14 and 3 days, respectively) and to the Directorate of the Norwegian Correctional Service (after 6 and 42 days, respectively). The Parliamentary Ombudsman is not aware that these limits are used systematically to ensure that all prisons comply with the same strict rules on administrative decisions pursuant to Sections 37 and 38 of the Execution of Sentences Act, and that other, less intrusive measures are implemented where possible. Furthermore, many who are held in solitary confinement for long periods are omitted from the figures, including inmates in self-isolation for whom no administrative decisions are made. This undermines the possibility of systematic prevention. 266 Directorate of the Norwegian Correctional Service (2002). Guidelines to the Execution of Sentences Act and its Regulations, section 37.6. Revised version of 2 April 2019. 267 See the Execution of Sentences Act Section 3 and the Regulations to the Execution of Sentences Act Section 3-35. For purposes of comparison, see the Regulations relating to rights and the use of force in child welfare institutions (the Rights Regulations) Section 1 268 The Directorate of the Norwegian Correctional Service has appointed a working group tasked with preparing a draft programme of measures to prevent the use of solitary confinement. The content of this programme was not yet known at the time of submitting this report. 12.2 The responsibility of prisons Accurate information is important, including in the ­preventive work of each individual prison. Such ­information requires that the actual scope is reflected in the figures, however. In that respect, too, lack of knowledge about the real scope of solitary confinement is clearly a problem. We have found examples in several prisons of the administration not knowing about the different sections’ actual routines for locking up inmates. There also ­appear to be differences between the prisons as regards the administration’s knowledge about ­administrative decisions on exclusion and the breakdown of such decisions by section. The failure to keep records of self-isolation also means that the prison administration is not fully informed about the real figures. One of the consequences of this is wide variations in the local prison administrations’ active efforts to reduce the extent to which inmates are locked up in their cells. Our findings show that some forms of solitary ­confinement are more predominant in some prisons than in others. There are wide variations between ­prisons, among other things, in the number of inmates who choose to isolate themselves (see section 8.1 Inmates who chose solitary c ­ onfinement (self-isolation)). The figures for exclusions and placement in security cells also vary widely between prisons, and some stand out more than others, including over time. There may be complex reasons for this. Some prisons have more challenging inmates than others, or have more prison places without communal facilities. The differences are so great, however, that it is likely that the local prison ­administration and local prison culture also have a ­bearing on the extent to which such measures are used.270 269 Directorate of the Norwegian Correctional Service, (*Information about suicides in prisons – follow-up information relating to the registration of suicide attempts’), letter of 8 August 2018 to the Parliamentary Ombudsman. 270 The need for prison governors to be assigned a clearer and more central role was also identified in a report submitted to the Parliamentary Ombudsman: Directorate of the Norwegian Correctional Service (2015). Rapport om utelukkelser i kriminalomsorgen (’Report on exclusions in the Correctional Service’), p. 11.