30 Gender and Security Toolkit Unless comprehensive statistics related to gender are gathered on a regular basis, there is no way of fully appreciating the size of any problem. This is not only an issue of accuracy, but also because certain people are rendered invisible, and the discrimination and inequality that they experience pass unnoticed. It is imperative that states collect statistics at the national and institutional levels on an array of topics, and specifically disaggregate these data by gender. The data should be collated monthly at a minimum and analysed regularly, with the findings feeding into policymaking. Such data should be collected and analysed in collaboration with organizations representing the population groups in question and those with lived experience within places of deprivation of liberty. Responsibility for oversight of a specific strategy to promote a gender-responsive approach should lie with a government body with the power to ensure that recommendations are implemented and evaluated. Although not an exhaustive list, governments should require individual institutions to collate gender-disaggregated information on:  staffing (in post, promotions and leavers), including at senior levels  staff participation in training opportunities  SGBV (prisoner on prisoner, prisoner on staff, staff on prisoner, staff on staff)  complaints (submitted and upheld)  use of force and/or restraint  use of segregation or solitary confinement. As with all aspects of working with people in places of deprivation of liberty, the staff group making up the internal inspection teams should reflect the diversity of the population which they monitor. Extra efforts should be made to reach out to groups that are less likely to report SGBV, and these groups will differ with context. By way of example, a report of the Office of the UN High Commissioner for Human Rights highlighted the barriers to people who are LGBTI reporting crimes against them: Victims are often reluctant to report their experiences for fear of extortion, breach of confidentiality or reprisals. In addition, prejudicial and inexact categorization of cases results in misidentification, concealment and underreporting. Failure to investigate, prosecute and punish violations when reported also contributes to incomplete assessments of the scale of violence.3 In terms of collecting data, estimates of the prevalence of SGBV can be derived from different sources, including surveys conducted among detainees or former detainees, official administrative records, and complaints submitted by detainees/prisoners to monitors, civil society organizations, domestic courts or international bodies. Different states have differing definitions of SGBV, and different methodologies mean that data are difficult to compare between states. However, the driving force for the collection of data should be to look inwards and understand the local context, and what can be reformed to improve gender equality. See Box 7 for an example of a regular SGBV data collection process.* * For more information on reporting and underreporting of SGBV prevalence see OSCE/ ODIHR (2019) Preventing and Addressing Sexual and Gender-based Violence in Places of Deprivation of Liberty: A Compilation of Standards, Approaches and Examples from the OSCE Region. For more on how monitoring bodies can improve their work with transgender prisoners see Association for the Prevention of Torture (2018) Towards the Effective Protection of LGBTI Persons Deprived of Liberty: A Monitoring Guide, Geneva: APT.