26 Jun The Problem with Independence in Torture Prevention
[Marie Steinbrecher is currently a teaching fellow at UCL’s Department of Political Science while completing her PhD at Royal Holloway’s School of Law.]
The Optional Protocol to the Convention against Torture (OPCAT) obliges State Parties to designate a National Preventive Mechanism (NPM). NPMs are mandated to visit all places of deprivation of liberty to prevent torture and ill-treatment. To date, 90 States have ratified OPCAT and must allow independent oversight of their detention facilities. In this post, I discuss the concept of independent oversight, and why independence is important yet challenging for NPMs to realise. First, I briefly introduce the meaning of preventive monitoring, as envisioned under OPCAT. Subsequently, I highlight the importance of independence, after which I discuss why independence is challenging for NPMs to achieve. Overall, I argue that OPCAT’s independence requirements are so vague that State Parties have no clear obligations, which risks NPMs lacking de facto independence. In particular, I highlight the complex relationship between the NPM and the State, and that despite OPCAT’s obligation, NPMs cannot be fully independent from the State.
What is preventive monitoring?
Places of deprivation of liberty may be an odd phrase, but it is chosen deliberately. The list of places where people can be deprived of their liberty is long, and the definition in OPCAT Article 4 includes all places that an individual cannot leave at their own will. This definition is sufficiently broad to include traditional places (e.g. prisons, immigration detention) as well as non-traditional places (e.g. psychiatric hospitals, homes for the elderly).
Preventive monitoring performs a crucial function: It helps to protect the rights of people deprived of their liberty and ensure that they are not mistreated. To this end, NPMs fulfil a number of functions, the most important of which is visits to places of deprivation of liberty. NPMs visit all places within the country – prisons, police facilities, orphanages, psychiatric institutions and many more. The idea behind this is that the authorities, such as the prison governor and guards, never know when the mechanism might enter. This should deter authorities from committing any acts of torture and it should allow NPMs to uncover both actual torture and any risks for this to happen in the future. Even more important are the recommendations that NPMs make. States must discuss their implementation and NPMs follow up in later visits to assess if the situation has improved. There should thus be a positive development over time.
Torture prevention takes many forms. For instance, in Norway, the NPM criticised the incarceration of children in an immigrant detention centre. The National Police Immigration Service then established a family unit in a separate building suitable for civilians. In another example in Austria, the NPM criticised the use of net beds in psychiatric facilities, which are often used to restrict movement and confine people. Net beds are now prohibited in Austria.
The importance of independence
OPCAT gives State Parties considerable freedom in the design and structure of their NPM. This has resulted in a plethora of bodies fulfilling the mandate: Many State Parties opt for designation of an existing body, such as National Human Rights Institutions or Ombudsman offices. Others create a new body designed specifically for OPCAT, or designate multiple institutions, sometimes in combination with non-governmental organisations (NGOs). While this flexibility allows each NPM to suit the context of their jurisdiction, OPCAT stipulates a few minimum requirements that NPMs must fulfil.
These requirements relate mostly to the powers (for example, NPMs must be able to visit all places and interview people at their own choosing) and transparency (annual reports must be published) relevant for the OPCAT mandate. The central requirement is independence, set out in Article 18: “The States Parties shall guarantee the functional independence of the national preventive mechanisms as well as the independence of their personnel.”
To be effective, NPMs must be independent from other stakeholders, most importantly from the government. This includes the power to enter any facility, view any document, and speak to any individual. It should also mean that the NPM can decide freely how to spend its budget, what priorities to set and how it disseminates its findings. Thus, independence protects the NPM from undue influence and ensures that it can work effectively to uncover the truth. Independence also protects people deprived of their liberty and ensures that civil society and the public are informed, thereby preventing censorship around torture and ill-treatment within the country.
The challenge of independence
Although independence is central for NPMs, and one of the few definite obligations for State Parties, the Article quoted above is too vague to give State Parties specific obligations. This creates difficulties for implementing OPCAT in national legislation and in the daily work of NPMs. Other bodies have aimed to provide further guidance, notably OPCAT’s treaty body, the Subcommittee on Prevention of Torture (SPT; see here and here), and the leading NGO in the field, the Association for the Prevention of Torture (APT; see here and here). They highlight important factors, such as NPMs being able to spend their budget freely, and to hire their own staff. They also make the helpful distinction between actual and perceived independence: The mandate of NPMs necessitates good working relationships with a variety of stakeholders, mainly the government and civil society.
While civil society can feed the mechanism information, and help disseminate its findings and raise awareness, the relationship with the government and State authorities is of special importance. NPMs do not have legal power – instead, they are required to draft recommendations. State Parties in turn are obliged to enter into dialogue with the mechanism regarding the implementation of recommendations. This creates a tension for NPMs with regards to their independence: On one hand, they are required to be independent and autonomous from any State authority. On the other hand, the State funds, staffs and allows the mechanism to operate in all detention places. Additionally, the NPM depends on the State for the implementation of its recommendations. Should a government be unwilling to implement recommendations, then this would have dire consequences for the effectiveness of the NPM.
States may have incentive to reduce independence and effectiveness of their NPM if they are not serious about torture prevention. Many strategies to influence NPMs exist, but one stands out, because it is in clear contradiction of OPCAT. Yet, Western liberal democracies, arguably the gold standard for human rights, repeatedly chose this model: In Italy, Switzerland, and the Netherlands, the NPM is an administrative organ of a Ministry. This has two implications for the mechanisms’ independence. Not only does ministerial subordination conflict independence as such, the Ministries hold competences relevant to the NPM mandate. For instance, in Switzerland, the NPM belongs to the Ministry responsible for military and police detention. If the government wishes to suppress criticism or forbid the NPM from visiting a place, then it has the power to do so. Even if the government never acts on this power, the risk for independence remains, leaving the NPM in a precarious position where it might have to weigh between a constructive working relationship with the Ministry and voicing criticism.
Much power, and responsibility, rests with State Parties in creating NPMs that work effectively. The text of OPCAT leaves considerable room for States to create mechanisms that look good on paper, whereas the reality may look much different. Despite additional guidance from bodies noted above, the relationship between State Party and NPM, and the tension regarding the mechanism’s independence, has not received sufficient attention so far. More guidance is needed to understand how NPMs may be independent from the State, and how State Parties ought to fulfil OPCAT.
OPCAT is not specific enough on independence, which means State Parties may interpret this obligation in a plethora of ways, and which increases the likelihood that State Parties will be in breach of OPCAT (or NPMs themselves). More specific guidance is provided by the SPT and APT and by some scholars, but while they all acknowledge that it is difficult for NPMs to have a relationship with the State while being independent from it, this challenge is not sufficiently addressed.