Cornelius Friesendorf, The OSCE in Central Asia: Debating Police-related Activities in:

Institute for Peace Research and Security Policy at the University of Hamburg (Ed.)

OSCE Insights 2020, page 1 - 16

Corona, War, Leadership Crisis

1. Edition 2021, ISBN print: 978-3-8487-7830-0, ISBN online: 978-3-7489-2233-9,


Bibliographic information
Cornelius Friesendorf The OSCE in Central Asia: Debating Police-related Activities OSCE Insights Institute for Peace Research and Security Policy at the University of Hamburg (IFSH), Centre for OSCE Research (CORE) 2020 | 01 DOI: 10.5771/9783748922339-01 The OSCE in Central Asia: Debating Police-related Activities Cornelius Friesendorf* Abstract This contribution calls for a debate on the effects of OSCE police-related activities in Central Asia. Drawing on a typology of internationally-supported police aid, it outlines three questions that deserve more scrutiny by participating States and civil society: Is the OSCE able to support democratic police governance in Central Asia? To what extent can the OSCE help improve human security? What are the limitations and risks of law enforcement support? Tentative evidence suggests that the OSCE faces significant challenges in translating its commitment to democratic policing into practice in Central Asia, mainly due to resistance from Central Asian governments, but there are other significant factors, including law enforcement support from other international actors, and institutional features of the OSCE such as short budget cycles that hamper strategic planning. This paper outlines how participating States that want the OSCE to support democratic policing can use opportunities, address limitations, and limit risks. Keywords OSCE, Central Asia, police-related activities, democratic governance, human rights To cite this publication: Cornelius Friesendorf, The OSCE in Central Asia: Debating Policerelated Activities, OSCE Insights Paper 1 (Baden-Baden: Nomos, 2020), at: Introduction The OSCE is committed to democratic policing and has been conducting policerelated activities for two decades.1 Democratic policing requires accountability and oversight: police must be accountable to the law rather than to government, and outside bodies such as parlia- * Dr. habil. Cornelius Friesendorf Head, Centre for OSCE Research (CORE) Institute for Peace Research and Security Policy at the University of Hamburg (IFSH) Hamburg, Germany ment and the media must be able to scrutinize the police. Democratic policing also requires that police protect human rights and are responsive to the public’s concerns.2 In practice, though, the OSCE is finding it increasingly difficult to strengthen police accountability. Shortly before losing his post as OSCE Secretary General in July 2020, Thomas Greminger stated that he was “concerned over the reduction of human rights and police accountability initiatives”.3 In Central Asia, the OSCE has faced some of its starkest challenges: Governments there seek law enforcement assistance, and welcome initiatives to 1 make police more service-oriented, but there is little appetite for democratic police governance. This paper calls for a debate on OSCE police support to Central Asian states. The first section offers an overview of OSCE police-related activities, and the second describes policing practices and structures in Central Asia. Subsequently, the paper asks the following questions: • Can the OSCE support democratic police governance in Central Asia? • Can the OSCE support states in honouring their commitments in the human dimension? • What are the limitations and risks of law enforcement support? The analysis highlights numerous obstacles to democratic policing, and to security sector governance and reform (SSG/R) more generally. Most importantly, personalized politics in Central Asia stands in contrast to liberal democratic norms. Other obstacles include Russian opposition to democratization, Chinese and Western ‘train and equip’ programmes, and the OSCE’s scarce resources and short planning cycles. These factors limit the OSCE’s ability to implement its comprehensive security agenda, and they make foreign support to law enforcement risky for local populations. The final section suggests how ‘liberal’ participating States can use opportunities and address the limitations and risks of police-related activities in Central Asia. These states should stimulate debate on policing, and proactively support democratic policing and civil society, as well as OSCE institutions, structures, and field operations, while limiting OSCE involvement in law enforcement. Moreover, the OSCE should evaluate its practices more rigorously. The OSCE and international police assistance The OSCE is a pioneer of democratic policing. Its 2008 Guidebook on Democratic Policing has become a reference document for police reform efforts around the world.4 Numerous other documents also reflect the OSCE’s commitment on this issue.5 A key agreement from 2012 states that police-related activities shall be guided by the norms, principles and standards defined by documents of the United Nations and the OSCE, such as the Charter of the United Nations, relevant UN conventions on police-related activities, the Helsinki Final Act, the Copenhagen Document, and various OSCE decisions on police-related activities. These documents emphasize, inter alia, the importance of the rule of law; respect for human rights and fundamental freedoms, including gender and minority issues; policepublic partnerships; [and] effective and accountable criminal justice systems.6 Police-related activities are primarily associated with the OSCE’s first (politico-military) dimension. However, they are also relevant for the second dimension (such as counter-corruption efforts) and the Cornelius Friesendorf 2 third (human rights-based policing). While the Permanent Council is the main decision-making body, activities are implemented by the OSCE executive structures: thematic units of the Secretariat, in particular the Strategic Police Matters Unit (SPMU) and the Border Security and Management Unit (BSMU), the field operations, and the OSCE institutions. With regard to the latter, the Office for Democratic Institutions and Human Rights (ODIHR) aligns policing with human rights standards, while the High Commissioner on National Minorities (HCNM) has promoted multi-ethnic policing, and the Representative on Freedom of the Media (RFOM) free coverage of security affairs. The first wave of OSCE police support activities focused on the Balkans. There, the OSCE has significantly contributed to democratizing the police since the early 2000 s, benefiting from factors such as broad mandates, EU membership conditionality, and a permissive attitude towards democracy promotion among participating States.7 Elsewhere in the OSCE area, conditions have been less permissive. Types of international police assistance The OSCE divides its police-related activities into two pillars: “general police development and reform”, and “threats posed by criminal activity”.8 But these categories do not distinguish between different degrees of support to police governance. This paper proposes an alternative typology that distinguishes between activities that emphasize: • democratic police governance (type 1), • better police protection of human rights (type 2), and • stronger law enforcement (type 3). These activity types vary in terms of their contribution to police oversight by external institutions such as parliament and civil society, which is important because internal oversight, by police superiors and the government, is often insufficient for addressing police misconduct. While external oversight is at the core of type 1 activities, it figures less prominently in type 2, and hardly plays a role in type 3. International actors may aid democratic police governance by supporting changes to the police legal framework, such as when they help national lawmakers in drafting bills that give parliament a stronger role. Less ambitiously, international actors may train parliamentarians in how to use parliamentary powers, or focus on civil society, such as by training journalists. Foreign reformers may also sponsor platforms where state and civil society representatives discuss police reform. International actors may improve police protection of human security, aiming to reduce police violence and corruption and thus increase protection from police, such as through anti-torture training. These activities also aim to improve police responses to issues such as domestic violence or human trafficking, i.e. increase protection by police. Type 2 activities involve pragmatic cooperation with The OSCE in Central Asia: Debating Police-related Activities 3 the police, and fostering police-public interaction, through community policing in particular. While community policing may include oversight, such as following up on how the police dealt with public complaints, it is less institutionalized than with type 1 activities. International actors may strengthen law enforcement through training and material aid, aiming to bolster the coercive capacity of the state in technical areas such as criminal investigation and border management. Democratic governance does not feature prominently in these type 3 activities, which also tend to involve transnational police interaction rather than state-society interaction. The public may indeed benefit, in that strengthening the state translates into better protection against third-party crime and violence. These three types of international support all have merits, but they also have limitations and risks. The pros and cons of specific approaches depend significantly on the local context in which they are applied. Central Asia presents particular challenges in this regard. Policing in Central Asia What constitutes democratic policing is controversial because underlying principles such as accountability require interpretation and are practised differently even across liberal democracies. Moreover, police misconduct is a problem across the OSCE area (as indicated, for example, by the Black Lives Matter protests in summer 2020). Nevertheless, democratic policing is more likely in more democratic states. Central Asian states rank low on indices of human rights and democracy. Freedom House, measuring “global freedom scores” in terms of political rights and civil liberties, classifies Turkmenistan, Tajikistan, and Uzbekistan as “not free”, in fact ranking them among the world’s least free states; Kazakhstan is also “not free” and Kyrgyzstan “partly free”.9 Methodological challenges and normative assumptions of such rankings notwithstanding, they help to understand non-democratic policing in Central Asia, which has been widely documented by international organizations, NGOs, and the media.10 Police misconduct has various causes, including institutional incentives driving the behaviour of police officers. Some Soviet-era systems remain in place, such as pressure on police to meet unrealistic crime-solving quotas, incentivizing violence against suspects and forcing confessions to achieve the numbers. Low salaries, equipment shortages, and familyunfriendly shifts for the rank and file also encourage police misconduct.11 Political pressure plays a role too: governments use the police as a tool against the political opposition. While political leaders give the police some discretion to extract illicit rents in exchange for political loyalty, corruption pyramids mean the rank and file must share these rents with their superiors. The overlapping powers of security agencies are also problematic, as indicated by the proliferation of special units. In Kyrgyzstan, the Ministry of Interior Cornelius Friesendorf 4 (MoI) has special units, as do the security service, the National Guard, and the border service.12 In Uzbekistan, the National Guard was given roles in the fields of public order and counterterrorism.13 In this system, external oversight of the police is weak. Even when formal structures are in place, their practical role is often stymied by informal norms privileging presidential administrations, national security councils, and law enforcement agencies. Thus, Central Asian parliamentarians have, over recent years, shown little inclination to challenge the executive branch. In Kyrgyzstan, oversight by the Zhogorku Kenesh was “non-systemic, incomplete and inconsistent, while parliamentary oversight of security and law enforcement agencies is even more limited and episodic”.14 MPs focused on lowlevel police misconduct rather than systematically improving police performance such as through post-legislative scrutiny.15 There were reportedly no systemic requests for documents from the police by the Uzbek parliament, not least due to the absence of an opposition party, and human rights advocates found it difficult to interact with MPs.16 The Tajik parliament largely rubber-stamped executive decisions, and MPs showed little inclination to discuss police violence, possibly because many of them were also connected to the security forces.17 Oversight by ombuds institutions was limited, too. The Kyrgyz ombuds office had significant powers and resources and investigated a large number of complaints. But there were doubts over the ombudsman’s independence, not least due to his former career in the intelligence services, and observers argued that parliament often ignored his recommendations.18 In addition, according to critics, the Uzbek and Tajik ombudspersons did not properly investigate complaints against the government and, in the case of Tajikistan, denied human rights violations.19 These conditions create opportunities but also limitations and risks for OSCE police-related activities. The following sections outline three questions that require debate. Can the OSCE support democratic police governance in Central Asia? Supporting democratic police governance aims at transforming how state institutions interact with one another and with society, making it the most ambitious type of police-related activity. Because it is so ambitious, it may be too tall an order for the OSCE in Central Asia. In fact, there is little evidence of systematic OSCE support for democratic police governance in Central Asia. In Kyrgyzstan, the government ended the OSCE Community Security Initiative (CSI) in 2016. This initiative, involving international police advisors, aimed at building trust between the police and the public and among ethnic Kyrgyz and Uzbek communities in the south, following inter-ethnic violence in 2010. During the early period of the CSI, the OSCE also provided vital support to police reform by facilitating discussions between the government and police reform experts from civil society.20 The OSCE in Central Asia: Debating Police-related Activities 5 Subsequent activities have been less ambitious. The Programme Office in Bishkek (POiB) continued to facilitate public council discussions but these produced what appeared to be rather moderate demands on the MoI. Tellingly, the POiB’s main partner for discussing oversight was that ministry, with the POiB stating that it “supported the MoI with furthering the parliamentary and civilian oversight” of law enforcement reform.21 The POiB also discussed SSG/R with the public administration academies of Central Asian states. But some Kyrgyz policing experts felt that the OSCE was too close to government and not sufficiently interacting with groups critical of the government, and that police reform was superficial.22 In Uzbekistan, President Shavkat Mirziyoyev sought a more active parliamentary role e.g. in budget oversight.23 This opened up opportunities for ODIHR, and for the Project Co-ordinator in Uzbekistan (PCUz), who co-organized a conference on “the democratization of legislation and law enforcement practice” in Uzbekistan, among other initiatives.24 But requests for technical police assistance outweighed requests for support on oversight, both from the OSCE and other international organizations.25 Understanding the limitations of the OSCE The preferences of the OSCE participating States’ governments shape the OSCE’s ability to support democratic policing. While the OSCE is more than an intergovernmental forum or instrument of states, it is a consensus-based organization, and democratic police governance tends to run counter to the interests of Central Asian governments. Resistance to democratic police governance is associated with domestic political logics. Henry Hale demonstrates how “patronal politics” based on personalized networks have dominated post-Soviet regimes, usually with presidents as their focal point, who are seen as being able to reward and punish individuals.26 Members of dominant networks hold official positions, but these networks also reach into non-state sectors, defying the distinction between state and society upon which SSG/R is based. Oversight by state institutions such as parliament tends to be ineffective because elites do not challenge patrons as long as these elites regard the patrons as strong. Informal norms revolving around personal acquaintances dominate over beliefs in abstract principles; patronal systems thus show high levels of corruption and weak rule of law. Political dynamics, even the violent toppling of governments, should not be mistaken for democratization; rather, such dynamics mean that shifting elite expectations bring new patronal networks to power. International actors, according to Hale, lack the leverage and linkages to change patronal politics in the post-Soviet space.27 For liberals, this is a depressing perspective on the prospects for democratic police governance. Western governments may try to identify inroads for democratic governance and provide extrabudgetary (ExB) funding that does not require con- Cornelius Friesendorf 6 sensus among all participating States. Nevertheless, they face formidable obstacles. A shift to effective control of the executive branch of government and its agencies would pose a challenge to dominant networks. Moreover, democratic police governance would require a strengthening of formal over informal norms, which is a slow, incremental process at best. As early as 2005, one assessment stated that the “climate for SSR in Central Asia is weak as a consequence of both the global ‘war on terror’ and the nature of political regimes that prevail across the region. Weak legislatures and judiciaries, emasculated medias and low levels of civil society activity have only reinforced the conservativeness of the Central Asian regimes”.28 Fifteen years later there was more rhetorical commitment to reform. Yet, police governance has not made great strides, supporting the view that patronalism is resilient. Controlling the OSCE Central Asian states have various options for preventing OSCE activities they do not want. Most importantly, as members of the organization with equal rights, they not only host field operations but make decisions about them and control the wording of their mandates. Mandates have become increasingly restrictive, as reflected in the revisions and modifications that also led to name changes: The OSCE Centre in Astana became the Programme Office in Astana in 2015. The OSCE Centre in Uzbekistan was transformed into the Office of the OSCE Project Co-ordinator in 2006. The OSCE Centre in Bishkek was transformed into the OSCE POiB in 2016, and the field office in Osh was closed. The OSCE Programme Office in Dushanbe replaced the OSCE Office in 2017. Turkmenistan still has the OSCE Centre in Ashgabat, but it has a limited remit. Central Asian governments not only control policy, but also its implementation. They give approval to activities, often down to the level of specific projects. They also issue interpretative statements that compel OSCE executive structures to only conduct activities explicitly covered by mandates, significantly limiting implementers’ autonomy. Russia, too, issues such interpretative statements.29 Governmental preferences also work indirectly: OSCE executive structures become risk averse. International staff at headquarters and in the field anticipate which activities Central Asian governments will endorse, and tend to err on the side of caution. Local field operations staff are exempt from rules prescribing maximum periods of employment. However, they have even more reason to be cautious than international staff. Legal protection by the OSCE, including tax exemption, is a perennial issue during negotiations of memorandums of understanding (MoUs) with host states; the OSCE’s weak legal status makes it difficult for the organization to exercise its duty of care.30 Local staff also have few chances to progress within the OSCE (a “non-career organization”) and tend to keep an eye on the job market, including government jobs, reducing their incen- The OSCE in Central Asia: Debating Police-related Activities 7 tives to advocate for politically risky governance initiatives. Another institutional obstacle to pursuing long-term objectives such as better governance relates to funding. The OSCE budget cycle is usually one year, and much funding is provided through ExB projects trickling in over the course of the year. This militates against strategic planning and the pursuit of ambitious goals such as democratic governance. An assessment of the OSCE’s SSG/R activities published in 2013 still rather accurately describes OSCE police-related activities in Central Asia: Projects are often ad hoc, based on requests from participating States and immediately available expertise, and shaped by the priorities of individual states which contribute extrabudgetary funding and seconded personnel. Consequently, projects are often not inserted within a ‘chain’ of activities that aim to achieve a broad goal – thereby considerably reducing their impact.31 Field operations may indeed plan several years ahead. But uncertainties over the extension of the mandate, the risk that host states may no longer consent even to programmes laid out in MoUs, and possible funding shortages create uncertainty. Bureaucratically, it is therefore rational to plan for the short term. If OSCE executive structures have little room for manoeuvre in promoting democratic police governance in Central Asia, what about efforts to improve police protection of human security? Can the OSCE help protect human security in Central Asia? Human security—the freedom of individuals from fear and want—falls primarily into the human dimension. Type 2 activities tackling human trafficking, prison reform, and gender-based violence touch less on the core of statehood than democratic governance, and are therefore tolerated or even sought by authoritarian states who see opportunities to gain legitimacy. Moreover, these activities can be flexibly adapted to the local context, which dovetails with the consensus against cookie-cutter solutions within the SSG/R community. Type 2 activities such as community policing also tend to be inclusive and thus play to the OSCE’s historical role of a convening power, creating forums where non-likeminded states or societal groups can find compromise solutions. For these reasons, type 2 activities have made up a large part of OSCE policerelated activities in Central Asia over recent years and were carefully tailored to government agendas. These included conferences, workshops, and roundtables, as well as training events, handbooks, awareness campaigns, and study trips for Central Asian officials.32 In Kyrgyzstan, the OSCE continued to support community policing after the end of the CSI by funding police vans (Mobile Police Reception vehicles) that facilitate police-public interaction in remote areas, and by supporting local public councils involved in the governance of this programme. The POiB also helped to improve road safety, provided training in Cornelius Friesendorf 8 new criminal justice laws, supported torture prevention and the rights of victims of human trafficking, and helped to implement a code of conduct for police. In recent years in Uzbekistan, the OSCE has supported the president’s efforts to improve state services and to address egregious human rights violations. The prevention of torture has thus become an important part of the agenda of the PCUz and other OSCE executive structures.33 Other activities covered human trafficking, and supported government efforts to create law enforcement media services and to address violent extremism and radicalization that lead to terrorism (VERLT). Even in Tajikistan, the OSCE kept human security concerns on the agenda by organizing activities focusing on issues including gender-based violence, torture prevention, and juvenile justice, and promoted community policing by funding and equipping model police stations, training police in responding to public requests, and supporting public councils bringing together police and community representatives. Challenges in improving human security Such activities provided real help to many people, as in Kyrgyzstan, where, according to the OSCE, many approached the OSCE-sponsored mobile police teams. However, empirical evidence and research findings from other fields raise questions as to whether the assumptions that seem to guide OSCE action are valid. First, OSCE support to civil society suggests that the organization assumes civil society can be empowered, and that civil society participation will change police behaviour. However, there are various issues with civil society support, including power asymmetries. Central Asian public councils serving as platforms for discussion on police reform tend to be dominated by MoI officials and the security forces. In Tajikistan, one assessment called such councils “government-run”.34 Moreover, civil society members may not represent vulnerable groups. The way those who hold power, such as male elders, deal with domestic violence, for example, may not be in line with liberal norms. The term “civil society” also implies a clear distinction between state and society that patronal systems defy. The possibility of changing police behaviour by changing their values is a second assumption that seems to guide OSCE efforts to improve human security; indeed the aspiration to change values has a long history in the CSCE/OSCE.35 The hope is that by interacting with civil society and international experts who diffuse international norms and best practices, police can be socialized into norm-compliant behaviour. Unfortunately, organization theory questions the prospect of changing police values through one-off events. Organizational cultures and routines are slow to change since they are produced and reproduced as officers internalize organizational norms at a young age, as well as through hierarchies, training, and peer pressure.36 Moreover, organizational behaviour is The OSCE in Central Asia: Debating Police-related Activities 9 shaped by politics. High-level corruption sends a signal that street-level corruption may be tolerated. Patronal networks are also likely to punish “change agents” calling for deep reform. A third assumption also seems to inform OSCE efforts: that small steps will develop their own dynamic. Police reformers may hope that model police stations will be rolled out across the country. However, the government may end or water down community policing projects, as illustrated by the Community Security Initiative in Kyrgyzstan. Police reform may go on for over a decade but not lead to democratic policing, as demonstrated by the case of Tajikistan. Central Asian states are skilled at accelerating or slowing down reform and at controlling its content. Type 2 activities therefore face these limitations, and carry with them the risk of buttressing authoritarian modernization (i.e. the efforts of illiberal states to gain legitimacy without changing patronal governance). These limitations and risks must be weighed against any improvements in human security, and are even higher with the next type of police assistance. What are the limitations and risks of law enforcement support? Law enforcement aid can improve the police’s ability to prevent and investigate crime more effectively. It may also align policing practices with human rights obligations, such as when police are enabled to rely on forensic evidence rather than on forced confessions. These type 3 activities are central to the OSCE’s work with police in Central Asia. The Annual Reports of the Secretary General on Police-Related Activities provide extensive lists of the OSCE’s involvement in fields such as organized crime, criminal investigation and analysis, crossborder cooperation in criminal matters, terrorist financing and VERLT, illicit drugs and chemical precursors, financial investigations, anti-money laundering and seizure of criminal proceeds, human trafficking, migration-related crime, border security, and cybercrime. The OSCE has provided support primarily through training and the provision of equipment. Moreover, international study tours are mechanisms for sharing best practices and for building transnational police networks. Type 3 activities are popular for several reasons. Most importantly, Central Asian governments seek law enforcement support: indeed, they complain that the OSCE does not deliver enough of it.37 Institutional drivers within the OSCE are also key. OSCE policy implementers, such as SPMU officials, often have a security background and are therefore inclined towards improving police capacity. Law enforcement support allows the Secretariat and field operations to report on activity, secure funding, and spend money quickly. Indeed, researchers have argued that instead of the OSCE socializing Central Asian states, the latter have socialized the OSCE as an instrument for preserving the status quo.38 Cornelius Friesendorf 10 Moreover, various other international actors deliver technical police aid, which puts the OSCE under pressure to do likewise in order to be seen as relevant by Central Asian states. Examples include European Union (EU) support to border management, the United Nations Office on Drugs and Crime (UNODC) police programmes, and bilateral U.S. law enforcement aid. Many programmes externalize Western domestic concerns about drug trafficking, terrorism, or illicit migration (in particular security risks emanating from Afghanistan), with the unintended effect of strengthening Central Asian power ministries.39 Russia also works closely with Central Asian security forces, while China is increasingly exporting its own policing practices there, many of which violate democratic policing norms. Law enforcement support as an inroad? The OSCE routinely includes human rights elements in its technical police aid, in line with its comprehensive security agenda. In fact, technical police aid can be an opportunity for mainstreaming human rights, such as by discussing international and national obligations not to torture detainees, or with practical guidelines such as not overtightening handcuffs. However, optimism with regards to translating comprehensive security into practice underestimates the capacity of Central Asian states to micromanage assistance, and may also overestimate the willingness of OSCE executive structures to risk antagonizing host states through creative mandate implementation. One could also argue that technical aid creates buy-in for democratic policing. However, David Bayley cautions international actors “to provide material assistance only to defray the operational costs of [democratic] reform for which there is local commitment rather than using it to induce commitment among people who are otherwise unwilling.”40 Moreover, training is not sustainable if newly learned skills are not immediately applied. In Central Asia, the domestic institutional setting militates against the application of internationally-sponsored human rights-based training. Type 3 activities may not only be ineffective for promoting democratic policing; they also involve significant risks under conditions of authoritarianism. Critics have pointed at the risk of OSCEsponsored law enforcement aid inadvertently reinforcing repression.41 Other risks are less visible, in particular buttressing the legitimacy of authoritarian governance. Central Asian states speak the language of democratic policing, making it easier for Western governments to authorize assistance. But embracing global models such as gender mainstreaming, especially without clear roadmaps and subsequent monitoring, is not the same as redistributing power. In Vienna in early 2018, for example, the Tajik interior minister presented the priority areas that needed further support from the OSCE participating States, specifically mentioning such issues as new police uni- The OSCE in Central Asia: Debating Police-related Activities 11 forms, the procurement of new equipment for the police, and the automatization of all communication processes in MoI structures. During the discussion that followed, the Minister touched upon issues of community policing and gender mainstreaming […].42 Conclusions and recommendations There is little public debate on the challenges to OSCE police-related activities in Central Asia outlined in this paper. While OSCE reporting is highly optimistic, there is no evidence that the organization has reflected on past criticism, and many assumptions remain untested. While we still lack knowledge about what works best in the field of foreign police assistance, various literatures do provide clues, if only about what does not seem to work well. Participating States concerned about OSCE policerelated activities in Central Asia could take the following steps: • Debate police-related activities. Participating States should discuss the opportunities, limitations, and risks of OSCE police-related activities within the Permanent Council, the Security Committee, and among the Group of Friends of SSG/R. They should underline that these activities are cross-dimensional and, as such, firmly anchored in the human dimension. Moreover, Western states should improve policing at home in order to create role models for other countries, and explore ways in which the OSCE can better support police reform west of Vienna, too. • Question concepts and claims of success. Rather than accepting the statements made in many publicly-available OSCE reports that downplay the difficulties of implementing OSCE commitments, reducing the concepts of democratic policing, SSG/R and capacity building to mere buzzwords, states should scrutinize whether OSCE rhetoric matches practices. States should stress, for example, that providing material and knowledge support to security forces without support for improved oversight violates the comprehensive security approach upon which the OSCE is based. In light of the challenges posed by Central Asian politics and Russian resistance to democratization, participating States should closely scrutinize whether OSCE rhetoric matches practices through better evaluation (see further below). • Identify change agents. International actors in Central Asia lack the leverage and linkages they have in other parts of the world. Moreover, democratic policing cannot take place in a silo: it depends on a broader SSG/R strategy and, indeed, on democratization, which is a long-term process even in contexts less patronal than Central Asia. Yet, the OSCE can support domestic change agents, in particular during critical junctures when political conditions are ripe (as the OSCE did in Kyrgyzstan after the 2010 revolution). The OSCE should identify the Cornelius Friesendorf 12 most promising change agents, and acknowledge that supporting them is a political, not a technical process, and as such requires a trade-off. Thus, civil society groups that share OSCE values may lack domestic influence while groups with domestic clout may be part of the regime. • Support civil society. Civil society is a crucial element for advancing the human dimension, notwithstanding cooptation and marginalization by the state. Participating States should make sure that field operations also work with civil society groups critical of the government. Central Asia has knowledgeable analysts and advocates of police reform that would benefit from systematic OSCE support such as advice on coalition-building, financial management, and report writing. Other inroads include training journalists on writing about policing, and sponsoring (and politically supporting) research on police reform and oversight – and SSG/R more generally, such as at the OSCE Academy in Bishkek. Political and financial support to Central Asian youth and educational initiatives merit special consideration as a long-term investment in institutional change. • Support OSCE executive structures. OSCE institutions, the Secretariat, and field operations need financial support for type 1 and type 2 projects, including through the ExB process. They also need political backing, such as assurances that OSCE staff, including local staff, will not be left alone if they displease host states, for example by working with reform-minded NGOs. • Limit OSCE law enforcement support. The risks of type 3 activities, which aim at strengthening the police, may outweigh the benefits for Central Asian populations. Making law enforcement aid conditional upon improved governance or human security protection, or ‘mainstreaming’ human rights into law enforcement aid, is no panacea. Central Asian states may put up democratic facades and micromanage projects, and the consensus principle leaves little room for EU-style conditionality. Reducing or ceasing political support and funding for ‘train and equip’ programmes is in line with the do no harm principle. Democratic states should also rethink law enforcement aid provided to Central Asia through other international organizations and bilateral programmes, to reduce competitive pressure on the OSCE. One might object that limiting law enforcement assistance deprives Western states of the chance to cooperate with Central Asian states against transnational security risks. But this view underestimates the negative consequences of unprincipled assistance, not only for human rights locally but for Western interests too. After all, repression creates a fertile ground for crime and political violence. There is a main precondition for debating and improving OSCE police-related activities as suggested above: evaluation. Participating States should scrutinize the The OSCE in Central Asia: Debating Police-related Activities 13 activities and underlying logics of the Secretariat and field operations, and systematically request, analyse, and debate the internal evaluations of the Office on Internal Oversight. Field operations’ selfevaluations also deserve more scrutiny, and extra-regular audits may additionally enhance transparency. Furthermore, the OSCE might benefit from more outside independent evaluations and innovative methodologies such as those used by ethnographers. Thick descriptions or surveys of public trust in the police raise questions of causality. Nevertheless, such approaches often produce better insights than box-ticking and new public management models. Participating States should also discuss how to publish information that would allow for an open and informed debate. The OSCE produces large amounts of information but publicly available documents leave many questions open. The public knows little about theories of change and underlying assumptions, which states provide how much extrabudgetary funding for which projects, or risk management plans. One way to address this problem is to create oversight bodies for specific projects and programmes that share their findings with the public. Most importantly, assessments should focus on whether OSCE police-related activities have changed police behaviour (outcome) and third-party crime and violence (impact). These measures are notoriously difficult to obtain due to contextual factors such as economic or demographic change. In the case of OSCE police-related activities, these difficulties are aggravated by further factors, including scarce reliable data provided by Central Asian governments, and police aid by other international donors. Soft power mechanisms such as norm diffusion, where the OSCE is strongest, are equally difficult to measure. Although evaluations of OSCE policerelated activities will necessarily remain inconclusive, the central rationale of these activities should not be beyond scrutiny. There is a risk in simply assuming that the OSCE cannot solve complex problems in a difficult political environment with a small budget, but that without the OSCE, policing in Central Asia would be worse. While this may be true, untested assumptions rarely make good policy. Notes 1 I thank the interviewees for sharing their views with me. For comments on an earlier draft, my thanks go to colleagues from the IFSH and to the external reviewers, and to Caroline Taylor for her editorial suggestions. 2 See David Bayley, Changing the Guard: Developing Democratic Police Abroad, Oxford: Oxford University Press, 2006, pp. 18-22. 3 Organization for Security and Co-operation in Europe (OSCE), Annual Report of the Secretary General on Police-Related Activities 2019, 2 July 2020, p. 2, at: /6/456052.pdf. 4 OSCE, Guidebook on Democratic Policing by the Senior Police Advisor to the OSCE Secretary General, 24 January 2007, available at: ecretariat/23804. 5 For an overview, see OSCE, Annual Report of the Secretary General on Police- Cornelius Friesendorf 14 Related Activities 2018, 8 August 2019, pp. 157-162, available at: https://www.osc 6 OSCE, Decision No. 1049: OSCE Strategic Framework for Police-Related Activities, PC.DEC/1049, July 26, 2012, p. 3, at: 7 Steffen Eckhard, International Assistance to Police Reform: Managing Peacebuilding, Basingstoke: Palgrave Macmillan, 2016, chapter 3. 8 OSCE, Annual Reports of the Secretary General on Police-Related Activities, available at: 108393. 9 Freedom House, Countries and Territories, Global Freedom Scores, at: https://fr d/scores?sort=asc&order=Total%20Score %20and%20Status, accessed on 5 June 2020. 10 See, for example, International Partnership for Human Rights: Torture in Central Asia: Time to Break the Vicious Circle, 26 June 2019, at: https://www.iphron ak-the-vicious-circle.html. 11 For an early analysis that still holds traction, see International Crisis Group, Central Asia: The Politics of Police Reform, 10 December 2002, at: https://d2071andv olitics-of-police-reform.pdf. 12 Author interview with security expert, Bishkek, 7 March 2019. 13 Umida Hashimova, The National Guard of Uzbekistan: Rising Profile, Lingering Problems, The Diplomat, 22 February 2019, at: 2/the-national-guard-of-uzbekistan-risingprofile-lingering-problems/. 14 Aida Alymbaeva/Rustam Burnashev/ Grazvydas Jasutis/Parviz Mullojanov/ Richard Steyne/Farkhod Tolipov, Parliamentary Oversight of the Security Sector: Case Studies from Central Asia, Geneva Centre for Security Sector Governance (DCAF), 2020, p. 28. 15 Author interview with Kyrgyz NGO, Bishkek, 11 March 2019. 16 Author interviews with an NGO member and a human rights group, Tashkent, June 2019. 17 See Alymbaeva et al., cited above (Note 14), pp. 47–54. 18 Ibid., p. 33; author interviews with Kyrgyz NGOs, Bishkek, March 2019. 19 Author meetings with human rights defenders, Tashkent, June 2019; communication with a Tajik researcher, 2019. 20 See Erica Marat, The Politics of Police Reform: Society against the State in Post- Soviet Countries, Oxford: Oxford University Press, 2018. 21 See OSCE, cited above (Note 5), p. 120. 22 Interviews with NGO representatives, Bishkek, March 2019. 23 Interview with Western embassy official, Tashkent, June 2019. 24 See OSCE, cited above (Note 5), p. 145. 25 Interview with UN representatives, Tashkent, June 2019. 26 Henry E. Hale, Patronal Politics: Eurasian Regime Dynamics in Comparative Perspective, New York: Cambridge University Press, 2015. 27 Ibid., pp. 442-445. 28 Organization for Economic Co-operation and Development (OECD), DAC Guidelines and Reference Series: Security System Reform and Governance, 31 May 2005, p. 24, available at: https://www.oec m-reform-and-governance_978926400788 8-en. 29 For an example of such a statement, see OSCE, Decision No. 1251: Programme Office in Dushanbe, PC.DEC/1251, 1 June 2017, available at: https://www.osce. org/permanent-council/322446. 30 See Mateja Steinbrück Platise, Carolyn Moser, and Anne Peters (eds.), The Legal Framework of the OSCE, Cambridge: Cambridge University Press, 2019. 31 DCAF, The Role of the Organization for Security and Co-operation in Europe (OSCE) in Security Sector Governance and Reform, 20 December 2013, p. 23, at: The OSCE in Central Asia: Debating Police-related Activities 15 publications/documents/OSCE_SSG_Ma pping_Study_Final_Report.pdf. 32 OSCE, cited above (Note 8). 33 On the OSCE’s training Uzbek officials on human rights in counter-terrorism investigations, see OSCE, cited above (Note 5), p. 140. 34 Olimjon Bakhtaliev, The Evolving Role of Civil Society in Peace and Security in Tajikistan: Challenges and Opportunities, Saferworld, 17 March 2020, at: https: // -civil-society-in-peace-and-security-in-tajik istan-challenges-and-opportunities. 35 Bruce Cronin, Creating Stability in the New Europe: The OSCE High Commissioner on National Minorities and the Socialization of Risky States, Security Studies 12:1, 2002, pp. 132-163. 36 Cornelius Friesendorf, How Western Soldiers Fight: Organizational Routines in Multinational Missions, Cambridge: Cambridge University Press, 2018, chapter 4. 37 See Sarvinoz Rukhullo, Programma reformirovaniya militsii: Voprosy bez otvetov, [Police Reform Programme: Unanswered Questions], Radio Ozodi, 23 February 2020, at 30449643.html. 38 Karolina Kluczewska, Benefactor, Industry or Intruder? Perceptions of International Organizations in Central Asia – The Case of the OSCE in Tajikistan, Central Asian Survey 36: 3, 2017, pp. 353-372. 39 Filippo De Danieli, Counter-Narcotics Policies in Tajikistan and Their Impact on State Building, Central Asian Survey 30: 1, 2011, pp. 129-145. 40 Bayley, cited above (Note 2), p. 64. 41 David Lewis, Reassessing the Role of OSCE Police Assistance Programing in Central Asia, Occasional Paper Series No. 4, Open Society Foundations, April 2011, available at: https://www.opensocietyfou e-osce-police-assistance-programing-centr al-asia. 42 OSCE, cited above (Note 5), p. 124. Cornelius Friesendorf 16


This contribution calls for a debate on the effects of OSCE police-related activities in CentralAsia. Drawing on a typology of internationally-supported police aid, it outlines three questionsthat deserve more scrutiny by participating States and civil society: Is the OSCE able to supportdemocratic police governance in Central Asia? To what extent can the OSCE help improve humansecurity? What are the limitations and risks of law enforcement support? Tentative evidencesuggests that the OSCE faces significant challenges in translating its commitment todemocratic policing into practice in Central Asia, mainly due to resistance from Central Asiangovernments, but there are other significant factors, including law enforcement support fromother international actors, and institutional features of the OSCE such as short budget cyclesthat hamper strategic planning. This paper outlines how participating States that want theOSCE to support democratic policing can use opportunities, address limitations, and limit risks.



The contributions to the 2020 edition of OSCE Insights examine the various crises the OSCE faced during that year. Themes include the efforts of the Minsk Group to manage the conflict over Nagorno-Karabakh and the implications of anti-coronavirus measures for the OSCE Code of Conduct on Politico-Military Aspects of Security. Furthermore, authors analyse OSCE conflict cycle tools, the OSCE’s role in the fight against antisemitism, the increasingly limited space for supporting democratic police governance in Central Asia, trust-building in the field of arms control, societal views on the conflict in Eastern Ukraine, relations between the OSCE and the Council of Europe, and Kazakhstan’s aspirations for hosting a connectivity center. With contributions by Andrew Baker, Cornelius Friesendorf, Frank Evers, André Härtel, Marietta Koenig, Sebastian Mayer, Michael Raith, Filip Ejdus, Alexandre Lambert, Thomas Schmidt, Marina Dolcetta Lorenzini, Anna Hess Sargsyan, Philip Remler, Richard Giragosian, Sergey Rastoltsev and Benjamin Schaller.


Die Beiträge der OSCE Insights untersuchen verschiedene Krisen, mit denen die Organisation während des Jahres 2020 konfrontiert war. Themen sind u.a. die Vermittlung der Minsk-Gruppe im Konflikt über Bergkarabach und die Frage, inwiefern Anti-Corona-Gesetze und Praktiken mit dem Verhaltenskodex zu Politisch-Militärischen Aspekten der Sicherheit vereinbar sind. Die Autorinnen und Autoren untersuchen weiterhin Instrumente der OSZE beim Konfliktzyklus, die Rolle der OSZE beim Kampf gegen Antisemitismus, den immer geringeren Spielraum bei der Unterstützung demokratischer Polizeiarbeit in Zentralasien, Vertrauensbildung in der Rüstungskontrolle, gesellschaftliche Wahrnehmungen des Konflikts in der Ostukraine, die Beziehungen zwischen der OSZE und dem Europarat sowie die Bemühungen Kasachstans um ein Konnektivitätszentrum. Mit Beiträgen von Andrew Baker, Cornelius Friesendorf, Frank Evers, André Härtel, Marietta Koenig, Sebastian Mayer, Michael Raith, Filip Ejdus, Alexandre Lambert, Thomas Schmidt, Marina DolcettaLorenzini, Anna Hess Sargsyan, Philip Remler, Richard Giragosian, Sergey Rastoltsev und Benjamin Schaller.