Second-Generation SSR or Unending Violence in Haiti?

In the literature on security sector reform (SSR), there is a debate between mainstream analysts of post-9/11 approaches, who argue that second-generation SSR (2GSSR) is emerging in some contexts, and post-colonial critics, who argue that despite discursive changes, Western-driven, securitized practices still dominate. This article bridges those views, building on Larzillière’s (2016) idea that the diffusion of global power offers southern actors room for maneuver to construct alternatives to first-generation SSR. Grounding the debate in the Haitian case, the paper explores the mixed record of police and rule of law reform in that relatively permissive environment. It concludes that while some stakeholders have used Haiti’s margin for maneuver to advance 2GSSR, powerful actors are privileging narrow security goals in a political and socioeconomic context that greatly constrains the consolidation of a new approach. The article concludes by situating the uncertain gains that have been made in Haiti in comparative thinking about the uneven prospects for 2GSSR in diverse fragile and conflict-affected states (FCAS).


Introduction
In the literature on security sector reform (SSR), there is a debate between mainstream analysts of post-9/11 approaches, who argue that second-generation SSR (2GSSR) is emerging in some contexts, and postcolonial critics, who argue that despite discursive changes, Western-driven, securitized practices still dominate. This article bridges those views, building on Larzillière's (2016) idea that the diffusion of global power offers southern actors room for maneuver to construct alternatives to first-generation SSR. Grounding the debate in the Haitian case, the paper explores the mixed record of police and rule of law reform in that relatively permissive environment. It concludes that while some stakeholders have used Haiti's margin for maneuver to advance 2GSSR, powerful actors are privileging narrow security goals in a political and socioeconomic context that greatly constrains the consolidation of a new approach. The article concludes by situating the uncertain gains that have been made in Haiti in comparative thinking about the uneven prospects for 2GSSR in diverse fragile and conflict-affected states (FCAS).

Economic Co-operation and Development's (OECD's) Guidelines on Security System
Reform and Governance (OECD 2004) was being selectively applied in FCAS. They suggested that the SSR model tended to privilege the technical reinforcement of state security agencies based on a Western model of the state, resulting in major conceptualcontextual gaps and a poor record of performance. What was needed, according to the authors in Sedra (2010), was a shift to 2GSSR, which in practice meant more context-appropriate, locally rooted, flexible and long-term approaches that could transform the governance of security institutions and change state-society relations.
These ideas gradually seeped into policy thinking. The OECD Development Assistance Committee (DAC) and various think tanks published guidance on how to close policyto-practice gaps by taking a system-wide approach and engaging local actors in more politically astute ways (DCAF-ISSAT 2015;OECD 2016). However, numerous reports documented enduring challenges, leading some to conclude that international agencies should reduce their expectations of what is realistic in FCAS (Ball and van de Goor 2013).
Within these diverse policy debates, evidence of 2GSSR also began to emerge. Researchers at the Centre for Security Governance, at the forefront of the 2GSSR concept, looked at the extent to which SSR was actually being practiced in four FCAS: Bosnia-Herzegovina (Marijan 2017), El Salvador (Rivard Piché 2017a), Sierra Leone (Bangura 2017) and Timor Leste (Dewhurst and Geising 2017). They concluded that hybrid 2GSSR -namely 2GSSR with enduring features of first-generation approaches -was emerging in some FCAS. In relatively permissive contexts like post-war Sierra Leone and Timor Leste, that trend was advanced, with the incorporation of local committees, traditional chieftains and other non-state actors in community-policing initiatives and longerterm processes of social change. In Bosnia, a more classical state-centric approach endures, yet paradoxically, the authors suggested that such an approach is appropriate in that context. In a more recent special issue of the Journal of Intervention and Statebuilding on this subject, Sedra (2018) and Donais (2018) concluded that despite shifts towards hybrid 2GSSR in some contexts, state-centric or "good enough" approaches remained the default mode in many FCAS.
This literature has been largely atheoretical. Early on, Hänggi and Bryden (2004) and Wulf (2004) suggested ways of conceptualizing the forms of SSR that are possible in different contexts along the war-to-peace and authoritarian-to-democratic continuums. Building on their work and adding ideas from constructivist theory, Baranyi and Salahub (2011) showed how SSR and its outcomes were negotiated at multiple levels and therefore varied according to the global/regional geopolitical context; the military, political, economic and cultural conditions on the ground; as well as the agency and power relations of stakeholders in distinct FCAS. That framework led them to cautious conclusions about the possibilities for substantive SSR, even in what appeared to be enabling contexts in Burundi, Haiti and South Sudan. Their conclusions also converged with critical analyses of the forms of hybrid peacebuilding and local inclusion emerging in distinct FCAS -from conservative to emancipatory approaches (Richmond 2011;Donais 2012). As discussed at the end of this section, the notion of a spectrum of hybrid outcomes helps us sharpen Sedra's idea (2018) that different forms of 2GSSR are also being built in FCAS.
Alongside the development of 2GSSR discourse, the post 9/11 context also generated critiques of international intervention as becoming increasingly securitized. Duffield (2001) cautioned that the merging of security and development practices reflected changing Western strategies of global governance. Drawing on the work of Foucault, he argued that post-Cold War, Western-led attempts to combine technologies and funds from military, development and diplomatic institutions essentially perpetuated 'biopolitical control' over populations in conflict-affected Southern societies. Those efforts were unlikely to benefit 'the governed' in the South or in the North. Though he did not focus on SSR per se, Duffield (2007) extended his analysis by arguing that said strategy lead to 'unending war' since it could not resolve the contradictions that perpetuated conflict, displacement and extreme inequities. Bendaña (2008), Chandler (2007) and Saliba-Couture (2012) leveled similar critiques at the securitization of development after 9/11. More recently, Larzillière (2016) concurred with Sedra et al. (2010) that since the US launched its War on Terror after 9/11, the technocratic priorities of training and equipping security forces for counterterror and other security operations have trumped priorities like democratic accountability, notably in the Middle East. She used the example of Lebanon to illustrate the social construction of those priorities through negotiations between powerful stakeholders -from the Lebanese military and Hezbollah to Western security agencies -despite the efforts of civil society organizations. Looking beyond Lebanon, Larzillière concluded that while this Western-led, technocratic and statecentric approach tended to predominate in Southern contexts since 9/11, changing international conditions offered room for maneuver: "The increasing rejection… of Western development policies, perceived as another way of implementing a Western geopolitical agenda, can be analysed from a counter-hegemonic perspective in which labels [such as SSR] used in development policies are transformed, adapted or rejected" (2016: 164; author's translation).
As illustrated in Figure 1, this article retains Larzillière's idea that global changes may open spaces for counter-hegemonic alternatives and links it to Sedra's notion that different hybrids of 2GSSR are emerging in FCAS. It refines their views by specifying a spectrum of possible outcomes -from conservative to transformative 2GSSR. Between those variables, it inserts the power-laden agency of key stakeholders -from core actors like senior security officials, to less powerful actors like human rights organizations. I retain post-colonial concerns about neocolonial continuity and biopolitical control to analyse the limits of current SSR practices. Yet our central goal is to understand the agency, institutional and contextual factors shaping different forms of 2GSSR and wider outcomes (hence the '+' in Figure 1), as well as the constraints preventing transformative change in distinct FCAS, without assuming either outcome a priori.
The article frames this perspective as ' critical institutional' rather than ' constructivist' for several reasons. Our earlier (Baranyi and Salahub 2011) use of constructivism assumed  that material relations of power were central to the social construction of SSR processes. Yet that understanding has been sidelined in the literature on FCAS, where scholars often use constructivism to underscore how formal and informal norms shape international processes (for example, Donais and McCandless 2016). Moreover, the notion of ' critical institutionalism' more clearly reveals its roots both in materialist critical theory (Cox 1987) and in the historical institutionalism of Evans et al. (1985). It builds on MacKay et al.'s (2011) feminist institutionalism and answers Krier's (2009) call for a critical institutionalism that recognizes the importance of institutions while examining the political and economic contexts (including the power relations) in which they evolve. In this article, we demonstrate how such an approach can illuminate the historical construction of 2GSSR in Haiti and in other FCAS.

The Haitian Case and Our Methodology
We ground this exploration in Haiti. Why? Compared to highly restrictive contexts like Iraq or Afghanistan (Sedra 2007), contemporary Haiti is a somewhat permissive context for 2GSSR. With deep governance, economic and environmental challenges, Haiti is ranked 12 th on the Fragile States Index (FFP 2018). Yet Haiti is not at war, some national leaders and institutions are committed to institutional reforms, and international intervention takes place under a UN umbrella in which somewhat counterhegemonic regional powers like Brazil have played key roles (Baranyi and Sainsine 2016). Like Lebanon, Sierra Leone and Timor Leste, Haiti is therefore a relevant setting in which to explore the latitude for 2GSSR. The literature on SSR in Haiti can be divided into a range of perspectives. The Haitian Police and the UN suggest that the Haitian National Police (HNP) has advanced over the past decade but that slower reforms in the judiciary, fiscal shortfalls and other governance challenges constrain progress (Maximé 2013;PNH 2016;UNSG 2017a and. Other academics and activists argue that despite large international investments, police and other rule of law reforms have failed because they reflect Western priorities, the securitization of development and neocolonial ' occupation' (Holly 2011;Brisson-Boivin and O'Connor 2013;Rehmonco 2017). Other scholars have used various analytical lenses to argue that while police reform has advanced, it is fraught with tensions between hard/technocratic and soft/social goals (Donais 2011;Fortin and Pierre 2011;Sainsiné 2015 and; that some elites have blocked reforms, perpetuating unequal access and violence (Berg 2013;Édouard 2013;Édouard and Dandoy 2016b); and that weak state regulation has left the field open for private security firms, community-based brigades and traditional actors, which vary in their commitment to the rule of law (Burt 2012;Hauge, Doucet and Gilles 2015;Édouard and Dandoy 2016b;Rivard Piché 2017b).
Using the critical institutional lens illustrated in Figure 1, the next section updates the evidence of uneven progress by the HNP, the institution at the center of SSR since 1995. The following section situates police development in Haiti's broader rule of law and political-economic contexts. In the last section, we conclude that while a transnational network of reformers has used Haiti's margin of maneuver to advance elements of 2GSSR, they are constrained by powerful actors who privilege hard security goals in a political and socioeconomic context that seems less and less favorable to a transformative approach. Those tendencies could lead to more violence over the coming years. We end by drawing on this case and on critical institutional theory to reflect on the varied prospects for 2GSSR in a range of FCAS today.
Research methods for this article included a review of official documents and their triangulation with data from scholarly publications, public opinion surveys, reports by the National Network for Human Rights (RNDDH) and the Office for the Protection of Citizens (OPC); media coverage; as well as 75 confidential interviews with key informants in the HNP, other government institutions, civil society organizations, the UN and selected bilateral partners. Half of the interviews and field-based observations were conducted between May 2014 and June 2016, with Professor Yves Sainsiné and his Haitian students -in the North, Northeast, South and West departments. Our original data on HNP performance was published in two French reports Sainsiné 2015 and and discussed with stakeholders at the OPC in 2015; at RNDDH and at the HNP academy in 2016; and at a conference organized by the Université d'Etat d'Haïti in 2017. These reports were discussed with officials, academics and others in Ottawa in 2016. The article is also informed by my ongoing engagement with the HNP and four other Haitian state and non-state institutions in the rule of law sector.
This article makes three main contributions to the literature on 2GSSR: it presents data on the HNP that has not yet been shared, in particular in English-language publications; it integrates that data with the analysis of stasis/change in other institutions like the judiciary and in the wider Haitian political economy (which are insufficiently examined together in the literature); and it also uses a critical institutional lens to guide the case study and its positioning on the spectrum of conservative to transformative SSR efforts in FCAS.

The HNP and Its Uneven Development A quick retrospective
Since much has been written about the history of security institutions in Haiti, this subsection merely summarizes key milestones. The HNP was created in 1995 as part of the Aristide government's effort to dismantle the vestiges of the Duvalier regime, including its Armed Forces, and build the democratic state enshrined in the 1987 Constitution (Mendelson-Forman 2006). Despite the HNP's advances, its development was undermined by incoherent Western influences and market-oriented socioeconomic policies, as well as President Aristide's attempt to use the HNP to maintain power (Fatton 2014). By the time Aristide was forcibly removed in February 2004, many HNP had left and only 5,000 remained on duty (Fortin and Pierre 2011).
Although the development of the HNP was situated as the core of a wider strategy in the immediate post-Aristide period, the contested environment following the US-led military intervention greatly constrained progress, even after the UN Mission for Stabilization in Haiti (MINUSTAH) was deployed. The relatively fair election of President Préval in 2006 enabled the relaunch of the HNP reform agenda, based on a plan designed by the UN and fine-tuned with the new Préval government. By 2010, HNP ranks grew to 9,500 personnel and experienced significant qualitative changes, contributing to decreasing crime rates and increasing public confidence. Yet human rights organizations and independent analysts voiced concerns about the risks of reinforcing the HNP's infrastructure and repressive capacities, without complementary changes in the domains of justice, democratic governance and socioeconomic inclusion (Donais 2011;Fortin and Pierre 2011). The catastrophic earthquake in January 2010 also negatively affected the HNP, damaging much of its infrastructure and human capital, while undermining progress on crime control (Muggah and Kolbe 2011;Dumas 2015). Within this context, a new HNP Development plan was launched in 2012.

The 2012-2016 HNP plan and its uneven implementation
The second HNP plan (CSPN 2012) was qualitatively different from its predecessor. Although its design was still led by MINUSTAH, senior HNP and other Haitian officials were consulted through workshops and drafting sessions. The plan was based on a logical framework with fairly clear objectives linked to 70 activities and outcome indicators. Those were grouped into eight areas: 1) institutional development -from decentralizing HNP units to reinforcing internal communications; 2) the recruitment and training of human resources, including women; 3) crime and violence prevention; 4) the Judicial Police; 5) the Inspector General; 6) strategic planning and budgeting; 7) operational capacities -from crowd control units to the coast guard; and 8) prison administration. A joint monitoring and evaluation process was established, through which MINUSTAH and the HNP General Directorate would periodically assess and report on progress. The plan is sometimes called a 'second-generation' document (Servilien 2017: 5), confirming the relevance of analyzing it through Sedra et al.'s (2010) prism of 2GSSR.
For the first years, MINUSTAH led the production of periodic reports. In 2015-2016, the Strategic Planning Unit of the HNP General Directorate led the preparation of the fifth report (PNH 2016). The HNP and MINUSTAH also co-managed the external summative evaluation (Servilien 2017). An analysis of those documents shows that despite progress, only about 65 per cent of the activities were completed or in an advanced state of execution by the end of 2016. Areas where there had been the most progress included building infrastructure and recruiting and training personnel (particularly for security operations like antinarcotics and crowd control), which fall within the 'train and equip' priorities of firstgeneration SSR. Other advances go beyond that frame, for example, on gender equality and strengthening the Inspector General's disciplinary function. Areas where there had been the least progress included the rollout of decentralization and community policing.
Those documents also revealed an increase in the analytical capacity and self-confidence of the HNP General Directorate. For example, the fifth report offered a perceptive analysis of 'institutional and social threats,' including possible attempts by the government to politicize the HNP; human rights violations against civilians and attacks on the police; as well as the persistent problem of under-financing by the Haitian state and by international donors (PNH 2016).
The new HNP Strategic Development Plan for 2017-2021, adopted by the Superior Council of the National Police in July 2017, includes measures to address such challenges. The HNP Directorate General led the formulation of the new plan, informed by a consultative process with senior Haitian officials and key international partners (CSPN 2017). The UN Secretary-General's latest report on Haiti (UNSG 2018b) suggests that during the first year of the plan's implementation, 100 of the 108 planned activities had been initiated, two had been completed and six had not been started. These public summaries are useful, but the full plan and first progress report have only been shared with a small network of officials, an issue to which we shall return in the section on national ownership.
Our research enables us to identify causal links between the HNP's progress on strategic planning, the institutional agency of senior HNP officials and their UN counterparts, and contextual factors like the UN's need to demonstrate results in the security sector to justify its reduced military presence in Haiti. Yet to understand the subtleties of HNP development, we need to look beyond those macro-level trends. As such, the remainder of this section focuses on five sets of challenges: 1) recruitment and training; 2) territorial deployment and community policing; 3) internal discipline and the use of force; 4) prison reform; and 5) national ownership. Gendered dynamics and local variations are integrated throughout the analysis. These are obviously not the only challenges facing the HNP or Haiti. Other priorities such as improving administration, communications, infrastructure, mobility and equipment, are extensively addressed in the 2017-2021 plan. We selected our discussion points based on our ability to gather sufficient high-quality data in those areas and because of their relevance to global debates. Baranyi: Second-Generation SSR

Recruitment and training
One of the most publicized targets in the 2012-2016 plan was to increase the number of uniformed police from about 10,000 in 2012 to 15,000 by the end of 2016. According to the final evaluation, the HNP achieved 78.4 per cent of this target (Servilien 2017: 12), as illustrated in Figure 2.
By mid-2017, the HNP had attained a ratio of 1.36 police officers for 1,000 citizens (UNSG 2017b: 4) -or 1 police for 733 inhabitants based on the HNP's calculations. The delay in reaching the recruitment target can partly be explained by improved human resource management systems, which generated more robust data and increased suspensions or early retirements from the force. The slow increase in personnel also reflected ongoing debates between HNP leaders and their international partners on how to balance quantitative growth with qualitative improvement.
Another concern was the difficulty of recruiting and retaining uniformed female personnel. Despite considerable efforts, women only represented 7.5 per cent of uniformed personnel by the end of 2016, short of the 11 per cent target and light years away from the 30 per cent quota enshrined in the 2012 Constitutional reforms (Servilien 2017: 12). According to Haitian and foreign gender equality champions whom we interviewed, machismo and poor material conditions for women within the force converged with wider social norms to deter many women from applying to -and staying with -the police service.
Before 2012, HNP officer training took place in the US, Canada and other foreign countries, perpetuating doctrinal fragmentation in the force. To remedy this problem, Canada helped the HNP build a new police academy, establish a competencybased and gender-sensitive curriculum and train Haitian trainers. Despite challenges such as the temporary suspension of teaching in 2015 due to budgetary difficulties, 325 officers including 26 women (eight per cent) had completed in-house training by 2018 (GAC 2018). Innovative work-study and ongoing graduate support programs were also established. Yet the low percentage of women graduates and the difficulty of securing stable funding from the Haitian state remained major challenges.
This analysis of HNP recruitment and training suggests that there has been progress towards 2GSSR -particularly in combining quantitative increases with qualitative advances such as the promotion of women's inclusion and gender-awareness. We clearly see the convergent agency of reformers at the HNP and of their international partners at work in this domain. However, we also observe constraints on their efforts in the resistance of some HNP officers to qualitative change (rooted in traditional social norms) and enduring budgetary constraints.

Territorial deployment and community policing
The 2012 HNP Development Plan acknowledged that greater efforts were required to reinforce the HNP's presence beyond the main cities, particularly in small communities lacking even the legally established minimum of three police agents. Based on that plan, special HNP units like the marine base in Les Cayes and the mechanical workshop in Port de Paix were refurbished. In 2015, about 250 members of the 25th promotion of police agents were also deployed outside the West department, 1 mainly to reinforce departmental maintenance of order units (UDMO) and motorized intervention brigades (BIM). Strengthening those units was seen as a way of providing extra monetary incentives to new recruits, addressing security problems in the provinces and demonstrating the HNP's capacity to fill the security gaps left by MINUSTAH's downsizing Sainsiné 2015 and. However, our field research confirmed the persistence of very low police-to-population ratios in the South (about 1:2666) and North (about 1:1785) departments even after increased deployments in 2015. It also suggested that the reinforcement of UDMO and BIM did not translate into more HNP deployments to small communities, or the deployment of more women to the field. RNDDH's research (2016) on 153 HNP posts in seven departments also documented the persistence of enormous human resource and material limitations beyond Port-au-Prince and the West department. The 2017-2021 Strategic Development Plan (CSPN 2017) used a "departmental functionality index" based on six indicators to assess the capacities of the HNP in each department. On that basis, it concluded that while the HNP was relatively capable in the West and South departments (respectively receiving 5.2/6 and 4.3/6), in most other departments the HNP was weak (about 3/6), while in the Northwest and Grande Anse, it was extremely weak (respectively receiving 2.2/6 and 2.7/6). The plan, thus, commits the HNP to increasing deployment beyond the capital and secondary cities and linking the deployment aims to improving material conditions to attract more female recruits.
Additionally, some HNP leaders and their international mentors recognized that the force should work in partnership with communities to adapt local policing strategies to prevent and reduce crime. The 2012-2016 plan set out the principles of community policing (CP) and committed the force to elaborate and implement a CP strategy at the national level. MINUSTAH and HNP reports (2014 documented advances during that period: increased training on community policing in the HNP school and HNP academy; the creation of the educational police (EDUPOL) and tourism police (POLITUR); piloting community policing in Jacmel, Les Cayes and Port-au-Prince. Furthermore, some specialized units such as the Gender-Based Violence Units and the Minors' Protection Brigades, conducted CP-style activities but received little political or material support. The lack of prioritization of and support to CP has been evidenced in previous research, which documented that some senior HNP officials saw community policing as a public relations exercise or as a distraction from the core business of getting tough on crime Sainsiné 2015 and.
Through increased deployments and the pursuit of CP, progress towards 2GSSR can be observed. These changes were due to the convergent agency of HNP reformers and their international partners. Yet here, too, we observe major constraints, such as the HNP and MINUSTAH tendency to privilege special-order forces in deployments outside the capital. These countertrends may be rooted in their bias towards robust police responses, but they were accentuated by the desire to demonstrate HNP capacity to ensure public security before the end of MINUSTAH's mandate in October 2017.

Internal discipline and the use of force
The work of the Inspector General (IG) is held up as a symbol of the HNP's advances. Operating from new headquarters in Delmas Baranyi: Second-Generation SSR or Unending Violence in Haiti?
Art. 2, page 9 of 19 2, in late 2013 a new IG and his team rebooted the vetting of HNP personnel and regularly recommended that personnel with dubious records be removed from service and/or prosecuted. They revived performance monitoring across the country, established a 'green line' for citizens to report problematic conduct by the police and held regular media briefings. Yet in 2017, the joint evaluation of the overall HNP plan concluded that despite those advances, the IG still has difficulties extending control in all regions and financial constraints continue to affect its effectiveness, particularly in distant regions (Servilien 2017: 18).
While praising the IG's professionalism, human rights organizations have also questioned its capacity to control certain HNP units, particularly when dealing with sociopolitical protests or criminal gangs. In the first category, we find incidents such as the excessive use of force by UDMO and BIM in February 2014 against citizens protesting the expropriation of peasants' lands for tourism development on Île à Vache in the South (RNDDH 2014); the excessive use of force by UDMO and BIM in December 2014 against residents of Ounaminthe and Fort Liberté in the Northeast who were protesting their limited access to electricity (Baranyi and Sainsiné 2016); and incidents in several parts of the country during the electoral year of 2015, when the Maintenance of Order Company and the new Departmental Operation and Intervention Brigade (BOID) allegedly used excessive force against members of political parties protesting allegations of fraud during the elections. In November 2015, the OPC called on the IG to discipline the accused units; the IG and HNP High Command responded by removing the BOID from crowd-control operations linked to the elections and sending them back to the HNP academy for more human rights training (Baranyi and Sainsiné 2016).
Such incidents decreased after more legitimate elections in 2016-2017, yet abuses linked to anti-gang operations continued. For example, in October 2017, in retaliation for the murder of a policeman, a BOID unit entered Lilavois (in the West) and allegedly killed three gang members, mistreated eight others and burned down five buildings. The IG investigated, three BOID members were detained and their files were forwarded to a local tribunal (RNDDH 2018 andUNSG 2018a). In November 2017, the HNP conducted an anti-gang operation in Grand-Ravine, also in the West. After losing two of their own to gunfire from gang members, they allegedly killed eight gang members and others -including personnel of the school in which the operation occurred. The IG investigated, an HNP agent was suspended and several HNP files were forwarded to the local tribunal. In August 2018, the accused HNP personnel had still not faced justice despite the advocacy efforts of human rights organizations and the UN (RNDDH 2018b and UNSG 2018b). The impunity of security personnel is a major obstacle to the enhancement of rule of law and 2GSSR in Haiti.
The measures taken by the HNP IG to increase the transparency of his unit's work and discipline police personnel accused of serious offenses, are further evidence of progress towards 2GSSR. He would not have been able to advance this far without the support of the HNP Director General, some high-level politicians and international partners. Yet his inability to enforce his writ on all special-order units or to ensure prosecutions, reveals constraints within the HNP and in the judiciary, as well as the limited influence of human rights organizations.

Prison reform
Based on modest targets for prison reform in the 2012-2016 HNP plan, one can observe certain advances in this domain. Officials in the Directorate for Penitentiary Administration (DAP) elaborated an action plan and secured its approval in 2016. New prisons were built with Canadian and US assistance, including the model facility in Croix-des-Bouquets and smaller prisons in towns like Cabaret and Hinche (Servilien 2017). Based on the 2017-2021 HNP plan, new initiatives to upgrade the DAP's autonomy and infrastructure are under way, as are efforts to reduce the number of persons in preventative and prolonged detention (PPD).
Notwithstanding those advances, the system's fundamental problems endure. In March 2018, the UN estimated (UNSG 2018a: 8) that of the 11,859 persons still in detention -including 394 women and 282 minors -75 per cent were still in PPD and awaiting trial. RNDDH's research suggests that this proportion has actually increased in recent years (RNDDH 2018a: 30). This partly explains the current overcrowding rate of 356 per cent based on UN norms, as well as the poor security and sanitary conditions in most prisons, apparently including in some new facilities like the women's prison in Cabaret (UNSG 2018a: 8;RNDDH 2018b).
A survey (n = 894) of conditions in the country's largest prison, the National Penitentiary in Port-au-Prince, adds texture to that picture. Édouard and Dandoy (2016a) confirm that over 4,000 men are detained in a facility built for 700 prisoners; many are regularly victimized and 57 per cent of inmates feel that they are in poor health. The study documents how the majority of these prisoners are young, impoverished and black, and how imprisonment increases prisoners' and their families' poverty. In brief, they conclude that Haiti's prisons ' act as a security system that exacerbates social inequalities' (Édouard and Dandoy 2016a: x, author's translation).
Our research in the South, North and Northeast departments suggests that the tendency of prisons to indefinitely detain the poor and reproduce their marginalization is not unique to the capital Sainsiné, 2015 and. A more recent study concludes that overcrowding in most prisons results in 'inadequate treatment and access to…justice, mental and physical health services, reintegration and professional training activities for detainees' (Nomo Zibi 2017: 7). In addition, 'women and girls are particularly affected by this state of affairs and are moreover vulnerable to discrimination and gender-based violence' (ibid.) The limited record of reform in Haiti's prisons is a microcosm of the challenges facing 2GSSR in the country. Senior officials, who are acutely aware of the system's problems, are trying to foster change and are motivated by pressures from human rights NGOs and the media. Yet they face multiple constraints including their limited autonomy from the HNP and the Ministry of Justice and Public Security (MJSP), a judiciary that seems adverse to the orderly release of persons under prolonged detention without trial, and the larger socioeconomic conditions that perpetuate extreme poverty, inequality and violence (Édouard 2013). Before addressing those constraints more systematically, this section concludes with an examination of the uneven national ownership of police reform processes in Haiti.

National ownership
When the 2012-2016 HNP plan was being drafted, there was a conscious attempt by the UN and other international partners, including La Francophonie, to involve HNP officials in the process. Our interviews with many senior HNP officials between 2014 and 2016 suggested that this worked fairly well with the HNP High Command and some departmental directors, but generated little buy-in at lower levels. Some officials argued that the shallowness of institutional ownership was a factor that constrained police reform Sainsiné 2015 and.
Motivated by that concern and facilitated by the increasing capacity of the HNP Strategic Planning Unit, broader consultations were initiated to inform the 2017-2021 plan. According to the HNP, that involved extensive consultations inside the force to "ensure the legitimacy of the new plan and…a greater appropriation of the plan by senior officials of the Institution" (CSPN 2017: 7, author's translation). It also involved consultations with other governmental, parliamentary, private sector and civil society stakeholders, as well as with international partners, in particular to "ensure political support to finance the new strategic plan" (ibid.). Those consultations complemented outreach to high-level politicians in private as well as in public spaces, such as the HNP anniversaries on June 12 every year.
That outreach persuaded the Moïse-Lafontant government to make public security as well as justice priorities for its mandate and to approve the new HNP plan in July 2017. Despite the growing fiscal crisis and the cutbacks applied in some sectors, the government allocated a ten per cent increase to the HNP in its 2017-2018 budget, almost half of which is capital investment to build or refurbish HNP infrastructure (MEF 2017). Those investments were partly motivated by the need to secure counterpart financing from international partners, who have covered the bulk of infrastructure improvement costs over the past decade. Yet from early 2018 onward, actual disbursements lagged far behind promises, causing severe budget shortfalls in the HNP.
Nonetheless, the HNP's efforts have begun to translate into broader support. Despite its criticisms of HNP performance on specific cases and issues, the respected human rights network RNDDH has consistently advocated better material conditions for the HNP and the judiciary. The Haitian public also seems relatively supportive of the HNP. The most recent public opinion data (collected in 2014) suggested that 58 per cent of Haitians trusted the HNP while only 42 per cent trusted the justice system (Gélineau and Zeichmeister 2016: 110-11). The difference may reflect advances in the HNP's crime reduction and outreach efforts -compared to the relative lack of progress in judicial reform documented in the next section.
Despite professional leadership by senior HNP officials and support from their international partners, HNP development has not gained enough executive, parliamentary, civil society or public support to consolidate a shift towards transformative 2GSSR. The following section considers how HNP reform fits into larger social and political dynamics.

The (Dis)enabling Context of SSR in Haiti
Despite widespread recognition that SSR should be harmonized with rule of law reforms as well as with enabling political and socioeconomic conditions, the Haiti-specific literature on those issues remains quite fragmented. To counter this tendency, this section provides a synthesis of recent studies on contextual dynamics affecting SSR in Haiti, to complement our focus on the HNP and its relations with core stakeholders. The main contextual dynamics affecting HNP reform in Haiti are understood to be: the limited reform of the justice system; the reinstatement of the army; and the wider socioeconomic crisis.

Justice
Based on his field experience and scholarly research, Berg (2013) offered a compelling analysis of limited judicial reform in Haiti. His main thesis was that over two decades of externally funded legal and procedural reform efforts have stalled, because they have not addressed the resistance of the judicial, political and economic elites who feel that they benefit from the system's dysfunctions. His counterpoint was that this path dependency could perhaps be broken through long-term efforts to support state and civil society reformers technically and politically, in the capital and in local contexts when/where space for reform is emerging. Nonetheless, he cautioned that the "evolution of the justice system in Haiti remains complex and uncertain, driven by numerous factors that lie beyond the control of any one actor, especially development agencies and donors from outside the country" (ibid.: 29).
This analysis can be extended to recent developments. The Superior Council of the Judicial Power (CSPJ) was established in 2012, but as Berg predicted, it was unable to fulfill its mandate under President Martelly due to budgetary and political constraints (RNDDH 2018a; UNSG 2018b). During that period, the Judicial Academy (EMA) was revived and graduated two promotions of new judges, yet it took several years for those magistrates to be placed in the system. Such contradictory tendencies led the UN Secretary-General to conclude that the functioning of justice institutions "hardly progressed" in the last period of MINUSTAH's mandate (UNSG 2017b: 4). In early 2017, the Moïse-Lafontant government supported the CSPJ's certification of reputable judges and the decertification of a few judges deemed incompetent or gravely corrupt. Yet the government also collaborated with parliament to appoint judges who had not graduated from the EMA or been vetted by the CSPJ (RNDDH 2018a).
Recurrent labor strife and work stoppages by judges and clerks -due to chronic underfunding and new budget freezes -also led to an historic low in the number of persons tried in 2017-2018 and undermined the implementation of Plan Themis, which aimed to reduce the number of persons in prolonged detention (RNDDH 2018a: 29). In its assessment of the Moïse-Lafontant administration's first year in office, the RNDDH concluded that the entrenched corruption of many judges, political interference by the executive and the legislature as well as tight fiscal policies converged to keep universal access to justice a distant dream for most Haitian citizens (RNDDH 2018a). Still, the UN and certain civil society organizations suggest that the appointment of a new CSPJ in 2018 (albeit with no female members), plans to unblock the adoption of new criminal codes and the appointment of new magistrates, reviving the Office of the Judicial Inspection, as well as initiatives like the establishment of legal assistance offices to reduce prolonged detention in the National Penitentiary, may help unblock some bottlenecks in the system (Collectif Défenseurs Plus 2018; RNDDH 2018b; UNSG 2018b).
Despite wide recognition that judicial reform is also required to consolidate progress towards 2GSSR, in practice there has been little movement in Haiti on this crucial link in the rule of law chain. Compared to the HNP, there is limited high-level agency for reform inside the justice system, although there are reformers scattered throughout the system. As UN officials working on justice have observed, there has also been much less international investment (financial and political) in justice reform over the past decade. Until these internal and external dynamics change, the prospects for transformative 2GSSR will remain limited.
The 'new' army Burt (2016) documents how, despite opposition from the UN, the US and Canada, the Haitian state has asserted its sovereignty by reinstating the army in recent years. In 2012, the Martelly government recreated the Ministry of Defense. In 2013, 150 personnel were recruited to form an engineering corps that received training from Ecuador and Brazil. Despite steady budgetary increases from US$ 2.9 million in 2012-2013 to US$ 6.8 million in 2015-2016, the Martelly government was unable to scale up the engineering corps (ibid.: 6).
The government's drive to reinstate the armed forces has been controversial. Western partners expressed concerns that this could draw scarce fiscal resources away from the consolidation of the police; human rights organizations worried that the army would become a tool to repress political opposition and social protest, much like the old Armed Forces of Haiti (FAd'H). Successive governments have retorted that this army will be apolitical, professional and law-abiding and that it will concentrate on missions like external defense and responding to natural catastrophes, which lie beyond the HNP's mandate or capabilities (Burt 2016).
Recent developments have stoked these controversies. In 2017, the Moïse-Lafontant administration launched a drive to recruit 500 new personnel. It increased the Ministry of Defense budget to US$8.5 million (MEF 2017: 27) and appointed a former FAd'H colonel as interim commander in chief. In 2018, it appointed six former FAd'H officers to the interim high command -including two officers charged with grave human rights violations or other crimes under the dictatorship of General Cedras in the early 1990s. The defense minister also promised to unveil a law to guide the army as well as plans to scale up the force to between 3,000 and 5,000 personnel by the end of the government's mandate. Clearly, that would require budget increases far superior to those the ministry has received to date.
For human rights organizations, those moves confirmed the perceptions that the government intended to recreate the FAd'H as a repressive instrument of the executive and its social allies (AlterPresse 2017;Charles 2018). The UN Secretary-General's observations in late 2017 reflect the respectful yet cautious attitude that had taken hold among official Western partners: I take note of the steps undertaken by the Government of Haiti to re-establish the Armed Forces, just as I have heard the voices of caution and concern raised by Haitian stakeholders…It is important that the process become a unifying national project, dissociated from personal or party politics, in order to meet the challenge of creating an apolitical body, as envisaged in the Constitution…Such an initiative should not harm efforts, or redirect resources otherwise available, to strengthen the national police or to pursue broader development priorities. (UNSG 2017b: 15) In theory, the creation of a new army could advance 2GSSR by relieving the HNP of expensive tasks like patrolling Haiti's coasts by air and by sea. However, the current economic austerity offsets these potential advantages, as the FAd'H competes with the HNP for scarce budgetary resources. Regardless of the economic pressure and the potential reputational risks of reinstating an armed force with a deeply problematic history, successive presidents, their allies in the Senate and former members of the FAd'H -supported by Brazil and Ecuador -have shown significant agency and expended political capital to reinstate the army despite opposition from crucial international stakeholders. Those actors' determination and relative power has enabled them to overcome considerable opposition from human rights CSOs, the UN and Western partners. Given the current political and economic crises in Haiti and partner country Brazil, the continued commitment to developing and expanding the army is uncertain. Recent experience, however, seems to indicate their agency and intention towards the further expansion of the armed forces.

The political and socioeconomic context
As in other domains of social life in Haiti, the economy has registered some advances and many challenges since the devastating earthquake in 2010. After a period of macroeconomic recovery, GDP growth slowed to 1.4 per cent in 2015 (Focus Economics 2018) and has not recovered since (MEF 2017: 3). Economists attribute this to decreased international assistance, the election-related crisis from mid-2015 to early 2017, as well as the destruction caused by Hurricane Matthew in October 2016 (IDB 2017). Global and national factors also converged to generate a 14 per cent inflation rate by 2016, adding fuel to the fire of social protests linked to the electoral crisis (Focus Economics 2018). Deeper problems undergirded those conflicts. They included the persistence of extreme poverty among 24 per cent of the population and of poverty among 59 per cent of the population (World Bank 2018), as well as of dramatic inequalities: in 2014, Haiti had a Gini coefficient of 0.61, one of the highest in the world (IDB 2017).
The macroeconomic slowdown aggravated a stringent fiscal situation, prompting budget cuts beginning in 2016. As the new government noted in its 2017-2018 budget, 'the situation of public finances is extremely worrisome, as revenues stagnated … in 2016-2017 despite the rise of inflation and the depreciation of the gourde …' (MEF 2017: 3). Notwithstanding the temporary stabilization of formal politics after the Moïse-Lafontant government took office in early 2017, inflation and fiscal cutbacks generated intense labor conflict as teachers, industrial workers, court clerks and even judges demanded that the administration pursue measures to catch up with inflation (RNDDH 2018a).
While promising to restore macroeconomic stability and fiscal probity, the Moïse-Lafontant government launched a 'Caravan of Change' of development projects across the country, scaled up the army and initiated an ambitious consultative process that is supposed to lead to a new social pactknown as the 'États Généraux et Sectoriels de la Nation.' The government and its international partners claimed that efforts to stimulate growth in sectors like agriculture, while controlling public expenditures and taking steps to increase public revenues, began to yield fruit in early 2018. They are hoped that the removal of fuel subsidies and the increase of value-added and public service taxes in mid-2018 -as part of an agreement with the IMF -would unlock increased budget support funding from international partners (Focus Economics 2018).
Yet, as pro-government Senator Youri Latortue warned at the opening of the new legislative session in January 2018, frustrations and discontent are more and more audible across the country (AlterPresse 2018a). When the Ministry of Finance announced a 38 to 50 per cent hike in fuel prices on July 6, riots exploded in the capital and other cities, until the government rescinded its decision the next day (AlterPresse 2018b). If the government reinstates those measures to honor its agreement with the IMF, it will probably drive recurrent social protest and political instability. The government may instruct the National Police to control those protests, further pressuring the HNP to deliver security services within a very restrained fiscal context. These factors combine to provide an environment that may not be conducive to 2GSSR in Haiti in the near future.

Conclusions
Shortly after the UN Security Council decided to replace MINUSTAH with the UN Mission for Justice Support in Haiti (MINUJUSTH), a group of diaspora activists issued this assessment: Its [MINUJUSTH's] principal objective is to help the Haitian state develop and professionalize the existing National Police…which will actually translate into more repression of the Haitian people … The power to maintain order…is really the power to defend the status quo, the power to keep intact the dominant order… One cannot pretend to 'reinforce' the rule of law when the state, by its nature and orientation, exists only to defend without compromise the interests of the dominant class and of a certain political class… One cannot institute a 'practice' of justice in state institutions to benefit the majority, when that state is subservient to the interests of foreign powers and of a subcontracting oligarchy. (Rehmonco 2017, author's translation) This article provides support for some of the concerns expressed by Rehmonco. It underscores the risk that a strengthened HNP could become the guarantor of an unjust political and economic order. Yet the use of a critical institutional framework also reveals how complex dynamics inside the HNP, in other parts of the state including the judiciary, in Haitian society and in those national actors' relations with distinct international partners, have opened space for the social construction of limited alternatives without consolidating a shift to transformative 2GSSR.
First, we have seen how Western and UN interest in stability has certainly ensured financial and political support to professionalize the HNP, which remains the core of SSR efforts (Rivard Piché 2017b). Yet there is a tenuous balance between those partners' support for 'hard' goals like increasing HNP ranks and strengthening counter-narcotics units, versus 'soft' goals like enhancing community policing and democratic accountability. Engagement by Latin American regional powers has also enlarged Haitian elites' room for maneuver, though not always favoring 2GSSR, as seen by their support for reinstating the army.
Second, we have seen how different Haitian actors have used that space to advance their own interests. Two successive Haitian governments have used Brazilian and Ecuadorian support to re-establish the FAd'H, despite Western opposition. The HNP High Command and its executive-level supporters have pursued a reform agenda reflecting their own mixed prioritiesnamely strengthening 'hard' and 'soft' capabilities while placing de facto emphasis on the former. The human rights and women's movements have campaigned for a distinct approach to HNP reform stressing community policing, gender equality and democratic accountability, yet they have insufficient power to redress the balance in favor of such goals. CSOs are also pushing for reforms in the justice system, yet they lack the power or support of politicians and magistrates at the apex of the system.
What does that mix of change and continuity imply for the debate on 2GSSR? In sum, it suggests that while it is too early to pronounce the consolidation of 2GSSR in Haiti, it is also too soon to pronounce its death. There are elements of 2GSSR taking root, notably in the HNP. Yet even there, secondgeneration reforms coexist with a tendency to privilege a tough-on-crime (and social protest) approach while treating accountability, community policing and gender equality as public relations exercises. Connecting those contradictory tendencies inside the HNP to the slow pace of reform in the justice system, the reestablishment of the army and especially the inability/unwillingness to reduce socioeconomic inequalities, leads us to conclude that the medium-term prospects for consolidating transformative 2GSSR are still tenuous.
This reading of the Haitian case through a critical institutional lens also adds texture to our cumulative understanding of 2GSSR in different FCAS. It confirms the view of 2GSSR optimists that there is margin for maneuver, for reformers and even for actors pursuing transformative changes, in relatively enabling contexts like Haiti, Sierra Leone and Timor Leste, where there is no war and where liberal democratic politics endure -at least formally. Yet even in those somewhat permissive situations, the mixed motives of civilian rulers as well as of senior officials in security and justice agencies, the relative weakness of human rights and women's movements, as well as the enduring weight of political and socioeconomic factors, greatly constrain the consolidation of transformative 2GSSR.
The fact that 2GSSR remains so difficult in permissive environments buttresses the belief of post-colonial critics and even of 2GSSR promoters like Sedra, that 2GSSR is almost impossible in more restrictive contexts. Those conclusions seem to apply to members of the g7+ grouping of FCAS like Afghanistan, the Democratic Republic of Congo, South Sudan and Yemen. In those societies, variants of authoritarian politics, war and external intervention trump 2GSSR, despite their governments' formal adherence to the New Deal for Peacebuilding and Statebuilding (Donais and McCandless 2016). 2GSSR also seems a chimera in middle-income FCAS like Iraq, Libya and Syria, which are plagued by war and intervention. In those contexts, what seems to exist in practice are variants of the biopolitical control strategies evoked by Duffield (2001).
Much remains to be researched with regard to SSR across the spectrum of situations from relatively permissive contexts like Haiti to much more restrictive contexts like Syria. In Haiti, there is a need for continued research on dynamics within the HNP, as well as between the police and other key actors -from the executive and key members of