Prioritization in human rights NGOs: The role of intra-organizational units of planning

Abstract Drawing on forty in-depth interviews with program managers in nineteen Western international human rights organizations, this article examines how human rights organizations make decisions about how to allocate resources and how to manage their commitments to specific causes, specific people, and specific areas. It argues that organizational routines shape the allocation of resources relatively independently of other factors and it pays particular attention to the role played by intra-organizational “units” of work and planning. Units of work and planning function as candidates for the allocation of resources within organizations. Resources are not allocated directly to issues or causes but rather are distributed on the one hand among a set of range of practices, such as reports and campaigns, and ways of responding, which are considered legitimate, and on the other hand among the thematic and geographical units, which structure human rights organizations. The article concludes by discussing some factors that play a role in the selection among these units. As human rights workers consider where their organizations can make a difference, other organizations and conditions for their work come into view, levers matter, and the way making a difference can be demonstrated plays a role.

, international courts (Hagan 2003), national courts and national political institutions (Wanyeki 2003;Sassen 2006;Nash 2007;Schjolden 2009;Mwambene 2018), universities, and the news media (McPherson 2012). But Western international human rights NGOs have shaped the imagination of what "human rights" are disproportionately; the debates between the "champions" of human rights and their critics in particular have taken these NGOs as a central reference point.
In the context of these debates, I treat questions about the allocation of resources as both an end and a means. Questions about the allocation of resources are important in themselves: There is a legitimate public interest in making more transparent how NGOs allocate resources. Questions about resource allocation are also a specific empirical entry point for developing an understanding of practices and taken-for-granteds in this particular set of organizations, which can then be used to answer other questions about the role these organizations play and do not play.
The article begins by introducing pertinent observations from previous research on human rights organizations. It then discusses some sensitizing concepts from the study's context in the sociology of knowledge and organizations and describes its data and methods. The core of the article analyzes the role of different "units" of work and planning, which are construed by routines within organizations and are available as candidates for the allocation of resources. Resources are distributed on the one hand among a set of acceptable practices and ways of responding, and on the other hand among the thematic and geographical units, which structure human rights organizations. These intra-organizational processes mediate the allocation of resources among different rights and violations. I then discuss some factors that play a role in the selection among these units. As human rights workers consider where their organizations can make a difference, other organizations and conditions for their work come into view, levers matter, and the way making a difference can be demonstrated plays a role.

International human rights NGOs and global civil society
This article builds on two waves of research on international human rights NGOs. Pioneering work on global civil society focused on the history of human rights NGOs as transnational actors, on their stated values, and on the degree to which they did or did not succeed in producing change in line with their stated values (Keck and Sikkink 1998). A second wave of studies of human rights NGOs has expanded on this work in three ways. It has, first, focused more closely on the ways international human rights organizations do their work in practice (e.g., Smith, Pagnucco, and Lopez 1998;Merry 2003;Hopgood 2006;Goodale and Merry 2007;Hopgood 2010;Wong 2012;Pruce 2015;Pruce and Budabin 2016). Scholars have, second, widened the scope of questions about the effects of NGOs. Scholars moved beyond a concern only with NGOs' stated goals and became interested in the range of effects of human rights work, including positive as well as negative, and intended as well as unintended effects (Barnett 2002;Kennedy 2004;Guilhot 2005;Meyers 2019). These studies, third, built on a recognition that human rights are not only "expressed" through NGOs, but are also given meaning by them, in a way that is necessarily specific and also selective (de Waal 2003;Merry and Levitt 2009).
Particularly in making this third point concerning the shaping of human rights by NGOs, previous research has established the foundation also for questions of prioritization. Under the heading of "agenda setting" this research has highlighted some factors that are directly relevant to the questions addressed in this article. The factors highlighted by this work can be discussed under two rubrics: This research has, first, drawn attention to cultural patterns in the ways human rights are applied to the world; it has, second, analyzed the role of a range of external actors who influence NGOs.
Regarding cultural patterns, scholars have observed, for example, that human rights NGOs have historically prioritized civil and political rights over social and economic rights (de Waal 2003;Hertel 2006). This can be linked to an affinity between human rights as a discourse and liberalism as a tradition that prizes individual autonomy and freedom "from" over freedom "to" (e.g., Douzinas 2014). Scholars have also highlighted the way gendered ideologies of vulnerability have shaped the agenda of international NGOs, for example, by making some noncombatants visible as civilians and not others (Carpenter 2005;Hahn and Holzscheiter 2013).
Second, researchers have identified the way different social groups and external stakeholders shape NGOs' agendas. External stakeholders include very prominently donors to human rights NGOs: Clifford Bob (2002; has memorably suggested that Western human rights organizations produce information for a "market in information" and are influenced by demand in this market by donors and political audiences. The influence of donors has also been an important theme in the literature on development (Hulme and Edwards 1997).
Scholarship has also clarified the influence of actors other than donors: Clifford Bob has shown how human rights NGOs are influenced by social groups that try to draw attention to their issues and try to market themselves as victims of human rights abuses (Bob 2005). In a series of studies of campaigns related to labor and economic rights in different regions of the world, Shareen Hertel has shown that partner organizations on the ground also shape which issues are taken up, and in what form (Hertel 2006). Researchers have shown that NGOs are embedded in networks of other NGOs, within which some NGOs are able to exert more influence than others (Carpenter 2007a;2011).
In what follows, I contribute to this conversation based on concerns from the sociology of organizations, which highlights the impact of mundane routines inside organizations. I draw in particular on a sociology of organizations influenced by phenomenology and the sociology of knowledge (Berger and Luckmann 1966). This work focuses on management practices, and pays specific attention to the knowledge involved in constructing a shared world, which then is a precondition for action within that world.

Shared practices in a field of organizations
Previous research on decision making inside organizations has drawn attention to the tacitly held assumptions and taken-for-granted routines that mediate between an organization's explicit goals and the situation in the world on the one hand and what the organization does on the other hand. Sociologists have highlighted how aspects of the practical organization of work in organizations shape organizational output (Cohen, March, and Olson 1972;Molotch and Lester 1975;Fishman 1978;Becker 1982;Vaughan 1996;Desmond 2007;Scheffer 2010;Medvetz 2013;Krause 2014).
Some work in the sociology of organizations has focused on how dynamics internal to one individual organization shape outcomes. Some work in the literature on international NGOs, on the other hand, has usefully drawn attention to what international NGOs share with all other organizations-including for-profit firms-in order to move beyond idealized conceptions of NGOs (Lindenberg and Bryant 2001;Cooley and Ron 2002;Sim eant 2005;Bob 2010;Prakash and Gugerty 2010).
Here I adopt a third route: It is useful not only to ask about what all organizations share but to ask about specific organizations and specific fields of organizations. Field theory encourages hypothesis about mediating practices that are shared across a set of organizations that see each other as relevant peers (Bourdieu 1995;Fligstein and McAdam 2012). Because of considerable uncertainty in terms of knowing and predicting their environment, organizations often look to peer organizations (Fligstein 2001) to develop a response (Meyer and Rowan 1977). This mutual adjusting of organizations who see themselves as peers is what DiMaggio and Powell (1983) called "institutional iso-morphism." This article focuses on what de Waal (2003) has called "secondary human rights organizations," which he distinguishes from primary mobilizations and local activist organizations. Secondary human rights organizations operate in different contexts across the globe, and they self-identify as human rights organizations rather than primarily or only as part of a specific social movement. International human rights organizations of this kind are today to some extent engaged in mutual referencing, as we would expect in a field of organizations-that is, they recognize each other as relevant peers (Fligstein 2001;Dezalay and Garth 2006).
In any field of practice, we expect some specific shared assumptions or doxa about "how things are done," which are usually tacitly accepted. I draw attention to two features shared among human rights organizations in this context: First, human rights organizations have a repertoire of practice. The idea of repertoire conveys that there is choice among acceptable ways of framing and responding to the world but the choice is not unlimited (Swidler 1986;Lamont 1992;Swidler 2001;Silber 2003). Repertoires tend to be specific to a specific social group (Lamont and Th evenot 2000) or in this case a specific set of organizations. Second, international human rights organizations plan work using regional and thematic categories.
I suggest that organizational practices are shaped by the ways in which units of work and planning are construed. Units of work and planning function as candidates for the allocation of resources within organizations. Resource are not allocated directly to issues or causes, but rather resources are distributed on the one hand among a set of range of practices, such as reports and campaigns, and ways of responding, which are considered legitimate, and on the other hand among the thematic and geographical units that structure human rights organizations. The construction of units of work and planning is a precondition for factors and mechanisms, which shape what in the concrete instantiation of the unit is selected.

Data and methods
The article is based on 40 in-depth interviews with staff members who work on the programming side of Western international human rights organizations. Respondents are in charge of either a specific country or a number of countries or work on a specific thematic area. Respondents worked for 19 different organizations, which were sampled to include diversity within the particular group of NGOs I targeted. Respondents worked for the largest and most prominent international human rights NGOs and organizations that introduce variation in term of size, national origin, political orientation, funding source, religion, and thematic focus. A list of organizations is included 1 ; I spoke to one to three people within each organization. Interviews lasted about an hour and were conducted in person whenever possible (n ¼ 31). Research was conducted between 2013 and 2017.
I interviewed program officers because they play a key mediating role between strategic planning in the organizations' senior leadership team and the day-to-day implementation of projects. Program teams are the center of the organizations' outgoing flows. Decisions are prepared here and they are implemented here and the most detailed knowledge of internal structures and events is located here, and not at the highest level in the organizational hierarchy. Program staff members already have a number of years of experience to look back on and hold significant responsibilities but are still involved in the everyday business of "doing human rights work." With my research design and interviewing technique, I stand in a specific tradition of expert interviewing, a tradition of interviewing influenced by the sociology of knowledge (Berger and Luckmann 1966;Bogner, Littig, and Menz 2009;Pfadenhauer 2009;Gl€ aser and Laudel 2004). Respondents are not interviewed primarily in the context of their biography, but in the context of their work and their membership in a specific group of experts. The expert, in this tradition, is not interviewed to give information about a subject area that he or she is knowledgeable about as an observer, but rather his or her practical knowledge of organizational processes, which he or she is involved in, is the target of the investigation.
My questions aim to target the practical knowledge and experience of professionals because it has consequences in the world. I asked my respondents about their everyday work. I asked what a typical day in the office looked like, how priorities are set, about debates they had with colleagues about priorities, and about work they were especially proud of and work that raised questions for them.
In analyzing the interview transcripts, my aim was to identify shared categories and assumptions. Each interview is first analyzed with regard to codes or headings emerging from it. While in the narrative interview the priority is on sequence or the inviolable logic of the single interview, here the focus is also on the search for common themes, unifying codes, and headings.
Results of thematic comparison are tested against the text and the full data set. Commonalities are probed, but deviations, contradictions, and particulars are also identified.
In interpreting interview data, we need to consider the artificiality of the interview situation and the possibility of a bias towards socially desirable responses. I would suggest that the problem of validity presents itself in a different form for interviews targeting practices as opposed to interviews targeting facts, values, or attitudes, which seem to be the targets of most critical discussion of interview research (e.g., Jerolmack and Khan 2014). My research design targets practices through questions that ask for stories about concrete instances of their everyday work life, looking for frames of reference that can be expected to remain intact even in attempts to present a selective and sanitized picture of their work.
In this research design, validity and reliability depend on the assumption that respondents share a social space with some shared assumptions. If the respondents' answers are strongly divergent, if they do not reveal shared assumptions about routines and procedures, or if the terms of the questions do not make sense to respondents, these assumptions would be falsified.
I complement the evidence produced in these interviews with other sources of evidence about what it is that organizations do and do not do, such as campaign publications, annual reports, other academic work, and published reflections by professionals and debates among them. All quotes from interviews are anonymized; if I mention the names of specific organizations I am drawing on publicly available information.
This article looks at practices and routines inside NGOs. The research design directly targets a key site of resource allocation in human rights work, the program management offices of international NGOs; the data do not allow me to quantify the effects of the observations about managers' practices, which I collect. The data do allow me to conceptualize mechanisms that affect the distribution of resources based on close examination of practices in this site.

Repertoires of practice
Human rights organizations share a set of assumptions about a set or repertoire of practices, which are considered legitimate as "human rights work"; resources, such as staff time, are distributed across these practices, not others. Previous work has highlighted the fact that human rights organizations, such as very prominently Human Rights Watch and Amnesty International, accord a central role to the "report" as a product of their activity in the planning of their work and the accounting for their work (Cohen 1996;de Waal 2003;Dudai 2012;Moon 2012). In what follows, I build on this observation about the report but suggest that we should not overlook "the campaign" as an element of the repertoire of practice, as well as the more narrowly legal aspects of human rights work.
As de Waal (2003) has pointed out, research and reporting has been a major activity for human rights organizations. He argues that secondary human rights organizations "use research, documentation and publication combined with the skilled use of the media and lobbying of politicians to make their concerns known. Their basic premise is that if people know about an abuse they will be moved to want to stop it." This is a premise that is not as often reflected as it could be (de Waal 2003: 239). Human rights organizations have developed a particular notion of "research," one that is heavily influenced by legal notions of research. In many organizations, researchers are looking to interview individual victims of human rights violations, and they are looking to establish facts by looking for corroborating evidence from different sources.
The information in human rights reports is heavily fact-checked. Several respondents have had the experience of research not being published because a superior in the organization judged that the information they had collected was not good enough or because there were not enough testimonies from individuals directly affected by a violation of human rights.
Human rights organizations have a repertoire of practices, which goes beyond research in this sense of "information for the public." "Campaigns" and "projects" are also important units of human rights work with potentially different implications for how work is organized: Campaigns emphasize an issue, a strategy, and a goal, rather than necessarily "new" information; projects can consist of a range of activities. Practices, such as writing a press release, sending advocacy letters, conducting research, or releasing extended pressers, are sometimes part of producing one of such units, but can in principle stand on their own.
A singular emphasis on reports or on information as products of human rights work would lead us to underestimate the legal aspect of human rights work, or at least would reduce what is particular about information when it is woven into particular legal and institutional contexts: Human rights organizations are regularly involved in writing briefs for court cases (see Van der Vet 2012; Van den Eynde 2013; Van der Vet 2014) or providing legal advice. Some human rights organizations have whole projects or units devoted to servicing particular legal instruments, such as the UN Human Rights Council's Universal Periodic Review or the Optional Protocol to the Convention against Torture.
Human rights NGOs also send observers to court cases, do monitoring visits, or send someone to accompany someone deemed at risk. "Training" is an important area of activity for human rights organizations. International human rights organizations provide training to human rights defenders and to activists working with local human rights organizations. They also organize workshops for teachers, police officers, and judges.
Though the repertoire has diverse elements, it still acts as a preselector among practices, and much of this preselection work is implicit in expectations about what is and what is not professional human rights work: When we compare international human rights NGOs to primary mobilizations, as de Waal (2003) has done, we can note the restrictions of the repertoire of practice by noting some of what human rights NGOs do not do: As secondary human rights NGOs, these organizations do not tend to provide long-term social services combined with public education in a place; they do not tend to turn out people for rallies, or invest significant staff resources in long-term organizing in particular communities, though they may work closely with organizations which do do that kind of work.

Thematic and regional divisions
Alongside a process that distributes resources among elements of the repertoire of practice, resources are distributed across geographical and thematic areas. Most international human rights organizations have both regional and thematic divisions: Regional programs cover large geographical areas, such as Africa or South East Asia. Within that, or standing on their own, organizations might create units focusing on specific smaller regions, such as the Caucasus or the Horn of Africa, or on individual countries. Human Rights Watch, for example, has regional programs as well as the following focus themes: children, migration, refugee rights, terrorism/counterterrorism environment, disability, United Nations, women's rights, business and human rights, lesbian, gay, bisexual, and transgender (LGBT) rights program, health and human rights, free speech, torture, arms, and international justice. 2 Another organization, the International Commission of Jurists, has a program that covers Europe and the Former Soviet Republics, a program for the Middle East and North Africa, a program for Asia, a program for Africa, and a regional office in Central America. The organization also has a program on the independence of judges, on the United Nations, on business and human rights, on Economics, Social and Cultural rights, on sexual orientation and gender identity, on women's human rights, and on global security and on the rule of law. 3 The themes that structure human rights work have a history, in the sense that they emerge and change over time. They may rise in prominence and also sometimes fall again. They also have a politics, which means they are championed by specific actors and coalitions relatively independently of the law (Carpenter 2007a(Carpenter , 2007bBob 2008;Carpenter 2010Carpenter , 2014. Though the Convenant on Economic, Social and Cultural Rights was passed in 1966, for example, human rights organizations were initially associated with a focus on civil and political rights (de Waal 2003). Amnesty International, for example, started to embrace reporting on economic and social rights in the early 2000s (Goering 2006;Hertel 2017).
The increasing attention to LGBTQ rights provides another example of historical change.
LGBTQ rights are today not only the focus of specific LGBTQ organizations, but have come to be included as an explicit focus by many mainstream international human rights organizations. This is the result of intense pressure and a long struggle by social movement actors (Mertus 2008).
Within themes, organizations put distinctive emphases. One respondent who worked for an organization focused on civil and political rights, for example, explained how they had not historically worked on freedom of religion, and have not engaged in the observation of elections; another respondent explained that she had a focus on LBT (lesbian, bisexual, and transgender) rights in her projects, a focus chosen to "highlight that gay men are often over-prioritised in projects."

Countries and focus countries
Human rights law applies across all signatory states. For some human rights professionals, "universalism" has meant that they aim to monitor the human rights situation in the entire world, and in each and every country. In reality, most organizations today have focus countries. One respondent from an American NGO, for example, explained a shift in her organization as follows: "It was felt that the list of countries or list of issues in the old days, five or more years ago only, was firstly too long but it was spread too thinly and also that it would chop and change too often, so just be too reactive. And so the idea is that we pick a few countries and really we stick with them for four years and then have a review. So I've been doing Bahrain and Egypt. Two and a half years ago I started with those. Indonesia too to some extent, and I got some extra staff so I may add another country or possibly even two next year." When countries are units of intervention, they can become units of strategic reflection, that is organizations, like the one just referenced, can think about how many and which it might be best to include in their programming. Organizations can try to create focus by limiting the number of countries, such as described in the preceding, or they can try to expand. One respondent, for example, spoke of her ambition to take a program focused on Russia as a basis for a larger program also involving Armenia, Belarus, the Ukraine and Kyrgyzstan: "So it's regionalizing it, and upscaling it, so that we're now dealing with many more stakeholders, many more activists, many more countries." One program officer has described to me how the formal focus on specific countries affects the way she "distributes" human rights: She 4 said, "If there is a lawyer arrested in Azerbaijan, because we have the project on Russia we will raise that issue, and probably the Azerbaijan lawyer won't be … we won't do anything about that Azerbaijan lawyer, even though for us from the objective point of view it doesn't matter. But then that's why I am saying there is a subjective factor with that, we have to … we are limited sometimes by the projects that we are implementing." The fact that human rights organizations have themes and focus countries means that it matters, for individual victims of human rights violation, which aggregate their violation is a part of. This is true in other contexts, where suffering meets response. Individual illness, for example, becomes part of an "emergency" that qualifies for humanitarian relief based on aggregate counts of cases per population and over time (Krause 2014). There is inequality in provision across and within national states, but this kind of inequality is a little more paradoxical for international efforts, which came into being to respond to the failings of national states and focus on individuals and their dignity.

Selecting among units of planning
Units of work constructed inside organizations shape the ground on which decisions are made. How do organizations decide among units of work? I next discuss a number of factors that come into view as program managers are charged with "making an impact." It should be noted that, philosophically speaking, human rights thinking has often been in opposition to utilitarianism (Hart 1979), and the general demand to demonstrate impact and show results has faced much more opposition in human rights work than in development aid and humanitarian relief (Desormeau and Ignatieff 2005;International Council for Human Rights Policy 2012;Schlangen 2014). There is a long tradition of "trying to do the impossible" or "simply doing what is right." Rhonda Schlangen quotes a human rights advocate she interviewed with a sentiment that expresses an important aspect of the anti-utilitarian culture of human rights NGOs. The advocate said: "I spent eight years defending political prisoners. There was no hope of their release. I lost every case. What was my [observable] impact? Zero. Should I have done it? I haven't found one person who says no" (Schlangen 2014: 6). Some of my respondents echoed a similar sentiment: "We would not say that, 'Okay, let's work in Georgia because that's where we can see progress instead of Uzbekhistan.' No, that is not how we are thinking. That's development thinking." But managers in human rights organizations do talk about impact all the time when discussing their own projects and programs. The respondent quoted earlier as contrasting "development thinking" with practices in human rights work, also said that she had "successfully argued to the management team that we should not focus on fundraising for these two regions for the coming three years because the impact we can have comparing to the work we would have to put into these regions is rather small." Speaking of staff as a scarce resource, another country representative told me, for example, "figuring out how to allocate that is a challenge, in a way that's going to have as much impact as possible."

Conditions of work and vulnerability of the target
Organizations do sometimes consider conditions on the ground that would make engagement meaningful or make success more likely. Countries differ in the ways in which they are hospitable to the repertoire of practice of human rights work. Ethiopia, for example, is very hard to access for foreign human rights workers. For that reason, it is difficult to produce reports in or about Ethiopia that hold up to the specific standards, which NGOs have developed based on their experience in other contexts.
The Association for the Prevention of Torture's Guidelines on Country Engagement, for example, state that an engagement "should be capable of having an impact in the national context. In particular there should exist the possibility to promote and pursue the three integrated elements for effective prevention of torture (transparency, legal framework, will and capacity to change)." 5 In the pursuit of impact, managers consider whether or not the responsible party or the "target of advocacy" is vulnerable to advocacy. In this logic, it was explained to me, it is difficult to make the case for a report on Ethiopia or Iran because this work would likely have no impact, given the intransigence of these governments.
One respondent explained how this discussion of vulnerable targets affects how his organization distributes attention among actors within Afghanistan. He explained that in Afghanistan there are three main actors engaged in human rights abuses: the Afghan government, the international community, and the Taliban. "So who are the worst of those three groups?" he asked. "Well, probably, arguably the Taliban. The Afghan government and the international troops and international embassies are not targeting large groups of civilians and intentionally blowing them up, but does the Taliban care at all what [my organization] says? We certainly haven't seen the evidence that they do." This focus on likely success is in tension with proportionality of attention to all cases: "This is kind of an ongoing argument I have with my boss because honestly I think we should be focusing more attention on the Taliban, I think that we have to keep talking about, out of a sense of proportionality, out of a sense that if you're committing human rights abuses we care and we're paying attention, even if, whether you're in all this, I mean it begins to feel a little bit skewed sometimes because, so the Taliban doesn't seem to care at all what we think. The Afghan government cares a little bit and the international community, particularly the US government is quite susceptible to embarrassment by a group like [my organization] because they take us seriously and they don't like to be spoken of badly by us. But we have to make sure that that doesn't mean that we spend all of our time shouting at the US government and none of our time talking about the Afghan government or the Taliban so it's very difficult, but should we spend a third a third a third of our time no probably not, so it's sort of a constant struggle I think to figure out how to balance that right. Yes, and I definitely try and advocate for a more frequent focus on the Taliban and I don't win that battle."

Levers of change and measurable results
Vulnerable perpetrators of human rights violation are a "lever" of some kind: something that can be used to have an impact. Levers attract resources toward them. Western governments and the law are also considered a lever within international human rights organizations, a tool that, properly influenced and used, can effect change in the whole world.
Several human rights organizations based in the United Kingdom have officers focused on the UK government. One program officer I spoke to had a focus on the UK Foreign Office as part of her job description. The US-American organization Human Rights First, for example, is a (universalist) human rights organization geared around a single leverage. It aims to "challenge America to live up to its ideals." This means that its unique strategy is to think about in which areas, on which issues, and through which means pressure on the US government can make the most difference for improving human rights outcomes across the world.
The institutions and instruments of human rights law play a role as one lever among others. Like the US government, the instruments of international human rights law attract attention as levers, as possible reinforcers, and as routes to success. Legal instruments command resources, through defining roles, projects, and programs that serve them. The OHCHR's Periodic Review Process relies on submissions by NGOs, including international NGOs on particular countries at particular times. The Association for the Prevention of Torture, for example, has a program servicing the Optional Protocol to the UN Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (OPCAT), which it also helped lobby for.
Legal processes also contain a promise to produce results, which are regarded as measurable results. The law, both national and international, emerges as a privileged result. The focus on results, defined in distinction to activities, has a long history in discussions of management across organizational sectors, dating back, for example, to progressive critiques of corrupt government, critiques of bureaucracies in the 1960s and the new public management of the 1980s and 1990s (see, e.g., Krause 2014). These discussions are also reaching human rights NGOs, particularly as they seek to access funds administered through governmental development agencies.
This pressure is brought to human rights work partly from donors, but it is not the same as donors' geopolitical self-interest; it is pushed by internal reformers as well (Desormeau and Ignatieff 2005;ICHRP 2012;Schlangen 2014). An important way to show results in human rights work is to demonstrate legal change in one way or another. Legal change could be a policy change in a specific state, a new international instrument, or the text of a constitution.
One legal officer explained how after the uprising during the Arab Spring, donors and NGOs were very keen to intervene and influence the new constitutions: "Especially concerning Economic, Social and Cultural Rights we are very keen on having more and more constitutions recognizing or guaranteeing explicitly a whole catalogue of rights and bill of rights in line with what we have now in international law. So you think all the new constitutions have … a whole bunch of economic, social and cultural rights compared to older constitutions. Each time we have a constitutional reform process … everybody just jumps on the thing and that was quite impressive in Tunisia." This respondent felt the attention on Tunisia was disproportionate.
Another respondent mentioned a pressure to focus on strategic litigation. "Some donors have discovered very much for themselves strategic litigation, so it means that they are very keen on seeing their money being used … [in this way] … because I guess it's very concrete. You go and you have a case and this is a straightforward case and you have a decision, hopefully a good one and a very systematic and structural one … there is this transformative effect of this kind of litigation … I think it's … quite a neat indicator if you get the decision." The legal focus of Western human rights NGOs has long been criticized in the debate on human rights (de Waal 2003;Shivji 2007;Moyn 2012;Hopgood 2013). The pursuit of both constitutional guarantees and strategic litigation can be in tension with what might otherwise be meaningful ways of organizing human rights work. This tension is highlighted in the account of the legal officer quoted in the preceding paragraph regarding the Tunisian revolution: Western NGOs, in accordance with their programming routines, had planned an orderly consultation meeting among "civil society representatives." Activists on the ground invited people via Facebook and the meeting was rather more tumultuous than expected.
Litigation involves building a relationship with possible clients, which can take many years and involves personal investments on all sides. This is in tension with the time frames in which donors usually expect NGOs to develop strategic litigation. While it is relatively easy to find funding for work that produces legal change, it is harder to find support for capacity building and training, to fund general operations, and to fund a capacity to give advice to people and react to current affairs.

Conclusions
This article has asked how international human rights organizations choose between the different things they could be doing, different things that could all be reasonably justified by reference to human rights and international human rights law. I have argued that underlying explicit discussions about priorities and strategies, the way in which "units" of work and planning are construed shape the distribution of resources. Resources are distributed on the one hand among a set of acceptable practices and ways of responding, and on the other hand among the thematic and geographical units, which structure human rights organizations.
The repertoire of practice includes very prominently reports but also campaigns, trainings, and legal briefs. This repertoire distinguishes international human rights NGOs from other sets of organizations, be it national armies, supermarkets, or humanitarian NGOs. It limits the way NGOs respond to the world, providing a measure of autonomy, for better or worse, vis-a-vis both donor demands and pressure from social movements.
The way international human rights organizations distribute resources across themes and countries has important implications for the way resources are distributed across rights, right bearers, and violations. Units of work mediate between organizations and resources on the one hand, and rights and violations on the other hand. Though human rights are accorded to each and every individual, any individual's rights or violation thereof becomes actionable for human rights NGOs as part of units tied to thematic or geographical aggregates. These units can become an object of organizational strategy, and rise and fall relatively independently of changes in the outside world.
The article has discussed some of the factors that shape how human rights NGOs select among the practices, countries, and themes that function as candidates for the distribution of resources: As human rights workers consider where their organizations can make a difference, conditions of work and the vulnerability of the target come into view, levers matter, and the way making a difference can be demonstrated plays a role.
My analysis is not a basis for commenting on the effectiveness of specific interventions; indeed, examining the sum of practices employed by the field of human rights organizations as a whole should make us skeptical of efforts to maximize effectiveness on the level of a single project or intervention, as they can lead to collectively problematic outcomes. By considering the experience of practitioners across organizations, we can gain some insights into the collective response, how it is distributed, what is hard to get done (but might be meaningful), and what the trends are (which might not be as meaningful).
I focused here on Western international NGOs; further research could explore the fuller range of practices that the much broader set of organizations that are or could be called human rights organizations employ, including social movement actors in all regions of the world.
The research design of this project targets a key site of resource allocation in human rights, the program management offices of international NGOs; this allows me to conceptualize mechanisms that affect the distribution of resources based on close examination of practices in this site. The data discussed here do not allow me to observe the overall effects of the mechanisms thus identified, however. This research could thus be usefully complemented with data about the patterns of distribution that result from the practices we examine. Building on this research, scholars can collect and examine data on patterns of distribution across countries, within countries, across different types of activities, and across different types of themes.