African Libraries in Development: Perceptions and Possibilities

Abstract Across Africa, libraries are contributing to development in their countries. However, this work remains largely unrecognized by international stakeholders in development. To bridge this gap, this research examines the current perceptions of libraries among development organizations and asks how libraries may gain visibility as partners. Interviews with development practitioners indicate that perceptions of libraries remain low and limited but allow for possible roles for libraries as community-embedded institutions and development resource hubs. To engage these roles, the collection of output data is important in order to demonstrate the capacity of libraries to participate meaningfully in development work.


Introduction
A problem of perception Across Africa, libraries are contributing to development in their countries in various ways. In Ghana for example, evidence shows that libraries are supporting UN Sustainable Development Goal (SDG) 3 on good health and well-being by engaging community outreach to enable healthy living (Dadzie et al., 2016) as well as SDG 4 on quality education by scaffolding skills for life-long learning (Atta-Obeng & Dadzie, 2020). The IFLA SDG Stories map (IFLA, 2020) offers additional examples from Kenya, Nigeria, South Africa, and Tunisia to showcase the work of libraries in development. From increasing employment opportunities for people incarcerated in Namibia (SDG 10 on reduced inequalities; Isaacks, 2020) to improving fish farming (SDGs 1and 2 on eliminating poverty and hunger; Imoni-Atebafia, 2020) and organizing recycling programs in Nigeria (SDG 12 on responsible consumption and production; Ogunjimi, 2020), it is clear that libraries' contributions cut across almost all SDGs as well as African Union (AU) Aspirations, demonstrating the capacity of libraries in development.
Despite efforts to amplify these successes, it is unclear whether this work is recognized by international stakeholders in development. Research carried out in 2012 (Fellows et al., 2012) indicated at the time that development organizations did not perceive libraries as partners, and that this was preventing them from gaining wider support. This becomes a larger issue as African libraries struggle to maintain support nationally in the form of funding and inclusion in national development plans. Library and information science (LIS) scholars and practitioners have called on libraries to raise their visibility through the use of data as a tool for self-advocacy (see Bundy, 2003;Sawaya et al., 2011). While libraries across the continent have affirmed the importance of this visibility, few studies have followed up to determine whether efforts have been successful at achieving it. Given the challenges that African libraries face in collecting and using data effectively (Lynch et al. 2020a;Moahi, 2019), the sector needs a more focused strategy that can speak to existing perceptions and understand which types of data would be most effective in shifting the narrative. This research thus illuminates the perception gap that exists between African libraries and development organizations and offers suggestions for how libraries can use data strategically for self-advocacy.

African libraries in development
Development can be defined as "improvement … of the social and material conditions for life" (Kingsbury, as cited in Appleby, 2010, p. 22), but it is a broad and contested term which also "defies definition" (Lawson, 2007, p. ix). With origins in colonial era Britain (Kanbur, 2006), international aid for development has been largely driven by the interests of the Global North and dispersed according to shifting philosophies over time (Lawson, 2007). The field of international development organizations has been similarly described as variable, a "fragmented … ecosystem and an ever-expanding cast of players" (Ingram & Lord, 2019, p. 5) with priorities that change to align with the latest socio-political trends. While some agendas look to partnerships with local institutions in order to increase local ownership of development (e.g. UN, 2015), critiques reveal that these relationships remain opaque, and the desired qualities of an ideal "development partner" shift over time in accordance with political agendas (Vavrus & Seghers, 2010). For libraries hoping to gain a foothold as recognized players in the field of development, this unstable ground is daunting to say the least.
Nonetheless, the work that African libraries do often matches the goals of development organizations expressed in agendas such as the United Nations SDGs and African Union Agenda 2063. From educational programming (Mahwasane, 2017) to community health (Emojorho & Ukpebor, 2013;Sow & Vinekar, 2012) to expanded access to information and communication technologies (ICTs; Ifijeh et al., 2016;Sears & Crandall, 2010), libraries and development organizations share common goals that make them natural partners in development.
However, African libraries have not historically been readily perceived as partners by the international development community (Fellows et al., 2012;Namhila & Niskala, 2013). As an example from Namibia, libraries were initially left out entirely from a World Bank 15-year strategic plan for education, in spite of the natural connection between schools and libraries (Namhila & Niskala, 2013). Advocates were able to mobilize quickly and lobby successfully for libraries to be added later, but specific plans for library development are also notably absent from the country's Vision 2030 development plan, despite an emphasis on "equal access to knowledge," to "transform Namibia … to a knowledge-based economy" (Office of the President, Windhoek, 2004). Though the situation of libraries varies widely across different African countries, they tend not to play central roles within long term national development plans (Benson et al., 2016;Kavalya, 2007) and this is both a symptom and cause of the lack of awareness of libraries in development.
Unfortunately, this tension between the potential of libraries and their lack of visibility is not new. African libraries have historically fought an uphill battle, wedged between external stakeholders and the communities they serve but not fully embraced by either. Many LIS scholars root these issues in the colonial history of African libraries; despite rich traditions of oral librarianship across the continent (Amadi, 1981a(Amadi, , 1981b, the libraries established by colonial administrations emphasized text-based practices and were built to replicate a European ideal, and this model gradually became accepted as the standard for modern African libraries (Mostert, 2001;Nyana, 2009). However, these libraries were not built to be relevant or accessible to the general population, and the negative effects continue until today (Albright & Kawooya, 2007;Kabamba, 2008;Moahi, 2019;Nyana, 2009;Sturges & Neill, 1998). While there is truth to this critique, the narrative of libraries as inherently colonial pigeonholes them and limits understandings of their role in supporting community-driven development. The perception of libraries as irrelevant and detached from the needs of their communities obscures their potential to act as intermediaries between communities and development organizations (e.g. Okojie & Okiy, 2019), to ensure that development is grounded in community needs. We need a better understanding of these recent efforts by libraries to re-align their programs to support their local communities in these ways, in spite of their history as colonial institutions.
Local perceptions, however, remain limited with real consequences for public libraries. Elbert et al. (2012) found that local government stakeholders in Ethiopia, Ghana, Kenya, Tanzania, Uganda and Zimbabwe perceived libraries as "lenders of books" and restricted in terms of technology, not necessarily as "agents for development and innovation." Baada et al. (2020) similarly note a "lack of recognition of the importance of public libraries" as a significant barrier in attaining adequate resources for public libraries in Ghana, as politicians may pay lip service to libraries while failing to allocate proper funds. A lack of a defined role in national development plans exacerbates this issue by leading many libraries to be reliant on inconsistent and short-term funding (Benson et al., 2016;Kavalya, 2007;Moahi, 2019), and decentralization policies which shift power to local governments also result in decreased administrative and financial support (Stranger-Johannessen et al., 2015, p. 88).
Such challenges understandably lead African libraries to look internationally for other opportunities. After all, the region of Africa received almost $52 billion US in official development assistance in 2017, according to the OECD (OECD, 2019). These numbers do not include development funding from so-called "emerging donors" such as non-OECD countries, estimated between $11 billion US and $41.7 billion US per year, almost ten years ago (Waltz & Ramachandran, 2011). Private funding, though notoriously difficult to track, is acknowledged to "dwarf" traditional forms of development funding (Ingram & Lord, 2019, p. 9); for perspective, Lawrence et al. (2015) calculate that U.S.-based foundations gave $1.5 billion US to Africa in 2012 alone. This represents a literal wealth of potential for African libraries to connect their work with the larger goals of development-oriented organizations and agendas in order to access wider support.
Historically, the connection between African libraries and development organizations was built on the hope that libraries could assist in the mass education of the general population in the postindependence era (Moahi, 2019;Mostert, 2001). This led some development organizations to initially give direct support to African libraries, notably UNESCO (Sturges & Neill, 1998). Mostert (2001) notes that within this relationship, solidified by the 1953 UNESCO Public Library Manifesto, libraries were "to provide information to development agents and agencies, to support formal and informal rural education programs through the provision of materials to both students and teachers, and to serve as centers of community education and cultural activities" (np), effectively positioning libraries as implementers of development. In the 1970s, philosophies shifted to emphasize the role of "information for development" to enable developing countries to "maximize the utilization of their local resources" in order to compete in the world market (Bouri, 1994, p. 160). Instead of increasing support for libraries, however, resources were directed toward new science and technology information systems within governments and elite spaces rather than libraries for the public, reducing library support (Bouri, 1994). The 1980s brought economic downturn which further limited the resources allocated to libraries (Maohi, 2019) in accordance with structural adjustment policies and neoliberal attitudes in development which de-emphasized funding for public institutions such as libraries (Kanbur, 2006, p. 7). As a result, the relationship between international development organizations and African libraries became more tenuous as support for public services waned.
More recently, there has been a resurgence of interest in the role of African libraries in development, driven by the belief that development can be empowered by access to information. Related studies position libraries as contributing to development by virtue of acting as public access sites (e.g. Kavalya, 2007;Onoyeyan & Adesina, 2014) which provide ICTs (e.g. Akintunde, 2004;Chisenga, 2000) and physical space for community-based programming which furthers development goals (e.g. Ashraf, 2018). Supported by organizations such as the International Federation of Library Associations and Institutions (IFLA), Electronic Information for Libraries (EIFL), and African Library and Information Associations and Institutions (AfLIA), international and regional advocates are attempting to amplify the successes of African libraries and increase roles for libraries in development worldwide (IFLA, 2015;AfLIA, 2019).
To advocate for libraries, Bundy (2003) and others (e.g. Sawaya et al., 2011) have emphasized that data is key to libraries being able to gain visibility and prove their community value in a world of neoliberal accountability. However, German and LeMire (2018) among others maintain that assessments of such value remain haphazard and challenging to implement, particularly in African contexts (Kabamba, 2008;Moahi, 2019). There is also a lack of consensus on which methods and types of data best demonstrate the value of libraries, with a general trend toward recommending surveys of users (German & LeMire, 2018;Kabamba, 2008). Many in the LIS field (e.g. Kabamba, 2008;IFLA, 2020) highlight the importance of narratives or stories to illustrate development successes, but this is accompanied by the awareness that anecdotes are not enough in the current neoliberal "climate of accountability and reduced public spending" (Hart, 2004, p. 114). Sawaya et al. (2011) recommend going further to combine "evidence from primary sources, including surveys and case studies, and secondary sources, such as government statistics and reports" (450) to comprehensively demonstrate how libraries contribute to funder goals. Young et al. (2020) have pointed out that official sources may lack necessary data, however, and so recommend novel data collection methods such as crowdsourcing to counter this challenge and enable libraries in Africa to fill gaps in official data sets. Overall, Moahi (2019) asserts that using any data effectively requires both qualitative and quantitative data, and crucially, analytic skills in order to weave together a convincing story about the value of African libraries in development.
However, evidence shows that African libraries face unique challenges in collecting and utilizing data for self-advocacy. A recent study indicated that library representatives from 16 African countries note challenges concerning data integrity in terms of completeness, accuracy, credibility, and relevancy; infrastructure; capacity; local investment in libraries; time; and participation of data collectors and respondents (Lynch et al., 2020a). Simply conducting a survey of library users requires more time and effort in order to build a relationship of trust with communities who have long been "socially excluded" from library spaces (Kabamba, 2008, p. 156). There is also concern that even if librarians have data, they themselves are not trained to "interpret data and integrate it with stories of impact" (Moahi, 2019, p. 242). Kabamba (2008) states that "many public libraries do not have the capacity for conducting systematic needs assessment surveys," but "it is understandable that not all librarians will possess these skills." (156) Instead, Kabamba and others (e.g. Igbinovia, 2016;Igbinovia & Osuchukwu, 2018) suggest that libraries solicit the skills and means of library schools, library associations, and development organizations to overcome challenges of capacity. This solution, however, still requires that groups such as development organizations want to work with African libraries as partners. Given the existing issues in library capacity, addressing this issue requires a more focused and strategic effort.
A 2012 study by Fellows et al. showed that the perception of global libraries among international development organizations was low, and public libraries "are typically overlooked as partners in development" (1). In interviews, many development leaders saw the "potential" of libraries but were wary of the perception of libraries as "repositories of books" that would challenge their efficacy as partners. This prompted us to ask if this is still the case and what insight more current perceptions may yield with regards to the use of data. We also sought to focus on the African context, given the unique challenges and possibilities for African libraries in development. To that end, this research asks the following questions to further investigate the issues of perception regarding African libraries in development: What are current perceptions among development organizations of African libraries as potential partners in development? How can libraries use data to effectively position themselves as potential partners?

Materials and methods
To answer these questions the Technology and Social Change Group (TASCHA), a US-based research institution, and African Library and Information Associations and Institutions (AfLIA), an African library organization, undertook research as part of a larger, collaborative project called Advancing Library Visibility in Africa (ALVA). This project broadly examines the relationship between public libraries and development across sub-Saharan Africa.
The research team conducted interviews to understand how international development organizations perceive public libraries in Africa and how libraries might position themselves to take advantage of potential partnerships. We invited adults (over age 18) who are currently working for a development organization that funds or implements projects in Africa to participate. For the purposes of this study, definition of "development organization" was based on the following criteria: Organizational desire to create change aligned with development goals, such as those outlined by the UN (Sustainable Development Goals) or the African Union (Agenda 2063) Provision of funding or services to realize these goals, including research and policy advocacy efforts Focus on long-term change rather than shortterm assistance (e.g. not humanitarian aid) Our definition of development organization is purposefully open to account for the malleable definitions and debates around the term "development" as well as the broad range of levels of involvement with this work. For example, some organizations may advocate at a policy level for the realization of development goals while others may provide direct service through material support on the ground. We wanted to acknowledge this reality of the development landscape and be able to elicit the perspectives of a broad range of actors involved with this work.
Based on this definition, we took a purposeful sampling approach and identified potential participants through (1) the preexisting contacts of the research team, based on their prior research experiences; (2) a survey of the websites of wellknown development organizations; and (3) the recommendations of existing participants in the study via respondent-driven or snowball sampling. We invited potential participants via email, and the overall response rate was rather low; we contacted 178 people total, 50 people (28%) responded overall, and we completed 21 interviews (12% of total contacted). Many potential participants did not respond or replied that they did not know anything about libraries and so were reluctant to participate. We encouraged participants with all levels of library knowledge and experience to participate, but we believe this issue speaks to the overall lack of awareness of the potential role of libraries in development, as will be detailed in our results section.
We interviewed 21 participants who work on behalf of bilateral, multilateral, and non-governmental organizations including private foundations and faith-based organizations. These individuals were based in various countries in North America, Europe, and sub-Saharan Africa, and they reported having worked in development between seven and 45 years. Therefore, though the sample size of 21 individuals is small, there is variety in terms of organization type, geographical location, and professional history. While all are employed at development organizations, we classified five participants as "library advocates" because their work at the organization involved working directly with or on behalf of libraries. We included them as participants because of their unique insights into interfacing between libraries and development organizations. The other 16 participants do not regularly work with or on behalf of libraries.
In terms of limitations, it is important to note that this study is not representative of all development agencies and countries and also had a limited number of responses given the vastness of the development field. Nevertheless, our sample includes participants from a organizations that vary in mission, size, and geographic base, and strong themes consistently emerged out of the interviews. This leads us to believe that this research provides a useful starting point for understanding the relationship between development organizations and libraries.
Interviews were semi-structured (see Appendix for guiding questions) and lasted approximately 60 minutes. Interviews were conducted in English via online video conferencing platforms and were audio recorded with participants' permission. Participants also gave informed consent for their words to be used in this research. Interview questions focused on the participant's professional background, their organization and how it selects local partners, their perceptions of libraries, and which types of data would be effective in shifting perceptions. Interviews were transcribed and anonymized to protect participant privacy, and in this paper, we have used numbers (e.g. Participant 1) as pseudonyms in order to attribute direct quotes.
Transcripts were then coded by two researchers simultaneously using the Dedoose software in order to cross-reference and triangulate analysis for consistency (Baxter & Eyles, 1997). Codes were developed inductively using a grounded theory approach (Glaser & Strauss, 1967) to identify common themes in interview responses. Given the research questions, we paid special attention to trends around current perceptions of libraries, data that development organizations use to assess potential partners, data that organizations want from libraries, opportunities for libraries to gain visibility as potential partners, and other factors that influence organizations' choice of partners. These trends were then analyzed to reveal insights, as discussed in detail below.

Results and discussion
The interview results and analysis revealed important insights for libraries interested in engaging with the development sector more formally. The first and second sections below detail current perceptions of development organizations. We find that these organizations do not readily think of libraries as partners but can imagine potential partnership roles when libraries are discussed. The final section then describes which types of data development organizations would need to be convinced that libraries could fill these partnership roles.

Current perceptions
Interview responses suggest that some development practitioners perceive libraries, but there is an overall limited perception of libraries and their role in development. Ten participants reported low or no professional experience with libraries, and nine participants reported low or no usage of libraries in their lives outside of work. In general, comments displayed a lack of awareness of libraries as potential partners in development.
Participant 14 exemplified this sentiment: [ … ] I don't even know if we're aware and I'm not sure we have any situational awareness around libraries in any place that would work, that's my hunch. I suspect, I wouldn't speak directly for some of our program managers, but I'm pretty sure most would say they had no idea.
Other comments echoed this lack of perception, stating that a "[library] really is a new stakeholder I had not thought of in the context of Sub-Saharan Africa" (Participant 20) and that they had "never heard them mentioned in any of the programs." (Participant 5). This lack of perception includes both an absence of any perception, as exemplified above, as well as the perception that libraries simply do not exist in the African countries where their organization works, as explained by Participant 13: I mean to be frank, I am not sure there are any libraries in [African city]. I don't believe that there are libraries in [city], because [country], I mean as you know, they're really, the government and infrastructure in [country] leaves very, very much to be desired and there are really very few public services provided to the citizens of [country], and that includes electricity and clean water and sanitation, just the bare minimum [ … ] and I am saying this not knowing if there is a library here, but I'm almost certain that there's not, but a library would be so far beyond that basic service that is already not provided that I don't think that it exists here. This general lack of perception of libraries extends to a lack of awareness of their role in development, with seven participants saying they had no perception of how libraries contributed to development. As an example, Participant 1 reflected, "It seems like there's got to be a connection, but I guess it's just something I never really thought much about, honestly until the email came [to request an interview], and I was thinking, wow, this is an interesting topic." Likewise, Participant 14 commented, "I think the biggest thing is [ … ] it would strike me as unusual, but not bad, but unusual for a library to take on that kind of operational role." In addition, many participants expressed that partnership decisions are highly dependent on the project, and, as for Participant 12, choosing a library partner may not "make sense," even if they are familiar with libraries as players in development. These comments illustrate that the role of libraries remains constrained in the eyes of development practitioners.
Those that did perceive libraries offered a limited perception, with a notable trend in associating libraries with books and literacy only, as mentioned by 13 participants. As a library advocate articulated, "Libraries are seen as areas, just spaces for academic -for reading and references. They're not seen as spaces for interaction, you know as open and flexible spaces. That perception I think is one of the weaknesses." (Participant 9) While for some participants, an association with literacy provided a natural connection to development goals and an "easy win" for libraries (Participant 15), for most participants, the notion of a library remained tied to physical books because "you walk into the library, you see books," which can be perceived as "antiquated" (Participant 10). As Participant 3 noted, however, the reliance on physical materials may be necessary in African libraries which experience electricity shortages and so this issue "needs to be looked at in context." Also, multiple participants recognized this as a "perception barrier to get over [ … ] this somewhat caricature image that many of us would have of libraries. They're just a place of books that you go to get books and return books, that's so distant from doing 'development work' and it's not even on your mental radar" (Participant 14). These responses resonate with previous research that libraries are overlooked as partners in development, and individuals who do perceive them tend to have limited perceptions of their work (Elbert et al., 2012;Fellows et al., 2012).
Overall, there is a persistent belief that libraries simply do not have capacity to be active partners in development. The most frequently mentioned perceptions of libraries among participants were related to issues of low capacity, citing outdated facilities, untrained staff, and lack of technology or technical expertise and other resources, among other potential challenges. Participant 14 summarized many of these concerns, explaining, I'll just take a real life example, if we went to our team that runs our out of school kids program in [region of Africa], and sort of asked the thought question or why are we not working or have we explored or should we explore working the libraries instead of just NGO implementing partners, and so there are people who would be like 'why libraries, they're not experts in the issues we care about, they don't do kind of on the ground delivery and implementation of developing programs, or they don't have the expertise we're after, it's a bunch of librarians, you know, they don't know how the development world works.' These concerns demonstrate that the perception of libraries as lacking in capacity limits the extent to which they can be seen as effective partners in development. Though some practitioners mentioned capacity building as an explicit or implicit part of their organization's mission in partnerships, more expressed a preference for partners "who can do the work" (Participant 17) and have a "track record" (Participants 14, 15) to prove it. When asked what qualities their organizations look for in potential partners, participants most frequently mentioned expertise in terms of development sector, local context, and technical aspects, as well as being generally capable and functional. These insights reveal that demonstrating the competency and existing success of libraries in development is crucial in presenting libraries as effective partners, and this will be discussed later.
In spite of a lack of visibility, however, many development practitioners saw the potential of libraries. 12 participants perceived libraries as existent but just not visible as partners; as Participant 17 mused, "You know, we don't really talk about it. I think that a lot of my colleagues, all of us are running around with multiple degrees, and I think we all probably grew up with a soft place in our hearts for libraries, but it's just not part of the conversation." Six participants also recognized the challenge of marketing libraries to development organizations, acknowledging that the issue is one of perception, as expressed by Participant 18: … this is probably more of a weakness on my side as opposed to the library side [ … ] I have a little bit of a stereotype of what libraries bring to the equation, in terms of partnership, and so I guess, my gut is that there is a pretty kind of narrow and specific role for libraries and so that's probably not the case.
Comments such as these and others demonstrate that it is possible to shift perceptions of libraries. At least six participants reported a change in views over time to see libraries as development partners. For some, such as Participant 10, this occurred personally as a result of the interview; they talked about seeing the invitation for interviews and "hopping onto it" because they were struck that working with libraries was a "totally unexploited partnership opportunity." For others (Participants 7,8,9,12), the change was a result of contact with colleagues who had knowledge of and experience with libraries in development. Seven participants expressed that libraries would be eligible partners if they were visible, and though raising visibility was generally recognized as a challenge, some were hopeful that this was "very doable" through "messaging and using [development organizations'] language" and "knowing how to leverage the space [ … ] and then knowing how to message it to the donors." (Participant 15) This outlook suggests that libraries can change their perception in the eyes of development organizations through strategic messaging which leverages their value, and we discuss some possibilities for this in the next section.

Roles for libraries
The optimism signaled above raises the question of how to exploit existing perceptions of libraries and work with what organizations are looking for in order to establish more concrete roles for libraries in development. This section offers the answers that emerged from our interviews. One possibility is promoting libraries as communityembedded institutions in order to gain visibility. When asked about the qualities desired in partners, the answers varied widely between participants and projects; however, one common theme was the importance of an established local presence. Participant 10 summarizes some of these related qualities: I looked at libraries because of what we're typically looking for in a partner is [ … ] that local context, right. We're looking for a credible partner as well to get over that kind of threshold of being able to say yes, we can do this, and we're working through trusted entities, right. Then that third piece, kind of that network community aspect of it is so crucial to the delivery of a lot if not all of what we do.
[ … ] Standing up something, like a local group that is unmoored from any of the community at large is really hard to do. No one wants to join. They have no association with it. It typically goes away. There's no sustainability plan for that group at the end of a project.
This desire for community-based partners, or those "steeped in certain communities" (Participant 15), was expressed by four other participants, and others mentioned linked qualities such as being trusted locally (3 participants), networked with local bodies (3 participants), and having local expertise (6 participants) as well as access to local data (2 participants). As Participant 14 explains, having insight into communities makes libraries effective in the eyes of development organizations: I think there is the argument to be made that libraries and more importantly the people who work there potentially have really interesting insights into both sort of lived realities of the communities in which they work, which is really important oftentimes, the work we all do. Also probably have interesting ways of sort of diagnosing challenges in those communities would bring theoretically a very multi or intradisciplinary approach to both problem identification and solution design.
Interviews signaled that development practitioners may be predisposed to see libraries as community-embedded institutions, given commonly held perceptions of libraries as community spaces (10 participants), spaces for informal learning (7 participants), and places for people to access resources such as information and internet (8 participants). At least five participants mentioned using libraries with their own families, and so leveraging these experiences of community value may be an opportunity for libraries to make connections and gain recognition for their work.
Another role available to libraries is a hub for resources that development organizations need in terms of space, data, and research. This role is an extension of being embedded in communities and can be built on other existing perceptions of libraries as "data intermediaries" (Participant 9) and "democratizing" (Participant 11) sites of information access; for example, Participant 18 discussed being "really impressed by the focus on kind of local materials and local resources" they observed in African libraries and "kind of the presence and availability of different research." These perceptions echo the fact that development practitioners already perceive libraries as places to access resources, as noted above, and this creates an opportunity for libraries to provide the resources needed by organizations for development work.
In some cases, participants indicated that the resource provided could be the library space itself. Many practitioners thought of a library as a "gathering place" or "venue" (Participant 5), "platform" or "collaboration space" (Participant 9), or simply "public space" that is sorely needed (Participant 15). Many were able to imagine possibilities for development programming that could take advantage of this space, for example, in hosting an "entrepreneurship club" for youth (Participant 16), a community-led radio station (Participant 6), or political information on citizen rights (Participant 18). This interest in the use of library space for development work is also reflected in the overall desire to know more about library capacity including available technology and connectivity (4 participants), staff training (4 participants), library networks (3 participants), other partnerships (3 participants), and general facilities data (2 participants). Given that the space met project needs, promoting libraries as venues for development activities would fit existing desires and perceptions.
It is also possible that libraries could provide the resource of data for development. When asked what data organizations would want from libraries, a common theme was the desire for data that addresses current gaps, expressed in terms of data that is unique from what is found in other official sources (4 participants), gender disaggregated (3 participants), locally focused (2 participants), and citizen-generated (1 participant). Put in other words by Participant 18, "I could see working with a library partner being really helpful and in that [ … ] they have like resources at their disposal related to data that we would not otherwise have access to or know existed." Participant 20 likewise mentioned interest in data that was "different" or "of higher quality" than that from national sources in-country. There is an opportunity for African libraries to meet this need through provision of data for data-oriented projects, as highlighted by 3 participants, such as Participant 16 who saw a role for libraries in "evidence generation" for governments working on disaster risk management and early warning systems. This role could also be fulfilled through joint data collection or "data collaborative work," emphasized by Participant 20 as a priority in their project. Libraries engaged in data collection could also leverage access to local populations as "places where a lot of people go" (Participant 5), given some organizations' interest in partners who serve specific populations such as women, youth, or refugees (4 participants). In this way, libraries could fill the gaps that exist in official data in order to furnish information development organizations want but find lacking in other official data sources, as African libraries have done via crowdsourced mapping (Lynch et al., 2020b;Young et al., 2020).
The role of libraries as a development resource hub could also be fulfilled via sharing development research. Three development practitioners cited disseminating research done by their organizations as an opportunity for libraries. As imagined by Participant 13, this could be done at the community level of "awareness raising" on research conducted by the organization on violence prevention. There were some examples of libraries successfully serving as institutional "custodians of knowledge" (Participant 9) and specifically African "local content and iIndigenous knowledge" (Participant 7) within their organizations. In this way, fulfilling this role has allowed their libraries to carve out an organizational niche for their work and gain recognition as contributors to development.

Data for advocacy
Of course, it is not enough for libraries to have the potential to fill important rolesthey also must prove to development organizations that they will succeed in meeting this potential. Overall, interviews emphasized the importance of library output data to advocate for libraries in development work. Output data, defined as information on the direct services and usage of the library, was seen by development practitioners as important to demonstrate the capacity and impact of libraries in development. In general, participants expressed a preference for impact data, but data which shows a direct connection between action and impact was acknowledged by many as a "holy grail" (Participant 14) and difficult to acquire reliably. Therefore, many expressed the utility of output data in order to make the case for impact. Both quantitative and qualitative data were discussed as useful, but there was a slight preference for quantitative data because, in the words of Participant 5, "the quantitative allows us to sell and to demonstrate undeniable success." As such, output data was generally accepted among practitioners as a necessary proxy for impact data.
Output data was also most frequently mentioned as the type of data that organizations would want from libraries, including data on usage (9 participants), number of people attending library programs (4 participants), content or collections (3 participants), and number of staff (1 participant). For usage, participants wanted to know both numbers as well as demographics of users. As described by Participant 16, development organizations want to know "Is the library a place that has a lot of traffic? Is it a place for where we can reach women or adolescent boys or adolescent girls?" Participant 13 similarly expressed wanting "to make sure that we were able to touch a lot of our target populations" including "marginalized groups" as a guaranteed "benefit of work through a library." Relatedly, interviewees reported using quantitative data (9 participants) and monitoring and evaluation data (9 participants) most frequently in order to assess partners, with attendance at events (Participant 13) and growth of program participants over time (Participants 5, 17) mentioned as just a few examples. Taken together, these responses suggest that quantitative data on library output, and in particular users and demographics, would be effective in order to address the concerns of development organizations and promote the viability of libraries. This information could be used in combination with data on library capacity, as described earlier, to demonstrate that libraries are capable of contributing to development work in varied ways.
As an example of how output data can be used for library advocacy, Participant 19 described their success in establishing an African digital library project for a development organization. By collecting the number of downloads, web page use, user location and the resulting geographical spread, this output data allowed them to successfully make the case for the broader development impact of the library project. Similarly, Participant 9 detailed collecting statistics for a library within a development organization on the number of visitors, number of people accessing materials online, number of people sharing materials on social media, and monthly events. They report that this output data has "made [ … ] senior leadership to see the potential of how we are serving the community" to the effect of doubling the library budget over time. As an important caveat, however, this participant had hired a social media specialist for the library and spent considerable energy acting as a "champion" with "passion," advocating and "fighting" for the library within their development organization. As other studies suggest (e.g. Lynch et al., 2020a), these opportunities are not available to all libraries, and the need to collect and use data for advocacy represents a significant burden for many African libraries as well as the necessity of capacity building to support these efforts if the benefits are to be realized.

Conclusion
Analysis of interviews suggests that perceptions of libraries among development organizations remain low and limited, associated with books and literacy and lacking capacity to participate meaningfully in development. However, there is evidence that these perceptions can and have changed over time, and many development practitioners are willing to see the potential of libraries as players in development. Combining current perceptions of libraries with what organizations want from potential partners, two possible roles exist for libraries to fill: trusted, local institutions with insight into their communities; and hubs for resources that organizations need for development such as space, data, and research. To position themselves in these roles, it would be strategic for libraries to collect quantitative data on their output (e.g. services) with a special focus on usage and demographics of their users in order to illustrate their impact on communities and ability to fulfill the needs of other organizations in development.
While this message is largely optimistic about the ability of African libraries to use data to shift perceptions, it is necessary to qualify this optimism by underscoring just how much variety there was in responses overall. There were very few points that were agreed upon by a majority of participants, and given that this was a small sample, the field of international development seems increasingly complex for libraries to break into. The challenge remains for libraries to navigate their way through the fractured, ever-shifting sea of international development agendas and players.
In addition, data is not a magic bullet; interviews intimated that other factors beyond data influence the partnership decisions of development organizations, including networking, referrals, previous relationships with an organization, and others. Our additional analysis (Young et al., forthcoming) describes these factors in more detail. In the end, African libraries may have data and passion, but without well-connected advocates and the ability to wield data effectively, their potential as development partners may stall. This underscores the need for additional research into what capacity African libraries currently have for collecting and utilizing data in addition to advocacy for building that capacity in order to gain visibility going forward. In addition, there is a need for additional, in-depth research on African libraries that have been successful in attaining development funding so that we can learn from them and share practical strategies with libraries on the ground.
This research then has varying implications for different stakeholders. For African library systems, we recommend to start by acknowledging the development work that libraries are already doing. For those who are interested in attracting development organizations, we recommend framing this work in terms of connections to the local community, attention to marginalized groups, and past successes in implementing projects. Library systems can also consider collecting quantitative data on library usage and user demographics, if these data are not already collected, but a variety of other types of data may be useful for different organizations and goals. They can also continue to network with development organizations, and for international audiences, consider promoting library strengths as community-embedded institutions and hubs for development resources including space, data, and research while also continuing to network and advocate for libraries at the local level.
For development organizations, we likewise recommend acknowledging the development work African libraries are already doing. Organizations can contact a library directly in a country where they work; for library location and contact information, our map of African libraries (librarysites.io) can be a useful tool that is still expanding as more locations are added. Given the power of personal contact to shift perceptions, we also recommend discussing the possibility of partnering with African libraries with development colleagues to make it a reality.
For those who support African libraries, we recommend enabling libraries to connect with development organizations directly. Library support organizations can arrange networking events for library representatives and/or put libraries into direct contact with development organizations. Simultaneously, they must continue to build the capacity of libraries to connect with development organizations through supporting skills such as marketing, storytelling, basic analyzing and reporting of quantitative and qualitative data, applying for grants, and writing proposals with a focus on using data effectively. This should be in addition to continuing to network and advocate for libraries at the local level. For researchers in this work, we recommend research on the data collection practices of public library systems across Africa, which we are currently conducting. We can also research and critically consider unintended consequences of aligning African public libraries with development organizations and how this affects libraries and their users, for example, how data about users are collected and used by development organizations. These efforts will help to sustain libraries' capacity to use data effectively for self-advocacy.