Public Libraries and Development across Sub- Saharan Africa: Overcoming a Problem of Perception

Public libraries and development organizations sharemany commongoals thatmake them strong potential partners in the Global South. In spite of these commonalities, libraries are often overlooked as development partners. This is unfortunate because these partnerships could allow librarians to more fully and effectively participate in the development of their communities. This paper explores strategies that libraries might employ to make themselves more visible to development organizations. It does so through analysis of interviews with practitioners who have funded or implemented development projects across countries in Africa. We find that these practitioners are open to working more with libraries, but that it will take a lot of work on the part of public libraries to accomplish broad shifts in perception amongst development agencies. Collecting and openly sharing data on the output and impact of public libraries is a first step in this work, but these efforts must also be augmented through increased marketing and networking by libraries. Libraries need to build internal capacity to use data to drive advocacy efforts and attract funding from development donors.


Introduction
Public libraries and development organizations share many common goals that make them strong potential partners in the Global South. Both groups seek to build strong community partnerships as they implement educational programs, expand access to information and communication technologies, support individuals to pursue new economic opportunities, and much more. In spite of these commonalities, libraries are often overlooked as development partners (Fellows, Coward, and Sears 2012). This is often the result of a perception problemlibraries are not framing their own work in terms of development, and development organizations therefore do not see the potential value in partnerships. This is unfortunate for public libraries, because partnerships with development organizations could provide much needed resources that would allow librarians to more fully participate in the development of their communities. Public librarians need tools and approaches that would enable them to overcome this perception problem and attract development partners and funding.
This paper explores strategies that libraries might use to make themselves more visible to development organizations, with a focus on what types of data they might collect about their current development impacts. The paper is based on analysis of interviews with practitioners from development organizations, who were asked about their general perception of libraries and about the types of evidence that might change their perception. These interviews revealed that development organizations continue to have very little perception of public libraries, in the sense that they do not readily think of them as potential partners or development actors. However, many of these participants also indicated that they would be open to partnering with libraries, once they started considering them. We found that data collection would be an important first step in attracting the attention of the development sector, but that librarians also need to be able to use these data to tell stories, network, and respond to specific proposals. We conclude that more applied research is needed to expand the capacity of Global South libraries to engage in these activities.

Public Libraries and Development
To understand the potential role of African public libraries in development, it is first important to understand the historical relationship between these libraries and international organizations. Naturally, the history of libraries varies across different African countries, and it is important to understand how these unique histories affect the ability of specific library systems to attract development partners. However, the shared colonial history of many sub-Saharan countries has produced general regional trends, meaning that a continental history can provide useful context for discussion. Amadi (1981aAmadi ( , 1981b argues librarianship took an oral form in many early African societies. However, colonial interventions ensured that text-based information dissemination practices gradually supplanted these oral cultures and defined modern African library models. Over time colonial libraries established libraries within their territories, primarily for the benefit of their military officers and families living in-country (Moahi 2019). These libraries were neither designed for nor accessible to the local communities in which they were built. As a result, the format of the library rarely made sense within the context of the needs or cultures of those communities (Albright 2007;Amadi 1981aAmadi , 1981bDavis 2015;Matare 1997;Nyana 2009;Strand 2016;Sturges and Neill 1998).
Gradually, an Anglo-American model gained increasing dominance across the continent given increasing influence of the United States across Africa. As colonies started moving toward independence, many sub-Saharan countries experienced an influx of expatriate librarians that were trained in the Anglo-American tradition and were eager to set up library systems in newly independent countries. Sturges and Neill (1998) argue that many of these expatriates had the good intention of setting up systems appropriate for the local cultural context, but often failed due to the lack of "blueprints" for setting up such systems. As a result, they reverted to their Anglo-American training. At the same time, African governments enthusiastically supported these efforts, with the hope that they could accelerate national development (Moahi 2019;Mostert 2001).
The Anglo-American tradition also gained hegemony at an international scale, shaping the early relationship between international organizations and libraries. Sturges and Neill (1998) identify the 1953 UNESCO Seminar in Public Libraries in Ibadan as being particularly important. The Seminar recommended the adoption of the UNESCO Public Library Manifesto, which itself pushed for the adoption of centralized public library systems across the continent (Sturges and Neill 1998). However, organizations like UNESCO tended to take a laissez-faire approach to establishment of these library systems, rather than pushing for and funding careful planning within central governments. Bouri (1994) argues that this approach was encouraged by the dominance of modernization theory within international development, which held the basic assumption that all societies follow the same basic economic development trajectory. African countries were viewed as being farther back on this trajectory, and international development organizations believed that they could help to accelerate these countries' path. In the library context, these organizations took a "development by demonstration" approach, which postulated that conferences, seminars, expert guidance, and pilot projects could jump start library systems in African countries (Bouri 1994).
This approach, unfortunately, has been widely regarded as a failure (Frimpong 2015). As Bouri argues, the "difficulties facing the establishment and sustained development of public libraries in the developing countries pointed to the much more fundamental need for national planning and government's direct commitment to the development of library services" (1994,160). International organizations gradually shifted from modernization theory to dependency theory, pushing for more thoughtfully planned development of endogenous institutions and resources within African countries (Bouri 1994). This represented an opportunity for the library sector, since information was viewed as a critical resource in this push for endogenous development (Bouri 1994). Unfortunately, though, these development organizations increasingly thought about information in a way that ultimately undermined public library systems. In contrast to the public library emphasis on intellectual, social, and political freedom, information systems were viewed much more technocratically as input into (exports-driven) economic development (Bouri 1994). Resources were therefore often directed into the development of elite science and technology information systems that catered to technocrats at the national level, rather than to libraries that served the general public. Libraries were further negatively impacted by economic declines in the 1980s, which further constrained their access to economic resources (Moahi 2019).
Through these processes, public library systems slowly lost support of international agencies and national governments while also becoming even less relevant to the general public. A 2012 University of Washington study found that public libraries "are typically overlooked as partners in development" (Fellows, Coward, and Sears 2012, 1). They key question for this paper is how public library systems can overcome this perception issue, to transform their relationship with development organizations and programs across the continent. And, in fact, current trends seem to be opening up space for this transformation. The focus on the United Nations Sustainable Development Goals (SDGs) has once again expanded international views of development beyond the narrow economic framing of the NATIS movement, offering libraries the chance to market their long-standing relationship with social and political forms of freedom. Moreover, international organizations like the International Federation of Library Associations and Institutions (IFLA) and Electronic Information for Libraries (EIFL) are lending support to this cause, by strongly advocating for an increased role for libraries in development. There is also growing excitement within the discipline of library and information science (LIS), with many academic papers calling for more research into how libraries can facilitate community-driven development (e.g., Fourie and Meyer 2016;German and LeMire 2018;Igbinovia and Osuchukwu 2018;Moahi 2019;Mutula 2003).
Unfortunately, this research has been largely aspirational in nature, and not grounded in empirical proof of how libraries can or do contribute to local development efforts. This research is often based on intuitive or conceptual claims about what libraries are hypothetically capable of, rather than in empirical studies of how libraries have been contributing to sustainable development. The most common argument made is that development is driven by information, that libraries are key sites for the transmission of information to the broader public, and that libraries are therefore positioned perfectly to contribute to development (e.g., Abdulla 1998;Hart and Nassimbeni 2016;Kavalya 2007;Onoyeyan and Adesina 2014). Similar arguments were made about the ability of libraries to grant access to information and communication technologies (ICTs; e.g., Akintunde 2004;Chisenga 2000) or lifelong learning programs (e.g., Abata-Ebire, Adebowale, and Ojokuku 2018; Akparobore 2011; Fourie and Meyer 2016) which can spur personal empowerment, broader community development, and democratization. Other researchers emphasize the physical space of the library as a unique community hub that brings people together, thereby allowing libraries to build close connections throughout the community and have broad impacts with their programming (e.g., Ashraf 2018). Authors also emphasize the values of public libraries, highlighting the importance of principles including empowerment, equality, diversity, and democracy, among others. Although these articles make a strong analytic case for why libraries could make strong development partners, they lack empirical evidence that libraries are actualizing this potential.
A handful of studies have attempted to provide more empirical proof of the role of libraries in development (Stranger-Johannessen 2014; Stranger-Johannessen, Asselin, and Doiron 2014), but they tend to focus on narrow case studies rather than documenting the wider impact of the library sector (e.g., Dent 2007;Dent and Yannotta 2005;Emojorho and Ukpebar 2013;Filson and Arthur 2017;Igbinovia 2017;Jones 2009;Kiconco 2018;Sow and Vinekar 2012;Strand 2016). This has led some researchers to call for the development of stronger and more generalizable analytic models for assessing the social and development impact of the library field. At the most basic level, several authors suggest that libraries should start by collecting stories about how they are making differences in the lives of their users (e.g., Igbinovia 2016;Kabamba 2008;Ngian 2017). The IFLA (2019) Storytelling Manual is one resource that libraries can use to help collect, frame, and tell these stories. However, Hart emphasizes that libraries should not stop at isolated stories, fearing that they may be too anecdotal to make a difference for libraries, particularly in the current "climate of accountability and reduced public spending" (2004,114). Instead, libraries need to adopt more rigorous forms of evidence-based advocacy (Economic Commission for Africa 2003).
Unfortunately, as German and LeMire point out, "little has been written about strategies for assessing library outreach efforts," (2018, 66) particularly at an international level. The most common approach to assessing library impact is through surveys of users of specific programs, but these are rarely carried out continuously or systematically (German and LeMire 2018). While these surveys may be useful for assessing the success of individual programs or services, they are less useful for examining long-term or regional trends related to sustainable development. Other researchers suggest that libraries consider engaging in long-term needs assessment surveys within their communities, followed by surveys of how well their services are meeting those needs (Gould and Gomez 2010; Kabamba 2008). Forsyth (2005) argues that libraries can further connect services to specific development goals, giving these surveys a more rigorous structure for demonstrating the role of libraries in development. Kabamba (2008) points out that this survey work can present difficulties, but that these can be overcome by building strong partnerships with the community and with research organizations. Of course, many of these recommended approaches also require consistent, longitudinal data collection and the support of advanced modeling software (Imholz and Arns 2007). It is therefore unclear whether they offer a sustainable approach given the current capacity of many public libraries in Africa, although they perhaps warrant further exploration. Moahi (2019) argues that, from the African context, considerable work is needed to build the capacity of public libraries to engage in any form of effective valuation. They will need both quantitative skills and the ability to weave statistics with qualitative stories about their impact.
Before concentrating resources and effort into building this capacity, however, it is important to understand what impact this capacity would have on the library sector. While these are important conversations within LIS, to a certain extent they beg the question of whether the development sector will care about data that describe the role of public libraries in development. If libraries want access to development organizations as funders and partners, then they need to be demonstrating their value in ways that these organizations understand. The LIS sector therefore needs to expand this nascent discussion to development researchers and practitioners, to understand what types of evidence would best position libraries to expand their role in the field of international development. This paper attempts to begin such a discussion between LIS and the development sector.

Methods
This paper emerged from a larger project being carried out collaboratively by the Technology & Social Change (TASCHA) group within the University of Washington Information School and the African Library and Information Associations and Institutions (AfLIA). This project, called Advancing Library Visibility in Africa (ALVA), broadly examines the relationship between public libraries and sustainable development across sub-Saharan Africa. Other components of the project have focused on understanding and building the capacity of African library systems to collect and share data generally (e.g., Lynch et al. 2020aLynch et al. , 2020bLynch et al. , 2020cYoung et al. 2020). In contrast, this paper draws on a set of interviews with development practitioners about what types of data libraries were most persuasive in driving their decision-making around partnerships. The long-term goal of the project is to combine these two research streams, to build the capacity of library systems to collect those forms of data that they can use for advocacy within the development sector.
The purpose of the interviews was to understand how international development organizations perceive public libraries in Africa, and to understand how libraries might position themselves to take advantage of potential partnerships with these organizations. Participants in the interviews were adults who work at development organizations that fund or implement projects in Africa. Potential participants were identified through (1) the pre-existing contacts of the research team, based on their prior research experiences; (2) a survey of the websites of well-known development organizations; and (3) the recommendations of existing participants in the study (e.g., respondent-driven or snowball sampling). Through this purposive sampling approach we recruited 21 participants for interviews. These participants were based across North America, Europe, and Africa, and they had anywhere from seven to 45 years of experience working in the development field. They all had direct experience funding or implementing projects in African countries. Five of the participants had positions within their organization that were directly related to library advocacy in some capacity, while the other 16 did not. We should note that the rate of response to our recruitment e-mails was quite low, a fact that we believe may be related to the lack of visibility that libraries have within the development field. We invited a total of 178 individuals to participate in an interview, but only 50 individuals responded (28%) and 21 (12%) eventually agreed to participate. Indeed, some of those that agreed to be participants at first expressed reluctance to participate in research about libraries, because they felt that they would not have much to contribute. This underscores the need to continue expanding LIS conversations with the development sector, and for enlisting development scholars (who have access to existing networks of development practitioners) in LIS research projects.
Interviews were semi-structured and conducted in English via a telecommunications platform of the participant's preference (e.g., Skype). Interview questions were grouped into three different sets. The first set of questions focused on gathering general information about the organization for which the participant works, including what types of projects they fund or implement, the organizations with which they partner, how they select partners, and how they evaluate the success of projects and partnerships. The second set of questions then shifted to asking more explicitly about libraries. This included questions about their general perception of libraries, whether they have ever funded or partnered with a library, and what types of data or skills libraries would need to demonstrate that they make for effective partners. The final set of questions gathered background information about the participant themselves, including their personal experiences with libraries and the amount of experience they have in the development sector. Interviews were audio recorded and later transcribed for analysis. These transcripts were anonymized to ensure the confidentiality of participants, and all quotes within this paper are attributed to pseudonyms.
The researchers used the Dedoose software platform to perform an inductive analysis of the interview transcripts using a grounded theory approach (Clarke 2003;Glaser 1978;Kitchin and Tate 2000). They looked for common trends that shed light on current perceptions of public libraries within the development field, data or other factors that influence partnership decisions being made by development organizations, and approaches that libraries could use to make themselves more visible to potential development partners. Two researchers analyzed the transcripts concurrently, with regular check-ins to triangulate coding schema and ensure consistency of analysis (Baxter and Eyles 1997). Results of this analysis are discussed in the remainder of the paper.

Discussion
Our interviews reaffirmed that libraries continue to face a perception problem with development organizations, but that most of our participants were very open to partnering with libraries in the future. Our participants believed that the increased collection and sharing of data could help libraries overcome the current perception problem, but that the development of this stronger data culture is likely only the first step in a broader project of re-marketing libraries. This confirms the need for further research into how library capacity might be expanded to play a stronger role within development partnerships in the Global South. These dynamics are explored below.

Current Perceptions of Libraries
The interviews strongly confirmed the findings of the 2012 University of Washington study, that libraries (and particularly public libraries) are generally not considered as partners by development organizations currently doing or funding work in the Global South. Participants confirmed that they generally do not perceive libraries in their work at all, meaning that they think very little about, and have little knowledge of, libraries in the context of development. Of our 21 participants, six individuals felt that they had no knowledge whatsoever about libraries operating across Africa and 11 participants had little to no professional or personal experience with public libraries. In fact, many of our participants expressed surprise that they were being asked to participate in an interview about libraries in the first place. As one participant said, "I don't ever recall discussing libraries with my colleagues. Now, I'm embarrassed to admit that. Also, I haven't even thought about it until you all brought it up.
[…] why have none of us thought about it?" (Participant 1) Similarly, Participant 20 confirmed that libraries would be an entirely new type of stakeholder for their organization, and several other participants confirmed that they have never reached out to libraries to explore possible partnerships. When asked questions about the strengths or weaknesses of libraries as actors within the development sphere, many of these participants felt uncomfortable even hazarding a guess. For example, when one participant was asked how they thought libraries might compare to their current partners, they said, "I don't know if I actually have a good answer just because we haven't worked with any" (Participant 14). Libraries were so far off of their radar that they weren't even certain how to start thinking about them within a development context. Even when our participants did themselves think about libraries, they still felt that libraries are not widely perceived within the development community. One participant said that the "whole spiel about [libraries] not being on the radar, I think is a pretty commonly held perception" (Participant 14), while another argued that "there are not development agencies that look to work with libraries as partners … that's basic knowledge" (Participant 7). Participants tended to identify different reasons for why libraries face this perception issue. For example, Participants 8 and 20 argued that the problem has emerged because libraries aren't visible within their own countries, meaning that national governments are unlikely to market their libraries to international actors. In contrast, Participant 15 argued that the development sector itself is too biased toward formal education to want to explore "the informal, non-measurable space" that libraries inhabit. Regardless of the reason, this feeling that libraries are not being perceived within the field was very widespread. A total of 15 of our participants either did not perceive libraries themselves or felt that libraries were not at all visible to most of their colleagues.
This lack of visibility led many of our participants either to believe that libraries do not exist at all in the African contexts in which their organizations work, or to make assumptions about libraries that were not grounded in any sort of empirical proof. Six participants believed that libraries did not exist in the types of places in which they do work. Some of these participants were quite confident in their assessment, saying things like, "There are no libraries in these communities" (Participant 6) and, "There aren't that many libraries, you know" (Participant 13). Others were more uncertain. Participant 17 said, "I have no idea if libraries exist in any of the countries I operate in. […] I literally don't know the landscape." While it is certainly possible that libraries do not exist within particular communities in which our participants have done work, our project has mapped a fairly dense network of libraries across the continent (e.g., Young et al. 2020). Furthermore, one of our participants that does advocacy for libraries within the development context indicated that they regularly have to counter the argument that libraries do not exist. They described one such situation: I remember having a conversation with a big ICT head in the [city] area years ago, and I was talking to him at a round table discussion about libraries in Africa. He kind of dismissed it saying, well, everyone knows that libraries in Africa are ill-equipped and there really aren't any. We used that as an opportunity to say, well, actually there is not a lot of data out there, but there are a number of libraries throughout the continent. (Participant 8)

Overcoming a Problem of Perception
As this participant describes, libraries are not only perceived as being largely non-existent, but also as being ill-equipped to serve their communities. And, in fact, our participants generally had the impression that libraries were not innovating in ways that would allow them to respond to pressing development needs. Thirteen of the participants admitted that they continue to view libraries as only being relevant in the area of book-lending and literacy. Yet, they also admit that this perception is not really based in evidence. As Participant 14 said: that's the perception barrier to get over if there is both the current or potential ability for libraries to play that kind of delivery role, it's getting over this somewhat of a caricature of an 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 or is optional. (Participant 14) Participant 9, another library advocate within the development field, bemoaned similar dynamics. They said, "Libraries are seen as […] just spaces for academics or reading and references. They're not seen as spaces for interaction, you know, as open and flexible spaces.
[…] That's why we really need to work on changing the perception of libraries as a key development partner." These views confirm many of the suspicions evident from the earlier literature review. Libraries are barely perceived at all within development circles, and when they are perceived it is in a negative light. This raises the question of whether these views can be changed and, if so, what methods might be used to change them.

Changing Perceptions with Data
Fortunately, though, many participants believed that libraries could be eligible partners if they were more visible to development organizations. In fact, several participants said that their invitation to participate in this project had already led them to think more seriously about libraries as potential partners. These participants agreed that, upon further reflection, they could easily imagine how libraries are contributing to development within their communities. Participant 13 does a nice job of summarizing transformations in their own thoughts: I don't know that anyone has an opinion on it […] but I do think that some sort of … just the reflection that I've done during this call, I think that if we did talk about a library and it existing in this context and how we could leverage that, I think that people would agree that it would be a useful partner, especially for sort of awareness raising and social cohesion, and also publishing conflict research and disseminating it. (Participant 13) Participant 10 similarly said that they chose to participate in the project because they felt that "this was a totally unexploited partnership opportunity" that they wanted to learn more about. Others agreed that it "seems like it's an area of such potential" (Participant 1) and that a library could be a "natural partner" (Participant 8) that is "critical to development" (Participant 1). Participant 12 indicated that their "complete ignorance" about libraries would make it easy for someone to convince them of the importance of partnering with libraries, since they don't actually have any negative thoughts about them. Participant 20 felt that libraries would likely have been eligible partners for past collaborations, if only they had been more visible to their organization. This is the most promising result of our studydevelopment organizations can imagine the relevance of libraries to their work, and they are eager to learn more.
Broadly speaking, our participants said that they would be open to looking at any type of data they could get their hands on to learn more about libraries. However, quantitative impact data were generally described as the most persuasive and desirable form of data. Participants extensively discussed the different applications of quantitative and qualitative data within the development field. Most participants believed that both types of data were important, and that was vital to combine many different types of data so that they are "able to have a really big picture" (Participant 16). Participants found qualitative data to be important for evaluating some of the more intangible types of impacts of their partners and projects, and for telling stories "that can complement and provide color to the depth of quantitative information you might be collecting" (Participant 12). However, most participants also agreed that quantitative information was more important in the field because it was what the largest donors expected to see. Participant 5, for example, said that "both [are important], but definitely the quantitative allows us to sell and to demonstrate undeniable success." Participant 13 indicated that their team tends to use qualitative data the most for their own internal evaluation purposes, but that their leadership members "are thinking about numbers and passing those on" because it's what their partners want. Participant 5 goes even further, telling us that "anything that's numerical is good." The takeaway for libraries is that any sort of data can be useful for attracting the attention of development partners, but that quantitative data are most likely to have sway with the large donors that drive work in the field.
Participants broadly preferred data that documents and evaluates the impacts that libraries are already having in their communities. As one participant put it, "I'd love to know not how many people did you train at this library, if libraries do like workshops or something, I don't know, but, you know, like what did your evaluation show, what was like the change in skill level six months later, or something" (Participant 12). Participant 5 emphasized that donors really want to see quantitative measurement and impact data. Nevertheless, many participants also recognized that rigorous evaluation or impact data are extremely difficult to collect. They are therefore willing to take many other types of data which might act as proxies for impact. Many participants identified output or usage data (e.g., number of library users, number of books that have been checked out, number of people participating in programs) as being valuable. Participant 14, for example, argues that output data may not have quite the same power as impact data, but they still give an indication of what libraries are capable of. They said: Yeah, I mean both are important. The impact stuff is, as you know, so hard for any of us to really defensively stand up in most cases, so if people have those and they're actually defensible, that's fantastic. Short of that, all the output stuff is definitely relative and it at least gives you a sense of scale or potential reach, not necessarily 'transformed for that', but you are you know impactful. Yeah, I would think both are valuable just … and of course we would all love to be able to have 'killer' impact stats … it's sort of a holy grail. (Participant 14) Participant 15 agreed, and indicated that output data is actually more commonly used throughout the field due to the difficulty of getting rigorous impact data. Many of the participants identified specific forms of output data that they thought libraries could easily collect, including number of patrons that come to the library, patron demographics, what types of services they regularly provide collections data, and patron usage of various types of content or services.
Ultimately, the participants wanted to see any data that could demonstrate that libraries are generally capable at engaging their communities and that libraries have a "track record" (Participant 14) of success in implementing development-oriented work. Specific types identified by participants included data on facilities, technology and connectivity data, staff training and expertise, and past partnerships. These data would help the development organizations assess the capacity of the library to carry out projects and also their ability to scale projects in partnership with other organizations or libraries. If libraries can begin to collect and openly share these data, they are more likely to catch the eyes of development partners and funders. Of course, this raises important questions: even if the data are available, how can libraries ensure that development organizations actually see them? What are the best methods to share or communicate their capacity and impact, and what tools or expanded capacity do they need to engage in these methods? These questions, and more, are explored in the next section.

But, Are Data Enough?
Unfortunately, several of our participants emphasized that collecting data and making it openly available may not be enough to widely change perceptions of libraries. Librarians also need to be able to use these data to respond to specific development needs, to tell stories about the impact they are having in their community, and to network within the broader development field. This can present significant challenges for libraries, since it requires additional skills beyond collecting data.
First, while our participants were interested in viewing a wide range of data, they also emphasized that these data need to be responsive to the needs of particular projects. They indicated that funders require different types of data from potential partners depending on the program they are trying to implement or the research questions that they are trying to answer. As Participant 1 said, "I'm having trouble giving a kind of general answer. […] Project by project, it's different. I mean, yeah, it's different from project to project." Collections data, for example, is unlikely to convince a development organization that a library would make a good partner for implementing a digital literacy program. This means that libraries cannot simply collect generic data, but need to have sufficient expertise to adapt their data collection strategies to respond to emerging development needs as expressed by donors. Participants continuously emphasized that development is a highly donor-driven field. Additionally, donors often don't just go around looking for data on partner organizations, but want those organizations to package their data in the form of a project proposal responding to a specific call for proposals. These calls are often competitive, and organizations strongly benefit from experience with and expertise in writing proposals in order to successfully navigate the process. Several participants suggested that libraries need to have specialized units responsible for writing proposals. Participant 20 suggested that if "they want to partner with outside funders there is a need probably to have some kind of a unit that allows them, because it's not easy to write proposals and to do the reporting that is required […] so then we need to sort of have […] almost like a consultant within the library itself." This seems like an impossible suggestion for many libraries and library systems, given their current capacity.
To make matters worse, sometimes data do not even play a very strong role in the decisions that development partners make about partners. They instead choose to work with organizations with whom they already have personal experience, or organizations for whom they have a strong referral from a trusted source. Participant 18 explained that their first choice is to go with partners with whom they have Overcoming a Problem of Perception already worked, so that they know that they work well together. If that is not possible, then they have to go through long processes of networking with new potential partners to ensure that it is a good fit. Ultimately, they emphasized that they make decisions based on "evidence from direct experience as opposed to going to some sort of, like, dataset or website or platform somewhere" (Participant 18). Many of the participants noted the influence of direct referrals from colleagues. Participant 14, for example, indicated that if they did not have direct experience with a partner, then they would look for referrals from others: I think that number 1, there is no certain silver bullet data point and I think the kind of information that tends to matter most, at least initially, is referrals. I mean, I can't overstate the importance of that and I'm not saying it's even a good thing, but I think in the real world, a referral from a trusted intermediary or interlocular, you know, that we have trust in is like overwhelmingly influential, in terms of who we would end up partnering with. (Participant 14) Naturally, this creates difficulties for librariesif they currently are not partnering with development organizations, then they do not have access to these types of referrals. In response to this limitation, participants suggested that libraries need to put a lot of time into developing data-driven stories that they can use for marketing and networking purposes. Participant 10 indicated that "a lot of it needs to be simply, just say, optics … I mean, this really is putting yourself out there, so people really understand that it's [partnering with libraries] a possibility more than anything else." Participant 14 agreed that this marketing was important, but also recognizing that it is a significant challenge for libraries: How do you flash those skills, expertise, and potential in front of folks and shops like us or the other thousand and one, you know, sort of intermediaries around the world trying to do this to sort of broaden their horizons and basically market those services so that it occurs in one's brain to pick up the phone, proverbially, and like call out to a library or at least have a long list of, you know, potential partners.
Libraries once again face a catch-22. They want to partner with development organizations because they need greater access to resources that they can use to expand their capacity and ability to create positive change in their communities. However, to gain access to development organizations, they need to have much greater capacity in a whole range of areasdata collection, data analysis, proposal development, marketing, networking, and more. How libraries can achieve this capacity, so that they can become visible, remains an open question. We've attempted to position our larger project as one answer to this question, and are exploring how we might develop tools and process that library systems might use to market themselves in the ways described above. We hope that this paper might also provide other LIS researchers and library support organizations with an applied research agenda for supporting libraries in these areas.

Conclusion
Across sub-Saharan Africa there is growing excitement within the LIS field about potential connections between libraries and development. Both the 2018 IFLA General Conference Assembly and the 2019 third AfLIA Conference & fifth African Summit, for example, had strong focuses on the work that librarians are doing to achieve UN Sustainable Development Goals and the goals of the African Union's Agenda 2063. Although libraries are well-positioned within communities across the continent to advance development, many of these libraries still require increased capacity, resources, and support to fully achieve their potential. One possible source for this support is from international development organizations. Unfortunately, these development organizations tend not to perceive libraries as partners, meaning that they are unlikely to fund or otherwise support the library sector.
This research explored the current perceptions that development organizations have of libraries, and asked practitioners what libraries might do to change these perceptions. We found that development practitioners are open to changing their minds about libraries, but that it will take a lot of work on the part of libraries to accomplish this shift in perception. Collecting and openly sharing data on the output and impact of libraries is an important first step in this work, but it must also be augmented through increased marketing and networking by libraries. They also need to build internal capacity to repackage or use collected data to drive advocacy efforts and attract funding.
Based on these findings, we make several recommendations for additional work for and with library systems across the continent. First, there needs to be more research into effective data collection practices within African library systems. Current data collection practices tend to be opaque, and data are not often shared openly. We need to know more about current practices, current data availability, and capacity to adjust or augment processes. How can data that is already collected be made more visible to development organizations? What tools or processes might help library systems produce more data, or more accurate data? How can current collection practices be adjusted to emphasize the types of data most valued by development organizations, including impact data? Answering these questions will require not only research, but also sustained efforts by libraries and library support organizations to build their own capacity and data culture. Second, more work needs to be done to highlight the capacity of libraries to perform and scale development partners. If development organizations are looking for referrals and evidence of past successes, then what efforts can help to get these things in front of these organizations? In the context of our project, we are working to build a Web platform that makes data about African libraries more easily accessible. However, it's unlikely that research platforms alone will succeed in making libraries visible. Libraries need to be doing more networking with local development actors and government partners; national government agencies need to be more vocal in advocating for libraries when they speak to donors about development; and development agencies that have funded book purchases for libraries need to expand their view of what libraries can accomplish. Third, libraries need more capacity to leverage their data within proposal applications, marketing, and networking. How can this capacity be effectively built? Should the focus be on building capacity within the national library system, individual libraries, or library support organizations? What tools or trainings would be most effective in building this capacity? To what extent can development organizations themselves step in to help build these capacities? These are difficult but important questions for researchers and practitioners alike.
Finally, research should consider the long-term consequences of aligning African public libraries with development organizations. As this paper's literature review suggests, external influence has not always been beneficial for the development of libraries across the continent. Too often this influence translates into the adoption of colonial norms, or to overt failures like NATIS. Libraries need to understand what unintended consequences might result from the decision to chase development funding. How might a focus on collecting impact data, for example, lead libraries to emphasize more formalized learning programs, and what would the effect be on their patrons? Will the expansion of data collection create opportunities to shut down libraries that are underperforming? It's also possible that libraries' efforts would be better served in improving their image within their own country, rather than focusing on international organizations. In this case, it may be worth tempering the excitement around aligning public libraries too strongly around international tools like the UN SDGs. Other studies have shown that libraries face similar challenges related to local perception (e.g., Baada et al. 2019;Elbert, Fuegi, and Lipeikaite 2012;Lynch et al. 2020a), and some of our participants also agreed that focusing on national-level advocacy could get libraries access both to national and international funding opportunities. As Participant 20 put it, "the question here is not only how can libraries be more visible to us, but how can libraries be more visible within their own country." Such an effort may require similar effortsnew forms of data collection, networking, and marketingbut framed in different ways. This approach may have the advantage of finding local or regional allies that are in closer alignment with local needs, although these local entities may also face greater resource constraints as compared to international donors. At any rate, these questions are just as important as the technical questions of how to implement new data collection practices. Our own project will continue to explore this broader research agenda, and we hope that other LIS and development researchers will join the conversation.