Introduction
Within the discipline of political science, scholarship and teaching on Africa faces a serious problem that demands urgent attention and redress. The problem is straightforward: the producers of knowledge on ‘African politics’ are overwhelmingly non-African academics located thousands of miles from the contexts about which they write.
As a non-African aspiring researcher based outside of Africa, I have come to realise more acutely the glaring inequality and inequity within the current system of knowledge production. As a graduate student preparing for courses and comprehensive exams in African studies and comparative politics, I became aware of the utter scarcity of African and Africa-based scholars in most of the assigned material. I also became well versed in the grand (hegemonic) theories in the ‘African politics’ literature (e.g. ‘neopatrimonialism’, the ‘rhizome state’), whose exclusive provenance was North American and European scholars based outside of Africa. Later, while conducting research in Mozambique and South Africa and then writing my dissertation, I discovered that most of the conceptual, theoretical and interpretive frameworks I employed were created by non-Africans. At the same time, as I interviewed and chatted with locals and read through materials written by Mozambican and South African academics, it was readily apparent to me that they possessed far more precise and revelatory theories and insight into how local politics worked in their communities and countries. Finally, when I was asked to teach a course called Civilizations of Africa, I worked diligently to ensure that my own selection of readings was less biased against African and Africa-based scholars. I still, reluctantly, found myself relying on textbooks written by non-Africans. A luta continua.
Let me be clear: I am not suggesting that only African or Africa-based scholars should speak to ‘African politics’. That would preclude me and many other dedicated academics from conducting research and teaching on the topic. Instead, the aim of this paper is to raise awareness and motivate actions to rectify the large underlying power asymmetries existing within the system of scholarly production, which clearly disadvantage African and Africa-based scholars from producing and spreading knowledge. As Olukoshi (2006, 535) wrote: these asymmetries ‘play to the advantage of non-African high priests of the discipline’ and they have been exacerbated by multiple factors. One important factor is the decline of higher education within Africa, which was accelerated by economic crises and structural adjustment programmes in the 1980s and the World Bank’s push to deprioritise higher education (in favour of primary and secondary) and to restructure and corporatise university education to meet the demands of the market (Mamdani 1993, 2007; Ndlovu-Gatsheni 2020). Currency devaluations, falling export revenues, and rising budget deficits also contributed to infrastructure and resource shortages. This confluence of factors has pushed an increasing number of Africa-based scholars to either leave the continent or continue to struggle under precarious conditions including inadequate salaries, and larger administrative and teaching burdens with less time and resources for research. Promising local scholars have also flowed out of local universities to more well-remunerated jobs in consulting or in international development agencies (Duffield 2014). The decline of African universities and African scholarship in the 1980s and 1990s also coincided with the rise to prominence of Euro-American descriptions of African states as broadly rent seeking, predatory, neopatrimonial or clientelistic (see Mkandawire 2013 for a probing review). Other factors that have deepened power asymmetries in knowledge production include: a shortage of resources among many Africa-based publishers; a deprioritisation of a culture of deep immersion in communities being studied;1 journal gatekeeping and publication biases that privilege particular methods, data, concepts and theories that are more likely to be employed by academics in the US or Europe. These problems extend well beyond the discipline of political science and are salient throughout the social sciences and humanities (also see Zeleza 1996; and Mama 2007, 4).
This paper’s (small) intended contribution to the effort to change the system of knowledge production and instruction is to increase awareness and put forth an ethical imperative to take individual and collective actions. Moreover, to the best of my knowledge, this is the first paper that provides a quantitative analysis of university course syllabi and PhD reading lists to illuminate one specific manifestation of the problem. This approach demonstrates how the knowledge production problem is not confined to academic publication but rather pervades the centres of higher education and affects the learning and training of millions of future practitioners, policymakers and academics. I focus on undergraduate and graduate training, which complements one study analysing two leading publications in African studies (Briggs and Weathers 2016) and many narrative theoretical articles (Ayoade 1980; Crowder 1987; Irele 1991; Yankah 1995; Zeleza 2002, 2009; Olukoshi 2006; Mama 2007; Robinson 2008; Mkandawire 2013) as well as media articles (Pailey 2016; Akinro 2019) and popular blog posts (Strohm n.d.).
The paper is structured as follows. The next section interrogates and problematises the concept or category of ‘African politics’ and the ‘Africanists’ who produce knowledge in this area. The third section, ‘Quantative data from undergraduate course syllabi and PhD reading lists’, presents and analyses the data from undergraduate syllabi and doctoral comprehensive exam reading lists. Finally, the section titled ‘Efforts to address the problem, and how to strengthen them’ discusses initiatives that are working to address the knowledge production problem and how others might contribute.
What is ‘African politics’ and who are the ‘Africanists’?
One way to begin to confront the problem is to challenge the concepts and categories that capture, undergird and perpetuate the power asymmetries within the discipline. Two are particularly problematic, and persist nonetheless. First, although widely used in academic publications, the category or sub-field of ‘African politics’ is misleading and impedes a more accurate and complete understanding of systems of government and political actions and behaviour across varied African contexts. It is linked to an ignominious history in which Africa and its diverse political and social configurations were homogenised, essentialised and distorted by Europeans and other Westerners (in this case, colonisers, explorers, missionaries and colonial anthropologists). During this time, African voices and rich oral traditions and histories were effectively suppressed and locked out of ‘knowledge’ production pertaining to their own complex societies and indigenous systems of local governance (Mkandawire 2005; Pailey 2016). Thus, the production of this ‘new knowledge’ regime and the hegemony of Euro-American epistemologies served to advance the goals of political domination and economic extraction during imperialism and colonialism (see Ndlovu-Gatsheni 2020).
While much progress has been made in correcting biased, inaccurate and simplified portrayals (owing to the work of Africans and non-Africans alike), the current system of knowledge production remains deeply problematic. One manifestation of this is the use of broad categories such as ‘African politics’ and their associated theories, which are often generalised to all of Africa based on single-country ‘field’ research or, increasingly, a field, survey or lab experiment conducted within one African country. This practice is carried out by many American and European ‘Africanists’ who regularly formulate and broadcast their authoritative arguments in the titles of their books and articles.2 Note, for example, the stacks of books and articles published in leading academic presses and political science journals titled using the formula ‘Democracy/Citizenship/Political Struggle in Africa: A Study of Country B’. This problem is made worse by the imbalanced selection of particular African countries (mainly large anglophone countries such as South Africa, Nigeria, Kenya and Ghana) in both research and teaching (see Kenski and Kenski 1976, 21). This approach disregards the cautiousness of scholars writing more than 40 years ago. For example, in 1976, as the five lusophone African countries were just gaining their independence, Kenski and Kenski (1976, 8) wrote: ‘The complex and dynamic politics of over 40 independent nations may not fit into a single mold or analytic framework that can be called “African”’. Similarly, in 1980, Ayoade (1980, 392) expressed: ‘African politics, in the sense of a unitary political setting and tradition, is a misnomer.’
To be sure, many ‘Africanist’ scholars are careful about generalising their findings from one or two country cases to 55 (the number of African countries). For example, many single-country articles published in African studies journals such as African Affairs, African Studies Review, Africa, Africa Today and Review of African Political Economy still rely on deep and lengthy qualitative research drawing on interviews and archives. These studies place high value on local voices and more intricate narratives rather than grandiose claims about the way African political systems work or how African political elites behave. Meanwhile, there are other scholars who are circumspect in their claims when using cross-national data sets to describe broader patterns (in particular regions) on the continent.
Although we, as academic researchers, are ultimately responsible for what we produce, I would be remiss to not highlight the pressures from journal and book editors and reviewers to generalise beyond our cases and help explain how politics and governance work broadly across the African continent. I have frequently struggled with this tension and have adjusted my own writing towards generalising and broader, more assured contributions than my natural inclinations would permit. For multiple manuscripts, I have received reviewer and editor comments that the arguments and findings are either too restricted to the country cases for the submission to constitute a valuable contribution or that they should be extended beyond the country case or sample of African countries in the paper. Beyond my own experience, I surmise that there is considerable pressure for other researchers to write authoritatively about ‘African politics’ and the contributions of their manuscripts. The pressures to conform to secure academic positions and promotions are powerful for non-African and African researchers alike. For example, there are many African doctoral students being trained in the US and the UK who may be compelled to employ the same narrow set of methodologies that are de rigueur in their departments, which are dominated by non-African faculty and graduate students. This pressure to generalise and to remove subtleties, qualifications and doubt seems detrimental to advancing our understanding of how politics operate in a nuanced and context-specific way.
The nature of the publication system and the generalising tendency of many ‘Africanists’ outside of Africa to speak decisively about ‘African politics’ seem to be second-order problems, which could be remedied by individual and collective efforts to restructure this system (e.g. how we execute our tasks as journal board members, editors, reviewers and researchers). However, there are interrelated, more first-order problems rooted in rigid power asymmetries that buttress the knowledge production regime in political science on Africa, which are connected to the earlier discussion of colonisation, historical misrepresentation and exclusion of African narratives.
Liberian academic and activist Robtel Neajai Pailey (2016) provides her take on what has changed since colonialism:
While the early writings and teachings about Africa are based on colonial expeditions, missionary exploits and anthropological ethnographies, contemporary scholarship is dominated by some non-Africans who have strategically positioned themselves as the authoritative voices in a 21st century scramble for influence, as if Africa were a tabula rasa with no intellectuals or knowledge production of its own.
Pailey (2016) continues: ‘It is clear that those who produce knowledge about something wield considerable power over it’ (also see Akinro 2019).
It is in this regard that the label ‘Africanist’ is particularly problematic. I argue that the self-referential use of the term ‘Africanist’ (and as a label for colleagues and co-authors) is an expression of power. It connotes broad Africa-wide expertise even when it is not necessarily merited. And while one cannot escape hearing this term tossed about at large conferences in the United States, I have never heard an African refer to herself as an ‘Africanist’. Anthropologist James Ferguson (in Schouten 2009, 2) speaks eloquently about the dangers of such labels and proclamations of expertise: ‘There’s a kind of arrogance that comes with expertise sometimes’. But, he adds, ‘one also needs an expertise that is conditioned on a kind of humility if you like’ and an awareness of ‘how little one knows’ (in Schouten 2009, 2, original emphasis). These comments are apropos for labels such as ‘Africanist’.
Who can claim to be an expert on such a diverse and complex continent of 55 independent countries, anyway? It is far rarer to hear academics use labels to authenticate their expertise on all of Europe, all of South America or even all 50 states in the federal republic. But it is done regularly with regards to Africa – why? I have no answers. One conjecture is that the term is useful in that it allows North American and European scholars to continue to project external interpretations of Africa as authoritative and scholarly, even though they are not from Africa and spend relatively little time there. No matter; being part of the ‘Africanist’ club lends external arguments greater credibility, particularly if one is in the top brass. Olukoshi (2006, 542) outlines the problem clearly:
Instead of engaging the local research community, there is a growing culture among Africanists of a massive self-referencing and the cross-referencing of a close-knit network of professional friends that ultimately feeds into the process of manufacturing and reproducing gurus and high priests of African Studies through a mutual reinforcement that eventually impoverishes the field as a terrain of serious knowledge production.
In other words, the dominant knowledge produced by non-African ‘Africanists’ – of highly complex and historically contingent processes – is overly simplified, incomplete and loaded with external concepts and paradigms (not to mention languages) (Ayoade 1980; Zeleza 2002; Mama 2007; Hountondji 2009). The dominance of external concepts, paradigms and languages poses multiple problems. First, many African people may be effectively excluded from the research process and outcome given that they do not understand the terms and languages used (see Mamdani 1993). Second, non-African researchers who conduct research in Africa bring external frames and ways of knowing. Because they rarely speak local languages (with some exceptions in anthropology, geography and other fields), they are thus less likely to engage with and understand local concepts and subtleties that are not easily translated.3 They are also less likely to draw on local language sources (written or oral).4
By contrast, there are troves of unpublished (or not widely available) research produced by Africans in Africa that demonstrate a detailed and sophisticated understanding of local political realities. As Mama (2007, 4) notes: ‘There is in Africa much gray matter to be excavated, developed, and disseminated’, and much of this work remains obscure and outside of the hegemonic regime of knowledge production. This wealth of knowledge is contained in university dissertations and reports commissioned by external donors and research organisations, often from North America or Europe. To be sure, there is enormous variation among the research outputs produced by local university students and among the publications generated by researchers and research institutes with donor support. Many African graduate students struggle to obtain the training, advisory support and financial resources to carry out high-quality research in their own countries. Local researchers and research institutes are heavily dependent on external support and outputs depend on the specific nature of the research partnership. Much of the work commissioned by development donors is strongly influenced by their own orientations and their preordained ‘research’ conclusions (see Zeleza 1996). However, in certain partnerships, local researchers lead teams in designing and implementing rigorous research programmes in collaboration with foreign universities and think tanks.
To give just one example of high-quality research in Mozambique, local scholars and graduate students often conduct deep (ethnographic) fieldwork in their home communities in local languages. Unlike most foreign researchers, they can draw on extensive networks and relationships of interpersonal trust with key informants as well as a rich understanding of historical factors relating to both place and people. Unfortunately, their work does not reach mainstream publication outlets in Africa or outside of Africa. While conducting my doctoral research on Mozambique, for example, I was only able to obtain access to these publications through local university library archives and visits to local research organisations. Most of this research is written in Portuguese (and sometimes French, as some promising Mozambican researchers earn scholarships to pursue doctorates in France) and it is rarely cited by North American and European scholars. Despite the exceptional quality of this work, it almost certainly does not reach institutions of higher education outside of Africa, where students wish to learn about Africa and gain exposure to diverse African perspectives as well as new concepts, theories and epistemologies. Instead, students are often shortchanged; as Ayoade (1980, 397) writes: ‘The use of borrowed concepts has forced the teaching of African politics into the focus, guidelines, and even values of the developed countries.’ What if, by analogy, students were required to study ‘American politics’ (nearly) exclusively from the perspectives and paradigms constructed by Africans (and written in African languages) who are based far from America? It would be equally startling and unedifying. Yet this is the reality for the study of ‘African politics’ in many of the top universities and colleges in North America, as will be shown by the analysis of the syllabus data in the third section of the article.
Having laid out the problems, what actions can we all take on an individual level? A first step is to critically examine our approaches to studying political dynamics in Africa (e.g. our theories, epistemologies, methods and data). This examination should include questions such as who guides our research questions and designs, what methods do we employ to answer our questions, who do we interview and how, which documents do we review, which theoretical frameworks do we rely on, how do we interpret our findings and how do we seriously and credibly engage with issues of external validity to other African countries. This reflexive examination and questioning will require serious engagement with local African academic communities (Olukoshi 2006).
The same applies to teaching; when we select source materials and deliberate over the themes and messages to convey and discuss, we should consider all of the questions above. We should also carefully weigh how we structure and facilitate those discussions so that African voices play a central role in shaping discourses about politics and governance. As Amina Mama (2007, 7) emphasises, these are not only practical questions but also important ethical questions. ‘Perhaps the best we can do’, she writes, ‘is become more conscious of the ways in which our identities, who we are, influence what we do and how we do it, so as to make more informed ethical choices’ (Mama 2007, 7). This process of reflexivity and conscientisation requires us to reflect on the institutions to which we belong and the resources at our disposal, as well as how we are positioned to contribute to knowledge production vis-à-vis other scholarly communities. Her questions bear repeating here (Mama 2007, 7):
What does our research and knowledge contribute to the various contexts and peoples we study? How do our research activities affect those we study? Can we develop the study of Africa so that it is more respectful toward the lives and struggles of African people and to their agendas, studies that contribute to the good of Africa?
This effort could be well integrated into recent advances in experimental and quasi-experimental research methods, as well as other methods including archival research or in-depth case studies and ethnography, which are producing valuable knowledge on governance in different African countries. However, at present, it is largely academics at well-endowed institutions in North America and Europe who have the access to funding,5 training in experimental methods, and networks to carry out this type of research. In the absence of efforts to support African academics based in Africa, the current system as well as publication biases will continue to privilege the mostly non-African ‘Africanists’ outside of Africa. As a result, what we learn from experimental findings will be substantially rooted in North American and European research agendas, theoretical frameworks and interpretations, instead of being driven by, and based on, the expertise and insights of local researchers and other people from those places. This is an enormous missed opportunity to partner with and learn from people who speak local languages and who possess far more intimate knowledge of day-to-day politics and the intersecting socio-cultural landscape. Moreover, these people are effectively shut out of producing more insight that could be used to transform lives and societies in ways that African people see fit (see Ayoade 1980; Zeleza 2002; Olukoshi 2006).
Quantitative data from undergraduate course syllabi and PhD reading lists
Analysis of course syllabi
In this section, I present authorship data from 1260 ‘readings’ (including academic and media articles, reports, book chapters, books, films and TED talks) from 24 course syllabi for courses on African politics or Africa and international relations taught between 2014 and 2019. The first stage of sampling was to search for syllabi available online from the top 50 undergraduate programmes as ranked by US News & World Report (n.d.). Two research assistants were able to locate and hand-code 12 syllabi from 12 universities. To augment the sample, I carried out a second stage of sampling and included other syllabi that I could locate online through a simple Google search. I was able to locate and hand-code another 12. Of the total 24 course syllabi, 23 are drawn from courses taught at US universities and one is from McGill University, Canada. Internet availability might introduce some bias, but it is not clear whether it would be significant or systematic in one direction. Nonetheless, the small sample size implies that the findings should be viewed as preliminary. Moreover, the sample of syllabi (and of PhD reading lists) is restricted to North American universities and thus does not include university courses in Europe, Asia, Latin America, or Africa. I hope the following analysis motivates more extensive research on the topic, which could examine additional materials from university courses outside of North America and test whether the patterns generalise.
Some additional coding notes are in order. I backchecked all the data inputted by the research assistants for errors, mainly with regards to nationality and location. The method of classification for nationality of authors was to search the Internet for information from publicly available CVs, university profiles, personal websites or other media articles (using the most up-to-date information). We classified an author as ‘African’ if we could confirm with a high degree of confidence that a person held nationality of an African country and ‘non-African’ if we could obtain confirmation of a non-African nationality. When we could not confirm nationality with a high degree of confidence, we would classify authors as ‘likely African’ or ‘not likely African’ based on names, time spent in Africa, location of degrees, or fluency in African languages. Location of base was also determined by examining different sources of publicly available information.6
This imperfect process inevitably involved a number of subjective judgement calls in coding and backchecking the data. All of these were made towards biasing the number of African and African-based authors upward. At least this way it is clear in which direction the data are biased: the percentages are over-estimated. For example, Frantz Fanon and Walter Rodney are classified as based in Africa, since they spent many years in Algeria and Tanzania, respectively. An author with British-Nigerian nationality would be classified as African. Even in cases when there was no evidence that the author was African, if we could not confidently ascertain nationality, we would code the author as ‘not likely African’ rather than ‘not African’. For example, an author whose degrees are all from England who teaches in England would still be coded as ‘not likely African’ if we could not determine nationality.
Before turning to the results, it is important to highlight a few other limitations of this paper’s empirical approach. The first limitation relates to the focus on author nationality, which is an imperfect measure of researcher orientations and biases. Many of the authors classified as ‘African’ are in fact dual nationals (e.g. Nigerian-British or Nigerian-American) and many have completed most of their academic training at universities outside of Africa. Some of these researchers are likely to employ the same hegemonic methodologies and frameworks that are prized by the leading academic outlets in North America and Europe. Moreover, even within African research institutes, external funding may shape the research conducted and the types of knowledge that are prioritised. To be clear, a radical transformation of scholarship and teaching on ‘African politics’ must go far beyond increasing the share of materials authored by African and Africa-based scholars on university syllabi (see Ndlovu-Gatsheni [2020] for a thorough discussion of decolonisation). The data on authorship presented below do not sufficiently capture the larger structural problems and historical dependencies, nor do they reflect the various political agendas held by social scientists. Rather, the data provide a contemporary window into how North American undergraduate and graduate training in ‘African politics’ is dominated by scholarship produced by non-Africans based outside of Africa, who are predominantly male.
Tables 1–6 present descriptive statistics that are unadjusted for co-authorship. Table 1 shows that only 15% of readings are written by Africans. We could not classify 16% of the sample by nationality, and 9.3% of these were coded as ‘not likely African’. Thus, the non-African figure is likely much higher than the reported 69%. In parallel, only 9% of the readings are written by authors based in Africa. We were only unable to classify 4% of author locations, leaving a shocking 87% of the readings as coming from outside of Africa. Also striking is the exclusion of women from course readings; only 15% of source materials are authored by women. These core findings illustrate the dramatic under-representation and exclusion of the scholars, researchers and activists who are closest to (and most familiar with) the political battles and actions being carried out across Africa. Table 2 shows that the results barely change when weighting the data based on co-authors.
Descriptive statistics for authors (unadjusted for co-authorship; %).
| Categories | Yes | No | Likely | Not likely | DNK/NA |
|---|---|---|---|---|---|
| Author is African | 15.00 | 68.81 | 1.75 | 9.37 | 5.08 |
| Author is based in Africa | 8.72 | 87.07 | 0.40 | 0.71 | 3.09 |
| Author is female | 15.33 | 80.46 | – | – | 4.21 |
Note: There were 1260 observations of authors. DNK = do not know; NA = not applicable.
Descriptive statistics for full sample (adjusted for co-authorship; %).
| Categories | Yes | No |
|---|---|---|
| Author is African | 15.64 | 67.53 |
| Author is based in Africa | 9.66 | 86.65 |
| Author is female | 14.99 | 80.37 |
Note: There were 994 observations of source materials (e.g. articles, books, films). For focus, the adjusted numbers for likely, not likely, and do not know/not applicable categories are not reported. Those numbers also barely change.
Although the samples are small, in the second stage of analysis, I examine whether authorship changes substantially depending on the nationality or gender of the instructor. Table 3 shows that for the 24 syllabi, three of the instructors were African (12.5%) and nine (37.5%) were female. Perhaps surprisingly, there are no substantial differences if the syllabi are disaggregated by African vs non-African instructors. African instructors do not include substantially more African or Africa-based authors. The only difference is that non-African instructors include slightly more female authors, but the percentages are both extremely low (14% and 16%; see Table 4). These results cut against the expectation that African instructors would be more familiar with African scholarly output and thus more likely to include it in their syllabi. Perhaps it is important to bear in mind that the African instructors in the sample teach at US universities (and have done so for many years) and thus their networks and the literature they engage with in leading academic journals are still slanted against research produced from Africans and Africa-based scholars.
Descriptive statistics for instructors (%).
| Categories | Yes | No | Likely | Not likely | DNK/NA |
|---|---|---|---|---|---|
| Instructor is African | 12.50 | 79.17 | 0 | 8.33 | 0 |
| Instructor is female | 37.5 | 62.5 | – | – | – |
Note: There were 24 instructor syllabus observations. DNK = do not know; NA = not applicable.
Authorship disaggregated by African vs non-African instructors (unadjusted for co-authorship; %).
| Categories | African instructor | Non-African instructor |
|---|---|---|
| Author is African | 14.94 | 14.31 |
| Author is based in Africa | 9.68 | 8.31 |
| Author is female | 14.19 | 16.35 |
Note: There were 1153 observations of authors.
Table 5 shows that disaggregating by the gender of the instructor also does not reveal substantial differences in the rates of inclusion of African or Africa-based scholars. Female instructors include nearly the same percentage of African authors and slightly more authors based in Africa (10% compared to 8%). The only noticeable difference is that female instructors’ syllabi contain nearly twice as many female authors (22% vs 12%).
Authorship disaggregated by gender of instructor (unadjusted for co-authorship; %).
| Categories | Female instructor | Male instructor |
|---|---|---|
| Author is African | 15.49 | 14.73 |
| Author is based in Africa | 10.18 | 7.91 |
| Author is female | 22.17 | 11.51 |
Note: There were 1260 observations of authors.
Finally, it is important to highlight individual exceptions in the small data set that diverge from the average patterns. One clear exception is Professor Pearl T. Robinson’s syllabus. This is not too surprising given that she has been a committed champion of African voices and has written on this very issue of knowledge production for many years (see Robinson 2008). Still, her syllabus is quite exemplary in its diversification: 46% of the authors in her course are African, compared to just 15% in the full sample (see Table 6). Moreover, twice as many authors in her syllabus are based in Africa (at 18% compared to 9%). And, finally, her syllabus has a female authorship rate of 57% compared to only 15% in the entire sample of syllabi. Perhaps all of us can put in more effort to follow her lead in diversifying our syllabi.
Descriptive statistics for Pearl T. Robinson’s syllabus (unadjusted for co-authorship; %).
| Categories | Yes | No | Likely | Not likely | DNK/NA |
|---|---|---|---|---|---|
| Author is African | 46.43 | 21.43 | 0 | 10.71 | 21.43 |
| Author is based in Africa | 17.86 | 71.43 | 0 | 0 | 10.71 |
| Author is female | 57.14 | 32.14 | – | – | 10.71 |
Note: There were 28 observations of authors. DNK = do not know; NA = not applicable.
To conclude, this section provided preliminary evidence that the vast majority of assigned ‘readings’ in undergraduate courses at leading North American universities on ‘African politics’ are produced by non-African academics based in the US and Europe, and primarily men. This stark hegemony (and exclusion of African voices) does not vary by whether the instructor is African or non-African, or by whether the course is taught by a man or a woman. With rare exceptions (see Pearl T. Robinson), the source material we are imparting to our students is grossly biased against Africans and Africa-based researchers. In addition, the small percentage of female authors reflects how women are substantially disadvantaged and under-represented in the US and African academies (see Mkandawire 2005). This exclusion poses an additional barrier to improving scholarship and to understanding the perspectives and aspirations of African women across the continent.
PhD comprehensive exam reading lists
In this section, I turn to graduate training and provide a descriptive analysis of PhD comprehensive exam reading lists in comparative politics. The sampling method for reading lists was as follows. I selected the top 15 universities and colleges in comparative politics as ranked by US News & World Report (2017). I chose the top universities and colleges because these are the institutions that are most likely to produce future academics. I then tried to locate their reading lists on the Internet using key phrase searches as well as browsing each institution’s PhD programme website. For most institutions, I could not locate a reading list online. Yale University (ranked 4th), Columbia University (7th) and the University of Chicago (13th) have online PhD comprehensive exam reading materials. The sample is thus limited to these top-ranked universities. Future work can investigate whether the findings from these three programmes are representative of a broader sample.
Below, I provide some descriptive analysis of the contents of the reading lists. For each list, I identify all source materials involving African countries and whether authors are African or based in Africa. The Internet-based method of classification for nationality and base location of authors was similar to the one employed for the undergraduate syllabi. The key difference in the following analysis is that I did not classify authors using the likely/not likely African or likely/not likely based in Africa. As a result, this preliminary descriptive analysis is more prone to errors such as not correctly identifying Africans or Africa-based scholars.
All reading lists are broken down by topic. For ‘methodology’ sections, it is important to note that there are no materials pertaining to African cases or written by African or Africa-based authors. Methods within comparative politics at the PhD level are dominated by US and European academics writing on their American and European cases. For the following substantive topics, there are again no articles or books related to African cases or written by Africans or Africa-based scholars: ‘Models of politics, collective action & applications’, ‘Bargaining & applications’, ‘Groups & coalitions’ (all at Columbia University). Ditto for ‘Political regimes’ and ‘Political economy’ at Yale University. In other categories, there are materials involving African cases but none or very few of these materials are produced by African or Africa-based scholars. For example, for ‘State building’ at Yale, there are two important treatises that make fairly broad claims about the ‘African state’. Both are written by academics based in the US and the UK: Catherine Boone’s (2003) Political topographies of the African state and Jeffrey Herbst’s (2000) States and power in Africa: comparative lessons in authority and control. In parallel, for ‘Political institutions’, there is one source material pertaining to Africa: Robert Bates’s (1981) book Markets and states in tropical Africa: the political basis of agricultural policies. While not denying the value of these books in helping academics and students to understand states in Africa, I argue the perspective that is put forward is greatly limited when there are no works by African or Africa-based authors included in this list. The only source by an African on the entire reading list at Yale University is a publication co-authored with James Habyarimana, who is a professor at Georgetown University and completed his graduate training in economics at Harvard University (Yale University n.d.).
Professor Habyarimana also appears as a co-author (in the ‘Identity politics’ section) on the list of Columbia University, but the few materials pertaining to Africa are authored by American or European academics (see reading list at Columbia University n.d.). In ‘Identity politics’, there are four materials with seven authors, but he is the lone African. It is also disheartening that there are no women authors. In parallel to the Yale University reading list, Herbst’s (2000) book States and power in Africa is presented as the authoritative source on Africa for two sections: ‘The state, institutions & state strength’ and ‘Non-democratic systems’. On ‘Accountability & citizen politician linkages’, there is only Nicolas Van de Walle’s (2001) book African economies and the politics of permanent crisis, 1979–1999.
The pattern is similar for the University of Chicago’s reading list (see University of Chicago 2018). There is a Comparative Exam Core List (from March 2018) and a list for Comparative Political Economy – Developed Economies, which have no African cases or African authors. In the Comparative Politics Exam Reading List: Comparative Institutions, there is one article related to African cases, which is Daniel Posner’s (2004) article on the Chewas and Tumbukas, appearing across university reading lists, but there are no African authors. For the Order and Violence Comparative Politics Exam List and Political Regimes and Transitions Reading List, there are five books featuring African cases but, to the best of my knowledge, not one African or Africa-based author. (The authors are Wood 2000; Autesserre 2010; Reno 2011; Arjona, Kasfir, and Mampilly 2015; Roessler 2016). Encouragingly, there are at least some women authors for these topics.
To summarise, as with the undergraduate syllabi presented earlier, graduate-level training in comparative politics in a small sample of top US universities is biased heavily against African and Africa-based scholars. The review above points to two main observations. First, for various topics within the graduate comparative politics reading lists, there are zero (or at best a few) books or articles based on African cases or written by Africans or Africa-based scholars. Second, the relatively few books or articles about African cases are predominantly written by non-African academics based in the US or the UK.
The future crop of comparativist scholars is being trained using materials that are overwhelmingly written by North Americans and Europeans, and the minuscule proportion of required readings drawing on African cases is produced almost exclusively by non-Africans. Most strikingly, the varying perspectives and contributions of Africans and Africa-based scholars to epistemology, concept formation, theoretical frameworks, and research methods are nearly entirely absent. In addition to fixing the underlying knowledge production problem, a dramatic shift in undergraduate and graduate training is required to generate a more expansive body of knowledge about politics and governance in all its varieties across the African continent.
Efforts to address the problem, and how to strengthen them
This paper has presented data from undergraduate and graduate training to more clearly demonstrate the consequences of the knowledge production problem with regards to understanding ‘African politics’. In the second section of the paper, I discussed actions that North American and European academics could undertake to correct for some of the inequities in the current research and publication system as well as in the classroom. Furthermore, I agree with Amina Mama and others that there is an ethical imperative to stand in solidarity with African and Africa-based intellectuals to help change the conditions that prevent many scholars from designing and carrying out their own research on the continent (Mama 2007, 4). Rather than just putting out our own ideas, academics at well-resourced universities in North America and Europe have a responsibility to support and participate in efforts to correct the power asymmetries that privilege certain scholars at the expense of others. In what follows, I describe a few institutional efforts to address the ‘African politics’ knowledge regime problem and how these efforts might be strengthened.7 Like Zeleza (2002, 10), I believe that such efforts should aim towards ‘building vibrant and cohesive African research communities whose relations with other communities in the rest of the world are less dependent and mutually productive’.
First, there is the Working Group in African Political Economy (WGAPE), which hosts meetings where they provide research methods training and networking opportunities for participants from African institutions; as well as opportunities to present work, and to serve as discussants. This is a laudable effort and it can perhaps be strengthened by including more Africans in leadership roles and providing scope for the use of different research methodologies. By my count, of the 14 members of the Executive Committee and four organisers, all scholars are based in the US, with very few Africans (CEGA n.d.).
Second, the Evidence in Governance and Politics (EGAP) network focuses on generating experimental research and rigorous evidence on governance, politics and institutions. It is encouraging that there are 50 EGAP researcher members in the Africa region. However, there are, by my count, only four Africans, including Dr Franklin Oduro, who is the Head of Research and Programs and Deputy Director of CDD – Ghana, which is the only organisation in Africa listed as an organisational member. None of the 13 directors listed on the EGAP website are African. Among their supported Metaketa I and II studies, there were seven experiments in Africa (in Benin, Burkina Faso, Uganda (2 projects), DR Congo, Malawi and Nigeria) but to the best of my knowledge there are no Africans on any of the research teams. This would be a fruitful area for the involvement of African academics and more partnerships with African research institutions, who can shape this experimental work.
Third, greater support could be provided to institutions within Africa, such as the Dakar-based Council for the Development of Social Science Research in Africa (CODESRIA), which has several initiatives and is a leading publisher of research by African scholars in and outside the continent. Founded in 1973 with donor funding, CODESRIA has historically served as a fount for enlightening debates and symposia with a wide range of African intellectuals (including Mahmood Mamdani, Thandika Mkandawire and Samir Amin) and the production of cutting-edge research on topics such as the politics of knowledge production, the university in Africa, and academic freedom and social responsibility (Mamdani 2007; Ndlovu-Gatsheni 2020). There are other important African-led publications, such as Feminist Africa, founded by Nigerian feminist scholar Amina Mama, and the Journal of West African History, founded by Nwando Achebe (Pailey 2016).
Fourth, additional assistance could be provided to expand social science training opportunities for African researchers, such as the Partnership for African Social and Governance Research, the Mawazo Institute’s PhD Scholars programme (a Nairobi-based research institute that helps East African women launch careers as academics), and the Center for Effective Global Action (CEGA)’s East Africa Social Science Translation Collaborative (EASST) (Manda and Strohm 2019). These operate on a small scale but have demonstrably positive (multiplier) effects. For example, through his participation in CEGA’s EASST Fellowship, Constantine Manda highlights the acquisition of research skills, networking opportunities and positive impacts on African scholars and their respective countries. Manda has carried this forward by training over 250 students, faculty and government officials across the continent and taking a directorship at the Impact Evaluation Lab at Tanzania’s Economic and Social Research Foundation (ESRF).8
While these are some small, piecemeal efforts aiming to address this problem, what is needed is a large-scale collaborative effort and the infusion of far greater resources (time, energy and financial) to correct for the substantial biases in who generates knowledge about ‘African politics’ and how it is generated. The system of knowledge production must be transformed so that Africans and Africa-based academics are driving research agendas and questions while shaping the theoretical frameworks, methods and data sources we use to answer them (see Hountondji 2009). Such an effort would expand the range of possibilities for learning about Africa, which are all heavily circumscribed as a result of the status quo hegemony of non-African academic institutions and scholars. In doing so, we can all help reconstruct the epistemological underpinnings of such knowledge in a way that is Africa centred and grounded in the dynamic aspirations of Africans across the continent.