The research imperative—that academics do research and in measurable terms (numbers of grants, publications, etc.)—has become dominant in the higher education and research institutions of many high-income countries such as the UK, Australia and the Netherlands, driven by the so-called new public management, intent on getting value from public universities by making them more accountable for their “research productivity” (Ryazanova and Jaskiene, 2022; Shore, 2008; Altbach et al., 2019). This research imperative is also an increasingly prominent aspect in the aspirations of universities in low-income countries (LICs) to establish research-led or research-intensive institutions (Kadikilo et al., 2023; Rambe, 2024; Uwizeye et al., 2022; Abrahams et al., 2010) and to enhance their participation in the global knowledge economy. It is an aspiration evident, too, in the mission statement of Eduardo Mondlane University, the first university of Mozambique, founded in 1962. Eduardo Mondlane University has had research as part of its core mission since 2013 (Ndaipa, 2023, p. 22). As aspiration this is, of course, laudable since knowledge production is at the core of improving the lives of citizens and the economic conditions of all countries including low-income ones. But it comes with the requirement to create the conditions which make the research imperative a realistic and achievable demand. In this article we therefore explore those conditions and the key objective is to analyse what enables and disables the conduct of research from local academics' perspective in the context of two public Mozambican universities located in Maputo, the country's capital.

The starting-point for this article was an email the first author received from one of the co-authors, stating that it had been impossible for her to do any research since she had completed her PhD. This person was a middle-aged woman, well-established as a university lecturer who had, as is not unusual in LICs, completed her PhD when already ensconced in her academic career. The pursuit of a PhD when already embarked on an academic career is common in LICs because it is often only those who are in already established academic positions and hence can get sponsorship or a grant to do a PhD who actually manage to complete such a degree (see Griffin, 2023; Nota, 2022). This is also evident in statistics detailing the percentage of adults with PhDs in high-income vs. low-income countries (see OECD, 2024, Table A1.1, p. 60).

The PhD is, however, only one initial stage of an academic career that involves research. Once completed it is meant to be followed by a steady programme of research, measured through research outputs such as conference papers and publications, the latter preferably in international, peer-reviewed journals. In this article we discuss the possibilities or otherwise of undertaking such work.

A brief excursus on higher education in Mozambique, which has been written about in a number of publications (e.g., Langa, 2017; Nota, 2022). Portuguese colonial rule over Mozambique ended in 1975 after armed struggle by Frelimo, the Front for the Liberation of Mozambique (Frente de libertação de Moçambique). Frelimo became and remains the dominant force (political and otherwise) in Mozambique, with membership creating privileged positions (Mabongo, 2015). The Portuguese created the first university in Mozambique only in 1962 and only for the children of the Portuguese settlers. By 1975 less than two percent of the students enrolled (40 students) were of Mozambican origin (Mario et al., 2003). Following independence Mozambican higher education gradually expanded, initially as a Soviet-style state-directed public sector designed to raise the general level of education among the indigenous population with an emphasis on training teachers, and in recent years also allowing private sector higher education to develop. According to Nota (2022), in 2022 Mozambique had 52 HE institutions (HEIs) of which 19 were public and 33, private. All HEIs struggle with many challenges including brain drain due to the lack of a graduate labor market, low levels of pay, lack of adequate resources and of adequately trained staff. Mozambican HEIs remain highly dependent on external donors (Cloete et al., 2017). The number of academic staff with PhDs remains very low, particularly outside of the capital city, Maputo. The issue of how to decolonize Mozambican HEIs is gradually beginning to be addressed.

Below we begin with a brief literature review concerning the issues related to conducting research in higher education in African countries and in particular in Mozambique. We then discuss our theoretical frameworks and present our methods and materials before discussing our findings, offering conclusions, providing recommendations regarding the conditions that are required to make the conduct of research a sustainable reality in the Mozambican context, and in particular in the two universities on which we report. We also discuss the limitations of this research and suggest possible lines of further research.

There is an extensive literature on African higher education and on research in African universities. This can be divided between texts that have Africa as a continent in focus (e.g., Mushemeza, 2016; Ndlovu-Gatsheni, 2021) and those centring on higher education in a specific African country such as Nigeria, Ghana, Uganda and South Africa (e.g., Amoako et al., 2021; Appiagyei et al., 2022; Dlamini and Dlamini, 2024; Jacob et al., 2021; Jacob and Garba, 2021; Rwothumio et al., 2021; Kerr, 2022). This research—as does work conducted in countries as diverse as Poland (Szromek and Wolniak, 2020), Turkey (Yalcin and Altun Yalcin, 2017), and Vietnam (Phuong et al., 2022) for example—testifies to the myriad challenges that academic staff face when trying to do research and not just in LICs.

These challenges can be grouped together in different ways. According to Nafukho et al. (2019, p. 44), “Individual characteristics (gender, rank, terminal degree, and experience) and institutional characteristics (number of undergraduate students enrolled, percentage of PhD students enrolled, and funding allocated for research function) are significantly associated with faculty research productivity.” Oringo and Muia (2016), discussing productivity constraints among the University of Nairobi's staff, point to the institution's research culture, institutional factors, resource factors and the research environment as key factors impacting on research productivity. North et al. (2011), discussing research productivity at the University of KwaZulu-Natal, focus on the relation between research productivity and demographic factors such as age, race and gender, and academic factors such as having a PhD. They find that older academics are less productive than younger ones, women are more productive than men, and there is almost no difference in research productivity among black, Indian or white people. Having a PhD, however, makes a significant difference regarding research productivity. Amoako et al. (2021, p. 34) found that “weaker foundational knowledge and skills in research methodology, lack of writing experience, deficient information technology skills and lack of mentorship constituted the challenges” college staff in Ghana faced, and that there was a need for “improving and building more infrastructure, specifically, offices for academic staff and technological infrastructure to enhance more research-based activities.” Jacob et al. (2021) identified eight challenges that beset academic staff at Nigerian universities: poor motivation; teaching large classes; poor staff development programs; inadequate infrastructural facilities; strike actions; insecurity; poor computer literacy; and poor research programs. In another article Jacob and Garba (2021, p. 108) highlight “inadequate funding, lack of strategic manpower planning, braindrain, poor motivation, unconducive working environment and corruption” as challenges for Nigerian academics. The result is low research productivity (Abe and Mugobo, 2021), a situation that is diagnosed across African countries, higher education institutions, and disciplines.

No matter where in Africa research on research productivity was conducted, as the texts above make clear, there are both largely shared and more country-specific issues. The strike actions and corruption referenced in the Nigerian context, for instance, were not mentioned in other African contexts, nor were lack of writing skills that came up in relation to the Ghanaian colleges of education staff. This does not mean that they do not exist in other African countries but rather, that the relevant authors did not identify them as prominent challenges.

At the same time inadequate infrastructure and funding concerns were raised by almost all authors. It is, of course, the case that focussing on one set of issues does not imply that other problems do not also exist. But it is both the fact that some different challenges are mentioned in diverse country-specific contexts, and that focussing on a particular context provides finer-gained insights into the particular challenges that make doing research either a possibility or an impossibility in a given institution. This is the epistemic gap which our research addresses and which prompts us to focus on the challenges that academic staff in two public Mozambican universities (identified as U1 and U2 in this text) face. This decision is reinforced by the fact that there is limited research on the difficulties Mozambican academics have to deal with when conducting research although research on Mozambican higher education more generally is growing (Langa, 2006, 2017; Langa and Zavale, 2015; Mendonça, 2014; Miguel et al., 2022; Ndaipa, 2023; Nota, 2022). Our key research question therefore is: what makes it possible or impossible for Mozambican academics to conduct research?

For the purposes of this article we draw on three theoretical frameworks, a recent and two well-established ones: decolonial theory, professionalization theory, and (self-)motivation theory. These theories, as we will show, speak to each other and to our study. Here, we briefly discuss each in turn. Decolonial theory has become increasingly widely used to discuss higher education and research in African countries (Mbembe, 2016; Ndlovu-Gatsheni, 2015, 2017, 2021; Ocholla, 2020; Adams, 2014; Chinn, 2007). One concern of decolonial theory in the context of African universities has been the decolonization of the curriculum, or epistemic decolonization (Ndlovu-Gatsheni, 2017; Ibrahima and Mattaini, 2019). Ndlovu-Gatsheni (2017) discusses this in terms of “epistemic disobedience to Eurocentric thinking” (51), linking it to “cognitive/epistemic justice as an essential prerequisite for social justice” (71), and argues: “Epistemological decolonisation is about opening rather than closing the academy to a plurality of knowledge.” (71) These knowledges involve, pre-eminently, “indigenous knowledge or African knowledge” that is “experiential” Owusu-Ansah and Mji, (2013, p. 1) and bottom-up, thus countering the power differentials that have plagued, and continue to shape, diverse knowledge systems (Afrocentric, Eurocentric) and their place in the global knowledge economy. In line with these arguments, our research is bottom-up, based on Mozambican academics' reported experiences. (Owusu-Ansah and Mji 2013) suggest that African epistemology “as a knowledge system […] is characteristically oral” (p. 2) and, further, that this oral knowledge system is “looked down upon” because “it is people-centered and sometimes not so easily ‘measurable‘” (p. 2). It is hence the opposite of statistics-centered, numbers-driven Eurocentric research that legitimates itself through quantity and abstraction. Again, in line with this, we use a qualitative approach in this study Thambinathan and Kinsella, 2021). Celebrating Afrocentricity in research, (Owusu-Ansah and Mji (2013) argue that it “encourages cultural and social immersion as opposed to scientific distance in research as well as the use of tools and methods indigenous to the people being studied.” (p. 2) As (Zembylas 2023) puts it: “the decolonial project is a form of epistemic and political disobedience,” designed to make visible “suppressed colonized knowledges and experiences” (p. 124). The idea is to “dismantle colonial structures of society and the delegitimization of local knowledge systems and languages.” (Zembylas, 2023, p. 126) Epistemic disobedience or “refusing the (colonial) university… marks both an ethos and a praxis that denies, resists, reframes and redirects the colonial and neoliberal logics of the contemporary university, while investing in new possibilities and lifeworlds that invoke decolonial practices” (Zembylas, 2023, p. 128). It is in this sense that we use a “critical qualitative methodology” (Zembylas, 2023, p. 129) in this article in order to gather information about the experience of academic staff attempting to do research in two Mozambican universities which in their institutional structures maintain professional and epistemic hierarchies that have existed in these institutions since colonial times. We shall elaborate further on this in our “Materials and methods” and in our “Findings and discussion” section.

An early version of professionalization theory, like decolonial theory is now, was concerned with power differentials (e.g., Forsyth and Danisiewicz, 1985) and their associated inequalities, focussing on the power wielded by professional organizations on the one hand, and that by individual professionals on the other. Education was recognized as a professional arena within this research, as was the notion of the socialization into professional beliefs and values, but not so much into professional practices which became prominent in later work on professionalization theory (e.g., Evetts, 2011). Lorenz (2015) describes the notion of the profession thus:

The focus on autonomy here is in stark contrast to the notion of profession and professionalization in many African countries where higher education is imbricated in state policies that direct these institutions' activities so that they respond to the prevailing economic and labor force needs (see Zeleza, 2009), where academic appointments may be related to political affiliation (see Kaizer, 2022), and the impact of international donors and transnational organizations is pertinent and pervasive (Federici and Caffentzis, 2004).

Much of the research and theorizing on professionalization in African countries, and particularly in sub-Saharan Africa, has centered on the professionalization of the public sector (e.g., Kamdjo and Ngamaleu, 2024; Kaizer, 2022; Mathonsi et al., 2022; Maphosa and Mudzielwana, 2014) to which higher education belongs. But the public sector is here often understood as the civil service or state organizations such as the police, with limited reference to higher education. Where the focus is on the professionalization of research it centers on research management, mostly in the South African context (e.g., Kerr, 2022; Williamson et al., 2020, 2024; Mlambo et al., 2022; Mosoge and Pilane, 2014) but not in Mozambique.

Our specific focus in this article is on professionalization in relation to the conduct of research (rather than teaching, for example) as the acquisition of values, beliefs, and, importantly, practices, inculcated during academics' socialization into their professional practice as researchers (Becher and Trowler, 2001), a process that is a significant aspect of completing a PhD and impacts on academics' working life beyond that point.

Academics' socialization into their professional practice as researchers is, of course, linked to their motivation which, according to Lohela-Karlsson et al. (2022, p. 2), involves both intrinsic and extrinsic motivation. Motivation theory describes motivation in a number of different ways (Ryan and Deci, 2000). Daumiller et al. (2020) define motivation as “the overall processes that give rise to faculty members initiating, sustaining, and regulating goal-directed behaviors.” (3) Within these processes different perspectives on motivation may be distinguished. A prominent one of these is self-efficacy, “an individual's belief that they can carry out a task” (Thomas et al., 2009, p. 161). Another is self-determination (Ryan and Deci, 2000), described in terms of competence, autonomy, and relatedness. We agree with Thomas et al. that “motivational attributes do not fully account for scholastic success” (Thomas et al., 2009, p. 161) and, utilizing as well as questioning Deci and Ryan's (1985) notion of self-determination, we also consider “the effects of immediate contextual conditions that either support or thwart the needs for competence and autonomy” (Ryan and Deci, 2000, p. 58). We argue that immediate contextual conditions play a significant role in academics' ability to perform research as work in various African countries shows (Bigirimana and Sibanda, 2016; Matimbwa and Ochumbo, 2019; Mirwoba, 2024). Below we present our methods and materials and our ethical considerations before discussing our findings and suggesting recommendations.

This qualitative research is based on information collected between December 2024 and February 2025 from seven academics working in Mozambican higher education in Maputo. Since diverse academic disciplines have different professionalization processes and practices, it is important to note that all the informants were located in social and life sciences disciplines. They were selected purposively on the basis that they were in full-time academic employment, meaning that conducting research was one of their core tasks. Three participants were in the process of completing their PhDs at the time of writing this article. Certain of their core characteristics relevant to this study are detailed in Table 1. These characteristics will be further commented on in the “Findings and discussion” section.

The small sample size was determined by five factors. The first was that at the time of this study being conducted and during the writing of this article, Mozambique was in a situation of heightened political tension and unrest due to contentious election results (see Tembe and Booty, 2025; Tembe and Rukanga, 2025). This led to informants being hard to reach. Second, we decided to undertake an interpretative phenomenological analysis (Eatough and Smith, 2017, p. 193) which is concerned with “the detailed examination of personal lived experience” as articulated by the informants. Such an approach, according to Polkinghorne (1989), means that “the participation of five (5) to twenty-five (25) individuals who have experienced the phenomenon [is] adequate enough for undertaking the study” [in Sintayehu and Hussien (2021, p. 4)]. The purpose of interpretative phenomenological analysis is not to arrive at a representative sample but to detail the particular experiences specific individuals report so as to gain a rich and in-depth understanding of the phenomenon under study (Brailsford, 2010). We wanted to get a bottom-up perspective of the experiences of academics seeking to undertake research in Mozambican universities, and such a perspective is effectively gained by collecting information from those directly experiencing the phenomenon. The third factor involved in deciding on the sample size was that this sample size is “reflective of the [still] very small number of PhD students relative to population size in Mozambique” (Griffin, 2023, p. 139). Fourth, the decision was also influenced strongly by our desire to decolonize research as already discussed in the theoretical frameworks section above. One of the concerns regarding social sciences and humanities research in African contexts is that because of many African universities' colonial histories, and because of the continuing dominance of western research paradigms in African universities, we see a fetishization of numbers and statistics as the only valid knowledge paradigm for research. As Mbembe (2016, p. 30) puts it:

Such decolonization—which is partly why we opted for qualitative research—goes hand in hand with our fifth reason, namely the issue of how workload models, to the extent that they exist, are arrived at and measured in higher education. In 1999 the UK higher education system introduced TRAC (Transparency Approach to Costing) which required universities to systematically collect data on the “full economic cost” of their activities (see https://www.trac.ac.uk/about/, accessed 18/02/2025). Staff were asked to provide information about their activities in fixed rubrics and as percentage points of their total professional activities. Quite apart from the fact that those rubrics did not cover all the work actually undertaken by staff, the very fact that work time was reported as per centages meant that the only thing staff could ever record was 100% of their work time rather than actual hours spent on given tasks. This involves a gross distortion of what actually goes on in academe, partly because academic staff do not usually keep time by clocking in and clocking out, partly because not all work was even recognized as such in the rubrics, and partly because the per centages bore no relation to actual time spent. Our intention in this article is to get at the actual time spent as reported by our participants in and on their work, and to gain bottom-up insights into what makes it difficult for them to conduct research. This is here best done through qualitative research.

We employed qualitative methods to arrive at our data through a process of participative co-production of knowledge (Redman et al., 2021; Conde and Walter, 2022) in the research design, the data collection and its interpretation. This involved a series of online focus groups lasting about 1 h each with the selected participants in which we discussed the focus of the study, the issues we wished to cover, how to collect the relevant data, and what those data told us. The data were collected through the following steps: participants were asked to identify in written free form how they spent their work time and how many hours they actually spent on different aspects of their work. This resulted in a number of individual and individualized accounts of their workloads. Following a subsequent focus group meeting to discuss these, we then created a grid that included all the work-related headings they had generated in the first round and they re-checked their information and amended it as necessary. The headings (e.g., “Teaching,” “Research,” “Admin,” “Outreach”) were largely standard headings for academic work in higher education. The headings functioned as themes under which the participants then recorded more detailed information such as “5 h seminars per week.” We performed a grid analysis, discussing the grid data in a further focus group meeting and the participants again amended their information. Some participants had, for example, given very precise hours, others less so. Participants also found that in the focus group discussions they remembered work they did that they had simply forgotten about such as sitting on committees, or attending particular meetings. The sheer number of different tasks they undertook could be hard to recall individually but surfaced in collective discussions. It also became clear that there were issues that impacted on the participants' ability to do research that had less to do with the actual hours spent and much more to do with how and by whom their worktime was structured. We discuss this in detail in the “Findings and discussion” section. The data interpretation proceeded through further focus group meetings between which the write-up of the research and its refining were conducted. Here the first author took the lead, with focus group members reading the text as it progressed and providing both individual comments and discussion points in the ongoing focus groups meetings. This intense, iterative process over several months ensured equal participation by those involved in the research and secured the veracity of the research from the participants' perspective.

This research involved human subjects but not the collection or processing of any personal or sensitive data. All participants were fully involved in discussions and decisions about all aspects of the research throughout the process and gave their consent regarding the use of their data as these appear in this article. In conducting the research we followed best research ethics practice in line with the Swedish Ethics Review Authority's recommendations (https://etikprovningsmyndigheten.se/en/) since Mozambique has no national ethics review board for Social Science and Humanities research. One of the universities we report on had established such a board in mid-2024 but information about its composition and requirements was not readily available within the institution. All participants were fully informed about the research and contributed throughout its process, agreeing to the use of the data they had provided in the production of this article. Participation was voluntary and unremunerated. Participants could withdraw from the research at any time without explanation. All data were pseudonymized.

The first thing to note is that of the seven research participants (see Table 1) only one was a professor and the rest were university assistants, the equivalents of lecturers, despite lengthy terms of service. This is indicative of a common phenomenon, namely that progression to professor is reserved for a small number of academic staff (Finkelstein and Altbach, 2014). Secondly, as the data from 2021 of one of the universities show (Table 2), while at the more senior level (professor) it is common for academics to have a PhD, this is relatively uncommon among university lecturers.

However, as one of the informants said, Table 2 does not indicate the total number of academic staff per rank or gender overall so that it is quite possible that at all levels there are staff who do not have a PhD. The likelihood of conducting research is by implication significantly reduced. These issues regarding Table 2 also highlight another problem, also relevant to the conduct of research: official data of interest for this study are not readily publicly available, even to senior staff within the institutions themselves. This suggests that this kind of information is not a priority for the institutions concerned. It has to be seen in a context where Mozambican higher education continues to be primarily devoted to raising the overall educational standards of the country through a focus on teaching rather than research. The focus on raising overall educational standards in the population at large is a function of the Portuguese colonial policy of not allowing indigenous people to participate in education at any advanced level (see Nota, 2022, pp. 30-35; Menete, 2020). Research has hence been secondary in importance at higher education level relative to teaching, and is in consequence also not as highly rewarded fiscally as teaching, meaning that you earn less as a researcher than as a teaching academic. Nonetheless, research is increasingly recognized as an important component of academic work as UEM's core mission shows (Ndaipa, 2023, p. 22).

The universities in which the participants worked employ a 40-hour week model of full-time work. Although we had intended to use a workload model that specifies working hours in greater detail and possibly by academic rank (for comparison see Phuong et al., 2022, p. 2) as a baseline against which to compare our data, it was in fact difficult to obtain such information even for local staff who, in any event, were adamant that there was a major difference between what the university guidelines suggested and what their working realities were. In this they are not unlike academics elsewhere (see e.g., Griffin, 2022). One of the universities had the following table (Table 3), detailing how academics' work was to be distributed by rank and by three major academic work dimensions: teaching, research, other professional activities.

Table 3 suggests relative equality in workload model within the professorial range of assistant to full professor, with assistants having a somewhat different distribution of working hours across the three categories from the professorial ranks. Each academic rank, however, has, rather surprisingly in view of the comments above, double or more hours devoted to research and other professional activities than to teaching. However, this distribution does not map onto the work experiences of our informants as detailed in their self-reported activities schedule (see Table 4). These show, first of all, many more categories than the institutional workload model indicates, in particular supervision as a separate category, outreach, unplanned activities, and outside work. These will be discussed further below. Secondly, Table 4 is not fully standardized in terms of how the informants responded to each category. This is deliberate, to give the reader a sense of how the informants interpreted the headings. Third, the self-reported working hours are clearly more, sometimes very significantly so, than the prescribed 40 h, and within that frame, research is clearly a minor activity despite its prominence in the official workload model. Only one informant's workload looked different (Ricardo) and this was because of a special dispensation to complete his PhD for which international funding was running out.

Table 4 clearly shows that on top of the actual teaching and administrative loads, staff at all academic ranks have numerous additional, officially expected and simultaneously officially unrecognized (meaning: not included in the workload model) as well as unplanned tasks to do. One participant detailed hers as follows:

The first point in this quote refers to the large class sizes which multiply much of the work to be done, for instance around examinations, student support etc. One unusual dimension of this was the requirement that Master's and PhD students publish two or more articles before submitting their theses/dissertations. Such publication had to be led by the supervisor and senior staff; where those staff were highly conscientious this could take many hours per week as in Mariana's case (see Table 4). As is evident from all this, these numerous additional tasks, expected and unrecognized ones, occupy significant amounts of working time that obviously reduces time to do research.

Beyond the list produced by the informants above, other factors cut into research time. Mozambique is a patriarchal, very hierarchical culture in which junior staff and females are expected to do the bidding of senior staff, especially senior male staff. Being at the beck and call of such staff means that junior academics and female academics have very limited control over their work time since, as one of the informants said, they can be called to unplanned meetings (a reasonably regular occurrence) which they are required to attend at any time, and these meetings can occupy many hours. As a result, junior staff and female academics especially lack autonomy in organizing their work time. This lack, also diagnosed in other research (see Lyer et al., 2023), means that setting aside time for research is difficult to manage. Unplanned activities were thus one workload category that impacted significantly on the informants' work time, and on their ability to organize their time so as to be able to conduct research.

The take-up of time by unplanned activities was compounded by additional tasks that went unrecognized in the workload model. Of these supervision and examinations in the form of sitting on defense committees or juries as they are called of PhD students, Master's dissertations and undergraduate projects was one. Such examinations involve, beyond the actual oral examination, the reading and commenting on the project at hand at each of these degree levels, all of which takes significant amounts of time. But beyond that time, the fact that these activities went largely unremunerated was an additional factor in creating resistance and resentment among the informants. This has to be understood in a context where academics draw basic salaries and are then recompensed separately for certain but not all additional tasks, a necessary income supplement when incomes are low as will be discussed further below.

One noticeable dimension of academic work in Mozambique is the high degree of bureaucracy that pervades academe there. Permissions have to be sought in writing for all manner of standard academic activities such as attending conferences, initiating research projects, being absent from campus for reasons of academic activities such as meetings being conducted elsewhere etc. A large constituency of administrators who, rather than serve, arbitrate in relation to academics, draws salaries that are not linked to income generation. These administrators frequently occupy a higher status than academics or researchers and draw larger salaries. Their presence, as within the new public management structures that have governed much of high-income-country academe since the 1990s, is “justified” through the numerous bureaucratic procedures that they preside over and through which they wield power, for instance through not responding to emails, not signing papers in a timely fashion so that processes get delayed etc. To mitigate this situation, academics are endlessly called upon to seek out the relevant administrators in person and attempt to chivvy them along—another time-consuming activity. As Zavale (2023) importantly states in this context: “Public HEIs are not self-governing institutions in matters of leadership. Vice-chancellors and deputy vice-chancellors are not elected by peers but appointed by the central government. Deans, directors, deputy deans, heads of department, and directors of academic programmes are also appointed by vice-chancellors, not elected.” (p. 92) Administrative appointments including in the administrative units of every public university are thus a matter of grace and favor associated with membership of the dominant party, Frelimo. This grace and favor culture translates into a related attitude when it comes to managing academics and university processes where preferment rather than merit is key to understanding how the system works, and staffing may be a function of preferment rather than job-related competence.

Beyond the various conventional activities that make up academic workloads and which may or may not be accounted for in official workloads models, and the socio-cultural factors just outlined, there are also material conditions that impact on Mozambican academics' ability to conduct research. We discuss these now.

A factor largely unaddressed in high-income country research on academics' workload (other than in terms of the importance of attracting research funding) are the material conditions one faces at work. This is because having appropriate facilities is largely taken for granted. One's material work conditions are, however, of crucial importance, including in low-income countries where the research infrastructure remains inadequate, and not simply for institutional reasons. A key dimension of research in the knowledge economy is internet access (Adeniyi et al., 2024). This requires a stable internet, continuous electricity provision, access to computers or other digital devices, and the capacity of institutions to acquire and maintain online resources. All of this is not a given in the Mozambican higher education context. The same goes for laboratory and other material facilities and materials (Mubai, 2021, p. 619), from books and journals, to pipettes and preserving solutions. Lack of material facilities makes conducting research much more difficult here than in high-income countries.

Another important dimension of the material conditions of work is remuneration since everybody needs to earn a living. Academic wages in Mozambique are low by international standards (Nota, 2022, p. 28). As Zavale (2023) puts it: “The basic monthly salary of a full professor does not reach USD 1,000, and if all supplements are included, it does not reach USD 2,000. This is the reason why most academics in Mozambique, as elsewhere in Africa, […] are engaged in other income-generating activities, such as teaching night classes or in private HEIs, consultancy in companies and NGOs, and political activities in government…” (96). Zavale's account highlights Mozambican academics' need to engage in income-generating activities outside of their immediate academic work in order to make ends meet (see also Mubai, 2021, p. 611). And, indeed, this was evident in our informants' accounts—they also engaged in a whole variety of activities outside the public university where they were employed, for economic reasons, from working at private universities to running their own businesses or consultancies. Again, this clearly reduces any time to do research. Or, if they were engaged in research for NGOs or through consultancies, the academics in question typically did not own that research since they conducted it in a service capacity and hence could not use the data they had generated in it for publications of their own.

As Zavale also states: “Research is barely funded by the government; it is occasionally funded by donors.” (2023, pp. 92–93) Lack of research funding resources was mentioned by our informants as an important detriment to research; it went together with a lack of incentives to conduct research, and resulted in low productivity. As Zavale, citing UNESCO and Kahn (2021) indicates:

These figures are, of course, not disaggregated by academic rank or gender but they demonstrate clearly that, and to some extent why, academic research productivity remains low in Mozambique. This is also evident in the numbers of publications of our informants (see Table 1). Those with the longest time in academe had the most publications but at six publications within 10 years of PhD completion, even the most productive person had less than one publication per year.

How are we to interpret these findings? In terms of decolonial theory Mozambican higher education presents itself as still deeply entrenched in colonial epistemic heritages, dominated by unproductive professional hierarchies, colonial epistemic traditions, high degrees of stifling bureaucracy, and socio-cultural norms that undermine academics' ability to conduct research. Whilst the academics who manage to complete a PhD, often abroad, are inducted through the PhD process into the professional practices associated with their discipline/s, their subsequent work environment is not conducive to them pursuing research, even though they are clearly equipped to carry out this task in terms of the competences they acquired by doing a PhD. Their research professionalization and practices are thus undermined by the conditions they have to work in, despite the fact that their self-efficacy, in terms of conducting research in principle, is not in question. But there is also no structured or obligatory continuing professional development in the form of workshops, training courses, and mentorships by senior staff to help those with a PhD develop their research careers, something our informants asked for in the discussions we had. Additionally, their “immediate contextual conditions” (Ryan and Deci, 2000, p. 58), not least the socio-cultural ones, tend to thwart them in their research efforts, with their numerous demands, many of which are unaccounted for in their workload model and not sufficiently remunerated to enable the academics to work solely within their employing academic institution despite being in full-time employment there. These immediate contextual conditions also and importantly include the material research environment (computer access, library provision, laboratory facilities, etc.) which remains under-resourced. Given these conditions and the remuneration structures which favor administrative and teaching posts, individual academics have to be highly self-motivated to conduct research. One main motivation for conducting research against these odds is that, as Nota (2022) indicates, the promotion of academics to professorial ranks demands—at least theoretically, i.e., when based on merit—a research profile. The lack of such a profile accounts for the fact that all our participants had already had long careers in academe without, with the exception of the longest-serving member, progressing significantly up the professorial ranks. As the numbers of staff with PhDs in Mozambique is growing, so the encouragement of emerging academics to pursue such degrees locally and in larger numbers should be encouraged to increase the overall cohorts of PhD holders and from an earlier stage in their careers.

Our recommendations address three levels: the micro one (individual academics), the meso one (the institutional level) and the macro one (the governmental level). At the micro level, there is the question of how individual academics can make it possible for themselves to do research under highly constraining circumstances. Three concrete recommendations are (a) to collaborate with others and thus share the work. These others could be local or come from farther afield. (b) Conduct research on issues that are immediately accessible—a recommendation more easily done in Social Sciences and Humanities subjects perhaps, than in other disciplines—although the problem of access to the internet and search engines is important in all disciplines. The research one might conduct could center on how to decolonize the university, curricula, research itself, but could also center on particular so-called substantive issues, or on methodological concerns. (c) Create specific (co-)writing spaces—they need not be very long (e.g., 2 h) but it would be good to make them a regular weekly or bi-weekly occurrence. With specific goals for each writing session, and in the presence of others, this can be highly productive (Kornhaber et al., 2016). It helps to create what Wiebe et al. (2023) describe as a community of practice, a mutually supportive research environment where it becomes easier to conduct and publish research. This would clearly support the professional practices academics are meant to engage in.

At the meso level, it is imperative for higher education institutions—in their own interests as research institutions, and in the interest of the scientific, economic and cultural development of their country—to support their researchers in conducting their work by reducing barriers to research. This includes serious pushes to improve access to the internet and to search engines, sustained efforts to reduce bureaucratic procedures, a re-thinking of pedagogical practices (for instance reducing the number and length of oral examinations), a reduction in meeting times and in unplanned events, planned actual research time as part of concrete realistic workload models (for instance whole-institution writing retreats of 1 week each twice per year), and at least symbolic recognition of research efforts. Universities should also actively engage in reducing unproductive sociocultural hierarchies within their institution by, for example, rewarding women and junior staff who conduct research appropriately, and through structured career development plans, research training workshops, and mentorships that facilitate the transition from PhD student to newly fledged researcher and beyond. In addition, and importantly, there need to be efforts and incentives to encourage administrators at all levels to see themselves as supporting, and not arbitrating about, researchers, with concrete evidence of such support being symbolically and if possible also otherwise (e.g., through promotion) being rewarded.

At the macro, governmental level, and in the interests of the country's economic, scientific and sociocultural development, more needs to be done to ensure stable and extensive internet access and energy provision. Researchers should also be more prominently recognized, symbolically as well as through incentives such as structured, transparent fiscal remuneration, research time as a reward for research effort, and dialogues about the research needs of the country.

There are, of course, many more recommendations one could make but we are also very clear that the economic conditions of Mozambique and the educational attainment of the population at large still require much effort. Mozambique's spend on research as a percentage of GDP was 0.31 in 2015 (latest figure) compared with Sweden where it was 3.41 in 2022 (latest figure).¹ This in turn impacts on what can be done for and in higher education institutions in the country but should not deter institutions from working toward resolving the issues that have been highlighted in this article.

We focus on two limitations here before suggesting possible lines of future research. The most obvious limitation of this study may well be the rather small number of informants. The reasons for this are explained in the methodology. A second limitation of the study is that it centers on the capital of Mozambique, and the institutions there are better resourced than those elsewhere in the country, in particular those in rural areas. However, we also know from existing research (e.g., Nota, 2022) that the problems which staff encounter in Maputo in relation to conducting research are not dissimilar to the ones academic staff in HEIs elsewhere in Mozambique face.

One future line of research could thus be to research the issue of the problems with conducting research across either a selected range of HEIs across Mozambique e.g., from different regions, of different types (private, public), and/or a range of disciplines to establish whether or not academics in particular disciplines experience greater problems than others. This could also be further refined by correlating these data with demographic factors such as age, gender, political affiliation, marital status etc. One concern that might hamper such research is the issue of maintaining appropriate confidentiality in relation to the informants. Another line of research might examine the relationship between academics' research in HEIs relative to research they conduct for NGOs, their own consultancies etc. One might also research what factors facilitate the conduct of research in relation to those who are most successful at doing that work. And, finally, it would be useful to undertake participative action research in HEIs to see how they try to improve their research profile. As this list suggests, there is much to be done.