This research review fills three main gaps identified within the literature. Firstly, meta-literature reviews on the broad economics of education-which discuss global impacts of investment on growth-largely predate 2020 and lack a primary focus on SSA (Benos & Zotou, 2014; Cui & Martins, 2021; Awaworyi Churchill et al., 2017). Secondly, this review highlights the need to synthesize nascent and specific evidence from within SSA across 2020-2025, which focuses on the moderating impact of governance, the economic effects of brain drain, and the impact of the financing of tertiary education and digitalization-all of which remain largely fragmented in current regional literature.

Thirdly, none of the extant reviews examine the specific methodological quality of recent research in this area, necessitating a rigorous review that assesses the current state of policy-ready evidence through PRISMA systematic literature review techniques.

Accordingly, this systematic review is conducted based on eighteen selected articles relevant to these factors, guided by four distinct research objectives.

To avoid the duplication and overlap that characterize broad descriptive reviews, the review was organized around one focused question and four non-overlapping objectives:

Focused review question: What does empirical evidence published between2020 and2025reveal about the effect of education and human capital on economic growth in Sub-Saharan African countries, and under what conditions is this effect strengthened or weakened?

The specific objectives were:

1) To synthesize the overall direction, significance, and magnitude of the effect of education and human capital on economic growth in SSA countries.

2) To assess how governance quality and institutional conditions moderate the education-growth relationship.

3) To identify complementary and attenuating factors, including public education expenditure, tax structure, remittances, brain drain and human capital flight, higher education financing, and digitalization, that condition the effectiveness of human capital investment.

4) To draw country-specific and policy-relevant lessons from single-country evidence (Ethiopia, Rwanda, and South Africa) and to evaluate the methodological rigor of the recent evidence base.

The origins of contemporary interest in the economics of education lie largely in the work of Becker (1964), who conceptualized education as a type of capital investment with returns akin to those of physical capital, flowing over the life cycle. Lucas (1988) explicitly formalized the role of human capital in endogenous growth by suggesting that the growth of productivity could be fueled by the accumulation of skills. Similar emphasis was echoed by Romer (1990), who argued that fundamental growth, in the long run, is underpinned by ideas and knowledge creation produced by educated workers.

With the evolution of these theoretical underpinnings, Mankiw, Romer, and Weil (1992) extended the framework to cross-country regressions and argued that an augmented Solow growth model incorporating human capital as part of the production function predicts growth much more effectively.

Over decades of cross-country and multi-regional study, the empirical evidence has, by and large, supported this positive social return to education; however, it has also highlighted heterogeneity across regions, time periods, and estimation models (Benos & Zotou, 2014; Cui & Martins, 2021; Ogundari & Awokuse, 2018; Zhou et al., 2022). Consequently, this theoretical and empirical literature base sets an initial baseline expectation in this review: that it is highly probable to find empirical evidence that supports the idea that the increase in stocks and flows of education to growth could be contributing to increased production in Sub-Saharan Africa.

A second thematic avenue of investigation focuses on whether public education finance, per se, stimulates growth efficiently. In the empirical growth literature, the debate is equally lively and nuanced. While meta-studies indicate positive average impacts of public education spending on growth (Awaworyi Churchill, Ugur, & Yew, 2017), the cross-country variation in magnitude and significance remains high; similarly, cross-country analyses argue that the impacts are much less evident in low-income economies compared to high-income economies (Idrees & Siddiqi, 2013; Appiah, 2017).

The crux of the argument points towards the “efficiency” with which the public investment is utilized: a dividend from education on growth is dependent on how, where, and when that spending occurs (Agasisti, 2014; Dissou, Didic, & Yakautsava, 2016), as well as on what other areas complementary to the investment occur, such as in Research and Development (Diakodimitriou, Tsioutsios, & Papageorgiou, 2025; Bambi & Pea-Assounga, 2024; Çetin, 2023; Tomić, 2015).

Moreover, distribution matters, as education spending, as an engine of growth, interacts with considerations of inequality (Sylwester, 2000; Artige & Cavenaile, 2023). History and regional context matter for the analysis of outcomes, from the USA context (Carpentier, 2006; Baldwin & Borrelli, 2008; Murova & Khan, 2017; Bania, Gray, & Stone, 2007) and Europe’s welfarist education context (Beraldo, Montolio, & Turati, 2009), to general-equilibrium examinations applied to countries like Tanzania and Zambia (Jung & Thorbecke, 2003). All these studies tend to indicate that spending level itself cannot be an optimal growth indicator; in the context of our investigation into these issues, this efficiency debate shapes the direction of the study through the attainment of Objective 3 of this study.

Increasingly, the literature argues that institutions are not just an alternative channel for growth, but a modulating variable, in that institutional efficacy matters in conditioning the growth from educational access. This is especially due to the fact that dysfunctional or poor governance through variables like corruption, fiscal misspending, and policy ineffectiveness is hypothesized to stifle the schooling-to-growth transformation, whereas good institutional quality can accelerate this relationship (Odhiambo, 2024; Abu Alfoul et al., 2024). Placed into structural growth comparisons, this has the ability to provide explanations-both through good institutions and by showing that good governance in tandem with digital transformation can actually contribute to some of the observed wealth inequality between Sub-Saharan Africa and other parts of the world, like the Middle East (Correa & Esquivias, 2025).

Relevant to our focus, research such as that by Karambakuwa, Ncwadi, and Phiri (2020), which examined what influences human capital impacts on growth within SSA, emphasized that factors which act as complements, for example, institutions and macro-environment stability, are imperative and vital components of growth. These factors are important given our interest in Objective 2 of this review.

As pertains to the specificity of Sub-Saharan Africa, the literature also brings in an issue regarding human capital leakage, where a significant portion of educated personnel exit from the region. According to studies such as Usman, Ozdeser, Cavusoglu, and Aliyu (2022), remittances can contribute to sustainable economic growth in this region, but overall human capital loss, essentially brain drain, on the other hand, leads to largely offsetting or cancelling the gain from that human capital gain. Furthermore, evidence by Temilade Akinola et al. (2024) indicates a negative growth influence, as the net outflow leads towards brain drain and related output decline in growth. Therefore, where public systems form the basis of the supply side, the eventual loss is really the export of the state return from investments, a fact not usually captured under the standard Human Capital Theory (Becker, 1964).

There is still another subset of scholarly studies that examines what comprises the quality of this human capital in the SSA context. Research (Amin & Ntembe, 2021) outlines how a lack of sufficient funding in higher education institutions hinders growth from its potential role in helping foster productive and modern workforce development across various sectors in SSA countries. A review of single-country empirical evidence in the sub-Saharan African context provides a different picture. Studies specifically concerned with the linkages between economic growth and educational outcomes, like Rwigema (2020) and Gu (2025), both provide some evidence of long-run co-integration of education expansion and growth in GDP for Rwanda, while studies like Cyiza et al. (2025) look at attainment alongside demography and fertility as a dependent variable, as opposed to macroeconomic outcome variables which require explicit output measures for growth.

A systematic search for primary research published from January 2020 to August 2025 was conducted. The relevant literature was identified using databases, namely: Scopus, Web of Science, Google Scholar, and the African Journals Online (AJOL) platform. The searches were carried out in real time, and relevant reference lists of the identified studies were also hand-searched to ensure inclusivity.

To facilitate the search, the Boolean operators (AND, OR) were used to join three specific conceptual categories: i) population (“Sub-Saharan Africa” OR “Africa” OR country names like “Ethiopia”, “Rwanda”, “South Africa”); ii) exposure (“human capital” OR “education” OR “educational attainment” OR “schooling” OR “education expenditure”); and iii) outcome (“economic growth” OR “GDP growth” OR “inclusive growth” OR “economic development”).

Truncation and wildcards were used appropriately as per database function. Only peer-reviewed academic journal articles published in the English language were included in the search. A total of 312 records were retrieved. After 41 duplicates were removed, 271 unique articles remained for screening.

List of acronyms used in this review: ARDL = autoregressive distributed lag, ECM = error correction model, GMM = generalized method of moments, AMG = augmented mean group; AJOL = African Journals Online.

To make an assessment on eligibility, the Population, Interest, Context, Study design (PICoS) was applied. Eligible studies had to meet the following criteria: a) published during the review period 2020-2025; b) the population of interest to include Sub-Saharan African countries, including studies of specific individual SSA countries; c) the study should examine education or human capital formation or education expenditure as an explanatory variable in the growth models; d) the study reported the growth outcome in terms of GDP growth or GDP per capita or inclusive growth or economic sustainability; and e) quantitative study with empirical model (including panel estimation models and time-series method like ARDL/ECM or robust case study econometrics). Studies were excluded if: the studies were published before 2020; the study focus is on non-African countries or non-SSA countries without specifically highlighting results for an SSA country; the study does not cover the outcome (growth); the study relies on a data sample used already in the analysis in this current review without any variation in specification/model to contribute newer findings; and conference abstract/opinion papers/studies evaluated as having weak methodology in full-text review were excluded.

Two authors (E.N.B. and J.N.) Independently screened studies, including duplicate screening, blinded via the software Rayyan QCRI. A first round of screening of titles/abstracts of the 271 non-duplicated items was performed by applying the pre-specified inclusion/exclusion criteria. The inter-rater agreement measured by Cohen’s kappa coefficient (κ), before consensus, was κ = 0.84 (95% CI: 0.78 - 0.90), showing strong agreement. Both authors independently assessed the full-text reports of the 52 eligible studies for eligibility. In case of disagreement about the eligibility of an article at the title/abstract level or at the full-text level, the conflict was adjudicated by one author (L.N.), who scrutinized the article independently and made the final decision.

For every study that met the inclusion criteria, we extracted information and summarized it in a table consisting of the following headings: authors/year, title of the study, context/time, econometric method, variables (both in terms of human capital/education measurement and output variables), and results/recommendations. A critical appraisal checklist was modified to evaluate whether the methodology adopted was sufficiently robust or if bias was likely (Higgins & Green, 2011). Our tool covers 8 criteria, including clearly stated research questions, appropriate sampling methods/representativeness of data, use of relevant econometric techniques with adequate diagnostic testing, robustness and sensitivity checks, corrections for endogeneity bias (e.g., via IV or GMM models), data availability, consistency between findings and conclusion, and the identification of study limitations.

For every study that met the inclusion criteria, we extracted study details into an extraction matrix (Table 1). Quality was scored, where studies with scores 7 - 8 were high, 5 - 6 were moderate, and 4 were rated as low (the 4 low-quality papers identified at the appraisal stage of full-text review were rejected). Table 2 provides the quality profile for the 18 studies.

The selected 18 studies for the analysis were published between 2020 and 2025 (Figure 2), with the year 2024 seeing the most publications (six studies). This growth in the literature is indicative of a growing academic focus on human capital and economic growth. Geographically (Figure 3), 12 out of the selected studies used multi-country SSA panels. In contrast, 3 studies used panel studies in specific countries (2 each for Ethiopia and Rwanda, and 1 for South Africa).

One study also compared SSA with the Middle East. Methodologically (Figure 4), eight studies used a similar empirical regression but with additional/extension or interaction variable analyses in their approach to the human capital-economic growth nexus. Six studies used various panel econometric tools, such as dynamic GMM, augmented mean group, and static panel estimation methods. Three studies employed time-series ARDL/ECM econometric approaches. Only one study was written in the form of a country case study. Most of the sample periods covered data from the mid-80s or 2000 up to 2018-2020, implying that 2020-2025 was the period within which analysis of the latest data was reported in journals.

Figure 3 illustrates the geographic distribution of the 18 included studies. The majority (12 studies) focus on Sub-Saharan Africa as a multi-country region, followed by Rwanda and Ethiopia with two studies each, while South Africa and a comparative SSA-Middle East analysis account for one study each.

Figure 4 presents the types of methodologies employed by the 18 selected studies. As the graph shows, empirical regression with interaction/extension was the most common type used, applied by 8 studies, or 44% of the total studies. Panel econometrics models, i.e., GMM, AMG, and static panel, were used by 6 studies, or 33% of the total studies. Additionally, the ARDL/ECM time-series approaches were applied by 3 studies (17%), whereas the case study approach was used in 1 study (6%).

Table 1 presents the characteristics of the 18 studies in terms of authors, year, title, methods used, and geographical coverage. The majority of the studies make use of panel data estimation methods such as dynamic GMM, ARDL bounds testing, and the augmented mean group estimator. Further, the majority of the countries of focus in the studies are from the Sub-Saharan African region, with case studies concentrating on Rwanda, Ethiopia, and South Africa.

Table 1. Extraction matrix of included empirical studies (n = 18).

Note: Table content reproduced from the supplied extraction matrix; bibliographic details and empirical findings have not been independently verified.

Table 2 summarizes the critical appraisal of the 18 studies included. Each study was appraised against seven methodological criteria: Research Question, Data and Sample, Specification and Diagnostics, Control of Endogeneity, Robustness Checks, and Results consistent with Conclusion. The total score ranges from 5 to 8 (Table 2). Accordingly, studies are likely to be of “high” (score of 7 - 8) or “moderate” (score of 5 - 6) quality.

Table 2. Critical appraisal of the included studies (n = 18).

Note: Scoring Key: 2 = fully met, 1 = partially met. Quality thresholds: Total 7 - 8 = High Quality; 5 - 6 = Moderate Quality.

Across cross-country literature, findings broadly confirm that human capital has a statistically positive impact on economic growth in SSA. The multi-country panel evidence is overwhelmingly supportive: Using the dynamic Generalized Method of Moments (GMM) alongside static panel estimations for 36 SSA countries between 1986 and 2018, Akinlo and Oyeleke (2020) conclude that the accumulation of human capital could lead to an increase in economic growth, and their dynamic model further provides evidence of persistence in growth dynamics. The results in the studies by Bashir Elsharif et al. (2024) and Bekele et al. (2024) arrive at similar outcomes using panel estimation, with the former deploying standard panel analysis whilst the latter employing the Augmented Mean Group estimator, arguing this through the lens of economic sustainability, which includes growth.

Further, using Autoregressive Distributed Lag (ARDL) bounds testing, as well as an error correction model, the study by Matashu and Skhephe (2022) indicates long-run co-integration of human capital and economic growth with a significant adjustment process in the short run to the long-run equilibrium. Importantly, through examining inclusive growth in Africa, the study by Owopetu, Oyedele, and Tella (2024) indicates that human capital, particularly through education, has been a key contributor to inclusive growth within the region. Time series and case studies focused on one country are broadly consistent with the cross-country panel estimates in this set of studies. In Ethiopia, both Dirir (2023) and Ferede et al. (2023) find evidence of a positive long-run co-integrating relationship between public education spending and real GDP per capita, based on ARDL bounds tests. In South Africa, Odhiambo (2021) shows a likely positive long-run link between the level of schooling and output, though its size is somewhat model-dependent. In Rwanda, Rwigema (2020) and Gu (2025) find positive structural long-run linkages between the education stock and economic development. Since these observational and time series studies estimate long-run cointegrating relations rather than using exogenous sources of variation as in quasi-experimental designs, caution must be exercised in terms of establishing strong macroeconomic causality.

However, this study might arguably be considered supporting evidence in the face of our findings, though certainly aligned. Out of 18 included papers, 16 primary empirical studies investigated the direct effect of human capital/education on economic growth (IDs 1, 2, 3, 4, 5, 7, 8, 9, 10, 11, 13, 14, 15, 16, 17, 18). Results demonstrated that there was either a significant positive direct coefficient or a significant positive interaction with growth, with the exception of the other 2 papers (IDs 6 and 12) focusing on human capital flight/brain drain as dampening factors for human capital. Further, for consistency and clear mapping in Table 3, all 16 papers focused on the primary direct impact of human capital on growth under Objective 1, sub-mapping moderation (IDs 5, 8, 18) under Objective 2, and conditioning/dampening factors (IDs 6, 7, 12, 15, 17, 18) under Objective 3.

The second objective is focused on whether the impact of education/human capital is moderated by the quality of governance within countries in SSA. Two studies specifically address this, and both confirm the proposition that institutions matter for the realization of education’s contribution to growth. Odhiambo (2024) argues that education may lead to growth only in countries with high levels of governance and quality institutions, given the results showing that education significantly and negatively affects growth in instances with poor governance levels.

There is, however, evidence from studies such as the one by Abu Alfoul, Bazhair, Khatatbeh, and Ariana (2024), who found consistent results with institutional quality as a moderator variable in a human capital growth model that assumes institutions behave as a complementary input to educational investments.

Similar results to this perspective were reported by Correa and Esquivias (2025) regarding the growth differentials in SSA relative to the Middle East, attributing these gaps specifically to the differences in the level of digitization and institutional quality as drivers for translating the existing education stock into realized growth. In total, based on these three pieces of evidence, it can be concluded that institutions do act as a moderator.

Regarding the third objective on factors that affect the growth-enhancing impact of education/human capital, four such factors arise in the empirical literature:

1) Brain drain/human capital flight: Usman et al. (2022) examine sustainable growth in SSA and indicate that remittances provide support for inclusive growth, but human capital flight and brain drain impact growth significantly, as these losses are not easily recovered, while Temilade Akinola, Awolaja, and Onakoya (2024) find an attenuating direct negative effect of human capital flight on growth in Nigeria.

2) Fiscal structure/revenue policy: Oyinlola and Adedeji (2022) find that the nexus between human capital conditions and tax structure matters for inclusive growth.

3) Efficiency of expenditure: Ferede, Mersha, and Teshager (2023) reinforce findings for Ethiopia that growth is conditioned by expenditure on education and that spending on its own (as noted in section 2.2) is not sufficient, hence supporting the “quality” versus quantity debate in expenditure.

4) Technology: The study by Amin and Ntembe (2021) indicates that financing education in terms of Higher Education can also constrain growth if not utilized efficiently and when funding mechanisms are poor, and also that when considering digitalization as a tool (Correa & Esquivias, 2025), it can complement growth where education levels exist.

From the country-specific lessons, ARDL bounds evidence from Ethiopia (Dirir, 2023; Ferede et al., 2023) points toward a predictable, long-run budgeting approach to education funding to leverage the benefits of co-integration with economic output. For South Africa (Odhiambo, 2021), the lesson is that while positive effects exist, there might be diminishing marginal returns of additional years of schooling as countries advance. The studies from Rwanda (Rwigema, 2020; Cyiza et al., 2025) demonstrate that schooling impacts growth in combination with wider economic structural and demographic changes.

On methodology, although the literature is large and expanding, three of the many studies (Akinlo & Oyeleke, 2020; Usman et al., 2022; Correa & Esquivias, 2025) used state-of-the-art estimators explicitly accounting for endogeneity and cross-sectional dependency in panels.

While a heavy bias exists in favor of interaction-based institutional papers (to address Objective 2), there is a shortage of quasi-experimental or truly instrumented research (to establish deeper, more exogenous causal mechanisms). Table 3 provides the summary of findings mapped to the four objectives.

Table 3. Synthesis of reviewed studies mapped to research objectives.

Note: Study IDs correspond to the numbered studies in Table 1. Studies 6 and 12 focus specifically on human capital flight/brain drain as attenuating structural drains on the human capital stock.