View Chapter:
Downloads:
Chapter 2
The Role of EdTech
Technology-led interventions have proliferated in education systems globally. In both rich and poorer economies, revolutions in computing and the diffusion of mobile phones have led to the rapid development and deployment of programmes that could support student learning in schools or at home. The COVID-19 pandemic, which forced education systems globally to invest in methods for remote instruction during school closures, reinforced the focus on technology-aided interventions. Estimates of the current size of global spending on Education Technology (EdTech) vary widely but are unanimous in forecasting rapid increases in the next decade.[1]
The sharp rise in global interest in EdTech has also coincided with a growing body of research evidence in this area, summarised in three successive reviews by Bulman and Fairlie (2016), Escueta et al. (2020) and Rodriguez-Segura (2022). Our attempt in this VoxDevLit is to (i) update the existing evidence base with studies that have appeared over the past four years; (ii) broaden the categories considered to include a wider range of technology-based interventions relevant to education systems in developing countries, and (iii) take a forward-looking view of the types of evidence that we most need in this space given the current state of research.
We focus exclusively on technology-enabled interventions focused on the education of children enrolled in pre-primary (kindergarten) through secondary school in low- and middle-income countries (LMICs). This excludes, for instance, interventions targeting college students or potential jobseekers, but includes home-based interventions involving parents as well as those seeking to improve governance and school functioning.
What do we mean by EdTech?
In this review, we take a broad definition of education technology to include all interventions that rely fundamentally on information and communication technology (ICT) tools, including computers or tablets, mobile phones (smart or otherwise), radio, and television. These interventions may target students, teachers, or parents but, to be within the remit of this discussion, they must either directly affect student learning outcomes or serve as essential enablers of interventions that do so.
Conceived this way, EdTech is best understood as a delivery system for educational inputs rather than as a direct input into student learning.[2] Whether a given EdTech intervention improves student achievement is determined by (i) whether it enables the delivery of additional inputs (e.g. more specialised instruction); (ii) the productivity of these inputs; (iii) whether it raises the productivity of existing inputs (e.g. teacher effort); and (iv) the extent to which it displaces inputs and the productivity of the inputs it displaces (such as regular class time).
This framework suggests three important organisational principles that will guide our review. First, given the diversity of interventions under the umbrella of education technology, it is not meaningful to think in terms of an “average effect” across EdTech interventions. Second, even the same EdTech intervention (e.g. computer-aided instruction) is likely to differ in its effects depending on how it is implemented and what it displaces. Third, the potential for EdTech interventions to improve productivity will differ across education systems, depending on substantial differences in the productivity of existing inputs and in the binding constraints to learning.
What are key constraints to learning in LMICs?
The central challenge for education systems in LMICs is that, although school enrolment and grade progression have both increased rapidly, student achievement often remains very low — a situation often dubbed as a “global learning crisis” by international organisations (World Bank 2017). A vast literature now examines interventions aimed at addressing this learning crisis (see, e.g. Glewwe and Muralidharan 2016, Angrist et al. 2024 for reviews). EdTech, as a class of interventions in LMICs, seeks to address the same underlying learning constraints and factors that limit student achievement in these settings. While many of the EdTech’s advantages and trade-offs generalise across contexts, some challenges are particularly acute in LMICs. It is useful to review these below to clarify EdTech’s potential to improve student outcomes in these settings.
Curriculum mismatch and within-class differences in ability
A core challenge in many LMIC education systems is the mismatch between official curricula and the actual academic preparation of most students (Banerjee et al. 2007, Pritchett and Beatty 2015, Muralidharan et al. 2019). Further, even within the same classroom, the range of achievement can span several grade levels of actual ability (Muralidharan et al. 2019). A key advantage of technology-enabled instruction is that it can provide personalised instruction targeted to students’ actual levels of achievement rather than the expectations of the official curriculum.
Low level of teacher skills
Teachers in many LMIC education systems often have a weak mastery of the content they are expected to teach (Bold et al. 2017). This constraint is particularly severe in some countries; however, even where it is not a system-wide problem, shortages of skilled teachers can be acute in certain areas. For instance, qualified teachers in STEM subjects or English language instruction, fields that command high labour-market returns in many settings (Azam et al. 2013), may be particularly scarce, even in contexts where most teachers are otherwise well-qualified and experienced. This challenge is also more common in remote and rural areas or in post-primary education levels, where remedial instruction is unlikely to be provided by community volunteers. Further, even within the same schools, there is substantial variation in teacher quality which is not well-predicted by commonly-observed teacher characteristics (Bau and Das 2020).
Inadequate parental ability to support education
The universalisation of primary (and increasingly post-primary) education is a relatively recent phenomenon in most low- and lower-middle income countries. Consequently, a substantial share of parents in many LMICs have themselves received limited education. This limits their ability to directly support children’s curricular development, even when they are highly motivated to do so. Since parental education tends to be highly correlated with other dimensions of socioeconomic disadvantage, this constraint particularly affects students from poorer households within any given setting.
Weak governance in schools
Schools in developing countries are, on average, much less effectively managed than those in developed economies (Bloom et al. 2015, Lemos et al. 2024). Teacher absence rates are high in many contexts, with roughly a quarter of teachers absent during surprise school inspections in India and Uganda (Chaudhury et al. 2006), and as many as 45% in Mozambique (Bold et al. 2017). Even when present in the school, teachers are frequently absent from classrooms. Interventions that align teacher pay to performance have demonstrated substantial positive effects in a range of LMIC settings, including Mexico, China, India and East Africa (see, e.g. Muralidharan and Sundararaman 2011, Behrman et al. 2015, Loyalka et al. 2019, Mbiti et al. 2019, Leaver et al. 2021, Gilligan et al. 2022, Mbiti et al. 2023). However, these interventions have generally been difficult to scale up, and issues of teacher absence and low effort have persisted in most settings.[3] These weaknesses in governance are also visible beyond merely teacher attendance. For example, Singh (2024) and Singh and Berg (2024) show, in two Indian states, that administrative data central to various education reforms are frequently undermined by widespread manipulation and misreporting.
The promise of EdTech for improving education in LMICs
How can ICT systems help address these constraints? We distinguish between four broad categories of EdTech interventions, distinguished by their primary aim and who they target.
Access to technology
This class of interventions reflects a primary focus on reducing barriers to accessing technology: representative interventions focus on, for example, providing laptops or access to the internet. The theory-of-change often treats the technology itself as a pathway to self-directed learning — e.g. through searching out relevant information online — but does not mandate specific instructional activities.
Direct instructional support for students
Perhaps the leading use case for EdTech in LMICs, this category includes educational software that directly provides instruction through computers, smartphones, or tablets, as well as lower-tech programmes that provide instruction through video broadcasts, radio, or basic feature phones. The promise of EdTech is two-fold. First, it can standardise the quality of educational materials at a high level and enable its consistent dissemination, directly addressing constraints related to teacher knowledge and variation in instructional quality. During the COVID-19 pandemic, this area expanded substantially to include a new class of interventions focused on online tutoring and lower-tech remote learning initiatives. Second, with the growing availability of computers and smartphones, it can potentially personalise instruction to students’ actual levels of academic preparation, thus side-stepping the constraints of undifferentiated, curriculum-level instruction. Technology-aided interventions can also allow for real-time feedback and tailored support for students at all ability levels.
Delivery mechanism for information to parents and teachers
The spread of mobile phones (and, increasingly, smartphones with internet access) has substantially broadened both the scope and frequency of information that can be shared with parents. This includes, for example, information about their children’s attendance and performance (see, e.g. Bettinger et al. 2022), details about schooling options available in their neighbourhoods (Allende et al. 2019, Agte et al. 2024), and materials to help parents support their children’s learning. This also holds true for teachers: in many educational systems, centralised digital repositories now provide easy access to curricular resources and administrative information.
Technology to improve assessments
Student assessments play an important role in education systems. They are central to (i) diagnosing student-and school-level needs for supplementary interventions; (ii) certifying skills at key educational milestones; and (iii) allocating students to scarce opportunities such as college admissions. Technology-aided assessment solutions can improve the reliability of these measures by reducing manipulation and providing more granular information (Singh 2024). Moreover, adaptive testing can improve the informativeness of student tests, especially in settings where the curriculum substantially outpaces student ability, by catering to a wider range of achievement levels and, therefore, providing more precise estimates of what students can do, rather than simply indicating a failing grade.
Technology to improve school governance, accountability and management
Technology-aided solutions can improve both the frequency and the reliability of metrics that are used for monitoring and incentivising effort in schools. These applications extend beyond test scores to include, for instance, technology-based tools for monitoring and incentivising teacher attendance (see, e.g. Duflo et al. 2012) or student effort (Hirshleifer 2017). Technology could also improve the design and delivery of teacher training and coaching programmes. Beyond schools, and as in large bureaucracies elsewhere, technology-driven systems can help in strengthening the managerial aspects of the education systems, such as maintaining accurate records for financial transfers, personnel management, and payrolls. While these dimensions relate directly to the overall quality of management in schooling, their link to student achievement is much more indirect, and evidence remains scarce; as such, these fall outside the principal remit of our review.
We summarise the evidence on EdTech in the same categories as presented above.
For full reference list see the end of the conclusion chapter.
Contact VoxDev
If you have questions, feedback, or would like more information about this article, please feel free to reach out to the VoxDev team. We’re here to help with any inquiries and to provide further insights on our research and content.