Most children in the Global South now attend school. But:
- There’s often no teacher in the classroom.
- When there is a teacher, they often aren’t teaching.
- When they are teaching, they often lack crucial knowledge and skip key lesson features.
This means students learn less than we would hope (issues I explored in my previous post).
So what’s the good news? Well, over the last couple of decades, programmes which dramatically improve teaching have emerged in the Global South. The best approaches achieve an impact per dollar spent almost as great as anything we can do to increase student learning (Angrist et al., 2020). These programmes have solved challenges which face all professional development providers: working at scale, transferring to new countries, and sustaining impact.
There’s no compelling synthesis of the evidence around effective professional development in the Global South. One barrier is the lack of trials: in a recent, careful search, I found fewer than 30 randomised controlled trials of teacher professional development, ever (compared to 120+ in the last two decades in the North). One recent review (Popova et al., 2021), collected a enormous variety of studies (from remedial computer education to improving boarding school management) then looked at the features which correlated with impact. Features which correlate with impact may not cause it. This perhaps explains apparent contradictions in the findings: for example, initial face-to-face sustained training is associated with greater impact – but more time in training and longer programmes aren’t. Lesson enactment helps – but more time practising doesn’t. The other review (Mitchell et al., 2024) focused solely on papers published by authors with African affiliations. It offers valuable insights about how countries usually approach professional development, but omits the programmes with the best evidence of student outcomes, covering the most promising strand of research in just one sentence.
So what follows is instead my attempt to explain three crucial strengths of the most promising programmes – and a crucial caveat. Based on this, I suggest five questions to consider in designing professional development in the Global South.
1) The best programmes scale
It’s (relatively) easy to design a programme which works in one or five or ten schools. It’s far harder to make the same programme work across a hundred schools, or a thousand.
My favourite example is Thinking, Doing, Talking Science (in England). A trial with 21 schools achieved an impressive impact on student learning: an effect size of 0.22 (Hanley, Slavin and Elliott, 2015). Logically, you take successful programmes to more schools, so a larger trial was launched, with 102 schools. Almost everything was the same: goals, training, resources. One thing changed: in the smaller trial, the programme designers delivered the training; for the larger trial, the designers trained others to deliver the training. The designers thought the facilitators did a good job, the teachers thought the training was effective – but the result was an effect size of just 0.01 (Kitmitto et al., 2018). This is a common problem. Individual programme designers are often highly effective. Building a team of effective teacher educators is far harder. If you can’t scale a programme, your impact will only ever be local and limited.
In the Global South, we find programmes which have scaled successfully. Perhaps the most impressive is Tusome (‘we read’, in Swahili). Several carefully-run pilots (under the name PRIMR) allowed designers to test and refine their approach. This informed the development of a national literacy programme to operate across 23,000 primary schools. Among the features of the successful scale up were:
- Communicating expectations in ways that made sense to teachers: the reading fluency students should achieve, in words per minute
- Providing the resources needed: books for every child; a guide for every teacher
- Getting government education officers visiting schools, monitoring student fluency and teacher progress and providing feedback (and checking that officers were doing so too; Piper, 2018a)
Unusually, researchers tested the role that individual components of the programme played. Was professional development and coaching enough? Did adding textbooks make a difference? What about teacher guides? The best results came from including all of these components (Piper, 2018b). Offering comprehensive guidance and support to students, teachers and those supporting teachers is a Swiss cheese model. Even where there are gaps (a teacher ignores the teacher guide, for example), the variety of support (teacher guide, teacher support, student materials), makes it hard for students to fall through them. Tusome saw effect sizes of 0.5-1.0 – extremely high for large-scale programmes – with the proportion of students reaching benchmarks in English and Swahili doubling (Piper et al., 2018a).
The best professional development in the Global South is effective at scale.
Highly structured programmes offering a range of support can dramatically improve student learning at scale. We find similar effects in the Teaching at the Right Level programmes (which we’ll discuss below). This brings us to our next strength.
2) The best programmes replicate and transfer
Just as programmes often fail when they scale up, programmes which work in one place often fail elsewhere. This is a problem: until you can come up with something which can be written down and replicated elsewhere, you have clever individuals doing impressive things – not a programme.
In the Global South, we see programmes replicating and transfering. Teaching at the Right Level (TaRL) exemplifies this. The premise is simple. Most Third Grade students don’t have the knowledge and understanding needed to learn the Third Grade curriculum. Most teachers plough through the Third Grade curriculum anyway. TaRL addresses this by placing students ‘at the right level’: grouping them based on current achievement, and teaching them accordingly.
This sounds obvious. As with any change in schools, making it work is extremely difficult. TaRL programmes have:
- Recruited teaching assistants – but found that schools didn’t deploy them as intended
- Sent teachers materials – but found just a quarter of schools used them
- Asked teachers to regroup students – but found just one in ten schools did so (Bihar and Uttarakhand; Banerjee et al., 2016; see also Duflo, 2020).
Making the programme work involves careful design, support and accountability.
TaRL was first tested in India. It has been replicated there across different states (each with its own education system). Having succeeded there, several versions – using teachers and teaching assistants, during and after class – were successfully replicated in 400 schools in Ghana (Duflo et al., 2020). Another replication, in Zambia, reached 1800 schools and led to substantial improvements in literacy and numeracy (Vramont et al., 2021).
We see something similar in the replication of a structured pedagogy approach which succeeded in Telangana (India). Researchers ran it (with some adaptations for context) in both Gambia and Guinea-Bissau, creating large learning gains in both (Eble et al., 2021; Fazzio et al., 2021).
Clearly, there will always be a degree of contextualisation needed: I’m not suggesting a one-size-fits-all intervention is going to work in every country, system or village. In particular, designers often have to vary who teaches (teachers or teaching assistants) and where (in or out of school) to make it work (see, e.g. Duflo et al., 2020; Fazzio et al., 2020, n.11) But the success of these programmes across countries suggests are converging on a model for effective professional development:
- Establish exactly what you want teachers to do: take a structured approach to teaching basic skills
- Provide comprehensive support to help them do so: student resources, teacher guides and ongoing monitoring and feedback
(This incorporates many of the mechanisms we’ve found effective in supporting teacher change in the Global North.)
The best professional development replicates, transferring across countries and continents.
3) The best programmes last
A particularly depressing barrier to effective teacher development is that positive results often dissipate after the programme ends. For example, a yearlong programme in El Salvador successfully increased teachers’ mathematical knowledge. By the next year, the gains had disappeared (Brunetti et al., 2023).
In the Global South, we find programmes with lasting effects. In South Africa, researchers examined a programme which aimed to improve primary pupils’ reading through teachers professional development or monthly instructional coaching (Cilliers et al., 2022). The initial programme was effective. Coming back the next year, the researchers found that – without further coaching or training – teachers were still more likely to be:
- Using reading books
- Using lesson plans
- Using specified activities
Students were still learning much more – particularly among teachers who received coaching – than those in the control group.
We’re also seeing lasting effects at an institutional level. For example:
- In Zambia, Teaching at the Right Level became both more effective (student gains from participating increased) and more cost effective (the cost per child fell from $291 in the pilot to $11 in the full-scale programme) as officials gained knowledge and skill in running it (Vramont et al., 2021).
- In Ghana, TaRL was run almost entirely by the government, without NGO support; its effects persisted a year after the intervention ended (Duflo et al., 2020, 9).
- In South Africa, Funda Wande developed an effective programme and built legitimacy and support, developing relationships at multiple levels with the education ministry. It adapted to maintain effectiveness when ministers and policy priorities changed (Samji and Kapoor, 2022).
The best professional development has lasting impact, in schools and ministries.
Why do these programmes work?
Many students in the Global South aren’t getting the teaching they deserve. But we can be increasingly optimistic: the best programmes work at scale, across countries, with lasting effects. Drawing together the approaches these programmes take, if we were designing or assessing a potential intervention in the Global South, we might begin by asking:
- How structured is the teaching approach? The most successful approach is often called ‘structured pedagogy’ – for good reason. Where teacher knowledge and expertise is limited (as we’ve seen), we must offer very clear guidance about what to teach and how best to teach it.
- How comprehensive is the programme? A range of support seems important: if in doubt, provide student resources and teacher guides and training and ongoing support. (This echoes our findings of our systematic review of teacher professional development in the Global North: the more mechanisms you use, the better.)
- How are teachers offered ongoing support, and held accountable? Teachers need continuing support: they need feedback and further training, as they make sense of new teaching approaches, and ongoing support holds them accountable for their presence and performance.
- How are we working with (and building) state capacity? The government does much of the work of education (Samji and Kapoor, 2022). We’ve seen the barriers students face in many countries: teacher absenteeism, delayed pay, strikes. Many authors are sanguine about the ability of states to overcome vested interests and drive change in their countries (for example, Bold et al., 2018, Fazzio et al., 2021). But, in some contexts, programme designers have found effective ways to work with the state and increase its capacity.
There’s also a meta-question:
- How are we iterating our approach? We may dream of overnight success, but the best programmes build on tough years of trial and error. Tusome ran multiple, large-scale pilots, which tested programme features and informed the design of the national programme (Piper et al., 2018a). Through piloting, Funda Wande made crucial tweaks – like filming models in classrooms like those of the teachers they wanted to reach (Samji and Kapoor, 2022). Most strikingly, the success of Teaching at the Right Level builds on years of struggle:
- In the first attempt, three programmes were tried. Only one succeeded (Banerjee et al., 2010).
- In the second attempt three versions of the initially successful programme were tested in both Bihar and Uttarakhand. Only one succeeded and only in Bihar (Banerjee et al., 2016).
- In the third attempt, two versions were attempted (Banrjee et al., 2017 – Haryana and Uttar Pradesh). Both succeeded.
In the process, researchers refined their methods (from a range of community involvement approaches to a specific approach to teaching), the people involved (from volunteers to teachers) and their approach to training and supporting those teachers.
A caveat
Another strand of research offers an interesting caveat. These promising programmes, they suggest, achieve effects which are impressive, but shallow. Ursula Hoadley (2024) spent time observing how teachers responded to a structured pedagogy programme. The programme was effective: student learning increased. But Hoadley found that teachers changed what they taught (using the more demanding texts they were provided), but struggled to change how they taught or to understand why. A similar examination of a very different programme found teachers following new procedures but not attending to pupil learning. Teachers “were confident that pupils understood the lesson,” but their discussion of what had happened was “theoretical”: none of them referred to any concrete incidents (Miyazaki, 2016, 13-14).”
So, a final question: how well do we really understand teachers’ beliefs and actions – before, during and after the programme? We need economists (designing large-scale randomised trials) and qualitative researchers (looking at teachers’ thoughts and beliefs) to interact and inform one another’s work.
Conclusion
Children in the Global South still face huge hurdles to getting a good education. But we are nearer than we’ve ever been to overcoming these hurdles. Increasingly, we have programmes which:
- Scale
- Transfer
- Last
They do so through structured teaching, comprehensive support, ongoing feedback, and – often – working effectively with government. They iterate to achieve this success. If we can sustain and expand these approaches – and better understand and reach teachers – perhaps we can crack the learning crisis in the Global South.
If you enjoyed, this you may appreciate…
My previous post, on barriers to learning in the Global South.
The findings of our research on elements of effective professional development in the Global North.
Broader thoughts about designing effective professional development: three questions every professional development designer must answer.
References
Vromant, N., Kuppens, L., Hazemba, M., Mwamba, F. K., & Cupito, E. (2021). Scaling Teaching at the Right Level. In Munene, I. Ensuring All Children Learn: Lessons from the South on What Works in Equity and Inclusion: Rowman & Littlefield.
I really appreciate your summary, especially on why programs like Tusome and TARL work.
I want to build on your caveat, which I think is very important. I have found that teachers’ beliefs are often indicative of what key stakeholders across the education system believe, inspectors being a key stakeholder. To teachers, what inspectors think is far more important than what the evidence says or what a well-meaning actor from the Global North says, even if the program comes with incentives.
If teachers are not subject knowledge experts in what actors from the Global North call subject knowledge, I think it’s helpful to ask, what are they experts in? For example, teachers may be likely to score badly on phonics knowledge assessments but are far more likely to score well on the official government primary exit exam. (Here I’m just reacting to the research that says teachers in the Global South lack expertise, which is valid, but lacks nuance. What if what they think and believe about teaching is just different to what actors in the Global North are looking for? Rather than only focusing on what they don’t know, how about finding out and sharing what they do know as well?)
What is important to teachers has to be important to program design.