The World Bank’s 2018 World Development Report on Education: a sceptic’s review

October 18, 2017

     By Duncan Green     

Guest post from Prachi Srivastava (@PrachiSrivas), Associate Professor in the area of education and international development at the University of Western Ontario.

When the World Bank announced that the 2018 World Development Report (WDR) would be on education, I was sceptical.

I’m not denying the Bank’s research expertise. It devotes substantial money and staff and has a trove of reports that are accessible in the public domain.

It’s also open to criticism – and receives lots of it, especially on education.

So why the scepticism?

To begin with, the Bank is one of the largest external donors for education in developing countries. ‘[M]anaging a portfolio of $US 9 billion with operations in 71 countries as of January 2013’, yet the 2018 WDR is the first since the series launched in 1978 to focus solely on education. What took it so long?

Secondly, how the Bank frames problems and assesses ‘evidence’ matters. It’s no secret that the Bank, even if implicitly, favours certain disciplines (i.e., economics) and methodological approaches (usually ‘big n’ quantitative studies). I’m not debating their virtues or vices. But this propensity has implications for the evidence that gets filtered in and out, and consequences for the ‘take-aways’. This is important because it has direct policy influence and signalling power, not only to governments and institutions with years of sectoral expertise, but also to a new community of actors, including private sector donors, with less experience in education.

Scepticism aside, the 2018 WDR Learning to Realize Education’s Promise does a few welcome things.

1. It invokes education as a basic right, and makes a strong moral case for prioritizing education and learning. The Report starts with the bold statement, ‘Education is a basic human right, and is central to unblocking human capabilities’ (p. 1), referencing Amartya Sen’s now seminal capabilities approach. Coming from the Bank, this is big.

It has been heavily criticised (famously by the late Katarina Tomasevski, the first UN Special Rapporteur on the Right to Education) for distancing itself from the rights-based approach in favour of an instrumental approach, and eschewing the moral imperative to act in education. While the instrumental approach remains, there are also rights-based and moral assertions. The Report even mentions examples of using litigation to claim the right to education in India and Indonesia (Box 11.3), and states that such action can work in the interests of the disadvantaged.

These are very welcome additions to the business case for education that has been taking hold. We have all heard the oft-repeated mantra (including by senior Bank staff) that education isn’t ‘just’ the right thing to do, it’s the ‘smart’ thing to do. (I’m always left wondering why simply doing the right thing isn’t enough). While the business case has become important in mobilising the will to increase education financing from donors, the Report’s opening messages on rights and moral obligations may help to reset the tone, or at least open these arguments up to a broader audience. I would have liked to see a stronger attempt at building a moral case for financing education, especially by OECD-DAC donors, many of which have not fulfilled their commitments, but this is a start.

2. It highlights the learning crisis. The Report presents evidence on learning disparities from a range of countries, particularly highlighting differences between children from richer and poorer backgrounds. It underscores the message that enrolment isn’t enough. While that message isn’t new to education researchers and policy wonks, the WDR reaches audiences (internal and external) that are likely to be unaware of the sheer extent of the learning crisis. It could have more fully assessed the evidence on the potential impact of the learning crisis on other skills (e.g., citizenship, critical thinking, creativity, ‘21st century skills’), which it begins to do (Spotlight 3), but this could have been more fully integrated.

3. It demonstrates the link between neuroscience and education. The Report devotes a substantial amount of space to presenting scientific evidence on the link between deprivation, early brain development, and learning, a point also addressed in previous Education for All Global Monitoring Reports. It highlights the consequences of malnutrition and brain development for children from deprived backgrounds on learning outcomes. ‘So even in a good school, deprived children learn less’ (p. 10), and early learning deficits are generally magnified over time (p. 7). The MRI images of the brain structure of an infant in Bangladesh whose growth was stunted compared with one whose wasn’t, are sobering (p. 115). The discussion and powerful imagery may help to press the urgency of early interventions and a more integrated approach to education delivery for deprived groups, especially in early and primary years.

Given the well-drawn out arguments in the Report, it is curious that this line of thinking isn’t embedded in analysing the evidence presented on school feeding programs in Burkina Faso, Kenya, and Peru (p. 148). It is also curious that evidence on India’s Midday Meal Scheme, the largest school feeding program in the world, is not cited. Young Lives analysis showed significant positive results of the Midday Meal Scheme on learning (vocabulary) in Andhra Pradesh, contrary to the studies cited.

4. It does a relatively decent (if condensed) job of presenting evidence on private sector provision and private schooling. Cards on the table—I was invited to meet with some members of the WDR Team to discuss this issue during informal consultations on the Report. Firstly, despite the unclear position on user fees the Bank has sometimes been known to take, the WDR is explicit about the negative impact of school fees and costs (p. 117). A range of policy interventions to help with schooling costs and fees are presented (e.g., non-merit scholarships, conditional cash transfers), as well as country experiences of eliminating fees (Figure 5.4, p. 118).

While it supports the theoretical assumption that competition increases quality, despite insufficient evidence in developing countries, regarding outcomes, the Report is clear: ‘there is no consistent evidence that private schools deliver better learning outcomes than public schools’ (p. 176). Huge.

The Report addresses concerns about ‘cream-skimming’ of students. Also leaving me pleasantly surprised, it is open about the potential of commercially-oriented providers to exploit incentives for profit, ‘Some private suppliers of education services…may, in the pursuit of profit, advocate policy choices not in the interest of students (p. 13).

It is refreshingly forthright in admitting difficulties in overseeing systems with a myriad of private schools, suggesting ‘governments may deem it more straightforward to provide quality education than to regulate a disparate collection of providers that may not have the same objectives’ (p. 177). Finally, the report considers systemic effects, albeit briefly, that private schooling expansion ‘can undermine the political constituency for effective public schooling in the longer term’ (p. 177). Massive.

So- the 2018 WDR makes some welcome interventions. However, to fully realise its potential, here are my top two suggestions for what it could have done better (there are others—I haven’t touched on teachers—but this post is now resembling an article):

Make the case for financing education less ambiguous. This has been raised as the main point of contention in nearly all the online reviews of the 2018 WDR (e.g., David Archer of Action Aid and Education International). And I would have to agree.

The ambiguity is incongruous with recommendations of high-level fora. For example, the Education 2030 Incheon Declaration and Framework for Action for SDG4 is unequivocal: The ‘Least developed countries need to reach or exceed the upper end of these benchmarks [4-6% GDP, 15-20% national budgets] if they are to achieve the targets’ (emphasis mine).

The discussion is also somewhat ahistorical. An earlier analysis by Mehrotra of what he termed, 10 ‘early high-achievers in education’ (Barbados, Botswana, Costa Rica, Cuba, Kerala (India), Malaysia, Mauritius, South Korea, Sri Lanka, Zimbabwe) concluded that high public expenditure as a proportion of GDP and as a proportion of national budgets were among the factors that contributed to expansion of relatively better quality primary education in early post-colonial contexts. If the aim now is to universalise education to secondary by 2030 (SDG Target 4.1), surely stable, secure, and increased financing by donors and domestic governments is pivotal, especially in countries that do not meet even the minimum benchmarks.

Focus on learning processes within schools and classrooms and their potential effects on learning outcomes. The Report acknowledges: ‘Learning is a complex process that is difficult to break down into simple linear relationships from cause to effect’ (p. 178). And it rightly attributes poor learning outcomes to poor quality provision. No one would argue otherwise.

However, ‘quality’ is influenced by a host of factors, many of which may be normative, socio-political, and micro-political (i.e., informal institutions). The learning outcomes that are the subject of the WDR are produced through learning processes structured in formal schooling processes. And formal schooling processes are embedded in the overt and hidden curriculum of the schools and classrooms (i.e., values and the reproduction of those values in formal schools) that children of different backgrounds have access to, and how those children, in turn, are positioned within them.

For example, research in India  shows that broader societal caste-based practices continue to affect how children experience schooling even within universalising initiatives. Based on emerging analysis from my current study of roughly 1500 school-aged children in Delhi, I argue that silent exclusion reflects broader societal exclusion and will impede meaningful learning even if children are enrolled (see also here).

This is messy stuff. Which means that improving quality will be harder than ‘aligning all the ingredients’ (Box 9.2). These are deep-seated issues that cannot easily be overcome by the ‘proper’ incentives.

Acknowledgements: I am grateful to David Evans, a Lead Economist on the 2018 WDR Team, for sharing insights and clarifying questions on the WDR process. Any errors or misunderstandings are mine.

October 18, 2017
Duncan Green