This blog was written by Keith M Lewin, Chair of the UKFIET Board of Trustees, Emeritus Professor of International Development and Education at the University of Sussex

“Learning to Realize Education’s Promise” is the first time the World Bank has devoted an entire 240-page World Development Report (WDR) to education and learning. It is surprising that it has taken so long, given that the main purpose of the Bank is to finance development, and low-income countries generally spend more on education than anything else apart from defence. The authors are to be congratulated on the encyclopaedic range of their deliberations and the plethora of descriptive analysis provided. Inevitably, not all expectations have been met, and how “education’s promise” can be realised remains elusive.

The main proposition of the WDR 2018 is that there is a new learning crisis and that most countries need to “show that learning really matters to them”. If they succeed they will extract themselves from “the low learning trap” equilibrium of poor learning levels linked to low levels of accountability (p.195). The evidence for this “trap” is patchy, given the lack of longitudinal data on learning. If there is a learning trap, it did not hold back countries that achieved high levels of performance with 20th Century educational investment strategies. If those countries who come later have to play by different rules, what is the new game and what are its new rules?

The WDR argues that three actions are essential at system level:

  • Assess learning—to make it a serious goal
  • Act on evidence—to make schools work for all learners
  • Align actors—to make the whole system work for learning

Phil Coombs, who first wrote about The World Education Crisis in 1968, would have agreed. So would John Dewey, Ralph Tyler, Lee Cronbach, Ben Bloom, Jean Piaget, Maria Montessori, Howard Gardner, and many other well-known educators who “took learning seriously”, but are not cited. Nor are any non-Western thinkers on learning. It is especially odd that the “learning crisis” described makes no reference to China, which has raised the learning levels of more learners than any other country in history over the last 30 years. Phil would have asked the WDR authors, what is new about this global learning crisis, who are the educators who (still) do not believe learning is a serious goal, and what have we (development partners) learned since 1968 about effective pedagogies and whole system change?

This contribution to the dialogue around the report addresses the most striking silence in the WDR. In 240 pages, only 5 pages are spent explicitly addressing questions of finance. If there is a learning crisis then in large part it is a financing crisis. I share Steve Klees’ astonishment that such a substantial report, full of interesting detail, can simply gloss over the central issue of financial sustainability. It is now 18 years since the then President of the World Bank declared in Dakar, at the World Education Forum in 2000, that it would ensure “no government with a credible strategy for achieving Education For All will be allowed to fail for lack of resources”. EFA was about learning as well as access.

There are seven issues related to financing that the WDR could have addressed but did not:

  1. Additional funding?

Recent modelling for the Global Partnership for Education (GPE) (Lewin, 2017) indicates that if schooling were to be universalised in GPE developing country partners, the amounts needed for education would be: 6.2% of GDP in low-income countries (LICs) and 6.3% in low-middle income countries (LMICs) [This scenario would still leave almost half of all children in LICs without access to upper secondary; providing universal access to pre-school would add about 15% to the total cost.] Total public expenditure on education across LICs is about US$19 billion and US$68 billion for LMICs, representing 3.8% and 4.8% of GDP respectively. This includes current aid contributions. To reach 6% of GDP would cost at least another US$13 billion per year for LICs and US$22 billion for LMICs, totalling over US$35 billion a year. There is no prospect of such large volumes of additional recurrent finance becoming available, so what is plan B? A WDR on “Learning” should explore the costs of learning and how they can be met within credible plans.

  1. More efficient learning

The Education 2030 Framework for Action, to which the Bank subscribes, urges “adherence to the international and regional benchmarks of allocating efficiently at least 4 – 6% of Gross Domestic Product and/or at least 15 – 20% of total public expenditure to education” (p.v). But currently, 40% of LICs and LMICs spend less than 4% of GDP on education (of which about a third is aid-related), and less than 15% allocate more than 6% of GDP. Fewer than 20% of LICs and LMICs spend more than 20% of their government budgets on education. If the share of the government budget for education were 20% and the amount of tax collected from domestic revenue were the LIC/LMIC average of 16% of GDP, then this would result in education expenditure being only 3.2% of GDP, i.e. 20% of 16%. It would need more than 30% of the government budget to provide 6% of GDP. This could only be achieved with improbably large cuts to government spending by other Ministries. Learning must be delivered more efficiently and effectively, but the WDR does not indicate how, nor what kind of fiscal reforms are most promising.

  1. Increased balance of domestic revenue to education

Aid to education has plateaued since 2010, at about US$12 billion a year, and there is no convincing sign that the appetite to increase aid to education is returning. The recent Replenishment Conference of the World Bank’s sister organisation, the GPE, raised about US$2.3 billion for disbursement from donors over 3 years, or about US$800 million a year. This fell well short of its aspirations, and was not much greater than in 2014. The amount is only about 2% of the additional amounts needed for recurrent financing for Education 2030. If it were distributed evenly across 50 countries, it would amount to only US$16 million a year per country. Pledges by countries at the GPE replenishment to increase their spending, amounted to US$110 billion, dwarfing the amount of external assistance, and were much more than the US$26 billion pledged in 2014. However, past experience has been that the delivery on pledges is patchy, governments have not honoured many pledges, and some aid commitments have not been delivered. If there is a learning crisis, it was not prevented by the impact of the 2014 replenishment, or the US$50 billion of World Bank investment in education since 2000, so what lessons can be drawn for future aid?

  1. Innovative financing

Though there is a small industry around identifying alternative methods of financing educational investment in LICs, it has yet to demonstrate how to generate the volume of recurrent finance necessary to meet needs. This is not surprising. No high-enrolment high-performance national education system uses innovative finance to fund the bulk of their school systems. Nor is much of their core financing from the private sector. Private sectors in LICs and LMICs are small, and unlikely and unwilling to finance education systems delivering services to those near or below the poverty line. Various reports on innovative finance identify a variety of more or less exotic financial instruments that include debt buy downs, leveraging increased borrowing by multilateral development banks, development bonds, philanthropic contributions, and crowd funding. None of these approaches provide reliable long-term investment to meet the recurrent costs of learning. The WDR should spell out the financing options and the impact of different choices on learning and who learns what.

  1. Rise in public spending in education

Whether there really is a low “learning trap”, leading to a low learning equilibrium, remains to be demonstrated. However, there is evidence of a LIC public expenditure equilibrium for investment in education. This has proved very resilient. According to Coombs in 1985, developing countries as a group increased spending from an average of 2.3% of GDP in 1960 to around 4% by 1979. The proportion of pubic spending allocated to education in developing countries increased 12% to 15% from 1960 to 1975. At the time of the Jomtien Conference, our analysis (Colclough and Lewin, 1990) indicated that on average, LICs were allocating between 4–5% of GDP to education and about 15% of public expenditure. Over the next three decades, up to the present, UNESCO Institute of Statistics data show that the averages have hovered around 4% of GDP and 15% of public expenditure on education. This is the level at which many systems have equilibrated over the long term. Setting arbitrary targets for expenditure on education, independent of demands for spending in other sectors, ignores the obvious. If the education budget, as a percent of GDP, goes up, then something else must come down. If the learning crisis is in part financial, the WDR needs a theory that explains this “resistance to change” to finance learning by increasing its share of the budget despite hundreds of billions of dollars of external assistance for development.

  1. GDP as indicator of spending

The quickest way to increase the proportion of public spending allocated to education is to have an economic recession! This can be seen in the UK. The amount of public expenditure allocated to education as a proportion of GDP increased from 4.9% to 5.8% between 2008 and 2010. To the casual observer this could have signalled a sudden enthusiasm to commit 18% more public resources to education. In fact, the domestic allocation only increased from about £79 billion to £84 billion, or around 6%, at a time when inflation was increasing from about 4% per annum. The apparent increase in educational investment is largely explained by a fall in UK GDP of as much as 20%. The message is simple. The volatility of GDP is much greater than the volatility of most educational expenditure. Measuring the educational effort of governments, and rewarding them with aid for meeting arbitrary and isolated targets, is at the very least risky and may be unwise. The WDR should have devoted more space to developing smarter indicators.

  1. National revenue raising systems

The good news is that national revenue raising systems are modernising. This is slowly transforming the landscape of educational financing and the “gaps” between what is currently financed and what is needed. Aid flows peaked in the early 1990s and most LICs have experienced substantial economic growth. Aid to Africa was greater than tax receipts from 1986 to 1995. Since then, it has fallen relative to GDP and tax revenues are now twice the value of aid. This is what is supposed to happen when countries develop and when aid programmes are effective. As countries develop, direct taxes become a larger share of revenue and more difficult to avoid with better biometric identification and tracking of transactions.

Finance, Tax and Sustainable Educational Development

The evolution of LICs towards becoming “fiscal states” that have the capacity to borrow to invest and grow without reference to aid and its conditionalities, has immense significance. It creates new avenues for financing on scale, especially in Africa. Twice as many African countries (20) took Eurobonds in 2015 as did in 2004. There is the opportunity to make more use of the resources of Africa currently held in pension funds (at least US$334 billon) and sovereign wealth funds (at least US$164 billion), more of which could be invested in LICs and LMICs. In addition, corporate tax evasion and tax fraud is estimated at US$50 billion to US$100 billion, which is a leakage that should no longer be tolerated.

Tax, not aid, is the dominant source of public finance in most countries and this will be even more so in 2030. If there is a new learning crisis it will be located and resolved within the political economies and national curricula of governments accountable to their taxpayers for investing fairly and effectively. The only sustainable solutions will be domestically driven. Achieving substantial increases in educational access and quality, leading to greater learning achievement that is sustainable requires serious fiscal reform, much more effective revenue collection, and awareness of the costs of learning. These should all be a major focus of discussions around “realising learning’s promise”, otherwise learning gains will not be sustained.

The World Bank is good at financing, but not necessarily at learning. It should be good at developing sustainable systems of finance for learning. The WDR needs to explain how this can happen, so that Phil Coombs’ academic descendants will not write another book on the Global Education Crisis in 2030.

This blog is an abridged version of the original article, which can be found on Keith’s website.

 It was also posted simultaneously on Education International’s Worlds of Education website.