The “data revolution” has become a buzz term in the development community these last few months. It may be French cynicism, but when I think about revolution, it reminds me of a quote from “The Leopard”, a great book by Giuseppe Tomasi di Lampedusa: "If we want things to stay as they are, things will have to change."[1]

So let’s be clear about which things should not stay as they are. According to UNESCO’s Institute for Statistics (UIS), nearly 58 million children of primary school age, and around 63 million young adolescents (between 13 and 15 years old) worldwide were not enrolled in school in 2012. This has to change.

According to the estimates of the Global Monitoring Report, about 250 million children worldwide fail to acquire the minimum level of learning by grade 4 or do not reach grade 4. Of these children, around 100 million are in partner developing countries of the Global Partnership for Education. This has to change, too.

How do we get out-of-school children in school and learning without data?

Targeting the out-of-school population is critical for policies to be effective. Poor children in remote areas, those living in conflict areas, disabled children, girls, ethnic minorities – they all require specific policies. But how are we going to put children in school if we don’t know who and where they are?

We also face a learning crisis as the 250 million children remind us. But because of the lack of learning data, this figure is more a ‘guestimate’ than an estimate. It gives us an order of magnitude of the crisis, but learning data from developing countries are critical to address this crisis and they are missing in too many countries, particularly the poorest.

Data are the basis for better policies

I want to make one point clear: investing in data are not meant to please statisticians, it is meant to inform policies to ensure that they are as effective as possible and reach children with the education they need.

Here are the two key challenges the education community is facing: (i) reaching the marginalized and (ii) improving learning. This won’t be achieved unless there is a dramatic change in the way we consider data.

So maybe the data revolution should in fact be a policy revolution, where solid evidence based on reliable data allows for the implementation of effective policies that ensure children are in school and learn.

To address new challenges we need new data

During the GPE replenishment conference a few weeks ago in Brussels, a parallel session was dedicated to the so-called data revolution. Jo Bourne from UNICEF introduced the discussion with the following questions: What’s wrong with data? Why do we need a data revolution?

Two key elements emerged from the discussion.

First, we still lack a lot of basic information in the education sector, which is particularly shocking when compared with the health sector. The explanation is straightforward, though: there was and there is an underinvestment in data in the education sector. There is also no doubt that the data gap has a negative impact on the level of international funding for education.

The second element introduced by Hendrik Van der Pol from UIS is that the world is changing quickly, and in order to address new challenges, we need new data.

So, if this data revolution is about real and significant change, what are the game changers for the education sector?  I suggest we focus on three big things:

1. Development of household surveys dedicated to education: This will help analyze equity issues and provide essential data for policies in order to reach marginalized children.  But to develop effective policies that reach minorities, disabled children, children in rural areas, poor children, girls, etc. who are not in school, we need more sophisticated data than what is currently provided by national Education Management and Information Systems (EMIS), particularly regarding out-of-school children. The only way to get such information is through household surveys. Unfortunately, existing surveys such as MICS and DHS[2] have not been designed for education and don’t provide the information needed for targeted equity policies.

Very few household surveys dedicated to education have been implemented and there is a need to systematize their use.

The cost of such surveys depends on country specificities but an average cost of $500,000 is probably a reasonable order of magnitude for a high quality survey.

2. Establishment of an international platform to support learning assessments: The learning crisis is at the top of the agenda for the education sector.

Unfortunately the lack of learning data affects the ability of national governments to develop meaningful policies and monitor progress. It’s a question of credibility for the education sector for the post-2015 agenda.

The work of the Learning Metrics Task Force fostered progress in the last two years. Ongoing work emphasizes two important gaps: (i) insufficient financing for learning assessments in low income countries, (ii) lack of coordination between key stakeholders, and (iii) data are not always used when they exist. An international platform could fill these gaps and support sound learning assessments in the poorest countries by raising additional funding and providing technical assistance. In addition, this platform could help track learning progress, both at country and international levels[3], as the development of national learning assessment systems is the key for progress. An investment of $50 million dollars over the next five years could help collect and use learning data in the 40 poorest countries in the world.

3. Creation of an international task force for education data to solve key data problems. Such a task force would bring together key international organizations, including civil society organizations in the education sector.

But it would also involve new partners, particularly from the private sector which could bring  expertise in new technologies and new perspectives for data collection, data processing, data dissemination and also more generally for information systems.

It is also essential to connect with initiatives and people working on data in other sectors to benefit from their experience.

Better data collection is worth the cost

Based on these initial estimates, the cost for significant improvements in terms of data for the 40 poorest countries in the world would not exceed $70 million dollars over the next four or five years. It is certainly not a small amount but it represents less than 3% of the amount raised in Brussels by the Global Partnership for Education for the same period, and it is less than 1% of the annual amount of ODA for education.

So if the education community wants to be serious about data, we need to start by investing in data strategically and consistently. Following the suggestions above would be a good start.

At a time where value for money is guiding most of the financing in development, it is worthy to remind us that investment in data are high value for money.

The Global Partnership for Education has taken a strong position on data through its data strategy which is reflected in its new funding model but also in the case for investment with very concrete steps. This is just the beginning and much more needs to be done.

The Global Partnership will continue to play an important role in the collective effort of the education community towards the data/policy revolution.