Education data in four charts

A simple but critical point for any data-driven transparency reform is that it can only be as strong as its source data. There’s no sense in publishing information if it is incomplete, unreliable, or impenetrable.

For some countries, making education data transparent is simply a matter of making existing datasets available to the public, or adapting or simplifying datasets to be more accessible and useful to citizens. In many countries, however—and particularly in developing countries—the first step will be to establish robust data collection processes to fill large data gaps and to strengthen existing systems to guard against poor quality data, i.e., “garbage in, garbage out.”

Data are often outdated and inadequately disaggregated or comparable. They are collected and shared mainly in non-machine readable formats (such as PDFs), which makes it difficult for users to analyze. Worst of all data are often overblown and inaccurate, either because governments or organizations lack the capacity to measure accurately, or overstate results due to local or international pressures.

To get a sense of how deep these issues run in education systems around the world, the Center for Universal Education at Brookings undertook a survey of ministry of education websites in low- and middle-income countries to understand the quality and availability of school-level data.

We found that, of 133 countries assessed, nearly half have no available data, because either ministries did not have websites, data were missing, or data were prohibitively difficult to access, i.e., a password was required. Of the remaining 72 countries, 43 have data only at the national level (non-disaggregated), leaving 29 countries with sufficiently disaggregated school-level data.

Figure 1: Level of disaggregation

Moreover, of data that are available, the majority is in PDF or non-downloadable format.

Figure 2: Data format

We also found that, while student data are commonly available, ministry websites primarily capture information on enrollment rather than indicators more closely linked to school quality, such as student assessment data or pass/dropout rates.

Figure 3: Student data

So too, when it is available, information on financing is focused on budgets rather than expenditures. These are “low-hanging fruit” datasets that have little influence over relationships between transparency and accountability and are often difficult for the public to understand without additional context. If the data are not meaningful or decipherable then it is unlikely that the public will access them much less use them to hold government to account.

Figure 4: Finance data

It is important to note that this exercise does not provide a definitive picture of data that are available, since they could be housed on a different website or simply not shared with the public. Similarly, this assessment does not provide any indication of the accuracy of the data or the strength of underlying data collection methodologies. However, it does provide a telling snapshot (as of April 2016) of the state of education data in developing countries—that is, as a whole, feeble.

Of course, the picture need not be so bleak. Our research revealed a number of shining examples of ministry websites that are both easy to navigate and also provide rich and timely datasets on school-level and national indicators, including Lebanon, India, Moldova, and the Philippines, among others. Lebanon’s Ministry of Education website, for example, includes an easy-to-navigate table of indicators about individual schools, such as examination scores, classroom characteristics, and licensing information of faculty members. India, too, provides report cards of individual schools with detailed, well-organized tables that document information about school facilities, staff, and examination results. The website also includes recently updated information about national indicators as well as summative annual reports.

A number of countries also have high potential in that they have systems in place to make school-level data more accessible, but are hampered by minor deficiencies (such as being difficult to navigate). Uganda, for instance, provides school and national level data on ownership, enrollment, classroom characteristics, and information on teachers. However, information is only offered in PDF format.

Underpinning transparency reform is a necessary first step ensuring that data are available, trustworthy, usable, and, ideally, that people actually care about the information. We find plenty of countries on a positive path, but satisfying all requirements simultaneously is a high—but necessary—bar that many have yet to satisfy.