Working with statistics
Statistics are a useful tool to give weight to your arguments and express key points clearly. They can also help in analysis, especially if you are comparing the impact of the same policy across regions, or the state of education in different places. They can help you develop a gendered analysis of the budget, or an analysis from the perspective of a particular group of people – for example, an ethnic or linguistic minority, pastoralists or people with a disability.
It can be very useful to access statistics produced by others (government, academics, donors, civil society groups) to aid in your analysis of an issue – or to support your analysis of budget allocations, implementation and impact. It is only through matching budget allocations to reliable demographic information that we can start to understand the levels of expenditure on education, and the potential impact of a budget.
However, it is important to recognise that statistics are notoriously unreliable. They may be presented as neutral but they are invariably highly political – they can hide as much as they show. When working with statistics produced by others you need to consider:
- Who compiled the statistics? What prejudices might the group have?
- What is being shown in the statistics and why were they compiled? This can solicit a purely descriptive answer, such as ‘enrolment rates’ or an analytical one, such as ‘that enrolment rates are increasing’.
- How are the statistics relevant to what we have been discussing?
- What comparisons can be made?
- What categories have been chosen in the statistics?
- What has been left out?
- Who decided on the categories? How do the definitions impact on the figures collected?
- How could the information have been collected?
- What sorts of decisions were made in compiling the statistics? What assumptions were made?
- Are the statistics neutral? What do they highlight or obscure?
- How accurate are the statistics? Are there aspects that you would challenge?
- How could they be used by different parties? How could they be used for advocacy, monitoring, campaigning or alliance building?
Because of this unreliability it is often useful to produce your own statistics, which can be used to challenge the data produced by others, as well as further your understanding. Of course, any statistics you produce will be open to the same criticism. It is important to be open and transparent about the methodologies used and assumptions made, and to point out gaps in information – what you are not able to conclude from the information you have. It is also useful to access statistics over a long time period (over five years) as this will enable you to judge their accuracy more reliably.
UNESCO Institute of Statistics recognises that countries have been known to manipulate statistics to give the appearance that they are closer to meeting internationally-agreed targets. One example of statistical manipulation relates to enrolment figures. These statistics an incredibly inaccurate indication of school attendance. They are compiled using registration at the beginning of the academic year and do not take actual attendance into consideration. The fact that many children repeat years means that the gross enrolment is frequently over 100 per cent. Another manipulation relates to national spending on primary education. Unless figures actually include the source of spending (e.g. government, donors, NGOs, parents) it is impossible to analyse real levels of expenditure on education.