Policy paper

Methodology note: number of children supported to gain a decent education since 2020

Updated 19 February 2024

This was published under the 2022 to 2024 Sunak Conservative government

Indicator description: Number of children supported by FCDO to gain a decent education since April 2020

Version: July 2023

Type of indicator

Output Indicator.

Results can be:

• cumulative (the results are added across years when different children are supported in different years) or
• peak year (the year with the highest results is used when there is overlap between the children supported in different years)

Overview

This indicator estimates the number of children supported by FCDO to gain a better-quality education. It tracks the full-time equivalent number of children FCDO has supported in education for at least a year. The use of full-time equivalent numbers of children supported provides a measure which is consistent across different countries and programmes and means one supported child is equivalent to roughly a year’s worth of education.

It consists of children who FCDO:

• fully educates or fully funds through school
• supports the majority of their education, such as if children are only in education due to FCDO support
• provides partial support to improve the education or attendance of children already in school, in which case a proportion of the child is counted based on the estimated proportion of their education attributed to FCDO

The estimate covers children in pre-primary, primary, and secondary education, children in both formal and non-formal schools, and children provided with vocational or skills education.^{[footnote 1]} Children are counted if supported for at least a year in a government school, or roughly its equivalent in non-formal or non-government education.

Aggregate FCDO results will be published on the gov.uk website on girls’ education results. Whilst individual programme results will not be published, results from underlying programmes should still be able to stand up to scrutiny from internal and external audiences. We ask that you follow the methodology described in this Note and provide supporting evidence for your calculations when submitting results.

Technical definitions

Key terms

Children

Count individuals who are 17 years and younger. This typically corresponds to pre-school to upper secondary aged children. Also count youth (up to the age of 24) attending education programmes designed for children. This allows us to capture young people just over the principal age threshold who are still learning at the upper secondary level.

Girls

Children identified as female, or youths identified as female who are attending education programmes designed for children.

Boys

Children identified as male, or youths identified as male who are attending education programmes designed for children.

Decent education

For a programme to count as supporting a decent education, it must do one or more of the following:

1. Provide a quality education
2. Improve the quality of existing education provision
3. Provide an education where there is no alternative education provision (for example, emergency settings)

Please indicate through which of these pathways a programme contributes to a decent education in the appropriate section of the results return template.

In the case of the first 2 – provision of quality education – this means a programme works to improve children’s learning outcomes. It is not necessary to demonstrate that children’s learning has improved because this is an output indicator. However, a short description of the action taken by a programme to improve learning outcomes must be reported in the results template. Where a programme is supporting education because there is no alternative provision, then action taken by the programme to make this provision must be described in the results template.

General principle

The general principle is to count all children whose education is benefiting from FCDO funding. The full-time equivalent number should be used when FCDO is only providing partial support. This is to ensure consistency between different types of education programmes, with different intensities of support. The full-time equivalent number is a proportion of children benefiting from FCDO support based on an estimate of FCDO’s contribution to their education over a year. This could be based on funding shares, learning outcomes, quality indicators or other relevant data sources. The approach taken will vary by programme depending on available data and the focus and expected impact of FCDO support. Estimates, proxies, or partial shares can be used when we don’t easily have the relevant information. When FCDO is clearly providing the majority of funding or learning experience, all children can be counted.

Children should be counted if we are supporting them for roughly the equivalent of at least one year enrolled in a government school (in terms of curriculum, attendance and focus on learning). This may be less than 9 months full time for non-government schools if the data collected is based on attendance rather than enrolment, or if the curriculum is more condensed. Relevant proportions should be taken when we are only providing a partial education (eg 20% of a child if we only support a child for a fifth of a school year).

Education funding provided through the education budget

Whilst no longer a commonly used instrument, when FCDO is providing budget or sector support, or financial aid through the government budget, the methodology is a pro-rata share of enrolment. The share is FCDO’s contribution to the education spend for the relevant schools.

First, the number of full-time equivalent children (or schools) covered by the FCDO programme is identified. Then the percentage of total education spend on these children funded by FCDO is estimated. This is usually the total FCDO spend divided by the total (government and donor) spend on the relevant children. Finally, this percentage of total education spend funded by FCDO is multiplied by the total number of children enrolled. This can also be expressed as the total cost per full-time child, multiplied by the FCDO contribution.

The same approach should be taken when funding non-government schools when FCDO funding goes through the private sector or NGO budget.

Education funding provided outside the education budget where FCDO is clearly providing the majority of funding or learning experience

In many cases FCDO programmes do not fund a child’s education through the education budget, but do provide influential partial support such as through technical assistance, targeted financial aid, cash transfers etc. We can count all children when we are clearly providing the majority of their funding or learning experience (more than approximately 75%). This could include children who are only in school, or only learning, because of FCDO support.

Other education funding provided outside the education budget

These programmes should all be included. However, a proportion of these children should be counted based on an estimate of FCDO’s contribution to their education. This ensures they are counted in a broadly comparable way to those that are fully funded through the government budget.

The most straightforward approach to calculate FCDO’s attribution to a child’s education is using funding shares. This calculation is broadly the same as with funding through the budget. First the total number of children benefiting from the FCDO programme is calculated. Then the percentage of their total education spend funded by the FCDO programme is estimated. When the relevant data is available, the recommended approach is to combine the FCDO spend with an estimate of national government unit costs. In this case the total education spend on these children would be the national government unit cost multiplied by the number of children benefiting, added to the additional spend from FCDO and other donors. Finally, the percentage of total education spend funded by FCDO is multiplied by the total number of children benefiting.

In many cases the relevant non-FCDO financial information will not be available. In these cases, partial or proxy estimates for unit cost data will be needed. For example, just using donor spend if there’s no information on the government spend, or unit costs from the region, similar projects in-country, or neighbouring countries if we feel they are a reasonable proxy.

For large one-off expenditure the same basic calculation should be used but including all unique children benefiting for up to the first 3 years. For a new school building or initial teacher training, for example, the children benefiting would be all children using the new school building, or being taught by the newly trained teachers, in the first 3 years. Then a reasonable estimate of FCDO’s share of their education costs over this period would be estimated when relevant.^{[footnote 2]} For smaller one-off costs, not significantly larger than annual running costs, it would usually make sense to focus on the children benefiting in the first year.

Funding shares are often not the best estimate of FCDO’s attribution to a child’s education. In many cases there is evidence that FCDO’s impact is greater than just funding shares. In this case other data and evidence can be used to estimate FCDO’s attribution when available. Examples include a programme’s estimated impact on learning outcomes or on an education quality index, or on increasing education spend from others. There should always be a clear evidenced based rationale to back up the use of other data and evidence.

Data calculations

For FCDO education funding provided through the education budget, the calculation is:

N = [(1)/(2)] x 3

Key:

N = Number of children supported by FCDO

(1) = FCDO spend on education

(2) = Total government and donor spend on education in relevant sector(s)

(3) = Total number of children enrolled in relevant sector(s)

In some countries household contributions to education may be large, which will overstate FCDO’s share of education spend. Unless exceptional circumstances, these contributions are excluded to ensure consistency between countries.

If FCDO is supporting private or NGO education we should use the same calculation, but restrict the total spend and enrolment to the private or non-government schools being supporting by FCDO. If we are funding 100% of the cost of the pupil through vouchers, we can count each pupil funded.

For education funding provided outside the education budget where FCDO is clearly providing the majority of funding or learning experience

We count all children where we have evidence that FCDO’s funding is clearly providing the majority of a child’s education funding or learning experience (at least approximately 75%). In this case these children can be regarded as being sufficiently supported to gain a decent education. A detailed calculation would not be proportionate, and all children can be counted. This approach can be used when there is not enough financial information to calculate full funding shares, but we have evidence that FCDO (or FCDO plus other donors) are covering the majority of a child’s education. This situation includes children who would not be in school or hardly learning at all without FCDO support, even if we do not fully fund their schooling. It also includes children supported to attend part-time schooling (such as in humanitarian responses) if they would not be learning at all without this support. This is partly a subjective judgement, and borderline cases should be agreed with the FCDO Girls’ Education Department to help towards greater consistency.

For FCDO education funding provided outside the education budget, the standard calculation would be the total number of children benefiting in some way from FCDO support multiplied by the percentage of their education experience attributed to FCDO. In most cases the attribution would be based on funding, and the calculation would be:

N = (1) x {(2) / [(1) x (3) + (4)]}

Key:

N = Number of children supported by FCDO

(1) = Number of children benefiting from FCDO intervention

(2) = FCDO spend on the benefiting children

(3) = Estimate of national unit cost for the children

(4) = FCDO (and other donor) spend. Usually equals (2)

The calculation above assumes that FCDO spend (and, if relevant, other donor spend) is additional to the national unit cost spend. When this is not a reasonable assumption (eg if FCDO spend is replacing some of the government spend) then the FCDO spend in the denominator (2) would be removed and the calculation just becomes FCDO spend divided by the unit cost: (2)/(3) = FCDO spend / unit cost.

The unit cost data (3) would normally be the standard government unit cost (eg national education budget divided by the number of children enrolled for given level of education). The data should be as close as possible to the timeframe, school level, type of school and geographical area as the children benefiting. However, out of date data or data for a different area or level can be used if needed (for example, usually only primary and secondary breakdowns should be sufficient). Partial unit cost data can be used when full standard unit cost data isn’t available (for example, just using donor funding).

Children covered by cash transfers should be included when there is an explicit focus on education, such as a condition on school attendance. When possible, we should count a proportion of these children based on funding shares (as above), treating the cash transfer as additional funds for education.^{[footnote 3]}

Alternative methods

Other methods can be used to estimate the FCDO attribution to a child’s education, such as based on an estimate of FCDO contribution to a child’s learning outcomes or indicators of the quality of education, if appropriate. In this case the basic calculation would be the same:

N = (1) x (2)

Key:

N = Number of children supported by FCDO

(1) = Total number of children benefiting from FCDO intervention

(2) = Proportion of the chosen indicator of education for the benefiting children attributed to FCDO

The chosen indicator of education (eg learning outcomes, quality index) should be as closely related to a child’s overall education as possible (eg learning outcomes should be as broad as possible given available data).

We can also include indirect results, where there is clear evidence that FCDO interventions have increased the education budget or influenced the programmes of others. As with other approaches, care should be taken to ensure that the results of influencing can be attributed to FCDO alone (with estimated proportions taken when achieved with others), and that it focuses on changes to the education of benefiting children compared to the status quo. Any results which FCDO is not delivering directly should not be claimed for more than one year. However, in most cases it will not be possible to quantify these indirect impacts and attribute to FCDO.

These alternative methods should be agreed with the Girls’ Education Department to ensure broad consistency across programmes.

Preventing double counting

Care should be taken to only count each child once. Smaller programmes should be excluded if there is a potential to double count children covered by larger programmes, unless the overlap can be accounted for (see below). This should consider pupils enrolled in more than one school (eg both a public and private school). However, if 2 programmes are only partly funding the same children, then the results can be added assuming the part funding calculations do not equal more than 100%.

Overlapping programmes or programme components

It is possible that posts will contribute to this indicator through several different programmes. Posts should provide disaggregation information showing the number of people supported through each programme as well as the overall contribution to this indicator. For example:

• beneficiaries of programme A
• beneficiaries of programme B
• total number of individual beneficiaries = C

Note that C does not necessarily equal A plus B if some individuals are beneficiaries of both programmes. If this is the case, you will need to estimate the overlap in the way that is most appropriate to your country context and based on your professional expertise. There is space in the results template to record your assumptions and disaggregated results.

Programmes may have separate components or interventions reaching the same beneficiaries. This creates a risk of double counting within a programme if beneficiaries from each component or intervention are added together. To reduce this risk, the approach is the same as for overlapping programmes: estimate the overlap in beneficiaries and reduce the total number supported by the number of beneficiaries shared by each component or intervention.

It is also possible that new programmes may operate in the same schools or areas as programmes that closed prior to April 2020 and which contributed results to the previous results reporting period (2015 to 2020). Where this is the case, you will need to estimate and account for the extent to which the same children have been supported by the old and the new programme(s). You may want to consider average school years attended by girls or boys in the country or province to estimate the overlap of cohorts between programmes.

Current programmes that also reported results in the period 2015 to 2020

Some current programmes will have submitted data when this result was calculated for the period 2015 to 2020. If you can identify with confidence the exact period covered by a programmes previous result; and can with confidence report only new results since then, record the new result in the results template and note in the appropriate place the period covered by your results return. For programmes that cannot with confidence identify the period covered by the previous results return, you must ensure you only return new results reported from April 2020. Whilst potentially creating a small underestimate, to eliminate the chance of double counting across the previous and current reporting periods, only report new results from April 2020.

For programmes spanning both the previous and current results reporting period and using the peak year method, perform the calculation. If the peak year fell before April 2020, then record the current result as zero and explain this in the Methodology Notes. If the peak year fell between financial year 2020 to 2021 and the present, then record the new result and note the previous result in the appropriate place in the Methodology Notes section of the results template.

Calculating the total result across years

The final numbers of children supported would be the sum of unique children supported from each education programme in a country.

When a programme is supporting the same children across different years, the peak year should be used to prevent double counting. However, additional children can be added to the peak year when it is easy to identify additional unique children in other years. Different peak years can be used for different programmes in the same country.

Data sources

Calculations would normally be done in whatever currency the government uses to reduce the impact of exchange rate differences.

The calculations should use the available spend and enrolment data that mostly closely aligns with FCDO’s programme in terms of school level and type and geographical area, and to the relevant FCDO financial year (April to March). However, there would normally be no need to pro-rata across years, types, areas or levels if there is some mismatch.

FCDO spend data can be found in Hera or AMP. This should include general budget support, education sector budget support, education projects and financial aid, and general projects and financial aid that include support to education.

Partner country expenditure data can be sourced from government systems (Ministry of Education or Ministry of Finance). For some countries the UNESCO Institute of Statistics (UIS) database may have data not available elsewhere. All relevant development partners’ education spend should be included in the calculation wherever possible, even if these are not going through the government budget. Actual spend (expenditure) rather than budget figures should be used whenever possible.

Similarly, unit cost data (eg average cost per child in primary, or secondary school) will usually be available from the Ministry of Education or can be derived from the government budget and enrolment figures. Estimates might also be available from UIS or in country multilateral organisations.

Data for the number of children enrolled should be taken directly from country Education Management Information Systems (EMISs), or from project-specific enrolment data. Where EMIS data includes enrolment in non-government funded schools, care must be taken to adjust total enrolments accordingly if needed.

For projects, enrolments and expenditure data should be available from project monitoring reports. For enrolment, this should ideally follow a similar methodology to the national EMIS to support comparability between countries and projects. Care should be taken to adjust according to the FCDO share of the project or programme if relevant. If children are only funded for a proportion of a full academic year, only that proportion of children should be counted.

Government expenditure, unit cost and enrolment data is also available from the UNESCO Institute of Statistics (UIS), but it takes up to 2 years for national data to be collected and processed by UIS. In addition, the data are then presented according to the International Standard Classification of Education (ISCED), which may not align to national definitions. Hence national expenditure and enrolment data is preferable.

Reporting roles

Posts and central teams select the most relevant data, perform the calculations, and provide results returns in response to central commissions, presently issued by Girls’ Education Department (GED). The lead GED statistics, results or evaluation adviser will then quality assure and prepare the results for aggregation, onward reporting, and use.

Worked example

Example 1

FCDO is providing £40 million a year on education sector support focused on primary education. The government expenditure on primary education is £800 million (including donor spend). Ten million children are enrolled in government primary schools.

The estimated proportion of pupil costs funded by FCDO is therefore 5% (=40/800), and the total number of children supported by FCDO is 500,000 (0.05 * 10 million).

Example 2

FCDO is supporting an education system strengthening programme which provides technical support to the Ministry of Education to improve the quality of teaching and curriculum reform in primary schools. The programme reaches 1 million children and costs £20 million a year. The standard government unit cost is £60 per child per year in primary school.

The estimated total education spend for these children is £80 million (= 1 million * 60 + 20 million). Hence FCDO is providing 25% (=20/80) of their total education costs. The total number of children supported by FCDO would be 250,000 (= 1 million * 0.25).

Example 3

FCDO has a £10 million programme to support 500,000 children to improve learning outcomes. A Randomised Control Trial demonstrates that the programme increases average reading and mathematics learning outcomes scores from 400 to 500. No other subjects were tested, and we have confidence that this is being replicated across the programme.

Hence FCDO’s contribution to our best estimate of learning outcomes is 20% (= [500-400]/500), and the total number of children supported by FCDO is 100,000 (= 500,000*20%).

Baseline data

Achieved programme results are reported from financial year 2020 to 2021 onwards. Some current programmes will have submitted data when this result was calculated for the period 2015 to 2020. These programmes must ensure they only return new results. For programmes that can identify with confidence the period covered by that programme’s previous result, new results may cover the period prior to April 2020. See the section Preventing Double Counting for further guidance.

Within GEC-R, results from financial year 2020 to 2021 will be added to DFID/FCDO results from the period 2015 to 2020.

Data dis-aggregation

Disaggregate results by:

• sex (female, male, unknown)
• disability (yes, no, unknown)
• school level (pre-primary, primary, secondary, unknown)

Only include disaggregation if you have supporting evidence, eg EMIS data sources. Do not assume splits for disaggregated categories (eg assuming a 50/50 split for sex).

Time period or lag

Time lag may vary. Where the time lag of your data is longer than one year, please detail the reason why in the methodology notes section of the results template and report results for the latest year you have.

Quality assurance measures

International data are quality assured by the UNESCO Institute of Statistics; partner country data and programme data will be subject to their own quality assurance arrangements put in place by the partner country/implementer.

Departments, Posts and SROs should use the best quality data they can find in their context. Data quality can vary substantially. When data quality is an issue, such issues should be reported in the results template.

There are 3 layers of FCDO quality assurance (QA) as part of this commission.

1. Statistics, results or evaluation advisers at Post or in reporting Departments quality assure results prior to submission and include supporting evidence for calculations. The results return is signed off by the PRO or SRO for each programme.
2. Statistics, results or monitoring, evaluation and learning (MEL) advisers in GED check results returns and calculations, follow-up with clarifying questions where required, record any issues in a QA log and resolve them with reporting teams.
3. Cleaned results are then confirmed with Posts and departments prior to any publication of country/department disaggregated results.

Methods must be documented sufficiently to enable quality assurance by the central team, using the Methodology Notes section of the return spreadsheet and by submitting a separate calculations spreadsheet or results note.

Central aggregation of results

The lead GED statistics, results or evaluation adviser will quality assure and then aggregate results from all Posts and Departments. During this process, the adviser will develop an approach to account for double counting between bilateral and multilateral programmes that operate in the same countries. This approach is being refreshed at the time of writing.

Results and accompanying documents will then be prepared and submitted to FCDO’s Chief Statistician for approval ahead of any publication.

Go back to about results estimates for children the FCDO supported to gain a decent education. You can also find out more about UK government support for girls’ education worldwide.

1. Children are counted as attending pre-primary education if the support is consistent with UNESCO ISCED 0 definition. This consists of education programmes for children from the ages of 0 to primary school entrance; with an intentional education component; which aim to develop the socio-emotional and some academic readiness skills necessary for participation in school and society; which are conducted through semi-structured group learning (usually based in a school or other institution – it excludes family based arrangements); and where education programmes are at least the equivalent of 2 hours a day for 100 days a year. ↩

2. In cases where the one-off expenditure is far larger than the ongoing yearly unit costs, it may be appropriate to just count all children benefiting, either in the first year or first 3 years. ↩

3. Although the funding is not all going to a child’s education, we still use this approach as an attempt to ensure some comparability between different levels of cash transfers (eg providing relatively small or large amounts), and with children funded in school. ↩