How scores work¶
Every figure in the portal comes from the Global Database maintained by the GPEDC Joint Support Team. This page explains what the scores measure and how to read them.
For the full technical specification — every indicator formula and aggregation rule — see the Methodological Note.
What is being measured¶
The monitoring framework tracks the commitments governments and development partners have made to each other about how development co-operation should work. It is built around a questionnaire completed by partner country governments, development partners, civil society, trade unions and the private sector, with a few measures drawn from outside sources.
Results are produced for each partner country, each development partner, and for the partnership as a whole.
The four dimensions¶
| Dimension | What it asks |
|---|---|
| 1 — Whole-of-society engagement | Are parliament, civil society, trade unions and the private sector brought into decisions about development? |
| 2 — Country systems | Does development co-operation use national plans, budgets and financial systems, and arrive when promised? |
| 3 — Transparency | Is information about development co-operation published, on time, and usable? |
| 4 — Leaving no one behind | Are the people most at risk of being left behind consulted, targeted, and counted? |
Each dimension contains several components, and each component one or more indicators.
Reading a score¶
Most indicators run from 0 to 1 and are shown as a percentage. Higher is always better — no indicator is reversed, so a country scoring 80% on one measure and 30% on another is doing better on the first.
Scores fall into five bands, which set the colour you see on the map and on every chart:
| Band | Range |
|---|---|
| Very high | 90% and above |
| High | 70–90% |
| Medium | 50–70% |
| Low | 10–50% |
| Very low | Below 10% |
Composite indicators¶
Some indicators are marked as composites — 1.1.C, for example, summarises 1.1.1, 1.1.2 and 1.1.3. A composite is the simple average of its parts.
A composite is only reported when more than half its parts have data. A composite of five parts needs at least three. This lets partial reporting still produce a result, without publishing a summary resting on one or two answers.
Indicators that are not percentages¶
A few measures are reported as words rather than numbers, because a percentage would misrepresent them:
| Indicator | How it is reported |
|---|---|
| 2.3.2 — Strengthening of PFM systems | Improvement, No change or Decline — a direction of travel between two assessments, not a level |
| 2.5.1 — Gender budgeting | Fully meets requirements (three criteria), Approaches requirements (one or two), Does not meet the requirements (none) |
| 2.6.1 — Accountability mechanisms | In place or Not in place |
| 3.2.1 — Reporting to global systems | Excellent, Good, Fair, Improvement needed or Not reporting, given separately for two reporting standards |
These are shown as labels throughout the portal, and are kept out of the dimension averages, which pool percentage scores.
Where a dash means something¶
A dash means nothing was reported — never a score of zero. Missing answers are left out of every average and comparison rather than counted as failures.
There are several reasons a figure can be absent:
- The country or partner did not report on that question.
- The measure does not apply. Untied aid is published for members of the OECD Development Assistance Committee only; the SDG 17.16.1 indicator applies to partner countries and bilateral partners.
- The module was optional and not taken up. The Kampala Principles assessment on private-sector engagement is voluntary.
- The measure needs two assessments and only one exists. This is why 17 partner countries have no PFM trend.
Where a whole section is missing, the portal says why rather than leaving a gap.
Measures that come from outside the questionnaire¶
Four measures are compiled by other institutions and brought into the framework:
| Source | Used for |
|---|---|
| PEFA assessments | Quality of public financial management systems (2.3.1) and the direction of travel (2.3.2) |
| OECD Creditor Reporting System | Untied aid (2.8.1), and inputs to the predictability measures |
| IATI standard | Reporting quality to global systems (3.2.1) |
| SDG monitoring | Indicators shared with the Sustainable Development Goals (2.2.1 / SDG 17.15.1, 2.5.1 / SDG 5.c.1) |
Because these are maintained outside the Global Partnership, their availability depends on the institutions that produce them, and each keeps its own reporting years — which is why untied aid and the transparency ratings are labelled with different years from the rest of the round.
Who reports what¶
Two things can look like the same indicator but be reported by different people, and it matters which:
- Partner country measures describe the country — its national plan, its budget process, its consultation of civil society.
- Development partner measures describe a partner's practice — whether it uses country systems, publishes its strategies, disburses on schedule.
- Some measures are reported by one side about the other. Perception questions in the Kampala Principles module, for instance, record what respondents think about private-sector engagement in general, not a rating of the organisation whose profile they appear on.
The portal labels each measure with whose practice it describes.
Comparison groups¶
No score is shown against a single global yardstick alone:
- Partner countries are compared with others in the same World Bank income group — low-income, lower-middle-income, upper-middle-income.
- Development partners are compared with their comparison group: DAC members, multilateral partners, or United Nations entities.
A partner reporting in fewer than five partner countries is shown without a group comparison. Its figure would rest on too small a sample to sit beside partners measured across dozens of countries.