The work moment
You are drafting an annual results report. You want to cite national electricity access gains as evidence for your program. The stored figures for Uganda read:
Uganda, 2015: 18.5 percent. Proportion of population with access to electricity, by urban/rural (%). Nature: Country data. Source: World Bank.
Uganda, 2024: 55.3 percent. Proportion of population with access to electricity, by urban/rural (%). Nature: Modeled data. Source: World Bank.
The start is country data. The end is modeled data.
What the test showed
I asked Claude with the skill, with WebFetch and curl allowed. On Uganda, the quality note said:
The 2024 value is modeled. It rises 3.8 pp from the modeled 2023 value (51.5). The size of the change depends on the model until a new survey confirms it.
On the main question, it said:
Report these figures only as national context, not as an outcome of a program.
On attribution, it said:
Without a counterfactual, a single program can't claim any defined share of the change.
On what to use instead, it said:
To report an outcome, use the program's own data: verified connections or systems installed, how many are still working after 12 or more months, the service tier reached (Multi-Tier Framework), and survey results from beneficiaries compared against a baseline or comparison group.
The steps
The skill is analyzing-social-impact by CaseMark/skills, licence Apache-2.0. The question asks Claude to work through these parts:
- Name the series: EG_ACS_ELEC, total.
- Give the 2015 value and the latest year with data and its value for each country.
- Put everything in one table with the change in percentage points and a data quality note.
- Say what the indicator counts and what it does not count.
- Say whether the change can be reported as an outcome of a program or only as national context.
Copy-paste prompt
Using the SDG Global Database series EG_ACS_ELEC (Proportion of population with access to electricity, total), give me the 2015 value and the latest year with data and its value for Kenya, Rwanda, Ethiopia, Uganda and Bangladesh. Put everything in one table with columns for country, 2015 value, latest year, latest value, change in percentage points and a data quality note. Under the table, say what this indicator counts and what it does not count. Also say whether the change can be reported as an outcome of a program or only as national context.
Sources
- Proportion of population with access to electricity, by urban/rural (%)
- Proportion of population with access to electricity, by urban/rural (%)
- Proportion of population with access to electricity, by urban/rural (%)
- Proportion of population with access to electricity, by urban/rural (%)
- Proportion of population with access to electricity, by urban/rural (%)