Farming, population and water stress in sub-Saharan Africa
One published global dataset, taken from a vague interest in farming's water use to a question, a real-world strategy, a counted sample of 20 countries and a chosen test.
What is the relationship between population growth (%) and the change in the agricultural sector's contribution to water stress (percentage points) across the countries of sub-Saharan Africa, 2005 to 2020?
- 1Issue
- 2Research question
- 3Strategy and tensions
- 4Your dataset
- 5Treat the data
- 6Analysis
- 7Evaluation
- 8Write-up
Find your issue, and check it is measured
An environmental issue: something happening, to something, because of something, with proof that somebody measures it.
Across sub-Saharan Africa, farming takes a growing share of the freshwater that countries renew each year. Where populations grow fast, the demand for food may be pushing that share close to what rivers and aquifers can replace.
- Proof it is measured
- The FAO AQUASTAT bulk CSV, opened, not just found. Indicator: Agricultural sector contribution to water stress (%): the water farming takes, as a share of the freshwater a country renews each year. It is one of the UN's Sustainable Development Goal indicators (6.4.2). One row is one variable, for one country, in one year. Worldwide, annual, to 2022.
- In the news
- UN News, 12 December 2025, reporting FAO's 2025 AQUASTAT Water Data Snapshot: agriculture accounts for about 72% of freshwater withdrawals worldwide. In sub-Saharan Africa, only a small part of the land farmed is irrigated.
- The system
- Stores: rivers, lakes and aquifers, such as the Niger and its Inner Niger Delta. Inputs: rain, and inflow from upstream countries. Outputs: agricultural, industrial and municipal withdrawal, and evaporation. The pressure: population growth raises the demand for food, and that raises irrigation. Taking out more water than is renewed runs the stores down.
Write your research question
One sentence naming the indicator and its units, the years and the places. No strategy in it.
What is the relationship between population growth (%) and the change in the agricultural sector's contribution to water stress (percentage points) across the countries of sub-Saharan Africa, 2005 to 2020?
- Independent variable
- Population growth, % change 2005 to 2020.
- Dependent variable
- Change in the agricultural sector's contribution to water stress (farming's withdrawal as a share of renewable freshwater), in percentage points, 2005 to 2020.
- The test sentence
- "Does the agricultural contribution to water stress change when population growth changes?" The one that responds is the dependent variable.
Find a real-world strategy and its tensions
One real-world strategy people are genuinely arguing about, two named positions, and what the tension does.
The African Union's Comprehensive Africa Agriculture Development Programme (CAADP), renewed by the Malabo Declaration in June 2014.
- The strategy
- The Malabo Declaration recommitted African Union members to CAADP, with seven commitments including ending hunger and halving poverty by 2025. It is run by the African Union Commission and its development agency, AUDA-NEPAD. It covers every country in the sample.
- Position 1 · economic, political
- The African Union, in its Framework for Irrigation Development and Agricultural Water Management (2020): rain-fed farming leaves food production exposed to climate shocks, so irrigated area has to grow.
- Position 2 · environmental, social
- Wetlands International (2016): Mali plans to add 330,000 ha to the Office du Niger, its state irrigation scheme. The plan depends on the Fomi dam upstream in Guinea. The dam would cut the water reaching the Inner Niger Delta by 23%, and cut fish trade by 31%, pasture by 28% and floating rice by 37%.
- The tension
- Feeding more people against leaving water in the rivers. A cubic metre taken out upstream never reaches the delta, so the same water cannot meet both goals. Both positions are about expanding irrigation, the part of CAADP this investigation measures.
- Link to the question
- CAADP's case for irrigation is that a growing population needs more water for farming. The question tests whether faster population growth went with a rising agricultural share of water stress.
- What it does
- Uneven costs: farmers who irrigate gain the harvest. Fishers and herders downstream lose the flood. Delay: the 4th CAADP Biennial Review found the continent largely off track for the 2025 targets.
- For the environment
- A smaller annual flood reaching floodplains and wetlands, and groundwater drawn down where farms pump it.
Select and document your data
A dataset you can defend: counted, documented, and cut to a rule fixed before you looked.
- Variables
- AQUASTAT codes 4568 (SDG 6.4.2. Agricultural Sector Contribution to Water Stress [%]) and 4104 (Total population [1000 inhab]). The bulk file is 88 MB and 936,332 rows. When you import it, choose Latin-1 as the text encoding (not UTF-8), and delete the duplicated year and element columns.
- Filters
- Sub-Saharan African countries as defined by the UN's standard list of regions (M49). Select them by numeric area code: this also drops AQUASTAT's own regional rows. Years 2005 and 2020 only.
- Inclusion rule
- Both variables present in both years, and the water stress value not identical in 2005 and 2020. AQUASTAT fills gaps by copying the last value forward, so an unchanged value means nobody measured it again, not that nothing changed. Written down before anything was plotted.
- How the numbers were made
- Every water stress value is estimated (E) or imputed (I), meaning filled in by calculation; none in the region is marked official. Population carries flag X, taken from the UN's estimates rather than measured by FAO.
- Controls
- Normalise: growth as a percentage, stress as a share of each country's own renewable water. Restrict: one region. Pair: each country is compared with itself.
- Credit
- FAO, AQUASTAT, CC BY 4.0, with the year of the last update and the date accessed.
Treat your data
Run a statistical test, leading to figures you built and a statistical result.
Spearman's rank correlation, two-tailed, at p = 0.05.
- Why this test
- A relationship between two measurements on the same 20 countries. The change in water stress is skewed (most countries near zero, a few moving a lot), so a rank test is safer than Pearson's. Spearman's needs at least 7 pairs. On this route the guide asks for at least 10 for a correlation.
- Decided before running it
- Two-tailed. Critical value of the coefficient about 0.45 for n = 20 (check it in the table you cite). Tied values take the average of their ranks. Free calculators: Social Science Statistics, Geography Fieldwork, Wessa. Run it through two and check they agree.
- Derived quantities
- Both variables are calculated rather than downloaded: (pop 2020 minus pop 2005) divided by pop 2005, times 100; and stress 2020 minus stress 2005.
The test has not been run and the scatter has not been drawn. Write the hypothesis first, then plot, then test. After that come step 6 (what it means), step 7 (most of the evaluation is about the E and I flags) and step 8.
The sources
Every one is public. The line under each says what it shows in this investigation. Follow the link to the original before citing it.
This outline plans an investigation rather than reporting one. The counts come from the FAO AQUASTAT bulk file as downloaded on 23 September 2026 and may change when FAO revises it. The relationship between the two variables has not been calculated. The choice of region, years and inclusion rule is the author's, not FAO's. The summaries of each source are also the author's own: follow the links to the originals before citing them. Checked 23 September 2026.