The Water Security Index is a cross-national water reference. It takes public figures about nations, organizes them by nation, and compares them across nations. That's the whole product: a number, a rank, a source, and a date. We present the data and stop. You draw the conclusion.
This is the idea the whole site is built to get right. A country can be water-abundant in total yet water-stressed per person, because the total is shared among many people. It can be water-scarce by nature yet water-secure in access, because it is rich enough to desalinate seawater. No single figure captures all of that, so we don't pretend one does. We show three true measures side by side and label each for the question it answers.
Total renewable freshwater counts the cubic metres of water a country's rainfall and internal rivers generate each year. It is a real fact, but it mostly measures how big and wet a place is — Brazil, Russia and Canada top it, which tells you little about whether their people have enough. Renewable freshwater per capita divides that resource by the people who depend on it, and reorders the list completely; this is the scarcity spine, read against the Falkenmark thresholds below. The water-stress ratio (SDG 6.4.2) is the demand side: withdrawals as a share of available renewable supply, which can pass 100% where a country desalinates or mines fossil groundwater. None of these is the "right" one. They count different things, and every view says which.
The per-capita figure becomes legible against a standard yardstick. Below 1,700 cubic metres per person per year, a country is said to face water stress; below 1,000, water scarcity; below 500, absolute scarcity. These are conventions, not laws of nature — a wealthy, efficient country below 1,700 may feel no strain, while a poor one above it may — but they give the number a frame, and the count of countries beneath each line is a clean, citable fact.
The scarcity-spine measures here count internal renewable water: the water generated inside a country's own borders. They deliberately exclude inflow from upstream neighbours. This matters enormously for some countries. Egypt generates almost no water internally yet depends on the Nile, which arrives from upstream; on internal water alone it looks catastrophically scarce. So the internal figure understates real availability for countries fed by transboundary rivers — but it captures their dependency risk, the water they don't control. We state this plainly wherever the internal figure appears.
To complete the picture we add two AQUASTAT measures beside the internal one. Total actual renewable water adds the net inflow from upstream — the rivers and aquifers that cross the border — to a country's internal generation. It's the fuller answer to how much water is physically there. The dependency ratio is the share of that total which originates abroad: 0% for a country living entirely on its own rainfall, near 100% for one living on someone else's river. Egypt's total actual supply is many times its internal supply, and almost all of the difference is Nile water arriving from upstream — which is why its dependency ratio sits near the top of the world.
Here is the rule we hold, the reason these are a Phase 2 addition rather than Phase 1: we never present the total actual figure without the dependency ratio next to it. Counting the water that crosses in from abroad and stopping there would over-credit a downstream country for a supply a neighbour could dam, divert or pollute first. The total says how much water reaches you; the dependency ratio says how much of it you control. Read together they're honest; the total alone is a flattering half-truth, and we don't ship it that way.
Availability is one half of water security; use is the other. AQUASTAT records how much water each country withdraws and what it withdraws it for — agriculture, industry, or municipal supply. We publish the agricultural share: the fraction of total withdrawal that goes to irrigation, livestock and aquaculture. Globally it's about 70%, and in the poorest farming economies it runs past 90%; the rich, industrial, service economies are where it falls away. That single number already draws the sharpest line on the page between the two kinds of country, and it sits directly on AQUASTAT's published total and agricultural figures.
We deliberately stop short of breaking the rest into industry versus municipal, and the reason is the same discipline that runs through the whole site. The sector volumes the source distributes don't reconcile with AQUASTAT's own published totals — for the United States they imply agriculture is 62% of withdrawal when AQUASTAT says 39.7%, because the industrial figure leaves out power-station cooling, the single largest non-farm use in much of the rich world. A three-way split built on those numbers would look precise and be wrong, and it would contradict the agricultural share we show right beside it. The old World Bank copies of the sector breakdown are no help either: they're frozen at 1987–2002. So we show agriculture against everything else, label the remainder for exactly what it is, and hold the industry-versus-municipal line until we can source it without it fighting the totals. Better an honest two-way split than a tidy three-way fiction.
The renewable-water and water-stress series originate with FAO AQUASTAT, the UN Food and Agriculture Organization's global water statistics. The drinking-water and sanitation access series come from the WHO/UNICEF Joint Monitoring Programme (JMP). GDP per capita comes from the World Bank. Most series we ingest through the World Bank's open data API, which redistributes them; the total actual renewable water and dependency-ratio figures we pull from FAO AQUASTAT's own dataset directly, carrying its native variable IDs and current vintages rather than the frozen mirror. Either way we stamp each figure with the body that actually produced it. Every figure carries its source and the exact year it was reported; the latest available year differs by country, so two rows in one table may carry different years, and we show each rather than force a single date.
A water-stress ratio over 100% means a country withdraws more than its renewable supply each year. There are two very different ways to do that. A desalination-rich Gulf state engineers its way past the natural limit by turning seawater into freshwater with energy and money — expensive, but sustainable as long as the energy lasts. A fossil-aquifer miner pumps ancient groundwater that does not refill on any human timescale — cheaper now, but spending a balance that won't be replenished. The number is the same; the futures are opposite. We label the cases rather than ranking them into a single "worst." Which countries are mining fossil aquifers that won't refill is now visible from space: see the water that doesn't refill.
A cross-country snapshot hides things. It can't see the difference between a wet region and a dry one inside the same country, and it freezes a moving picture into a single frame. Correlation is not causation, and we say so once, plainly, in the footer of every page rather than hedging every sentence. Where a relationship is largely a proxy for national wealth, we flag it on the page rather than trumpet it. Honesty about the limits is part of the product.