World Bank Water & Sanitation Topic Data

Most water panels make you choose: either the service-delivery story (who has taps and toilets) or the resource story (who is pumping what, from how much). This topic refuses the choice. It gathers every World Development Indicators series tagged to the water sector into one panel - and the span is the selling point.

Four families ride together. Access and service ladders carry the JMP-derived mnemonics: basic and safely managed drinking water (SH.H2O.BASW.*, SH.H2O.SMDW.*), basic and safely managed sanitation (SH.STA.BASS.*, SH.STA.SMSS.*), hygiene and handwashing (SH.STA.HYGN.*) and deaths attributable to unsafe water and sanitation (SH.STA.WASH.P5). Resource and withdrawal series sit under ER.H2O.* - total withdrawals split by agriculture, municipal and industrial use, renewable internal freshwater resources, and water stress expressed as withdrawal against available supply. Water productivity (ER.GDP.FWTL.M3.KD) prices a cubic meter in output terms. And the money side arrives as private participation in water and sanitation (IE.PPI.WATR.CD, IE.PPN.WATR.CD).

Provenance travels per series: the access ladders trace to the WHO/UNICEF Joint Monitoring Programme, the withdrawal and resource figures to FAO AQUASTAT. Get a sample of this dataset and start with the stress series - it is where the two halves of water data finally meet.

Each row is one country, one indicator, one year. The five series below lead the topic's own featured set, quoted with their mnemonics exactly as reported:

SH.H2O.SMDW.ZS : People using safely managed drinking water services (% of population)
ER.H2O.FWTL.K3 : Annual freshwater withdrawals, total (billion cubic meters)
ER.H2O.INTR.PC : Renewable internal freshwater resources per capita (cubic meters)
ER.H2O.FWST.ZS : Level of water stress: freshwater withdrawal as a proportion of available freshwater resources
IE.PPI.WATR.CD : Investment in water and sanitation with private participation (current US$)

One record's assembled shape, built from each field's documented example so the layout is visible:

country_name   : Brazil
country_code   : BRA
indicator_name : People using safely managed drinking water services (% of population)
indicator_code : SH.H2O.SMDW.ZS
year_columns   : 1960 ... 2025  (blank where no observation exists)

Two things worth noticing before any modelling starts. Units live inside the indicator name itself - percent of population, billion cubic meters, cubic meters per capita, current US$ - so unit errors surface at read time rather than in a review comment. And blank cells are genuinely empty rather than zero: a missing 1997 is a gap in collection, not a claim that Brazil withdrew nothing.

Five structural fields carry every observation - deliberately boring, deliberately joinable:

The pair worth respecting is indicator code against year column. The code is the contract: SH.H2O.SMDW.ZS means exactly one thing everywhere it appears, which makes a sample and a scheduled full delivery diff cleanly. The wide year layout is what you will reshape - every analyst eventually unpivots it into a long country-year spine, and requested samples can arrive already melted if that is where the work lives.

• Geographic: about 200 countries and economies, with the World Bank's regional and income aggregates riding alongside in the same table.
• Temporal: year columns back to 1960 wherever observations exist, with recent reference years running through 2025 depending on the series.
• Granularity: one annual observation per country per indicator - a clean three-key spine with no composite rows to unwind, and no sub-national detail anywhere in it.
• Volume: tens of thousands of populated country-indicator-year rows across the water-related series; the adjacent environment slice of the same catalogue ships roughly 39,000 rows across 147 indicators, which gives the sense of scale.

Read as a product: this is the broadest thematic span in the water utilities shelf. Six decades of annual observations support things shorter panels cannot - sanitation-ladder trajectories by decade, stress convergence across income groups, the investment cycle around privatisation waves - without stitching vintages together first.

API, files, or your warehouse. Daily, weekly, or hourly.

Market screening and prioritisation. Water stress, per-capita renewable resources and withdrawal mix give every economy a demand-and-scarcity picture on one spine. A sizer ranking candidate markets for treatment, irrigation-efficiency or metering products gets the shortlist arithmetic done in an afternoon - and defends it, because each input keeps its originating organisation attached.

Cross-country benchmarking. Service-ladder percentages apply identical definitions across all economies rather than national statistics patched at the borders, so the safely managed drinking water ranking of thirty countries holds up in front of a board. Stress ratios compare withdrawal pressure against actual supply instead of raw volumes that flatter big, wet countries.

Investment and ESG analysis. Private participation in water and sanitation, tracked in current US$ alongside the physical indicators, puts a capital-flows series next to the service gaps it is meant to close - useful for screening where utility concession activity concentrates and whether it moved the ladder.

Research, teaching and journalism. These are the numbers SDG 6 reports quote: who reached basic sanitation and when, where handwashing facilities lag, which economies withdraw more than their renewables replenish. Writers and lecturers get citable series with named origins instead of screenshots.

Development economists and market researchers get the widest thematic span per join in the shelf - one country code unlocks access, resources, stress and investment simultaneously, which is why it anchors water utilities market research work. Data scientists and ML engineers treat it as a source of slow-moving macro features - stress ratios, ladder percentages, per-capita resources - joined onto firm-level or transaction-level data by country code and year, as in water utilities data science. Policy researchers, NGO program teams and sustainability reporting analysts round out the pool, using the JMP-custodied ladders as the citation-grade backbone for program baselines and disclosure annexes.

Within the shelf this is the breadth anchor, answering a different question from everything beside it. World Bank - Safely Managed Drinking Water Services Indicator (SH.H2O.SMDW.ZS) goes deep on one access series with rural and urban splits; widen it with this panel when withdrawals, renewable resources, stress and investment flows enter the question - the two share one country-keying scheme, so the join is trivial. EEA Data and Maps - Water runs European, environmental and station-grained, strong on water quality where this panel stays at national annual grain. The head-to-head sits in EEA Data and Maps - Water vs World Bank Water & Sanitation Topic Data. Prototype on the granular sources; cite this one when a cross-country number has to survive scrutiny. The full ordering - this panel ranks fifth of ten with a quality score of 8 - sits on best water utilities datasets.

Stated plainly, because each one shapes an analysis:

• Annual grain only. One observation per country per indicator per year - nothing here resolves a month, a season or a service outage. Pair it with a high-frequency source when intra-year dynamics matter.
• No sub-national detail. Utility districts, river basins and cities stay invisible; every figure rolls up to the national economy or a named aggregate.
• Countries and aggregates share a spine. Regional and income aggregates ride alongside national rows, so unfiltered totals double-count unless the aggregate codes are dropped first.
• Recency varies by series. Year columns reach 2025, but individual indicators stop where their collection stops. Take the latest non-null year per series as the freshest fact, never the calendar year.

None of these arrive as surprises at delivery time. Samples ship with the caveats attached, and Datadory flags which analyses they bite before anyone commits to one.

Because the interesting cut is rarely all water series by all economies. It is stress and withdrawals for a drought-exposed watchlist; sanitation ladders plus private-participation investment for twelve programme countries since 2000. Requested samples come back shaped to that question - wide or long, aggregates filtered or kept - with the original indicator codes intact so the sample and the scheduled delivery diff cleanly. Scheduled drops land keyed on country code and indicator code, so consecutive passes append onto one spine instead of spawning versioned spreadsheets. The full catalog holds the rest of the water utilities shelf beside it.