The numbers behind Pakistan’s basic necessities crisis are not hidden, they are published, repeatedly, by the same institutions the government cites when it wants to talk about macroeconomic stability. Taken together, they make the AI data centre pitch look less like ambition and more like a country skipping several chapters of its own story.

Start with water, since it is the resource discussed least and running out fastest. According to Arab News, Pakistan’s per capita water availability has fallen sharply from around 5,600 cubic metres in 1956 to less than 800 cubic metres today, a level well below internationally recognised scarcity thresholds. A report in The Nation puts the country among the ten most water-stressed nations in the world, noting that more than 60 percent of the population lacks access to clean drinking water and relies instead on contaminated supplies. This is not a distant projection. It is the water situation Pakistan is living in right now, in the same year officials are discussing water-intensive server cooling as an export industry.

Electricity tells a parallel story of a resource that looks abundant on paper and scarce in practice. Per Digital Pakistan’s breakdown of the energy sector, Pakistan has roughly 46,600 megawatts of installed generation capacity, yet still experiences load shedding alongside circular debt that had climbed to over Rs 2.6 trillion, close to 9.3 billion dollars, by mid-2025, with distribution companies routinely reporting technical and commercial losses exceeding 20 percent against a global efficiency benchmark closer to 3 to 5 percent. 

Government’s own response has been to link outages directly to bill recovery. As ProPakistani reported, the energy minister told the National Assembly in May that the government would move toward a transformer-level load shedding mechanism, under which areas with poor bill recovery would face outages while better-paying areas received improved supply. Even under normal conditions, STI’s power crisis tracker recorded shortfalls of around 4,000 megawatts during peak demand in 2026, driven by weather disruptions, fuel shortages and technical faults.

And the burden of financing that dysfunction is not shared evenly. As of April 2026, NEPRA-approved slabs show lifeline consumers paying as little as Rs 3.95 per unit, while unprotected domestic consumers using over 700 units pay up to Rs 47.69 per unit, with the same source noting that NEPRA approved an additional Rs 3.82 per unit surcharge from March through June 2026 following an IMF advisory to raise electricity prices. The national average uniform tariff stood at Rs 33.38 per unit as of January 2026. Even at the subsidised end, the base rate is layered with fuel price adjustments, quarterly adjustments and surcharges that push the effective bill well past the headline slab rate, meaning the poorest connections are still absorbing the cost of a system built to serve far larger, and far less price-sensitive, industrial and commercial demand.

Food security sits on the same fault line. The World Food Programme’s Pakistan country page puts 7.5 million people, 21 percent of the population it assessed, in crisis-level food insecurity or worse, including 1.25 million in emergency-level conditions, alongside around 40 percent of children under five who are stunted. The picture is sharper still in the provinces least visible to national policy conversations. The latest IPC malnutrition analysis for Balochistan, Khyber Pakhtunkhwa and Sindh found over 2.71 million children aged six to fifty-nine months currently affected by acute malnutrition, including roughly 706,000 cases severe enough to carry a risk of death.

None of this is an argument that Pakistan should abandon every ambition beyond subsistence. It is an argument about what a state’s stated priorities say about who it believes it is governing for. A government citing circular debt figures in the trillions, malnutrition figures in the millions, and water availability figures below internationally recognised scarcity lines, while simultaneously pitching subsidised power and water to foreign server farms, is not describing two unrelated policy tracks. It is describing one track, where the same scarce resources are being asked to do two incompatible jobs at once, sustain a population already short of them, and attract an industry whose primary beneficiaries sit elsewhere. Until the water, electricity and food numbers move in the right direction for the country’s own citizens, the data centre conversation is not a growth strategy. It is a distraction with a technology sector’s branding attached to it.