A power crisis of geography

Pakistan’s latest power shortage in the north is not really a story of inadequate generation. It is a story of policy failure. Cheap power exists. A large part of it just happens to be sitting in the wrong place.

The problem is increasingly geographic. Over the past decade, substantial generation capacity was added to the grid in the south, including nuclear, imported coal, and Thar-based plants, without a matching build-out of the transmission system needed to move that power efficiently to demand centers in the north. The result is absurd but predictable: consumers continue paying for surplus capacity that cannot always be fully evacuated when and where it is needed.

That contradiction was on display last week. Even during peak hours, several low-cost plants in the south were running far below potential, despite tariffs of roughly Rs14 per unit or less. At the same time, expensive furnace oil generation in the north was reportedly being run hard to compensate for RLNG shortages, while hydel was drawn down under pressure to contain public anger. A system with surplus installed capacity ended up behaving like one under shortage. That is not a generation problem. It is a planning problem.

There is little point pretending this can now be fixed quickly through a late transmission push. If a major mid-country addition such as Chashma-5 comes online by 2030, some of today’s grid stability and regional load-balancing issues may ease on their own. Spending billions of dollars now on transmission assets that may only become operational around the same time deserves scrutiny, not blind repetition of old planning reflexes.

That does not mean the current situation should be accepted. It means the policy response must shift. Pakistan should stop thinking only in terms of moving electricity to demand and start thinking more seriously about moving demand to electricity.

The south already hosts a significant share of the country’s underutilized low-cost power base. If that power cannot be transmitted north efficiently, then energy-intensive industry should increasingly be encouraged to locate closer to it. This is the more rational long-term adjustment. Consumers are already paying capacity charges on idle southern plants. The least policymakers can do is ensure that this stranded efficiency is converted into productive economic use.

That requires industrial policy rooted in comparative advantage rather than administrative habit. Karachi and the broader southern corridor, with port access and relatively cheaper available power, should be the natural home for energy-intensive industry. If Pakistan ever wants to attract sectors that consume large and stable baseload power, this is where they belong. By contrast, labour-intensive manufacturing should be pushed toward regions where labour supply is abundant and land costs are lower, including parts of South Punjab. A garment cluster there makes far more sense than endlessly reproducing the same geography of concentration and congestion.

Pakistan does not need more vague talk about industrialization. It needs a different map. One cluster should be built around energy logic. Another should be built around labour logic. The country has spent too long pretending that growth can be organized from Islamabad while ignoring the hard economics of location, fuel, logistics, and grid structure.

The recent shortage in the north should therefore be read as a warning. It shows what happens when generation planning, transmission planning, fuel planning, and industrial policy are all handled in silos. The state added capacity without fully solving evacuation. It concentrated demand without aligning supply. It now asks consumers to finance both the excess and the inefficiency.

This is why the crisis is ultimately one of policy, not capacity. Pakistan has already paid for much of the power it is failing to use well. The next step is not merely to produce more electricity, but to finally organize economic activity around the system it has actually built.

Copyright Business Recorder, 2026