Prime Minister Shehbaz Sharif’s meeting on agricultural reform on Saturday dealt with wheat demand, cultivated area and projections for the coming crop. He approved changes at the National Agricultural Research Centre, asked for the Pakistan Agricultural Research Council to meet international standards, directed certified companies to ensure quality seed, and ordered a national system for estimating crop production. That last decision reaches beyond a bureaucratic exercise. Pakistan has too often tried to manage harvests after they have ripened, entered godowns or unsettled the market.
The wheat episode of 2024 explains why. Pakistan imported roughly 3.4 million tonnes of wheat to cover an anticipated shortfall. Government figures later put the 2024-25 harvest at a record 31.4 million tonnes, as a result of which farm-gate prices weakened, and growers were left carrying costs already incurred on fertiliser, diesel, credit and labour. A farmer chooses seed and acreage months before a food department fixes procurement or a ministry considers imports. Crop estimates therefore carry a direct bearing on the price offered at the mandi.
The FY26 figures offer little comfort. Agriculture grew 2.89 per cent, yet crop growth stood at 1.44 per cent. Wheat output reached 29.61 million tonnes after the area under cultivation increased 4.4 per cent to 9.48 million hectares, leaving yield per hectare effectively flat. The country has added land to wheat without securing the productivity gains that research, seed quality and extension services are meant to deliver.
Water makes the failure more expensive. Surface-water availability in FY26 stood at 11.1 per cent below average system use. Domestic edible-oil production meets around one-tenth of national demand, while the annual import bill is projected near $6 billion. Research priorities cannot continue to be dictated by procurement habits and entrenched crop lobbies when water is scarce, and foreign exchange is under pressure. Oilseeds, pulses, fodder, climate-resilient cotton and livestock genetics deserve the same policy seriousness accorded to wheat.
The latest agricultural census counted 11.7 million farm households, with the average holding down to 5.1 acres from 6.6 acres in 2010. On a farm of that size, poor seed, a pest outbreak or a missed irrigation turn can consume the margin for an entire season. The federal researcher, provincial extension officer and crop reporter need to work from the same figures and the same agricultural calendar. Devolution has made coordination harder. However, it has also made coordination indispensable.
A joint Chinese-Pakistani assessment of PARC and NARC identified outdated laboratories, chronic underinvestment, weak commercialisation and the loss of trained researchers. Chinese collaboration and overseas training can add capacity, provided the work reaches farms before the harvest is priced. By the next crop cycle, PARC should be able to show which varieties have moved beyond research stations, where they were planted and what they earned growers. Late crop reports and seed varieties confined to paper would mean Islamabad rearranged institutions while the farm stayed where it was.
