In central Punjab, one of Pakistan’s most representative farming regions, climate change has driven an average annual production loss of 40.6% for cotton, followed in order of impact by maize (35.1%), rice (33.4%), sugarcane (25.2%) and wheat (22.9%). These losses are reported by the farmers themselves and, according to the authors, are consistent with national 2025 figures: a 30.7% drop for cotton and 8.9% for wheat, against an overall 13.5% decline in yields.

The study “Climate Change and Agricultural Adaptation in Punjab, Pakistan,” by Shakeel and colleagues, reconstructs these figures from 100 smallholder farmers in Toba Tek Singh. Agriculture accounts for roughly 23.5% of Pakistan’s GDP, employs 37% of the labor force and supports more than 60% of the rural population, and for the authors the impact reaches well beyond the fields. The collapse in cotton hits the textile sector, country’s leading export industry, deepens dependence on imports and strains foreign-exchange reserves, while falling food-crop output fuels food insecurity, more volatile prices and income losses for smallholders.

To cope with the climate, farmers rely mainly on three strategies - shifting sowing dates, adopting drought-resistant seeds and using water-efficient methods - and the study measures which factors drive their adoption.

Education is the only factor with a positive, statistically significant effect across all three strategies: each additional year of schooling raises the likelihood of shifting sowing dates by 34.5%, of adopting drought-resistant seeds by 30.3% and of turning to water-efficient methods by 36.3%. Farm size, by contrast, matters only for water-management practices.

The strongest associations emerge from direct experience: farmers who have already suffered crop losses are 71 times more likely to adopt water-efficient methods (OR 71.2), while those aware of climate change are almost 16 times more likely to change their sowing dates (OR 15.8). The authors note, however, that these high values often remain only marginally significant: awareness alone is not enough without economic resources and institutional support.

Pakistan is among the countries most exposed to climate risk, as the floods of 2010, 2022 and 2025 have shown. Adaptation responses at the level of individual farms, however, remain little studied: research has so far focused on macroeconomic analyses and aggregate yield models, treating adaptation as an all-or-nothing choice and overlooking farmers’ concrete decisions.

At the household level, the sample reports an average annual income of 485,150 rupees (six in ten families below the average), an average of six acres farmed and 5.5 years of schooling.

Punjab is the country’s agricultural heartland, and its farming is especially sensitive to swings in temperature, rainfall and water availability. Despite the National Climate Change Policy 2023 and the target of shifting 60% of the energy mix to renewables by 2030, climate-smart agriculture struggles to take hold because of financial constraints and weak institutional coordination. Hence the study’s question: which economic, social and perceptual factors actually drive smallholders to adapt, and how to calibrate extension services, climate information systems and access to credit accordingly.

As a representative case, the researchers chose the district of Toba Tek Singh in central Punjab: 2.5 million inhabitants, 462 mm of rainfall a year and heavy reliance on canal irrigation and groundwater. The sample of 100 farms was drawn in successive stages - from district to tehsil, then to three Union Councils and finally one village each - using the standard formula for proportions (90% probability, 10% margin of error).

The data came from interviews based on a structured questionnaire, pre-tested on ten farms, covering socio-economic characteristics, farm data, climate perception and practices adopted. The analysis combines descriptive statistics with a binary logistic regression in which the dependent variable equals 1 when the farmer adopts a strategy. The three strategies were analyzed separately in relation to age, education, income, farm size, climate awareness, perceived effects on yields and production costs; results are expressed as odds ratios, which indicate how much the probability of adoption changes for each one-unit increase in the variable.