How Far Agricultural Estimates Are Correct

August 27, 2026

The Centre’s has been making tall claims of record production of major food grains. But the Annual Report on Improvement of Crop Statistics (ICS) Scheme, 2023-24’ published by the National Statistical Office (NSO) under the Union Ministry of Statistics and Programme Implementation (MoSPI) is saying otherwise. This is concerning as the advance and final production estimates released by the Union Ministry of Agriculture and Farmers Welfare are considered the cornerstone of India’s agricultural policy. Based on these very statistics, the procurement strategy of the Food Corporation of India and state agencies is determined; plans for procurement at the Minimum Support Price are formulated; the availability of food grains for the Public Distribution System is assessed; decisions regarding exports and imports are taken; and policies related to national food security are framed. In several states, the assessment of crop insurance claims and the evaluation of the agricultural sector’s performance are also based on these official estimates. Therefore, these crop production figures are not merely statistical data; they form the basis for crucial policy decisions affecting millions of farmers, consumers, and the nation’s food and agricultural economy. India’s final production estimate for a crop is prepared based on two factors: first, the total area under the crop, and second, the average yield per hectare. The average yield is estimated solely through Crop Cutting Experiments (CCE). Under this process, the crop from a scientifically selected small portion of a field is harvested, threshed, cleaned, and weighed; based on this, an estimate of the yield per hectare is calculated. This figure serves as the basis for determining the total production for the entire district, state, and country. The scope of CCEs in India has expanded rapidly over the past five decades. When the Improvement of Crop Statistics (ICS) scheme was launched in 1973-74, only 1.73 lakh or 0.173 million CCEs were proposed under the General Crop Estimation Survey (GCES). By the year 2023-24, this number had risen to over 11.62 lakh or 1.162 million. Based on these methodologies, crop yield assessments and the monitoring of CCEs were conducted across 19 states and Union Territories (UTs) maintaining land records—including three EARAS or Establishment of an Agency for Reporting Agricultural Statistics states (Kerala, Odisha, and West Bengal) and UTs such as Goa, Meghalaya, and Dadra & Nagar Haveli and Daman & Diu. However, according to the report, Crop Estimation Surveys are still not conducted in Nagaland and Sikkim. Conducting CCEs and compiling the data is the responsibility of the State Agricultural Statistics Authorities (SASAs), while the Field Operations Division under MoSPI provides technical guidance to the states and ensures that the experiments are carried out in accordance with prescribed scientific standards. During CCEs, supervisors oversee the entire process on-site and record their observations in Schedule AS 2.0. Based on this data, the NSO prepares estimates for the average yield, the statistical error (Percentage Standard Error), and the yield per hectare for 33 major crops. These ‘Quick Estimates’ are used to verify the data submitted by the states. Therefore, if there are procedural flaws, a lack of supervision, or recording errors in CCEs, the impact is not limited merely to agricultural statistics; it can also undermine the credibility of the country’s official crop production figures. Thus, if CCEs are not conducted properly, the foundation of India’s overall production estimates could be undermined. This is precisely why the quality of CCEs is reviewed separately in the ICS report. The most significant finding of the report is that only 72 per cent of CCEs in 2023-24 were conducted in accordance with the prescribed scientific procedure, whereas 28 per cent exhibited procedural lapses, violations of rules, or other errors. This simply means that approximately one in four CCEs in the country was not conducted in accordance with the prescribed standards. According to the report, procedural errors were also recorded in 11 per cent of the CCEs. Examples include the failure to correctly select the designated plot, non-adherence to the harvesting and weighing procedures, or the violation of other technical guidelines. Furthermore, serious errors were found in recording ancillary information in approximately 10 per cent of the cases. Such ancillary details—regarding irrigation, seed variety, fertilisers, etc.—play a crucial role in subsequent analysis. If this information is incorrect, both the analysis and comparison of production could be affected. The report also highlights another serious flaw regarding sample selection. According to scientific norms, once a village and a field have been selected for a CCE, they should not be changed. However, in 202324, selected villages or fields had to be replaced in most states, including Andhra Pradesh, Assam, Bihar, Chhattisgarh, Haryana, Himachal Pradesh, Karnataka, Kerala, Madhya Pradesh, Maharashtra, Odisha, Tamil Nadu, Uttarakhand, and Puducherry. Only Gujarat, Jammu & Kashmir, Jharkhand, Punjab, Rajasthan, Telangana, and West Bengal were the states where the village replacement rate did not exceed 10 per cent in even a single season. According to the report, the need to change the originally selected villages due to the unavailability of the experimental crop arose in 9-50 per cent of cases in Assam, 23-53 per cent in Karnataka, 1-47 per cent in Kerala, 26-29 per cent in Tamil Nadu, 7-37 per cent in Bihar, 15-27 per cent in Andhra Pradesh, and 8-23 per cent in Odisha. In Puducherry, this figure even reached 50 per cent during one season. The report identifies the lack of an updated GCES plan, the failure to select fields on time, and insufficient coordination with farmers as the primary reasons for this; consequently, it clearly recommends updating the GCES plan, ensuring timely field selection, and maintaining regular contact with farmers to uphold the credibility of the scientific sampling methodology. According to the report, the performance of several states in monitoring CCEs was also unsatisfactory. During the Kharif season, the CCE supervision rate remained below 80 per cent in Himachal Pradesh, Jharkhand, Karnataka, Madhya Pradesh, Maharashtra, Uttar Pradesh, and West Bengal. During the Rabi and summer seasons, Bihar (Garma), Himachal Pradesh, Jammu & Kashmir, Karnataka (Rabi), Madhya Pradesh, Maharashtra, Uttar Pradesh (Rabi), and West Bengal also failed to achieve the 80 per cent mark. In contrast, the national-level response rates were 85 per cent for Pre-Kharif, 81 per cent for Kharif, 80 per cent for Rabi, and 88 per cent for the summer season. This implies that, on average, one out of every five CCEs could not be effectively supervised. The entire CCE system is based on the principle that each experiment undergoes independent supervision to ensure that field selection, plot demarcation, harvesting, threshing, and the weighing of the produce are carried out in accordance with scientific guidelines.