- The new international guidelines for the 2030 round of agricultural censuses offer an opportunity to reestablish periodic agricultural censuses and modernize agricultural statistical systems in Latin America and the Caribbean.
- The new guidelines propose focusing the census on essential structural information and supplementing it with surveys, administrative records, and other data sources.
- Country experiences show that censuses generate greater value when they are part of ongoing statistical systems, rather than isolated operations.
Agriculture accounts for about 6% of GDP and 15% of employment in Latin America and the Caribbean. Between 1961 and 2021, agricultural production in the region grew at an average annual rate of 2.9%, driven mainly by improvements in total factor productivity (TFP), which increased by 1.7% annually—above the global average. These findings come from the Inter-American Bank (IDB) publication Agricultural Productivity in Latin America and the Caribbean: What We Know and Where We’re Headed.
This trend lost momentum during the last decade analyzed. Between 2010 and 2020, the region’s TFP grew about 0.9% annually and accounted for 40% of the increase in production; the remaining 60% came from input growth. In other words, recent growth has increasingly depended on higher input use than on productivity improvements.
The publication also underscores the importance of robust agricultural data systems for understanding these dynamics. Its review of the regional literature illustrates this: from an initial pool of more than 43,000 publications, it selected 82 studies that met all inclusion criteria. Thirty-six were from Brazil and 14 from Mexico; none were found for Caribbean countries. The most studied countries are those with agricultural censuses and surveys, and survey data were the most commonly used source in the analyzed studies. Where there is no data, there is no evidence.
The United Nations Statistical Commission approved new guidelines for the Global Agricultural Census Program 2030 (CAM 2030) in March 2026, proposing a more targeted approach for the censuses countries will conduct between 2026 and 2035. The new round incorporates lessons learned from the previous one, including reducing questionnaire overload, removing non-structural variables that can be collected through surveys, making better use of georeferencing, and shortening the time it takes to publish results.
Instead of concentrating all information in the census enumeration, CAM 2030 proposes enumerating a core set of 27 essential items and collecting the remaining 96 items primarily through sampling. It also incorporates a decision tree to help determine which information should be included in the census and which can be obtained through sample modules or supplementary surveys.
The logic is simple: the census can focus on the sector’s structural information, while intercensal surveys collect variables that change more frequently. The goal is not to add more questions to the questionnaire, but to use the most appropriate data-collection method for each type of information.
The experiences shared during the regional workshop on agricultural censuses, held in Antigua, Guatemala, from April 21 to 23, 2026—including Spain’s 2020 Agricultural Census—help us identify five lessons for countries preparing for the new census cycle.
1. Integrate administrative records into the census design. Spain demonstrates how administrative records can play a central role in census design. In the 2020 Agricultural Census, the National Institute of Statistics used various administrative sources to obtain information and focused direct data collection on farms most likely to be excluded from those records. This helped reduce costs by 38% compared to the previous census (from 25.4 to 15.8 million euros) and shortened the time required to release the results. Furthermore, the infrastructure developed continued to be used in subsequent surveys.
Uruguay is moving in a similar direction. The country built a georeferenced framework of producers based on administrative records, cross-checked that information during interviews, and used it to validate and complete the questionnaires. It is currently evaluating the possibility of moving toward an administrative census in 2029.
2. Articulate the census with an ongoing statistical program. Mexico articulates the census with an ongoing statistical program. Following the 2022 Agricultural Census, the National Institute of Statistics and Geography conducted the 2025 National Agricultural Survey and plans additional surveys in 2027, 2029, and 2031, before the next census in 2032. The framework adopts AGRIS, the FAO’s model for integrated agricultural surveys, and uses a multiple list-and-area sampling frame based on the 2022 census.
Peru demonstrates why this continuity also requires keeping sampling frames up to date. Its National Agricultural Survey has been conducted annually since 2014 and, since 2023, has used a new Master Sampling Frame provided by the Ministry of Agricultural Development and Irrigation (MIDAGRI).
3. Adapt data collection to different types of production units. Chile combined in-person data collection using mobile devices with web-based self-enumeration and adapted the questionnaires to the characteristics of the production units. It applied a comprehensive questionnaire to units with active operations and sales and a shortened version to units producing for self-consumption.
Additionally, it mapped the areas corresponding to those who responded online and excluded them from the field survey to avoid duplication. Brazil, for its part, plans to treat large agricultural enterprises differently, as they may require multiple respondents and specialized contact mechanisms. It also considers online data collection as an option.
4. Assess operational capacity needs from the outset. A census's quality depends not only on its methodological design but also on its operational capacity. For its upcoming agricultural census, Brazil plans to survey approximately six million farms, create more than 35,000 temporary positions, and use some 27,000 mobile devices.
The operation will also require transportation suited to rural areas, communications infrastructure, and hundreds of data collection points distributed throughout the country. This scale demonstrates that recruitment, training, transportation, connectivity, and equipment distribution must be planned and budgeted from the outset as central components of the census.
5. Plan for data dissemination and use. Paraguay demonstrates that the post-enumeration phase is just as important as the enumeration itself. The preliminary results of its 2022 National Agricultural Census were presented 19 days after the operation concluded, and the final results were published six months later.
The country also disseminated tabulations and maps through a geoportal, published anonymized microdata, and designed a Multiple Sampling Framework to conduct intercensal surveys. In addition, it conducted a Post-Census Evaluation Survey—commissioned to an external firm for design and implementation—to measure data coverage and quality.
Based on the discussions at the regional workshop in Antigua, Guatemala, the IDB proposed five strategies for strengthening agricultural statistics systems:
• Institutionalize annual surveys with representative samples.
• Develop panels that allow for tracking the same production units over time.
• Expand access to anonymized microdata.
• Incorporate variables on irrigation technology, plant health, and chemical input management.
• Strengthen sampling designs to produce representative information at the municipal or district level.
Countries in the region are already finalizing the schedule for their next agricultural censuses. According to the national plans available for 2026, Bolivia plans to conduct its census in 2026; Brazil and Colombia, in 2027; and Costa Rica, in 2028. Guatemala plans to conduct its 5th National Agricultural Census in 2030, and Mexico plans to conduct its next one in 2032.
Other plans are still being finalized. Ecuador is considering conducting its census between 2027 and 2028, while Uruguay is evaluating the possibility of conducting an administrative census in 2029. These dates correspond to national plans that are currently in effect or under evaluation and are subject to change. The CAM 2030 recommends conducting the census in 2030, or as close to that year as possible, so international comparisons are meaningful. Of the countries with known plans, only Guatemala will meet that deadline.
The quality of a census is determined long before the actual data collection begins. Developing the census framework, selecting the data to be collected through complete enumeration, integrating administrative records, mapping, budgeting, and hiring staff are all part of the pre-census phase. For countries planning censuses in the coming years, these preparations are an immediate priority.
Given the delays in the previous round, the region already has examples of how to move forward. The challenge now is to ensure the next census is not just a once-in-a-decade event, but the starting point for an integrated, continuous system of agricultural statistics.
Learn more about the methodological guidelines for the new round by consulting the FAO’s Guidelines for the 2030 Global Agricultural Census Program. To explore the relationship between agricultural productivity, information availability, and statistical systems in the region, download the IDB publication Agricultural Productivity in Latin America and the Caribbean: What We Know and Where We’re Headed.
This blog summarizes the discussions from the Regional Workshop on Agricultural Censuses (Antigua Guatemala, April 21–23, 2026), co-organized by the FAO and the IDB with support from Spain’s INE, AECID, and Guatemala’s INE, as well as the Fourth Meeting of the Community of Practice on Agricultural Statistics in Mexico, Central America, and the Dominican Republic (June 24, 2026), organized by ECLAC.