by Michael Rahija, Regional Statistician, FAO Regional Office for Latin America and the Caribbean

World Statistics Day reminds us of the power of trusted data to guide better decisions. At FAO, we see statistics as vital accelerators of the Sustainable Development Goals (SDGs) and as tools to transform agrifood systems towards better nutrition, better production, a better environment, and ultimately, a better life for all.

In recent years, countries across Latin America have made significant progress in modernizing their agricultural statistical systems. By adopting new technologies, updating methodologies, and creating spaces for regional cooperation, they are building stronger data foundations. Yet the demand for detailed agrifood statistics continues to outpace the growth of budgets and innovations that can reduce costs. To meet this challenge, national statistical systems must embrace an “all option and all hands on deck” approach.

Agriculture is a cornerstone of Latin America´s economy

Agrifood systems remain central to Latin America’s economy and society. Agriculture including forestry and fisheries accounts for more than 10% of the economies of Nicaragua, Bolivia, Belize and Honduras, and more than 25% of all jobs in Ecuador, Nicaragua, and Guatemala (Figures 1 and 2).

Small-scale producers and family farms generate much of this food and employment. Behind every tortilla made with Guatemalan maize, every cup of coffee brewed with Colombian beans, and every ceviche paired with a Peruvian pisco sour, there is a story that only reliable data can reveal:

• Is farmland distributed equitably, or increasingly concentrated among large commercial farms?
• How are farmers adapting to the changing climate and its impact on productivity?
• Are smallholders earning enough to escape poverty?

Without credible data, governments are left in the dark unable to design effective policies, strengthen food security, or honor international commitments such as the SDGs.

Regional progress in agrifood systems statistics

The foundations of agricultural statistics are regular censuses every 5–10 years and annual sample surveys. In both areas, Latin America has advanced by adopting methodologies and innovations that reduce costs while improving data quality.

During the 2020 World Agricultural Census round, every country in the region transitioned from costly, error-prone paper questionnaires to electronic data collection. Tablets and smartphones replaced pens and paper, and several countries even integrated web-based and telephone interviews. Compared to the 2010 round, when nine out of eleven countries still used paper, this was a major leap forward.

Table 1: Interview methods by country from last census to WCA 2020

Notes: Paper indicates an interview method in which responses are handwritten on paper questions. CAPI indicates the use of a tablet or smart phone to record answers. CAWI indicates a web-based self-administered inteview, and CATI notes a telephone-based interview administered remotely.

Sample surveys also improved through the adoption of multi-frame sampling designs and aspects of the AGRISurvey methodology. For instance, Chile, El Salvador, Mexico, and Peru incorporated area frames to better capture smallholder farms, which are difficult to maintain in list-based frames due to the sector’s dynamism. This innovation was so important that it inspired the creation of a regional working group under the Statistical Conference of the Americas, which produced a diagnostic study on multi-frame sampling to identify common challenges and build capacity across countries.

Meanwhile, countries such as Costa Rica, Colombia, and Ecuador advanced by adopting survey modules from AGRISurvey to collect economic, environmental, and social data at the farm level. These data are critical for policymaking and for monitoring SDG indicators related to smallholder productivity and income (2.3.1 and 2.3.2), sustainability (2.4.1), food loss (12.3.1), and equal access to land (5.a.1).

Next frontier: Every option on the table

To meet the rising demand for detailed agrifood systems data, national statistical systems in Latina America must work together and put every option on the table.

In this regard, administrative data is one promising area. Many countries in Latin America have developed robust administrative reporting systems for agricultural commodity prices, livestock vaccinations, agricultural land registry, and family farmer registries among many others. However, there is not widespread incorporation of data collected by these registries to produce agrifood system statistics. It is essential that countries combine their knowledge and experience to develop tools and identify best practices for exploiting administrative data for agrifood system statistics.

Another promising area for cost reduction in Latin America is artificial intelligence particularly large language models (LLMs). These tools show promise to streamline the production of training materials, help refine questionnaires, and even support remote interviews, making surveys faster and more affordable. Agentic LLMs could be built to assist enumerators in the field and respondents in self-administered interviews such as CAWI and CATI to improve data quality.

Finally, countries in the region are also experimenting with the incorporation of remote sensing data in statistical production processes. Remote sensing data has been shown to be useful for updating area frame stratification, and support forecasting indicators such as crop area and yield. However, further work is needed to build expertise in NSOs, develop technological infrastructure, and collect field level training and validation data,

Strength in numbers: All hands-on deck

Latin America’s national agrifood system statistical systems face many challenges, from limited budgets to fast-changing data needs. Yet, these challenges also present opportunities. The real strength lies in collaboration, creating spaces where countries can share experiences, consolidate both successes and lessons learned, and build capacity across borders.

Working Groups of the Statistical Conference of the Americas (CEA) and the Latin American and Caribbean Commission for Agricultural Statistics already provide important platforms to advance these efforts. Likewise, FAO continues to support countries by organizing technical webinars, regional knowledge-sharing networks, and targeted events on topics ranging from the 2030 Census Round to the use of administrative data and new technologies.

On this World Statistics Day, one lesson is clear: building the future of agrifood statistics in Latin America will require an “all hands-on deck” approach. Every institution, every innovation, and every partnership must be mobilized to meet the growing demand for reliable data. At the same time, countries must keep “all options on the table” from administrative data and remote sensing to artificial intelligence and new survey methods. Only by combining these tools and working together can we ensure that the power of data be maximized to accelerate sustainable food system transformation towards better nutrition, better production, a better environment, and ultimately, a better life for all.

Mr Michael Rahija is Regional Statistician in the FAO Regional Office for Latin America and the Caribbean.