Determinantes da adoção de tecnologias de agricultura digital por produtores de soja do estado de São Paulo

Abstract

Digital technologies have been changing production processes and business models in various industries by enhancing technical efficiency, fostering innovation, and enabling unprecedented access to data and information. This dissertation has two complementary objectives: (i) to identify and describe the main Digital Agriculture Technologies (DAT) adopted by soybean producers in the state of São Paulo; (ii) to identify the determinants of adoption and intensity of use of these technologies by farmers. First, a systematic literature review on digital technology adoption by grain farmers was conducted. Following this review and meetings with researchers and supply chain agents, a structured questionnaire was developed for microdata collection. A primary dataset of 150 soybean farms in the 2023/2024 crop year (cross-sectional survey) was obtained to perform empirical analysis. The data were collected in field research carried out between May and October of 2024. Econometric models of qualitative choice (logit regression) and count data (Poisson) were applied to identify the determining factors for the adoption and intensity in use of eight different DAT - grid soil sampling, variable rate seed application, variable rate fertilizer application, yield monitor, yield map, GPS auto-steer (autopilot), management software, and drone. The findings indicate that soybean area, farmer experience with soybean production, access to private consultancy, and education positively influence the adoption of most DA technologies, while cost perception negatively affects these farmers’ decisions. The results of Poisson model showed that more intensive adopters of digital technologies are more innovative and experienced, have higher levels of schooling, manage larger farms, hire consultancy services, and have a lower perception of digital technology costs. The empirical results of the work can contribute to farmers' resource allocation decisions and can also guide public policy design, helping agricultural extension agents, consultants, and technology developers in their strategies. Recommendations for future research include the investigation of the relationship between the adoption of digital agriculture technologies and their effects on total factor productivity (TFP) indicators. Additionally, the integration of climate-related variables into productivity analyses is suggested, as this may offer more robust insights into the environmental factors influencing agricultural efficiency and technological outcomes.