Access to Digital Extension Services and Crop Yield: Evidence From Paddy Cultivation in India

Abstract

We examine the influence of information access on seasonal paddy yields in 10 selected Indian states using decade-long (2009–2019) data from the Government of India's Kisan (farmers) call center (KCC) initiative. Based on 17 million farmers' queries at the KCC, we isolate two most frequent concerns—one related to pest pressure and the other related to weather predictions. Leveraging these farmers' queries and crop yield–related datasets at the district level, we assess the impacts of weather and paddy pest infestation information on seasonal paddy yields. Endogeneity concerns are addressed using an instrumental variable (IV) approach—previous years' paddy yields and percentage gross irrigated area as IVs for pest and weather-related queries, respectively. After confirming the validity of the IVs, the results reveal that an increase in seasonal pest attacks—proxied by a 1% increase in seasonal pest-related queries from districts—decreases paddy yields by 0.028 t ha ${}^{-1}$ (i.e., a decrease of 1.1% of average paddy yields). On the contrary, a 1% increase in district-level seasonal weather queries increases paddy yields by 0.008 t ha ${}^{-1}$ (i.e., an increase of 0.3% of average paddy yields). Our results point to the positive contribution of paddy yield gains, while addressing proactive types of queries (i.e., weather), as well as the limitations of digital extension services, which are unable to recover the potential yield while addressing reactive types of queries (i.e., paddy pests) in a developing country.