By Esther Raineau-Rispal
Widespread adoption of new technologies can take time, depending on agents’ perception of risk and the information they receive. It is especially true in the construction sector, yet high-performance materials are essential to ensure the efficiency and durability of the renovation and construction of dwellings. The depth of adoption is usually approached in one of two ways, either with inter-firm diffusion, which corresponds to new adoptions by firms over through time, or intra-firm diffusion, which measures the intensity of use by adopters. This paper presents an empirical procedure to estimate both dimensions simultaneously and account for geographical diffusion hubs. The estimator combines a panel data double-hurdle model and a spatial adoption index meant to capture word-of-mouth effects. A comprehensive dataset was built using scanner and geolocalized census data. The model was run using French data on an innovative gypsum board launched in 2017. Controls include both local and firm-specific features, as well as information regarding their purchase behaviors. Results suggest that inter and intra firm adoption are not driven by the same determinants, and that word-of-mouth is not the sole factor explaining the emergence of geographical clusters of adoption.