Competition and Multilevel Technology Adoption: A Dynamic Analysis of Electronic Medical Records Adoption in U.S. Hospitals

Abstract

This article develops and estimates a dynamic oligopoly model to study the multilevel adoption of electronic medical records in U.S. hospitals. I find substantial competitive effects in the adoption process, and hospitals of the same adoption level compete more intensely than hospitals of different levels do. Counterfactual experiments suggest that competitive effects prompt hospitals to engage in preemptive adoption and greatly deter the second adopter in a duopoly market. By evaluating the government's subsidy policy enacted in 2009, I find that a policy strategy accounting for market and hospital heterogeneity and competitive effects would have been more effective in stimulating adoption.

To follow the market or the parent system: Evidence from health IT adoption by hospital chains

“…I control for market characteristics (encapsulated in $${\mathcal{M}}_{it}^{j}$$) following the approach by Lin (2015) and Wang (2021). I include the following two market observables: total elderly population and the HHI.…”

“…I include the following two market observables: total elderly population and the HHI. I control for market unobservables using market‐level group dummies, similar to the method used by Collard‐Wexler (2013), Lin (2015), and Wang (2021). I construct these group dummies in the following steps.…”

“…This paper is related to four strands of literature. First, it contributes to the current studies on EMR adoption built on network effects theory (Lee et al, 2013;Miller & Tucker, 2009;Wang, 2021). Most prior studies explore the network effect from EMR adoption by studying how the fraction of nearby adopters affects adoption probability.…”

“…The location of each hospital can be directly linked to the corresponding health service area by the federal information processing standards code. Following prior studies, I control for market characteristics that are related to healthcare market competition and also affect hospital adoption decisions, using the following two variables: the size of the local elderly population and the Herfindahl‐Hirschman Index (HHI) (Lin, 2015; Wang, 2021). I obtain the former from the Area Health Resource Files.…”

“…Moreover, empirical evidence suggests that hospitals are more likely to exchange patient records if they share a common EMR system (Castillo et al., 2018; Downing et al., 2017; Everson & Adler‐Milstein, 2016). One exception is the paper by Wang (2021), who finds the competition effect dominates when considering different levels of adoption, but her analysis does not account for vendor heterogeneity.…”