“Lock-In and Unobserved Preferences in Server Operating System Adoption: A Case of Linux vs. Windows"

This paper attempts to distinguish state dependence (or lock-in) from unobserved preferences in the decision to adopt Linux or Windows as the operating system for computer servers. To this end, we use detailed survey data of over 100,000 establishments in the United States. Without accounting for unobserved heterogeneity in establishment-specific preferences for operating systems, we find a strong positive correlation between the current choice and the previous choice, suggesting potentially high switching costs and lock-in. To account for unobserved preferences for either operating system, we impose weak identifying assumptions and employ recently developed dynamic discrete choice panel data methods (Arellano and Carrasco 2003). The results show little or no evidence of state dependence, implying that unobserved preferences, rather than switching costs and lock-in, are more important factors in the adoption decision. Once taste heterogeneity is taken into account, we additionally find little evidence of network effects between server operating systems and non-server operating systems.