The Sourcing of Software Services

Abstract

Transaction cost economics (TCE) is the most prominent theory for studying questions surrounding information systems (IS) outsourcing. Its most widely tested argument is that client-specific services are rather kept in-house due to the high transaction costs which arise for safeguarding against opportunistic behavior. However, empirical support for this argument has so far been inconclusive. We seek to resolve this inconsistency by following two recommendations in prior literature. First, we extend TCE by considering trust differences between internal employees and vendor staff as a context factor. Second, we explicitly account for TCE's economic rationale by acknowledging that client-specific knowledge influences both transaction cost and production cost differences between in-house and outsourced IS services. Using data from 139 organizations on the sourcing of software development and maintenance services, we found support for our theoretical extension. For IS services requiring client-specific knowledge, in-house transaction cost advantages were particularly high if more trust was put in in-house personnel compared to vendor staff. Production costs were generally lower in-house when knowledge specificity was high. Both in-house transaction and production cost advantages were associated with lower degrees of outsourcing. The results explicate an important boundary condition of TCE and help shed light on the so far inconclusive empirical results regarding the role of asset specificity in explaining IS outsourcing decisions. The insights help decision makers by providing a better understanding of the economics of IS sourcing and the important role of trust.