Files
Abstract
Making correct software application choices is important because most modern organizations are not able to run an effective business without the support of information technology (IT). However, a software procurement decision is not a simple make-or-buy decision anymore. Organizations can choose from different alternative procurement options such as in-house development, software applications from software vendors, and open source software from communities. Each of these software procurement options has a different profile of risks and costs. For decision makers to make an informed decision, it is important that they understand the economical implications of each software procurement option and the factors that make one procurement choice better than others for a certain software project. Transaction Cost Economics (TCE) offers a theoretical framework to gain insights into these software procurement processes. Supported by empirical evidence collected by interviewing IT decisions makers, we propose that TCE needs to be refined and expanded to accurately apply to software procurement decisions. First, when a software application is embedded in a company's IT architecture, the costs of transaction failure include a readjustment cost that is in fact an important type of asset specificity not included in the original TCE framework. Second, if a failing software application is ingrained in a business's core processes, replacing this critical software application may also result in loss of revenue, decreasing productivity, as well as panic costs. These costs of transaction failure could go far beyond the costs of the initial investment and influence transaction costs as well. In the software procurement domain, the possibility of readjustment costs and losses due to criticality may affect transaction costs even in absence of other asset specificity dimensions such as human or physical asset specificity. Third, in TCE, the emphasis is on potential costs and risks of opportunism that increase transaction costs and may make market transactions unfavorable. However, in software procurement, safeguarding against transaction failure for reasons other than opportunistic behavior can be equally or even more important. Including the risk of transaction failure in the TCE framework might potentially change the favorability of a governance choice contrary to traditional TCE's prediction.