Terrence August, Hyoduk Shin and Tunay Tunca
Information Systems Research, December, 2013, 24(4), 1068-1086
We present a model that captures the economic incentives for firms to contribute to open source software development and then explore the strategic and policy decisions of a software originator and regulators, respectively. We explore joint development and pricing equilibria in software markets for both open source and proprietary software strategies. If a firm chooses an open source strategy, she makes her source code publicly available which has several effects. First, the software receives a development boost as outside contributors help write and refine the product's source code, as well as produce related modules and utilities. Second, the firm must rely heavily on services as the primary revenue source since her source code is freely available to users. Third, the availability of source code makes it easier for competitors in the services market to build expertise and improve the quality of their own offerings, which means the originating firm faces tougher competition in this market. For example, a large contributor such as IBM can dedicate vast resources to the development of an open source project in order to increase his mastery of the software and improve his services packages. This investment can both benefit and harm an originator operating in an open source environment. Alternatively, the originating firm may choose a proprietary option, in which case the source code of her product remains closed. In this case, the firm incurs the entire product development cost alone. On the upside, in addition to potentially charging for software and integration services, the firm retains the ability to charge a price per each copy of the product regardless of whether she provides the services aspect.
Using this model, we study three main issues: First, under both open-source and proprietary licensing strategies, we thoroughly examine and identify the equilibrium outcomes in a market for software where an originator and contributor invest in development and firms compete for revenues from providing services. Second, building on this equilibrium analysis, we explore how software development efficiency for both the originator and subsequent contributors impacts the originator's choice of source code licensing strategy. Third, we study the implications of an open approach on social welfare. The U.S. government has recently begun to advocate open source development in certain circumstances and, more importantly, continues to play a significant role in setting policy to help increase the value generated by information goods in the economy. Accordingly, we also study social surplus, and provide insights into how governance and policy can utilize the economic incentives influenced by competition in the software services market to improve welfare.
In essence, what lies at the core of our paper is the management of two goods that are complements, namely the software and the integration and services provided for it. The main trade-off in this software licensing decision with services is as follows: the originator can either develop a product alone and be able to charge for the software herself in return (proprietary strategy), or she can open up the source code to outside contributors to share the burden of investment in development but rely only on services and integration income for revenue, albeit with possibly higher quality of software and services thanks to the efforts of the contributor firm (open source strategy). There are a number of factors that play critical roles in determining the best licensing strategy such as the relative efficiencies of the originator and the contributor firms, and capabilities of the firms reaping benefits from each others' software developments.
Contributor efficiency is a double-edged sword for the originator. When the contributor is strong in development efficiency and utilizing developments to improve quality, it can pay off for the originator to open the source code since this can result in significantly increased total software and service quality at little cost to the originator. However, a strong contributor can also be a formidable competitor that can steal significant market share. The originator has to balance these two factors when making her licensing decision. We find that if the originator is sufficiently efficient in development, her licensing decision mainly depends on her ability to harness the contributor's development to improve quality: if the originator is adept at improving the quality of her software/service package by utilizing the contributor's development, then an open source strategy is optimal, otherwise the originator is better off keeping the software proprietary. We find that an increase in originator development costs can in fact decrease the originator's service price and increase the contributor's service price. This occurs because decreased originator development efficiency reduces originator effort in investment and increases differentiation between the originator and the contributor's offerings. Consequently, the contributor can be better off by increasing his efforts in development, quality and price. We also show that increased contributor efficiency can unexpectedly decrease welfare. This result can manifest when the contributor is highly efficient because if the originator opens up the source code, the originator can be squeezed out of the services market as the contributor uses his development efficiency and strategic pricing to open up a large gap between the overall attractiveness of his offering and that of the originator. In such circumstances, the originator may instead choose a proprietary strategy, which results in lower software and service quality and decreased welfare.
Invited Talks and Conferences
- INFORMS Annual Meeting 2011, Charlotte, NC, November 15, 2011
- Conference on Information Systems and Technology 2010 (CIST), Austin, TX, November 6, 2010
- Marshall School of Business, University of Southern California, April 30, 2010
- Graduate School of Management, University of California, Davis, October 28, 2009
- INFORMS Annual Meeting 2009, San Diego, CA, October 13, 2009
- ITO Seminar, Rady School of Management, University of California, San Diego, December 1, 2008
- Product Innovation and Technology Management 2008 (PITM), Rady School of Management, University of California, San Diego, May 12, 2008
- Workshop on Information Systems and Economics 2007 (WISE), Desautels Faculty of Management, McGill University, December 8, 2007