Assumption 1: A software industry operating on two business models: a) CSS: Software is proprietary and sold for a price and profit;
b) OSS: Software code is developed by for-profit companies, which then provide code for free.
In the latter case, companies, which develop OSS, earn the profit necessary to sustain this activity by other complementary activities, including:
b1) Provide OSS to create a market for their own proprietary software;
b2) Provide OSS to create a market for complementary services (customisation, training, maintenance, consulting, advertising etc) or products (hardware, digital media)
OSS can also be developed outside this market using business model c:
c) Software code is developed by non-for-profit entities (research, academia, non-commercial initiatives) competing for grants and funds.
Since we are looking to compare (a) to (b) ceteris paribus, these providers can be excluded. I.e. we assume that the amount of resources consumed and the amount of software generated through model (c) is independent from the mix of (a) and (b).
Assumption 2: Let market volume of (a) become zero; all software in the market is provided using business model (b). Obviously, (b1) must necessarily be also zero, by definition. All software (excluding c) is now produced through model (b2), i.e. to support cross-selling of other complementary services and products.
Assumption 3: The market has no structural deficiencies, i.e.
- no information asymmetries (i.e. all consumers and suppliers are immediately aware of the existence of all software);
- no network externalities (i.e. software is interoperable with other software and hardware, therefore interchangeable, and no consumer lock-ins are possible)
By assumptions (2) and (3) follows that:
- all market providers can offer to all potential consumers exactly the same corpus of OSS Code to support sales of their other products or services. Therefore:
- any competitive advantage in selling the non-software products or services and gained by developing software, will immediately dissipate, since it will be immediately available to competitors. Competitiveness in terms of cost reduction, value creation etc. will only be feasible though innovation in the non-software activities of these suppliers.
- In other terms, software becomes a freely available commodity with minimal (indirect) returns. Consequently:
- New software is developed only when absolutely necessary and development stagnates due to lack of resources. - Model (c) still generates innovation, but the lump sum of innovation is less than in any other combination of (a) and (b).
Conclusion: We need proprietary software to spur innovation, and we need open source software to lower the cost of developing proprietary software, therefore there must be an optimum mixture.
Discussion 1: Is the market stable enough to achieve an equilibrium? Is this equilibrium optimum from the point of view of innovation, consumer well being etc?
Discussion 2: Is there also a competition for resources between (a) and (c) as well as between (b) and (c); Do we need to bring a dynamic interaction with (c) into the hypothesis?
Engineer and Entrepreneur.