The Economics of Technology Sharing

In this post I will try to answer the question why programmers invest their own time in OS projects? What motivates them to contribute and what is in it for commercial companies?

Literature provides us with 6 possible reasons:

• Improve performance in paid work (e.g. system administrators looking for solutions for their companies) 
• Intrinsic pleasure (e.g. cool project) 
• Future job offers, shares in commercial open source based companies 
• Ego gratification from peer recognition 
• The promise of higher future earnings 
• Intellectual curiosity 

Signaling 

Other very important reason for investing programmers’ time in open source projects is signaling, i.e. proving high level of competence. Why this is important in OS? Because the ability to signal a high level of competence can be higher in an OS project. 

• Outsiders can see the contribution of each individual 
• The programmer takes full responsibility for a subproject 
• More knowledge can be transferred to new environments, because many elements of the source code shared across OS projects 

Let’s have a look at the Apache Project.

Instead of central leader they have a series of committees to resolve open issues. There are five levels of rank within the Apache Software Foundation (ASF): 

• Developer 
• Committer 
• Project management committee member 
• ASF member 
• ASF board member 

Advancement is made in recognition of an individual’s commitment and contribution to an Apache project. Empirical study (Hann et al 2004) showed that: 

• Sheer volume of contribution have little impact on salary 
• Moving into higher rank resulted in 14-29% increase 

Commercial Firms

Commercial firms may interact with open source projects in a number of ways:

• Benefit from complementary markets 
• Benefit from learning 
• Benefit from good public relations 
• Compete directly with open source providers 

Example: IBM and HP released codes to open source communities

Why?

Bait and hook: if the released code will be used more widely, profits in the complementary segment will grow (e.g. from consulting services)

Other very important reason is that many small developers are uncomfortable doing business with large firms, that may compete in the developers segment and reduce price in order to raise demand for the broad software platform.

By contrast, when a large firm makes its platform available on an open source basis, the small firm need no longer fear being squeezed.

However, we don’t know if the corporation will keep all source code in the public domain? Or if it will highlight important contributions adequately?

In this case important licensing plays important role.

Open source licenses

Permissive license

users retain the ability to use the code as they want. Common in projects with strong appeal to the community of contributors

Restrictive license 

common in projects geared for end users who are unlikely to appreciate the coding (e.g. computer games)

Quality

There is no consensus if open source software is superior to “off-the-shelf” commercial software. Advocates of OS code’s higher quality stress that:

• Users can enhance quality 
• Users can customize it to meet particular needs 
• There is a superior development process: 
• Workers in commercial firms may not report errors of fellow employees – however OS programmers do not have incentive to collude, and their project is more peer reviewed 
• Security flaws are more easily identified because many people are involved 

Counter arguments:

• The openness of the code allows hackers to figure out its weaknesses 
• Poorer documentation (due to the incentives structure) 
• Poorer user interface 

Public policy

Governments around the world encourage the development and use of open source projects. However, the impact on social welfare is not conclusive:

• OS projects have access to a larger talent pool 
• On the other hand: free riding problem 

According to static point of view, any potential user has access therefore OS increase social welfare.

Dynamic view states that developers may lack incentives to introduce new products therefore OS reduce social welfare.

Bibliography: Lerner, Josh, and Jean Tirole. The economics of technology sharing: Open source and beyond. No. w10956. National Bureau of Economic Research, 2004.

Innovation models

In organizational science there are two prevalent models of innovation:

(1) Private investment model
(2) Collective action model

Open source contains elements of both models, and represents private-collective model

The private investment model
In the private investment model an innovator earns from private goods thanks to intellectual property protection. Intellectual property laws are simply the innovators’ rights to their work. They are an incentive to create new knowledge for the innovators. However, intellectual property protection diminishes knowledge dissemination.

The collective action model
The collective action model works under conditions of market failure, where innovators collaborate in order to produce a public good.

Provision of public goods implies that if any user consumes it, it cannot be feasibly withheld from other users. This infers two main problems:

(1) Free riding problem
(2) Incentive problem

General solution to those problems is to provide monetary and reputation-based rewards to innovators.

Private-collective model
Private-collective model comprises of both models where incentives for private investment and collective action coexist.

First of all, software users rather than software manufacturers are the typical innovators; a fundamental incentive is using the software. Secondly, innovators freely reveal the proprietary code they developed. This implies (1) an increase of innovator benefits (e.g. sales of complementary goods) and (2) private losses will typically be quite low, rivalry with potential adopters is low – but the reward may be significant (e.g. reputation, reciprocity, building a community)

In private investment model it is assumed that free revealing of code leads to a loss of private profits. Conversely, in OS projects free revealing can result in net gains for the company. For example, free revealing can increase innovation diffusion and so increase an innovator’s profits through positive network effect.

Collective action model assumes a free rider can obtain equal benefits to those a contributor can obtain. Conversely, in OS projects contributors gain private benefits which are stronger than those available to free riders, e.g. learning and enjoyment, sense of ownership and control over the work, they can choose the project, the task and the technical approach to suit their own interest, and last but not least participating in a community. These rewards reduce the free riding problem.

It is worth noting, that free riders can be good for the company. There are two reasons for that (1) their adoption of the software increase its market share and help to set it as a standard in a marketplace, (2) some users do not write code – but contribute by reporting bugs.

Source: Eric von Hippel and Georg von Krogh (2003) "Open Source Software and the 'Private-Collective' Innovation Model: Issues for Organization Science" Organization Science 14, 209-223.

Open Source

What is open source?

Open source usually refers to software that is released with source code under a license that ensures that derivative works will also be available as source code, protects certain rights of the original authors, and prohibits restrictions on how the software can be used or who can use it.

You can find over 300k open software projects on Sourceforge.net - provider of free services to open source developer.

What is an open source project?

Open source project is a voluntarily collaboration of Internet-based communities of software developers to develop software that they or their organization need. Contributors agree to make all enhancements available to everybody. Many of contributors are not paid; the strucutre is often loosely structured, contributors are free to choose interest area.

Open Source is different from shareware, freeware and crowdsourcing. Shareware is a proprietary software provided for free (binary files), usually on a trial basis (the user pay for continued use/support). Freeware is a software provided for free (can be copyrighted or in the public domain). Crowdsourcing is basically an outsourcing of a task to a group of contributors.

Early history

1960s-1970s: sharing source code was commonplace (part of the research cultures). Software was mainly developed in academic and corporate labs by scientists and engineers

Early 1980’s: AT&T enforced its property rights of Linux, to which many academics and other corporate researchers contributed

1985: The Free Software Foundation was established by Richard Stallman, a programmer at MIT Artificial Intelligence Laboratory

"Software users should freely learn and create; software should be free to use, modify and distribute" The philosophy of The Free Software Foundation

1998: the Open Source Movement was founded by prominent hackers, replacing the term “free software”
and emphasizing the practical benefits of OS (economic and technological)

Watch a video of Richard Stallman talking about the open source:

How OS projects evolve?

A project is typically initiated by an individual or a small group having an idea, for an intellectual, personal or business reason. Anyone with the proper programming skills and motivation can use and modify any OS software written by anyone. The project initiators usually become the project “owners” taking responsibility for project management. Others can download, use and “play” with the code (most of them are free riders). However, some go on and modify the code and then they post it on the project website for others to use it and for feedback. In many projects the privilege of adding to the authorized code is restricted to only a few trusted developers a.k.a. “gate keepers”.

Who is using OS?

• 20% of Internet users use Firefox (source: www.statowl.com, 06-2012)
• Facebook uses PHP and MySQL and is the largest user in the world of memcached, an open-source caching system 
• Google has over one million servers running a customized Linux version as an operating system 
• 50% of web servers employ Apache 
• 60% of web servers use Linux as an operating system 
• PERL and PHP are the dominant scripting languages