Column : Case for open source isnt open and shut

Written by Bibek Debroy | Updated: Jul 12 2009, 03:38am hrs
The communications & IT ministry has just come out with a national policy on open standards for e-governance. This is designed to ensure seamless interoperability of various e-governance solutions developed by multiple agenciesand avoid vendor lock-in. Since the policy is still in draft form, theres time to think things through.

As of now, the draft allows multiple standards within any application, but forces a single standard whenever there is an interface. It also states a standard can only be chosen if it is royalty-free. Across several countries in the world, and not just developed countries, technology neutrality has been advocated as a desirable policy, because it drives choice, competition, improved service and innovation. And it is dangerous for governments to dabble in technology, because one never knows what the future might bring.

Ostensibly, Indian government is in favour of a technology neutral policy, such as in case of telecom. But the proposed draft policy on open standards is hardly technology neutral. Plus, multiple standards permit consumer choice and competition. There are cogent reasons for public policy to remain technology and vendor-neutral.

A government pre-fixing a standard also implies official knowledge about existing and future technology. Any government-mandated laws and rules that require use of specific technologies are likely to stifle innovation. This becomes even more important when the global trend is towards allowing a multiplicity of standards, with inter-operability remaining the key.

Consumer welfare Technology evolves through industry action and different firms within the same industry are often driven by competitive motives. So, standards are bound to be different. So, instead of straight-jacketing technological evolution, lets accept multiple standards.

But we should want these standards to be compatible and work with each other and this is the inter-operability issue. No one can anticipate advances in technology, least of all the government. Nor is it always the case that the best technology, however defined, always wins in the marketplace. VCRs offer the best counter-example.

The average life-cycle for technology is short, even more so for software. But the underlying data may need to last for decades and one shouldnt get stuck with a single standard. The sensible course for any government is to offer choice in standards, including choice in inter-operability.

There is a difference between inter-operability and open-source versus proprietary software angle. Government action reveals a preference for open-source, often perceived as free software, though quantification of both costs and benefits remain suspect. Open-source software cannot be equated with open standards. Nor can open-source software be equated with royalty-free software.

Yes, some governments have introduced legislation backing open-source software. This is bad economics and bad law. Policy has been based on incomplete understanding of technology and confusion between open-source software and open standards and confusion that inter-operability is best pushed by mandating standards. This is inappropriate on at least five counts. First, IT industries are in a position to develop standards, independent of government mandates. Second, if market-based development is a core principle, there is no reason to deviate, not unless market failure is established. Third, trade agreements may violate such preferences.

Fourth, governments can go wrong and state failure is more difficult to correct than market failure. Fifth, principles of equality before the law are violated. This has subsequently been recognised and before citing what some governments have done in early 1990s, one needs to ask whether they continue to do so. Or have they altered their public procurement policies subsequently, recognising that choice and competition are the key

There is little point in re-inventing the wheel. The experience of other countries shows that technology-specific preferential legislation doesnt work and is inefficient. Accordingly, countries like the UK, Ireland, Canada, New Zealand, Peru, Costa Rica and Malaysia have rolled back preferences in favour of open-source software in public procurement decisions.

Preferences, which are in the last resort arbitrary, can force uniformity of products and lock-in with specific vendors. This constrains choice, competition and innovation. Is it better to implement faulty policies and then retreat from them, incurring assorted costs in the process Or is better to learn from other countries and be clear about policies to start with Let us not adopt this draft policy in a hurry and then regret it later.

The author is a noted economist