Talk:Interoperability

Urs Gasser and John Palfrey led a discussion of interoperability focused around 10 core “strong statements” about interoperability, illuminated by historic, current, and prospective case studies. John Palfrey introduced the session by remarking, “Interoperability is like apple pie.” More interoperability, he noted, is generally thought to be better. Once we pass this generalization, however, the picture becomes more complicated. Crafting a working definition of “interoperability” is challenging in itself. What do we mean when we say we want interoperability in the technologies of the future? As Urs Gasser observed, interoperability is closely tied to iLaw’s core themes: What is the role of governments in enabling innovation? What are the downsides of increased interconnectedness? Our challenge is to be precise about how much openness and what sort of safeguards need to be in place in order to be responsive to challenges and responsible as we move forward.

10 Strong Statements about Interoperability

1. Interoperability is generally desirable, especially in fields related to information, because it can lead to innovation, consumer choice, competition, and systemic efficiencies. Case studies that illustrate this statement include the development of online music systems and the mass use of barcodes.

• Question for Further Discussion: Although we can conclude that interoperability is beneficial based on these case studies, it is difficult to find large-scale empirical evidence. How broadly can we generalize from the volumes of anecdotal evidence demonstrating a positive relationship between interoperability and innovation?

2. Interoperability comes in different degrees and occurs at multiple layers – not only the technology and the data layers, but also at the human and institutional layers. Case studies that illustrate this statement include the railway system, email, and the simplified form of English used by air traffic controllers and pilots.

• Questions for Further Discussion: How can interoperability principles be applied to transcend different institutional cultures? How do translation and interoperability relate to one another?

3. Interoperability is not an unalloyed good, because it can exacerbate problems such as concerns over privacy, security and reliability. Case studies that illustrate privacy concerns include GoogleBuzz and the smart grid. From a security standpoint, compatibility across systems may mean that the same weaknesses or vulnerabilities can be exploited across platforms. Notably, however, interoperability does not necessarily have to take the form of standardization. Consider the example of currency: the Swiss franc coexists with the Euro.

• Questions for Further Discussion: How do we arrive at the optimal level of interoperability, even if it is not the maximal level of interoperability? At what level do we, or should we, impose constraints on data flows?

4. There is no one single way to accomplish an optimal level of interoperability. It can be achieved through a broad range of means, involving both public and private actors.

• Question for Further Discussion: A conceptual interoperability framework spans unilateral approaches to collaborative approaches, and non-regulatory approaches (private actors) and regulatory approaches (state actors). How do we balance and integrate the roles of the state and private firms in creating a productive trajectory?

5. Some of the means of accomplishing interoperability, such as open standards processes, are better than other in certain circumstances, especially in terms of ensuring interoperability over time.

• Question for Further Discussion: How do we harness interoperability to preserve knowledge on day one, and over time?

6. While most interoperability is achieved through private sector cooperation, the state may have a role to play in achieving interoperability. Our puzzle is figuring out how to match our challenges with our available tools. Case studies suggest that interoperability may be best achieved through efforts led by private sectors and corporations. Market forces play a powerful role in finding the right balance but this does not necessarily mean, however, that government does not play an important role, particularly ex ante.

• Questions for Further Discussion: Which sectors or actors should lead us? Should the approach be laissez-faire or highly structural?

7. Law should be used to foster development of interoperability systems and to address problems that can arise from highly interoperable systems.

8. The state can use its convening power, its purchasing power, and the power to force disclosure of adoption of a standard ex ante, as well as the blunt instrument of ex post competition and antitrust laws, to adjust the playing field in favor of certain degrees of interoperability. Threatened regulation may, in some instances, be quite effective. The best course may be to create positive opportunities and up-front incentives to spur interoperability efforts early in a technology’s lifecycle.

• Questions for Further Discussion: Is there an ideal mix of ex ante and ex post regulation? How do we time our efforts to accommodate network effects in markets? What does the example of Apple – who maintains that its products have been able to succeed in part because they are not interoperable – teach us?

9. There is a bidirectional relationship between interoperability and the law. Legal systems can help to establish or maintain interoperability over time, and legal systems themselves sometimes need to be rendered interoperable. China’s WTO accession offers one important case study.

10. How we accomplish interoperability matters greatly in the context of information and communications technologies, but it’s relevant to any complex system, including global financial system, the SG, loT, cloud computing, and many other emerging architectures of the future. These systems present us with a series of moments where interoperability by design will matter a great deal, both to thinking about how different actors work together and to minimizing opportunities for lock-in over time to non-interoperable systems.