Tech for Good
We need to define who gets access to what type of data
India’s goal of unlocking value of non-personal data can benefit from transparent consultations on foundational questions related to value of data, incentives in data markets, competition laws, role of trusteeship, data trusts, communities and a framework for prioritization of public interest use cases
Existing data governance policies are driven by a market-centric approach, based on the concepts of data ownership, and treat data as property. This leads to concentration of profits and power in the hands of a few tech players. Moreover, the value of data remains trapped in silos, which can otherwise be utilized to make incredible advancements, such as healthcare breakthroughs.
The government-appointed committee for non-personal data governance, which released its draft report in July, has called out such systemic flaws. It has recognized the necessity for addressing collective harms, and suggested ways to unlock advances in AI and public interest. These deliberations place India at the forefront of global thinking on data governance, alongside the European Union.
But there are three main areas that require a pragmatic rethink—business implications related to incentives for innovation and compliance burden, need for a policy framework for data trusts, and adopting a practical approach to value of data.
First, the committee should reconsider mandatory data sharing across all sectors, which will distort the business incentive structures, create a first-mover disadvantage and raise the compliance burden. The report proposes all “data businesses” above a certain unspecified threshold must register and report their meta data on an open platform. Any entity can request raw data based on the publicly available declaration of meta data. The proposed regulator can adjudicate and make this data sharing mandatory.
Let’s take an example of a startup that has invested in drones to collect data about soil and farm conditions to provide crop insurance services. It will need to report all its meta data on the public platform. Another new entrant can ask for access to all the raw data for free, without making any infrastructure investments. Such an outcome is likely to upend the incentives for innovation. Commentators have arguably labelled this approach as “appropriation of intellectual property” and “nationalization of data”. While some criticisms are extreme, the sharp commentary has focused the spotlight on the distortion of incentives. In addition, regulatory burden is likely to increase with the need for extensive reporting and responding to data requests.
The goal of data sharing for public interest can be met by creating a bottom-up, narrow and well-defined framework for who gets access to what type of data and why. This will help identify specific sectors, considered critical to public and community interest, through an extensive consultation process. For example, it seems reasonable to suggest that private charter buses that operate public transport be mandated to share data for urban planning, as part of licensing requirements. The UK’s Bus Services Act of 2017 helped implement this and made data available to improve commuting across the country. In addition, legislation for portability for data—similar to mobile number portability—can help aid competition and incentivize innovation, without imposing high costs.
Second, the report introduces the concept of trustees and data trusts without clearly defining the role, accountability and legal basis of such institutions. The committee should recommend sandbox experiments at academic centres that can help create policy and tech framework for data trusts. These steps can fast track creation of data trusts with a solid foundation based on practical learnings.
Third, the committee’s conclusions on valuation of data are inconsistent and unnecessary. A narrow focus on public interest outcomes in specific sectors will obviate the need for a data marketplace. The report rightly reasons the theoretical and empirical framework for value of data is untested and complex. Yet, it has recommended creation of a digital marketplace to discover a fair price of data. This does not adequately account for costs of liabilities such as re-anonymization and privacy breach. Similarly, the recommended compensation on the basis of FRAND (fair, reasonable and non-discriminatory) is not a tested methodology for data markets; the standards, comparisons and jurisprudence to determine this compensation is sketchy.
India’s goal of unlocking value of non-personal data can benefit from transparent consultations on foundational questions related to value of data, incentives in data markets, competition laws, role of trusteeship, data trusts, communities and a framework for prioritization of public interest use cases. The intent to shift the Overton window on data governance should not translate into a haste to encumber business sentiments.
Sushant Kumar is principal at Omidyar Network India.
Catch all the Business News , Breaking News Events and Latest News Updates on Live Mint. Download The Mint News App to get Daily Market Updates.
MoreLess
First Published:10 Sep 2020, 08:46 PM IST
- Top Gainers
- Top Losers
- 52 Week High