Rebalancing the data asymmetry

Posted by Paul Mitchell
General Manager, Technology Policy Group

 A combination of ubiquitous data, analytics and machine learning is ushering in a transition from the information economy to a knowledge economy – where growth is driven primarily by the production of new ideas, insights, and knowledge. However, data-driven economies are reliant on a dependable supply of data to be sustainable. The current imbalance between the amount of data about individuals held by or accessible to institutions, and the inability of those same individuals to control the use of that data has created an asymmetry of power, resulting in a crisis of trust that was further exacerbated by last year’s revelations about government use of data and other events such as the recent data breaches from large retail outlets. All of this impacts the flow of such data. In the new World Economic Forum Global Information Technology Report 2014, a chapter on Rebalancing Socioeconomic Asymmetry in a Data-Driven Economy, co-authored by Microsoft, discusses some of the challenges that will need to be addressed to enable a thriving knowledge economy.

A key issue is how to incorporate the concept of fair value exchange into a data-driven world. In traditional economies, fair pricing is established when supply and demand are appropriately balanced. In data-driven economies, the suppliers – i.e., consumers – currently lack the necessary information to make rational decisions on whether or not their data should be used, especially if they are asked at the initial time of data collection. However, there are deeper issues. Most efforts to value personal data have primarily focused on estimating its commercial value and even then, there is very little agreement on how to do this. It is also difficult to estimate the social value of the use of data to different individuals. For example, how should the value of sharing personal health data to facilitate research into new drugs or improve medical protocols be estimated? Most economists would consider this an externality to be ignored.

The lack of either clearly understood commercial or social valuation of data has led the computer scientist Jaron Lanier to note that, “We’ve decided not to pay most people for performing the new roles that are valuable in relation to the latest technologies. Ordinary people ‘share,’ while elite network presences generate unprecedented fortunes.” He believes that this can lead to “massive disenfranchisement” and potential contraction of the economy. For the knowledge economy to thrive, there should be mechanisms for individuals to receive fair compensation for use of data about them. In other words, fair value exchange must be reset for the data-driven era.

How can a combination of technology and policy make this possible? Consider a data architecture where data is logically accompanied by a “metadata tag” that contains or refers to specific policies, e.g., who can access what data for which purposes in what context, and related provenance information. This could enable a “smart” data infrastructure that could track how the data is used to create new knowledge, and conceptually enable personalized control of the data use. When implemented as part of a trust framework, with policies that would penalize entities for disobeying or illegally modifying the policies specified, these techniques could start to address data’s crisis of trust.

Such mechanisms could also provide the underlying accounting necessary for data use and thus a better estimate of the monetary value of data. This could enable a sustainable “marketplace” of data that would satisfy the needs of all stakeholders involved – businesses, individuals, and government – while enabling use of data in ways that respect individual preferences.

Although recent news reports suggest that we have passed the peak of big data’s hype cycle, in reality data-driven economies are only just beginning to emerge, and generating more questions than answers. To address the questions, and ensure that the promise of a data-driven knowledge economy can be realized, will require strong partnerships between academic researchers, industry, consumer groups and policy makers on a global level. With a better understanding of the perspectives, challenges and opportunities ahead, innovative approaches to evidence-based policy making will be possible, and we will all realize the benefits. At Microsoft, we are collaborating closely with these multiple stakeholders and involving our Microsoft Research teams to participate in, as well as facilitate, these dialogues globally.