The troublesome concept of data ownership

As a member of a EU project consortium I have had the opportunity to closely follow the EU policy development in the area of digital economy and to attend a number of major EU conferences related to the ICT and data economy. At almost all these events speakers brought up the notion of data ownership. Some of them addressed data ownership as something that lacked a clear definition, whilst others were able to explain it in very detail. Often data ownership was not even problematized, but simply considered as something that everyone would understand and thus no further explanation was required. Also, I have noticed that business people would normally associate data ownership with data management practices, while lawyers would try to articulate their understanding of the concept by drawing parallels to more definite legal terms such as intellectual property or contractual duties. 

I have to admit that it was hard to reconcile all those different views, but at least there was one thing that I understood immediately, namely that that data ownership has become a buzzword in the EU.
The contrast between the popularity of the term, especially in the light of the exploding data economy in the EU, and its blurred nature, has been a nudge for me to deep dive into the concept, and to reflect on it from different perspectives. This short essay is the summary of my analysis and will hopefully result in a decent definition of data ownership.

Data through the lances of the traditional property law concepts

Data are an intangible asset, which in many ways match the public good character of information and knowledge (Arrow 1962; Nelson 1959), at least with respect to non-rivalry in use (Ingrid 2015). Like other information-related goods, they can be reproduced and transferred at almost zero marginal costs (OECD 2015). In contrast to the property rights in physical goods, where the owner typically has exclusive rights and control over the good, this is not the case for intangibles such as data. This explains why it has been argued that property in data would challenge traditional concepts of civil law, which since Roman times have attributed property to tangible goods. Roman concepts of usus utendi et abutendi are the most prominent examples of proprietary entitlements and, frankly, one can hardly imagine them being vested in bits of data.



Data ownership in the system of IPRs

Now, if the basic proposition states that there can be no traditional property rights vested in data, what is the point of this discussion?

Well, for data, intellectual property rights (IPRs) are typically suggested as the legal means to establish clear ownership. For structured databases, a sui generis database right was created by the EU Database Protection Directive (29/9/EC), which protects the “substantial investment in either the obtaining, verification or presentation of the contents” [Art. 7(1)]. Apparently, the database right is somehow at odds with the traditional intellectual property rights (and has received quite some criticism from both the judiciary as well as from the academia), but it well illustrates why arguments for ownership and trade in data often rely on “return on investment” justifications (Ingrid 2015).

Property rights in personal data

In principle, personal data is no different than any other data, as it is equally intangible and thus the same argument against vesting property rights in data would apply. However, since recently personal data has been described as the new oil and, as such, a rivalrous good, which could make it more tangible and explain the relevance of the property discourse. For example, the new business models on the internet and social media, for instance the so called freemium pricing strategies in which users get services for free but “pay” with their personal data often without their own knowledge (“If you don’t pay for the product, you are the product”), may form a rationale for the propertisation.
The idea of propertisation of privacy rights is actually nothing new. It emerged in the US in the 1970s and was first brought to the EU in the 2000s (Bygrave 2013). Property in personal data refers to the entitlement to exclude the other from personal data by default. Put differently, the default entitlement in favor of the individual implies that there is no disclosure, collection or use of personal data by default (Purtova 2015).

The propertisation theory was not intended to be an end in itself, but to offer a theoretical framework for legal solutions that would strengthen personal rights, in particular data protection rights. Under Lessig’s proposal, consumers who would hold the original property entitlements to their own personal information would be able to bargain with data users to determine when it would be advantageous to forfeit their privacy by selling their data. 
The EU data protection legislation is sometimes claimed to include some property entitlements. For instance Victor (2015) sees property features in the newly-established right to data portability and the right to be forgotten.

Ownership v. property

I have shown above that there can be property in data. But how does the ownership fit in this discussion? Smith (2009) defines ownership as the ultimate right to use and abuse the object in question. Ownership and property are often used as synonyms, although they are clearly not. The concept of property contains two separate types of interest - ownership proprietary interests, e.g. the right to use an object, and non-ownership proprietary interests, e.g., the right to substract monetary value from the object. A good example is a 999-year lease of the land. That lease is an ownership right – both of us can reasonably say that the land is ‘my’, but at the same moment it is also a non-ownership right held by you in my land (Cooke, 2012).

Data ownership and big data contracting

Although ownership as a concept has caused quite some trouble to the courts, Smith argues that it might be useful in cases where property principles are still blurred, such as information and data. This could be the answer to the question why talking about ownership in data has been so popular. It is a way to avoid bringing up the legally disputable concept of property and rather focus on ownership, which is, as Smith points out, easily understood by laymen too.
In addition, the fact that ownership is typically split among several people makes it suitable for complex data business models, where often a big number of players with different interests are involved throughout the data lifecycle.

While property law imposes obligations on all the parties regardless of their relations (erga omnes, in rem), contractual duty only applies between the parties of a contract (intra partes, in personam). The latter will normally mean weaker protection, since a data owner will not be able to claim a legal entitlement against third parties, but only to the party of the contract. Nonetheless, owing to the uncertainty of its ownership under applicable IP law, some claim that the ownership in relation to data can be, and is most often, designated and protected by a contract (Glazer, Lebowitz and Greenberg 2013). Proper evaluation, protection and ownership identification of data can be essential in crises and insolvency situations; evaluating and assessing ownership of data is a sine-qua-non step for determining the value of a company or of its assets, for instance, in case of bankruptcy (Bird&Bird 2015).

Instead of a conclusion

However, increasingly the truly valuable asset in the information business is not the data. It is the clever analytics (software, algorithms), which runs smart data analysis and enables previously undiscovered patterns to be revealed. Today, the relevant property is not what protects the mere aggregation of data, but the copyright (or perhaps even patent) that protects the business analytics software (Wienand 2014).
The troublesome concept of data ownership The troublesome concept of data ownership Reviewed by Helena Uršič on 9:40 AM Rating: 5

No comments:

Theme images by merrymoonmary. Powered by Blogger.