Factors That Can Work against Data Sharing Agreement
[55] OECD (2013), “Exploring the economics of personal data: A survey of methodologies for measuring monetary value,” OECD Papers on the Digital Economy, No. 220, OECD Publishing, Paris, dx.doi.org/10.1787/5k486qtxldmq-en. There is a false tendency to believe that any gain or loss of a profit equals or proportional gain or loss of investment incentives, but this belief greatly simplifies the decision-making process and the underlying economy and ignores the relevance of alternative investment opportunities. 23. ←. Federer et al. (2015[48]) show, through a survey of 190 clinical and basic researchers from the National Institutes of Health`s Intramural Research Program in the United States, that while 71% of respondents reported sharing data directly with colleagues, only 39% said they had ever used a data repository service to share data. See also Sposito (2017[49]) and Wallis, Rolando and Borgman (2013[91]). The case of Cambridge Analytica is just one of many incidents where data is reused in another context in violation of agreed terms and conditions. Violation of these terms may not always be the result of malicious intent. As demonstrated by the Copenhagen expert workshop, data access and exchange is about taking data out of one context and transferring it to another. Referring to Nissenbaum (2004[9]) on privacy as contextual integrity, experts argued that changing context makes it difficult to ensure that existing rights and obligations are not compromised, for example when the assumptions and privacy expectations implicit in the initial use no longer apply to subsequent uses.
This is consistent with the above observation that information derived from the data is contextual and, therefore, the risks associated with data reuse. Sources: DTP (n.d.[45]), Data Transfer Project, datatransferproject.dev/; (DTP, n.d.[46]), DTP Github website, github.com/google/data-transfer-project (accessed February 5, 2019). In particular, Rodricks talked about how companies react when studies emerge that appear to have a high potential to negatively affect their business interests. “Those who worry the most. are those that emerge from human studies in which the authors examine the causal relationships between exposure and disease,” he said. The industry gets [really] nervous when studies like this appear when it comes to their product. They are concerned about increased regulation and accountability. Ellen Silbergeld, a professor at the Johns Hopkins Bloomberg School of Public Health and editor-in-chief of Environmental Research, added that this situation would make it very difficult for her to conduct the kind of studies she conducts to get people to volunteer for studies. “If I told someone in the studies we`re doing with Native Americans, people in the Amazon, people in downtown Baltimore and workers in South Carolina that the data could become available through a lawsuit, they wouldn`t participate.” If you are new to data exchange agreements or would like to learn more about them, here are some useful resources: “In Consortium X, we start with a proposal in which we describe the objectives and main analysis, the expected results, etc. Then we try to send technical details to the statisticians so that we can get feedback on the methodology, and then we have several sets of drafts. (…) [Active] collaboration is much more challenging than just being a data provider. (Interviewee 7) Once the agreement is concluded, it is the responsibility of the principal investigator to ensure that the agreed terms are met.
Ideally, these additional concerns should be addressed in the data sharing agreement to allow for clear communication and take additional security precautions if necessary: “We worked with another research group in Australia. Recently, the data access contract expired (…) and we tried to renew it. The legal team [of our institute] said that it had to be the controller-to-controller version of the EU`s contractual clauses, and the legal team of the other institute said that they could not sign such an agreement. There have been disagreements essentially on the [exact wording] of the agreement, but [our institute] cannot easily change the wording of these clauses. We had to interrupt access to the data for the Australian team, even though we have been working together for ten years without any problems. (Interviewee 2) What resources are available for data sharing agreements? ← 14. Examples include supervised research data centres where authorized researchers analyze data in a physically secure location; and secure remote data access services, where authorised researchers access a secure portal (OECD, 2013[100]; OECD, 2016[10]). The continuation of the benefits of the EEAS must therefore be weighed against the costs and legitimate national, public and private interests, in particular the rights and interests of the relevant stakeholders (protection of their privacy, intellectual property rights and national security). This is especially the case when it comes to sensitive data.
In addition to the documented barriers to data exchange in general, there is literature that describes the difficulties of data exchange infrastructures in creating incentives for data contribution from those who generate data (`data generators`). For example, the WorldWide Anti-Malerial Resistance Network (WWARN) data platform reported that data generators only agreed to provide data after being promised co-authorship for work in high-impact factor journals for data use [13]. At a later stage, the publication of pooled analyses by WWARN followed the ICMJE guidelines more strictly, although this step was greeted with dissatisfaction by some data generators [14]. Murtagh et al. describe four areas where progress is needed to establish functional biomedical data platforms. Two of these areas focus on elements related to the recognition and recognition of scientists such as “[t]he recognition of scientists` investment [in data generation] … [and] the essential contributions of all to the construction, maintenance and operation of data infrastructures” [15]. ← 42. According to Hofheinz and Osimo (2017[87]), co-ownership essentially concerned the rights and obligations that a limited number of stakeholders have with each other. To better understand the impact on data ownership, the authors recommend comparing data “ownership” with the rights parents have over their children. “These studies are really complicated, they often have complex study designs and different shutdown dynamics.
Some [research] questions that were asked didn`t work, others worked better. Some have been understood, others have not been understood and it is very difficult to just hand this over to someone else without all the basic information. (…) I don`t mind people using my data, but [they should] understand how the data was generated and what the limitations were. “Take that [record] and do what you [want].” I am not opposed to that principle. I hope we can do that, but we are not there yet. (Interviewee 17) In addressing these policy issues, policymakers need to avoid the “data policy trap” which, according to discussions at the Copenhagen expert workshop, is the tendency to seek a single solution to a multidimensional problem. Flexible data governance frameworks that take due account of different types of data and the different contexts of its reuse, while taking into account domain- and culture-specific specificities, are essential. ← 13.80% of social media site users in the United States are concerned about third-party access by businesses and 70% about third-party access by governments.
[87] Hofheinz, P. and D. Osimo (2017), Making Europe a Data Economy: A New Framework for Free Movement of Data in the Digital Age, www.lisboncouncil.net//index.php?option=com_downloads&id=1338. Control of agricultural data by leading ATPs has given rise to controversial discussions about the potential harm that discrimination and financial exploitation could cause to farmers and who owns agricultural data (Bunge, 2014[69]; The Economist, 2014[70]; Poppe, Wolfert & Verdouw, 2015[71]; Wolfert, 2017[72]; Sykuta, 2016[73]). For farmers, the benefits of data-intensive equipment also became less obvious, as there was a sense that farmers would “degrade” to the benefit of local stewards of land, animals and equipment and would only act as an entrepreneur ensuring that the interactions between supply and demand in the agricultural system worked properly together (OECD, 2015[20]). This has been exacerbated by uncertainties regarding the issue of data ownership (Banham, 2014[74]; Igor, 2015[75]). The coexistence of privacy frameworks has led some experts to believe that several owners should be taken over (co-ownership), including in particular the data subject, “since neither the `data producer` nor the `data collector` could claim an exclusive right to the data”.42 These experts argue that often several stakeholders were involved in the contribution. Collection and control of personal data, including the data subject himself….