Data sharing is the primary ethical problem that stakeholders in digital farming frequently bring forwards. Simone van der Burg, leader of the work package on ethics in IOF2020 (WP7), therefore spent a significant part of her time on it. One of the goals of WP7 is to enhance reflection among stakeholders about the issue of data sharing. To offer input for that reflection, Van der Burg and her colleagues Elsje Oosterkamp and Marc-Jeroen Bogaardt first did a series of interviews with stakeholders (farmers, ICT developers, policymakers, insurance companies, banks, NGOs and government controlling bodies), asking for their reflections about the future of data sharing. Questions were asked ;such as: What are sensitive data? With whom would you (not) be willing to share these data? Are these data sensitive in everyone's hands? What do you see as a possible future for data sharing? How can data sharing be beneficial to you?

Based on the results of these interviews, Van der Burg and her co-workers identified four scenarios for future data sharing and governance and they listed values that play a role in a selection of these scenarios. The scenarios and the values were subsequently used as input for reflection of farmers and developers of smart farming technologies across the EU. Questions asked during the workshops included: Which values are important to you personally? And: What would your scenario look like if these values were respected in your  scenario? In addition, stakeholders are invited to re-think the scenarios in their role as citizens, because this might lead to a different perspective on data sharing scenarios than if they reflect as a representative of their company or business. Broadening the reflection about the data sharing future in this way, is characteristic of a responsible research and innovation approach: it not only asks people to express their values and preferences, but offers methods to enhance the reflection of  participants about their first intuitions. To do that, it helps to offer four scenarios, as there are then alternatives to think about, which invites participants to compare the pros and cons of each of them, based on values, as well as from their different roles as professionals or as citizens. Having to reflect about alternatives usually motivates participants to offer reasons for the choices they make. 

The four scenarios were:

During the workshops, participants reflected on the pros and cons of these various scenarios, but they also developed them further. For example, different compositions of the management team of the data library in scenario 2 were proposed by the workshop participants: some proposed the EU government to be in charge, others preferred to give the lead to a large company (such as GS1), or a multi-stakeholder management team that includes farmer’s representatives, developers of smart farming technologies and policy makers. And very often participants proposed to combine different scenario’s, such as scenario one and two or three and four. 

Workshops were carried out across the EU with farmers as well as developers of farm management systems and digital platforms focusing on farm data. The discussions with these various workshop participants shapes the background of this interview with Simone van der Burg. 

Opinions about that were quite diverse. Some participants thought it was fairer to pay farmers for their data, as in this way farmers would share in the benefits produced on the basis of their data by the company developing the smart farming technologies. So, offering money in exchange for data were sometimes thought to help realize a fairer distribution of the benefits. 
 
Another reason that participants brought forwards for the monetization of data is that it provides an incentive for farmers to share their data. This is also considered an advantage of monetization, for at present many farmers do not trust the businesses that are asking for their data. Farmers suspect that these businesses will use these data for their own benefit: they can develop new services and products, sell data to third parties, use them to inform their choices on the stock market, or use them for profiling of farmers and send them advertisements etc. As farmers do not particularly like these uses of their data, they sometimes feel reluctant to share their data. This reluctance can of course hinder the further digitalization of farming across the EU. 

In spite of these two reasons to support monetization of data, it is also important to mention that the market scenario (number 3) was considered the least attractive scenario according to our workshops. 

Well, some people initially chose it, because they thought it was the most realistic scenario. But a lot of participants were very critical about it. 

Especially the small and medium size companies developing farm management systems and services were critical. They thought that the market model would not be particularly helpful for them in the long run, as they have less room to invest in the purchase of data then larger companies do. Large companies are able to pay for data and are therefore able to make a better, more accurate, product.  Eventually this can lead to a situation in which a few large businesses dominate the market, such as is the case in North America. This prospect is not particularly attractive to a lot of the smaller companies, as it will be hard for them to compete with these large businesses. 

There were also other reasons to oppose the scenario. Some participants in the workshops brought forwards that smart farming also has a public role to play. If large agri-businesses will dominate the market, these public actors may no longer have access to the data, or only if they pay for it. This could make it more difficult to use the data for research, or to inform policy, which would not help the realization of the public goals of smart farming.   

The EU Code of Conduct for agricultural data sharing by contractual agreement was made by key stakeholders (farming organizations and agribusinesses) with the specific aim to allow farmers the freedom to choose what happens to their data. But what will happen to this freedom, if you offer an incentive to share data by offering payment?  This sounds perhaps strange, for of course everyone is free to choose whether to sell data or not. However, at present, the code of conduct gives data users the responsibility to be transparent about the ways they intend to use the data, end it asks them to return to farmers to ask their permission for eventual secondary or ‘ extra’ uses of their data. But what will happen when farmers sell their data to data users? When data are sold, then farmers transfer the ownership rights to the companies using the data. So, the new owner can do what he or she wants with the data and does not have to return to the farmer to ask whether this is OK. Therefore, it means that he will lose control over the subsequent uses of the data. This does not necessarily realize more trust in the sharing of farm data. 

Besides, if you ask someone for consent after giving information, this should be a free decision, meaning that the person giving consent is not coerced or manipulated into doing something. But offering payment could in fact be considered a kind of manipulation, or at least a gentle push in a the direction of data sharing. But is it right to do so, if the consequences of sharing data are not yet completely clear? As data use and re-use for various purposes is a practice that is still under development, selling data is perhaps not like other sales one has done; consequences are harder to foresee when the data economy is developing rapidly.

Monetization of data does not change anything about the protection of privacy and security of data. Data that fall under the General Data Protection Regulation (GDPR) still need to be protected. So there’s no interesting difference in this respect. However, the GDPR only protects data that refer to a person’s identity. Most farm data do not fall under the GDPR, as they cannot be linked to someone’s identity. It is precisely for this reason that key actors developed the EU Code of Conduct. This Code of Conduct was made to enhance trust in data sharing, by giving farmers and other ‘data originators’ the chance to control their own data. The EU Code of Conduct ascribes responsibility to data users to give information to the so-called ‘ data originator’ about what they are going to do with the data and ask their agreement. If, however, data originators sell their data, then they lose their rights over the data as the actor who bought them will then be the owner.  After that, the new owner will be the one who decides about the data. 

That is always a tricky question. People sometimes expect guidance from an ethicist, but this is not how I look at my role in IOF2020. The actors in IOF2020 are responsible for making the decisions that shape our digital future. At the end of the day, they should be able to look at themselves in the mirror and think: I did well. The only thing an ethicist can and should do is help them become more aware of the values and norms that underlie their initial intuitions and enhance reflection about those intuitions. My job is done well when actors do not decide too quickly, but fist pause and think: this is what I am inclined to do, but is this really good? When decisions are made after reflection about the values and norms that are at stake, then innovation can be called ‘ responsible’. That does not mean that the result is always good, but it is the best we can do at this moment.   

The decision about whether to monetize data or not, should take into account everything mentioned above: considerations about fair distribution of benefits, competitiveness of SME’s and large companies in the market, the benefits of data sharing vs the control that farmers have over their data use and the accountability of data users towards farmers.  

Besides, there’s one more topic that I would like to put on the table. One of the reasons why the EU invests in digitalization of farms is because it promises to respond to public concerns. Digitalization of farms should help to produce more, safer and better quality food, with less burden on the environment, which helps to feed the growing world population at a time when the environmental conditions to grow food will become worse. This shows that there is also a public interest to access the data, as data allow to monitor just how much more food production is realized, how much water is used, how many pesticides and what effects of food production are on the quality of the soil. These data are an important source for research which can inform food policy, as well as public awareness of consumers. When data are monetized, however, it is unclear how these public goals can continue to be served. What will happen to the accessibility of these data for public actors like researchers, policy makers and consumers? If they want to use the data, will they have to pay for them? And who decides about whether a particular actor is allowed to purchase the data? If a company who owns the data does not like the purpose for which a research institute or policy maker wants to use them, can he then refuse to sell? What effects will this eventually have on the effectiveness and accurateness of policy makers and the distribution of power in our societies?  

I think there are large questions there, which need reflection and debate. As this debate is just starting up, I would think it is quite premature to decide right now that monetization of data is the preferred option. So, if I would have to give any advice at all, it would be: slow down, reflect first about the pros and cons of the diverse available ways to foster data sharing before taking a decision.