Our interviews suggest that funders can generally be placed into two broad categories: those that aim to actively shape research in an area in order to achieve a specific goal or impact within society; and those that aim to fund quality research within a specific topic area or discipline, but are not as concerned about setting the agenda.
Our sample size is small, but not surprisingly philanthropic funders appear to fall in the former category, while the research councils fall in the latter. For instance, when asked what the future holds for the field of data-related research, one interviewee from a philanthropic organisation with a focus on social justice mentioned that they hope to see topics such as data justice more at the centre of the data-related research agenda. They further noted that their organisation seemed to be heading in that direction:
“I hope that it becomes second nature to think about power in conjunction with data… I am really thinking and hoping that we are going to see a shift there that […] we’re not going to keep talking about data-driven technology without talking about how power is being distributed and or redistributed” (Participant B, philanthropic funder).
The choices of this organisation in terms of what to fund next therefore seem to be guided by overarching, thematic, internal objectives that correspond to the types of impact they want to have on the field and the changes they want to see in the world.
On the other hand, people from organisations that are more concerned with funding quality research than with setting the agenda for a topic or research area, seemed relatively less interested in anticipating the future of data-related research within their part of the landscape. Often, it seems, decisions on the types of research to fund next are driven by external trends that are emerging in the field rather than the result of an active role the organisation wants to have in setting the agenda of that field. One participant, for instance, noted that their own agenda is shaped through collaboration with, and awareness of, the research agendas of external research communities: “ we work closely with research communities, [and] we understand what research agendas are coming up” (Participant J, UK research council). Ultimately, it seems that it is the breadth of topics within a research area that really matters. As one participant from a research council stated in an interview, “it’s more about getting that breadth – [making sure] that the research area is covered.” (Participant D, UK research council).
When asked about the future of data-related research and funding, participants generally mentioned: (1) that they are seeing an increase in funding and research about public engagement, transparency, and fairness, and (2) that the future of funding is highly dependent on factors such as spending reviews and financial priorities, and that makes it quite hard to anticipate the direction of the data-related research ecosystem.
To the first point, one interviewee said “I believe we will see an increase in substantive research into approaches to algorithmic audit” (Participant L, philanthropic funder). Other interviewees mentioned the increasing interest in transparency and public engagement. For instance,
“I think the future will increasingly need solid public engagement around [this field] as it expands. And I think that there have been a couple of instances where the public have perceived things to be done with their data by the government […] that they were not on board with. […] And I think the pandemic has perhaps made people more aware of data. But yeah, I think that that’s becoming increasingly important and understood by people in the field. So I think public benefit and transparency are going to be really key” (Participant E, UK research council).
Someone else from a UK research council stressed the importance of increasing people’s awareness and trust in data-driven systems, saying:
“… part of our remit is into public engagement, which I think actually is a low hanging fruit, we need to do more to make sure we can engage people in how their data is used and bring some acceptance towards technologies, such as artificial intelligence, because otherwise they are not gonna have an impact” (Participant I, UK research council).
In addition to public engagement activities, it seems reasonable to say that in order for data to be trustworthy and for people to subsequently trust it, there needs to be a high degree of curation of that data. About that, one participant said that “…we definitely need to fund more efforts around data curation and data quality, building on and implementing fair principles” (Participant K, UK research council).
To the second point, one participant mentioned that “there is always a difficult tension between what we think is important and what we can actually fund.” (Participant H, UK research council). Although this is something that applies to all kinds of research, it seems important to highlight how budget allocation and financial priorities play a role in setting future research agendas.
The next section presents the findings of our quantitative analysis. In particular, we attempt to provide answers to eight questions: