The final meeting of term focused on catalyzing change and the importance of improved data sources for modelling scenarios and researching solutions to fundamental questions regarding food supply.
We are jointly hosting these three meetings with the Joint Research Centre (JRC) of the European Commission and we are hoping to co-produce a brief white paper for EC policy makers identifying future research gaps and opportunities in this area.
Our first witness is Theirry Negre, the Head of the Food Security (FOODSEC) Group within the Monitoring Agricultural Resources (MARS) Unit at the European Commission’s Joint Research Center (JRC) who will be coming over from Ispra, Italy.
He will join Professor Jaideep Prabhu, the Jawaharlal Nehru Professor of Indian Business & Enterprise at the University of Cambridge Judge Business School and Dr Drew Purves, the former head of the Computational Ecology and Environmental Science group (CEES) at Microsoft Research who has just started working for Google Deepmind.
Wicked problems and questions generated by the discussion
What is the role of government and private institutions in data collection? Should governments be doing more? In many ways national institutions seem to be lagging behind private companies in gathering data. A reliance on the free market for data and, in turn, solutions for food production poses difficulties concerning adequate access, appropriate information gathering, confidentiality issues and short-termism. The potential for international bodies akin to meteorological services may be provide long-term perspectives.
Who should take ownership of data? Given how much big data is being gathered, is it appropriate for private or governmental institutions to have ownership of such an important resource? Is more research needed into how data is stored and traded? How does ownership of data impinge on creating open and transparent data sources? Sharing data openly may create issues regarding confidentiality, but this must be balanced with the potential benefits from new applications and viewpoints that may not have otherwise been discovered if data was not freely shared.
What levels of data are required? Amalgamating big data and local level information is crucial to building up a layered analysis. Ignoring certain levels of a system can risk overlooking key elements. Sometimes working at a community level, for example, makes a problem easier to solve.
Is there too much data? Drew Purves observed that in some cases a kilobyte of data may be more valuable than a petabyte of another kind of data. Even for experts huge amounts of data can be difficult to utilise and this problem is magnified for end-users such as businesses. There is a skill shortage in this area and more data scientists are needed. Additionally, sometimes data is considered valuable for its own sake, and if this data is not shared it may impede innovation.
Is a commercial strategy appropriate to data collection? By persuading people of the commercial benefits of surveys and data collection, they will be more inclined to provide accurate information. More work needs to be done to demonstrate the impact of such data on practical outcomes for end-users (such as farmers), so that people can understand its importance. Public trust concerning confidential data is a big issue.
How can information be passed onto its end-users? Farmers (or anyone affecting the supply chain) need specific information concerning how to improve their practice. Large-scale conclusions based on global data may not provide practical information for a farmer to act on.
What is the relationship between data analysis and resulting policy? There is little evidence regarding a linear connection between someone’s ability to access information and changes to their subsequent behaviour and decision-making processes. Understanding the impact of information on decision makers is extremely important. Sometimes politicians are hamstrung by external influences that will not be mitigated by improved knowledge and this suggests some element of the policy advice system needs changing, whether it is the system itself or the method of communication between scientists and policy makers.
For more information about these meetings, please follow the links on the right or e-mail Dr Rosamunde Almond (r.almond@damtp.cam.ac.uk)