The BBC’s Shared Data Unit (SDU) is part of the Local News Partnerships programme in which the public corporation participates in collaborative journalism by working with regional news titles around the UK. With the SDU, the BBC is facilitating projects that create news specifically sourced from data sets, by teaching data skills to staff from partner organisations. I wondered if this model could be used to transfer skills in other sectors.
SDU Assistant editor Pete Sherlock, told an audience at the Open Data Institute that the aim of the SDU is to improve data literacy skills throughout the journalism industry. This is so that journalists have the skills to interrogate data sets and explain its implications, in the spirit of accountability and transparency.
The partnership between the SDU and the News Media Association - which represents the UK regional press - means that the data stories published by the unit are often picked up by between 40 and 50 local news providers around the country.
Sherlock said that high point of his career was the publication of a story on the diminishing bus network of the UK's transport infrastructure. The investigation, carried out by three reporters from different regional papers, found that there had been a loss of 134 million miles of bus routes in the last 10 years.
Sherlock explained that to investigate the issue and create the story, journalists from three UK publishing companies came together; Annette Belcher from Trinity Mirror, Paul Lynch from Johnston Press and Claire Wilde from Archant. He said: “They had to spend three months on secondment at the BBC in Birmingham learning data journalism techniques. At the same time the story was shared with 800 regional media outlets who have signed up to receive free content on our local news partnership.”
The secondees get both formal day to day training as well as mentoring from the core SDU team, during which time they learn how to clean, analyse and visual data, as well as a bit of coding and scraping.
This sounds like it could be a great way for data skills to be diffused in other industries. But is media such a singular industry and the BBC such a unique organisation that this specific form of data skills transfer wouldn’t work anywhere else?
Sherlock has learnt several lessons since the SDU was set up in 2017. These are; collaboration is now mainstream, it is possible to maintain autonomy in a partnership, successful partnerships benefit both parties, collaboration can lead to the cross-pollination of ideas and techniques, and finally partnership and collaboration fits into the SDU’s wider remit of being a data advocate, wanting to spread data skills throughout the industry.
These lessons could well be applied to other sectors, but the model might not segway as easily.
Companies in the private sector guard their data fiercely and probably won't see any benefit in helping competitors to get better at analysing their own data. So I think that this type of model might be difficult to replicate here. After all, there is no common public good that all commercial organisations would be working towards that would encourage them to share and disseminate skills.
Government, local authorities and other organisations in the public sector are not driven to be competitive however upskilling their staff in data skills could lead to them being more efficient which would be a definite plus. Nesta's programme of Offices of Data Analytics is faciltating data sharing among regional bodies. If data skills sharing is also taking place, then ODAs are a great example of diffusion of skills.
Charities and other non-governmental organisations in the third sector could well benefit from having a look at data of peer organisations that would better inform them about their users and lead to catering to their needs in a more comprehensive way. But would they have the spare human resource to take on such a project?
Say this were the case and public and third sector organisations were willing to share knowledge and expertise about data with each other. Which organisation would be the benefactor imparting its data skills expertise? Amongst charities, this could be Charity Commission or the Fundraising Regulator. But would they have the depth of data knowledge in the first place to be able to spread it around? Would knowledge exchange between equals be more appropriate than knowledge transfer?
Perhaps I was so enthralled with the idea of altruistic data skills sharing for the public good that I hoped it could be applicable in other circumstances. Clearly, I don’t have any answers, only questions. However, it is worth thinking about.