The relationship between open data and journalism is a complex one. No one questions the desirability of good journalism and the value of open data but both are under threat. British print circulation is falling, and for many titles by over 10% in a year and traditional revenue streams are under pressure. At the same time the momentum behind the British government’s open data movement shows signs of weakening as its impact is being questioned. They could help each other out. Harder stories based on real evidence (including open data) could make a contribution in halting the decline in journalism and open data. Sadly, in the UK, this seems unlikely to happen. If the private sector won’t seize the opportunity we should look to public sector broadcasting to set standards in quality and exciting journalism.
This blog is prompted by Jonathan Stoneman’s presentation at the ODI lunchtime talk on Friday. Here are the slides. Jonathan knows his stuff after years spent at the BBC and with links with the Reuters Institute for which he recently produced a working paper on “Does open data need journalism?” The paper is worth a read.
I have always felt that data and journalism should be natural bedfellows. A news story without data (or evidence) can easily be anecdote or feel like comment, and data without a story is, well, just data. But many journalists are not interested in data and some journalists are not interested in something that’s open (preferring a leak, hack or secret source, when not being spoon fed at a press briefing). Even at the best media houses, data or charts can be dropped because they ruin a good story. As Stoneman put it, data lovers and journalists generally have a “different approach to life”.
While media has its issues, open data has failed to take off as enthusiasts had hoped. Last month the Economist (“Out of the box“) reckoned the problem could be put down to four features: some released data has been less than useful (boring topics?), it’s often hard to navigate, there are too few people able to use it, and anxieties persist about privacy. The low skill levels is a problem for the mainstream media. One wonders how many could even use excel with any confidence. It could also be argued that open data has not delivered as key data sets need to be released – for example, data on budgets, legislation, statistics, spending, timetables, companies, elections, emissions, maps, postcodes – to provide a structure within which other data fits. This might be called a national information infrastructure (NII) and is something that the coalition failed to deliver despite plenty of prompting (not least from ODUG). The same issues have also played a part in the non-appearance of the armchair auditor (see this blog from Ben Worthy).
There are, of course, some examples of very good data journalism. The Open Knowledge Foundation’s excellent Data Journalism Handbook sets out some stories based on open data and the Global Editors Network has awards. A summary of the Guardian’s data journalism shows some great examples, but tellingly dates from 2011.
Stoneman did set out his favourite sites (often where open data has been structured and can be used to mine stories) including OpenPrescribing.net (information about GP drug prescribing in England), work by Tom Forth, Justice at Stake (chasing fairer courts), maplight (about money and politics in the US), CityMetric, and CitizenAudit (US non-profit tax data). Most are in the US where Stoneman said it was a “richer environment”.
He also said that Freedom of Information had been a greater tool for journalists than open data but that the relationship between them was interesting. The two together, operating effectively, could lead to considerable transparency and be a great source of fresh stories. Open data might be losing momentum but FoI is under real threat in the coming year given the review currently being undertaken. One option for the review team, is to deliver greater value from the effort that government puts into FoI. They could do this by ensuring that all responses are publicly available and could be re-used. That would be a win for openness if, as expected, the review restricts the use of FoI.
Stoneman ended by speculating about open journalism. It is a poorly -defined term that means different things to different people. The Guardian has one take on it but that dates from 2012 and perhaps their philosophy has slipped a bit. Stoneman pointed out, for example, that the datablog is not what it used to be and no longer makes the relevant data readily available allowing users to develop angles for themselves. Anyway, for Stoneman open journalism was more about where you get to when you write stories using open data.
It might be unreasonable to expect the privately run media organisations to devote time and money to this if it’s not going to help their advertising revenues but what about public sector broadcasters? Perhaps this is something to point out to the BBC in responding to its trust’s enquiry into statistics? It would be an inventive way to boost local news coverage (within the national context), deliver useful public information, be innovative and enhance democracy, all consistent with their remit.