Riots are an opportunity for long-form data journalism
It is easy to think of “data journalism” as being about the automatic computer analysis of large datasets, but good data journalism has story-telling at the centre. Over the coming days, weeks, and months there is a lot of data journalism to be done about this week’s riots and looting in the UK.
There are so many data points - from the locations of the incidents, to the numbers of people arrested and charged, their demographic data, and the sentences they receive. Several media outlets, including the BBC, Guardian, Telegraph have begun to map and collate these.
Using data is vital here. Aditya Chakrabortty has described the riots as “a kind of grand Rorschach test in which members of right and left would peer into smouldering suburbs and shopping streets – and see precisely what they wanted to see”. This is evidenced by contributions to the debate from Mel, Max, Damian, Sunny and Seumas, or the ludicrous set piece posturing for the moral high ground between expense fiddlers Michael Gove & Harriet Harman on Newsnight.
Really good data journalism will help us untangle the truth from those prejudiced assumptions.
But this is data journalism that needs to stay the course, and seems like an ideal opportunity to do “long-form data journalism”. How long will these looters serve? What is the ethnic make-up and age range of those convicted? How many other criminals will get an early release because our jails are newly full of looters? How many people convicted this week will go on to re-offend?
The impact of the riots is going to be felt in data-driven stories for months and years to come. I’ve no doubt that experienced data crunchers like Simon Rogers or Conrad Quilty-Harper will factor it into their work, but I anticipate that in six months time we’ll be seeing stories about a sudden percentage rise in crime in Enfield or Central Manchester, without specific reference to the riots. The journalists writing them won’t have isolated the events of the last few days as exceptions to the general trend.
(If you don’t think that is the case, you only have to ask yourself why London crime data over the last ten years can be used to suggest you are statistically most likely to be murdered in Kings Cross or Aldgate.)
There can be genuine social consequences to the misinterpretation of data. If the postcodes in Enfield become marked as a place where crime is now more likely as a result of one night of violence, then house prices could be depressed and insurance costs will rise, meaning the effects of the riots will still be felt long after broken windows are replaced.
It is the responsibility of the media to use this data in a way that helps us understand the riots, not in a way that prolongs their negative impact.