Think before you retweet: why the "scrounger spike" is illusory

What happened between 2004 and 2010 is more people started writing about welfare.

This chart seems to have quickly acquired mythological status on Twitter:

It appears to show the number of times the word "scrounger" appeared in the UK press from 1994 to the present, with a huge unprecedented rise coinciding with the arrival of the coalition in power.

It’s easy to see why people have seized on it, given the sustained campaign against benefit claimants in parts of the print media and the government's willingness to stoke the flames by selective, and in some cases deceptive use of official data. But the picture it gives of trends over time is almost certainly exaggerated.

The underlying data in the chart comes from a database called LexisNexis. This is an indispensable resource for looking at media coverage over extended periods, but as any researcher who has worked with it will tell you, it needs to be treated with caution, and especially when trying to construct a time series.

The main reason is that titles appear in the database at different times: if you're looking at the mid-1990's, the only titles included are the Mirror, Mail, Times, Guardian and Independent. By the time you reach 2013, the Sun and Express are included, as well as the Telegraph. That is hardly a minor difference when you're doing searches on words like "scrounger".

When Kate Bell, Ben Baumberg, Dan Sage and I looked at media coverage of benefits last year for the charity Turn2Us, we commissioned a purpose-built database to iron out data problems, as well as manually cleaning the articles we'd sourced from LexisNexis to remove irrelevant material. Analysis of media coverage over time is a demanding task. Even with a customised database, producing and coding a series from 1995 to 2011 represented several weeks work.

We didn't just look for the word "scrounger", we constructed a set of word-lists intended to capture broad categories of negative vocabulary such as fraud and dependency. "Scounger(s)" was included in a word-list we called "non-reciprocity", along with terms like "handout(s)", "shirker(s)" and "something for nothing". The chart shows some of the results for a consistent set of titles from 1995 to 2011:

So has there been an unprecedented rise in the use of "scrounger" and related terms? Yes – see the purple curve on the chart – but it's on quite a different scale to that shown in the other chart.

The difference is, I presume, mainly to do with our use of a consistent set of titles. We just don't have data for the Sun and Express for the 1990's, and these are by far the most virulent titles when it comes to coverage of benefits.

What is perhaps of more interest is that overall use of negative vocabularies is very similar in 2010-11 and 1998-9 (the dotted red curve). But there has been a pronounced shift in the language used to convey negative messages away from fraud (although this remains very important) and towards non-reciprocity: from "cheats" to "scroungers".

Using consistent data which has been checked and cleaned makes for a less striking chart, but it allows us to put recent coverage into a longer-term context. On this evidence, negativity in press coverage of benefits has not been at unprecedented levels: the first years of New Labour's period in office were very similar to the coalition's first years in this respect.

The reason is that media coverage is strongly skewed towards negativity in all periods, and in both these periods there was simply a lot of coverage.

But since 2008, negativity has increasingly taken the form of accusations of "scrounging", while in New Labour's first years in office the focus was on fraud, largely because this was a theme the government chose to highlight.

It's widely argued that there has been a shift in public attitudes towards benefit claimants over recent decades, with no signs of an increase in sympathy when unemployment rose in 2008/9, in contrast to earlier periods. And it's tempting to see negative media coverage as the culprit here. Our report did find evidence that attitudes are influenced by which newspaper people read, but when it comes to explaining shifts in attitudes over the longer term, increased negativity in the press seems an unlikely candidate, just because UK media coverage of benefits has been strongly negative for as far back as we can go using this data.

It may be that the press and the political right have become more efficient in tapping into public anxieties and grievances – maybe "scroungers" is a stronger framing than "cheats". It may be that New Labour's borrowing of political vocabulary from the right ("something for nothing", "dependency culture") recast the terms of debate, so that negative attitudes which were always to some extent in play were legitimised and made respectable.

But a simple story of media manipulation, evidenced by counting occurrences of negative language, seems to be ruled out. That doesn't mean there is no manipulation of course. There is a lot, and most of it is clearly politically inspired. But this is only part of a larger story. Those worried about public attitudes to benefits and claimants need to ask why negative coverage finds such a ready audience.

An earlier version of this piece was originally posted on Declan's blog l'Art Social, and is reposted here with his permission.

Declan Gaffney is a policy consultant specialising in social security, labour markets and equality. He blogs at l'Art Social

Reuters/New Statesman composite.
Show Hide image

When it comes to social media, we all have a responsibility to avoid sharing upsetting images

If Twitter is the new journalism, we are all editors – and responsible for treating our fellow humans with dignity.

“I wish I hadn’t seen that”, my colleague says from across the desk. It’s been an hour since the first reports came in of a shooting outside Parliament, and the news agency Reuters has started posting photographs of injured people, knocked down by the terrorist as he drove across Westminster Bridge.

In one, a brunette woman leans over a victim whose blood is beginning to stain the wet pavement. Lying on her back, she is framed by scattered postcards sold for tourists which have been knocked to the floor. She is clutching the arm of the woman helping her, but her eyes are staring dead into the photographer’s lens.

Another photograph – the one that my colleague is referring to – disturbs me even more: a man who has fallen (or been pushed?) off the bridge onto a stairwell. He is face down in a pool of blood, his left leg at an unnatural angle. It is impossible to tell if he is alive or not.

Briefly, before I scroll past, I wonder if someone, somewhere is seeing the same picture and experiencing a shock of recognition as they recognise their friend’s clothes.

And then there is one picture which I now cannot find on Twitter, but which, lying in bed last night, I could not stop thinking of: a woman’s legs extended from under the wheel of a bus, her skirt hiked up to show her underwear, her shoes missing.

We are a desk of journalists covering an attack on the Houses of Parliament, so I keep scrolling. It is only later, in an article by the Telegraph, that I learn a junior doctor has declared the woman dead.

Of course, the shock of seeing images like these is nothing compared to what war reporters, doctors or police go through on a regular basis. But a 2015 study at the University of Toronto found that extended exposure to violent or disturbing material can have a severe effect on journalists’ mental health.

The impact can be particularly confusing when one does not anticipate seeing violence.On social media, we increasingly encounter images this way: without warning and without a chance to steel ourselves. This is particularly a problem when it comes to members of the public, whose jobs don’t require them to look at shocking material but who can nevertheless be exposed to it just by virtue of using a social media network.

It is for this reason that, shortly after Reuters published their photographs of the Westminster victims, prominent journalists began posting asking their colleagues not to retweet them. Some protested the fact that Reuters had published them at all.

In today’s media landscape, news moves fast and social media faster. Where a picture editor would have previously had until their print deadline to decide which images to run, now photographers are able to send their work back to the office almost instantaneously, and editors must make a snap decision about what to release.

Deciding what images to use can be a difficult call – especially under pressure. On the one hand, there is the urge to not turn away, to bear witness to the full magnitude of what has happened, even if it is shocking and upsetting. On the other, there is the need to treat fellow human beings with dignity, and particularly to avoid, where possible, showing images of victims whose families have not yet been informed.

Social media makes this process even more difficult. Once released online, photographs of the Westminster attack were quickly saved and re-posted by private individuals, stripped of context or warning. One can choose not to follow the Reuters Pictures account, but one cannot necessarily avoid seeing an image once it is being retweeted, reposted and recycled by private accounts.

As the line between traditional news and social media blurs and we increasingly become participants in the news, as well as consumers of it, our sense of responsibility also shifts. On Twitter, we are our own editors, each charged with making sure we extend dignity to our fellow humans, even – especially – when the news is dramatic and fast-moving.

I was glad, this morning, to encounter fewer and fewer photographs – to not see the girl lying under the bus again. But at 3am last night, I thought about her, and about her family; about them knowing that journalists on desks across Britain had seen up their loved one’s skirt during the last moments of her life. It was, without putting too fine a point on it, no way to encounter a fellow human being.

Over the next few days, we will find out more about who the victims were. The media will release images of them in happier times, tell us about their jobs and careers and children – as is already happening with Keith Palmer, the policeman who we now know died on the Parliamentary Estate.

It is those images which I hope will be shared: not just as a way to resist fear, but as a way of acknowledging them as more than victims – of forging a different connection, based not in horror and voyeurism, but in a small moment of shared humanity.

There is no shame in being affected by graphic images, however removed one “ought” to feel. If you would like someone to talk to, Mind can provide details of local services.

The BBC also provides advice for those upset by the news.

Find out how to turn off Twitter image previews here.

Stephanie Boland is digital assistant at the New Statesman. She tweets at @stephanieboland