Running the numbers, it seems that the BBC's articles are most likely to skew positive about Brexit, while the Independent is the most negative.
In the middle, the Guardian and the Mail (the latter possibly due to MoS vs DM stances).
In the middle, the Guardian and the Mail (the latter possibly due to MoS vs DM stances).
So, how did I reach this conclusion? Readers of previous threads will know that I’ve been analysing the headlines of 2,000 Brexit articles and body text of 1300 items published since Jan 2018.
To keep the sentiment analysis very simple, I’ve looked for certain keywords relating to Brexit. Positive terms often used by pro-Brexit politicians (Opportunity, Success, Prosper, Ambitious, etc.) and Negative terms used by opponents (Risk, Damage, Disruption, Concern, etc.).
In aggregate, across the news items analysed, these keywords appeared 2400 times (with duplicate instances in the same article removed), split almost *exactly* 50:50 down the middle, which makes any skews across news outlets very easy to see.
So, first off, how often do news outlets use these positive or negative terms when talking about Brexit? Overall, 3/4s of articles used one or more of these keywords.
How does this break down by news outlet? Over 30% of BBC News #Brexit items had a positive keyword (but no negative), while the figure was 19% in the Guardian and Independent and 15% in the Mail. 
Many news outlets strive for some form of balance, so we can see in the Guardian that 34% of #Brexit items had both positive keywords and negatives, while in the Independent and BBC, the proportion of items with both sentiments was at ~20%.
The Mail was actually highest in this category (44% had both +ve and -ve keywords) – likely due to the structure of its articles. Its pieces tend to be longer, and thus have more opportunity for both negatives and positives.
This raised a methodological problem - longer items might skew towards appearing balanced, which would favour certain news outlets (like the Mail).
So by looking at the ratio of the number of unique +ve keywords in each article to the unique -ve keywords in each article, we can compensate for that…
Broadly-speaking, the trend holds. Nearly 60% of the keywords in BBC #Brexit items were positive, while at the other end of the spectrum, just 45% of keywords in the Independent were positive.
On this same measure, the Guardian skews marginally negative, while the Mail slightly positive - but both roughly in the middle.
So what’s going on here? Is the BBC really massively pro-Brexit? Well I don’t think so, but I believe it’s part and parcel of the same issues I’ve written about before – namely the absence of senior politicians who are challenging the Government narrative.
I previously found that across both Left and Right news outlets, Shadow Cabinet members appeared to be largely absent from the Brexit debate. See:
That’s not necessarily a problem for most newspapers – their editorial teams can still add opposing viewpoints and challenge the Government position…
…However, if the BBC’s approach to balance in politics is to tend to seek a counterpoint from an equivalently positioned commentator/politician (who in this particular case are missing), then its coverage will inevitably skew to the prevailing Government position on Brexit.
Perhaps something for the BBC editorial team to consider – how do they balance political debate in the unusual case of one side avoiding the discussion?
(Upping the excellent Reality Check article volume perhaps?)
(Upping the excellent Reality Check article volume perhaps?)
Data caveats: Unlike previous pieces, which had simpler data, this is piece is open to critique around ‘well what about keyword x’ or ‘what if a positive sentiment was preceded by the word ‘not’’…
…All entirely possible I admit, but given the size of the sample of data and the clarity of the trend, it’s probably unlikely that this would significantly change the overall pattern.
Previous pieces here:
