The boycott of The Sun in Merseyside, the county covering Liverpool, after the 1989 Hillsborough disaster cut survey-measured support for leaving the EU there by about 11 percentage points compared with other northern English areas, according to a study in the American Political Science Review. Its difference-in-differences designs, which compare how two groups move over time rather than where each one starts, use public opinion data spanning three decades supplemented by referendum results and “show that the boycott caused EU attitudes to become more positive in treated areas”.
The paper is by Florian Foos and Daniel Bischof. When it was published in 2021 it listed Foos at the London School of Economics and Political Science and Bischof at Aarhus University and the University of Zurich.
The disaster was a human crush at Hillsborough stadium in Sheffield on 15 April 1989, in which 97 Liverpool Football Club supporters lost their lives, the paper says. A footnote records that initially 96 people died in the disaster and that the number was later updated to 97, citing a 2021 Guardian report. The paper describes “biased and slanderous reporting” by The Sun on the disaster and its victims, which led to a sudden boycott of the paper in Merseyside.
So does any of that reach the 2016 referendum?
The attitude estimates come from the British Social Attitudes survey, an annual cross-sectional survey of opinion in Britain, used here from 1983 to 2004, the last year it carried parliamentary constituency identifiers. Of 35,204 respondents, 908 lived in one of the 15 constituencies inside Merseyside, the paper says.
Two different 11-point numbers sit in this study, and they are not the same thing. Self-reported Sun readership in Merseyside fell from 18 to 7 per cent after Hillsborough, an 11-percentage-point drop against “a precisely estimated 0 percentage points” of change in other northern English counties.
The second is the opinion effect. Across the models the estimated fall in Euroscepticism ranges from 7 to 17 percentage points, and the authors’ preferred specification, model 6, reports an 11-percentage-point decrease, on 10,384 observations across 172 constituencies covering 1985 to 2004. Most of the decline occurs between 1989 and 1994, the paper says.
In Merseyside, respondents born after 1972, who were 16 or younger when the boycott began, were significantly less likely to read The Sun and to be Eurosceptic. Outside the Merseyside region that cohort was overall more likely to do both, the paper says.
The fall in Euroscepticism was largest among unskilled and semiskilled working-class respondents, the group most likely to have read The Sun before Hillsborough.
“This is highly notable, as this is the part of the population who would usually have read The Sun in Merseyside,” Foos says in the Aarhus BSS write-up of the study.
One thing the design cannot do is separate two possible causes. Readership in Merseyside shifted toward pro-EU papers, and the Daily Mirror became by far the most popular national paper there after Hillsborough, so the models cannot empirically distinguish non-readership of The Sun from substitution with pro-EU material in The Mirror. The authors also say they cannot identify individual-level mechanisms with repeated cross-sectional data.
They call it “highly unlikely” that the drop in Euroscepticism was in large part due to the Mirror’s pro-EU coverage rather than the absence of The Sun’s Eurosceptic content, noting the Mirror published one editorial containing European integration keywords between 1996 and 2000 against The Sun’s 62.
The bigger question for the design is what else changed in Merseyside after 1989. The paper calls this the “no compound treatment assumption” and names three threats to its inference: rising Labour support under Tony Blair, EU structural funds and deindustrialisation.
On each it offers a response. Merseyside was already a Labour bastion by the late 1980s and the results survive matching respondents on party identification, the paper says. Merseyside became a priority region for EU structural funds only from 1994, after the largest drop in Euroscepticism, and controlling for the offshorability of respondents’ jobs leaves the findings unaffected.
For the referendum itself, the paper estimates that Merseyside counting areas were around 8 to 9 percentage points less Leave voting in 2016 than other northern English counting areas, measured against the 1975 vote. In its conclusion the authors write that the boycott “might have decreased” the Leave vote share in Merseyside.
A boycott that began as a refusal to buy a paper may have shifted how a region felt about the EU. The study is careful about what it can say on the 2016 vote, and I want the title to be just as careful.
The Aarhus BSS write-up carries the same caution, in a line attributed to Bischof: “We do not have the scientific evidence to proclaim that the UK would still be part of the EU if The Sun did not exist. But we can safely say that the Merseyside boycott made local citizens considerably less Eurosceptic.”
What the sources cannot settle: The Sun does not release circulation data at county level or below, so the authors are unable to estimate effects on actual readership in Merseyside and the measure is what people said about themselves. They note self-reports can be a function of social desirability bias, and say that in this case such bias would confirm a strong social norm against reading The Sun in Merseyside. Guardian figures quoted in the paper, a fall from 55,000 copies to around 12,000 in Merseyside, “cannot be independently verified”.
The opinion series also stops in 2004, and the estimates are local to Merseyside and the North of England. Whether they travel elsewhere “must ultimately remain an empirical question”, the paper says.
To measure the boycott today, the authors ran a telephone survey of newsagents and cornershops in Merseyside, Lancashire and Cheshire in January and February 2021. It found 62 per cent of Merseyside newsagents reported not selling The Sun, rising to 86 per cent in the city of Liverpool, while 90 per cent of Lancashire newsagents and 88 per cent in Cheshire reported selling it. The response rate was 25 per cent, on a sample of 165 shops.
The Sun apologised in 2012 under the headline “Hillsborough: The real truth”, calling its reporting the “gravest error” in the tabloid’s history. The paper says the boycott is ongoing to this day.