In this “Conversation with Authors,” we spoke with APSR authors Florian Foos and Daniel Bischof about their article, “Tabloid Media Campaigns and Public Opinion: Quasi-Experimental Evidence on Euroscepticism in England.”

APSR: What was the genesis of this piece?

Florian Foos: Daniel and I were both Ph.D. students in England when the report of the Hillsborough Independent Panel was released, and the case got a lot of media attention in this context. So that’s really where my interest in the case came from. That also resulted in the apology that The Sun issued many years after their false reporting on the disaster, just a couple of days after the commission panel report was released.

A couple of years later, Daniel and I ended up in Zurich together as postdoctoral researchers. In 2015-2016, in the context of the Brexit referendum campaign, the debate on Euroscepticism was everywhere. We used to watch a lot of football together, Daniel and I–and you know, obviously we are political scientists, so we started talking about politics. The Sun was extremely active in the referendum campaign, with lots of important front pages. And then I made this connection, I guess, between the fact that people in Liverpool really don’t read The Sun and what that could mean for their political attitudes towards the EU.

Daniel Bischof: It was on a car drive to your parents’ place that you brought it up for the first time. We wanted to go to the region in Germany where Florian’s parents are from – Rhineland-Palatinate – that’s quite famous for wine to attend the annual wine festival, so it was on that trip where we came up with the idea!  

APSR: As you were writing this piece, was there something that most surprised or confounded you? Did you come in with any strong priors, or was this an exploratory piece?

Florian Foos: I think the prior was that The Sun would fuel Euroscepticism, since it’s the most important tabloid, widely read in the U.K, that had been campaigning on leaving the EU for a long period of time. And that is what we find.

Empirically, one thing that surprised me, to be honest, was that most of the effect materializes in the first two years, directly following Hillsborough and the onset of the boycott. I expected a more gradual decrease in Euroscepticism over time. The way we make sense of it after the fact is that there was a lot of Eurosceptic coverage in the tabloid directly surrounding Hillsborough, due to the Sun’s feud with Jacques Delores, the then-Commission President, and Thatcher’s famous “No, No, No” speech in 1990. The late 1980s/start of the 1990s was the first time the issue really became highly salient in the UK. There was little positive coverage at the same time, so the Eurosceptic tabloids had the field pretty much for themselves. Once attitudes are formed, they would be more difficult to change going forward.

Another thing that surprised me was just the attention that the paper received when we first presented a draft at the American Political Science Association Meeting (2018). There was actually a Financial Times journalist who contacted us who had gone through the conference program, looking at papers relating to the UK or Brexit. He contacted us to write a piece about the article, and when that was published we received quite a few emails. From people based in Liverpool, telling us their experiences with the boycott–in particular, there was one academic based in Liverpool who contacted us after having seen the presentation at a conference in 2019, Stuart Wilks-Heeg. He was very helpful in providing resources on the ground, putting us in touch with research assistants who then conducted the survey of news agents that we did in Merseyside, Lancashire, and Cheshire. And also sending us materials, helping us to contextualize everything, since he fully understood the local dimension to the story. I think the attention it got and the help we received from people in Liverpool was quite remarkable.

Daniel Bischof: I was quite amazed at how strong the norm of the boycott was even today. Basically, for the R&R we were asked by the reviewers that we ourselves should go into the field and try to measure which kind of shops actually were selling The Sun and what their reaction is if asked. And some of the reactions we found we put in the paper, and I found it quite striking. I knew the boycott was key to Liverpool’s identity, but I did not expect the emotional response to be so strong. This was something we realized when we went into the field for the more qualitative part of the beginning of the paper.

APSR: Can you talk a little bit about the change from the first submission to the article that is eventually published in the APSR? How is the piece stronger after peer review?

Florian Foos: I think the main empirical analysis did not change much–the attitudinal data and the data on the referendum. What really the referees and editors contributed was to ask us for additional data collection on the strength of the boycott–so the news agent survey. And I really think that does make the empirical part stronger.

Daniel Bischof: Also, the content analysis of The Sun.

Florian Foos: Yes, the content analysis–which was hand-coding of the articles which were published in The Sun. I think the whole narrative became much richer and more detailed through that. I also think that this must be the first piece with quantitative data available on the strength of the boycott on Merseyside. You can see very clearly that that boycott is stronger in the city, in Liverpool. Basically, the further away from Anfield you move, the weaker it is. And I think it justifies some of the core assumptions that the boycott is geographically limited.

We also did news agent surveys in adjacent counties and you see that 90% of news agents do stock The Sun there. So, it’s really the validity of that key assumption that the boycott is limited geographically that we can now support empirically, with the survey. And that was based on reviewer’s comments.

Daniel Bischof: Also, we worked a lot on the theory, so I think this piece got much better due to reviewers’ comments specifically asking about the mechanisms, the key theoretical ideas, the specificity of the topic, etc.

APSR: This piece uses a difference-in-difference (DiD) design, and there has been a little bit of a discussion about how robust DiD findings are. Can you talk a little bit about the threats to causal inference using DiD, and what are some of the best practices people can use as they move forward trying to use this design?

Florian Foos: I think a lot of the statistical discussion at the moment is really about differential treatment timing, which doesn’t apply to our case, because it’s really a straightforward before/after type of comparison. So it’s a relatively simple DiD setup in our case.

In general, I think one key to a valid DiD is excludability: whether you can credibly argue that it is really the boycott that is causing that decline and not anything else that could have happened at the same time. We go to really some length in the paper to substantiate that. You know, things that might have happened differently in Liverpool than in other counties in the UK, around or just after Hillsborough, for instance, things to do with European subsidies. These are some of the questions I think you’d have to address.

Daniel Bischof: Specifically with regards to our case, it’s just one region. So people come up easily with arguments that apply to the region that can drive the effect. That’s a lot different from other DiD where a treatment occurs in 5 or 6 different constituencies, which is completely different from the story that we have here from a methods point of view.

Florian Foos: Yes, that does make uncertainty estimation difficult and we acknowledge that. We do everything we can, clustering at different levels, bootstrapping etc, but it would be better had there been 10 independent boycotts. Unfortunately, there weren’t. On the other hand, a strength of the design is that the treatment is plausibly quasi-random – that’s not strictly speaking necessary for a DiD to be valid, but it makes our case stronger. The independence assumption is quite credible, simply based on the fact that the boycott was not a function of the Eurosceptic coverage. Qnd that it doesn’t seem to be the case that people were reading The Sun at the time mainly because of their Euroscepticism, but rather they were reading it for entertainment value, celebrity coverage, football. In terms of actual politics, it only takes maybe 10% of the coverage of the paper, but it’s very outspoken when it does cover politics.

Daniel Bischof: And when they do cover politics it’s lengthy, it’s a couple page spread with their particular take on the E.U. It requires no knowledge from the person, you can know nothing about the E.U but you still have a negative image about it in your head when you see the cover image.

Florian Foos: Yes, the title pages are very famous, it’s what The Sun is famous for. Lots of graphic photos, quite scandalous in how it’s all made up.

APSR: Right, and the people who read The Sun might not be the highest information voters, so you just have this perfect storm of low information voters being presented with attention-grabbing graphics in the absence of strong counter frames.

Florian Foos: It’s also a case study in how issues become politicized. The salience of some issues is driven by political entrepreneurs who keep emphasizing them. And as you say the scope conditions are such that they should be favorable for effects to materialize.

APSR: Is there anything else you want your readers to know about this article and where do you see you taking it in the future?

Florian Foos: I can think of two angles as a researcher as to how this can be taken forward. One of them is that I’m interested in media influence more broadly, because I think you have a lot of interesting cases at the moment where you have entrepreneurs taking over papers, or where papers are expanding into the TV market. And an increasing number of papers that have a political agenda at the moment, such as the case of Fox news in the US or Australia with Murdoch. To study media influence a bit more comparatively, and with a focus on for instance other European countries (especially in Eastern Europe) is one thing I would like to do.

And the other part is really what persuasion or attitude change is and what it means. There’s a big debate going on about under which conditions persuasion is possible and what persuasion actually is. So the question is, under what conditions does influence actually materialize and transform into something that lasts. It’s a big question that a lot of people are asking and working on right now.

Daniel Bischof: For me it’s more about the social norm aspect. I just received two grants to study social norms, and I was always interested in social norms and how they differ across localities, and this paper really steered and pushed my interest into this even more. Trying to understand how specific localities have very different ideas about how they interact with each other, how they interact with elites and how they interact with political institutions. This is really something I want to study in the next five years: specifically trying to understand what role social norms play for democratic attitudes and support for democracy.

So the typical example of the storming on the capitol in the Us: You read the statements of the people that were involved, it’s quite striking what kind of idea they have about what democracy is and what freedom is. I want to understand where this comes from. What role do peers play, do political elites, does Donald Trump play in their idea of democracy? Whether they actually have a structured idea of what democracy is, that’s a storyline I want to follow up on. What are social norms and what role do they play in democracies?

Florian Foos: If you’re asking what message is in the paper but which hasn’t received as much attention, I would say it’s that public opinion is malleable, it can be shaped by political actors, and the media are one of them. It’s also a bit of a story about Euroscepticism and where it comes from. Obviously there has been a lot of coverage in the UK on structural factors. Is this about culture or the economy? What role does immigration play in the formation of Eurosceptic attitudes, especially amongst working class voters? I think we show there is a different dimension of that story. We are not arguing that Euroscepticism is solely because of the type of media coverage present in England, but clearly I think the Eurosceptic media in the UK have contributed to the formation of these types of attitudes. I think it’s a story that a lot of people intuitively believe but there was not that much robust evidence on that. Also, the DiD where we show that effects on working class respondents who were more likely to read The Sun before the boycott, are stronger is really important to that debate. I think that provides, basically, if not an alternative story, but a complimentary story to what other people are telling about the reason for the rise in Euroscepticism in the UK.

Daniel Bischof: I think it adds an important dimension to this research on populism or Euroscepticism in general. Most of the time what the current research does is it tries to give a lot of agency to the “people.” It suggests, “Well it’s migration and xenophobia more generally; that’s the reason they become Eurosceptics; people basically listen to their fears.”

But what we show is that part of this agency is really with the elites and how they communicate specific issues and how they pressure it onto the forefront of voters’ minds. In current research,  quite often there is this idea that there is some external threat of change that is building up and politicians should listen to that fear, so in that model the politicians are more like listeners to what the public wants. But here what we show is that elites drive what people think.

Florian Foos: Public opinion is not only something that exists sui generis, in a vacuum. There are many actors who try to shape it.

It’s also a cautionary tale, really, for pro-European elites in the UK and elsewhere. If you don’t engage, and you just leave the field for twenty years to the Eurosceptic media then it’s very difficult to counter that in a year, even if you have a balanced referendum campaign.

Daniel Bischof: It also shapes what other people think about the UK. So me, as a German, I hear often: “Oh they voted for Brexit, it’s their fault,” but there was a reason they voted for it, and even if voters have agency it’s not only their fault. It’s how elites shape these attitudes for twenty or thirty years. And this is something that is not on people’s forefront when people think about Brexit–they think the voters wanted it, they get what they deserved. But that is just too short of a story.

– Daniel Bischof, Aarhus University

– Florian Foos, London School of Economics