What are we talking about when we talk about migrants?

As a journalist I’ve long been fascinated by the interplay between public opinion, politics and migration policy, and you can’t unpack the complexities of public opinion without talking about how migrants are portrayed in the media, as Amy Clarke noted here last month. What we read or hear feeds our world view, but the exact mechanics of that relationship are multi-directional, multi-faceted and under-studied.

One of the most interesting projects putting media handling of migration under the microscope is the University of Oxford’s Migration Observatory, part of the Centre on Migration, Policy, and Society (COMPAS). For some of the background on their work, I spoke to researcher Will Allen. He investigates the language used in the UK national press to talk about migration, as well as how civil society organizations use data and evidence about migration to communicate their messages.

The Migration Observatory has spent a lot of time looking specifically at how newspapers cover the migration story. Why?

The Observatory’s interest in newspaper coverage of migration came partly out of one of the most popular reports on our website that focused on public opinion. Former director Dr Scott Blinder did a survey in 2011 that asked people in the UK “Who do you normally think of when you think of immigrants coming to and living in Britain?” Almost 2/3 think of asylum seekers and less than a third think of students, even though students by the actual numbers are much more common. That leads naturally to wondering where people get their perceptions about who immigrants are. One possible answer is through media.

To explore this, we needed to describe what the media actually said. In the UK, newspapers play a bigger role in public debate than for example in the United States. When we scoped out previous work done on this, we found small studies over specific time periods. However, there were few systematic overviews of a large part of the UK press over a long period, with exception of a key linguistic study done at Lancaster University. So our ‘Migration in the Media’ project grew out of a desire to provide a large evidence base on what the UK press had actually said, rather than rely on assertion.

 How did you do it?

We downloaded data in monthly batches from Nexis UK, looking at all national newspapers in the UK for three years, starting in January 2011. Replicating the model used by researchers at Lancaster, we adopted a wide search string including all references to immigrants, migrants, asylum seekers and refugees, across 20 national papers including tabloids and broadsheets.

We included all sections of the papers: sports, arts, letters to editor, politics. Our argument is that people get their information about migrants from many different areas – for example, a review of a play that interviews a refugee, or a sports story about an athlete who comes from another country, as is the case with the British Olympic gold medallist Mo Farah. Then we used linguistic software to look for patterns of words. For example, to find out what words tend to be associated with immigrants, we searched for all the words that modified the word “immigrant.”

Will Allen. Photo credit: The Migration Observatory

Will Allen. Photo credit: The Migration Observatory

What did you find?

What we found is that the most common way of describing the word immigrant is “” It was the most frequent phrase across all publication types. The mid-markets – those that lie between tabloids and broadsheets – used it more than the others. In some ways the study confirms a common assumption, that characterizing immigrants as illegal is fairly common. But it challenges the assumption that this only happens in the tabloids and mid-markets, because we found that it is shared among all the publications, including the broadsheets.

We also found that the quantity of coverage of different groups tends to match public opinion. For example, Scott’s survey shows people don’t tend to think of students, and the press generally doesn’t mention students in relation to immigration unless it is the context of illegality.

Since the 2013 report we’ve more than doubled the dataset. We’ve also started using a more advanced software called The Sketch Engine that allows us to be much more nuanced and to cut the data more finely.

What do you do with that data?

We now have a comprehensive, systematic data set of tens of millions of words, which is a pretty large evidence base to work from. However, I don’t want to oversell the quantitative approach to texts. It’s not just about letting the computer run with statistical approaches to text: the team at the Migration Observatory, as well as COMPAS where we are based, also bring a lot of knowledge about the debate and the nature of the press in the UK.

What analyses have come out of this?

We just released a new report last month on media portrayals of Romanians and Bulgarians in the year leading up to the removal of labour controls in those countries. We found that language related to Romanians and Romania was strongly associated with criminality, homelessness, and poverty. We didn’t find those associations with Bulgarians. We also found a range of numbers used in association with these groups that appeared to show different estimates of the anticipated number of migrants that would come to the UK once controls were removed, which showed a lack of clarity and consistency.

Who’s your audience for this data, and how do you reach them?

All of the work we do at the Migration Observatory is intended to inform debate about international migration to the UK. That debate happens in lots of different spheres – policy, Parliament, non-governmental organisations, media, and public attitudes and perceptions. Our outreach aims to work with different stakeholder groups – civil society, politicians and policymakers, other academics, journalists, bloggers and people who write about immigration issues in their professional work. The data is useful for academics and policy makers and also for civil society organizations in their advocacy work. They can point to solid research that documents what is said in an empirical way. We post everything to our web site, and our team is frequently doing interviews and news spots, as well as presenting work at gatherings like the EU Commission.

You’re also doing research on visualizing data. Tell me more about that.

We’re always looking for new ways of communicating our work. We try to make sure all of our briefings have a chart or table and certainly a link to the original dataset so people can access it. We try to write in an accessible way and visualize the data in a way that’s shareable, understandable and visually appealing. For example, we have produced a set of maps from the 2011 UK Census that show the distribution of non-UK born people across local authorities in England and Wales.You’re also doing research on visualizing data. Tell me more about that.

What we’ve found, however, is that while charts and tables are an important part of how we operate, there’s a lot of scope to use other kinds of visualization to communicate our research into migration. One of the projects I’m involved with right now, called ‘Seeing Data’ looks at, among other aspects, the emotional impact of visualization. What feelings do people have when they look at certain charts? Why do some types work better than others? These are questions we’ve thought about, but we haven’t had solid research to back up why a certain visualization might be better than others.

But what ‘better’ looks like depends on what someone is trying to accomplish. Visualizations can be used for different purposes. One could be a call to action – for example, a visualization of gun crimes in the US might be a way of making people aware of the issue. In the case of the Observatory, we might define a good visualization by how easily people can drill down and use the data themselves. We want people to be able to use our reports. That influences how we view effectiveness. Can we do it better? Probably. That’s why we’re doing this research.

Migration in the news visualisation screenshot

Visualizing migration in the news. Image credit: The Migration Observatory

Are people talking more about how we discuss migration?

From an academic point of view, it has definitely grown. With new ways of collecting data – being able to access social media, for example – social scientists can look at public opinion from a much broader perspective. ‘Big Data’ approaches open the possibility of collecting data in a more systematic way from a lot more sources. Also, in the UK researchers are increasingly being asked to show the impact of their work. Working with non-academic users is encouraged. On the topic of migration, that often involves working with civil society groups. Understanding how certain stories about migration are publicly discussed is highly important for these groups.

I think it’s important to examine the links between where people get their ideas about migration and how these ideas are shaped by different sources. Then secondly, how do these perceptions inform changes occurring at policy and electoral levels? Finally, for me, the biggest, broadest level is examining how these perceptions and different conceptions of migration contribute to wider social change. Understanding the intersection and interplay of media perception, public opinion and policy outcomes is crucial.

What reaction have you had from journalists?

For the ‘Migration in the Media’ project, we had a good amount of coverage – including in The Guardian and the Press Gazette, which is an industry magazine. Since that report was fundamentally a study that put the press under a microscope, we had to have realistic expectations. Other Observatory outputs have appeared in tabloids, broadsheets, and television media, including recently on the front page of the Financial Times. Our Head of Media and Communications Rob McNeil plays a crucial role not only in liaising with journalists directly but also in actively placing the Observatory as a key independent source in the UK’s migration debate.

Do you have specific questions you hope your audience will bring with them when looking at your data?

We’re hoping people will engage with the data and set aside pre-conceived ideas that they might hold about how media talk about migration. I also hope they hold us and other organizations to account and say ‘show us your data’. Are the methods they use sound? And is the analysis sound as well? This is important when talking about a contested issue like migration. If someone is making a statement for or against a policy, or promoting a certain agenda, I would hope that people will ask what evidence they have for their claim, as well as if they have considered any trade-offs or limitations involved in implementing their position. Perhaps it is a bit idealistic, but that’s why I do this work: I would like to see debate about migration, as far and as appropriately as possible, to be based on sound evidence and deliberate debate.



William Allen @williamlallen

Scott Blinder @ScottBlinder

Migration Observatory @MigObs



Further Reading

Allen, William (2012). Whose Dreams Are Alive? Thinking about News Media as Stories. COMPAS Blog: University of Oxford.

Allen, William and Scott Blinder (2013). ‘Failed’ Asylum Seekers and the UK Press. Border Criminologies: Centre for Criminology, University of Oxford.

Vicol, Dora-Olivia and William Allen (2014). How Has the UK National Press Described Bulgarians and Romanians? British Politics and Policy: London School of Economics and Political Science.

One comment

  1. […] Taylor interviewing Will Allen (October 2014) ‘What are we talking about when we talk about migrants?’ The […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: