What we talk about when we talk about immigration

In 2003, the Press Complaints Commission determined that the term "illegal asylum seeker" was legally inaccurate and should no longer be used. It is impossible for an asylum seeker to be illegal, because under human rights legislation everyone has the right to seek asylum. The ruling has not been entirely effective, and the term continues to slip into use - recently the conservative Australian immigration minister Scott Morrison told his staff to call asylum seekers who arrive by boat "illegal maritime arrivals".

Situations like this make you wonder how the language we use to talk about immigration affects how we think about immigration.

I work at a research centre that focusses on immigration, where quite naturally we tend to keep up with news on immigration. Following press coverage of immigration has taught me a lot about how the media works. I've become frustratingly aware of how the media is 'news' focussed, leading to media frenzies over things like the Lampedusa boat tragedy or the stowaway who fell from a plane over London. News like this is exciting, but it doesn't really tell you much about immigration in general. Most immigrants are not illegal, and most come through traditional transportation routes.

An academic at my centre did a research project on public opinion toward immigration. One of the findings was that the immigrants people are most concerned about are asylum seekers, while they show little concern about students. This concern seems misplaced when you look at actual numbers of migrants. In recent years, there have been fewer and fewer asylum seekers, but high and increasing numbers of students. This lack of concern might be because people assume students don't stay, but in fact some studies show that up to 20% of students are still in the country five years after graduation.

When public opinion seems out of sync with reality, you wonder why. Once again research offers some insights. My colleagues recently finished a big project on migration in the news, focussing on the language British newspapers use to describe immigration. They used big data techniques to process large amounts of data (58,000 newspaper items containing about 43 million words pieces appearing in 20 national UK publications from 2010 to 2012) and to minimise human bias.

The results were probably as one would expect - among other findings, they were able to demonstrate that "illegal" was the most common modifier of immigrants (even though most immigrants are not illegal) and "failed" was the most common modifier of asylum seekers.

Fortunately "illegal" as a modifier of asylum seekers has become less common, but newspapers continue to call asylum seekers "illegal immigrants" and to imply that immigrants are predominately illegal.

It would be nice if the industry would try harder to eliminate the more blatant forms of press bias, but we can't really do anything about the 'news' focus. The press likes to tell stories about extreme situations because we like to read those stories. But we need to remember that most news is news because it is new and, more importantly, unusual.

To fight the news bias I find it helps to develop a sense of context. Context is a general sense of the world built by applying memory and critical thinking to the news and analysis you consume. Context tells you that not all Muslims are terrorists, even though the ones in the news mostly are. Context tells you that not all people on benefits are undeserving scroungers, even though the Daily Mail tells you they are. And context should tell you that not all immigrants are illegal, even though the news coverage gives that impression.

But sometimes you also have to fight against context - sometimes you have to check to make sure your perceptions, developed from a consistent diet of faulty news coverage, are actually real. Academic research is good for that.

Research tells us that 13% of people in the UK are immigrants, not 31% as on average the UK public believes. And Black and Asian people make up 11% of the population, not 30%. And Muslims make up 5% of the population, not 24%.

We can't just blame the press. The information is out there, we need to look for it.