‘The problem in the Mediterranean’ and the problems with the Irish Times Poll

By El Roto

The poll results published by the Irish Times today in its story headlined ‘Poll: Majority against taking in fleeing migrants’ present an appalling picture of Ireland’s attitudes toward what the paper refers to as ‘the problem in the Mediterranean’.

‘The problem in the Mediterranean’ relates to the fact that this year alone, between 1,500 and 1,750 women, men and children have died in attempts to cross the Mediterranean Sea from North Africa into Europe. The Secretary-General of the International Maritime Organization, Koji Sekimizu, predicts that half a million people could cross into Italy this year, by comparison with 170,000 last year, of whom 10,000 died in the attempt. (source)

The headline and poll data, on the surface, suggests a striking depth of animosity felt towards migrants in Irish society. There is no point seeking to improve it, or redress it, in accordance with the terms in which it is presented. However, that does not mean we should accept the data at face value.

It’s worth highlighting several key features in the poll and the accompanying report.

The question that provides the headline is ‘Should Ireland offer to resettle migrants rescued in Mediterranean?’ The Irish Times says 52% of those polled said Ireland should not offer to do so, and 48% said that it should. As others have pointed out, there is no ‘Don’t Know’ option for this question.

The absence of a ‘Don’t Know’ option is worth reflecting on. A government minister is unlikely to know what to do in many situations in the absence of further information and advice. There would be nothing wrong with that: it is in the nature of problems that we start out not knowing how to solve them.

But the Irish Times question, thus presented, suggests that everyone already knows what to do about the issue. Or, to put it differently, it suggests that out of 1,000 people, not a single one did not know what to do when it came to resettling migrants.

There is a far more accurate term for ‘rescued migrants’: refugees. These are people seeking refuge -as a New York Times article in April puts it- from war and poverty in Africa and the Middle East, or, as Peter Sutherland in an Irish Times report on May 11th puts it, ‘escaping persecution’.

The question does not mention refugees, and it does not mention what it is these migrants are fleeing –including the responsibility of EU member states in producing the situation they are fleeing. It is, as with previous Irish Times surveys, a loaded question.

The survey seeks to gauge the level of support or opposition to resettling refugees rescued in the Mediterranean on the basis of how many people the respondent thinks it would be appropriate to resettle. But since it does not quantify how many ‘fleeing migrants’ there are needing resettlement, a low number is not necessarily indicative of strong opposition, and a high number is not necessarily indicative of strong support.

That is, if you think there have been 1,000 refugees rescued in the Mediterranean, then 100 -one tenth of the overall figure- does not necessarily reflect a reluctance to accept refugees, bearing in mind the collective responsibility of all EU member states in this regard. One’s familiarity with the figures involved, as well as the overall background story, will have an influence.

The survey data presented gives no indication of the contextual information provided to the respondents in order for them to give their answer. One piece of contextual information, for instance, might be the overall death toll to date this year. Another might be details of the persecution such people are fleeing. Another might be the fact that the Charter of Fundamental Rights of the European Union guarantees the right to asylum. I will leave it to others to judge whether this is likely to have been provided.

It is worth considering the conception to politics that such an approach to polling –where the respondent is not given data on which they can make their decision- might reveal. In democratic societies, citizens are supposed to decide upon how society is to be run. For this they need reliable information that they can evaluate, and they need time to deliberate and discuss. Under the approach to polling demonstrated here, however, what matters is not whether the respondent has evaluated the information and reached a decision accordingly, but what the implications are for the political parties that seek the respondent’s vote, since it is they who will ultimately decide the policy to pursue.

The accompanying editorial says that ‘Fine Gael and Labour show significant positive support (61 and 56 per cent respectively) while Sinn Féin supporters are most strikingly resistant (70 per cent against)’. This gives a false impression of what various party supporters actually think. There were 1000 people surveyed. This is enough to give a reliable enough estimate of overall support for a political party. But it is not enough to give a reliable picture of what voters for that political party actually think.

For example, the poll finds that Labour is on 7%. That means that 70 respondents said they would vote Labour. Whereas Sinn Féin is on 21% support, and so 210 respondents said they would vote Sinn Féin.

To get an adequate picture of what the 220,000-odd Labour supporters think, using, for example, a 3% or so margin of error, you would have to survey around 1,000 Labour supporters, and roughly the same number of Sinn Féin supporters.

So, in neither case is the number of respondents adequate for a rigorous survey of Labour and Sinn Féin supporter opinion, or a rigorous survey of the opinion of supporters of any other party. That does not prevent the Irish Times from presenting it as such. It may well be true that the ‘Government is not substantially out of step with the majority of its support base’, as the Irish Times claims, but the survey data does not provide reliable enough evidence for this claim.

There is a more insidious aspect however. It ties in with assumptions underpinning the current referendum campaign, namely, the idea that the rights of persecuted minorities are a matter for majority opinion, without any reference to fundamental guarantees enshrined in law. Similar assumptions underpinned a survey from October last year, also covered by Stephen Collins, that sought the opinion of the electorate, and had the headline Majority are in favour of direct provision, poll finds.

Here, too, the Irish Times cites what ‘the electorate’ thinks. But this electorate does not include a large number of people living in Ireland who might have an opinion on the matter, i.e. migrants from other countries.

Such migrants are by and large irrelevant to the matter of how the Government can go about its business, because they are unable to vote in legislative elections. And so the Irish Times ignores them outright, whilst affecting to be worried about ‘the sort of anti-immigrant party that has poisoned the politics of many European states’.

The grim irony is that the basic standpoint underpinning the Irish Times poll questions and conclusions –one of a national electorate that vests a legislature with the power to decide which categories of people deserve to live and which can be left to die- is precisely that of the anti-immigrant parties it decries.