"Today’s Ipsos MRBI/Irish Times “Perceptions Research” poll that finds we, the citizenry of the State, are seriously deluded on a range of public policy issues about key facts which inform decision-making must be a matter of concern. The poll repeatedly begs but does not answer the question: “now that we have given you the correct answer, would you change your view on the substantive issue?” Presumably the answer for many is Yes, although, also for many more, it’s surely the case that where they have guessed at an answer, that guess probably reflects their prejudices and a desire to validate them. In that sense the poll is as much a snapshot of political prejudices as of public perceptions."
Today’s Irish Times is really irritating me. I read the poll results (at least in headline, abbreviated form with no detailed data, which I don’t really expect to be forthcoming) on the IPSOS/MRBI Twitter before I saw the gloss put on them by the political editor Stephen Collins, but I could see from the questions the obvious ideological bent. Their UK sister firm, IPSOS MORI, did a ‘Perils of Perception’ survey last year which went deeper into the perceptions of the British public about media-driven issues like immigration and welfare, which was an intriguing read that prompted me to think a bit about how ‘perception’ can be aggregated and measured. This survey is much more narrow, however, and sticks to investigating whether Irish people have the ‘correct’ interpretation of the economic and political situation.
I say ‘interpretation’, since although the items can be claimed as matters of factual knowledge, the framing and inclusion of the question versus the likelihood of accurate statistical knowledge, on the part of the average respondent, speaks volumes to me. Two questions are asked about the amount of state spending on welfare recipients: the first comparing the total spend to that on public servants or politicians, the second between the unemployed, pensioners and child benefit. In each case the correct answer is clearly the largest group, even though individually they are paid less - and it’s that confusion that seems to be borne out by the answers, not a conspiracy to not recognise the high cost of the welfare state. To the extent that the same results could have been achieved even with a better explanation of the question, would I think simply show the lack of genuine statistical understanding in the public discourse over ‘costs’ and individual burdens.
That’s the spending side: on the tax, it’s the old chestnut of the small group of high earners contributing (rightly) the largest share of income tax under a ‘progressive’ system. That people underestimate this - that the top 10%, according to the figures the survey is based on, pay 59% rather than 28% of all income tax, yet begin at €75,000 in gross income rather than €153,000 as suggested - to me illustrates an underestimation of income inequality, not of rich people’s unrecognised support for the state. That distinction is I suppose a matter of political debate, but that’s exactly what the use of this poll shuts down by implying that voters make wrong decisions because they don’t fully appreciate the facts, rather than the way the facts are shaped being part of the problem.
* * *
Previously, I tried to suggest that the ‘perils of perception’ were as much in the readings of opinion polls as in the opinions themselves. It’s too easy to take evidence of mistakes in statistics as evidence of mistaken belief, without considering other sources of error and difference. For example, I would be dubious about this apparently sincere concern:
"The findings on perceptions of ethnic background are perhaps the most worrying of the poll. Although only 12 per cent of the population consists of non-nationals, more than one in three of us believe that the non-national community represents over 30 per cent of the population. That view is shared by as many as 47 per cent of women and 42 per cent of unskilled workers and the unemployed. Fifteen per cent of respondents even said they believe one in two of us are non-nationals. Such misperceptions are deeply dangerous and could, if unchallenged by the Government, be the basis for the sort of worrying upsurge in racism seen elsewhere in Europe."
Or as Stephen Collins rather bluntly put it elsewhere in the paper, “poorer” people (as presumably measured by ABC1/C2DE social class) are more likely to be ‘wrong’ on this subject. But immigration in Ireland is not uniform - quite likely if you’re living in an inner-urban area, or a less-well-off suburb, immigrants will make up a larger proportion of the local population. Probably not quite as large as the perceived figures would suggest, and I don’t discount the involvement of a certain alarmism or social disquiet, but another factor to be considered I think has to be education: if your schooling, particularly in terms of mathematical literacy, is not as developed as it may have been in higher-class areas, how does that affect the ability to make a statistical judgement such as assigning a proportion to national demographics? It’s hard for anyone to visualise ‘30 per cent of the population’, and most will default to personal experience at some point, so what are we really measuring here?