Erroneous and “fake” opinion polls have been withdrawn in recent weeks, but not before they had time to raise fundamental questions about the use of opinion polls: How many polls do “we” need? And what do they do with the political conversation? In a comment in M24, Andreas Halse in Tankesmien Agenda advocates toning down the “tyranny of polling”, which he claims “bombards” the election campaign and “reduces politics to a sporting event”. In Dagsnytt Atten he increased the criticism and characterized sloppy reporting of opinion polls as “a democratic problem”. At the same time, figures from Pollofpolls show that the “bombarding” is returning. The database has recorded a decreasing number of surveys with questions about municipal or county council elections in local election years: 2011: 523 2015: 486 2019: 439 2023: 337 – Even if we take into account the fact that there are fewer municipalities and that there are still some days of the election campaign, it looks like there will actually be fewer polls, says Lars Øy, who is one half of the two-team team behind Pollofpolls. The other half is Johan Giertsen: – Surveys raise the temperature and engagement in election campaigns, which is good. Surveys also provide a lot of important knowledge about the journeys of voters, he says. – Reminds me a little too much of sports journalism. Other political scientists and election researchers news has been in contact with are less enthusiastic, and point out that “random noise in the data” can lead to big headlines that have no basis. Like Halse, several of them take as a starting point three traditional objections to the use of opinion polls: They are based on a basic metaphor that politics is a kind of horse race, and appeal to the lowest form of journalism. Symbolized by margins of error being under-communicated. They shape the same landscape they set out to observe, and can influence voters who may have a conscious or unconscious desire to belong to the “winning team”. The methodology is not “neutral”, but in reality is based on choices and discretionary assessments. On top of that, voters have an annoying tendency to think incorrectly when they state which party they last voted for. The result is that the “weighting” comes out skewed. In Dagsnytt Atten last week, election researcher Bernt Aardal stated that “there are probably a bit too many measurements”. Photo: Terje Bendiksby / NTB scanpix Peter Egge Langsæther, who researches voter behavior at the Department of Political Science, tells news that he “agrees in parts with Halse’s criticism”: – There is sometimes too much focus on individual polls and it is a bit too a lot about sports journalism, he says. – They would like to cut down a little on the amount of cases that only deal with individual measurements with small or inconspicuous changes from the previous measurement. He adds that he “has an anecdotal impression” that the media have become better at presenting margins of error and holding back when they present opinion polls. – I would probably like to cut down a bit on the amount Peter Egge Langsæther, political scientist – My impression is that journalists and newspapers have become better over time at presenting margins of error and holding back when they present opinion polls, even if there is still a lot of sin against this. Opinion polls have their place – when used correctly, they can help us understand why voters move as they do and why elections end as they do. I think the media can use opinion polls to give readers relevant and interesting information when they are willing to dive a little deeper, compare information from several opinion polls, and beyond. At its best, the media here make us understand political developments better. But they could probably cut down a bit on the amount of cases that only deal with individual measurements with small or inconspicuous changes from the previous measurement. Jørgen Bolstad, political scientist – Most people will probably agree that voters should ideally focus on political matters and principles, and not care too much about who is going forward or backward in the short term. In this sense, today’s focus on opinion polls may seem somewhat exaggerated. At the same time, it is valuable both for voters and politicians to get a general picture of what election outcomes are likely. It also helped to create a certain predictability for everyone who has to take care of political authorities. What is potentially problematic is when small differences from one measurement to another get a lot of attention – both because small changes are not really that interesting and because they are not necessarily real either. Eirik Løkke, Advisor at Civita – I generally think there is too much focus on opinion polls. It is obviously interesting for the public to know which way the wind blows; nevertheless, the extent of opinion polls means that the cases and arguments end up in the shadows, to the advantage of which party wins and which for some of those who lose. The election campaign should be much more about what the parties think and their rationale. That said, I think that the voters have a greater understanding of opinion polls today than before, but there is equally reason to believe that opinion polls influence the voters in the form of them wanting to be on a winning team. The scope and use of opinion polls during election campaigns is one of the reasons why I have argued that we should hold elections every four years, and not every other year, through a joint election day for parliamentary elections and local elections. This will mean that the party can think more long-term and less about the next election and the next opinion poll. Aslak Bonde, commentator – When I was younger, I was busy criticizing the measurement tyranny. With many of the same arguments that Halse uses. Now I’m not so sure about the meaning of these measurements anymore. There are so many of them, and we have had this phenomenon for so long that I think the vast majority of voters know what they are measuring and what they are not measuring. At the same time, it will always be the case that opinion polls provide a basis for tactical voting, but one would hardly have gotten away with that if there were no opinion polls either. Johan Giertsen, Poll of polls – I think Norwegian media generally cover surveys seriously and in a way that inspires confidence. Lots of great measuring material in local newspapers around the country. And surveys drive up the temperature and engagement in election campaigns, which is good. Measurements also provide a lot of important knowledge about the journeys of voters. Bjarte Folkestad, political scientist – The weighting problem is well known. This concerns, among other things, that individual groups of the population are systematically underrepresented in surveys (FRP voters, for example). The fact that we have fragmented party systems around the country (many parties) does not make it any easier to measure party support either. In general, opinion polls are good at capturing national trends, and the closer you get to election day, the more precise they become. But this applies especially at national level, and the observations also become more precise when you combine the various measurements Professor of political science, Jørgen Bølstad, has programmed the website Estimite, which collects and analyzes public opinion data. The aim is to have a “more truthful speech” about which way the political winds are blowing. Photo: Caroline Utti / Caroline Utti – The scope of opinion polls causes the cases to end up in the shadows In Dagsnytt Eighteen last week, election researcher Bernt Aardal stated that “there are probably a bit too many polls”. He added that there are “many sources of error even if everything is done according to the textbook” and that the media at a general level are not good enough to take the margins of error into account. – It is something many people understand, but do not take the consequences of. Within the margin of error, but beyond comprehension, to put it somewhat polemically. He added that he wanted more surveys about what issues people are concerned with, and not necessarily which party they lean towards. – Then we would have a better basis for understanding why things happen. Andreas Halse in Tankesmien Agenda argues for downplaying the “tyranny of polling”, which he claims is “bombarding” the election campaign. – Not so sure about the significance of these measurements Eirik Løkke in the liberal think tank Civita, points to the same: – It is obviously interesting for the public to know how the election wind is blowing. But the scale of opinion polls means that the cases and arguments end up in the shadows, to the advantage of which party wins and which for some of those who lose. Commentator in Morgenbladet, Aslak Bonde, has followed Norwegian politics for a short lifetime and has been out on a winter night before. – When I was younger, I was busy criticizing the measurement tyranny, he says to news. He adds: – Now I’m not so sure about the meaning of these measurements anymore. There are so many of them, and we have had this phenomenon for so long that I think the vast majority of voters know what they are measuring and what they are not measuring. – We need more surveys Svein Tore Marthinsen, election analyst – I think the media should spend much more money and far more resources on buying good and far more surveys. And train skilled journalists with an understanding of numbers who are able to present them in a good way. That’s because good surveys create a living democracy that gives voters great opportunities to give party and politicians feedback between elections. And because measurements can provide important route information for both politicians and voters. Unfortunately, the field is far too low-priority in the media industry, there is not enough developed expertise and the media is often the cheapest solution – as we have seen – rarely the best. And a number of municipalities are rarely or never measured. We therefore need more measurements, more expertise in the field and more focus on averages and trends over time. It is not always easy to know how the various elections work. Ellen Karin Moen answers the things that several people have asked about.
ttn-69