How many immigrants per year should Switzerland be prepared to welcome? Do the figures put forward by political parties and conveyed by the media play a role in influencing public opinion? Psychologists from the University of Geneva (UNIGE), Switzerland, set about testing a well-known reasoning bias - anchoring bias - that consists in providing a deliberately low or high figure for information purposes before respondents give their opinion on a subject. The researchers found that the figure supplied played a vital role in shaping respondents' opinions, regardless of their political orientation or the political party responsible for the figure. The research, which is published in the Journal of Applied Social Psychology, points out the dangers of number-based political votes, since public opinion is easily swayed by the first number conveyed in the media.
The issue of immigration has dominated the political landscape in recent years. Some countries, such as Germany, have opened their doors, while others such as Hungary have been more reluctant. But what is the situation in Switzerland? In 2016, the country was home to about 8 million so-called permanent residents, including 2 million foreign nationals. In addition, there were 90,000 non-permanent residents and 68,000 asylum seekers, around half of whom had arrived the previous year and were still in the asylum process.
"We calculated that the median point between the number of new immigrants arriving annually in Switzerland and the total number of immigrants in the country is around 50,000," explains Fanny Lalot, a researcher in the Psychology Department in UNIGE's Faculty of Psychology and Educational Sciences (FPSE), and first author of the study.
The importance of anchoring bias
The UNIGE team investigated whether the figures communicated in the media about the annual number of immigrants arriving in Switzerland influenced public opinion. How many refugees would the Swiss be prepared to receive?
"A widely-known reasoning bias has been analysed since 1974, namely anchoring bias," says Fanny Lalot. "This means that when you ask someone to evaluate a situation, you can influence his response by giving him a low number (or, by contrast, a high number) that will at first serve to 'anchor' his thinking." If the set-point number is low, the interviewee's response will also be low. But if the number supplied is high, the respondent's evaluation will also be high. "To address our question about immigration, we started with the figure of 50,000 to choose our anchor points: 1,000 for the low number, and 100,000 for the high figure," continues the psychologist. The researchers then submitted the following questions to a group of 300 people: "Politicians from one political party have proposed that Switzerland should welcome 1,000 immigrants. Do you think the country should welcome more or fewer immigrants? How many exactly?" A second group of 300 people were given the same questions but with the figure of 100,000 immigrants.
The results were clear-cut. "The average for the first group was around 20,800 immigrants a year. But this figure was six time higher in the second group, with an average of 126,000 immigrants a year! It just goes to show the importance of anchoring bias," says Fanny Lalot.
And what role do political affinities play?
Would the results be different if the anchor figure was clearly issued by the Swiss People's Party (UDC) or Socialist Party? The psychologists divided the participants into four groups. The first group was given a recommended figure of 1,000 immigrants issued by the UDC. The second group also received the figure of 1,000 but from the Socialist Party. The third group received a figure of 100,000 immigrants from the UDC, and the last group a figure of 100,000 immigrants from the Socialist Party. "To our surprise, the averages of participants' assessments were almost the same between groups 1 (20,000 immigrants) and 2 on the one hand (15,000 immigrants), and 3 (140,000 immigrants) and 4 on the other hand (130,000 immigrants). This proves that the political source of the anchor figure does not matter for the public's assessment; only the number itself counts," says Fanny Lalot.
There was another surprise in store for the researchers: people leaning towards the right answered with a figure that was higher than the one initially suggested, respectively 18,000 (25,000 for left-wing participants) and 100,500 (160,000 for left-wing participants). "This result was striking. In general, participants who were given a high anchor figure tend to answer with a lower number than the one proposed in the preamble. Here, it was higher!" says Fanny Lalot. "One can imagine that this issue provoked a somewhat more humanistic reaction, which encouraged people to defend higher figures than those allegedly proposed by a politician."
Be careful with manipulated numbers!
The UNIGE study shows that the Swiss public is ready to accept as many, if not more, immigrants as political parties advise. But the results mostly point out the dangers of votes based on numbers. Because, regardless of political affinities, citizens are strongly influenced by the numbers conveyed in the media. "Since this is a relatively well-known principle, political parties could use it to try to influence public opinion by being the first to communicate figures that best serve their interests," says Fanny Lalot. "It's up to us to take this bias into account and not let ourselves be swayed, especially on matters as sensitive as immigration!"