Whether polls influence or not is one of the most common questions usually asked during election times.
A quick answer is yes, they influence, for example, politicians, who make decisions and prepare their speech considering survey results; and it is also evident that they influence political commentators, who guide much of their analysis by what the polls show.
But how much and how they influence voters is something on which polling science has yielded mixed and inconclusive results.
For decades, what has been used most when talking about the influence of surveys is the binomial of the effects bandwagon and underdog.
The first means jumping on the winning bandwagon: some voters decide to support whoever is ahead in the polls. Based on this belief, not entirely proven, politicians spread propaganda surveys in their favor, and some candidates end up believing their own data.
The other effect means supporting the candidate who is trailing in the polls. If the effect bandwagon It is difficult to prove, this is much more, and it makes less sense in an environment of electoral competition, but it is there as part of the demographic imagination.
Almost no theory of voting gives weight to polls in their explanations of why and how people vote. The perspective of strategic voting, or useful voting, as it is usually called in our country, is perhaps the one that gives them the most weight, proposing that some voters, when they perceive that their preferred option does not have many chances of winning, support a second option. as long as the least desired option does not win.
But even from that perspective, the influence is not exerted directly by the polls, but by the parties and their candidates by coordinating or sending signals to their electorates to consider this strategic behavior. Without such signals, surveys alone don’t do the trick.
Poll legislation, in Mexico and other countries, is based on the assumption that they can influence the electorate and that is why their publication is prohibited a few days before an election.
Influence is often taken for granted but rarely documented. And the famous effects bandwagon and underdog They seem insufficient today in a political world in which polls have grown in centrality and visibility.
That is why fresh analyzes like the one published by the Cambridge Elements series in 2021 are welcome, The Power of Pollseither The power of polls: a comparative experimental analysis of the effects of electoral pollsby a group of two Canadian and one American researchers, James Roy, Shane Singh and Patrick Fournier, whose analysis offers data from several countries, in addition to their own.
Roy and his colleagues wonder if surveys influence voters to become politically motivated and seek more information during electoral campaigns, thus contributing to a more informed and even more participatory citizenry.
The latter is very useful, since one of the long-standing concerns is that unfavorable polls for a party or candidate can demobilize their followers, causing an effect of abstentionism among them, another aspect that is not entirely proven.
But I think the first aspect is more innovative: surveys provide the electorate with very valuable and important information, and that influences voters to seek less other types of information about the candidates and their proposals. It is a somewhat unexpected effect.
It is also one of the most ignored possible effects until now: the influence on the motivation to obtain information. For some voters, both partisan and nonpartisan, polls are such a primary source of information that following them reduces the search for other data.
The informative role of polls is of vital importance in democratic elections. The notion that among its possible effects is the decision not to be informed helps us understand why parties and their candidates are turning campaigns into survey-centric debates. Focusing on who is beating who takes attention away from who is proposing what and why.
Remembrances
Another common question these days is whether the advantage that the polls show today in favor of the ruling party is irreversible or not. Here is a piece of information from our EF surveys from six years ago: AMLO led by 8 points in a hypothetical confrontation in September-October, with still no definition of candidates in the PAN and the PRI. He won by 30 points. The popular will moves.
