A politician wants to vote for the decriminalisation of marijuana. But she knows that opinion polls suggest that her electorate is against easing access to weed. Although she cares about the issue, she also cares about being re-elected. What should she do?

Should she:

  1. Heed the pollsters and vote against decriminalisation;
  2. Appeal to her electorate, citing evidence and appealing to their values; or
  3. Just vote for it, since her electorate will support her anyway?

I think most political advisers would suggest the first or the second option. But a fascinating new experiment by David E. Broockman and Daniel M. Butler suggests politicians have broad latitude to shape public opinion without any electoral cost.

Broockman and Butler, political scientists at University of California, Berkeley and Washington University in St. Louis respectively, write that the traditional model of democracy views citizens as “issues voters”. According to this view, politicians lose support when they defy public opinion and they gain support when they embrace popular ideas. Last year, when Ed Miliband, the Labour leader, pledged to freeze UK consumers’ energy bills, his advisers referred to the policy’s popularity in focus groups; they assumed that, in part, UK citizens are “issues voters”.

Opponents of the “issues voters” view say that politicians are able to shape public opinion, typically through appeals to the values of their electorate. If politicians “frame” issues in the right way, then they can win round sceptical voters. George Lakoff is perhaps the best known exponent of this idea. In Don’t Think of an Elephant!, and other books, he argues that politics is a game of metaphors. US Republicans, he says, are clever at finding persuasive phrases (e.g. “tax relief“). Last week at the Conservative party conference, George Osborne, the chancellor, tried to establish a frame for the next UK election – the idea that only by securing the economic recovery can the country afford to spend on the NHS and other nice things.

But what if citizens aren’t voting for politicians based on issues and what if they aren’t forming their opinions based on frames or appeals to their values? What if, as Broockman and Butler argue, “citizens tend to agree with their favored political leaders because citizens simply adopt politicians’ positions as their own”?

The experiment

Advocates of “position adoption” – the idea that voters adopt the views of politicians they trust without needing to be persuaded of the merits of those views – have been around for decades. A famous study of a 1976 US presidential debate between Jimmy Carter and Gerald Ford showed that voters changed their views on unemployment insurance based on those of their favoured candidate (as opposed to changing their views on their candidate because the politician didn’t share their views on policy). The flow of causality seemed to go from opinion of politician to opinion of policy.

However, we don’t know why this seems to be the case in some situations. Is it because the politicians do in fact appeal to the reason or the passions of voters? Is it because voters trust their candidates? Is it because they don’t care that much?

Thanks to the study by Broockman and Butler, we have a better idea. The two researchers did something rare in political science – they created a natural experiment with the blessing of active politicians, in this case seven legislators from an unnamed midwestern state in the US. The study:

  1. Asked the seven politicians to list all the issues they are working on;
  2. Polled citizens in each district to establish their views on each issue;
  3. Identified the four issues with the least public support in each district;
  4. Sent voters one of three possible letters from their legislator – a “placebo” letter with no reference to an issue; a second letter in which the politician pithily wrote that he would be supporting an unpopular policy; and a third letter in which the politician also explained why he was taking that position;
  5. Polled citizens again on their views on each issue and on their legislator.

The study has a full description of the methodology and examples of the letters that were sent. Here are the sorts of issues that the politicians took positions on:

The researchers found that “the constituents who were sent letters with the legislators’ positions were about 5 percentage points less likely to disagree with the legislator and about 5 percentage points more likely to agree with the legislator.”

But the content of the letter did not seem to matter to the recipients. Broockman and Butler found that there was no difference in citizens’ likelihood of supporting a position based on whether the constituents received the letter explaining the politicians’ decision, or whether they received the pithy letter. What is more, this latter group was no more likely to have negative views of their politician. The authors conclude that there is “strong evidence that legislators can significantly shape constituents’ views on issues by merely staking out their positions.”

The potential implications

This is one study, in one part of a big country, using one rank of politician from one party (Democrats). But other results also support the idea that voters will change the positions based on those adopted by their political leaders. What is going on?

One possible explanation is that only trusted politicians can easily shift public opinion. Writing about David Cameron’s speech at the Conservative party conference, in which the prime minister pledged to cut taxes without explaining how they would be funded, Philip Collins argued in The Times that: “the identity of the person doing the promising matters as much as the content of the promissory note.” Mr Cameron has higher personal poll ratings than Mr Miliband, and the Conservatives have higher poll ratings on economic competence. Together, these would have suggested to the prime minister that he had room to make his pledge.

Nevertheless, the Broockman and Butler study features local politicians, all of whom lack the profile of a senator, never mind a president or a prime minister. The voters involved in the study probably didn’t know much about their legislators – and yet they were still influenced by his policy changes, regardless of his reasons.

As the researchers write, “no one believes any of these perspectives [issue voters, elite persuasion, or position adoption] explains politics all the time.” However, studies such as this one imply that there are broader implications. They suggest that the constraints on public opinion are weaker than commonly assumed by politicians.

Get alerts on Opinion when a new story is published

Copyright The Financial Times Limited 2019. All rights reserved.
Reuse this content (opens in new window)

Follow the topics in this article