PRESIDENT Donald Trump’s use of Twitter has sparked much controversy among political commentators. Many regard the habit as reckless. Others believe that worrying about it is largely a distraction. But what is the effect of the president’s tweets on domestic politics? There are limits to the power of politicians’ statements to shift behaviour. However, as partisanship grows, committed members of political parties seem increasingly inclined to change their attitudes to match those of their parties’ leaders. 

Research suggests that partisanship trumps other factors when people form political opinions. James Druckman, a political scientist from Northwestern University, looked with his colleagues at the comparative effect of substantive information and party endorsement on what to do with undocumented Americans who had arrived in the country as children. Substantive information changed both Democrats’ and Republicans’ opinions on the issue—some to more and some to less sympathetic positions—as long as they did not know their party leadership’s opinion. Once told this, partisans on both sides adopted the party position and became unmoved by information presented to them.

One area in which this mechanism seems to have operated in America over the past couple of years is public opinion about Russia. For decades, Republican and Democratic attitudes to Russia tracked each other closely. In 1990, according to a survey by the Pew Research Centre, a think-tank, 33% of Democrats named Russia as “the greatest danger” to America. The Republican figure was 34%. By 2012, those numbers were down to the low single digits for partisans of either party. But by 2017 only 21% of Republicans saw Russia as the greatest threat, compared with 39% of Democrats. In 2015, just 12% of Republicans and 15% of Democrats had a favourable view of Vladimir Putin, Russia’s leader. By contrast, at the start of 2017 32% of Republicans viewed Mr Putin favourably, and only 10% of Democrats agreed. This suggests that Mr Trump’s vocal admiration for his Russian counterpart has filtered down the ranks.

The impact of Mr Trump’s statements is not limited to views on Russia. In 2009 more Republicans than Democrats were generally in favour of free-trade agreements. Over the past two years, the president has repeatedly voiced disdain for such agreements, though. His base appears to have been impressed: by 2017, Republican support for them had fallen to 36% compared with 67% for Democrats. A similar process seems to have taken place regarding trust in the media. Surveys suggest that the gap between Republicans and Democrats who trust national news organisations climbed from 12 percentage points to 23 between 2016 and 2017, with Democrats displaying much higher levels of trust.

Even opinions on traditionally apolitical topics are taking on a partisan character. Since Mr Trump was elected, Democrats have become less confident about the economy, as Republicans’ confidence has increased. This may not be in itself surprising—what is interesting is the extent of their differences. The partisan gap in an index of consumer expectations reported by the University of Michigan last autumn suggests Democrats are braced for a recession while Republicans may believe Mr Trump’s talk of 6% annual GDP growth  

Of course, opinions reported to surveyors may reflect the need to demonstrate partisan loyalty as much as they do underlying beliefs. Research for the National Bureau of Economic Research by Atif Mian Amir Suf and Nasim Khoshkhou suggests that partisan-driven change in reported sentiment is not reflected in survey answers to questions about whether now is a good time to buy a car or household items, for example. The paper also finds no evidence of a larger rise in actual car purchases after the election in counties that voted more heavily for Mr Trump.

Nonetheless, the effect of heightened partisanship on public opinion is worrying. As Republicans model their opinions on a leader who is not afraid to spread falsehoods, the result may be an increasingly mis-informed electorate. At the end of 2016, 62% of Trump supporters believed that millions of illegal votes were cast in the presidential election, for example, though there is no evidence to support this. Knowing the power of the bully pulpit, political leaders should aim to use it responsibly.