The unexpected victory of Donald Trump over heavily favored Hillary Clinton, who US polls had predicted would win the presidential election, has thrown the future of the country’s polling industry into question.
Although most polls correctly predicted Clinton's narrow popular-vote victory, virtually all surveys failed to forecast Trump's rise to the presidency through securing the most votes from the Electoral College.
The prevailing question now is which polls will survive and what adjustments will be required to avoid another blunder. Many public polls have already disappeared as news organizations ended them amid deteriorating finances.
Of the 20 major polling institutions, only the USC/LA Times presidential election poll, consistently gave Trump the edge.
On Election Day, which was on Tuesday November 8, the RealClearPolitics polling data aggregator showed Clinton ahead by 3.3 percentage points nationally. But just hours later, the polling community was offering a collective apology.
“The industry is definitely going to be spending a lot of time doing some soul-searching about what happened and where do we go from here,” says Chris Jackson, head of US public polling at Ipsos, the polling partner of Reuters.
FiveThirtyEight.com, a website that focuses on opinion poll analysis and was created by respected election forecaster Nate Silver, had forecast that Clinton had a 71 percent chance of beating Trump.
Many pollsters weight their samples based on the electorate as it was composed in prior election contests, but that was their mistake because polls simply underestimated the number of quiet, poll-avoiding Trump supporters, according to Larry Sabato, a University of Virginia political professor and director of the school's Center for Politics.
White voters living in rural America had a higher than expected turnout, which helped Trump, Sabato said, while African-American and younger voters who mostly supported Clinton had a lower than expected turnout.
The problem came down to the models the pollsters used to predict who would vote - the so-called likely voters, said Cliff Young, president of Ipsos Public Affairs US.
“The models almost universally miscalculated how turnout was distributed among different demographic groups,” Young said. And turnout was lower than expected, a result that generally favors Republican candidates.
Political analysts also say pollsters may not have understood the depth of resentment towards the former first lady, US senator and secretary of state, whom many saw as a corrupt member of the elite Washington establishment.
Trump stunned the world on Wednesday by defeating heavily favored Clinton in Tuesday's presidential election, sending the United States on a new, uncertain path.