Depending on who you asked ahead of Election Night, the 2018 midterms were a referendum on President Donald Trump and the Republican-controlled Congress. Or they were a test for progressive Democrats. Or the elections were all a single issue: health care, immigration, the economy — the list went on and on. Now that the exit polls are in, the picture is becoming more clear on why Democrats were able to win control of the House and how Republicans increased the majority in the Senate.
Exit polls aren’t perfect. But they do provide information on what mattered most to voters, and how they viewed key leaders and issues, based on their income, age, race, and other groupings.
Exit polls can help people better understand how they fit into the bigger political picture and their moment in history, by gauging how much sway political campaigns and policy-making hold over a voter’s decision when they cast their ballot, said Courtney Kennedy, director of survey research at Pew Research Center.
“You just voted. How did you react to everything that has happened in the past few months?” Kennedy said.
Here’s a look at how exit polls work, the issues that appeared to matter most to voters in 2018, and how to interpret these findings.
Was 2018 a referendum on President Trump?
More than half of all exit poll respondents — 54 percent — said they disapproved of President Donald Trump, according to National Election Pool exit polling from House races. That sentiment was hyper-partisan with nine out of 10 respondents who identified as Democrats saying they disapproved of the president, while roughly as many Republicans said they approved of the president.
Americans cared about health care
More than 300,000 Americans stand on the brink of receiving health care coverage under Medicaid expansion after this week’s midterm election. Idaho, Nebraska and Utah all voted to expand Medicaid.
The issue of health care drove a plurality of voters — 41 percent — to the ballot box, according to exit poll results. By and large, those voters who said health care was their most important midterm issue identified as Democrats. Roughly a quarter of Republicans — 23 percent — agreed.
A voter concern that had been the top motivator for roughly a decade — the economy and jobs — was unseated by health care in the exit polls, Kennedy said.
Trump’s inflammatory anti-immigrant rhetoric — deemed racist by many — has recently included misinformation about a migrant caravan traveling by foot through Central America and Mexico for the U.S. border, along with his questionable threats to upend birthright citizenship. Did the president’s pre-election push on immigration issues impact voting?
When asked at exit polls, 23 percent of voters said immigration was the most important issue to them. Of those respondents, three-quarters said they were Republicans, and Kennedy said it reflects the effectiveness of campaign messages about immigration leading up to this week.
The president frequently touts his success in boosting the economy and lowering joblessness during his campaign rallies, press pool interviews and on Twitter. In fact, the final federal economic report before the election offered undeniably good metrics, including a significant wage boost for workers. At the same time, 52 percent of wealthier voters — with household incomes above $50,000 — said they disapprove of Trump in a recent PBS NewsHour/NPR/Marist poll.
For voters, roughly a quarter this week said the economy brought them to the polls. Along party lines, nearly two-thirds of those respondents were Republicans, and another third were Democrats.
What are exit polls?
Exit polls are different than the thousands of polls conducted leading up to a major election, which help pundits read the tea leaves on what choices constituents will make on Election Day.
To conduct exit polls, an interviewer randomly and systematically asks voters after they cast their ballots. They ask voters how they voted, what attitudes and issues were most important in compelling them to vote and if they affiliate themselves with a given party, along with their age, gender, race and highest level of education, according to the American Association for Public Opinion Research.
Major news outlets began this practice in the 1970s, with CBS News, ABC News, CNN and NBC News eventually merging into National Election Pool. This week, the Associated Press, Fox News and NORC at the University of Chicago launched an exit polling service for the 2018 midterms.
How do we interpret exit polls?
Exit polls are a tool for understanding how and why people vote — but should not be viewed as complete knowledge. This is one lesson many took from the 2016 presidential election when exit polls wrongly projected that former Secretary of State Hillary Clinton would beat President Donald Trump.
“While the general public reaction was that ‘the polls failed,’ we found the reality to be more complex,” according to a report written and issued by polling experts to investigate how projections suggested that Clinton had a 90-percent chance of winning the presidency.
Polling experts have since worked to improve their methods and how they collect responses from voters for exit polls, Kennedy said. While exit polls are typically conducted in person on the big day, she said interviewers increasingly go to polling stations to capture early voter opinions and call up respondents to follow up over the phone.
This year’s election supports that modified approach. More than 30 million Americans voted early, before Nov. 6, in what experts have called an historic voter turnout. That is consistent with early voting trends identified since 2000, AAPOR said. Those numbers can affect exit poll accuracy if pollsters did not alter their methods.
Exit poll results can be skewed, depending on who chooses to answer or pass on pollster questions. Selective refusal happens if a cluster of respondents don’t want to participate in the survey. For example, if Republican voters are more likely to say no than their Democrat peers to the questionnaire, AAPOR explains this could produce a “slight Democratic overstatement.”
Premature leaks of exit poll data, as well as election regulations dictating how close exit pollsters can stand to a polling station also can influence poll results.
People also need to pay attention to an exit poll’s margin of error. Since this poll is not a census of all voters and their opinions, the margin of error attempts to indicate how much those opinions represented in the poll could vary from reality.
So far, these bias concerns haven’t emerged with the 2018 exit polls in substantial ways, Kennedy said.