By Dante Chinni
If Hispanic voters turn out in greater force in November than 2012, it could mean millions more votes.
As Election Day nears, pollsters shift from measuring opinions among registered voters to surveying those who are not only registered but considered likely to cast ballots.
Just who is a likely voter? In the remarkably atypical 2016 election cycle, that’s a good question.
Consider the differences between the 2012 presidential race and this year’s contest. The first African-American president is not on the top of the ballot. For the first time in U.S. history, one of the major presidential parties has nominated a woman. The Republican nominee has energized a particular segment of the electorate — working-class, white voters. And the Democratic and Republican presidential candidates have the highest negative ratings in history.
Any one of those things could have an impact on who comes out to vote in November. Together they create unusual uncertainty about the nature of electorate.
Are African-Americans less likely to vote this year? Will women be more likely to vote? Will Republican Donald Trump’s unique resume and unconventional approach draw new voters to the polls? Considering the high negatives for both candidates, will overall turnout drop? How much will racial diversification shape the voter pool?
Somewhere in that mix of questions lies the answer about the composition of the 2016 vote.
The argument for demographic change to play a big role is strong. This is the composition of the voter pool in the last four presidential races, according to the exit polls.
Composition of the Presidential Vote
| Election | White | African-American | Hispanic | Asian |
|---|---|---|---|---|
| 2000 | 81% | 10% | 7% | 2% |
| 2004 | 77% | 11% | 8% | 2% |
| 2008 | 74% | 13% | 9% | 2% |
| 2012 | 72% | 13% | 10% | 3% |
That’s a pretty clear pattern: Whites are dropping as a share of the electorate, while other groups are growing. The trends mirror the changes in the broader U.S. population. Since 2000, whites have declined from 69% of the population to about 62%. Hispanics have climbed from 13% to 18%. Asians have climbed from 4% to 6%.
But whites have fallen more as a share of the vote tally than they have in the general population.
In other words, the demographic trends in the population as a whole aren’t driving all the changes in the table above. It could be that white turnout increases this year and that African-American turnout fall in the first post-Obama election. Those developments, should they occur, could have a big impact on the vote tally.
Mr. Trump leads Hillary Clinton among white voters by 9 percentage points on the ballot that includes third-party candidates in the latest Wall Street Journal/NBC News poll. But Mrs. Clinton leads Mr. Trump by 71 points among African-Americans.
With margins like those, if African-Americans slip just one point in their composition of the 2016 electorate, it could cost Mrs. Clinton about 1 million votes.
What if Mr. Trump drives up the vote among whites without a college education? They made up about 36% of the votes cast in 2012, but as recently as 2004 they made up 43% of the total vote. If their share of the electorate returned to 2004 levels, it could boost Mr. Trump’s tally by millions of votes. The Republican leads among whites without a college education by 20 percentage points in the latest Journal/NBC News poll.
Getting near that 2004 number likely wouldn’t be easy given that racial diversity has increased. But such a change is not impossible.
What of the Hispanic vote? Mr. Trump’s support among that group is at 16% in the latest Journal/NBC News poll in the four-way race. That’s 11 points worse than Mr. Romney’s 27% share in 2012. What if that group of voters turns out in force in opposition to Mr. Trump, and its share of the vote bumps up from the 10% in 2012 to 11% or 12%? That could be worth more than a million votes for Mrs. Clinton.
The point is that small changes in a demographic group’s share of the electorate can produce meaningful changes in the vote tally. That’s why different predictions about who is likely to vote can yield very different results in opinion polls. And the unique profiles of this year’s candidates mean that turnout is hard to predict.
As long as the race stays reasonably close, differences of opinion about turnout will mean that pollsters produce varying results in the coming weeks.
Leave a Reply