The folks over at Public Policy Polling (PPP) have released a poll documenting the presidential preferences of the citizens of Virginia. This survey (1,032 likely voters, margin of error 3.1 percentage points, results here) shows that in a hypothetical contest between the presumptive presidential nominee of the Republican Party, Donald Trump, the presumptive presidential nominee of the Democratic Party, Hillary Clinton, the presidential nominee of the Libertarian Party, Gary Johnson, and the presumptive presidential nominee of the Green Party, Jill Stein, Clinton leads Trump by 3 percentage points, 42%-39%, with Johnson polling at 6% and Stein at 2% (10% respond “Undecided”).
PPP asked Virginia respondents what their presidential choice would be if only Trump and Clinton were on the ballot. In that imaginary two-person contest, Clinton still leads by 3 percentage points by a margin of 48%-45% (7% respond “Not Sure”).
Since one of the mantras of this blog is that it is trends that matter and not a single poll, can we see if there are any trends that this poll and the past surveys tell us?
The first trend that pops up is the fact that Clinton’s lead in the Old Dominion has been shrinking since the first poll was taken at the start of 2016. The initial 2016 poll taken in Virginia was done by Roanoke University (as mentioned earlier in this post). This January survey showed Clinton leading Trump by 17 percentage points (52%-35%).
In April, Christopher Newport University released a poll (as mentioned earlier in this post) which showed Clinton’s lead was now 9 percentage points (44%-35%).
In the next month*, Gravis Marketing released a survey (as mentioned earlier in this post) that showed Clinton’s lead now down to 4 percentage points (45%-41%).
Now, in June, this latest poll from PPP shows Trump trailing Clinton by 3 percentage points. Are these surveys a case of Clinton losing or Trump gaining? The answer is the latter. In January, Clinton’s support of the Virginia electorate was 52%, but is now 48% (a drop of 4 percentage points) in June. Trump’s support has increased 10 percentage points from 35% in January to 45% in June.
In comparing simply the January Roanoke poll (PDF topline results here) to the June PPP survey, are there any other trends that explain Trump’s rise in Virginia? Well, there’s the favorability factor.
In January, Trump had a favorability rating of -41% (22% respond “favorable” versus 63% “unfavorable). Half a year later, Trump’s favorability “jumps” to -28% (32% “favorable” versus 60% “unfavorable”). In six months, more residents of the Old Dominion have a favorable opinion of the Republican Party’s presidential nominee. For comparison, Clinton’s favorability rating in January was -15% (36% “favorable” versus 51% “unfavorable). Of the respondents to the June PPP poll, 39% viewed her as favorable and 53% view her as unfavorable giving her a favorability rating of -14%. Clinton’s favorable number increased only 3 percentage points, but her unfavorable numbers increased also.
There are no other demographic breakdowns from the January Roanoke College poll, so I can do no other comparisons to this June’s PPP poll.
Moving on to the April survey from Christopher Newport University (CNU), the factor of party identification plays a major part in Trump’s climb in Virginia’s polls. In the CNU survey, only 68% of the respondents who identified themselves as members of the Republican Party said they would vote for Donald Trump. Of the respondents who said they had voted in the Virginia Republican primary, a full 28% said they would not vote for Trump (13% would vote for Clinton, 13% would vote for a third-party candidate, 3% would not vote). For comparison, 87% of respondents who identified themselves as members of the Democratic Party said they would vote for Hillary Clinton. Of the respondents who said they had voted in the Virginia Democratic primary, only 9% said they would not vote for Clinton (4% would vote for Clinton, 3% would vote for a third-party candidate, 2% would not vote). The director of the organization that produced this survey called this a “loyalty gap” between the two candidates.
Jump forward to June and the loyalty gap is long gone. Of the respondents who identify as members of the GOP, now 84% say they would vote for Donald Trump (in the two-person question), an increase of 16 percentage points. Clinton’s support among fellow Democrats in June remains steady at 88% (+1 percentage point).
In addition to grabbing his fellow Republicans, Trump has increased his support among Independent voters. In April, Trump won 36% of the Independent vote (versus 26% for Clinton, 27% for “Neither”, and 11% for “Undecided”). Two months later, Trump now commands (in the two-person race) 47% of the Independent vote while Clinton takes 39% (and 14% respond “Not sure”). It is true that Clinton’s share of the support increased by 13 percentage points, which is greater than Trump’s increase of 11 percentage points among Independents, but the numbers favor the Republican Party nominee when the contest is expanded to four people. In the Clinton/Johnson/Stein/Trump ballot, Trump wins 42% of the Independent vote (+6 percentage points from April; -5 from the June two-person race) while Clinton receives 29% (+3 from April; -10 from the two-person race). Johnson takes 10% of the Independent vote and Stein wins 6%. Since Clinton’s drop among Independents from the two-person race to the four-person race (10 percentage points – 39% versus 29%) is greater than Trump’s decrease (5 percentage points – 47% versus 42%), it appears that the addition of third-party presidential candidates in Virginia draws support away from Clinton among Independents.
The caveat to the above statement is that in the four-person contest, 76% of Republicans would support Trump while 83% of Democrats would stick with Clinton. That’s a drop of 8 percentage points (76% versus 84%) for the Republican Party presidential nominee, but only a 5 percentage point drop (83% versus 88%) for the Democratic Party presidential nominee. Part of the drop in Trump’s numbers is that the number of undecided GOP respondents jumps 6 percentage points from 7% in the two-person race to 13% in the four-person contest.
The third (and final) factor I will discuss that helps explain Trump’s rise is his shrinking of the gender gap. In April, Clinton led over Trump with female voters by 23 percentage points (52% Clinton, 29% Trump). In June, Clinton still led among that demographic, but now only by 11 percentage points (51% Clinton 40% Trump). In the four-person race, Clinton’s lead over Trump is still 11 percentage points (45% Clinton, 34% Trump, 4% Johnson, 3% Stein). In two months, Trump’s support among female voters, in a two-way contest, has increased 11 percentage points.
Another caveat should go here and it relates to the male vote. To a much smaller extent, Clinton has cut into Trump’s lead among men. In April, Trump led among male voters by 6 percentage points (43% Trump, 37% Clinton). In the two-person contest in June, Trump’s lead among male voters is 4 percentage points (49% Trump, 45% Clinton)**.
Trump is gaining on Clinton in the Old Dominion and part of the gain can be ascribed to Republicans coming around to support their party’s nominee, voters starting to see him in a favorable light (relatively speaking), and his narrowing of the gender gap.
All in all, the 13 votes in the Electoral College that Virginia holds looks more up for grabs than it did in January.
*In the chronology of Virginia polls, there was a May survey conducted by Roanoke College (as mentioned earlier in this post) that showed a 38%-38% tie between Trump and Clinton. I skip that survey in this post because I consider it an outlier. It has the fewest number of respondents (610 likely voters) and the highest margin of error (4 percentage points) of all the surveys taken in Virginia this year.
**In the four-person contest, Trump’s lead among men is 7 percentage points over Clinton (46% Trump, 39% Clinton, 7% Johnson, 2% Stein), which is a slight increase (+1 percentage point) over the April two-person race numbers.