New CNN poll 8/15
CNN Poll: Democratic advantage is growing
01:35 - Source: CNN
CNN  — 

While more people say they are planning to vote for a Democratic candidate over a Republican candidate in a generic vote question for Congress in November, almost half (48%) still say that Republicans will be in control of the House and Senate after the election, according to a new CNN poll conducted by SSRS.

That number is fairly divided by party identification lines, with Democratic subgroups (such as those who disapprove of Trump, nonwhite voters, and women) more likely to believe Democrats will take Congress, while Republican subgroups have confidence in their party.

However, the one set of demographic groups that don’t follow their usual partisan lines is age. Usually, those under the age of 45 tend to be more liberal, while those over 45 are more conservative. For example, in this poll, President Donald Trump’s approval is at 36% among those under 45 and 46% among people over 45.

In opposition to usual trends, most of those under 45 (53%) say that Republicans will maintain control of Congress. This is in contrast to the 45+ age group, where people are split in their beliefs. Forty-four percent say that Republicans will control Congress and 42% say the Democrats will.

Essentially, some younger Americans support the Democratic candidates, but don’t believe they’ll win, according to this poll.

There’s no one explanation for the discrepancy among age groups versus partisan lines, but it’s curious since elsewhere we’ve seen partisan lines growing deeper as almost every issue is divided by Democrats versus Republicans and their respective demographic subgroups.

One theory is that younger people have seen fewer elections and were much more hopeful or disenfranchised by the surprise outcome of the 2016 election. Those over 45 have seen this before and are less trusting that the more likely outcome will be the one to occur.

The CNN poll was conducted by SSRS August 9-12 among a random national sample of 1,002 adults and 921 registered voters reached on landlines or cellphones by a live interviewer. Results for the full sample have a margin of sampling error of plus or minus 3.9 percentage points, it is larger for subgroups.