Who are Independent VotersThe differences in people who identify with a party “not very strongly”, and those who identify as independent but “are closer to” a party.
Cameron LopezBlockedUnblockFollowFollowingMar 16The data we are using is polling conducted by YouGov Blue and from the progressive data organization Data For Progress, it consists of 3,215 voters and then is weighted by “age, sex, race, education, urban/rural status, partisanship, marital status, and Census region” to reflect the American voting public, the weights were attached to data from Catalist of the 2018 voter share.
They publically published the data and for that I am thankful.
The data I am going to show stem from the Party Identification questions, although I wish they were also asked what the person’s actual registered party is, because coming from Kentucky, in the Kentucky 6th district, registered Democrats only voted for Amy McGrath at around 65% according to the New York Times Upshot Polling (or to reference Harry Enten’s Twitter Bio “Party ID does not equal Party Registration”) and it would be useful to look into how campaigns can get past the data flaw and find voters who identify as their party but are registered as the opposite party.
Nevertheless, the questions asked were “Generally speaking, do you think of yourself as a …?” and a follow-up question for those who answered a party was “Would you call yourself a strong (insert party) or a not very strong (insert party)?” and the subsequent question for those listed themselves as Independent, Other, or Not Sure were “Do you think of yourself as closer to the Democratic or the Republican Party?”.
A simple bar graph leads us to this spread.
The benefit of having a larger sample size is that we can look inside answers, and having at least 200 responses for all of the categories (besides Not Sure Who I’m closer to, but we will ignore them for now) we can look at the behaviors of these groups.
The groups I want to focus on are highlighted below:First, we will look at the ideology of each of these five categories.
We will call “Closer to GOP” GOP Independents, and the same for Democrats.
We will label GOP + Not Very Strong as GOP Weak Partisans and same for Democrats again.
As we see, GOP Weak Partisans and GOP Independents are more likely to call themselves Conservative than Democrats are to call themselves Liberal.
We also see that there is a subgroup of “Closer to Dems” that is to the left of Democrats that slightly outsize their conservative counterparts.
There are more Democrats than Republicans overall, so when looking at which districts should work on turning out your base versus running to the middle and turning out the moderates for Democrats.
Even when ideologies differ, the voting patterns are largely ubiquitous for party ID.
When we look at who each of these groups voted for in the 2018 House Races, we get the first look of the overarching trends of partisanship (or lack thereof) voting behaviors.
In this graph, we compare the Party IDs vs 2018 Vote:It gets a bit cluttered with all the “lean ______”, and secondary options, so here’s the two-party vote with lean Democrat and Lean GOP and Undecideds.
Here we see that Strong Partisans, Weak Partisans, and Partisan Independents usually voted for their own party, but when we compare Weak Partisans to Partisan Independents, we see that Partisan Independents are more partisan than Weak Partisans.
When looking at the crosstab for the all-party votes in 2018, we see that Republicans only received 77% of the Not Very Strong Republicans, but Democrats received 88% of the Not Very Strong Democrats, which is a nice boost for 2018 Democratic campaigns.
When looking at policy preferences for Partisans vs Non-Partisans we also see similar trends, but we see more similarities between Weak Partisans and Partisan Independents than Strong Partisans.
Here are what each group think of Black Lives Matter by Party ID.
We see Weak Partisan Democrats and Partisan Independent Democrats view BLM in a more similar light than Strong Democrats, but Partisan Independent Republicans view BLM more like Strong Partisans, which leads us to a little detour to look at the Racial Resentment Scores (Page 10 in the Code Book) for Party ID’s.
Here we see racial resentment of Weak Partisans are more to the total median, while racial resentment of Partisan Independents is extremely close to Strong Partisans.
Maybe this helps us explain the next crosstab of Donald Trump support by Party ID, as we see, Weak Partisans are much more willing to break rank than Partisan Independents.
Republicans also have a lot more who are only willing to say “Somewhat” than “Very Favorable” which may lead to turnout problems in the future.
Although, Republicans might turn out more on negative partisanship than anything else.
We also see Racial Resentment scores of the 60 Democrats (Strong, Weak, Partisan Independent) who said they favored Trump had a racial resentment score average of .
3941, similar levels to Republicans.
Of the 114 Republicans (Strong, Weak, Partisan Independent) who said they did not favor Trump, had a racial resentment score of .
3088, the 51 who had very unfavorable scores had an average racial resentment score of .
2333, similar to Weak Democrats.
These are too small of sample sizes, but they would be something you could look for in a larger dataset.
The next piece of data views the 2016 Election by Party ID, since our polls are based on 2018 turnout, it’s important to remember that this sample is about 6% higher for in Democrats than 2016, so Donald Trump’s numbers will look a bit better than they really were.
The most interesting aspect to me is Trump’s ability to grab almost 10% of Weak Dems.
Partisan Independents, while not voting for Trump, also didn’t vote for Hillary.
Then, of course, there are the “true independents”, who we will look at next, where Hillary was closer to Gary Johnson than Trump.
Let’s unravel the “True Independents” and see how true to their word they are.
Their voting behavior was a bit more erratic but mostly predictable.
They are plurality Republican, and those that voted Republican in 2018 mostly voted for Trump, with an almost equal amount of the remaining votes going towards Libertarian Gary Johnson, Hillary Clinton, or Not Voting.
“True Independents” who voted for a Democratic House candidate voted for Hillary about half the time, with the other half being distributed to Gary Johnson (12%), Trump (8.
5%), Jill Stein (8.
12%), Not Voting (7.
51%), Other (6.
51%), and Evan McMullin (4.
59%).
The Independent who is switching party vote is a small portion of Americans who vote.
Total of 6.
87% of total Americans switched from Obama voter->Trump, or Romney->AntiTrumpOnly 2.
13% changed from Trump Voter->Don’t Re-Elect Trump or Hillary Voter-> Re-Elect TrumpThis past graph is pretty damning that you probably shouldn’t be appealing to Trump Voters but to your base in 2020.
The people largely don’t change their opinion on the same person.
85% vote for the same party for Senate and House (2nd Dem/Dem + Dem/Dem + GOP/GOP)Here are some more tables:I don’t really like the colorful texts eitherThis is betterGreen Jobs particularly poll well with Non-Strong Partisans, 76% — 16% (Grand Total Line)Labor Unions Poll poorly, 44–48, but with total voters, it is 49–41A little under 75% of Non-Strong Partisan voters that voted in 2018 voted for the same party in 2016, even with non-strong partisan ties.
The 2 types of non-strong partisans that didn’t vote their expected party both times come to about 11% of the 2018 electorate.
Even less are those that voted for the opposing party.
Around 3% of this part of the electorate voted for the opposite party.
It’s more useful to appeal to the 8% who voted differently, than the 3% of the other party.
.. More details