Today, the New York Times featured some of our polling data and proprietary algorithm results and their implications for the Republican primary race ““ a mostly under-the-hood look at Donald Trump’s coalition.
We wanted to give a bit of a different look at the same data at a local level, showing how support for the Republican candidates has changed over the past five months ““ maybe even in the district in which you live.
This map shows a week by week look at which Republican primary candidate has a plurality — more support than any other candidate — for that week, in every Congressional district across the country, over the past five months.
Here are a few things it’s telling us:
- Carson had a plurality of support in a vast majority of Congressional districts (the geographic level of the map) by the start of November, but his national support only slightly exceeded Trump’s (by about 1.7% at his peak).
- Trump’s support has been remarkably consistent despite marked changes in his competitors’ support. (You can read more about what the data says about Trump’s coalition in the New York Times.)
- In November, we start to see Ted Cruz trending upward, gaining traction first in his home state of Texas and eventually expanding toward having a small plurality out west in areas that at one point was Carson’s territory.
- And, of course, we see that candidates tend do better in their home states. Look at Arkansas, where Huckabee — who hasn’t registered above the single digits at any time during the campaign — has a consistent plurality of support. Even those candidates that don’t earn a plurality on the map tend to do better in their home states.
January is a long month in primary season, so we’re just as curious to see how things shake out in the coming weeks ahead of the Iowa caucus and New Hampshire primary.
How Did We Calculate These Estimates?
Right now, most of the polls you’re probably reading about in the news are reporting a national or single-state topline of the race ““ usually with a margin of error of ~5% or higher.
What we do is a bit different than what you’ve been reading:
1. We collect a lot more data.
We’ve gathered an enormous amount of survey data on the GOP primary so far. Back in August, we released data from an initial poll of 757 self-identified Republicans from a total sample of 3,007 Americans (you can read more about the findings in the New York Times). Since that initial survey starting August 10, 2015, we’ve collected over 10,000 more survey responses to the GOP Primary horserace question on our ongoing weekly national tracking survey of 2,000+ respondents (if you’d like to add your own question, learn more here).
2. We’re using math that isn’t typically used in election analytics.
To build the maps you’re looking at, we’re running tens of thousands of simulations using proprietary Bayesian algorithms that leverage all of that data to make estimates of survey responses in small geographies or demographic subgroups (if you’re interested in learning more, check out multilevel regression and post-stratification).
Using these methods we’re able to confidently generate estimates within 8.7 percentage points at the Congressional level which is 5.2 times better than what we could do with surveys alone.
Methodological note: The above figures employ discussed methods and show our estimated support for candidates over the 20 weeks between August 10, 2015 and December 27, 2015. We ran models for each of the top ten candidates — Bush, Carson, Christie, Cruz, Fiorina, Huckabee, Kasich, Paul, Rubio and Trump — and an additional model for those candidates polling at less than one percent — Gilmore and Santorum — together with those candidates who have dropped out — Graham, Jindal, Pataki, Perry and Walker.
Civis Analytics conducted 40,050 live telephone interviews of adults in the United States contacted on telephones from August 10, 2015 to December 27, 2015. Among respondents of these surveys there were 11,441 self-identified Republican or lean Republican adults. These respondents were asked their candidate preference in the GOP primary. Undecided respondents are not considered as part of the analysis, map, or trend lines.
Respondents are sampled from nationally representative voter and consumer files.
Results were weighted based to a modeled 2016 Republican electorate.