2

I am curious about the correlation between Senate and President. I have found that Susan Collins winning by a relatively large margin appears to be an outlier.

The senate races seem closer to the president in 2016, which is my hypothesis. Based on preliminary data, excluding Georgia's regular, what is the correlation between two party Dem vote share for President and Senate in each state? And was my hypothesis correct? (The correlation was about 91%.)

I also feel the US is a quasi parliamentary democracy right now in the sense that people vote for and based on parties and not so much based on the individual candidates. That is a topic for a different question but I put this in for context.

Improve this question
4
  • 1
    Exclude Arkansas too. – Number File Nov 6 '20 at 18:06
  • And Louisiana it has multiple candidates. – Number File Nov 6 '20 at 18:19
  • 1
    It also probably is a factor that Susan Collins is one of the most liberal leaning of the Republicans. – JohnFx Nov 6 '20 at 20:04
  • 1
    It is not so much that she is more liberal but that New England is weird in the fact that they still have the concept of liberal Republicans which don't exist anymore in my opinion. I would not vote for the governor of Vermont for another example because I am a straight ticket voter. – Number File Nov 6 '20 at 20:29
7

Looking at the results so far, as one might expect, there is a clear correlation between the vote in Senate races and the Presidential election. In the figure below, I've excluded Georgia's special election and Arkansas' election in which the Democrats did not field a candidate. In Louisiana's blanket primary, I've summed the votes of candidates representing the Democratic & Republican parties.

The correlation coefficient of the two variables is ~0.95, indicating a strong positive correlation, and the trend line shown on the graph has a gradient of ~0.88.

Looking at outliers, the only result with a studentized residual greater than three is the Nebraska result, where Democrat candidate Chris Janicek won ~27.8% of the two-party vote for the Senate, while Joe Biden won ~40.1% of the two-party presidential vote. The studentized residual of Susan Collins' result in Maine was around 2.78, so although close to being categorized as an outlier, it does not fit the definition in this case.

On the graph below, the red line indicates the trend line, while the blue line indicates the straight-ticket vote line.

enter image description here

The figures above are based on preliminary data recorded at 10:20 GMT 9th Nov. as reported by Dave Leip's Election Atlas.

Improve this answer
0

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .