A couple of weeks ago, I said the one thing I have learned throughout the 2016 primary season is to not overreact to the results of a single poll or a particular primary contest. The more I think about it, in addition to not overreacting, the main thing I would say I learned this year is that conventional wisdom is anything but.
Donald Trump had no chance at the Republican nomination; until he won it. Hillary Clinton was going to win the Democratic nomination easily; until Bernie Sanders gave her a much tougher fight than many expected.
Conventional wisdom has been swatted away in terms of statistics as well. Remember when Donald Trump had a ceiling of around 35%? Turns out that wasn’t true.
I am sure people still believe Bernie Sanders wins more often than he loses when Independents are allowed to vote in primaries. This has also been disproven.
This does not only pertain to the presidential level. It happens on the state and local level as well.
In the Maryland Senate race, many were of the impression that high African American turnout would help Donna Edwards defeat Chris Van Hollen in their primary.
Well, African Americans made up 46% of the electorate, a plurality, and Van Hollen still won by 14%; aided in large part by garnering 37% of the African American vote.
And in the wake of Jamie Raskin’s victory in the primary to replace Van Hollen in the 8th Congressional District, many are of the belief that David Trone’s entry into the race split the vote between he and Kathleen Matthews; giving Raskin the victory.
This is what I want to focus on in this post. Remember, this is the year where conventional wisdom has been blown up so many times. The CD8 example is no different.
What I did was reallocate David Trone’s votes, giving some to Raskin and Matthews. Obviously there would be a few more variables at play had Trone not gotten in the race; with other candidates receiving some share of his votes. But I thought simply dividing them up proportionally between the two top tier candidates would still tell the story.
I ran through six different scenarios to find out how many of Trone’s votes would have had to go to Matthews in order for her to overtake Raskin and win the primary.
Turns out, it is more than people think.
Here are the results of the actual primary.
The four charts below shows what happens to the vote totals when I give Matthews between 60% and 80% of Trone’s votes in Carroll and Frederick Counties; which is not inconceivable since these were Raskin’s weakest areas. They also show Matthews getting between 20% and 40% of Trone’s votes in Montgomery County.
The yellow shaded sections show what percent of Trone’s votes each candidate would get, and also the additional number of raw votes they would receive. The unshaded sections show how the additional votes would affect a hypothetical race.
As you can see, it takes Matthews receiving 80% of Trone’s votes in Carroll and Frederick, and 40% of his votes in Montgomery before she even starts to gain votes on Raskin. In the other scenarios, Raskin actually increased his margin over Matthews.
Then I played out a hypothetical where Matthews received 90% of Trone’s votes in Carroll and Frederick. In the primary, Raskin only received around 12% of the vote in each county, so that number is not inconceivable either. I also gave Matthews half of Trone’s votes in Montgomery. Take a look:
Even in this situation, Raskin still wins by over 3,500 votes.
My quest came to an end when I finally plugged in the numbers that put Matthews over the top. She still gets 90% of Trone’s votes in Carroll and Frederick, but this time she gets 60% of Trone’s votes in Montgomery.
First things first; Matthews getting 60% of Trone’s votes in Montgomery County is highly unlikely. Raskin has pretty strong roots here so it is hard to see that many of the county’s voters not having him as their second choice.
Second thing, even in this highly unlikely scenario, Matthews only wins by a little under 1,200 votes.
The bottom line is Raskin won this race by running up the score in Montgomery County. Considering how strong he was in the county, it is nearly impossible to see Matthews, or Trone if for some reason he were in the race instead of Matthews, overtaking him in a two way race.
Again, it is not quite as simple as what I laid out. But at the same time, it is also not as simple as saying Matthews lost because of David Trone. He certainly didn’t help her, but he didn’t cost her the election outright either. And yet, this line of thinking has prevailed among many. Including me until I looked into it.
Conventional wisdom. It is always right until it’s not.