As one of the states that lingered on counting votes for weeks after election day, Arizona had an interesting little side story as many were watching it more closely. Some noticed an oddity with Mark Kelly (D) not only appearing to have won the Senate race in Arizona and ousting incumbent Martha McSally (R) but also garnering more votes than Joe Biden.
Related: A Little Red State with a Big Blue Crown – Chapter 1
We’d like to thank John P. Christie for this Op-Ed. If you have an Op-Ed or LTE
you would like us to consider please submit it to Skip@GraniteGrok or Steve@GraniteGrok.com.
This is unusual because typically, by party, there is a top of the ticket down-ballot drop off that occurs where the presidential candidate gets more votes than the senate candidate for their respective party. The presidential and senate race in each state are comparable in the sense that they are the only statewide federal races that are ever on the same ballot. Congressional seats are not statewide, and governors’ races are not federal. This result from Arizona was a seemingly newsworthy nationally because Kelly managed to attain 102% of Biden’s votes.
Back here in NH Jeanne Shaheen got 106% of Biden’s votes.
I had not seen or heard any coverage of this truly historic accomplishment in the media. Maybe you did but more likely you had not, until now. So, that large level of overperformance of a senate candidate over their party’s candidate for president was apparently quickly put aside, if acknowledged at all, to get back to business with Shaheen’s third term. However, Sen. Shaheen also made an even larger historical accomplishment that is also apparently not getting much (if any) of the major press it would seem to deserve.
Jeanne Shaheen got the most votes of any federal level race in the history of New Hampshire (almost 451k). That is more than any senate candidate and any presidential candidate, ever in the state. Re-read that, sounds like a pretty big deal – yet the crickets chirp. I would think that Shaheen outperforming the likes of Obama (384k & 369k), Hillary Clinton (348k), and even the new record vote holder for president in NH, Joe Biden (424k) would be something even an opposing party would understand a bit of extended celebrating for and would still be dominating the news of the state weeks later.
Oh well, guess not.
Actually, I have scrolled a good amount through the headlines following election day on news aggregators and cannot cite a single headline calling out either of these major historical accomplishments. I’m sure they exist, somewhere, right?
To put a little more context into the Shaheen vote total, the last time she won she garnered 251k, so she got 200k more votes this time than last time. 2014 was a mid-term election but it was also quite a competitive race against former Senator Scott Brown. The first time she ran back in 2008 on the wings of Obama she had about 359k, so she got almost 100k more in 2020 than in the ‘Democratic Peak’ year. In New Hampshire a vote total change of 100 thousand is huge, 200 thousand is astronomical. Some will say Messner was a less than great candidate and we know the Democrats thoroughly painted him as an out-of-stater, a classic political play that they also used on Mowers – though I think most of NH’s residents are from somewhere else at this point (and it didn’t seem to matter to them when the Clinton’s moved to NY solely for Hillary’s senate run). The nativist otherization by the NH Democratic party in 2020 is another ironic story of its own.
Next, lets continue as we did in the presidential race with a brief Benford’s Law analysis.
In the chart above we have again plotted the Benford distribution line and the histograms associated with the vote data for Sen. Shaheen. I will repeat the disclaimer from chapter 1 that Benford’s Law gives you an expected distribution of values. It can be used as an indicator of potential fraud but is not proof of fraud on itself. In this case, the 1, 2, and 8 position would stick out as being over-represented in the distribution. This does not mean that they are fraudulent, but it would be a good first level screening tool to refine where to look to next if fraud were suspected. Perhaps coincidentally, these are the same number positions that were also overrepresented in the Biden vote distribution. Also, like Biden’s result, it is safe to at least say they do not fit the line very well. What about Messner’s results?
Messner’s plot also does not fit the expected line very well, with most points over and under. He also went over on point 1 and 8 but not by as much as Shaheen did. So, the Benford analysis is less interesting than the raw vote totals. Which we should actually compare those because when looking at the percent of senate candidate votes relative to the presidential candidate by party, not only did Shaheen smash historical records in the relative percentage and total votes, but Messner did unusually low with 89% of Trump’s vote (typically Republican senate candidates in NH garner high 90’s percentages) and this was while being specifically endorsed by Trump. In wrapping up the Benford discussion we can also now conclude that of all the federal level state-wide races (Shaheen/Messner/Biden/Trump); Trump’s Benford chart is the closest to the expected distribution.
Since we are comparing only years with Presidential races, the first and last are both Shaheen but the middle is Maggie Hassan which still establishes high turn-out party preferences for the federal office. If you looked at only Shaheen races, her second would be artificially low as it occurred in a mid-term. So, the comparative data here shows us that not only did the unusual ratcheting up of 2020 appear similar to Biden’s, but also Messner did not extend the trend line that would have continued, even though Trump did just that perfectly. The story goes here that a historically low number percentagewise that supported Trump did not support his party’s senate candidate, even with his endorsement. Whether or not those votes went to Shaheen instead would not account for the total votes (even if 100% of them did there is still a vast gulf of support that is less clear in origin in terms of party voters).
Even in that less likely event of some Trump voters deciding to also support Shaheen we end up with the same question as when reviewing the presidential race; where did all that additional support come from? Trump kept on pace and increased support, so we were able to conclude republicans didn’t vote for Biden. Not only did Trump get the highest votes for any republican presidential candidate in NH history (which he also did in 2016, so he outperformed his previous record) but despite the loss, Messner also earned the second highest votes for any republican senate candidate in state history; Kelly Ayotte still retains the title from 2016.
In summary, Chapter 2 has told us the story of the trendlines for senate party preference apparently performing a complete flip as shown with the sharp divergence in the line graph above. While that is interesting enough the story from the numbers also tells us of the massive vote totals acquired by Sen. Shaheen in terms of 106% of Biden’s votes and the highest federal office vote total ever in the state.
Key historical events are shown by the data, yet these major stories seem to have been left well alone by the media, which is perhaps the oddest part of the tale. These highlights are low-hanging fruit practically falling off into the media’s hands and yet go unharvested.
Why?
