Without a doubt, the superlative early voting effort by Democrats and allied groups is largely responsible for Democrat Liz Mathis' landslide victory over Republican Cindy Golding. Although the results are still unofficial and precinct-level demographics are not yet available, sufficient detail exists to draw some preliminary conclusions from the early reporting.
The reported absentee split tells the story, with Mathis garnering 71.56% to Golding's 27.79%. As of 8 Nov, the returned absentees broke down as follows:
She's (Mathis) taking 15% of the GOP vote from Golding, while losing only 9% of the Democratic vote.
the adjusted figures are 3765 Democratic and 2041 Republican. Noting that Golding received only 2106 actual votes by absentee suggests that the independent vote broke overwhelmingly for Mathis, even if PPP's estimates are not accurate for early voting. This supports the early observations made here of not just elevated Dem absentee turnout relative to 2010, but of a robust independent response as a result of Dem efforts, particularly in the precincts that skew younger.
The PPP poll put Mathis at 52% with a 6 pt advantage over Golding. One possibility for the conservative estimate is that perhaps the poll did not capture sufficient younger voters judging by the age breakdown found in the poll's internals. I found the structure of the voter screen problematic, given the importance of early voting in this election.
they pulled a list of voters who had voted in at least one of the last three general elections, and the call script began by asking respondents to "hang up now" if they were not planning to vote on Tuesday.
Voters in the district surely were inundated by GOTV contacts, so it's not difficult to imagine that the response by an early voter to "hang up now if you do not plan to vote on Tuesday" would be to do exactly that, rather than thinking "oh, they must also mean have I already voted?" The precinct-level demographics will ultimately reveal whether Dems and independents enjoyed a boost in under-35 voting relative to 2010. The absolute domination by Mathis of the early NP vote as compared to PPP's estimate of an underwater 45%-50% performance, suggests a significant youth component, the demographic most receptive to her campaign, during early voting. I also suspect that Dems captured a greater share of the crossover vote during early voting, almost certainly better than 15% estimated for total turnout.
As predicted, the GOP fell short of 2010 early voting performance by a considerable amount. Ignoring the small contribution from split Cedar Rapids precincts, the partisan performace relative to 2010 is as follows:
Total turnout in this special election is approximately 80% of 2010 turnout. If Dem and NP early voting had tracked 2010 performance, the GOP effort would actually be in line with their 2010 performance as well. This suggests that ascribing the breakdown to "depressed" GOP turnout is perhaps less relevant than crediting the performance of the Dem early voting campaign, which put the election out of reach for Golding.
It is interesting to note that the early voting fraction this time was about one third as compared to one quarter in 2010, while the actual number of early voters is almost the same but with very different partisan distribution. When statistical demographics become available, the question of how to allocate reduced GOP enthusiasm vs unexpectedly high Dem early voting turnout will be answered.
Below are the election day results by precinct. The second figure is a more detailed view of the heavily populated area of the district, which includes Marion. Republican Cindy Golding prevailed in the election day match-up 50.95% - 48.36%.
Senate district 18 polling day results
Polling day results in the Cedar Rapids suburbs
There are few surprises. North Marion tends to vote more Republican than South Marion. Hiawatha 1 is routinely more blue than Hiawatha 2, and the same holds for Cedar Rapids 46 vs. Cedar Rapids 47. Robins, Liz Mathis' home precinct, is a Republican stronghold. One could almost say that the overall split reflects the lean of the district, and that was the problem for Cindy Golding. No surprises on Election Day meant no opportunity to overcome the formidable early voting advantage accrued by the Democratic team. The overall partisan split by precinct will be estimated when the precinct-level demographics are available. Unfortunately, election offices are only required to report absentees results by precinct (instead of a central absentee precinct) for gubernatorial and presidential contests.