With the initial returns last night, Democratic candidate James Thompson appeared to be in a position few anticipated: he was leading Republican Ron Estes by twenty points in Sedgwick County (Wichita) and by sixteen points in neighboring Harvey County. It was only after the most rural counties rolled in, and the election day votes too from these larger areas, that Estes zoomed into the lead. Estes would finish the race just shy of a seven point margin, a double-digit decline from President Donald Trump’s performance there just five months back. Here are a few things we’re taking away from last night’s results.
First, Democrats are energized, as they should be: they’re the out party and are determined to climb their way back into a majority. Republicans went through their own wilderness period in the two years following the 2008 Presidential election, losing several open house seats like NY-23 and PA-12, before their landslide flip of the House in 2010. Democrats so far have yet to flip a Congressional District but may have better luck in Georgia next week, where Jon Ossoff has managed to amass nearly all of the Democratic support in the region behind him.
Second, we notice that Republicans really delivered on election day after trailing badly in early voting here. In Sedgewick, the early vote broke for Thompson by 5,663 votes. But Estes won the election day vote by 3,789 votes. In Harvey, election day voting was enough to swamp Thompson’s early vote win, and the pattern grew more in the Republican’s favor the further from Wichita you traveled. This is important to keep in mind for next week’s race in Georgia, as early voting figures have drifted from being majority Democratic on a daily basis to majority Republican. If Republicans there turn out strongly on election day, we could see a similar pattern: Ossoff starting out ahead of the fifty percent mark only to fall under as the GOP voters come through.
Third, local forces matter, parties that respond to them reap the benefits, but they are one factor, not the end-all for close contests. Kansas Democrats knew Governor Sam Brownback was deeply unpopular. Thompson latched onto this far more than he challenged Trump, and Estes’ initial response to the attacks was muddled. Local Republicans managed to sound the alarm and get national help to avoid a surprise upset, while Democrats who may be disappointed they failed to capture the seat should take considerable pride in the incredible margin reduction they accomplished. But you can’t associate the closeness of this contest entirely with Brownback: when he narrowly won re-election in 2014, Republican Congressman Mike Pompeo won his race here by thirty-three points.
Fourth, the temptation is strong for people to over-analyze this special election because both parties are looking at every result that comes down the river as proof they are doing swell/on the march/holding their own/insert narrative here. There may be a temptation to apply a generic surge rate based on previous races to the upcoming contests in Georgia and the rest- one that would convert narrow losses in some to anticipated Democratic victories- without accounting for the fact that these are irregular elections staggered out over the course of many months. There will be a temptation to dismiss this contest as a fluke or oversold because the Republican still won: it may have been close, but it was no one-pointer. In the end, turnout rates matter, turnout operations matter, and each race is unique. If Ossoff wins next week outright, that doesn’t mean Quist becomes the favorite in Montana; and if Ossoff fails to clear 41%, that doesn’t mean Democratic efforts to win back the House have stalled.
Fifth, we are very proud of the election night team here at DDHQ, which provided a continuous stream of data throughout the night. While I was personally skeptical of the race at its onset, we as a group insisted on giving this contest our full attention, and I am glad we did so. You never know which races will turn out to be surprises (like the Virginia 7th Republican primary in 2014) or duds (like the Wisconsin DPI contest this year), so we approach as many contests as possible with the idea that every one of them matters, their voters matter, and the public wants the information quickly and accurately. We can’t wait to show you what we have in store for Georgia 6.