Chantilly, VA – Susan Falconer, Campaign Manager for Barbara Comstock For Congress released the following statement after Monmouth University released a Monmouth Modeling Experiment today:
“The Monmouth “modeling experiment” was wrong in June and it’s wrong again. There were no polls consistent with their June “modeling experiment” and none since. In 2016 this group predicted a Hillary Clinton/Tim Kaine victory of 6 points which included erroneous predictions they would win Florida, Ohio, Pennsylvania, and Wisconsin. We are used to the media and Democrats celebrating their victories pre-election day.
“We have consistently led in our internal polls this year and Barbara has always overperformed both our own internal polls as well as the pundit predictions. In 2014 – a good Republican year – Barbara overperformed Democrat predictions by 14-15 points. In 2016 – a tough year for Republicans – Democrat predictions were off by 11 points. So good year or bad year, Barbara Comstock has overperformed and the Democrat and media predictions have been wrong. Barbara hasn’t won 7 elections in 9 years by listening to the pundits. Has anyone predicted she would win every inside the Beltway precinct in Fairfax? She did in 2016. There is not a single Democrat – not Ralph Northam, not Terry McAuliffe, not Mark Warner, and not Tim Kaine – who has ever received more votes than Barbara Comstock in Virginia’s 10th District. Last week, “The Hill” newspaper reported that internal polls on both sides have this race closer than reported and that is why Republicans continue to strongly support this race in the closing weeks. Republicans also know things never reported by the press like the fact that her opponent has been an underperformer in her state senate district by 8-10 points historically
“The media rarely reports on the nature of the Monmouth modeling experiment. Unlike real pollsters, they say in print that the margin of error doesn’t apply to the sample they released. From their own methodology statement they explain: ‘For results based on this sample, one can say with 95% confidence that the error attributable to sampling has a maximum margin of plus or minus 5.1 percentage points (unadjusted for sample design). Sampling error can be larger for sub-groups.’ This is the same type of experimental modeling that Hillary Clinton’s campaign used in 2016 to their own detriment.
“Past erroneous Democrat polls that were released at this point in previous Comstock campaigns in 2014 and 2016 have been wildly off. On September 30, 2014 House Majority PAC released a poll to the media tried to prop up Democrat candidate John Foust claiming he was within two points of Comstock — 41% to 39. On Election Day in 2014 Barbara won by 16 points against Foust and 3 independents by a margin of 56-40. In 2016, the media trotted out phony polls anointing Luann Bennett the winner by 5 points only to be proven 11 points off when a few weeks later Barbara won by 6 points – 53-47. “
“With Northern Virginia televisions slammed with a slew of ads comparing Comstock to Trump, it looked as if it would be difficult for the GOP freshman to survive Trump headwinds in the D.C. suburbs. But Comstock ended up winning comfortably, running a whopping 11 points ahead of Trump in the district…” – Josh Kraushaar, National Journal, December 20, 2016