Clinton significantly underperformed her polls across the Midwest and Rust Belt in the 2016 election, costing her the Electoral College.

Judged against certified 2016 results and final 2016 polling averages (e.g. RealClearPolitics, FiveThirtyEight, HuffPost) for Midwest and Rust Belt states (MI, WI, PA, OH, IA, MN). True: Clinton's actual margins fell well short of her final polling averages across most of these states (misses of several points), and losing Michigan, Wisconsin and Pennsylvania was decisive for the Electoral College, meaning she would have won with them. Mostly true: the broad pattern holds but the underperformance was small or absent in one or two of the named states. Mixed: the underperformance was clear in some key states but not in most of the region, or it was not decisive for the Electoral College outcome. Mostly false: the polling misses were small (about 1 to 2 points) and not region-wide, so 'collapsed' overstates them. False: Clinton matched or beat her polls in these states, or the results there did not decide the Electoral College.

Set before the research began.

1. E1 AN EVALUATION OF 2016 ELECTION POLLS IN THE UNITED STATES AAPOR Ad Hoc Committee on 2016 Election Polling

   “State-level general election poll errors were much larger, with an average absolute error of 5.1 points and an average signed error of 3.0 points. This was exacerbated by state polls, which indicated the wrong winner in Pennsylvania, Michigan and Wisconsin – states that collectively were enough to tip the outcome of the Electoral College.”

2. E2 7 experts try to explain how the polls missed Donald Trump’s victory Vox · 9 November 2016

   “The whole region was something of a black hole for pollsters. In the home stretch, Clinton was projected to be up by around 7 points in Wisconsin. She lost it by a point. In Iowa, she was only down by 3 points in most polls. She lost it by a whopping 9 points.”

3. E3 7 experts try to explain how the polls missed Donald Trump’s victory Vox · 9 November 2016

   “Clinton lost the state. Grossmann wasn’t alone in badly missing it — most polls had her up by around 4 points in the state.”

4. E4 7 experts try to explain how the polls missed Donald Trump’s victory Vox · 9 November 2016

   “The RealClearPolitics average for Clinton in Pennsylvania, for instance, only had her up by 1.5 points — well within the margin of her error in her loss.”

The AAPOR committee report [E1] says state polls named the wrong winner in Pennsylvania, Michigan and Wisconsin, and that those three states together were enough to tip the Electoral College, so the decisiveness part of the claim holds. Vox [E2, E3] reports large misses in the region: Clinton led late Wisconsin polls by about 7 points and lost by about 1; she trailed by about 3 in Iowa polls and lost by 9; and in another state, which the quote does not name, most polls had her up by about 4 before she lost it. In Pennsylvania, however, the RealClearPolitics average had her up only 1.5 points, a miss Vox calls within the margin of error [E4], and the evidence gives no figures for Ohio or Minnesota. The overall pattern holds, but the miss was small in at least one named state, which fits "mostly true".