Margin of Error

The post-election autopsy is well underway. The obvious question, of course, is why was everyone so wrong about the eventual outcome? (Even some Trump advisors were skeptical he would win. So…) Specifically, why were the polls so wrong when almost every single poll had Hillary Clinton ahead 3-4 percentage points.

Consider that most polls have a margin of error of 2-3 percentage points. Nate Silver of FiveThirtyEight.com contemplates a scenario where 2 percentage points shifts to Clinton’s favor. This would be the equivalent of 1 in 100 voter changing from Trump to Clinton. The result:

screen-shot-2016-11-09-at-4-10-35-pm

It’s a very different electoral map.

When you consider that Hillary is winning the popular vote and, if current trends hold, should eventually end up winning 1 to 2 points, were the polls really that far off? Were they within the margin of error? Maybe. That margin of error, however, translates to a drastically different electoral landscape. History, thus, is created in the margins.