A month after Trump’s upset victory, the aftermath is still sending shocks through the United States and the world. In addition to a hard rightward move on policy, Trump, Pence, and Company appear to be bent on uprooting many institutions. The risk to the American system of government and life has been noted by both liberals and conservatives.
Hillary Clinton’s narrow loss to Donald Trump was influenced by many causes in the home stretch: complacency driven by conventional wisdom and polls (and yes, poll aggregation), which led to the media assumption that she would win, which in turn was a likely driver of the tone of coverage. And of course there is so much to say about the candidates themselves.
In mid-October, I said I didn’t think Trump would clear 240 electoral votes, a statement I paid for later by eating a bug on CNN. My error seems to be accounted for by two events: (1) undecided Republican voters coming home, and (2) FBI Director Jim Comey’s letter to Congress about Clinton’s email.
From opinion data alone, it is possible to estimate when a change occurred. This can test between alternative explanations, which include not only the Comey letter (October 28th) but preceding events such as the announcement a hike in Affordable Care Act premiums (October 24th). However, it is not possible to see the shift using the averaging methods used by other aggregators. They tend to smear results out over time. For example, the Huffington Post does not allow a sudden shift to be seen.
I calculated a day-by-day margin using polling data from the Huffington Post. These polls were done on multiple days, which I converted to individual dates using the following procedure:
- Made a large table with one date per column (spreadsheet here);
- For each survey, entered the Clinton-Trump margin across all dates covered by the survey sample;
- Calculated the median Clinton-Trump margin by date;
- Shifted the time axis by 3 days, to correct for the fact that each survey also covered earlier and later dates.*
The above graph shows the result. After the Affordable Care Act premium hike announcement, opinion did not move for days, arguing against this as a main driver of the late swing in opinion. It could still be a factor, as is the case for many events. But such an effect would have to be gradual.
However, the big change does coincide well with the release of the Comey letter. Opinion swung toward Trump by 4 percentage points, and about half of this was a lasting change. This was larger than the victory margin in Michigan, Pennsylvania, Florida, and Wisconsin. Many factors went into this year’s Presidental race, but on the home stretch, Comey’s letter appears to have been a critical factor in the home stretch.
*Based on reader feedback and some correspondence with Marcy Wheeler, this post has been modified. The original version is here. The graph is the same, it’s just shifted by three days to avoid “time travel,” in which the effect appeared to show up before the cause. This is a bit of a kludge: a fixed-interval correction is not possible because each survey was done over a different time period. A better alternative would be to calculate medians using only the final sampling date. I’ll try that out later – traveling at a conference now.