Debrief · 2016 cycle
The 2016 Debrief · post-mortem · permanent archive

How did the polling industry do this cycle?

The 2016 industry MAE landed at 6.5pt — graded across 454 historical races with a known outcome — and 454 polls were analyzed across 81 pollsters. Signed bias −4.55pt: pollsters again systematically over-predicted Democrats.

INDUSTRY MAE
6.5pt
vs midterm avg
SIGNED BIAS
−4.55pt
D-leaning
POLLS · POLLSTERS
454
81 firms
RACES SCORED
454
w/ known outcome
CYCLE 2012 2014 2016 2018 2020 2022 2024
Industry mean absolute error · 2012–2016
2012 4.7pt 2014 4.8pt 2016 6.5pt 2018 4.5pt 2020 5.9pt 2022 4.6pt 2024 4.1pt MAE · 2-WAY MARGIN
Per-pollster scorecard · 2016 · sorted by MAE
Signed bias · ±5.7pt
MAE · 0–7.5pt
data-orbital
−1.96pt
2.0pt
surveyusa
+0.31pt
2.4pt
insights-west
−1.06pt
2.4pt
nbcwsjmarist
−1.86pt
3.1pt
quinnipiac-university
−3.40pt
3.4pt
emerson-college
−1.58pt
3.8pt
clarity-campaign-labs
−3.86pt
3.9pt
the-times-picayunelucid
−3.97pt
4.0pt
cnnorc
−2.16pt
4.4pt
public-policy-polling
−4.59pt
4.6pt
cbs-newsyougov
−3.79pt
4.8pt
wmurunh
−5.15pt
5.2pt
monmouth-university
−5.12pt
5.4pt
breitbartgravis-marketing
−4.22pt
5.4pt
loras-college
−2.24pt
7.0pt
Biggest-miss races · 2016
Race Expected Actual Miss Direction
us-senate-2016-ok D+26.00 D+43.16 -17.2pt R-margin underestimate
us-governor-2016-nd D+40.00 D+57.13 -17.1pt R-margin underestimate
us-senate-2016-ut D+27.50 D+41.09 -13.6pt R-margin underestimate
us-senate-2016-ak D+28.50 D+15.20 13.3pt R-margin underestimate
us-senate-2016-nd D+48.50 D+61.51 -13.0pt R-margin underestimate
us-senate-2016-id D+26.00 D+38.40 -12.4pt R-margin underestimate
us-senate-2016-ar D+13.00 D+23.60 -10.6pt R-margin underestimate
us-governor-2016-wv D+3.50 R+6.79 10.3pt R-overestimate