USA Today ran a story a few days ago summing up the latest polling in the Kentucky gubernatorial contest:
Bluegrass Poll: Conway maintains lead over Bevin
Democratic candidate Jack Conway was up by five points and Republican candidate Matt Bevin had some huge weaknesses when it came to questions voters had about him. The poll — sponsored by WHAS11, The Courier-Journal, The Lexington Herald Leader, and WKYT — had a margin of error of 3.5 points, and found that Conway got 45% of voters, while Bevin trailed with 40%.
Now here’s the lede from a Courier-Journal story about the election:
Matt Bevin became the first governor from Louisville in a century and Lt. Gov.-elect Jenean Hampton became the first African American to win statewide office in Kentucky on Tuesday as Bevin led a near-Republican rout of state constitutional offices.
Democrat Jack Conway failed to roll up the large margin in Jefferson County he needed and couldn’t minimize losses elsewhere as Matt Bevin grabbed large margins throughout the rest of the state.
With 93 percent of the vote counted, Bevin led Conway 53 percent of the vote to 44 percent – a near landslide.
So while the percentage supporting Conway turned out to be about right, the percentage supporting Bevin was off by 13 percentage points.
So much was off in Kentucky. Someone sent around this snarky and obseqious Rachel Maddow bit in favor of “rising Democratic star” Adam Edelen. Edelen wasn’t just going to win re-election as auditor, that was just a brief stepping stone to his challenge of U.S. Sen. Rand Paul in next year’s Senate race. He raised a whopping $900,000 and ran TV ads in the last week. He ended up losing to someone who cobbled together $37,000 for his campaign.
All this is just the latest example of how conventional wisdom, centered around failed poll predictions, is wrong. And it’s a big deal that it’s wrong, because so much of our political journalism follows the lead of polling.
polling is just abysmal and yet it drives so, so much of our political coverage — and determines presidential debate participants!
Peter Hamby (@PeterHamby) November 4, 2015
A quick word in defense of even Kentucky polling, which helped strategists on both sides figure out that this was a close enough race to pour resources into at the end. Conway was winning for much of the race and the numbers broke in Bevin’s favor a bit late. That’s why you saw so much recent investment in the race, even from a reluctant Republican Party.
But back to why polls are given way too much weight by political journalists. Remember how the bluegrass poll had a margin of error of 3.5 percent. It was based on a survey of 798 “likely voters.”
When you’re extrapolating so much from 798 people, you have to make a lot of assumptions about what the demographic makeup of an electorate will be and what type of people will be voting. One of the assumptions most pollsters made was that they used a model based on a previous off-year election when Gov. Steve Beshear trounced his weak Republican opponent by more than 20 points. So some pollsters were using a model that assumed an electorate that was more inclined to vote Democratic than it actually was.
It’s difficult to determine what model to use and what assumptions should go into that model. Which is why polling is at least as much art and science, and some pollsters do better gauging certain electorates than others. What very few pollsters figured out in this election was how deeply unpopular President Obama is in Kentucky and how that affected voters’ willingness to support his party, by voting for a Democrat or even by showing up to vote.
Polling is like everything else in life — people are afraid to get too far away from the herd. If you stick to the convention wisdom and get it wrong, you’re in pretty safe company. If you have different assumptions, you’re taking a risk. Again, all this is fine and explains why polling is such a big business in political campaigning.
However, if you’re a journalist, you simply need to be aware of how tiny errors in assumptions can yield absolutely huge errors in poll results. Polling is an excellent tool for campaign consultants but no journalist should allow himself to be manipulated by polling when polling is so frequently in error.