Do a poll
Carter and I disagree on most every issue, but we agree on one big thing: Good polling is essential in a campaign.
Which is why I’m obsessed with why Hillary Clinton’s campaign stopped state-level polling in the final weeks of 2016. I think about that decision every time Trump does something dumb, despicable or dangerous – that is, every day.
Stanley Greenberg, Bill Clinton’s pollster, charges Hillary’s campaign with “malpractice and arrogance” for stopping polls in states like Wisconsin, Pennsylvania and Michigan. Trump won all three.
Now, Politico’s Steven Shepard dives into the decision – and into the division between old polling heads like Greenberg and a new generation that believes in data analytics. Shepard writes:
“In October, about a month before Election Day, campaign officials pulled the plug on its deeper-dive state surveys — the messaging polls — over the objection of the campaign’s pollsters, who argued at the time that it was too risky to stop listening to voters beyond just the horse race, according to three sources associated with the campaign. The campaign continued, however, with its analytics polling, calling thousands of voters across battleground states every night to ask basic questions about candidate preference.
“The campaign’s pollsters argued against that decision at the time — analytics polling, they argued, might tell the campaign’s leaders whether Clinton was ticking higher or sinking lower but it couldn’t tell them why.
“Those pollsters were overruled by senior Clinton campaign leaders….”
Now, this may sound like the typical post-election finger-pointing by the losers. But the debate is important, because a lot of campaigns – national, state and local – will have the same debate during the 2018 and 2020 election cycles.
Data analytics can be a powerful campaign tool. But 2016 is powerful evidence that traditional polling is still vital.
A national pollster who was doing surveys in North Carolina told me the Friday before Election Day that Clinton was sinking here – and pulling down other Democrats. She started dropping when James Comey announced the FBI was reopening its email investigation.
Her campaign wasn’t polling, so they were flying blind. They didn’t know she was falling behind in states she couldn’t afford to lose in the Midwest. Instead, they held the closing rally in North Carolina, a state Clinton didn’t have to win and, in the end, couldn’t win.
So, whenever a candidate asks for advice, I’m sticking to four things that I know work – and that Carter and I agree on: (1) Raise money. (2) Do a poll. (3) Give voters information. (4) Repeat.
Do a poll
Carter and I disagree on most every issue, but we agree on one big thing: Good polling is essential in a campaign.
Which is why I’m obsessed with why Hillary Clinton’s campaign stopped state-level polling in the final weeks of 2016. I think about that decision every time Trump does something dumb, despicable or dangerous – that is, every day.
Stanley Greenberg, Bill Clinton’s pollster, charges Hillary’s campaign with “malpractice and arrogance” for stopping polls in states like Wisconsin, Pennsylvania and Michigan. Trump won all three.
Now, Politico’s Steven Shepard dives into the decision – and into the division between old polling heads like Greenberg and a new generation that believes in data analytics. Shepard writes:
“In October, about a month before Election Day, campaign officials pulled the plug on its deeper-dive state surveys — the messaging polls — over the objection of the campaign’s pollsters, who argued at the time that it was too risky to stop listening to voters beyond just the horse race, according to three sources associated with the campaign. The campaign continued, however, with its analytics polling, calling thousands of voters across battleground states every night to ask basic questions about candidate preference.
“The campaign’s pollsters argued against that decision at the time — analytics polling, they argued, might tell the campaign’s leaders whether Clinton was ticking higher or sinking lower but it couldn’t tell them why.
“Those pollsters were overruled by senior Clinton campaign leaders….”
Now, this may sound like the typical post-election finger-pointing by the losers. But the debate is important, because a lot of campaigns – national, state and local – will have the same debate during the 2018 and 2020 election cycles.
Data analytics can be a powerful campaign tool. But 2016 is powerful evidence that traditional polling is still vital.
A national pollster who was doing surveys in North Carolina told me the Friday before Election Day that Clinton was sinking here – and pulling down other Democrats. She started dropping when James Comey announced the FBI was reopening its email investigation.
Her campaign wasn’t polling, so they were flying blind. They didn’t know she was falling behind in states she couldn’t afford to lose in the Midwest. Instead, they held the closing rally in North Carolina, a state Clinton didn’t have to win and, in the end, couldn’t win.
So, whenever a candidate asks for advice, I’m sticking to four things that I know work – and that Carter and I agree on: (1) Raise money. (2) Do a poll. (3) Give voters information. (4) Repeat.