The Topline: The Trump administration is not poll-driven

Say what you will about the Trump administration -- and with disapproval in the Gallup daily tracker hitting 60 percent last week, people have a lot to say -- but they are not committing that cardinal political “sin” of governing by poll. In fact, they frequently seem fully committed to the opposite: finding the popular route and going the other way.


Gallup daily presidential approval poll, 3-day average.

Ditching the polls has long been a key part of Trump’s brand and appeal to his base of voters, who see him as different than other politicians who talk in measured, focus-grouped soundbites.

During the 2016 campaign, Trump famously scoffed at others for hiring pollsters, although he later brought one in and made another (KellyAnne Conway) his third and final campaign manager. His poll cherry picking became so infamous that Australian Prime Minister Malcolm Turnbull included it in his impersonation of Trump that leaked this week.

Since he became president, polls have not shown much good news for Trump, and he has mostly had little to say about them. But it’s in the arena of policy where the Trump administration’s apparent disregard for the polls is most apparent. A wide reading of the polls show Trump and his administration consistently taking the less popular path.

Medical marijuana - Recent polls have shown medical marijuana among the very few issues where Americans are nearly unanimous in their support. Polls have found support between 83and93 percent over the last few months. While Republicans are more likely than Democrats to oppose medical marijuana, they still want to keep it legal, by a margin of 70 percent to 26 percent in a recent Yahoo / Marist poll. Even on recreational marijuana, which Sessions has also targeted for increased federal enforcement, he is swimming against a growing wave of public sentiment in favor of legalization.

Healthcare - Opinion on the American Health Care Act is perhaps the most poignant example of going against public opinion. Trump pushed the House to pass the bill, and celebrated its passage in the Rose Garden, before telling Senators he thought it was “mean” and hoped the Senate version would be less so. As Senate Republicans work out their own version behind closed doors, support remains very low, averaging 29 percent in recent national polls. The New York Times’ Upshot estimates that opinion is tilted against the AHCA in all 50 states.


Climate change - On climate, polls have shown majorities of Americans believe climate change is real and at least partially man-made, and they oppose Trump’s decision to withdraw from the Paris agreement.

Budget matters - Voters perennially support cutting “wasteful spending,” but Pew polling finds that, when pressed, most don’t actually want to cut much other than foreign aid. Even so, the Trump administration approach to the budget has been more has been more meat ax than scalpel.

Twitter - A recent Morning Consult/Politico tracking poll shows 69 percent think Trump uses Twitter too much. Even many of his own supporters wish he would cool it.

Not governing by polls is a time-honored claim of politicians, even as they pour over crosstabs behind the scenes. There’s a reason they do: polls help politicians get reelected. They also help leaders understand the views and wishes of their constituents, even if they choose a less popular course. Whether Trump is truly ignorant of what polls say, or if he is aware but choosing to ignore it, his policy strategy seems to be to do the exact opposite of what the public says it wants.


CommonWealth Magazine - Dissecting Baker’s stance on millionaire’s tax.

“Governor Charlie Baker hasn’t formally said he’s against the so-called ‘Fair Share’ ballot question, but the totality of his public comments and statements from his office certainly suggest more opposition than support. He points to tepid, 1 to 2 percent income growth as evidence the state needs to ‘continue to live within our means.’

“But this argument, focused on the average growth rate, masks a major disparity between the highest earners, who would be subject to the proposed tax, and everyone else who would not. Income gains in recent years have been heavily skewed toward those at the top of the economic ladder.”



As mentioned above, Donald Trump hit a record 60 percent disapproval in Gallup’s daily tracker last week. Gallup’s Frank Newport notes he also hit a new low in approval in the poll’s weekly average.

WaPost’s Philip Bump looks at the latest polling and thinks it’s getting close to crisis mode for the Trump administration.

But Trump is holding Chris Christie’s beer as he plumbs the depths of terrible approval ratings. A recent Quinnipiac poll finds New Jersey votersdisapprove of his performance 81 to 15 percent. Harry Enten of FiveThirtyEight says this is the fifth worst poll of all time for a sitting governor.

Meanwhile, a HuffPost/YouGov poll finds that Barack Obama’s approval numbers have gone up since he left office.

Last month, we ran a Topline looking at Jon Ossoff’s primary run. The runoff is next Tuesday, and  polls have him in adead heat against Republican Karen Handel.

Your pastor is probably partisan, according to a new study. The New York Times report breaks down the findings by denomination, location, gender, and age.

Our polling partner, WBUR, has a new podcast! Ron Suskind and Heather Cox Richardson will look at the national political landscape through a historical lens in Freak Out and Carry On.

The snap elections British Prime Minister Theresa May called snapped back on her, resulting in a hung parliament. Nate Silver argues that the surprise result was not entirely surprising, in part because British polls have been so inaccurate in the past.

Online pollster YouGov had mixed results predicting the final margin, but their seat-by-seat model was one of the few to predict a hung parliament. Here’s their post mortem.

YouGov used a method called Multi-level Regression and Post-stratisfication (MRP), which uses statistics to predict local results from national polling and demographics. (It’s the same method The Upshot used for the 50-state health care poll cited in the lede above.) The Monkey Cage blog has an explanation of MRP, which showed promise in the 2016 presidential election, as well.

Finally, a disturbing number of Americans think that chocolate milk comes from brown cows.

--------------------------------NERD ALERT RIP-------------------------------

We come to bury Huffpollster, and to praise it.

The Huffington Post announced a round of layoffs this week, including one of the two remaining staff members working on the HuffPollster site. The remaining editor, Ariel Edwards-Levy, announced the site will stop aggregating poll data other than Donald Trump’s approval rating and the 2018 generic congressional ballot. Ariel will also continue to produce some polling and analysis, but this narrowed scope is a big blow for anyone who cares about polling and politics.

In its heyday, HuffPollster aggregated and tracked essentially every state and national race that had a decent amount of polling, and even many local races like the 2013 Boston Mayoral contest. The wealth of polling data this poll tracking produced was a boon to pollsters and researchers and spawned many research and news articles over the years, including some by your loyal correspondents. The site started as the Mystery Pollster blog in 2004, then became, and was acquired by The Huffington Post in 2010.

We hope Huffington Post will reconsider, or find a new home for the site, the terrific analysis for which it is known, and the reams of valuable data it houses.