The Polls Put Democrats in a Trap of Their Own Making

Remember the heady days of June 2016, when Hillary Clinton led Donald Trump in every one of twelve major polls — by a margin of 10 points, according to ABC and the Washington Post?

Or August 2016, after the Khan imbroglio, when Clinton was beating Trump by 7 points, according to USA Today/Suffolk?  It was hardly an outlier.

Or October 2016, after the exposition of the "Billy Bush tapes," when, coming up on the election, Clinton was beating Trump in virtually every poll?  Of the 78 polls started after September 21, 2016, 73 had Clinton winning, four had Trump winning, and one was a tie.

At the time, I said, in two articles that Trump would be the 45th president of the United States.  I explained why.

Now, in June 2020, the same press corps has pronounced the Trump presidency dead, has held the funeral, and has lowered the coffin into the grave.

The problem?  No corpse.

THE SIX PERCENT RULE: In 2016, I predicted that Trump could lose the popular vote by as much as 4% and still win the Electoral College.

This is because roughly 4,000,000 Democrat voters in California, New York, and Illinois are in excess of the plurality needed to win those states.  The difference between winning California by 50.0001% and winning it by 99% is meaningless in a presidential election.

In fact, Trump lost the popular vote by 2.1% in 2016 but won the Electoral College, 304 to 227.

In June 2020, you can add an additional 2% to the margin by which Trump can be losing in the polls without being behind in the presidential race — for a total of 6% (in the case of national polling).  The reason is that, in late summer or early fall, pollsters' samples will shift from "registered voters" to "likely voters."  This will give Trump an almost automatic 2% bump.

The bottom line?  If Donald Trump is less than 6% behind in national polls, he is probably leading in the presidential race — even if you ignore the methodological problems with the polls. And while the battleground state polls aren't governed by electoral college dynamics, most of them are reduced to "margin of error" once the 2% bump is taken into account.

By these metrics, Trump was on course to re-win the presidency in every major poll in March.

THE POLLS THEMSELVES: What does it say that the polls have shifted in the last month?  Well, polls have not proven particularly reliable in matters relating to Donald Trump.  But some are unreliable-er than others.

CNN: This past week, the Trump campaign demanded that CNN retract its outlier poll purporting to show that Trump trailed Biden by 14%.  It claimed that the poll's sampling included too few Republicans — and that the timing of the poll was intended to maximize the impact of bad news and minimize the effect of good news, such as better-than-expected employment numbers.  (The "deliberate timing" issue was particularly significant in skewing the results of the erroneous polls in 2016.)

The critique of CNN's methodology was done with the assistance of the polling firm McLaughlin and Associates.  It was probably well taken — with respect to both the methodology and the intention of the pollster.

ABC/WASHINGTON POST: Similarly, the recent ABC/Washington Post poll showed a 10% gap in favor of Biden.  The Post, in particular, in polling neighboring Northern Virginia, has a history of producing wildly misleading numbers early (and sometimes late) in a campaign.  For example, Virginia gubernatorial candidate Ken Cuccinelli was found by the Post to be 11 points behind his Democrat opponent on the eve of his election.  Many Republicans had already washed their hands of the race as a result of the polls.  But the final results were a relative squeaker: 48% to 45.5%.

Monmouth: This is one of the most consistently biased of the polls.  Even if you discount Monmouth, it weighs down the polling average — which now stands, for Trump, at an 8.1% loss to Biden.  By contrast, in October 2016, Monmouth had Trump losing 50% to 38%.

Notwithstanding the weight of polls like this — with a long history of methodological bias — RealClear's 8% Trump deficit was within the margin of error of the 6% that would put Trump on track for victory in November.

ISSUE POLLS: Issue polls are even easier to jimmy than head-to-head polls between popular figures.  This is because you can word these polls to reflect any bias you want.

For example, USA Today published a poll purporting to show that the American people favored doing everything possible to ensure that individuals didn't "get sick and die" as a result of the coronavirus.  The alternative was doing everything possible to avoid a "recession."  Notice how the newspaper makes one option seem clinical and mild and the other option personal, inevitable, and horrific.

On June 11, USA Today published another poll asking whether respondents supported "protesting peacefully outside the White House."  Not surprisingly, most Americans support "peaceful protest" — if the pollster neglects to mention the protest fire–damaged historic St. John's Church, facing the White House just across Lafayette Square.

This is important because people's attitudes toward racial unrest may be decisive in the presidential race.

One poll found that roughly one sixth of Americans support the rioters who took to the streets in the wake of the death of George Floyd.  One sixth!  This number doesn't sound remotely credible.  But if one sixth of Americans truly approve of major burning and looting in 40 cities, it is not the one-sixth that will decide the November election.  And it is an indication that America has become truly polarized.

Union Democrats in the Western tier of counties in Pennsylvania — and in Macomb County, Michigan — will surely be alienated by these images.  And Biden's virulently anti-gun stance was already slated to hurt him in these areas.  There's a reason why veteran Michigan Democrat representative John Dingell was pro-gun — or at least pretended to be.

We know the identity of the demographic that carried the Rust Belt in 2016, and Joe Biden has just done everything he can to minimize his chances of reaching them.

In fact, it's hard to believe that suburban white women in Montgomery County, Pennsylvania (outside Philadelphia) will respond positively, either — particularly given the stated objective of bringing the damage to these wealthy enclaves.

As a result, as of June 3, CNBC had Trump four points up in the Keystone State.

The one state in which motivated black turnout could really help Biden is North Carolina.  "The Tarheel State" has a tradition of black early-voter churchgoers being bused to polling places after church on Sunday.  But neither is this a demographic that will be easily motivated by images of looting.

Ohio, another state that Democrats perceive is in play (even in the face of an 8-point Trump victory in 2016), is, like Michigan and Pennsylvania, filled with pro-gun union Democrats.

As for Florida, it remains to be seen how much Trump's dalliance with sentencing reform will cost him in the Sunshine State.  It does not seem to have the effect of pulling a large segment of the minority vote into the Republican column.  But the good news for Trump is that the "O'Biden-bama administration" worked hard to alienate Florida's Cuban-American community through its flirtation with the Castros.

If Trump loses Florida, it is his own fault.  And if he wins Florida, he wins reelection.

So where does that leave us

A lot will happen between now and November:

  • Will Biden botch the debate, as expected?
  • How soon and permanently will the US pull out of the coronavirus pandemic?
  • Will the dearth of on-campus classes stymie the 2018 Democrat voter registration model?

At this point, I'd have to give the advantage to Trump, assuming he doesn't do something really stupid — like going  after guns in the hope of carrying Broward County.

The big problem for Democrats is that if you lie to gain advantage, the one thing you have to remember — if only in your own mind — is what is the lie and what is the truth.

Michael E. Hammond is the former executive director of the Senate Steering Committee and was often called the "101st Senator" when he worked on Capitol Hill.  Today he serves as the general counsel of Gun Owners of America.

Image: Gage Skidmore via Flickr.

Remember the heady days of June 2016, when Hillary Clinton led Donald Trump in every one of twelve major polls — by a margin of 10 points, according to ABC and the Washington Post?

Or August 2016, after the Khan imbroglio, when Clinton was beating Trump by 7 points, according to USA Today/Suffolk?  It was hardly an outlier.

Or October 2016, after the exposition of the "Billy Bush tapes," when, coming up on the election, Clinton was beating Trump in virtually every poll?  Of the 78 polls started after September 21, 2016, 73 had Clinton winning, four had Trump winning, and one was a tie.

At the time, I said, in two articles that Trump would be the 45th president of the United States.  I explained why.

Now, in June 2020, the same press corps has pronounced the Trump presidency dead, has held the funeral, and has lowered the coffin into the grave.

The problem?  No corpse.

THE SIX PERCENT RULE: In 2016, I predicted that Trump could lose the popular vote by as much as 4% and still win the Electoral College.

This is because roughly 4,000,000 Democrat voters in California, New York, and Illinois are in excess of the plurality needed to win those states.  The difference between winning California by 50.0001% and winning it by 99% is meaningless in a presidential election.

In fact, Trump lost the popular vote by 2.1% in 2016 but won the Electoral College, 304 to 227.

In June 2020, you can add an additional 2% to the margin by which Trump can be losing in the polls without being behind in the presidential race — for a total of 6% (in the case of national polling).  The reason is that, in late summer or early fall, pollsters' samples will shift from "registered voters" to "likely voters."  This will give Trump an almost automatic 2% bump.

The bottom line?  If Donald Trump is less than 6% behind in national polls, he is probably leading in the presidential race — even if you ignore the methodological problems with the polls. And while the battleground state polls aren't governed by electoral college dynamics, most of them are reduced to "margin of error" once the 2% bump is taken into account.

By these metrics, Trump was on course to re-win the presidency in every major poll in March.

THE POLLS THEMSELVES: What does it say that the polls have shifted in the last month?  Well, polls have not proven particularly reliable in matters relating to Donald Trump.  But some are unreliable-er than others.

CNN: This past week, the Trump campaign demanded that CNN retract its outlier poll purporting to show that Trump trailed Biden by 14%.  It claimed that the poll's sampling included too few Republicans — and that the timing of the poll was intended to maximize the impact of bad news and minimize the effect of good news, such as better-than-expected employment numbers.  (The "deliberate timing" issue was particularly significant in skewing the results of the erroneous polls in 2016.)

The critique of CNN's methodology was done with the assistance of the polling firm McLaughlin and Associates.  It was probably well taken — with respect to both the methodology and the intention of the pollster.

ABC/WASHINGTON POST: Similarly, the recent ABC/Washington Post poll showed a 10% gap in favor of Biden.  The Post, in particular, in polling neighboring Northern Virginia, has a history of producing wildly misleading numbers early (and sometimes late) in a campaign.  For example, Virginia gubernatorial candidate Ken Cuccinelli was found by the Post to be 11 points behind his Democrat opponent on the eve of his election.  Many Republicans had already washed their hands of the race as a result of the polls.  But the final results were a relative squeaker: 48% to 45.5%.

Monmouth: This is one of the most consistently biased of the polls.  Even if you discount Monmouth, it weighs down the polling average — which now stands, for Trump, at an 8.1% loss to Biden.  By contrast, in October 2016, Monmouth had Trump losing 50% to 38%.

Notwithstanding the weight of polls like this — with a long history of methodological bias — RealClear's 8% Trump deficit was within the margin of error of the 6% that would put Trump on track for victory in November.

ISSUE POLLS: Issue polls are even easier to jimmy than head-to-head polls between popular figures.  This is because you can word these polls to reflect any bias you want.

For example, USA Today published a poll purporting to show that the American people favored doing everything possible to ensure that individuals didn't "get sick and die" as a result of the coronavirus.  The alternative was doing everything possible to avoid a "recession."  Notice how the newspaper makes one option seem clinical and mild and the other option personal, inevitable, and horrific.

On June 11, USA Today published another poll asking whether respondents supported "protesting peacefully outside the White House."  Not surprisingly, most Americans support "peaceful protest" — if the pollster neglects to mention the protest fire–damaged historic St. John's Church, facing the White House just across Lafayette Square.

This is important because people's attitudes toward racial unrest may be decisive in the presidential race.

One poll found that roughly one sixth of Americans support the rioters who took to the streets in the wake of the death of George Floyd.  One sixth!  This number doesn't sound remotely credible.  But if one sixth of Americans truly approve of major burning and looting in 40 cities, it is not the one-sixth that will decide the November election.  And it is an indication that America has become truly polarized.

Union Democrats in the Western tier of counties in Pennsylvania — and in Macomb County, Michigan — will surely be alienated by these images.  And Biden's virulently anti-gun stance was already slated to hurt him in these areas.  There's a reason why veteran Michigan Democrat representative John Dingell was pro-gun — or at least pretended to be.

We know the identity of the demographic that carried the Rust Belt in 2016, and Joe Biden has just done everything he can to minimize his chances of reaching them.

In fact, it's hard to believe that suburban white women in Montgomery County, Pennsylvania (outside Philadelphia) will respond positively, either — particularly given the stated objective of bringing the damage to these wealthy enclaves.

As a result, as of June 3, CNBC had Trump four points up in the Keystone State.

The one state in which motivated black turnout could really help Biden is North Carolina.  "The Tarheel State" has a tradition of black early-voter churchgoers being bused to polling places after church on Sunday.  But neither is this a demographic that will be easily motivated by images of looting.

Ohio, another state that Democrats perceive is in play (even in the face of an 8-point Trump victory in 2016), is, like Michigan and Pennsylvania, filled with pro-gun union Democrats.

As for Florida, it remains to be seen how much Trump's dalliance with sentencing reform will cost him in the Sunshine State.  It does not seem to have the effect of pulling a large segment of the minority vote into the Republican column.  But the good news for Trump is that the "O'Biden-bama administration" worked hard to alienate Florida's Cuban-American community through its flirtation with the Castros.

If Trump loses Florida, it is his own fault.  And if he wins Florida, he wins reelection.

So where does that leave us

A lot will happen between now and November:

  • Will Biden botch the debate, as expected?
  • How soon and permanently will the US pull out of the coronavirus pandemic?
  • Will the dearth of on-campus classes stymie the 2018 Democrat voter registration model?

At this point, I'd have to give the advantage to Trump, assuming he doesn't do something really stupid — like going  after guns in the hope of carrying Broward County.

The big problem for Democrats is that if you lie to gain advantage, the one thing you have to remember — if only in your own mind — is what is the lie and what is the truth.

Michael E. Hammond is the former executive director of the Senate Steering Committee and was often called the "101st Senator" when he worked on Capitol Hill.  Today he serves as the general counsel of Gun Owners of America.

Image: Gage Skidmore via Flickr.