2020 US Elections: Handicapping the race--polls, registration and the shy Trump voter

• Biden leads in the RealClearPolitics nationwide average but gap has narrowed.
• In 2016 Clinton's lead was 6.4%, Biden's is 8.6%.
• In swing states Trump is +0.4% over 2016 performance.
• Non-polling metrics and surveys show a much tighter race.
• Markets have no clear favorite, though Wall Street contributions favor Biden.

In the final two weeks of the US Presidential campaign, polling indicates that Democrat Joe Biden has a commanding lead, but neither of the campaigns is acting like the race is over and there are excellent reasons to think the decision point lies ahead.

The RealClearPolitics (RCP) average of national polls gave Joe Biden an 8.6% lead over President Donald Trump on October 20^th. That is down 1.7% over last nine days from his 10.3% edge on October 11^th. In 2016 Hillary Clinton had widened her lead from 6.0% to 6.4% over the same period.

Local polls, particularly in the six swing states that will decide the election are far more indicative of the election result. In those states, Pennsylvania, Michigan, Wisconsin, Florida, North Carolina and Arizona, all of which went for Trump last time, the polling is much tighter than the national average. In these states Biden is leading by 4.0% and Clinton was ahead by 4.4% two weeks before the election.

Electoral College and the popular vote

American elections are decided by voting in each state for the members of the electoral college equal to the number of senators and representatives, of which 270 are required to win the presidency.

In 2016 Donald Trump won the Electoral Collage vote 304 to 227 while losing the popular vote 46.1% to 48.2%, 62,984,828 to 65,853,514. Clinton's victory in the popular vote in California, 4,269,978, was greater than her margin in the rest of the states combined, 2,868,686. 

Polling

Polling is a difficult art. Like all modeling it depends on assumptions to carry the data forward to its goal of accurately representing a vastly larger population than the recorded sample. The two most important considerations for a national elections are the population chosen, adults, registered voters or likely voters and the skew of the sample, that is what percentage of Republicans, Democrats and independents to include in the representation.

Most pollsters began months ago with the widest audience, adults, moved to registered voters in the summer and now are using a likely voter screen. That opens another area for assumption. How to determine who is going to vote, who is that likely voter? Each firm has it own formula.

Gallup, the oldest US polling firm, tracks party affiliation. Official party registrations tend to change very little but what people tell pollsters is their party identification is quite volatile.

In October 2016 Democrats held a five-point lead in affiliation. In June of this year the lead was seven points, 33% to 26% with independents at 38%. By mid-September that had become a one-point advantage for the Republicans, 28% to 27% with independents leading at 42%. Does this predict voting. No? Pollsters use these figures to help determine the makeup of their likely voter pool. But the swing in the number of people who consider themselves Republicans, in the midst of a volatile campaign, is at least indicative of an electorate that makes up its own mind.

The most crucial choice for a survey is what percentage of Democrats and Republicans to include in the study. Nationally, registered Democrats outnumber Republicans but just taking those numbers is deceptive.

A poll can be made to read almost any result by changing the percentages of each party represented, oversampling one party or the other. Many polls are conducted by organizations with political connections. The best way to judge the accuracy of a polling firm is to look at its record in 2016.

The two most accurate firms doing national polling that year were Trafalgar Group, which was the only firm that predicted a Trump victory, and Rasmussen.

Alternative polling, young voters, shy voters

In a poll conducted from September 14-28 of registered voters, Gallup asked people if they were better off than they were four years ago. In response 56% said that they were “better off”, 32% they were worse off. That score tops the percentages for Barack Obama (45%) George Bush (47%), George H.W. Bush (38%) and Ronald Reagan (44%) at the end of their first terms. Except Bush senior, all won reelection.

Young voters who trend heavily liberal and Democratic are far less interested in participating this year according to Hill-Harris X polling.

The phenomenon of shy Trump voters, that is poll respondents who say they are undecided or voting for Biden when they really intend to vote for the president, is by its nature difficult to quantify. Nonetheless it certainly accounted for some of the rampant inaccuracy of the polling in 2016.

How large this effect may be was considered by a Harvard/Harris survey which asked registered voters who they believed their neighbors were voting for. A plurality of 43% said Trump, 36% said Biden and 21% said they were unsure. The same September poll showed Biden leading Trump 47% to 45%

Trump approval

The latest RCP average for President Trump's job approval, including seven polls over the last three weeks, is -9%, 44.3% approve and 53.3% disapprove. Historically, approval rating at 47% or higher give a high probability for reelection. The spread on the surveys is wide, from -15 in an Economist/YouGov poll to -1% in the Rasmussen.

The same caution about the sampling makeup as in the election polls applies here.

Voter registration

Republicans have been far more energetic in registering voters in the crucial states of Florida, North Carolina and Pennsylvania and have been going door to door for months. The Biden campaign did not have a registration effort at all until about one month ago.

According to a New York Times article Republicans have reduced the 330,000 Democratic registrant edge of 2016 in Florida to just 134,000 out of 14.4 million voters. In North Carolina total registered Democrats have fallen by 6.2% while Republicans have increased 3.5%. In Pennsylvania Republicans have gained 134,000 voters over 2016, though there were still almost 800,000 more registered Democrats.

The significance of the gains is underlined by the narrow margin of Trump's victory in Florida, 112,911 votes and in Pennsylvania, 44,292. North Carolina was not as close though the edge there was but 173,315.

See 2020 US Elections: Handicapping the Presidential Race

Campaign enthusiasm and effort

The Biden campaign has elected to keep their candidate under wraps. His appearances have been limited, closely scripted and totally lacking in the large venues and crowds preferred by the Trump campaign.

As this goes to post Biden in the midst of a four-day hiatus, having called off all events from Monday until after the debate on Thursday evening. Biden's organizers have called their so-called 'lids', the end of campaigning for the day, on over a quarter of the days in the past month.

It is difficult to gauge the effect Biden's absentee campaign will have on voters and the amount of enthusiasm for a candidate who does so few personal outings. No American presidential candidate has ever tried to wage an election bid almost completely through the media.

The contrast to the Trump campaign is striking. The president revels in crowds, large raucous events, boat and car parades, unscripted speeches and flaming rhetoric. Whether the rallies will persuade enough voters to back him a second time is unknown but the energy commitment and enthusiasm of his supporters cannot be doubted.

Biden was not the choice of Democratic voters. His campaign was an also-ran having won no primaries until the Democratic establishment settled on him before the South Carolina contest as the alternative to socialist Bernie Sanders. His running mate, California Senator Kamala Harris, did not even make to a primary. She withdrew before the first vote despite having been the front-runner in early considerations.

Democrats will need the enthusiasm and participation of more than just their base to provide a victory to the Biden-Harris ticket. Can the Democrats motivate their less committed voters with a virtual campaign? It is doubtful.  What is certain is that the Biden campaign is not even trying to stir up excitement.

Markets

Wall Street contributions have favored the Biden campaign almost five-to-one over Trump in the third quarter. According to the Center for Responsive Politics, the financial industry has given more than $50 million to back the Democratic candidate compared with about $10 million for Trump.

In one sense the payments are proof of Trump's outsider status to both the Washington and New York establishments. But the strongest factor behind the preference are the odds of a Biden victory. More than anything else Wall Street wants to back the winner and from the numbers, that seems certain to be Biden.

If Trump defies the polls again, financial executives will have foregone any extra influence their contributions might have purchased, but their main businesses, the economy and markets, will prosper nonetheless.

Whether Trump or Biden is in the White House come January, reviving the economy will be the main priority. Stimulus will be the first item on the agenda.

Conclusion

The imponderables loom large in this election. Mr. Trump's victory in 2016 demonstrated the fallibility of polling.

Have the models been corrected to repair the clear deficiencies?

It is notable that the only major poll to correctly predict the outcome in 2016, Trafalgar Group, is the one outfit that makes a concerted effort to account for the “shy Trump vote.” According to their lead pollster Robert Cahaly, the reluctance of people to tell the truth about their preferences for fear of being judged disapprovingly is especially common among Trump voters. Trafalgar expects this vote to be a huge factor this year.

Is it plausible that people would lie to pollsters? The answer for 2016 was clearly yes. Are the evident differences between the enthusiastic and participatory Trump campaign and the Biden virtual effort going to play out at the voting booth on November 3^rd? Will young voters be motivated to make the the effort for a campaign and a candidate that has done almost nothing to woo their commitment?

President Trump has an advantage in the Electoral College and the Republican registration effort can only have augmented that edge.

But perhaps the biggest question is much simpler. An election is a choice. That on offer is between Donald Trump and Joe Biden. Will the voters, many of them Democrats, who gave the victory to the upstart Donald Trump in 2016, desert his cause for Joe Biden and Kamala Harris?

The polls are tightening much as they did in 2016. The race is far from over.

 

Please see our other campaign coverage:

2020 Elections: Trump is showing signs of a comeback, will the dollar follow?

2020 Elections: Seven reasons why this is not 2016, time to focus on the Senate.

2020 Elections: Ginsburg and the Supreme Court

2020 Elections: How stocks, gold, and the dollar could move in four scenarios, nightmare one included.