If
the US presidential election were held today polls suggest Joe Biden would be
elected. Not only would it be a landslide popular vote victory but he would
win a clear electoral college decision. Right now the Economist gives Joe Biden a 93%
chance of winning.
But the election is not today. It is still nearly four months out.
Surveys or polls are not predictions, they are snapshots in time
and are no guarantee of the future. Additionally, national polls are
worthless when it comes to presidential elections. It is not that they
are usually wrong—they did predict Hillary Clinton would win by about two
percentage points in 2016 and she did. But instead the road to the White
House is through the electoral college and the only polls that really matter if
at all are the ones in the critical swing states they will decide the
election. What happens in Arizona, Florida, Michigan, North Carolina,
Pennsylvania, and Wisconsin is a lot more important than what happens in
California, New York, or Alabama. Despite
how much Democrats are salivating, realistically Georgia and Texas, while
competitive, are not swing states. At best, forcing Trump to defend these states
means less resources for the states really in play. For Democrats, don’t repeat the mistakes of
2016 and campaign where there is no realistic prospect of winning.
A lot can happen between now and election days, with it then
either being November 3, 2020 (actual election day) or in the case of some
states such as Minnesota, when as early as September 19, 2020 early
voting takes place. Approximately
38 states allow some absentee or early voting, complicating predictions because
ballots are being cast over a period of many days or weeks.
Political scientists develop models to predict presidential
elections. Central to the models are presidential approval and national economic
performance several months before the election. Based on these variables
Donald Trump should lose. But these models too are flawed in that they overlook
the finery of the swing states and the electoral college and they also seem to
ignore the fact that campaigns really matter. Hillary Clinton
should have won four years ago but ran a lousy campaign. In
1988 Michael Dukakis at one time had a 17-point lead over George Bush but lost
because of bad campaigning, complacency, and race-baiting (remember
Willie Horton) by the latter.
The point is that four months until November 3, is an
eternity. Four months ago, was mid-March, just before the pandemic
kicked in. The US economy was at 4.4% unemployment for March, up from 3.5% in
February. Joe Biden was struggling to shake off Bernie Sanders for the
Democratic Party nomination and it still looked like Trump was favored to win
re-election. The major issue—beyond Trump himself—was the fallout from
the Senate trial and failed impeachment.
Since then everything and nothing has changed. By
that, this election was always going to be about Trump. Unlike in 2016
where the election was a referendum on Hillary Clinton and she lost because
people did not like her, in 2018 and now in 2020 the election is a referendum
on Trump. The pandemic and the economy have and have not changed
that. Yes, they have changed the election in that they are
now issues and voters are judging Trump on how he is handling both. Evidence
suggests his mishandling of them are impacting some support among his base. But
they have changed nothing in the sense they both issues are merely surrogates
for how voters think or feel about Trump.
Think about it. It has taken double-digit
unemployment, a record crash of the economy, and 3,500,000 infections and 138,000 deaths to
change the political dynamics of the 2020 presidential election. Without
the pandemic and the collateral damage, it has impacted on the economy, Biden
might have had little chance of winning. Even now, while the models say
Trump will certainly lose, variables such as changes in the economy, a
lull in the pandemic, a horrible Biden campaign, or a ramped-up on steroids
racial appeal by Trump that makes Willie Horton look tame. And Democratic
Party control of the Senate is not certain.
It takes a lot to defeat a sitting president. Bush
lost in 1992 because of a three-way split in the vote with Ross Perot running
as a strong third-party candidate. Prior to that, Jimmy Carter in 1980
lost to Ronald Reagan and in 1976 Carter defeated Gerald Ford. In both
those cases sitting incumbents lost because of extraordinary circumstances
(1980 it was the Iran Hostage crisis and oil embargo and in 1976 the
Watergate backlash the pardoning of Richard Nixon). Prior to that it was in
1932 the last time a sitting presidential lost in a general election and that
was when the Great Depression brought down Herbert Hoover to Franklin
Roosevelt. Only five incumbents have failed to win a second
term.
Trump is highly vulnerable but this is an extraordinary election.
Partisanship or polarization is so high it is, as noted, taking a national
emergency a catastrophe to even begin to melt his base. Voting in the
age of Covid-19 might produce distorted results that could change the
election. Already the litigation and court fights over early voting or
voting rights portend how fragile franchise rights are and how the Supreme Court may impact how the 2020
elections are held.
If the 2020 presidential election were held today arguably Joe Biden would
win. But it is not being held today and if history tells us
anything, a lot can still change, especially at the electoral college level,
between now and November 3.
