AMAC Exclusive – By Andrew Abbott
The insular world of election forecasting received a shock earlier this month when ABC News announced that it had hired up-and-coming pollster G. Elliott Morris to replace FiveThirtyEight founder Nate Silver – once considered a “wunderkind” of the polling industry. However, as the history of election polling shows, changing the pollsters will likely do little to change the baked-in biases of the coastal elite media class.
Officially, Disney, which owns ABC News, chalked Silver’s ouster up as part of a broader push to cut 7,000 jobs across the company. Yet the subsequent hiring of Morris suggests that executives at ABC may have simply grown weary of Silver’s poor predictions in recent election cycles.
Originally a baseball data analyst, Silver first rose to prominence in the world of political polling after he accurately called the elections in 2008 and 2010. His popularity exploded in 2012 after he correctly predicted the winner in every state and the District of Columbia in the 2012 presidential election.
As Silver rose in popularity, so too did the belief that the tech disruptors revolutionizing the world would soon turn elections into a game of data science.
So prevalent was this belief that Robby Mook, Hillary Clinton’s 2016 campaign manager, structured the Clinton campaign around the principles of data science. Campaign resources were deployed according to what the polling dictated would be their most effective use.
Of course, Silver’s models would fall hopelessly flat in 2016. After a modestly successful cycle in 2018, Silver struggled again in 2020 and 2022. Finally, ABC decided that Silver’s cost had exceeded his value.
Many pundits identify Hillary Clinton’s surprise loss in 2016 as a watershed moment for the polling industry. After virtually every pollster (save a few like the Trafalgar Group) predicted an easy Clinton victory, many Americans lost trust in election polling entirely.
However, 2016 was hardly the only case in history where the pollsters got it hopelessly wrong.
Aside from 2016, arguably the most famous failure occurred in 1948, when polls all predicted that New York Governor Thomas Dewey would handily defeat incumbent President Harry Truman, with most having Dewey up by 10 or more points. When the final tally was counted, however, Truman defeated Dewey by over two million votes. The iconic image of a victorious President Truman holding a newspaper with the headline “Dewey Defeats Truman” has haunted pollsters ever since.
Polling apologists dismiss these events as flukes and maintain that polling is an exact science. The data is not flawed, they insist, merely the individuals interpreting it.
Yet 1948 and 2016 are more the norm than most in the media class would like to admit. Another infamous polling flop occurred in 1980 when polls showed incumbent president Jimmy Carter as a clear favorite against his challenger, California Governor Ronald Reagan.
In October 1980, most polls had Reagan trailing Carter by as many as eight points. Yet Reagan would go on to win a landslide victory, trouncing Carter by nearly 10 percent.
Overall, of the last ten Democrat presidential primaries in which a Democrat incumbent was not on the ticket, only four of the frontrunners in January of an election year became the party’s nominee.
After 2016, polling experts assured Americans that the flukes had been worked out of the system and that the polls would hold true in future elections. In an interview with NPR in October 2020, polling expert Ashley Kirzinger stated, “There’s a lot of incredibly smart people trying to do this right… Nobody wants to get this wrong.”
Nevertheless, the pre-election polling of 2020 was the most inaccurate in 40 years. Despite most polls accurately predicting Joe Biden would become President, a post-election analysis conducted by several prominent political scientists found that “polling during the two weeks before the election overstated support for then-Democratic nominee Joe Biden by 3.9 percentage points, which was the largest polling error since 1980.”
Following 2020, some pollsters genuinely threw up their hands and said it was “impossible” to know why the polls were wrong yet again.
Ironically, Nate Silver, who himself has become a victim of hopelessly flawed poll results, has provided perhaps one of the most insightful analyses of just why pollsters get it so wrong so often. In a 2017 interview, he noted, “People looking at the polls are mostly in newsrooms in Washington and Boston and New York. These are liberal cities, and so people tend to see evidence (in our view, it was kind of conflicting polling data) as pointing toward a certain thing… People weren’t using the more thoughtful sides of their brains; they were using the more emotional sides of their brains.”
For as long as polling operations and the pundits who editorialize them are the purview of coastal elites, the shortcomings of polling will likely persist. And as long as those shortcomings persist, the public at large will continue to trust the media and their polling allies less and less.
Andrew Abbott is the pen name of a writer and public affairs consultant with over a decade of experience in DC at the intersection of politics and culture.
Read the full article here