Four surprising data points to consider as we head down the stretch

Image by Nick Youngson CC BY-SA 3.0 Alpha Stock Images

In contrast to my previous posts on predicting elections that demonstrated a strong possibility of a Biden victory, here are a few data nuggets that signal a potential upset.   

1. Battle ground Florida – party registration and the Hispanic vote

The gap between Florida registered Democrats and Republicans narrowed in 2020 – Democrats hold a slight advantage of 134,000 voters.  In 2016, the Democrat advantage was 327,000 and in 2008 it was 658,000. Recall Trump won the state in 2016. 

According to a recent UNF survey, Biden maintains a 7-point edge over Trump among Florida Hispanics – 50% to 43%.  Cuban-American support for Trump is noteworthy, as approximately 59% intend to vote Republican.  Biden receives 25% of the Cuban-American vote.    

Compared to 2016, Trump is over-performing.  For example, he won 35% of Florida Latinos four year ago and his support today improved 8 points to 43%.

Compared to 2016, Trump over-performing in Florida among Latinos

Year	Democrats	Republicans
2016 exit polls	62%	35
2020 UNF survey	50	43

2. African American Support for Trump

Compared to 2016, Trump has gained ground among Black voters.  To be sure, he still lags considerably behind Biden, but the gains could make a difference in battle ground states like Michigan, Pennsylvania, or Wisconsin.  The improvement is especially evident among Blacks age 18-44 and among Black men. 

Compared to 2016, Trump is over-performing among Black voters nationwide

	2016 Gap between Clinton – Trump	2020 Gap between Biden – Trump	Trump Improvement
Overall	– 82	– 71	+ 11
18 – 44 yrs old	– 76	– 57	+ 19
Men	– 72	– 57	+ 15
College Degree	– 85	– 75	+ 10

  

3. Are you better off now than four years ago?

This question taps the idea of retrospective voting, which argues that voters are heavily influenced by economic conditions.  If people feel better off financially today than four years ago, they are likely to support the incumbent.  If they have less in their pockets, then the incumbent party will be punished.   

A chief reason for Bill Clinton’s unexpected victory over George Bush in 1992 was voters’ retrospective assessments of the economy.  The expression “It’s the economy, stupid” was repeated over and over and became the slogan for Clinton’s successful campaign. 

Recently, Gallup showed that American’s views of the economy are remarkably positive.  A majority (54%) of Americans rated Trump positively for his job performance on the economy.  Likewise, a majority (55%) considered themselves and their family better off now than they were four years ago.  In June, 53% rated their personal financial situation as either excellent or good. 

The ‘better off’ figure is the highest recorded at comparable times in an incumbent president’s first term – see table.  Given these percentages, most incumbents win reelection – among the incumbents listed, GHW Bush is the only one that lost, and his economic numbers were especially poor. 

Job approval on economy and are people better off today (%)

Presidents	Job approval of economy	Better off today
Trump	54%	55
Obama	45	45
GW Bush	46	47
Clinton	57	—
GHW Bush	18	38
Reagan	50	44

It’s reasonable to ask why voters would return Biden to office if they feel better off now than when he ended his tenure as VP four years ago.     

Then again, this is not a typical election. We may learn retrospective evaluations are less important than in the past – as the president’s handling of the pandemic overcomes other typical considerations.       

4. Voters expect Trump will win

In some surveys, people are asked, “Regardless of whom you support, and trying to be as objective as possible, who do you think will win the election in November.”  Gallup found that 56% believed Trump would win.   

A comprehensive study examined the accuracy of citizens win expectations compared to opinion polls and quantitative prediction models.  Examining the 1988 – 2012 elections, citizen’s expectations of the winner provided more accurate forecasts than other methods – for details see Polly Vote:  citizen forecasts.  

Gallup also showed that for every presidential election year from 1996 to 2012, citizens expectations accurately predicted the winner of the popular vote – though not Electoral College.     

Studying Gallup’s findings in the table below, two are noteworthy.  First, Independent’s expectations strongly favor Trump – 17-point spread.  Second, compared to Republicans, Democrats exhibit far less confidence in their candidate’s chances – nearly a quarter of Democrats think Trump will win. 

Finally, given the possibility that a certain portion of Trump supporters are unwilling to acknowledge their preferences publicly, some pollsters think the win expectations question could be a more accurate indication of support. 

If so, the election could be closer than polls suggest.    

Bottom Line

No one expected Trump to garner this level of support among Latinos and African Americans.  Perhaps the numbers reflect Joe Biden’s weakness as a candidate as much as Trump’s unusual strength among minority groups.   

Strong ratings on the economy and citizens’ expectations for a Trump victory also run counter to conventional wisdom.  After all, the economy nearly collapsed in the spring, followed by record unemployment and on-going turmoil in the markets.  And polls have consistently showed a Biden victory – even before the pandemic arrived.  Yet, people believe Trump will win. 

The relatively large share of Democrats that foresee a Republican victory may be still dazed from 2016 and bracing for a repeat.  That is of course rank speculation on my part.  And, it does not explain why a significant majority of Independents think Trump will win as well. 

However, clearly, some Democrats are cautious, and perhaps reminded of the famous Yogi Berra quote: “It ain’t over till it’s over”

Without question, the weight of available evidence remains solidly in Biden’s favor.  However, it is important to acknowledge other indicators that are not considered conventional wisdom — and therefore largely ignored.

2016 humbled many.  As the data above show here, there is nothing conventional about this election cycle.