In the final Gallup poll of March, President Trump’s public approval tied his previous high of 49% and improved 5-points from another Gallup survey fielded two weeks before.
Trump quickly highlighted the achievement. He also touted a 60% public approval of his handling of the pandemic. And on twitter, Trump bragged about the lofty television ratings of his daily coronavirus task force briefings.
How should we evaluate the uptick in Trump’s public approval?
For those that like Trump, the rise validates their feelings and reflects solid leadership. For those that dislike him, Trump’s impulsive self- aggrandizement is frustrating, especially in light of the pandemic.
Regardless of whether you like him or not, this post offers a means to evaluate presidential approval.
Generally, observers of political affairs propose two standards to gauge presidential approval.
First, the majority standard.
Majorities confer legitimacy in democratic processes. Approval ratings below 50% diminish a president’s authority, especially in negotiations with the legislature. In general, majority thresholds are common in electoral processes and if reached strengthen elected officials claims of a public mandate.
In Trump’s case, approval remains underwater. More people still disapprove of the President than approve.
Second, many analysts employ the approval ratings of past presidents to evaluate a current president. Trump’s 5-point increase appears trivial compared to the substantial surge in approval of George Bush after 9/11, Bush senior during the Gulf War, Reagan following an attempted assignation, and Carter during the Iran hostage crisis. More recently, after U.S. forces killed Osama Bin Laden, Obama’s approval increased by 8 points.
Though persuasive, the typical benchmarks neglect context. As mentioned in my last brief, we are now in an unprecedented climate of party polarization. Public approval of presidents is low by historical standards and exhibits an unusual consistency.
Across nearly four years of polling data, on average 86% of Republicans approved of Trump while only 8 percent of Democrats. Likewise, during the Obama administration, 81 percent of Democrats and 14% of Republicans approved.
Partisans are resolute in their feelings toward the opposition. Accordingly, majority approval might not be realistic target – or at least the rare event. If true, we should not assume a 5-point boost in Trump’s approval is comparable to a 5-point increase during Ronald Reagan or George Bush’s administrations.
In short, typical fluctuations in approval experienced by Presidents before partisan polarization may now be unattainable. The vastly different political conditions that characterize different presidencies require a standardization of approval ratings.
A z-score converts approval percentages into comparable units – measured in standard deviations. Scores can be negative, positive, or zero. They demonstrate the deviation of a specific presidential approval score from the president’s average approval.
The average rating becomes the benchmark. It signifies the typical performance for a given president in a certain period. A monthly approval scores can be fairly typical, close to the president’s average, or substantially higher or lower. The direction and extent of deviation from the mean provides data to compare presidents and assess changes in presidential approval.
Let’s take a closer look at Trump’s 49% approval.
George Bush saw his approval skyrocket after 9/11. The first Gallup poll in September 2001 reported approval at 51%. The first poll in October, 87 percent, a historic 36% increase.
For the first 10 months of George Bush’s first term, he enjoyed an average approval rating of 59.4. The October 87% approval converted to a z-score of 2.70. In other words, the post-attack approval was highly unusual nearly 3 standard deviations above the typical level for Bush.
Finally, Trump’s 49% approval translated to a z-score that matched Bush’s 9/11 boost. Compared to Trump’s average approval of 40.8, 49% is unusually high, just as 87% was exceptionally high for Bush. Indeed, 49% is a genuine outlier, nearly 3 standard deviations above Trump’s typical rating.
Across the first term – so far
To broaden the focus, let’s compare Bush to Trump using the first 3 years and 3 months of their first terms.[i] The benchmark average increases for Bush – from 59.4 to 64.7, considering 39 months of approval scores – as opposed to 10 months as calculated for analysis above.[ii]
The graph below plots z-scores across the number of months in office. The peaks draw attention. For Bush, 9/11: For Trump, the SARS-COV-2 pandemic. Of all the month-to-month changes, for both Presidents, Trump’s latest approval ratings are the most unlikely – exhibiting the largest z-scores.
All president’s desire high public approval. Like Bush after 9/11, Trump’s recent approval scores will make it difficult to maintain above average public support. After 9/11, Bush sustained higher than average approval for 14 months. That would be a remarkable achievement for Trump. Trump’s monthly approval floats above, then below, the average, never demonstrating a genuine consistency. It seems unlikely that Trump can sustain – or even increase – his current approval. However, the twin health and economic crises make predictions very uncertain.
Does Trump have a reason to celebrate? Yes. His very low approval average (40.8%), combined with modest month-to-month variation in approval scores (standard deviation = 2.9) suggests 49% is a big deal. In Trump’s ultra-partisan world, 49% is an achievement, as unusual as 87% approval for Bush nearly 2 decades ago.
However, there is a vast difference between 87% approval and 49 percent. Bush immediately exploited the extraordinarily high ratings, winning Congressional passage of several landmark bills including the Patriot and Homeland Security Acts. He was also able to maintain exceptional approval levels for some time.
At 49%, Trump cannot hope to do the same. Democrats remain steadfast. Half the country disapproves of Trump and that promises prolonged legislative battles and partisan discord.
[i] Approval ratings extracted from Gallup archives. I used the first poll of each month to calculate averages and measures of dispersion. The first Gallup poll in March 2020 was 44% – which I included in this analysis. In a later March poll, Trump’s approval rose to 49%. To stay consistent with the methodology, I did not include that high point in the analyses. Rather, I used the 49% high Trump received in February to calculate z-score and compare to Bush.
[ii] The higher average approval for Bush will undoubtedly lower the z-score for 9/11 approval. It does so by about .6.