Trump, Covid-19, and the Rally ‘Round The Flag Phenomenon

As the number of confirmed Covid-19 cases goes up, so too does President Trump’s approval rating.  This is not a coincidence. As of today, the RealClearPolitics “poll of polls” has Trump’s approval at 47.3%, the highest since he took office, and just 2% below his 49.3% disapproval number.  That gap is the smallest it has been since Trump’s inauguration. One week ago, the gap was -7.9%.

RCP “Poll of Polls”

For the denizens of Twitter, and for those whose primary source of news is cable talk shows and editorials in the NYC-Washington DC media axis, the upward trend in Trump’s approval may seem baffling, particularly given their steady drumbeat of stories criticizing the administration’s response to the corona virus. But it shouldn’t be. Trump is benefiting from a “rally ‘round the flag” effect  – the same phenomenon that has boosted support for previous presidents in times of national crisis.  First documented by political scientist John Mueller in a study focusing predominantly on Cold War military events. Mueller’s finding has subsequently been confirmed, and developed, in several additional studies that provide a clearer portrait regarding the basis of the rally effect.

The primary source of this phenomenon is rooted in presidents’ relatively unique position in the American political system. At the most basic level, presidents – as the only elected official with a national constituency – are the closest we have to the individual embodiment of national sovereignty.  The impact of that role is heightened by the fact that in the U.S., the President plays a ceremonial function in additional to his (someday her) partisan political position.  As such, when circumstances threaten the nation’s sovereignty, he benefits from his stature as both political head of government and chief of state by becoming the focal point of public concern about events. 

But the impact of rally events is not felt universally across all members of the public. Matthew Baum finds that the most partisan members of the public are the least likely to respond to a rally event. The reason is that they are the most politically aware, and their  opinions are more likely to exhibit greater ideological constraint, which lessens their likelihood of changing attitudes toward the President in response to events.  More moderate, less politically aware voters, in contrast, are more responsive to the rally event.  Interestingly, Baum finds that partisans are more likely to rally behind a president of the opposing party in response to a unifying event.  The reason is that support from this group has more room to grow, since more of them have a lower opinion of the president to start.

William Baker and Jon O’Neal, meanwhile, find that the rally effect is greater when the President actively solicits support for his action through public statements, and if that action receives bipartisan support. So it is not the event itself so much as the President’s framing of that event, and the reaction among other elites to his framing, that seems to drive the size of the rally response.

We can see, then, why Trump’s approval ratings have gone up.  First, he has appeared on an almost daily basis, often in prime time, to give press conferences documenting his administration’s response to the corona virus. Keep in mind that although that response has been harshly criticized on cable shows, only about 12 million people watch these shows regularly. That is far fewer than the almost 140 million who voted in the 2016 presidential election – most of whom do not use op eds or talk shows as their primary source of political information. 

Moreover, Trump has been flanked at these press conferences by non-partisan medical experts, including Tony Fauci, the director of the National Institute of Allergy and Infectious Diseases, and Dr. Deborah Birx, a physician who previously served as President Obama’s U.S. Global Aids coordinator.  While Trump often uses the events to take jabs at political rivals, and to praise his own administration’s response, most of the information that is transmitted by others at these press gatherings centers on specific policies the administration has taken to combat the virus.  Most people are reacting to the information related to how to combat the virus, rather than the partisan frame in which it is discussed on cable television by Trump and his critics.   

Most importantly, Trump’s policies, if not his framing, have attracted support at all levels of government.  The most visible example is the $2.2 trillion economic relief package, which sailed through Congress with overwhelming bipartisan support and was signed into law by Trump yesterday.  Despite Trump’s best efforts to step on his own bipartisan message, for now at least he is reaping the benefits of the administration’s highly visible response to the Covid-19 pandemic, one that in its broad outlines is attracting generally positive reviews nationally, as measured by polls. 

Why does it matter if his approval ratings go up? Studies show that as presidents’ popularity increases, so too does their likelihood of winning reelection. For example, Alan Abramowitz’s incumbent-centered election forecast model estimates that Trump will gain about 2.5 electoral votes for every 1% increase in his approval rating, as of the June before the election.  Other forecast models show a similar positive relationship between a president’s approval ratings and his reelection prospects.

Of course, that assumes that Trump’s higher approval ratings will persist until June.  There is good reason to suspect that won’t be the case. George W. Bush received an initial boost in approval after invading Iraq and removing Saddam Hussein from power.  But as the Iraq war dragged on, and U.S. casualties mounted, his support dropped steadily, as shown here.

Bush Rally Events During His First Six Years

More generally, studies document that most rally effects are short-lived, and barring additional events, presidential approval typically reverts to the pre-event level.  In the event of a sustained rise in the death toll caused by the coronavirus during the next several months, one could envision a similar drop in Trump’s approval, particularly if that leads to a renewal of the partisan polarization among political elites that Trump has confronted through most of his first term in office. This is almost certain to happen as the presidential election campaign comes back into focus, and Joe Biden ramps up his attacks on Trump’s handling of the pandemic.  

Note also that the jump in approval Trump is experiencing, while significant, is not as large (as yet) as what other presidents have experienced after rally events.  For comparison purposes, Paul Brace and Barbara Hinckley find that highly publicized international incidents of short duration, coupled with a presidential statement explaining the U.S. role in that event, can boost a president’s approval rating by about 8%. (Note that their study predates the increase in party sorting since the 1990’s which may dampen the size of a rally event.)

An additional consideration is how the administration’s response to the coronavirus will impact the economy. Most election forecast models include a measure of aggregate economic performance as one of their explanatory variables.  Abramowitz, for example, finds that a 1% drop in GDP can cost incumbents nearly 20 electoral votes.  Should the economy fall into an extended recession, despite the passage of the stimulus bill, it could very well jeopardize Trump’s reelection chances, assuming past performance is a reliable indicator.

Most forecast models don’t kick in until midway through the election year, or later, so it is far too early to make useful predictions. Moreover, the public response to a a pandemic may not be the same as how it historically has responded to a military incident abroad, or domestic terrorism at home. So we must be cautious in predicting its electoral effects. For now, Trump appears to be benefiting from a rally-‘round-the-flag effect, at least as measured by approval ratings.  Whether, and how long, it will last, remains to be seen. 


  1. Trump’s cohort has definitely been defined by reassuring titles. There appear to be, however, two matters that should stand out not only those op-ed informed: The present “task force” members have not been consistent (in terms of exactly who appears and how their messages compare to Trump’s). The Q&A portions have been remarkably rocky.
    Local networks cut off the broadcast at certain points. I wonder if this has coincided with the main speech’s conclusion.

    Also, any thoughts on a Cuomo ascent?

  2. Are the pollsters factoring in any allowances for differences in the respondents to their polls since the shelter in place orders went into effect? For example, many more people may be home and answering land lines now. Do pollsters adjust for those demographics (i.e. it’s not a retiree answering the landline during the day but a millennial working from home due to the pandemic?)

  3. Sam,

    I agree re: the inconsistency of the message. I think part of the explanation is that policymakers, including the experts (Fauci, Birx) are adjusting to the changing numbers on the ground – it’s a fluid situation, and as Fauci has acknowledged from the start, the U.S. response has been hampered by regulations that were not designed to deal with this type of pandemic. (See the CDC rules on testing, for example.) So they’v had to adjust their response as bottlenecks and other obstacles appear. Note also that the press statements are often emphasizing different aspects of the pandemic – one day it is testing, the next production of ventilators, etc. It is a reminder of something I teach in bureaucracy: it is very difficult in a federalist system of this size to run things from the White House. Most of the key decisions will be made on the ground, by state and local officials, dealing with the specific circumstances they are confronting. The president cannot – and probably should not – try to run things from the White House.

    Cuomo is getting a lot of face time right now which certainly enhances his name recognition. (In that vein, a post I wrote after the death of his father four years ago has suddenly been resurrected because a lot of people are mistaking by discussion of Mario as an analysis of Andrew!) So he certainly has high visibility. But note that much of the push behind his candidacy is coming from social media and op ed columns – we shouldn’t blind ourselves to the fact that these aren’t particularly good barometers re: Cuomo’s level of support in the rust belt, for instance. And there are daunting hurdles to overcome logistically – how does he actually win enough delegates, for example, at this late stage to get the nomination? Is he counting on a brokered convention? All in all, I find a draft Cuomo scenario highly unlikely.

  4. Cecilia,

    Great question. The best pollsters are doing their best to weight the surveys responses in a way that accurately mirrors the underlying population. This means adjusting for gender, age, income, education, etc. This is an inexact science of course, which is one reason I recommend relying on multiple trends over time before coming to any conclusion regarding shifts in public opinion. One potential issue is how pollsters deal with partisan identification. Some weight by party ID, to make the survey match what they believe to be the actual distribution of partisan support among the general population. Others do not, arguing that party ID is not a fixed demographic variable, and so they simply adjust for other population variables (age, etc.) and take the partisan distribution that the survey results show. My point is that it is certainly possible that some surveys are simply picking up more Republican-leaning respondents. Interestingly, however, in looking at the cross-tabs of several surveys, it appears that Trump’s approval bump is mostly among Democrats and liberals, which is consistent with what prior research would suggest. That also suggests to me that this bump might not be sustained.

  5. Hey Mike,
    Mainstream Media is Correct: The Orange Man IS bad. And that’s been true for decades regarding this mendacious, corrupt grifter, tax cheat and sexual predator.

    And now The Orange Man’s need for ego stroking and his pathological narcissism has cost thousands of American lives.

    He can never accept accountability for his own mistakes and pathetically poor judgment. Those who continue to desperately cling to him and his destructive policies will soon enough come to see how they were taken to the cleaners by this criminal con man.

Leave a Reply

Your email address will not be published. Required fields are marked *