Days after attending a European Union leadership summit on Dec. 10 and 11, French President Emmanuel Macron developed coronavirus disease symptoms — quickly sending many of Europe’s leaders into quarantine.

An estimated two dozen or more sitting global leaders have acknowledged contracting the novel coronavirus, including the president of Brazil and the prime ministers of Russia and Britain. Also on the infected leaders list is former president Donald Trump. How did the virus affect their popularity?

FiveThirtyEight’s Nathaniel Rakich raised this question before the November U.S. elections. All leaders want to be popular. U.S. public opinion scholar James Stimson surmised, “If the real power of the presidency is not directly proportional to the most recent Gallup popularity rating, it is not far from it.”

During the pandemic’s first wave last March and April, many world leaders saw spikes in their approval ratings. These “rally-’round-the flag” effects stem from heightened patriotism in the face of national threat.

A sickened leader could represent such a threat. Do leaders who contract the virus receive a “personal” popularity rally?

How we did our research

Researchers at the Pollitik public opinion lab at Georgia State University analyzed the net approval (approval minus disapproval) of five incumbent executives who contracted the novel coronavirus. For data, we use weekly estimates from the Executive Approval Project and public health outputs from Our World in Data. We find that personal leadership rallies — a bump in popularity after covid-19 diagnosis — are, at best, modest. And factors that might explain this common outcome fall short, as they vary across our cases.

Here’s what we found:

Boris Johnson, British prime minister

Johnson became ill in March, as cases in Britain began a steep rise. His March 27 infection appears to have produced a personal covid-19 rally of 16 to 18 points. Our graphic shows that the rapid growth of new covid cases and deaths throughout March 2020 boosted Johnson’s popularity — a general covid-19 rally.

It’s not clear how much Johnson’s illness and hospitalization contributed to a personal covid-19 rally, however. Several events — emergency funding for the National Health Service, Prince Charles’s testing positive for the virus, the Clap for Our Carers campaign — boosted a sense of patriotism and national threat, rallying Britons well before Johnson’s diagnosis. Indeed, YouGov’s Adam McDonnell finds many cabinet members who did not contract the coronavirus also enjoyed rallies, some stronger than Johnson’s.

Most likely a broader covid-19 rally felt across the British political system dwarfed Johnson’s personal covid rally.

Mikhail Mishustin, prime minister of Russia

Mishustin replaced Dmitry Medvedev as prime minister in January 2020, two weeks before Russia’s first diagnosed case of covid-19. Russian President Vladimir Putin tasked Mishustin with the government’s pandemic response.

Mishustin’s net approval ticked up two points immediately after his positive test on April 30. Statistical modeling confirms this small personal rally was ephemeral, if it existed at all.

Jair Bolsonaro, president of Brazil

Unlike many leaders, Bolsonaro’s popularity nosedived with the arrival of the pandemic. By stubbornly downplaying the seriousness of the virus, he may have removed the key condition for the public to rally around the flag: a stark national threat.

But did Bolsonaro’s three positive coronavirus tests — July 7, 16 and 21 — induce a personal covid rally? According to our graphic, Bolsonaro’s popularity continued its upward trend the week after he contracted the virus. From there, it tapered slightly before resuming its upswing after Bolsonaro tested negative. Both the bump and the slump were minuscule.

Donald Trump, former U.S. president

Trump’s net approval in early 2020 was fairly typical of that throughout his presidency. In April, as Americans grew more concerned about spiking infection numbers and covid-19 death counts, Trump’s net approval rate rose to highs not seen since he took office in January 2017. In comparison with his counterparts around the globe, however, Trump’s general covid-19 rally was weak.

Trump tweeted Oct. 2 that he had contracted the virus. Our figure shows that weekly estimates of his net approval barely flinched in the aftermath. In this, Trump is no different from other leaders. The next month, he lost his reelection bid.

Emmanuel Macron, president of France

After his country beat back the first wave of the coronavirus earlier in the year, Macron tested positive for the virus on Dec. 17, during France’s second wave. It’s too early to say with confidence how the diagnosis will affect Macron’s popularity.

However, daily Morning Consult readings indicate Macron may see a modest personal covid bounce. Between Dec. 16 and Jan. 3, the French president’s disapproval fell from 62 to 55 percent and approval rose from 31 to 36 percent. This matches the minimal personal covid rallies we found for Macron’s counterparts.

So why would a covid-19 diagnosis lead to only limited personal covid rallies?

The comparison of what happened in these five cases suggests weak to nonexistent personal covid rallies are the norm — we don’t see evidence of shared explanations among these five cases. We dismiss institutional factors that vary across our cases, such as degrees of democracy and state centralization; electoral and party systems; constitutional structure and the executive’s role.

But consider this explanation: Highly polarized societies and politics may dull citizen responsiveness to leaders’ actions. Populist rhetoric (featuring anti-elitism, people-centrism, and a Manichaean worldview) exacerbates a polarized public opinion, galvanizing leaders’ approval among their followers and hardening opponents’ disapproval.

If our cases fit this narrative, the scope of our conclusions is fairly restricted. To investigate, we compare our cases across indicators of societal polarization, political polarization and party populist rhetoric from the V-Dem (v. 10) and V-Party (v. 1) data sets.

The United States and Brazil are highly polarized, while Britain, Russia and France are not. This suggests societal and political polarization alone cannot account for the unimpressive personal covid rallies in these five cases.

We also rule out populist rhetoric. Mishustin, a technocrat, is not a member of Putin’s United Russia party, whose populist rhetoric is on par with that of Johnson’s Conservatives. And Macron’s La République En Marche! scores significantly lower on the rhetoric scale than all of these other leaders’ parties.

These five leaders received minimal approval rallies from contracting covid-19, and it’s likely an essential feature of the experience. If more global leaders test positive, we may learn more about this curious result.

Ryan E. Carlin is professor of political science, core member of the Executive Approval Project, and director of the Pollitik public opinion lab at Georgia State University. Students involved in the Pollitik lab (see full list here) helped research and write this article. Follow this research on Twitter @Pollitik_Lab and @ExecApproval.