The Easter Bump

I’m going to start off with my conclusions, then I’ll try to convince you that they’re true.

1. There does appear to have been a small increase in coronavirus spread in Georgia that coincides with the timing of Georgian Easter.  However, it’s important not to read too much into this.

2. If there is an Easter bump, it probably can’t be blamed on the Georgian Orthodox Church – in fact I think it’s more likely that private family gatherings caused the bump as opposed to Easter masses – although I think the rhetoric of the Church towards the virus and the example the Church set were both harmful.

3. The Georgian government’s measures around Easter time probably prevented a much larger Easter bump.  However, the government has not done a good job at communicating its strategy or rationale or decision-making process regarding the Easter situation.

Let’s take these conclusions one by one.

************

1. There does appear to have been a small increase in coronavirus spread in Georgia that coincides with the timing of Georgian Easter.  However it’s important not to read too much into this.

Here’s a graph of coronavirus cases in Georgia from civil.ge (sorry I’ve cut off the y axis, but the graph starts back in February and we don’t need the whole thing).  The green line is total confirmed cases, and the blue line is total confirmed active cases.

The black box is the plateau that we appeared to be on before the Easter bump – a period when the number of daily recoveries approximately equaled the number of daily new infections.  The red box is the period that I am arguing represents an Easter bump – when daily new infections exceeds daily recoveries.  To be exact: April 19th was Georgian Easter.  From April 19th to 24th, the total number of active cases went up by only 3.  From April 24th to May 3rd (that is, five to fourteen days after Easter) that number went up by 52.

So what it looks like is that the suppression measures taken by the Georgian government were working, and the number of active cases was leveling off, and then something happened and the number of active cases took off again.  However, it’s notable that we were specifically looking for an “Easter bump” because of the massive controversy here in Georgia over the dispute between Church and State and this could be random statistical noise, or an increase in testing, or some other event that no one noticed.  While the “Easter bump” is noticeable, it’s also well within the bounds of other increases in infection numbers that don’t have any obvious cause.  We have to be careful not to let our biases and preconceptions influence our perception of the data too much.

************

2. If there is an Easter bump, it probably can’t be blamed on the Georgian Orthodox Church – in fact I think it’s more likely that private family gatherings caused the bump as opposed to Easter masses – although I think the rhetoric of the Church towards the virus and the example the Church set were both harmful.

My main argument here is that the plateau (in the black box) where it looked like coronavirus active infections might be peaking occurred between six and thirteen days after Palm Sunday.  Remember that the controversy over churches refusing to close really came to a head when the churches insisted on holding their Palm Sunday masses in person – that was essentially a preview of Easter.  Palm Sunday mass was held, there was much hand-wringing (including headlines about an alter boy who got infected), and then… nothing.  Now, Easter is bigger than Palm Sunday in a normal year, but that’s because Easter gets even very casual Christians to come to church.  Those casual Christians might be more included to stay home and watch the Easter mass on TV – whereas the Christians who go to Palm Sunday mass are more devout and might be less likely to stay home on Easter.  So it’s at least reasonable to compare the results of Palm Sunday and Easter masses.  And since Palm Sunday mass seems to have had no effect at all, it’s reasonable to guess that Easter mass may have had little or no effect as well.

Easter is not just a time for gathering in churches – it’s also a time for gathering in families.  People travel to their family homes in the countryside and have big family dinners and catch up with cousins and neighbors and whatnot, and then on Monday they visit the cemeteries where their ancestors are buried and pay their respects.  You can tell people to wear a mask and stand six feet apart at a church, but can you tell them that at a family dinner or at a cemetery?

I don’t want to let the Church off the hook, here.  Their refusal to cooperate with the Georgian government undoubtedly undermined containment efforts.  Their claims that God would protect Georgia from coronavirus and that the faithful could not get infected from attending services were irresponsible.  If they had set an example and told people to stay home – like their counterparts in other Orthodox countries – many people would have listened and many fewer risks would have been taken.  However, the evidence I have in front of me is not consistent with the scenario of widespread infection due to attending masses.

************

3. The Georgian government’s measures around Easter time probably prevented a much larger Easter bump.  However, the government has not done a good job at communicating its strategy or rationale or decision-making process regarding the Easter situation.

Here is a photo from Tomas Pueyo’s essay “Coronavirus: The Hammer and the Dance“.  You may have seen the essay.  This table stood out for me:

The takeaway of this table is that if your goal is to eliminate the virus from the population, you don’t have to take every measure available to stop the virus from spreading.  You only have to take enough measures to keep the transmission rate below 1 – in other words, to make sure that each infected person, on average, infects less than one other person.  Keep in mind that elimination is not the only goal – we also want to minimize suffering and death – and that the numbers on this chart are made up for the purposes of illustrating an example, rather than based on empirical data about the impact of various measures on infection rates.  In other words, don’t rely on this chart for creating real world policy or trying to predict real outcomes.

However, what’s important is the concept that this chart demonstrates: that measures taken to reduce transmission rates add up, and can be traded off against each other.  So if you can’t ban large gatherings, perhaps you can implement travel restrictions instead and get about the same amount of benefit.  This, in fact, is exactly what the Georgian government did.  Again, it’s worth mentioning that banning religious gatherings *and* restricting travel might have prevented the Easter bump entirely, or at least reduced it.  However, the travel restrictions and other draconian measures around Easter time at least kept the Easter bump small enough that it’s well within the range of other transmission bumps that don’t have an obvious explanation – in other words, it’s small enough that we might have mistaken it for random noise if we didn’t know that these transmissions occurred during the most restrictive stage of Georgia’s lockdown period.

So the Georgian government was smart to compensate for extra transmissions around Easter by placing additional travel and work restrictions on the populace – however, the fact that I have not seen *anyone* recognize this suggests that the government isn’t doing a good job of justifying its decisions to the public.  Instead, I’ve heard that the restrictions were designed to prevent people from getting to church without actually closing the churches down.  I’ve also seen complaints that allowing congregations to congregate means that the Georgian government is implicitly acknowledging that no social distancing measures are necessary and that therefore coronavirus has been overhyped in order to seize control over the whole society or subdue the population during the run-up to the next election so people can’t protest Georgian Dream.

Governments need all the credibility they can get when they’re asking people to endure economic and social hardship in order to protect society at large from a deadly threat.  The government won’t get compliance with these restrictions if it isn’t transparent enough to reassure the people that it is on their side.  And yet throughout the entire confrontation with the Church, the Georgian government has been evasive and incoherent in a way that has diminished the people’s trust in their intentions and motivations.  This was an unforced error in an otherwise exemplary pandemic response.

************

In the end, neither Church nor State is coming out of the Easter season looking totally, er, immaculate – but then again, it could have been a heck of a lot worse.  Honestly I’m more concerned about the government deciding to start lifting restrictions in the absence of a clear peak and subsequent drop in transmission rates.  It would be a shame if the extreme nature of the Easter lockdown – and the small size of the Easter bump – caused lockdown fatigue and an early release on restrictions, which would inevitably lead to a much larger surge in infections.  Coronavirus isn’t over, and we need to stay the course.