Thread -Pls RT- 🚨New #COVID19 paper🚨
We all know the stories of people✈️to Ischgl⛷️ and returning home🤒with #COVID19
I investigate how the EXACT timing of the winter school-break 2020 led to regional outbreaks, which persist❗️ #vinterferie #sportlov
economics.ku.dk/research/publi…
We all know the stories of people✈️to Ischgl⛷️ and returning home🤒with #COVID19
I investigate how the EXACT timing of the winter school-break 2020 led to regional outbreaks, which persist❗️ #vinterferie #sportlov
economics.ku.dk/research/publi…
Why should the timing of the school-breaks matter? #vinterferie #sportlov✈️⛷️
The number of Covid-19 cases was in the single-digits in early February 2020.
Not until the very end of the month did the number of (confirmed) daily cases skyrocket📈 #Week9
2/n
The number of Covid-19 cases was in the single-digits in early February 2020.
Not until the very end of the month did the number of (confirmed) daily cases skyrocket📈 #Week9
2/n
Many regions had school-breaks in late February that cluster travel ⛷️✈️ by region and time! #vinterferie #sportlov
This creates the potential for large-regional outbreaks!
➡️Multiple independent simultaneous introductions of #COVID19 to the same geographic location❗️❗️
3/n
This creates the potential for large-regional outbreaks!
➡️Multiple independent simultaneous introductions of #COVID19 to the same geographic location❗️❗️
3/n
I use WITHIN-COUNTRY variation in school-break timing to analyze this!
What do I find?
Regions that had a school-break in week 8, 9 or 10 (Feb 17th-March 6th 2020) had on average 6⃣0⃣% higher spread of #Covid19 in spring 2020 than regions with a break in week 7 or before
4/n
What do I find?
Regions that had a school-break in week 8, 9 or 10 (Feb 17th-March 6th 2020) had on average 6⃣0⃣% higher spread of #Covid19 in spring 2020 than regions with a break in week 7 or before
4/n
What about persistence?
Since I run a separate regression per-month we can see the development over time👇
From the fall ➡️same regions systematically worse hit with the spread 30-50% higher❗️
Evidence of strong degree of underlying persistence following large-outbreaks
5/n
Since I run a separate regression per-month we can see the development over time👇
From the fall ➡️same regions systematically worse hit with the spread 30-50% higher❗️
Evidence of strong degree of underlying persistence following large-outbreaks
5/n
Digging deeper!
What do Belgium, Bavaria, Southern-Netherlands and Stockholm have in common?
All had a school-break in week 9 (24 February -1 March 2020) and have been (relatively) heavily impacted by Covid-19!
6/n
What do Belgium, Bavaria, Southern-Netherlands and Stockholm have in common?
All had a school-break in week 9 (24 February -1 March 2020) and have been (relatively) heavily impacted by Covid-19!
6/n
What about specific week numbers then❓
Did it matter if the #vinterferie #sportlov was in week #⃣8, #⃣9 or #⃣10?
Short answer: YES
Week #⃣9 regions were the most heavily affected both in the spring and fall.
➡️90% higher in the spring❗️
➡️70% in the fall❗️
7/n
Did it matter if the #vinterferie #sportlov was in week #⃣8, #⃣9 or #⃣10?
Short answer: YES
Week #⃣9 regions were the most heavily affected both in the spring and fall.
➡️90% higher in the spring❗️
➡️70% in the fall❗️
7/n
Data? #COVID19 case data from around 650 European regions and 14 countries
Testing variation? Cross-country yes➡️relatively small within a country-month pair as used in paper!
The results are also very similar using Covid-19 deaths from (@rki_de) and hospitalizations (@rivm) 👇
Testing variation? Cross-country yes➡️relatively small within a country-month pair as used in paper!
The results are also very similar using Covid-19 deaths from (@rki_de) and hospitalizations (@rivm) 👇
What are the implications?
1. Once a large-outbreak has occurred, it persists in the region even during the summer months when case numbers are very low!
2. Rapid measures to avoid large regional outbreak(s) are therefore fundamental to long-term containment! #COVID19
9/n
1. Once a large-outbreak has occurred, it persists in the region even during the summer months when case numbers are very low!
2. Rapid measures to avoid large regional outbreak(s) are therefore fundamental to long-term containment! #COVID19
9/n
3. Vaccination strategy?
Regions that have recently experienced large outbreaks are more vulnerable to further outbreaks!
Suggests that regional prioritization may be warranted!
10/n
Regions that have recently experienced large outbreaks are more vulnerable to further outbreaks!
Suggests that regional prioritization may be warranted!
10/n
4. New (more contagious) strains?
Results highlight the importance of initial containment measures, especially in regions/countries that are far from reaching #herdimmunity
5. Support for (longer) region-specific containment measures!
11/n
Results highlight the importance of initial containment measures, especially in regions/countries that are far from reaching #herdimmunity
5. Support for (longer) region-specific containment measures!
11/n
6. Paper provides an exogenous regional measure of initial exposure to Covid-19 for a number of European countries!
➡️important to account for when evaluating the effectiveness of containment measures or other policies
#research #COVID19
12/n
➡️important to account for when evaluating the effectiveness of containment measures or other policies
#research #COVID19
12/n
Thanks to @jrmunch, @eikonomics_eiki, and @miriamwuest for helpful feedback.
Shout-out to @AsjadNaqvi for his efforts on harmonizing Covid-19 data!
github.com/asjadnaqvi/COV…
13/n
Shout-out to @AsjadNaqvi for his efforts on harmonizing Covid-19 data!
github.com/asjadnaqvi/COV…
13/n
• • •
Missing some Tweet in this thread? You can try to
force a refresh
