Abstract

o Since local governments were authorized to send out WEAs in September 2019, they have sent out WEAs on natural disasters and other pending issues in their respective regions.
- The number of WEAs sent to the Seoul Metropolitan City has grown dramatically since the spread of COVID-19. In January 2020, before the full-blown spread of COVID-19, only 10 different messages were sent out, but the number increased five times to 51 in March and 7.7 times to 393 over the previous month.
- Decisions on whether a WEA should be sent out as well as its content are up to the local government concerned. Therefore, the number of outgoing WEAs and their content varied with the local governments.
- We could confirm that the more confirmed cases in a region, the more messages are sent out. However, we also found that some regions reporting similar confirmed cases showed significant differences in the number of WEAs sent out.

o The number of WEAs did cause significant changes in population variability in the regions where they were sent out, and the direction and extent of changes differed depending on age groups.
- It turns out that sending a WEA caused changes in variability for the 20s population and the 60s and 70s population (baseline effect: negative).
- On the other hand, WEAs increased the variability of the 30s-50s population (baseline effect: positive). As this generation represents the most economically-active population and therefore inevitably has to go to work, it can be inferred that population movements of this age group were concentrated during the evening rush hours as private gatherings were canceled, consequently increasing its population variability.
- In the meantime, it is confirmed that the more WEAs get accumulated, the less effect they tend to have across all age groups (cumulative effect: negative). It proves indirectly that under the prolonged pandemic, people started to feel tired of getting repetitive messages with similar content and started to react less.

o Population variability of people in their 60s and 70s population decreased when a WEA included date and order. In contrast, the variability of lower age groups in their 20s and 30s increased when a disaster message included place, symptom, and online information.
- As date and order information is included in messages intended to give information about confirmed cases, these results show that the overall mobility of higher age groups decreased as they reacted to the messages reporting of confirmed cases.
- Messages containing places increased population variability for the age group of the 20s and 30s. It can be inferred that when a natural disaster occurs confirmed cases/cluster infection cases are found in a specific area, people in this age group avoid visiting that area, which causes population variability to increase in its surrounding areas.
- The higher the number of WEAs sent outgrows, the less impact they have on population variability, regardless of types of disaster information.

o Hands-on personnel responsible for sending WEAs in the metropolitan and other local government agencies consider the text messaging process as effective by and large.
- There is a wide range of disasters, including natural and social disasters, and disasters are getting more and more diversified. Nevertheless, their evaluation is positive in that they can collect disaster information quickly and accurately through smooth connections and interaction with related departments, and they can also respond quickly as the approval process is getting more simplified.
- Limitations on the length of text (90 Korean letters) and access restrictions (depending on authorization levels) hinder the complete gathering of disaster information while working as obstacles in the process of re-checking disaster information to enhance accuracy. Therefore, the authors pointed out these restrictions as an improvement opportunity.

o In order to continue to improve and advance the disaster information delivery system, it is necessary to build a feedback system that allows Korean people (information recipients) to send their opinions to information providers as well as to review achievements of the disaster information delivery program.
- However, there are no specific performance indicators to measure what has been achieved by communicating disaster information. Performance is measured only in indirect methods, such as sharing cases by government agencies and media reports, which means that improvement efforts are required in this area.

o When the government advances its WEA system for disaster information in the future, it needs to make a detailed policy review on disaster message blocks.
- There is no data on indicators related to ratios of received and blocked messages due to limitations in the current CBS method. However, we need to find ways of measuring and collecting such data by improving the delivery system as this data will give an idea of how widely disaster information reaches the Korean public.
- If a user blocks the receipt of disaster messages once, the current system allows the block to continue until the user unblocks it. If this feature remains as it is now, it will become an obstacle to communicating disaster information in an unpredicted disaster situation. Therefore, the authorities should look into how this could be improved, such as requiring users to change message blocks periodically to build improvements into the disaster text messaging system.