Publication of KISDI basic research (21-01) ‘Empirical analysis of data utilization in a disaster situation’.

“Changes in the volume of people movement in areas where disaster text messages are sent”

.................................................................................................................

The results of an empirical analysis of the effect that disaster text messages have on population mobility, with population mobility shown to decrease in the case of people in their 20s, 60s and 70s.

.................................................................................................................

“The mobility of the elderly changed in response to news about confirmed cases in their residential districts, whereas the mobility of young people changed in response to information about symptoms, locations, and online information.”

Ways of improving and refining the disaster information delivery system:

▲ Establish a feedback system that sends the opinions of disaster information recipients to information senders.

▲ Develop detailed performance indicators for inspecting the performance of disaster information delivery systems.

▲ Periodically initialize the settings of disaster text message blocking.

KISDI (President Kwon Ho-yeol) recently published KISDI’s basic research (21-01) report: ‘Empirical analysis of data utilization in a disaster situation’.

Amid the COVID-19 pandemic, the South Korean government obtained the information it needed to respond to the disaster from collected public data and quickly delivered it to the general public via numerous disaster information delivery channels. To establish an effective and inclusive disaster relief plan at a time when disasters are becoming increasingly frequent and diverse, it is critical to check whether disaster information is being delivered to the public in an effective manner. Furthermore, to better prepare for disasters that could happen in the future, the entire process, ranging from the collection to the broadcasting of disaster information, must be reviewed. In addition, new means of collecting and delivering the information needed for the handling of disasters must be explored.

The report presents an empirical analysis of the effectiveness of providing information using text messages during the COVID-19 pandemic, which is a typical disaster situation. Text messages are one of the most efficient means of delivering disaster information, and will continue to be a pivotal means of delivering disaster information in future disaster situations. Since most text messages contain warnings to abstain from gatherings and other social distancing guidelines, their effectiveness was measured by observing the changes in population mobility in the area(s) where disaster text messages were sent. In addition, the information delivery effectiveness of text messages was measured for each type of information after first classifying the type of disaster information contained in disaster text messages. Furthermore, KISDI contacted employees of metropolitan and local governments who are in charge of sending out text messages in disaster situations and listened to their opinions about the overall process of transmitting disaster text messages and the feedback process. Next, KISDI suggested policies for improving the efficiency of the disaster information delivery process. The key findings of the study are summarized as follows:

First, since municipal governments were authorized to send out disaster text messages in September 2019, they have been sending out text messages concerning natural disasters and other pending regional issues. Since the start of the COVID-19 pandemic, the number of disaster text messages sent by the Seoul Metropolitan Government has increased rapidly. Municipal governments can decide at their discretion whether to send disaster text messages and choose the contents accordingly; and, consequently, the number of text messages sent and the contents of such messages vary from one municipal government to another. There was a clear pattern of more text messages being sent in regions with a lot of confirmed Covid cases. However, due to the diverging standards adopted by different municipal governments, two nearby regions could have very different numbers of sent messages even if they have similar numbers of confirmed cases.

Second, disaster text messages caused statistically significant changes to population mobility in the area where the messages were sent. The direction and size of the changes differed by age group. The sending of disaster text messages reduced (- basic effect) the population mobility of those in their 20s, 60s and 70s in the region concerned. On the other hand, the sending of disaster text messages increased (+ basic effect) the population mobility of those in the 30-50 age group. This could be explained by the fact that as those in the 30~50 age group are the segment of the population mostly heavily engaged in economic activities and also caught in the unfortunate circumstance of having to go to work in the morning as breadwinners, they choose to skip after-work gatherings and go straight home, causing a peak in population mobility at around 6PM time. Meanwhile, the longer text messages are accumulated in a person’s smartphone, the more the impact of text messages on population mobility decreases (- cumulative effect) for all age groups. This is evidence of the fatigue effect; in other words, too many similar disaster text messages received repeatedly during a seemingly never ending pandemic lead people to become complacent and disinterested in the messages.

Third, it was discovered that there are differences among age groups in the kind of disaster information that changes population mobility. For those in the 60-70 age group, population mobility was affected when the date and case number information were included in the text message sent in the event of a confirmed case of Covid-19 in their district. For those in the 20-30 age group, population mobility was affected when the location, symptoms and online information were included in the text message. In all types of disaster information, as the number of disaster text messages sent increased, the impact on population mobility tended to decline.

Fourth, the text message managers of local government offices viewed the disaster text message sending process to be very efficient because disaster information could be collected quickly and accurately through seamless communication with the related government departments. However, they said that the imposition of a 90-character limit on text messages and the restriction of information access to everyone except those managers involved undermined the integrity of disaster information collection. They suggested that the system needs to be improved because these limitations act as impediments when information needs to be reconfirmed to improve the accuracy of disaster information.

Lastly, the study confirmed the existence of demand for policy decisions on the following issues: development of a feedback system for improving or refining the disaster information delivery system, in which the opinions of disaster information recipients are communicated to information suppliers; development of performance indicators for evaluating the efficiency of disaster information delivery; measurement of the information reception rate and information block-out rate; and revision of the system for blocking out disaster text messages.

Assistant research fellow Yoon Sung-wook said, “This study empirically analyzes what kinds of changes in citizens’ movement patterns were induced by disaster text messages that deliver disaster-related information to the public via their smart phones. The study also sought answers to the question of which kinds of information were the causes of statistically significant changes in population mobility.” He added, “Since the kinds of disaster information that cause changes in movement patterns differ by age group, the composition of the population in disaster-stricken areas should be considered when composing disaster information to be sent as a disaster text message.”