COVID-19 as Reflected in University President Bulk Email

TL;DR

This study analyzes how internal crisis communication via central bulk emails from university presidents evolved during the COVID-19 pandemic. Using 198 emails from eight U.S. universities across four stages (Baseline, Early Covid, Covid Crisis, Covid Era), it defines metrics for information quantity/clarity, positivity, collectiveness, personal humanity, leadership centralization, and government references, while accounting for inter-university variability. The findings show that during crises emails become more informative, clearer, and more positively framed in closing; they adopt more personal and collective language, reference local government more, and reduce mentions of other university leaders. The work contributes a public dataset and provides evidence that internal crisis communications align with established crisis-communication strategies, informing the design of digital tools to automatically identify crisis moments and guide leadership in bulk-email communications.

Abstract

E-mail ``Messages From the President'' to university students, staff, and faculty have long been used to keep campus communities aware of the latest policies, events, and news. But during the COVID-19 pandemic, as universities quickly closed facilities, sent students home, and canceled travel, these messages took on even greater importance. We report on a content analysis of bulk emails from different universities' presidents to their students and employees before and in three stages of the pandemic. We find that these messages change as universities move towards and through closure. During the pandemic, 1) presidential bulk emails tend to be more informative, positive, clearer than before; 2) they tend to use more personal and collective language; 3) university presidents tend to mention more local political leaders and fewer other university leaders. Our results can inform research on digital crisis communication and may be useful for researchers interested in automatically identifying crisis situations from communication streams.