Overview and Context
In the United States, state and local governments have shouldered most of the responsibility for setting policy in regards to COVID-19. These include closing schools and businesses, social distancing measures, testing requirements, and when to relax these restrictions. Because of the fragmented response, several basic questions become difficult to answer. What restrictions are being put in place? How are they communicated with the public? How do they correlate with changes in mobility and disease transmission? The UC Davis DataLab is building a comprehensive dataset of all official state communications regarding COVID-19 to help answer these questions.
The CovidDocs project aims to collect and catalog all official state communications related to COVID-19, such as executive orders, emergency declarations, public health orders, and guidance documents. These documents are tagged with relevant metadata, such as what restrictions are being called for in these documents. CovidDocs provides data for analyses by DataLab’s data scientists and collaborating UC Davis faculty. The goal of the study is to create a well-documented data set that can inform research into the pandemic and the public health response.
Inclusion Criteria and Metadata
The CovidDocs project aims to collect all of the state-level communications since January 1, 2020---even ones that are not explicitly related to COVID-19---in order to ensure a comprehensive dataset. For example, some of the earliest impacts of the pandemic can be seen in changes to trucking regulations, which in and of themselves don’t explicitly mention coronavirus but indicate socio-economic changes that may be relevant for pandemic researchers. Example documents include executive orders, public health orders, emergency declarations, and other guidance documents. Thus far a total of over 7000 documents have been collected from all 50 states.
For each document, the metadata identifies the state, the date of issue, the effective date, the expiration date, and some indicators of how the document affects widely-used non-pharmaceutical interventions. These interventions include social distancing, limiting gathering sizes, stay at home orders, face coverings, elections, or restrictions on schools, businesses, parks, and recreation. There is also metadata indicating whether each document is explicitly related to the pandemic.
DataLab is working on a first release of the CovidDocs data. This is a living dataset as we will be adding documents as state governments release new communications. While our goal is to provide a contemporary dataset for DataLab’s data scientists and faculty collaborators, we plan to release versions of this archive periodically. These releases will happen once we determine the dataset is complete for a given time period, and the documents have undergone quality checks. We will also provide our DataLab and collaborating UC Davis faculty-led research teams the first opportunity to work with it before a public release.