CoronaNet: COVID-19 Government Response Event Dataset

Principal Investigator(s): View help for Principal Investigator(s) Cindy Cheng, Hochschule für Politik at the TU Munich; Joan Barceló, NYU Abu Dhabi; Allison Hartnett, University of Southern California; Robert Kubinec, NYU Abu Dhabi; Luca Messerschmidt, Hochschule für Politik at the TU Munich

Version: View help for Version V1

Name	File Type	Size	Last Modified
coronanet_release.csv	text/csv	19.2 MB	07/25/2020 04:05:AM

Project Citation:

Cheng, Cindy, Barceló, Joan, Hartnett, Allison, Kubinec, Robert, and Messerschmidt, Luca. CoronaNet: COVID-19 Government Response Event Dataset. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2020-07-25. https://doi.org/10.3886/E120342V1

Project Description

Summary:

The CoronaNet Research Project compiles a database on government responses to the coronavirus. Our main focus is to collect as much information as we can about the various fine-grained actions governments are taking to defeat the coronavirus. This includes not only gathering information about which governments are responding to the coronavirus, but also who they are targeting the policies toward (e.g., other countries), how they are doing it (e.g., travel restrictions, banning exports of masks), and when they are doing it.

Together with 500 political, social, and public health science scholars from all over the world, we present an initial release of a large hand-coded dataset of more than 20,000 separate policy announcements from governments around the world visible since December 31, 2019. Data collection is ongoing.

The data yields detailed information on:

The level of government responding to the coronavirus crisis (e.g., national, regional/state, local/municipal)
Specific actions taken (e.g., travel bans, investments in the public health sector, etc.)
Geographical areas targeted by these measures
Who or what they are targeting (e.g., foreigners, ventilators)
Compliance mechanisms (e.g., mandatory or voluntary)
Timing of policy responses.

Given the exceptional times, we have decided to release a version of the dataset that has not undergone extensive data cleaning. We aim to improve the data day by day but can not assure full accuracy among the policies. For the most up to date version of the data, please visit https://www.coronanet-project.org.

Scope of Project

Subject Terms: COVID-19; public policy; health policy; public health; governance

Geographic Coverage: World-wide coverage

Universe: This data repository aims to collect all government policies made in response to the COVID-19 pandemic across all countries since December 31, 2019, when the government of China first reported the outbreak in Wuhan to the WHO. Data collection is ongoing.

Data Type(s): event/transaction data; observational data

Methodology

Response Rate: The data collection procedure for this manuscript is as follows: (1) The PIs designed a Qualtrics survey with survey questions about different aspects of a government policy action to streamline the CoronaNet data collection effort across different research assistants who collect the data. The questions included in the Qualtrics survey was based in large part on policies the PIs coded with regards to policies adopted by the Taiwanese government from December 31, 2019 to March 21, 2020 as well as travel policies adopted by a cross-section of countries in mid-March, 2020. (2) Research assistants (RAs) watch a 2 hour training video for how to use the Qualtrics survey to document policies (3) RAs document the policies they find in the Qualtrics survey. Each RA is responsible for tracking government policy actions for at least one country. RAs were allocated depending on their background, language skills and expressed interest in certain countries. At the time of the initial deposit to ICPSR, there were more than 500 RAs collecting data on government policies made in reaction to COVID-19. (4) RAs are organized by regions, or if they are collecting subnational data, by countries. Regional and country managers oversee the quality and progress of the data collection effort. (5) Questions and mutual feedback between RAs, reigional and country managers and the PIs of this project are answered via a Slack channel set up for that purpose. All RAs are required to join the Slack channel before joining the data collection effort.

Sampling: Multiple Coding Validation

As we collect the data, we also engage in post-data validation checks.

Before validation, we use a team of RAs to check the raw data for logical inconsistencies and typographical errors. In our latest data release, we have cleaned all observations until April 1st.

We randomly sample 10% of the dataset using the source of the data (e.g. newspaper article, government press release) as our unit of randomization. We use the source as our unit of randomization because one source may detail many different policy types. We then provide this source to a fully independent RA and ask her to code for the government policy contained in the sampled source in a separate, but identical, survey instrument. If the source is in a language the RA cannot read, then a new source is drawn. The RA then codes all policies in the given source. This practice is repeated a third time by a third independent coder. Given the fact that each source in the sample is coded three times, we can assess the reliability of our measures and report the reliability score of each coder.

For more information on the results of these validation checks, please see: https://www.coronanet-project.org/validation_website.html

Data Source: Government policies announced in reaction to the COVID-19 pandemic were coded from a variety of information sources, including: government press releases and websites, videos/transcripts of government press briefings, and newspaper articles, among others. These sources were identified using the following methodology: (1) The PIs partnered with the machine learning company Jataware to automate the collection of more than 200,000 news articles from around the world related to COVID-19. Jataware employs a natural language processing (NLP) classifier using Bidirectional Encoder Representations from Transformers (BERT) to detect whether a given article is indicative of a governmental policy intervention related to COVID-19. They then apply a secondary NLP classifier to categorize the type of policy intervention (e.g. "declaration of emergency", "quarantine", "travel restrictions", etc.). Next, Jataware extracts the geospatial and temporal extent of the policy intervention (e.g. “Washington DC” and “March 15, 2020”) whenever possible. The resulting list of news sources is then provided to research assistants for manual coding and further data validation. (2) RAs further systematically check the following platforms to identify relevant policies: (i) the ACAPS COVID-19 dataset on government policies to check the completeness of our data (ii) the information page on COVID-19 policies of the US. Embassy website of a particular country (iii) the Wikipedia page on a particular country’s response to the COVID-19 pandemic (iv) the relevant government websites of a particular country (e.g. executive office, health ministry) (v) newspaper coverage of a particular country (via e.g. LexisNexis or Factiva).

Collection Mode(s): web scraping; web-based survey

Unit(s) of Observation: Each unique observation corresponds to a government policy made in response to the COVID-19 pandemic (type_sub_cat) initiated by a particular government on a particular day. Please see our codebook for more information on how the policies and their sub types are defined: https://www.coronanet-project.org/data/codebook/CoronaNet_Codebook.pdf

Geographic Unit: The geographic unit of analysis ranges from country, province, municipalities as well as other geographical units.

Related Publications

Download this project

Published Versions

V1 [2020-07-25]

Export Metadata

Dublin Core

DDI 2.5

Report a Problem

Found a serious problem with the data, such as disclosure risk or copyrighted content? Let us know.

This material is distributed exactly as it arrived from the data depositor. ICPSR has not checked or processed this material. Users should consult the investigator(s) if further information is desired.