There are several projects tracking government responses to the Covid-19 pandemic using public sources. One of the largest of these projects is the CoronaNet Research Project. Over 500 research assistants worldwide contribute to this comprehensive dataset comprising of over 100,000 entries as of February 2022. “A good reason” was inspired by the CoronaNet project: I participated as a research assistant with the task of keeping track of policy changes in Honduras and later in the German state of Schleswig-Holstein. Initially, I meant to use CoronaNet data to write a short report about stay-at-home orders in different countries. Sadly, I found out that despite the enormous efforts of hundreds of volunteers, the dataset was still very incomplete at that time (late 2020). CoronaNet researchers are consistently working on updating and completing their dataset and I hope my project can contribute as an additional source of information for them as well as for other researchers.
Perhaps the most comprehensive and most cited policy tracker is the Oxford COVID-19 Government Response Tracker (OxCGRT). It provides data on 23 indicators of government response collected by over one hundred Oxford University students and staff from different parts of the world. One of the included indicators are stay-at-home requirements which are classified among two dimensions: First, countries are put into four different categories depending on the severity of stay-at-home requirements and coded like this:
0 – no measures
1 – recommend not leaving the house
2 – require not leaving the house with exceptions for daily exercise, grocery shopping, and “essential” trips
3 – require not leaving the house with minimal exceptions (e.g. allowed to leave only once every few days, or only one person can leave at a time etc.)
Second, it is noted whether the policy is targeted or not meaning whether the policy only applies to a sub-region or jurisdiction.
While this is generally a very well-maintained dataset, it does hardly allow for a more nuanced analysis of the effects of stay-at-home orders in different countries. Most European countries have been grouped into category 2 for most of the time even though the exceptions did not in all cases include “daily exercise, grocery shopping, and “essential” trips” or the definition of what is an essential trip varied dramatically. Spain and Austria have been grouped in the same category for instance when Austrians could leave their house “for recreational purposes” at all times while in Spain physical exercise was not treated as “essential” and even going for a short walk was banned for many weeks. Assigning Spain and Austria into the same category neglects the potentially very different effects lockdown policies had in these countries, especially on the mental health of their populations.
The Oxford COVID-19 Government Response Tracker further does not distinguish between curfews and all-day stay-at-home orders. In reality, it makes a huge difference whether people are free to move during the day and are “only” forced to stay home for example from midnight to 5 AM as was the case in most of Germany in May 2021.
In my dataset, I included a variable that checks for any discrepancies between my data and OxCGRT’s data on stay-at-home restrictions, i.e. day-country combinations for which their data indicates mandatory stay-at-home restrictions and mine doesn’t or vice versa. That is currently (February 2022) the case for 9 percent of day-country entries for which both trackers have data. This potentially means that in nearly 1 of 10 cases either I or the OxCGRT team have classified a country incorrectly. Given that I’m all alone and they are a big team of researchers speaking many different languages, I assume that more mistakes are to be found on my side, especially for countries whose language I don’t understand. On the other hand, I have a very narrow focus on stay-at-home restrictions, and I already identified some mistakes in the OxCGRT data in the past, so please don’t take either data at face value. If you find any mistake in my data, please send me an e-mail!
The most important source for my dataset (around 25 percent of all URLs in the dataset) was the Covid-19 Health System Response Monitor (HSRM) of the European Observatory on Health Systems and Policies that collects data on Covid-19 policies in European countries.
The Oxford Supertracker provides an overview of Covid-19 policy trackers and surveys. On 2 May 2021, 151 policy trackers were listed, most of them specialised on distinct regions or countries and on specific types of policy intervention. There are a number of policy trackers that cover stay-at-home orders and intend to do so worldwide or at least for all European countries, but a quick scan through these revealed that none of them offers complete data covering all European countries for the entire period from the beginning of 2020 until now.
The described limitations of the CoronaNet and Oxford datasets are the main motivation for this project. The core question my data aims to address is: Could you go for a walk? This question might sound almost silly in pre-2020 terms. There is little research on this because it was seen as self-evident to most researchers at least in Europe that you could just open your door and go out to take a breeze, no matter at what time and with which motivation. As we all know, this fundamental freedom has been taken away from many people, but there is little structured evidence so far. Established datasets such as the Oxford Government Response Tracker take into account whether a country imposed a stay-at-home order or not but they do not address the question of what this effectively meant if you lived there and just wanted to go out to take a walk.