Analysing the closure of bars and restaurants in Cataluña

On October 16th 2020, with the increase of COVID19 cases, the regional government of Cataluña, Spain announced that there would be a 15 day closure of bars and restaurants across the region.

Like Comment
Read the paper

Introduction:

On October 16th 2020, with the increase of COVID19 cases, the regional government of Cataluña, Spain announced that there would be a 15 day closure of bars and restaurants across the region. They stated that it was absolutely necessary to avoid a total lock-down such as the one experienced across the whole of Spain at the beginning of the pandemic. At the time, the vaccine was unavailable as a prevention method and non-pharmaceutical interventions (NPIs) was one of the few interventions that the government had available in order to try to slow down the rising number of cases.

At the time COVID19 figures were increasing at a then-unprecedented rate with 1,620 cases reported in the 24 hours previous to the closure of bars and restaurants with Barcelona having 400 daily cases per 100,000 inhabitants with many countries at the time reporting just 50 cases per 100,000. Other restrictions included, shopping centers only being able to operate at 30% capacity, gyms at 50% and the government asking businesses where possible to allow workers to work from home and also requested that universities revert back to online-only lectures.

The decision to close the hospitality sector continued until November 23rd 2020 (a 5-week total closure) after continuous policy renewals of the proposed closure of bars and restaurants. The decisions at the time were considered very unpopular with the local population but Spain was hit particularly hard for a second time after the first wave, despite having one of the most severe lock-downs in Europe back in March 2020 in which it was closed completely for 3 months.

Economically, tourism and the hospitality industry is one of Spains and Cataluñas most popular sectors and a sector in which many people work partly in-black. Therefore when such a sector is forced to close, there is little financial help available to the people that rely on these sectors.

Mobility and COVID19:

 

We obtained, in collaboration with the Spanish Ministry of Transportation and Urban Agenda (MITMA) data on anonymized mobile phone population mobility paterns which allowed us to compute the movements to, from and within municipalities and districts of a given region. For each region we also obtained the number of COVID19 cases. Here a region is defined as a basic health area (Área Básica Salud) and data on COVID19 cases was collected on a daily frequency. 

 

Since the spread of COVID19 is from person-to-person, then increased interaction of people would suggest an increase in the number of COVID19. We use a proxy for the interaction of people through mobile phone mobility data and the aim of the investigation was to measure the reduction in mobility levels after the public policy intervention of closing bars and restaurants was introduced. In order to do so we used each Autonomous Community of Spain as a control group. Here the control group should act as a “what-if” scenario. What would the mobility of people in Cataluña look like if Cataluña did not introduce the policy of closing down bars and restaurants?

 

Figure 1 shows the mobility data for Cataluña and Madrid. Each bar corresponds to the total trips for a given day and the periodic drops correspond to weekends. The shaded area corresponds to the time period in which the policy was introduced in Cataluña (5 weeks in total). We can clearly see that there is a shift downwards to the fitted regression line for the mobility of people in Cataluña immediately after the policy was introduced. 

 

Cataluña vs Madrid Mobility Levels
Caption

Figure 1: Number of trips: Shaded region corresponds to the period in which the policy was introduced in Cataluña.

Relation to COVID19:

With increased mobility we can expect that there may be an increased risk of COVID19 incidence and thus, we aimed to relate the policy of closing bars and restaurants to a reduction in the COVID19 incidence, once a time-lag has been taken into consideration. Figure 2 shows a growth rate ratio of cases and normalized mobility. There does appear to be a strong relationship between the increase in mobility and increase in cases once an optimal lag of 21 days for the COVID19 cases data was taken into consideration.

Conclusions:

After the strict lockdown that Spain endured during the first months of the pandemic, policy makers were keen to find alternative ways in order to help slow the spread of COVID19 incidence. The unpopular nature of a full lockdown in Spain was not just difficult politically but would have further pushed the country into economic hardship. The regional government's plan to target a specific sector of where social interaction is highest did appear to have a positive effect on the slow-down in the incidence of COVID19 in Catalunña. It is often difficult to accurately measure the direct impact of a policy decision on the slow-down of COVID19 however, the methods used in this research are aimed at exactly this, measuring the impact of policy decisions.

Mobility and COVID19 cases relationship
Caption

Figure 2: Growth rate ratio and normalised mobility correlation.

Matthew Smith

Researcher, BSC

Barcelona Supercomputing Center researcher