The coronavirus is challenging societies around the world, with Italy one of the main centers of attention. Due to Italy's early and severe exposure to the virus, many measures that later became commonplace around the world were first implemented there. The first measures were taken on a municipal level at the end of February, quarantining some of the most affected areas in Lombardy. The full region of Lombardy and several northern provinces were put under lockdown on March 8, with a national quarantine going into effect on March 10. In this post, we look at the effects of these measures.
The chart below is an overview of the traffic flows in Italy relative to their January averages. Measured the number of daily trips between and within cities compared to the average number of daily trips in January. Select one of the bars to see the geographic breakdown of the mobility reduction for that week in the map below.
The map shows the remaining mobility per region in the selected week relative to January traffic levels. Each colored region on the map represents a corresponding urban area. The lower the percentage value, the greater the reduction in mobility. The colored arcs represent mobility between urban centers. Hover over the colored areas to see details.
A first reduction of mobility can be observed in the most affected areas in the week from February 22 to February 28. Brescia, Milan and Bergamo saw their residual mobility levels reduced to around 80% - 85% of January traffic. This trend continues into the first week of March during which we see a reduction in traffic between northern cities to 40% - 70% of January levels.
The week of the lockdown shows a stark contrast between Italian cities and the cities from neighboring countries included in this analysis: While Italian inner city traffic was reduced to 45% - 60% of January levels throughout the country, traffic between cities dropped to 10%-50%. Figures from the end of March show incredibly low levels of mobility, with most connections between cities reduced to less than 10% and most inner city traffic below 25% of January traffic flows.
The two graphs below show the distribution of residual mobility in the selected week within and between the selected cities. The first graph depicts the percentage of the selected cities whose residual mobility falls inside the bucket noted on the x axis. The second graph shows the percentage of city connections whose residual mobility falls inside the respective range of percentage points in the given week.
For this particular analysis we asked ourselves the following questions: Do we see an overall reduction of mobility in Italy following the implementation of restrictions in response to COVID-19? If yes, how strong was the reduction? How quickly did people comply? Furthermore, we wondered if we would find regional differences. It also seems reasonable to expect that local trips to the supermarket stay relatively frequent while longer distance travel between cities would drop off sharply.
We were also curious about cross-border traffic. Every country adopted a slightly different response to COVID-19 and it seems reasonable to investigate whether these differences are visible in people's mobility patterns.
The data for this analysis stems from the TomTom Origin/Destination API. This API processes huge amounts of de-identified floating car data to provide estimates of traffic flows between and within geographic regions.
We want to be clear that while these results are interesting to study, we should be careful with their interpretation. If a city shows less of a mobility reduction than another city, this does not mean that there is less compliance with the restrictions. It might just as well mean that this city hosts more essential industry and simply cannot reduce mobility further. To make the most out of this data, it should be correlated and combined with other sources and carefully interpreted.