Mobility Data From US, Europe, Asia

By in
Mobility Data From US, Europe, Asia

It seems like not an hour goes by now that we don’t hear about fresh curbs on economic activity as countries and cities around the world try to curb the spread of the virus. Measuring the impact of these initiatives on output/labor conditions is difficult, and not just because these efforts are different everywhere around the globe. We also have to consider how much a given country or city tried to return to “normal” over the last few months. Are they locking down from pre-pandemic levels, or from already depressed ones?

We’ll use Apple Mobility data to address this question for 4 countries/cities around the world. Apple makes this information freely available (link below), primarily as an aide to government policymakers. It is based on users’ rerouting requests to Apple Maps: the more reroutes, the greater the likely congestion on streets, roads, and in mass transit. The data here is from the start of 2020 and benchmarked (those percentages you’ll see in the lower left corner) to January 13th.

Example #1: The United States. This graph shows that US driving/walking patterns did get back to some level of normal over the summer but are slowly trending down now. Mass transit usage, most common in the US’ largest cities, never recovered.

Example #2: Germany. Compare this to the US chart and you’ll see that the country’s recent initiatives to curtail the virus spread are more visible in the data. All forms of transportation are showing lower levels than January right now.

Example #3: France. Here we see an even more pronounced effect from recent lockdown efforts. We are, in fact, very near the trough levels of April/May.

#4: Hong Kong. Apple doesn’t provide Mobility data for China, so this is as close as we can get. The data here looks nothing like the US or European charts above (no recovery), but it does look very much like Seoul and Singapore (use the link in the Sources section below to see those, or any other large cities/countries).

Summing up: we’ve been doing this traffic-based analysis on a weekly basis in these notes for +6 months now and can share 2 lessons with you as you evaluate these/other Apple Mobility charts.

  • The first is that you don’t see much of a drop in new cases until you get a sharp reduction in mobility (April-May in the first 3 charts above). By that measure, Europe should start to see results soon, but the US may not for several weeks.
  • The second lesson is that this sort of crowd-sourced data is remarkably accurate in the aggregate, either at a city or country level. It is absolutely predictive of things like viral spread, but also a population’s mobility. This relates directly to economic variables like leisure/hospitality spending, in-store shopping, and energy/fuel consumption.

    That’s why we always publish this analysis in “Disruption” rather than “Data”, for it is a novel way to track economic activity. We look forward to using it in 2021 and beyond for cheerier topics.

Source:

Apple Mobility Data: https://covid19.apple.com/mobility