Comparing Apples and Alphabets: tracking the return to normality
Mobility data from Google and Apple can help us look through the rhetoric on the return to economic normality and see what is really happening on the ground.
On Sunday, Boris Johnson set out a ‘first sketch of a road map for reopening society’. Several other countries and some states in the US have already taken their first steps on this road towards restoring their economies to something like normality, but it will not be a straight path anywhere.
There are plenty of open questions. Will strict quarantine measures have to be re-imposed if there is a second wave of COVID-19? And even if people are permitted to revert to their pre-lockdown work and personal lives, how long will it take for them to feel comfortable enough to commute and socialise as they did before?
To help us begin to answer the second of those two questions, we monitor a number of high-frequency metrics. For example, Google and Apple both publish mobility data based on the requests they receive for travel directions through their popular apps. These services are admittedly used more for longer or idiosyncratic trips – we tend not to need directions for our way to work or the local shops – but they can provide some insight into how much people are venturing outside. The mobility trends are also in line with other data sources we track, such as restaurant bookings, cinema takings and labour-market activity.
Google’s data are more granular, separating travel to work from ‘retail and recreation’ movements. The latter category seems a good proxy for consumer spending. However, Google only releases its data with a one-week lag. Apple’s data, on the other hand, are daily but only split travel into driving, walking, and – for some countries – public transport.
We’re concentrating on Apple’s data for now given their timeliness. The data are available for individual regions and cities, so we can assess the impact on activity and virus trends at a granular level. They are also available for all major economies in the world except China.
We first adjust for weekly seasonality before calibrating the data onto GDP based on a deep dive into the impact on the French economy conducted in late March by the country’s National Institute of Statistics and Economic Studies.
This reported that the economy was operating 35% below normal levels at that time. This was a particularly bleak assessment based on alternative data sets such as credit-card usage and surveys of different sectors. We shouldn’t take it too literally, but it gives us an anchor against which to benchmark economic activity given Apple’s mobility index for France was down 82% at the time.
Based on our analysis of this relationship between mobility and economic activity, and acknowledging the caveats already mentioned, the US is holding up better than elsewhere (from an economic perspective) and is gradually reopening.
We’ve also seen signs of economic improvement in Southern Europe as some restrictions have been lifted, although the trend in Spain has deteriorated recently.
One caveat to this approach applies to emerging markets. Given that Apple users in countries like India and South Africa are likely to reflect a narrow elite, the data may not provide an ‘apples-to-apples’ comparison for such regions versus developed economies.
Overall, though, this information about mobility trends on the ground is one more source of insight into the likely nature and duration of the current downturn. Along with the other data sources we monitor, this helps build a nuanced picture of the global economy and inform our views of the macro environment, the probable direction of risk sentiment in the market, and the outlook for businesses.