- Google mobility reports show where people are visiting, using a tracking app. Six places are reported: grocery and pharmacy; parks; transit stations (public transport); retail and recreation; residential; workplaces. The data helps health authorities monitor lockdowns. Can the data also say something about the economy?
- The data relies on people letting themselves be tracked. There is likely to be a selection bias in age and income level of those followed. Bias in phone-based tracking has been a problem in the past (e.g. with Pokémon Go).
- Retail visits will not match retail sales. Online retail sales will appear in the "residential" category. Retail visits data in economies with high online sales will miss this.
- Workplace visits will not match how many people have stopped working. The data presumably covers office and factory locations. That will not record people working from home, farmers, delivery drivers, shop workers, teachers, people like plumbers who work in houses, etc. Looking at the detailed US Occupation Employment Statistics, about 38% of the workforce seems likely to work at Google "workplaces".
- Google mobility data is helpful at a time when there are few reliable economic statistics. It is not a direct guide to the economy, and should be used with care.