High-frequency Data links Economy and Pandemic
Confine people to their homes; shut the shops; spending goes down; GDP suffers. It may seem like common sense that stronger Covid-19-related restrictions, while essential for public health, would result in a lower GDP, but anecdata isn’t real data. Luckily, real data is available in the form of a model from the Bank of England which demonstrates how high-frequency data can help make predictions during challenging times.
The model tracks the stringency of government measures and the mobility of populations to predict GDP. Using Oxford University’s Coronavirus Government Response Tracker, which tracks the stringency of government measures, like closures and travel bans, from around the world, we can see that, broadly speaking, GDP fell more in countries with more stringent restrictions.
We can also see daily measures of population mobility thanks to Apple’s Mobility Trends Reports. This data complements government rules by including people voluntarily restricting their movement in response to risks. Falls in GDP were also greater in countries where households were less mobile.
The Bank of England combines this timely data on mobility and rules, controlling for country-specific characteristics, to generate GDP predictions. The model correctly predicted that the UK would suffer a bigger fall in GDP than Germany, which needed fewer control measures. This kind of high-frequency data can complement more traditional measures of economic activity during these challenging and unpredictable times.
Read more here.
