Abstract
Motivation
Academicians and policy-makers grapple with monitoring the economic impact of crises such as COVID-19 when survey data are scarce.
Purpose
We show how a set of publicly available real-time indicators—nitrogen dioxide emissions, night-time lights, mobile phone mobility tracking, internet searches, and food prices—tracked changes in GDP across multiple countries in 2020.
Approach and methods
We first describe the extent to which real-time indicators captured the COVID-19 crisis. We employ linear models, selected using the least absolute shrinkage and selection operator (LASSO), to examine the capacity of these indicators to track GDP growth during the crisis.
Findings
Google Mobility and staple food prices both sharply declined in March and April 2020, followed by a rapid recovery returning to baseline levels by July and August 2020. Mobility and staple food prices experienced a milder decrease in low-income countries.
Nitrogen dioxide and night-time lights showed a similar pattern, with the steepest fall followed by a swift recovery in lower middle-income countries.
In April and May, Google search terms reflecting economic distress and religiosity spiked in some regions but not in others.
Linear models explain about half of the variation in annual GDP growth in 68 countries. In a smaller subset of higher-income countries, real-time indicators explain about 65% of the variation in quarterly GDP growth.
Policy implications
Real-time indicators offer several advantages in crisis monitoring, being readily available, cost-effective, and not requiring face-to-face interactions, which are particularly valuable during a pandemic.
Academicians and policy-makers grapple with monitoring the economic impact of crises such as COVID-19 when survey data are scarce.
Purpose
We show how a set of publicly available real-time indicators—nitrogen dioxide emissions, night-time lights, mobile phone mobility tracking, internet searches, and food prices—tracked changes in GDP across multiple countries in 2020.
Approach and methods
We first describe the extent to which real-time indicators captured the COVID-19 crisis. We employ linear models, selected using the least absolute shrinkage and selection operator (LASSO), to examine the capacity of these indicators to track GDP growth during the crisis.
Findings
Google Mobility and staple food prices both sharply declined in March and April 2020, followed by a rapid recovery returning to baseline levels by July and August 2020. Mobility and staple food prices experienced a milder decrease in low-income countries.
Nitrogen dioxide and night-time lights showed a similar pattern, with the steepest fall followed by a swift recovery in lower middle-income countries.
In April and May, Google search terms reflecting economic distress and religiosity spiked in some regions but not in others.
Linear models explain about half of the variation in annual GDP growth in 68 countries. In a smaller subset of higher-income countries, real-time indicators explain about 65% of the variation in quarterly GDP growth.
Policy implications
Real-time indicators offer several advantages in crisis monitoring, being readily available, cost-effective, and not requiring face-to-face interactions, which are particularly valuable during a pandemic.
Original language | English |
---|---|
Journal | Development Policy Review |
DOIs | |
Publication status | Published - 3 Jul 2024 |
Keywords
- COVID-19
- GDP impact estimation
- big data