Nowcasting Cost of Living Using Google Trends: The Case of Selected Asian Countries

Yee Sheng Mak, Siok Kun Sek, Mohsen Ayyash
Statistika, 106(3): 332–352
https://doi.org/10.54694/stat.2025.56

Abstract

This paper examines the cost of living (COL) in six selected Asian countries: Indonesia, Japan, Korea, Malaysia, Singapore, and Thailand. Motivated by rising inflationary pressures and limitations of traditional economic indicators – such as reporting delays and limited granularity – this study explores whether real-time data from Google Trends searches can improve the tracking and prediction of COL patterns in several country-specific settings. The analysis covers the period from 2010 to 2024 and employs Mixed Data Sampling (MIDAS) regression. The results demonstrate that MIDAS models provide superior flexibility and responsiveness by accommodating high-frequency Google Trends data alongside lower-frequency macroeconomic variables. Notably, MIDAS models yield improved predictive accuracy and timeliness in capturing COL dynamics. The findings also highlight the significance of search terms like debt, inflation, and unemployment as behavioural proxies for public economic concerns. This study underscores the potential of integrating search engine data into COL analysis to improve the timeliness and sensitivity of economic monitoring.

Keywords
Cost of living, Google Trends, macroeconomic indicators, Mixed Data Sampling (MIDAS) regression, real time data