Consumption during unemployment: a Synthetic Diff-in-Diff Approach

Article tags:
  • Transaction data research

Jeannine van Reeken-van Wee - economist - ABN AMRO Bank

Co-authors: Nora Neuteboom (IMF), Bas van der Klaauw (VU Amsterdam), George Kapetanios (Kings College)

Consumption during unemployment: a Synthetic Difference in Difference Approach using Bank Transaction Data

This study investigates household financial behavior during unemployment spells in the Netherlands using high-frequency transaction data. In our approach, we contrasted treated individuals, who underwent job loss, with a synthetic control group consisting of non-treated individuals possessing comparable financial characteristics. The initial onset of unemployment triggers a substantial surge in income, primarily attributed to severance pay, but swiftly drops post-unemployment, with unemployment benefits covering slightly over half of former salary earnings. Despite a re-employment rate of around half within six months, the treatment group experiences a persistent average monthly earnings reduction of approximately 600 EUR by month. Spending patterns fluctuate significantly, surging before unemployment due to severance pay and declining below non-treated individuals post-unemployment, indicating challenges to fully smooth consumption after job loss. Furthermore, our study shows different spending patterns for people who receive severance pay versus similar amounts of liquidity holdings. People with severance pay spend more around job loss, contrasting a much smaller relative spending of people with similar amounts of liquidity holdings.

Jeannine van Reeken-van Wee is economist at ABN AMRO Bank. She focusses on analysing bank transaction data to study economics or policy-related matters. Currently, she is pursuing a part-time PhD in applied microeconomics at the Vrije Universiteit Amsterdam.