The consumption response to labour income changes

We study the heterogeneity in the consumption responses to income changes originating from the magnitude, sign, and dynamic nature of the income change.

While empirical studies have investigated the heterogeneity stemming from household characteristics, little is known about the effect income change types has on the consumption response. Additionally, these studies have reported a large spread of responses due to incompatible methods, definitions, datasets, and policy events. We tame this zoo of consumption responses with a three-pronged approach.

First, we define a framework that organises income changes into positive/negative recurrent, permanent, and transient changes and develop an algorithm that can identify them in individual's income time-series. Second, we leverage administrative banking data spanning twelve years and millions of clients to systematically estimate how consumption responds to these three types of labour income changes while controlling for household-level characteristics. Third, using the COICOP standard for consumption, we delve deeper into the source of consumption response by disentangling between the response in non-durable, semi-durable, and durable consumption.

We find that consumption responses to positive recurrent labour income changes are almost twice as large as the responses to transient positive labour income changes. We also find these responses originate more from changes in semi-durable consumption than from durable or non-durable consumption. Since policies to bolster economic resilience of households often target labour income, which makes up the main source of their income, understanding how policy design affects the consumption response is crucial in formulating appropriate policies.

"Milan van den Heuvel is the CEO and Research Director at Prometis Lab, a non-profit research institute that aims to enables more informed decisions by all levels of society, companies, and government by researching socioeconomic challenges leveraging anonymised financial data from BNP Paribas Fortis. He is also a visiting Professor at the Economics department of Ghent University, the Co-Chairholder of the BNP Paribas Fortis Research Chair: ""Models for a changing world."", and an affiliate researcher at the Center for Applied AI at Chicago Booth"