Taming the Zoo of Consumption Responses to Labour Income Changes

As policies aimed at enhancing household economic resilience routinely target labour income —the primary source of household income— understanding the consumption responses elicited by diverse labour income changes is essential for effective policy design. We study the heterogeneity in the consumption responses to labour income changes originating from the magnitude, sign, and dynamic nature of the income change (transient, recurrent, or permanent). While prior empirical research has explored how household characteristics shape consumption responses, the impact of different types of income changes has received little scholarly attention. The literature reports a wide range of often conflicting estimates — a “zoo” of responses — driven by variations in methodologies, definitions, datasets, and policy contexts. To tame this zoo, we propose a three-pronged approach. First, we establish a framework that categorizes income changes as recurrent, permanent, or transient, and develop an algorithm to identify these income changes in individual income time series at scale. Second, we use comprehensive administrative banking data, spanning twelve years and covering millions of households, to systematically estimate consumption responses to these different types of labour income changes. Third, we apply the COICOP consumption classification to disentangle how nondurable, semidurable, and durable consumption respond to each type of income change. We find that consumption responses to positive recurrent labour income changes are nearly twice as large as the responses to transient positive labour income changes. These consumption responses are primarily driven by changes in semidurable and durable consumption, rather than nondurable consumption.