A Quantitative Agent-Based Model of Household Consumption and Saving: Application to French Data

This paper develops a bottom-up framework for household consumption and saving, which departs from the standard dynamic optimization approach. Instead, we adopt a behavioural and a data-driven approach which embeds micro heterogeneity into the macroeconomic transmission of policies and shocks. Starting from the Household Finance and Consumption Survey (HFCS) data, we adapt a credit-augmented consumption framework to the microeconomic level.
In this setting, credit and liquidity constraints, as well as wealth, are reflected in consumption decisions since the marginal propensity to consume out of current income is derived from households’ balance sheets. We complement the consumption block with a portfolio module that models the allocation of saving as a function of wealth, uncertainty, and asset returns. Moreover, we micro-found subjective beliefs about future income growth and income uncertainty within a Bayesian learning framework calibrated against Consumer Expectations Survey (CES) data.
Finally, we incorporate this consumption-saving framework into a larger agent-based model of the French economy to assess its properties.

Recommended citation: Gurgone, A., & Muellbauer, J. (2026). A Quantitative Agent-Based Model of Household Consumption and Saving: Application to French Data. In preparation.