Bachelor Thesis · University of Pisa

Heterogeneity and Local Interactions in Macroeconomic Dynamics

by Mattia Pasqualini · University of Pisa · 110/110 summa cum laude

StatusCompleted, 2022
FieldMacroeconomics
MethodAgent-Based Modeling
Grade110/110 Summa cum Laude

Abstract

This thesis develops an agent-based approach to macroeconomic modeling, focusing on the dynamics of labor and goods markets under different institutional arrangements. Rather than relying on the representative-agent assumption common to standard macro models, the analysis explicitly models heterogeneous agents interacting locally, allowing aggregate dynamics to emerge from the bottom up.

Applying insights from behavioral economics and search theory, the model analyzes how the economy responds to negative productivity shocks, and how local interaction structures shape the propagation and persistence of these shocks at the aggregate level.

Leveraging MultiVeStA — a Java-based tool for automated statistical analysis of agent-based models — the thesis identifies the emergence of self-sustaining shock resistance under certain parameters and conditions, offering a novel mechanism through which local heterogeneity can generate aggregate resilience.

Research Question

How do local interaction structures and agent heterogeneity in labor and goods markets shape the economy's aggregate response to negative productivity shocks?

Key Contributions

01
Departs from representative-agent macro models by explicitly modeling heterogeneous, locally-interacting agents in labor and goods markets.
02
Integrates behavioral economics and search theory into the agent decision rules, improving the model's predictive and explanatory power.
03
Discovers the emergence of self-sustaining shock resistance: under certain parameter regimes, local heterogeneity generates aggregate resilience to negative shocks.
04
Uses MultiVeStA for rigorous, automated statistical analysis of simulation outputs — a methodologically advanced approach for an undergraduate thesis.

Methodology

1

Agent-based market design

Labor and goods markets are modeled with heterogeneous agents operating under different institutional arrangements, departing from the representative-agent assumption of standard macro models.

2

Behavioral & search-theoretic foundations

Agent decision rules incorporate insights from behavioral economics and search theory, improving the realism of matching and price-formation processes in the model.

3

Statistical analysis via MultiVeStA

Simulation outputs are analyzed using MultiVeStA, a Java-based statistical model-checking tool, enabling rigorous, automated analysis of how the economy responds to negative productivity shocks across parameter regimes.

Keywords

Agent-based modeling Macroeconomics Heterogeneous agents Search theory Behavioral economics Productivity shocks MultiVeStA
Read full thesis View all research

Request more information

Interested in the full working paper, data, or methodology? Send a message and I'll get back to you directly.

This opens your email client addressed to m.pasqualini@tue.nl with your message pre-filled.

Opening your email client...