Full-information estimation of heterogeneous agent models using macro and micro data

Laura Liu; Mikkel Plagborg-Møller

doi:10.3982/QE1810

Full-information estimation of heterogeneous agent models using macro and micro data

Laura Liu, Mikkel Plagborg-Møller

Princeton University

Research output: Contribution to journal › Article › peer-review

Abstract

We develop a generally applicable full-information inference method for heterogeneous agent models, combining aggregate time series data and repeated cross-sections of micro data. To handle unobserved aggregate state variables that affect cross-sectional distributions, we compute a numerically unbiased estimate of the model-implied likelihood function. Employing the likelihood estimate in a Markov Chain Monte Carlo algorithm, we obtain fully efficient and valid Bayesian inference. Evaluation of the micro part of the likelihood lends itself naturally to parallel computing. Numerical illustrations in models with heterogeneous households or firms demonstrate that the proposed full-information method substantially sharpens inference relative to using only macro data, and for some parameters micro data is essential for identification.

Original language	American English
Pages (from-to)	1-35
Number of pages	35
Journal	Quantitative Economics
Volume	14
Issue number	1
DOIs	https://doi.org/10.3982/QE1810
State	Published - Jan 2023

ASJC Scopus subject areas

Economics and Econometrics

Keywords

Bayesian inference
C11
C32
E1
data combination
heterogeneous agent models

Access to Document

10.3982/QE1810

Cite this

@article{a5da8f1ab96346e9a970998fd280d292,

title = "Full-information estimation of heterogeneous agent models using macro and micro data",

abstract = "We develop a generally applicable full-information inference method for heterogeneous agent models, combining aggregate time series data and repeated cross-sections of micro data. To handle unobserved aggregate state variables that affect cross-sectional distributions, we compute a numerically unbiased estimate of the model-implied likelihood function. Employing the likelihood estimate in a Markov Chain Monte Carlo algorithm, we obtain fully efficient and valid Bayesian inference. Evaluation of the micro part of the likelihood lends itself naturally to parallel computing. Numerical illustrations in models with heterogeneous households or firms demonstrate that the proposed full-information method substantially sharpens inference relative to using only macro data, and for some parameters micro data is essential for identification.",

keywords = "Bayesian inference, C11, C32, E1, data combination, heterogeneous agent models",

author = "Laura Liu and Mikkel Plagborg-M{\o}ller",

note = "Publisher Copyright: Copyright {\textcopyright} 2023 The Authors.",

year = "2023",

month = jan,

doi = "10.3982/QE1810",

language = "American English",

volume = "14",

pages = "1--35",

journal = "Quantitative Economics",

issn = "1759-7323",

publisher = "The Economic Society",

number = "1",

}

TY - JOUR

T1 - Full-information estimation of heterogeneous agent models using macro and micro data

AU - Liu, Laura

AU - Plagborg-Møller, Mikkel

PY - 2023/1

Y1 - 2023/1

N2 - We develop a generally applicable full-information inference method for heterogeneous agent models, combining aggregate time series data and repeated cross-sections of micro data. To handle unobserved aggregate state variables that affect cross-sectional distributions, we compute a numerically unbiased estimate of the model-implied likelihood function. Employing the likelihood estimate in a Markov Chain Monte Carlo algorithm, we obtain fully efficient and valid Bayesian inference. Evaluation of the micro part of the likelihood lends itself naturally to parallel computing. Numerical illustrations in models with heterogeneous households or firms demonstrate that the proposed full-information method substantially sharpens inference relative to using only macro data, and for some parameters micro data is essential for identification.

AB - We develop a generally applicable full-information inference method for heterogeneous agent models, combining aggregate time series data and repeated cross-sections of micro data. To handle unobserved aggregate state variables that affect cross-sectional distributions, we compute a numerically unbiased estimate of the model-implied likelihood function. Employing the likelihood estimate in a Markov Chain Monte Carlo algorithm, we obtain fully efficient and valid Bayesian inference. Evaluation of the micro part of the likelihood lends itself naturally to parallel computing. Numerical illustrations in models with heterogeneous households or firms demonstrate that the proposed full-information method substantially sharpens inference relative to using only macro data, and for some parameters micro data is essential for identification.

KW - Bayesian inference

KW - C11

KW - C32

KW - E1

KW - data combination

KW - heterogeneous agent models

UR - http://www.scopus.com/inward/record.url?scp=85147013993&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85147013993&partnerID=8YFLogxK

U2 - 10.3982/QE1810

DO - 10.3982/QE1810

M3 - Article

SN - 1759-7323

VL - 14

SP - 1

EP - 35

JO - Quantitative Economics

JF - Quantitative Economics

IS - 1

ER -

Full-information estimation of heterogeneous agent models using macro and micro data

Abstract

ASJC Scopus subject areas

Keywords

Access to Document

Other files and links

Fingerprint

Cite this