Concept-Based Bayesian Model Averaging and Growth Empirics

J.R. Magnus, W. Wang

Research output: Working paperDiscussion paperOther research output

274 Downloads (Pure)

Abstract

Abstract: In specifying a regression equation, we need to determine which regressors to include, but also how these regressors are measured. This gives rise to two levels of uncertainty: concepts (level 1) and measurements within each concept (level 2). In this paper we propose a hierarchical weighted least squares (HWALS) method to address these uncertainties. We examine the effects of different growth theories taking into account the measurement problem in the growth regression. We find that estimates produced by HWALS provide intuitive and robust explanations. We also consider approximation techniques when the number of variables is large or when computing time is limited, and we propose possible strategies for sensitivity analysis.
Original languageEnglish
Place of PublicationTilburg
PublisherEconometrics
Number of pages39
Volume2012-017
Publication statusPublished - 2012

Publication series

NameCentER Discussion Paper
Volume2012-017

Fingerprint

Uncertainty
Bayesian model averaging
Empirics
Growth regressions
Approximation
Growth theory
Sensitivity analysis
Weighted least squares
Least square method

Keywords

  • Hierarchical model averaging
  • Growth determinants
  • Measurement problem

Cite this

Magnus, J. R., & Wang, W. (2012). Concept-Based Bayesian Model Averaging and Growth Empirics. (CentER Discussion Paper; Vol. 2012-017). Tilburg: Econometrics.
Magnus, J.R. ; Wang, W. / Concept-Based Bayesian Model Averaging and Growth Empirics. Tilburg : Econometrics, 2012. (CentER Discussion Paper).
@techreport{889f1e526cc4470e87ce2420e409cf06,
title = "Concept-Based Bayesian Model Averaging and Growth Empirics",
abstract = "Abstract: In specifying a regression equation, we need to determine which regressors to include, but also how these regressors are measured. This gives rise to two levels of uncertainty: concepts (level 1) and measurements within each concept (level 2). In this paper we propose a hierarchical weighted least squares (HWALS) method to address these uncertainties. We examine the effects of different growth theories taking into account the measurement problem in the growth regression. We find that estimates produced by HWALS provide intuitive and robust explanations. We also consider approximation techniques when the number of variables is large or when computing time is limited, and we propose possible strategies for sensitivity analysis.",
keywords = "Hierarchical model averaging, Growth determinants, Measurement problem",
author = "J.R. Magnus and W. Wang",
note = "Pagination: 39",
year = "2012",
language = "English",
volume = "2012-017",
series = "CentER Discussion Paper",
publisher = "Econometrics",
type = "WorkingPaper",
institution = "Econometrics",

}

Magnus, JR & Wang, W 2012 'Concept-Based Bayesian Model Averaging and Growth Empirics' CentER Discussion Paper, vol. 2012-017, Econometrics, Tilburg.

Concept-Based Bayesian Model Averaging and Growth Empirics. / Magnus, J.R.; Wang, W.

Tilburg : Econometrics, 2012. (CentER Discussion Paper; Vol. 2012-017).

Research output: Working paperDiscussion paperOther research output

TY - UNPB

T1 - Concept-Based Bayesian Model Averaging and Growth Empirics

AU - Magnus, J.R.

AU - Wang, W.

N1 - Pagination: 39

PY - 2012

Y1 - 2012

N2 - Abstract: In specifying a regression equation, we need to determine which regressors to include, but also how these regressors are measured. This gives rise to two levels of uncertainty: concepts (level 1) and measurements within each concept (level 2). In this paper we propose a hierarchical weighted least squares (HWALS) method to address these uncertainties. We examine the effects of different growth theories taking into account the measurement problem in the growth regression. We find that estimates produced by HWALS provide intuitive and robust explanations. We also consider approximation techniques when the number of variables is large or when computing time is limited, and we propose possible strategies for sensitivity analysis.

AB - Abstract: In specifying a regression equation, we need to determine which regressors to include, but also how these regressors are measured. This gives rise to two levels of uncertainty: concepts (level 1) and measurements within each concept (level 2). In this paper we propose a hierarchical weighted least squares (HWALS) method to address these uncertainties. We examine the effects of different growth theories taking into account the measurement problem in the growth regression. We find that estimates produced by HWALS provide intuitive and robust explanations. We also consider approximation techniques when the number of variables is large or when computing time is limited, and we propose possible strategies for sensitivity analysis.

KW - Hierarchical model averaging

KW - Growth determinants

KW - Measurement problem

M3 - Discussion paper

VL - 2012-017

T3 - CentER Discussion Paper

BT - Concept-Based Bayesian Model Averaging and Growth Empirics

PB - Econometrics

CY - Tilburg

ER -

Magnus JR, Wang W. Concept-Based Bayesian Model Averaging and Growth Empirics. Tilburg: Econometrics. 2012. (CentER Discussion Paper).