Premium
Maintaining (Locus of) Control? Data Combination for the Identification and Inference of Factor Structure Models
Author(s) -
Piatek Rémi,
Pinger Pia
Publication year - 2015
Publication title -
journal of applied econometrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.878
H-Index - 99
eISSN - 1099-1255
pISSN - 0883-7252
DOI - 10.1002/jae.2456
Subject(s) - inference , locus of control , econometrics , identification (biology) , scarcity , latent variable , computer science , monte carlo method , sample (material) , economics , statistics , psychology , machine learning , artificial intelligence , mathematics , social psychology , microeconomics , botany , biology , chemistry , chromatography
Summary Factor structure models are widely used in economics to extract latent variables, such as personality traits, and to measure their impact on outcomes of interest. The identification and inference of these models, however, highly depend on the availability of rich longitudinal data. To overcome the common problem of data scarcity, this paper proposes to combine datasets that each identify some part of the likelihood, thereby recovering the identification of the complete model. The performance of the approach is demonstrated by a Monte Carlo experiment. We apply this technique empirically to study the impact of locus of control on education and wages. Our strategy allows us to elicit the distribution of pre‐market locus of control from a sample of young individuals, and to measure its impact on education and wages in a sample of adults. Our findings indicate that the effect of locus of control on wages mainly operates through education. Copyright © 2015 John Wiley & Sons, Ltd.