A dynamic factor model approach to incorporate Big Data in state space models for official statistics

Research output: Working paperProfessional

60 Downloads (Pure)

Abstract

In this paper we consider estimation of unobserved components in state space models using a dynamic factor approach to incorporate auxiliary information from high-dimensional data sources. We apply the methodology to unemployment estimation as done by Statistics Netherlands, who uses a multivariate state space model to produce monthly figures for the unemployment using series observed with the labour force survey (LFS). We extend the model by including auxiliary series about job search behaviour from Google Trends and claimant counts, partially observed at higher frequencies. Our factor model allows for nowcasting the variable of interest, providing reliable unemployment estimates in real time before LFS data become available.
Original languageEnglish
PublisherarXiv.org at Cornell University Library
Publication statusPublished - 31 Jan 2019

Keywords

  • high-dimensional data analysis
  • state space
  • factor models
  • nowcasting
  • unemployment
  • Google Trends

Cite this

@techreport{db4473100a764b4cb3afa58f829b29fe,
title = "A dynamic factor model approach to incorporate Big Data in state space models for official statistics",
abstract = "In this paper we consider estimation of unobserved components in state space models using a dynamic factor approach to incorporate auxiliary information from high-dimensional data sources. We apply the methodology to unemployment estimation as done by Statistics Netherlands, who uses a multivariate state space model to produce monthly figures for the unemployment using series observed with the labour force survey (LFS). We extend the model by including auxiliary series about job search behaviour from Google Trends and claimant counts, partially observed at higher frequencies. Our factor model allows for nowcasting the variable of interest, providing reliable unemployment estimates in real time before LFS data become available.",
keywords = "high-dimensional data analysis, state space, factor models, nowcasting, unemployment, Google Trends",
author = "Caterina Schiavoni and Franz Palm and Stephan Smeekes and {van den Brakel}, Jan",
year = "2019",
month = "1",
day = "31",
language = "English",
publisher = "arXiv.org at Cornell University Library",
type = "WorkingPaper",
institution = "arXiv.org at Cornell University Library",

}

TY - UNPB

T1 - A dynamic factor model approach to incorporate Big Data in state space models for official statistics

AU - Schiavoni, Caterina

AU - Palm, Franz

AU - Smeekes, Stephan

AU - van den Brakel, Jan

PY - 2019/1/31

Y1 - 2019/1/31

N2 - In this paper we consider estimation of unobserved components in state space models using a dynamic factor approach to incorporate auxiliary information from high-dimensional data sources. We apply the methodology to unemployment estimation as done by Statistics Netherlands, who uses a multivariate state space model to produce monthly figures for the unemployment using series observed with the labour force survey (LFS). We extend the model by including auxiliary series about job search behaviour from Google Trends and claimant counts, partially observed at higher frequencies. Our factor model allows for nowcasting the variable of interest, providing reliable unemployment estimates in real time before LFS data become available.

AB - In this paper we consider estimation of unobserved components in state space models using a dynamic factor approach to incorporate auxiliary information from high-dimensional data sources. We apply the methodology to unemployment estimation as done by Statistics Netherlands, who uses a multivariate state space model to produce monthly figures for the unemployment using series observed with the labour force survey (LFS). We extend the model by including auxiliary series about job search behaviour from Google Trends and claimant counts, partially observed at higher frequencies. Our factor model allows for nowcasting the variable of interest, providing reliable unemployment estimates in real time before LFS data become available.

KW - high-dimensional data analysis

KW - state space

KW - factor models

KW - nowcasting

KW - unemployment

KW - Google Trends

M3 - Working paper

BT - A dynamic factor model approach to incorporate Big Data in state space models for official statistics

PB - arXiv.org at Cornell University Library

ER -