Zero-truncated panel Poisson mixture models

Estimating the impact on tourism benefits in Fukushima Prefecture

Masaki Narukawa, Katsuhito Nohara

Research output: Contribution to journalArticle

Abstract

This study proposes an estimation approach to panel count data, truncated at zero, in order to apply a contingent behavior travel cost method to revealed and stated preference data collected via a web-based survey. We develop zero-truncated panel Poisson mixture models by focusing on respondents who visited a site. In addition, we introduce an inverse Gaussian distribution to unobserved individual heterogeneity as an alternative to a popular gamma distribution, making it possible to capture effectively the long tail typically observed in trip data. We apply the proposed method to estimate the impact on tourism benefits in Fukushima Prefecture as a result of the Fukushima Nuclear Power Plant No. 1 accident.

Original languageEnglish
Pages (from-to)238-246
Number of pages9
JournalJournal of Environmental Management
Volume211
DOIs
Publication statusPublished - Apr 1 2018

Fingerprint

Gaussian distribution
Nuclear power plants
Accidents
tourism
panel data
nuclear power plant
accident
Costs
distribution
method
travel cost

Keywords

  • Contingent behavior
  • Count data model
  • Inverse Gaussian distribution
  • Random-effects model
  • Travel cost method
  • Zero truncation

ASJC Scopus subject areas

  • Environmental Engineering
  • Waste Management and Disposal
  • Management, Monitoring, Policy and Law

Cite this

Zero-truncated panel Poisson mixture models : Estimating the impact on tourism benefits in Fukushima Prefecture. / Narukawa, Masaki; Nohara, Katsuhito.

In: Journal of Environmental Management, Vol. 211, 01.04.2018, p. 238-246.

Research output: Contribution to journalArticle

@article{1ae0391690fc4841bbf11572da537799,
title = "Zero-truncated panel Poisson mixture models: Estimating the impact on tourism benefits in Fukushima Prefecture",
abstract = "This study proposes an estimation approach to panel count data, truncated at zero, in order to apply a contingent behavior travel cost method to revealed and stated preference data collected via a web-based survey. We develop zero-truncated panel Poisson mixture models by focusing on respondents who visited a site. In addition, we introduce an inverse Gaussian distribution to unobserved individual heterogeneity as an alternative to a popular gamma distribution, making it possible to capture effectively the long tail typically observed in trip data. We apply the proposed method to estimate the impact on tourism benefits in Fukushima Prefecture as a result of the Fukushima Nuclear Power Plant No. 1 accident.",
keywords = "Contingent behavior, Count data model, Inverse Gaussian distribution, Random-effects model, Travel cost method, Zero truncation",
author = "Masaki Narukawa and Katsuhito Nohara",
year = "2018",
month = "4",
day = "1",
doi = "10.1016/j.jenvman.2017.12.082",
language = "English",
volume = "211",
pages = "238--246",
journal = "Journal of Environmental Management",
issn = "0301-4797",
publisher = "Academic Press Inc.",

}

TY - JOUR

T1 - Zero-truncated panel Poisson mixture models

T2 - Estimating the impact on tourism benefits in Fukushima Prefecture

AU - Narukawa, Masaki

AU - Nohara, Katsuhito

PY - 2018/4/1

Y1 - 2018/4/1

N2 - This study proposes an estimation approach to panel count data, truncated at zero, in order to apply a contingent behavior travel cost method to revealed and stated preference data collected via a web-based survey. We develop zero-truncated panel Poisson mixture models by focusing on respondents who visited a site. In addition, we introduce an inverse Gaussian distribution to unobserved individual heterogeneity as an alternative to a popular gamma distribution, making it possible to capture effectively the long tail typically observed in trip data. We apply the proposed method to estimate the impact on tourism benefits in Fukushima Prefecture as a result of the Fukushima Nuclear Power Plant No. 1 accident.

AB - This study proposes an estimation approach to panel count data, truncated at zero, in order to apply a contingent behavior travel cost method to revealed and stated preference data collected via a web-based survey. We develop zero-truncated panel Poisson mixture models by focusing on respondents who visited a site. In addition, we introduce an inverse Gaussian distribution to unobserved individual heterogeneity as an alternative to a popular gamma distribution, making it possible to capture effectively the long tail typically observed in trip data. We apply the proposed method to estimate the impact on tourism benefits in Fukushima Prefecture as a result of the Fukushima Nuclear Power Plant No. 1 accident.

KW - Contingent behavior

KW - Count data model

KW - Inverse Gaussian distribution

KW - Random-effects model

KW - Travel cost method

KW - Zero truncation

UR - http://www.scopus.com/inward/record.url?scp=85041555251&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85041555251&partnerID=8YFLogxK

U2 - 10.1016/j.jenvman.2017.12.082

DO - 10.1016/j.jenvman.2017.12.082

M3 - Article

VL - 211

SP - 238

EP - 246

JO - Journal of Environmental Management

JF - Journal of Environmental Management

SN - 0301-4797

ER -