Page 107 - Contributed Paper Session (CPS) - Volume 4

CPS2135 Sumonkanti Das et al.

Multilevel time series modeling of mobility
trends in the Netherlands for small domains
2
1
1
Sumonkanti Das , Harm Jan Boonstra , Jan van den Brakel
1 Maastricht University
2 Statistics Netherlands

Abstract
The purpose of the Dutch Travel Survey is to produce reliable figures about
mobility of the Dutch population. In this paper, multilevel time-series models
have been developed to estimate reliable mobility trends at several
aggregation levels, accounting for discontinuities induced by two different
redesigns, and outliers due to less reliable outcomes in one particular year.
The model is fitted to annual input series of direct estimates and standard
errors at the most detailed breakdown into 504 domains defined by the
combination of sex, age-class, motive and mode for the period 1999-2017.
The standard errors of the direct estimates are smoothed through Generalized
Variance Function (GVF) method. The model is fitted in a hierarchical Bayesian
framework using Markov Chain Monte Carlo (MCMC) simulations. Global-local
priors are considered for regularization purposes. Predictions for higher
aggregation levels are obtained by aggregation of the most detailed domain
predictions, resulting in numerically consistent set of trend estimates.

Keywords
Generalized variance function; Global-local priors; Hierarchical Bayesian
approach; MCMC simulation; Small area estimation; Survey redesigns

1. Introduction
The purpose of the Dutch Travel Survey (DTS) is to produce reliable
prediction of mobility trends (such as average distance per journey) of the
Dutch population. The target variable in this paper is the average number of
journey parts per person per day (pppd). A journey with a specific motive (e.g.
traveling to work) can be made by more than one transportation mode. In
such case, journey parts are defined as the breakdowns of the journey for a
specific motive into separate sections made by the transportation modes. Thus
journey parts are characterized by journey motive and transportation mode.
In this study, the target parameter is estimated at the most detailed level based
on the cross-classification of sex (male, female), age-class (0-5, 6-11, 12-17,
18-29, 30-39, 40-49, 50-59, 60-69, 70+), motive (work, shopping, education,
other), and mode (car driver, car passenger, train, BTM (bus/tram/metro),
cycling, walking, other). Annual direct estimates and their standard errors for

96 | I S I W S C 2 0 1 9