Longitudinal Analysis of Trip Rates in Britain: Partial Effects of Migration, Technology, and Housing Costs

Longitudinal Analysis of Trip Rates in Britain: Partial Effects of Migration, Technology, and Housing Costs


Reza Tolouei, AECOM, Esra Suel, Imperial College London, Marina Triampela, AECOM


The effects of migration, housing costs, and technology on the observed trends in trip rates in Britain were studied.


The UK Department for Transport (DfT) commissioned a research study in early 2015 to investigate recent travel trends and identify the main drivers of these trends. Whilst trends in various travel demand measures were of interest, the main focus of the study was trip rates. The project team drew upon complementary AECOM and Imperial College London’s experience in undertaking this study, the outcome of which is the subject of this paper.
Recent analysis of Great Britain National Travel Survey (NTS) data suggests that average trip rates across all modes have decreased between 1988 and 2010 for the majority of trip purposes. An initial study by DfT showed that cross-sectional variation in trip rates by well-documented traditional factors such as car ownership, working status, age, sex, gross income, and area type, failed to explain the observed trends in trip rates.
A literature review was undertaken by the study team to identify possible drivers of observed trends in trip rates. Using the outcome of this, it was sought to focus on factors which may have materially influenced trip rates, for which there was good quality data over reasonable time series, which were likely to change and influence future trip rates, and where the influence was not fully understood in the literature. Accordingly, some key areas were identified to be the main focus of the study. These included income after allowing for housing costs, migration, technology, and possible change in under reporting of trips.
In summary, the following specific research questions were identified and addressed:
• How the existing NTEM segmentation could be extended to include more factors, providing greater forecasting accuracy?
• Does income after housing expenditures provide statistical salience in explaining trip rates beyond gross income, both cross-sectionally and trends over time?
• What is the effect of migration on trip rates and to what extent changes in migration can explain observed trends in trip rates?
• What is the effect of online shopping and internet use on trip rates and to what extent changes in these factors can explain observed trends in trip rates?
We investigated the above questions, separately for each trip purpose, through a disaggregate analysis of a cross-sectional time-series data. No single study dataset was identified to provide time-series information on trip rates and all the above factors. The standard NTS datasets (2002-2012) were therefore augmented with information on housing expenditure from the UK’s Living Costs and Food Survey and information on migration-status from the 2001 and 2011 Census.
Negative binomial regression models (and zero-inflated version of these models) were used to estimate trip rates, separately by trip purpose and working status categories. It was found that income after housing costs provides better statistical performance than gross income for the majority of the estimated models. Findings suggest that higher income is associated with higher levels of trip making for recreational and holiday purposes, with the suggestion that this may be at the expense of trip making for the purpose of visiting friends and relatives. It was found that migrants tend to make fewer recreational/social, visiting, and holiday trips. These results are consistent with the hypothesis that migrants are disconnected from their social network. Given recent and likely future levels of inbound migration to the UK, this result has potentially important policy implications. Internet use for shopping in general was found to be associated with substantially higher levels of non-work related trip making, especially for recreational and holiday purposes and correspondingly lower levels of commuting. When the effects are statistically significant, in most of the cases trip rates tend to be lower for those who do online shopping more frequently.
The findings from this study have been used to develop a trip rate forecasting tool, providing forecast trip rates for input into DfT’s National Trip-End Model (NTEM) in developing forecast scenarios.

Reza Tolouei, AECOM
Esra Suel, Imperial College London
Marina Triampela, AECOM
Scott Le Vine, Imperial College London
John Polak, Imperial College London
Paul Hanson, AECOM
Pawel Kucharski, Department for Transport


Association for European Transport