Regional Land Transport Demand Model

Technical notes for practitioners

June 2024

June 2024 Publication Dr Eilya TorshizianAlina Fehling

Abstract

The National Land Transport Demand Model (NLTDM) was developed between 2011 and 2012. It was intended to be a tool for considering how transport demand may evolve, rather than a tool for providing point estimates of demand.

The model takes a hybrid approach to forecasting transport demand. This simplifies the relationship between transport demand and macroeconomic aggregates. However, it combines top-down relationships with additional details of behavioural parameters and often includes reduced (that is, simplified) forms of conventional regional transport models.

With policy discussions focusing more on equity, emissions and inclusion, there is an increasing demand for more flexible policy models and tools. The Regional Land Transport Demand Model (RLTDM) provides a useful framework for current policy discussions. However, it was developed using the MATLAB software package, which is rarely used by practitioners. New Zealand Transport Agency Waka Kotahi (NZTA) commissioned Principal Economics Limited to recode the RLTDM in Stata, using its matrix language (Mata), and provide further notes for practitioners on how to use the model.

The hybrid approach means the NLTDM can be manipulated by people with differing degrees of modelling expertise. This enables researchers and policy advisors to further investigate transport-demand factors, which is useful given the inherent uncertainty with transport modelling. The model evaluates transport-demand scenarios 30 years into the future, taking account of mega-trends in:

  • population growth
  • spatial demography
  • technology
  • income and economic growth
  • industrial composition
  • policy and prices in relevant areas

This report provides a technical tagline: of the Regional Land Transport Demand Model (RLTDM), which is a hybrid approach to forecasting transport demand across New Zealand regions. The model’s outputs include deterministic and stochastic forecasts of a wide range of economic and transport series. We re-coded the model in Stata and Mata.