Fusing household travel surveys and external data for impact assessment under complex surrounding conditions: the case of introducing fare free public transport during the pandemic

In 2020, the German city of Monheim am Rhein implemented fare-free public transport (FFPT) amid the Covid pandemic, necessitating an impact evaluation despite the strong impact of Covid on mobility behaviour. Evaluating such transport policy measures typically relies on panel or cross-sectional surveys conducted before and after implementation. However, an important requirement of this approach is that other conditions remain largely unchanged. Distinguishing treatment from control groups to assess the influence of further factors is often difficult in comprehensive transport policies. To address the challenge of differentiating between the impact of FFPT and the impact of Covid on mobility behaviour, we utilized additional data sources alongside a representative household survey conducted from 2018 to 2023. The study faced significant hurdles due to both the pandemic and ongoing construction affecting public transport during the second year of FFPT. We integrated various external data sources to contextualize the household travel survey results: daily Covid incidence rates in North Rhine-Westphalia, mobile phone travel demand changes compared to 2019, nationwide travel surveys (including the German Mobility Panel), local passenger counts, and quarterly national public transport figures. Our multi-method approach aimed to disentangle the effects of external factors on travel behaviour changes —particularly Covid—from those attributable to FFPT. Descriptively, household survey results were contextualized using external data trends. Quantitative evaluations allowed for a comparison between local ridership data in Monheim am Rhein with national averages post-FFPT introduction. Multivariate analyses indicated that day-specific Covid incidences were not effective proxies for pandemic-related travel demand changes; thus, we focused on mobile phone data for more accurate insights, which also had its limitations. Ultimately, our approach allowed for a robust evaluation of FFPT by effectively combining various data sources with household surveys, mitigating uncertainties related to Covid impacts and offering transferable insights for future mobility survey evaluations.