Abstract
Efforts to promote cycling generally fall into two categories: hard measures, such as new infrastructure and traffic management, and soft measures, such as education and training. While the effectiveness of each approach has been studied individually, the interaction between the two remains largely unexplored. This study provides the first empirical test of whether hard infrastructure measures modify the effectiveness of soft behavioural interventions, analysing the impact of Low Traffic Neighbourhoods (LTNs) alongside adult cycle training in Lon-don. We operationalise effectiveness as the proportion of trainees who transition to regular cycling (at least once per week), measured via pre-and post-training surveys. Using a Difference-inDifferences (DiD) approach, we analyse longitudinal survey responses from 1,470 adult trainees trained across two cohort blocks, 2017-2019 (pre-LTN implementation) and 2022-2024 (post-LTN implementation), categorised into three spatial classes: inside an LTN, in a borough with LTNs but outside the specific zone, and in a borough without LTNs. The DiD estimator reveals no statistically significant interaction effect (DiD estimate: +4.15 percentage points, OR = 1.22, 95% CI: 0.62–2.37, p = 0.564), with wide confidence intervals indicating the study was underpowered to detect modest 1 to moderate effects. This suggests that among highly motivated adult trainees, the presence or absence of LTN infrastructure does not meaningfully alter the impact of cycle training. The findings speak to this specific, self-selected population and do not address the broader relationship between LTN provision and cycling uptake more generally.