• Parking search strategies and parking location choices are modeled within the same framework in unregulated urban parking environments. • Explainable machine learning is used to identify the factors which influence parking search behavior. • Discrete choice models quantify the trade-off between search times, cost, and walking distance for car drivers. • Search time is identified as the most significant factor in influencing parking search strategies and location choice. • Walking distance is found to be statistically insignificant compared to search time and parking cost in parking location decisions. • The findings support demand-responsive pricing and policies that reduce inefficient cruising for parking. This paper examines on-street searching for parking strategies and parking location choices using a stated-preference survey and two modeling approaches: interpretable machine learning models and discrete choice models. While most existing studies focus on regulated or priced parking environments, limited work has analyzed how drivers search for parking together with how they choose a parking location in largely unregulated contexts. The aim of this study is to address this gap by identifying the factors that influence how drivers search for parking and how they trade off search time, walking distance, and parking cost when choosing where to park. The best machine learning model developed achieves an accuracy of 84% in classifying search strategies. Feature importance analysis shows that average parking search duration, parking distance from destination, and vehicle size are significant factors influencing parking search strategies. In the parking location choice models, search duration and parking cost have statistically significant negative effects on choice probability, while walking distance is not statistically significant. Behavioral metrics, including elasticities and willingness-to-pay, reveal that drivers place greater value on reducing search time (€2.54 per minute) than on reducing walking distance. These results support policies that prioritize reducing search effort, such as demand-responsive pricing, residential parking policies, and measures that improve the balance between overutilized on-street parking and underutilized off-street facilities.
Vakrinou et al. (Mon,) studied this question.