Impact of On-Street Parking Pricing on Public Transport Use
Parking pricing is used to manage scarce on-street spaces and can influence mode choice, yet its potential to shift travel from private cars to public transport depends on local conditions and individual preferences. This study investigates possible responses to changes in Thessaloniki’s controlled on-street parking charges, focusing on visitor-type trips to the city centre and the potential transition from a fixed hourly fee to dynamic pricing.
An online stated-preference survey conducted in December 2023 collected responses from 208 individuals. Each respondent considered six hypothetical car-versus-bus travel scenarios, yielding 1,248 choice observations. Scenarios varied parking charges and parking-search times while incorporating travel times and costs tailored to respondents’ areas of residence. They assumed a typical weekday afternoon trip with an approximately two-hour stay in the city centre. At the time of the survey, Thessaloniki’s public transport network consisted of buses, before the metro began operations.
A mixed-logit model was developed to account for differences in travellers’ sensitivity to parking charges and relative travel times and costs. The model also incorporated existing frequencies of car use, public transport use and controlled-parking use. Results indicated a pre-existing preference for private cars and a strong relationship between established travel habits and stated choices. Cost-related preferences displayed considerable heterogeneity. Survey participants generally viewed €1.50 per hour as a normal parking charge, while approximately €1.75–€2.00 per hour was perceived as expensive.
Four hourly pricing scenarios (€1.00, €1.70, €2.00 and €2.50) were subsequently assessed with model-based predictions using a synthetic dataset. Indicative absolute modal shares, calibrated to Thessaloniki’s strategic traffic model for city-centre trips unrelated to work or education, suggest that public transport share could increase from 31.0% at the €1.70 baseline to 33.8% at €2.00 and 37.9% at €2.50.
The study offers evidence to support parking-pricing decisions while highlighting distributional sensitivities and the importance of complementary public transport improvements. The findings are based on hypothetical choices and modelled scenarios, not on observed effects following implementation of dynamic prices.
Supporting evidence
The study analysed 208 online survey responses collected in Thessaloniki in December 2023, generating 1,248 stated-preference choice observations. Each participant considered six scenarios comparing car and bus travel to the city centre. Parking prices, parking-search times and relative travel costs were assessed using a mixed-logit model that incorporated prior mobility habits and individual variation in cost sensitivity.
In the stated-choice scenarios, public transport was selected by 22.6% at a two-hour parking charge of €2 and a five-minute search time, compared with 59.6% at €5.50 and two minutes of searching. The mixed-logit model achieved an adjusted pseudo-R² of 34.82%. Monthly public transport use was positively associated with selecting buses (β=3.1542, p<0.001), while monthly controlled-parking use was negatively associated with the public transport alternative (β=−1.6617, p=0.0302).
For one-hour parking, modelled public transport share rose from 34% at €1.70 to 40% at €2.50. Indicative absolute shares for city-centre non-work/non-education trips rose from 31.0% at €1.70 to 33.8% at €2.00 and 37.9% at €2.50. The mean estimated value of travel time was €7.20/hour, compared with a median of €3.49/hour. These figures describe stated preferences and scenario predictions, not actual ridership changes.
Key findings
- Parking prices are associated with stated willingness to use public transport. In the modelled one-hour parking scenarios, public transport’s share of car-versus-bus choices increases from 34% at €1.70 to 40% at €2.50. Calibrated estimates for non-work/non-education city-centre trips place the public transport share at 31.0% under current pricing, 33.8% at €2.00 and 37.9% at €2.50. These are indicative scenarios rather than verified policy outcomes.
- Existing mobility habits strongly shape responses. Respondents who already use public transport at least monthly were more likely to choose it in the experiment. Frequent users of controlled parking were more inclined toward the private car. This suggests pricing policies should be analysed alongside established travel patterns, not solely through average fare sensitivity.
- Cost sensitivity differs substantially between travellers. Mixed-logit estimates show heterogeneity in preferences related to parking and travel costs. The value of travel time has a mean of €7.20/hour and a median of €3.49/hour, illustrating a skewed distribution. Uniform prices may therefore affect population groups differently, raising affordability and equity considerations.
- Parking duration matters for policy design. At €2.50/hour, the estimated public transport share in the two-mode comparison is 31% for a 30-minute stay, 40% for one hour and 55% for two hours. Charges differentiated by duration may consequently be relevant when defining alternatives to a fixed hourly tariff, although these estimates do not establish the effects of a real-time dynamic-pricing system.
- Pricing should be combined with complementary measures and further evaluation. Better public transport services and targeted affordability protections may help address unintended consequences. The study considered car and bus choices for a fixed destination, omitting other modes and potential destination changes; its survey preceded the metro opening. Broader samples, updated travel conditions and post-implementation monitoring are needed before applying the predictions as citywide forecasts.
