Modeling Single Occupant Vehicle Behavior in High-Occupancy Toll (HOT) Facilities

Goodall, Noah J.; Smith, Brian · 2009 · ROSA P / University of Virginia. Center for Transportation Studies

archive: archived pipeline: cataloged verified

Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)

Summary

This study investigates the behavior of single-occupant vehicle (SOV) drivers in High-Occupancy Toll (HOT) lanes, specifically examining how dynamic toll rates and real-time traffic conditions influence lane choice. The research addresses a critical gap in transportation planning: while HOT facilities are increasingly deployed as active traffic management tools, there is limited empirical evidence regarding how drivers actually respond to minute-to-minute changes in pricing and congestion. The authors aim to develop a predictive model for SOV behavior to assist operators in designing effective dynamic tolling algorithms and to inform the feasibility of future HOT projects. The methodology relies on empirical data from the I-394 MnPASS facility in Minneapolis, Minnesota, one of the few HOT systems utilizing dynamic pricing. The study analyzed 27,831 records of three-minute intervals collected over a 16-month period. Data sources included traffic detector stations measuring speed and volume in both HOT and general-purpose (GP) lanes, as well as tolling databases recording transaction times and rates. The authors calculated the "cost per hour of travel time saved" for SOVs, adjusting for scenarios where GP lanes operated at free-flow speeds. To isolate price-sensitive behavior, the analysis first identified "regular" users who utilized the HOT lanes even when no time advantage existed (GP speeds > 55 mph). These baseline usage rates were subtracted from the total data to isolate the behavior of drivers responding to real-time congestion and pricing. The results reveal that SOV behavior is dominated by two distinct groups. Approximately 87.5% of SOV users are "regular" drivers who use the HOT lanes at predictable rates throughout the morning peak, regardless of toll levels or travel time advantages. The remaining "price-sensitive" drivers utilize the lanes more frequently when the cost per hour of time saved is lowest. A predictive model incorporating time savings, time of day, and toll rates achieved an R² value of 0.684. Notably, a model assuming all users are regular drivers (ignoring price sensitivity) achieved a nearly identical R² of 0.675. This indicates that the dynamic pricing structure at the MnPASS facility had a negligible influence on aggregate driver behavior. The significance of these findings lies in their challenge to the premise that dynamic pricing is an effective tool for managing HOT lane congestion. Since the majority of users are insensitive to price fluctuations, the tolling mechanism fails to significantly restrict SOV entry during high-demand periods. This suggests that HOT facilities may not function as effectively as intended for active traffic management, implying that planners and operators must reconsider the reliance on pricing alone to maintain service levels. The study provides essential empirical evidence for refining driver behavior models and evaluating the operational viability of future HOT lane implementations.

Key finding

Eighty-seven point five percent of single-occupant vehicle users utilized the HOT lanes at predictable rates regardless of travel time advantages, indicating that the dynamic pricing structure had a negligible influence on overall driver behavior.

Methodology

dataset

Sample size: 27831

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• traffic density