Meta-Analysis Confirms AV Deployment Will Increase VMT ~6%; Empty Miles Could Double Daily Travel (March 2026)

Headline Finding

A peer-reviewed meta-analysis published in March 2026, drawing on 26 earlier studies, found that widespread AV adoption is expected to increase vehicle miles travelled (VMT) by approximately 6% — even for shared autonomous vehicles. Non-shared AVs increase VMT by 6.9%; shared AVs by 5.33%. A separate comprehensive review of 48 studies (2019–2025) found that empty or “deadhead” VMT (“eVMT”) could as much as double the number of miles a city’s population drives in a day. Both studies confirm that without proactive policy intervention, AV deployment will generate net VMT growth, not reduction.

Detail

Meta-analysis 1 (University of Texas-Arlington, published via ScienceDirect, March 2026):

• Source: “Assessing automated vehicle-induced VMT: meta-analysis of current research,” published in ScienceDirect journal
• Method: Meta-analysis of 26 previously published studies
• Key finding: VMT increases ~6% across all AV scenarios; non-shared AVs increase VMT 6.9%; shared AVs increase VMT 5.33%
• Even pooling implies that the widely-held assumption that shared AV fleets will reduce VMT is not supported — sharing reduces but does not eliminate VMT growth
• Reported by Streetsblog USA (March 5, 2026) and Planetizen (March 2026)

Meta-analysis 2 (comprehensive literature review, 2026):

• Review of 48 studies published 2019–2025
• Key finding: empty vehicle miles (“eVMT”) — deadhead miles while driverless vehicles reposition to pick up passengers — could “as much as double the number of miles an average city population drives in a day”
• Drivers of VMT growth identified: reduced perceived travel time cost, induced demand from new user groups (elderly, non-drivers), modal shift away from transit, and eVMT
• Policy interventions that can mitigate or reverse VMT increases: road pricing, transit integration, and ride-sharing mandates

Combined implication: VMT growth is consistent across the literature. The question is not whether AVs increase VMT but by how much and through which mechanisms. eVMT is the largest single contributor and is directly controllable by policy (idle charging zones, depot requirements, dynamic road pricing on empty repositioning trips).

Relationship to existing evidence: This meta-analysis extends EV23 (VMT Induced Demand) with more current data and a specific eVMT finding. The 2026 studies are the most up-to-date literature available.

Governance Implication

The eVMT finding is the most governance-relevant element. Two implications for the CAM paper:

1. eVMT as a city-controllable variable: Cities can directly regulate eVMT by: (a) requiring vehicles to park at licensed depots between rides rather than cruise; (b) pricing empty repositioning trips at peak rates via congestion/distance pricing; (c) requiring fleet operators to minimise repositioning distance as a licence condition. The meta-analysis provides the quantitative basis for arguing that eVMT regulation is not speculative — it addresses a documented and quantifiable harm.

2. Demand failure mode is now better evidenced: The paper’s Demand failure mode chapter should cite the March 2026 meta-analysis as the latest and most methodologically rigorous source for the VMT growth claim. The 5–7% VMT increase range, and the potential doubling of daily city travel from eVMT, are powerful numbers.

Sources

• “Assessing automated vehicle-induced VMT: meta-analysis of current research,” ScienceDirect (Transportation Research Part D or equivalent), published February–March 2026
• “Study: AVs Will Super-Charge VMT,” Streetsblog USA, March 5, 2026
• “Study: Self-driving Cars Could Raise VMT by 6%, Increasing Congestion,” Planetizen, March 2026
• “Autonomous Vehicles Could Create More Gridlock — Study Finds,” Auto Connected Car News, March 2026
• “The Rebound Effect of Autonomous Vehicles on Vehicle Miles Traveled,” MDPI Sustainability, 2025 (background study)