Video summary

Mexico City has a lot to teach the world about contraflow bus lanes

Main summary

Key takeaways

Educational

Main ideas / concepts / lessons

  • Contraflow (counterflow) bus lanes can significantly improve bus service on one-way streets by allowing buses to travel opposite to general traffic.
  • The video challenges an earlier claim (from an old 1980s study) that buses are faster by about ~20% in counterflow lanes, and tests whether similar improvements hold with expanded/better data.
  • Curb-running bus lanes are common but often problematic because the curb space has other competing uses, including:
    • turning cars (right turns),
    • stopping trucks and deliveries,
    • space needs for bikes and pedestrians.
  • Contraflow lanes are presented as a curb-lane “upgrade” (generally cheaper and easier than center-median bus lanes), but they require street and signal adjustments and may face driver unfamiliarity.

Methodology / study approach (detailed steps)

1) Start from prior research and form a key hypothesis

  • An earlier (1980s) study reportedly found buses move ~20% faster in lanes that run opposite traffic (counterflow).
  • Core questions:
    • Do contraflow lanes really reduce travel time?
    • Why aren’t they standard, especially on one-way streets?

2) Attempt a quick replication in Mexico City (initial test)

  • Location: Mexico City, during filming of earlier transit-related videos.
  • Route: Trolleybus line 1
    • A long route with a continuous contraflow lane through the urban center.
    • An adjoining with-flow curb lane on the same street for comparison.
  • Data collection:
    • One weekday, PM peak hour (5:00–8:00 p.m.)
    • No special events; date was Dec 18 (week before Christmas; no school).
  • Sample size: 13 bus trips
  • Result: contraflow lane buses were ~14% faster than with-flow in this limited snapshot.

3) Expand into a more robust study with strict selection criteria

The study targeted contraflow lanes meeting three requirements:

  • Comparable route: the same road has one contraflow and one with-flow lane for comparison (ideally both directions have frequent buses).
  • Sizable route: long enough to matter (not just a short connector).
  • Data access: reliable arrival/departure timing data (real-time preferred).

4) Identify study locations and handle data constraints

  • Mexico City
    • Became the main source due to extensive contraflow usage.
    • Constraint: no real-time data, so data was collected manually in person.
  • Outside Mexico City
    • Found two London-area examples (“Fake London” and “London London”):
      • Both had real-time GTFS data, enabling automation via Python.

5) Data collection plan across routes

Mexico City (manual, camera-assisted)

  • STE line 1 (north–south)
  • STE line 2 (east–west)
  • Collected AM and PM peak hour data across multiple weekdays
  • Used a camera placed at one end of the study zone to capture movement/travel-time information.

London examples (automated, real-time GTFS via Python)

  • Used a Python script to collect travel times along corridors.
  • Data collection lasted for several weeks.

6) Analyze travel-time effects across time periods

  • Compare contraflow vs with-flow performance in:
    • AM peak
    • PM peak
  • Identify which periods show consistent improvement.

7) Analyze reliability as a secondary (co-benefit) outcome

Reliability was assessed in two ways:

  1. Headway adherence / waiting-time likelihood

    • Measures: likelihood of having a bus arrive within a few minutes of expected headways (headway-based punctuality).
    • Also examines distribution patterns to address:
      • bus bunching (clusters arriving too close together),
      • vs a more “tall, narrow” distribution near target intervals.

  2. Travel time variability through the day (consistency)

    • Used non-peak range data in London examples to examine fluctuations across service hours.
    • Compared range/variability between contraflow and with-flow operation.

8) Conclude and interpret both travel time + reliability together

  • The video argues that even if travel time savings are slightly less than the “20%” claim, reliability improvements amplify the overall service benefit.

Results / findings (with key numbers)

Travel time savings (contraflow vs with-flow curb lanes)

Overall framing:

  • Contraflow lanes are generally faster, though not uniformly in every period/route.

Key results:

  • Fake London
    • Contraflow faster in AM by ~13%
    • Contraflow faster in PM by ~19%
  • London London
    • Contraflow faster in AM by ~15%
    • Contraflow faster in PM by ~34%
  • Mexico City STE line 2
    • Contraflow faster in AM by ~10%
    • Contraflow faster in PM by ~12%
  • Mexico City STE line 1
    • Contraflow faster in PM by ~6%
    • Contraflow slightly slower in AM by ~3%

Summary across periods:

  • Contraflow lanes were faster in 7 out of 8 studied AM/PM periods.
  • The video’s conservative estimate: average time savings ~10–20%, later refined toward ~15% on average.

Reliability improvements (key “co-benefit”)

  1. Better likelihood of buses arriving near expected times

    • Mexico City line 1: passengers ~11% more likely to have a bus within a few minutes of expected headways.
    • Mexico City line 2: passengers ~5% more likely.

  2. Reduced bus bunching / improved headway distribution

    • Contraflow helps shift the frequency distribution closer to the desired pattern (near target headways), reducing:
      • long gaps, and
      • clustering/bunching.

  3. More consistent travel times throughout the day

    • London UK:
      • Contraflow fluctuation: ~30 seconds
      • With-flow fluctuation: ~4× that amount
    • Fake London:
      • Similar pattern, especially during the PM period.

Interpretation:

  • Reliability gains may be small per trip, but across many trips over time they can accumulate into meaningful system-level improvements.

Practical implications / recommendations

  • Contraflow lanes should be the default for new curb-running bus lanes on one-way streets.
  • Safety/operational concerns exist, but the video suggests they can be addressed through:
    • good signage and education,
    • street and signal reconfiguration,
    • restricting left turns for cars,
    • understanding local driver behavior (including likely rule-breaking where it occurs).
  • The video recommends pilot testing rather than prolonged debate because contraflow lanes are described as cheap to install and uninstall, enabling cities to run experiments on selected corridors using measurable outcomes.

Potential downsides mentioned

  • Driver confusion about vehicles traveling the “wrong” way.
  • Possible safety concerns due to unexpected opposing-direction traffic.
  • Requires operational changes beyond standard curb bus lanes, such as:
    • traffic signal retiming,
    • restricting left turns,
    • street redesign elements.

Speakers / sources featured (as named or referenced)

  • Narrator / channel host: speaker in the video (no name provided in subtitles)
  • Transit World: channel/brand referenced (“Welcome to Transit World…”)
  • Prior research source: an old 1980s Chicago study (not further specified)
  • Cities/routes used as sources/examples
    • Mexico City: STE line 1 and STE line 2
    • “Fake London”: ~2 km contraflow installation with GTFS real-time data
    • “London London”: near Elephant & Castle, ~0.6 km, with GTFS real-time data
  • GTFS real-time: data standard referenced for London examples
  • Python script: tool used to extract travel times (no specific author named)

Original video