Safety

Safer for everyone on the road, starting with drivers

Safety is the reason for this project. Most of the crashes a road diet prevents are between cars: rear-ends, sideswipes and left turns. The people walking and biking to our schools benefit too.

Why this road crashes

Boston Post Road squeezes four travel lanes into a width that barely fits them. Each lane is about 9.5 feet wide, where the standard is 11. That leaves little room for error.

Then add the turns. Driveways, side streets and parking spaces line the corridor, and every driver waiting to turn left stops in a travel lane. The drivers behind brake hard or swerve into the next lane, and on 9.5-foot lanes both go wrong often.

The four crashes a road diet is built to prevent

A road diet is one of the most effective safety countermeasures available for a road like this one, and this is why: it goes after the exact crash types FHWA reports falling most when four lanes become three. Each one below is drawn as it happens today, and again under the new layout. Fire trucks and ambulances have their own page.

What the research says

19–47%

fewer crashes overall

After four-lane roads were converted to three, in studies the Federal Highway Administration cites.

Proven

safety countermeasure

FHWA's label for safety treatments with a proven record of reducing crashes.

3 kinds

of crash drop most

Rear-end, left-turn and sideswipe: the exact crashes that four narrow lanes and constant left turns produce.

The evidence keeps getting stronger

Road diets have been studied for decades, and the research has not stood still. The national estimate everyone quotes came from 45 sites. The newest work covers more roads and more recent years, and it separates the crashes that hurt people from the ones that only bend metal. It finds bigger reductions, not smaller.

Cover of the FHWA Road Diet Informational Guide, with photographs of three-lane roads.

2014 · Federal Highway Administration

The national guide: 19 to 47% fewer crashes

FHWA gathered the research into one guide and put the headline on page one: converting a four-lane undivided road to three lanes cuts overall crashes by 19 to 47 percent. The range comes from a 2010 FHWA study of 45 sites using the empirical Bayes method, which corrects for changes in traffic and for roads picked right after a bad year. Fifteen small-town sites in Iowa came in at 47%, thirty suburban sites in California and Washington at 19%, and all 45 together at 29%.

−29%total crashes, all 45 sites
45sites studied
Transportation Research Record Safety Effectiveness of the Road Diet Treatment in Rhode Island Y. Zhou, S. Himes, T. Le,
J. Gooch, K. Northup,
P. Pasvao
2022

2022 · Rhode Island

A Northeastern state with roads like ours, and a significant drop in crashes

Rhode Island’s DOT wanted to know whether road diets were worth adopting across the state, so it had the crashes measured on its own roads — empirical Bayes again, compared against untreated four-lane undivided roads with similar characteristics. Total crashes fell 29%. The crashes that killed or injured someone fell 37%. The roads it studied are the closest match to ours in any of this research: four-lane undivided streets carrying 3,390 to 22,500 vehicles a day, averaging 13,459 — squarely inside the 10,000 to 15,000 Boston Post Road carries.

−29%all crashes
−37%crashes that killed or injured
13,459average traffic on the roads studied
Transportation Research Record Development of Road Diet Segment and Intersection Crash Modification Factors Linda Lim
Michael D. Fontaine
2022

2022 · Virginia

A large, long-term study, and a significant reduction

Researchers at Virginia's transportation research council measured crashes before and after at 36 stretches of road and 39 intersections converted between 2009 and 2018, by the same method. Total crashes fell by about a third. The crashes that killed or injured someone fell by far more, which is the number that matters on a road where five people have died. Every result was statistically significant, and the roads studied carried 2,600 to 17,000 vehicles a day — the band Boston Post Road sits in.

−38%all crashes, along the road
−64%crashes that killed or injured
−35%all crashes, at intersections
−46%killed or injured, at intersections

Where it has been built

Sources
  1. FHWA, Road Diet Informational Guide, ch. 2
    Four-to-three conversions: 19–47% fewer total crashes, including rear-end, left-turn and sideswipe. The range is from FHWA’s 2010 empirical Bayes study of 45 sites: 47% at 15 small-town Iowa sites, 19% at 30 suburban sites in California and Washington, 29% for all 45 combined. Published as the Road Diet Informational Guide, FHWA-SA-14-028, November 2014.
    highways.dot.gov
  2. Zhou, Himes, Le, Gooch, Northup and Pasvao, “Safety Effectiveness of the Road Diet Treatment in Rhode Island,” Transportation Research Record, 2022
    Empirical Bayes before-and-after evaluation for the Rhode Island Department of Transportation, against reference groups of untreated four-lane undivided roads with similar characteristics. Total crashes −29% (CMF 0.71); fatal and injury crashes −37% (CMF 0.63). Treatment sites were four-lane undivided roads carrying 3,390 to 22,500 vehicles a day, averaging 13,459, using crash data from 2011 to 2019. Carried out to help the department decide a statewide direction for the countermeasure.
    Transportation Research Record
  3. NYSDOT traffic counts, Boston Post Road
    Average annual daily traffic at the most recent count (2024).
    The 45-year trend
  4. Lim and Fontaine, “Development of Road Diet Segment and Intersection Crash Modification Factors,” Transportation Research Record, 2022
    Empirical Bayes before-and-after study of 36 segments and 39 intersections converted from four lanes to three in Virginia, 2009–2018. Segments: total crashes −38%, fatal and injury crashes −64%. Intersections: total −35%, fatal and injury −46%. All results statistically significant. Study sites carried 2,600–17,000 vehicles a day.
    Transportation Research Record