Data Study

2018 School Safety Study

2018 School Safety Study
Share on
Key takeaways

Zendrive School Safety Snapshot:

Back to School 2018

How dangerous are the drivers around your kid’s school? Find out here.

Traffic crashes are the leading cause of death for all school-aged kids in the United States. Distracted driving is on the rise.

This is unacceptable.

To prevent these casualties, we need to prevent the driver behaviors that contribute to them, and that means we need to measure them. This is Zendrive’s specialty.

Find Your Kid’s School:

Here’s what we learned:

  • The worst hours to be a kid are pickup and dropoff times: 7-8 AM and 3-4 PM.
  • Driver behavior failed to improve around 90% of schools.
  • Driver behavior in cities and suburbs saw the worst decline from 2017-2018.
  • On average, city school roads are 4x more dangerous than rural school roads.
  • 9 of the 10 most dangerous schools are in rural areas near busy streets.

Fifty-seven million kids walk, bike, take the bus or get a ride to school everyday in the U.S. It’s a necessary journey, and sometimes it can be dangerous. At Zendrive, we believe traffic injuries and deaths are preventable. Our goal is to build safe streets and healthy communities, ones where every kid can get to school without worrying about the drivers around them.

Add your voice to the movement to make roads safe for everyone! Read our detailed analysis of driver behavior around schools here.


What does Zendrive do?

  • Zendrive’s mission is to make the roads safe through data and analytics. We use smartphone sensors to measure driver behavior in real time, which our machine-learning algorithms then turn into actionable safety insights for an individual, fleet, or road.
  • With over 150 Billion miles of driver behavior measured analyzed, Zendrive’s driver safety scores are 6x more predictive of future crashes than the industry standard. Our technology is used by fleets who want to measure and improve driving safety, consumer apps to offer collision detection and emergency response, and insurance apps to measure driver risk. Cities use Zendrive’s anonymized data to make city planning safer. Together, we can eliminate road collisions.

What did this study measure?

We mapped out areas around over 125,000 schools nationwide and analyzed over 10.5-billion miles driven within them, giving us a safety snapshot of April 2018, which we compared to data analyzed from April 2017. Since schools, their playgrounds, fields and community spaces are seven-day-a-week destinations, we looked traffic throughout the week; this spanned nine million drivers’ anonymized aggregated data.

What are the stats?

125,703 schools (and all surrounding roads within one-quarter of a mile)

3,094 counties (measured as averages of their school area scores)

10.5-billion miles driven and analyzed within school areas

9.1-million anonymous individual drivers in this study

1-billion trips

4.6-billion unsafe driving events

  • Phone use while driving (any handheld phone use while the vehicle’s moving, like texting or emailing)
  • Speeding
  • Rapid acceleration
  • Hard braking

Timespan analyzed:

  • April 2018
  • Monday, Tuesday, Wednesday, Thursday, Friday, Saturday, Sunday
  • 24-hours
  • “Morning peak” dropoff times are 7-8 am
  • “Afternoon peak” pickup times are 3-6 pm

What do the grades mean?

The grades measure the relative safety of the roads around a school. The more unsafe driving events per trip (like hard braking, phone fiddling, etc), the lower a grade the school will get.

The grades work on a curve. So when you put all 125,703 schools together, the top percentile, or the absolute safest school areas, gets an A+.

The next percentile is an A, and so on.

At the state- and county-levels, the grades were also given out on a curve. However, to account for outliers, only the 90th-percentile of driver behavior around schools within each area was used.

Why did my school not show up?

School areas with insufficient driving data were not included in the final analysis. As we get more data, we’ll release new Safety Snapshots with more info. The schools dataset is taken from the National Center of Education Statistics (NCES) datasets on public and private schools, found here. The NCES gathers their public school data from Common Core Data and private school data from the Private School Survey. If your school is not in the NCES database, then it won’t be in our analysis.

How did you calculate this?

School Areas

  • The schools dataset is taken from the National Center of Education Statistics (NCES) datasets on public and private schools, found here. The NCES gathers their public school data from Common Core Data and private school data from the Private School Survey.
  • Some religious schools, correctional facilities and other institutions that may only be open on the weekends; public learning centers, etc. were included in the study.
  • School area boundaries were defined as all road segments around a school’s location within a standard fixed geohash. Geohashes are equally sized grids on the earth’s surface. We used geohashes with the precision level of 6, which is equal to rectangular grids with size 1.2 km by 0.6 km, or about one-quarter of a mile.

Safety scoring

  • School safety scores and rankings are based on unsafe driving events detected by the Zendrive platform within school areas.
  • We looked at the number of trips that crossed through each school area and the number of events that occured within each school area, then we rated safety based on the number of events with the school area per trip that crossed through the school area.
  • Transportation infrastructure and traffic characteristics around schools – such as highway type, number of lanes, average number of vehicles travelling – were not used to determine Zendrive scores.
  • Unsafe driving behavior data was normalized by sample size weights to account for potential sample and selection biases.
  • Schools within the same boundary definition got the same ranking.

Have any more data to share?

Here are the 2018 Gold Star Schools, which had zero risky driving events during the study.

- Tonto Basin Elementary

- Skull Valley Elementary School

- Hume Lake Christian Academy

- Laguna ElementaryMarinCalifornia

- Rio Lindo Adventist Academy

- South Central Calhoun Elementary

- Harrison Elementary

- Deer Isle-Stonington Elementary School

- Mill Creek Elementary School

- Webber School

- Kimball Elementary School

- Denton High School

- Denton School

- Bozeman Christian

- The Hunter School

- Dugan-Tarango Middle School

- Tobe Turpen Elementary

- Scenic View School

- Conneaut Valley Elementary School

- Laurel Run School

- Sugar Grove School

- Utica Elementary School

- Mercedes Academic Academy

- Anthony Center

- Drain School

- Elk Mountain Elementary

- Lance Creek Elementary

Bottom 10 schools

- Oaks Montessori School

- Bolton-Edwards Elementary and Middle School

- Amikids Georgetown Middle and High School

- Luther High School

- Buckeye Local High School and Ohio Valley Energy Technology Academy

- Fessenden Elementary School

- H. W. Byers Elementary, Middle and High School

- Luther Branson School

- St. Paul Elementary School

- Pleasant Grove

Top 10 counties

- Lake of the Woods, Minnesota

- Alpine, California

- Costilla, Colorado

- Highland, Virginia

- Logan, Nebraska

- Blaine, Nebraska

- Columbia, Washington

- Grant, Nebraska

- Nelson, Virginia

- Hardin, Illinois

Bottom 10 counties

- Marshall, Mississippi

- Allendale, South Carolina

- St. Clair, Missouri

- Tom Green, Texas

- Callahan, Texas

- San Francisco, California

- Simpson, Mississippi

- Laclede, Missouri

- Millard, Utah

- Ascension, Louisiana

State Ranking (safest to least safe)

1. Wyoming

2. Hawaii

3. Vermont

4. Montana

5. Maine

6. New Hampshire

7. South Dakota

8. West Virgina

9. North Dakota

10. Oregon

11. Idaho

12. Iowa

13. Washington

14. Virginia

15. Massachusetts

16. Minnesota

17. Rhode Island

18. Maryland

19. Kansaa

20. Nebraska

21. Wisconsin

22. North Carolina

23. Ohio

24. South Carolina

25. Kentucky

26. New Mexico

27. Connecticut

28. Utah

29. Tennessee

30. New Jersey

31. Missouri

32. New York

33. Colorado

34. Delaware

35. Georgia

36. Arkansas

37. Nevada

38. Alabama

39. Texas

40. Pennsylvania

41. Indiana

42. Oklahoma

43. Michigan

44. Arizona

45. Mississippi

46. Louisiana

47. District of Columbia

48. Illinois

49. Florida

50. California

Rural schools = safer roads. Counties with more urban areas and higher population are generally more dangerous. Which makes sense: a greater density of people walking and driving means a higher potential for crashes to occur.

Drivers should exercise extreme caution from 3 – 6 pm. Nationwide, the most dangerous time of day near schools is 3:00 – 4:00 pm… so take extra attention and care when in areas where kids walk, bike, or get off buses.

Be wary of afternoon pickup times. Drivers around schools during the afternoon peak pick-up time of 3:00 – 6:00 pm are more dangerous than the morning peak of 7:00 – 10:00 am.

Who are these drivers anyway?

Zendrive has anonymized, aggregated driver data from drivers from our customers and partners all over the country, such as HopSkipDrive and more. Some are commercial drivers, some are individual drivers, and all have the power of the Zendrive platform to keep them safe.

Wait, my school's roads aren't as dangerous as you say. What's up?

We took a snapshot of data from April 2018. If your school’s had safety improvements since then, they won’t be reflected in the data.

My school is wrong on the map. Can you fix it?

We used the school dataset from the National Center of Education Statistics (NCES). The NCES gathers their public school data from Common Core Data and private school data from the Private School Survey. If your school is not on the map, please contact NCES here.

What's next?

We’re going to be releasing periodic Safety Snapshots as we analyze more miles in more areas. Stay tuned!

Get notified of new studies
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.