Speeding Drops By More Than Half After Waze Police Alerts Appear

Navigation App May Be Teaching Drivers They Can Beat the System, Study’s Author Warns

After a Waze alert flagged enforcement at a roadside spot in Australia, passing drivers went from averaging slightly over the speed limit to averaging about 2.4 mph under it. A new study clocked 9,052 vehicles with a laser speed gun of the kind police use, and the share of vehicles over the limit fell from 54.8% to 25.8% once the alert appeared.

Speeds dropped at all 33 sites, which ranged from school zones to roads with 62 mph limits. Because only vehicle speeds were recorded, the study cannot confirm which passing drivers were actually using Waze, but the timing points to the alerts as the likely reason for the slowdown.

But the study’s author cautions that a slowdown at one spot may not be good news for road safety. If drivers learn that a heads-up lets them dodge a ticket, they may feel freer to speed everywhere else. That worry comes from criminology theory, and the study itself did not test it.

Testing Waze Police Alerts on 33 Real Roads

A research team set up at 33 sites across Southeast Queensland, Australia, including highways, neighborhood streets, commercial areas, and active school zones. Posted speed limits ranged from about 25 to 62 mph (40 to 100 km/h).

At each site, an unmarked research car sat parked in one spot while a handheld laser speed gun recorded the speeds of passing vehicles. That device, a ProLaser 4, is widely used by police agencies in Australia and elsewhere. Meanwhile, the team watched Waze, Google Maps, and Apple Maps to catch the moment a user-submitted report of police or a speed camera first appeared. Waze was nearly always the first to show one.

That moment split the data in two. Speeds recorded before the alert were compared with speeds recorded for an equal stretch afterward, so a site where the alert surfaced 14 minutes in got a 14-minute “before” window and a 14-minute “after” window. Because the research car never moved, drivers saw no visual change between the two periods. No identifying information about drivers or vehicles was collected.

Numbers Behind the Waze Slowdown

Measured as the difference from the posted limit, vehicles ran about 1.2 km/h over (under 1 mph) before an alert appeared and about 3.9 km/h under (roughly 2.4 mph) afterward. A statistical model put the typical drop at around 5 km/h, or 3 mph. Site by site, the shift ranged from less than 1 km/h to more than 22 km/h (about 14 mph).

Counted vehicle by vehicle, 2,255 of the 4,115 clocked before an alert were over the limit, compared with 1,276 of the 4,937 clocked after.

Results varied by road type. Roads with a 100 km/h limit (about 62 mph) showed the biggest change: average speeds went from slightly over the limit to about 6 mph under it, and the share of speeding vehicles fell from 64.0% to 12.1%. Active school zones with a 40 km/h limit (about 25 mph) saw the smallest shift. Even after the alert, 38.5% of vehicles there were still over the limit, the highest leftover rate of any road type, down from 64.3% before.

Why Beating a Waze Alert Isn’t the Same as Driving Safer

Criminologists describe an idea called punishment avoidance: each time someone breaks a rule and gets away with it, the belief that they will be caught weakens. Research cited in the paper also holds that deterrence depends more on what people believe about the odds of getting caught than on how many officers are actually out on the road.

Applied to Waze, the author’s argument runs like this: a driver who gets a warning, eases off the gas, and passes the spot without a ticket has had a small experience of beating the system. Earlier research cited in the paper includes a case in which a driver who used enforcement-alert technology described feeling “probably slightly more invincible.” Waze has reported more than 150 million monthly users worldwide, so the worry is that drivers get better at dodging detection while staying just as committed to speeding.

Compliance that rests only on fear of getting caught, rather than belief in the speed limit itself, tends to be unstable, the author writes, drawing on earlier criminology research. Speeds in this study were recorded only at the alert spot, though, so whether drivers actually speed back up afterward remains untested.

Alerts also appeared quickly. On average, one showed up on Waze about 14 minutes after the research team arrived, and it stayed on the app for roughly 15 minutes after the team left. In the author’s reading, that leaves only a short window in which approaching drivers go unwarned.

Neither the speed of an alert nor its number of user “likes” lined up with how much speeds fell. Sites where more users confirmed an alert did not show bigger slowdowns.

Police Could Turn Waze Alerts to Their Advantage

Anderson also presented the findings to frontline road policing officers and asked for their reactions. Officers raised the idea of treating an alert as a cue to relocate. Once an enforcement team shows up on Waze, drivers ahead are already warned, so a single crew could pack up and move farther down the road, creating the appearance of several enforcement teams when only one is working.

Officers also discussed making enforcement highly visible on purpose, since a visible stop is more likely to be reported on Waze. An alert could then spread caution among drivers who never pass the checkpoint at all. None of these ideas was tested in the study, but the author’s takeaway is that shorter, more spread-out deployments may make better use of limited police resources.

Speeding contributes to roughly one-third of fatal crashes in Queensland, and earlier research has estimated that a 5% cut in average speed could lower fatal crashes by about 20%. A few kilometers per hour shaved off at one roadside spot is a modest gain on its own. Whether it grows into lasting safer driving or fades once drivers clear the alert zone is the question this study could not answer.

---