This is another piece of junk science and junk journalism as so often produced by the American Association for the Advancement of Science (AAAS)!
This indoctrination is driven by a leftist obsession with purported racism claims!
Notice it is Black versus white drivers!
Notice that this study is only limited to Florida!
Perhaps the biggest and most devastating flaw of this study: The study authors did not bother to directly sample the ethnicity/race of the 200,000 Lyft drivers, but by inferring ethnicity/race of these drivers by "fit[ting] a model that relates the self-reported race or ethnicity of the 45.7% of drivers who are registered voters in Florida to a picture that each driver submits when they apply to drive on Lyftâs platform. This fitted model is then used to infer the race or ethnicity from the picture submitted by the 54.3% of drivers who are not registered voters.".
One may also question the motives of these researchers: If researchers are obsessed with purported racism, then they will find it despite facts to the contrary.
There are several, serious doubts about this study's finding of racial profiling of drivers by police officers or racial bias in policing (It appears none of the following questions were investigated in this study):
- What was the skin color of the police officers? How many of the police officers were non-white?
- How do drivers respond to and interact with police officers during a traffic stop? Will some drivers more likely insult a police officer as racist?
- What were the driving records of the drivers at the time of the traffic stop?
- Where (location) did the traffic stop occur? (That this study focused narrowly on the Florida Highway Patrol is probably not good enough. Potentially, this particular selection of the FHP may have introduced a bias by itself.)
The abstract indicates that the authors have taken into account accidents and reoffenses. It states "no evidence that accident and reoffense rates explain these estimates"). However, when you look at how the authors defined and measured accidents and reoffenses (see the Methods section of the study), then you find out their approach was rather dubious and peculiar!
"The phrase âdriving while Blackâ entered the American lexicon in the 1990s, when the people learned that police officers were explicitly targeting racial and ethnic minorities for traffic stops. Still, the argument has persisted that higher numbers of detainments and citations for non-white drivers are not due to bias, but rather, a higher likelihood for those drivers to engage in traffic offenses. Now, using a clever dataset, researchers have clearly demonstrated the latter is untrue.
A team analyzed data from more than 200,000 Lyft drivers in Florida who were tracked using high frequency GPS pings from their smartphones. In that way, the team was able to determine whether and by how much a driver was speeding when they were pulled over, regardless of what was written in a police report. They also assessed how often that driver engaged in such behaviors. The data were unequivocal: There were no discernible differences in driving behavior between white and minority drivers, yet police were up to 33% more likely to issue a citation and charged up to 34% more if the driver wasnât white.
âOur findings suggest that police racial profiling of drivers is due to police having animus or prejudice against minority drivers,â the team writes. Properly translating these findings into policy actions will require additional work, write Dean Knox and Jonathan Mummolo in a related Perspective. âFor example, it may be just as important to study bias in how officers are assigned to work those locations and times in the first placeâestimates that may reveal patterns of, for example, overdeployment in minority neighborhoods causing disparate impact.â"
"Data limitations have long stymied research on racial bias in policing. To persuasively demonstrate bias, scholars have sought to compare officer behavior toward minority versus white civilians while holding constant all other factors in the police-civilian encounter that might provide alternative explanations for enforcement disparities. ... On page 1397 of this issue, Aggarwal et al. (2) report using data from the ridesharing service Lyftâhaving obtained vehicle location on more than 200,000 drivers using high frequency GPS pings from their smartphonesâto analyze speeding enforcement by the Florida Highway Patrol (FHP) and to show how such data offer a path forward for addressing both challenges.
One challenge to establishing all-else equal comparisons in studies of policing is that standard police datasets contain one sided officer accounts of civilian behavior, which past work has shown do not always accurately measure actual driver behavior. For example, prior research (3) has shown that in the same FHP context, officers gave white drivers a âdiscountâ on tickets by reporting lower speeds relative to the speeds reported for minority drivers. Aggarwal et al. use the Lyft data to construct an objective measure of speeding behavior.
The other challenge is that traditional police-generated datasets are inherently selective: For example, they do not contain every police-civilian encounter in which an officer could have cited a speeding driver but rather only the subset in which an officer chose to pull vehicles over and therefore had to fill out forms documenting the stop. ... The approach of Aggarwal et al. resolves this challenge too, by allowing researchers to observe all times when Lyft drivers are active rather than only the selected sample of those where officers chose to detain them."
From the editor's summary and abstract:
"Editorâs summary
US police detain racial or ethnic minority drivers (Asian and Pacific Islander, Black, and Hispanic) more than white drivers. However, a longstanding debate remains unsettled: Does this necessarily reflect police bias? Hypothetically, if minorities were more prone to disobey speed limits and traffic laws, then their traffic stops may be warranted. To put this debate to rest [???], Aggarwal et al. examined rideshare data from Lyft in the state of Florida to compare minority drivers with their white counterparts. Lyft objectively measured driversâ locations, driving speed, and location speed limits ... White and minority drivers showed no discernible differences in speeding behaviors or traffic violations. However, when both drove at identical speeds, police were still 33% more likely to issue speeding citations to minority drivers and charged 34% more expensive fines, unequivocally revealing bias. ...
Abstract
Prior research on racial profiling has found that in encounters with law enforcement, minorities are punished more severely than white civilians. Less is known about the causes of these encounters and their implications for our understanding of racial profiling. Using high-frequency location data of rideshare drivers in Florida (N = 222,838 individuals), we estimate the effect of driver race on citations and fines for speeding using 19.3 million location pings. Compared with a white driver traveling the same speed, we find that racial or ethnic minority drivers are 24 to 33% more likely to be cited for speeding and pay 23 to 34% more money in fines. We find no evidence that accident and reoffense rates explain these estimates, which suggests that an animus against minorities underlies our results."
Caveat: I did not read the entire study.
ScienceAdviser