New Jersey DOT Documents Accident Severity Spike with Red Light Cameras

Oh, they simply have to "fine tune it" to get some more revenue.

Can you doubt the motives of government when it comes to revenue???

Fred

HawaiianFlyer titled this thread by saying:

"New Jersey DOT Documents Accident Severity Spike With Red Light Cameras"

Then, HawaiianFlyer gives us some quotes from a thenewspaper.com article commenting on a recent report on Red Light Camera by the NJDOT by posting these quotes:

"Traffic accident injuries spiked at the intersections where red light cameras were installed in New Jersey, according to a report released Monday by the state Department of Transportation (NJDOT)."

"Despite the mediocre results, NJDOT was not ready to call the program a failure."

This statement was based on a thenewspaper.com article dated 11/28/2012 found at http://thenewspaper.com/news/39/3960.asp.

I suspect that HawaiianFlyer did not actually read the report.

In order to see whether of not the words "spiked" and "mediocre" were appropriate, one would have to read the actual report which can be found at http://thenewspaper.com/rlc/docs/2012/nj-rlcreport2.pdf.

After telling readers that traffic accident injuries had spiked in New Jersey and that the results of the NJDOT report were mediocre, thenewspaper.com article tries to influence readers opinions by saying:

"NJDOT officials attempted to paint the program in the most positive light possible. The report claimed the right-angle crashes were "more severe" than rear end collisions, implying the cameras may have provided a modest safety benefit. The actual data suggest the opposite is true. Before cameras were installed, there were no accidents at the monitored intersections serious enough to merit an "A" rating, which generally means a crash victim was admitted at a hospital for treatment of a disabling injury. Once ticketing commenced, the only "A" accident recorded was a rear end collision -- not an angle collision -- at the intersection of Blackwood‐Clementon Road and Cherrywood Drive in Gloucester Township."

This whole paragraph needs review - but first it is noted that New Jersey classifies accidents by the following scale developed by the National Safety Council developed for five (5) categories of injuries: O - fatality, A - disabling injury, B - evident injury, C - possible injury, and O - property damage only (no injury).

I found the statement "NJDOT officials attempted to paint the program in the most positive light possible" interesting because thenewspaper.com is obviously trying to twist the NJDOT report in the most negative way possible. Indeed, thenewspaper.com has incorrectly cited numbers in the report. It states that the only "A" accident occurred after ticketing had begun - when, in fact, it occurred Pre- Camera. I checked the http://www.state.nj.us/transportation/publicat/lmreports/pdf...
to verify that the specific intersection cited by thenewspaper.com showed that the "A" accident occurred Pre-Camera, just as the http://www.state.nj.us/transportation/publicat/lmreports/pdf... report stated.

Here is the table from the report.

Statewide ‐ RLR Locations
Right Angle Crashes
Severity.................... K A B C O Total
Pre‐Camera............... 0 0 1 20 39 60
Year 1...................... 0 0 5 26 20 51
Same Direction Crashes
Severity.................... K A B C O Total
Pre‐Camera............... 0 1 1 72 212 286
Year 1...................... 0 0 4 80 259 343

Note that thenewspaper.com article said that "[t]he report claimed the right-angle crashes were "more severe" than rear end collisions, implying the cameras may have provided a modest safety benefit. The actual data suggest the opposite is true."

In the case of the "A" accident, thenewspaper.com misstated the data in order to make their false claim that "the opposite was true."

Additionally, note that the NJDOT report did not really claim that "right-angle crashes were 'more severe' than rear end collisions." Indeed, the report used the modifiers "in general" and "tend" when comparing right-angle crashes to rear end collisions (in the above table NJDOT refers to rear end collisions as same-direction crashes). Here is what the report actually said.

"National studies that focus exclusively on raw numbers and associated percentage changes are missing the critical factor of crash severity. For example, at a location where right-angle crashes decreased by two (2) but same-direction crashes increased by three (3), it might be concluded that RLR was ineffective, as the total number of crashes increased. However, in general, right-angle crashes tend to be much more severe when compared to other crash types. As a result, crashes must be analyzed not only numerically but also by severity."

Recall that thenewspaper.com article started out with an attention getter.

"Traffic accident injuries spiked at the intersections where red light cameras were installed in New Jersey, according to a report released Monday by the state Department of Transportation (NJDOT)."

Then, in order to make the casual reader get a feeling of statistical significance in the results they will report, thenewspaper.com says:

"Under the state law that authorized photo ticketing program in 2008, NJDOT officials must closely monitor the 25 municipalities and 83 intersections using automated ticketing machines."

Are there really 83 intersections in the data? NO! The NJDOT report states:

"As of May 1, 2012, there were eighty-three (83) intersections in twenty-five (25)municipalities authorized for program participation. Based on the established reporting parameters, monitoring systems at two (2) RLR locations in Newark now have two (2)years of data for study analysis. Additionally, twenty-four (24) intersections within nine (9) municipalities in six (6) counties have been recording violations for at least one (1)full year."

So - the NJDOT report is presenting two year's worth of data on only 2 locations and one year's data on only 24 locations. One would have hoped that thenewspaper.com would have noted the total number of intersections with actual comparable data was limited. However thenewspaper.com choose to highlight only the 83 "authorized" locations.

But note how thenewspaper.com article misleads, even on the 24 locations, when it says:

"[T]he total number of accidents at these intersections increased a statistically insignificant amount one year after the devices were installed -- from 577 before the devices were installed to 582 collisions with ticketing in place. The number of rear end collisions increased a significant 20 percent."

What thenewspaper.com article decided to do was "edit" what the NJDOT report said to make it seem less positive. Consider what the NJDOT report actually said:

"Combining all RLR locations and comparing the Pre-Camera installation 12-month time period versus the Year 1 installation 12-month time period, the data indicate that total crashes are up 0.9%, however, more severe right-angle crashes are down 15%, while same-direction (rear-end) crashes are up 20%."

Note that thenewspaper.com decided that a 20% increase in rear end collisions was "significant" and - thus - decided not to mention the 15% (and supposedly insignificant) decline in the more severe right-angle crashes.

Contrary to the way that thenewspaper.com treats things, the NJDOT report was careful to treat the data with discreetness. They say:

"The data from the two sites having two full years of operation shows a reduction in all types of crashes, however, the data are still too limited to draw any definitive conclusions about the pilot program at this time. The Department therefore recommends continued data collection and monitoring of RLR program intersections"

even though the data for the two year analysis is very positive for both years:

"For the two locations with two (2) years of data, when the Pre-Camera year crash data is compared to Year 2 crash data, right-angle crashes are down 86%, same-direction crashes are down 42%, total crashes are down 57%, and estimated severity costs have been reduced by $268,900. Crash severity cost is the overall cost attributable to running red lights, which include such things as vehicle damage, property damage, emergency response, and medical care."

and

"the Year 1 (2010) data revealed that total crashes were down 45%, right angle crashes were down 57%, and same-direction crashes were down 50%. Additionally, crash severity costs were reduced by an estimated $149,000, and both locations experienced a decreasing trend line regarding the number of citations issued."

Consider the "spike". What prompts thenewspaper.com to use such language. Well, they say:

"The next level of severity, "B," is applied when victims suffer contusions, large lacerations or other visible injuries. Prior to camera use, there was one rear end and one angle accident meriting a "B" rating. After cameras installed, these figures shot up to 5 angle and 4 rear end "B" accidents. The only category seeing an accident reduction at the photo enforced locations were angle collisions that produced minor, property-damage-only collisions."

Now recall that the "A" accident was a pre-camera value of 1 and the Year 1 value was 0. So, the more severe accidents (where the injury was disabling), went down to zero. But - was that "significant"? The numbers are just too small to make any claim - such as "more severe accidents plummeted to zero"? No it does not. Likewise, the 1 to 5 right-angle and 1 to 4 rear end accident increases are not significant.

The readers of this post are encouraged to read the report and decide for themselves if "spiked" is an appropriate way to describe the actual data. The readers should also decide if there is enough data to be statistically significant.

Assuming you read the NJDOT report, you will know that it concludes:

"Conclusions and Next Steps The Department’s focus remains solely on the potential safety benefits provided through RLR. Following that focus, the data shows decreases in right-angle crashes; decreases in the number of citations issued; and a continuation of those trends for the locations at which RLR has operated for two years—however, it is not prudent at this time to draw any final conclusions. Of the fifty-nine (59) locations that had been authorized as of June 1, 2011, only 41% managed to initiate RLR operation to the point of achieving one
full year of data by the end of 2011. Also, of the authorized municipalities, only nine (9)
municipalities (representing 36% of the overall program) are included within this Second Annual Report.

As such, further sustained analysis is needed and the Department recommends that the Traffic Control Signal Monitoring Systems Pilot Program continue."

It is clear that FZbar did not read the report (see his comment in the post above) because the trends in the data presented clearly indicate that the number of violations (which generate revenue) are decreasing. The NJDOT report states:

"Combining all twenty-four (24) locations statewide, the overall decrease in citations issued in the first month of RLR operation as compared with the amount issued in the twelfth month was 50%, with 35,644 violations issued in month one versus 17,934 citations in month twelve.

The overall goal regarding RLR citations is decreasing trend lines of citation issuance. While there were some regional and individual location exceptions, and most
intersections experienced a monthly fluctuation, Chart 2 reveals a nearly constant decrease in RLR citations issued statewide."

We know the reports presented by The Newspaper.com are both fair and balanced as well as accurate.

Agreed: MSNBC has a similar reputation.

As a motorcycle rider, any increase in rear end collisions, however temporary, makes me nervous.

The back of my car can absorb a lot of damage before I get anything more than a sore neck.

The back of my motorcycle, not so much.

The Mafia controled the garbage trucks, now its the traffic cams

Red light cams = Cash registers

And boy he eats a lot!

I guess you didn't get the Twinkie Memo?

'splain to me why they are (honestly) sad that Twinkies have gone the way of the dodo bird?

I bet they tasted about the same...

I did not see a problem with thenewspaper.com's reporting of this study and I am glad they are there to highlight this information.

Did you read the study?

I read the article, the report, and your dissertation on it as well.
It looks like cherry picking is common on both sides. Lucky for us thenewspaper.com provides a link to the report, so we can all make up our own minds.

For me as a motorcyclist, at least now there is more documented proof of the additional danger I face in redlight camera happy municipalities.

As I recall, you read last year's report from NJ with the same rosey outlook when it showed the same trend, jgermann. The report is easily available, we can figure it out.

What was not said was that tailgating was the problem. Rear end accidents are caused by idiots who follow to close and not cameras.
Money made from these cameras are not a question here but idiot drivers who do not know the law.

That's a good theory until you see someone literally slam on their brakes the moment a light turns yellow, and realize that following distances sufficient to allow for that are practically impossible in an urban or suburban environment.

Don't believe that happens? Visit the DC area and Maryland suburbs. I still see it here and we've had the cameras for many years.

@JD4x4, if you are suggesting that I cherry picked from the NJDOT 2012 report, I would appreciate your pointing out why you think so.

Also, do you agree with the fact that thenewspaper.com claimed that the "A" accident occurred "once ticketing commenced" was a false statement and that the "A" same direction accident had occurred pre-camera.

Concerning your statement that I "read last year's report from NJ with the same rosey outlook when it showed the same trend", are you saying that I had previously made a comment on this site last year? Or, are you just objecting to the fact that I included comments from the 2012 report on what the NJDOT found in their 2011 report in my " dissertation" (as you call it)?

In case you are interested, here are links to the NJDOT 2011 report:

November 24, 2011 NJDOT released its REPORT ON TRAFFIC CONTROL SIGNAL MONITORING SYSTEMS

http://www.state.nj.us/transportation/publicat/lmreports/pdf...

http://www.state.nj.us/transportation/publicat/lmreports/pdf...