In late August, the nonprofit National Safety Council (NSC) published some preliminary statistics that raised the interest of The Newtown Beeexpert Neil Chaudhary, PhD, his interest was also piqued., and upon reaching out for some perspective from local traffic safety

Dr Chaudhary holds a PhD in experimental psychology, and is CEO at Trumbull-based Preusser Research Group, Inc (PRG), a full-service research firm specializing in transportation, highway safety, and issues related to drug and alcohol abuse. For 14 years, he has helped supply the federal government and many states, including Connecticut, AAA, and the Insurance Institute for Highway Safety with research involving behavioral traffic safety-related issues.

That initial report stated that motor vehicle fatalities for the first six months of 2016 had increased a staggering 45.3 percent.

But after reviewing that initial data, the Connecticut Department of Transportation (ConnDOT) has revised those stats down measurably to 18.3 percent. An official response memo provided to The Newtown Bee from ConnDOT, the state agency states that the NSC report is based on incomplete statistics for Connecticut. While the state's numbers are up, they are not as high as the NSC report states, according to the memo.

"We recognize that Connecticut is experiencing the same upward trend as most states," the memo reads. "However, the CT data reported in the recent NSC report was preliminary and is significantly different from the final number of fatalities that CT actually experienced. CT did not experience the 45 percent increase listed in the NSC report."

The memo goes on to clarify that ConnDOT submits its official Fatal Accidents Reporting System (FARS) data to the national FARS data base compiled by US DOT, which is subsequently used for all national reporting on fatalities.

The Connecticut FARS data recorded 115 fatalities in the first six months of 2015 and 136 in 2016, so the observed increase was 21, which is a 18.3 percent increase, the memo states.

"This is substantially less than the 45 percent reported by NSC. It is still premature to conclude the increase for 2016 will be 18 percent since the last 6 months of 2016 are not yet reported," the ConnDOT memo says.

Fatalities Increasing Nationwide

Offering further explanation as to why there is a discrepancy, and what the actual trend is, ConnDOT said that recent data suggests that fatalities are starting to increase in states across the country after many years of decline. It was first noticed in 2015, but early indications for 2016 suggest a continuation of the upward trend.

About half of the increase appears to be due to increasing vehicle miles of travel (VMT) as the economy rebounds from the recession and the recent drop in gas prices, the ConnDOT memo continues. It is too early to determine what other factors might be contributing to the increase beyond that of increasing VMT.

"In a small state like Connecticut, even a single year is inadequate for drawing meaningful conclusions," the memo goes on to clarify. "The wide variations in year to year fatalities experienced are the reason CT DOT adopted a 5-year moving average as a performance measure. The 5-year time period yields much more reliable information on trends in fatalities and fatality rates."

What gets lost in the exclusive focus on the first six months of 2016, the memo goes on to say, is the overall good safety record in Connecticut.

Connecticut ranks ninth or tenth best among all states when measured by the number of fatalities per million miles of vehicle travel. By adjusting for the volume of travel, the fatality rate makes it possible to compare each state to other states regardless of differences in the size of the state or traffic volume in a state.

For example, the memo notes that New York is a large state with more than 1,000 traffic deaths each year, but it has the same fatality rate as Connecticut. Connecticut typically has about 250 or so traffic deaths, which is one fourth of the New York total.

Despite the large difference in the number of fatalities, both states have the same fatality rate of 0.80 deaths per million.

Connecticut Ranks Low

The bottom line: Connecticut's rate of 0.80 fatalities per million miles of vehicle travel is the ninth or tenth lowest rate among all states. It is also well below the US average of 1.08 deaths per million VMT, ConnDOT reports.

Dr Chaudhary explained that the NSC's late August numbers came from the data released by NHTSA a day or two earlier, and based on FARS reports submitted to the government.

"Keep in mind this release by NHTSA is the annual file - which means that it is not complete and subject to change," he said before ConnDOT provided its memo to the newspaper. "The final release usually does not appear for a full year after the annual file comes out. I believe the state's internal tracking is likely more accurate than the federal number, although they should match after the final file is released. The state tends to be aware of fatal crashes where the police are still investigating - they can include those in their tally but perhaps can't submit an incomplete report to NHTSA."

Dr Chaudhary said that sometimes the fatality reports take a long time to be closed, particularly if there are charges pending related to the crash.

Somewhat Relieved

Following the release of the ConnDOT memo, Dr Chaudhary was somewhat relieved, but countered that an 18 percent increase is "still real and something to be concerned about, but not like the incredible 45 percent increase."

"The state suggests in this release that the first six months are not an appropriate gauge of change since last year and point to a general downward trend," he said. "They are doing well indeed compared to other states in terms of fatalities per vehicle mile traveled."

Generally speaking, Dr Chaudhary said, for every mile driven in Connecticut, the state experiences a lower chance of a fatality than roughly 80 percent of the other states.

"Where I somewhat disagree with them is their take on 'small state' and the ability to draw conclusions. Unlike typical statistical analyses where we have a sample of a population we are interested in, FARS is a census - it is all fatalities. So any change is a real change whether statistically significant or not - there is no margin of error so to speak," he said. "I do agree with them that changes from one year to the next using the six months of the year, and even year-to-year changes, are not a great indicator of trend."

He praised CT DOT for a nice job using the five-year moving average to show that despite occasional increases in a year, the state is headed in the right direction in terms of decreasing fatalities in motor vehicle crashes.

"Additionally, they are right that a lot could change over the next six months - but that change could be up or down relative to the previous year," he said. On a related matter, he noted that seat belt use rate in the state was just released, and it stands at 89.4 percent usage rate, which could also have a positive relation to vehicle deaths.