Car accidents are on the rise in legal states, and some data crunchers are pointing their fingers at recreational marijuana. But legalization does not lead to more car crashes or to more fatal road collisions, according to a new study published in the American Journal of Public Health by researchers from the University of Texas at Austin.

The study compared the rate of fatal car crashes in Colorado and Washington before and following legalization in 2012. After delving through years of data, researchers concluded that there is no evidence to suggest that marijuana has made roads unsafe.

“We found no significant association between recreational marijuana legalization in Washington and Colorado and subsequent changes in motor vehicle fatality rates in the first three years after recreational marijuana legalization,” the authors wrote.

And there's no evidence to suggest that these states are seeing more car accidents due to cannabis. “[W]e also found no association between recreational marijuana legalization and total crash rates when analyzing available state-reported nonfatal crash statistics.”

Those findings contradict another recent study on the effect legalization has had on road safety. Earlier this week, the Highway Loss Data Institute reported that car crashes are on the rise in states that have legalized recreational marijuana. Their study sifted through years of insurance claims related to collisions and found that accidents rose 16 percent in Colorado, 6.2 percent in Washington state and 4.5 percent in Oregon after those states okayed recreational marijuana.

"We're concerned about what we're seeing," Matt Moore - Senior Vice President of HLDI - told CNBC. "We see strong evidence of an increased crash risk in states that have approved recreational marijuana sales."

There's only one problem with his conclusion: nothing connects those accidents to marijuana.

As Gregor Bister of CNBC noted, "While Moore's research finds a greater crash risk, his study does not say if the increase in collisions in the three states were directly caused by drivers who were high."

So there have been more car crashes in legal states, but there's no reason to believe marijuana is the culprit. Unless Moore can find data pinning these accidents on cannabis, the relationship between the two issues is coincidental - not causal. Basically, you could use Moore's logic to blame these statistics on the Denver Broncos because their five consecutive AFC West Divisional Championships (2011-2015) coincided with the rise in car crashes. But NFL football didn't have anything to do with those accidents, and neither did marijuana according to the data.

But Moore insists that there's reason to connect the two issues. And he suspects we'll see more car accidents in other legal states once they've been selling marijuana as long as Colorado.

"Colorado has had legal pot sales the longest and it is showing the greatest effect," Moore noted. "Meanwhile, Oregon has had pot sales for the shortest amount of time, so its increase is the lowest, but that could change over time."

Yes, Colorado did begin selling cannabis about 20 months before Oregon. But if that makes a huge difference, then Washington state should be seeing more crashes by now. Colorado started selling recreational marijuana only 6 months before Washington, but its 16 percent rise in car accidents is more than double the rate in the Evergreen State. Meanwhile, Washington's 6.2 percent increase in traffic accidents is only 1.7 percent higher than Oregon's 4.5 percent, even though Oregon began selling recreational marijuana over a year after its northern neighbor.

In other words, the numbers don't bear out Moore's idea that car crashes become increasingly common as legalization gets underway. And there's no reason to scapegoat marijuana legalization as the culprit.