Regardless of where you stand on the issue, the legalization of marijuana is well underway across America. As of today, 31 states and the District of Columbia have laws legalizing cannabis for medical and/or recreational use. Nine of those states and the District of Columbia have legalized recreational marijuana, and at least six more states are expected to follow suit within the next two years.

Currently about 72 million out of 323 million Americans live where cannabis is legal, so the ramifications — good and bad — affect nearly a third of our nation's population.

Aside from the fact that marijuana is still classified as a Schedule 1 controlled substance (same as heroin and cocaine) by the federal government regardless of how it used — subjecting those who partake to potential criminal prosecution — there's an even larger and more deep-seated problem looming in states where pot is legal.

In a nutshell, it is the fact that there is no reliable, efficient way to test for stoned drivers.

States with the longest history of legalized marijuana — including Washington and Colorado — have seen worrisome increases in weed-related traffic deaths (Colorado by 48 percent). Marijuana-related nonfatal accidents and traffic citations have jumped as well.

Conversely, drunk-driving fatalities in the U.S. have fallen by a third in the last three decades, owing largely to the combination of increased education and deterrence provided by breathalyzer testing.

Marijuana, however, metabolizes much differently than alcohol. It does not reach its peak effect while it is in a person's bloodstream. It does so when it's in the fatty tissues of the body, and there's simply no proven roadside test available to determine that.

There are blood and urine tests that can ascertain that someone has cannabis's main active ingredient, tetrahydrocannabinol (THC), in their system. But it is much more difficult and much more costly to determine whether it got there two hours or two weeks ago. It requires an expert toxicologist to pinpoint timing, and most states don't have the resources to hire such experts — who come at an average fee of $5,000 per case.

Even more fundamental, science still lacks data correlating the presence of THC and actual impairment. This makes it difficult to establish legal limits for marijuana impairment in the same manner as for alcohol.

"It's simply not possible today to determine whether a driver is impaired based solely on the amount of the drug in their body," noted Marshall Doney, head of the American Automobile Association (AAA), following an extensive study on the issue by the association in late 2016.

The lucrative promise of big profits has spurned wide-ranging research into the development of a suitable device similar to a breathalyzer, and several are reported to be in the advanced stages of testing. Human trials, however, have yet to be undertaken and could take years to complete.

This predicament leaves law enforcement officers having to rely on their noses when stopping suspected stoned drivers or investigating accidents. Such "sniff tests" are unreliable, and in some cases they could be unjust, given that it might well be a passenger rather than the driver doing the puffing.

Beyond that, there are a wide variety of cannabis-containing edible products on the market, usually in the form of candies and baked goods that emit little or no odor.

All of this raises questions about how law enforcement officials are dealing with this pressing issue of public safety.

Let's take a look at what's happening in New Jersey as it edges toward legalizing marijuana. A couple of potential, partial solutions surfaced during a recent conference bringing together representatives of the AAA, law enforcement, court officials and attorneys.

According to Cathleen Lewis, director of public affairs for AAA Northeast, the lack of a practical scientific test leaves determination of impairment to Drug Recognition Experts (DREs) — police officers extensively trained to spot signs of drugged driving.

"The reality is, that's the best tool we have today," Lewis told the Asbury Park Press. "The good news is New Jersey has the second-most trained DREs in the country (behind California). The problem is that's not enough. Even if every police department had one, that one officer isn't going to be working 24 hours a day."

It is important to note as well that DRE officers' determinations are always open to scrutiny because the tests they perform are largely subjective rather than scientific — a big problem for prosecutors when DUI cases go to court.

"Our best hope may be education," says Lewis. "That starts with shooting down a big lie pushed by some marijuana advocates — that when you're high on weed, you're a better driver. In most cases that's simply not true. The vast majority of studies indicate marijuana impairs drivers, albeit to varying degrees."

In an ideal world, a significant portion of the massive revenue going to states that legalize recreational marijuana would bolster development of reliable testing, better enforcement (more DREs) and education.

In the meantime, America must acknowledge the obvious: There remains a significant downside to lighting up.