Jeff Sessions (Jacquelyn Martin/Associated Press)

So Attorney General Jeff Sessions took to the pages of The Washington Post to write an op-ed last weekend. Sessions is rescinding an Obama administration policy that instructed federal prosecutors to avoid seeking mandatory minimums in some drug cases.

In Sessions’s defense, he did get one thing right, although he seemed to utterly miss the significance of it. And then he got a lot of things wrong. So many, in fact, that only a line-by-line review will do the whole thing justice.

So let’s get to it. Sessions begins:

Drug trafficking is an inherently violent business. If you want to collect a drug debt, you can’t, and don’t, file a lawsuit in court. You collect it by the barrel of a gun.

So this is the thing Sessions got right. Drug trafficking is violent. It is violent because courts and other traditional nonviolent means of settling disputes aren’t available to anyone involved. And it isn’t just debts. Where purveyors of legal products compete for customers by offering a better product, a cheaper product or better service, drug traffickers win customers, or “turf,” by killing one another. This has always been true — of drugs, and of every other product sold on the black market.

It’s encouraging that Sessions realizes this. What’s puzzling is how Sessions can (a) acknowledge that black markets cause violence, (b) claim to worry about said violence, and yet (c) work behind the scenes to expand black markets. Sessions not only opposes legalizing drugs, but he also wants to return states that have already legalized recreational marijuana — and who seem to be doing just fine — to the days when marijuana was available only on the black market. Or to put it as Sessions does: If pot retailers in Colorado, Washington and the other legalization states need to collect on a debt today, they do what any other retailer does. They use the legal system. If Sessions had his way, pot dealers in these states would to back to collecting debts “by the barrel of a gun.”

Why does Jeff Sessions want people in Washington, Colorado, and the other states that have legalized marijuana to experience increased violence — violence that he himself acknowledges would be inevitable if he were to get his way? Is it really that important to make it more difficult for people to get high? What for Sessions would be an appropriate “dead bodies”-to-“euphorias prevented” ratio?

For the approximately 52,000 Americans who died of a drug overdose in 2015, drug trafficking was a deadly business.

About 18,000 of those deaths involved prescription opioids, which are legally available. About 8,000 involved benzodiazepines, which are also available legally. Both of those types of drugs are made by pharmaceutical companies, prescribed by doctors and sold by pharmacies. Does Sessions believe those are all inherently violent industries? The Journal of the American Medical Association estimates that 88,000 people die each year from alcohol-related deaths. Does Sessions believe that Anheuser-Busch, Diageo and E & J Gallo run “deadly businesses”? What about the 480,000 people who die each year from smoking? Is tobacco a “deadly business”?

Moreover, there’s solid and mounting evidence that marijuana may be an effective substitute for opioids when it comes to treating pain. States that have legalized marijuana have seen a drop in hospitalizations for opioid addiction and overdose, suggesting that if it’s easily available, people prefer to treat pain with marijuana rather than with opioids. Which means that under Sessions’s preferred policy of pot prohibition, we’d almost certainly see much higher numbers of opioid addiction and overdose deaths.

Yet in 2013, subject to limited exceptions, the Justice Department ordered federal prosecutors not to include in charging documents the amount of drugs being dealt when the actual amount was large enough to trigger a mandatory minimum sentence. Prosecutors were required to leave out objective facts in order to achieve sentences lighter than required by law.

This isn’t an accurate characterization of the memo issued by former attorney general Eric Holder. The memo states only that in cases in which a defendant was in possession of enough drugs to trigger a mandatory minimum federal sentence, federal prosecutors should decline to charge the crime that would trigger that sentence if the defendant meets a number of criteria, including not having committed an act of violence in association with the crime, not being the leader or organizer of a trafficking organization, not having ties to cartels or major drug traffickers, and not having a significant criminal history.

I suppose in some sense the memo required prosecutors to “leave out” facts in that it asked them to charge less than what federal law permits. But prosecutors have always had the discretion to bring lighter chargers than what they could conceivably bring. Moreover, when you consider the unfair, irrational way in which federal authorities measure drug quantities for the purpose of charging and sentencing, the Holder policy at best made the playing field slightly more level.

This was billed as an effort to curb mass incarceration of low-level offenders, but in reality it covered offenders apprehended with large quantities of dangerous drugs. The result was that federal drug prosecutions went down dramatically — from 2011 to 2016, federal prosecutions fell by 23 percent. Meanwhile, the average sentence length for a convicted federal drug offender decreased 18 percent from 2009 to 2016.

There are any number of reasons the number of federal drug prosecutions might drop. But note the range of years Sessions chooses here. The Holder memo was in 2013. Why go back to 2011? Because that’s when federal drug prosecutions peaked. But there was a big drop between 2011 and 2012, which wouldn’t have been affected by the Holder memo at all. There’s also a big drop between 2013 and 2014. You might argue the Holder memo played a role there, but the memo wasn’t issued until August of 2013. And between 2014 and 2016, the number of prosecutions dropped, but only slightly. I don’t see a handy table for average sentence length by year, but I’ll guess that Sessions chose 2009 as his baseline for that statistic.

Of course, if the Holder memo did reduce drug sentences for nonviolent offenders, good. That’s exactly what it was supposed to do. The evidence for this isn’t overwhelming, but the Sentencing Commission did report in March of 2016 that federal prosecutors were focusing less on low-level offenders, and more on serious and violent offenders. One would think that this is a positive trend, too.

For Sessions to show that this was somehow detrimental to public safety, he needs to show a link between lenient sentences and crime. And here his argument falls to pieces.

Before that policy change, the violent crime rate in the United States had fallen steadily for two decades, reaching half of what it was in 1991. Within one year after the Justice Department softened its approach to drug offenders, the trend of decreasing violent crime reversed. In 2015, the United States suffered the largest single-year increase in the overall violent crime rate since 1991.

A two-year uptick after a 20-year decline isn’t exactly a sign that the trend “reversed.” The crime rate was never going to keep falling forever. Sessions’s sentence about 2015 is technically true — but only because in 20 of the 24 years between 1991 and 2015, crime didn’t rise it all; it dropped. The murder rate in 2016 was up over previous years but it puts us back only to where the murder rate was in 2008, which at the time was a 40-year low.

But even if crime were up significantly, there’s simply no research whatsoever to suggest that federal prosecutors not charging mandatory minimums in drug cases would have anything to do with it. By one estimate, about 530 federal defendants in 2012 would have received shorter sentences had the Holder memo been in effect. If that’s true, it probably wouldn’t vary much from year to year. And it’s rather doubtful that 500 or so people are somehow responsible for the increase in crime in places like Chicago and Baltimore. It’s just as doubtful that drug dealers are suddenly more brazen and violent because they know that if they get caught, there’s a chance they may not get the maximum sentence allowed under law. (This argument is particularly inept given that the Holder memo specifically exempts drug offenders who have a history of violence, or who committed a violent act along with the drug offense in question.)

And while defenders of the 2013 policy change point out that crime rates remain low compared with where they were 30 years ago, they neglect to recognize a disturbing trend that could reverse decades of progress: Violent crime is rising across the country. According to data from the FBI, there were more than 15,000 murders in the United States in 2015, representing a single-year increase of nearly 11 percent across the country. That was the largest increase since 1971.

This is a picture-perfect example of how to lie with numbers. I wrote about that “largest increase since 1971” talking point last year. Instead of re-writing the same arguments, I’ll just re-post a relevant excerpt:

One of the problems with measuring crime as a rate is that rate increases tend to loom larger as the base number falls. For example, let’s say the murder rate in City X was 20 murders per 100,000 people in 1994. If the rate went up by one murder per 100,000 the following year, that would amount to a 5 percent increase. Now let’s say that City X saw a steep drop in crime over the next 15 years, and dropped to a murder rate of, say, just 5 people per 100,000 in 2009. After bottoming out, the rate then went up by 1 murder per 100,000 the following year. That would make for a 20 percent increase. Measuring crime by its rate in the population will always be infinitely better than using raw numbers. But it isn’t without its limitations.

These problems then get magnified when headlines frame them in ways such as, “U.S. sees largest crime increase in 30 years.” That may be true. But that’s only because (a) crime was falling for 25 or more of those years, and (b) any increase from a historic low is going to look larger than increase from a period when crime was more common. … City X isn’t less safe now than it was then. It’s quite a bit more safe. But context and long-term perspective are harder to fit into a headline.

In that same post, I noted that others had pointed out that the murder rate in 2015 was actually the sixth lowest in a half-century, and the overall crime rate was the third lowest since 1971. So yeah, it’s all about how you use the data. You have to have a pretty dim worldview — or a pretty strident agenda —  to take the year with the third lowest crime rate in 45 years and portray it as an ominous sign of trouble ahead.

The increase in murders continued in 2016. Preliminary data from the first half of 2016 shows that large cities in the United States suffered an average increase in murders of nearly 22 percent compared with the same period from a year earlier.

Perhaps this is the place to make a concession: There is good data suggesting that some of America’s cities are in the throes of a surge in violent crime. There are lots of theories as to why, many of which we’ve explored here at The Watch. But as far as I know, no one has suggested that the reason violent crime has risen in these cities is that … federal prosecutors no longer seek mandatory minimum sentences in some drug cases that meet Holder’s checklist of criteria. Jeff Sessions would appear to be the first and only person to make that connection.

As U.S. attorney general, I have a duty to protect all Americans and fulfill the president’s promise to make America safe again. Last month, after weeks of study and discussion with a host of criminal-justice participants, I issued a memorandum to all federal prosecutors regarding charging and sentencing policy that once again authorizes prosecutors to charge offenses as Congress intended. This two-page guidance instructs prosecutors to apply the laws on the books to the facts of the case in most cases, and allows them to exercise discretion where a strict application of the law would result in an injustice. Instead of barring prosecutors from faithfully enforcing the law, this policy empowers trusted professionals to apply the law fairly and exercise discretion when appropriate. That is the way good law enforcement has always worked.

The Holder memo wasn’t really a departure from that. It, again, was a pretty minimal effort at de-carceration that affected only a small percentage of drug cases. It didn’t eliminate prosecutorial discretion. It didn’t set violent offenders free. Its most important effect was likely symbolic. It stated that the highest-ranking law enforcement officer in the country had finally recognized what has been apparent to most of the rest of the country for a long time — that the drug war had grown excessive, and that its excesses were counterproductive. Likewise, Sessions’s revocation is also largely symbolic. It states that the highest-ranking law enforcement official in the country still thinks the drug war is a right good fight. We’ll lick drugs for good any day now.

Defenders of the status quo perpetuate the false story that federal prisons are filled with low-level, nonviolent drug offenders. The truth is less than 3 percent of federal offenders sentenced to imprisonment in 2016 were convicted of simple possession, and in most of those cases the defendants were drug dealers who accepted plea bargains in return for reduced sentences.

Fair enough. It’s true that most federal drug prisoners aren’t in for simple possession. It’s also true that it’s preposterously easy to rope minimal offenders in on conspiracy, racketeering and money-laundering charges associated with drug trafficking. Sessions isn’t wrong that some people doing time on possession charges plea bargained down from more serious charges, but it’s also true that federal prosecutors stack charges against defendants — charges for which there may not be much evidence — in order to get them to plead guilty, or to give up bigger fish.

Federal drug offenders include major drug traffickers, gang members, importers, manufacturers and international drug cartel members. To be subject to a five-year mandatory sentence, a criminal would have to be arrested with 100 grams or more of heroin with the intent to distribute it — that is 1,000 doses of heroin.

True, but other mandatory minimums can kick in for far smaller quantities. Use of a firearm while committing any drug crime, for example, brings a five-year minimum. (And “use” of a firearm can simply mean having one on you, or even having one in your car.) Importing or exporting any quantity of drugs — or being part of any conspiracy to do so — also brings mandatory minimums. There are plenty of horror stories about girlfriends or siblings getting roped into conspiracy charges after minimal participation. And if you happen to have sold or given someone an illicit drug that results in death or serious injury of a user, you’re also looking at a 20-year minimum, regardless of the quantity.

The truth is that while the federal government softened its approach to drug enforcement, drug abuse and violent crime surged.

And — again — the truth is that there’s zero evidence that the former had anything to do with the latter. The overwhelming majority of drug prosecutions take place at the state and local level, not the federal level. Which means that even if there were some link between not seeking mandatory minimums and an increase in violent crime, that still wouldn’t do much to bolster Sessions’s argument, because the Holder memo had no bearing on the vast majority of drug prosecutions in America.

The availability of dangerous drugs is up, the price has dropped and the purity is at dangerously high levels. Overdose deaths from opioids have nearly tripled since 2002. Overdose deaths involving synthetic opioids rose an astonishing 73 percent in 2015.

It depends on the drug. It’s largely true that heroin purity and potency have increased. It’s also true that we’ve seen some particularly toxic combinations, like heroin laced with fentanyl. These are also byproducts of prohibition. During alcohol prohibition, black market booze was notoriously potent, and often toxic. Deaths and hospitalization from alcohol poisoning soared. No one drinks bathtub gin or wood alcohol anymore — because no one has to.

Fentanyl is a synthetic opioid that is quickly becoming a major contributor to the U.S. addiction crisis. Here are the top things to know about the drug. (Amber Ferguson/The Washington Post)

As for the overdose deaths, again, a large percentage of those deaths involved legal prescription drugs.

My fear is that this surge in violent crime is not a “blip,” but the start of a dangerous new trend — one that puts at risk the hard-won gains that have made our country a safer place.

So far, there’s no reason for this fear. The rise in violent crime seems to be driven by large upticks in about a dozen large cities. That in itself is worrying, and it is something that needs to be discussed. It is not a reason to abandon efforts to make federal drug sentencing fairer, more just and more proportionate.

Some skeptics prefer to sit on the sidelines and criticize federal efforts to combat crime. But it’s not our privileged communities that suffer the most from crime and violence. Minority communities are disproportionately impacted by violent drug trafficking. Poor neighborhoods are too often ignored in these conversations.

Studies by academics such as Bruce Western have shown that low-income neighborhoods have been absolutely devastated by mass incarceration. Imprisonment imposes consequences on not only the incarcerated but also their children, families, friends, extended networks and even entire neighborhoods. Children of incarcerated parents are more likely to be depressed, to drop out of school and to commit crimes of their own.

It seems doubtful that the 500-plus people who may have benefited from the Holder memo are going to wreak havoc on poor communities in the years they’d otherwise have been in prison. It seems probable that the status quo is going to be far more detrimental to those communities — continuing to incarcerate at the levels we have been, continuing to make people in those communities unemployable because of arrest and prison records, continuing to make jobs in the illicit drug trade the only potential source of income for those people, and so on.

Regardless of wealth or race, every American has the right to demand a safe neighborhood. Those of us who are responsible for promoting public safety cannot sit back while any American communities are ravaged by crime and violence.

Doing anything is often worse than doing nothing. There’s zero indication that the policies promoted by politicians such as Sessions have done much to make those communities safer. In fact, as noted earlier, Sessions himself concedes that black markets create violence — which means his continued advocacy for marijuana prohibition would likely make many communities more dangerous. But at risk of being repetitive, there’s still no basis for connecting the Holder memo to any increase in violent crime. Stating over and over again that there is doesn’t make it true.

There are those who are concerned about the fate of drug traffickers, but the law demands I protect the lives of victims that are ruined by drug trafficking and violent crime infecting their communities. Our new, time-tested policy empowers police and prosecutors to save lives.

Returning to the old policy will only enable prosecutors to “save lives” if you believe that (a) the drugs the defendants were distributing were ending lives, (b) the defendants would have returned to distributing drugs in the years the Holder memo took off their sentences, and (c) the people who would have ruined their lives by ingesting drugs they obtained from the defendants wouldn’t have or couldn’t have simply found another dealer to supply them with the very same drug.

(a) may well be true for some people. (b) seems less certain. (c) would be contrary to basic economics, everything we know about human behavior, everything we know about black markets and just about all of human history.

Certainly, drug trafficking lowers the quality of life in a community. Turf wars between drug gangs can make those communities more dangerous. But again, Sessions himself concedes that prohibition itself creates these problems. It’s pretty rare that liquor store employees erupt in gun fights over turf. And if prohibition begets violence, the only way the solution to an increase in violence can be more prohibition is if the new prohibition wipes out drug trafficking entirely. Otherwise, more prohibition usually just means more violence. Knock out one major dealer, and new dealers will emerge and go to war to take his place.

We all know that rescinding the Holder memo isn’t going to end drug trafficking. It isn’t going to affect the opioid crisis. It isn’t going to move the needle either way on the violence in Chicago or Baltimore. The most likely outcome is that a few hundred more nonviolent offenders spend a lot more time in federal prison than they otherwise would have. I suppose it will also give Sessions the satisfaction of having rolled back one of the few substantive criminal-justice reforms of the Obama administration. But the crime rate and the violence in America’s cities will rise or fall independent of the Holder memo.

The one thing we can all depend on — the one sure thing: Illicit drugs will continue to be available to pretty much anyone who wants to use them.