No, thousands of "crackheads" aren't going to start flooding America's streets Tuesday.

That's just one of several myths that have surrounded the U.S. Sentencing Commission's vote in June to make federal sentence reductions retroactive for current prisoners convicted of crack possession or use.

What happens Tuesday is that some eligible federal prisoners who have petitioned for reduced sentences under rules Congress passed last year can begin being released. Those rules sought to address a disparity that meant crack offenders were given the same mandatory five-year minimum sentence as were offenders in possession of 100 times as much powder cocaine.

Since the so-called 100:1 ratio was imposed in 1986 — shortly after the cocaine-related death of college basketball star Len Bias — it has come to be widely regarded as racially discriminatory. That's because the great majority of those convicted of crack possession are African-American — about 84 percent, according to Justice Department statistics. By contrast, African-Africans make up only about 30 percent of those convicted of possession of powder cocaine.

In June, the Sentencing Commission voted unanimously to redress what it called the "fundamental unfairness" of the old law by allowing prisoners convicted before it was changed to seek to reduce their sentences to be in line. 

The new policy applies only to those convicted in federal court — the tens of thousands of crack cocaine convicts in state prisons aren't affected. And it effectively applies only to those federal prisoners convicted after 2007, when the Sentencing Commission similarly allowed federal crack prisoners to seek retroactive reductions after a different adjustment of the guidelines.

That's a narrow subsection, comprising prisoners convicted in federal court of crack possession since the last adjustment. The commission projects it covers about 12,000 inmates in 116 federal prisons across the country.

Not all of those 12,000 prisoners will have their sentences reduced. For one thing, there's no way to know how many will actually seek reductions, particularly those who are near the ends of their sentences anyway. 

And the reduction isn't automatic; prisoners must go before federal judges, allowing for potentially dangerous or violent offenders to be screened out. When the courts went through the same process three years ago, they rejected more than a third of petitioners. 

The average reduction is projected to be about three years. Even with that reduction, the average sentence will still be about 10 years; that means many of those who win sentence reductions will still have several more years to serve.

• U.S. Sentencing Commission statement on new guidelines (.pdf)

All told, projections are that between 1,000 and 2,000 prisoners across the country will be eligible for immediate release when the policy takes effect Tuesday.

History does tell us that at least some of them will re-offend. But if the 2008 release is any indication, it won't be because they were let out early.

Source: U.S. Sentencing Commission, May 2011

Federal statistics show no difference in recidivism between crack defendants who were released early under a 2007 program and those who finished their sentences.

In May, the Sentencing Commission published its analysis of what happened to the approximately 16,000 prisoners who went free when their sentences were reduced after the earlier policy went into effect in March 2008. It compared their outcomes against those of a similar number of crack cocaine offenders who completed their sentences.

Over the two-year period, 30 percent of the early releases were arrested again for a new crime, the statistics show.

And what about the control group? More of them — 33 percent — were arrested again. 

It's tempting to say the statistics show that early release makes crack offenders less likely to re-offend, but in fact the difference is within the statistical margin of error. What it does show is that there's no appreciable difference in recidivism between the two groups. (See chart above.)

• Read the entire U.S. Sentencing Commission analysis (.pdf)

(Regardless of whether they are released early or not, crack offenders are about half as likely to be arrested again as are federal criminal offenders overall, 59 percent of whom the Justice Department says are re-arrested within two years of release.)

More reality check: The new policy doesn't, in fact, wipe out the disparity in cocaine sentencing. It's the result of a compromise as Congress debated the new sentencing guidelines last year. True, the crack possession-to-powder possession isn't a whopping 100:1 anymore. But it's still 18:1 — meaning you can have 18 times more powder cocaine than crack in your possession and still wind up with the same minimum sentence. 

There's one last misperception, perhaps the biggest of them of all. It's the persistent belief that the 1986 law disproportionally cracking down on crack was passed because Len Bias died after a night of crack-fueled celebration over his having been picked second in the NBA draft. His death came at the peak of the 1980s concern over crack and its role in drug gang violence that drove homicide rates into the hundreds a year in several major cities.

But Len Bias did not die from smoking crack cocaine.

At their annual seminar on sentencing, federal prosecutors reported last year: "Ironically, Len Bias's death was later shown to have been from a powder cocaine overdose — not a crack cocaine overdose as initially believed."

The paper is titled "Still Haunted by Len Bias." (.pdf)