14 March 2014

Most Federal Drug Sentences Will Be Probably Shorter As Of November 1, 2014

The Obama Administration supports a modest, across the board reduction in drug trafficking sentences for offenses not subject to mandatory minimum sentences, relative to current sentencing guidelines.  The changes will officially take effect November 1, 2014, unless Congress acts to countermand the U.S. Sentencing Commission (which is unlikely), but federal prosecutors have been directly to start implementing the changes immediately.
Attorney General Eric Holder testified today at a hearing of the U.S. Sentencing Commission on proposed amendments to the sentencing guidelines, one of which is to lower drug sentencing guidelines by two levels. He supports the change. Once approved by the Commission, unless Congress rejects the proposed amendments, it will go into effect Nov. 1. 
In a press release today, DOJ says:
Until then, the Justice Department will direct prosecutors not to object if defendants in court seek to have the newly proposed guidelines applied to them during sentencing.
. . . 
The problem is, of course, the reduction would not affect mandatory minimum sentences. The only way out from a mandatory 5, 10 or 20 year sentence remains limited to those who either cooperate with the government or are eligible for the Safety Valve. 
In order to qualify for the safety valve, a defendant cannot have more than one criminal history point. A DUI, for example, is one point. If the person violates his probation, or is charged with the drug offense while on probation for a DUI, he or she will have more than one point and be ineligible. 
Since mandatory minimums trump the guidelines, the change won't affect those charged with a mandatory minimum whose guidelines are now less than 60, 120 or 240 months, unless they either cooperate or have no more than one criminal history point and meet the other conditions. 
A bill is pending in Congress (S. 1410) that would amend the Safety Valve, to make more defendants eligible for it. It has already been weakened, and would only apply to some of those with up to two criminal points. It passed the Senate Judiciary Committee in January, 2014, but has yet to be voted on by the full Senate. DOJ has supported expanding the safety valve.
Via Talk Left.

What Does A Two Offense Level Reduction Mean?

Translating the proposed two offense level reduction into actual prison sentences requires resort to a chart that sets guideline ranges based on offense level and criminal history.

For a given starting guideline sentence range under the old sentencing guidelines, a two offense level adjustment results in roughly the same amount of sentencing reduction, regardless of criminal history.

Thus, an offender facing a 51-63 month drug offense sentence before the change (about 5 years) (regardless of the combination of offense level and criminal history necessary to produce this guideline sentence under the old guidelines), would be reduced to about 41-51 months (about 4 years) after the guideline change.

In another example, someone who would have faced a 18-24 month presumptive sentence under the old guidelines would face a 12-18 month presumptive sentence under the new guidelines. This would be equivalent under Colorado's state law criminal sentencing system to downgrading a felony from a Class 5 to a Class 6 felony (the lowest grade of felony under Colorado law).


Drug offenders make up 50.1% of the prisoners incarcerated with the Federal Bureau of Prisons (about 98,554 prisoners, as of January 25, 2014), so an across the board downward tweak in the presumptive prison sentences for drug offenses has the potential to materially reduce federal incarceration.

About 30% of drug offenders serving prison sentences (as opposed to jail sentences of generally less than one year) are in federal prison (which tends to have longer sentences for comparable drug offenses), while about 70% of drug offenders servicing prison sentences are in state prisons.  In contrast, in 1980, about 20% of drug offenders serving prison sentences were in federal prison, and 80% were in state prison.  The federal "market share" of drug offense incarceration has increased by about 50%.
The effect, the Justice Department says, would be to reduce by 11 months the average sentence of a drug dealer and would trim the federal prison population by roughly 6,550 inmates at the end of five years. The proposal would affect about 70 percent of drug trafficking offenders.
From here.

The 30% of drug trafficking offenders who are subject to mandatory minimum sentences and don't qualify for the "safety valve" would not be affected.

The percentage reduction in federally incarcerated drug offenders as a result of the change, after five years, would be about 6.6%.

 This change is relative to an already substantial reducing in sentences for comparable drug possession quantities for crack cocaine that has already been made law (although crack cocaine offenses continue to be punished significantly more severely than comparable powder cocaine offenses).

But, while the sentencing guideline change is big enough to be notable relative to the status quo, this is still only an incremental baby step.

Even with a two offense level reduction, the number of federal prisoners in prison for drug offenses in 2019 will still be more than 8.5% greater than in 2005, The U.S. population has grown by 6.2% from 2005 to 2012, and is projected to be about 10% greater than it was in 2005 by 2019. Thus, the impact of all current changes in federal drug crime offense sentencing (and the proposed sentencing guideline impacts) combined, will reducing the federal drug offense incarceration rate from 2005 to 2019 by only 1.5%.

In part, the very modest reduction in overall drug offense incarceration flows from the fact that lots of the growth comes from drug offenders serving very long, already imposed drug sentences leaving the federal prison system more slowly than new offenders serving moderately shorter new drug sentences enter the system.  If the percentage of prisoners serving old regime sentences that is released each year is small, and similar numbers of new prisoners are added to the system each year, reduced sentences for new prisoners has only a limited impact on total federal prison populations.  In the federal system, only about 28% of inmates at any given time are serving sentences of less than five years, and 3% are serving life sentences; the median inmate is currently serving a sentence of almost 9 years in prison.

This proposal does not even remotely return incarceration rates for drug offenses to those of 1980.

The U.S. population has grew by about 37% between 1980 and 2011, so we would expect the raw number of incarcerated drug offenders in 2011 to be about 26,000 in state prisons, 6,500 in federal prison, and 23,600 in jails, if the dramatic "drug war" increases in prison sentences and enforcement efforts between 1980 and 2011 had not taken place.  The actual number of state drug offense prisoners is about 9 times as great, the number of federal drug offense prisoners is about 14 times as great, and the number of jail inmates serving drug offense sentences is about 7 times as great, as they would be if the 1980 status quo had been maintained.

The population adjusted growth in drug offense incarceration since 1980 accounts for about 443,000 of the persons incarcerated in prisons and jails in 2011.  Increased incarceration for drug offenses accounts for about 29% of the growth in incarceration since 1980, and the growth in incarceration for drug offenses accounts (without regard to the 1980 baseline drug offense incarceration rates) accounts for about 20% of the total number of people incarcerated today.

The proportion of growth in total jail incarceration and growth in total prison incarceration, in each case, in excess of population growth, due to increased incarceration of drug offenders, is about the same for prisons and for jails.

"War on Drugs" or "War On Crime"

The "War on Drugs" was announced by the President of the United States in 1982, roughly coinciding with the surge in U.S. incarceration rates.

In all there 501,886 people in prisons and jails in 1980 (about 22,000 in federal prison, 298,000 in state prisons and 182,000 in jails), and there would have been 687,600 in 2011 had incarceration rates remained constant.  Instead, there were 2,228,400 people in prisons or jails in 2012, an increase of about 1,540,800 people incarcerated.  The overall U.S. incarceration rate per population has grown by 324% (i.e. has a bit more than tripled).

Parole is used modestly less frequently than it was in 1980.  In 1980, the number of people on parole from felony prison sentences was equal to 69% of the population serving such sentences; by 2012 that percentage had dropped to 57%.  If parole practices today were the same as there were in 1980, about 7% of inmates who are currently incarcerated (about 102,000) would be on parole instead.

The shift in the relative frequency of jail sentences and probation sentences has changed more modestly.  Both jail and probation are typically alternatives to each other in the case of misdemeanors and less serious felonies.  In 1980, 86% of the combined number of jail and probation sentences were to probation; in 2012, 84% of the combined number of jail and probation sentences were to probation.  If the proportions were the same in 2012 as they were in 1980, there would be about 93,000 fewer people in jails, i.e. the percentage increase in jail incarceration would have been 356% from 1980 to 2012 instead of 409%.

Overall, shifts away from probation and parole sentences account for about 13% of the increase in the incarceration rate.

Another factor in the increased incarceration rate is increased use of life imprisonment.  About 12% of state prison inmates (about one in eight) and 3% of federal prison inmates are serving life sentences and more than 30% of those life sentences are life in prison without parole sentences.  These figures, moreover, understate the reality, because significant numbers of prisoners are serving consecutive sentences for multiple offenses that combined exceed their reasonable life expectancy (as a prison inmate), and a significant subset of those prisoners have first dates when they are eligible for parole that also exceed their life expectancy (as a prison inmate) or arisen when they are very advanced in age.

Incarceration rates for women have grown 50% faster than for men since 1980, largely do to increased sentences and increased exercise of executive discretion to incarcerate people for offenses like drug offenses, illegally entering the country after being deported, and white collar crimes.

Thus, increased incarceration for drug offenses is an important factor, but only one of several important factors causing the United States to have the world's highest incarceration rate.

The U.S. has an incarceration rate 50% higher than third highest in the world Russia (#2 Rwanda has a slightly higher incarceration rate than Russia), and about five times as high as the nations like Spain that are on the high end of incarceration rates in Western Europe.  The U.S. incarcerates six times as many people per capita as Canada (the incarceration rate for whites in the U.S. is about 2.2 times that of the overall incarceration rate in Canada; the incarceration rate for black men in their 30s in the U.S. is about 10%).

In contrast, in 1980, the U.S. had an incarceration rate about 24% higher than Spain's today, and about 53% lower than Russia's current incarceration rate.

The total U.S. incarceration rate peaked in 2010 plus or minus a year and depending on how you count it, and has declined only modestly since then.

One way to think about the concept is that the average threshold of criminal deviance necessary to cause someone to be incarcerated has declined dramatically in the U.S. from 1980 to the present.

Of course, 71% of the growth in incarceration is due to increased numbers of people serving time for offenses other than drug offenses.  Much of what is described as a "war on drugs" is to a great extent simply a long term overall trend towards imposing long sentences for offenses.  While growth in drug offense incarceration has been more rapid than growth in non-drug offense related incarceration, there hasn't been a huge difference.

But, sentences for drug offenses have definitely grown more severe, particularly at the federal level.  According to a Sentencing Project report, linked above and cited throughout this post when other authority is not identified:
[I]n 1986, released drug offenders had  spent an average of 22 months in federal prison. By 2004, federal drug offenders were expected  to serve almost three times that length: 62 months in prison. . . . Most of these people are not high-level actors in the drug trade, and most have no prior criminal record for a violent offense.
The U.S. Sentencing Commission two offense level reduction proposal reduces the average drug offender's sentence by something less than 16% (i.e. 11/63+ months).

Now, just over 50% of federal prison inmates are drug offenders compared to 19% in 1980, while currently about 18% of state prison inmates are drug offenders, compared to about 6% in 1980, and about 22% of jail inmates are drug offenders, compared to about 9% in 1980.

Increased incarceration of drug offenders has had a far more dramatic impact on the character of federal prisons than state prisons, however.  In 1980, only about 8% of all people serving prison sentences were in federal prisons, while in 2012, about 14% of people serving prison sentences were in federal prisons, almost twice the "market share" in a shift driven mostly by increased incarceration for drug offenses, immigration offenses, and weapons offenses.

Overwhelmingly, the growth in the federal prison population is a policy choice that is not strongly driven by federalism concerns.  Roughly 75%-85% of federal prison inmates are incarcerated for offenses that would also be violations of state in almost every states and Indian Reservation.

This includes most federal convictions for drug trafficking offenses (50% of federal prisoners), weapons and explosive and arson offenses (16%), sex offenses (6%), about half of federal extortion, fraud or bribery crimes (half of 5%), burglary, larceny and property offenses (4%), robbery (mostly bank robbery) (4%), homicide, assault and kidnapping (3%), and racketeering (0.2%).

Many of the plain vanilla sex offenses, burglaries, larcenies, homicides, assaults and kidnappings end up in federal prison because the federal system handles federal crimes committed on Indian Reservations and in federal territory that is not subject to state criminal justice jurisdiction (including lots of federal parks in the American West).  But, the federal prosecutions for drug trafficking, weapons and explosive and arson, child pornography, kidnapping, robbery and racketeering overwhelming are instances where the federal crime actively prosecutes criminal activity within state boundaries that also constitute federal crimes (although the federal criminal sentences for the same conduct are frequently much more severe).

The largest category of federal prisoners who couldn't be convicted of crimes under state laws are immigration offenders (11% of  federal prisoners).

Prospects for S. 1410 and for a Guideline Change Rollback

Support for S. 1410, discussed by Talk Left, which would expand the availability of the mandatory minimum sentence "safety valve" from the Department of Justice and the Democratic majority on the Senate Judiciary Committee is only a beginning point.

In addition to needing support from a bipartisan group of 60 Senators to clear the Senate, also needs support from a majority in the relevant subcommittees and committees in the House, a majority of the rank and file members of the U.S. House of Representatives, and support from the GOP leadership controlled rules committee in the House that controls which bills make it to floor votes and on what basis the measure and amendments to it can be debated.

 The House Rules Committee is a particularly daunting challenge, since it tends to refuse to bring up bills that lack majority GOP support, even if they are backed by Democrats and a minority of Republicans who together have a majority of rank and file members of Congress in the House.

Republicans have been much more receptive to sentencing reforms in the last few years than at any other time since the early 1980s, or so.  But, it isn't at all obvious that reducing mandatory minimum drug offenses sentences will play well with the GOP base in a mid-term election year as we move into primary season. So, S. 1410 is still far from being a sure thing, or even from being more likely than not to pass, and it may be diluted further if it does pass.

 On the other hand, there is no indication that there is support in Congress sufficiently strong to roll back the proposed sentencing guideline changes for drug offenses.

Has it been worth it?

Crime rates that peaked in the mid-1990s have recently reached record lows.  Criminologists have made a wide variety of efforts to explain this phenomena, but there is very little doubt that some significant portion of that drop is due to incarceration of convicted criminals who would have committed more crimes had they not been in prison.

The low end estimates of this effect tend to be on the order of one-seventh of the total reduction in crime rate, while the high end estimates attribute the lion's share of the reduction to this effect.  I suspect that the truth is in the middle.

Conservative leaning criminologists and economists comparing crime rates and incarceration rates make the case that empirically, at the forest level of analysis, there is not good evidence to suggest that who you incarcerate matters nearly as much as how many socially deviant people you incarcerate.  For example, their models looking at links between U.S. crime rates and institutionalized population sizes tend to show (counter-intuitively) that mental health institutionalization which involved a population with a large proportion of women and many mentally ill people who were not a threat to others, had a similar effect to increasing the number of institutionalized people in prisons for crimes by the same number of people per population.

Liberals looking at the issue on more of a case by case basis, are inclined to think that we could receive similar crime reduction benefits with a much more selective approach, by focusing on incarcerating people who are psychopaths or violent or extreme repeat offenders, while taking rehabilitative and community oriented corrections approaches towards non-violent economically motivated offenders whose economic circumstances drive their woes.

Some liberals argue that any incremental benefits from incarceration are offset by the harm that mass incarceration does to whole communities breaking down their social fabric in a way the dramatically worsens crime in those communities.  For example, a recent study noted that victims of domestic violence in mandatory arrest of perpetrator jurisdictions (particularly minority women) had lower life expectancies than those in jurisdictions that did not, possibly in part because of the harm that those arrests disrupt fragile households that are adding value for those victims in a permanent way.  Conservatives are more comfortable accepting a "born that way" explanation of racial and ethnic differences in crime rates, while liberals are much more strongly inclined to look to context and environmental factors such as "broken communities'  for explanations.

Liberals also note that in international comparisons, the link between the size of the institutionalized population and the crime rate is not nearly as strong as it sees in the U.S. time series statistics.

Conservatives would attribute many of these international differences to ethnic culture as revealed by proxies like race, Hispanic origins, religious affiliation, and social class.  For example, racial and ethnic mix, and urban neighborhood character, are much more powerful predictors of homicide rates in geographic subdivisions prepared using regression models than economic indicators such as poverty rates, unemployment rates, and education, which can confound race and ethnicity variables.

Poor whites, poor Asians (e.g. Southeast Asian refugees) in communities with high rates of unemployment and low levels of education commit and are victims of far fewer homicides than poor blacks in communities that produce comparable economic statistics.  Communities of rural Yankee whites have less crime than economically comparable communities of rural Southern whites.  Central cities have much higher serious violent crime rates than economically comparable suburbs and rural areas.  Men without at least some college with a mental health problem and/or substance abuse problem are at very high risk of committing felonies that lead to prison sentences.  Women and men who have at least some college who are employed are more than an order of magnitude less likely to commit felonies that lead to prison sentences, even when they have mental health and/or substance abuse problems.

Raw incarceration and crime rates for Hispanic men are an intermediate case between black men and white men.  But, models looking at incarceration and crime rates for Hispanic men after controlling for education, unemployment, income and age explain a large percentage of the discrepancy between Hispanic men and white men in likelihood of committing felonies leading to prison sentences, for example, by doing so.  In contrast, controlling for these same factors cannot by itself explain the discrepancies between black men and white men in these statistical measures of crime involvement.  It could be that racism in the criminal justice process is dramatically influencing the outcomes of that process (although this is less plausible, or at least less directly a factor, in the case of metrics like homicide rates).  But, a race blind analysis does not explain the data.

Equally important, liberals tend to weigh the harm caused by property crimes and vice that does not involve violence less seriously than conservatives.  On the other hand, liberals sometimes underestimating the problems associated with having a class of people, who can give rise to organize crime organizations, who are powerfully driven to reoffend because their ordinary means of support involves illegal economic conduct and property crimes.  Impulsive murders are much less likely to reoffend after a modest prison sentence than professional drug dealers or people convicted of trafficking in stolen goods.

Conservatives and liberals alike, however, are increasingly recognizing that a criminal justice system that incarcerates and places people under correctional supervision at such unprecedentedly high rates is very expensive for taxpayers, and has other serious economic costs in the private sector as people waste away in idleness and become unemployable even if they were employable before they went to prison.

No comments: