Booker and Beyond Analyzing Sentencing Reform and Exploring New Research Directions

DARYL FOX: Good afternoon, everyone. Welcome to today’s webinar, “Booker and Beyond: Analyzing Sentencing Reform and Exploring New Research Directions,” hosted by the National Institute of Justice. For notification about today’s webinar, the opinions, findings, and conclusions or recommendations expressed here are those of the presenters themselves and do not necessarily reflect the positions or policies of the National Institute of Justice of the U.S. Department of Justice. At this time, it’s my pleasure to introduce Dr. Nancy Merritt, Senior Policy Advisor within the National Institute of Justice, for some welcoming remarks and introductions of today’s speakers.

NANCY MERRITT: Thank you, Daryl. And thank you all for joining us for the webinar. I’ll be moderating the discussion today along with Raven Lewis, a PhD candidate at Rutgers who we’re happy to have working with us at NIJ as a research assistant.

Today’s webinar is entitled “Booker and Beyond: Analyzing Sentencing Reform and Exploring New Research Directions.” The U.S. Supreme Court’s 2005 Booker decision provides the starting point for today’s conversation, but the discussion will address criminal justice sentencing reform research in general. The webinar presentation will address selected research findings from the last 15 years with particular attention paid to the analytic methods used. We’ll then discuss the strengths and weaknesses of these methods and implications for future research. Importantly, the webinar and follow up discussion will not be limited to Booker research. One of the primary goals of the webinar is to encourage audience members to think about new ways of approaching reform research and how future research efforts could be improved through the use of new analytic tools and methodologies, improve training, and enhance access to data.

We’re very fortunate to be hearing from three experts today, all of whom have extensive experience studying and reporting on sentencing reform. We will begin with a presentation by Dr. Jeffrey Ulmer, Professor of Sociology and Criminology at the Pennsylvania State University. Dr. Ulmer received his PhD in sociology from Penn State and conducts research in areas including courts and sentencing, criminological theory, and symbolic interactionism, religion and crime, and race disadvantage, and violence rates. He’s received funding for his research from the National Science Foundation, the National Institute of Justice, and other organizations in addition to numerous awards for his research and publications.

Dr. Ulmer will be followed by Dr. Mona Lynch with the School of Social Ecology at UC Irvine, where she serves as interim dean and chancellor’s professor of criminology, law and society with an appointment in the school of law. Dr. Lynch is a social psychologist by training. Her current research addresses plea bargaining, criminal sentencing, and punishment processes with a focus on institutionalized forms of bias within the criminal legal system. As an NIJ W.E.B. Du Bois scholar in race and crime, she’s currently studying the effects of implicit bias instructions on jury decision-making. She’s published widely and is the recipient of numerous research grants and publication awards.

Finally, we’ll hear from Dr. Paul Hofer. Dr. Hofer has a PhD in research psychology from Johns Hopkins University and his JD from the University of Maryland School of Law. Dr. Hofer worked at the Federal Judicial Center in Washington, D.C. for a decade before moving to the U.S. Sentencing Commission as special projects director in 1995. After working as a Soros Justice fellow, he returned to federal service in 2009 with the Federal Public Defender Sentencing Resource Counsel. He retired from the federal courts in March, but has continued his 35 years as adjunct assistant professor in the Department of Psychological and Brain Sciences at Johns Hopkins.

I’ll ask the presenters some brief follow up questions after the presentations, then open up the floor to the audience for additional questions. Please forward your questions as instructed at the beginning of the webinar and Raven will pass them on for discussion. We will address as many of the questions as possible during the webinar. Thank you again for your participation. And now I’ll turn it over to Dr. Ulmer. Dr. Ulmer.

JEFFREY ULMER: Thank you. So as Nancy said, I’m a professor of sociology and criminology here at Penn State University where we have a long tradition of studying courts and sentencing, and where also the Pennsylvania Commission on Sentencing is based. So we have here at the university, the Sentencing Commission of the—of the state of Pennsylvania. I’d like to talk to you today about four studies of federal sentencing primarily, that start with the issues raised by the Booker decision and claims about disparity increasing after the Booker decision, but then moves on to other more recent research that looks at other ways that the guidelines work and don’t work, and other mechanisms of disparity, other locations of disparate—of racial and ethnic disparity primarily, that are much more based in the social context of district courts. To start with the issues raised by Booker and in a pair of articles by myself and Michael Light, and John Kramer.  John Kramer was the executive director of the U.S. Sentencing Commission in the late ’90s. But he’s also long been a professor, now professor emeritus of sociology and criminology here at Penn State.

I’m following the Booker/Fanfan decision and the Gall decision and others, this expanded judicial sentencing discretion relative to the period right before the Booker decision, and also the period of the ‘90s and early 2000s. Many feared that this increase in judicial discretion would cause an increase in disparity. And the 2010, 2011 U.S. Sentencing Commission reports found that racial disparity in sentence lengths did indeed increase post-Booker and post-Gall, especially post-Gall 2007. We, my colleagues and I, provided an alternative analysis to the U.S. sentencing report, which pursued some different kind of methodologies, some different statistical assumptions. And we also went somewhat beyond the U.S. Sentencing Commission report in the two studies of 2011. We widened the comparison time periods to not just the period of time that the U.S. Sentencing Commission report compared to - which was the period of time from 2003 to 2005 under the more restrictive guidelines regime, under the PROTECT Act and the Finney Amendment, which restricted judicial discretion even more so than it had been before. The guidelines were, if you want to think about it this way, more mandatory than ever before during that period, and then their report compared the post-Booker and post-Gall period with the pre—with this 2003 to 2005 period we went all the way back to pre-2003 and even into the ’90s. And we found that levels of racial, ethnic, and gender disparity, while present, was generally comparable to the pre-2003 levels when the guidelines were in fact still mandatory and not just advisory. It was just different from the 2003 to 2005 period when the guidelines were really restrictive. And we found that American males’ odds of imprisonment increased significantly post-Gall, but sentence lengths did not. And we found that immigration cases accounted for a significant portion of the sentence length disparity affecting Black males. And then questioning the focus on just judicial discretion, we found that government-sponsored departures below guidelines, that is departures below guidelines that were endorsed by the U.S. attorneys’ offices, the federal prosecutors were a greater source of racial disparities than judge-initiated deviations from the sentencing guidelines.

So the upshot is we argued that these findings raised questions about whether the decisions, Rita or Gall, Booker, caused increased disparity. And ultimately raising questions whether the guidelines must be mandatory in order to  limit racial, ethnic, and gender disparities.

Let’s get at the second study. This was a study done by myself and Brian Johnson at the University of Maryland based on a survey of federal judges we did from 2005 to 2007. And when we fielded the survey, unbeknownst to us, we captured the period of time right after Booker. We captured the immediate aftermath of Booker in our 2005 survey. And then we looked at 2005 to ’07 sentencing data. And we had qualitative interviews as well with U.S. attorneys, judges, and defense attorneys. It captured this period of uncertainty after Booker, but before later decisions clarified what the advisory guidelines really meant. And it was a time of great variation among judges in the perceived constraint and normative authority of the guidelines. And that’s exactly what our survey captured. We found that during this period, at least, guideline conformity - in other words, judges following the guidelines versus departing, deviating from the guidelines - were strongly influenced by judges’ perceptions of the guidelines as normatively legitimate. In other words, their perception of the guidelines as the right—normatively right sentencing standards versus being inappropriate sentencing standards - being too severe, too harsh, not legitimate. And they also depended on the perceptions of the constraint from their circuit courts in terms of appeals of guidelines issues. And their perceptions of their local U.S. attorneys’ offices and their actions, and their likelihood of appealing, or their pursuit of government-sponsored departures or their plea agreement practices. So these three kind of informal factors strongly influenced judges’ conformity to the guidelines, and in fact influenced their conformity to the guidelines more than anything else - more than any other attitudes of the judges or characteristics of the judges.

And so a key theme of this was that guidelines—maybe guidelines do not primarily exert influence through constraint, through force, but rather through normative influence - through being seen as appropriate benchmarks, best practices. And by making sentencing easier and reducing uncertainty, thus, perhaps guidelines, the perceived legitimacy and appropriateness of the guidelines is very important to judges following the guidelines in an area of advisory guidelines. Which by the way, other sentencing guideline systems such as Pennsylvania have always been advisory in the way that the federal courts guidelines are advisory.

Here’s a study by myself and Noah Painter-Davis and Leigh Tinik in 2016. This study looked at federal court data and Pennsylvania State court data, Pennsylvania being the guideline state. And we wanted to see how much of disparity, racial and ethnic disparity, came from factors that are determined before sentencing. So we looked at the raw differences in Black, Hispanic, or Latino men versus White men and the sentences they get, and then you start controlling for or holding constant the different things that under guidelines affect sentencing and mandatory minimums, and then how much of the disparity, - the differences between these groups - is due to these things that are decided before sentencing begins. And we talked about these distinctions using characterizations developed by Shawn Bushway and Brian Forst of type A discretion - which is the discretion exercised by local judges and courts. And type B discretion - the discretion that’s of policymakers that’s built into sentencing guidelines and other sentencing policies.

We found that we were able to explain most of the disparity in the federal data with things that happen before sentencing begins. The guideline factors, mandatory minimums, case processing factors—how one was convicted whether by trial by guilty plea, and other extralegal factors, importantly, such as citizenship and education. By the way, in the Pennsylvania analysis, we could not explain as much of the variation the Black, White difference—and Hispanic differences in state court data with things that happened prior to sentencing, thus indicating that there was still some disparity or unexplained differences at the sentencing stage.

Two—the two last studies I’ll talk about are—the theme is that disparity, racial, ethnic and other forms of disparity is complicated and it’s bound up in the social context of local courts, something of a theme of mine throughout my career. In an article by myself and Mindy Bradley, we looked at the sentencing of Native American defendants, and the complexities of tribal and federal justice—of federal jurisdiction crime in Native American lands. The article goes into a great deal of detail about the nuances and complexities of this—of this tribal-federal justice relationship and this organizational coupling between tribal justice and federal justice through different federal funding mechanisms, like the 2010 Tribal Law and Order Act, the Major Crimes Act, federal jurisdiction and so forth. We found that the more tightly bound, the more tight the relationships between tribal courts and federal justice, federal law enforcement, federal—the Federal U.S. attorneys’ offices, the tighter the participation in these programs, the more sentencing severity increased for Native Americans. So Native American defendants got harsher sentences, the more tribes cooperated with federal justice. Maybe that was intended, maybe that was not, but that was the effect. Also, interestingly, because of tighter the linkage between federal law enforcement, U.S. attorneys’ offices and tribes, we saw greater substantial assistance and government-sponsored departures below the guidelines for Native Americans. These are sentencing guidelines departures that are controlled by or initiated by the prosecutors to foster substantial assistance to law enforcement, or reward other behavior, the government-sponsored departures that the prosecutors would like to move for a more lenient sentence. So this makes sense that the U.S. attorneys’ offices may have been using these mechanisms to get more cooperation, more assistance for Native American defendants in the prosecution of other Native American defendants.

Finally, this is a very recent study, myself and Brandy Parker. We looked at Hispanic defendants’ sentencing disadvantage in different types of immigration destinations. In sociology, sociologists have studied the differences in the—in the types of immigration destinations and the reception that Hispanic immigrants receive there. We thought that perhaps Hispanic citizens and non-citizens would receive harsher sentences in some immigrant destinations than others. We found that there was no Hispanic-White disparity in traditional immigrant destinations in either time period - early 2000s or 2010 to ’12. Traditional destinations being New Mexico, Southern California, South Texas, South Florida, places like that. There was greater disparity in the 2000s, in around the 2000 census. Around the turn of the 21st century, Hispanic citizens and non-citizens received longer sentences in places that were new destinations, places where there was a new wave of Hispanic immigration - places like districts in North Carolina, Indiana, Missouri, Arkansas, Tennessee, Pennsylvania, and so forth. Places that saw new immigration—Hispanic immigration in the 1990s. And also in non-destinations - places where there was little Hispanic immigration - Hispanic defendants were harshly treated there. By the early 2000s, there was no disparity in the traditional destinations and in what you might think of as new destinations, emerging destinations. Places that had seen a lot of Hispanic immigrations from 2000 to 2010. But again, especially for Hispanic non-citizens, in the new destinations around the 21st century and especially in the non-immigrant destinations - the places where there was little immigration - Hispanic non-citizens were sentenced much more harshly. And this was especially true of undocumented non-citizens. So Hispanic-White sentencing disparity and for Hispanic citizens and non-citizens varies greatly by—is affected, correlated strongly with, levels of local immigration in a district.

So the upshot of these four studies I’ve explained to you and I can expand later but federal sentencing disparity existed before Booker, it exists after, and it’s tightly bound to social context, thus defying simplistic explanations of disparity that rely on guideline changes that affect judicial discretion. You know, if we restrict judicial discretion, we will get less disparity. This research really problematizes that. Thank you.

NANCY MERRITT: Thank you very much, Jeff. That was a great presentation. It really shows us the importance of looking at court and community context in studying this decision-making process. I’m going to save my questions for you until we get to the end of all the presentations and turn it over now to Dr. Lynch.

MONA LYNCH: Well, thank you. First, let me just to thank you for having me and for being part of this and to the candidates for coming. This is a long time coming, we’ve been trying to organize this for a while and really happy to see it come through and really pleased to be with two people whose work I’ve admired for a very long time and who I’ve learned from over the years as I’ve kind of dabbled into the federal sentencing world.

So let me just begin by saying my approach to—the way I came to Booker was really through my interest in law and society scholarship and thinking about how policy gets put into action on the ground. And what are the ways that institutional actors respond, resist, kind of, you know, reinterpret what the policy is supposed to do, and I’ve done this kind of work in a number of different kinds of settings and came to the federal court sort of late in my career of doing this. So when I approached the first project—on the Booker case, I was thinking about a couple of things. One was, how are policy changes responded to and resisted in varied and creative ways by legal actors? And how do those variations manifest as local norms and practices that transcend the policy change? So kind of a tension point that I thought about is what—in what ways do local actors in their local context try to maintain some stability in the face of policy change? And Booker is a really nice example or a nice case to study because it was really imposed upon these disparate district courts - I think 92 or 94 districts, there’s divisions within districts that have their own local norms and so on, and this came from the Supreme Court to sort of say, okay, now a lot of discretion is going from prosecutors to judges in federal sentencing. So, Jeff talked a little bit about these issues around kind of local norms and local contexts and local culture. Those are things that also really resonated with me in this approach.

So the first the study that I did was a quantitative study of the federal sentencing data and what I was looking at was just drug cases. Now, I was interested in drug cases particularly because they’re really sort of subject to two sets of sentencing logics. One is the sentencing guidelines that was prior to Booker. Basically mandatory judges were stuck with the sentencing guidelines except under very limited conditions. And then the mandatory minimum statutes, which set a baseline for some sentences, and the mandatory statutes or what we call mandatory minimums are most prevalent in drug cases. The mandatory minimums were not affected by Booker, so those bottoms still sat there and as certainly as Jeff’s work has shown and others have shown, that really mattered in terms of what happened after Booker. So I was interested in how these two kinds of sentencing logics were working after Booker and I asked a series of questions about how drug cases were adjudicated and outcomes pre- and post-Booker. So, the first question I asked was, do within-district drug case sentencing patterns demonstrate stability across different policy periods indicating the influence of local norms? And the question there is if you look at a given district, is it behaving like itself over time versus being impacted really strongly by these outside policy changes and then kind of flipping its behavior?

The second question is, are the mechanisms for getting to sentence outcomes changing in response to policy reforms? And I suspected that, yes, there would be this sort of motivation or sort of push towards stability that would mean that legal actors would have to figure out how to get to where they wanted to get but they’d have to work within these new rules. I know certainly a lot of research across in a number of different institutions that would suggest that that’s the case. And then third, are cases that are not subject to mandatory minimums, - the ones that are simply being freed from prosecutorial control in terms of charging and allowing judges to have a lot more discretion in terms of sentencing - are those more likely to vary from the guidelines compared to those that are subject to mandatory minimums?

So I published that work with, at that point, my graduate student Marisa Omori, who is now an assistant professor at University of Missouri-St. Louis, back in 2014. What we found is that, yes, in fact there is a fair amount of stability across the different policy periods within districts indicating that local norms did matter and that there this pressure to maintain outcomes in drug cases even with Booker. And we looked at several different policy periods as Jeff just described that there are a number of periods in terms of how restrictive the guidelines were that we looked at and basically what we found is that there is a pressure to maintain stability. And we did find some evidence, as did others, that the mechanisms for getting to those sentencing outcomes did change in response to policy reform. We found the real effect of using mandatory minimums as a way to get to those outcomes or prosecutors charge more mandatory minimums and to control the judicial discretion. And finally on the question of - are cases that are not subject to mandatory minimums more likely to vary from the guidelines compared to those that were subject to them? We found no, actually no evidence of that because there was this sort of overarching pressure to maintain the norms.

So here is just a really simple graph that shows the percentage of the guideline minimum sentence that’s imposed nationally beginning back in the early guidelines period, ’92 all the way through 2012, the end of the period for which that study was done. And you can see there’s really not that much change—if the guideline minimum is a hundred percent, drug cases are pretty much being sentenced to about 85 percent of the guideline minimum across all periods. So one kind of uptick is that PROTECT Act period that Jeff just mentioned. So again, a lot of stability and within that, if you just break it out even regionally, you’ll see that the South has the highest or the closest to the guideline minimums across all periods. The Midwest is second in terms of that. The West is third and then the Northeast is really the most generous in terms of giving low sentences or breaks on the guideline minimum. So you see the most change in—or the most reduction from the guideline minimum in the Northeast —so there’s a sense of both regional norms as well as district level norms.

So I followed that study by doing a qualitative study in four districts that I selected for their distinct features and I did observations of how drug cases were adjudicated and I did interviews with legal actors. And part of that is really pretty retrospective at this point, it was in 2013 that I did the study, so Booker had started to take effect and shape things by 2005. But there were a number of actors who were there during that period who I was able to talk to about how the Booker case changed local practices. I asked a series of questions to those actors who had been in the system at that time about specific legal strategies that had changed at three different stages of the criminal process - how charging changed, how plea negotiation changed, and how sentencing had changed.

So, while Professor Ulmer’s study done in 2005 captured that there’s a lot of anxiety and uncertainty about how Booker is going to shake out, by this point when, I was interviewing people, people were very clear about sort of that trajectory what had happened and how it had changed norms and culture. And in fact, they did mention that there was a lot of uncertainty at first and there was a fair amount of—kind of fidelity to the guidelines in the early period right after Booker in all four districts.

So what I found was that there are multiple effects of Booker and there was a fair amount of local variation within that. So across all four districts, there is an increased charging of mandatory minimum charges and mandatory minimum enhancements which is another way of getting at these sort of mandatory bottoms that prosecutors have, - especially in drug cases and in child porn cases. So that was really prosecutors’ effort to create a bottom and sort of stop that discretion from going lower than what they hoped it would be. But an even more important finding in terms of doing qualitative work was that threats to use mandatory minimums and mandatory enhancements were used quite frequently in plea bargaining - particularity in drug cases and child pornography cases, which have a set of laws that allow for those threats for very similar kinds of conduct. So those threats don’t show up in those outcome data that we used to look quantitatively at what’s going on in the federal system but it became very clear that this is a common practice in plea bargaining - which is to use the threat of mandatory minimums or enhancements to—for prosecutors to get where they want to go in terms of sentence severity. And within that, there is quite a bit of bargaining in the shadow of the assigned judge, so judges became much more important since they were less constrained and so defense attorneys would have the kind of upper hand in certain kinds of cases where they knew that the judge was sympathetic and would be likely to sentence below the guidelines. Prosecutors held the upper hand in cases where the judges were more, say, guidelines-centric. But generally, there was also a move towards a bit more use of what are called binding pleas to tie the judge’s hands if the plea is accepted. So the judge would have to accept the binding plea but if the judge was willing to accept the plea, he or she would be limited in terms of the sentence that could be imposed—by the plea bargain that’s made.

And then finally, a really interesting finding - and this was only in my Northeastern districts, so I did kind of a regional variation. There was a real move to forgo plea agreements at all. So the prosecutor sort of got knocked out of the process and with the right judge, a defense attorney would just take the case, you know, plea straight up, take the case to the judge, put on a very robust sentencing hearing and see what the judge did on that. And by the time I was doing my field work, this was relatively common. This had become a new cultural norm where that hadn’t existed at all prior to Booker. And again, these are the kinds of things that would be harder to detect in the quantitative data but are really important for understanding how adaptation happens and how policy lives on the ground shall we say.

So the last set I want to talk about is one that doesn’t have anything to do with Booker, but it was really motivated by the same kinds of questions and it’s to ask the question, what about internal policy changes? So again, the Booker case is sort of an exogenous policy change and forced upon these district actors who may or may not agree with it and may or may not chafe at being told what—told about these new rules. But the Department of Justice also has its own internal policies that changed -often sometimes dramatically - with changing administrations. And we had a perfect example of this in 2017 under then Attorney General Jeff Sessions who basically undid a number of internal policy reforms that Holder had instituted under Obama. And the most significant one at least for drug cases was that Holder created a policy that asked U.S. attorneys not to charge mandatory minimums in cases that met certain kinds of guidelines for being “low level drug cases.” That policy was instituted in 2013, then Sessions came in and rescinded that and encouraged prosecutors to use all tools—use all mandatory minimums, they were told to always use mandatory minimums, to use mandatory enhancements and plea bargaining and so on. So, this is a real change from the kind of move that the Obama administration was trying to do to lessen the number of people in federal prison for drugs. And so what happened after Sessions did that? Well, the use of drug mandatory minimum shot up and my last slide will show a picture of this. The likelihood of prison increased for drug cases, drug sentence lengths increased, and the effects—and this is the most interesting thing I think we found was the effects were significantly driven by the appointment of a Trump-appointed U.S. attorney in a district, so that person could kind of impose a new policy in the local office. And what this told us is that really takes translation of the policy to the office and that sort of convergence around the ideals of the policy to really make a policy sort of be wholeheartedly taken up and put into place. Now there was a lot of variations based on other factors or regional variations or variations around population issues and so on, but this was a very interesting finding for us and an important one to raise the question about. You know, what role middle-level actors play in terms of policy implementation? So I’ll just end with the slide here of our findings. So this is a proportion of drug cases sentenced to binding mandatory minimums and you can see prior to the first, there was an early Holder memo that basically gave more local control to prosecutors and rescinded some of the prior Bush policies around drug cases. There’s a drop in the mandatory minimums that’s really become dramatic after the Holder 2013 memo and then this very quick shooting up of the use of mandatory minimums to actually exceed where they were just a tiny bit under the Bush era. So with that, thank you.

NANCY MERRITT: Thank you very much, Dr. Lynch. I The last slide really shows the importance of looking, not only at changes in the law, but changes in the policy over time. As you pointed out, the adaptations that occur under different circumstances can change the way implementation of the law plays out.

I’m going to turn now to Dr. Hofer who is going to give our last presentation which will be followed by some questions and answers. Dr. Hofer?

PAUL HOFER: Well, thanks, Nancy and NIJ for organizing the webinar and greetings to our online participants and Jeff and Mona whose work I’ve followed for many years.

As Nancy mentioned in her intro, I’ve been analyzing sentence reform for my entire career because I happened to begin at the Federal Judicial Center in 1986 - right as the Commission and the guidelines were coming online, and instead we’ll be looking to the Commission and to the public defenders, I’ve been analyzing the system from several different angles. My remarks today are going to elaborate on chapter in the 2019 volume of Crime & Justice series, edited by Michael Tonry, and a number of other papers, one in criminology and public policy, Booker as a natural experiment, and a number of articles in the federal sentencing report where I reviewed different reports by the Commission.

My main conclusion from analyzing sentencing reform from all these different angles is that it’s never really been tried in the federal system. The original conception of the SRA was as a good government progressive reform, it set up an independent expert agency to develop policy based on research. But right away even in the original act itself, congressional micromanagement began to creep in and it’s continued throughout the guidelines era, I mean, especially in drug cases and sex offenses against children, economic crimes, immigration offenses, and of course most notably the sentencing reformat was betrayed before the guidelines even took effect in 1986 with passage of the Anti-Drug Abuse Act, it created these mandatory minimums for drug offenders based on the quantity of a mixture of substance containing a detectable amount of various drugs.

This stopped the research-driven process of drug development in its tracks. Those policies were incorporated into the guidelines and in some ways made worse. And this was the single greatest contributor to the tripling of the federal prison population. And the growing gap between sentences imposed in Black and White defendants. It opened up right after supposedly sentencing reform took effect. Most defendants today are not sentenced under guidelines that were developed by the Commission in what Supreme Court case of Kimbrough called the exercise of its characteristic institutional role, in other words as an independent expert agency.

When Booker came along, it kindled the hope that increased judicial discretion might revive some aspects of the original vision of reform. One hope was it would allow more departures based on offender characteristics that the Commission had deemed not ordinarily relevant to the sentencing decision and, indeed, judge sponsored below range sentences did increase by nearly 10 percent as shown here on this chart at the time of the Booker decision, the vertical blue line. They continued a gradual increase until 2015 and then had largely leveled off. This was a general loosening of restrictions on the kinds of departures that judges were doing or wanting to do before Booker. Departures based on the defendant’s family circumstances or age or physical or emotional conditions. For many of us, the biggest hope created by Booker was the judges might initiate a new kind of review of the guidelines, looking not just for unusual factors in the case or about the defendant, but reviewing the pedigree and the research support for the various guidelines, the most important of which were not based on a—the research has not really validated as being fair and effective. The idea was the judges could reject unsound guidelines, even in routine cases. We did see some of this review of the child porn guidelines for some types of drugs such as crack and ecstasy. Although, it’s notable that in the entire period, even when the Commission was saying the hundred to one ratio between powder and crack cocaine was unfair, and the Department of Justice was not defending it, there wasn’t a single circuit court that found that the quantity ratio was unreasonable. It only got changed later, when the Commission and eventually Congress in 2010, lowered, or raised rather, the quantity threshold for crack to make the ratio 18 to one instead of hundred to one.

The third big hope was created by Booker, was that Commission and Congress might be driven to amend the guidelines that were not developed and validated with research. I think Booker did help empower the revision of that crack guideline, but it largely stopped there. I can’t say that judicial revolt against unfair guidelines has led to additional amendments after that one, so the optimism fostered by Booker has been largely extinguished. The USSC itself took some steps to reinforce the guidelines, they created this three-step process where you need to calculate the guideline range and then consider the recommendations in the manual for possible departures and only then get to a third stage where you might evaluate the guideline recommendation itself in light of the purposes of sentencing.

That kind of administrative-like review was discouraged by many appellate courts, who argued that judges weren’t in a position like the Commission to develop sentencing policy, even though the Commission often doesn’t get final say on sentencing policy. And I have to say there was some apathy among district judges, many of whom have only known the guideline system and just assume that there must be some foundation for the guideline recommendations or some evidence that they are fair and effective. It’s worth noting, I think, that in the last quarterly report from the Commission, there’s only 43 percent of defendants who are sentenced within the guideline range. The Commission has kind of obscured this fact in recent years because now they report a new number of cases sentenced under the guidelines manual, which means they were within the range or it was a departure for a reason that’s discussed in the manual. But despite this increase in sentences below the range, it has not translated into any reduction in the portion of defendants going to prison. As we can see here, there’s a long-term trend towards increasing imprisonment. And, frankly, in 2005, Booker decision didn’t create a blip. Nor has it resulted in a marked reduction in average sentence length. This chart shows in blue, means sentence lengths from 2001 through 2017. And in orange, it’s the guideline minimum recommended by the guideline manual. Even though there were additional sentences below the guideline range after Booker, which you’d think would reduce sentence lengths, those reductions were offset by increases in the sentences recommended by the guidelines in drug cases and in white-collar crime cases. The average sentences began to fall only years later, not due to Booker, but mostly due to reductions in revisions and guidelines themselves. Notably, two points off of the Commission applied to old drugs and drug quantity table and the reduction of the threshold for—or the increase in the quantity threshold for crack cocaine.

So in summary, the mandatory minimums and guidelines continued to dominate the trends of the types and lengths of sentences imposed, since this move in tandem with the guideline minimums. Many have noted that this reflects the continuing gravitational pull of the guideline recommendations. The SRA did not give control of the sentencing policy to an independent expert agency. But it did create a centralized system sentencing rules that Congress, and to some extent the Commission, has used to control—largely control sentencing policy, Booker didn’t reverse that central fact. This Booker is—this webinar is about Booker and Beyond, and I’ve kind of entered the beyond phase of my career. Now, I’m also somewhat beyond hope that the guidelines can be salvaged. To do so, it would require repeal of the mandatory statutory minimums, repeal of the scores of specific directives that Congress has given to the Commission to increase sentences or set sentences as certain level and it would require a reinvigorated Commission dedicated to evidence-based policymaking and comprehensive amendment of the most used guidelines.

Now, some have suggested that even if the policymaking role of the Commission has failed, its research mission remains viable. Certainly the Commission collects and makes available a great deal of data. But the Commission’s failed in its research mission, in my opinion, and increasingly so in recent years. This means that outside researchers must rise to the challenge and do the critical evaluations of federal sentencing policy that the USSC has lately been avoiding. It’s worth remembering what a strong research mandate the USSC gave the Commission. It was charged with developing means of measuring the degree to which sentencing penal and correctional practices are effective in meeting the purposes of sentencing. In 2004, I authored the Commission’s 15-year review, and it was explicitly focused on evaluating the success of the sentencing reform and achieving the purposes of sentencing reform. The question of how well the guidelines actually reflected fair, proportionate, and effective sentencing was supposed to be the next phase of evaluation, but it’s really never happened. If one reviews recent Commission reports, you’ll see that they largely declined to critically evaluate the guidelines or even the statutory minimums. The reports are full of charts and factoids, most of which are of unclear relevance to any policy decision and are often confusing or downright misleading. I don’t have time here to detail all the problems. But somebody needs to do this and I think the research community should rise to the occasion.

For example, there was a report just this past March on armed career criminals. These are people convicted of a gun offense, who if they have two prior, violent, or drug offenses receive some—a 15-year mandatory minimum and some of the longest sentences in the system. Three quarters of these defendants are African American. This was intended as an incapacitation measure. So its success should depend on how well it is actually targeting dangerous offenders. But this report reported that 59 percent of these defendants are re-arrested after eight years, that’s lower than your re-arrest rate for a defendant that’s just in category three out of six criminal history categories under the guidelines. In other words, it’s not targeted anywhere near on the most dangerous or most high-risk defendants but there’s not a critical word or recommendation in the report suggesting that this guideline is poorly targeted. We found a similar problem with a 2016 report on career offenders. Although there the Commission did recommend some statutory change, it recommended that radically increased sentences of the career offender guideline shouldn’t occur when the predicate offenses, the past offenses are only drug offenses. But it didn’t make a strong case for it, it didn’t lay out the strongest possible evidence and nothing has happened. I could go on with more examples, the Commission issued a fact sheet on a report, the Commission issued last year on the supposed deterrent effect of 10 year sentences, I direct you to find the factsheet from the Federal Public Defender’s.

The bottom line is that outside researchers need to do this critical evaluation that the USSC neglects. A research-focus on which defendants are at highest risk of future violence, and which relatively low risk offenders are sentenced far too long. It’d be nice to have a crimes averted model that could measure the incapacitation effects of alternative policies. I started working on such a model when I was at the Commission, but nothing has happened to it subsequently. We need proportionality studies. What guidelines over-punish relative [INDISTINCT] of the crime. We’ve talked a lot about the drug guidelines. And I got to tell you, this is a national embarrassment and disgrace. I have a paper at SSRN that demonstrates that the current quantity thresholds found in the guidelines of mandatory minimums lead to nothing less than bizarre results. Where less harmful drugs, for example, Sudafed - which you can be prosecuted for if you possess with an intention to make methamphetamine, it’s punished more harshly than something like heroin. Heroin by far the most harmful drug by most public health measures. As you can see here, smaller numbers of doses of things like crack, or meth, or MDMA, ecstasy receive five-year sentences and 3,333 doses of heroin. The sentences imposed don’t reflect the harmfulness of the drugs and it doesn’t reflect the social costs of different drug crimes. I’ve got a fact sheet that used some data developed by the Department of Justice on the cost of different kinds of crimes. And from it, you can learn that powder cocaine, trafficking offenses with social costs of between a 120- and 475,000 dollars receive a five year sentence - the same as an economic crime, the social costs of seven to eleven, or eight—twelve million dollars. So it’s not that drug crimes are being sentenced too harshly, it’s that they’re sentenced—drug crimes are being sentenced far more harshly than other types of crimes, given the social cost involved in them all.

I’m out of time. I’ve got some more things that I could say about sentencing disparity. That’s certainly the topic that’s received the most attention from research scholars. I would argue that we don’t really need a lot more multiple regression studies of discrimination on the part of judges because what we’re asking is in the rules themselves. Local cultures are important, absolutely. But the actual policies represented by the guidelines and laws have driven this huge gap between White and Black defendants. It has come down in recent years, largely because the revision to the crack guideline, and we need more work demonstrating other places where laws with extreme adverse impacts, but little effectiveness are driving this form of racial disparity. And I’ll stop there.

NANCY MERRITT: Thank you very much, Dr. Hofer. That gives us a lot to think about. One thing, while I’ve got you right here - your presentation repeatedly called upon the public to conduct additional analyses using sentencing guidelines data - so that there are other sources of analyses, rather than just the Sentencing Guidelines Commission itself. Are you familiar with any training resources for new researchers in this area? I have been told that the data sets can be somewhat complex. In the absence of someone attending a university where they have a mentor who can help them, how would you suggest that a new researcher acclimatize themselves to the data sets and learn how to manipulate the data for accurate research?

PAUL HOFER: Well, it’s certainly true if a researcher starts by going to the Commission’s website and downloading the annual individual data files that are made available there, they could easily get around because those data files are incredibly lengthy, complicated, and unnecessarily so, frankly. I mean, in a recent year, one of the data files had over 14,000 variables. Most of these variables are of absolutely no interest to researchers. They occasionally are needed by the Commission in order to recreate exactly how the guidelines were applied to a particular case. But I have, and the Federal Public Defender’s, boiled down those datasets to a few hundred variables that are really important. And that’s really enough for most—I wish the Commission or somebody would make available a research dataset that avoided some of the really administrative types of various variables that are in that flat file that you download from the Commission. Most of whose cells are empty.

If it’s possible to get someone to give you a boiled down dataset, preferably combining across years, and a good codebook that explains the variables clearly, that would be a tremendous start and could really potentially save somebody a month of work trying to figure out what they need and how to make it workable. The Commission sponsored a workshop for researchers and has a number of publications that can help, but they aren’t as clear as would be nice. And they haven’t sponsored a workshop for, I think, over 10 years. The NIJ sponsored a workshop a few years ago. And I think that would be very helpful to give someone an orientation, someone who has some experience, if you can get a mentor or find someone to work with who can help you avoid some of the pitfalls that would be a good idea. I’m retired now. I mean, you can drop me an email if you really have some questions, I might be able to help because I have been working with the dataset for quite a while.

One thing I would say is it’s really important to get familiar with how the law of sentencing works and how the guidelines work. You can’t just look at the variables. You’ve got to understand the guideline system. There have been some excellent researchers who’ve made some tragic mistakes, working with the dataset, not understanding sentencing. There was an economist a few years ago, who joined in the racial disparity debate and refined the control variables that were used to control for legally relevant differences among defendants. And he started using the cell of the sentencing table as the control, but he didn’t realize that sometimes a mandatory minimum comes in and trumps that cell and so he didn’t account for that. So he was controlling—he was expecting—his model expects the defense to be sentenced at the bottom of the cell that was the predicted regression equation, predicted to be sentenced at the bottom of the cell. When a lot of defendants and, they are disproportionally African American, have a—trump that comes in and the judge hasn’t any choice to sentence even within the cell in some cases has to impose the mandatory minimum. If you don’t properly control for that you get race effect that jumps out that’s actually a mandatory minimum effect and you blame judges for disparity that’s actually created by Congress and the mandatory minimums. So you really got to understand how the system works, and one last thing I would say is that some of the most important research that needs to be done, we need data that the Commission is not making available. They have a dataset that shows the recidivism of defendants. There’s nothing more valuable to evaluating the effectiveness of the different sentencing policies than to have that recidivism dataset. The Commission argues they can’t make it publicly available. I don’t see why they can’t somehow mask the identity of the defendants. We’ve tried to get them to release it, so far they’ve refused. I’d like to see continuing pressure to make additional available—data available in addition to the individual datasets that are now downloadable from the Commission’s website.

NANCY MERRITT: Okay. Thank you. It sounds like the best bet is to find a learned mentor in terms of working with the data.

PAUL HOFFER: It could…[MISSING]

NANCY MERRITT: thank you for that. We’ve been talking mainly about the federal guidelines here, but let me jump over to Dr. Ulmer for a second and ask --- Dr. Ulmer you talked about your federal research, but I know you’ve done a lot of work with state level sentencing guidelines to perform research. Could you tell to us why it’s important to look at both state and federal sentencing structures rather than just one or the other? What the different findings are—and what each would show us?

JEFFREY ULMER: Yes. Well, the federal and state systems, institutions are both very impactful and important in—but in different ways. The federal system is, first of all, highly institutionally important and symbolically important. It’s what the federal justice and the federal government does. It’s what the Department of Justice—it’s what their U.S. attorneys are, you know, acting on. It covers the entire country and secondly the federal punishments are very severe and deeply impactful to defendants who get very long imprisonment sentences. So the impact on individual’s lives is very large. And third it’s sociologically interesting because you have this one system, one legal set of policies and laws that tries to fit over top of 50 states and 94 federal district courts in all their variation and distinctiveness throughout the country. You know, every place from Vermont to Guam and on the state side, state courts are very important. Number one because that’s where most defendants are. The large majority of criminal defendants are sanctioned in state courts and so it’s impactful in terms of the sheer numbers and size and state courts handle everything from minor misdemeanors to first degree murders. And states do a variety of things, some states have guidelines, some do not. Some have traditional sentencing systems where there’s virtually no structure to sentencing. Others have flat sentencing or statutory minimums and maximums. And then two, states and state courts at the county level or municipal level courts have an enormous amount of less visible discretion. Less visible discretion over everyday lives of millions and millions of people, so that’s—in that way the state court systems are very important study.

NANCY MERRITT: Thank you. That really shows why we need to study state and federal sentencing, even though we focused on federal sentencing for this presentation. Dr. Lynch, I wanted to ask you --- the research that you’ve discussed today, and a lot of your research, shows that you qualitative analysis along with your quantitative analysis - using key stakeholder, interviews, observation, etcetera. Why do you think it’s important to use qualitative as well as quantitative data, and what kind of additional insights does this provide you?

MONA LYNCH: Thanks for that question, that’s a great question. And it kind of goes to the points that Paul just made about people not knowing what they’re looking at when they’re using quantitative data. I know a fair amount about how courts work and about things like plea bargaining and so—and I worked in the system. My husband is a criminal defense attorney so I knew enough to understand sort of how that mechanics work behind, but I still found myself after doing the—that first Booker study. When I first went into the field I got a grant after that to go into the field and do the observations and interviews. I felt like, “Wow, I wish I had known more.” And had some of the knowledge that I had after doing that, when I did the quantitative study. And it really actually triggered me thinking about using the data in completely different ways. So I’ve done subsequent studies that really think about those data away from just, you know, the standard multiple regression, thinking about, you know, places and important—how do you—how do we reconsider a place, how do we reconsider getting at the processes which is really the tricky part. It’s really, you know, you could—you could well measure outcomes and you can find, you know, you can kind of specify some things that look like processes, but how can we do that better? So that’s the first reason to do it - because it actually gives you that nuance, understanding of what you’re doing with the—with the datasets. And I went through the training—I went to the NIJ Sentencing Commission Training at ISPCR, so I wasn’t completely naive going into it. But more than that, it’s really—it really gets sort of how do we measure social and social psychological processes and I’ve—I think Jeff will—Jeff knows I’ve been frustrated with some of the assumptions made using quantitative analysis about what’s underlying, what are the psychological and sociological processes happening? What are judges thinking? What are prosecutors thinking and so on, that you just can’t get to with the—with the administrative data and while interviews are—have their own limitations using multiple data sources and multiple approaches really triangulates around the problem, right? So you bring in—you add to your understanding with different forms of data and with the observations, those are very good to see—to be able to play off of what you’re hearing in interviews or what people are telling you. What’s it—what’s it actually look like in these different settings? What does it look like when a defense attorney is talking to a prosecutor about resolving a case? What’s it look like when they’re in front of the judge and the—in that kind of formal setting and what are they telling you about their strategy there? Those things can really give you a clear sense about how law actually works in action.

NANCY MERRITT: Thank you. That’s very helpful. I think that really gives us a better understanding of how the two types of analysis fit together.

We have about 12 more minutes. So let me turn to some of the audience questions. The first one, I think pertains to any of the panelists. And the question is --- in your experiences, is there much inter-judge variation in response to legal changes within jurisdictions, relative to between jurisdictions? Have any of you had the chance to really look at that?

MONA LYNCH: Jeff you can jump in. I can talk about like qualitative, kind of, qualitatively and let that…

JEFFREY ULMER: Please do, please do.

MONA LYNCH: Okay. So…

JEFFREY ULMER: You have the inter-judge more than I have.

MONA LYNCH: Yeah. And there are some good studies that have looked at this that have been able to get—so the data from the Commission doesn’t code for—I mean they do—they have secret cases as Paul referred to that have actual level data. So you could get to the judges, but we’re not—most people are not privy to that. Some researchers have been able to get that and really look at that and then there have been some open data sources, Massachusetts District, or Massachusetts in particular, where people have looked at sort of the inter-judge disparity. So there is some quantitative work on there, but qualitatively the—I think that two things operate together within a given jurisdiction even within a division, there’s real—there’s differences in norms. There’s, sort of, a range of ways that judges think about cases, but overlaying that is, sort of, you know, sort of what is imaginable as a, you know, incredibly harsh outcome versus an incredibly good outcome and those things differ by place. So, you know, what’s a good outcome? And I was struck by this. I went from a very—my northeastern district to a southern district where the norms are completely different and I was shocked to hear what was a good deal, you know, what was a good judge outcome. Even with the variation in judges that just the ranges are constrained by that kind of local context. So it’s really a matter of within district variation as constrained by, you know, the district context.

NANCY MERRITT: Okay. Great. And do any of the rest of you have comments or can we move on? Okay. I’ll move on. Oh, go ahead, Dr. Hofer, were you going to say something?

PAUL HOFER: I would say that there is some quantitative research on inter-judge disparity. It happens to be, kind of, a neat methodology sentence cases are assigned to Judges randomly in many courthouses. We have the opportunity to look at whether the differences among judges in the same random assignment pool change in response to policy changes and I just have to study for the 15-year review and it found that the guidelines did modestly reduce of some of these differences among judges even within the same courthouse. And there’s been some work since Booker, unsurprisingly, some of those differences have come back with the additional—some judges are more willing to use other departure power or to even challenge, you know, whether a guideline is good in a routine case. I still think that that kind of disparity is far less problematic than the one—the racial disparity for example that’s built-in by grossly adverse impactful and ineffective and unfair rules.

NANCY MERRITT: Okay.

PAUL HOFER: Because local culture is important, but actual rules also too make a big difference. I mean if we could get them amended it would really help.

NANCY MERRITT: I understand. You’re pointing out something that’s very important - that to study these reforms well, and really understand them, you have to both look at the local context and the rules and how they work together. I think that’s very important in this realm of research. Another question we have has to do with disparity in different types of cases. This person asks whether there are differences in adherence to the guidelines or the reform in terms of victim versus non-victim crimes? And then, within the subset of victim crimes, is there a disparity across those crimes and whether any of you look at that or are familiar with research in that area?

MONA LYNCH: I can say that the…

JEFFREY T. ULMER: I can’t think of anything that’s looked at that breakdown.

MONA LYNCH: Yeah. And I would say that the federal system data is not the best place to look at that because of—particularly the violent crimes are very odd subset of—it’s just not—they’re not—your kind of regular street victim crimes are not likely to come to federal court except you’ve got the Indian—yeah, the Indian country cases where you do have it. So you have these regional differences, you have this population differences, and jurisdictional differences that really matter. I just want to say that Paul’s earlier point, I totally agree on the—on the rules. The rules to me are the framework for all of that—that shape all of that and against which people sort of operate, so the rules really do construct what the realm of possibility is. So I’m always talking within those rules.

NANCY MERRITT: The next question I think speaks to that, and also some of what you were talking about earlier Dr. Lynch. The question may be difficult to respond to—to provide a simple response, but the questioner wants to know how you would recommend learning about sentence structures. Given the complexity and the differences across structures, and the need to really understand them before you jump into the research, what’s the best way to familiarize yourself with those structures?

MONA LYNCH: Jeff should start with that one because he’s trained so many students and has been so on the ground doing this stuff qualitatively and quantitatively.

JEFFREY ULMER: Well, thank you. But, yeah, your—you’ve done as much and more. It’s hard to—it’s hard to point to one source, you know, one thing you should look at, but I suppose you should start with review articles. Articles that talk about, you know, articles written by independent scholars, like, Paul has talked about that—such as Richard Frase, Michael Tonry.

PAUL HOFER: Kevin Reitz.

JEFFREY ULMER: And others. And that talked about—first the development and—oh, Kim [INDISTINCT] the legal scholar Kim [INDISTINCT] and Jose Cabranes, their work, but the evolution and structure of, you know, how the guidelines came to be what they were, then review articles like in the 20—2019 Crime and Justice volume that Paul talked about. That Paul has a piece in and I have a piece in. And there are reviews of disparity research in that volume. And then third I, you know, just read the key studies. Read the Booker research studies that they did back and forth, the impact of Booker, you know, read the—you know, read the reports, the U.S. Sentencing Commission reports and compare and contrast those with other research people did. Finally I would say, like, Mona said and, you know, there’s just a ton of, you know, the federal sentencing world and the guidelines and how attorneys plea bargain with the guidelines and, you know, the role of probation and it’s just so incredibly complicated. You know, one of the best things you might do is talk to a federal defender or two. You know, I’m convinced that people like federal defenders, and assistant U.S. attorneys, and judges, and tribal court liaison. I’m convinced that these are the—only people that truly understand how this all works. So federal probation people too, so I would say, you know, talk to people that actually do it.

MONA LYNCH: I would echo that. It’s a starting point. I—there are—there are times that one can read articles and you didn’t know that the person actually never even sat and watched what goes on in federal court. You learn a lot through observation and talking to people and it’s really important to understand the kind of basic ground rules for how law gets put into action. That you can’t just get from articles because, because the articles aren’t, you know, written for that purpose. And so really getting to know your subject is important.

NANCY MERRITT: Thank you. I would agree with all of that. Did you have anything you wanted to add to that Dr. Hofer?

PAUL HOFER: No, I think that covers it.

NANCY MERRITT: I think that’s great advice. We are at—just about at the end of our session, so I want thank you all, you’ve been great.  And Raven, thank you so much. You’ve made my job easier working in the background. I want to remind the audience that this webinar has been recorded and will be posted on the NIJ website.  So, you’re welcome to come and take a look at that. Also, while you’re there please take a look at the funding and training opportunities as well as the NIJ research that’s been completed. I do hope you join us for future webinars. Thank you all and goodbye.

PAUL HOFER: Thank you. Bye.