Note: this post is adapted from remarks I recently delivered to a group of federal criminal practitioners.
Post-Booker, sentencing in the federal system is nominally governed by 18 U.S.C. § 3553(a), but I often get a sense of frustration from judges and lawyers about this statute. It seems that there’s really no “there there.” The statute provides a long laundry list of sentencing considerations, including just about anything that anyone could possibly regard as relevant to the business of selecting a punishment. It seems as if a judge could pick out a sentence anywhere within a wide statutory range and find a justification for it somewhere in 3553(a).
For lawyers who perceive 3353(a) this way, on either the prosecution or the defense side, I could readily understand why there would be some cynicism regarding post-Booker sentencing and perhaps even a disinclination to exert oneself much on sentencing advocacy.
But I’d like to suggest today that there is a plausible way of seeing 3553(a) such that it has a little bit more analytical structure, a little bit more definite content, than a simple laundry list. In order to develop this point, I’ll need to unpack some of the history of 3553(a).
This is a story in three parts.
Part I: Intellectual Origins of Federal Sentencing Reform
Through the middle decades of the 20th century, the dominant view of punishment in the United States was the view embodied in the Model Penal Code, which was adopted in 1962. In this view, the primary purpose of punishment is rehabilitation of the offender, with incapacitation a secondary aim for the minority of offenders who could not safely be given rehabilitative treatment through a probationary sentence in the community. This view of punishment was intimately bound up with discretionary, indeterminate sentencing and the parole system; incarceration was something that was supposed to be based on each individual offender’s particular circumstances and needs.
This view of punishment collapsed with remarkable suddenness in the 1970′s. It was repudiated not so much because conservatives finally vanquished liberals, but because the liberals themselves came to reject the rehabilitative paradigm. Their basic critique was that the commission of a criminal offense should not become an occasion for tinkering with a person’s psyche for an indefinite period of time, subject only to the whims of corrections officials who had uncertain personal agendas and biases. So, what the liberals said was, there has to be a cap on this — no more holding people indefinitely until they meet someone’s definition of being rehabilitated. But what was the alternative going to be?
The most influential response to this dilemma was arguably the limiting retributivism theory (LR) of Norval Morris.
Morris, one the most distinguished American criminal law professors of the second half of the twentieth century, first proposed LR in 1974. In essence, Morris’s idea was that the moral gravity of an offense established corresponding limits on the severity of the punishment. Morris conceded that moral gravity was not something that could be determined with scientific precision, but he argued that we could nonetheless usefully think about offense severity in terms of a range: this particular offense has at least X level of severity, but no greater than Y level. Within the range, other factors besides offense severity might be taken into account in order to determine the specific sentence to be imposed. Morris also argued in favor of the parsimony principle; on humanitarian and cost-effectiveness grounds, he felt that sentences should presumptively be at low end of the desert range, unless other considerations like deterrence or incapacitation clearly pointed to a sentence higher on the desert range. Finally, Morris was skeptical that there was much that could be said at the time about what other purposes required, so he contemplated that at least until the science of deterrence, incapacitation, and rehabilitation advanced considerably, the practical import of LR was that sentences would generally be at the low end of desert range.
Part II: LR and the Earliest Versions of 3553(a)
The primary legislative author and proponent of federal sentencing reform in the 1970′s and 1980′s was Senator Ted Kennedy. Kennedy introduced a series of sentencing reform bills, beginning in 1975 and continuing until sentencing reform was finally adopted in 1984.
How do we get from Morris to Kennedy?
Kennedy himself said that the original model for his legislation was a proposal developed by the Yale Law School sentencing workshop. Reflecting the liberal critique of rehabilitation and Morris’s theory of limiting retributivism, the Yale proposal authorized sentencing judges to consider all of the traditional purposes of punishment, but included an express prohibition on sentences that were disproportionate to the gravity of the offense (proportionality being a retributive or just deserts idea).
Kennedy then took these concepts from the Yale proposal and translated them into legislative language that would later become 3553(a).
It is fascinating today to read the original 1975 Kennedy bill, which includes the same mandate that we find in 3553(a) for the sentencing judge to “consider” a list of various purposes and types of information. But the list was much shorter in 1975 than what Congress actually adopted in 1984.
The primary purpose listed in 1975 was retributive: “The need for the sentence imposed to reflect the seriousness of the offense and promote respect for law by providing just punishment for the offense.” This is, of course, almost verbatim what we see in today in 3553(a)(2).
What’s missing from the 1975 list? The incapacitation, specific deterrence, and rehabilitation purposes that are in the current list.
So, going back to the beginning, we see a much clearer and narrower focus, not on everything under the sun, but on retribution—just as Morris was advocating.
The only other clear, mandatory purpose that we see identified in the 1975 bill was general deterrence. However, it is not clear that deterrence would have been seen by Kennedy as a competing purpose that might trump retribution. A point commonly made by influential sentencing reform advocates in the 1970s, like Alan Dershowitz, was that deterrence was more a function of the certainty of punishment than the severity. Kennedy himself expressed the view in a 1979 law review article that sentencing disparities were a concern because they robbed punishment of its certainty and thereby undermined deterrence. So, it is possible that the real thrust of the deterrence provision in the ’75 bill was to reduce sentencing disparities. This purpose is today codified at 3553(a)(6), but the ’75 bill contained no precursor to this language — perhaps because the author felt that the deterrence mandate sufficiently addressed disparity concerns.
These observations about the ’75 bill give us some basis for seeing retribution as not just another item on the laundry list, but as something that lay at heart of the original intent of 3553(a).
We should note, too, that it is particularly appropriate to go back and look at the ’75 bill post-Booker, because the system originally envisioned by Kennedy was basically the post-Booker system: advisory guidelines with sentences subject to reasonableness review on appeal.
Part III: The Actual Language Enacted in 1984
In some ways, the final language moved away from Morris’s LR ideal.
The list of sentencing considerations was much longer in 1984 than 1975, and so retribution no longer stands out as it once did. And Morris’s bête noir, rehabilitation, has been added to the list.
On the other hand, added also has been the language now appearing at the start of 3553(a), which reflects Morris’s parsimony principle and gives a special, elevated status to the (a)(2) factors, including retribution. The language reads, “The court shall impose a sentence sufficient, but not greater than necessary, to comply with the purposes set forth in paragraph (2) of this subsection.”
We can also see Morris in 18 U.S.C. § 3582, which tells us that imprisonment is not an appropriate means of providing correction and rehabilitation. This is diminishing the status of rehabilitation relative to the other (a)(2) factors.
The Argument in a Nutshell
As originally conceived, 3553(a) was the centerpiece of federal sentencing reform, and core purposes of 3553(a) were (1) to knock rehabilitation off its pedestal as the dominant purpose of sentencing, and (2) to elevate retribution as an important, perhaps even the most important, purpose of sentencing.
Beginning in the late 1970’s, 3553(b) was added to legislation. The idea of mandatory guidelines emerged as the centerpiece of federal sentencing reform, and 3553(a) was relegated to secondary importance.
Since Booker has knocked out 3553(b), though, doesn’t it make some sense to look back to what the legislation was trying to accomplish before 3553(b) was added? If I’m right about that, then we could plausibly see 3553(a) today, not as an unstructured laundry list, but as an articulation of LR, with just punishment based on offense severity as the key determinant of sentence severity. Among other things, this would have important implications for the role of offender characteristics (including criminal history) in the 3553(a) calculus.