Blog
Understanding the Federal Sentencing Guidelines
The federal sentencing guidelines do not sentence anyone. A judge does, in a courtroom, after considering a set of calculations that most defendants encounter for the first time in the weeks following indictment.
Most people who search for information about sentencing guidelines are not engaged in academic research. They are reading because a number has been attached to someone they care about, or to themselves, and the number is larger than they expected. The guidelines did not produce that number in isolation. A probation officer produced it, a prosecutor endorsed or contested portions of it, and a judge will consider it alongside factors the guidelines do not capture. That sequence matters more than any chart.
The Sentencing Reform Act of 1984
Before 1984, federal sentencing was indeterminate in a manner that permitted sentences for identical conduct to differ by years depending on the judge, the district, and factors that no one recorded. Congress perceived this as a failure of uniformity.
The Sentencing Reform Act created the United States Sentencing Commission, an independent agency within the judicial branch, and charged it with constructing a system that would constrain judicial discretion without eliminating it. The original proposal contained three hundred and sixty offense levels. The Commission reduced that figure to forty three, organized along a vertical axis and crossed with six criminal history categories along the horizontal. The resulting table, first published in 1987, assigned a sentencing range in months to every intersection of offense level and criminal history. For a defendant at offense level 20, criminal history category I, the range was 33 to 41 months. At the same level with a category VI history, the range became 70 to 87 months.
The system was mandatory. Judges were required to sentence within the applicable range unless specific departure grounds existed, and those grounds were narrow. For nearly two decades, the guidelines operated as a binding constraint on judicial discretion, which was precisely what Congress intended.
How the Calculation Works
Every federal offense carries a base offense level, which functions as the starting integer in a sequence of adjustments that will determine the final number. For drug trafficking, the base level is keyed to type and quantity. For fraud, it begins at a level reflecting the nature of the scheme and then ascends according to loss amount, with the loss table alone occupying over one hundred pages of the guidelines manual. For violent offenses, the base level corresponds to the specific statute of conviction.
From that starting point, the level moves. Specific offense characteristics add or subtract levels: the use of a firearm, the number of victims, the sophistication of the scheme, the quantity of contraband beyond what the base level already captured. Chapter Three adjustments account for the defendant’s role in the offense (a leadership role increases the level; a minimal role reduces it), whether the defendant obstructed justice, and whether the defendant accepted responsibility, which in practice means whether the defendant entered a guilty plea and cooperated with the presentence investigation.
The criminal history category is calculated separately. Each prior conviction receives points. A sentence exceeding thirteen months adds three points. Sentences between sixty days and thirteen months add two. Shorter sentences add one. Additional points accrue if the defendant committed the current offense while under supervision or within two years of a prior release. The total determines the category, which ranges from I (zero to one point) to VI (thirteen or more).
Once both numbers are fixed, the sentencing table produces a range. The intersection is mechanical. A defendant at level 24, category II, faces 57 to 71 months. At level 24, category IV, the range is 77 to 96 months. The difference between categories, at the same offense level, can be measured in years. The difference between offense levels, even a shift of one or two, compounds.
What the table cannot convey is the degree to which each of those adjustments is contested. The base offense level is rarely in dispute. Everything after it frequently is. The government’s calculation of loss, the defendant’s role, the application of enhancements for vulnerable victims or sophisticated means, and whether the conduct the government attributes to the defendant qualifies as relevant conduct at all: these are the arguments that determine the range, and they occur before the judge has exercised any discretion whatsoever. In our practice, the objections to the presentence report’s guideline calculation are often the most consequential filings in the case.
The Presentence Investigation Report
After a conviction or a guilty plea, a probation officer is assigned to prepare a presentence investigation report. The report is, if we are being precise, not a neutral document. It is a recommendation dressed as an assessment. The officer interviews the defendant, reviews financial records, contacts family members, visits the defendant’s residence, and compiles a criminal history. The officer then applies the guidelines to the facts of the case and arrives at a proposed offense level, criminal history category, and sentencing range.
The defendant and the government each receive a draft. Objections are filed. In many districts, the probation officer resolves what can be resolved and flags the remaining disputes for the court. The final report, with its addendum of unresolved objections, is what the judge reads before sentencing.
For the defendant, the presentence report is the first encounter with what the system perceives them to be. Somewhere in those pages, a life has been converted into points: three for a conviction at nineteen that resulted in fourteen months, two for the probation sentence in another state, two more for committing the current offense while on supervision. The total is a category. The category is a row on a table. The row intersects a column, and the column produces a range of months that may bear little resemblance to what the defendant understood the plea agreement to mean.
Whether the probation officer’s judgment on contested enhancements is correct or a product of institutional conservatism is a question worth considering.
We prepare our objections to the draft presentence report with a specificity that the guidelines themselves invite. Every adjustment is traced to the applicable commentary. Every enhancement is tested against the facts the government can actually prove, not the facts the government has alleged. The line between relevant conduct and charged conduct is where many of the most productive challenges occur, and in districts where probation officers tend to defer to the government’s version of events, those challenges are the mechanism by which the actual guideline range is established at sentencing, not in the pages of the draft report.
After Booker
In January 2005, the Supreme Court decided United States v. Booker and rendered the guidelines advisory. The constitutional holding was straightforward: the Sixth Amendment requires that any fact increasing a sentence beyond the range authorized by a jury verdict or guilty plea must be found by a jury beyond a reasonable doubt. The mandatory guidelines permitted judges to find such facts by a preponderance of the evidence. That could not stand.
The remedy was less straightforward. Rather than require jury findings for every enhancement, the Court excised the provision making the guidelines mandatory. Judges must still calculate the guidelines. They must still consider them. They are no longer bound by the result.
The expectation in some quarters was that sentences would decrease across the board. They did, though the picture is uneven. The rate of within-guidelines sentences fell from somewhere above eighty percent to something in the vicinity of half, depending on the district and the year. Judges began to exercise a discretion they had not possessed since 1987. The guidelines remained the starting point for every sentencing, which is a different thing from being the ending point.
Variances and the Section 3553(a) Factors
The argument that matters most at sentencing, in the years since Booker, concerns whether the guidelines produce a just result for this particular defendant.
Section 3553(a) of Title 18 directs the sentencing court to consider seven factors: the nature and circumstances of the offense; the history and characteristics of the defendant; the need for the sentence to reflect the seriousness of the offense, promote respect for the law, provide just punishment, afford adequate deterrence, and protect the public; the kinds of sentences available; the applicable guideline range; any pertinent policy statement; and the need to avoid unwarranted sentencing disparities. The court must impose a sentence “sufficient, but not greater than necessary” to serve these purposes.
In fiscal year 2024, courts applied variances in thirty two percent of sentencing proceedings. Departures, by contrast, appeared in only four percent. The practice of sentencing had already shifted toward variance as the primary mechanism for non-guidelines sentences, and the Commission’s 2025 amendments acknowledged that shift by eliminating most departure provisions from the operative text of the guidelines. The sentencing process is now two steps: calculate the range, then consider the 3553(a) factors.
The simplification is welcome. It does not change the substance of what a defense attorney must prepare. A variance argument is not a list of sympathetic facts appended to a sentencing memorandum. It is a theory of the case reframed for a different audience. The trial, or the plea, asked whether the defendant committed the offense. The variance argument asks whether the guideline range, correctly calculated, produces a sentence that is just.
We begin constructing the variance argument at the moment of engagement, not in the weeks before sentencing. The client’s employment history, family circumstances, medical records, contributions to the community, the conditions that preceded the offense: all of this is gathered and organized into a narrative that the court can weigh against the number the guidelines produce. A variance argument that has been constructed with care does not ask the court to ignore the guidelines. It provides the court with a reason, grounded in the statute, to impose a sentence below the range.
Some judges are receptive. Some are not. In the Eastern District of North Carolina, the variance rate was seventeen percent in 2024. Certain districts exceeded forty percent. The disparity between districts is itself one of the tensions the guidelines were created to resolve, and it persists.
I am less certain about whether the 2025 simplification will reduce that disparity than the Commission’s language suggests. The elimination of formal departures changes the vocabulary of sentencing. It does not change the judge.
The 2026 Proposed Amendments
The Commission has proposed amendments for the cycle ending May 2026, and several concern economic crimes. The loss table under Section 2B1.1, which governs fraud, theft, and embezzlement, would be restructured from sixteen tiers to eight. Broader tiers mean fewer disputes over whether a loss figure falls on one side of a threshold or the other. The DOJ has described that trajectory as sending the wrong message.
The tension is familiar. The Commission sees a system producing disproportionate outcomes. The Department sees a system producing deterrence. Both positions contain something the other does not.
There is also a proposed amendment addressing post-offense rehabilitation (conduct between conviction and sentencing that demonstrates commitment to change), which would create a formal mechanism for courts to credit that conduct. Defense attorneys have always argued for the credit. The proposed amendment would move the argument into the structure of the guidelines, which may or may not make it more effective. The Commission is also revisiting methamphetamine purity distinctions, which have operated on assumptions about culpability that the drug market no longer supports.
These amendments, if approved and not rejected by Congress, take effect November 1, 2026.
The Space the Guidelines Cannot Occupy
The federal sentencing guidelines are, at their best, a system that aspires to treat similar conduct with similar consequences. At their most mechanical, they are a system that converts human behavior into integers and cross references those integers on a table, as though a table could capture what a sentencing hearing is meant to resolve.
The gap between those two descriptions is not a flaw. It is the system itself. Every sentencing hearing is an attempt to close that gap, to argue that the number the table produces does or does not account for what happened and who was involved. The guidelines provide the frame. The judge, the prosecutor, the defense attorney, and the defendant occupy the space within it.
The question is never whether the guidelines apply. The question is what they miss.
A consultation is where that conversation begins. The first call costs nothing and assumes nothing. It is the beginning of an assessment that the guidelines, for all their precision, cannot perform on their own.