Trump’s Circumvention of the Justice Department Clemency Process

Jack Goldsmith, Matt Gluck
Tuesday, December 29, 2020, 1:49 PM

An updated tally of Donald Trump’s grants of clemency shows how extensively the president circumvented the normal pardon process to serve his personal and political goals and whims.

President Trump in the Oval Office (Official White House Photo by Shealah Craighead).

Published by The Lawfare Institute
in Cooperation With

Last week, President Trump granted 49 pardons and commutations. This brings his total number of grants of clemency to 94. The new grants confirm a judgment we made six months ago (after Trump had granted clemency to 34 people): The clemency system is dominated by insider access to the president and almost exclusively serves the president’s personal and political goals and whims.

We have significantly updated our chart of pardons and commutations in light of these new ones. The chart seeks to track patterns about the likely aims of and explanations for the president’s pardons—including whether they advance Trump’s political agenda, were influenced by a personal connection to Trump, had a connection to something Trump saw on television or were based on Trump’s admiration for a celebrity. Our judgments are based on (and thus are limited by) what is in news reports; our categorizations thus could shift based on additional reporting. While the general pattern of a corrupt pardon process is clear, judgments in individual cases are often qualitative and inferential. We have linked the stories on which we rested these judgments, and readers can decide for themselves.

Here, we seek to drill down on the scale of backdoor access and self-serving pardons by looking at the extent to which Trump’s 94 pardons and commutations circumvented the standard pardon attorney process in the Justice Department.

(Note that Trump has actually granted clemency to just 87 people. But he has granted clemency 94 times, sometimes to the same person in stages. For example, on July 10, he commuted Roger Stone’s 40-month prison sentence, two-year supervised release and $20,000 fine; and then, on Dec. 23, he issued Stone a full pardon.)

It is not news that Trump has disregarded the role usually played by the Office of the Pardon Attorney. The Washington Post reported in February 2020 that White House advisers, led by Trump’s son-in-law Jared Kushner, had taken “more direct control over pardons and commutations, with President Trump aiming to limit the role of the Justice Department in the clemency process.” And the Post reported just last week that “rather than consult with the Justice Department’s pardon attorney for recommendations, Trump has routinely subverted the process and largely favored political allies and those who are well-connected for clemency.” Lots of evidence, detailed in columns D-G in the chart, indicates that personal contacts with Trump explain a lot of the pardons.

What is missing, however, is a focused judgment in each of the 94 cases whether the pardon attorney weighed in with a positive recommendation. As we show below, and in column C in the chart, in most cases there is concrete evidence—based on information in a public Justice Department database—that the pardon attorney did not recommend clemency, even before one looks at the evidence about the characteristics of the clemency grant and how it served Trump’s goals.

The Justice Department Pardon Process

Since the 19th century, the Justice Department has processed pardon requests and advised the president on various forms of clemency, including full pardons and sentence commutations. The current system is governed by Justice Department regulations and is managed by the department’s pardon attorney, who reports to the White House through the deputy attorney general. This process is not constitutionally compelled. Presidents have the authority to issue pardons without consulting the Justice Department (or anyone else). Many prior presidents have circumvented the pardon attorney or significantly altered the process—and there’s nothing necessarily wrong with that. This typically happens, for example, in clemency actions that relate to high-stakes presidential policy decisions, such as Gerald Ford’s pardon of Richard Nixon, or Barack Obama’s clemency initiative (which in part led the pardon attorney to resign). Other times, the president circumvents the pardon attorney in more questionable contexts, such as many of Bill Clinton’s last-day self-serving pardons. The aim of the pardon attorney process is to regularize and standardize pardon considerations, especially for cases where the sentence is completed; to ensure integrity in the consideration of pardons; and to protect the White House from being inundated with pardon requests and from issuing unvetted and potentially controversial pardons.

The regulations specify that “a pardon is granted on the basis of the petitioner’s demonstrated good conduct for a substantial period of time after conviction and service of sentence.” The Justice Department has a strong (but waivable) presumption against even considering an application for pardon until “at least five years after the date of the release of the petitioner from confinement or, in case no prison sentence was imposed, until the expiration of a period of at least five years after the date of the conviction of the petitioner.”

In determining whether a petition for a pardon should be granted, the pardon attorney considers four factors: (1) outstanding post-conviction conduct, character and reputation; (2) the seriousness of the offense (with more serious offenses requiring longer waits, and special hesitation pardons for “a prominent individual or notorious crime”); (3) the extent to which petitioner has accepted responsibility for the crime, displayed remorse and sought atonement; and (4) the “purpose for which pardon is sought may influence disposition of the petition.”

The process for vetting a pardon petition for favorable consideration is a lengthy one and typically takes more than a year and often several years. The FBI has to investigate and write a report on the case. The U.S. attorney in the district of conviction and the sentencing judge, and occasionally the assistant attorney general in charge of the Justice Department component responsible for the case, are consulted and they weigh in as well. “The views of the United States Attorney or Assistant Attorney General are given considerable weight in determining what recommendations the Department should make to the President,” according to the regulations. A similar process of consultation takes place with commutation petitions before the pardon attorney considers making a favorable recommendation. After receiving this input, the pardon attorney makes recommendations to the deputy attorney general. The petitions that the deputy approves are “presented for action to the President with a report and recommendation from the Department, and the substance of the recommendation by the United States Attorney or Assistant Attorney General is included in this report.” There are other layers of review. (We thank Margaret Love for guiding us on the intricacies of the pardon attorney office and process.)

Many people think this Justice Department process is broken—that its criteria for clemency are misplaced or too demanding, and that the process takes far too long. Some criminal justice experts have applauded Trump for skirting the pardon attorney to grant clemency in at least a few instances. We take no position on these normative questions here; we simply use news reports and the known factors about the Justice Department process to assess which of Trump’s pardons likely were based on pardon attorney recommendations. Obviously, one might think that the Justice Department process is broken and that Trump’s insider method is unattractive as well.

Assessing the Justice Department’s Involvement

The Justice Department has a searchable database of all pardon or commutation petitions filed since 1989. Each petition is assigned a docket number. The date of the petition is not indicated, though the docket numbers are in chronological order. When a petition for clemency is filed with the pardon attorney, it is automatically and invariably entered into this database. If a case is not docketed in the database, then it has not been filed with the pardon attorney—and the pardon attorney thus could not have recommended clemency to the president in that case.

Of 94 grants of clemency made during Trump’s tenure, 41 lacked a petition in the database (marked “never in database” in the chart). These include some of the most prominent and controversial grants of clemency, including those issued to Roger Stone, Paul Manafort, George Papadopoulos, Michael Flynn, Charles Kushner and the four Blackwater contractors—Nicholas Slatten, Paul Slough, Evan Liberty and Dustin Heard. Persons in this category clearly were not recommended by the pardon attorney. This leaves 53 out of 94 cases to be explained.

There are other cases in which the person granted clemency had a petition in the system, but that petition did not support the clemency granted. First, the person pardoned might have had a petition in the database that was marked denied, and then Trump subsequently granted a pardon without submission of a new petition. This is what happened, for example, with Patrick Nolan, whose petition for clemency was denied in 2008 but who received a pardon from Trump on May 15, 2019, without submitting a new petition.

Similarly, the person pardoned might have at one time submitted a petition that was “administratively closed” as a final action (which means that Justice Department consideration stopped for some reason, typically for lack of follow-up by the petitioner) and did not submit a new petition before Trump granted the pardon. This is what happened, for example, with Conrad Black, whose petition for commutation was administratively closed in 2009 and who received a pardon from Trump on May 15, 2019, without submitting a new petition. We also include the pardon grants to Dwight and Steven Hammond—who had petitions for clemency in the database, and were granted clemency, but who did not petition for pardons, which they also received—in this category. And we include both Alice Johnson’s commutation and her later pardon in this category since her earlier petitions had been denied and neither of her clemency grants was supported by a petition from the pardon attorney. We designate these cases as “no pending petition” and concluded that they could not have been recommended by the pardon attorney. There have been 15 such cases. This leaves 38 cases to be explained.

In addition, Trump sometimes granted partial clemency to someone who had a petition pending, but then later granted clemency again without the submission of a new petition. An example is Crystal Munoz, whose first petition for commutation was denied in 2016. Munoz applied again during the Trump administration and on Feb. 18, 2020, received a commutation of the remaining 14 months of her prison sentence. She then received a second commutation (issued on Dec. 22, 2020) of her supervised release without submitting a new petition. This last grant of clemency was not supported by a new petition and thus was almost certainly not recommended by the pardon attorney. There are three such “no new petition” examples in the chart. That leaves 35 cases to be explained.

We conclude that an additional four cases were probably not recommended by the pardon attorney based on the following logic. As noted above, docket numbers for petitions are not dated. But we are reliably informed that any case assigned a docket number higher than 291,000 would likely have been filed after June 2020 and thus far too late (labeled as “too late” in the database) for a possible pardon to be fully investigated and a recommendation sent to the White House in the ordinary course. This is a very conservative estimate: The pardon process typically takes longer than a year, and our ignorance of actual filing dates for petitions numbered up to 291,000 means we cannot capture in this analysis the petitions filed in the first half of the year that are also almost certainly too late.

Of the people in the database who were granted pardons or commutations, four (Christopher Wade, Cesar Lozada, Joseph Martin Stephens and Christopher II X) had docket numbers higher than 291,000 and thus were too late to be processed in the ordinary course. Of course, the process can go more quickly in the exceptional case if the White House intervenes and insists. But given the timing, and the White House’s articulated disdain for the pardon attorney process—making it less likely that the White House would find value in a quick Justice Department approval—we conclude that these four were likely not recommended by the pardon attorney. That leaves 31 cases to explain.

Seven of these 31 cases, we think, were probably recommended by the pardon attorney. CNN reported that the pardon attorney recommended five people whom Trump pardoned on July 29, 2019—Michael Tedesco, Roy Wayne McKeever, John Richard Bubala, Chalmer Lee Williams and Rodney Takumi. In addition, these five decisions satisfy the traditional pardon attorney criteria, and the petitions for these decisions were filed with enough time for the pardon attorney process to run its course.

We think two other clemency decisions were also probably recommended by the pardon attorney, but here we are making a judgment call. Trump’s Dec. 22, 2020, pardon of Alfred Crum was supported by former Deputy Attorney General Rod Rosenstein and the relevant U.S. attorney, which is what would happen in a normal case. The facts of the case meet the traditional Justice Department criteria: Crum is an elderly man who pleaded guilty a long time ago to a relatively trivial crime and maintained a clean record for almost 70 years. And Crum has no apparent connection to Trump. We draw a similar conclusion for similar reasons about Trump’s Dec. 23, 2020, pardon of Russell Plaisance, though we are a bit less certain here. This pardon also involved an old conviction for a relatively trivial crime, it was supported by the prosecuting attorney, and we have found no connection between Trump and Plaisance. While Plaisance’s pardon was posthumous, which is unusual for a pardon attorney recommendation, the fact that he died less than a month before the pardon was granted suggests that his case might have been already pending at the White House with a favorable recommendation. Neither of these cases had the kind of support that might suggest direct advocacy to the White House on their behalf.

That leaves 24 cases. These are the hardest cases about which to reach firm conclusions, even after a thorough examination. These pardons could in theory have been recommended by the pardon attorney because they were in the Justice Department database, and the applications were filed in time to be investigated in the ordinary course. That said, various characteristics of each clemency grant—including the close connections to Trump, the general nonsatisfaction of Justice Department criteria, and the existence of Trump’s alternate vetting process in the White House—make it unlikely that the pardon attorney would have recommended these grants, especially the ones in recent years. Scattered reporting suggests (but only suggests) that many grants of clemency among these 24—concerning, for example, Kristian Saucier, Dwight and Steven Hammond, Crystal Munoz, Tynice Hall, Judith Negron and otherswere unsupported by the pardon attorney. Relatedly, “a source close to the pardons office did not believe that the pardon attorney had given [the applications of Bernard B. Kerik and Rod R. Blagojevich] full-throated support,” which suggests that the pardon attorney might have sent their files to the White House without recommendation.

We doubt that any of these 24 cases had traditional pardon attorney support. (We are least sure in this judgment about the pardon for Rebekah Charleston.) The pardon attorney surely would not have picked out and made Trump-convenient recommendations in these cases in the ordinary course of business, so if any of these cases were accompanied by pardon attorney analysis, that almost surely happened on White House command.

We mark these cases as “probably not” in the chart, but in truth it is hard to make firm judgments here.

In sum, we believe, based on current news reporting, that seven of the 94 Trump grants came on recommendation from the pardon attorney, and that the rest either clearly did not or probably did not. (See column C in the chart.) Our other data (in columns D, E, F and G in the chart) indicate that at least 85 out of 94 Trump pardons had a personal or political connection to the president (that is, at least one hit in columns D, E, F or G). This conclusion reveals anomalies in the chart for two grants. We assess that Joseph Martin Stephens (too late) and William J. Plemons Jr. (never in database) did not receive recommendations from the pardon attorney, and yet we cannot clearly identify any connection between these recipients and Trump. Which raises the question: How did these candidates come to Trump’s attention? The answer, we suspect, lies in the fact that both grants of clemency were supported by “business associates.” If we knew more about the identity of those associates, we could draw a connection to Trump. But we do not.

(As always, please get in touch if you discover errors or have additional relevant information.)

Correction: This article originally stated that 84 out of Trump's 94 pardons had a personal or political connection to the president. The correct number is 85.

Jack Goldsmith is the Learned Hand Professor at Harvard Law School, co-founder of Lawfare, and a Non-Resident Senior Fellow at the American Enterprise Institute. Before coming to Harvard, Professor Goldsmith served as Assistant Attorney General, Office of Legal Counsel from 2003-2004, and Special Counsel to the Department of Defense from 2002-2003.
Matt Gluck is a research fellow at Lawfare. He holds a BA in government from Dartmouth College.

Subscribe to Lawfare