Trump’s pardons: the economics of clemency

Firms report fees for pursuing Trump pardons, while some recipients obtain relief from financial obligations. For victims, a presidential signature can change the prospects of recovering money. Three cases reveal how the economics of clemency work.

The filing is signed July 18, 2025. In it, J M Burkman & Associates reports $960,000 in lobbying income for its client Joseph Schwartz. The form calls for a good-faith estimate rounded to the nearest $10,000. The figure represents income reported by the firm for its work. Identifying the recipient of the money is essential to understanding it. [1]

A separate registration, effective April 22, 2025, identifies the assignment: seeking a federal pardon. It names Jack Burkman and Jacob Wohl as lobbyists. [2]

Schwartz, a former nursing-home operator, had been sentenced to three years in prison in a case involving unpaid employment taxes and a missing retirement-plan disclosure. The New Jersey U.S. attorney’s office announced the sentence on April 17, 2025. [4] On November 14, Donald Trump granted him a full and unconditional pardon for the federal case identified in the warrant. [3]

The commercial engagement and the pardon are documented. The firm’s exact role in the decision remains unknown. That leaves an economic question: what value does an applicant place on the prospect of being heard, and who bears the consequences of the eventual decision?

Why pardons are back in the news

In the Justice Department register consulted on September 24, 2026, the latest entry is a September 8 pardon for Emory Clash Jones. The register also lists 23 pardons and six commutations on September 3. The page was updated September 9. It documents those decisions without guaranteeing that no subsequent, unlisted action has occurred. [5]

On September 9, Trump said he would consider pardoning his former lawyer Michael Cohen. Reuters was reporting a possibility, not a completed grant. [6] On September 13, CBS examined paid clemency intermediaries. [7]

These developments put access to presidential decision-making back in the spotlight. The register identifies completed grants; statements and reporting shed light on how applications reach the president.

Presidential authority extends beyond the administrative process

Article II of the Constitution gives the president clemency power over offenses against the United States, except in cases of impeachment. That authority comes directly from the Constitution. [8]

The administrative route starts with an application to the Office of the Pardon Attorney. Its rules normally require a five-year wait after release from custody, or after conviction when no prison sentence was imposed, for a post-sentence pardon application. But the regulation expressly states that these rules are advisory and do not constrain presidential authority. Using another route is therefore not, by itself, evidence of illegality. [9]

The instruments also differ. A pardon expresses presidential forgiveness. A commutation reduces some or all of a sentence without changing the conviction. Neither is a declaration of innocence. [10] The official application also makes clear that a pardon does not expunge the conviction. [11]

That latitude gives alternative routes for presenting a case potential value. The criteria the president actually used still require examination in each case.

What an intermediary can sell

An economic analysis should distinguish work on an application from a claimed ability to get it considered.

The first might include assembling records, drafting the petition, presenting the applicant’s circumstances and finding supporters. The second involves claiming to know people who can bring the case to the decision-maker’s attention. This is an analytical distinction: a single invoice may not identify the price of each service, much less verify a promise of access.

An applicant may place an enormous value on a small improvement in the odds: years of freedom, a chance to resume a career and, sometimes, relief from financial obligations. A provider does not have to guarantee success to find a customer. The customer only has to believe that the assistance improves the chances sufficiently.

Waiting can strengthen that willingness to pay. The DOJ application warns that consideration can take months or years, with no guaranteed outcome. [11] For someone waiting, the prospect of receiving attention sooner may have value in its own right, even when the ultimate decision remains uncertain.

Measuring lobbying effectiveness would require comparing similar applications with and without intermediaries, accounting for their age, the offenses and the arguments submitted.

Looking only at recipients would introduce selection bias: successes would be visible while paying clients who received nothing would disappear from the account. A provider who accepts only promising cases could also accumulate successes without substantially improving their original chances.

The $960,000 disclosed in the Schwartz case therefore describes one particular engagement. Typical fees and the return on money spent remain unknown. [1]

Ozy: restitution to victims falls within the commutation

The fraud case involving media company Ozy Media moves the analysis from private fees to the financial consequences of public action. [13]

In a February 16, 2025 order, the federal court in Brooklyn set restitution owed by Carlos Watson and Ozy Media at $36,769,153.97. Their liability was joint and several: both defendants were responsible for the same obligation. These were not two independent claims to be added together. [13]

On March 28, 2025, Trump commuted Watson’s sentence. The warrant expressly removed further fines, restitution, probation and other conditions of the sentence covered by the grant. [12] A separate warrant, dated the same day, extended a commutation to Ozy Media, Inc., the corporate entity. Clemency can therefore reach a company as well as an individual. [14]

Restitution is a compensatory obligation imposed through a criminal proceeding. When clemency remits the unpaid portion, the consequence extends beyond relief for the recipient: it affects an avenue through which victims could recover money. The decision changes the financial relationship between debtor and creditors.

The legal reasoning predates this administration. In a June 19, 1995 opinion, the DOJ’s Office of Legal Counsel concluded that a full and unconditional pardon could remit criminal restitution not yet received by the victim under the statutory framework it examined. That is an executive-branch legal opinion, not a court ruling settling every possible arrangement. [15] The Ozy and Watson warrants provide more direct evidence here because they expressly address restitution. [12] [14]

The $36.77 million is the amount ordered by the court. The economic loss to victims depends on the balance outstanding at commutation, payments already received and realistic prospects for collection.

A claim against an insolvent debtor may have limited economic value while retaining significant legal value. Conversely, removing an obligation that could still be collected may deprive victims of money they would actually have received. The amount on the face of a judgment does not settle that distinction.

Ozy illustrates the financial reach of a commutation. Schwartz documents lobbying income in a separate case. No transaction connecting the two cases is documented here.

Walczak: civil tax claims can survive a pardon

Paul Walczak’s case illustrates another boundary. On April 11, 2025, the health-care business operator was sentenced to eighteen months in prison and $4,381,265.76 in restitution to the United States for tax offenses. [16] A full and unconditional pardon followed on April 23, twelve days later. [17]

Clemency followed almost immediately after sentencing. Its financial effect nevertheless depends on the kind of obligation involved.

Measuring the financial effect requires separating the criminal penalty from civil tax claims. The DOJ explains that a pardon does not end noncriminal civil proceedings, including IRS or Securities and Exchange Commission actions. Remission of a financial penalty also covers only the unpaid portion. [10]

The amount of a tax fraud, court-ordered restitution and an outstanding tax balance cannot automatically be treated as alternative versions of the same number. They may involve different legal grounds, periods and collections.

The same precision is needed when a business and its executive are involved. A warrant must be read by reference to its named beneficiary and case. Relief for an individual cannot simply be presumed to extend to every company associated with that person. The two separate Ozy warrants make the need to check the scope particularly concrete. [12] [14]

The restorative purpose of clemency

An analysis focused on money could end up treating every request for clemency as a suspicious attempt to bypass justice. That would overlook what the process is also meant to assess.

The DOJ’s Justice Manual identifies rehabilitation, post-conviction conduct, the seriousness of the offense, acceptance of responsibility and the need for relief among the relevant considerations. It also gives considerable weight to the prosecutor’s views and seeks the sentencing judge’s opinion. [18]

The strongest argument for discretion is that a person’s circumstances do not stop evolving when judgment is entered. Clemency can provide a way to consider personal change or punishment that no longer appears justified. This is a rationale for the institution, not automatic validation of an individual grant.

Equally, the possibility that ordinary procedures are slow or imperfect does not establish a justification for advantages tied to financial resources. The useful question concerns consideration: were relevant reasons examined, and did people in comparable circumstances have a comparable opportunity to be heard?

The White House defends its process. In Reuters’ investigation published June 11, 2026, it described a rigorous review involving White House counsel, the DOJ and the president. It also said that spending money on pardon lobbying was pointless. [19]

A service can be sold without delivering its promised effect. Assessing the sales pitch and the official response requires access to the file and the communications preceding the decision.

The missing records needed to assess financial influence

The investigation starts with the recipients of funds. Advisory fees, political donations, commercial payments and personal benefits to decision-makers are different transactions. Grouping them under a vague heading of “money around Trump” would strip away the character of each flow.

It then requires reconstructing the decision-making chain: who proposed the name, passed along which arguments, spoke to whom and received what response? The task is to connect financial flows to specific interventions and the reasons for a decision.

It also requires documenting failures, applications without paid representation and cases supported through other channels. Without a comparison group, even striking sequences of events cannot reliably measure the advantage money provides.

There is an institutional limit to that investigation. The official application explains that petitions, background investigations and DOJ recommendations are generally unavailable under freedom-of-information and privacy law. [11] The quality of that review is consequently difficult to assess from outside.

On the financial side, each case requires the applicable judgment, the precise clemency instrument, previous payments and surviving rights. Adding all the fines and restitution listed beside recipients would produce a nominal total. It would not automatically measure amounts relinquished, losses actually suffered by victims or a budgetary cost.

How clemency could change the calculation of risk

The economics of law suggests a further hypothesis, which should remain clearly identified as one. If some actors expected to obtain relief after conviction through an intervention available for a fee, the expected cost of offending could fall. Their calculation would include not only the risk of prosecution and conviction but also the prospect of securing a later reduction.

Measuring that effect would require observing people’s expectations and the resulting changes in behavior. Clemency warrants and lobbying disclosures describe decisions and reported services.

The opposite possibility also deserves attention: a process seen as unpredictable could prompt applicants to spend heavily without delivering any systematic advantage. The main beneficiaries of the market would then be the providers rather than the clients who believe they are improving their chances.

Intermediaries report income. Clemency can remove a financial obligation. The money’s actual influence on the decision remains to be measured. [1] [12] [14]

For victims, the question ultimately comes down to a concrete amount: the money they could still realistically have recovered before the president signed.

Related reading: Trump, the CLARITY Act and conflicts of interest and Trump’s promised $5,000 dividend and its funding.

Sources

Documents checked September 24, 2026. LDA filings are the provider’s disclosures; warrants and the court order are primary records.

1. LDA · Lobbying income disclosure. 2025-07-18. Reported income rounded to USD 10,000; no annual total inferred.
2. LDA · Schwartz engagement registration. 2025-05-06. Effective April 22; signed May 6, 2025.
3. DOJ · Joseph Schwartz pardon warrant. 2025-11-14.
4. DOJ, New Jersey · Schwartz sentencing announcement. 2025-04-17.
5. DOJ · Second-term clemency register. 2026-09-09. Register update date; latest listed grant: September 8, 2026.
6. Reuters · Statement concerning Michael Cohen. 2026-09-09. Wire report also checked at MarketScreener.
7. CBS · Reporting on clemency intermediaries. 2026-09-13.
8. National Archives · Constitution, Article II, Section 2.
9. eCFR · Administrative clemency rules.
10. DOJ · Clemency effects FAQ. 2026-09-09.
11. DOJ · Pardon application. 2024-05.
12. DOJ · Carlos Watson commutation warrant. 2025-03-28.
13. Brooklyn federal court · Ozy/Watson post-trial order. 2025-02-16. Case 23-CR-00082, document 388, order pp. 45–46 and 52–53 (PDF pp. 47–48 and 54–55). Court record published by GovInfo.
14. DOJ · Ozy Media corporate commutation warrant. 2025-03-28.
15. DOJ, Office of Legal Counsel · Legal effects of a pardon. 1995-06-19. Executive-branch opinion, section III; scope limited to the legal framework examined.
16. DOJ · Paul Walczak sentencing. 2025-04-11.
17. DOJ · Paul Walczak pardon warrant. 2025-04-23.
18. DOJ, Justice Manual · Pardon review criteria.
19. Reuters · June investigation and White House response. 2026-06-11. Reuters investigation also checked at BreakingNews.ie.

Limitations

This document-based analysis includes no original interviews or access to bank accounts or confidential case files. Lobbying income figures are reported estimates. The records do not support calculating a price for a presidential decision, a market size or an intermediary’s average effectiveness. Amounts are nominal U.S. dollars, without currency conversion or discounting. Court-ordered restitution is historical: outstanding balances, previous payments and recoverable value at clemency remain unknown. The incentive scenarios are economic hypotheses.