Consider the plaintiff who knew, to a moral certainty, which human being would read the first page of the complaint.<\/p>\n
In late 2022 a group of physicians and medical associations organized themselves into an entity called the Alliance for Hippocratic Medicine. They incorporated in Amarillo, Texas. Then they filed suit against the Food and Drug Administration, asking a federal court to undo the agency’s year 2000 approval of mifepristone, a drug used in more than half of the abortions in the country. The suit landed, as suits filed in the Amarillo division of the Northern District of Texas almost invariably did, on the desk of Judge Matthew Kacsmaryk. There was one active judge sitting in that division. The odds of drawing him were not long. They were roughly one hundred percent.<\/p>\n
Judge Kacsmaryk ruled in April 2023. He suspended the FDA’s decades-old approval, a remedy of astonishing breadth from a single trial judge in the Texas panhandle. The Fifth Circuit narrowed the order. The Supreme Court stayed it while the appeal ran its course. And then, in FDA v. Alliance for Hippocratic Medicine<\/em>, 602 U.S. 367 (2024), a unanimous Court held that the plaintiffs never had standing to bring the case at all. Justice Kavanaugh, writing for everyone, explained that these doctors did not prescribe mifepristone, were not injured by other people’s prescriptions, and could not manufacture an Article III controversy out of ideological objection. The whole edifice came down for want of a plaintiff with a stake.<\/p>\n I teach professional responsibility. I have taught it long enough to be suspicious of my own outrage reflex. So let me say at the outset what I think the mifepristone episode is really about, because it is not mainly about abortion, and it is not mainly about the FDA. It is about a structural feature of the federal courts that lets a sophisticated litigant, with a little planning, hand-pick the individual who will decide the most consequential questions in American law. The polite name for the practice is forum shopping. The blunter name, the one that entered the newspapers in 2023 and 2024, is judge shopping.<\/p>\n Start with a statute almost nobody outside the judiciary has read. Under 28 U.S.C. 137, each district court decides for itself how to divide its business among its judges. Congress left the housekeeping to the houses. Most large districts assign civil cases by a district-wide random draw, a genuine lottery, so that a lawyer filing in Manhattan cannot know whether she has drawn a Reagan appointee or an Obama appointee until the clerk tells her.<\/p>\n But many districts are carved into divisions, and some of those divisions are staffed by exactly one active district judge. Amarillo. Wichita Falls. Galveston, for stretches. Victoria. When you file in a single-judge division, you have not entered a lottery. You have made a reservation.<\/p>\n None of this is new, and that is the part the outrage cycle tends to forget. Divisional structure is a relic of the era before automobiles and interstate highways, when a farmer bringing a federal claim should not have to travel three hundred miles to a courthouse. The single-judge division was a convenience of geography. What changed is not the map. What changed is the sophistication of the litigants who learned to read it.<\/p>\n The history here runs in both partisan directions, and any honest account has to say so. Judge Reed O’Connor, who sits in the Fort Worth and Wichita Falls divisions, became the destination of choice for red-state attorneys general challenging federal policy in the 2010s. He struck down the entire Affordable Care Act in December 2018 in Texas v. United States<\/em>, a ruling the Supreme Court later dispatched on standing grounds in California v. Texas<\/em>, 593 U.S. 659 (2021). The pattern was unmistakable. Coalitions of state officials filed suit after suit in venues where the assignment was, for practical purposes, predetermined.<\/p>\n And the mirror image existed too. During the first Trump administration, litigants challenging immigration and environmental rollbacks gravitated toward the Northern District of California and the Ninth Circuit, where the reception was warmer and the nationwide injunctions came readily. Anyone who tells you judge shopping is a one-party sin has not been paying attention. The tool is neutral. Only the users have a politics.<\/p>\n What made the Texas single-judge divisions distinctly powerful was the combination of near-certain assignment and the remedy on offer. A predictable judge is one thing. A predictable judge who will issue a nationwide injunction, binding the executive branch as to every person in the United States, is a different order of leverage. Pardon me, of advantage. The two ingredients together are what turned a quiet quirk of divisional structure into a national argument.<\/p>\n On March 12, 2024, the Judicial Conference of the United States, the policymaking body of the federal courts, announced new guidance. The Administrative Office, then led by Judge Robert Conrad Jr., described a policy urging district courts to assign civil cases through district-wide random selection whenever the suit seeks to bar or compel state or federal action, whether by injunction or declaratory judgment, and whether the relief sought reaches beyond the parties.<\/p>\n The theory was tidy. If you cannot predict your judge, you cannot select your judge, and the incentive to incorporate a plaintiff in Amarillo evaporates. Senator Charles Schumer praised it. He had been complaining about single-judge divisions for a year. The reform had the clean logic of a rule written to solve exactly one problem.<\/p>\n Then came the reading of the fine print, and the fine print mattered.<\/p>\n The Judicial Conference does not command the district courts. It advises them. Section 137 still puts case-assignment authority with each district, and nothing in the March 2024 announcement changed the statute. Within days, the policy revealed itself as a recommendation dressed in the syntax of a directive. Judge Jeffrey Sutton, in his capacity discussing the guidance, and other officials acknowledged that districts retained their discretion.<\/p>\n The Northern District of Texas noticed. Chief Judge David Godbey indicated that his court would not overhaul its assignment practices to conform. The Southern District of Texas signaled similar reluctance. Senator Mitch McConnell, from the other side of the aisle, wrote in defense of the districts’ independence and against what he framed as a partisan effort to close a venue that had proven inconvenient to the sitting administration. The Judicial Conference, cornered, clarified that the guidance was indeed guidance. Districts could adopt it. Districts could decline.<\/p>\n So the great reform of March 2024 arrived as a suggestion, and several of the districts most responsible for the underlying phenomenon suggested, in reply, that they would pass. I do not say this to sneer at the Conference. I say it because the episode teaches something about the limits of soft law inside a judiciary jealous of its own decentralization. The federal bench is not a hierarchy in the corporate sense. It is a confederation of life-tenured officials who guard their prerogatives, and a policy that depends on voluntary compliance from the least willing members is a policy that has already conceded its weakest point.<\/p>\n There is a deeper irony worth naming. The people best positioned to fix single-judge assignment were the very chief judges least inclined to, because their districts benefited, in prestige and in docket, from being the country’s chosen forum. Reform by request runs aground on that reef. You are asking the house that wins the auction to stop holding the auction.<\/p>\n If you want to see what concentrated case-assignment looks like when the stakes are commercial rather than constitutional, drive south from Fort Worth to Waco.<\/p>\n Judge Alan Albright took the bench in the Waco division of the Western District of Texas in 2018. He had practiced patent litigation. He understood the field, he made no secret of wanting the cases, and he ran his docket in a way that patent plaintiffs found congenial: brisk schedules, a reluctance to stay proceedings, an unwillingness to send cases elsewhere. Word travels fast among the patent bar. Filings followed.<\/p>\n By 2021 the numbers had become genuinely strange. A single judge in a mid-size Texas city was drawing something on the order of a quarter of all patent cases filed in the entire United States. Roughly a fifth of the world’s patent litigation, on some counts, flowed through one courtroom. No serious person believed that Waco had become the natural center of the American innovation economy. The plaintiffs went there because filing there meant drawing Judge Albright, and drawing Judge Albright meant a favorable path to a jury.<\/p>\n This happened, I should stress, after the Supreme Court had supposedly curtailed patent forum shopping. In TC Heartland LLC v. Kraft Foods Group Brands LLC<\/em>, 581 U.S. 258 (2017), the Court read the patent venue statute narrowly, holding that a domestic corporation resides only in its state of incorporation for venue purposes. The decision was meant to break the grip of the Eastern District of Texas, where Judge Rodney Gilstrap in Marshall had long presided over a plaintiff-friendly patent docket built on the older, permissive venue reading. And TC Heartland<\/em> did redistribute cases. It just redistributed a great many of them a few hours west, to Waco, because large technology companies have offices in Austin and the Western District, which meant venue there was proper.<\/p>\n The appellate courts pushed back through the mandamus power, the same tool they had used a decade earlier when the Eastern District’s dominance drew scrutiny in cases like In re Volkswagen of America, Inc.<\/em>, 545 F.3d 304 (5th Cir. 2008), and In re TS Tech USA Corp.<\/em>, 551 F.3d 1315 (Fed. Cir. 2008). The Federal Circuit issued a run of mandamus decisions ordering Judge Albright to transfer cases to more convenient districts, faulting his handling of the venue-transfer analysis in matters involving Apple, Google, and others. The message from the reviewing court was audible. Stop keeping cases that belong elsewhere.<\/p>\n Then the district itself acted, and it acted where the Judicial Conference two years later would only advise. On July 25, 2022, Chief Judge Orlando Garcia of the Western District of Texas issued an order requiring that new patent cases filed in the Waco division be assigned randomly among a dozen judges across the district. Not one judge. Twelve. The single-judge magnet was, by administrative fiat, degaussed.<\/p>\n The effect was immediate and instructive. New patent filings in Waco fell off. Plaintiffs, denied the certainty of Judge Albright, adjusted. Some filings migrated back to the Eastern District of Texas and Judge Gilstrap. Some went to the District of Delaware, the other great patent forum, where many technology companies are incorporated. The litigation did not vanish. It relocated, again, in search of a favorable draw.<\/p>\n That last point is the one I ask my students to sit with. The 2022 Waco order proves that a district can fix its own assignment problem overnight, with a stroke of the chief judge’s pen, no act of Congress required. And it proves, in the same breath, that fixing one forum does not cure the underlying appetite. The water finds the low ground. You can dam Waco, and the current runs to Marshall or to Wilmington. Assignment reform inside a single district treats a local symptom of a national condition.<\/p>\n Return to the mifepristone case and read its ending as a professor of remedies might.<\/p>\n The Supreme Court did not scold the Amarillo plaintiffs for shopping. It did not need to. It held that they had no injury, and an entity with no injury cannot invoke the judicial power, whatever forum it selects. FDA v. Alliance for Hippocratic Medicine<\/em> is a standing decision, and standing is the doctrine that polices exactly this kind of manufactured controversy. The plaintiffs assembled themselves, chose their courthouse, secured their remedy at the trial level, and then discovered at the end of the road that the front door had been locked all along.<\/p>\n There is a lesson in the sequence that cuts against easy reform enthusiasm. The system, in the mifepristone case, corrected itself. It corrected itself slowly, at enormous cost, after a nationwide disruption of a drug approval that had stood for more than twenty years. But it corrected itself through ordinary doctrine, applied by nine justices who did not agree about much else. If your worry is that judge shopping produces lawless outcomes, the counterexample is the very case that made judge shopping famous.<\/p>\n And the remedy problem got its own answer, later and larger. In Trump v. CASA, Inc.<\/em>, decided June 27, 2025, the Supreme Court held that federal courts likely lack the equitable authority to issue universal injunctions, the sweeping orders that bind the government as to everyone, party and non-party alike. Justice Gorsuch had been building toward this for years. Read his concurrence in Department of Homeland Security v. New York<\/em>, 589 U.S. 1112 (2020), where he warned that universal injunctions had escaped their historical banks. Read Justice Thomas in Trump v. Hawaii<\/em>, 585 U.S. 667 (2018), tracing the pedigree and finding it thin. CASA<\/em> turned that scholarship into a holding.<\/p>\n Here is why CASA<\/em> belongs in any discussion of judge shopping. The whole reason to reserve a table in Amarillo, or to camp in Waco, is the size of the prize. A trial judge who can only bind the parties before him is a modest figure. A trial judge who can, by one order, freeze federal policy for three hundred and thirty million people is a national power center, and litigants will spend real ingenuity to choose which national power center hears their case. Shrink the remedy and you shrink the incentive. If a single district judge can no longer hand you the country, the appeal of picking that district judge cools considerably.<\/p>\n I find this more persuasive than assignment reform, and I say so as a gentle contrarian to the Schumer position. The March 2024 policy attacked the second half of the equation, the selection, while leaving the first half, the prize, untouched. CASA<\/em> went at the prize. Which is to say the Supreme Court, in the ordinary course of deciding a remedies question, did more to deflate judge shopping than the Judicial Conference’s guidance ever could, because it changed what the shopping was for.<\/p>\n Now to the part of this I actually teach, and the part that gets the least careful attention in the press coverage.<\/p>\n Is judge shopping a violation of a lawyer’s professional duties? The reflexive answer is yes, and the reflexive answer is wrong, or at least far too quick. A lawyer owes her client competent and diligent representation. Model Rule 1.1. Model Rule 1.3. A lawyer who knows that filing in one proper venue will draw a more favorable judge, and who files elsewhere out of some abstract commitment to randomness, has arguably shortchanged the very client she is bound to serve. Zealous advocacy, within the bounds of the law, includes choosing the most advantageous forum the rules allow. We do not discipline lawyers for winning the coin toss. We do not discipline them for declining to flip it when the rules give them the choice.<\/p>\n The ethical line falls not at forum selection but at candor and manufacture. Rule 3.3 governs candor to the tribunal. Rule 8.4 forbids conduct involving dishonesty, fraud, deceit, and misrepresentation, and separately forbids conduct prejudicial to the administration of justice. Incorporating a shell organization in a single-judge division, weeks before suit, to conjure venue where none organically exists, moves toward that second category. Not because the lawyer chose a favorable forum, but because the choice depended on an artifice built for the occasion.<\/p>\n That is a narrow charge, and I want to be honest that it is hard to prove and rarely brought. Bar discipline for forum selection is close to nonexistent, and it should be uncommon, because the alternative is a regime in which every venue decision invites a second-guessing inquiry into the lawyer’s motives. We do not want that. Motive-hunting inside the ordinary work of advocacy is corrosive in its own right.<\/p>\n So if the ethics rules do not reach most of this conduct, why does it trouble me? Because professional responsibility is broader than the disciplinary code. It always has been. The organized bar has long understood that lawyers hold a public trust, that the legitimacy of adjudication depends on the appearance as well as the fact of impartiality, and that a practice which is individually rational can be collectively destructive.<\/p>\n Here is the destruction, stated plainly. When the public learns that a case was filed in Amarillo because the plaintiff wanted Judge Kacsmaryk, or in Waco because the plaintiff wanted Judge Albright, the public draws an inference about the judge and about the outcome. The inference may be unfair to the judge, who took an oath and may well have ruled the same way anywhere. It does not matter. The perception forms. And the perception is that the outcome was baked in before the first brief was filed, that the assignment was the case, that a federal judge is less an impartial arbiter than a known quantity to be acquired like any other litigation asset.<\/p>\n The federal judiciary has no army and no purse. Alexander Hamilton said as much in Federalist No. 78, and the point has aged into cliche precisely because it keeps being true. The courts hold their authority on credit, and the credit is public confidence that judges decide cases rather than that litigants decide judges. Every well-publicized instance of successful judge shopping draws down that account. The single-judge division does not corrupt the judge. It corrupts the story the public tells about the judge, and in an institution that runs on belief, the story is not a trivial thing.<\/p>\n Which brings me to a position I hold with some conviction and offer without apology. The March 2024 policy was the right instinct aimed at the wrong lever, and pitched at the wrong altitude. Assignment is a matter for Congress, not for a non-binding conference resolution that the resisting districts could and did shrug off. Section 137 is a congressional statute. Congress can amend it. Congress can require district-wide random assignment for cases seeking to enjoin state or federal law, and Congress can make the requirement mandatory rather than hortatory. That the reform came instead as guidance, immediately softened into a suggestion, tells you the reformers preferred a gesture they controlled to a statute they could not guarantee.<\/p>\n The Western District of Texas showed the better path in July 2022. It did not wait for permission. It fixed its own house by order of its own chief judge, and the Waco patent flood receded within months. Any district with a single-judge division could do the same tomorrow. Most have not. Their reluctance is the actual scandal here, not the ingenuity of the lawyers who exploit what the districts decline to close.<\/p>\n I will end where the doctrine did, because the doctrine got closer to the truth than the policy did. The way you kill judge shopping is to make the judge worth less. Not by insulting the office. By confining the remedy. A trial court that binds the parties before it, and no further, is a court no rational litigant will cross an ocean of incorporation paperwork to select. Trump v. CASA<\/em> understood this in 2025, and the understanding is older than the case: the equitable power was never meant to let one judge govern the nation. Restore the modest injunction and the extravagant forum shopping loses its object. Leave the extravagant remedy in place and no amount of random assignment will stop clever people from finding the low ground, because the water always does.<\/p>\n","protected":false},"excerpt":{"rendered":" A professional responsibility teacher on single-judge divisions, the March 2024 Judicial Conference policy, Waco’s patent docket, and whether choosing your judge corrodes the courts.<\/p>\n","protected":false},"author":1,"featured_media":57,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[15],"tags":[],"class_list":["post-37","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-courts-cases"],"yoast_head":"\nthe mechanics that make it possible<\/h2>\n
what the Judicial Conference actually said in March 2024<\/h2>\n
Waco and the patent gold rush<\/h2>\n
the standing coda and what it teaches<\/h2>\n
the professional responsibility question<\/h2>\n