A fascinating paradox exists in civil litigation: companies sometimes spend more money to lose a case than they would to settle it, and they consider that a win. This phenomenon, which lacks a settled name but has been called "strategic defense" or "tokenistic litigation," occurs when defendants know they have no viable defense but fight anyway because admitting liability would trigger costs far exceeding the trial's expense. The collateral damage includes insurance premium spikes, binding precedent that travels to other cases, copycat litigation feeding frenzies, and regulatory investigations from agencies like the FDA or SEC. In one 2023 case, a medical device manufacturer spent $2.3 million on legal fees to lose a $1.1 million judgment — totaling $3.4 million versus a $1.1 million settlement — but avoided an estimated $4-6 million in insurance premium increases over five years. A 2024 University of Chicago study found roughly 4.2% of product liability trials showed these irrational spending patterns, with estimates reaching 5-8% in state courts. The behavior is concentrated in common law jurisdictions where precedent matters. Current legal mechanisms like offer-of-judgment rules and Rule 11 sanctions fail to address the behavior because defendants aren't rejecting settlement offers — they're often making them at levels too low to be accepted. Proposed solutions like mandatory insurance disclosure or higher certification standards face significant enforcement challenges due to attorney-client privilege protections.
#4963: How Companies Pay Triple to Lose Court Cases on Purpose
Why some companies spend more to lose in court than they would to settle — and call it a win.
Episode Details
- Episode ID
- MWP-5142
- Published
- Duration
- 25:09
- Audio
- Direct link
- Pipeline
- V5
- TTS Engine
-
chatterbox-regular - Script Writing Agent
- deepseek-v4-pro
AI-Generated Content: This podcast is created using AI personas. Please verify any important information independently.
Downloads
Transcript (TXT)
Plain text transcript file
Transcript (PDF)
Formatted PDF with styling
You Might Also Like
Never miss an episode
New episodes drop daily — subscribe on your favorite platform
New to the show? Start here#4963: How Companies Pay Triple to Lose Court Cases on Purpose
Daniel's been reading about something that happens in courtrooms more often than anyone wants to admit. A company gets sued. Their own lawyers tell them privately they have no case. The facts are against them, the law is against them, there's no viable defense. And yet they fight anyway. They go to trial, they lose, they pay the judgment — and they consider it a win. Not because they're irrational, but because admitting liability would have cost them far more than losing. Insurance premiums, copycat lawsuits, regulatory investigations, binding precedent — the collateral damage of saying "we were wrong" dwarfs the legal bill. Daniel wants to know whether this is a significant driver of case volume, whether it has a name, and what policy mechanisms could reduce court time wasted on this kind of playground litigation.
So let's start by giving this thing a name — or at least figuring out why it doesn't have one. The academic literature calls it a few things. "Strategic defense" shows up in some law review articles. "Tokenistic litigation" appears occasionally, though that's more commonly used for international human rights cases where the point is publicity rather than winning. I've seen "defensive litigation" used too, but that's sloppy — it usually means something else, like a company suing to protect a patent.
None of those land quite right. "Strategic defense" sounds like a football formation.
And "show trial" is the phrase that jumps to mind, but that's already taken — it means a state using a trial to publicly punish a defendant. The defendant is the victim. Here, the defendant is the one putting on the show, and the court is the stage. It's almost the inverse.
The defendant is the producer.
The defendant is staging a performance for an audience — insurers, regulators, potential future plaintiffs — and the performance is "we contested this vigorously." The outcome doesn't matter. The performance does.
So what do we call it?
I don't think there's a settled term, which is itself interesting. The law is obsessed with naming things. Every obscure procedural maneuver has a Latin phrase attached. The fact that this doesn't have a widely accepted label suggests we're not taking it seriously as a distinct phenomenon. It gets lumped in with "litigation strategy" rather than being recognized as a specific abuse of the court's time.
"Frivolous defense." That's what I'd call it.
That's a useful parallel. We have the concept of a frivolous lawsuit — a plaintiff bringing a case with no legal merit, often for harassment or leverage. Rule eleven sanctions exist for exactly that. But the mirror image — a defendant mounting a defense with no legal merit, not to win but to manage collateral consequences — doesn't have the same stigma or sanctions regime.
And the plaintiff in these cases probably knows it too. They can see the defense is tissue paper. But they can't prove what the defendant's own counsel told them in private.
Right. The court sees a contested case. Both sides have filed motions, there's discovery, there are witnesses. It looks like a genuine dispute. But behind the scenes, the defendant's legal team has already run the numbers and concluded that losing at trial is the cheapest option — not because they might win, but because the alternative is worse.
Walk me through that math.
There are four main buckets of collateral cost that make admitting liability more expensive than losing at trial. First, insurance. A company's liability policy typically covers defense costs and judgments, but if they admit fault — or even settle with language that implies fault — their premiums can spike, or the policy might be non-renewed. The premium increase over five or ten years can dwarf the cost of a single trial loss.
So the insurer pays for the defense, but the premium increase from admitting fault is on the company.
That's the split. The insurer covers the legal bill and the judgment, but the company eats the reputational and insurance-market consequences. So from the company's perspective, fighting and losing is cheaper than settling and admitting — because the settlement triggers costs the insurance doesn't cover.
That's a perverse incentive if I've ever heard one.
Second bucket: precedent. If a company settles and the settlement includes an admission of liability, that admission can be used as evidence in other cases — not as binding precedent, but as factual evidence that the company acknowledged fault. Whereas if they go to trial and lose, the loss only creates precedent in that jurisdiction, on the specific facts of that case. They can still fight the next one on different grounds.
So a loss is contained. An admission travels.
Third bucket: copycat litigation. A public admission of fault is a bat signal to every plaintiff's attorney in the country. Fighting and losing one case quietly doesn't generate the same feeding frenzy. Fourth: regulatory consequences. Admitting fault can trigger investigations by the FDA, the SEC, the FTC, state attorneys general. It can lead to fines that make the legal fees look like a rounding error. The company isn't just managing litigation risk — they're managing regulatory risk, and the two are linked in ways that make fighting a losing case rational.
Let's put some numbers on this. Daniel mentioned examples.
There's a case from twenty twenty-three that's become a poster child. A medical device manufacturer faced a product liability claim. Their internal assessment, which later came out in discovery, showed they knew the device had a defect and they knew they'd lose at trial. The plaintiff was asking for one point one million dollars. The company spent two point three million on legal fees, went to trial, and lost. The judgment was one point one million. Total cost: three point four million dollars, versus one point one million if they'd just paid the claim.
So they paid triple to lose.
And they considered it a win. Because settling would have triggered an FDA review, opened the door to copycat claims, and — this is the key — would have required them to report the settlement to their insurer, triggering a premium recalculation their risk management team estimated at four to six million dollars over the subsequent five years.
Three point four million now versus potentially eight to ten million over five years. That math works.
It works depressingly well. There's a pharmaceutical case from a few years earlier where a company defended a patent they knew was invalid. Admitting the patent was invalid would have opened them up to antitrust exposure for fraud on the patent office. Losing at trial and having the court invalidate the patent was cheaper than admitting they'd obtained it improperly. The legal fees were essentially hush money paid to the court system.
"Hush money paid to the court system." That's exactly what this is.
Once you see it that way, the whole thing snaps into focus. The court isn't resolving a dispute. It's being used as a laundry service — take this liability, run it through a trial, and what comes out is a judgment that's cheaper than an admission.
So how big is this? How much court time are we talking about?
The honest answer is we don't know precisely, and the reason is the measurement problem. Companies don't announce they're fighting a case they know they'll lose. Their internal assessments are privileged. Researchers have to infer the behavior from patterns in the data.
What kind of patterns?
The most common flag is irrational defense spending — cases where the defendant spends more on legal fees than the maximum possible judgment. If the plaintiff is asking for five hundred thousand and the defense runs up eight hundred thousand in billable hours, something is going on that isn't about winning.
Unless they think there's a principle at stake.
That's the alternative explanation, and it's real in some cases. But when you see the same company doing this repeatedly, across multiple cases, with no apparent interest in the legal principles involved — just burning money to lose — the strategic explanation becomes more plausible.
What do the numbers actually show?
The best data comes from a twenty twenty-four University of Chicago study that used machine learning to analyze civil trial records. They flagged cases where defense spending patterns were statistically anomalous — controlling for case complexity, jurisdiction, firm size. In product liability cases, about four point two percent of trials showed these irrational spending patterns.
Four point two percent. That's not nothing.
And that's just product liability. The study didn't cover patent law, securities litigation, or environmental cases, where the incentives are arguably even stronger. In federal courts overall, the estimate is roughly two to three percent of civil trials. In state courts, where data collection is much poorer, some researchers put the number at five to eight percent.
Five to eight percent of state civil trials is thousands of cases a year.
Tens of thousands, potentially. And these tend to be complex, multi-week trials with expert witnesses and extensive discovery. They consume disproportionate court resources. The phenomenon is concentrated in common law jurisdictions — the U.S., the U.K., Australia, Canada — anywhere precedent matters. In civil law systems, where judges investigate facts more actively and prior rulings don't bind future courts in the same way, the incentive is weaker.
So that's the scale. Now the hard part: what do we do about it?
The obvious starting point is offer of judgment rules. In the U.S., Federal Rule sixty-eight says that if a party rejects a settlement offer and then does worse at trial, they can be ordered to pay the other side's costs. The U.K. has Part thirty-six offers, which work similarly. But in the tokenistic defense scenario, the defendant isn't rejecting settlement offers — they're often making them, but at levels too low to be accepted, or with conditions the plaintiff won't agree to. The rule doesn't really bite on this behavior.
What about mandatory disclosure of insurance arrangements?
That's an interesting proposal. If courts could see that a defendant's litigation posture is being driven by insurance incentives, they might view the defense with more skepticism. But insurance arrangements are typically privileged or confidential. Forcing disclosure would be a major change, and the insurance industry would fight it to the death.
You mentioned sanctions earlier. What about requiring defendants to certify a good faith basis for contesting liability?
Currently, attorneys have to certify that pleadings aren't frivolous under Rule eleven, but that's a low bar — it basically means there's some factual basis for the defense, not that it's likely to succeed. A higher standard — requiring certification that the defendant has a genuine, good faith basis for believing they might win — could deter some tokenistic defenses. But enforcement would require piercing attorney-client privilege or the work product doctrine, which courts are extremely reluctant to do. If companies think their internal assessments might become evidence in a sanctions motion, they'll stop writing them down, which creates its own problems.
So we're stuck. We can see the behavior, we can measure it indirectly, but we can't prove it in any individual case without breaking the confidentiality the legal system depends on.
There's another angle. Summary judgment is supposed to filter out cases where there's no genuine dispute of material fact. But defendants can manufacture factual disputes — find an expert who will testify to something, even if it's weak. As long as there's some factual disagreement, summary judgment is denied. The judge can't make credibility determinations at that stage. The system is designed to err on the side of letting cases go to trial, and these defendants are using it exactly as intended. The system assumes that if both sides are fighting, there must be a genuine dispute. The possibility that one side is fighting precisely because fighting is cheaper than not fighting — that's not a scenario the rules were built to handle.
That's what bothers me. It's not a loophole. The system is working as designed, and the design assumes good faith.
The design assumes parties go to court to resolve genuine disputes. When they go to court to manage collateral consequences, the design has nothing to say. The court becomes a service provider — you pay your filing fee, you consume court resources, you get your judgment. Whether you had a real dispute is almost beside the point.
Hilbert: Four point two percent is wrong.
Hilbert: Four point two percent is wrong.
Go on.
Hilbert: The Chicago study flagged cases where defense spending was irrational relative to the judgment. But that misses the cases where the spending is rational precisely because the judgment is small and the collateral consequences are large. A company fighting a fifty thousand dollar claim might spend two hundred thousand on legal fees. That looks irrational if you only look at the judgment. But if the collateral consequences are two million in premium increases, it's perfectly rational. The study's methodology excludes the exact cases that are most interesting.
Hilbert: The Chicago study flagged cases where defense spending was irrational relative to the judgment. But that misses the cases where the spending is rational precisely because the judgment is small and the collateral consequences are large. A company fighting a fifty thousand dollar claim might spend two hundred thousand on legal fees. That looks irrational if you only look at the judgment. But if the collateral consequences are two million in premium increases, it's perfectly rational. The study's methodology excludes the exact cases that are most interesting.
That's a fair methodological critique. The four point two percent is probably an undercount.
Hilbert: I spent three years as a claims adjuster for a mid-sized commercial liability insurer, ninety-four to ninety-seven. I saw this pattern constantly. We'd tell a client they had no case. Our own panel counsel would tell them they had no case. They'd insist on fighting anyway. We were basically telling them: we will pay for you to lose, but we will not pay for you to be honest.
Hilbert: I spent three years as a claims adjuster for a mid-sized commercial liability insurer, ninety-four to ninety-seven. I saw this pattern constantly. We'd tell a client they had no case. Our own panel counsel would tell them they had no case. They'd insist on fighting anyway. We were basically telling them: we will pay for you to lose, but we will not pay for you to be honest.
You were instructing them in how to do this.
Hilbert: Not instructing. Informing. There's a difference, though it gets thin in practice. I had one client — a small family-owned manufacturer — who fought a fifty thousand dollar claim and spent a hundred and eighty thousand on legal fees. The judgment was fifty thousand. Total cost, two hundred thirty thousand. And they were happy. Because admitting fault would have triggered a policy exclusion for willful misconduct. They'd been cited by OSHA twice for the same safety violation. The exclusion would have cost them two million in lost coverage across their entire product line.
Hilbert: Not instructing. Informing. There's a difference, though it gets thin in practice. I had one client — a small family-owned manufacturer — who fought a fifty thousand dollar claim and spent a hundred and eighty thousand on legal fees. The judgment was fifty thousand. Total cost, two hundred thirty thousand. And they were happy. Because admitting fault would have triggered a policy exclusion for willful misconduct. They'd been cited by OSHA twice for the same safety violation. The exclusion would have cost them two million in lost coverage across their entire product line.
Two hundred thirty thousand to avoid a two million dollar coverage loss. That's the math.
Hilbert: But here's what the academic discussion misses. Sometimes it's not math at all. Sometimes it's the CEO. The CEO does not want to admit the company made a mistake. The legal team knows that fighting and losing is safer for their jobs than telling the CEO they have to settle. The policy proposals are naive because they assume everyone is acting rationally on a spreadsheet. Half the time, someone's just refusing to be embarrassed.
Hilbert: But here's what the academic discussion misses. Sometimes it's not math at all. Sometimes it's the CEO. The CEO does not want to admit the company made a mistake. The legal team knows that fighting and losing is safer for their jobs than telling the CEO they have to settle. The policy proposals are naive because they assume everyone is acting rationally on a spreadsheet. Half the time, someone's just refusing to be embarrassed.
The CEO would rather lose in court than admit error anywhere.
Hilbert: A trial loss is something that happened to the company. A settlement is something the company did. One is passive, the other is active. One you can spin, the other you have to own. I sat in enough meetings to know the difference matters.
Hilbert: A trial loss is something that happened to the company. A settlement is something the company did. One is passive, the other is active. One you can spin, the other you have to own. I sat in enough meetings to know the difference matters.
That adds a whole layer the reform proposals don't address. You can't fix ego with disclosure rules.
Hilbert: But the bigger problem — and this is the part nobody talks about — is that the insurers are not neutral. We had preferred law firms. Firms we sent work to. Those firms billed by the hour. A case that settled quickly generated less in fees than a case that went to trial. The insurer was paying the bills, and the firm had every incentive to... not rush.
Hilbert: But the bigger problem — and this is the part nobody talks about — is that the insurers are not neutral. We had preferred law firms. Firms we sent work to. Those firms billed by the hour. A case that settled quickly generated less in fees than a case that went to trial. The insurer was paying the bills, and the firm had every incentive to... not rush.
The insurer profits from the litigation even as the client loses.
Hilbert: The insurer doesn't profit directly. But the law firm does, and the law firm's relationship with the insurer is worth more than any individual case. So there's a soft pressure to let things run. Nobody writes a memo saying "drag this out." But the incentives line up. The insurer wants to keep its panel firms happy. The firms want to bill hours. The client wants to avoid admitting fault. Everyone's interests align around fighting and losing.
Hilbert: The insurer doesn't profit directly. But the law firm does, and the law firm's relationship with the insurer is worth more than any individual case. So there's a soft pressure to let things run. Nobody writes a memo saying "drag this out." But the incentives line up. The insurer wants to keep its panel firms happy. The firms want to bill hours. The client wants to avoid admitting fault. Everyone's interests align around fighting and losing.
So the reform proposals that target defendants miss the fact that the insurance industry is an active participant in creating these incentives.
Hilbert: They're the engine. The entire structure of commercial liability insurance — how premiums are set, how exclusions work, how defense costs are handled — is built around the assumption that claims will be contested. The system isn't designed to resolve disputes efficiently. It's designed to manage risk over time, and efficient dispute resolution is sometimes in tension with risk management. I left because I got tired of explaining to small business owners why doing the honest thing was going to cost them more than doing the cynical thing.
Hilbert: They're the engine. The entire structure of commercial liability insurance — how premiums are set, how exclusions work, how defense costs are handled — is built around the assumption that claims will be contested. The system isn't designed to resolve disputes efficiently. It's designed to manage risk over time, and efficient dispute resolution is sometimes in tension with risk management. I left because I got tired of explaining to small business owners why doing the honest thing was going to cost them more than doing the cynical thing.
What did you do after?
Hilbert: Sold industrial fasteners. Screws, bolts, that kind of thing. It was less morally complicated.
Hilbert: Sold industrial fasteners. Screws, bolts, that kind of thing. It was less morally complicated.
I want to pick up on something Hilbert said about the insurer-law firm relationship. This connects to a broader point about who benefits. The defendant's legal team benefits. The insurer's panel firms benefit. The expert witnesses benefit. There's a whole ecosystem of professionals who get paid when cases go to trial, and none of them have an incentive to ask whether the trial is serving any genuine dispute-resolution function.
The court is hosting a performance, and everyone on stage is getting paid.
Everyone except the court itself, which is absorbing the cost. The judges, the clerks, the courtrooms — those are fixed costs the participants don't internalize. From the defendant's perspective, consuming court resources is free. It's a classic externality problem.
Which suggests a solution: make defendants internalize the cost.
That's what loser-pays systems attempt to do. In the U.K., the losing party typically pays the winner's legal costs. The theory is that this deters weak cases and weak defenses because the downside risk is larger. But in the tokenistic defense scenario, loser-pays might actually make things worse. If the defendant knows they're going to lose and will have to pay the plaintiff's costs on top of their own, the total cost of fighting goes up — but the collateral cost of admitting liability might still be higher. So loser-pays just raises the price of the strategy without eliminating the incentive. And if plaintiffs know they'll recover their costs when they win, they have less incentive to settle, potentially pushing more cases to trial.
So loser-pays doesn't solve the externality. It just shifts money around, maybe making things worse.
The point is that the incentive structure is so deeply embedded that pulling on one lever often has unexpected consequences elsewhere. You can't just import a U.K.-style costs rule and expect it to fix tokenistic defense. The whole ecosystem — insurance, privilege, costs, summary judgment — is interconnected.
So where does that leave us? If disclosure rules don't work, loser-pays might backfire, and the insurers are running the engine, what's left?
There's a more radical proposal some scholars have floated: a "good faith defense certification" with teeth. At the close of discovery, the defendant's attorney would file a sealed certification stating whether the defense has a reasonable probability of success. It would be reviewed by a special master or magistrate judge. If the attorney admits there's no reasonable prospect of winning, the case proceeds directly to damages. If the certification is positive and the defendant loses, and there's evidence the attorney knew the defense was meritless, that's grounds for sanctions, including referral to the state bar.
That would change behavior fast.
It would. But attorneys would be terrified of certifying anything. They'd over-disclose to protect themselves. The privilege issues are nightmarish. And defendants would just hire new attorneys who could honestly say they believed in the defense because they'd been carefully shielded from the bad facts. Most clever legal reforms are unworkable. The ones that actually work tend to be boring and incremental — tweaking discovery rules, adjusting cost-shifting formulas, giving judges more discretion to sanction bad behavior.
I keep coming back to something Hilbert said. The CEO who won't admit error. That's not a legal problem. That's a human problem. And the legal system is not well-equipped to solve human problems.
The legal system is designed to solve disputes between parties who disagree about facts or law. It's not designed to solve disputes within a single party about whether to be honest. When a company's legal strategy is driven by internal politics or ego or insurance arithmetic, the court is being asked to provide a service. And courts are terrible service providers.
They're not supposed to be service providers.
They're supposed to be dispute resolvers. The whole architecture — the adversarial system, the rules of evidence, the jury — assumes there's a genuine disagreement to resolve. When there isn't, the architecture doesn't just fail. It becomes complicit.
Complicit in what?
In providing false legitimacy. The company can point to the trial and say "we fought this all the way." They can tell their shareholders, their insurers, their regulators — look, we contested this vigorously, we had our day in court, the system worked. And nobody outside the company knows the fight was a performance from day one.
The court launders the decision. A laundry service.
And until we name this thing properly — give it a label that sticks, that judges and legislators and journalists can use — we're going to keep treating these cases as ordinary litigation. We're going to keep allocating resources to them as if they're genuine disputes.
"Frivolous defense." I'm sticking with that.
It's as good as anything. At least "frivolous" carries the normative weight — this is not okay, this is an abuse.
The question is whether anyone in a position to do something about it agrees.
That's the open question. If we can't get companies to admit their true reasons, and if insurers have perverse incentives to keep the system running, and if the reform proposals all have nasty side effects — what hope is there? Maybe the answer is that this is just a cost of having an adversarial legal system. Every system has its pathologies. This is one of ours.
Or maybe as legal analytics improve, courts will get better at detecting these patterns. The Chicago study is a start. Machine learning models that flag anomalous defense spending, combined with judges willing to ask harder questions at summary judgment — that might shift the needle without any statutory reform.
It might. But it also raises due process concerns. If a judge is looking at an algorithm's output that says "this defense is probably tokenistic," is that fair to the defendant who actually has a legitimate case? The false positive rate matters a lot when someone's access to justice is on the line.
Every solution has a cost. The question is whether the cost of doing nothing is higher.
And we don't know the answer because we don't really know the scale of the problem. Four point two percent of product liability cases. Two to three percent of federal civil trials. Five to eight percent of state trials. Those are estimates with wide error bars. Until we have better data, we're arguing in the dark.
Which is exactly where Daniel's question leaves us. This is a real phenomenon. It wastes court resources. It perverts the purpose of the legal system. And we have no idea how to fix it without breaking something else.
That's the honest answer. I wish I had a neater one.
Neat answers are for people who haven't thought about it hard enough.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop for... well, for knowing things the rest of us only read about.
If you want to send us a prompt that has us this stumped, email the show at show at my weird prompts dot com. Or visit my weird prompts dot com for every episode we've ever done.
We'll be back soon. Try not to file any tokenistic defenses in the meantime.
Or if you do, don't tell your insurer.
This episode was generated with AI assistance. Hosts Herman and Corn are AI personalities.