Two young siblings approach a parent.
“Tracy took my hairbrush,” says Alex.
“I did not. It’s not her hairbrush. It’s mine. She took it out of my room, so I took it back.”
The parent must resolve the dispute. Who has the burden of proof?
Under our justice system, that’s easy to answer. Alex does. Alex claims ownership. Tracy has possession. Alex has accused Tracy of a tort: conversion. In our system, the accuser has the burden of proof, and the opposing party’s possession is presumed to be rightful until proven otherwise.
But who should have the burden of proof? That’s not so easy to answer, at least not on the civil side of the justice system.
Not having heard any evidence, the parent has no basis to give even the slightest benefit of the doubt to either sibling. Two children are arguing over ownership of a hairbrush. Each makes the same claim: “I am the true owner. My sibling is falsely claiming that I am not.” Only one can be telling the truth, but the truth cannot be discerned from a mere accusation and denial.
Actually, there isn’t a single accuser. Each child is an accuser, accusing the other of having taken something of theirs. And each is a denier, denying having done anything wrong.
Of course, only one child has possession, and a convenient maxim says that possession is nine-tenths of the law. But why should that matter? If Tracy had come to the parent while the brush was still in Alex’s room, their roles would be reversed. Tracy would be claiming theft, Alex would be in possession, and Tracy would have the burden of proof.
Does Tracy’s entering Alex’s room and taking the hairbrush justify shifting the burden of proof to Alex?
It seems that our civil justice system has developed procedural rules of convenience, rules like “the plaintiff has the burden of proof” and “if the plaintiff fails to persuade you by a preponderance of the evidence, the defendant wins.” But if you think those rules are driven by truth-finding or fairness, they’re not. Rather, they seem to exist for the sake of having a common framework on which every stakeholder in the dispute can rely as providing the terms of engagement.
Let’s look at this more carefully.
Why do we have a burden of proof? Why does it belong to the party on the left side of the v? Is this a good thing? Is there a better way to do it?
These questions are easy to answer on the criminal side. On the civil side, not so much.
In a criminal case, the burden of proof is on the government, and it’s a heavy one. The government might have a strong case and great evidence to support it, but if the government fails to foreclose a simple doubt, that doubt, if reasonable, will allow a defendant who committed a crime to go free.
This serves an important policy: Depriving a person of liberty is so destructive to the defendant’s humanity and potentially so harmful to the community’s conscience that we need ample protection against mistaken outcomes. We also need this protection to curb zealots within the government from using the levers of a prosecution to terrorize those who are innocent but disliked. So we guarantee defendants a lawyer, a trial, the right to confront and interrogate their accusers, the right to remain silent, the right not to be convicted by improperly obtained evidence, and the right to have their guilt or innocence determined by a neutral jury. And on top of that, we give the government the burden of proving the case beyond a reasonable doubt, something approaching a moral certainty.
These safeguards no doubt allow some criminals to walk. But we tolerate some mistaken acquittals for the sake of avoiding mistaken convictions. A wrongful conviction is an offense against morality. Not so for a wrongful acquittal.
That explains why the government has the burden of proof in a criminal case and why the burden is so high. But it doesn’t explain why, in a civil case, the burden of proof belongs to the plaintiff.
Let’s look at one of the oldest and most famous legal disputes. When two women went to King Solomon, each claiming to be the mother of the same baby, it would have been terribly unfair had Solomon assigned the burden of proof to either of them. To whom should he have assigned it? To the first woman who approached him, claiming to be the true mother? To the woman who happened to be holding the baby when the claim was made?
Had there been burdens of proof back then as we now know them, Solomon might have said to the first woman: You’re claiming to be the true mother and that the other woman is not. So you have the burden of proof. If you fail to persuade me that your claim is true, or if I find the evidence evenly balanced, then you’ll lose, and I’ll declare the other woman to be the true mother.
But that would have been arbitrary. Were the evidence inconclusive, the woman would have lost simply for coming forward first. And had Solomon given the burden of proof to the other woman, it would have been equally arbitrary, with an equal chance of an unjust outcome. After all, what hung in the balance was the awesome responsibility of uniting the child with the true mother. Should such an outcome turn on how the burden of proof is assigned?
Solomon didn’t think so. Determined to get his decision right, he neither needed nor wanted a burden of proof. Applying a burden of proof would have risked producing an outcome dependent on process. Instead, he only needed evidence.
And he found it. By threatening to cut the baby in half, he caused the true mother to scream in horror and to offer to abandon her claim to spare the baby’s life. That reaction was the evidence, for it revealed the strong maternal bond on which Solomon could make the right decision.
Assigning a burden of proof would have had no relationship to discovering the truth. Not then. And not today.
For reasons unrelated to truth-finding, the burden of proof advantages one side at the expense of the other. It tilts the playing field against the party who has it. If that party fails to persuade the fact finder that the party’s contention is true, that party loses, even if the contention is true. The party without the burden of proof gets the benefit of the doubt.
This isn’t to say that burdens of proof are necessarily bad. It’s only to say they’re unnatural. They’re a human invention, not driven by truth-finding. They serve other purposes.
As we’ve seen, in criminal cases they’re meant to keep innocent people out of jail. But what purpose do they serve in civil cases?
Sometimes, a statute or common-law rule will assign issue-specific burdens of proof for policy reasons. For example, to benefit accident victims, a statute might make the vehicle owner vicariously liable for the driver’s negligence, unless the owner proves that the driver was not operating the vehicle for the owner’s benefit. Or to protect individuals from undeserved reputational harm, a party asserting fraud, whether as a claim or defense, must prove it with clear and convincing evidence.
But aside from specially created burdens of proof on specific issues, what truth-seeking, justice-promoting ends are served by placing a general burden of proof on plaintiffs?
Consider how cases were decided in the early 18th century before burdens of proof evolved to what they are today. Back then, judges would tell juries to let their conscience guide them as to whose evidence made the most sense and to rule for that party. They would say to weigh the evidence, weighing whatever evidence favored a proposition against the evidence that opposed it, without reference to a burden of proof. Jurors would decide the case by crediting whatever evidence they found the most believable, pretty much like the Solomonic model.
Today though, as the law has evolved, the burden of proof belongs to the accuser. That favors the accused, the defendant. The accuser may have the truth on her side, but if she can’t persuade the adjudicator to believe her, then the accused, without having to prove anything, will escape any consequence and justice won’t be done.
So how did we move from the evenly balanced instructions of the early 18th century to the tilted instructions we use today?
Some procedural rules are codified in the Federal Rules of Civil Procedure and its state analogues. But they don’t mention the burden of proof.
Other procedural rules are nested in case law. Those rules evolve. They mutate. They creep into a case through skillful advocacy. That case then becomes a precedent, and a precedent often repeated becomes a norm. Over time, it becomes a legal article of faith.
That’s how the civil burden of proof got to where it is today and, according to scholarly research, where it has been since probably the middle to latter part of the 19th century.
Some argue that the burden of proof belongs to plaintiffs to discourage frivolous suits. That might be a post hoc argument to preserve the status quo, but it doesn’t seem related to how the burden of proof evolved. Nor is the burden of proof a particularly potent weapon against frivolous suits. We have so many other mechanisms: Rule 11, abuse-of-process claims, anti-SLAPP statutes, bad-faith fee-shifting statutes, the Rules of Professional Conduct, dispositive motions, and the generally high cost of litigation, to name a few.
For those reasons, eliminating the burden of proof would hardly open the floodgates to more frivolous litigation than what the current machinery stops. If a suit is frivolous, if the facts are weak or the case law hostile, the suit will fail and get flushed away, regardless of who has the burden of proof.
So what justice-promoting purpose does today’s burden of proof serve? What relationship does it have to the delivery of true civil justice? To understand just how much the civil burden of proof disadvantages the party who has it, consider the combined effect of two common jury instructions. The first goes like this:
The plaintiff has the burden of proving its claim by a preponderance of the evidence. For the plaintiff to meet this burden, you must be persuaded by the evidence that the plaintiff’s claim is more probably true than not true.
And the second goes:
In deciding the facts, you may have to decide which testimony to believe and which testimony not to believe. You may believe everything a witness says, or part of it, or none of it. What is important is how believable the witnesses were, and how much weight you think their testimony deserves.
The second instruction gives the jury license, if it wants to, to disregard all the testimony offered by the party who has the burden of proof. For that reason, that party not only must offer enough evidence to prove that the claim is more probably true than not, but also must make jurors care enough to do their job and to disregard the judge’s permission to be arbitrary about whether to credit any testimony at all.
Put another way, the party with the burden of proof has the practical burden of moving jurors from a state of indifference to a state of caring about the outcome. That may be a far weightier burden than simply offering sufficient, logical evidence on each element of a claim. A failure to meet this weightier burden will doom the case.
In combination, those two instructions can make it easier for defendants to win. Under those instructions, jurors who lean toward avoiding confrontation or avoiding the cognitive dissonance that comes with making the wrong decision could easily conclude that the evidence is equally balanced. But if they do that, the default rule takes over and the defendant wins, not because the jury thought the defendant was in the right but because the jury found it cognitively difficult to choose between two conflicting versions of the truth.
What would be fairer? Let’s start with the burden of production, who goes first. Instead of a rule that automatically puts the burden of production on the plaintiff, we could do what the National Football League does: The winner of a coin toss could decide whether to go first or let the other side go first. As a consolation, the loser of the toss could decide whether to present the closing argument first or last. Some modifications might need to be made for multiparty cases or bifurcated trials, but variations could be developed to handle special situations.
Now the bigger challenge: Other than when some policy reason allocates the persuasion burden for certain issues (fraud, for example), what if we dispensed with the general burden of persuasion altogether? Why do we need it? Yes, it establishes who wins in the event of a tie, but how fair is that? In tennis, ties are unacceptable. The match simply isn’t over until one player wins on the merits. Shouldn’t we expect the same from the civil justice system?
Wouldn’t it be an improvement if the rules left no room for a tie and required adjudicators to choose between one side or the other, based on the totality of the evidence and on which side had the better case, without trying to apply a burden of proof? Is there a reason not to be Solomonic in this sense?
And consider the practical effect of the persuasion burden. It handicaps the plaintiff and gives the defendant a big lead. If you doubt that, do this thought experiment: Say you represent only defendants. How much more concerned would you be if, in your next trial, no one had the burden of proof and jurors were told that they had to evaluate all the evidence, credit the evidence that made the most sense to them without regard to which side offered it, and apply the law to the facts they found based on the evidence so credited? Now imagine you represent only plaintiffs. If the trial were conducted under those same rules, how much happier would you be?
We often assume that whoever wrote the rules was right. When rules have been around a long time, they become like gospel. We assume they serve a salutary purpose and the interests of justice, without asking how we got them or whether they make sense. So no one should begrudge a defense counsel for harnessing the benefits that the civil burden of proof offers.
But if a commission were formed to figure out how to deliver better justice, and if the commission were to list some rules for reexamination, maybe the commission should consider adding the civil burden of proof to the list.
