In a criminal case, in all common law jurisdictions that I know of, an accusation must be proven to a high level which is commonly stated as being "beyond reasonable doubt", whereas in a civil case the burden of proof of the accusation is typically lower, expressed as something like "preponderance of evidence" or perhaps "more likely than not".

Is there any explanation in the legal literature of what justifies having a distinction in burden of proof?

  • 1
    Don't have time to cite specific literature right now, but typically the reasoning is that a government when prosecuting someone has more resources and should be able to obtain the evidence necessary. Additionally a civil trial doesn't result in imprisonment.
    – Viktor
    Commented Nov 1, 2016 at 0:08

2 Answers 2


James Whitman argues that they are the result of historical developments in the late 18th century which are now applied to a role for which they were not intended.

His thesis is that the "beyond reasonable doubt" standard was originally constructed to protect jurors, not the accused. Specifically, it was protect Christian jurors from eternal damnation for falsely convicting an innocent person, particularly in capital cases.

Theologically, if a juror had doubts, the only way to avoid sin was not to act i.e. to acquit. Indeed John Adams was explicit about this in his (successful) defence during the Boston Massacre trial. The rule was therefore introduced to make it easier for a jury to convict - exactly the opposite of the way it is used today.

In civil cases, where the accused's life or liberty was not at stake, such a rule was not needed; a juror could have any number of doubts and still go with the more convincing argument without imperilling his (they were always men) immortal soul.

As theological interpretation has changed and a more pragmatic and secular view of the law has arisen over the last 2 and a half centuries, the meaning has changed and is now seen as a way to protect the accused where their life and liberty are at stake in a criminal trial.

  • Yet even from the era the arguments were better that ten murderers go free than one innocent hanged.
    – Joshua
    Commented Oct 17, 2023 at 18:48

Because in a civil case you have two equally involved sides. If I claim you damaged my car which cost $10,000 to repair, it's not only that you lose $10,000 if you lose the case, but I lose $10,000 if you win the case. Therefore the burden of proof should be equal in both directions.

Assume a kid at an expensive boarding school is accused of causing $10,000 damage to the headmaster's car, and instead of suing for damages, the headmaster took the $10,000 out of the kid's school fees and tells the parents that the kid will be thrown out if they don't pay another $10,000. We have the exact same argument, but now the headmaster would be in court and accused.

If the burden of proof was "proven beyond reasonable doubt", in the first case the parents would have to pay if it was proven beyond reasonable doubt that the kid was responsible; in the second case they would only get their money back if it was proven beyond reasonable doubt that the kid was not responsible.

  • It sounds like you're saying that the "51% likely" standard ensures that the case is resolved in favor of one party. A higher standard leaves a gulf where the answer is, "Plaintiff doesn't win, but defendant doesn't prevail either?" That seems at least logically compelling, but I have a suspicion that civil actions don't always reduce to "either Plaintiff must win or else Defendant must win," because the defendant isn't (in the simplest case) making any claims. E.g., if A suffers damages $X and claims B is liable, the answer could be "yes to the first stipulation, no to the second."
    – feetwet
    Commented Nov 1, 2016 at 13:16
  • I understand that reasoning, but I'm looking for e.g. law reviews or case law (even from judges of the 1600's) that set forth some reasoning. Have you seen that argument in print? The problem is that with a criminal case, you have two equally involved sides: society and the accused.
    – user6726
    Commented Nov 1, 2016 at 15:55
  • @feetwet, as a separate matter, "51% likely" is arguably a misapplication of the preponderance of evidence instruction. From the epistemological POV, you have three states regarding persuasion: persuaded, dissuaded, and agnostic. "Burden of proof" addresses the situation where there isn't sufficient evidence to reach any conclusion, thus the plaintiff must persuade you, not just "not dissuade you". But that would be another thread...
    – user6726
    Commented Nov 1, 2016 at 15:59
  • @user6726 - Right, I was just trying to distill this answer. If I understand it correctly (which is doubtful) it incorporates a premise that a case can have only two outcomes, when in reality it can have three.
    – feetwet
    Commented Nov 1, 2016 at 16:11
  • @feetwet, specifically: there can be only two legal outcomes ('for A; for B'), but the fact-finders make that choice based on three epistemological states, and a rule about who has the burden of moving you out of that middle state. The 3-to-2 conversion problem is a huge problem in jury instructions, IMO.
    – user6726
    Commented Nov 1, 2016 at 16:29

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .