This is the case in most common law countries.
What is the commonly cited justification for this? Are jury trials shown to be more accurate than bench trials (particularly for criminal cases)?
Jury trials in common law jurisdictions are simply a fact, and don't need or get justification. Jury trials are ancient. In England the Scandinavians had an assembly, the þing ("thing") for deciding matters, such as guilt. Under Norman rule this became systematized, to the point that the Magna Carta Art. 39 states the law that
No free man shall be captured, and or imprisoned, or disseised of his freehold, and or of his liberties, or of his free customs, or be outlawed, or exiled, or in any way destroyed, nor will we proceed against him by force or proceed against him by arms, but by the lawful judgment of his peers, and or by the law of the land
There was a period subsequently where the King gained more power and trials were more arbitrary and oppressive through the Star Chamber, leading to various reforms such as the birth of the US. The right to a jury trial is for reasons of historical precedent part of the US constitution.
Because of the legal right to a jury (especially in criminal cases), the only question that can reasonably come up is for a defendant to ask if they will have better chances with a bench trial versus a jury trial. All it takes is one person on a jury to not be convinced to avoid a conviction, so superficially you'd think an accused person would prefer a jury trial to increase their odds of not being convicted. This is encouraged by a mind-set "the state versus the citizen", where the prosecution and the judge are both agents of the government. On the other hand, the average citizen is more likely to misunderstand the import of evidence, and be less able to disregard statements made during the trial that were held inadmissible.
Civil and criminal trials are fundamentally different, at least as far as the idealized versions go: in a civil trial, you have a disagreement between two "regular citizens" who are equal. There is no imbalance of power.
In a criminal trial, you have a regular citizen vs the entire government and society as a whole. There is a substantial imbalance of power, and almost all of the rules and procedures for criminal trials are there to make sure this imbalance of power cannot be abused or even accidentally disadvantage the defendant.
The idea of "a jury of your peers" ties into that as well: its intent is to make sure that the people determining the defendant's guilt are "regular citizens" as well, that they have similar life experiences, similar education, similar background, etc. It could be argued, for example, that the subset of people who go to law school and become judges and prosecutors are not a representative sample of society, especially when you look at lower income neighborhoods. In the US, for example, the percentage of judges who are people of color is significantly lower than in the entire population, whereas the percentage of defendants of color is significantly higher.
Of course, the idealized versions may not necessarily be congruent with reality. For example, civil suits between a consumer and a mega corporation are most definitely not a disagreement between equals, and when you look at a criminal trial against some well-connected billionaire with friends in very high places, you have to ask yourself which direction the imbalance of power is actually pointing in reality.
It also gets more complicated when the government is one of the two parties in a civil suit, or even different parts of the government are suing each other.
But at its core, these assumptions: "a disagreement between equals" vs. "the entire society against one person", underly a lot of the differences between civil and criminal trials.
The underlying justification for the jury trial is essentially to maintain the popular legitimacy of the justice system.
It should be noted that in its original incarnation, the jury were expected to know the defendant and the victim, so as to better judge their overall characters. In very early times, juries were also expected to do the legwork of investigation for themselves, rather than simply listening to evidence submitted in court under the supervision of a judge.
In the case of aristocratic defendants tried by aristocratic peers, it is of course an obvious point that a jury of aristocrats would tend to know an aristocratic defendant. Defendants drawn from the common masses tended to be tried by a jury of reputable figures drawn from the local community - in that sense not necessarily what true criminals and vagabonds would consider their peers, but certainly the jury would be drawn from the lowest ranks of local people who would be regarded as having some stake in the social order and were embedded in local life.
There are notorious examples in history of the judicial process being operated despotically, such as the Star Chamber, and of terrible consequences following indirectly for the despot (in that case, the beheading of Charles Stuart). Accumulated experience persuaded those at the highest levels that the principle of deferring to the judgment of the multitude on matters of justice, was the safest guarantee against tyrannies, and a guarantee against judicial proceedings triggering popular rebellions.
It is only later, perhaps as late as Victorian times (although I can't be sure offhand of the exact era), that juries were expected to be strangers to the defendant and victim, and to reach a verdict only on the evidence adduced in court. Investigation was increasingly handled by professional law officers and police forces.
At least some of the impetus for this was because growing populations and migrations meant juries could not always be expected to know both defendant and victim, and because more cases occurred in which a jury knowing one or more of the parties would lead to bias.
So the "accuracy" of the jury system - in the technocratic sense of delivering verdicts that correctly apply the existing law to the facts - is of only secondary concern. A primary concern is that accountability for the verdict is passed to the jury, to redirect potential grievance away from the judiciary when correct verdicts cause grievance amongst at least some of the local populace.
It also acts to relieve steam from the system when a judge may undoubtedly have convicted a defendant, but for whatever reason a jury (as representatives of broader community opinion) do not think a conviction is consistent with a just outcome, or have a different systematic perception of the evidence than a judge would.
Notably in recent times, when Kenny Noye was infamously acquitted in 1985 for stabbing a policeman to death, it acted as the strongest possible signal of the public's loss of trust and confidence in the police (in contrast to the professional judiciary who habitually took the word of the police as gospel), and confirmed the need for the police reforms (which were already underway at the time, but followed a period of significant lawlessness and corruption amongst police officers in the 1970s).
Noye was later jailed for an unconnected knife murder, if that bears on the likely "accuracy" of the jury's verdict on him in 1985.
But many serious convictions dating from the 1970s and 80s were later reversed following eventual acceptance amongst the judiciary that the police had often been liars, and jury cases like that of Noye essentially helped to shock the system non-destructively, and served to stem the further accumulation of miscarriages generally.
This is more political than legal, but...
I believe the rationale is not “accuracy”, it is actually just reverse. Jury trials, like https://en.wikipedia.org/wiki/Blackstone's_ratio is all about the state proving that the accused is guilty, and not just executing someone because a single person said so, whether that person be a lord or a judge or whatever.
Jury trials ensure that guilty people go free. They also reduce the chances of innocent people being punished, possibly fatally. They also ensure that there are innocent people, because if nobody is ever found not guilty there’s no reason to bother, and the accused become guilty simply by being accused.
Watch some of the 1st Amendment auditors on YouTube and see cops saying that people are engaged in suspicious behavior and need to show they evidence (typically in the form of a government issued I’d) that they are not engaged in criminal activity and if they don’t comply they will be arrested. Sometimes they even are arrested and charged with hindering an investigation to determine what crimes they have committed. Typically such arrest gets tossed out and occasionally result in law suits and settlements. Imagine if they weren’t, if a cop just saying you didn’t help him determine what crimes you had committed was enough to get you years in jail?
Jury trials, and jury nullification, is one way society tries to reign in government overreach.