Judges face bureaucratic pressure to maintain high conviction rates. According to the following paper, it can have negative effects on a Judge's career if he appears to be lenient:

http://www.rasmusen.org/published/Rasmusen-01.JLS.jpncon.pdf

Only in a system where judges are elected, like the US. Many, if not most European countries (to discuss the examples I'm familiar with) appoint judges and have had a jury trial system until the beginning of the 20th century. It was abandoned after it was found to be easy to manipulate on a regular basis. Jury trials degenerated into character trials - prosecution would try to discredit the accused, and defense would try to paint him as a pillar of the community. Juries placed too much weight on their personal feeling towards the accused, and too less on what happened and what the evidence was.

We like to hope that the US system solved this problem, of course.

I believe only a few US states elect judges; they're usually appointed.

Even if a judge is appointed, there could still be public outrage over the perception of low conviction rates. This would cause another judge to be appointed. The same thing goes on with cabinet positions of all levels of government where the public demonstrates outrage.

I think the line of thought is that a jury is not as easily bought, harder to threaten, and better at ironing out the moral ambiguities that can be the failings of a single person.

However, a judge and no jury is certainly more expedient, neither is infallible in terms of their judgement. That is the reason for appeals.

"How does the judge/jury distinction come into play? Is there empirical evidence that trials with juries have more accurate outcomes?"

Juries may be less accurate and the OP's point still stands - there's a difference. I can easily imagine that juries tend to convict less often. That doesn't mean they're more accurate.

The line I've heard is that experienced lawyers would rather be tried by a judge if they were innocent, but a jury if they were guilty.