Mediation Actually Works (Statistically Speaking, at least for DOJ)
I like mediation. I mediate for a couple of the Maryland Circuit Courts and have done a few private mediations. I think it works. That's why I'm a beleiver. But I never found any good statistics on it. (And as anyone who's read my posts about bad faith statistics, I like statistics.)
(I suspect the people who coordinate the ADR programs at the local courts have tons of good data; they just can't or don't share it.)
Anyway, the Justice Department has put out some of their data. This is data is about the results of the Justice Department adopting ADR in its cases, not macro statistics about ADR.
You can read the results yourself. The bottom line is that the Justice Department saved a massive amount of money by paying for mediation and came out ahead as a result of the mediation (closer to resolution, narrower issues, etc.) around 75% of the time in bad years. (i.e. in a "bad" year, they resolved half of cases and got some benefit from the ADR in about half the remaining cases.)
This confirms what I think everyone already believed based on anecdotal evidence: ADR saves everyone --even or maybe especially repeat players-- small fortunes.