Stephen Dubner has a good "football freakonomics" post laying out the evidence that home field (or court) advantage is a real phenomenon across American professional sports and that it's overwhelmingly due to referee bias, rather than "intangibles" about crowd support or home cooking.
What he doesn't do is illustrate the sound business strategy principles at work here. Researchers emphasize that the bias appears to be completely unconscious, which is true. But it's also so evident in the statistics that you might think league officials would be working diligently to extirpate the bias from the system. In practice, they're totally complacent. And with good reason! The issue is that team-level sports revenues are driven, at the margin, by wins and losses. Obviously different teams have different structural levels of revenue built in (a .500 Mets team will earn more than a .500 Royals team) but in marginal terms more wins = more ticket sales. And home-biased officiating is a neat way to get around the problem that every win requires a loss. If officials systematically have their thumbs on the scale of the home team, then most teams can win most of their home games. And home games are, after all, where you sell your tickets (and parking passes, beer, jerseys, etc.) so this is what matters. Fans want to root, root, root for the home team and see a win. Biased officiating delivers the goods.
Note that this means that technology-driven improvements in official accuracy (instrant replay and such) create a business model dilemma. If you flagrantly refuse to use accuracy-improving new technology, you develop a credibility problem that can end up threatening to undermine the whole sport. But making the close calls more accurate isn't actually desirable.