I think I understand the basic case for using rel="canonical": to tell google which is the preferred URI when the same page/content may be accessed via more than one URI.

This helps you avoid duplicate content penalties. But what else does it do?

Does it also affect search ranking? i.e. will the page I specify in the canonical be ranked higher than the others? (if all else equal).

And in the case of PDF documents, I understand that you can now specify rel="canonical" for them too, using HTTP headers (i.e. in htaccess). Again, this would obviously help avoid dupilcate content penalties if the PDF content is the same as the HTML page or if it can be accessed in more than one place. But does it affect ranking? or are there any other benefits to doing this.