Some preprints on arXiv have 1.5- or double-spaced lines, making it less pleasant to read and more wasteful to print.
Is there a reason for it, except for the laziness/sloppiness? (If drafts were >1-spaced, one need to change only one option.)
I think the real underlying cause is just what Anonymous Mathematician said, that having a double-spaced version of a paper is more convenient for going through and making notes on a printed copy. But it's not just a phenomenon of the dark ages; many people still do this today during the revision process when a paper is being prepared for submission.
In fact, at least in physics (my field), the tradition of double-spacing is perpetuated by the fact that at least one of the major journal publishers' LaTeX class defaults to a
It's worth noting that arXiv recommends not submitting in double-spaced mode. But probably very few submitters have the patience to read through the instructions in their entirety.
I don't like it either. My understanding is that it's for historical reasons: in the dark ages, publishers asked for double-spaced manuscripts (produced on typewriters) because it would leave space for the typesetters to mark up the manuscript. Some authors got used to doing this and have continued the tradition to this day, despite the fact that publishers no longer care, and a few younger authors even imitate it. I think a lot of people find it a little annoying, but they typically don't care enough to complain to the authors, so it will take a while for this practice to die out.
P.S. It's not just the arXiv, but also journal submissions.