Many publishers (including many European publishers: e.g. Elsevier /
North Holland, Springer / Kluwer) provide authors with a set of LaTeX
macros that will make the submission look exactly like it should if
accepted and published in that journal. Indeed, our own Stata Journal
does the same, providing a set of LaTeX macros that makes it possible
for the author to produce a camera-ready article. Some would carp
that this implies the publisher is pushing the work back on the
author, but in reality it makes it more likely that the typesetting
errors that creep in from de novo typing of an entire manuscript are
prevented.
My review of the last submission I received as an associate editor of
the Stata Journal was delayed by two weeks by the authors' providing
a MS Word document with embedded equations that turned into gibberish
in my perfectly fine copy of MS Word. If and when accepted, the paper
will have to be retyped in LaTeX, the Stata Journal's standard.
Clive Nicholas is correct in noting that most social science journals
accept MS Word and, possibly, LaTeX. In contrast, it is my
understanding that most mathematical journals will not accept
submissions in anything but LaTeX (or its variant amsTeX). And as
Nick Cox points out, there are excellent implementations of LaTeX for
all OSs. My coauthor uses Windows, and his MikTeX (http://miktex.org)
and WinEDT (http:winedt.com) environment creates PDFs with one click
(as does my TeXShop in Mac OS X). The fact that we both use LaTeX
makes it much easier to exchange documents, Stata LaTeX-formatted
output (thanks to Ben Jann's estout) and graphics to be included in
our research papers.
