Sauropod Vertebra points to a 2013 paper today, “arXiv e-prints and the journal of record: An analysis of roles and relationships”. It seems to indicate that not all articles in certain fields are free on arXiv, as is popularly imagined. Much may be slipping through the sieve, even in fields professing to adore arXiv…
“Even in mathematics, the field that is most committed to arXiv, only a feeble 21.5% of published papers are also available on arXiv! In physics, it’s 20%, and ‘Earth and Space’ it’s a smidge under 12%. For everything else, it’s virtually nothing.”
Nor is everything on ArXiv reaching the Web of Science, it seems. I did a quick search and found a long abstract of another paper by the same research team, “On the scientific impact of ArXiv: A case study of astrophysics”. Using a large 1990 to 2012 trawl of arXiv, they found that…
“slightly less than 50% of arXiv submissions [on astrophysics] were also found in WoS [Web of Science]”
Admittedly, a March 2012 sample is now four years out-of-date, and things may have changed since.