There was an interesting comment on open access in the latest issue of Nature. A previous comment suggested that open access models based on “pay to publish, free to read” hurt science in developing countries. I very much agree with the recent reply by Massimo Sandal which suggests the opposite. Massimo’s reply is based on the argument that scientists read many more articles than they publish and must first read journals before they can publish them. Of course one would have to look at the average price per article read versus the price per article published to do a full economic analysis. For my own research, I’m sure I’ve read thousands of articles, and published a handful of papers. Some have been in pay-to-publish journals, but most were not. OFf the top of my head, my reading costs have exceeded my publication costs by nearly an order of magnitude.
I think this is still an open question, and it would be interesting to see some data on it.