Importance of free access to research articles on decision to submit to the BMJ: survey of authors

Introduction

The traditional publishing model based on subscriptions from readers has been heavily criticised for restricting access to important scientific information while generating large profits for publishers.1 2 Realising the importance of free unrestricted access to scientific information, the BMJ introduced an open access policy in 1998, whereby all readers could gain free access to all journal content online with no restrictions.3 Free access to content, however, comes with a financial cost to the BMJ Publishing Group through a potential loss of subscriptions. Reduced revenue from subscriptions was one of the key reasons for putting some content (including editorials and education and debate articles) behind access controls in January 2005.4 The BMJ, however, has retained free access to all original research articles on bmj.com. While free access to research articles is important to BMJ editors we do not know how much our authors value it and how they would perceive the introduction of access controls for this material.

Methods

I emailed the first author of 479 research articles (papers, primary care papers, and short reports) published in the BMJ between 1 January 2003 and 31 March 2005 an invitation from the BMJ to complete an electronic survey (see bmj.com). I excluded authors of papers who were not categorised initially on our manuscript tracking system.

Results

Sixty four email addresses were incorrect, and 51% (211/415) of the remaining authors responded. Most (75%, 159/211) said the fact that all readers would have free access to the specific paper they had published on bmj.com was very important or important to their decision to submit their paper to the BMJ (table).

I asked authors whether closure of free access to research articles would influence the likelihood of their submitting research articles to the BMJ in the future. Most (53%, 111/211) said they would be slightly less likely to submit and 14% (29/211) would be much less likely to submit. A third (71/211) said this would not influence their decision.

Authors were equally divided in their opinion as to whether the closure of access to parts of the journal since January 2005 had affected their view of the BMJ: 40% (84/211) said it had and 38% (80/211) said it had not. Around a fifth (47/211) were not aware that we had closed access to parts of the journal, possibly because they have institutional subscriptions allowing automatic full access. In contrast, two thirds of authors (141/211, 67%) said their view of the BMJ would change if we closed access to research articles, 20% (42/211) said it would not change their view, and 13% (28/211) were not sure.

The box gives some illustrative sample quotes of how authors' views of the BMJ have been affected since we closed access to parts of the journal and how their views would be affected if we closed access to research papers. Comments largely focused on disappointment with a regressive step in the era of open access publishing, loss of a unique feature of the BMJ “that sets you apart from most other major journals,” a perceived reduction in the journal's usefulness as a resource and global influence, restricted readership, less attractive to publish in, and the negative impact on the journal's image. None of the quotes were negative about open access. All the comments received from authors are available on bmj.com.

Discussion

Authors clearly value free access to BMJ research articles and consider this an important factor in deciding whether to submit to the journal. Closing access to research articles would have a negative effect on authors' perceptions of the journal and their likeliness to submit.

This study was limited by a low response rate (51%) and unfortunately I cannot compare responders and non-responders in terms of demographics and research experience as this type of information about individual authors is not kept. One possible reason for the low response rate was that the BMJ was simultaneously conducting another online author survey and authors may have felt overburdened. The response rate, however, is comparable with rates of other surveys with professionals (published surveys of physicians have a mean response rate of 54%).5 Responding authors may have tried to emphasise a particular message to the publishing group and may have been advocates of open access publishing in general. Regardless, the results show that the issue was important to many authors, even if all the non-responders were indifferent.

The individual comments from participants suggest that closure of access to research articles is likely to have a considerable negative impact on the image, and therefore potentially the strategic and long term financial success and viability, of the BMJ. The publishing group has agreed to keep free access to research articles for now.