Introduction
Science works best when everyone has free access to the data, publications, and deliberations arising from scientific studies. Recent advances in information technology make it possible to implement some of Open Science principles, but their adoption has lagged due to structural and social limitations in the scientific enterprise. One notable limitation is the current science publishing system, which arose at a time when ‘closed science’ policies of private data, confidentiality, and restricted access were the norm. The current system impedes open science through its ever-increasing demands for long sexy stories and drives some of the biggest problems in science today. These include data irreproducibility, delayed publication, omission of key data, inability to publish orphan observations or experimental reproductions, biased peer review, excessive rounds of review and revision, irrational parameters of scientific impact, and the harmful exclusion of reviewer and community feedback from the early and middle phases of scientific studies.
The Academic Publishing Industry
Academia is facing serious well-documented challenges from current models for scholarly publishing: problems in peer review, poor evaluation criteria for research and researchers, absence of fair credit for contribution, non-communication and redundancy, long delays in publication, and irreproducibility of research data and results. There are several reasons why these challenges exist including primarily the exaggerated commercial interest of publishers, who act as middlemen in what should actually be a peer-to-peer evaluation process. This is compounded by a “publish or perish” culture, rendering these problems systemic and deeply ingrained, so much so that both science and wider society suffer.
At present, scientific and medical research is typically submitted to be peer reviewed and then published in academic journals or at conferences. According to the International Association of Scholarly Publishing’s market report from 2015, the total market is estimated to be worth over $25 billion USD a year, of that over $10 billion USD is in the English language. Adjusted profit margins for some of the largest publishers are in the region of 37-40% after tax, exposing the industry for the highly exploitative practices that it employs. The research publishing industry is dominated by a small number of publishers that charge high fees for authors to publish. It has been described as ‘oligarchic’ (
Lariviere et al., 2015), and the largest publisher Elsevier has been reported for abuse of a dominant position within a dysfunctional scholarly publishing market (
Eve et al., 2016). The cost to publish is on average around $5000, based on the total annual revenue for scholarly publishers divided by the total number of articles published each year (
Schimmer et al., 2015). Estimates show that less than 10% of submitted articles are eventually published, and the peer review process is often very slow. Moreover, journals charge individual subscription fees of $200 to $300 a year to access the published research, with institutional fees often being around $10000 to $15000 per journal.
The fallout is a situation where publicly funded researchers and other taxpayers must pay to access research that was originally publically funded and peer reviewed gratuitously by those same publicly funded researchers. The vast majority of research still remains inaccessible to most people on this planet (
Piwowar et al., 2018). This has deep and negative consequences for the progress of research, and its wider use in society to help resolve major issues that affect economies, the environment, and health on a global scale.
The current academic research evaluation system also needs a much needed overhaul (e.g.,
DORA, the
Leiden Manifesto). At present, it is a profoundly unscientific practice, with far-reaching negative consequences (e.g.,
Brembs et al., 2013). At the moment it is geared towards explicitly rewarding the publication of research articles in ‘high impact’ journals that are often owned by commercial publishers. This creates a perverse system of incentives, which leads to a ‘publish or perish’ culture, and the current ‘reproducibility crisis’. Consequently, the production of other outputs (e.g., code, data, hardware) is seen as secondary or lesser in value, which creates an output-driven bias in research.