Innovation in the public sphere: reimagining law and economics to solve the National Institutes of Health publishing controversy

摘要

Eugene Garfield and his company, the Institute for Scientific Information (ISI), originally mapped references for all journals in science and technology to determine the criticality of journals to their respective fields, ‘impact factor'.10 By 1980, Thomson Reuters had purchased ISI and began using ‘impact factor' as a marketing tool to sell the most cited journals from their collection.11 University tenure boards quickly began using the same measurement to evaluate tenure applicant contributions, counting ‘high-impact' articles published,12 effectively incentivizing scientific authors to publish in ‘high-impact' journals.13 This focus on impact factor mostly favored for-profit, large publishing companies with extensive distribution, over smaller publishing companies and university presses. As large publishing companies gained market share, these companies purchased remaining high-impact journals owned by smaller companies, retiring competitor journals and retaining highest impact titles to maximize returns.14 Publishing companies also negotiated long licensing periods with authors as a condition of publication,15 extending the length of time they could charge an access fee for each work. These activities cemented large commercial publishers in the publishing industry and dramatically reduced competition. Historically, large commercial publishers employed traditional publishing models, offering journal copies in print form, employing full paywalls for online versions of articles,16 using a selection process (typically peer review), and negotiating lengthy licensing terms with authors. Traditional publishing models use subscription fees to cover the cost of marketing, operations, and distribution, focusing on attracting individual downloads and subscriptions to cover costs. Traditional publishing companies today offer a hybrid publishing model, in particular dropping paywalls for archived articles or after embargo periods or adopting OA titles.17 Other journals give the authors a choice: authors may publish OA if they pay an upfront fee.18Modern scientific publishing includes a variety of OA options including self-archiving repositories (green OA), author-pays and funded gold OA (peer-reviewed or non-peer-reviewed), and OA data repositories (often green OA).27 OA models, in particular gold OA, can be for-profit or non-profit, and owned by a small university press, non-profit organization, or large publishing company. While in 2005, the NIH relied on voluntary PI deposits in PubMed Central, the Consolidated Appropriations Act of 2007 made depositing in PubMed a requirement.49 Researchers had voluntarily deposited at an initial rate of 7 per cent,50 but mandatory deposit requirements increased participation to 75 per cent by 2012.51 In November 2012, the NIH published a notice on their website that the institutes plan to delay funding payments until authors comply with the publishing terms and may ‘pursue actions' to enforce NIH interests, depending on the ‘degree and duration of non-compliance', beginning in March 2013.52All government grants, including NIH grants with PIs, are governed by a specific set of federal regulations for government contract rights. Until 1887, the US government could claim sovereign immunity and avoid contractual litigation. In 1887 Congress passed the Tucker and Little Tucker Acts, waiving a sovereign immunity defense in contract when non-governmental entities have exhausted government administrative remedies.54Assuming Thermalon establishes a contractual common law standard, the NIH will need to sufficiently characterize whether the depository requirement is a term or condition within the contract. For example, if the NIH intended to treat this requirement as a condition precedent to funding deposit,60 the NIH would not be required to perform required activities (depositing additional funding) until the funding recipient deposited a published article. Another complication includes the NIH's limitations on copyright licensing. Traditionally, an author would retain rights and transfer them at his/her discretion,63 or the work would be covered under an employment contract with the NIH, conferring complete copyrights as a work-for-hire agreement. In this case, the NIH funding agreement requires the transfer of non-exclusive licensing rights to the US government, limiting the PI's legal right to transfer exclusive licensing rights to a publishing company through a future contract. If the PI promises to confer non-exclusive license to the US government, then later promises to confer exclusive licensing rights to the publisher, and both are treated as contracts, the PI may breach the NIH contract and be liable for damages. While much empirical research has analyzed citation dynamics and bibliometrics fo