Ending the NIH embargo accelerated public access to funded research, but substantial delays remain

arXiv:2609.14842 · cs.DL, cs.CY · Submitted 2026-09-13 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Computer Science. The day in computer science: software, systems and theory on arXiv.

Tom: I'm Tom.

Jane: And I'm Jane, here with Lu, Meng and our in-house large language model, Lalam.

Tom: We're looking at the paper: "Ending the NIH embargo accelerated public access to funded research, but substantial delays remain".

Jane: Here's the gist of the paper. Ending the NIH embargo accelerated public access to funded research, but substantial delays remain On 1 July 2025 the National Institutes of Health began requiring that funded articles be publicly available in PubMed Central on the day they are published, ending the 12-month embargo that had applied since 2008 Comparing NIH-supported articles with articles in the same journals and months that reported no US federal support, before and after the policy, the share available in PubMed Central within a week of publication rose by 5.6 percentage points more for NIH-supported articles (95% CI 1.9 to 9.6) The change was concentrated in journals outside the Directory of Open Access Journals, where the share of NIH-supported articles in PubMed Central within a week doubled from 11% to 22%, with no differential change in DOAJ-listed journals For articles published in early 2026, the policy had made NIH-supported research available sooner, and most NIH-supported articles in journals outside the Directory were still not in PubMed Central a week after publication Study Design and Cohorts The study compared two publication cohorts of NIH-supported journal articles indexed in PubMed, one published January to April 2025 and one published January to April 2026 Cohorts are defined by publication period; the policy applies by acceptance date Each NIH-supported article was matched to an article published in the same journal and calendar month that reported no US federal support The outcome is whether the article was publicly available in PMC within 7 days of its publication date, taken as the earlier of the PubMed and OpenAlex dates to reduce misclassification from issue dates The estimate is the 2025-to-2026 change among NIH-supported articles minus the change among comparison articles in the same journal-month cells, with journal-cluster bootstrap confidence intervals Key Findings on Policy Impact The policy raised availability within a week by about six percentage points The difference-in-differences was 5.6 percentage points (95% CI 1.9 to 9.6; Fig.

Tom: 1C) The same contrast on the day of publication was 5.0 points (1.6 to 9.0) and at 30 days it was larger at 8.7 points (4.2 to 13.7) The increase was concentrated in non-DOAJ journals Among the 266 journals not listed in the Directory of Open Access Journals (DOAJ), the share of NIH-supported articles in PMC within 7 days rose from 11.0% to 21.6%, while the comparison share rose from 9.3% to 11.5%; the difference-in-differences was 8.4 points (3.8 to 12.9) Among the 61 DOAJ journals, there was a difference of 0.0 points (−6.0 to 6.2) Limitations and Future Questions The remaining gap is the more consequential finding Among non-DOAJ journals, 78% of NIH-supported articles in the matched sample were still not in PMC a week after publication Whether this gap closes through faster manuscript deposit, publisher deposit, or a shift of NIH-funded work toward open-access venues is the question the next cohort will answer The estimate is a publication-cohort contrast for articles in the matched journals, not the effect of the policy on any individual article Availability in PMC is not readership, and an article outside PMC may be readable elsewhere; the analysis says nothing about who paid Methodological Details NIH-supported articles were identified in PubMed by publication types “Research Support, N.I.H., Extramural” or “Research Support, N.I.H., Intramural” with “Journal Article”, for January to April of 2024, 2025, and 2026 Comparison candidates were OpenAlex articles with a PMID in the same journal, year, and month, excluding any with a US government research-support publication type or an NIH or NSF grant agency in PubMed The primary contrast (2025 versus 2026) comprised 1,880 NIH-supported and 1,880 comparison articles per year in 327 journals and 638 journal-month cells The pre-policy contrast (2024 versus 2025) was built the same way on the larger set of journals with NIH-supported articles in both of those years: 4,757 and 4,757 articles per year in 698 journals The difference-in-differences estimate is defined as X j wj hy¯ N j,2026 − y¯ C j,2026 − y¯ N j,2025 − y¯ C j,2025 i All sampling used a single fixed seed The public repository contains the article-level data, the script that regenerates every estimate reported in the main text and in this supplement together with the cumulative-availability curves and bands in Fig.

Jane: 1, the figure script, and a provenance inventory listing the source, snapshot date, and query for each input The code is available at https://github.com/Wang-Haining/nih embargo The author designed the study, performed the analysis, and wrote the paper All inputs are public (PubMed, PMC, OpenAlex, Crossref, DOAJ) The 7-day horizon was fixed before any 2026 outcome was observed The definition of the publication date was settled after the audit described in the supplement Independent Support Identification NIH support was also identified from publisher-deposited funder metadata Crossref funder coverage of the PubMed-identified NIH cohort was 41.0% in 2025 and 39.7% in 2026 The difference-in-differences was re-estimated with NIH support defined by this independent source and comparison articles restricted to those without NIH funding in Crossref Among the 1,880 matched comparison articles in each year, 12 (0.64%) in 2025 and 13 (0.69%) in 2026 carried NIH funding in Crossref; removing them gives 5.8 points (2.1 to 9.8) Restricting the NIH-supported group to articles with an explicit NIH grant agency in the PubMed record gives 5.8 points (2.0 to 10.0) The analysis assumes that, without the policy, the gap between NIH-supported and comparison articles in the same journal would have stayed constant The pre-policy contrast over the preceding year, on a broader set of journals, provides supporting evidence References

1: National Institutes of Health.

Tom: Revised policy on enhancing public access to archived publications resulting from NIH-funded research.

Paper summary: Tom: This paper is looking at the impact of that July 1st, two thousand twenty-five requirement to make funded research public on day-of-publication <ref:2609.14842#pg1>. Jane The core thesis here is comparing NIH-supported articles with those in similar journals that didn't have federal support before and after the policy was introduced >

Lu: The comparison is done across different publication periods, specifically January to April twenty twenty five and January to April twenty twenty six, matched by journal and calendar month >

Meng: They used a specific method where they matched each NIH-supported article to an article published in the same journal and month that reported no US federal support >

Tom: The outcome they are measuring is whether the article was publicly available in PubMed Central within seven days of its publication date, using the earlier of the PubMed and OpenAlex dates to avoid misclassification >

Jane: The estimate they're reporting is a difference-in-differences calculation, looking at how much more likely NIH articles were to be in PMC within seven days compared to the same change happening for comparison articles >

Lu: The results show that the policy raised availability within a week by about six percentage points overall, with a difference-in-differences of five point six percentage points over the whole study period >

Tom: But there's a big caveat here because they split their findings based on journal type, which is where things get interesting >

Jane: They found this effect was concentrated in journals that were not listed in the Directory of Open Access Journals, or DOAJ >

Lu: In those non-DOAJ journals, the share of NIH-supported articles in PubMed Central within a week doubled from eleven percent to twenty two percent >

Meng: While the effect wasn't seen in the DOAJ-listed journals at all, they showed zero differential change there >

Conclusion: Tom: So looking at the title, "Ending the NIH embargo accelerated public access to funded research, but substantial delays remain", it suggests a mixed outcome >

Jane: It shows that while the policy did speed things up for some research, there are still significant delays happening in other areas of funded work >

Lu: The authors found that availability within seven days increased by about six percentage points for NIH-supported articles when compared to non-federal support articles in the same journals >

Meng: That increase was heavily skewed toward those journals not in the DOAJ, which is a key piece of information for how this policy actually affects access across different publishing styles >

Lalam: The biggest concern they raise is that seventy eight percent of NIH-supported articles in their matched sample were still not in PubMed Central a week after publication >

Tom: So what this means for the world, it suggests that just having a policy change isn't the whole answer when it comes to making funded research freely available >

Jane: The question they leave open is whether this gap shrinks through faster manuscript deposit or publisher deposit or if NIH-funded work shifts toward venues that are already open access >

Lu: And the authors admit that their estimate is a publication-cohort contrast for articles in the matched journals, not necessarily the effect of the policy on any single article >

Meng: It means we still need to figure out how to get this research into public repositories more reliably if we want real access for everyone >

Haining Wang

Indiana University School of Medicine

cs.DL, cs.CY

Submitted: 2026-09-13

Updated: 2026-10-31

Code: https://github.com/Wang-Haining/nih_embargo

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 83/100

The gist: Ending the NIH embargo accelerated public access to funded research, but substantial delays remain On 1 July 2025 the National Institutes of Health began requiring that funded articles be publicly…

Key concepts

NIH Embargo
This was the previous rule requiring funded research articles to be hidden from the public for 12 months after publication. Ending this embargo meant researchers could make their work publicly available much sooner.
PubMed Central (PMC)
PMC is a free digital repository that hosts full-text versions of biomedical and life sciences journal articles. The study measured how quickly NIH-funded articles appeared in PMC after the new policy began.
Non-DOAJ Journals
DOAJ stands for Directory of Open Access Journals, which are journals that commit to being open access. The study found that the positive impact of ending the embargo was much stronger in journals not listed in DOAJ, suggesting a greater need for this change there.

Terminology

Summary

Ending the NIH embargo accelerated public access to funded research, but substantial delays remain On 1 July 2025 the National Institutes of Health began requiring that funded articles be publicly available in PubMed Central on the day they are published, ending the 12-month embargo that had applied since 2008 Comparing NIH-supported articles with articles in the same journals and months that reported no US federal support, before and after the policy, the share available in PubMed Central within a week of publication rose by 5. p. 0

Where the paper says it

edu ABSTRACT On 1 July 2025 the National Institutes of Health began requiring that funded articles be publicly available in PubMed Central on the day they are published, ending the 12-month embargo that had applied since 2008. Comparing NIH-supported articles with articles in the same journals and months that reported no US federal support, before and after the policy, the share available in PubMed Central within a week of publication rose by 5.

6 percentage points more for NIH-supported articles (95% CI 1. p. 0
Where the paper says it

Comparing NIH-supported articles with articles in the same journals and months that reported no US federal support, before and after the policy, the share available in PubMed Central within a week of publication rose by 5. 6 percentage points more for NIH-supported articles (95% CI 1.

9 to 9.6) The change was concentrated in journals outside the Directory of Open Access Journals, where the share of NIH-supported articles in PubMed Central within a week doubled from 11% to 22%, with no differential change in DOAJ-listed journals For articles published in early 2026, the policy had made NIH-supported research available sooner, and most NIH-supported articles in journals outside the Directory were still not in PubMed Central a week after publication

Study Design and Cohorts

The study compared two publication cohorts of NIH-supported journal articles indexed in PubMed, one published January to April 2025 and one published January to April 2026 Cohorts are defined by publication period; the policy applies by acceptance date Each NIH-supported article was matched to an article published in the same journal and calendar month that reported no US federal support The outcome is whether the article was publicly available in PMC within 7 days of its publication date, taken as the earlier of the PubMed and OpenAlex dates to reduce misclassification from issue dates The estimate is the 2025-to-2026 change among NIH-supported articles minus the change among comparison articles in the same journal-month cells, with journal-cluster bootstrap confidence intervals

Key Findings on Policy Impact

The policy raised availability within a week by about six percentage points The difference-in-differences was 5.6 percentage points (95% CI 1. p. 0

Where the paper says it

Comparing NIH-supported articles with articles in the same journals and months that reported no US federal support, before and after the policy, the share available in PubMed Central within a week of publication rose by 5. 6 percentage points more for NIH-supported articles (95% CI 1.

9 to 9.6; Fig. 1C) The same contrast on the day of publication was 5. p. 1
Where the paper says it

6 percentage points (95% CI 1. The same contrast on the day of publication was 5.

0 points (1. p. 0
Where the paper says it

Comparing NIH-supported articles with articles in the same journals and months that reported no US federal support, before and after the policy, the share available in PubMed Central within a week of publication rose by 5. 6 percentage points more for NIH-supported articles (95% CI 1.

6 to 9.0) and at 30 days it was larger at 8. p. 1
Where the paper says it

What the number means For articles published in early 2026, about one additional NIH-supported article in eighteen was in PMC within a week of publication that would otherwise not have been, and about one in twelve in journals outside DOAJ. The 30-day estimate is larger than the 7-day one.

7 points (4. p. 0
Where the paper says it

Comparing NIH-supported articles with articles in the same journals and months that reported no US federal support, before and after the policy, the share available in PubMed Central within a week of publication rose by 5. 6 percentage points more for NIH-supported articles (95% CI 1.

2 to 13.7) The increase was concentrated in non-DOAJ journals Among the 266 journals not listed in the Directory of Open Access Journals (DOAJ), the share of NIH-supported articles in PMC within 7 days rose from 11. p. 1
Where the paper says it

0 points; 3. The increase was concentrated in non-DOAJ journals Among the 266 journals not listed in the Directory of Open Access Journals (DOAJ), the share of NIH-supported articles in PMC within 7 days rose from 11.

0% to 21.6%, while the comparison share rose from 9. p. 1
Where the paper says it

The increase was concentrated in non-DOAJ journals Among the 266 journals not listed in the Directory of Open Access Journals (DOAJ), the share of NIH-supported articles in PMC within 7 days rose from 11. 6%, while the comparison share rose from 9.

3% to 11.5%; the difference-in-differences was 8. p. 0
Where the paper says it

For articles published in early 2026, the policy had made NIH-supported research available sooner, and most NIH-supported articles in journals outside the Directory were still not in PubMed Central a week after publication. Keywords open access · public access policy · PubMed Central · science policy · difference-in-differences Since 2008, NIH has required that peer-reviewed articles arising from its funding be deposited in PubMed Central (PMC), but publishers could delay public release for up to 12 months [1].

4 points (3. p. 0
Where the paper says it

Comparing NIH-supported articles with articles in the same journals and months that reported no US federal support, before and after the policy, the share available in PubMed Central within a week of publication rose by 5. 6 percentage points more for NIH-supported articles (95% CI 1.

8 to 12.9) Among the 61 DOAJ journals, there was a difference of 0. p. 1
Where the paper says it

0 points; 3. The increase was concentrated in non-DOAJ journals Among the 266 journals not listed in the Directory of Open Access Journals (DOAJ), the share of NIH-supported articles in PMC within 7 days rose from 11.

0 points (−6. p. 0
Where the paper says it

Comparing NIH-supported articles with articles in the same journals and months that reported no US federal support, before and after the policy, the share available in PubMed Central within a week of publication rose by 5. 6 percentage points more for NIH-supported articles (95% CI 1.

0 to 6.2)

Limitations and Future Questions

The remaining gap is the more consequential finding Among non-DOAJ journals, 78% of NIH-supported articles in the matched sample were still not in PMC a week after publication Whether this gap closes through faster manuscript deposit, publisher deposit, or a shift of NIH-funded work toward open-access venues is the question the next cohort will answer The estimate is a publication-cohort contrast for articles in the matched journals, not the effect of the policy on any individual article Availability in PMC is not readership, and an article outside PMC may be readable elsewhere; the analysis says nothing about who paid

Methodological Details

NIH-supported articles were identified in PubMed by publication types “Research Support, N.I.H., Extramural” or “Research Support, N. p. 2

Where the paper says it

Ending the NIH embargo accelerated public access to funded research Preprint Materials and Methods NIH-supported articles were identified in PubMed by the publication types “Research Support, N. , Extramural” or “Research Support, N.

I.H., Intramural” with “Journal Article”, for January to April of 2024, 2025, and 2026 Comparison candidates were OpenAlex articles with a PMID in the same journal, year, and month, excluding any with a US government research-support publication type or an NIH or NSF grant agency in PubMed The primary contrast (2025 versus 2026) comprised 1,880 NIH-supported and 1,880 comparison articles per year in 327 journals and 638 journal-month cells The pre-policy contrast (2024 versus 2025) was built the same way on the larger set of journals with NIH-supported articles in both of those years: 4,757 and 4,757 articles per year in 698 journals The difference-in-differences estimate is defined as X j wj hy¯ N j,2026 − y¯ C j,2026 − y¯ N j,2025 − y¯ C j,2025 i All sampling used a single fixed seed The public repository contains the article-level data, the script that regenerates every estimate reported in the main text and in this supplement together with the cumulative-availability curves and bands in Fig. p. 4
Where the paper says it

The primary contrast (2025 versus 2026) comprised 1,880 NIH-supported and 1,880 comparison articles per year in 327 journals and 638 journal-month cells. The pre-policy contrast (2024 versus 2025) was built the same way on the larger set of journals with NIH-supported articles in both of those years: 4,757 and 4,757 articles per year in 698 journals.

1, the figure script, and a provenance inventory listing the source, snapshot date, and query for each input The code is available at https://github. p. 6
Where the paper says it

com/Wang-Haining/nih_ embargo) contains the article-level data, the script that regenerates every estimate reported in the main text and in this supplement together with the cumulative-availability curves and bands in Fig. 1, the figure script, and a provenance inventory listing the source, snapshot date, and query for each input.

com/Wang-Haining/nih embargo The author designed the study, performed the analysis, and wrote the paper All inputs are public (PubMed, PMC, OpenAlex, Crossref, DOAJ) The 7-day horizon was fixed before any 2026 outcome was observed The definition of the publication date was settled after the audit described in the supplement

Independent Support Identification

NIH support was also identified from publisher-deposited funder metadata Crossref funder coverage of the PubMed-identified NIH cohort was 41.0% in 2025 and 39. p. 0

Where the paper says it

ENDING THE NIH EMBARGO ACCELERATED PUBLIC ACCESS TO FUNDED RESEARCH, BUT SUBSTANTIAL DELAYS REMAIN Haining Wang Department of Biostatistics and Health Data Science Indiana University School of Medicine Indianapolis, IN 46202 hw56@iu. edu ABSTRACT On 1 July 2025 the National Institutes of Health began requiring that funded articles be publicly available in PubMed Central on the day they are published, ending the 12-month embargo that had applied since 2008.

7% in 2026 The difference-in-differences was re-estimated with NIH support defined by this independent source and comparison articles restricted to those without NIH funding in Crossref Among the 1,880 matched comparison articles in each year, 12 (0. p. 5
Where the paper says it

The difference-in-differences was re-estimated with NIH support defined by this independent source and comparison articles restricted to those without NIH funding in Crossref, in the journal-month cells of the main analysis that contained both groups in both years, with the same publication-date rule. This replication sample comprised 1,118 NIH-supported and 877 comparison articles in 2025 and 1,055 and 880 in 2026, across 171 journals and 271 cells.

64%) in 2025 and 13 (0. p. 0
Where the paper says it

ENDING THE NIH EMBARGO ACCELERATED PUBLIC ACCESS TO FUNDED RESEARCH, BUT SUBSTANTIAL DELAYS REMAIN Haining Wang Department of Biostatistics and Health Data Science Indiana University School of Medicine Indianapolis, IN 46202 hw56@iu. edu ABSTRACT On 1 July 2025 the National Institutes of Health began requiring that funded articles be publicly available in PubMed Central on the day they are published, ending the 12-month embargo that had applied since 2008.

69%) in 2026 carried NIH funding in Crossref; removing them gives 5. p. 5
Where the paper says it

64%) in 2025 and 13 (0. 69%) in 2026 carried NIH funding in Crossref; removing them gives 5.

8 points (2. p. 0
Where the paper says it

Comparing NIH-supported articles with articles in the same journals and months that reported no US federal support, before and after the policy, the share available in PubMed Central within a week of publication rose by 5. 6 percentage points more for NIH-supported articles (95% CI 1.

p. 0
Where the paper says it

Comparing NIH-supported articles with articles in the same journals and months that reported no US federal support, before and after the policy, the share available in PubMed Central within a week of publication rose by 5. 6 percentage points more for NIH-supported articles (95% CI 1.

1 to 9.8) Restricting the NIH-supported group to articles with an explicit NIH grant agency in the PubMed record gives 5. p. 5
Where the paper says it

8 points (2. Restricting the NIH-supported group to articles with an explicit NIH grant agency in the PubMed record gives 5.

8 points (2.0 to 10.0) The analysis assumes that, without the policy, the gap between NIH-supported and comparison articles in the same journal would have stayed constant The pre-policy contrast over the preceding year, on a broader set of journals, provides supporting evidence

References

[1] National Institutes of Health. Revised policy on enhancing public access to archived publications resulting from NIH-funded research. p. 2

Where the paper says it

References [1] National Institutes of Health. Revised policy on enhancing public access to archived publications resulting from NIH-funded research.

Notice NOT-OD-08-033, 2008 [2] National Institutes of Health. p. 0
Where the paper says it

ENDING THE NIH EMBARGO ACCELERATED PUBLIC ACCESS TO FUNDED RESEARCH, BUT SUBSTANTIAL DELAYS REMAIN Haining Wang Department of Biostatistics and Health Data Science Indiana University School of Medicine Indianapolis, IN 46202 hw56@iu. edu ABSTRACT On 1 July 2025 the National Institutes of Health began requiring that funded articles be publicly available in PubMed Central on the day they are published, ending the 12-month embargo that had applied since 2008.

2024 NIH public access policy. p. 0
Where the paper says it

Keywords open access · public access policy · PubMed Central · science policy · difference-in-differences Since 2008, NIH has required that peer-reviewed articles arising from its funding be deposited in PubMed Central (PMC), but publishers could delay public release for up to 12 months [1]. In December 2024 NIH removed the embargo [2], and in April 2025 it moved the effective date forward to manuscripts accepted on or after 1 July 2025 [3], following the 2022 White House directive that all federal agencies do the same [4].

Notice NOT-OD-25-047, 2024 [3] National Institutes of Health. p. 2
Where the paper says it

[3] National Institutes of Health. Revision: Notice of updated effective date for the 2024 NIH public access policy.

Revision: Notice of updated effective date for the 2024 NIH public access policy. p. 2
Where the paper says it

[3] National Institutes of Health. Revision: Notice of updated effective date for the 2024 NIH public access policy.

Notice NOT-OD-25-101, 2025 [4] Alondra Nelson. p. 2
Where the paper says it

Accessed 13 September 2026. [4] Alondra Nelson.

Ensuring free, immediate, and equitable access to federally funded research. p. 2
Where the paper says it

[4] Alondra Nelson. Ensuring free, immediate, and equitable access to federally funded research.

Memorandum, Office of Science and Technology Policy, Executive Office of the President, 2022 [5] Caitlin R. p. 2
Where the paper says it

Ensuring free, immediate, and equitable access to federally funded research. Memorandum, Office of Science and Technology Policy, Executive Office of the President, 2022.

Ryus, Caroline Raymond King, and Edward R. p. 2
Where the paper says it

[5] Caitlin R. Ryus, Caroline Raymond King, and Edward R.

Melnick. The NIH 2025 public access policy: Immediate access, unequal costs. p. 2
Where the paper says it

Ryus, Caroline Raymond King, and Edward R. The NIH 2025 public access policy: Immediate access, unequal costs.

PLOS Med., 23(6):e1005124, 2026 [6] Jason Priem, Heather Piwowar, and Richard Orr. p. 2
Where the paper says it

, 23(6):e1005124, 2026. [6] Jason Priem, Heather Piwowar, and Richard Orr.

OpenAlex: A fully-open index of scholarly works, authors, venues, institutions, and concepts. p. 2
Where the paper says it

[6] Jason Priem, Heather Piwowar, and Richard Orr. OpenAlex: A fully-open index of scholarly works, authors, venues, institutions, and concepts.

arXiv [Preprint], 2022 [7] Crossref, REST API documentation. p. 6
Where the paper says it

01833 (accessed 13 September 2026). Crossref, REST API documentation.

https://api. p. 6
Where the paper says it

Crossref, REST API documentation. https://api.

crossref.org [8] National Library of Medicine, PMC OAI-PMH service. p. 6
Where the paper says it

org (accessed 12 September 2026). National Library of Medicine, PMC OAI-PMH service.

https://pmc. p. 2
Where the paper says it

All inputs are public (PubMed, PMC, OpenAlex, Crossref, DOAJ). Code, article-level data, and every result table are at https://github.

ncbi.nlm.nih.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Focus on non-DOAJ journals for targeted research funding strategies; since The increase was concentrated in non-DOAJ journals, AI systems can prioritize tracking research output from these specific journal clusters to identify emerging high-impact areas outside established open access venues.

  2. Develop predictive models for NIH funding impact based on publication venue; the finding that the 78% of NIH-supported articles in the matched sample were still not in PMC a week after publication suggests an AI can predict the likelihood of delayed public access for new research submissions based on their journal type and funding source.

  3. Create automated compliance monitoring tools for researchers and institutions; by comparing NIH-supported articles with articles in the same journals and months that reported no US federal support, AI can flag when NIH-funded manuscripts are not deposited in PubMed Central within the required timeframe, mitigating potential policy violations.

Abstract

NIH-funded articles accepted from 1 July 2025 must be public in PubMed Central (PMC) on the day of publication, ending a 12-month embargo in place since 2008. Whether ending the embargo has made articles available sooner has not been measured. Using 1,880 NIH-funded articles published January to April 2026 and 1,880 from 2025, each matched to a same-journal, same-month article that PubMed does not list as US government supported, we compare the change in availability, dated by the day each article went live in PMC. By four months, availability had risen 19.4 percentage points more among NIH-funded articles than among comparison articles. Within a week of publication, however, 6 to 8% of articles in every group and year were available, only 0.1% of NIH-funded articles on the day itself, and the policy effect was 0.3 percentage points. Articles reach PMC by two routes: publishers deposit the final version, or the accepted manuscript is deposited with the author's approval. Two thirds of the gain came through author manuscripts, whose median wait from publication, among those live by the data cut, fell from 374 days in 2025 to 70 in 2026; final versions from publishers moved from 44 days to 36. Agencies ending their embargoes should count days to public access, not only eventual deposit, and test publisher-triggered submission at acceptance.

Sources

Related papers