Causal evidence of racial and institutional biases in accessing paywalled articles and scientific data

2025-09-10

AI Summary

This study investigates whether knowledge-access inequities faced by Global South researchers—arising from paywalls and restrictive data-access policies—are rooted in racial and institutional biases. Employing a mixed-methods approach, it integrates observational analysis of 250 million publications, semi-structured interviews, and two-dimensional randomized email audit experiments targeting 18,000 and 11,840 fictitious doctoral candidates. For the first time, it disentangles and quantifies the distinct causal effects of racial identity versus institutional affiliation on scholarly access outcomes. Results show that race significantly reduces response rates to requests for paywalled articles, whereas institutional prestige predominantly determines dataset approval decisions. Informal gatekeeping practices constitute a structural barrier, exacerbating constraints on epistemic breadth and widening disparities in academic influence. The findings provide empirical grounding for advancing scientific equity and identify precise policy levers—particularly around transparent access protocols and bias-mitigating review criteria—for intervention.

Technology Category

• Philosophy and Ethics of AI: Bias, Fairness & Equity
• Application Domains: Humanities & Computational Social Science
• Computer Vision: Bias, Fairness & Privacy

Application Category

• Security and Privacy: Data transparency and provenance
• Social Networks and Social Media: Fairness and bias in social network and social media analysis
• Semantics and Knowledge: Provenance, trust, security and privacy, and ethical issues in managing semantic data

Abstract

Scientific progress fundamentally depends on researchers' ability to access and build upon the work of others. Yet, a majority of published work remains behind expensive paywalls, limiting access to universities that can afford subscriptions. Furthermore, even when articles are accessible, the underlying datasets could be restricted, available only through a "reasonable request" to the authors. One way researchers could overcome these barriers is by relying on informal channels, such as emailing authors directly, to obtain paywalled articles or restricted datasets. However, whether these informal channels are hindered by racial and/or institutional biases remains unknown. Here, we combine qualitative semi-structured interviews, large-scale observational analysis, and two randomized audit experiments to examine racial and institutional disparities in access to scientific knowledge. Our analysis of 250 million articles reveals that researchers in the Global South cite paywalled papers and upon-request datasets at significantly lower rates than their Global North counterparts, and that these access gaps are associated with reduced knowledge breadth and scholarly impact. To interrogate the mechanisms underlying this phenomenon, we conduct two randomized email audit studies in which fictional PhD students differing in racial background and institutional affiliation request access to paywalled articles (N = 18,000) and datasets (N = 11,840). We find that racial identity more strongly predicts response rate to paywalled article requests compared to institutional affiliation, whereas institutional affiliation played a larger role in shaping access to datasets. These findings reveal how informal gatekeeping can perpetuate structural inequities in science, highlighting the need for stronger data-sharing mandates and more equitable open access policies.

Problem

• Examining racial and institutional bias in accessing paywalled scientific articles
• Investigating disparities in response rates to data and article access requests
• Analyzing how informal gatekeeping perpetuates structural inequities in science

Innovation

• Randomized email audit experiments design
• Large-scale observational analysis of citations
• Semi-structured qualitative interviews combination