China's high share of domestic citations may say less about researchers favoring work from home than about the sheer size of the country's research output, according to the analysis. The study finds that China's relative home-citation bias has been shrinking, even as its overall domestic citation rates rose with publication volume.
The study asks whether China's elevated home citation rate reflects genuine preferential citation or the mechanical effect of its rapidly expanding publication volume. Its answer is qualified: the analysis reports a remaining home bias, but says the excess is declining and is not stronger than that seen in other major research-producing countries. The paper defines the home-reference rate as the share of a paper's assignable references attributed to its country, and the home-citation rate as the share of assignable incoming citations from that country.
A bigger network changes the picture
The researchers examined Web of Science articles and reviews published from 1980 to 2025. After removing records with incomplete metadata, they built a directed network with more than 44 million publication nodes and more than 1.8 billion citation links. Each paper was assigned to one country using the first author's first listed institutional affiliation.
They then created a comparison baseline by randomly rewiring citation links while preserving the publication years of the citing and cited papers, the number of references made by each paper, and the number of citations received by each paper. The reshuffling was repeated 10 times. This was designed to ask how much domestic citation would be expected from the network's size and structure alone.
The expected Chinese home-reference rate rose rapidly after 2010, and the expected home-citation rate followed a similar path. Because those expected trends closely matched what was observed, the paper interprets the raw increase in domestic citations as being driven substantially by expanding output, rather than by preferential citation behavior alone.
The relative bias moved toward the baseline
To separate scale from preference, the researchers compared the observed rate with the rate expected under the reshuffled network. An observed-to-expected ratio above 1 means the observed pattern is higher than that baseline. China's ratio for home references declined steadily over roughly the past decade and a half, remained above 1, and moved closer to the random-citation level.
The corresponding relative home-citation ratio was comparatively low against other major countries. It also declined and stayed above 1. In the paper's interpretation, that pattern points to shrinking, not intensifying, home bias. The ratio remaining above 1 means the analysis still found an excess over the reshuffled expectation.
The study also says home bias is not unique to China. All countries in the comparison showed relative home bias above the random baseline, while China's bias was no stronger than that of other major producers and declined over time.
More signs of a changing research profile
The analysis reports that China's reference disruption increased since 2000 and converged with Western countries. Citation disruption still lagged Western levels, but the gap narrowed. Overall persistent disruption approached Western levels.
Persistent disruption is intended to capture research that shifts attention away from earlier work rather than simply building on it. The paper calculates disruption at the level of links between a focal paper and its references, then averages those values across a paper's references and citing papers. High reference disruption combined with low citation disruption is treated as persistent disruption, while lower citation disruption indicates stronger consolidation and a retained foundational role.
The paper also reports that China was the largest source of publications in the dataset. For papers published in Nature, Science and Cell, it ranked second behind the United States, and its category-normalized citation impact, or CNCI, improved.
What the analysis cannot settle
The results do not reveal why any individual researcher cited a domestic paper. A high home citation rate may reflect warranted expertise rather than deliberate preference. The country assignment also reduces each paper to the first author's primary affiliation, so it does not represent the full pattern of international collaboration.
The reshuffled baseline accounts for publication volume and network degree structure, but not field-specific citation norms. These constraints limit the result to the indexed Web of Science corpus and its first-author assignment rule.
The authors interpret the falling relative ratios as evidence that Chinese research is becoming more internationalized, and the convergence in persistent disruption as evidence of improving originality and research impact. The analysis does not identify a single cause for that shift, and it cannot establish whether any particular domestic citation reflected preference or warranted expertise.
Data and disclosure
The raw Web of Science data were obtained under an institutional license and cannot be publicly shared under its terms.
Funding is reported from national S&T innovation strategic research project 2024GH055 for China's 15th Five-Year Plan and from the Shandong Provincial Natural Science Foundation project ZR2023QG129. The authors declare no competing interests.
Paper data and sources
Original title: China's Shrinking Home Bias and Rising Disruptive Impact: Evidence from a Global Citation Network Analysis
Authors: Nan Deng, Zheng Ma, Xiaohui Liu
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-28
DOI: Not available
Original paper · Full text