Institutional broadening of the Japanese astronomy and astrophysics research community during 1996-2025

arXiv:2610.16633 · astro-ph.IM · Submitted 2026-10-12 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics. The day in astrophysics: observations, models and what the data forces.

Vera: We're looking at the paper: "Institutional broadening of the Japanese astronomy and astrophysics research community during 1996-2025".

Paper summary: Jocelyn: The authors look at bibliometric data to see if the institutional base for Japanese astronomy and astrophysics got wider between one thousand nine hundred ninety-six and two thousand twenty-five <ref:2610.16633#pg1>.

Vera: They did a count of Japan-affiliated publications, and that number went from five hundred eighty-eight in nineteen ninety six to two thousand one hundred fifty-six in twenty twenty five.

Jocelyn: And they also counted the unique Japanese institutions involved, which grew from one hundred twenty three to two hundred twenty seven.

Vera: That jump tells us there was a nominal broadening of the base, but they immediately point out that this isn't uniform participation.

Subrahmanyan: I’m curious about that concentration part. If you have more institutions involved, does it mean everyone has an equal shot at contributing to the big picture?

Jocelyn: Not really. The paper shows that even with more participating institutions, the links between publications and institutions stayed concentrated among the leading ones.

Vera: It means a lot of those papers are still coming from a core group of major players.

Subrahmanyan: So, if we look at the top tier—say, the top five or top ten contributors—the share of those contributions actually went down slightly over time, moving from about fifty to fifty-five percent in the late nineteen nineties down to around forty-five percent by early twenty twenty five.

Jocelyn: That’s a modest weakening of concentration among those very top institutions, which is something interesting.

Vera: But wait, they also used this measure called Shannon entropy, which gives them an effective number of institutions based on frequency weighting.

Subrahmanyan: That effective number increased over the period, meaning participation got more diverse in a weighted sense than just the raw count of unique institutions.

Jocelyn: The paper says this increase in effective numbers was more gradual than the nominal count of unique institutions they found.

Vera: So, to put it simply, what they found is an expansion on the edges of the community around a still-strong center, rather than everything becoming equally distributed across all participants.

Subrahmanyan: That suggests that while new groups are joining in, the established players are still driving most of the output.

Jocelyn: It’s about seeing a broadening of the periphery instead of a complete decentralization of participation across all these institutions.

Vera: That's what their analysis points toward, and it sets up what they do next by looking at those institutional structures more closely.

Conclusion: Vera: So, wrapping up this look at "Institutional broadening of the Japanese astronomy and astrophysics research community during one thousand nine hundred ninety-six–two thousand twenty-five" <ref:2610.16633#pg1,Institutional broadening of the Japanese astronomy and astrophysics research community during 1996>. The core finding is that we see a nominal expansion in the number of institutions involved, going from one hundred twenty three to two hundred twenty seven.

Jocelyn: But what matters more than that raw count is how those publications are linked back to the institutions themselves.

Vera: The paper concludes that this broadening happened, but it wasn't a complete leveling out of participation across all these entities.

Subrahmanyan: It confirms what we often see in large science communities: you get more people involved, but the established leaders still dominate the publication landscape.

Jocelyn: Exactly. They show that concentration among the top institutions weakened modestly, but it didn't disappear entirely; those major players are still heavily represented in the output.

Vera: So what does this mean for us looking at cosmology and structure formation?

Subrahmanyan: It means that while new groups are joining the research effort, if you’re trying to build a massive theoretical framework, you still need to keep your focus on those established centers of gravity.

Jocelyn: It tells us that follow-up work should probably look at how these newer institutions are connecting to that core group.

Vera: It shows the institutional structure is evolving, but the influence distribution remains uneven across the whole system.

Jocelyn: The key thing here is recognizing that participation diversity can be measured in two ways: just counting unique players, or measuring how frequently they show up in the actual publication records.

Frontier Research Institute for Interdisciplinary Sciences, Tohoku University · Publications of the Astronomical Society of Japan

astro-ph.IM

Submitted: 2026-10-12

Updated: 2026-10-12

Comments: 6 pages, 4 figures; Author's Original Version. This article has been accepted for publication in Publications of the Astronomical Society of Japan Published by Oxford University Press

DOI: 10.1093/pasj/psag154

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The gist: Using bibliometric data from publication records, we investigate the institutional broadening of Japanese astronomy and astrophysics over 1996–2025 The observed change is best interpreted as an…

Key concepts

Institutional Base Expansion
This refers to the increase in the total number and variety of Japanese institutions involved in astronomical and astrophysics publications over time. It shows that more different organizations are contributing to this field, suggesting a wider reach for research.
Concentration Among Leaders
This measures how much publication activity is dominated by the very top institutions. Although participation broadened, the analysis showed that the share of publications from the top five or ten institutions only slightly decreased, meaning major players still hold a significant portion of influence.
Frequency-Weighted Diversity (Effective Number)
This metric uses Shannon entropy to measure how diverse institutional participation is based on publication frequency. It shows that while the nominal count of unique institutions grew, the actual distribution of research links became more varied in terms of how often different institutions contributed.
Institutional Periphery vs. Core
The findings suggest that the observed change is not a complete decentralization but rather an expansion of the outer layer (periphery) around established leading institutions (the core). This means new institutions are joining, but they are still largely linked to the major existing centers.

Terminology

Summary

Using bibliometric data from publication records, we investigate the institutional broadening of Japanese astronomy and astrophysics over 1996–2025 The observed change is best interpreted as an expansion of the participating institutional base around a persistent core of leading institutions

Institutional Base Expansion

The study analyzed bibliometric data to determine if the institutional base of Japanese astronomy and astrophysics broadened over the period 1996–2025 The number of Japan-affiliated publications increased from 588 to 2,156, and the number of unique participating Japanese institutions increased from 123 to 227, indicating a nominal broadening of the institutional base However, institution–publication links remained concentrated among leading institutions.

Concentration Among Leaders

The analysis revealed that while the number of participating institutions increased, concentration among the very top institutions weakened modestly The top-5 and top-10 shares showed a gradual decline from the late 1990s to the 2020s, indicating a modest weakening of concentration among the very top institutions. Furthermore, the top-15 and top-20 shares remained higher throughout the study period, showing that a large fraction of institution–publication links continued to be associated with the broader upper tier of leading institutions. p. 3

Frequency-Weighted Diversity

A complementary measure, the effective number of institutions derived from Shannon entropy, showed that institutional participation became more diverse in a frequency-weighted sense The effective number increased over the study period, indicating that institutional participation became more diverse in a frequency-weighted sense. However, this increase was more modest than the nominal number of unique institutions shown in the bottom panel of Figure 1. p. 3

Contextual Interpretation

The results suggest that the observed pattern is best interpreted as an expansion of the institutional periphery around a still-concentrated core, rather than as a complete equalization or decentralization of institutional participation This broadening reflects an increase in the number and variety of institutions appearing in publication records, while the frequency of participation remained uneven and concentrated among major institutions.

Limitations

Several limitations should be noted regarding the study's methodology SciVal Institution IDs represent SciVal-normalized institutional entities and do not necessarily correspond one-to-one to legal organizations, university corporations, campuses, or research institutes. Additionally, SciVal sector metadata and the manually assigned academic institution types should be interpreted as analytical classifications rather than official institutional categories. p. 4 Finally, because the dataset is based on Scopus and SciVal, the results may differ from analyses based on other bibliographic databases such as ADS. p. 4

Conclusion

The overall pattern is therefore best described as a broadening of the participating institutional base accompanied by a persistent, though modestly weakened, concentration among leading institutions The broadening observed here reflects an increase in the number and variety of institutions appearing in the publication records, while the frequency of participation remained uneven and concentrated among major institutions.

Acknowledgments

This work was supported by JSPS KAKENHI Grant Number JP26K03645.

References

Abt, H. A. 1981, PASP, 93, 207

Abt, H. A. 1996, PASP, 108, 1059

Bornmann, L. 2012, EMBO Rep. p. 5, 13, 673

Fujiwara, H. 2025, PASJ, 77, 6, 1192

Fujiwara, H. 2026, PASJ, 78, 2, 405

Hill, M. O. 1973, Ecology, 54, 427

Ministry of Education, Culture, Sports, Science and Technology (MEXT) 2025, School Basic Survey (Gakkou Kihon Chousa), Portal Site of Official Statistics of Japan (e-Stat)

Moed, H. F. 2010, Journal of Informetrics, 4, 265

Shannon, C. E. 1948, Bell Labs Technical Journal, 27, 3, 379

Waltman, L. 2016, Journal of Informetrics, 10, 365

--- Page 5 ---

Table 1 summary: The index is normalized to 1.00 in 1996 Counts of Japanese higher education institutions were compiled from the School Basic Survey conducted by MEXT and obtained via e-Stat (MEXT 2025). p. 4

Table 1 data comparison: The number of unique Japanese institutions increased from 123 to 227, whereas the number of Japanese higher education institutions decreased slightly from 1,236 to 1,162 over the same period. p. 3

Table 1 data comparison: The academic sector accounted for the largest component of the increase in unique Japanese institutions in the bottom panel of Figure 1. p. 4

Table 1 data comparison: To examine this component in more detail, we manually classified academic-sector institutions into national universities, public universities, private universities, two-year colleges, and technical colleges. p. 4

Table 1 data comparison: The comparison indicates that the increase in unique academic-sector institutions was not driven solely by national universities. p. 4

Table 1 data comparison: The broadening of the academic sector was not driven primarily by an expansion of national universities, but also involved public universities, private universities, technical colleges, and other nonnational-university institutions in the publication records. p. 4

--- Page 4 ---

Figure 4 summary: The black line shows the total number of Japanese institutions classified as academic in the SciVal sector metadata, corresponding to the academic component in the bottom panel of Figure 1.

Figure 4 summary: The stacked areas show manually classified higher education categories. p. 4

Figure 4 summary: The comparison indicates that the increase in unique academic-sector institutions was not driven solely by national universities, but also involved public universities, private universities, and technical colleges. p. 4

--- Page 2 ---

The Scopus database used for retrieval via SciVal was last updated on June 24, 2026 Because data retrieval was conducted several months after the end of 2025, the 2025 publication records were considered sufficiently indexed for the purposes of this longitudinal analysis.

The resulting institution list, hereafter referred to as the Japanese institution list, was used as the basis for all institution-level analyses in this study. p. 1

We did not manually reconstruct historical institutional entities or reassign publications to predecessor institutions. p. 1

--- Page 1 ---

The number of Japan-affiliated publications increased from 588 to 2,156, and the number of unique participating Japanese institutions increased from 123 to 227 The resulting institution list, hereafter referred to as the Japanese institution list, was used as the basis for all institution-level analyses in this study.

--- Page 4 ---

Selected parent-level umbrella institutions were excluded from the Japanese institution list before computing institution-level indicators.

When multiple authors in the same publication were associated with the same Institution ID, or when the same Institution ID appeared more than once within a publication record, that Institution ID was counted only once for that publication. p. 1

For each year t, we computed the number of publications, Npub(t), the number of unique Japanese institutions, Ninst(t), and the number of institution–publication links for institution i, Li(t). p. 1

The total number of Japanese institution–publication links in year t is Ltot(t) = X i Li(t). p. 1

For each year, institutions were ranked by Li(t), and the top-k share was defined as Sk(t) = P i∈top k Li(t) / Ltot(t). p. 1

The Shannon entropy is then H(t) = - X i p i(t)log2 p i(t), (4). p. 1

and the corresponding effective number of institutions is Neff (t) = 2H(t). p. 1

--- Page 5 ---

The effective number can be interpreted as the order-one Hill number and provides a frequency-weighted measure of institutional diversity, complementing the nominal count of unique institutions.

This is equivalent to the order-one Hill number and provides a frequency-weighted measure of institutional diversity, complementing the nominal count of unique institutions. p. 1

--- Page 4 ---

The top-5 share decreased from roughly 50–55% in the late 1990s and around 2000 to about 45% in early 2020s, while the top-10 share declined from about 65% to about 60%.

The declines in the top-15 and top-20 shares are less pronounced. p. 3

--- Page 4 ---

The results show that the institutional base broadened while the distribution of institution–publication links remained uneven The number of participating institutions increased, and concentration among the very top institutions weakened modestly, but publication participation remained strongly associated with the upper tier of leading institutions.

--- Page 4 ---

The observed change is therefore best interpreted as an expansion of the institutional periphery around a persistent core of leading institutions, rather than as a complete equalization or decentralization of institutional participation.

--- Page 5 ---

The effective number increased over the study period, indicating that institutional participation became more diverse in a frequency-weighted sense This means that the increase in participating institutions was not purely nominal; the distribution of institution–publication links also became somewhat less concentrated.

--- Page 5 ---

The overall pattern is therefore best described as a broadening of the participating institutional base accompanied by a persistent, though modestly weakened, concentration among leading institutions The observed change is therefore best interpreted as an expansion of the institutional periphery around a persistent core of leading institutions, rather than as a complete equalization or decentralization of institutional participation.

--- Page 4 ---

The observed change is therefore best interpreted as an expansion of the institutional periphery around a persistent core of leading institutions, rather than as a complete equalization or decentralization of institutional participation The observed change is therefore best interpreted as an expansion of the institutional periphery around a persistent core of leading institutions, rather than as a complete equalization or decentralization of institutional participation.

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. Improved AI systems can distinguish between nominal institutional broadening and frequency-weighted diversity by calculating both measures; this allows them to interpret institutional expansion as an expansion of participation around a persistent core of leading institutions, rather than as a complete equalization or decentralization of institutional participation.

  2. The improved system can quantify the relationship between publication output growth and institution count growth by analyzing the Number of publications increased substantially over the study period, from 588 in 1996 to 2,156 in 2025, corresponding to an increase by a factor of 3.7, while simultaneously noting that the growth in participating institutions was more moderate than the growth in publication output.

  3. The AI can identify patterns of institutional concentration by assessing the annual share of links accounted for by the top k institutions, noting that Persistently high top-k shares indicate that institutional participation remained concentrated among leading institutions despite the broadening of the overall institutional base.

Related papers