arxiv: 2601.05093 · v2 · submitted 2026-01-08 · 💻 cs.SI

Measuring Structural Political Fragmentation

Yuan Zhang , Laia Castro , Frank Esser , Alexandre Bovet This is my paper

Pith reviewed 2026-05-16 16:15 UTC · model grok-4.3

classification 💻 cs.SI

keywords political fragmentationnetwork polarizationE-I indexeffective number of communitiesmultiscale fragmentationeffective branching factoronline political networkscommunity detection

0 comments

The pith

Political fragmentation in networks must be measured by separating the strength of group separation from the number of groups and adding a metric for how groups split at different scales.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper shows that standard measures such as modularity and the original E-I index blend two distinct aspects of fragmentation: how strongly groups are cut off from one another and how many such groups exist. This mixing makes it difficult to compare fragmentation levels across different countries or platforms. The authors identify the pairwise adaptive E-I index as a cleaner gauge of separation strength and the effective number of communities as a better count of fragments. They add a new metric, the effective branching factor, that tracks how larger fragments split into smaller ones across multiple scales. When these three measures are applied to online networks from Brazil, Spain, and the United States, the country orderings remain consistent.

Core claim

The two aspects of network fragmentation are best captured by the pairwise adaptive E-I index and the effective number of communities, while other measures confound the strength of separation and the number of fragments. The effective branching factor is introduced to capture multiscale fragmentation by measuring how fragments at one level split into smaller fragments at the next level. Applying these metrics to empirical datasets from Brazil, Spain, and the United States yields consistent country rankings.

What carries the argument

The effective branching factor (EBF), which quantifies the average number of child fragments produced by each parent fragment across successive levels of a community hierarchy.

If this is right

Cross-country and cross-platform comparisons of fragmentation become interpretable once separation strength, fragment count, and branching are tracked separately.
Multiscale political structures can be described by the rate at which groups subdivide from one level to the next.
Earlier studies that relied on confounded metrics may produce misleading rankings when reanalyzed with the separated dimensions.
Changes in fragmentation over time can be monitored by watching each dimension independently in longitudinal network data.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same three metrics could be tested on offline political networks or survey-based affiliation data to check whether online patterns generalize.
A high effective branching factor might indicate greater difficulty in forming stable coalitions, an implication worth testing against legislative voting records.
If the separation-strength and branching measures move together in response to external shocks, they could serve as early-warning signals for rising political instability.
The consistent cross-dataset rankings suggest that some structural properties of fragmentation are comparable across different democratic systems.

Load-bearing premise

Online interaction networks faithfully represent structural political fragmentation without platform-specific artifacts or selection biases in the data collection.

What would settle it

Recomputing the three metrics on the same datasets but with an alternative community detection method that yields substantially different country rankings would show the measures are not robust.

Figures

Figures reproduced from arXiv: 2601.05093 by Alexandre Bovet, Frank Esser, Laia Castro, Yuan Zhang.

**Figure 2.** Figure 2: Community merging patterns based on similarity in ideological positions. Heatmaps (left column) [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Figure 3: Community merging patterns based on similarity in social identity. Heatmaps (left column) showing [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

**Figure 4.** Figure 4: Multi-level pie charts showing the proportions of different social identity categories for each com [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗

**Figure 5.** Figure 5: Correlation heatmap between ideological position and social identity in Brazil, Spain, and the United [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗

**Figure 6.** Figure 6: Comparison of the distributions of demographic variables - age, gender, ethnicity, religion, income, [PITH_FULL_IMAGE:figures/full_fig_p014_6.png] view at source ↗

**Figure 7.** Figure 7: The optimal clusters for multi-scale community detection in Brazil (BR), Spain (ES), and United [PITH_FULL_IMAGE:figures/full_fig_p015_7.png] view at source ↗

**Figure 8.** Figure 8: Distribution of community sizes and percentages of singleton communities–those with a single node– [PITH_FULL_IMAGE:figures/full_fig_p015_8.png] view at source ↗

read the original abstract

Political fragmentation denotes the differentiation of a political system into multiple groups and the extent of separation among them. It often manifests structurally in online interaction behaviors. To measure and compare political fragmentation across contexts, previous scholarship has often relied on network measures of polarisation such as modularity and the Krackhardt E-I index. Here, we show that these metrics combine two aspects of fragmentation: the strength of separation and the number of fragments. These two aspects have not been clearly distinguished in previous work, making comparisons across varied systems difficult to interpret. In addition, none of them is designed to capture the multiscale fragmentation structures that characterize real-world multi-dimensional political spaces. We compare several network measures and show that the two aspects of network fragmentation are best captured by the pairwise adaptive E-I index and the effective number of communities (ENC), while other measures confound the strength of separation and the number of fragments. Furthermore, we introduce a novel metric for multiscale fragmentation, the effective branching factor (EBF), capturing how political fragments at one level split into smaller fragments at the next level. Applying EBF to two empirical datasets spanning Brazil, Spain, and the United States yields consistent country rankings across datasets. Overall, these results clarify three complementary dimensions of structural political fragmentation: strength of separation, number of fragments, and between-level branching. They support a more holistic characterization of structural political fragmentation.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

This paper clarifies how standard network metrics mix separation strength with fragment count and introduces an effective branching factor for multiscale fragmentation, with consistent country rankings but thin validation details.

read the letter

The main thing to know is that this paper separates two things that often get lumped together in polarization measures—the strength of separation between groups versus how many groups there are—and it adds a new effective branching factor to track how fragments split across levels. It does well by comparing several common metrics, showing that the pairwise adaptive E-I index isolates separation strength cleanly while the effective number of communities tracks fragment count without the usual confounding. The EBF looks like a straightforward extension for multiscale structure, and applying it to the Brazil, Spain, and US datasets produces consistent country rankings across the two collections, which is the right kind of check for a methods piece. The soft spots sit mostly in the empirical grounding. The abstract and summary give no equations, error analysis, or sensitivity checks, so it is hard to tell how much the rankings depend on specific network construction choices or data filtering. The assumption that these online interaction networks represent structural fragmentation without major platform or selection artifacts is stated but not probed in depth. Nothing in the argument looks circular or internally inconsistent. This paper is for researchers who work on political networks and need sharper tools for cross-context comparisons. A reader focused on measurement refinement would get practical value from the distinctions and the new metric. It deserves a serious referee because the core separation of concepts is useful and the EBF addresses a real gap, even if the methods section will need more detail and robustness tests.

Referee Report

0 major / 3 minor

Summary. The manuscript claims that standard network measures of political fragmentation such as modularity and the Krackhardt E-I index conflate two distinct aspects: the strength of separation between groups and the number of fragments. It identifies the pairwise adaptive E-I index and the effective number of communities (ENC) as isolating these aspects separately, introduces the effective branching factor (EBF) to capture multiscale hierarchical fragmentation, and reports that application to online interaction networks from Brazil, Spain, and the United States produces consistent country rankings across two datasets.

Significance. If the derivations and empirical results hold, the work clarifies three complementary dimensions of structural political fragmentation and supplies more interpretable metrics for cross-context comparison. The introduction of EBF for between-level branching and the empirical consistency check across datasets are positive features that could improve reproducibility and applicability in political network analysis.

minor comments (3)

[§3] §3 (Methods): the definition and computation of the pairwise adaptive E-I index and EBF should be accompanied by explicit equations or pseudocode to allow direct replication; the current presentation leaves the adaptation rule and branching calculation underspecified.
[§5] §5 (Empirical results): the country-ranking tables report consistent orderings but omit uncertainty estimates or statistical tests for the differences; adding bootstrap intervals or permutation tests would strengthen the consistency claim.
[Figure 3] Figure 3 (EBF illustration): the schematic of hierarchical splitting is helpful but the axis scales and community-size thresholds used to compute EBF are not labeled, reducing interpretability.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive summary of the manuscript, the recognition of its contributions, and the recommendation for minor revision. No specific major comments were listed in the report.

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper distinguishes strength of separation from number of fragments using comparisons of standard network metrics (modularity, Krackhardt E-I index) and proposes the pairwise adaptive E-I index plus effective number of communities (ENC) as better isolators of each aspect. The effective branching factor (EBF) is introduced as a new multiscale measure based on hierarchical splitting. None of these reduce by construction to fitted parameters, self-referential definitions, or load-bearing self-citations within the provided text; the empirical consistency check across Brazil/Spain/US datasets supplies external validation rather than circular support. The derivation chain therefore remains self-contained against external network-theory benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Metrics rest on the assumption that online interaction graphs encode political fragmentation; no free parameters or invented entities are described in the abstract.

axioms (1)

domain assumption Network representations of online interactions accurately reflect structural political fragmentation.
Invoked when applying metrics to empirical datasets from Brazil, Spain, and the United States.

pith-pipeline@v0.9.0 · 5540 in / 1153 out tokens · 21478 ms · 2026-05-16T16:15:46.205240+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

We compare several network measures and show that the two aspects of network fragmentation are best captured by the pairwise adaptive E-I index and the effective number of communities (ENC)
IndisputableMonolith/Foundation/ArithmeticFromLogic embed_strictMono_of_one_lt unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

we introduce a novel metric for multiscale fragmentation, the effective branching factor (EBF)

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

45 extracted references · 45 canonical work pages

[1]

L. A. Adamic and N. Glance. The political blogosphere and the 2004 us election: Divided they blog. InProceedings of the 3rd International Workshop on Link Discovery, 2005

work page 2004
[2]

Y . Y . Ahn, J. P. Bagrow, and S. Lehmann. Link communities reveal multiscale complexity in networks. Nature, 466(7307):761–764, 2010

work page 2010
[3]

Arnaudon, D

A. Arnaudon, D. J. Schindler, R. L. Peach, A. Gosztolai, M. Hodges, M. T. Schaub, and M. Bara- hona. Algorithm 1044: PyGenStability: A multiscale community detection framework with general- ized markov stability.ACM Transactions on Mathematical Software, 50(2):1–8, 2024

work page 2024
[4]

C. A. Bail, L. P. Argyle, T. W. Brown, J. P. Bumpus, H. Chen, M. B. Hunzaker, J. Lee, M. Mann, F. Merhout, and A. V olfovsky. Exposure to opposing views on social media can increase political polarization.Proceedings of the National Academy of Sciences, 115(37):9216–9221, 2018

work page 2018
[5]

Bakshy, S

E. Bakshy, S. Messing, and L. A. Adamic. Exposure to ideologically diverse news and opinion on facebook.Science, 348(6239):1130–1132, 2015

work page 2015
[6]

Barber ´a, J

P. Barber ´a, J. T. Jost, J. Nagler, J. A. Tucker, and R. Bonneau. Tweeting from left to right: Is online political communication more than an echo chamber?Psychological Science, 26(10):1531–1542, 2015

work page 2015
[7]

J. Bright. Explaining the emergence of political fragmentation on social media: The role of ideology and extremism.Journal of Computer-Mediated Communication, 23(1):17–33, 2018

work page 2018
[8]

Chao, C.-H

A. Chao, C.-H. Chiu, and L. Jost. Phylogenetic diversity measures and their decomposition: a frame- work based on hill numbers.Biodiversity Conservation and Phylogenetic Systematics, 14:141–172, 2016

work page 2016
[9]

T.-H. Y . Chen, A. Malkam¨aki, A. Faqeeh, E. Palosaari, A. Kotkaniemi, L. Funke, and T. Yl ¨a-Anttila. Climate policy elites’ twitter interactions across nine countries.arXiv preprint arXiv:2412.15545, 2024

work page arXiv 2024
[10]

Cinelli, E

M. Cinelli, E. Brugnoli, A. L. Schmidt, F. Zollo, W. Quattrociocchi, and A. Scala. Selective exposure shapes the facebook news diet.PLOS ONE, 15(3):e0229129, 2020

work page 2020
[11]

Clauset, C

A. Clauset, C. Moore, and M. E. Newman. Hierarchical structure and the prediction of missing links in networks.Nature, 453(7191):98–101, 2008

work page 2008
[12]

Conover, J

M. Conover, J. Ratkiewicz, M. Francisco, B. Gonc ¸alves, F. Menczer, and A. Flammini. Political polarization on twitter. InProceedings of the International AAAI Conference on Web and Social Media, volume 5, pages 89–96, 2011

work page 2011
[13]

Delvenne, S

J.-C. Delvenne, S. N. Yaliraki, and M. Barahona. Stability of graph communities across time scales. Proceedings of the National Academy of Sciences, 107(29):12755–12760, 2010

work page 2010
[14]

J. N. Druckman, S. Klar, Y . Krupnikov, M. Levendusky, and J. B. Ryan. Affective polarization, local contexts and public opinion in america.Nature Human Behaviour, 5(1):28–38, 2021

work page 2021
[15]

M. P. Fiorina and S. J. Abrams. Political polarization in the american public.Annual Review of Political Science, 11(1):563–588, 2008

work page 2008
[16]

M. P. Fiorina, S. J. Abrams, and J. C. Pope.Culture war? The myth of a polarized America. 2004

work page 2004
[17]

Flamino, A

J. Flamino, A. Galeazzi, S. Feldman, M. W. Macy, B. Cross, Z. Zhou, M. Serafino, A. Bovet, H. A. Makse, and B. K. Szymanski. Political polarization of news media and influencers on Twitter in the 2016 and 2020 US presidential elections.Nature Human Behaviour, 7(6):904–916, 2023. 10

work page 2016
[18]

Fortunato

S. Fortunato. Community detection in graphs.Physics Reports, 486(3–5):75–174, 2010

work page 2010
[19]

Garimella, G

K. Garimella, G. De Francisci Morales, A. Gionis, and M. Mathioudakis. Political discourse on social media: Echo chambers, gatekeepers, and the price of bipartisanship. InProceedings of the 2018 World Wide Web Conference, pages 913–922, 2018

work page 2018
[20]

Garimella, T

K. Garimella, T. Smith, R. Weiss, and R. West. Political polarization in online news consumption. InProceedings of the International AAAI Conference on Web and Social Media, volume 15, pages 152–162, 2021

work page 2021
[21]

G. V . Golosov. The effective number of parties: A new approach.Party Politics, 16(2):171–192, 2010

work page 2010
[22]

Goodwin, K

A. Goodwin, K. Joseff, M. J. Riedl, J. Lukito, and S. Woolley. Political relational influencers: The mo- bilization of social media influencers in the political arena.International Journal of Communication, 17:21, 2023

work page 2023
[23]

Hanretty

C. Hanretty. Party system polarization and the effective number of parties.Electoral Studies, 76:102459, 2022

work page 2022
[24]

Harteveld

E. Harteveld. Fragmented foes: Affective polarization in the multiparty context of the netherlands. Electoral Studies, 71:102332, 2021

work page 2021
[25]

Himelboim, M

I. Himelboim, M. Smith, and B. Shneiderman. Tweeting apart: Applying network analysis to detect selective exposure clusters in twitter.Communication Methods and Measures, 7(3-4):195–223, 2013

work page 2013
[26]

Huddy, L

L. Huddy, L. Mason, and L. Aarøe. Expressive partisanship: Campaign involvement, political emotion, and partisan identity.American Political Science Review, 109(1):1–17, 2015

work page 2015
[27]

Iyengar, Y

S. Iyengar, Y . Lelkes, M. Levendusky, N. Malhotra, and S. J. Westwood. The origins and consequences of affective polarization in the united states.Annual Review of Political Science, 22(1):129–146, 2019

work page 2019
[28]

Iyengar, G

S. Iyengar, G. Sood, and Y . Lelkes. Affect, not ideology: A social identity perspective on polarization. Public Opinion Quarterly, 76(3):405–431, 2012

work page 2012
[29]

effective

M. Laakso and R. Taagepera. “effective” number of parties: a measure with application to west europe. Comparative political studies, 12(1):3–27, 1979

work page 1979
[30]

Lambiotte, J

R. Lambiotte, J. C. Delvenne, and M. Barahona. Random walks, markov processes and the multiscale modular organization of complex networks.IEEE Transactions on Network Science and Engineering, 1(2):76–90, 2015

work page 2015
[31]

Lijphart.Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries

A. Lijphart.Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries. Yale University Press, 1999

work page 1999
[32]

Mart ´ın-Guti´errez, J

S. Mart ´ın-Guti´errez, J. C. Losada, and R. M. Benito. Multipolar social systems: Measuring polariza- tion beyond dichotomous contexts.Chaos, Solitons & Fractals, 169:113244, 2023

work page 2023
[33]

i disrespectfully agree

L. Mason. “i disrespectfully agree”: The differential effects of partisan sorting on social and issue polarization.American Journal of Political Science, 59(1):128–145, 2015

work page 2015
[34]

L. Mason. A cross-cutting calm: How social sorting drives affective polarization.Public Opinion Quarterly, 80(S1):351–377, 2016

work page 2016
[35]

Mason.Uncivil Agreement: How Politics Became Our Identity

L. Mason.Uncivil Agreement: How Politics Became Our Identity. University of Chicago Press, 2018

work page 2018
[36]

McCarty, K

N. McCarty, K. T. Poole, and H. Rosenthal.Polarized America: The Dance of Ideology and Unequal Riches. MIT Press, 2016

work page 2016
[37]

G. E. Noether. Sample size determination for some common nonparametric tests.Journal of the American Statistical Association, 82(398):645–647, 1987

work page 1987
[38]

S. C. Phillips, K. M. Carley, and K. Joseph. Why do people think liberals drink lattes? how social media afforded self-presentation can shape subjective social sorting.arXiv preprint arXiv:2404.02338, 2024

work page arXiv 2024
[39]

R. H. Pildes. The age of political fragmentation.Journal of Democracy, 32(4):146–159, 2021

work page 2021
[40]

D. N. Posner. The political salience of cultural difference: Why chewas and tumbukas are allies in zambia and adversaries in malawi.American Political Science Review, 98(4):529–545, 2004

work page 2004
[41]

C. R. M. A. Santagiustina, J.-P. Cointet, and P. R. Morales. Expressing one’s identity online: Left-right and cross eu-country variation in self-representation in social media.arXiv preprint arXiv:2501.05927, 2025. 11

work page arXiv 2025
[42]

F. B. Soares and R. Recuero. Hashtag wars: Political disinformation and discursive struggles on twitter conversations during the 2018 brazilian presidential campaign.Social Media + Society, 7(2):20563051211009073, 2021

work page 2018
[43]

F. B. Soares, R. Recuero, and G. Zago. Asymmetric polarization on twitter and the 2018 brazilian presidential elections. InProceedings of the 10th International Conference on Social Media and Society, pages 67–76, 2019

work page 2018
[44]

V . A. Traag, L. Waltman, and N. J. Van Eck. From louvain to leiden: guaranteeing well-connected communities.Scientific Reports, 9(1):1–12, 2019

work page 2019
[45]

V oltaLula,

S. ¨Unal, A. Metin, C. Anckar, and B. Tekin. Form of government, electoral system, and party system fragmentation: A global comparison.Asian Journal of Comparative Politics, 9(2):308–323, 2024. A Identification of Political Influencers Based on Users’ Public Self- presentation Political influencers on Twitter/X in this study were identified using data don...

work page 2024