🌀 nonlinear sciences

Corruption as a self-sustained collective state in political systems

This paper proposes a minimal compartmental model demonstrating that systemic political corruption can emerge as a self-sustaining, dynamically stable state driven by reinforcing interactions between power concentration and relational structures, thereby explaining its persistence across electoral cycles without requiring centralized coordination.

Original authors: Nuno Crokidakis, Jaime L. C. da C. Filho

Published 2026-07-23

📖 7 min read🧠 Deep dive

CC0 1.0

Original authors: Nuno Crokidakis, Jaime L. C. da C. Filho

Original paper dedicated to the public domain under CC0 1.0 (http://creativecommons.org/publicdomain/zero/1.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where social scientists and physicists are having a late-night conversation over coffee, trying to figure out why some things just won't go away. They aren't talking about a stubborn stain on a carpet, but something much bigger: why do bad habits, like political corruption, seem to stick around even when we try to scrub them out? This field of study is called sociophysics. It's a bit like using the rules of physics—the study of how particles move and interact—to understand how people behave in groups. In this corner of science, researchers treat society like a giant, complex machine where individual choices (microscopic interactions) can add up to create huge, predictable patterns (macroscopic behavior), much like how water molecules coming together create a wave. The big question here is: Is corruption just a bunch of bad apples acting alone, or is it a self-sustaining system that keeps feeding itself? Understanding this matters because if we think it's just "bad apples," we might keep trying to pull them out one by one, only to find the tree is still rotten. But if it's a system, we might need to change the soil, not just the fruit.

Now, let's dive into what Nuno Crokidakis and Jaime L. C. da C. Filho actually did in their new paper. They didn't go out and interview politicians or count bribes. Instead, they built a tiny, simplified mathematical model—a sort of "toy universe" made of equations—to see how corruption behaves when it interacts with the networks of friends and allies that protect it. Think of it like a video game simulation where you have two main characters: Corruption (let's call him "C") and Support Structures (let's call her "S"). "C" is the act of being corrupt, and "S" is the web of connections—like secret handshakes, favors, and alliances—that helps "C" hide and grow.

The authors set up a race between two forces. On one side, you have dissipation, which is like the "cleaning crew" of society: laws, investigations, and public outrage that try to wash corruption away. On the other side, you have reinforcement, where corruption helps build more support structures, and those structures, in turn, make it easier for corruption to survive and spread. The paper suggests that these two forces are locked in a tug-of-war. If the cleaning crew is strong enough, corruption fades away. But if the support structures get too good at protecting the corruption, something surprising happens: the system flips into a new state where corruption becomes "self-sustaining." It's like a campfire that, once it gets hot enough, starts generating its own wind to keep burning, making it impossible to blow out with a simple breath.

The researchers found that there is a specific "tipping point" or threshold in this tug-of-war. They call this the critical threshold. Below this line, the cleaning crew wins, and the system stays "clean." But once the reinforcing interactions get strong enough to cross that line, the system flips into a "captured state." In this state, corruption isn't just a few isolated mistakes; it becomes a stable, collective condition. The paper suggests that once a system enters this captured state, it becomes incredibly hard to change. Even if you throw a big institutional crisis at it—like arresting a few leaders or changing the government—the system naturally relaxes back to its corrupt state. It's as if the system has a memory and a muscle that pulls it back to the same spot, no matter how hard you push it away.

The authors explicitly argue against the idea that corruption is just the result of isolated individuals making bad choices or that it's simply a failure of specific institutions. Instead, their model suggests that corruption can emerge as a collective phenomenon driven by the feedback loop between the bad acts and the protective networks. They don't claim to have solved the problem of corruption or proven that every corrupt system works exactly this way. Rather, their simulations and mathematical analysis suggest that this "self-sustaining" mechanism is a plausible explanation for why corruption persists across many election cycles and political crises.

To make this even clearer, imagine a playground where kids are playing a game. If a few kids start cheating, and the teacher (the cleaning crew) catches them, they stop. That's the "clean" state. But what if the cheating kids start forming a secret club? They share snacks to keep each other quiet, and they make a rule that anyone who snitches gets kicked out of the club. Now, the cheating isn't just about one kid; it's about the whole club protecting itself. The teacher might catch one kid, but the club just recruits a new one, and the cheating continues. The paper suggests that in real politics, this "club" is the relational structure (alliances, patronage, protection deals). The model shows that if the club gets strong enough, the cheating becomes the normal way the game is played, and the system becomes "captured."

The paper also introduces a concept called the "effective reproduction number," which sounds like something from a biology class about viruses, but here it's about corruption. It's a simple ratio that tells you if corruption is going to die out or take over. If the number is less than 1, the cleaning crew wins. If it's greater than 1, the corruption wins and becomes self-sustaining. The authors point out that this isn't just about how many corrupt people there are, but about how well they are connected and protected.

One of the most interesting parts of their findings is the idea of "phase transitions." In physics, this is like water turning into ice. You can cool water down a little, and it's still liquid. But once it hits a specific temperature, it suddenly freezes. The authors suggest that political systems might work the same way. You can have a little bit of corruption here and there, but once the support structures get strong enough to cross that critical threshold, the whole system suddenly shifts into a "captured" phase. The paper notes that once you are in this captured phase, small changes—like firing one official or passing one new law—won't fix things. The system is too deep in the "captured" zone. To fix it, you'd need to fundamentally change the rules of the game to weaken the support structures, not just the individuals.

The authors use the political dynamics of Rio de Janeiro, Brazil, as a qualitative example to illustrate their point. They mention that despite repeated arrests, removals, and political turnover, the underlying structure of corruption seemed to remain intact. They aren't saying their model perfectly describes Rio, but rather that the pattern of "leaders change, but the system stays the same" fits the behavior of a system deep inside the captured phase. It's like trying to fix a broken toy by changing the batteries when the gears are actually jammed; you need to fix the gears (the structures) to make it work.

In the end, this paper offers a fresh perspective. It suggests that the persistence of corruption might not be a mystery of human nature, but a predictable outcome of how systems work when they are reinforced by their own internal structures. It's a reminder that sometimes, to fix a problem, you can't just fight the symptom; you have to break the cycle that keeps the symptom alive. The authors suggest that future work could add more complexity, like how people adapt or how random events play a role, but for now, this simple model gives us a powerful new way to think about why some systems just won't let go of the bad stuff.

Technical Summary: Corruption as a Self-Sustained Collective State in Political Systems

Problem Statement
Traditional interpretations of political corruption often attribute it to individual misconduct, isolated institutional failures, or rational choice incentives. However, the persistence of corruption across successive electoral cycles and institutional crises suggests a more complex phenomenon: systemic corruption as a self-sustaining governance arrangement. The paper addresses the lack of formal frameworks describing how corruption emerges as a collective, dynamical regime driven by feedback between interaction structures and macroscopic social variables. Specifically, it seeks to explain how corruption-supporting relational structures (e.g., patronage networks, clientelistic ties) and corruption itself mutually reinforce one another to create a stable, "captured" state that resists standard institutional dissipation.

Methodology
The authors propose a minimal nonequilibrium dynamical model based on a system of coupled ordinary differential equations (ODEs). The model treats the political system as a coarse-grained complex system where microscopic interactions are aggregated into macroscopic variables.

The core of the model consists of three variables normalized to the interval $[0, 1]$ :

$C(t)$ : The level of systemic political corruption, interpreted as an order parameter measuring the degree of institutional embedding of corrupt practices.
$A(t)$ : The density of corruption-supporting relational structures (alliances, protection arrangements, reciprocal support networks).
$P(t)$ : An emergent macroscopic observable representing the concentration or stability of political power.

The dynamics are governed by the following coupled equations:

$\frac{dC}{dt} = \alpha A(1 - C) - \beta C$
$\frac{dA}{dt} = \gamma C(1 - A) - \delta A$
$\frac{dP}{dt} = \eta C - \mu P$

Here, $\alpha$ and $\gamma$ represent the strengths of reinforcing interactions (corruption fostering structures, and structures fostering corruption), while $\beta$ and $\delta$ represent effective institutional dissipation rates (enforcement, oversight, political competition). The variable $P(t)$ is treated as a consequence of the $C-A$ dynamics rather than an independent driver, allowing the system to be reduced to a two-dimensional feedback loop between corruption and support structures.

The analysis employs standard techniques from nonlinear dynamics and statistical physics, including:

Derivation of stationary states (fixed points).
Linear stability analysis via the Jacobian matrix.
Global stability analysis using the Bendixson-Dulac criterion and the Poincaré-Bendixson theorem to rule out periodic orbits.
Definition of an effective reproduction number ( $R_0$ ) to characterize the phase transition.

Key Contributions and Results
The paper identifies a nontrivial phase transition separating two distinct dynamical regimes: a "clean" state and a "captured" state.

Stationary States: The system admits two qualitatively distinct stationary solutions:
- Clean State: $C^* = 0, A^* = 0, P^* = 0$ . This represents the absence of self-sustained corruption.
- Captured State: $C^* > 0, A^* > 0, P^* > 0$ . This represents a regime where corruption and its supporting structures coexist and reinforce each other.
Critical Threshold and Phase Transition: The existence of the captured state is governed by a critical threshold defined by the parameters. The transition occurs when the reinforcing interactions overcome the dissipative mechanisms. The authors define an effective reproduction number:
$R_0 = \frac{\alpha \gamma}{\beta \delta}$
- If $R_0 < 1$ (or $\alpha < \alpha_c = \frac{\beta \delta}{\gamma}$ ), the clean state is stable, and corruption decays.
- If $R_0 > 1$ , the clean state becomes unstable, and the system evolves toward the captured state.
Stability and Dynamics:
- Local Stability: Linear stability analysis confirms that the captured state is a stable attractor when $R_0 > 1$ . Small perturbations to the macroscopic variables naturally relax back to the stationary state.
- Global Stability: Using the Bendixson-Dulac criterion, the authors prove that no periodic orbits exist in the physically admissible phase space. Consequently, all trajectories asymptotically approach the unique stable equilibrium (either the clean or captured state), ruling out persistent oscillatory dynamics.
- Scaling: Near the critical point, the stationary corruption level scales linearly with the distance from the threshold ( $C^* \sim R_0 - 1$ ), indicating a continuous phase transition with a critical exponent $\beta' = 1$ .
Phase Diagram: The results are summarized in a phase diagram in the $\alpha$ vs. $\delta$ plane. Systems deep within the captured phase ( $R_0 \gg 1$ ) exhibit high resilience; macroscopic perturbations (such as anti-corruption operations or leadership changes) that temporarily reduce corruption levels do not alter the underlying interaction structure, causing the system to rapidly relax back to the captured state.

Significance and Claims
The paper argues that the persistence of corruption is not merely a result of individual strategic choices but emerges spontaneously from the reinforcing feedback between systemic corruption and the relational structures that sustain it.

Mechanism of Persistence: The model provides a mechanism explaining why corruption persists across electoral cycles and institutional crises. Once a system enters the captured regime ( $R_0 > 1$ ), the corruption-supporting infrastructure acts as a self-sustaining governance arrangement. Institutional interventions that fail to alter the fundamental balance between reinforcement ( $\alpha, \gamma$ ) and dissipation ( $\beta, \delta$ ) are insufficient to shift the system out of the captured phase.
Qualitative Illustration: The authors cite the political dynamics of the state of Rio de Janeiro, Brazil, as a qualitative example. They note that repeated arrests, impeachments, and leadership turnovers have not led to a structural reduction in corruption, consistent with a system operating deep within the captured phase where the relational infrastructure remains intact despite changes in personnel.
Theoretical Bridge: The work establishes a conceptual bridge between political science, statistical physics, and nonlinear dynamics. It suggests that long-term political capture can emerge without centralized coordination or complex strategic behavior, driven instead by simple reinforcing interactions.
Modesty of Claims: The authors explicitly state that the model is a "minimal" and "coarse-grained" description. They do not claim to provide a quantitative fit to specific real-world data or to distinguish between specific types of actors (e.g., politicians vs. criminal groups). Instead, the framework serves as a conceptual tool to understand how macroscopic patterns of corruption can emerge from microscopic interaction rules, highlighting that robust collective structures can arise from simple dynamical ingredients.

Technical Summary: Corruption as a Self-Sustained Collective State in Political Systems

More like this