Phase Transitions in Economic Inequality:Taxation and Extremal Replacement Dynamics
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Physics. The day in general physics: instruments, methods and results across the field.
Kai: We're looking at the paper: "Phase Transitions in Economic Inequality".
Paper summary: Kai: So looking at the conclusion of "Phase Transitions in Economic Inequality:Taxation and Extremal Replacement Dynamics," it boils down to this: the paper identifies generic mechanisms connecting taxation, redistribution, and inequality patterns <ref:2604.28168#pg1>.
Mira: They show how the emergence of never-poor agents provides a clear microscopic mechanism for persistent wealth concentration and its resistance to redistribution <ref:2604.28168#pg3>. This is a structural reason why inequality patterns persist even with dynamics at play <ref:2604.28168#pg3>.
Lev: From an engineering view, it means that if we want to build models of wealth dynamics, we need to account for these kind of emergent features—bistability and hysteresis—even when the underlying rules seem minimal <ref:2604.28168#pg3>.
Kai: The paper explores how taxation structure and extremal replacement dynamics interact to generate these qualitatively distinct inequality regimes in a way that is hard to see elsewhere <ref:2604.28168#pg3>.
Mira: It's a minimal setting that allows us to investigate how those specific interactions lead to persistent, non-trivial collective behavior in an economy <ref:2604.28168#pg3>.
Lev: The work is interesting because it shows that even minimal stochastic redistribution rules can produce things like non-ergodicity and abrupt changes in inequality <ref:2604.28168#pg3>.
Kai: That's the core finding. It sets up a framework for understanding why inequality might behave so differently depending on the tax system you have in place <ref:2604.28168#pg3>.
Conclusion: Kai: So, we've been looking at how this minimal agent-based model shows that even simple rules can create these really abrupt changes in inequality <ref:2604.28168#pg1>.
Mira: Yeah, this paper looks at how taxation and the way people replace the poorest agents together lead to two totally different long-term states for wealth distribution <ref:2604.28168#pg1>.
Lev: I'm looking at it from an engineering standpoint, and they found that these transitions are first-order like, meaning you don't just smoothly cross a line; there’s a jump between the two states <ref:2604.28168#pg1>.
Kai: They call this hysteresis, which means if you change the tax strength parameter slowly, the system can get stuck in one inequality level and only jump to another when you push it harder than you started <ref:2604.28168#pg1>.
Mira: The authors are focusing on how regressive taxation causes this big jump into a much lower inequality state once the tax strength parameter crosses a certain threshold <ref:2604.28168#pg1>.
Lev: They also showed that for progressive taxation, like when the tax rate is higher than one, you don't see that big abrupt change at all; it just evolves more smoothly <ref:2604.28168#pg1>.
Kai: The main implication is that these minimal stochastic redistribution rules alone are enough to generate things like non-ergodicity and bistability in a wealth system <ref:2604.28168#pg3>.
Mira: It's showing how the structure of the tax system interacts with extremal dynamics to produce these qualitatively different inequality regimes <ref:2604.28168#pg3>.
Lev: It’s a minimal setting, which is good because it means we can isolate exactly what taxation and replacement dynamics do without all the noise from complex market interactions <ref:2604.28168#pg3>.
Kai: They point out that this happens because of the formation of never-poor agents, which are basically agents who permanently escape replacement and keep wealth concentrated <ref:2604.28168#pg3>.
Mira: It gives us a microscopic way to understand why inequality patterns can be so persistent and resistant to redistribution in these models <ref:2604.28168#pg3>.
Lev: So, the paper is basically saying that even very simple rules for moving money around can create rich collective dynamics, including things like abrupt changes in wealth inequality <ref:2604.28168#pg3>.
Kai: It sets up a framework where we can study how the structure of taxation and how agents are replaced together interact to generate these distinct inequality patterns <ref:2604.28168#pg3>.
Mira: This minimal model is useful because it lets us see exactly how those interactions lead to persistent, non-trivial collective behavior in an economy <ref:2604.28168#pg3>.
Lev: We can use this kind of framework to look at other systems where simple redistribution rules might be at play, like in some types of asset exchange models <ref:2604.28168#pg3>.
Statistical and Interdisciplinary Physics Group, Centro AtⒶmico Bariloche (CNEA) · Instituto Balseiro, Universidad Nacional de Cuyo · Instituto de Física, Universidade Federal do Rio Grande do Sul · Instituto Nacional de Ciçncia e Tecnologia de Sistemas Complexos, CBPF
physics.soc-ph
Submitted: 2026-04-30
Updated: 2026-08-31
Journal ref: Phys. Rev. E 114 (2026) 054310
DOI: 10.1103/jrnm-3f84
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Importance score: 91/100
The gist: Phase transitions in economic inequality are studied using a minimal agent-based model to investigate how taxation and extremal replacement dynamics lead to abrupt transitions between distinct…
Key concepts
- Tax Collection Step
- At each time step, a total tax T is calculated based on the average wealth of the system. This tax is then collected from agents with positive wealth. The amount of tax collected scales with the system's total wealth and a strength parameter, directly influencing redistribution.
- Wealth Replacement Step
- The poorest agent in the system has their entire wealth replaced by a new value drawn randomly from a uniform distribution between zero and twice the collected tax. This mechanism drives extremal dynamics, constantly pushing wealth towards the poor.
- Phase Transition
- This refers to an abrupt, discontinuous change in the system's behavior when a parameter (like tax strength) crosses a critical threshold. In this model, it means moving suddenly from one stable state of inequality (e.g., very unequal) to another stable state (e.g., more homogeneous).
Terminology
Summary
Phase transitions in economic inequality are studied using a minimal agent-based model to investigate how taxation and extremal replacement dynamics lead to abrupt transitions between distinct stationary regimes The system self-organizes into two distinct stationary phases when changing the total tax collected, characterized by a non-ergodic, high-inequality regime and a more homogeneous ergodic phase This research demonstrates that minimal stochastic redistribution mechanisms alone can produce discontinuous transitions, metastability, and non-ergodicity through the interplay between taxation and extremal replacement dynamics in a non-conservative setting
Model Description
The system is composed of N agents characterized by their wealth w, initially drawn from a uniform distribution in the range [0, 1), and subsequently rescaled so that the total wealth satisfies W = PN i=1 wi = 1 The system evolves in discrete time through two sequential processes: a tax collection step and a wealth replacement step
A. Tax collection step
At each timestep, a total tax T = γ⟨w⟩2 (1) is collected from all agents with positive wealth, where γ > 0 is called the tax strength parameter and ⟨w⟩ = 1/N∑i=1 wi = W N (2) is the mean wealth of the system, proportional to the total wealth W The magnitude of the tax depends on the strength parameter and scales with the system’s wealth
B. Wealth replacement step
Following the principles of extremal dynamics, the poorest agent in the system, with wealth wp(t), is selected Its wealth will be replaced by a new value drawn from a uniform distribution in the interval [0, 2T), wp(t + 1) = U[0, 2T) = U[0, γ⟨w⟩) (4) The parameter γ controls the strength of redistribution
Characterization of Dynamics
The system reaches a stationary state, as defined by the Gini-based criterion introduced in the previous section, after a transient whose duration depends on the parameters After the system reaches a steady state, the dependence of inequality on the model parameters can be explored
-
For regressive taxation (α < 1), the Gini index exhibits a discontinuous jump from a high-inequality state to a much lower value once γ exceeds a parameter-dependent threshold
-
For progressive taxation (α > 1), the Gini index remains close to zero up to γ ≈ 1, reaching almost perfect equality in this regime
-
For γ > 1, the Gini index displays a slow but monotonic increase with γ, reflecting increasing fluctuations in individual wealth
Mean-Field Theory and Spinodals
The mean-field approximation replaces stochastic quantities by their ensemble averages to derive a deterministic drift equation for the rescaled wealth ui The drift f(u) has a nontrivial fixed point f(u∗) = 0 at uα−1∗ = 2 µ Mα γ (13) This bifurcation at α = 1 provides a dynamical interpretation of the three qualitative regimes identified in Sec. III A
The mean-field solution clarifies structural features such as the lower edge ua of the support, which emerges from self-consistency conditions (17)–(19) The upper edge γ of the support is a direct consequence of the injection mechanism, since replaced agents are drawn uniformly from [0, γ)
Evidence of Phase Transition
The abrupt change in the Gini index for regressive taxes suggests a phase transition between two distinct stationary regimes A hysteresis experiment shows that the transition from the high- to the low-inequality phase and back does not occur at the same γ value, indicating a first-order-like phase transition
The width of the hysteresis loop quantifies the range of γ over which both states remain dynamically stable, and this area is non-monotonic with respect to α The spinodal analysis recovers both the existence of two coexisting states and the nonmonotonic dependence of their persistence on the taxation structure
Discussion
The results demonstrate that even minimal stochastic redistribution rules can generate rich collective dynamics, including non-ergodicity, bistability, hysteresis, and abrupt changes in inequality Under regressive taxation (α < 1), a non-ergodic regime emerges for small γ due to the formation of never-poor agents that permanently escape replacement
Under progressive taxation (α ≥ 1), no abrupt transition is observed as wealth distributions remain comparatively homogeneous, consistent with continuous evolution or crossover between regimes The mean-field theory ties together stationary distributions, the location and order of the transition, and the resilience of wealth concentration
Conclusion
The present framework identifies generic mechanisms linking taxation structure, redistribution, and collective inequality patterns The emergence of never-poor agents provides a clear microscopic mechanism for persistent wealth concentration and its resistance to redistribution This framework offers a minimal setting in which to investigate how taxation structure and extremal replacement dynamics interact to generate qualitatively distinct inequality regimes
Acknowledgments
L.G. acknowledges support from ACAL through a mobility fellowship that enabled a research visit to UFRGS SG and JRI acknowledge financial support from CNPq through projects 309560/2025-2, 304101/2020-6, and 406820/25- (Note: The reference list provided in the text contains multiple entries for acknowledgments; this token points to the general acknowledgment section).
Appendix A: Finite-size analysis
The curves collapse across system sizes within error bars, and the location of the transition remains unchanged as N increases The magnitude of the jump in the Gini index remains stable as N increases, suggesting that the observed bistability persists in the thermodynamic limit
Appendix B: Bistability and dependence on initial conditions
The system exhibits dynamical bistability for a fixed value of γ = 1.75 and α = 0, where two stationary states are accessible depending on initial microscopic fluctuations
Microscopic structure of the bistable stationary states
The microscopic differences between the two dynamically stable stationary states are illustrated by comparing the stationary wealth distributions obtained for γ = 1.75 and α = 0, when approached from below or above in the hysteresis cycle Along the increasing-γ branch, the distribution is more dispersed, while along the decreasing-γ branch, it is more concentrated around a typical wealth scale
Appendix C: Saddle-node theory of the upper spinodal
The upper spinodal γ upc(N) is defined as the largest γ for which a condensed configuration is dynamically self-sustained The finite-N upper spinodal is non-monotonic, rising, reaching a maximum near α ≈ 0.34, and then falling This provides a direct, quantitative explanation for the nonmonotonic hysteresis area of Fig. p. 9 10
Appendix D: Large-N limit
As N → ∞ the tension dissolves, because a vanishing share can still carry a diverging wealth, leading to the condition µ(γ) = 0 This limit recovers γ upc(∞)(α) = 3 − α/(1 − α) The finite-N upper spinodal is below γ upc(∞), and the downward correction increases with α, producing the nonmonotonic γ upc(α) of Fig. 13
References
[1] J. Rawls, A Theory of Justice: Original Edition (Harvard University Press, 1971)
[2] T. Piketty, Capital in the Twenty-First Century (Harvard University Press, 2014)
[3] V. Pareto, Cours d’Economie Politique (F. Rouge, Lausanne, 1897)
[4] V. M. Yakovenko and J. p. 1 B. J. Rosser, Colloquium: Statistical mechanics of money, Reviews of Modern Physics 81, 1703 (2009)
[5] B. K. Chakrabarti, A Chakraborti, S R Chakravarty, and A Chatterjee, Econophysics of Income and Wealth Distributions (Cambridge University Press, Cambridge UK, 2013)
[6] R. N. Mantegna and H. E. Stanley, An Introduction to Econophysics: Correlations and Complexity in Finance (Cambridge University Press, 2000)
[7] T. Lux, Applications of statistical physics in finance and economics, The European Physical Journal Special Topics 161, 181 (2008)
[8] J. Angle, The surplus theory of social stratification and the size distribution of personal wealth, Social Forces 65, 293 (1986)
[9] J. Angle, The inequality process and the distribution of income to blacks and whites, Journal of Mathematical Sociology 17, 77 (1992)
[10] A. Dr˘agulescu and V. M. Yakovenko, Statistical mechanics of money, European Physical Journal B 17, 723 (2000)
[11] A. Chakraborti and B Chakrabarti, Statistical mechanics of money: How saving propensity affects its distribution, The European Physical Journal B - Condensed Matter and Complex Systems 17 (2000)
[12] A. Chatterjee, B K Chakrabarti, and S S Manna, Pareto law in a kinetic model of market with random saving propensity, Physica A 335, 155 (2004)
[13] J.-P. Bouchaud and M M´ezard, Wealth condensation in a simple model of economy, Physica A 282, 536 (2000)
[14] J. R. Iglesias, S Gon¸calves, G Abramson, and J L Vega, Correlation between risk aversion and wealth distribution, Physica A: Statistical Mechanics and its Applications 342, 186 (2004)
[15] F. Slanina, Inelastically scattering particles and wealth distribution in an open economy, Physical Review E 69, 046102 (2004)
[16] S Pianegonda, J Iglesias, G Abramson, and J Vega, Wealth redistribution with conservative exchanges, Physica A: Statistical Mechanics and its Applications 322, 667 (2003)
[17] B. M. Boghosian, A Devitt-Lee, M Johnson, J Li, J A Marcq, and H Wang, Oligarchy as a phase transition: The effect of wealth-attained advantage in a fokker–planck description of asset exchange, Physica A: Statistical Mechanics and its Applications 476, 15 (2017)
[18] K. K. L. Liu, N Lubbers, W Klein, J Tobochnik, B M Boghosian, and H Gould, Simulation of a generalized asset exchange model with economic growth and wealth distribution Physical Review E 104, 014150 (2021)
[19] P. Bak and K Sneppen, Punctuated equilibrium and criticality in a simple model of evolution, Phys. p. 0 Rev. Lett. 71, 4083 (1993)
[20] J R Iglesias, S Gon¸calves, S Pianegonda, G Abramson, and J L Vega, Wealth redistribution in our small world Physica A 327, 12 (2003)
[21] M Bisi, G Spiga, and G Toscani, Kinetic models of conservative economies with wealth redistribution Communications in Mathematical Sciences - COMMUN MATH SCI 7 (2009)
[23] J.-P. Bouchaud, On growth-optimal tax rates and the issue of wealth inequalities Journal of Statistical Mechanics: Theory and Experiment 2015, P11011 (2015)
[24] T Dias and S Gon¸calves, Effectiveness of wealth-based vs exchange-based tax systems in reducing inequality, Physica A: Statistical Mechanics and its Applications 641, 129721 (2024)
[25] B Hayes, Follow the money, American Scientist 90, 400 (2002)
[35] N C Kakwani, Measurement of tax progressivity: An international comparison, The Economic Journal 87, 71 (1977)
[36] K Binder, Theory of first-order phase transitions, Reports on Progress in Physics 50, 783 (1987)
[37] K Binder, Finite size scaling analysis of ising model block distribution functions, Physical Review Letters 47, 693 (1981)
As N → ∞ the tension dissolves, because a vanishing share can still carry a diverging wealth, leading to the condition µ(γ) = 0 This limit recovers γ upc(∞)(α) = 3 − α/(1 − α) The finite-N upper spinodal is below γ upc(∞), and the downward correction increases with α, producing the nonmonotonic γ upc(α) of Fig.
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
-
Automated Identification of Inequality Regimes: The improved AI system can accurately classify economic systems into distinct stationary regimes, such as
non-ergodic, high-inequality regime
andmore homogeneous ergodic phase,
by monitoring indicators like the Gini index and Top 1% wealth share. -
Discontinuous Transition Detection: The system can detect abrupt changes in inequality by identifying the discontinuous jump in the Gini index when regressive taxation is varied, a feature characteristic of first-order-like transitions.
-
Hysteresis Quantification: The AI can quantify the resilience of inequality by measuring the hysteresis area and its dependence on progressivity, which reflects
the persistence or resilience of the unequal regime
and identifies bistable regions based on control parameter values. -
Microscopic Structure Mapping: The system can visualize the internal structure of coexisting stationary states by analyzing complementary observables like Shannon entropy and Binder cumulant, distinguishing between states based on their
concentration
andextreme wealth fluctuations.
-
Spinodal Boundary Prediction: The AI can predict the limits of stability by locating the upper and lower spinodals, which define the bistable region where two stationary states coexist, providing a quantitative explanation for observed hysteresis in terms of mean-field theory.
Related papers
- Transient Signal Propagation in Heterogeneous Networks under Various Inputs
- New rules, old habits: across seven team sports, each league keeps its own way of using the rules
- URBAN-SPIN: A street-level bikeability index to inform design implementations in historical city centres
- Logarithmic dispersion at marginal stability in collective attention
- Glassy dynamics of metropolitan human mobility frozen near the gravitational equilibrium
- Two preregistered datasets on production-type labelling in the ENTSO-E Transparency Platform: multi-fuel reporting, register vintages, persistent misclassification, and cross-product consistency