pith. sign in

arxiv: 2310.15861 · v2 · submitted 2023-10-24 · 💰 econ.TH

Context-based Imitation and the Evolution of Behavioral Rules

Enrique Urbano Arellano , Xinyang Wang This is my paper

Pith reviewed 2026-05-24 06:54 UTC · model grok-4.3

classification 💰 econ.TH
keywords imitation dynamicsbehavioral rulesCondorcet winnersocial choicebelief evolutioncontext-dependent behaviorenvy minimization
0
0 comments X

The pith

Multiple contexts turn imitation dynamics into a weighted social choice problem where consensus occurs exactly when some rule is a Condorcet winner.

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

The paper examines agents who copy behavioral rules from better-performing peers within one context and then apply the copied rule across all contexts. This modeling choice converts standard discrete-time imitation into an equivalent context-weighted social choice problem. The population reaches consensus on a single rule if and only if that rule beats every alternative in pairwise comparisons under the context weights; without such a winner, persistent non-convergence or cycles can occur. Among protocols that imitate within contexts, imitate-if-better is the unique rule that minimizes envy. The same framework yields conditions under which imitation produces rational expectations and conditions under which beliefs and consumption keep fluctuating in unchanging environments.

Core claim

Agents copy rules used by better-performing peers in the same context and apply them across contexts. This turns the dynamics into a context-weighted social choice problem. The population converges to consensus if and only if some rule is a Condorcet winner; otherwise, persistent non-convergence can occur. Among same-context imitation protocols, imitate-if-better uniquely minimizes envy. The framework provides a new account of belief evolution, characterizing when imitation selects rational expectations and showing how persistent belief and consumption fluctuations can arise in stationary environments.

What carries the argument

Context-weighted social choice equivalence produced by same-context imitation followed by cross-context application of the copied rule.

If this is right

  • Consensus on a rule occurs if and only if that rule is a Condorcet winner under the context weights.
  • Imitate-if-better is the unique same-context protocol that minimizes envy.
  • Imitation selects rational expectations precisely under the conditions identified by the equivalence.
  • Persistent fluctuations in beliefs and consumption arise even when the underlying environment is stationary.

Where Pith is reading between the lines

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

  • Social choice theorems on Condorcet winners and cycles can be imported directly to predict stability of behavioral rules under imitation.
  • Laboratory experiments that vary the number and frequency of contexts could test whether observed rule adoption tracks the predicted Condorcet condition.
  • The framework suggests examining how changes in context frequencies alter the speed or likelihood of convergence without altering the qualitative result.

Load-bearing premise

Agents copy rules used by better-performing peers in the same context and apply those rules across different contexts.

What would settle it

A simulation or observation in which the population converges to consensus despite the absence of any Condorcet winner under the context weights, or fails to converge despite the presence of one.

Figures

Figures reproduced from arXiv: 2310.15861 by Enrique Urbano Arellano, Xinyang Wang.

Figure 1
Figure 1. Figure 1: Single-period learning from an individual’s perspective [PITH_FULL_IMAGE:figures/full_fig_p009_1.png] view at source ↗
read the original abstract

We study the evolution of behavioral rules in environments with multiple contexts. Agents copy rules used by better-performing peers in the same context and apply them across contexts. Multiple contexts turn discrete-time imitation dynamics into a context-weighted social choice problem: the population converges to consensus if and only if some rule is a Condorcet winner; otherwise, persistent non-convergence can occur. Among same-context imitation protocols, imitate-if-better uniquely minimizes envy. The framework provides a new account of belief evolution, characterizing when imitation selects rational expectations and showing how persistent belief and consumption fluctuations can arise in stationary environments.

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.

Referee Report

2 major / 1 minor

Summary. The manuscript studies the evolution of behavioral rules in environments with multiple contexts. Agents copy rules used by better-performing peers in the same context and apply them across contexts. It claims that this converts discrete-time imitation dynamics into a context-weighted social choice problem: the population converges to consensus if and only if some rule is a Condorcet winner; otherwise persistent non-convergence can occur. Among same-context imitation protocols, imitate-if-better uniquely minimizes envy. The framework also characterizes when imitation selects rational expectations and how persistent belief and consumption fluctuations can arise in stationary environments.

Significance. If the central equivalence and characterizations hold, the paper provides a parameter-free link between imitation dynamics and social choice theory, offering a new microfoundation for consensus formation, non-convergence, and belief evolution. The envy-minimization result and the account of rational expectations selection are notable strengths that could influence work on evolutionary game theory and behavioral economics.

major comments (2)
  1. [Abstract] Abstract: The if-and-only-if claim that convergence occurs precisely when a Condorcet winner exists is load-bearing for the central contribution, but the abstract presents it without reference to the supporting theorem, model definition, or derivation; the cross-context application of copied rules is the key modeling step that produces the social-choice structure and requires explicit verification.
  2. [Main results (inferred from abstract claims)] The manuscript states the equivalence between the imitation process and context-weighted social choice as following directly from the rule-copying mechanism, but without the full model setup, payoff definitions, or proof steps visible, it is not possible to confirm the absence of gaps in the derivation of the consensus condition.
minor comments (1)
  1. [Abstract] The abstract would benefit from a one-sentence statement of the basic agent and context setup before stating the equivalence result.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their thoughtful report and positive assessment of the paper's potential contributions to evolutionary game theory and behavioral economics. We address each major comment below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The if-and-only-if claim that convergence occurs precisely when a Condorcet winner exists is load-bearing for the central contribution, but the abstract presents it without reference to the supporting theorem, model definition, or derivation; the cross-context application of copied rules is the key modeling step that produces the social-choice structure and requires explicit verification.

    Authors: We agree that the abstract would benefit from a reference to the supporting result. We will revise the abstract to include a parenthetical citation to the main theorem establishing the if-and-only-if equivalence. The cross-context application of copied rules is defined explicitly in Section 2: agents observe and compare performance only within the realized context but adopt the copied rule for use in all contexts. This modeling choice is what induces the context-weighted social choice structure, which is then formally derived in Section 3. revision: yes

  2. Referee: [Main results (inferred from abstract claims)] The manuscript states the equivalence between the imitation process and context-weighted social choice as following directly from the rule-copying mechanism, but without the full model setup, payoff definitions, or proof steps visible, it is not possible to confirm the absence of gaps in the derivation of the consensus condition.

    Authors: The complete model setup—including the finite set of contexts, the space of behavioral rules, context-dependent payoff functions, and the imitate-if-better protocol—is given in Section 2. Section 3 contains the full derivation: the multi-context dynamics are shown to aggregate into a weighted tournament whose Condorcet winner (if it exists) is the unique asymptotically stable consensus state. The key steps are (i) context-by-context performance comparison, (ii) frequency-weighted aggregation across contexts, and (iii) equivalence to the Condorcet criterion. To improve visibility we will insert a short proof outline in the introduction, but the existing sections already supply the requested elements. revision: partial

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper defines imitation as agents copying rules from better-performing peers within the same context and applying copied rules across contexts; the claimed equivalence to a context-weighted social choice problem (convergence iff Condorcet winner) is presented as a direct consequence of this modeling choice rather than a separate derivation. No equations or steps reduce by construction to fitted inputs, self-citations, or renamed empirical patterns; the central claims rest on the explicit imitation protocol plus standard social choice concepts without circular self-reference or load-bearing self-citation chains.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The model rests on domain assumptions about copying behavior and cross-context application; no free parameters or invented entities are mentioned in the abstract.

axioms (2)
  • domain assumption Agents copy rules used by better-performing peers in the same context
    Core process stated in the abstract that enables the social choice framing.
  • domain assumption Copied rules are applied across contexts
    This assumption converts single-context imitation into the weighted social choice problem.

pith-pipeline@v0.9.0 · 5616 in / 1282 out tokens · 25186 ms · 2026-05-24T06:54:09.259983+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

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

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

40 extracted references · 40 canonical work pages

  1. [1]

    Acemoglu and A

    D. Acemoglu and A. Ozdaglar. Opinion dynamics and learning in social networks. Dynamic Games and Applications, 1: 0 3--49, 2011

    work page 2011

  2. [2]

    Alos-Ferrer and K

    C. Alos-Ferrer and K. H. Schlag. Imitation and learning. In The handbook of rational and social choice. Oxford University Press, 2009

    work page 2009

  3. [3]

    R. J. Aumann. A synthesis of behavioural and mainstream economics. Nature human behaviour, 3 0 (7): 0 666--670, 2019

    work page 2019

  4. [4]

    A. V. Banerjee. A simple model of herd behavior, 107q. J. ECON, 797: 0 798, 1992

    work page 1992

  5. [5]

    W. M. Baum. Rules, culture, and fitness. The Behavior Analyst, 18: 0 1--21, 1995

    work page 1995

  6. [6]

    W. M. Baum, P. J. Richerson, C. M. Efferson, and B. M. Paciotti. Cultural evolution in laboratory microsocieties including traditions of rule giving and rule following. Evolution and Human Behavior, 25 0 (5): 0 305--326, 2004

    work page 2004

  7. [7]

    Blume and D

    L. Blume and D. Easley. Evolution and market behavior. Journal of Economic theory, 58 0 (1): 0 9--40, 1992

    work page 1992

  8. [8]

    Blume and D

    L. Blume and D. Easley. If you're so smart, why aren't you rich? belief selection in complete and incomplete markets. Econometrica, 74 0 (4): 0 929--966, 2006

    work page 2006

  9. [9]

    L. E. Blume and D. Easley. Optimality and natural selection in markets. Journal of Economic Theory, 107 0 (1): 0 95--135, 2002

    work page 2002

  10. [10]

    Dasaratha and K

    K. Dasaratha and K. He. Network structure and naive sequential learning. Theoretical Economics, 15 0 (2): 0 415--444, 2020

    work page 2020

  11. [11]

    L. I. de Castro, M. Pesce, and N. C. Yannelis. A new approach to the rational expectations equilibrium: existence, optimality and incentive compatibility. Annals of finance, 16: 0 1--61, 2020

    work page 2020

  12. [12]

    Eliaz and R

    K. Eliaz and R. Spiegler. A model of competing narratives. American Economic Review, 110 0 (12): 0 3786--3816, 2020

    work page 2020

  13. [13]

    G. W. Evans and S. Honkapohja. On the local stability of sunspot equilibria under adaptive learning rules. Journal of Economic Theory, 64 0 (1): 0 142--161, 1994

    work page 1994

  14. [14]

    G. W. Evans and S. Honkapohja. Learning dynamics. Handbook of macroeconomics, 1: 0 449--542, 1999

    work page 1999

  15. [15]

    M. Feldman. An example of convergence to rational expectations with heterogeneous beliefs. International Economic Review, pages 635--650, 1987

    work page 1987

  16. [16]

    J. P. Flynn and K. Sastry. The macroeconomics of narratives. Available at SSRN 4140751, 2022

    work page 2022

  17. [17]

    Frick, R

    M. Frick, R. Iijima, and Y. Ishii. Welfare comparisons for biased learning. 2021

    work page 2021

  18. [18]

    Friedman

    M. Friedman. The methodology of positive economics. 1953

    work page 1953

  19. [19]

    Ghosh and M

    A. Ghosh and M. A. Khan. On a diversity of perspectives and world views: Learning under bayesian vis- \'a -vis degroot updating. Economics Letters, 202: 0 109839, 2021

    work page 2021

  20. [20]

    Gilboa, A

    I. Gilboa, A. Postlewaite, L. Samuelson, and D. Schmeidler. Economic models as analogies. The Economic Journal, 124 0 (578): 0 F513--F533, 2014

    work page 2014

  21. [21]

    J. D. Hey and A. Morone. Do markets drive out lemmings—or vice versa? Economica, 71 0 (284): 0 637--659, 2004

    work page 2004

  22. [22]

    Hofbauer and K

    J. Hofbauer and K. H. Schlag. Sophisticated imitation in cyclic games. Journal of Evolutionary Economics, 10 0 (5): 0 523--543, 2000

    work page 2000

  23. [23]

    Kandori, G

    M. Kandori, G. J. Mailath, and R. Rob. Learning, mutation, and long run equilibria in games. Econometrica: Journal of the Econometric Society, pages 29--56, 1993

    work page 1993

  24. [24]

    J. G. Kemeny and J. L. Snell. Finite Markov chains: with a new appendix ``Generalization of a fundamental matrix". Springer, 1976

    work page 1976

  25. [25]

    Kogan, S

    L. Kogan, S. A. Ross, J. Wang, and M. M. Westerfield. Market selection. Journal of Economic Theory, 168: 0 209--236, 2017

    work page 2017

  26. [26]

    Lobel and E

    I. Lobel and E. Sadler. Information diffusion in networks through social learning. Theoretical Economics, 10 0 (3): 0 807--851, 2015

    work page 2015

  27. [27]

    G. J. Mailath and L. Samuelson. Learning under diverse world views: Model-based inference. American Economic Review, 110 0 (5): 0 1464--1501, 2020

    work page 2020

  28. [28]

    M. M. Mobius, M. Niederle, P. Niehaus, and T. S. Rosenblat. Managing self-confidence: Theory and experimental evidence. Technical report, National Bureau of Economic Research, 2011

    work page 2011

  29. [29]

    Mossel, M

    E. Mossel, M. Mueller-Frank, A. Sly, and O. Tamuz. Social learning equilibria. Econometrica, 88 0 (3): 0 1235--1267, 2020

    work page 2020

  30. [30]

    Pesce, N

    M. Pesce, N. Urbinati, and N. C. Yannelis. On the limit points of an infinitely repeated rational expectations equilibrium. CSEF Working Papers, 677: 0 1--27, 2023

    work page 2023

  31. [31]

    Sandroni

    A. Sandroni. Do markets favor agents able to make accurate predictions? Econometrica, 68 0 (6): 0 1303--1341, 2000

    work page 2000

  32. [32]

    K. H. Schlag. Why imitate, and if so, how?: A boundedly rational approach to multi-armed bandits. Journal of economic theory, 78 0 (1): 0 130--156, 1998

    work page 1998

  33. [33]

    R. J. Shiller. Conversation, information, and herd behavior. The American economic review, 85 0 (2): 0 181--185, 1995

    work page 1995

  34. [34]

    R. J. Shiller. Speculative asset prices. American Economic Review, 104 0 (6): 0 1486--1517, 2014

    work page 2014

  35. [35]

    R. J. Shiller. Narrative economics. American economic review, 107 0 (4): 0 967--1004, 2017

    work page 2017

  36. [36]

    P. D. Taylor. Evolutionarily stable strategies with two types of player. Journal of applied probability, 16 0 (1): 0 76--83, 1979

    work page 1979

  37. [37]

    R. M. Townsend. Market anticipations, rational expectations, and bayesian analysis. International Economic Review, pages 481--494, 1978

    work page 1978

  38. [38]

    Tran and V

    L. Tran and V. Vu. The" power of few''phenomenon: The sparse case. arXiv preprint arXiv:2302.05605, 2023

    work page arXiv 2023

  39. [39]

    X. Vives. Social learning and rational expectations. European Economic Review, 40 0 (3-5): 0 589--601, 1996

    work page 1996

  40. [40]

    H. Yan. Natural selection in financial markets: Does it work? Management Science, 54 0 (11): 0 1935--1950, 2008

    work page 1935