Statistics of opinion domains of the majority-vote model on a square lattice

The existence of juxtaposed regions of distinct cultures in spite of the fact that people’s beliefs have a tendency to become more similar to each other’s as the individuals interact repeatedly is a puzzling phenomenon in the social sciences. Here we study an extreme version of the frequency-dependent bias model of social influence in which an individual adopts the opinion shared by the majority of the members of its extended neighborhood, which includes the individual itself. This is a variant of the majority-vote model in which the individual retains its opinion in case there is a tie among the neighbors’ opinions. We assume that the individuals are fixed in the sites of a square lattice of linear size $L$ and that they interact with their nearest neighbors only. Within a mean-field framework, we derive the equations of motion for the density of individuals adopting a particular opinion in the single-site and pair approximations. Although the single-site approximation predicts a single opinion domain that takes over the entire lattice, the pair approximation yields a qualitatively correct picture with the coexistence of different opinion domains and a strong dependence on the initial conditions. Extensive Monte Carlo simulations indicate the existence of a rich distribution of opinion domains or clusters, the number of which grows with $L^2$ whereas the size of the largest cluster grows with $\ln L^2$. The analysis of the sizes of the opinion domains shows that they obey a power-law distribution for not too large sizes but that they are exponentially distributed in the limit of very large clusters. In addition, similarly to other well-known social influence model – Axelrod’s model – we found that these opinion domains are unstable to the effect of a thermal-like noise.

💡 Research Summary

The paper investigates a spatial majority‑vote model on a two‑dimensional square lattice, motivated by the persistent coexistence of distinct cultural or opinion domains in societies despite repeated interpersonal interactions that tend to homogenize beliefs. Each lattice site hosts an agent holding a binary opinion (±1). At each update a randomly selected agent considers its “extended neighbourhood”, defined as the agent itself together with its four nearest neighbours, and adopts the opinion that is in the majority within this five‑site set. In the case of a tie (exactly two agents for each opinion), the agent retains its current opinion, thereby introducing a self‑preservation rule absent in the classic majority‑vote model.
The authors first apply a single‑site mean‑field (MF) approximation, deriving an ordinary differential equation for the global opinion density ρ(t). This analysis yields three fixed points (ρ=0, 1, ½) but stability calculations show that only the consensus states (ρ=0 or 1) are stable, implying that the single‑site MF predicts inevitable takeover of the lattice by a single opinion, irrespective of initial conditions.
Recognizing that this approximation neglects spatial correlations, the study proceeds to a pair‑approximation MF framework. Here the nearest‑neighbour correlation C(t)=⟨σ_iσ_j⟩ is introduced, leading to coupled evolution equations for ρ and C. Numerical integration of these equations reveals a richer phase portrait: depending on the initial values of ρ and C, the system can settle into a multitude of metastable configurations characterized by the coexistence of several opinion domains. Positive correlations sustain clusters of like‑minded agents, preventing global consensus.
To test these analytical predictions, extensive Monte‑Carlo simulations were performed on lattices of linear size L ranging from 32 to 256, with both random (ρ₀≈0.5) and biased (ρ₀≈0.8) initial conditions. The simulation data exhibit several striking statistical regularities. First, the number of distinct opinion clusters N_c in the final frozen state scales linearly with the total number of sites, N_c∝L², indicating that, on average, each lattice site belongs to a separate small cluster. Second, the size of the largest cluster S_max grows only logarithmically with system area, S_max∝ln(L²), confirming that no macroscopic domain dominates the system. Third, the cluster‑size distribution P(S) follows a power law P(S)∝S^{−τ} with τ≈2.1 for intermediate sizes (S up to roughly 10⁴), but crosses over to an exponential tail P(S)∝exp(−S/S₀) for very large clusters (S≫10⁴). This dual scaling mirrors the predictions of the pair approximation and demonstrates that the system self‑organizes into a scale‑free regime bounded by a finite‑size cutoff.
The robustness of these domains was further examined by introducing a thermal‑like noise parameter ε. With probability ε an agent flips to the opposite opinion regardless of its neighbourhood, mimicking spontaneous opinion changes or external perturbations. As ε increases, the authors observe a sharp transition at ε_c≈0.03: below ε_c the cluster structure persists, while above ε_c the domains rapidly dissolve and the system approaches a disordered, mixed state. This noise‑induced destabilization parallels findings in Axelrod’s cultural dissemination model, where cultural regions are similarly fragile to random perturbations.
In conclusion, the study demonstrates that (i) a naïve single‑site mean‑field treatment overestimates consensus formation, (ii) incorporating nearest‑neighbour correlations via a pair approximation captures the emergence and persistence of multiple opinion domains, (iii) large‑scale simulations confirm that the number of domains scales with system area while the largest domain grows only logarithmically, (iv) cluster sizes obey a power‑law regime with an exponential cutoff, and (v) modest levels of stochastic noise can eradicate the domain structure. These results provide a quantitative framework for understanding how local majority‑rule interactions, spatial correlations, and random fluctuations together shape the rich tapestry of opinion diversity observed in real societies.