Evolution of family systems and resultant socio-economic structures

The study asks how family systems—specifically parent-child residence (nuclear vs extended) and inter-sibling inheritance (equal vs unequal)—evolve under socio-ecological conditions and how these systems shape socio-economic structures. It situates family traits as culturally transmitted, slow-changing features influenced by ecological and social factors (e.g., subsistence strategies, heritable resources), yet with unclear causal directions between family systems and societal structures. The authors aim to construct a minimal, long-term evolutionary model appropriate for pre-industrial agricultural contexts where human labour, land, property, and diminishing returns to labour are central, to explain the emergence of four family system "phases" and link them to wealth distributions and modern ideologies, then validate with cross-cultural data.

Prior work shows: (1) family traits are vertically transmitted and slow to change; (2) correlations exist between exposure to the Western Church and nuclear families; (3) evolutionary anthropology explains family relations via parental investment and intra-family competition; (4) heritable resources in agricultural/pastoral societies lead to sibling competition and patriliny; (5) greater wealth inequality in agricultural societies is linked to lower polygyny; (6) subsistence patterns and socio-ecological conditions shape family traits; (7) phylogenetic comparative methods reveal historical diffusion and ancestry effects on family traits. However, causal directions remain debated. Multi-level selection has explained group-level structure evolution and was previously applied by the authors to kinship structures; here they focus on residence and inheritance patterns and briefly on polygyny.

Model: An agent-based, multi-level evolutionary model (hierarchical Moran process) with societies composed of families. Competition occurs at both family and society levels. Societies split when family count doubles initial size, replacing a random society to keep total societies fixed; thus faster-growing societies prevail. Key environmental parameters: (i) land capacity c (relative capacity per society), limiting available land per family when the number of families exceeds capacity; production per family then scales inversely with family count; (ii) wealth required for survival e, the minimum wealth needed to establish a new family; families below e fail to reproduce. Family state and strategies: Each family has population and wealth w, and two heritable strategy parameters: s (probability children remain to produce in an extended family before separation) and λ (inheritance inequality among siblings; shares proportional to exp(-λ m) by birth order m; λ=0 is equal division, larger λ biases toward earlier-born heirs). Strategies transmit culturally with mutation rate μ (Gaussian noise). Wealth decays by factor d each generation. Production and life cycle: With probability 1−s, siblings separate before production (nuclear path): inheritance is divided per λ, each new family pays e to found, then produces independently. With probability s, siblings remain to produce together as an extended family, then separate afterward, dividing inheritance and paying e. Production follows diminishing returns: output increases with log(1+labour) and scales with current wealth (linear feedback) and a capacity factor r=min(1, c×(initial or target family size)/current families), with multiplicative noise ηN(0,σ²). Under diminishing returns, siblings’ total production is higher if working independently, but when capacity is limited, society-level production can be lower for all-nuclear societies due to inefficient land use, creating a family-society selection conflict. Demography: Number of children per family is Poisson with mean b + f w (baseline b and wealth effect f). After reproduction, wealth decays by d. Societies split when family count reaches 2×initial count. Parameters (Table 1): example defaults include b=1.5, f=0.3, d=0.5, Ns=30 (initial families/society), N=50 (societies), μ=0.03, σ=0.1; c and e varied; s and λ evolve. Algorithm (summary of equations): wealth decay w_{i}^{t+1}=(1−d)w_{i}^{t}; inheritance shares proportional to e^{-λ k}/Σ e^{-λ k}; founding cost e per new family; production w gain ∝ r(1+η)(1+w) log(1+labour) with labour equal to 1 per nuclear offspring or summed for extended production; births N_i^{t+1} ~ Poisson(b+f w_i^{t}); strategy transmission s_{i,j}^{t+1}=s_i^{t}+ζ, λ_{i,j}^{t+1}=λ_i^{t}+ζ with ζN(0,μ²). Initial strategies s=0.5, λ=log 2 for all families. Extensions: (1) Marriage and polygyny: families track male/female populations, separate inheritance rules for sons/daughters, and polygyny allowed if bridewealth affordable; polygyny increases production and reproduction. (2) Alternative subsistence scenario with relaxed diminishing returns (production linear in labour) to represent labour-extensive patterns (e.g., foraging/horticulture). Simulations typically run 2000 steps; outcomes averaged over 100 trials per condition; classification thresholds: extended if s≥0.5; unequal inheritance if λ≥log 2.

• Evolution of family systems by environment: Increasing land capacity c selects for nuclear families (lower s). Increasing wealth required for survival e selects for more equal inheritance (lower λ). Four emergent phases: stem (extended + unequal) for small c, small e; community (extended + equal) for small c, large e; absolute nuclear (nuclear + unequal) for large c, small e; egalitarian nuclear (nuclear + equal) for large c, large e.
• Robustness and parameter effects: Phase diagram qualitatively robust across parameter ranges. Larger society-level competition (large number of societies Ne or small families per society Nf) favours nuclear strategies even under limited c (selfish family-level advantage). Higher baseline birth rate b yields more children per family, smaller per-child shares, reduced wealth accumulation, and broader regions of equal inheritance.
• Marriage/polygyny extension: Son-biased parental investment evolves broadly; polygyny frequency <20% in agricultural (diminishing returns) settings and largely independent of c and e; marriage modeling minimally affects s and λ outcomes. With linear returns (labour-extensive), polygyny exceeds 20% broadly and extended families remain favourable even with sufficient c.
• Wealth distributions: Within societies, wealth distributions show power-law poor tail (w^{-α}) and exponential rich tail (exp(−β w)). Smaller c → smaller α (heavier poor tail); smaller e → smaller β (heavier rich tail). Extended families associate with heavier poor tails (more poor), while unequal inheritance associates with heavier rich tails (more rich). Heir vs non-heir distributions diverge more as e decreases (stronger inequality λ), accelerating heir wealth accumulation; low c (extended) leads to poorer younger siblings; high c (nuclear) benefits younger siblings.
• Empirical validation (SCCS): Among 91 agricultural societies with movable-property inheritance: 14 stem, 30 community, 17 absolute nuclear, 30 egalitarian nuclear. Spearman correlations consistent with model: Communality of land (proxy for larger c) correlates with extended vs nuclear (Corr. −0.31, P<0.05); Land shortage (proxy for smaller c) correlates with extended (Corr. 0.26, P<0.1). Frequency of internal warfare (proxy for larger e) correlates with equal inheritance (Corr. 0.36, P<0.1); Acceptability of violence also positive (Corr. 0.27). Number of poor (proxy for smaller α) correlates with extended families (Corr. 0.29, P<0.1); Number of rich people (proxy for smaller β) is negatively correlated with equal inheritance (Corr. −0.37, P<0.01).

The model demonstrates how environmental constraints on land and survival wealth shape family residence and inheritance strategies, resolving the directionality between socio-ecological conditions and family systems. It links micro-level family strategies to macro-level wealth distributions and, by extension, socio-economic structures and ideologies. Historical-geographical patterns align: absolute nuclear regions (England, Netherlands) had ample land and lower survival thresholds, yielding heavy rich tails and lighter poor tails, facilitating capital accumulation and liberal-capitalist ideologies. Egalitarian nuclear regions (France, Spain, Italy) had sufficient land yet higher survival needs, yielding lighter tails on both sides and values of freedom and equality. Stem family regions (Japan, Germany, parts of rural Western Europe) with limited land and low survival thresholds exhibited heavy tails on both sides and stratification. Community family regions (China, Russia, Northern India) with scarce land and high survival needs had heavy poor tails and light rich tails, consistent with widespread poverty and support for communism/social egalitarianism. The framework is general under diminishing returns and multi-level selection and offers a mechanism for historical transitions: as populations grow and conflicts intensify, c declines and e rises, shifting systems from nuclear to stem to community. Parameter dependencies on group size and competition explain variation in family forms under similar ecologies, and alternative subsistence economizes explain higher polygyny and extended residence where labour returns are linear.

A minimal multi-level evolutionary model explains the emergence of four canonical family systems in pre-industrial agricultural societies as adaptive responses to land capacity (c) and survival wealth (e), and links these systems to characteristic wealth distributions and socio-economic structures validated by cross-cultural data. The study integrates insights from evolutionary anthropology, demography, and socioeconomic history and shows that micro-level family strategies can shape macro-level societal patterns and ideologies. Future work should incorporate social stratification (elite-commoner interactions), explicit intra-family competition at the individual level (three-level models), broader subsistence ecologies, and world-system interactions, and leverage larger, longitudinal datasets to infer causal dynamics and temporal transitions.

• Social stratification is not modeled; landlord-tenant relations and class interactions could change environmental parameters and family systems across strata.
• Intra-family competition is implicit; explicit modeling of parent-sibling conflict and birth-order effects at the individual level is absent.
• Scope restricted to pre-industrial agricultural societies with diminishing returns; other subsistence patterns and modern contexts require alternative production functions and interactions.
• Empirical analysis is correlational, limited by sample size (SCCS), variable coverage, and lack of chronological data; causal inference and richer validation remain open.
• The model abstracts marriage details (despite an extension) and does not include broader economic/political networks (towns, international trade) characteristic of world-systems.