Inequality in relational wealth within the upper societal segment: evidence from prehistoric Central Europe

The study investigates long-run inequality dynamics within a specific societal segment—the upper societal segment—of prehistoric Central Europe, focusing on relational wealth (an individual’s endowment with social ties and networks). Existing work often measures inequality at the level of entire societies and emphasizes material wealth, leaving gaps regarding intra-elite dynamics and non-material wealth. The authors assemble a long-run dataset of ~5000 single-funeral burial mounds (first four millennia BCE) and use burial mound volume as a proxy for relational wealth to analyze inequality and the relative size of the upper societal segment over time. They hypothesize that mound volume reflects the economic-political capacity to mobilize labor and resources, and that inequality within this segment varies with technological, environmental, demographic, network, and sociopolitical changes across the Neolithic to Iron Age.

The paper situates itself within research on inequality in modern, preindustrial, and ancient contexts (e.g., Alfani, Milanovic, Piketty, Scheidel) and specifically ancient inequality drivers (Bogaard; Fochesato; Kohler; Windler). It underscores the need to study inequality within elites because intra-elite disparities can destabilize societies (Turchin). Drawing on anthropological concepts, relational wealth encompasses social networks and the capacity to mobilize collective labor (Borgerhoff Mulder; Beck & Quinn). Archaeological and ethnoarchaeological studies indicate monument construction involves broader kin groups and functions as social signaling and long-term investment (Jeunesse; Miller; Wunderlich; Bliege Bird & Smith; Quinn). The study also integrates literature on technological change, climate events, demography, trade/exchange networks, and sociopolitical transformations as potential drivers of inequality and network restructuring (Boserup; Childe; O’Brien & Shennan; Bond events; population studies by Zimmermann; trade networks by Kristiansen et al.; political economy by Furholt et al.).

• Data: Compiled a dataset of 4986 single-funeral burial mounds across Central Europe covering the first four millennia BCE (Neolithic to pre-Roman), plus counts of individuals in flat and collective burials to estimate the share interred in mounds (proxy for the size of the upper societal segment). Primary archaeological catalogs and additional sources are listed in the Supplementary Information (SI). - Dating and binning: Burial mound dating intervals vary widely (10–4700 years). The analysis restricts to mounds with maximum dating uncertainty ≤600 years to ensure precision (robustness checks show thresholds >600 years yield similar results). The period is divided into 200-year bins; each mound is assigned to a bin by the mean of its start/end dating. Robustness to 250- and 300-year bins is reported. - Proxy construction (volume from floor area): Because mound volumes are not directly observed, the authors compute volumes from recorded ground areas and assumed shapes. Shapes are categorized as round (round, round-oval, oval) and rectangular (rectangular, trapezoid). For round mounds, assuming hemispherical form and height proportional to radius, volume scales with area^1.5. For rectangular mounds, assuming a cuboid with length and height proportional to width, volume similarly scales with area^1.5. Since inequality indices are scale-invariant, proportionality constants do not affect results. Only single-funeral mounds are used to tie monument size to an individual’s relational wealth. - Inequality measurement: They compute the Gini index and Generalized Entropy Measures (GEM: I_c, I_0, I_1) on mound volumes for each 200-year bin across Central Europe (spatially aggregated for a time-granular perspective). - Statistical inference: Distributions are heavy-tailed, making bootstrap/asymptotic standard errors unreliable. The authors implement permutation tests (as per Dufour et al.) to assess the statistical significance of differences between inequality estimates across periods. Samples are rescaled by their means; K random permutations are drawn; a standardized difference in indices serves as the test statistic; two-sided p-values are computed from the permutation distribution. - Upper segment size: The share of individuals buried in mounds is computed as single-funeral mounds divided by the sum of individuals in burial mounds, flat graves, and collective graves, providing a proxy for the relative size of the upper societal segment. - Sample coverage and caveats: Some bins (0–200 BCE, 2800–3000 BCE, 3000–3200 BCE) have too few mounds for credible analysis. In intervals dominated by collective burials (e.g., 3300–2800 BCE), the single-mound proxy is not valid for inequality, and no indices are computed.

• Persistent, high inequality within the upper societal segment: Across most of the first four millennia BCE, inequality in relational wealth (via mound volumes) among those with burial mounds is high. - Nonlinear, wave-like dynamics with an overall increasing trend: Inequality fluctuates within recurrent two-phase patterns and trends upward over time, especially during the last 1200 years BCE. - Changing size of the upper societal segment: The share of people buried in mounds exhibits wave-like expansions and contractions, indicating temporal shifts in how many individuals could express relational wealth via mound construction. - Temporal patterning (P1–P4): The authors identify repeating two-phase patterns linked to innovations, network restructuring, and external triggers. Phase (1) typically features centralized networks around few nodal individuals; Phase (2) shows more localized networks enabling broader participation in the upper segment. - Specific statistical result: The increase in inequality from 2400–2200 BCE to 2000–1800 BCE is statistically significant (p = 0.002). Some later changes (e.g., 1000–800 to 800–600 BCE) are not statistically significant due to sample size differences. - Contextual correlates: Inequality dynamics correspond with technological changes (animal traction, wheel, tin bronze, iron), climate events (Bond events), demographic shifts (growth and restructuring of collectively acting groups), and shifts in long-distance trade and sociopolitical organization (e.g., Corded Ware, Bell Beaker, Únětice, Tumulus, Urnfield, Hallstatt). - Proxy limitations period: During 3300–2800 BCE, widespread collective burial means the mound-based inequality proxy is not applicable; nevertheless, size differentiation in collective monuments suggests inequality between burial communities.

Findings address the research question by showing that intra-segment (upper societal segment) inequality in relational wealth was high, dynamic, and generally increasing, rather than stable. The observed wave-like patterns align with periods of technological innovation, climatic variability, demographic restructuring, and reconfiguration of trade and social networks, which alter network centrality, connectivity, and the capacity of individuals to mobilize labor and resources. The study suggests that advances in agriculture, metallurgy, and transport increased overall connectivity and population densities, fostering competition among elites and generating higher inequality at the top. It also highlights episodes in which disruptive events (e.g., migration, cultural shifts) localized networks and temporarily broadened participation in the upper segment. The role of conflict is discussed as both an expression and facilitator of network access and elite consolidation, with archaeological indicators such as weaponry in burials, fortified settlements, and battlefield evidence. Overall, the results underscore the importance of analyzing inequality within socially distinguished segments alongside whole-society measures to understand ancient social complexity and its drivers.

The paper contributes a novel, long-run, quantitative perspective on inequality in relational wealth within an upper societal segment in prehistoric Central Europe, using burial mound volume as a proxy. It documents persistently high and increasingly trending inequality with wave-like fluctuations, alongside changes in the size of the upper segment. The results are linked to technological, environmental, demographic, and sociopolitical processes that restructure networks and opportunities for accumulating relational wealth. The authors propose future research directions: (1) expand evidence on inequality within specific societal segments to better contextualize overall inequality trends; (2) investigate the role of inter- and intra-group violence in sociopolitical transformations; (3) measure and compare multiple wealth dimensions beyond material forms using quantitative approaches; (4) conduct regional analyses by leveraging the dataset’s spatial granularity and integrating other regional data sources.

• Proxy constraints: Using mound volume as a proxy for individual relational wealth excludes periods/regions dominated by collective burials (e.g., 3300–2800 BCE), where inequality among individuals cannot be assessed; inequality likely manifested between burial communities instead. - Sampling and dating: Some 200-year bins have low sample sizes (e.g., 0–200 BCE; 2800–3000 BCE; 3000–3200 BCE), limiting inference. The analysis restricts to mounds with dating uncertainty ≤600 years to improve precision; results are robust to alternative thresholds and bin widths but remain sensitive to data availability. - Shape/volume reconstruction: Volumes are inferred from recorded areas and assumed shapes (area^1.5 transformation), introducing model-based uncertainty about mound geometry; scale invariance of indices mitigates multiplicative constant issues but not shape assumptions. - Heavy-tailed distributions: Standard bootstrap/asymptotic SEs are unreliable; permutation tests are employed, which depend on rescaling and permutation counts and still reflect finite-sample limitations. - Scope and representativeness: The study aggregates Central Europe to gain temporal granularity; subregional heterogeneity is not analyzed here. Burial practices bias the observable upper segment (mounds were reserved for socially distinguished individuals), so results do not represent whole-population inequality. The share of mound burials approximates the relative size of the upper segment but depends on the completeness and comparability of burial records across time and space.