Cliodynamics

We provide empirical evidence that the distribution of empires' lifetime is exponential with a typical time-scale of 300 years. After introducing and computing a proxy measure for the dimension of a large polity we interpret the data by means of a simple dynamical model for the empires' evolution. The resulting theory suggests that the behavior of a long-lived empire is qualitatively different from a short-lived one. They all experience an initial expansion phase but while the former goes through a contraction period over the last $20\%$ of its life, the latter stays essentially unchanged after reaching its maximum size. In both cases, a sudden collapse occurs when the empire's size is still close to its own historical average.

This study presents a civilizational dynamics model that applies Newtonian mechanics to Toynbee's challenge-response theory. By defining quantifiable indicators of societal response (α) and challenge (β), the model constructs a civilizational sustainability index S(t)=lnα(t)−lnβ(t). The first, second, and third differences of S(t) are interpreted as civilizational velocity, acceleration, and jerk, respectively, offering dynamic insights into the non-linear changes of civilizational states. Simulations applied to Roman civilization (BC 100 – AD 100) identify critical transitions within this historical period, providing empirical support for Toynbee’s hypothesis that civilizations collapse when challenges are either too weak or overwhelming. The model offers a novel quantitative method to analyze civilizational resilience and decay by capturing structural transformations through mathematical dynamics. While acknowledging the limitations of quantitative modeling in the humanities, this interdisciplinary approach demonstrates significant potential for extension to fields such as economics, politics, and systemic risk analysis.

The paper aims to suggest Data Envelopment Analysis (DEA), as a preliminary tool of military efficiency. This approach examines battles as a decision-making unit (DMU) characterized by multiple inputs and outputs. Data: Data are measurable indicators such as battles won, territorial expansion, casualties, and historical impact. By assigning quantitative values, the study facilitates cross-temporal comparisons of military effectiveness. Analysis: The strategic performance of Alexander the Great at the Battle of Granicus River (334 BCE) serves as a primary case study of DEA employment. The scope of analysis extends to a range of historically significant conflicts, from the Trojan War to World War II to compare with and be used as benchmarks. Results: The study demonstrates the practical utility of quantifying historical events into a single comparable metric, facilitating clearer comparisons across diverse battles ranked from highest to lowest values. 

In the study of the causes of the Russian Revolution, the problem of the standard of living of the population plays an important role. This problem, in turn, is linked to the question of agricultural productivity. In modern historiography, both domestic and foreign, the thesis of the growing productivity of Russian agriculture in the post-reform period is considered proven - and in particular the increase in grain yields. This thesis is based on the well-known works of V.G. Mikhailovsky, V.M. Obukhov and A.S. Nifontov, in which time series of grain yields in European Russia were constructed on the basis of official statistics. Meanwhile, the opinion of the experts of the 1901 Commission is well known, who believed that the increase in yields recorded in official statistics was explained by the improvement in the system of collecting harvest data. Reforms to improve the survey system were carried out in 1870, 1883 and 1893. The author examines the dynamics of 4-year averages and shows that when 4-year periods containing the years indicated are excluded from consideration, the yield in the remaining time intervals does not increase. In other words, the increase in yield in the time series shown was explained by more detailed accounting.

    The period 1893-1914 is considered separately, when it is assumed that the yield data were quite accurate and no new reforms were made in the field of their collection. Previously it was assumed that the yield, calculated by the regression coefficient of the linear model, increased by 12% during this period. The author conducts a more detailed analysis and shows that the regression coefficient used previously is statistically insignificant. Thus, the claim of an increase in returns over this period cannot be statistically substantiated. Perhaps the return was a random variable independent of time.

The paper also examines the dynamics of gross cereal yields per capita and shows that average per capita yields did not increase between 1893 and 1914.

       Thus, the prevailing opinion about the growth of agricultural productivity in Russia in the post-reform period needs to be revised

The previous research with incomplete data revealed that zemstva expenditure on education per capita were higher in regions with low level of education, but these spending did not make much of a difference – human capital in these regions remained relatively low (Popov, Konchakov, Didenko, 2024). The results reported in this paper provide additional and more rigorous proof that zemstva activities and the increase in their spending for education in 1897-1913 contributed to the spread of primary education and to the decline in the inequality of the distribution of human capital not only between the regions< but also within the regions (ratio of secondary to primary education enrollment). 

But we also show that there were more powerful forces at play – education for tuition fees, central government and city/town administration financing – that were pushing the development in an opposite direction, increasing the secondary education enrollment in most regions faster than the primary education enrollment. The result was the widening gap between low and high educated individuals that could have contributed to the formation of the intelligentsia phenomenon – educated intellectuals that were not able to find the proper place in the national economy to apply their knowledge. Intelligentsia opposition to the tsarist regime, however, did not take violent forms – regions with fast growing educational disparities registered lower, not higher increases in peasants’ unrest, industrial strikes and crimes against persons.

This report describes the current canonical dataset named Polaris2025, a subset of Seshat: Global History Databank data. This dataset focuses on a well curated list of polities and variables that have been already used in publications or will be used in forthcoming ones. The report provides an introduction to the methods and procedures of the Seshat project relating to the curation and release of the Polaris2025 dataset. The Polaris2025 dataset builds upon an earlier release, the Equinox2020 dataset, and is now the most current and complete set of publicly available Seshat data. The Equinox2020 dataset remains valuable as a replication dataset. This report also describes important changes in the Seshat Databank itself, notably the new SQL database, the community tools, and the GIS portal.