Crisis Databank

While the scope and interconnectedness of our current crises are perhaps unique to these turbulent times, societal collapse, polarization and fragmentation, eruptions of civil violence, and even environmental catastrophe have been recurrent challenges since humans began to form large-scale societies over 6,000 years ago. The Crisis Databank project, led by Peter Turchin and Daniel Hoyer, seeks to draw out contemporary lessons from a rigorous assessment of past periods of sociopolitical instability, documenting the range of outcomes that societies have experienced.

Utilizing the innovative longitudinal dynamic analyses pioneered by Seshat: Global History Databank, we aim to pinpoint the leverage points that can tip the scales from the more devastating to the less disruptive consequences. Our data-driven approach not only reveals common patterns in societal dynamics, but also highlights the critical differences between societies and the unique challenges they each face. Our research is geared towards academics, policy-makers, social justice movement leaders, and other stakeholders through detailed analyses and guidelines that we will develop to translate the insights from historical study to useful, much-needed tools to help navigate our present.

Population booms and busts in prehistory

It is sometimes assumed that crisis and social breakdown are hazards that face only societies with large-scale political and economic organization – states or empires. The fall of the Western Roman Empire, the collapse of the Classic Maya, and the disintegration of the Bronze Age Mediterranean around 1200 BCE are well-known examples. In The Collapse of Complex Societies (1998), Joseph Tainter explored “collapse” as a phenomenon that occurs when a society is no longer able to sustain the level of complexity it has achieved: infrastructure falls into disrepair, information flow becomes restricted, and there may be territorial fragmentation and population decline. In contrast, prehistoric periods – especially before the Bronze Age – are presented as comparatively idyllic, times of stability and steady population growth.

Yet evidence is accumulating that life before the emergence of states and empires was also characterized by periodic crises and population declines. One very dramatic example is the Cucuteni-Trypillia culture in the fourth millennium BCE. Settlements associated with this culture in modern-day Romania, Moldova and Ukraine reached sizes of 10,000 or more inhabitants, considered by some as the first cities in Europe, but after a long period of occupation, a complete depopulation is seen in the region, with barely any evidence for human habitation for several centuries. Similarly, for the first farmers in western and central Europe, associated with the Linearbandkeramik culture, there is evidence of repeated cycles of population booms and busts. What causes this pattern?

Possible drivers of population trajectories include the variety of challenges faced by early farmers like changes in climate, overexploitation of resources, conflicts, warfare and diseases. There is no shortage of plausible ideas, but we need a way to judge which hypotheses best explain the observed patterns.

To this end, Dániel Kondor and Peter Turchin have been building agent-based models of prehistoric farming societies in Europe, and evaluating them against archaeological data on population levels. This modelling work suggests that external drivers such as climatic changes are insufficient to account for the temporal and spatial scale of population booms and busts in prehistoric Europe. The best-supported model incorporates density-dependent conflict, including cultural shifts between peaceful and confrontative attitudes among farming settlements.

The results support the hypothesis that internal conflict is the main driver of demographic patterns. Like the better-known states and kingdoms of written history, it seems that our Neolithic forebears also underwent cycles of social cohesion and disintegration. Importantly, our research highlights that the very earliest farmers of Europe formed complex regional systems, even though political authority was limited to much smaller scales, possibly to individual settlements, or closely aligned groups of settlements.

Historical CrisisDB

No one can talk about history for any length of time without considering its cycles. In Ages of Discord, Peter Turchin showed that, even in vastly different eras of human history, societies tend to oscillate predictably between two opposing social forces: 1) the economic and social interests of elites and 2) the promotion of popular well-being. And importantly, that these oscillations lead to correlated undulations of societal unrest and political violence. But to what extent is history actually repetitive? Is it possible to identify why some societies are able to right the ship while others descend into chaos and bloodshed?

To answer these questions, we need a historical database. We are working now to bring together the information we need to study how political crises and state breakdowns have played out across roughly 300 historical case studies ranging from the Egyptian Old Kingdom to the twentieth-century United States. Combining data on demographics and elite numbers, coin hoards, skeletal evidence of violence, state expansion and contraction, epidemics, famines and other variables helps create a nuanced, dynamic picture of crisis and recovery for each case study.

Our hope for the Historical CrisisDB project is that a systematic look at historical crises can reveal unsuspected patterns—and maybe even suggest possibilities for responding to the situation we find ourselves in today.

Contemporary CrisisDB: Our Ages of Discord

The Contemporary CrisisDB project – what we call “Our Ages of Discord” – is an effort to extend our dataset on crises into the present day. We will chronicle economic, political, and social variables to track conditions for elites, non-elite populations, and state institutions across a host of modern countries. This growing databank currently houses data on over a dozen countries (and growing), covering over 100 variables—measured annually since 1900—compiled from 50+ datasets. These variables are intended to allow unprecedented detail in the investigation of Structural Demographic Theory’s ability to explain recent social crises and forecast future ones.

© Peter Turchin 2023 All rights reserved

Privacy Policy