Project Description

Why do government structures change? Traditionally, the answer refers to political parties exercising control over a state bureaucracy. Yet government structures have been changing ever since they emerged alongside the modern state and thus already before parties existed. The recent decline of party government worldwide also begs the question whether other actors and mechanisms matter that have been underestimated so far. Meanwhile, it became a truism that “structure shapes policy”, but existing research focuses on other effects of structural change in government, such as democratic and economic outcomes or cabinet governance. We lack systematic analyses of how structural change affects government policy.

To address these gaps, STATE-DNA studies the change of units inside ministries and agencies as the ‘building blocks’ of the modern state. It submits a novel theory of evolutionary government that regards structural change in government as an interplay between a unit’s structural features and its organizational and environmental environment. This notion of multiple levels allows to study more than ‘party-centric’ causes of change and, for the first time, to analyze the effects on government policy explicitly, as evolutionary consequences managed by distinct units.

State-DNA utilizes genetic measures to assess changes within government structures, compiling a comprehensive dataset that spans from the early 19th century to the present. Our study includes six countries—France, Germany, Japan, the Netherlands, Norway, and the United Kingdom—from 1815 to 2025.

We gather data from a variety of primary and secondary sources, including government directories, organizational charts and budgets. Additionally, we employ web scraping techniques and make use of the increasingly available government APIs, to ensure comprehensive and up-to-date data collection. This approach allows us to capture detailed information on government structures and their changes over time.

By integrating historical data on legislative activity, organizational structures, and environmental factors, we employ established methods in computational social science, such as natural language processing (NLP) and webscraping. Additionally, we aim to incorporate advanced techniques from biology, like data assimilation, to trace and predict the evolution of government structures. Our empirical analyses will provide new insights into the origins and consequences of these changes, offering a predictive framework for understanding future governance.