For rigorous empirical study of religion to move forward, improved data on religious demographics of states’ populations, with more extensive and consistent coverage, become essential. The Religious Characteristics of States (RCS) data set project provides annual data on religious demographics for ninety-seven denominations, from 2010 back to as early as 1800 (approximately 6.7 million data points). Unlike existing data sets, RCS matches the temporal coverage of the Correlates of War and Polity data sets and is structured for easy merger with them. Illustrations of RCS’s capabilities include measurements of religions’ associations with regime type, wealth, and conflict. Religious characteristics emerge as significant state-level variables and rival other factors conventionally thought to influence political processes and outcomes. The results encourage further research using RCS on a wide range of problems within international relations and comparative politics.

Political secularization theories have predicted religion's decline in public and political life, and desecularization theories have predicted the reverse trend. However, there is little agreement on the timing of either phenomenon or even their existence. Until now, deep empirical tests of any of these were hampered by lack of historical country‐level data on religious preferences of governments (previously used data sets go back only to 1990). However, the new Government Religious Preference data set (GRP) measures state religion from 2015 back to the 1800s. Using GRP data, this article offers the first long‐term quantitative measurement of political secularization and in doing so, weighs in on competing claims regarding its timing. This article finds strong support that political secularization happened gradually over the long 19th century, accelerated after World War II, and peaked in the 1970s or 1980s. In contrast, the article finds only tepid support for the existence of political desecularization overall.

Symmetry has been a fundamental tool in the exploration of a broad range of complex systems. In machine learning, symmetry has been explored in both models and data. In this paper we seek to connect the symmetries arising from the architecture of a family of models with the symmetries of that family's internal representation of data. We do this by calculating a set of fundamental symmetry groups, which we call the intertwiner groups of the model. Each of these arises from a particular nonlinear layer of the model and different nonlinearities result in different symmetry groups. These groups change the weights of a model in such a way that the underlying function that the model represents remains constant but the internal representations of data inside the model may change. We connect intertwiner groups to a model's internal representations of data through a range of experiments that probe similarities between hidden states across models with the same architecture. Our work suggests that the symmetries of a network are propagated into the symmetries in that network's representation of data, providing us with a better understanding of how architecture affects the learning and prediction process. Finally, we speculate that for ReLU networks, the intertwiner groups may provide a justification for the common practice of concentrating model interpretability exploration on the activation basis in hidden layers rather than arbitrary linear combinations thereof.Preprint. Under review.

Effects of religious norms on interstate armed conflict are understudied compared to such effects on intrastate conflict. Yet religion exerts no less influence on states’ preferences and outcome of interstate interaction than other factors. Religion fosters emotion, in turn generating cognitive and motivational biases in states’ leaders. Religion acts as a political ideology that inculcates worldviews and programs political agendas. Religion accomplishes these things through scripture, clergy, and historical narrative. Through these media, religion instills standards of behavior, including for resort to (political) violence. Thus, religions influence states’ propensities to initiate interstate armed conflicts, even when belligerents’ goals are not necessarily sacralized. However, different norms are ingrained in different religions, and median war ethics of different religions range from permissive to restrictive. Therefore, over time and space, different religions are expected to have different effects on initiation of interstate armed conflicts. Regressing that outcome against the newly introduced Government Religious Preference (GRP) scores for Christianity, Islam, and Buddhism reveals that such is the case; different religions are differently associated with war and peace.