Elephants, Donkeys, and Colonel Blotto

TL;DR

This paper introduces a new neural network-based method to analyze political discourse and extends the Colonel Blotto game model with stochastic activation to replicate observed political dynamics.

Contribution

It develops a novel empirical analysis method using CNN classifiers and extends the Colonel Blotto game with stochastic elements to match real-world data.

Findings

Neural network classifiers effectively label political statements.

Extended Colonel Blotto model reproduces empirical political dynamics.

Model demonstrates dynamics similar to actual US political party behavior.

Abstract

This paper employs a novel method for the empirical analysis of political discourse and develops a model that demonstrates dynamics comparable with the empirical data. Applying a set of binary text classifiers based on convolutional neural networks, we label statements in the political programs of the Democratic and the Republican Party in the United States. Extending the framework of the Colonel Blotto game by a stochastic activation structure, we show that, under a simple learning rule, the simulated game exhibits dynamics that resemble the empirical data.