A theory of political belief system dynamics should incorporate causal connections between elements of the belief system and the possibility that belief systems are influenced by exogenous factors. These necessary components can be satisfied by conceptualizing an individual’s belief system as a network of causally connected attitudes and identities which, via the interactions between the elements and the push of exogenous influences, produces the disparate phenomena in the belief systems literature. We implement this belief systems as networks theory in a dynamic Ising model and demonstrate that the theory can integrate at least six otherwise unrelated phenomenon in the political belief systems literature, including work on attitude consistency, cross-pressures, spillover effects, partisan cues, and ideological differences in attitude consensus. Our findings suggest that belief systems are not just one thing, but emerge from the interactions between the attitudes and identities in the belief system. All code is available: https://osf.io/aswy8/?view_only=99aff77909094bddabb5d382f6db2622 .