The need for a translational science of democracy