The goal of this monograph is to study the indicator function for a set of permutations mapping one finite sequence of positive integers to another from a representation theoretic, combinatorial and probabilistic perspective. The degree of a function of permutations is the size of the largest pair of sequences required when expressing it as a linear combination of these indicators. This notion of degree, implicit in work of Diaconis, is critical for many applications of representation theory to extremal combinatorics, machine learning, probability and statistics. We use the term local to indicate bounded degree and initiate the study of low degree class functions, which encode probabilistic data for permutation statistics on each cycle type simultaneously. We begin with a self contained treatment of for functions of permutations, developing its theory using language familiar to enumerative and algebraic combinatorialists. This leads naturally to a novel basis for symmetric functions we call the path power sum symmetric functions. The most technically challenging part of our work is the path Murnaghan-Nakayama formula, which expands path power sums into Schur functions. By combining the the path Murnaghan-Nakayam formula with the classical theory of character polynomials, one obtains a structural characterization for moments for permutation statistics conditioning on cycle type. We then analyze asymptotic properties of these moments. In doing so, we introduce the novel family of regular permutation statistics, which include almost all reasonable weighted pattern counting statistics. We show a large family of regular statistics satisfy a law of large numbers on a given cycle type depending only on the proportion of fixed points and a have variances depending only on fixed points and two cycles.