Modern city governance relies heavily on crowdsourcing to identify problems such as downed trees and power-lines. Two major concerns are that (1) residents do not report problems at the same rates and (2) agencies differentially respond to reports, leading to an inefficient and inequitable allocation of government resources. However, measuring such under-reporting and differential responses are challenging statistical tasks: ground truth incident and risk distributions may differ by area, and, almost by definition, we do not observe incidents that are not reported. First, we develop a method to identify (heterogeneous) reporting rates, without using external (proxy) ground truth data. We apply our method to over 100,000 resident reports made to the New York City Department of Parks and Recreation, finding that there are substantial spatial and socio-economic disparities in reporting rates, even after controlling for incident characteristics. Second, we develop a method to audit differential response rates, even when incident occurrence varies spatio-temporally and the agency faces capacity constraints.