A Generalized Model for Including Equity in the Siting of Emergency Services

Abstract We propose a method to identify inequity in access to emergency services using logistic regression and present a model to address inequity by siting additional facilities, recommending both quantity and location. We classify emergencies by the median income bracket of their Zip code tabulation areas. We use logisitic regression to determine whether the probability of emergency responses classified by income bracket are within the average response, defined by time or distance, and quantify overall equity using mean absolute deviation. To address inequity, we iterate a bounded maximal covering location problem over coverage distance and additional facilities to provide a set of possible solutions. We evaluate the mean absolute deviation of each solution and identify the tradeoff of adding facilities and minimizing deviation.We use 2019 data from the United States Internal Revenue Service and San Francisco Fire Department as an example case for our model. In this example, we find the distribution of emergency rooms leave the low income population of San Francisco significantly underserviced and recommend to decision makers additional locations that reduce the average distance to an emergency room while minimizing mean absolute deviation—providing more equitable access to emergency rooms across income groups in San Francisco.