Our paper on functional form of police performance index is now Peer Reviewed and Published

Officials pay attention to how their agencies are ranked. Think of the influence of the U.S. News college rankings. Bearfield, Maranto and Wolf ranked cities by an index that includes both homicides and police-related civilian deaths. But my new paper shows that how high a city ranks depends on the mathematical function used to combine the individual variables into a single score. Functions that are equally plausible can make the rankings more or less sensitive to police-related civilian deaths.

The publisher allows me to give away some free downloads, so you can get the paper at https://www.tandfonline.com/eprint/HGTQKSQXDRZR5IAUMGWW/full?target=10.1080/10999922.2021.1886439

David S. Reed (2021) Police Performance Rankings Depend on the Functional Form of the Index: A Comment on Bearfield, Maranto, and Wolf, Public Integrity, DOI: 10.1080/10999922.2021.1886439

Police Performance Rankings Depend on the Functional Form of the Index

Bearfield, Maranto and Wolf (2020) advise policy-makers to measure policing outcomes using a metric that includes rates of homicide, police-related civilian deaths (PRCD) , and poverty. They present such an index, which they call the Police Performance Index (PPI). But alternative functional forms that are equally plausible can lead to different rankings of police departments, and therefore different policy conclusions. My comment presents one such alternative index, under which changes in police-related civilian deaths have a greater potential effect on a city’s ranking: https://doi.org/10.31235/osf.io/r8yst