Police Unions Kill
Came across this interesting study, with the headline figure being police unionisation is linked to 14% of non-White civilian deaths.
Collective Bargaining Rights, Policing, and Civilian Deaths by Jamein P. Cunningham and Rob Gillezeau. Published in volume 18, issue 3, page
Using an event-study design, we examine whether states' introduction of duty-to-bargain requirements for police unions led to meaningful changes in civilians killed by law enforcement. We find that the introduction of collective bargaining rights significantly increases the number of civilians killed by police, particularly among non-White civilians, while having limited effects on officer safety. Our results indicate that the adoption of bargaining rights for law enforcement can explain 14 percent of all non-White civilian deaths by legal intervention between 1959 and 1988. Given our findings, the idea that police unions exacerbate violence is empirically grounded.
I don't want to be construed as anti-union here (I am a dues paying member of a union!), but I think this is a decent demonstration that unions are not a sort of ideal "pro-labour" form of organisation. They are pro-member, with those members being labourers in some capacity.
What is good for the membership may not be good for society writ large. This is fairly evidently true for police unions, but can also hold true for, say, teacher unions or nursing unions. There's a decent number of examples of unions clashing with other unions due to competing interests (see longshoremen and teamster unions feuding, as an example).
On the Left there's a fairly uncritical pro-union-in-all-circumstances-except-police stance, at least outside of some of the more niche Marxist debates about the revolutionary potential of unions (which I won't go into here). This makes sense to a degree given the hostility faced by unions over the past few decades, but I do think there is a need to recognise that they are another form of special interest group that has specific interests to a select membership. This needs to be balanced against other interest groups and the interests of the public more broadly.
Good news is that there are lots of different models for how unions can operate (contrast America's with the sectoral bargaining model in many European countries, for example). So there's room to be pro-union while still finding ways to balance their special interests against broader ones.
One of the key graphs from the study: you can see that in the years following the implementation of duty to bargain (where the government has to bargain with a police union) that non-white civilian deaths gradually rise. Such a rise isn't prevalent amongst white civilian deaths or when the government isn't obligated to entertain a union.














