First there is a mountain, then there is no mountain. Then there is.

An interesting but frustrating piece, (Mis)measuring the Shoplifting Crisis by Charles Fain Lehman.  The subheading is "Is there a wave of retail theft? Data tell a more complicated story than the headlines."

Most of the points he raises are good points, particularly with regard to the weak integrity of the data.  But data has always been weak in this regard.  He is trying to treat two propositions equally.  One proposition is that property crime, shoplifting in particular, has risen due to rampant smash-and-grabs.  The other proposition is that there is no such rise and that the perception of an increase is a product of non-typical videos with high sharing and engagement. 

Even though he self-consciously tries to be fair, it still comes across somewhat as an effort to rescue the narrative that there is no real rise in crime.  However, he does introduce a lot of data and suitably caveats the strength and weakness of the data.

Two striking things that Lehman does not address.  One, he acknowledges that many bad trends in big cities started increasing in 2015.  Two, the disconnect between big city data and suburban/smaller city/country data.

He has enough crumb trails to suggest that the trends capturing attention are big city trends but he does not elaborate why they are confined to big cities.  If, indeed, the trends are limited to select big cities, then that is meaningful information.  

Command-F on either Ferguson (2014) or Floyd (2020) returns zero results.  Yet the inflection points he is discussing correlates with the dates and aftermaths of both the Ferguson and Floyd riots, both of which were nationwide but typically concentrated in select large cities.  

Command-F on defunding also yields null results.  He does allude to police pulling back from enforcement but only in the context of cities being less willing to bring charges against perpetrators.  There is no discussion about the decline in big city police levels, the increase in churn, and the overall pullback of police enforcement out of concerns that Mayors or City Councils will throw police officers under the bus as has happened repeatedly since 2020.

All-in-all, despite the protestations that this is a complex story with many facets, all of which need to be examined in terms of validating or invalidating data, which is all true, this ends up being a salvage job on the proposition that there is no real crisis.  If the crisis is sourced in defunding and de-policing and you are not discussing the consequences of that, then the piece is not serious, despite the volume of data and the validity of some of the arguments.