Math and reality conspire against the preferred SJW narrative

Yesterday I commented in Insight Backed By Data that 60% of homicides are committed in the 62 largest cities which have about 20% of the US population.

In today's Washington Post there is an article, Violent crime is rising. But that’s not the most provocative finding in the FBI’s big new report. by Max Ehrenfreund.

Murders in the United States jumped 11 percent last year, according to federal data released Monday, but nonviolent crimes declined, an unusual divergence that's puzzling criminal justice experts.

While an increase in homicide is usually associated with more minor crimes as well, that was not the case in 2015. The number of murders nationally increased by the largest percentage in decades, but violent crimes overall increased just 4 percent and property crimes declined 3 percent.

[snip]

Still, what to make of the sudden increase in homicides is not clear. Some criminologists say the data is evidence against the "Ferguson effect" -- a popular theory that suggests homicides have increased because police have become reluctant to interact with potential criminals on the street. According to this argument, cops fear becoming involved in a violent altercation that could result in protests such as those in Ferguson, Mo., Baltimore and Charlotte, N.C.

Those of the left have long been in denial that there is a Ferguson Effect despite the striking rise of homicides in cities that have reined in policing activities. I see the evidence as pretty strongly indicative that there is a Ferguson Effect but Ehrenfreund raises an interesting objection.

There is no inherent reason that violent crimes (murder) have to be correlated with non-violent crime (such as burglaries) but I am willing to take it as a reasonable assumption that there is a general correlation.

Police can deter potential criminals not just by being a watchful presence on patrol. They can stop people who appear to be involved in criminal activity, talking with them to gather information or to disperse people who are fighting. Police can also search civilians for firearms, knives or tools for breaking locks and windows.

These activities make committing all kinds of crimes more difficult, not just homicide. If the increase in homicides were due to hesitance on the part of police to stop civilians, some criminologists say they would expect an increase in other "street crimes," including burglary and robbery.

However, just 3.1 percent more cars were stolen last year, and the number of robberies increased just 1.4 percent. The number of burglaries declined 7.8 percent, according to the new federal data.

That is a reasonable objection to the Ferguson Effect.

Is there an explanation? There might be others but I suspect that it has something to do with yesterday's post.

Let's take it as settled that the predominant portion of murders occur in city environments. The numbers suggest 60% of murders occur where only 20% of the people live. What about property crimes? Violent crime is a fraction of all crimes (fortunately).

Here is my rough speculation. If most of the violent crime is happening in a circumscribed area among only a small portion of the population, you might have quite different correlations between property crime and violent crime in the two areas.

I am guessing that the secular trend of declining crime in the rest of the country continues. Policing practices there have not changed materially and the secular trend dominates anyway. Both violent crimes and property crimes decline.

In the dense areas within city limits, where there has been a trend to reduce policing, there has been a sharp increase in murders. Because those areas dominate the number of murders, whatever happens to the murder rate in cities is going to dominate the rate for the nation as a whole. A number-based example likely will better illustrate what I am thinking.

Let's assume for purposes of an example, that there are 100 murders in the country and that 90% of them occur within city limits and 10% of them occur everywhere else and there is an overall secular decline in all crime, including violent crime. Cities are 20% of the population and everywhere is 80% of the population. Let's further assume that police in the cities, because of Ferguson-like riots and protests, adjust their policing practices so that they no longer stop-and-frisk, no longer proactively attempt to identify suspects, reduce presence on the street, etc. As a consequence, the murder rate within the city rises by 10% whereas in the rest of the country, the secular trend continues and there is a reduction of 10% in the number of murders outside the city.

City murders therefore increase from 90 to 99 and murders everywhere else fall from 10 to 9. Total number of murders is now 108, i.e. an 8% increase in the murder rate while 80% of the population continues to experience a 10% decrease in violence.

Now what about property crime? Let's assume that there is the same positive correlation between violent crime and property crime everywhere but the ratio of violent crime in the city is different from that everywhere else. Let's assume that 50% of property crime is in the city and 50% is everywhere else. In year 0, in the city, there are 90 murders and 50 property crimes whereas everywhere else, there are 10 murders and 50 property crimes. In the city, there is a ratio of 1.8 violent crimes to property crimes whereas everywhere else there is a ratio of 0.2 violent to property crimes.

If there are the same trends as above (city increases by 10% and elsewhere decreases by 10%) then property crimes increase to 55 in the city and decrease to 45 elsewhere but the overall volume of property crime remains the same. Property and violent crime are equally correlated in their respective areas but are differentially consequential.

Total crime (violent and property) has gone from 140 (90+50) in the city to 154 (99+55) in the city where 20% of the population lives whereas 80% of the population sees their total crime going from 60 (10+50) to 54 (9+45). Overall crime for everyone goes from 200 incidents (140+60) to 208 (154+54), an overall 4% increase.

In this hypothetical example, the mystery of Ehrenfreund's paradox is resolved. Reduced policing in the cities can lead to an increase in overall crime and especially an increase in violent crime without a corresponding increase in property crime. It all depends on the relative rates of crime between property and violent in the city as well as the degree of concentration of crime between city and everywhere else.

Now whether or not the actual numbers bear this out is another question for which I do not have time at the moment to document. I am pretty comfortable with the estimate from yesterday (60% of violent crime occurs among 20% of the population). I am also pretty comfortable with the assumption that the ratio of violent crime to property crime is higher in cities AND that the ratio of total property crime is probably more balanced between the city and everywhere else.

Under the circumstances then, the statistics are absolutely feasible that violent crime (concentrated in the cities) rises while overall crime and particularly property crime decreases.

That also remains consistent with the more causal explanation of the Ferguson Effect, i.e. cities which reduce their policing do actually see an increase in the violent crime rates.

It seems to me that the Ferguson Effect is alive and well despite the efforts to disguise or hide it.