Concentrate Where the Murders Are Concentrated
That such a principle is well established is illustrated by t Edgar K. Browning and Jacquelene M. Browning’s textbook, Public Finance and the Price System, which I used when teaching my first such class over four decades ago and which said, “The key issue here is the geographic area over which persons necessarily benefit [or are harmed],” which requires that “care is needed in determining what types of policies are more suitable for local governments.”
However, that principle is often honored in the breach today, as politicians at higher-level governments are always trying to regulate and legislate issues that are more local in character. Why? It lets politicians in areas where the problems are greatest pretend they are a national problem rather than ones tied to their jurisdictions and policies. Further, the power to vote on national-level plans gives politicians representing other areas the leverage to “rent” their support for such programs in exchange for more of what they want through the legislative pork barrel.
Just think how many times a single event in one place starts trending, then immediately gives rise to proposals for new state or national policies as “the solution,” as is so common with issues of crime. The Monterey Park mass shooting is a good illustration. The same day it was reported in the Los Angeles Times, they ran an editorial about mass murder shootings becoming “a sickeningly frequent occurrence in America” arguing that mass shootings “have one thing in common: They have guns” and asserting that we must limit the Second Amendment in the US Constitution—not only federal law, but the highest law of the land—because “national suicide is not the compulsory price of freedom.”
The result of such broad, national responses is also poor “target efficiency,” because too little attention focuses on the more local reasons for where the problems are worse.
An excellent example of this is provided by recent research on the US murder rate by the Crime Prevention Research Center, and its president, John R. Lott Jr., whom I have known since we overlapped many years ago in the UCLA Economics PhD program. I would note that John’s work is often controversial, which also makes him a frequent subject of ad hominem attacks, because the empirical data he develops can strongly contradict what others are “selling” as the truth in some area, particularly with regard to crime. However, I have never seen him abuse logic and statistics to get a particular answer he set out to find (or was paid to, as many “researchers” are). His focus, which strongly reminds me of the work of Harold Demsetz, who taught both of us, is on designing empirical tests to differentiate among alternative explanations, then following where the evidence leads, rather than torturing evidence to create the “right” wrong answer.
Increases in homicide rates tend to be treated by state and federal politicians as if they are broadly distributed national problems to scare Americans into supporting overly broad-brush “solutions.” But Lott’s research shows instead that “homicide rates have spiked, but most of America has remained untouched.” Or as David Strom summarized the results, “There are vast swathes of the country where violent crime is very, very rare, and small areas of the country where it is common.” If that is true, we should focus our attention on those small areas, not on national policies poorly focused on where the actual problems are most severe.
Lott’s research, which used 2020 homicide data, examined the concentration of homicides in particular areas to see whether America’s increasing homicide problem is national or local. He let that data tell its story.
First, he focused on county-level data rather than national data. Some of the dramatic results he found:
- The worst five counties (Cook, Los Angeles, Harris, Philadelphia, and New York) accounted for about 15 percent of homicides.
- The worst 1 percent of counties (31), with 21 percent of the US population, accounted for 42 percent of the homicides.
- The worst 2 percent of counties (62), with 31 percent of the population, accounted for 56 percent of the homicides.
- The worst 5 percent of counties (155), with 47 percent of the population, accounted for 73 percent of the homicides.
- In contrast, over half of US counties (52 percent) had zero homicides in 2020, and roughly one-sixth of the counties (16 percent) had only one.
Continuing his investigation, Lott looked at even finer-scale zip code data for Los Angeles County. He found that the worst 10 percent of zip codes in the county accounted for 41 percent of the homicides, and the worst 20 percent accounted for a total of 67 percent of the homicides.
From such data, Lott concluded that: “Murder isn’t a nationwide problem.” Instead, “It’s a problem in a small set of urban areas, and even in those counties murders are concentrated in small areas inside them, and any solution must reduce those murders.”
Despite the constant political and media drumbeat to portray homicides as a national problem that threatens everyone everywhere, and thus demands national solutions in line with what the political Left wants, the evidence points us in a far more local direction.
That may well explain the political reason for the volume and persistence of that drumbeat. It provides camouflage for those whose policies (and those who support them) would come under far greater scrutiny if people recognized just how concentrated homicides are and then asked what is different in those places, rather than the “blame America first” bromides they are routinely misdirected toward today.
But that means if we really cared about those most harmed by the murder rate, rather than imposing broader-than-necessary restrictions on Americans, it is important to follow the evidence so many would prefer to keep hidden.