The proposed Ordinance No. 2019 raises some troubling questions, particularly regarding subsection 9.942. This subsection makes a civil infraction of any calls to the county’s 911 system — apparently only having jurisdiction over situations where the GRPD is dispatched — where the caller has engaged in bias and where no clear
crime has been committed.
Among many other problems with this legislation, it seeks to literally fine
some uses of 911. It is an ironclad law of economics that when you tax,
fine or penalize something, you get less of it. When you incentivize
something, you get more of it. Fining the reporting of crime will result in
less reporting of crime, which will result in more crime.
The language of the ordinance also states that using “protected class
status” as criteria in a judgement on whether or not to report a crime “in
whole or in part” would leave the crime reporter open to a $500 fine. As I
will show, this amounts to literally outlawing rationality, proclaiming
that some of the strongest prior information regarding the likelihood of
criminal activity must be ignored.
The ordinance also appears to saddle ordinary citizens with a more onerous
burden in evaluating potential criminal acts than the responsibilities of
the police themselves. Indeed, requiring people to make expert judgements
about the legality of events they are witnessing that, in some cases, may
be life or property threatening is shifting one of the chief functions of
the police onto ordinary citizens. Specifically, the language of the bill requires ordinary citizens to correctly determine whether or not reasonable suspicion is present before they call 911 or any city authority. Failure to accurately assess whether reasonable suspicion is present, something that requires police to undergo intense study and field experience to master, will result in a $500 fine.
And a $500 fine in a country where half of all Americans couldn’t make rent or eat in the event of a $500 emergency is a serious disincentive to ever use emergency services. This is a highly effective means of dramatically increasing violent crime, bringing to mind
the grisly consequences of the Ferguson Effect, which radical leftists love nothing more than to pretend doesn’t exist.
There appears to be no American precedent for such an ordinance, although
there is choreographed recent push among radical neo-Marxists to push
through similar legislation elsewhere. It’s a bad idea.
A medical analogy
The Mayo Clinic’s website says the following on lung cancer screening
tests:
“People with an increased risk of lung cancer may consider annual lung
cancer screening using low-dose CT scans. Lung cancer screening is
generally offered to people 55 and older who smoked heavily for many
years and are otherwise healthy.”
Why would a test that costs hundreds of dollars to perform be done on
otherwise healthy people? A naïve observer might conclude that it was a
simple ploy for doctors and hospitals to rake in more cash for unnecessary
tests.
But it turns out that such testing has a firm rational basis. On top of
being one of the deadliest forms of cancer, lung cancer is also one of the
most prevalent. The chance of being diagnosed with lung cancer in a given
year is 73 in 100,000, more than twice the chance of an American being killed in a car crash.
But lung cancer isn’t car crashes. Virtually no 10 year olds get lung
cancer. Even those in their 30s who are heavy smokers almost never develop
the disease. So, who gets lung cancer? Those over the age of 55 who are also smokers. For this unfortunate group, their chances of being diagnosed with lung cancer in any given year are an astonishing 2.1 percent.
If doctors started ordering all 10-year-olds or even heavy smokers under
the age of 30 to undergo low-dose computed tomography, we would rightly
suspect them of padding their income with unnecessary testing because
virtually no one in these groups will get lung cancer. On the other hand,
when concentrating on those 55 years and older who are heavy smokers, if
even a quarter of those with cancer get their disease caught in time for a
cure, it means that, given a $250 test, each life saved will have cost just
$50,000. In the medical screening literature, this indicates a highly
cost-effective screening method. Considering that the typical course of
lung cancer treatment runs around $150,000, it is clear that this screening
method is an easy sell on a cost-savings basis alone.
Here, the rational choice is to continue screening smokers 55 years and
older. And in fact, this is what medical facilities across the country are
doing, saving thousands of lives each year in the process.
Bayesian inference and general diagnostics
What happened above may seem obvious, but it bears breaking down further to
illustrate the general principles at play. An expensive test was ordered across an entire population. This test has real costs, as the vast majority, 98 percent, of those who undergo the test will be determined not to have cancer. The cost, approximately $250 per test, is
nontrivial. In a badly strained healthcare system that struggles to even
deliver basic care to most Americans, it is not at all obvious that
spending $250 per person on a group of people where the test will have
effectively been a retrospective waste 98 percent of the time is a good
deal. But through rigorous analysis and careful statistical logic, it turns
out that ordering the test is not only defensible, it produces astonishing
savings while sparing many from a terrible, protracted death.
This can be slightly restated as a general principal of diagnostics, namely
that group characteristics gave a strong enough index of suspicion that an
expensive screening test was called for. Specifically, only two group characteristics were needed in order for our doctors to recommend the test: being over 55 and being a heavy smoker.
This is the concept of Bayesian inference, which allows the diagnostician
to use all available information to make the best possible judgement about
the given circumstances. In the case above, it would have been ridiculous
profiteering to order the test for all patients. But the fact that any
given patient was over 55 and a heavy smoker gave doctors a strong Bayesian prior, statistical evidence strong enough to warrant ordering an expensive test.
Keep in mind that 98 percent of those who underwent the test were found not
to have cancer. Taken alone, this number might seem like weak evidence. But
it becomes incredibly strong evidence when we consider the entire decision
tree. In decision-theoretic terms, we can say that, given our Bayesian
priors, administering the test has a higher expected value than not testing. And in this case, the value of testing is far higher than not testing.
Swapping variables
So, why am I boring you with this disquisition on lung cancer? Because the
concepts of Bayesian inference, expected value and their role in
diagnostics are powerful tools that are used everyday to address an
astoundingly wide array of problems. It turns out that the specific pattern
articulated above can be generalized and adopted to a practically infinite
number of problems by just swapping out a few variables.
Take the lung cancer example. Instead of the $250 LDCT test, lets replace
it with a 911 call. Instead of people in a doctors’ office, lets imagine
all people who might be in a suspicion-inducing circumstance. And instead
of those over the age of 55 who smoke heavily, let’s use black men between
the ages of 15 and 40. Finally, instead of lung cancer, lets use the
horrifying outcome that we are trying at any reasonable cost to avoid as a shooting.
But first a few words on relative risk and black crime.
The shocking realities of black urban violent crime
Even in the 1890s, black writer W.E.B. DuBois noted the highly elevated rates
of black crime within American cities, which he notes stretched all the way
back to the 1860s in Northern cities.
Today, that trend not only persists, in some cases, it has gotten far
worse. In 2013, California blacks committed all categories of violent crime
at 5.35 times the white rate. In 2014, black arrestees in New York City for the very
serious crime of shooting — an incident where a firearm is discharged and
at least one person is hit by the bullet or fragments — were
overrepresented by a phenomenal factor of 98 times the white rate.
While New York City suffered from a particularly violent black population,
other cities saw huge multiples for the crime of shooting as well.
The Federal Bureau of Prisons reports that in 2013 blacks were incarcerated
at 13 times the white rate for the neighborhood-destroying crime of armed
robbery. In fact, across the entire country, blacks are vastly
overrepresented as arrestees, suspects and convicts in every major crime
category. What’s more, all of these are likely to be underestimates because crime clearance rates in the neighborhoods where most blacks reside tend to run from substandard to dismal.
Here in Grand Rapids, we have a conspicuous problem with shootings, the
same category for which blacks are overrepresented in New York by the
jaw-dropping factor of 98 times. Offender information is not readily
available, and I don’t have time to conduct a research project. But a quick
glance at what information is available indicates that the expected
national trends apply to Grand Rapids as well: heavily white East Grand Rapids has a crime index of 110.7 while considerably less-white Grand Rapids has an index of 315.6, indicating nearly three times as much crime. Every area that is whiter than Grand
Rapids has a lower crime index.
While statistics from New York City are probably among the most accurate in the country , the black overrepresentation among shooting arrestees of 9,800 percent is
abnormally high. Blacks in New York are particularly violent. However,
because 1) the clearance rate in black neighborhoods for shootings can be
assumed to be almost nonexistent in many major cities 2) I personally know
for a fact, having been witness to various shooting incidents, that most
don’t even make the news in Grand Rapids and 3) the fact that statistics
from other cities strongly confirm the general trend for urban black
violence — for example, Chicago blacks commit murder at around 30 times the
white rate — we can assume that the same general trends apply here, if at a
lesser magnitude.
Relative risk
All this means that for the crime of shooting, we can reasonably
conclude, for the sake of illustration, that blacks are offending at
somewhere in the vicinity of 35 times the white rate. It should also be
noted that the crime of shooting is extremely serious. It makes
neighborhoods completely unlivable. Most middle-class people of all
races find regions where shootings are common to be completely intolerable,
taking extraordinary measures to flee. And their reasoning is sound, around
13 percent of people who are shot die. To most residents, risking death
simply by leaving one’s home is unacceptable.
So, when someone sees a generic black person walking down the road in their
neighborhood, they know that statistically, they are 35 times more likely
to encounter the potentially fatal circumstances that a shooting entails
than if that generic person were white. Are they acting reasonably if their
natural reaction is fear and suspicion?
Many people encounter deer while out walking in wooded, secluded or even
suburban neighborhoods. Few elect to call the authorities when the presence
these relatively harmless animals interrupts a stroll. However, those who
encounter black bears generally dial the authorities.
Black bears are almost certainly more dangerous, one-on-one, to humans than
deer. But a one-on-one encounter with a deer is far more dangerous than
many realize — deer are actually the most dangerous animal in the country,
by a very wide margin, but this is due to mostly to auto accidents.
Yet, in just one month, there were four serious reported deer attacks
in California, one of which was fatal. There are 100 times as many deer in
the U.S. as black bears. But black bears have only attacked seven people
since 2010.
It is safe to say that black bears are more dangerous than deer but
probably nowhere near 35 times more dangerous. Should people, then, who
call the police when sighting a black bear be chastised for doing so?
Should they be fined $500 since they almost certainly wouldn’t make the
same call when spotting a deer in their neighborhood? The question is
absurd.
But the relatively docile and timid black bear is, in actuality, only
moderately more dangerous than deer. To get anywhere near a factor of 35
times as dangerous as deer, we would need to turn to the black bear’s
dramatically more dangerous cousin: the grizzly bear.
With just one tenth the black bear population, they have been responsible
for 15 attacks since 2010, with attacks often producing fatalities and
involving multiple victims. This would put the grizzly at a minimum
relative risk of 20 relative to the black bear, and this would probably put
grizzly encounters in the ballpark of 35 times the relative risk of deer,
although these are very rough estimates, and the real factor could be
considerably higher or lower.
But whether it’s 35 times or 80 times or 15 doesn’t matter all that much.
The bottom line is that grizzly bears are superlatively violent and
dangerous when compared with deer. So, the question: Should people who
encounter a grizzly bear in their neighborhood be fined $500 for calling
the authorities? Is there any possible conceivable scenario in which fining
people for calling the authorities on one of nature’s perfect killing
machines would be justified because that same person would not call the
authorities when spotting a whitetail?
Disaggregating the data
But all this has a very unseemly feel. Are we really comparing black people
to grizzly bears? No. The fact is that “black people” is not a useful
category because, like any other group, black people comprise a wide and
colorful tapestry of heterogeneous humanity. If someone claimed a fear of
well-dressed middle-aged black ladies, that fear could rightfully be
labeled as irrational, phobic or even racist.
Prime crime age
But this leaves an even bigger problem: If college-educated professional
black men and 50-year-old black church ladies are committing virtually no
crime, then how in the heck do we end up with a black shooting multiple of
35 times the white rate?
The answer is that underclass black males between the ages of about 15 and
40 commit violent crime out of all proportion. This prime-crime age has
long been recognized by criminologists, roughly mirroring in criminal males
the progression of fertility in females — a sharp rise in the teenage
years, peaking around 20, followed by a slow dropping off throughout the
20s, which accelerates in the 30s, finally reaching the age of so-called
criminal menopause at some point in the 40s. Beyond this point, only the
most hardcore offenders continue recidivating.
This is both good and bad news. It means that the large majority of black
people have a right to expect that they will not be legitimately feared or
treated with skittishness or suspicion. However, it also means that black
men in the 15-40 age range pose such a standout risk to public safety that
people viewing them with deep suspicion, particularly where other factors
are present, is not only justifiable: It would be wholly irrational not to view them as potential threats.
The full 911 call cost-benefit model
So now, let’s take a look at how the exact same model that doctors use to
help diagnose lung cancer plays out when applied to the ability of citizens
to call 911. In the model, we are substituting the group of heavy smokers
who are 55 and over with men between the ages of 15 and 40 who are black.
We are substituting low-dose computerized tomography with 911 calls. And we
are using as our Bayesian prior the strong information that we have
concerning the above demographic’s propensity for criminal shootings.
One last aside: the relative risk of lung cancer is “just” 15 times in
smokers what it is for non-smokers. The relative risk of a black male
between the age of 15 and 40, let’s call them black youths from
this point onward, perpetrating the dyscivilizational crime of shooting
versus a random white person doing the same is massively higher than that,
probably well over 100 times. Furthermore, I estimate that the average
violent offender commits 2.28 serious violent crimes per year*.
This means that a black male of prime-crime age on any given day has a 1 in
160 chance of violently victimizing someone — again, I’m just using
shootings as a representative violent crime. This is about .6 percent, less
than the 2 percent per-instance chance of a member of our 55-plus, heavy
smoker cohort above, but not a whole lot less. Note: this is a very rough estimate that could be off in either direction by an
order of magnitude — I simply don’t have time to research and build an
accurate estimate. But it’s within a realistic ballpark and will serve for
illustration.
Remember that above, the final step in justifying the diagnostic measure of
ordering a costly test was constructing what decision theorists refer to as the payoff
matrix. Simply put, not running the test over the entire population
resulted in far less savings than running the test gained in the aggregate.
So, what’s the payoff matrix here? Well, it’s the same answer that Anton
Chigurh gave when the rural store clerk in “No Country for Old Men” asked
what he stood to gain by correctly calling the coin toss.
“Everything.”
But for now, we’ll ignore that crucial conclusion, which is the entire root
cause of the decades-long white flight across the country.
We could formulate things this way: The median age of a U.S. citizen is 38.
The median retirement age is 67 and the median income is $31,000 per year.
In 2018, Chicago had 2962 illegal shootings that resulted in 495 deaths , giving a kill rate of 17 percent.
For simplicity, let’s just use the shooting mortality rate as a proxy for chance of death in all violent crimes. Let’s also stipulate for simplicity that the homeowner is equivalent to any potential neighborhood victim in order to avoid having to deal with complexities like the value of social capital — that is, the homeowner is exactly as concerned about the lives of all his neighbors as he is about his own.
Our suspicion-stricken white homeowner from above — unisex here, men expect to lose more — can calculate that their expected loss from not summoning police intervention, in purely monetary terms, is expected to be (.17*.006*31,000*29)/3= $305 (let’s just say that the criminal encounters three potential victims each day) on each and every
encounter with the prime-crime black that they don’t call the police
outright. This is their payoff. And this is with no further information. Again, $305 is the expected loss incurred by not calling the police based solely on the fact that a black male of prime-crime age has been spotted in the neighborhood.
So, what happens when the homeowner calls 911? What is the cost, and what
is the payoff? To our homeowner, the cost is effectively zero. To our
prime-crime wanderer, probably effectively nothing as well. But our
homeowner expects a long-term stochastic gain, or savings, of $305 every time he calls the police.
The correct move is for the homeowner to call 911, every single time. And it’s not even close.
Conclusion
The above argument made some ridiculous assumptions, including that life is
worthless other than future income. In reality, any violent crimes that
frequently result in death are antimatter to civilized society.
Law-abiding, productive people will not tolerate areas where these crimes
are endemic and, history shows, will go to extreme measures to completely
extract themselves from such regions.
Our argument also neglected the near certainty that anyone who is involved
in a violent incident is likely to incur serious costs, including both
monetary and psychological, the 83 percent of the time that the encounter
doesn’t result in death.
The subsection of the proposed ordinance that motivated this letter states:
“Sec. 9.942 – Biased Crime Reporting 1. No person shall knowingly or
recklessly report to a City police officer, City dispatcher, or other City
personnel that an individual who is an actual or perceived member of a
protected class as identified in Sec. 9.935 of this Ordinance has
committed, or may or will commit, a crime, if such report is based in whole or in part on the individual’s membership in a protected class and not on a reasonable
suspicion of criminal activity in consideration of all available facts and
the totality of the circumstances. 2. This Section is enacted for the
purpose of ensuring the public health, safety, and welfare of the City’s
residents.”
The argument above is making very weak assumptions while assuming that the
ordinance is being violated in the strongest possible way, namely that the sole criteria used in deciding to call the authorities are the suspect’s skin color and gender (age is also used, but the argument will still hold even if age is left unspecified due to the
dramatically higher crime rates of men versus women).
This means that if we conclude that the above person is justified in
preemptively calling the police, then they must be at least as justified in calling the police in every other possible scenario covered by the ordinance. That is, if a person can be
shown to be justified in calling the police on a suspect based solely on
their skin color, gender and age, then there is no possible circumstance in
which criteria additional to those two in the deliberation on whether or
not to call the authorities could be construed as unsound.
Above, we indeed showed that not only would a homeowner be completely
economically justified in calling the police based solely on the criteria
of skin color, gender and age in the case of black males between the ages
of 15 and 40, but that doing so at every opportunity
would provide significantly higher long-term expected value than not doing
so, namely $305 per 911 call.
We also showed that any law that fined people $500 for calling the
authorities when they spotted a grizzly bear is beyond comically absurd
even if that same group would rarely call the authorities when sighting a
deer. However, black males of prime-crime age are almost certainly even more dangerous relative to a generic white person than
grizzly bears are to deer. If this is true, then it must be even more
absurd to enact a law fining people $500 for calling the police on this
demographic even when no other indicators of potential criminality are present.
This may sound outrageous, but it is in fact what underpins the phenomenon
of white flight — people generally find violent crimes, especially armed
robbery and murder, to be so intolerable that they are willing to vacate
homes which have been in families for generations, taking a huge losses and
incurring all the hardships of uprooting their lives, just so that they can
get away from these civilization-destroying crimes.
The very obvious subtext of this proposed ordinance is anti-white: It’s aim
is not to discourage Korean store clerks from calling the police on Indian
cardiologists. The aim is to prevent whites from exercising rational
judgement and calling the police on black males of prime-crime age. And
this is precisely the type of hairbrained radical leftist effort that has
reliably produced countless cataclysmic urban hellscapes, from Detroit to
Camden, New Jersey, to St. Louis to the Chicago.
Creating an ordinance whose stated goal is to penalize the reporting of
crime — what you fine, tax or penalize you get less of — and thereby increase crime is a boneheaded abuse of power. But to outlaw
Bayesian inference in the purpose of keeping people’s neighborhoods livable
is to outlaw rationality in the service of pure evil. If that is the
directive of Grand Rapids city leadership, then they are unfit to lead.
Napkin math used in example
*Here, we need to determine a rough figure for how likely a single
encounter with a black youth is to result in a violent victimization. This
would require fairly in-depth research. So, I am going to use extremely
crude napkin math. First, one third of
black men have felony records. Some of these are expungable, but let’s
ignore that. This is a case where the number should theoretically be
actuarially stable — the population is so large that black felons who die
are replaced by equal numbers of new black felons at the bottom end of the
age range with high predictability and regularity.
Let’s also take an old policing axiom, “Either you don’t do crime or crime
is all you do.” What this means is that those who commit one violent felony
while in the prime-crime age range are likely to be active criminals.
Because criminals typically show high versatility — engaging in many types
of crimes — and most criminals who have offended seriously enough to land
in prison are habitual, it is probably safe to assume that around 75
percent of felons are at least capable of violence. All of this means that
around 25 percent of black men between the ages of 15 and 40 are likely to
engage in violence, given the right circumstances.
I’m not an actuary and I’m not getting paid to write this. So, I’m going to
sacrifice accuracy to get an estimate without resigning myself to 40 hours
of work.
80 percent
of prisoners are rearrested within nine years of release. This leaves a
chance of not getting caught in any given year of 83 percent
(1-(.83^9)≈.8). This means the average criminal will get caught every 3.72
years (.5=.83^3.72). The clearance rate for murder in Chicago is 12
percent. A general trend is that more serious crimes tend to have about
twice the clearance rates of
less-serious crimes
.
This means that the average criminal is probably committing something like
100/6=16.7 violent crimes before they get caught. And the average violent
criminal has (50-15)/3.72+3.58=4.79 periods of potential criminality (3.58
years is the average sentence served for violent crimes and I chose age 50
to adjust for the fact that some criminals go on offending until death).
This all adds up to 4.79*16.7 violent crimes committed over the course of a
35-year criminal career or 2.28 violent crimes per year.
Dialectic Rec Yard 