Fallacies

From high school though university, so much of that which is taught in formal schooling on the topics of “fallacies” and philosophy condemns students to the rejection of good instincts and the broad dismissal of general rules or conclusions. From lectures on “correlation doesn’t equal causation” to the notion that all things are subjective in a world of artificial constructs, the “lessons” often spoil the students’ minds, distance them from physical reality, and keep them preoccupied in a dizzying field of abstractions where nothing is known and nothing can be known. 

Ultimately, despite the need for some basic understanding of fallacies in order for students to think critically, so much of the focus overemphasizes the ways in which ideas and conclusions can be quickly dismissed or “rationalized” away, not for the identifiable errors within the ideas or conclusions but for their mere imperfections. Such is the case for essentially any idea or conclusion. This point is proven by the way in which people often treat data and statistics, leaning on their rough understandings of “fallacies” and other such tools in order to prevent themselves from having to otherwise accept inconvenient truths. 

Indeed, one such example is observable in the way that people often reject statistics by claiming that “correlation doesn’t equal causation”, or — in a specific example of clever “rationalizing” — the claim that “people commit more crimes simply because they live in bigger cities”. While, at first look, one might say, “Hey, isn’t that an instance of conflating causation with correlation?” And yes, it is, but you see, this is a common technique among the incredulous who, in identifying any unpleasant “correlations” (in this case, ones dealing with uncomfortable figures on culture and demographics), set out to overcome them with their own fallacies. You see, it is not a matter of the incredulous adhering to the principles consistently or universally, but rather using them selectively to overcome or overrule ideas and conclusions they do not like. This is consistent with the way they generally conduct themselves: rules for thee, but not for me.

Let us address the proposition directly: the idea that “people commit more crimes simply because they live in bigger cities.” But first we need to deal with the idea that “correlation doesn’t equal causation”.

First, the principle is that “correlation doesn’t equal causation”; it is not that correlation never aligns with causation. If we are to accept that correlation can never equal causation, then statistical analysis can never become actionable. While it is certainly true that correlation doesn’t independently equal causation, that does not imply that causation cannot be reasonably inferred. Indeed, statistical analyses are actionable only for the inferences that we are able to draw from them. Data and statistics, in and of themselves, lend nothing more than a measure of certainty; they don’t, in and of themselves, tell the story. 

As for the specific case at hand, there is absolutely no reason that we ought to accept that people in bigger cities will necessarily murder at higher rates than citizens of smaller cities or towns; and, were it true that the size of a city’s population generally predicts its crime rates, then we would see crime rates scale along with population size across the globe, but (of course) this is not consistently the case. While it is true that, on balance, murder is generally more likely in bigger cities, that says nothing about the causes, let alone the many exceptions. Put another way, it is reasonable to expect that there will be a greater number of crimes in the city, but (all else equal) there is no reason to assume that the rate of crime should be higher — unless there is a distinction to be drawn between the types of people who live in bigger cities as compared to those who live in smaller towns. After all, the rate already affords an advantage to the bigger cities, given that they have far more people to offset the exceptional criminal event. 

If the rate is already assumed to fairly compare towns and cities irrespective of population size (which is the very purpose of the metric to begin with), then we can safely assume that there is some distinction between smaller and larger cities that can account for the propensity of criminal behavior. After all, the exceptional event in a small town can constitute a rather significant percentage of the total population, whereas residents of the larger city must take it upon themselves to overcome the inherent scale of their population to match, let alone exceed, that same rate. So, what is it about those particular cities that makes it more likely for people to commit crime there? Is it just the fact that people are more likely to bump into each other in the city? Or is it that people who live in the city are just inherently more violent? Neither. If either of these were necessarily the case, then the rule would be more definitive, with similar crime rates being shown among all cities of similar population size. This leaves the following question, so far as we examine the human aspect alone: Are there cultural or demographic distinctions that can explain the difference? The data suggest that the answer to this question is a resounding ‘yes’.

Ultimately, there are many cities in the United States, and across the world, which suffer a disproportionate amount and frequency of crime, and there is hardly a perfect correlation between the rate of crime and the population of the city. Indeed, in the United States, there is a far more conspicuous correlation between demographics and crime that cannot be discussed here, or anywhere in the light of day, without running the risk of insulting somebody. 

The truth is that, in America, demographics do a better job of predicting criminal outcomes than the bare size of any city’s population. For instance, in America, a white person is far more likely to die at the hands of a black person than a black person is likely to die at the hands of a white person. Put another way, blacks murder significantly more whites, by both quantity and percentage, than whites murder blacks. 

In fact, white offenders are responsible for 8 percent of all black homicide victims. Meanwhile, black offenders are responsible for 16 percent of all white homicide victims. Upon adjusting for the population differences between the two races, we reach a factor of 11: this means that the average black person, in the event of any homicide, is eleven times as likely to murder a white person than his white counterpart is likely to murder a black person. 

Take into consideration black-on-black murders in addition to those described above and you’ve already accounted for a majority of the murders that take place in the United States: black-on-black homicide accounts for 90 percent of all black homicide victims. What’s more, despite the fact that blacks constitute only 12 percent of the population in the United States, they are responsible for roughly half of all homicides. 

All in all, over the long term, the data bear out a very different tale than that which has been parroted by people who either control the popular narrative or prefer the comforts of “political correctness”. Despite the data and despite the most reasonable inferences, people are incredulous and steadfast in their preferred beliefs, even in the face of incontrovertible fact. Meanwhile, their high school and college experiences have equipped them to “rationalize” their positions so that they can continue to conveniently dismiss inconvenient truths while fancying themselves enlightened all the while.