The purpose of the crime lab is to bring social science to bear on crime in O‘ahu to help its residents and visitors understand what’s going on. So, before we get into the monthly updates and analyses of local crime data, let’s start with some basic lessons from criminology and sociology about how to think about crime rates.
As students in my introductory criminology class can tell you, there are some key things to keep in mind when we talk about crime rates:
Variation in the crime rate for any given locale or jurisdiction is normal. We need to step back and look at the trends over time rather than focusing on what crime did in the last month, season, or year.
The advanced version: When I look at a change in the crime rate, I want to know if it’s within one or even two “standard deviations” of previous years’ crime rates (or really the mean of previous crime rates). Standard deviations are a statistical measure that help us distinguish between signal and noise—basically, is it within the margin of error (or the normal/expected fluctuation) or is it not.
Variation in the crime rate across locale or jurisdiction is also normal. There tends to be big variation across cities and between urban, suburban, and rural areas. Unfortunately, people sometimes think that just because the crime rate is really high in one place, that’s true of all places or all places “like” it. In fact, some politicians and journalists sometimes extrapolate from a few key places like Chicago or New York City to suggest crime is up or down countrywide.
Reported crime rates are not necessarily the actual crime rate. When we say “the crime rate,” we really mean the reported crime rate. But the reported crime rate depends on whether victims report crimes and how law enforcement agencies interpret incidents (different jurisdictions can count the same crime differently, but we can also find biases based on where or when a particular crime happens or based on other circumstances like victim or perpetrator characteristics. To get around these issues, we can also look at victimization surveys instead of just those crimes reported to law enforcement agencies. Homicide rates are also a particularly useful measure because there is less reporting bias (if someone is murdered, we’re more likely to see that reported than, for example, a theft or a sexual assault or embezzlement or hate crime). Indeed, some categories of crime are highly underreported or people can disagree or misunderstand about what counts as a given crime.
Things get really tricky when we look at the different stages a criminal act can go through as it progresses through the criminal justice system. Something that you and I might identify as a particular type of crime, say, burglary, might be charged as something that police or prosecutors think they are more likely to be able to prove such as criminal trespass. In some cases, crimes may shift between categories: a violent crime may become a property crime, and vice/versa. And as with the reporting, we also need to see that charges aren’t dropped at some point. So there is a fair bit of fuzziness even in official data just because of the nature of modern criminal justice.
“Crime rate” is kind of misleading. There is a big difference between violent crimes and property crimes or drug crimes, but often we just refer to the “crime rate.” Some sources use “index crimes,” or a list of about seven major crimes that we care about, like murder and burglary, to give a more holistic look at the crime rate. And, as you can imagine given the above-noted variation from place to place, even “national crime rates” are not always representative.
“Crime rates” and victimization rates also vary substantially by demographic categories. People’s actual crime risk varies a lot not just by whether they live in the countryside, a suburb, or the city, or in a given state, but also by age, gender, and race, among other factors. So even when the crime rate is up or down, different people in the same jurisdiction may be more or less safe, so generalizations can be tricky and misleading.
Each of these factors are not unique to crime rates, but are true of lots of “rates” including things like COVID case rates. You can think through the logic of this one by thinking about how most of the above points are also true for covid cases, or whatever is your favorite example.
The Hawaii Crime Lab 