Understanding crime data isn’t easy. Friends, news stories – even mayors and police chiefs – often present crime statistics in an oversimplified way that allows easy conversation. But it’s important to understand the forces at work behind crime stats to have a truly accurate picture of what’s really going on.
New York City crime rates are a classic case of tricky numbers. Over the years, various officials have claimed responsibility for lower overall crime rates, lower homicide rates, lower violent crime rates, and many other measures of success. In 2015, New York City’s commissioner tweeted and talked about record-low crime rates. But an investigation found crime rates were actually up over the previous year. How is this possible?
As it turns out, the New York City commissioner, police chief, and watchdog agencies were all using slightly different methods for tracking and reporting crime stats. For example, the commissioner’s office routinely excluded victims of gunshot violence whose clothing – but not bodies – were pierced by a bullet. This type of cherry-picking leads to juicy sound bites but doesn’t present an accurate picture of the data. Here are some factors to consider when learning how to interpret crime statistics.
More Neighbors, More Crime?
You might assume that an upscale, gated neighborhood would have less crime than a rougher neighborhood nearby. But when it comes to crime data, assumptions are often incorrect. The size – or more specifically, the population – of the neighborhood is important. Perhaps the gated community is huge and having more people simply means more potential criminals and crime victims. At the same time, it’s not safe to assume that a larger community will always seem to have more crime than a small one.
Compare the violent crime rates in Syracuse, NY and Walla Walla, WA. Syracuse has 110,000 more people than Walla Walla, yet Walla Walla has a higher per capita crime rate. Did you catch that “per capita”? It means population size was factored in. Syracuse might have more crimes in number, but Walla Walla has more crimes per capita – for its population size. Understanding how crime relates to population is an essential part of interpreting crime statistics.
The Effect of Vigilance
Supreme Court Justice Louis Brandeis wrote, “Sunlight is said to be the best of disinfectants; electric light the most efficient policeman.” Shedding light on crime, in the hopes of resolving and preventing it, can be a good thing. But it has an effect on crime data. When your neighbors report more crime, crime stats go up. It makes sense when you think of it this way, but it’s hard to remember when reading about, or discussing, crime rates.
Imagine looking at neighborhood crime rates as you choose your next home to purchase. Do you want to live in a neighborhood with a high reported burglary rate? Well, it depends. Perhaps the residents of the neighborhood are vigilant about watching for burglars and reporting burglaries to the police. Wouldn’t that be better – and safer – than a neighborhood where burglaries go unnoticed and unreported?
Data Collection Methods
It’s also vital to understand how data is collected and shared. Law enforcement professionals across the United States track crime in local precincts, report the crime to the state and share data with the Federal Bureau of Investigation (FBI).
In fact, the gold standard of crime statistics resides with the FBI in its Uniform Crime Reporting (UCR) program, which has collected and distributed crime information since 1929. It publishes four databases of crime stats annually, with data from 18,000 agencies. But even this trustworthy source comes with a caveat: It’s slow. Local reporting agencies, like police stations, often take months to provide information to the UCR, which in turn often takes until at least the fall of any given year to combine and produce nationwide data for the previous year. So it’s sometimes hard to know if you’re looking at up-to-date information.
There’s also the matter of variation in tracking and reporting.The FBI UCR has reporting standards, but not all agencies follow them to the letter. Each community – and each police precinct – has its own habits, culture, and standards for recording crime. For example, in a sparsely-populated rural area, an assault and battery in a local restaurant might be rare and certainly reported in a crime tally. But in a bigger city, where dozens of bar fights happen daily, police might simply break up the scuffle and leave it unreported.
It’s worth noting that some of the most powerful factors in crime are hard to record, making them almost invisible in crime data. A study by the Yale Department of Sociology found that demographic indicators like race, age, gender, and poverty – which are often included in crime stats – were far worse predictors of gun violence than a person’s social network – which isn’t generally a part of crime stats because it’s hard to measure.
So whenever you encounter crime statistics, run through a checklist in your head. Is it a trusted source? Where are they getting their information? Are they cherry-picking the data? Are there terms like “per capita” with a deeper meaning? Is a higher rate of something really bad, or does it just mean vigilant citizens are doing their duty by reporting a crime?
Crime Data Terms to Know
- Assault: Legally, an assault is a threat to carry out bodily harm without actually doing so. It is sometimes confused with battery, which is making physical contact.
- Average: Average usually refers to the mean – where a set of numbers is added, then divided by how many numbers there are. Averages can be deceptive because even widely spaced numbers create a middle ground. The average of 1 and 99 is 50. But there’s a big difference between 1 and 99 murders.
- Battery: Sometimes confused with assault, a battery is bodily harm against another person. “Assault and battery” means a threat was both made and carried out.
- Burglary: A burglary occurs when a person unlawfully enters a structure with the intent to commit a crime. It is often confused with robbery, theft or larceny.
- Demographic: Characteristics of a population like age, education, race, income, and gender are demographics.
- Descriptive statistics: These statistics quantify something, like the number of burglaries in a certain county.
- Firearms-related deaths/gun deaths: Look for the data behind these hot-button terms when you see them. Are suicides included? Self-defense situations?
- Home invasion: This is forceful entry into a private residence with intent to commit a crime upon the occupants, like rape, assault, murder or kidnapping.
- Homicide: The words homicide and murder are often used interchangeably, however, there is a difference. Homicide is when one person kills another. Murder is the legal charge of one person intentionally killing another, a form of criminal homicide. Some homicides are excusable or justifiable, like self-defense or a police officer killing someone in the line of duty.
- Inferential statistics: Unlike descriptive statistics, inferential statistics are an attempt to draw conclusions and interpret data.
- Larceny: The terms theft and larceny can generally be used interchangeably. Both are the unauthorized taking of property by theft or extortion – but not by robbery or burglary.
- Mean: The mean is the mathematical way of finding an average. See “average” above.
- Median: Often confused with mean and average, the median is the exact middle of a set of numbers. In the set of numbers 1, 3, 5 and 7, the middle numbers are 3 and 5, so the median is 4.
- Per capita: This is a term that accounts for the size of a population. It comes from the Latin “by heads” and means something is being stated per person in the group.
- Population: A population is simply a group of people. When looking at data, it’s important to know exactly which population is being discussed. Is it the U.S. population? The city population? A small population for a specific research study?
- Robbery: When someone uses fear or force to take personal property, it’s robbery – often confused with burglary, larceny or theft.
- Sample size: Every research study has a sample size – a representative group inside the larger group. For example, in a study of American attitudes about crime, the researchers might survey 20,000 citizens – not every person in the U.S. So the sample size is 20,000.
- Sampling process: The sampling process is the method researchers use to choose a sample group to interview or survey. It is a controversial part of research and data collection because it leaves room for bias, errors, and creation of misleading statistics. Faulty polls may have errors in the sampling process.
- Theft: Theft is a general term that includes the crimes of burglary, robbery, and larceny.
- Violent crime: Although this term is used to mean many things, the FBI gives it a very specific definition. Violent crimes are murder, nonnegligent manslaughter, forcible rape, robbery, and aggravated assault.