Earlier today, the FBI released the latest crime figures, including 2017 estimates for rates of reported crime based on information provided by police departments across the country. 

The Uniform Crime Reports (UCR), as it is officially known, is an important bellwether for crime and safety in the United States. In recent years, the release of the FBI data has been accompanied by warnings that violent crime is increasing and has been used to call for hard-line policies to combat spikes in offending. However, as Vera has reported in previous years, the overall trend has been (and continues to be) one of a steady decline in crime rates. If we take the long-view, the major crime categories tracked by the UCR that receive the most attention (property crime, violent crime, and homicide) are at much lower rates than their peak in the early 1990s. Property crime per 100,0000 people is down 54 percent (from 5,140 in 1991 to 2,362 in 2017), violent crime per 100,000 people is down 46.5 percent since its peak (from 716 in 1991 to 383 in 2017) and homicides are down 45.9 percent.  Compared with last year, the national rate for property crime has declined by 3.6 percent, the violent crime rate is down by 1 percent, and the homicide rate has decreased by 1.9 percent.

However, while the national crime rate is an easy metric to track over time, most of the factors that affect crime are local. And so, as you might expect, crime rates vary from city to city. Focusing exclusively on the big picture can obscure local trends. To address this concern, and provide a more nuanced description of crime rates in America, Vera has updated our earlier analysis of FBI data using the crime data released today, with emphasis on the changes from 2016 to 2017.

The latest round of data from the FBI clearly shows that we continue to live in a period of safety that is unprecedented within many of our lifetimes (the last time we saw crime rates this low was in 1967). This is essential to remember when considering headlines, political speeches, or tweets that claim otherwise—claims that erroneously tout the dangers posed by immigrants to the United States and urge, against all evidence, for the adoption of punitive policies. In short, these statements are not supported by what we know about changes in reported crime over the past three decades.

Our analyses focused on 285 cities with populations of 100,000 or greater, classified as small, medium, large, or very large cities. Here’s what we found:

Property crime: Small decreases in the average city-level rate of property crime in 2017 continued a long-term decline that began in the late 1980s, and this was true for small, medium, large and very large cities. Vera’s analysis of cities revealed wide variation (see below for details). However, the number of cities experiencing decreases or minimal change far outweighed the number experiencing increases.

Violent crime: Comparing 2016 and 2017, violent crime rates declined slightly in three of the four groups included in the analysis: very large cities, small cities, and medium-sized cities. Violent crime was slightly higher in 2017 for large cities. For all four groups of cities, the averages of city-level rates of violent crime remained at or near their lowest values in more than three decades. The number of cities experiencing decreases or minimal changes far outweighed the number experiencing increases.

Homicide: Compared to 2016, the averages of city-level rates of homicide were slightly higher in 2017 for large cities and small cities, but slightly lower in 2017 for very large and medium-sized cities. For all four groups, the averages of city-level rates of homicide remained at or near their lowest values in more than three decades. Again, the number of cities experiencing decreases or minimal change far outweighed the number experiencing increases.

Below you can see a more detailed description of results for property crimes, violent crime, and homicide, including trend data going back to 1985 and an exploration of the number of places (cities with more than 100,000 residents) that experienced shifts in the crime rate between 2016 and 2017. The presentation differs from most other published summaries of crime trends in several important respects:

Measures of change. Most reports focus heavily on year-to-year percentage change. However, percentage change tends to exaggerate the actual change in risk to public safety for jurisdictions with low crime rates, and underestimate the implications for jurisdictions with very high rates. Instead, we focus on the actual change in risk to public safety (as measured by absolute change in rates per 100,000 population).

Emphasis on city-level rates. Many reports emphasize average rates for national-, regional-, or state-level aggregates. Those reflect primarily the number of people affected and are strongly influenced by a few cities with very large populations. Our analysis focuses on city-level rates to emphasize the number of places affected and differences among cities. In the following charts, we include average city-level rates for 203 small cities (with populations of 100K to 250K), 48 medium-sized cities (250K to 500K), 23 large cities (500K to 1M), and 11 very large cities (1M and more).

Long-term context. Some reports focus exclusively on the most recent year-to-year change. However, a certain amount of year-to-year fluctuation is normal. Like several others, our analysis places recent changes in the context of long-term trends so readers can judge whether a single year’s change is outside the ordinary.

Number of cities. Most reports focus on 30 to 60 of the largest cities or the cities with highest crime rates. However, there are many more cities with smaller populations, and smaller cities also tend to have lower rates of crime (especially for violent crimes). We include summaries for up to 285 cities with populations of 100,000 or greater for which data were reported in enough years to support long-term trend analyses.†

Property Crime

In every year since 1995, average city-level property crime rates were highest for large cities and lowest for very large cities and small cities (see Figure 1).

In all four population groups, the 2017 declines in average city-level property crime rates were very small in relation to the long-term decline. These changes resulted in the lowest rates in more than three decades.

Changes in City-Level Property Crime Rates 

Across 276 cities with adequate data, the 2016 to 2017 changes in property crime rate ranged from a decrease of 1,106 crimes per 100,000 people to an increase of 1,244 per 100,000.

Changes for most cities were concentrated in a much narrower central range.

100 of 276 cities experienced minimal change—plus or minus 125 crimes per 100K (represented in Figure 2 in the two bars adjacent to zero).

124 of 276 cities experienced greater than minimal decreases in property crime rates.

52 of 276 cities experienced greater than minimal increases in property crime rates.

Violent crime

For small, medium-sized, and very large cities, the average city-level rates of violent crime declined.

For large cities with populations between 500,000 and 1 million, the average city-level rate of violent crime increased from 853 to 885 per 100,000.

In all four population groups, the 2017 changes in average city-level violent crime rates were very small in relation to the long-term decline, and rates remained at or near their lowest levels in more than three decades.

Changes in violent crime rates

Across 273 cities with adequate data, the 2016 to 2017 changes in violent crime rates ranged from a decrease of 230 per 100,000 people to an increase of 257 per 100,000.

Changes for most cities were concentrated in a much narrower central range.

In the context of the very wide overall range (a range of 487 crimes per 100,000), 119 of 273 cities experienced minimal change (plus or minus 25 crimes per 100,000, represented in Figure 4 in the two bars adjacent to zero).

75 of 273 cities experienced greater than minimal decreases in violent crime rates.

79 of 273 cities experienced greater than minimal increases in violent crime rates.

Homicide

The average city-level homicide rate declined slightly from 12.6 to 12.1 homicides per 100,000 across the 48 medium-sized cities and from 10.9 to 10.4 homicides per 100,000 across the 11 very large cities.

The average city-level homicide rate increased slightly from 12.7 to 13.4 homicides per 100,000 across the 23 large cities and from 5.9 to 6.5 per 100,000 across the 204 small cities. In all four population groups, the 2017 changes in average city-level homicide crime rates were very small in relation to the long-term decline, and rates remained at or near their lowest levels in more than three decades

Changes in Homicide Rates

Across 273 cities with adequate data, the 2016 to 2017 changes in homicide rates ranged from a decrease of 22.1 homicides per 100,000 people to an increase of 17.7 homicides per 100,000.

Changes for most cities were concentrated in a much narrower central range.

In the context of the wide overall range (a range of 39.8 homicides per 100,000), 173 of 280 cities experienced minimal change (plus or minus 2 homicides per 100,000, represented in Figure 6 in the two bars adjacent to zero).

46 of 280 cities experienced greater than minimal decreases in homicide rates.

61 of 280 cities experienced greater than minimal increases in homicide rates.

†The number of cities with adequate data to support a given analysis was different for different analyses, ranging from a low of 273 to a high of 285. Most of the differences occurred within the group categorized as “small cities.”