Trendspotting: Gallup Says Crime Rate Fell in 2020

NewsWire: 12/07/2020

• According to the latest survey from Gallup, American adults are reporting the lowest crime victimization rate in 20 years. Just 13% of Americans say they were the victim of a crime (such as burglary, robbery, or assault) within the past 12 months. (Gallup) 
• NH: What's happening to crime in 2020? Good question.
• In late April, following weeks of Covid lockdown, the evidence pointed to a dramatic drop in nearly every type of crime (see "Crime Drops during Covid-19 Pandemic"). Later, after the racial unrest following the death of George Floyd (see "How the Floyd Protests are Different"), many police departments in large cities were reporting higher levels of violence. We wrote about a July study by the Council on Criminal Justice, updated in September, that showed mixed changes in big cities during the summer: a rise in homicide and violent crime, a fall in drug and property crimes, and no change in other crimes. (See "America's Rising Crime Rate.")
• Gallup has now released its "Crime Victimization Survey" for 2020. And the news is good. Compared to last year, the victimization rate reported by households declined by 4 percentage points, and the rate reported by persons declined by 2 percentage points. The 2020 rate wasn't just below what it was in 2019. It was the lowest ever reported by Gallup since it began the survey in 2000.

• Victimization rates are usually deemed superior to arrest or conviction rates because many crimes go unreported. The gold standard for determining victimization rates is the National Crime Victimization Survey (NCVS), carried out yearly since 1973 by the Census Bureau (N=50-100K). Unfortunately, the NCVS report for 2020 won't be released until late in 2021. So the Gallup survey is our best early glimpse of what's going on. Here's the NCVS through 2019.

• Note that the NCVS rate, like the Gallup rate, has trended down gradually since 2000. Through most of the 1990s, before the Gallup survey began, the NCVS rate declined very steeply in what is sometimes called "The Great Crime Decline." In any given year, the absolute number for the NCVS is lower than the Gallup number (7.3% versus 15% in 2019, for example). That's because NCVS asks after a more restrictive number of crimes: rape, sexual assault, robbery, and aggravated assault. Gallup also asks after crimes like burglary, car theft, and vandalism.
• If the NCVS rate, when it appears, follows the same trend as the Gallup rate, then the official violent crime rate for 2020 will clock in at the lowest ever recorded since the NCVS survey began in 1973.
• How likely is that? Hard to say. Gallup's crosstabs show that property crimes like burglary and car theft are responsible for the biggest declines. Gallup data on violent crimes--the focus of NCVS--are more mixed: down slightly for robbery and up slightly for other types of aggravated assault. This suggests that the 2020 outcome may be uncertain.
• On the other hand, if you look at real-time police data from major cities (maintained by U Penn's Quattrone Center), there is some basis for optimism. Arrests for aggravated assault and robbery are clearly down since the April lockdown in most of them. Here let's look at the numbers for Chicago, New York City, and Philadelphia.

• Now let's return to the 2020 Gallup report. By demographic group, Gallup reports declines in just about every category. Cities, interestingly, are reporting the largest decline, and suburbs the least. The West the most, the South the least. Low-income households the most, affluent households the most. So in case you think the problem with the Quattrone Center data is that police in NYC and Chicago just aren't arresting people, despite rising violence, here's evidence that answers your doubts: Low-income city dwellers are less likely to report any violence.
• To be sure, it's also true that nonwhites are reporting the least decline and that Millennials under 35 are the only group reporting a slight rise in victimization. Altogether, however, these numbers don't lend much quantitative support to the widespread media impression that racial unrest in northern cities is triggering a nationally significant rise in violence.

• So where does the media impression come from? It comes in part from the media anchors themselves, who often depict violent protest as examples of justifiable rage. (CNN's Chris Cuomo asks: "Who says protests are supposed to be polite and peaceful?")
• It also comes from the Whitehouse. During his reelection campaign, President Donald Trump argued very openly that he alone could protect the American people from a rising tide of rampage and mayhem. ("You won't be safe in Joe Biden's America.") Judging by the exit polls, Trump's strategy worked: Protests over police violence were an "important" factor for 67% of voters. It just didn't work well enough to get him elected.
• How well Trump's strategy worked is suggested by another recent Gallup survey on Americans' perceptions of crime.

• Several findings in this chart are worth pointing out. First, Americans' perceptions of whether crime is rising or falling are only loosely tied to trends in their own reporting of victimization. Second, Americans' perceptions of crime "in the U.S." is always much more dire than their perceptions of what's happening "in your area."
• What's really interesting, though, is what happened this year. From 2019 to 2020, the share of Americans believing crime is up nationally jumped by 14 points, from 64% to 78%. That's the second-largest gain ever reported. The first largest was the rebound in 2002, though in that year, at least, Americans also reported that crime was rising locally. In 2020, however, the share reporting a local rise in crime actually declined. In fact, a record share of respondents in 2020 (62%) said that crime locally was either "not too serious" or "not serious at all."
• If you suspect this cognitive dissonance might be politically motivated, well, Gallup has you covered. Gallup breaks out the respondents by political party. The share of Democrats saying crime has risen nationally in 2020 rose by only 4 points. The share of Republicans saying the same thing rose by 24 points. (Notice that these two groups moved in the opposite direction in 2017, the year Trump assumed office.)

• Did you expect any better? The two parties are no more objective in evaluating crime trends than they are in evaluating economic trends. (See "If You Love Trump, You Love the Economy. If Not, Not.")
• Let's sum up. Official survey numbers on criminal victimization in 2020 won't be released until next September. But based on Gallup's survey, plus real-time data from urban law enforcement agencies, it appears that crime did not rise much if at all this year. Indeed, it may have declined. If it has declined, this would extend a generational trend toward less violence that has been with us now for just over 25 years.