Is the new crime wave upon us? Don’t believe the headlines

The FBI released preliminary estimates for crime in the year 2005 earlier this week. This is how the headlines read:

From The Independent, a British newspaper, writes:

“Violent crime on the increase, says FBI report”

The first line of the article says “The United States is experiencing its biggest jump in violent crime in 15 years…”
From CNN:

“Violent crime takes first big jump since ’91”

From NPR:

“US violent crime rises at pace unseen in 10 years”

From the Chicago Sun-Times:

“FBI:Violent crime on rise in 2005”

Based on these headlines, you might think that the next great crime wave arrived in 2005.

Perhaps you remember James Alan Fox from Freakonomics. He is the one we write about who did the report for Janet Reno that predicted a “bloodbath” of youth violence, even as crime was starting to plunge in the 1990s. He echoes the concerns of these headlines (but with a little more moderation than he used back in the 1990s:

Criminal justice experts said the statistics reflect the nation’s complacency in fighting crime, a product of dramatic declines in the 1990s and the abandonment of effective programs that emphasized prevention, putting more police officers on the street and controlling the spread of guns.

“We see that budgets for policing are being slashed and the federal government has gotten out of that business,” said James Alan Fox, a criminal justice professor at Northeastern University in Boston. “Funding for prevention at the federal level and many localities are down and the (National Rifle Association) has renewed strength.”

Still, Fox said, “We’re still far better off than we were during the double-digit crime inflation we saw in the 1970s.”

In light of this media reporting, I think it is interesting to read what the FBI news release actually says:

Preliminary figures indicate that, as a whole, law enforcement agencies throughout the Nation reported an increase of 2.5 percent in the number of violent crimes brought to their attention in 2005 when compared to figures reported for 2004. The violent crime category includes murder, forcible rape, robbery, and aggravated assault. The number of property crimes in the United States from January to December of 2005 decreased 1.6 percent when compared to data from the same time period in 2004. Property crimes include burglary, larceny-theft, and motor vehicle theft. Arson is also a property crime, but data for arson are not included in property crime totals. Figures for 2005 indicated that arson decreased 2.2 percent when compared to 2004 figures.

So the actual increase in violent crime from 2004 to 2005: 2.5%. Given that violent crime has fallen 40-50% since its peak, this hardly seems like reason to panic. And I find it very interesting that none of the headlines I could find made any mention of the fact that property crime fell 1.6 percent. I guess after so many years of falling crime, more falling crime just isn’t newsworthy.

While the pundits are eager to start explaining the many reasons for this “sharp rise” in violent crime, my view is that all of this is nonsense. Prison populations are growing slightly, and my guess is the number of police is growing slightly as well. By now we have gotten almost all of the expected benefits of legalized abortion and crack-related crime is flat. Other demographic shifts are too small to matter on an annual basis. So all in all the factors that I believe to influence crime basically predict no big shifts in crime one way or the other. Given the inherent variability in crime, it is not at all surprising to see shifts up or down a few percent.

But that doesn’t make for exciting headlines.


qualityg

Steven,

From Post above - "Given the inherent variability in crime, it is not at all surprising to see shifts up or down a few percent."

"Finally someone speaks about "Normal Variation."

On a lighter (deeper) note the FBI may use the same violent crime stats over and over like the do in Detroit looking for Jimmy Hoffa.

WEEK ENDING 5/27/06
FBI CIRCUS AT FARM PUTS ON BIG SHOW: Barn shredded in Hoffa search.

http://qualityg.blogspot.com/2006/06/qualitygs-wacko-week-see-week-ending.html

SteveSailer

The abortion rate among whites fell from 19 in 1991 to 11 in 1999, according to the Alan Guttmacher Institute of Planned Parenthood.

Should we thus soon expect an upturn in crime rates among white 14-17 year olds?

Seth_Roberts

In the 1950s, the economist Oscar Morgenstern complained that changes in GDP and the like that economists took to be significant and interpreted were actually measurement error. I wonder how long it will be before these survey numbers come with an indication of the measurement error.

lfstevens

Thanks for commenting on the new numbers. Perspective is a beautiful thing.

Trogdor

"... and the (National Rifle Association) has renewed strength ..."

I'd like an explanation how this leads to an increase in violent crime. I'd also like to see how the government "controlling the spread of guns" leads to a decrease in violent crime.

These are big claims ... so where's the big proof to back them up, hmm?

And I agree that perspective is beautiful, and in mentioning measurement error ... I'd have to think 2.5% (and admittedly 1.6%) can't be far from it.

PatHMV

I looked at the numbers in more depth at my own blog, StubbornFact.us. I went to the 2004 UCR report to get more detailed data, since the preliminary report today only disclosed the trend numbers, not the actual increases. Extrapolating the 2005 numbers by applying the percentage change to the 2004 numbers, the increase looks even more minor. Violent crimes per 1,000 people rose from 4.65 in 2004 to (gasp) 4.77 in 2005. By the by, the graphs in the NPR story are even more misleading than the headlines.

A bigger source of error is that the UCR covers only crimes reported to the police. The US DOJ conducts an annual National Crime Victimization Survey, which surveys a large portion of the country to count both reported and unreported crimes. Those numbers are probably more accurate, especially for rape crimes, because those are less likely to be reported to the police.

Read more...

leliathomas

If we were to all use the sort of logic that was used in these reports, we might come to the conclusion that rapes are on the rise, too. However, is it that more rapes are taking place or that more rapes are getting reported? Just because there has been a 2.5 percent increase in violent crime does not mean there's more crime. Quite possibly, it just means more people are reporting situations, maybe through anonymous tip lines or emergency usage of their cell phones.

It's getting hard to read the news these days. Hardly any article is without an obvious agenda, and reporters are often oblivious to it or can't do anything about it, because their employers won't let them.

Urban Shocker

it's the meth crisis :)
http://www.sentencingproject.org/news.cfm

"Important findings of the report include:

Methamphetamine is among the least commonly used drugs.
Methamphetamine remains a rare occurrence throughout most of the country and is not indicative of a nationwide problem.
Methamphetamine use is declining among our nation's youth."

markohmark

Speaking of reporting/measurement errors,
here's a big one from Minneapolis:

http://www.startribune.com/462/story/487441.html

"The FBI released data Monday indicating that violent crime surged 35.5 percent in Minneapolis last year, but the city's Police Department said a computer glitch grossly exaggerated what was actually a 15 percent jump from 2004.
A 15 percent increase is still six times the 2.5 percent national rise in murders, rapes, robberies and aggravated assaults in 2005, figures disclosed in preliminary data from the FBI's Uniform Crime Report. Still, the volume of violent crime in Minneapolis is about 30 percent below levels in the mid-1990s, when the number of annual murders nearly hit triple digits.

The nationwide increase in violent crime reversed a five-year downward trend, but these crimes are more than 25 percent below the numbers reported in 1992, the start of a years-long decline."

Read more...

qualityg

Steven,

From Post above - "Given the inherent variability in crime, it is not at all surprising to see shifts up or down a few percent."

"Finally someone speaks about "Normal Variation."

On a lighter (deeper) note the FBI may use the same violent crime stats over and over like the do in Detroit looking for Jimmy Hoffa.

WEEK ENDING 5/27/06
FBI CIRCUS AT FARM PUTS ON BIG SHOW: Barn shredded in Hoffa search.

http://qualityg.blogspot.com/2006/06/qualitygs-wacko-week-see-week-ending.html

SteveSailer

The abortion rate among whites fell from 19 in 1991 to 11 in 1999, according to the Alan Guttmacher Institute of Planned Parenthood.

Should we thus soon expect an upturn in crime rates among white 14-17 year olds?

Seth_Roberts

In the 1950s, the economist Oscar Morgenstern complained that changes in GDP and the like that economists took to be significant and interpreted were actually measurement error. I wonder how long it will be before these survey numbers come with an indication of the measurement error.

lfstevens

Thanks for commenting on the new numbers. Perspective is a beautiful thing.

Trogdor

"... and the (National Rifle Association) has renewed strength ..."

I'd like an explanation how this leads to an increase in violent crime. I'd also like to see how the government "controlling the spread of guns" leads to a decrease in violent crime.

These are big claims ... so where's the big proof to back them up, hmm?

And I agree that perspective is beautiful, and in mentioning measurement error ... I'd have to think 2.5% (and admittedly 1.6%) can't be far from it.

PatHMV

I looked at the numbers in more depth at my own blog, StubbornFact.us. I went to the 2004 UCR report to get more detailed data, since the preliminary report today only disclosed the trend numbers, not the actual increases. Extrapolating the 2005 numbers by applying the percentage change to the 2004 numbers, the increase looks even more minor. Violent crimes per 1,000 people rose from 4.65 in 2004 to (gasp) 4.77 in 2005. By the by, the graphs in the NPR story are even more misleading than the headlines.

A bigger source of error is that the UCR covers only crimes reported to the police. The US DOJ conducts an annual National Crime Victimization Survey, which surveys a large portion of the country to count both reported and unreported crimes. Those numbers are probably more accurate, especially for rape crimes, because those are less likely to be reported to the police.

Read more...

leliathomas

If we were to all use the sort of logic that was used in these reports, we might come to the conclusion that rapes are on the rise, too. However, is it that more rapes are taking place or that more rapes are getting reported? Just because there has been a 2.5 percent increase in violent crime does not mean there's more crime. Quite possibly, it just means more people are reporting situations, maybe through anonymous tip lines or emergency usage of their cell phones.

It's getting hard to read the news these days. Hardly any article is without an obvious agenda, and reporters are often oblivious to it or can't do anything about it, because their employers won't let them.

Urban Shocker

it's the meth crisis :)
http://www.sentencingproject.org/news.cfm

"Important findings of the report include:

Methamphetamine is among the least commonly used drugs.
Methamphetamine remains a rare occurrence throughout most of the country and is not indicative of a nationwide problem.
Methamphetamine use is declining among our nation's youth."

markohmark

Speaking of reporting/measurement errors,
here's a big one from Minneapolis:

http://www.startribune.com/462/story/487441.html

"The FBI released data Monday indicating that violent crime surged 35.5 percent in Minneapolis last year, but the city's Police Department said a computer glitch grossly exaggerated what was actually a 15 percent jump from 2004.
A 15 percent increase is still six times the 2.5 percent national rise in murders, rapes, robberies and aggravated assaults in 2005, figures disclosed in preliminary data from the FBI's Uniform Crime Report. Still, the volume of violent crime in Minneapolis is about 30 percent below levels in the mid-1990s, when the number of annual murders nearly hit triple digits.

The nationwide increase in violent crime reversed a five-year downward trend, but these crimes are more than 25 percent below the numbers reported in 1992, the start of a years-long decline."

Read more...