Cooking the Gun Homicide Numbers at the NYT

In his New York Times column "Obama's Gun Play," Charles M. Blow lays out a familiar but inaccurate talking point: we need increased gun control laws because the United States is the murder capital of the planet.  Mr. Blow writes:

[T]he U.S. is in a league of its own, and not in a good way. We have nearly 9 guns for every 10 people, and about 9 out of every 10 of our homicides are committed with one of those guns. No other country even comes close.

A column is accompanied by a large graphic which illustrates American gun violence as an enormous ball by itself way up at the top, while a dozen other tiny spheres representing OECD countries are clustered at the bottom.

New York Times readers, however, ought to hesitate before scurrying to speed-dial their Congress reptiles and demand an end to this preventable epidemic of gun ownership.  To begin with, that repeated phrase "9 out of 10" has a euphonious ring, but Blow's data is simply inaccurate.  I was unable to find any U.S. homicide statistics in the source cited, the United Nations Small Arms Survey, which for the most part discusses the proliferation of guns in places like Timor and Yemen.  According to the Department of Justice, the percentage of homicides for guns in the U.S since 1977 has risen slightly from 62% to 68% -- less than seven out of ten, not nine out of ten.

More importantly, although these figures may be interesting for certain analyses, they don't tell us much about societal violence.  Incredibly, Blow never addresses the obvious relevant figure: the overall homicide rate.  He is able to claim that the U.S. is in "a league of its own," with a country sphere at the top of the chart, only by basing his analysis on an irrelevant statistic -- "percent of homicides by firearms."  A homicide is a homicide, whether it's done with a Glock or a garden hose.  In the U.S., sharp instruments -- e.g., knives -- are the second-most popular murder weapon.  They have dropped from around 20% to 12% of the total.  For comparison, in Canada, guns and knives both account for around 34% of the total murders.  In England and Wales, where guns are strictly controlled, 28% of homicides are committed with a knife and only 6% with a firearm.  British gun murders in some years rank below death by blunt object, kicking, and strangulation.  But so what?  During the Rwandan genocide, the gun homicide rate was extremely low, given the preference of Hutus for death by machete.  Is the percentage of homicides by kicking any less indicative of violence?  Homicidal people seem to find the means necessary, and if nothing else is available, strangulation with one's bare hands is quite effective.

A second statistic presented in the New York Times chart repeats this same distortion.  The circles representing countries are "sized relative to firearm homicides per capita" rather than overall homicide rates.  Thus, the country sphere for the U.S., with an overall homicide rate of 5.0 per 100,000, is much larger than Mexico's, which has 15.0 homicides per 100,000, or triple that of the U.S.  The country sphere for Turkey is miniscule despite Turkey's homicide rate of 6.9, 28% higher than that of the U.S.  Turks and Mexicans apparently find ways to kill each other without using guns.  In the case of the Mexican drug gangs, beheadings are trending upward.

Blow also selects only OECD, which are thirty wealthy countries with relatively low crime rates countries for his data.  OECD members Mexico and Turkey, as mentioned, have higher homicide rates than ours, and when we look at the world as a whole, the U.S. ranks 45th, below El Salvador (71 murders per 100,000, or 14 times the U.S. rate), Honduras (67), Jamaica (58), Guatemala (52), etc.

In addition, Blow chooses data from 2005, although 2009 data is readily available.  In this 2005-2009 period, the U.S. homicide rate dropped by 11%, from 5.6 to 5.0 -- with a one-year drop of 8% from 2008 to 2009 -- while Mexico's murder rate rose by 36%, from 11.0 in 2005 to 15.0 in 2009.  None of these trends holds any interest for Mr. Blow since they don't fit with his pre-established paradigm that increased gun ownership leads to increased homicide.

Mr. Blow, who is black, unfortunately injects race into his plea for gun control, noting that "gun owners are almost twice as likely to be white as nonwhite, are more than three times as likely to be male as female and are more likely to live in the South and Midwest than in the East or West."

Once again, Blow uses what appear to be statistical facts -- "three times as likely," etc. -- to make a point about gun violence.  One might ask, however, if white Southern redneck guys own a lot of guns, then why is the homicide rate for American blacks, according to Department of Justice statistics, thirteen times that of whites?  The 2005 figures are 3.5 homicide offenders per 100,000 for whites and 26.5 for blacks.  For black males in the 18-24 age group, the figure is -- tragically for their predominantly black victims -- 203.3 per 100,000, or 58 times that of the overall white population.  These enormous disparities among groups clearly cannot be explained by gun ownership.

It is true that the American murder rate is much higher than that of any country in Western Europe.  Finland has the highest rate, 2.2, while neighboring Norway has the lowest of 0.6.  Should we strive to lower our murder rates to European levels?  Of course we should.  But singling out irrelevant and inaccurate statistics about gun homicides and gun ownership to make America look more violent in order to argue for greater controls on guns does not help the debate.

Gullible New York Times readers are likely to be bamboozled by such deceptive statistical games, but Blow's propaganda cannot withstand serious scrutiny.
In his New York Times column "Obama's Gun Play," Charles M. Blow lays out a familiar but inaccurate talking point: we need increased gun control laws because the United States is the murder capital of the planet.  Mr. Blow writes:

[T]he U.S. is in a league of its own, and not in a good way. We have nearly 9 guns for every 10 people, and about 9 out of every 10 of our homicides are committed with one of those guns. No other country even comes close.

A column is accompanied by a large graphic which illustrates American gun violence as an enormous ball by itself way up at the top, while a dozen other tiny spheres representing OECD countries are clustered at the bottom.

New York Times readers, however, ought to hesitate before scurrying to speed-dial their Congress reptiles and demand an end to this preventable epidemic of gun ownership.  To begin with, that repeated phrase "9 out of 10" has a euphonious ring, but Blow's data is simply inaccurate.  I was unable to find any U.S. homicide statistics in the source cited, the United Nations Small Arms Survey, which for the most part discusses the proliferation of guns in places like Timor and Yemen.  According to the Department of Justice, the percentage of homicides for guns in the U.S since 1977 has risen slightly from 62% to 68% -- less than seven out of ten, not nine out of ten.

More importantly, although these figures may be interesting for certain analyses, they don't tell us much about societal violence.  Incredibly, Blow never addresses the obvious relevant figure: the overall homicide rate.  He is able to claim that the U.S. is in "a league of its own," with a country sphere at the top of the chart, only by basing his analysis on an irrelevant statistic -- "percent of homicides by firearms."  A homicide is a homicide, whether it's done with a Glock or a garden hose.  In the U.S., sharp instruments -- e.g., knives -- are the second-most popular murder weapon.  They have dropped from around 20% to 12% of the total.  For comparison, in Canada, guns and knives both account for around 34% of the total murders.  In England and Wales, where guns are strictly controlled, 28% of homicides are committed with a knife and only 6% with a firearm.  British gun murders in some years rank below death by blunt object, kicking, and strangulation.  But so what?  During the Rwandan genocide, the gun homicide rate was extremely low, given the preference of Hutus for death by machete.  Is the percentage of homicides by kicking any less indicative of violence?  Homicidal people seem to find the means necessary, and if nothing else is available, strangulation with one's bare hands is quite effective.

A second statistic presented in the New York Times chart repeats this same distortion.  The circles representing countries are "sized relative to firearm homicides per capita" rather than overall homicide rates.  Thus, the country sphere for the U.S., with an overall homicide rate of 5.0 per 100,000, is much larger than Mexico's, which has 15.0 homicides per 100,000, or triple that of the U.S.  The country sphere for Turkey is miniscule despite Turkey's homicide rate of 6.9, 28% higher than that of the U.S.  Turks and Mexicans apparently find ways to kill each other without using guns.  In the case of the Mexican drug gangs, beheadings are trending upward.

Blow also selects only OECD, which are thirty wealthy countries with relatively low crime rates countries for his data.  OECD members Mexico and Turkey, as mentioned, have higher homicide rates than ours, and when we look at the world as a whole, the U.S. ranks 45th, below El Salvador (71 murders per 100,000, or 14 times the U.S. rate), Honduras (67), Jamaica (58), Guatemala (52), etc.

In addition, Blow chooses data from 2005, although 2009 data is readily available.  In this 2005-2009 period, the U.S. homicide rate dropped by 11%, from 5.6 to 5.0 -- with a one-year drop of 8% from 2008 to 2009 -- while Mexico's murder rate rose by 36%, from 11.0 in 2005 to 15.0 in 2009.  None of these trends holds any interest for Mr. Blow since they don't fit with his pre-established paradigm that increased gun ownership leads to increased homicide.

Mr. Blow, who is black, unfortunately injects race into his plea for gun control, noting that "gun owners are almost twice as likely to be white as nonwhite, are more than three times as likely to be male as female and are more likely to live in the South and Midwest than in the East or West."

Once again, Blow uses what appear to be statistical facts -- "three times as likely," etc. -- to make a point about gun violence.  One might ask, however, if white Southern redneck guys own a lot of guns, then why is the homicide rate for American blacks, according to Department of Justice statistics, thirteen times that of whites?  The 2005 figures are 3.5 homicide offenders per 100,000 for whites and 26.5 for blacks.  For black males in the 18-24 age group, the figure is -- tragically for their predominantly black victims -- 203.3 per 100,000, or 58 times that of the overall white population.  These enormous disparities among groups clearly cannot be explained by gun ownership.

It is true that the American murder rate is much higher than that of any country in Western Europe.  Finland has the highest rate, 2.2, while neighboring Norway has the lowest of 0.6.  Should we strive to lower our murder rates to European levels?  Of course we should.  But singling out irrelevant and inaccurate statistics about gun homicides and gun ownership to make America look more violent in order to argue for greater controls on guns does not help the debate.

Gullible New York Times readers are likely to be bamboozled by such deceptive statistical games, but Blow's propaganda cannot withstand serious scrutiny.