Charting Gun Violence and Median Income Across the United States
This second spreadsheet tests the assumption that gun crime may be related to income. Again it shows the gunfire homicide rate per 100,000; but it is confined to the United States.
The other statistic is the median income. “Median” indicates that half the reported incomes were above this value, and half were below.
The firearms murder rates are for 2006; the income numbers are the household averages from 2002-2004. One could make the case that it’s reasonable to look for a cause that pre-dates the effect. Honestly, the 2004 income figures were readily available.
Why does the chart show “Median Income per $10,000” rather than the direct number? The graph now shows similar numbers for “Murders per 100,000” and “Income per $10,000”. Without adjusting the income, the blue “Murder Rate” line would have hugged the x-axis of the chart. Now, however, the lines share the same area nicely. One should plot the graph with two different scales, but this illustrates the point well enough.
The statistics are incomplete, because the data set missed Florida’s murder rate for gun crimes and the income in the District of Columbia. It was tempting to include the murder rate for DC, since it is by far the highest.
Once again, the spreadsheet is sorted by increasing firearms homicides; and once again, the second line “bounces around” rather than tracking along or against the murder trend line. It seems obvious that there is little relationship between a state’s median income and it’s firearms homicide rate.
A sociologist may want to use median income among convicted killers within each state, rather than the overall median income. Of course, there are any number of other ways to differentiate the states. One might examine the per-capita spending on public schooling, for example. For a discussion on gun control, it would be instructive to rate each state for its restrictions on firearms ownership, and plot gun crime rates against that metric.
Gun Statistics: Other Approaches
Many other statistical approaches are possible and potentially useful in gun crime statistics, of course. Types of weapons, and changes in crime rates over time, are just two methods. Pure “gun statistics” would show the rate of ownership: how many people own a gun; and how many guns does each owner possess?
One Canadian study compared the use of firearms and knives in committing a variety of violent crimes in 2006. Knives had a slight “edge” for almost all the crimes of violence, with the exception of attempted murder. Robbery and assault had the largest gap for knives, at 18.9% versus 13.9% and 6.0% versus 1.0% respectively.
Do these statistics indicate that gun control is so effective in Canada that criminals must turn to other weapons? One could also argue that people will commit homicide with whatever is at hand, so there is no point to a ban on assault weapons.
The same report claimed that, in Canada, the “rate of firearm use in violent crime has remained stable since 2003,” and warned that one mass murderer could skew a trend line by adding a significant number of gun crimes into one year.
Decoding Science. One article at a time.