| Torture numbers, and they'll confess to anything- | | | | city over 100,000, and that is Baltimore, with 637,556. |
| Gregg Easterbrook | | | | It wasn't a perfect match, but it was close enough. |
| On Aug. 25, I read an interesting letter to the editor | | | | But Virginia was in a special category. Since it does |
| of one of the major daily newspapers that serves | | | | not have any large cities, I had to aggregate the |
| the Greater Washington, D.C., Area. The writer | | | | population data for Norfolk, Richmond and Newport |
| stated that Virginia, with its relatively lax gun laws, | | | | News to come up with a total of 611,666. The crime |
| has a lower crime rate than Maryland and the District | | | | data for these three smaller cities, I concluded, should |
| of Columbia, both of which have stricter laws. From | | | | collectively be comparable to similar data for the |
| this statement readers were expected to conclude | | | | District and for Baltimore. |
| that more guns on the street mean less crime. | | | | I now had three separate geographic areas that |
| At first I was inclined to dismiss the man's argument | | | | were comparable, at least in population totals. |
| as something merely copied from a National Rifle | | | | I then went to the FBI's Uniform Crime Report and |
| Association press release. But instead I decided to do | | | | discovered that data are arrayed for all cities with |
| some research about crime rates in the three | | | | over 100,000 population in nine different crime |
| geographic areas mentioned by the letter writer. | | | | categories, not just one. Further, when rates for |
| The first thing I discovered was that the | | | | comparable areas are considered, Virginia doesn't look |
| demographics of Virginia, Maryland and the District | | | | so good. It has higher rates of forcible rape and |
| are quite different. Virginia has vast rural and | | | | larceny than Baltimore. Also, when rates are |
| semi-rural areas, whereas the District is a densely | | | | compared with those in the District, Virginia ends up |
| populated city. Maryland is similar to Virginia with one | | | | having more cases of forcible rape, property crimes, |
| exception: it has a city, Baltimore, that rivals the | | | | larceny and burglary. Thus, the original letter writer |
| District in size and composition. Virginia, of course, | | | | had no basis for arguing that lax gun laws equate to |
| does not have a city of comparable size. Virginia's | | | | less crime. Reversing the writer's logic, an argument |
| largest city, Norfolk, has a population which is less | | | | can be made that in many categories just the |
| than half that of the District. | | | | opposite is true. |
| I learned in my college statistics class that samples (in | | | | The lesson in all this is that readers should be |
| this case, cities) must be similar before comparisons | | | | skeptical when they read letters, reports, news |
| can be made. With this in mind, I tried to identify | | | | stories and articles (including this one) that contain |
| comparable cities or areas in each of the three | | | | statistics. Numbers can be manipulated to prove |
| jurisdictions. I discovered that the District has an | | | | almost anything. |
| urban population of 581,530. Baltimore has only one | | | | |