gaithersblog.net

As has been mentioned elsewhere (e.g. The Examiner and Mocoprogressive), the County Police have started to publish a list of outstanding traffic warrants. In looking through the list, however, it struck me that some of the names (both first names and last names) were quite common therein. I wanted to check that perception matched reality, so I pulled the records out of the PDF and did a little sorting and counting.

Now, there were, by my count, a total of 5877 records in the August 22 list. There were a total of 3213 surnames (last names) represented in that list. Below is a table of the most common 32, or 1% of those names, with a count of the number of warrants for people with those names. Note that those counts add up to exactly 1000, or 17% of the warrants:

I did the same analysis for first names on the list. There were fewer unique first names, 1880 in total. Here are the nineteen (again, 1%) most common:

This 1% of first names accounts for 1453 of the 5877 warrants, or 25%. The first three, which are the only names which appear on more than 100 warrants, add up to 628, or about 11% of all the warrants in the County’s list. I also looked a little at the age distribution, using the listed birthdates. It was, I think, pretty unsurprising that the counts took a steep climb from age 18 to around age 25, and then slowly tailed off between the ages of around 32 and, well, old. (There were some records for people who are either over 100 or under 10. They could be errors, they could be old warrants that have never been cleared, or we could just have some centenarian scofflaws out there).

Update: It occurred to me that someone might wonder how overrepresented some of these names might be if some of the people on the list had multiple warrants, so I did a little more analysis to try to adjust for this. Understand that I can’t really adjust for it, because the list provides no way to tell with certainty that two warrants are for the same person. So what I did do was to collapse into a single count any records where the first name, last name and birthdate matched; there were a total of 5319 unique first name/last name/birthdate combinations in the file. This modification could be wrong in multiple directions — an alias, typo or alternative spelling for example could result in a single person still being listed twice, while there is no reason you couldn’t have two people with the same name born on the same date. But at least this is a one way to try to get an idea of how multiple warrants affects the list.

In fact, however, after doing this the name lists change relatively little. On the surname list, we pick up Fuentes and Jackson, losing Guzman and Pineda. On the given name list, we lose Hector and pick up Anthony. The orders also change a little bit: Robert and William trade places, for example, and Oscar moves up a few to before John. In the surname list, Martinez and Lopez switch, as do Flores and Johnson; there’s maybe a dozen such small changes. But the first three names on the given name list — which don’t change order — still amount to about 11% of of the unique name/birthdate combinations on the list.